Apoptosis in Normal Development and Cancer
Apoptosis in Normal Development and Cancer Edited by MELS SLUYSER Netherla...
38 downloads
1045 Views
3MB 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
Apoptosis in Normal Development and Cancer
Apoptosis in Normal Development and Cancer Edited by MELS SLUYSER Netherlands Cancer Institute
UK Taylor & Francis Ltd, 1 Gunpowder Square, London EC4A 3DE USA Taylor & Francis Inc., 1900 Frost Road, Suite 101, Bristol, PA 19007 This edition published in the Taylor & Francis e-Library, 2005. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.” Copyright © Taylor & Francis Ltd 1996 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, electrostatic, magnetic tape, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. ISBN 0-203-48315-4 Master e-book ISBN
ISBN 0-203-79139-8 (Adobe eReader Format) ISBN 0-7484-0444-9 (Print Edition) (formerly 013-320-599-1 (Print Edition)) Library of Congress Cataloging Publication Data are available Cover design by Jim Wilkie The picture of apoptotic cells on the cover was supplied by Dr N.Mori (Research Institute for Advanced Science and Technology, Osaka Prefecture University, Japan) and Dr M.van der Valk (Department of Molecular Genetics, The Netherlands Cancer Institute, The Netherlands).
Contents
Editor’s Introduction
vi
1
Apoptosis: a 20th Century Scientific Revolution B.V.Harmon and D.J.Allan
1
2
The Enzymology of Apoptosis F.M.Hughes and J.A.Cidlowski
18
3
Molecular Mechanisms of Apoptosis S.Tanuma
34
4
Identification of Genes Associated with Cell Death C.A.Lamb and J.J.Cohen
54
5
Genes Involved with Apoptosis N.J.McCarthy, E.A.Harrington and G.I.Evan
63
6
The Role of p53 in Apoptosis S.W.Lowe
86
7
Bcl-2 and the Regulation of Programmed Cell Death J.C.Reed
112
8
Molecular and Genetic Control of Apoptosis in Drosophila J.M.Abrams
151
9
Evolution of Apoptosis D.L.Vaux
166
10
Hormonal Control of Apoptosis in Gonadal Tissues H.Billig and A.J.W.Hseuh
177
11
Apoptosis in the Immune System C.D.Gregory
196
12
Suppression of Apoptosis by Monoclonal Antibodies in Mouse Malignant T-lymphoma Cells N.Fujita, S.Kataoka and T.Tsuruo
223
13
Apoptosis as a Predictive Factor for Cancer Therapy F.Pezzella, G.Gasparini, A.L.Harris and K.C.Gatter
234
14
Control of Cell Death in the Nervous System
245
v
H.C.Drexler and J.-Y.Yuan Index
264
Editor’s Introduction
Like people, cells die in different ways: accident, murder, old age, even suicide. Cells that are chosen to die —either because they are superfluous, diseased, or have served their useful purpose—don’t just fall apart and expire: they go through a predictable, well-choreographed series of events. The cells round up, their outer membranes form bulges called blebs, nuclear membranes and some internal structures break down, the nuclear DNA is fragmented by enzymes, and finally the cell breaks into pieces that are devoured by still vital neighboring cells. This mode of cell death, in which single cells are deleted in the midst of living tissues, is called ‘apoptosis’, a Greek word for the dropping of leaves from trees. Apoptosis, first defined in the 1970s, is important for many aspects of life. During embryonic development it serves to remove cells which have lost their usefulness. Apoptosis is also responsible for the physiological death of cells in the course of normal tissue turnover. When cells of the immune system are selected during an immune response, other cells have to be removed and this is accomplished by apoptosis. As well as being induced by developmental control processes, apoptosis can also be induced by toxic insults and, in particular, by agents that damage DNA. It is frequently seen in tumors and may play a key part in the kinetics of tumor growth. Many genotoxic cancer treatments may also exert their effects through the enhanced induction of apoptosis. Certain genes are responsible for regulating apoptosis, and importantly, these same genes also play a role in cancer. Cancer therapy by irradiation, chemotherapy, and hormone treatment all induce apoptosis in tumor cells, although high doses can also cause cell destruction by other means. An understanding of the mechanisms of apoptosis is therefore central to gaining a better insight into developmental biology and cancer, and may ultimately lead to ways of treating cancer more effectively. The book Apoptosis in Normal Development and Cancer is a state-of-the-art presentation of this rapidly developing field. After a historical introduction it reviews our current knowledge of enzymology and molecular mechanisms. The genetic basis of apoptosis is discussed in detail. After a review of evolutionary aspects, hormonal control and the role of apoptosis in the immune system, new insights are presented into how apoptosis may be used to improve cancer diagnosis and therapy. We hope that this book will be a useful guide to workers in the field, and will also serve as an overview to those interested in the mechanisms that are the basis of cellular life and death. Mels Sluyser Amsterdam, 1996
1 Apoptosis: a 20th Century Scientific Revolution BRIAN V.HARMON and DAVID J.ALLAN School of Life Science, Queensland University of Technology, Brisbane, Queensland 4000, Australia
1.1 Introduction History is made without knowing of its making Jean-Paul Sartre The apoptosis concept, that death of selected individual cells within a tissue is an active process mediated by the cell’s own biochemical mechanisms [1], represents one of the most important milestones in cell and tissue research this century. Prior to its introduction, all cell death was considered to be the outcome of injury and to be degenerative in nature. The discovery of apoptosis owes much to careful electron microscopic observations, the fortuitous coming together of minds and, as so often happens with major discoveries, a good deal of luck. However, the most important thing about the work of Kerr, Wyllie and Currie was not so much the new facts they discovered but the new way they developed of looking at and interpreting their findings and the published work of others. The ‘revolutionary’ idea that a cell could actively participate in its own death was not to be generally accepted until some 20 years had passed. The genesis, extension and ultimate acceptance of the apoptosis concept shows many similarities to the historical development of other major scientific concepts. Our interest in apoptosis dates back to the early years. We shared the good fortune of working with John Kerr in the Pathology Department, University of Queensland, at the time when many of the events we are about to describe were happening (see note at end of chapter). John Kerr’s recently published reminiscences [2] of the period leading up to the first apoptosis paper were an invaluable source of information. However, we have tried where possible to rely on our own recollections of the events that took place and the interpretations we have given are ours alone. In this chapter, an historical perspective, we examine the events leading up to the discovery of apoptosis in 1972 and then follow its evolution as a concept until around 1990, when the explosion of interest in apoptosis began.
2
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
1.2 The Cell Before Apoptosis Although cells were first observed in thin slices of cork by Robert Hooke as far back as 1665, for many years few scientists realized the enormous significance of this finding. It was not until much later (in 1838), when the German botanist Matthias Schleiden stated that the cell was the basic unit of life, that the ‘cell theory’ was finally born. Theodor Schwann, a German physiologist, advanced the same theory the following year, and both Schleiden and Schwann are generally credited with establishing the ‘cell theory’ [3]. While dividing cells seem to have been noticed as early as 1832 by Barthelemy Charles Dumortier, a Belgian botanist, division of cells by mitosis was not widely recognized until some years after the cell theory had come out [4]. Rudolf Virchow, a German pathologist, was able to incorporate this new information on cell division into his revolutionary Cellularpathologie doctrine [5], proposing that the basis of all disease was injury to the smallest viable constituent unit of the body, the cell. It was to be a further 114 years before cell death by apoptosis was defined by Kerr, Wyllie and Currie [1], and, like the discovery of cells themselves, few scientists were to grasp immediately the importance of this finding. Like all living things, cells must die. Every minute, over 3 billion cells in our body die and, to balance this enormous loss, an equal number of cells must be produced by mitosis. While not all cells that are lost from the body undergo the classical morphological changes of apoptosis (e.g. cells sloughing off the intestinal villi do not), an enormous number of other cells in our body are undergoing this process (apoptosis) at any one time. How is it then, that it took so long for apoptosis to be defined and for the importance of this mode of cell death to be accepted? To answer this question, we need to look at the context in which observations of cell death were being made in the days before apoptosis. Interestingly, scientists were observing various aspects of apoptosis without realizing it for a very long period of time, often believing that they were describing changes unique to the particular circumstances under consideration. Some of the names used for the changes observed included acidophilic and Councilman bodies in the liver [see 6], tingible bodies in lymphoid germinal centres [see 1, 7], karyolytic bodies in gut crypts after X-irradiation [see 8] and Civatte bodies and Sunburn cells in the skin [9]. One of the main reasons the generalized nature of the process they were observing was not appreciated is that most of the scientists working on cell death at that time tended to confine their studies to narrow fields of investigation (e.g. radiobiology or embryology). Furthermore, there existed a firmly entrenched view that all cell death was a degenerative phenomenon produced by injury. To compound matters, morphological descriptions of cell death were confusing as they attempted to accommodate all changes (both apoptotic and necrotic) within the framework of a single type of cell death, necrosis. The cryptic nature of the apoptotic process itself would also have contributed. Necrosis usually affects large groups of contiguous cells and, because the necrotic debris formed remains in the tissue for an appreciable length of time, necrosis is easily observed in tissues. Apoptosis on the other hand affects scattered single cells or small groups of cells, it is extremely rapid, and the apoptotic bodies remain visible in tissues for only a matter of hours after they are formed [10]. This, taken in conjunction with the small size of many apoptotic bodies and the absence of inflammation, means that only small numbers of apoptotic cells may be observed at any one time point, even when many cells are being deleted by this process [10]. 1.3 Genesis of the Apoptosis Concept The apoptosis story began in 1962 when John Kerr, a young Pathology Registrar from the Royal Brisbane Hospital in Australia, went to England to undertake a PhD at the University College Hospital Medical
APOPTOSIS: A 20TH CENTURY SCIENTIFIC REVOLUTION
3
School under the supervision of the distinguished Australian pathologist Sir Gordon Roy Cameron. Cameron had left Australia permanently to work in England and is best remembered for his monumental monograph Pathology of the cell [3], a work described by Singer and Underwood [4] as ‘one to which Virchow himself might well have been pleased to have set his name’. The thoughts Cameron espoused on cell death in this work are, in hindsight, interesting. He stated ‘We are urgently in need of chemical and physical investigations of cellular death, for our knowledge at the moment is crude and full of gaps’ [3, p. 405]. When he suggested that Kerr should work on ischaemic liver injury produced in rats by ligating branches of the portal vein, he set in train a fascinating journey of discovery that was to yield far more insight into cell death than he could have ever imagined. Joe Smith, Cameron’s deputy, introduced Kerr to the recently developed histochemical methods for delineating lysosomes in frozen sections of tissues and suggested that he might use these methods in his study. Lysosomes had only recently been discovered by DeDuve and his colleagues in Belgium and it was thought that rupture of these organelles with release of their digestive enzymes might be a critical event in the production of cell death, an idea that was not in fact supported by later work. Nevertheless, the lysosomal changes Kerr observed in ischaemic liver were interesting. Patches of coagulative necrosis developed in the central parts of lobules, and in the swollen necrotic cells present in these patches lysosomal enzymes were found dispersed throughout the cytoplasm. In areas bordering patches of confluent necrosis and in the periportal parenchyma however, scattered small shrunken round masses of cytoplasm containing intact, discretely staining lysosomes were observed. The latter appeared to represent a type of cell death distinct from the swollen cells of coagulative necrosis and the name ‘shrinkage necrosis’ was proposed for this process [11]. Importantly, ‘shrinkage necrosis’ seemed to be involved in the useful function of adjusting the mass of the liver to a reduced blood supply as, once the lobes had become moderately atrophic, the cell death ceased. Back in Australia, Kerr took up a position as Senior Lecturer in Pathology at the University of Queensland Medical School and found that ‘shrinkage necrosis’ was not limited to the liver ischaemia model he had been studying; similar dying cells were also present in areas bordering patches of confluent necrosis in livers poisoned with the hepatotoxins heliotrine and albitocin [12, 13]. Electron microscopic studies carried out with the help of firstly David Collins and later Brian Harmon (who successively ran the departmental electron microscope unit) convinced Kerr that ‘shrinkage necrosis’ was indeed a distinct mode of cell death. Over the next few years he published detailed descriptions of the ultrastructural features of this type of cell death [12, 13] culminating in the paper ‘Shrinkage necrosis: a distinct mode of cellular death’ [6]. The small round masses of cytoplasm observed by light microscopy were found to result from condensation and budding of hepatocytes into large numbers of membrane-bounded cellular fragments. In contrast to classical necrosis, organelles within the fragments remained well preserved until they were phagocytosed and digested by surrounding cells (see Figure 1.1). Parenchymal as well as Kupffer cells participated in the phagocytic process, there was no associated inflam mation and cells undergoing this process (shrinkage necrosis) were also found to be present in normal untreated liver. Shrinkage necrosis was subsequently observed by Kerr in a number of other normal tissues, and in a major breakthrough, found to be widespread in basal cell carcinomas of skin [14]. Several other researchers [15–17] had just come to the realization that spontaneous cell loss (calculated to be more than 95% in some tumours) was an important parameter in neoplastic growth, and Kerr and his colleague Jeffrey Searle [14] suggested that much of the cell loss in these tumours might be occurring by shrinkage necrosis. The process was not limited to basal cell carcinomas however, similar dying cells were soon found to be present in other types of malignant tumours [18].
4
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 1.1 Large number of apoptotic bodies (arrowed) in rat liver lobe undergoing atrophy following ligation of its hepatic portal vein. This is the same experimental model as the one used by Kerr in the original ‘shrinkage necrosis’ publication [6].×2500
In 1970, a lucky twist of fate landed A.R. (later to become Sir Alastair) Currie of Aberdeen, Scotland in Brisbane as Mayne Guest Professor in Pathology at the University of Queensland. Kerr spoke to Currie of his findings and showed him electron micrographs of cells undergoing shrinkage necrosis. Currie was excited by these findings as he and Andrew Wyllie (his PhD student) had seen similar small cell fragments histologically in the adrenal cortices of rats given prednisone. Currie invited John Kerr to spend a sabbatical period in the Aberdeen Pathology Department looking more closely into this interesting phenomenon, an offer Kerr took up the following year. Over in Aberdeen, Kerr, Wyllie and Currie embarked on further studies of this new type of cell death. Electron microscopy of the dying cortical cells in adrenals of prednisone treated rats showed the same stereotyped sequence of morphological events that Kerr had observed in livers and tumours. Moreover, the death could be completely prevented if ACTH were given to rats at the same time as the prednisone injection. These findings added a whole new dimension to their studies as they had now established that the death not only occurred under both physiological and pathological conditions, but also could be switched on or off in an endocrine dependent tissue by manipulating levels of trophic hormones. A final important piece of the jigsaw was given to Kerr, Wyllie and Currie by Allison Crawford, a zoologist undertaking a PhD in the Aberdeen Pathology Department at the time. She drew their attention to the extensive literature on programmed cell death in the normal embryo. While the seminal work of Glücksmann at Cambridge in the early 1950s had established the important and essential role that this type of cell death played in embryogenesis [19], few outside the field of embryology knew of its occurrence. Examination of electron micrographs of embryonic cell death in the literature was enough to convince the trio that the death was the same as they had been studying.
APOPTOSIS: A 20TH CENTURY SCIENTIFIC REVOLUTION
5
Thus, what had started out as three essentially separate fields of investigation of cell death, Kerr’s shrinkage necrosis in pathological tissues, Wyllie and Currie’s controllable cell death in endocrine tissues and programmed cell death in the embryo came together to form the basis of the generalized concept on cell death that was soon to be termed apoptosis. The major factor that allowed the link to be made was the clearly recognizable ultrastructural features of the cell death. By themselves, it is unlikely that any of the three findings would have had a long-term impact on the study of cell death. Together, however, they were to form a concept that would revolutionize not only the cell death paradigm but also scientific thinking on a whole host of medically and biologically important processes. After extensive further studies, a paper was written for the British Journal of Cancer and published the following year [1]. The term ‘shrinkage necrosis’ had undesirable connotations as it suggested the new type of cell death was simply another variant of necrosis, which was clearly not the case. The name apoptosis was proposed to stress the importance of this ‘hitherto little recognized mechanism of controlled cell deletion which appears to play a complementary but opposite role to mitosis in the regulation of animal cell populations’ [1]. The footnote from the paper is worth reproducing in full as it describes clearly both the origin and suggested pronunciation of the word ‘apoptosis’: We are most grateful to Professor James Cormack of the Department of Greek, University of Aberdeen, for suggesting this term. The word ‘apoptosis’ is used in Greek to describe the ‘dropping off’ or ‘falling off’ of petals from flowers, or leaves from trees. To show the derivation clearly, we propose that the stress should be on the penultimate syllable, the second half of the word being pronounced like ‘ptosis’ (with the ‘p’ silent), which comes from the same root ‘to fall’ and is already used to describe the drooping of the upper eyelid. Like leaves falling from trees, apoptosis had been found to affect single cells or small groups of cells scattered throughout tissues, not large numbers of contiguous cells as occurs with coagulative necrosis. It was well recognized by the authors at the time that understanding factors that might regulate this controllable mechanism of cell deletion would be of great significance. In support of the apoptosis concept, Kerr, Wyllie and Currie had assembled a great deal of evidence in the short time they had been working together. The morphological changes were essentially the same as had been described earlier by Kerr for ‘shrinkage necrosis’. However, the death (apoptosis) was now shown to be involved in cell turnover in many healthy adult tissues and to be responsible for elimination of cells during normal embryonic development [1]. It occurred spontaneously in many malignant neoplasms and was enhanced in regressing tumours following some types of therapy [1]. Furthermore, it had been shown to participate in normal physiological involution (e.g. endometrium) and pathological atrophy of various tissues and organs (e.g. castration-induced involution of rat prostate) [1]. Finally, apoptosis could be triggered by noxious agents, both in the embryo and adult animal [1]. 1.4 Morphological Features of Apoptosis The morphological features of apoptosis described in the 1972 paper were nuclear and cytoplasmic condensation and breaking up of the cell into a number of membrane-bound ultrastructurally well preserved ‘apoptotic bodies’. This was followed by phagocytosis and degradation of these apoptotic bodies within phagosomes, with both parenchymal cells and histiocytes participating in the phagocytosis. While more detailed descriptions of the morphological changes were given by the authors in subsequent publications,
6
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 1.2 Diagram illustrating the sequence of ultrastructural changes in cells undergoing apoptosis. The features shown are essentially the same as those depicted by Kerr, Wyllie and Currie in the original apoptosis paper [1] but include slightly more detail. An early apoptotic cell (1) is shown surrounded by normal cells. Early apoptosis is characterized by compaction and margination of nuclear chromatin into sharply circumscribed masses that abut on the nuclear envelope, convolution of the nuclear outline, overall condensation of the cell, and in tissues, separation of the dying cell from its neighbours. In the next phase (2) the nucleus fragments, extensive cell surface protrusions develop and membrane-bounded apoptotic bodies of various size and composition are formed (3). Apoptotic bodies formed in vivo are rapidly phagocytosed by either epithelial cells (4) or specialized mononuclear phagocytes (5) and degraded within phagolysosomes. The structural integrity of organelles within apoptotic cells or bodies is usually maintained in vivo until the process of lysosomal degradation begins.
especially in relation to changes in nuclear chromatin [20–23], the key morphological features of the process had all been accurately defined in this ‘first’ apoptosis publication. Before proceeding with the historical overview, we will briefly outline the cardinal morphological features of apoptosis. The
APOPTOSIS: A 20TH CENTURY SCIENTIFIC REVOLUTION
7
Figure 1.3 (a) Scanning electron micrograph of a ‘budding’ apoptotic cell 24 h after addition of actinomycin D to cultures of HeLa cells at a final concentration of 15 µg ml−1. The apoptotic cell, which has pulled away from its neighbours and rounded up, has extensive surface protuberances and one cluster of apoptotic bodies (at bottom right) appears to have just separated. (b) Transmission electron micrograph of a ‘budding’ apoptotic cell at a comparable stage to the one illustrated in (a). The apoptosis in this case is occurring in murine EMT6 sarcoma nodules 2 h after heating at 44°C for 30 min. Note the characteristic condensation and margination of nuclear chromatin and overall condensation of the apoptotic cell. ×2800
morphology still remains the most accurate and reliable method of identifying the process. The description that follows, a present day account, is based on our studies and the detailed published reports of others [10, 20–25 ]. The sequence of ultrastructural changes is summarized in diagrammatic form in Figure 1.2. Electron micrographs illustrating different stages of the process are shown in Figures 1.3–1.6. The earliest changes in apoptosis occur in the nucleus where chromatin is compacted and segregated into sharply circumscribed masses that abut on the inner nuclear envelope (Figures 1.1 and 1.6). Concomitant with these changes, the cell condenses, rounds up, and in tissues, pulls away from its neighbours. Convolution of nuclear and cell outlines then ensues and is followed by ‘budding’ of the cell (Figures 1.3a, b) into a number of discrete membrane-bounded fragments called apoptotic bodies (Figures 1.3a, 1.4). The size of apoptotic bodies formed varies considerably, as does their composition, and while one body may contain several nuclear fragments, others may completely lack a nuclear component. The extent of ‘budding’ (see Figures 1.3a, b) varies markedly with cell type, often being relatively restricted in small cells like lymphocytes with a high nucleocytoplasmic ratio (see Figure 1.6). Apoptotic bodies formed in vivo are rapidly engulfed and phagocytosed by surrounding resident cells or macrophages and degraded within phagolysosomes (Figure 1.5). There is no associated inflammation. Apoptotic bodies formed in vitro are seldom phagocytosed (Figure 1.6) and eventually undergo degenerative changes that resemble necrosis ultrastructurally; the term ‘secondary necrosis’ is often applied to this change [20]. 1.5 Further Development of the Apoptosis Concept, 1972–1980 The definitive apoptosis paper [1] went largely unnoticed by the wider scientific community, and in those areas where it was noticed often received a somewhat hostile reception. As mentioned earlier, there was at the time a firmly entrenched view that all cell death was a degenerative process caused by injury. Pathologists found the concept that cell death could be an active controlled process difficult to accept, and many were among the harshest critics of apoptosis. Embryologists had been studying programmed cell
8
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 1.4 Apoptosis in mouse small intestinal crypt 2 h after injection of cytosine arabinoside at a dose of 250 mg kg−1 body weight. Dying cells have ‘budded’ into a large cluster of markedly condensed apoptotic bodies and several bodies have been extruded into the crypt lumen. Another apoptotic body (arrowed) has been phagocytosed by an adjacent crypt epithelial cell. ×3200
death for more than 20 years and were well aware that cell death was a controllable process. As such, apoptosis was not seen as being particularly novel and did not have much immediate impact on this group of scientists. Moreover, despite the claims by Kerr, Wyllie and Currie that apoptosis was the mechanism of programmed cell death, few ultrastructural studies of cell death in developing embryos had been carried out and there was a lingering doubt among embryologists that the two processes might not be exactly the same. In a review of cell death during embryogenesis published many years later, Hinchliffe [26] likened the morphological changes that occur in embryonic cell death with ‘the apoptotic scheme of cell fragmentation’ but still did not concede they were one and the same process. The other major group of scientists, the biologists, had long considered cell death to be a pathological process with no relevance to normal structure and function. Indeed, standard biological textbooks at the time made little or no mention of cell death. As apoptosis had been defined by pathologists, and as most of the early papers on the topic had been published in pathology journals, most biologists were not even aware of the revolution going on around them. Among those who were, the majority had no understanding of basic pathological processes such as inflammation and scarring and could not appreciate that the way a cell died might be significant. Others were extremely dubious about the claim that apoptotic bodies could be phagocytosed by parenchymal cells, which were not at that time thought to have phagocytic potential. In the account that follows, we give our view of what we believe are some of the important milestones in apoptosis research in the 18 years following its definition. It is not possible to list every contribution made, especially in more recent times, and readers wanting a detailed coverage of apoptosis research over this period are referred to the many comprehensive reviews or book chapters available on the topic [10, 20–25].
APOPTOSIS: A 20TH CENTURY SCIENTIFIC REVOLUTION
9
Figure 1.5 Many apoptotic bodies within an intra-epithelial macrophage in rat ventral prostate 2 days after castration. Although the bodies are at various stages of lysosomal degradation the apoptotic origin of most can still be recognized by the presence of typical nuclear fragments. ×5000
At the end of his study leave in Aberdeen, Kerr returned to Brisbane, and in collaboration with a great number of local scientists set about the task of extending the known range of occurrence of apoptosis. Over the next 8 years they published many papers on the topic, some of the more important findings being that apoptosis was responsible for regression of the tadpole tail during metamorphosis [27], for the involution of rat prostate following castration [28], that it was induced in proliferating normal and neoplastic cells by cancer chemotherapeutic agents [29], that it was enhanced in liver allograft rejection [30] and that it was the mechanism of T-cell killing [31]. Sanderson and Glauert [32] published electron micrographs of T-cell killing the same year as Kerr’s group but did not define the death as being apoptotic in type. Wyllie and Currie, who were both to move to Edinburgh University, published detailed studies with Kerr in 1973 on apoptosis regulated by hormones in the adrenal cortex [33, 34]. Wyllie completed a short review on cell death in normal and neoplastic cells the following year [35] and in 1975 was awarded a PhD degree from the University of Aberdeen. While most of the apoptosis research up to about 1980 centred on either Kerr or Wyllie’s groups, one scientist quick to grasp the importance of the apoptosis concept was Chris Potten, Head of Epithelial Biology at the Paterson Institute for Cancer Research in Manchester. In 1977, following a visit to his laboratory by Jeffrey Searle (one of John Kerr’s colleagues from the University of Queensland Pathology Department), Potten showed the existence of a small group of cells in adult intestinal crypts that were extremely sensitive to apoptotic induction by irradiation [36]. This publication in the influential journal Nature, and Potten’s early acceptance of apoptosis were great boosts to apoptosis research. The apoptosis concept was to take root far earlier in the United Kingdom than in America and it was to be many years later before any interest was shown by American scientists. Apart from the British Journal of Cancer, which published the original apoptosis paper [1] and one other on apoptosis later that year [37],
10
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 1.6 Cells showing the early nuclear changes of apoptosis (condensation and margination of chromatin) and apoptotic bodies containing multiple nuclear fragments are numerous in this culture of human Burkitt’s lymphoma cell line BM 13674 4 h after heating at 43°C for 30 min. ×2500
the journal that deserves most credit for ensuring the survival of apoptosis research in the early difficult years is the Journal of Pathology, which published no fewer than seven papers on the topic from 1971 to 1975 [6, 14, 29, 33, 34, 38, 39]. Publication of apoptosis papers was for a long time fraught with difficulty and many submitted papers were rejected. Wyllie and Currie published little on apoptosis for a few years following their early papers. In 1980 however, Wyllie published a finding in Nature that undoubtedly played a major role in having the apoptosis concept eventually recognized by the wider scientific community, the association between apoptosis and endogenous endonuclease activation [40]. Wyllie proposed that in thymocytes undergoing apoptosis following glucocorticoid treatment, activation of an endogenous endonuclease was responsible for cleavage of DNA at the linker regions between nucleosomes into oligonucleosomal sized fragments [40]. This resulted in the formation of a characteristic ‘ladder’ pattern in agarose gels stained with ethidium bromide (see Figure 1.7); on the other hand, DNA breakdown in cells undergoing necrosis was random and resulted in a diffuse smear. Until this time, the hypothesis that apoptosis was a discrete mode of cell death with distinct biological significance had been based on its characteristic morphology and its circumstances of occurrence. Wyllie’s finding provided strong corroborative biochemical evidence for the discrete nature of the apoptotic process. Moreover, it helped to draw into the apoptosis field a whole new group of researchers who, while having no real interest in the morphological changes, were intrigued by these striking DNA changes. Indeed, many researchers who were later to enter the field did so after they had encountered internucleosomal DNA cleavage in their own areas of research. Wyllie’s paper also marked the introduction of glucocorticoid-treated thymocytes as an in vitro model for biochemical studies of apoptosis.
APOPTOSIS: A 20TH CENTURY SCIENTIFIC REVOLUTION
11
Figure 1.7 Agarose gel electrophoresis of DNA extracted from control and heated cultures of the human Burkitt’s lymphoma cell line, BM 13674 shows the characteristic ‘ladder’ pattern of apoptosis in Lane 3. Lane 1: DRlgest III DNA fragment size marker; Lane 2: control culture; Lane 3: culture showing extensive apoptosis 6 h after 43°C heating for 30 min
Wyllie, Kerr and Currie reunited the same year (1980) to prepare an updated apoptosis review for the International Review of Cytology [20]. Significant new developments, stressed for the first time in this review, were the importance of apoptosis in immune killing and the finding that active mRNA and protein synthesis seemed to be required in many (but not all) circumstances for apoptosis to proceed. New gene products and control mechanisms of apoptosis were also proposed as being important future areas of research. Further applications of apoptosis were being recognized at this time. For example, a chapter was written by Kerr and Searle for inclusion in Radiation biology in cancer research [8]. Up until this point, most of the work on apoptosis had been published by either Kerr or Wyllie’s groups. Having greatly extended the known range of occurrence of apoptosis, having a striking biochemical marker for the process (the DNA ‘ladder’) and some evidence that it was indeed an active type of cell death marked the end of the first era in the development of the apoptosis concept. 1.6 Applications and Mechanisms 1981–1990 The next phase of the apoptosis story was to be marked by the gradual entry of new investigators into the field and significant advances in our understanding of the role that apoptosis plays both in the immune system and in cancer. However, for the first couple of years after the 1980 publications appeared, little seemed
12
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
to be happening. One event worthy of mention was the publication, in 1981, of the first book dealing specifically with cell death to include a chapter on apoptosis [see 41]. Kerr and his group continued with studies further defining the incidence of apoptosis and its importance in medically and biologically important processes, and in 1985 suggested that the fact that apoptotic bodies are phagocytosed and degraded without extracellular release of their contents might be of considerable importance in containing viral infections [42]. Wyllie and colleagues on the other hand were more focussed on studies of the apoptotic mechanism. They showed, among other things, that the cellular condensation that accompanies apoptosis was accompanied by an increase in density [43] and that changes occurring in the nature of carbohydrates exposed on the surface of apoptotic bodies might be responsible for their rapid phagocytosis [44]. The late 1970s and early 1980s had seen a number of publications showing that apoptosis was the mode of cell death involved in T, K and NK cell killing [reviewed in 10] and it was the role of apoptosis in the immune system more than anything else that was to be the major driving force behind apoptosis research in the 1980s. John Cohen and Richard Duke, two immunologists from Colorado, USA, were among the first of the new investigators to enter the field. They published a number of apoptosis papers in the early to mid-1980s on topics which included inhibition of apoptosis by mRNA and protein synthesis inhibitors [45], mechanisms of immune killing [46, 47], and the induction of apoptosis in IL-2 dependent T cells by withdrawal of growth factor [48]. As mentioned earlier, the apoptosis concept had been embraced far earlier in Britain than in the United States, and the papers by Cohen and Duke were welcomed by others working in the field as evidence that, at last, American scientists were beginning to see the importance of this concept. In the cancer field, a notable development in 1984 was the finding that apoptosis was markedly enhanced in preneoplastic foci and nodules developing in liver after the administration of carcinogens [49, 50]. One of the most significant findings however, came in 1987 when Wyllie and his colleagues showed that increased apoptosis in tumours could result from processes intrinsic to the tumour cells themselves, different levels of apoptosis being found in otherwise similar fibroblast tumours expressing different oncogenes [51]. This was the first study implicating oncogenes in the regulation of apoptosis and as such was the forerunner of one of the most exciting areas of apoptosis research in the 1990s, the role of oncogenes and tumour suppressor genes in regulation of apoptosis. However, at the time its significance was not widely appreciated. By 1987, many new scientists had entered the field and apoptosis papers were being published more widely. Important contributions were made by many of these investigators and are detailed in a number of apoptosis reviews and book chapters published around this time [21–23, 52]. Nevertheless, despite the increasing interest in the field, many presentations of apoptosis papers at scientific conferences were still subject to bitter attack and submitted papers were still rejected out of hand by some referees. Towards the end of the 1980s, interesting new findings relating to the apoptotic mechanism began to appear, including, among other things, the roles of tissue type transglutaminase [53], elevated cytosolic Ca2 + levels [54], gene expression [55–57], tumour promoters [58] and the APO-1 antibody [59]. A novel approach to the study of the genetic control of programmed cell death was the use of Caenorhabditis elegans [60]. The findings that generated most interest around this time however, were those relating to the physiologic roles that apoptosis subserves in the immune system [reviewed by 61, 62]. It was shown that apoptosis was responsible for deletion of autoreactive T-cells in the thymus during the development of selftolerance [63] and for the selection of B-cells in lymphoid germinal centres during the humoral immune responses [64]. The first international meeting to highlight apoptosis, Modulating factors in multistage chemical carcinogenesis, was held in Sardinia in September, 1989. The following year a meeting designed
APOPTOSIS: A 20TH CENTURY SCIENTIFIC REVOLUTION
13
specifically to explore molecular approaches to the mechanism and regulation of apoptosis was organized by David Tomei and Fred Cope and held at the prestigious Cold Spring Harbor Laboratory. Many of the participants were invited to contribute to a book Apoptosis: the molecular basis of cell death, which was published soon after [65]. The search was now well under way for putative ‘death genes’, but the breakthrough, when it came, was not a ‘death gene’ but one that prevented death, the bcl-2 gene [66]. In what is undoubtedly one of the most significant apoptosis publications since Wyllie’s DNA endonuclease paper [40] some 10 years earlier, Hockenberry et al. [66] showed that bcl-2 operated in a completely different way from other oncogenes that promote mitosis by blocking apoptosis. Vaux et al. [67] had reported two years earlier that bcl-2 cooperated with another oncogene c-myc to immortalize pre-B cells but had not shown that it did so by blocking apoptosis. It now seems clear that prolonged survival of cells overexpressing bcl-2 is a factor in predisposition to malignancy [68, 69, and reviewed in 25] and there is now evidence that bcl-2 can greatly increase the resistance of tumours to anticancer drugs [70–72]. The bcl-2 findings and results published soon after, showing that p53 tumour suppressor gene might play a role in regulating apoptosis [73], generated enormous interest, and knowledge of the mechanisms involved in its regulation has grown rapidly in the last few years. Apoptosis is intimately involved in the pathogenesis of cancer, AIDS and autoimmune disease, and studies into its role in chronic neurodegenerative diseases and even ageing are now well under way. Implicit in the concept that apoptosis is an active gene-directed process of self-destruction is the possibility that it might be able to be selectively regulated. The ability to control apoptosis in patients suffering from these diseases would revolutionize approaches to treatment. 1.7 Apoptosis—a Scientific Revolution Scientific work is said by Kuhn [74] to come in two sorts of historical episodes. The first termed ‘normal science’ is characterized by extended periods of investigation into various aspects of accepted paradigms. These extended periods are punctuated by brief periods of ‘revolutionary science’ which arise when researchers recognize a significant body of experimental or observational anomalies that will not fit within the existing paradigm and so challenge its validity. Usually, the challenge thrown out by these observational anomalies fails to make an impact or is rebuffed and the status quo is maintained. The ‘apoptosis revolution’ began in 1972 and it was to be a further 20 years before it was finally successful. Articulation of the apoptosis concept led initially to hostile rejection by adherents of the existing paradigm that had dominated since the time of Virchow. The failure of most scientists and journal editors to support the new concept and the lack of research funding are expected parts of scientific revolution [74]. Why was the ‘apoptosis revolution’ successful? Firstly, Kerr, Wyllie and Currie and the small group of scientists they were able to gather around them were almost passionate in their belief that apoptosis was an extremely important concept. As such, they were able to push on with apoptosis research through some very difficult years. Secondly, with the passing of time, striking applications of the concept were extended into an ever increasing number of areas, thus attracting a new generation of scientists into the apoptosis field. At the same time, more and more of the older, conservative adversaries were retiring, creating a less hostile environment. Finally, the discovery that apoptosis could be regulated by certain oncogenes and the p53 tumour suppressor gene opened up the possibility of finding the means of controlling its occurrence. Apoptosis has now become ‘normal science’. The following chapters will explore some of the exciting new findings on apoptosis in normal development and in cancer. The challenge for the future is to ensure that the careful morphological observations and good scientific method that underpinned the early
14
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
development of the apoptosis concept are maintained in the new era of studies of the molecular mechanisms of apoptosis. Acknowledgements We thank Clay Winterford for his excellent technical assistance with electron microscopy, artwork and photography. The work was supported by a Research Encouragement Grant from the Queensland University of Technology. Footnote about Kerr, Wyllie and Currie John F.R.Kerr graduated in medicine from the University of Queensland, Australia in 1958 and completed his PhD at the University of London in 1964 under the supervision of Sir Roy Cameron. He has been Professor of Pathology at the University of Queensland since 1974 and his recent review ‘Apoptosis: its significance in cancer and cancer therapy’ [25] will be one of his last as he retired at the end of 1994. Andrew H.Wyllie graduated in medicine from the University of Aberdeen in 1967 and completed his PhD at the same university in 1975 under the supervision of Professor Alastair Currie. He completed postdoctoral training at Cambridge in England before rejoining Alastair Currie at Edinburgh University. He has been Professor of Experimental Pathology at Edinburgh University since 1992 and continues to work at the forefront of apoptosis research. Alistair R.Currie graduated in medicine from the University of Glascow in 1944. He served as Professor of Pathology at the Imperial Cancer Research Fund Laboratories in Oxford, the University of Aberdeen and the University of Edinburgh and chaired a number of major scientific committees in the United Kingdom. He was awarded a knighthood for his services to medical sciences in 1979. His interest in apoptosis continued in retirement until his death early in 1994. Note about the authors: Brian Harmon worked as a Research Assistant for John Kerr from 1968 to 1974 (the period spanning the first apoptosis paper) and both authors undertook PhDs under the supervision of John Kerr at the University of Queensland Pathology Department in the 1980s. References 1 2 3 4 5 6 7 8
. KERR, J.F.R., WYLLIE, A.H. & CURRIE, A.R. (1972) Apoptosis: a basic biological phenomenon with wide ranging implications in tissue kinetics. Br. J. Cancer 26, 239–257. . KERR, J.F.R. (1994) Apoptosis—past and future. In: Milestones of Australian medicine, J.PEARN (ed). Amphion Press, Brisbane, pp. 153–166. . CAMERON, G.R. (1952) Pathology of the cell. Oliver & Boyd, Edinburgh. . SINGER, C. & UNDERWOOD, E.A. (1962) A short history of medicine, 2nd edn. Oxford University Press, London. . VIRCHOW, R. (1858) Die Cellularpathologie in ihrer Begründung auf physiologische und pathologische Gewebelehre. A.Hirschwald, Berlin. . KERR, J.F.R. (1971) Shrinkage necrosis: a distinct mode of cellular death. J. Pathol. 105, 13–20. . SWARTZENDRUBER, D.C. & CONGDON, C.C. (1963) Electron microscope observations on tingible body macrophages in mouse spleen. J. Cell Biol. 19, 641–646. . KERR, J.F.R. & SEARLE, J. (1980) Apoptosis: its nature and kinetic role. In: Radiation biology in cancer research, MEYN, R. & WITHERS, R. (eds). Raven Press, New York, pp. 367–384.
APOPTOSIS: A 20TH CENTURY SCIENTIFIC REVOLUTION
9 10
11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
28 29
30
31
32
15
. WEEDON, D., SEARLE, J. & KERR, J.F.R. (1979) Apoptosis. Its nature and implications for dermatopathology. Am. J. Dermatopathol. 1, 133–144. . KERR, J.F.R. & HARMON, B.V. (1991) Definition and incidence of apoptosis: an historical perspective. In: Apoptosis: the molecular biology of cell death, TOMEI, L.D. & COPE, F.O. (eds). Cold Spring Harbor Press, New York, pp. 5–29. . KERR, J.F.R. (1965) A histochemical study of hypertrophy and ischaemic injury of rat liver with special reference to changes in lysosomes. J. Pathol. Bact. 90, 419–435. . KERR, J.F.R. (1969) An electron-microscope study of liver cell necrosis due to heliotrine. J. Pathol. 97, 557–562. . KERR, J.F.R. (1970) An electron microscopic study of liver cell necrosis due to albitocin. Pathology 2, 251–259. . KERR, J.F.R. & SEARLE, J. (1972) A suggested explanation for the paradoxically slow growth rate of basal-cell carcinomas that contain numerous mitotic figures. J. Pathol. 107, 41–44. . IVERSEN, O.H. (1967) Kinetics of cellular proliferation and cell loss in human carcinomas. A discussion of methods available for in vivo studies. Int. J. Radiat. Biol. 46, 609– 623. . REFSUM, S.B. & BERDAL, P. (1967) Cell loss in malignant tumours in man. Eur. J. Cancer 3, 235–236. . STEEL, G.G. (1967) Cell loss as a factor in the growth rate of human tumours. Eur. J. Cancer 3, 381–387. . SEARLE, J., COLLINS, D., HARMON, B. & KERR, J.F.R. (1973) The spontaneous occurrence of apoptosis in squamous carcinomas of uterine cervix. Pathology 5, 163–169. . GLÜCKSMANN, A. (1951) Cell deaths in normal vertebrate ontogeny. Biol. Rev. 26, 59–86. . WYLLIE, A.H., KERR, J.F.R. & CURRIE, A.R. (1980) Cell death: the significance of apoptosis. Int. Rev. Cytol. 68, 251–306. . KERR, J.F.R., SEARLE, J., HARMON, B.V. & BISHOP, C.J. (1987) Apoptosis. In: Perspectives on mammalian cell death, POTTEN, C.S. (ed). Oxford University Press, Oxford, pp. 93–126. . WYLLIE, A.H. (1987) Cell death. Int. Rev. Cytol Supp. 17, 755–785. . WALKER, N.I., HARMON, B.V., GOBÉ, G.C. & KERR, J.F.R. (1988) Patterns of cell death. Meth. Achiev. Exp. Pathol. 13, 18–54. . ARENDS, M.J. & WYLLIE, A.H. (1991). Apoptosis: mechanisms and roles in pathology. Int. Rev. Exp. Pathol. 32, 223–254. . KERR, J.F.R., WINTERFORD, C.M. & HARMON, B.V. (1994) Apoptosis: its significance in cancer and cancer therapy. Cancer 73, 2013–2026. . HINCHLIFFE, J.R. (1981) Cell death in embryogenesis. In: Cell death in biology and pathology, BOWEN, I.D. & LOCKSHIN, R.A. (eds). Chapman and Hall, London, pp. 35–78. . KERR, J.F.R., HARMON, B. & SEARLE, J. (1974) An electron microscope study of cell deletion in the anuran tadpole tail during spontaneous metamorphosis with particular reference to apoptosis of striated muscle fibres. J. Cell Sci. 14, 571–585. . KERR, J.F.R. & SEARLE, J. (1973) Deletion of cells by apoptosis during castration-induced involution of the rat prostate. Virchows Arch. B Cell Pathol. 13, 87–102. . SEARLE, J., LAWSON, T.A., ABBOTT, P.J., HARMON, B. & KERR, J.F.R. (1975) An electron microscope study of the mode of cell death induced by cancer-chemotherapeutic agents in populations of proliferating normal and neoplastic cells. J. Pathol. 116, 129–138. . SEARLE, J., KERR, J.F.R., BATTERSBY C., EGERTON, W.S., BALDERSON, G. & BURNETT, W. (1977). An electron microscopic study of the mode of donor cell death in unmodified rejection of pig liver allografts. Aust. J. Exp. Biol. Med. Sci. 55, 401–406. . DON, M.M., ABLETT, G., BISHOP, C.J., BUNDESEN, P.G., DONALD, K.J., SEARLE, J. & KERR, J.F.R. (1977) Death of cells by apoptosis following attachment of specifically allergized lymphocytes in vitro. Aust. J. Exp. Biol. Med. Sci. 55, 407–417. . SANDERSON, C.J. & GLAUERT, A.M. (1977) The mechanism of T cell mediated cytotoxicity. V Morphological studies by electron microscopy. Proc. R. Soc. Ser. B 198, 315–323.
16
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49
50
51
52 53 54
55
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. WYLLIE, A.H., KERR, J.F.R. & CURRIE, A.R. (1973) Cell death in the normal neonatal rat adrenal cortex. J. Pathol. 111, 255–261. . WYLLIE, A.H., KERR, J.F.R., MACASKILL, I.A.M. & CURRIE, A.R. (1973) Adrenocortical cell deletion: the role of ACTH. J. Pathol. 111, 85–94. . WYLLIE, A.H. (1974) Death in normal and neoplastic cells. J. Clin. Pathol. Suppl. (R. Coll. Pathol.) 7, 35–42. . POTTEN, C.S. (1977) Extreme sensitivity of some intestinal crypt cells to X and gamma irradiation. Nature 269, 518–521. . CRAWFORD, A.M., KERR, J.F.R. & CURRIE, A.R. (1972) The relationship of acute mesodermal cell death to the teratogenic effects of 7-OHM-12-MBA in the foetal rat. Br. J. Cancer 26, 498–503. . KERR, J.F.R. (1972) Shrinkage necrosis of adrenal cortical cells. J. Pathol. 107, 217– 219. . KERR, J.F.R. & SEARLE, J. (1972) The digestion of cellular fragments within phagolysosomes in carcinoma cells. J. Pathol. 108, 55–58. . WYLLIE, A.H. (1980) Glucocorticoid-induced thymocyte apoptosis is associated with endogenous endonuclease activation. Nature 284, 555–556. . WYLLIE, A.H. (1981) Cell death: a new classification separating apoptosis from necrosis. In: Cell death in biology and pathology, BOWEN, I.D. & LOCKSHIN, R.A. (eds). Chapman and Hall, London, pp. 9–33. . CLOUSTON, W.M. & KERR, J.F.R. (1985) Apoptosis, lymphocytotoxicity and the containment of viral infections. Med. Hypotheses 18, 399–404. . WYLLIE, A.H. & MORRIS, R.G. (1982) Hormone-induced cell death. Purification and properties of thymocytes undergoing apoptosis after glucocorticoid treatment. Am. J. Pathol. 109, 78–87. . DUVALL, E., WYLLIE, A.H. & MORRIS, R.G. (1985) Macrophage recognition of cells undergoing programmed cell death (apoptosis). Immunology 56, 351–358. . COHEN, J.J. & DUKE, R.C. (1984) Glucocorticoid activation of a calcium-dependent endonuclease in thymocyte nuclei leads to cell death. J. Immunol. 132, 38–42. . DUKE, R.C., CHERVENAK, R. & COHEN, J.J. (1983) Endogenous endonuclease-induced DNA fragmentation: an early event in cell-mediated cytolysis. Proc. Natl. Acad. Sci. USA 80, 6361–6365. . COHEN, J.J., DUKE, R.C., CHERVENAK, R., SELLINS, K.S. & OLSON, L.K. (1985) DNA fragmentation in targets of CTL: an example of programmed cell death in the immune system. Adv. Exp. Med. Biol. 184, 493–508. . DUKE, R.C. & COHEN, J.J. (1986) IL-2 addiction: withdrawal of growth factor activates a suicide program in dependent T cells. Lymphokine Res. 5, 289–295. . BURSCH, W., LAUER, B., TIMMERMANN-TROSIENER, I., BARTHEL, G., SCHUPPLER, J. & SCHULTE-HERMANN, R. (1984) Controlled death (apoptosis) of normal and putative preneoplastic cells in rat liver following withdrawal of tumor promoters. Carcinogenesis 5, 453–458. . COLUMBANO, A., LEDDA-COLUMBANO, G.M., RAO, P.M., RAJALAKSHMI, S. & SARMA, D.S.R. (1984) Occurrence of cell death (apoptosis) in preneoplastic and neoplastic liver cells. Am. J. Pathol. 116, 441–446. . WYLLIE, A.H., ROSE, K.A., MORRIS, R.G., STEEL, C.M., FOSTER, E. & SPANDIDOS, D.A. (1987) Rodent fibroblast tumours expressing human myc and ras genes: growth, metastasis and endogenous oncogene expression. Br. J. Cancer 56, 251– 259. . ALLAN, D.J., HARMON, B.V. & KERR, J.F.R. (1987) Cell death in Spermatogenesis. In: Perspectives on mammalian cell death, POTTEN, C.S. (ed). Oxford University Press, Oxford, pp. 229–258. . FESUS, L., THOMAZY, V. & FALUS, A. (1987) Induction and activation of tissue transglutaminase during programmed cell death. FEBS Lett. 224, 104–108. . MCCONKEY, D.J., HARTZELL, P., DUDDY, S.K., HAKANSSON, H. & ORRENIUS, S. (1988) 2, 3, 7, 8Tetrachlorodibenzo-p-dioxin kills immature thymocytes by Ca2+— mediated endonuclease activation . Science 242, 256–259. . BUTTYAN, R., ZAKERI, Z., LOCKSHIN, R. & WOLGEMUTH, D. (1988) Cascade induction of c-fos, c-myc, and heat shock 70 K transcripts during regression of the rat ventral prostate gland. Mol. Endocrinol. 2, 650–657.
APOPTOSIS: A 20TH CENTURY SCIENTIFIC REVOLUTION
56 57 58
59
60 61 62 63
64 65 66 67 68
69 70 71 72
73 74
17
. WADEWITZ, A.G. & LOCKSHIN, R.A. (1988) Programmed cell death: dying cells synthesise a co-ordinated, unique set of proteins in two different episodes of cell death. FEBS Lett. 241, 19–23. . BUTTYAN, R., OLSSON, C.A., PINTAR, J., CHANG, C., BANDYK, M., NG, P.-Y. & SAWCZUK, I.S. (1989) Induction of the TRPM-2 gene in cells undergoing programmed death. Mol. Cell Biol. 9, 3473–3481. . KIZAKI, H., TADAKUMA, T., ODAKA, C., MURAMATSU, J. & ISHIMURA, Y. (1989) Activation of a suicide process of thymocytes through DNA fragmentation by calcium ionophores and phorbol esters. J. Immunol. 143, 1790–1794. . TRAUTH, B.C., KLAS, C. PETERS, A.M.J., MATZKU, S., MOLLER, P., FALK, W., DEBATIN, K.-M. & KRAMMER, P.H. (1989) Monoclonal antibody-mediated tumor regression by induction of apoptosis. Science 245, 301–305. . YUAN, J. & HORVITZ, H.R. (1990) The Caenorhabditis elegans genes ced-3 and ced-4 act cell autonomously to cause programmed cell death. Dev. Biol. 138, 33–41. . COHEN, J.J. (1991) Programmed cell death in the immune system. Adv. Immunol. 50, 55–85. . GOLSTEIN, P., OJCIUS, D.M. & YOUNG, J.D.-E. (1991) Cell death mechanisms and the immune system. Immunol. Rev. 121, 29–65. . SMITH, C.A., WILLIAMS, G.T., KINGSTON, R., JENKINSON, E.J. & OWEN, J.J.T. (1989) Antibodies to CD3/T-cell receptor complex induce cell death by apoptosis in immature T cells in thymic cultures. Nature 337, 181–184. . LIU, Y.-J., JOSHUA, D.E., WILLIAMS, G.T., SMITH, C.A., GORDON, J. & MACLENNAN, I.C.M. (1989) Mechanism of antigen-driven selection in germinal centres. Nature 342, 929–931. . TOMEI, L.D. & COPE, F.O. (eds) (1991) Apoptosis: the molecular biology of cell death. Cold Spring Harbor Press, New York. . HOCKENBERY, D.M., NUNEZ, G., MILLIMAN, C., SCHREIBER, R.D. & KORSMEYER, S.J. (1990) Bcl-2 is an inner mitochondrial membrane protein that blocks programmed cell death. Nature 348, 334–336. . VAUX, D.L., CORY, S. & ADAMS, J.M. (1988) Bcl-2 gene promotes haemopoietic cell survival and cooperates with c-myc to immortalize pre-B cells. Nature 335, 440–442. . HOCKENBERY, D.M., ZUTTER, M., HICKEY, W., NAHM, M. & KORSMEYER, S.J. (1991) BCL2 protein is topographically restricted in tissues characterized by apoptotic cell death. Proc. Natl Acad. Sci. USA 88, 6961–6965. . KORSMEYER, S.J. (1992) Bcl-2 initiates a new category of oncogenes: regulators of cell death. Blood 80, 879–886. . COLLINS, M.K. L., MARVEL, J., MALDE, P. & LOPEZ-RIVAS, A. (1992) Interleukin 3 protects murine bone marrow cells from apoptosis induced by DNA damaging agents. J. Exp. Med. 176, 1043–1051. . LOTEM, J. & SACHS, L. (1992) Hematopoietic cytokines inhibit apoptosis induced by transforming growth factor beta1 and cancer chemotherapy compounds in myeloid leukemic cells. Blood 80, 1750–1757. . MIYASHITA, T. & REED, J.C. (1992) Bcl-2 gene transfer increases relative resistance of S49.1 and WEHI7.2 lymphoid cells to cell death and DNA fragmentation induced by glucocorticoids and multiple chemotherapeutic drugs. Cancer Res. 52, 5407–5411. . YONISH-ROUACH, E., RESNITZKY, D., LOTEM, J., SACHS, L., KIMCHI, A. & OREN, M. (1991) Wildtype p53 induces apoptosis of myeloid leukaemic cells that is inhibited by interleukin-6. Nature 352, 345–347. . KUHN, T.S. (1970) The structure of scientific revolutions, 2nd edn. University of Chicago Press, Chicago.
2 The Enzymology of Apoptosis FRANCIS M.HUGHES, Jr and JOHN A.CIDLOWSKI Molecular Endocrinology Group, Laboratory of Integrative Biology, National Institute of Environmental Health Sciences, P.O. Box 12233, Research Triangle Park, NC 27709, USA
2.1 Introduction Apoptosis, or programmed cell death, is a critical process in many areas of biology such as morphological modeling of tissues during development, removal of autoreactive immune cells, hormonal or age-related tissue atrophy and normal cell turnover. During apoptosis, cells, by virtue of their own genetic makeup, die and are deleted for the overall good of the organism. This process occurs in many plant and animal species, in diverse cell types and in response to a large number of different inducing agents. However, programmed cell death is not the only process by which cells die. Cells may also die by a catastrophic process called necrosis which usually results from physical trauma to the cell causing the plasma membrane to be compromised. While necrosis is used by organisms in a few physiological processes, such as complementmediated lysis, its occurrence is rare and apoptosis is, by far, the predominant form of physiological cell death. In all species studied to date, apoptotic cells display a remarkably conserved set of events that encompass specific morphological and biochemical alterations. When a cell becomes phenotypically apoptotic it detaches from its neighboring cells and shrinks, resulting in a loss of intracellular volume. Electron microscopic analysis has shown that commensurate shrinkage occurs in all organelles, with the exception of mitochondria, which maintain a normal appearance [1]. Apoptosis is an energy-requiring process, and functional mitochondria are necessary to provide adequate energy for the process to reach completion. The chromatin condenses and eventually the entire nucleus fragments and disperses as vesicles within the cytoplasm. The rapid loss of fluid within an apoptotic cell results in a crenalated plasma membrane that eventually pinches or ‘buds’ off into small spherical structures, known as apoptotic bodies, which contain bits of cytoplasm, organelles and pieces of the fragmented nuclei. These bodies (which eventually constitute the entire cell) are recognized and endocytosed by neighboring cells or resident macrophages where they are degraded and the macromolecules recycled. In contrast to the changes seen in apoptosis, necrosis is characterized by a gross swelling of the cell and its organelles, which is apparently precipitated by the loss of integrity of the plasma membrane. The influx of water eventually causes the cell to rupture, spilling its contents into the surrounding fluid. One of the hazards of necrotic death is that exposure of intracellular constituents to the extracellular milieu will precipitate an inflammatory immune response. Apoptosis avoids
ENZYMOLOGY OF APOPTOSIS
19
this potentially dangerous side-effect by encapsulating all intracellular components in membrane bound vesicles that are resistant to leakage (i.e. apoptotic bodies). The necessity of cell death to the proper functioning of an organism is indisputable and clearly apparent in many of the documented examples of apoptosis. The goal of apoptosis is, therefore, to remove the selected cells in a timely and orderly fashion without arousing the immune system or causing other potentially harmful side-effects. In this review we will discuss the enzymes which act as motors to drive the progression of apoptosis. Undoubtedly many enzymes are involved in the signal transduction cascades which relay the various apoptotic signals; however, these enzymes are likely to be stimulus- and/or cellspecific. Rather than discuss individual signalling pathways, attention in this review will be focused on those enzymes that constitute the common pathways in all cells as they undergo apoptosis. 2.2 Proteases Intuitively, a whole-cell catabolic phenomenon such as apoptosis is quite likely to employ a number of proteases which destroy proteins and recycle amino acids late in the apoptotic process. However, recent data suggest that proteases may also play a causative role in the apoptotic process, perhaps even in the triggering mechanism. A clear example of the involvement of proteolysis in apoptosis is found in the cellular cytotoxicity of the immune system. Cytotoxic cells (cytotoxic T lymphocytes, and natural killer cells— referred to collectively as cytotoxic lymphocytes or CL) kill their targets by vectorially secreting the contents of granules which deliver a ‘lethal hit’ to the target cell, ultimately triggering the target cell’s own endogenous apoptotic program. The involvement of proteolysis in this form of cellular killing was first suggested by Chang and Eisen in 1980 [2] when they observed that N-tosyl-L-lysyl-chloromethylketone (TLCK), a serine protease inhibitor, prevented the death of target cells attacked by CL. However, at that time, little was known about apoptosis, and it had not been appreciated that CL kill their target cells by this process. This clue was nearly forgotten when the first granule protein, perforin, was isolated and cloned [3–5]. Monomers of perforin were shown to assemble in the target cell membrane to form a large waterfilled pore which appeared to be freely and non-specifically permeable to ions [6–8]. It was hypothesized that these pores allowed water to enter the cell, eventually causing an osmotic lysis in a manner homologous to that seen with complement (i.e. necrosis). While this scenario was intellectually pleasing, target cells were known to undergo DNA fragmentation typical of apoptosis, and subsequent studies showed that purified perforin alone was unable to evoke this particular component of the apoptotic response [9], although it would evoke membrane lysis and loss of viability. Further purification of granule components have since identified a host of different constituent proteins (primarily proteases) potentially involved in CLmediated cytotoxicity [10]. Thus, the role of perforin may simply be to allow these molecules access to the cytoplasm and/or nucleus of the target cell where they exert their deadly effects [10–13]. These perforin pores may also allow the influx of Ca2+ which is used in downstream events such as nuclease activity. The CL granule proteases, which currently number at least 12 discrete proteins, are collectively known as ‘granzymes’ and can be grouped into three main categories based on their substrate specificity [10]. These enzymes are ser-ases (cleaves after serine residues; examples are granzymes A, D, E, F, H, BLT-esterase), asp-ases (cleaves after aspartic acid; examples are granzyme B, fragmentin, cathepsin-D) and met-ases (cleaves after methionine; the only identified member is Met-ase). Although preliminary evidence has suggested that granzymes can activate membrane receptors [14], the majority of data suggest that granzymes act intracellularly in target cells to activate the apoptotic program. For example, purified granzyme A [15, 16] or fragmentin [12, 17] can cause DNA fragmentation and apoptosis only when applied
20
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
with sublytic concentrations of perforin, suggesting that diffusion of these proteases through perforinformed pores and the cleavage of intracellular proteins are both necessary steps in the induction of CLmediated killing. While the evidence appears clear that proteolysis is involved in CL-induced target cell apoptosis, the question remains as to whether this enzymatic activity is important in other models of apoptosis. Early work from our laboratory suggested that glucocorticoids could stimulate proteolysis [18]. For example, steroid treatment of lymphocytes was associated with an increase in a 17 kd serine pro tease [19]. While the identity of this protease is unknown, several other studies have examined a role for proteases in apoptosis through the use of inhibitors. For example, Kaufman et al. [20] found that the fragmentation of DNA and the morphological alterations seen during apoptosis in HL-60 cells treated with etoposide (a topoisomeraseII inhibitor) or -irradiation, or rat thymocytes treated with glucocorticoids, could be delayed by a variety of protease inhibitors such as N-ethylmaleimide, iodoacetamide, p-chloromercuribenzenesulfate, N-tosyl-Lphenylalanine-chloromethylketone (TPCK) and TLCK. Importantly, other inhibitors (phenylmethylsulfonyl fluoride, phosphoramidone and antipain) were unable to prevent the apoptosis, suggesting some degree of specificity. Interestingly cell death in these studies was never prevented in the inhibitor-treated cells but only delayed. A loss of viability in the absence of apoptotic characteristics was also seen in cells treated with inhibitor alone (i.e. no apoptotic stimuli) suggesting that, while the extended inhibition of proteolysis can delay apoptosis, this treatment eventually causes a necrotic demise. Similar results were obtained by Bruno et al. [21] using a topoisomerase-I inhibitor as the apoptotic agent in these cells. Protease inhibitors also prevent tumor necrosis factor-induced cytotoxicity [22–24] suggesting that the use of proteolysis in the apoptotic program may be widespread and part of a underlying common process. While these results are intriguing, protease inhibitors are notoriously non-specific and may affect many cellular processes. In an effort to narrow the spectrum of potential proteases, Squier et al. [25] has examined the effects of an inhibitor specific for calpain, a calcium-dependent neutral protease. This inhibitor prevented both glucocorticoid and -irradiation-induced apoptosis of thymocytes as well as cycloheximide-induced apoptosis of metamylocytes. Studies of this type are beginning to elucidate the identity of the specific endogenous proteases involved in apoptosis. Calpain may play such a role in many, although maybe not all models [25]. This field is not without controversy, however, since another study [26] presents contradictory data with a T cell hybridoma model system. In this study cys-protease inhibitors, ser-protease inhibitors and calpain inhibitors all prevented anti-TCR or anti-thy-1-induced apoptosis but enhanced the DEX-induced death of these cells. It is not clear whether these results are due to the use of a hybridoma, whose physiology is likely to stray far from that of a normal cell. Consistent with other reports, these investigators did demonstrate that the TCR-induced death of normal peripheral T cells was blocked by the protease inhibitors but, unfortunately, did not examine the steroid effects on these cells. However, another study showed that leupeptin, a calpain inhibitor, failed to block glucocorticoid-induced death of thymocytes [27]. Perhaps future studies will address the apparent discrepancies among these model systems. Identification of the specific proteases involved in apoptosis may lead to a better understanding of their mechanism of action and provide us with better tools with which to explore this mechanism. Currently several attractive scenarios exist for a critical role for proteases in apoptosis. For example, we have previously proposed that the nuclease responsible for internucleosomal cleavage of DNA during apoptosis is maintained in non-dying cells in an inactive high-molecular-weight complex, possibly via interaction with a specific proteinaceous inhibitor [28]. One potential mechanistic role for proteolysis in the induction of apoptosis is to degrade this putative inhibitor, freeing the nuclease and activating the apoptotic cascade. Another possibility would be the degradation of ADP-ribosyltransferase which may lead to derepression of the nuclease (see below).
ENZYMOLOGY OF APOPTOSIS
21
2.3 Nucleases By far the most studied enzymatic activity associated with apoptosis is the nuclease activity responsible for cleavage of DNA in the internucleosomal regions [29, 30]. Degradation of the genome in this manner has received considerable attention recently and is clearly important in the overall process. In addition to internucleosomal cleavage of DNA, several studies have found that dying cells also cleave their DNA into large (50–300 kb) fragments [31–34] and it has been further suggested that such fragments may provide the substrate for subsequent internucleosomal cleavage [33, 34]. Finally, one type of nuclease activity that has not been well investigated is the potential for RNase activity in apoptosis. Interestingly, in an apoptotic cell the rate of RNA degradation has been shown to increase [35] while specific mRNAs may increase, decrease or remain unchanged [36–41]. The degradation of genomic DNA in an apoptotic cell by cleavage in the internucleosomal regions has been detected by literally hundreds of independent investigators in a vast number of different experimental models (see [42] and references therein). The near universal correlation of this activity with apoptosis would seem to underscore its importance in the overall process. In contrast, most models of necrosis are characterized by the random degradation of DNA without apparent regard to chromatin structures [43–45]. During necrosis, DNA degradation is accompanied by simultaneous activity of proteases, perhaps released by the osmotic disruption of lysososmes. Proteases degrade the histones and other proteins that maintain chromatin structure, exposing the entire length of DNA to subsequent degradation by necrotic nucleases. The form that a particular cell death takes (apoptotic or necrotic) can be distinguished by analysis of DNA integrity on an agarose gel. When the gel was stained with ethidium bromide, the internucleosomal cleavage of DNA during apoptosis will be seen as a ‘ladder’ pattern with each specific band corresponding to DNA fragments which are multiples of the basic nucleosomal unit (180–200 bp). In contrast, the random cleavage of DNA seen during necrosis appears as a smear on these gels. Thus, the ability to distinguish between necrosis and apoptosis using the integrity of DNA as a biochemical parameter is an important tool in this field and is best used to complement morphological characterization. It should be noted that two recent studies [46, 47] have claimed to detect internucleosomal fragmentation during necrosis; however, the models employed may not be exclusively necrotic in nature. In the first study [46], the cell population contained a low level of morphologically apoptotic cells, and the second study [47] also displays several classically apoptotic morphological changes such as cell shrinkage and detachment from neighboring cells. Interestingly, a model of necrosis not displaying these morphological changes also did not possess internucleosomal fragments of DNA [47]. Together, the available evidence has failed to conclusively identify internucleosomal cleavage of DNA in necrosis, and therefore this type of cleavage would appear to remain an integral component of apoptosis. Almost 18 years of research have passed since internucleosomal DNA cleavage was first described [29] and associated with apoptosis [30], yet the causative role for this process and its mechanism of action are still unknown and actively debated. Activation of nuclease activity has been shown to occur prior to morphological changes typical of apoptosis [30] and its activity has been proposed to be the first irreversible event which leads to a cell’s inevitable demise [48–50]. Intuitively it would seem that extensive degradation of the genome would overwhelm all repair processes and ensure the imminent death of the cell. DNA fragmentation might then precipitate or ‘trigger’ the resulting morphological changes. Indeed, if this cleavage is mimicked by treating thymocyte nuclei with micrococcal nuclease, the chromatin condenses in a manner that recapitulates that seen in an apoptotic cell [50]. Moreover, preventing DNA fragmentation during apoptosis by altering chromatin structure with the polyamine spermine has been shown to prevent subsequent morphological changes [51]. In contrast, other studies [52] suggest that activation of
22
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
internucleosomal cleavage is not associated with the chromatin condensation stage of the morphological changes. Indeed, several reports have also dissociated internucleosomal DNA fragmentation and morphological alterations by finding cell lines and experimental conditions which induce an apoptotic morphology in the absence of oligonucleosomal DNA fragments [52–56]. These interesting exceptions may eventually provide us with some insight into whether redundant systems may be involved in apoptosis. Such systems would ensure the ultimate goal of removing the cell, even if one step in the process (for example internucleosomal cleavage of DNA) is inactive due to mutation, etc. At this time no laboratory has conclusively identified the nuclease responsible for internucleosomal cleavage during apoptosis, although several candidate molecules have been proposed. Several attempts have been made to purify a Ca2+/Mg2+— dependent internucleosomal cleavage enzyme from various tissues such as thymus [57, 58], porcine and rat liver [59, 60] and bull seminal plasma [61], although very few of these efforts have resulted in purification to homogeneity. In separate studies, Hashida et al. [61] purified a 36 kd nuclease from bull seminal plasma while, more recently, Ribeiro and Carson [62] purified a 22–26 kd protein from human spleen. In the latter study the purified protein was shown to posses internucleosomal cleavage activity. Recently, our laboratory [63] isolated an 18 kd nuclease (NUC18) active in apoptotic thymocytes but inactive in control cells. NUC18 activity appears with the same kinetics as both 50 kb and oligonucleosomal DNA fragments in vivo, is induced in a steroid-specific manner, is present in the nucleus, and is dependent upon Ca2+ and Mg2+ for activity. In addition, this enzyme is inhibited by Zn2+ and aurintricarboxylic acid, substances known to inhibit many examples of apoptosis. In non-apoptotic cells, this enzyme is maintained in an inactive high-molecular-weight complex of >100 kd, while the native molecular weight of the active enzyme in apoptotic cells appears to be 25 kd (18 kd under denatured reducing conditions). Finally, NUC18 has a pH optimum in the physiological range (7.0–8.5). This enzyme has recently been purified to homogeneity and partially sequenced [64]. Comparison of sequence data revealed a high homology to cyclophilin A, the intracellular binding protein for the immunosuppressant drug cyclosporin A. Cyclosporin A is a widely used drug credited with the prevention of rejection in organ transplant patients, although considerable toxicity has also been noted. This surprising homology between NUC18 and cyclophilin A implicates the cyclophilin family of related proteins in the endonucleolytic activity seen in apoptosis. Prior to these findings cyclophilins were only known to possess a cis-trans peptidyl-prolyl isomerase (rotamase) activity which led to the proposal that they were primarily important in protein folding [65, 66]. However, recent studies have shown [67] that the alteration of rotamase activity could not account for the immunosuppressive effects of this drug. Thus, identification of a potential new function for cyclophilin as a nuclease should lead to a reevaluation of how cyclosporin A exerts its immunosuppressive effects. Our laboratory has shown [64] that purified, recombinant and even crystallized and resolubilized cyclophilin A possesses inherent nuclease activity that is enhanced by clyclosporin A. Together, these results identify NUC18 as a member of the cyclophilin family of immunophilins and implicate this protein family in the DNA catabolism associated with apoptosis. Ongoing studies are aimed at further clarifying the role of cyclophilins in apoptosis and determining their mechanism of regulation. In addition to NUC18 and the cyclophilins, at least two other known proteins, DNase I and II, have been implicated in the internucleosomal cleavage underlying at least some models of apoptosis. DNase I has been proposed by two independent laboratories [54, 68, 69] as a candidate enzyme. This enzyme can be activated by Mg2+ alone [70] but is primarily dependent on Ca2+ and Mg2+ and is inhibited by Zn2+ but not by aurintricarboxylic acid [70]. Interestingly, the molecular sequence of DNase I predicts a secreted enzyme [71] and, indeed, immunological data have failed to find the enzyme in the nucleus where it would be required for apoptosis. Rather, studies have localized it to the cytoplasm, apparently associated with the endoplasmic reticulum [69]. It has been proposed [70], however, that dissolution of the ER during apoptosis
ENZYMOLOGY OF APOPTOSIS
23
may allow diffusion of DNase I into the nucleus to cause internucleosomal cleavage fragmentation. Given the highly efficient and organized nature of apoptosis, it seems unlikely that this process would employ random diffusion as the primary method of ensuring DNA degradation. Moreover, when incubated with intact chromatin, DNase I does not degrade this substrate specifically into oligonucleosomal fragments but cleaves randomly, generating a smear on an agarose gel similar to that observed during necrosis [69]. In these studies, specificity for internucleosomal regions could only be imparted by an unidentified ‘internucleosomal specificity factor’ in serum. Perhaps future studies will address these inconsistencies and provide further insight into DNase I as a possible apoptotic nuclease. Work from another laboratory has provided yet another candidate for apoptotic DNA degradation in the acidic enzyme DNase II. Investigators implicated this enzyme in the DNA degradation of apoptosis after they failed to detect a Ca2+/ Mg2+ -dependent nuclease in apoptotic Chinese hamster ovary cells (CHO cells). These investigators did, however, measure an acidic activity seemingly equivalent to the known enzyme DNase II [72]. Purified DNase II was able to cleave DNA in the internucleosomal regions but was not inhibited by inhibitors of apoptosis such as Zn2+ and aurintricarboxylic acid. Moreover, cleavage by this nuclease generates 5 -hydroxyl ends on the DNA fragments, while apoptosis has been associated with the production of 3 -hydroxyl ends [73, 74]. Previously, DNase II was thought to be primarily a lysosomal enzyme where its pH optimum (5.5) would be ideally suited for activity. Although the authors did detect a decrease in intracellular pH during apoptosis [72], the lowest pH an apoptotic cell could be expected to achieve is 6.3–6.6, a level barely permissible for DNase II activity. Thus, for a cell to use this nuclease so far outside its optimal pH range, the cell would have to produce enormous amounts of protein to have sufficient activity to cause the extensive and rapid DNA degradation seen during apoptosis. Finally, no conclusive evidence has been presented to demonstrate that DNase II is present in the nucleus where it might cause the DNA degradation during apoptosis. Given the primarily lysosomal location of DNase II, it seems plausible that the role of DNase II in apoptosis is in the neighboring cell or macrophage where it functions to degrade DNA in the endocytosed apoptotic bodies. The studies presented above demonstrate that a number of different laboratories have attempted to isolate the enzyme responsible for internucleosomal cleavage during apoptosis, and out of these studies several candidates have been proposed. In addition to the single activities detected above, it should be noted that other studies have identified multiple nuclease activities in apoptotic cells. For example, Nikonova et al. [75] characterized three independent nuclease activities in radiation-induced apoptotic thymocytes from rats: (1) a 40 kd Mn2+ -activated enzyme (DNase I is 40 kd); (2) a 37 kd acidic nuclease (DNase II is 38–44 kd), and (3) a 22-kd Ca2+/Mg2+ -dependent enzyme (NUC18 and cyclophilins are 18–25 kd). Thus, it is possible that individual signalling pathways activate specific nucleases which precipitate an apoptotic cascade; however, such tailor-making of enzyme to a particular signal would surely amount to a great cost to the cell. Therefore, it is likely that, if multiple enzymes can bring an apoptotic genome to an appropriate end, then the number of such readily available nucleases is small. Given the redundancy of biological systems, and the importance of maintaining a functional apoptotic pathway in a given cell, such a limited but multiple enzyme hypothesis should be considered. While the cleavage of DNA into oligonucleosome fragments during apoptosis is well established, investigators have recently discovered that DNA is also degraded into very large fragments centering primarily around 50 kb and 300 kb in size [32–34, 76]. Although generation of such DNA fragments was documented as early as 1979 by Kokileva [31], it was not linked to apoptosis at that time. Recently, pulsedfield gel electrophoresis has allowed researchers to reliably separate very large fragments of DNA [77], and studies of chromatin structure [78] have shown that the lowest level of chromatin organization is the nucleosome, which is composed of DNA wrapped twice around a core group of histones. These
24
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
nucleosomes are then coiled together into solenoids which form a structure known as the 30 nm fiber, which can be visualized in the electron microscope. The 30 nm fiber is attached to the nuclear matrix through identified but imprecisely defined matrix attachment regions (MARs) [79], where it ‘loops’ out and back to the matrix where it is attached by the next MAR, approximately 50 kb away. Six loops encircle the matrix, each offset by approximately 60°, forming a structure resembling a ‘rosette’. The appearance of DNA fragments of 300 kb and 50 kb in length in apoptotic cells appears to correspond to cleavage of these loop and rosette structures [34]. Initiation of apoptotic DNA degradation has been proposed [33, 34] to begin with the cleavage of these structures, although it does appear that some cells do not generate the 300 kb fragments [34]. Furthermore, these large fragments have been suggested to provide the substrate for internucleosomal cleavage [33]. While such a scenario is intellectually logical, it should be noted that these conclusions are based solely on data showing that DNA fragmentation into large fragments precedes internucleosomal cleavage. Unfortunately, in these studies comparisons are made between ethidium bromide-stained pulsed-field and conventional agarose gels, and considerably more ethidium bromide is bound to a 50 kb DNA molecule than to a 200 bp species. Thus, the sensitivity of large fragment detection is significantly greater than that of small fragments, and a precursor-product relationship between large fragments and small fragments must remain tenuous. The number of systems in which this form of DNA degradation has been examined is limited, although large DNA fragments have been seen in a few instances where internucleosomal DNA degradation was not detected [34, 80]. One must remember that internucleosomal cleavage of DNA was also initially found in all cells examined and required extensive searching to identify the exceptions discussed above. Thus, the universal nature of this form of DNA degradation remains a very tentative conclusion. Another important question currently being considered in many laboratories is whether both forms of chromatin degradation (into small and large fragments) arise from the action of a single nuclease or multiple nucleases (clearly a question similar to that addressed throughout this section). Although no conclusive experiments have been published, recent data [33, 76] have provided evidence for multiple enzymes based on differential ionic requirements. Specifically, the generation of oligonucleosomal fragments during in vitro DEX-induced apoptosis of rat thymocytes [33] was apparently suppressed by 1 mM Zn2+ while the accumulation of large fragments was increased, presumably by preventing their further internucleosomal degradation. While intriguing, these results suffer from the same sensitivity issue as discussed above and could result from a single nuclease (which generates large fragments first and then uses these as substrate to produce small fragments) whose activity was severely reduced but not completely abrogated by this concentration of Zn2+. Indeed, it takes significantly fewer (250-fold less) cuts in chromatin to generate a 50 kb fragment than it takes for a 200 bp fragment. Thus, reducing the activity of a single nuclease with Zn2+, over a limited timeframe, coupled with the enhanced sensitivity of large fragment detection, may reveal large DNA fragments while precluding the detection of small fragments. In addition, Sun and Cohen [76] have shown that isolated thymocyte nuclei incubated with the correct ions can spontaneously generate large DNA fragments. Such a phenomenon (known as autodigestion) has previously been documented with the internucleosomal fragmentation of DNA and is interpreted to be the result of ionic activation of a constitutively present enzyme [27, 81–84]. When these experiments were carried out in the presence of Mg2 + alone the accumulation of large DNA fragments was clear while the authors state that no oligonucleosomal fragments were seen. While these results also suggest separate nucleases that mediate each form of DNA degradation, caution must again be exercised because of sensitivity issues and the fact that autodigestion in isolated nuclei may not adequately represent the in vivo apoptotic process. One type of nuclease activity which has not been well characterized during apoptosis is the action of RNases. Very little has been done on this subject despite its potential importance. Older data from our
ENZYMOLOGY OF APOPTOSIS
25
laboratory [35] showed that overall RNA degradation is increased in glucocorticoid-stimulated apoptotic rat thymocytes which is a result one might expect given the catabolic nature of apoptosis. However, the glucocorticoid-stimulated death of thymocytes requires both transcription and translation for lethality. Thus at least some RNA degradation must be staved off during the earlier portions of the process. Indeed, the fate of specific transcripts may be highly variable and may possibly rely on the necessity of their role downstream in the apoptotic process. For example, in renal cell apoptosis EGF mRNA levels rapidly decrease while TGF- levels increase [39]. Other investigators have identified messages that show early transient increases and decreases in later stages of apoptosis [37, 40, 41]. While the role of such fluctuations is unknown, these studies suggest an intricate system of controls over RNA degradation From the work presented in this section it is clear that nucleases, specifically endodeoxyribonucleases, represent an extensively studied enzymatic activity associated with apoptosis. Despite recent controversy regarding the requirement of DNA degradation in apoptosis, it remains an early and near-universal phenomenon, a fact which would appear to underscore its importance. Perhaps as interest in this process increases so too will our knowledge of the nuclease activities involved and how they fit into the cause and effect relationships that drive this process. 2.4 ADP-ribosyltransferase ADP-ribosyltransferase (ADPRT) is an enzyme whose role in apoptosis may range from the triggering event in the activation of the apoptotic nuclease and degradation of the genome [85] to only being incidentally activated by DNA breaks as if to prepare for a repair process [86]. ADPRT transfers the ADP-ribose moiety of NAD+ to nuclear proteins [87] and is found in nearly all types of cell [88, 89], placing it in a good position for part of the common apoptotic pathway. We have previously shown that the ADPRT inhibitors aminobenzamide and methylnicotinomide can directly enhance apoptosis alone while accentuating the effects of glucocorticoids on apoptosis in thymocytes. In addition, others have shown [85] that C-nitroso compounds (6-nitroso-1, 2-benzopyrone and 3-nitrosobenzamide) that inhibit ADPRT promote apoptosis in leukemic cells and other types of malignant cell. In these studies, evidence was presented that the Ca2+/Mg2 + -dependent nuclease in these cells is maintained in an inactive state by polymers of ADP-ribose. Derepression of this enzyme at the induction of apoptosis was brought about by inactivation of ADPRT. In this regard, others have shown [20] that ADPRT is cleaved early in apoptosis of HL-60 cells induced by several chemotherapeutic agents, suggesting a proteolytic mechanism of inactivation. Cleavage and inactivation of ADPRT (followed by derepression of a nuclease and degradation of the genome) would certainly be consistent with the early role for proteolysis discussed above. Interestingly, ADPRT-deficient cell lines were found to be hypersensitive to some apoptotic agents but resistant to others [90]. Thus, the potential exists for this enzyme to play an important role in apoptosis. 2.5 Lipid Modifying Enzymes The regulation of lipid structure in an apoptotic cell has not been extensively studied. While one may not, a priori, expect lipid alterations to be involved in apoptosis, the few studies that have focused on lipids have detected a dramatic loss of plasma membrane phospholipid asymmetry that results in the exposure of phosphatidylserine on the cell’s surface [91–94]. The aberrantly revealed phospholipid is subsequently recognized by unidentified phosphatidylserine receptors on macrophages which trigger the rapid
26
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
phagocytosis of the apoptotic cell. Normally, phosphatidylcholine and sphingomyelin predominate on the extracellular side of the membrane which allows for fairly tight packing of the phospholipid head groups. In contrast, phosphatidylserine and phosphatidylethanolamine are found primarily on the inner leaflet, endowing this side of the membrane with a looser packing [95, 96]. In at least some types of cell this asymmetry is maintained by an ATP-driven phospholipid translocase [94, 97, 98]. Therefore, the induction of apoptosis might in some way inactivate or adversely affect this translocase, bringing about the loss of phospholipid asymmetry and extracellular localization of phosphatidylserine. Besides a direct inactivation of the translocase by post-translational modifications, other possibilities for disabling the translocase may exist. The actual mechanism involved is unknown and further studies are needed to resolve this issue. The importance of membrane lipid alterations in apoptosis was also suggested in a recent study [94] which showed that membrane changes occur prior to DNA degradation. Whether any causal relationship exists between these two events remains to be explored; however, such early labeling of a cell destined for apoptosis may promote its rapid recognition and degradation in vivo, preventing the accumulation of apoptotic bodies which may leak their contents. In vitro, where there are no phagocytic cells, these apoptotic bodies do lose their membrane integrity in a process known as secondary necrosis. Rapid recognition and removal of apoptotic cells in vivo ensures that this process of secondary necrosis occurs nearly exclusively in culture. Interestingly, there are also in vivo models in which apoptotic bodies accumulate, such as in atretic ovarian follicles. In this model, accidental leakage of ovarian cellular components might precipitate an autoimmune reaction. Interestingly, premature ovarian failure is associated with an autoimmune etiology [99– 101], suggesting the intriguing possibility that defective lipid alterations during apoptosis may have a previously unrecognized clinical relevance. While the use of lipid or membrane modifications to promote phagocytosis of apoptotic cells is quite interesting, it may not be the only means of recognition. Indeed, macrophages have been shown to recognize apoptotic cells through specific carbohydrates, the vitronectin receptor, thrombospondin and other mechanisms [102–108]. However, it is not clear whether changes in membrane lipids may aid or in other ways be involved in these alternative mechanisms. 2.6 Transglutaminases When an apoptotic cell begins its morphological degeneration it shrinks and ‘blebs off’ into the small, typically spherical apoptotic bodies. It has been shown in a number of in vitro and in vivo models [109–117] that these structures are stabilized by extensive protein cross-links to the point that they are resistant to dissolution by detergent or chaotropic agents [118], a process similar to that of cornification of terminally differentiated keratinocytes [119]. The cross-links in these structures are formed by the specific activity of the enzyme tissue transglutaminase which catalyzes the formation of ( -glutamyl)lysine isopeptide bonds and some -glutamyl-bis-spermidine cross-links [112]. These linkages are so resistant to proteolysis that the engulfing cell is unable to degrade them and eventually expels the ( -glutamyl)-lysine isodipeptide after degrading the surrounding protein. This peptide is detectable in the medium from cultured cells undergoing apoptosis [120] and is also found in normal rat serum where its levels are increased by experimentally inducing apoptosis in the liver [121, 120]. In addition to activation of constitutively expressed transglutaminase, apoptosis stimulates the synthesis of de novo protein. Several studies have noted increases in both protein and mRNA for tissue transglutaminase [109, 110, 112, 115, 116]. While the precise role of this enzyme in apoptosis has not been decisively ascertained, such cross-linking would reduce the risk of
ENZYMOLOGY OF APOPTOSIS
27
leaking intracellular constituents from the apoptotic bodies [118] and thereby protect against inflammation and autoimmune diseases. Studies employing molecular biological approaches have provided evidence for such a role for tissue transglutaminase in apoptosis. Evidence from Fesus’s laboratory (reported in [118]) suggests that transfection of cultured granulocytes with antisense oligonucleotides to tissue transglutaminase leads to increased cellular fragility and leakage of intracellular proteins. Conversely, when tissue transglutaminase is overexpressed in BALB/c 3T3 fibroblasts [122] the cells showing the highest level of expression often demonstrated extensive membrane blebbing and cellular fragmentation in a manner reminiscent of apoptosis. However, these cells did not display condensed chromatin or the nuclear fragmentation characteristic of apoptotic cells. Thus, high expression of transglutaminase was only partially able to recapitulate the properties normally seen in apoptosis. Taken together, the available data are consistent with transglutaminase playing a role in apoptosis by cross-linking proteins in apoptotic bodies to form a rigid protein scaffold that stabilizes these bodies and protects against leakage of the intracellular components into the extracellular space. Although transglutaminase activity is a late event in the apoptotic cascade, activation of this enzyme may become the first universally accepted step in the common pathway. Clearly the formation of apoptotic bodies is one of the few changes associated with apoptosis that is not controversial. Therefore, the mechanisms that govern the formation of these structures might truly be part of a common pathway. To our knowledge, increased transglutaminase parameters (mRNA, protein and/or activity) has been reported in every example of apoptosis in which it has been examined [118]. 2.7 Conclusion In this review we have discussed some of the enzymatic activities associated with apoptosis. Certainly in a complex phenomenon such as apoptosis, many enzymes are involved and our list cannot be considered exhaustive. Together, it appears that these, and probably other, enzymes form a cascade which begins with signal transduction pathways and converges on one (or a few) triggering events which activate the changes characteristic of apoptosis. In this cascade, stimulation of an enzyme will precipitate downstream events but will be unable to activate upstream changes (e.g. transglutaminase activation leads to cell blebbing, etc. but is unable to cause upstream nuclear fragmentation). This cascade may also bypass a step that has been inactivated by mutation to allow the ultimate goal of removing the cell. With the realization of the importance of apoptosis in many physiological and pathological processes, work on the motors that drive this process should progress quickly. Acknowledgements This work was supported by DK32078, DK09150 and HD07315. References 1 2
. WYLLIE, A.H., KERR, J.R. & CURRIE, A.R. (1980) Cell death: the significance of apoptosis. Int. Rev. Cytol. 68, 251–306. . CHANG, T.W. & EISEN, H.E. (1980) Effects of N-tosyl-L-lysyl-chloromethylketone on the activity of cytolytic T lymphocytes. J. Immunol. 124, 1028–1033.
28
3 4 5 6 7 8 9 10 11 12 13
14 15 16 17
18 19 20
21
22 23 24 25
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. LICHTENHELD, M.G., OLSEN, K.J., LU, P., LOWREY, D.M., HAMEED, A., HENGARTNER, H. & PODACK, E.R. (1988) Structure and function of human perforin. Nature 335, 448–451. . SHINKAI, Y., TAKIO, K. & OKUMURA, K. (1988) Homology of perform to the ninth component of complement (C9). Nature 334, 525–527. . TSCHOPP, J. & NABHOLZ, M. (1990) Perforin-mediated target cell lysis by cytolytic T lymphocytes. Annu. Rev. Immunol. 8, 279–302. . HENKART, P.A. (1985) Mechanism of lymphocyte-mediated cytotoxicity. Annu. Rev. Immunol. 3, 31–58. . PODACK, E.R. (1985) Molecular mechanism of lymphocyte-mediated tumor cell lysis. Immunol. Today 6, 21–27. . TSCHOPP, J. & JONGENEEL, C.V. (1988) Cytotoxic T lymphocyte mediated cytolysis. Biochemistry 27, 2641–2646. . DUKE, R.C., PERSECHINI, P.M., CHANG, S., LIU, C.C., COHEN, J.J. & YOUNG, J.D. (1989) Purified perform induces target cell lysis but not DNA fragmentation. J. Exp. Med. 170, 1451–1456. . TRAPANI, J.A. & SMYTH, M.J. (1993) Killing by cytotoxic T cells and natural killer cells: multiple granule serine proteases as initiators of DNA fragmentation. Immunol. Cell Biol. 71, 201–208. . PODACK, E.R., HENGARTNER, H. & LICHTENHELD, M.G. (1991) A central role of perform in cytolysis? Annu. Rev. Immunol. 9, 129–157. . SHI, L., KRAUT, R.P., AEBERSOLD, R. & GREENBERG, A.H. (1992) A natural killer cell granule protein that induces DNA fragmentation and apoptosis. J. Exp. Med. 175, 553–566. . SMYTH, M.J., BROWNE, K.A., THIA, K.Y., APOSTOLIDIS, V.A., KERSHAW, M.H. & TRAPANI, J.A. (1994) Hypothesis: cytotoxic lymmphocyte granule serine proteases activate target cell endonucleases to trigger apoptosis. Clin. Exp. Pharmacol. Physiol. 21, 67–70. . TSCHOPP, J., SCHAERER, E. & SUIDAN, H.S. (1993) Granules of cytolytic T lymphocytes: a multitask organelle. Abstract in: EMBO Workshop on Cell-Mediated Cytotoxicity, Neve Ilan, Israel. . HAYES, M.P., BERREBI, G.A. & HENKART, P.A. (1989) Introduction of target cell DNA release by the cytotoxic T lymphocyte granule protease granzyme A. J. Exp. Med. 170, 933–946. . SHIVER, J.W., SU, L. & HENKART, P.A. (1992) Cytotoxicity with target DNA breakdown by rat basophilic leukemia cells expressing both cytolysin and granzyme A . Cell 71, 315–322. . ALLBRITTON, N.L., VERRET, C.R., WOLLEY, R.C. & EISEN, H.N. (1988) Calcium ion concentrations and DNA fragmentation in target cell destruction by murine cloned cytotoxic T lymphocytes. J. Exp. Med. 167, 514–527. . MACDONALD, R.G., MARTIN, T.P. & CIDLOWSKI, J.A. (1980) Glucocorticoids stimulate protein degradation in lymphocytes: a possible mechanism of steroid-induced death. Endocrinology 107, 1512–1524. . MACDONALD, R.G. & CIDLOWSKI, J.A. (1981) Glucocorticoid regulation of two serine hydrolases in rat splenic lymphocytes in vitro. Biochim. Biophys. Acta 678, 18–26. . KAUFMANN, S.H., DESNOYERS, S., OTTAVIANO, Y., DAVIDSON, N.E. & POIRIER, G.G. (1993) Specific proteolytic cleavage of poly (ADP-ribose) polymerase: an early marker of chemotherapy-induced apoptosis. Cancer Res. 53, 3976–3985. . BRUNO, S., DEL-BINO, G., LASSOTA, P., GIARETTI, W. & DARZYNKIEWICZ, Z. (1992) Inhibitors of proteases prevent endonucleolysis accompanying apoptotic death of HL-60 leukemic cells and normal thymocytes. Leukemia 6, 1113–1120. . RUGGIERO, V., JOHNSON, S.E. & BAGLIONI, C. (1987) Protection from tumor necrosis factor cytotoxicity by protease inhibitors. Cell. Immunol. 107, 317–325. . SUFFYS, P., BEYAERT, R., VAN-ROY, F. & FIERS, W. (1988) Involvement of a serine protease in tumournecrosis-factor-mediated cytotoxicity. Eur. J. Biochem. 178, 257–265. . KUMAR, S. & BAGLIONI, C. (1991) Protection from tumor necrosis factor-mediated cytolysis by overexpression of plasminogen activator inhibitor type-2. J. Biol. Chem. 266, 20960–20964. . SQUIER, M.K., MILLER, A.C., MALKINSON, A.M. & COHEN, J.J. (1994) Calpain activation in apoptosis. J. Cell. Physiol. 159, 229–237.
ENZYMOLOGY OF APOPTOSIS
26
27 28
29 30 31 32 33 34
35 36
37 38
39
40 41
42 43 44 45 46 47
29
. SARIN, A., ADAMS, D.H. & HENKART, P.A. (1993) Protease inhibitors selectively block T cell receptortriggered programmed cell death in a murine T cell hybridoma and activated peripheral T cells . J. Exp. Med. 178, 1693–1700. . MCCONKEY, D.J., HARTZELL, P., NICOTERA, P. & ORRENIUS, S. (1989) Calcium-activated DNA fragmentation kills immature thymocytes. FASEB 3, 1843– 1849. . SCHWARTZMAN, R.A. & CIDLOWSKI, J.A. (1993) Mechanism of tissue-specific induction of internucleosomal deoxyribonucleic acid cleavage activity and apoptosis by glucocorticoids. Endocrinology 133, 591–599. . SKALKA, M., MATYÁSOVÁ, J. & CEJKOVÁ, M. (1976) DNA in chromatin of irradiated lymphoid tissues degrades in vivo into regular fragments. FEBS Lett. 72, 271–274. . WYLLIE, A.H. (1980) Glucocorticoid-induced thymocyte apoptosis is associated with endogenous endonuclease activation. Nature 284, 555–556. . KOKILEVA, L.B. (1979) DNA degredation in isolated rat liver nuclei. Molec. Biol. Rep. 5, 171–174. . SZABÓ, G. JR., BOLDOG, F. & WIKONKÁL, N. (1990) Disassembly of chromatin into 50 Kb units by detergent. Biochem. Biophys. Res. Commun. 169, 706–712. . BROWN, D.G., SUN, X.-M. & COHEN, G.M. (1993) Dexamethasone-induced apoptosis involves cleavage of DNA to large fragments prior to internucleosomal fragmentation. J. Biol. Chem. 268, 3037–3039. . OBERHAMMER, F., WILSON, J.W., DIVE, C., MORRIS, I.D., HICKMAN, J.A., WAKELING, A.E., WALKER, P.R. & SIKORSKA, M. (1993) Apoptotic death in epithelial cells: cleavage of DNA to 300 and/or 50 kb fragments prior to or in the absence of internucleosomal fragmentation. EMBO J. 12, 3679–3684. . CIDLOWSKI, J.A. (1982) Glucocorticoids stimulate ribonucleic acid degradation in isolated rat thymic lymphocytes in vitro. Endocrinology 111, 184–190. . DOWD, D.R., MACDONALD, P.N., KOMM, B.S., HAUSSLER, M.R. & MIESFELD, R. (1991) Evidence for early induction of calmodulin gene expression in lymphocytes undergoing glucocorticoid-mediated apoptosis . J. Biol Chem. 266, 18423– 18426. . OWENS, G.D. & COHEN, J.J. (1992) Indentification of genes involved in programmed cell death. Cancer Metastasis Rev. 11, 149–156. . TILLY, J.L., KOWALSKI, K.I., SCHOMBERG, D.W. & HSUEH, A.J. (1992) Apoptosis in atretic ovarian follicles is associated with selective decreases in messenger ribonucleic acid transcripts for gonadotropin receptors and cytochrome P450 aromatase. Endocrinology 131, 1670–1676. . WALTON, G., BUTTYAN, R., GARCIA-MONTES, E., OLSSON, C.A., HENSLE, T. W. & SAWCZUK, I.S. (1992) Renal growth factor expression during the early phase of experimental hydronephrosis. J. Urol. 148, 510–514. . DELIC, J., COPPEY-MOISAN, M. & MAGDELENAT, H. (1993) Gamma-ray-induced transcription and apoptosis-associated loss of 28S rRNA in interphase human lymphocytes. Int. J. Radiat. Biol. 64, 39–46. . GUENETTE, R.S., CORBEIL, H.B., LEGER, J., WONG, K., MEZL, V., MOOIBROEK, M. & TENNISWOOD, M. (1994) Induction of gene expression during involution of the lactating mammary gland of the rat. J. Mol. Endocrinol. 12, 47–60. . SCHWARTZMAN, R.A. & CIDLOWSKI, J.A. (1993) Apoptosis: the biochemistry and molecular biology of programmed cell death. Endocrine Rev. 14, 133–151. . KERR, J.F.R. (1971) Shrinkage necrosis: a distinct mode of cellular death. J. Path. 105, 13–20. . WYLLIE, A.H. (1981) Cell death: a new classification separating apoptosis from necrosis. In: Cell Death in Biology and Pathology, LOCKSHIN, B.A. (eds), Chapman and Hall, London, pp. 9–34. . SEARLE, J., KERR, J.F.R. & BISHOP, C.J. (1982) Necrosis and apoptosis: distinct modes of cell death with fundamentally different significance. Pathol. Annu. 17, 229–259. . COLLINS, R.J., HARMON, B.V., GOBÉ, G.C. & KERR, J.F.R. (1992) Inter-nucleosomal DNA cleavage should not be the sole criterion for identifying apoptosis. Int. J. Radiat. Biol. 61, 451–453. . FUKUDA, K., KOJIRO, M. & CHIU, J.F. (1993) Demonstration of extensive chromatin cleavage in transplanted Morris hepatoma 777 tissue: apoptosis or recrosis. Am. J. Pathol. 142, 935–946.
30
48 49
50 51 52
53
54
55
56 57 58 59 60. 61
62 63
64
65 66 67
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. COMPTON, M.M. & CIDLOWSKI, J.A. (1986) Rapid in vivo effects of glucocorticoids on the integrity of rat lymphocyte genomic deoxyribonucleic acid. Endocrinology 118, 38–45. . MCCONKEY, D.J., HARTZELL, P., AMADOR-PÉREZ, J.F., ORRENIUS, S. & JONDAL, M. (1989) Calcium-dependent killing of immature thymocytes by stimulation via the CD3/T cell receptor complex. J. Immunol. 143, 1801–1806. . ARENDS, M.J., MORRIS, R.G. & WYLLIE, A.H. (1990) Apoptosis: the role of the endonuclease. Am. J. Pathol. 136, 593–608. . BRUNE, B., HARTZELL, P., NICOTERA, P. & ORRENIUS, S. (1991) Spermine prevents endonuclease activation and apoptosis in thymocytes. Exp. Cell. Res. 195, 323– 329. . OBERHAMMER, F., FRITSCH, G., SCHMIED, M., PAVELKA, M. PRINTZ, D., PURCHIO, T., LASSMANN, H. & SCHULTE-HERMANN, R. (1993) Condensation of the chromatin at the membrane of an apoptotic nucleus is not associated with activation of an endonuclease. J. Cell Sci. 104, 317–326. . COHEN, G.M., SUN, X.M., SNOWDEN, R.T., DINSDALE, D. & SKILLETER, D.N. (1992) Key morphological features of apoptosis may occur in the absence of internucleosomal DNA fragmentation. Biochem. J. 286, 331–334. . UCKER, D.S., OBERMILLER, P.S., ECKHART, W., APGAR, J.R., BERGER, N.A. & MEYERS, J. (1992) Genome digestion is a dispensable consequence of physiological cell death mediated by cytotoxic T lymphocytes. Mol Cell. Biol. 12, 3060–3069. . TOMEI, L.D., SHAPIRO, J.P. & COPE, F.O. (1993) Apoptosis in C3H/10T1/2 mouse embryonic cells: evidence for internucleosomal DNA modification in the absence of double-stranded cleavage. Proc. Natl Acad. Sci. USA 90, 853–857. . ZAKERI, Z.F., QUAGLINO, D., LATHAM, T. & LOCHSHIN, R.A. (1993) Delayed internucleosomal DNA fragmentation in programmed cell death. FASEB J. 7, 470–478. . NAKAMURA, M., SAKAKI, Y., WATANABE, N. & TAKAGI, Y. (1981) Purification and characterization of the Ca2+ plus Mg2+ -dependent endodeoxyribonuclease from calf thymus chromatin. J. Biochem. 89, 143–152. . NIKONOVA, L.V., ZOTOVA, R.N. & UMANSKY, S.R. (1990) Isolation of Ca2+/ Mg2+ -dependent nuclease from calf thymus chromatin. Biokhimiya 54, 1397–1405. . ISHIDA, R., AKIYOSHI, H. & TAKAHASHI, T. (1974) Isolation and purification of calcium and magnesium dependent endonuclease from rat liver nuclei. Biochem. Biophys. Res. Commun. 56, 703–710. STRÄTLING, W.H., GRADE, C. & HÖRZ, W. (1984) Ca2+/Mg2+ -dependent endonuclease from porcine liver. Purification, properties and sequence specificity. J. Biol. Chem. 259, 5893–5898. . HASHIDA, T., TANAKA, Y., MATSUNAMI, N., YOSHIHARA, K., KAMIYA, T., TANIGAWA, Y. & KOIDE, S.S. (1982) Purification and properties of bull seminal plasma Ca2+, Mg2+ -dependent endonuclease. J. Biol. Chem. 257, 13114–13119. . RIBEIRO, J.M. & CARSON, D.A. (1993) Ca2+/Mg2+ -dependent endonuclease from human spleen: purification, properties and role in apoptosis. Biochemistry 32, 9129– 9136. . GAIDO, M.L. & GIDLOWSKI, J.A. (1991) Identification, purification and characterization of a calciumdependent endonuclease (NUC18) from apoptotic rat thymocytes: NUC18 is not histone H2B. J Biol Chem 266, 18580–18585. . MONTAGUE, J.W., GAIDO, M.L., FRYE, C. & CIDLOWSKI, J.A. (1994) A calcium-dependent nuclease from apoptotic rat thymocytes is homologous with cyclophilin: recombinant cyclophilins A and B exhibit nuclease activity. J. Biol. Chem. 269, 18877–18880. . FISCHER, G., WITTMANN-LIEBOLD, B., LANG, K., KIEFHABER, T. & SCHMID, F.X. (1989) Cyclophilin and peptidyl-prolyl cis-trans isomerase are probably identical proteins. Nature 337, 476–478. . TAKAHASHI, N., HAYANO, T. & SUZUKI, M. (1989) Pedtidyl-prolyl cis-trans isomerase is the cyclosporin A-binding protein cyclophilin. Nature 337, 473–475. . SIGAL, N.H., DUMONT, F., DURETTE, P., SIEKIERKA, J.J., PETERSON, L., RICH, D.H., DUNLAP, B.E., STARUCH, M.J., MELINO, M.R., KOPRAK, S.L., WILLIAMS, D., WITZEL, B. & PISANO, J.M. (1991) Is
ENZYMOLOGY OF APOPTOSIS
68 69
70 71 72 73
74 75 76 77 78 79 80
81 82 83 84 85
86
87 88
31
cyclophilin involved in the immunosuppressive and nephrotoxic mechanism of action of cyclosporin A? J. Exp. Med. 173, 619–628. . PEITSCH, M.C., HESTERKAMP, T., POLZAR, B., MANNHERZ, H.G. & TSCHOPP, J. (1992) Functional characterization of serum DNase I in MRL-lpr/lpr mice. Biochem. Biophys. Res. Commun. 186, 739–745. . PEITSCH, M.C., POLZAR, B., STEPHAN, H., CROMPTON, T., MACDONALD, H.R., MANNHERZ, H.G. & TSCHOPP, J. (1993) Characterization of the endogenous deoxyribonuclease involved in nuclear DNA degradation during apoptosis (programmed cell death). EMBO J. 12, 371–377. . PEITSCH, M.C., MANNHERZ, H.G. & TSCHOPP, J. (1994) The apoptosis endonucleases: cleaning up after cell death? Trends Cell Biol. 4, 37–41. . POLZAR, B. & MANNHERZ, H.G. (1990) Nucleotide sequence of a full length cDNA clone encoding the deoxyribonuclease I from the rat parotid gland. Nucleic Acids Res. 18, 7151. . BARRY, M.A. & EASTMAN, A. (1993) Identification of deoxyribonuclease II as an endonuclease involved in apoptosis. Arch. Biochem. Biophys. 300, 440–450. . ALNEMRI, E.S. & LITWACK, G. (1990) Activation of internucleosomal DNA cleavage in human CEM lymphocytes by glucocorticoid and novobiocin. Evidence for a non-Ca2+ -requiring mechanism(s). J. Biol Chem. 265, 17323–17333. . GAVRIELI, Y., SHERMAN, Y. & BEN-SASSON, S.A. (1992) Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J. Cell Biol. 119, 493–501. . NIKONOVA, L.V., BELETSKY, I.P. & UMANSKY, S.R. (1993) Properties of some nuclear nucleases of rat thymocytes and their changes in radiation-induced apoptosis. Eur. J. Biochem. 215, 893–901. . SUN, X.-M. & COHEN, G.M. (1994) Mg2+ -dependent cleavage of DNA into kilobase pair fragments is responsible for the initial degradation of DNA in apoptosis. J. Biol Chem. 269, 14857–14860. . BIRREN, B. & LAI, E. (1993) Pulsed field gel electrophoresis: a practical guide. Academic Press, New York. . FILIPSKI, J., LEBLANC, J., YOUDALE, T., SIKORSKA, M. & WALKER, P.R. (1990) Periodicity of DNA folding in higher order chromatin structures. EMBO J. 9, 1319–1327. . BOULIKAS, T. (1993) Nature of DNA sequences at the attachment regions of genes to the nuclear matrix. J. Cell Biochem. 52, 14–22. . ORMEROD, M.G., O’NEILL, C.F., ROBERTSON, D. & HARRAP, K.R. (1994) Cisplatin induces apoptosis in a human ovarian carcinoma cell line without concomitant internucleosomal degradation of DNA. Exp. Cell. Res. 211, 231–237. . HEWISH, D.R. & BURGOYNE, L.A. (1973) Chromatin substructure. The digestion of chromatin DNA at regularly spaced sites by a nuclear deoxyribonuclease. Biochem. Biophys. Res. Commun. 52, 504–510. . COHEN, J.J. & DUKE, R.C. (1984) Glucocorticoid activation of a calcium-dependent endonuclease in thymocyte nuclei leads to cell death. J. Immunol. 132, 38–42. . JONES, D.P., MCCONKEY, D.J., NICOTERA, P. & ORRENIUS, S. (1989) Calcium-activated DNA fragmentation in rat liver nuclei. J. Biol. Chem. 264, 6398–6403. . NIETO, M.A. & LOPEZ-RIVAS, A. (1989) IL-2 protects T lymphocytes from glucocorticoid-induced DNA fragmentation and cell death. J. Immunol. 143, 4166–70. . RICE, W.G., HILLYER, C.D., HARTEN, B., SCHAEFFER, C.A., DORMINY, M., LACKEY, D.A. III, KIRSTEN, E., MENDELEYEV, J., BUKI, K.G., HAKAM, A. & KUN, E. (1992) Induction of endonucleasemediated apoptosis in tumor cells by C-nitroso-substituted ligands of poly (ADP-ribose) polymerase. Proc. Natl Acad. Sci. USA 89, 7703–7707. . CHAUDUN, E., ARRUTI, C., COURTOIS, Y., FERRAG, F., JEANNY, J.C., PATEL, B.N., SKIDMORE, C., TORRIGLIA, A. & COUNIS, M.F. (1994) DNA strand breakage during physiological apoptosis of the embryonic chick lens: free 3 OH end single strand breaks do not accumulate even in the presence of a cationindependent deoxyribonuclease. J. Cell Physiol. 158, 354–364. . HAYAISHI, O. & UEDA, K. (1982) ADP-ribosylation reactions. Academic Press, New York. . IKAI, K., UEDA, K. & HAYAISHI, O. (1980) Immunohistochemical demonstration of poly (adenosine diphosphate-ribose) in nuclei of various rat tissues. J. Histochem. Cytochem. 28, 670–676.
32
89 90 91
92
93 94
95 96 97
98
99 100 101 102 103
104 105 106
107
108
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. COUNIS, M.-F., CHAUDUN, E. & COURTOIS, Y. (1985) Nuclear ADP-ribosylation in the chick lens during embryonic development. Biochem. Biophys. Res. Commun. 126, 859–866. . CHATTERJEE, S., CHENG, M.F. & BERGER, N.A. (1990) Hypersensitivity to clinically useful alkylating agents and radiation in poly (ADP-ribose) polymerase-deficient cell lines. Cancer Commun. 2, 401–407. . FADOK, V.A., VOELKER, D.R., CAMPBELL, P.A., COHEN, J.J., BRATTON, D.L. & HENSON, P.M. (1992) Exposure of phosphatidylserine on the surface of apoptotic lymphocytes triggers specific recognition and removal by macrophages. J. Immunol. 148, 2207–2216. . FADOK, V.A., LASZLO, D.L., NOBLE, P.W., WEINSTEIN, L., RICHES, D.W. H. & HENSON, P.M. (1993) Particle digestibility is required for induction of the phosphatidylserine recognition mechanism used by murine macrophages to phagocytose apoptotic cells. J. Immunol. 151, 4274–4285. . SCHLEGEL, R.A., STEVENS, M., LUMLEY-SAPANSKI, K. & WILLIAMSON, P. (1993) Altered lipid packing identifies apoptotic thymocytes. Immunol Lett. 36, 283–288. . MOWER, D.A. JR., PECKHAM, D.W., ILLERA, V.A., FISHBAUGH, J.K., STUNZ, L.L. & ASHMAN, R.F. (1994) Decreased membrane phospholipid packing and decreased cell size precede DNA cleavage in mature mouse B cell apoptosis. J. Immunol. 152, 4832–4842. . VAN HOEVEN, R.P. & EMMELOT, P. (1972) Studies on plasma membranes. XVIII. Lipid class composition of plasma membranes isolated from rat and mouse liver and hepatomas. J. Membr. Biol. 9, 105–126. . ZWALL, R.F.A., ROELOFSEN, B. & COLLEY, C.M. (1973) Localization of red cell membrane constituents. Biochem et Biophys Acta 300, 159–187. . ZACHOWSKI, A., FAVRE, E., CRIBIER, S., HERVE, P. & DEVAUX, P.F. (1986) Outside-inside translocation of aminophospholipids in the human erythrocyte membrane is mediated by a specific enzyme [published erratum appears in Biochemistry 1986 Nov 18; 25(23): 7788]. Biochemistry 25, 2585–2590. . BEVERS, E.M., TILLY, R.H., SENDEN, J.M., COMFURIUS, P. & ZWAAL, R.F. (1989) Exposure of endogenous phosphatidylserine at the outer surface of stimulated platelets is reversed by restoration of aminophospholipid translocase activity. Biochemistry 28, 2382–2387. . COULAM, C.B., KEMPERS, R.D. & RANDELL, R.V. (1981) Premature ovarian failure: evidence for the autoimmune mechanism. Fertil. Steril. 36, 238–240. . ALPER, M.M. & GARNER, P.R. (1985) Premature ovarian failure: its relationship to autoimmune disease. Obstet. Gynecol. 66, 27–30. . LABARBERA, A.R., MILLER, M.M., OBER, C. & REBAR, R.W. (1988) Autoimmune etiology in premature ovarian failure. Am. J. Reprod. Immunol. 16, 115–122. . DUVALL, E., WYLLLIE, A.H. & MORRIS, R.G. (1985) Macrophage recognition of cells undergoing programmed cell death (apoptosis). Immunology 56, 351–358. . SABOURIN, L.A. & HAWLEY, R.G. (1990) Suppression of programmed cell death and G1 arrest in B-cell hybridomas by interleukin-6 is not accompanied by altered expression of immediate early response genes. J. Cell Physiol. 145, 564–574. . SAVILL, J., DRANSFIELD, I., HOGG, N. & HASLETT, C. (1990) Vitronectin receptor-mediated phagocytosis of cells undergoing apoptosis. Nature 343, 170–173. . DINI, L., AUTUORI, F., LENTINI, A., OLIVERIO, S. & PIACENTINI, M. (1992) The clearance of apoptotic cells in the liver is mediated by the asialoglycoprotein receptor. FEBS Lett. 296, 174–178. . SAVILL, J., HOGG, N., REN, Y. & HASLETT, C. (1992) Thrombospondin cooperates with CD36 and the vitronectin receptor in macrophage recognition of neutrophils undergoing apoptosis . J. Clin. Invest. 90, 1513–1522. . DINI, L., FALASCA, L. LENTINI, A., MATTIOLI, P., PIACENTINI, M., PIREDDA, L. & AUTUORI, F. (1993) Galactose-specific receptor modulation related to the onset of apoptosis in rat liver. Eur. J. Cell Biol. 61, 329–337. . SAVILL, J., FADOK, V., HENSON, P. & HASLETT, C. (1993) Phagocyte recognition of cells undergoing apoptosis. Immunol. Today 14, 131–136.
ENZYMOLOGY OF APOPTOSIS
109 110
111 112
113 114 115
116
117
118 119 120
121
122
33
. FESUS, L., THOMAZY, V. & FALUS, A. (1987) Induction and activation of tissue transglutaminase during programmed cell death. FEBS Lett. 224, 104–108. . SCHULTE-HERMANN, R., BURSCH, W., FESUS, L. & KRAUPP, B. (1987) Cell death by apoptosis in normal, preneoplastic and neoplastic tissues. In Chemical Carcinogenesis, FEO F. et al. (eds). Plenum Press, New York, pp. 263–274. . FESUS, L. & THOMAZY, V. (1988) Searching for the function of tissue transglutaminase: its possible involvement in the biochemical pathway of programmed cell death. Adv. Exp. Med. Biol 231, 119–134. . FESUS, L., THOMAZY, V., AUTUORI, F., CERU, M.P., TARCSA, E. & PIACENTINI, M. (1989) Apoptotic hepatocytes become insoluble in detergents and chaotropic agents as a result of transglutaminase action. FEBS Lett. 245, 150–154. . THOMAZY, V. & FESUS, L. (1989) Differential expression of tissue transglutaminase in human cells. Cell Tissue Res. 255, 215–224. . KNIGHT, C.R., REES, R.C. & GRIFFIN, M. (1991) Apoptosis: a potential role for cytosolic transglutaminase and its importance in tumour progression. Biochim. Biophys. Acta 1096, 312–318. . PIACENTINI, M., AUTUORI, F., DINI, L., FARRACE, M.G., GHIBELLI, L., PIREDDA, L. & FESUS, L. (1991) ‘Tissue’ transglutaminase is specifically expressed in neonatal rat liver cells undergoing apoptosis upon epidermal growth factor stimulation. Cell Tissue Res. 263, 227–235. . PIACENTINI, M., FESUS, L., FARRACE, M.G., GHIBELLI, L., PIREDDA, L. & MELINO, G. (1991) The expression of ‘tissue’ transglutaminase in two human cancer cell lines is related with the programmed cell death (apoptosis). Eur. J. Cell Biol. 54, 246–254. . DAVIES, P.J.A., STEIN, J.P., CHIOCCA, E.A., BASILION, J.P., GENTILE, V., THOMAZY, V. & FESUS, L. (1992) Retinoid-regulated expression of transglutaminases: links to the biochemistry of programmed cell death. In: Retinoids in Normal Development and Ter at agenesis, MORRIS-KAY G. (ed). Oxford University Press, Oxford, pp. 249–263. . FESUS, L., DAVIES, P.J. & PIACENTINI, M. (1991) Apoptosis: molecular mechanisms in programmed cell death. Eur. J. Cell Biol. 56, 170–177. . TARESA, E., KEDEI, N., THOMAZY, V. & FESUS, L. (1992) An involucrin-like protein in hepatocytes serves as a substrate for tissue transglutaminase during apoptosis. J. Biol. Chem. 267, 25648–25651. . FESUS, L., TARCSA, E., KEDEI, N., AUTUORI, F. & PIACENTINI, M. (1991) Degradation of cells dying by apoptosis leads to the accumulation of ( -glutamyl)lysine isodipeptide in culture fluid and blood. FEBS Lett. 284, 109–112. . COLUMBANO, A., LEDDA-COLUMBANO, M., CONI, G., FAA, G., LIGOURI, C., SANTA-CRUZ, G. & PANI, P. (1985) Occurrence of cell death (apoptosis) during the involution of liver hyperplasia. Lab. Invest. 52, 670–679. . GENTILE, V., THOMAZY, V., PIACENTINI, M., FESUS, L. & DAVIES, P.J. (1992) Expression of tissue transglutaminase in Balb-C 3T3 fibroblasts: effects on cellular morphology and adhesion. J. Chem. Biol. 119, 463–474.
3 Molecular Mechanisms of Apoptosis SEI-ICHI TANUMA Department of Biochemistry, Faculty of Pharmaceutical Sciences, Science University of Tokyo, Shinjuku-ku, Tokyo, Japan, and Research Institute for Bioscience, Science University of Tokyo, Noda, Chiba, Japan
3.1 Introduction Cells in multicellular organisms can kill themselves by activating a suicidal genetic program in response to a wide variety of signals, including hormones, cytokines, ionizing radiation, and chemotherapeutic agents. The process of this cell suicide is called ‘apoptosis’, which usually occurs under physiological conditions [1–6]. This type of cell death is considered to be different from necrosis, which is the result of severe injury. The phenomenon of apoptosis has long been known as ‘programmed cell death’, which is fundamental for embryonic development, such as in metamorphosis, morphogenesis and synaptogenesis [5–7]. It also occurs in many adult tissues in normal cell turnover, thymic negative selection, cell mediated cytotoxicity, and so on [1–4, 6, 14, 22–24]. The fate of cells in multicellular organisms is decided by at least two different external signals, i.e. survival and promoting (growth, differentiation, apoptosis and transformation) signals (Figure 3.1). The survival factors are involved in metabolic events for the maintenance of the living cells. The promoting factors divert the cell towards proliferation, differentiation, apoptosis or transformation. Apoptosis can be considered as a suicidal process that eliminates a cell unable to receive survival signals and/or able to receive death signals from other cells. Apoptosis also occurs in cells that cannot repair damage adequately. Transformed cells are less dependent on survival signals from other cells and produce autocrines themselves in order to proliferate. Thus, apoptosis is a gene-regulated process like proliferation, differentiation and transformation. The physiological significance of apoptosis stems from the active removal of unwanted cells from the cell community [8–14] and the social control of the cell community for survival of the organism via regulation of physiological systems (neural, endocrine and immune systems) [15–25]. Defective regulation of apoptosis may play a part in the etiology of cancer, AIDS, autoimmune diseases, viral infection, cardiovascular disease, degenerative neural diseases, malformation, osteoporosis and ageing [26–30]. The molecular mechanisms of apoptosis are still unknown. The signal transduction and the determination of apoptosis are complex and dependent on the type and state of cells. However, the execution events leading to cell death are a common process irrespective of the initial apoptotic stimulus. A hallmark of apoptosis is enzymatic nucleosomal fragmentation of chromosomal DNA [31–39]. This fragmentation is
MOLECULAR MECHANISMS OF APOPTOSIS
35
Figure 3.1 Cell death or survival
thought to be involved in commitment to the irreversible death program. Elucidation of the molecular mechanisms of apoptosis is critical for understanding the biomedical significance of the cell elimination mechanism in control of the cell community and for developing the pharmacological manipulation of apoptosis to establish new therapeutic strategies for prevention and treatment of apoptosis-related diseases. In this laboratory we have focused on the molecular mechanisms and the biological significances of apoptosis during ontogenesis and oncogenesis. Here, I report some insights into the common execution pathway, especially the nature of the endonuclease involved in DNA fragmentation during apoptosis [38, 39]. In this review, I also discuss the molecular mechanisms and biological roles of apoptotic cell death. The studies on molecular mechanisms of apoptosis presented here may provide new insights into this superb system for survival rather than death. 3.2 Characteristics of Apoptosis Apoptosis was originally defined on morphological, biological and functional grounds [1–3]. Morphological features still provide the most reliable markers of apoptosis [1–3]. Unlike necrosis, apoptotic cell death passes through a series of morphologically distinct alterations (Figure 3.2, upper panel). Necrosis, another form of cell death, due to noxious stimuli or severe injury, such as hyperthermia, hypoxia, ischemia, lytic viral infection, and physical or chemical trauma, also shows characteristic morphological and biochemical features [1, 40–42]. Necrotic cell death is characterized by swelling of mitochondria and the endoplasmic reticulum. The cells are unable to maintain ion transport systems in their plasma membrane. This is followed by cell swelling and lysis, which provokes inflammatory reactions in the surrounding cells in the tissue. The histologically visible part of apoptosis has a short duration like that in cell division. In the initial stage, an individual cell, embedded in normal tissue, loses contact with its neighbors. By light and electron
36
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 3.2 Processes of apoptosis and programmed cell death
microscopy, chromatin in the nucleus is seen to become clumped and collapsed at the nuclear envelope, and then to break down into one or several nuclear fragments. Cells undergoing apoptosis shrink by reduction of their cytoplasmic volume without change in content of intact organelles. The shrinkage is thought to be rapid, but the time of this morphological change is still unknown. Furthermore, the exact pattern of morphological change may differ in different types of cell and in different states. In the next stage, membrane ruffing and blebbings lead to cellular fragmentation. The resulting fragments, so called apoptotic bodies, contain mostly nuclear remnants and intact organelles. Finally, the apoptotic bodies are recognized and rapidly engulfed by phagocytic cells, macrophages and adjacent cells, and removed. As a result, there is no accompanying inflammatory response with this process, unlike the situation in necrosis. The biochemical feature most commonly associated with apoptosis is the cleavage of chromosomal DNA, resulting in nucleosomal fragments recognizable as a DNA ladder on agarose gel electrophoresis (AGE) (Figure 3.3). This DNA fragmentation, induced preferentially in the linker regions between nucleosomes, is generally believed to be catalyzed by an apoptotic-specific Ca2+/Mg2+ dependent endonuclease. The morphological changes in the apoptotic nucleus are often associated with internucleosomal DNA fragmentation. However, in some non-lymphoidal cells, no typical nucleosomal DNA cleavage is
MOLECULAR MECHANISMS OF APOPTOSIS
37
Figure 3.3 Schematic diagram of internucleosomal DNA fragmentation in apoptosis
observed [43–47]. At present, there is no consensus of opinion on the role of DNA fragmentation in the process of apoptosis or its physiological implication. Functional characteristics of apoptosis are that it occurs under specific physiological conditions and does not necessarily involve severe damage or strong stimuli. Furthermore, apoptosis is regulated by internal cellular situations and external signals from other cells, stimulation by hormones and withdrawal of survival factors, and suppression by mitogens. Apoptosis apparently occurs in individual cells that are unwanted or are dangerous cells for the cell community. These progressive morphological and biochemical changes suggest the existence of a set of genetically regulated biochemical events. Direct evidence that cell death in animals is regulated by a genetic program comes from genetic studies in the nematode Caenorhabditis elegans (Figure 3.2, lower part) [5, 48, 49]. Several cell death (ced) genes involved in programmed cell death, such as ced-3, ced-4 and ced-9, have been identified. The morphological characteristics of programmed cell death resemble those in apoptosis of mammalian cells. However, biochemical studies show that the dying cells in C. elegans do not cleave genomic DNA themselves. Moreover, no formation of apoptotic bodies is seen. During evolution, the cell death program is considered to have been modified in a cell-, tissue- and species-specific manner to accomplish a number of physiological functions. 3.3 Molecular Mechanisms of Apoptosis Figure 3.4 illustrates the critical pathways of apoptosis. The induction of apoptosis occurs when cells receive various signals at cell-surface receptors. During signal transduction in plasma membranes, second
38
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 3.4 Molecular mechanism of apoptosis
messengers (cAMP, inositol-1, 4, 5-triphosphate (IP3), diacyl glycerol (DG), cGMP, etc.) are produced through the signal transduction systems, such as adenylate cyclase, phospholipase C, guanylate cyclase and ionic channel systems, which are essential in living cells (Figure 3.5). It is still unclear what happens downstream of the apoptosis-specific Fas/APO-1 (CD 95) ligand and tumor necrosis factor (TNF). Nevertheless, the cross-talk of the second messengers produced results in the expression of apoptosis regulator genes, the activation of pre-existing apoptotic gene products and/or suppression of the expression of survival genes. However, a cell with competence to undergo apoptosis is able to reverse these processes to survival until determination occurs. In the determination process, cells pass through a ‘point of no return’ for apoptosis. Several oncogenes, such as c-myc, p53, bcl-2, c-fos, c-jun, and Rb, have been shown to be important in commitment events. Both induction and determination processes of apoptosis are diverse in different cell types and states. However, once the apoptotic program is initiated, execution appears to occur irreversibly by a common pathway. In many types of cell a sustained increase in the concentration of intracellular Ca2+ occurs early in apoptosis [31–34]. This may be a trigger for the activation of a nuclear apoptotic-specific endonuclease that cleaves genomic DNA to nucleosomal oligomers [50–58]. ICE-like proteases, other kinds of proteases such as carpain [59, 60] and a tissue transglutaminase that cross-links cytoplasmic proteins are also activated [61–63]. The series of execution events converts a cell to an irreversible apoptotic state. Apoptotic cells or their fragments (apoptotic bodies) are rapidly recognized and phagocytosed by macrophages or their neighbors. The process of apoptosis requires energy from ATP, and sometimes depends on new gene expression after induction. Moreover, in normal lymphocytes it can be suppressed by inhibition of RNA or protein synthesis [12, 64]. These findings suggest that the transcription and translation of some genes, including those of activator proteins, are required for progression of apoptosis. In contrast, in neutrophils,
Figure 3.5 Schematic representation of molecular events in apoptosis. R=receptor; G=GTP-binding proteins; ACase=adenylate cyclase; PCL=phospholipase C; PKA=protein kinase A; PKC=protein kinase C; PTK=protein tyrosine kinase; GCase=guanylate cyclase; PIP2=phophatidylinositol-4, 5-diphosphate; IP3=inositol-1, 4, 5triphosphate; DG=diacyl glycerol; ER=endoplasmic reticlum
MOLECULAR MECHANISMS OF APOPTOSIS 39
40
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
gastrointestinal crypt cells and other cell lines, such as the human promyelocytic leukemia cell line, HL-60, apoptosis is rapidly induced by inhibitors of RNA or protein synthesis [65–67]. In these cases, the apoptotic machinery is already constructed but suppressed by protein inhibitors with short lives. 3.3.1 Induction of Apoptosis Various biological agents are known to induce apoptosis. Almost all apoptotic factors, such as the Fas/ APO-1 ligand, TNF and glucocorticoid, act specifically through either cell surface or nuclear receptors [68–71]. These factors function to produce intracellular signals as second messengers and facilitate transcription of apoptotic genes whose products are involved in the initiation process. Fas/APO-1 was identified as an apoptotic receptor homologous to TNF/nerve growth factor (NGF) superfamily receptors. Antibodies against Fas/APO-1 or its ligand induce apoptosis in immature T lymphocytes and cells expressing Fas/APO-1. The endogenous Fas/APO-1 ligand and its gene have recently been identified [72, 73]. A genetic deficiency of the lpr gene causes lymphoproliferation and an autoimmune disease that resembles systemic lupus erythematosus (SLE). The lpr gene has recently be shown to encode Fas/APO-1 [74]. Mutation of Fas/APO-1 does not result in deletion of autoreactive lymphocytes in mice. However, patients with SLE do not have mutations in the region homologus to the lpr gene. Lymphocytes of SLE patients have been shown to express high levels of bcl-2 [75]. Apoptosis also develops after loss of survival factors from other cells that normally suppress expression of the apoptotic program. Numerous cytokines, hormones and neurotrophic factors prevent apoptosis. Like apoptotic factors, many survival factors can act through receptors. Moreover, many cells show loss of survival factors but do not all die as a consequence. This modulation of apoptosis may thus be important for understanding the mechanism by which apoptosis is initiated. Viral infections, such as with human immunodeficiency virus 1 (HIV-1) and influenza virus also induce apoptosis [30]. In HIV-infected patients, depletion of CD4+ T cells leads to lymphopenia and immunodeficiency [30, 76, 77]. The mechanism of the acquired immunodeficiency syndrome (AIDS) is complex, but is known to cause apoptotic cell death of CD4+ T cells. This may be mediated by the HIV-1 gp-120 glycoprotein which binds to the CD4+ T cells. The binding of gp-120 to CD4+ T cells followed by cell-clustering has been suggested to trigger apoptosis of normal CD4+ T cells. Cell-mediated cytotoxicity both for CD8+ and CD4+ cytotoxic T cells can target cell death via apoptosis [78, 79]. The molecular mechanisms of the induction of cell-mediated cytotoxicity have been shown to involve Fas/APO-1 receptor, perfolin/granzyme and fragmentin [80]. Non-physiological agents, such as radiation, heat, anticancer drugs and toxic substances, also induce apoptosis [15–25]. The failure of the damaged cells to progress through the cell cycle restriction point for apoptosis in the G1 phase or checkpoint in the G2 phase may lead to apoptosis. Cells with unrepaired damage will undergo apoptosis from these points. If not, the cells will proliferate abnormally causing transformation. 3.3.2 Determination of Apoptosis There has been some recent progress in identifying important genes that regulate commitment to apoptosis. By genetic analysis of ced mutants in C. elegans, several genes that regulate the programmed cell death pathway have been identified (Figure 3.3, lower part) [5, 48, 49]. Some gene products act as permissive
MOLECULAR MECHANISMS OF APOPTOSIS
41
Figure 3.6 Regulation of apoptosis after DNA damage. DNA damage activates a common pathway that can lead to apoptosis or proliferation
elements to determine which cells with express the death program. Other ced genes initiate the death process by downregulating ced-9, whose product is an inhibitor of cell death. Suppression of the expression of ced-9 is necessary for initiation of cell death. A third class of ced gene products are essential for killing (Ced-3 and Ced-4) and engulfing (Ced-1, 2, 5, 6, 7, 8, 10). DNA in the dead cells is digested by phagocytic cells having Nuc-1. The existence of apoptosis-regulating genes has also been demonstrated in vertebrates by the recent discovery that c-myc, p53 and bcl-2 can act in this way [81–91]. Changes in such gene products can be responsible for initiation of apoptosis that is dependent on induction of de novo synthesis of either RNA or protein or on suppression. These findings support the idea that the cell becomes commited to a suicidal genetic program in response to the induction signals. Figure 3.6 shows the putative interactions between cmyc, p53 and bcl-2, which may be involved in determination of cell death or survival. The c-myc proto-oncogene, which is known to be important for cell proliferation control, plays a part in the regulation of apoptosis [81]. When fibroblasts are cultured in medium with a low serum concentration, expression of c-myc induces apoptosis. The expression of c-myc in the absence of other growth stimuli is considered to be abnormal and to commit the cells to the apoptotic program. The observation that c-myc antisense oligonucleotides prevent the induction of apoptosis by anti-CD3 antigen in T cell hybridoma [82] supports this idea. Adenovirus E1A may behave in the same manner as c-myc [83]. These oncogenes are involved in determining whether a cell undergoes proliferation or apoptosis. Determination of the apoptotic pathway after DNA damage may also be regulated by intrinsic factors related to the cell cycle, such as cyclin-cdK (cdc) and RB. The suppressor oncogene p53 has been shown to have dramatic effects on apoptosis [84, 85]. Overexpression of wild-type p53 in a myeloid cell line induces apoptosis [85]. Results in p53 knock-out mice show that DNA-damaging agents, such as irradiation and alkylating agents, require p53 to induce apoptosis in thymocytes, while glucocorticoids do not [86]. The gene product, p53, can delay cell cycle progression before initiation of DNA replication. The true function of p53 may be to arrest a DNA damaged cell in the G1 phase while the damage is repaired. When a cell cannot repair the damage within a critical period the cell will undergo apoptosis. Thus p53 is an essential element in the pathway leading from DNA
42
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
damage to apoptosis (Figure 3.6). The interactions between p53 and some other proto- and viral-oncogene products inactivate the p53 function. Mutations or deletions in p53 are observed in many human cancers [87]. Suppression of p53 function may promote cancer by permitting cells to replicate DNA before repair is complete (Figure 3.6). Translocation of bcl-2 to the immunoglobulin heavy chain locus [t(t14:18) transition] in human follicular B cell lymphoma results in prolonged cell life [88]. In Burkitt’s lymphoma, Epstein-Barr virus infection increases bcl-2 expression [89]. Overexpression of bcl-2 prevents apoptosis induced by irradiation, glucocorticoid or c-myc [90]. Thus, the putative oncogene bcl-2 is considered to be an apoptosis suppressor gene [91]. However, bcl-2 does not prevent apoptosis induced by cytotoxic T lymphocytes or Fas/APO-1 ligand [92]. The effect of Bcl-2, like that of c-Myc and p53, depends on the cell type and state or the stimuli. Bcl-2 can be detected in tissues in which apoptosis is temporarily suppressed (bone marrow cells, intestinal epithelia, the basal layer of skin), overridden (memory B cells) and in neurons. It is normally located in mitochondrial and nuclear membranes, but its mechanism of action is still not clear. Bax, a partner of Bcl-2, which is also a member of the Bcl-2 family, acts in opposition to Bcl-2 [93]. The ratio of Bax to Bcl-2 could be critical for the determination of cell death or survival in response to apoptic stimuli. 3.3.3 Execution of Apoptosis Apoptosis is usually an active process, requiring ATP as well as RNA and protein syntheses. This feature represents an important distinction between apoptosis and necrosis, which reflects the disintegration of a cell that has lost metabolic integrity and energy stores. In apoptosis, energy stores are needed to maintain cellular integrity and to synthesize macromolecules that are critical to progression of the processes. Signal transduction via receptor and formation of second messengers, cAMP, IP3, DG, etc. and sustained increases in intracellular Ca2+ levels may be important triggers of apoptosis (Figure 3.5). These changes are followed by rapid and selective export of ions and water from the cell leading to condensation of the cytoplasm and an increase in cell density. Clumping and fragmentation of chromatin in the nucleus occur in parallel with shrinkage of the cytoplasm in apoptotic cells. The cleavage of chromosomal DNA into nucleosomal fragments of 180–200 bp or their multiples is the biochemical feature most commonly associated with apoptosis. This inter-nucleosomal cleavage is preceded by the fragmentation of DNA to form large pieces of 20–50 kbp in length. The identification of the endonucleases responsible for fragmentation is therefore of prime importance for understanding the mechanism regulating apoptosis. One of the best characterized systems for apoptosis is immature thymocytes, which undergo apoptosis when exposed to irradiation, glucocorticoids or antibodies to the CD-3 T cell receptor complex [17, 31–34]. Studies using such systems have suggested that a sustained increase in intracellular Ca2+ during apoptosis may activate a constitutive endonuclease that mediates DNA fragmentation [31–34]. A likely candidate for such an enzyme is a Ca2+/Mg2+ -dependent endonuclease(s) [50–58], such as Nuc 18 [55] and DNase I [56]. The involvement of this type of endonuclease is supported by the observations that Ca2+ ionophores induce DNA fragmentation and cell death of thymocytes and that Ca2+ chelators can prevent both [32–34]. Furthermore, apoptosis in thymocytes is known to be inhibited by Zn2+ [32]. Thus, an endogenous Ca2+ -dependent endonuclease that is inhibited by Zn2+ is suggested to function in DNA fragmentation during thymic apoptosis. One promising approach to identification of this endonuclease is to study the nature of an endonuclease (s) purified from apoptotic cell nuclei and compare it with that of DNA fragmentation at a cellular level. We used rat thymocytes in order to examine the molecular mechanism by which chromosomal DNA is cleaved
MOLECULAR MECHANISMS OF APOPTOSIS
43
into nucleosomal oligomers [38, 39], Typical morphological changes in apoptotic thymocytes induced by irradiation were seen by transmission electron microscopy (Figure 3.7d). By this method, cell shrinkage, chromatin condensation and the disappearance of cell surface microvilli were observed by comparison with normal thymocytes (Figure 3.7a). Similar morphological changes characteristics of apoptosis were observed in dexamethasone-treated rat thymocytes. In both cases, agarose gel electrophoresis of cellular DNA from the apoptotic thymocytes showed ladder patterns, indicating the biochemical feature of apoptotic internucleosomal cleavage of chromosomal DNA (Figure 3.7e). We attempted to purify the endonuclease responsible for this internucleosomal cleavage from rat thymocytes. For detection of the endonuclease activity that cleaves linker regions of chromatin, we used HeLa S3 cell nuclei as substrate, because they contain little endogenous endonuclease activity. An endonuclease activity specific for the internucleosomal regions should produce nucleosomal ladders on agarose gel electrophoresis. Endonuclease activities present in isolated nuclei from rat thymocytes were solubilized with salts. Essentially complete solubilization of nuclear endonuclease activities were obtained when a step of sonication in 0.5 M (NH2)SO4 was included. The soluble enzyme preparation was used for subsequent chromatography. Three endonuclease activities were resolved in the third step of CM5PW HPLC (Figure 3.7c). These DNase activities, tentatively named DNase , and in order of their elution, catalyzed the cleavage of linker DNA of chromatin in HeLa S3 cell nuclei and also cleaved supercoiled plasmid DNA endonucleolytically. Interestingly, the induction of apoptosis by irradiation resulted in decreases in the activities of DNases and , without affecting DNase activity appreciably (Figure 3.7f). Similar results were obtained with apoptotic rat thymocytes induced by dexamethasone. The active fractions of DNases and from normal and DNase from apoptotic rat thymocytes were further purified by sequential HPLC steps on heparin5PW, G2000SW gel filtration and a second CM5PW column. There was no evidence for dissociable complexes of a single DNase or interconvertibility of the various forms in any of the chromatographies used. The physical and catalytic properties of these enzyme preparations were studied. The molecular masses of these three DNases were determined by the SDS-PAGE-renaturation method (activity gel system). This activity gel system was based on the ability of DNase to be renaturated after removal of SDS and to cleave DNA during incubation. As shown in Figure 3.8b, the localizations of DNases in gels could be detected by the disappearance of DNA fluorescence as dark bands on an ethidium bromide-fluorescent background. DNases and exhibited nonfluorescent bands corresponding to protein bands of 32 kDa. In contrast, DNase activity was detected as a protein of 33 kDa. The activities of these DNases detected with the activity gel system were correlated with those detected by DNA fragmentation assay (Figure 3.8a). To determine the apparent native molecular masses of these DNases, each enzyme fraction was subjected to TSKG-2000SW gel filtration HPLC. DNases , and were eluted in buffer volumes corresponding to molecular masses of 28, 30 and 31 kDa, respectively. These results suggested that all these DNases are monomeric polypeptides. The pH optima of both DNases and were approximately 5.6 in acetate-KOH or Mes-NaOH buffer. In contrast, the DNase activity was observed at neutral pH with a maximum at pH 7.2 in Mops-NaOH buffer (Table 3.1). The divalent cation profiles of DNases , and were distinct (Table 3.1). The DNase required both Ca2+ and Mg2+ for full activity, the optimal concentrations of both being 1–3 mM. DNase was sensitive to Zn2+: half-maximal inhibition was achieved with Zn2+ concentrations as low as 40 µM. In contrast, the activities of DNases and were not affected appreciably by these divalent cations at concentrations of up to 1 mM, but higher concentrations (10–30 mM) of these divalent cations inhibited and DNases as well as .
44
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 3.7 Changes in activities of DNases , and in apoptotic thymocytes. Cell morphology (2100×magnification) (a, d) and DNA fragmentation (b, e) of normal (a, b) and apoptotic (d, e) rat thymocytes were analyzed by electron microscopy and agarose gel electrophoresis, respectively. Apoptosis of rat thymocytes was induced by irradiation with X-ray (10 Gy) and incubation at 37°C. DNase activities from normal (c) and apoptotic (f) rat thymocyte nuclei were separated by CM-5PW HPLC and measured with a HeLa S3 nuclear assay system (agarose gel electrophoresis). The percentage fragmentation was determined by densitometry and plotted as a function of the fraction [38, 39]
Main inhibitors of endonuclease were tested on the purified DNases , and (Table 3.2). These three DNases were not inhibited by G-actin, an inhibitor of DNase I [38, 39]. Aurintricarboxylic acid (ATA), which has been shown to inhibit endonucleases [55, 57, 58], completely inhibited all these DNase activities at a concentration of 100 µM. For determination of the mode of DNA cleavage by these three DNases, the purified DNA from HeLa S3 cell nuclei digested with each DNase was analyzed by end-labeling methods (Figure 3.9, lower part) [39]. If DNA fragments have free 3 -OH and 5 -P ends, the resulting nucleosome ladders should be detected simply (without alkaline phosphatase pretreatment) by 3 end-labeling of the extracted DNA by terminal deoxynucleotidyl transferase and only one 32P-labeled nucleotide triphosphate, [ -32P]dCTP, and by 5 endlabeling by T4 polynucleotide kinase and [ -32P]ATP only after pretreatment of the DNA with alkaline phosphatase. In the opposite case, among the end-labelings, the 3 ends of DNA chains should not be labeled without alkaline phosphatase pretreatment. Among the end-labelings, the 5 ends of the fragments (Figure 3.9c, lane 4) could not be labeled without alkaline phosphatase pretreatment. Thus, the DNase produced 3 -OH/5 -P ends of DNA chains. In contrast, the DNA fragments formed by DNases and had 3 P/5 -OH ends as indicated by labeling of their 3 ends by terminal deoxynucleodidyl transferase only after alkaline phosphatase pretreatment and 5 end fill-in reactions without alkaline phosphatase pretreatment (Figure 3.9a, b). The modes of DNA cleavage during apoptosis induced by irradiation or dexamethasone were next investigated by using the end-labeling methods. The resultant
MOLECULAR MECHANISMS OF APOPTOSIS
45
Figure 3.8 Activity gel analysis of DNases , and . DNase activities were measured by HeLa S3 nuclear assay system (agarose gel electrophoresis) (a). The molecular masses of DNases (lane 1) and (lane 2) purified from normal rat thymocyte nuclei and that of DNase (lane 3) from irradiated apoptotic cells were analyzed with an activity gel system [38, 39] (b). The markers of molecular mass of proteins were phosphorylase b (97 400), BSA (66 200), ovalbumin (45 000), carbonic anhydrase (31 500), soybean trypsin inhibitor (21 500) and lysozyme (14 400) Table 3.2 Properties of DNases , and Condition
DNase activity (%)a
Complete
100 3/0 5.6 (MES-NaOH) 92
5.6 −Mg2+
100 3/0 7.2 (MES-NaOH) 94
100 3/3 (MOPS-NaOH) 16 1 96 94
−Ca2+ −2-Mercaptoethanol 115 119 −PMSF 81 102 2+ +Ca (3 mM) 101 104 +Mn2+ (3 mM), (−Mg2+/ 104 103 45 Ca2+) +Zn2+ (0.1 mM) 90 93 21 +G-actin (100 µg ml−1) 99 97 98 +Aurintricarboxylic acid (30 µM) 32 0 29 (100 µM) 0 0 0 a Enzyme activities were assayed as described in section 3.3.3 (Hela S3 nuclear assay) [38, 39], deleting (−) or adding (+) the compound to the optimum assay mixtures at the indicated concentration(s). Values represent means of three determinations. (Standard errors of means were 5%.)
autoradiograms revealed that in both cases, the 5 ends of the fragments (Figure 3.9d, e, lane 4) could not be labeled without alkaline phosphatase pretreatment. The same labeling patterns as seen in Figure 3.9d and
46
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
MOLECULAR MECHANISMS OF APOPTOSIS
47
e were observed when the purified DNA from HeLa S3 cell nuclei digested with the DNase was endlabeled (Figure 3.9c). Thus, the fragments produced at a cellular level had 3 -OH and 5 -P ends. These results suggest that the apoptosis induced in these conditions is catalyzed by DNase . These findings are important for identifying other factors participating in apoptosis and for elucidating the molecular mechanism of DNA fragmentation. The DNA fragmentation leads to nuclear fragmentation and the formation of apoptotic bodies. Apoptotic bodies have been shown to contain highly cross-linked proteins that are resistant to dissolution by detergents or chaotropic agents. Cross-linking of proteins in the apoptotic bodies is due to the formation of ( -glutamyl) lysine isopeptide bonds and some -glutamyl-bis-spermidine cross-links [61–63]. Although there are several transglutaminases, the cross-linking of proteins in apoptotic cells may be catalyzed by a specific intracellular transglutaminase, named tissue transglutaminase (Figure 3.5). The enzyme accumulated to high levels in some terminal differentiated cells is both induced and activated during apoptosis. Little is known about the factors that control the expression of this enzyme during apoptosis. While the role of transglutaminase in apoptosis has not been firmly established, it is likely that the crosslinking of proteins stabilizes the apoptotic bodies and limits the leakage of intracellular constituents into the extracellular space. The cell surface properties of apoptotic cells are altered during the apoptotic processes. Modified glycan structures in cell surface glycoproteins may render apoptotic cells susceptible to engulfing cells. The recognition and phagocytosis pro cesses of apoptotic cells are possibly mediated by several receptors on the macrophages. The vitronectin, thrombospondin and phosphatidylserine receptors have been suggested to be specifically involved in this recognition [95, 96]. Ascialic glycoprotein receptors participate in the phagocytosis of apoptotic cells in the liver. It is not yet clear how alterations in the distribution of ligands for these receptors on apoptotic cells mark them for elimination. Some apoptotic cells, particularly those derived from androgen-deprived prostate, express large amounts of a sulfated glycoprotein, SGP-2, that can inhibit complement-mediated cell lysis (Figure 3.5) [97, 98]. The primary sequence is identical to the gene product of testosterone-repressed prostate message 2 (TRPM-2), which was originally isolated from regressing prostate after androgen ablation [99]. This protein, which is a normal secretory product of several types of cells, presumably prevents complement-mediated lysis of the apoptotic cells. 3.4 Biological Implications of Apoptosis Over the last couple of years it has been recognized that apoptosis can play important roles in the control of the cell community in both developing and mature vertebrates [1–13]. Apoptosis, which is a cell suicidal process inherent in multicellular organism, is the functional opposite of mitosis. Physiologically, it is considered to be a process whereby an organism eliminates unwanted and dangerous cells, i.e. old, excessive, damaged or preneoplastic cells. Cell populations in the organism are modified and regenerated during development and postnatal life by elimination of some cells with survival of others. Apoptotic cell death to remove unneeded cells is regulated by interactions with other cells. Apoptosis apparently does not occur at random in all cells of a tissue, only in individual cells that are no longer needed. The reason for the development of apoptosis during evolution is considered to be the provision of the ability to establish a social control system in multicellular organisms and to produce many repertoires of cells. Apoptosis is induced by a wide range of physiological signals. Furthermore, non-physiological agents, such as X-ray and cytotoxic drugs mimic the effects of natural apoptotic factors [15–25]. These external
48
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 3.9 Modes of DNA cleavage by DNases , and and in apoptotic rat thymocytes. DNA was extracted from HeLa S3 cell nuclei digested by DNase (a), (b) or (c). DNA was extracted from apoptotic rat thymocytes 4 h after irradiation (10 Gy) (d) or dexamethasone (10−7 M) treatment (e). The DNA was incubated with (lanes 1 and 3) or without (lanes 2 and 4) alkaline phosphatase (APase) prior to 3 end- (lanes 1 and 2) or 5 end- (lanes 3 and 4) labeling. The 3 ends of the DNA fragments were labeled by incubation with 5 U of terminal deoxynucleotidyl transferase and 0. 83 mCi ml−1 of [ -32P]dCTP in the presence of 25 mM Tris-HCl (pH 7.6), 10 mM dithiothreitol and 1 mM CaCl2. The 5 ends of the DNA fragments were labeled by incubation with 5 U of T4 polynucleotide kinase and 0.83 mCi ml−1 of [ -32P]ATP in the presence of 100 mM Tris-HCl (pH 7.6), 20 mM MgCl2, 10 mM dithiothreitol and 0.2 mM spermidine. The phosphoryl groups in the ends of DNA chains were removed by pretreatment with 20 U of calf intestinal alkaline phosphatase in the presence of 36 mM Tris-HCl (pH 8.0) and 1 mM MgCl2. Aliquots of DNA were subjected to 2% agarose gel electrophoresis and autoradiography [39]
factors trigger input of signals for the determination stage of apoptosis. Other toxic signals, such as Ca2+ ionophore and hypoxia, may induce apoptosis by increasing intracellular Ca2+ -dependent apoptotic endonuclease and proteases. On the other hand, apoptosis can be suppressed genetically by expression of
MOLECULAR MECHANISMS OF APOPTOSIS
49
bcl-2 [88–91]. Malignant cells growing rapidly due to expression of a high level of c-myc may need to prevent apoptosis by bcl-2 expression [90]. In the Epstein-Barr virus, the E1B gene blocks apoptosis, resulting in immortalized cells. Chemical tumor promoters also seem to inhibit apoptosis and contribute to transformation. Mutations or deletions in the p53 tumor suppressor gene may promote human cancer development [84]. The genes actually involved in execution of apoptosis have yet to be identified. The central apoptotic gene product is thought to be an endonuclease that cleaves chomosomal DNA into nucleosomal oligomers. DNase purified from rat thymocytes is the most likely candidate for the apoptotic endonuclease [38, 39]. The apoptotic genes are considered to be highly conserved in higher animals since, in general, the execution process is a common pathway involving fundamentally critical events, including removal of apoptotic cells (apoptotic bodies). Apoptotic bodies are rapidly engulfed by phagocytic cells [96]. So cell death via apoptosis occurs without involving an inflammatory response in the surrounding cells in the tissue. Apoptosis can thus be considered as an altruistic cell suicidal mechanism for protecting neighboring cells. A benefit of apoptosis during tissue involution may be reduction of energy expenditure, unlike the situation in necrosis. Pharmacological strategies for controlling apoptotic processes may prevent apoptosis-related diseases. In AIDS or Alzheimer’s disease, a way of inhibiting apoptosis or maintaining homeostasis might help in preventing the cell loss. Conversely, drugs that promote apoptosis preferentially in malignant cells might be of therapeutic valuable and could also enhance the effect of anticancer drugs. Chemical compounds that specifically affect the regulatory mechanisms of apoptosis could thus attenuate these diseases and provide new therapeutic strategies. The pharmacological manipulation of apoptosis could also probably control normal ageing. The implications of control of apoptosis in therapy of cancer, AIDS and autoimmune and degenerative nerve diseases are obvious. Besides examination of the therapeutic value of regulation of apoptosis for therapy of various diseases, it is necessary to identify all the genes involved in expression of the apoptotic program. 3.5 Conclusion Using apoptotic rat thymocytes induced by X-ray irradiation or glucocorticoid treatment, we found that the novel DNase was as an apoptotic endonuclease [38, 39]. DNase is also present in nuclei from apoptotic calf thymocytes and rat splenocytes. Several putative apoptotic endonucleases such as Ca2+/Mg2+ dependent endonucleases [53, 54, 58], Nuc 18 [55], DNase I [56] and Ca2+/Mn2+ -dependent endonuclease [57] differ from DNase by their physical and catalytic properties (Table 3.1). As the ends of DNA fragments cleaved by the DNase are the same as those produced in apoptotic rat thymocytes, and Zn2+ in the µM order, known to prevent apoptosis, inhibited DNase , this enzyme is considered to be the endonuclease responsible for DNA fragmentation during thymic apoptosis. The role of DNase in thymic apoptosis is supported by the observations that only DNase activity was retained in apoptotic thymocyte nuclei. However, in some cell lines, DNA fragmentation occurred without increase in the intracellular Ca2+ concentration [106, 107]. These observations imply that the apoptotic endonuclease responsible for DNA fragmentation may not always be the Ca2+/Mg2+ -dependent endonucleases and may differ in different cell types and states. Nevertheless, it will be essential to investigate whether the intranuclear Ca2+ concentration increases and is sufficient for DNase activation in apoptotic cells. Recently, Barry and Eastman identified a DNase II-like endonuclease from Chinese hamster ovary (CHO) cell nuclei that caused similar DNA fragmentation to that observed in apoptotic cells [104]. The activity of this acidic CHO endonuclease was not dependent on Ca2+ or Mg2+, and was insensitive to Zn2+
50
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
(Table 3.1). These properties are similar to those of DNase II, which is known to be a lysosomal enzyme [105], and our DNase and . The cognate forms of DNases and are also present in calf thymocytes and rat splenocytes. The physiological significances of DNases and are still unknown, but these enzymes are in general important roles for nuclear functions. Studies on changes in their activities on various stimuli (unpublished data) may provide some clues to their significance. Confirmation of the essential role of DNase in apoptosis in vivo requires further studies. Studies are also needed on whether DNases , and are produced by separate genes or by alternative splicings. These problems could be resolved best by cloning these DNases. Furthermore, the mechanisms of activation of these three nuclear DNases in thymocyte nuclei are unclear. Possibly the DNases interact with endogenous inhibitors or activators, and/or are modified post-translationally by poly(ADP-ribosyl)ation [108, 110] or phosphorylation. In fact, Ca2+/Mg2+—or Ca2+/Mn2+ -dependent endonuclease activities have been shown to be regulated by poly(ADP-ribosyl)ation [111, 112]. Studies on these possibilities may provide important clues for understanding the biological significance of these nuclear DNases in cell death or survival. There are several possible explanations for the physiological significance of DNA fragmentation. First, selective DNA degradation may occur at specific DNA domains involved in cell proliferation such as replication origin and telomere. Second, integrated c-oncogenes and proviral genes may be cleaved for prevention of transformation by their transfection. Third, protection from liberation of large amounts of DNA causing autoimmune reactions is necessary, since antibodies against double-stranded DNA are associated with SLE. Fourth, the formation of apoptotic bodies leading to DNA fragmentation facilitates their engulfment by phagocytes. Although speculative, these explanations indicate the possible importance of DNA fragmentation during apoptosis and suggest ways in which it may have very distinct spatial and temporal significance. The elucidation of the mechanisms of apoptosis is now progressing rapidly [113, 114] at organ, cellular and molecular levels. These studies should provide new insights into the fundamental significances of apoptosis in ontogeny and phylogeny of multicellular organisms. We still do not know how to commit the elimination of unneeded cells in a cell community. We also do not know how to determine the extent of cell damage that is repairable or that is sufficient for apoptosis. Abnormal apoptosis may prevent removal of genetic variants. Ultimately, the purposes of apoptosis are considered to be elimination of individual cells damaged sufficiently to increase risk of heritable genetic changes, and maintenance of genetic fidelity. Thus it is principally a genetic program for gene survival via gene elimination [114]. References 1 2 3 4 5 6 7 8 9 10 11 12
. KERR, J.F.R. (1971) J. Pathol. 105, 13–20. . KERR, J.F.R., WYLLIE, A.H. & CURRIE, A.R. (1972) Br. J. Cancer 26, 239–257. . WYLLIE, A.H., KERR, J.F.R., MACASKILL, I.A.M. & CURRIE, A.R. (1973) J. Pathol. 111, 85–94. . WYLLIE, A.H., KERR, J.F.R. & CURRIE, A.R. (1980) Int. Rev. Cytol. 68, 251– 306. . ELLIS, R.E., YUAN, J. & HORVITZ, H.R.A. (1991) Rev. Cell Biol. 7, 663–698. . TOMEI, L.D. & COPE, F.O. (eds) (1991) In Apoptosis: the Molecular Basis of Cell Death. Cold Spring Harber Laboratory Press, New York. . SAUNDERS, J.W. (1966) Science 154, 605–612. . UMANSKY, S.R. (1982) J. Theor. Biol. 97, 591–602. . COWAN, W.M., FAWCETT, J.W., O’LEARY, D.D.M. & STANFIELD, B.B. (1984) Science 225, 1258–1265. . KERR, J.F.R., HARMON, B. & SEARLE, J. (1974) J. Cell. Sci. 14, 571–585. . COHEN, J.J. (1991) Adv. Immun. 50, 55–85. . GOLSTEIN, P., OJCIUS, D.M. & YOUNG, J.D.E. (1991) Immunol. Rev. 121, 29–65.
MOLECULAR MECHANISMS OF APOPTOSIS
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
51
. GERSCHENSON, L.E. & ROTELLO, R.J. (1992) FASEB J 6, 2450–2455. . RAFF, M.C. (1992) Nature 356, 397–400. . BERGER, N.A. (1985) Radiat. Res. 101, 4–15. . CARSON, D.A., SETO, S., WASSON, D.B. & CARRERA, C.J. (1986) J. Exp. Med. 163, 746–751. . UCKER, D.S. (1987) Nature 329, 62–64. . SMITH, C.A., WILLIAMS, G.T., KINGSTON, R., JENKINSON, E.J. & OWEN, J. J. (1989) Nature 337, 181–184. . DIVE, C. & HICKMAN, J.A. (1991) Br. J. Cancer 64, 192–196. . ARENDS, M.J. & WYLLIE, A.H. (1991) Int. Rev. Exp. Pathol. 32, 223–254. . GROUX, H., TORPIER, G., MONTE, D., MOUTON, Y., CAPRON, A. & AMEISEN, J.C. (1992) J. Exp. Med. 175, 331–340. . WILLIAMS, G.T., SMITH, C.A., MCCARTHY, N.J. & GRIMES, E.A. (1992) Trends Cell Biol. 2, 263–267. . COHEN, J.J., DUKE, R.C., FADOK, U.A. & SELLINS, K.S. (1992) Ann. Res. Immunol. 10, 267–293. . COHEN, J.J. (1993) Immunol. Today 14, 126–130. . TANUMA, S., SHIOKAWA, D., TANIMOTO, Y., IKEKITA, M., SAKAGAMI, H., TAKEDA, M., FUKUDA, S. & KOCHI, M. (1993) Biochem. Biophys. Res. Commun. 194, 29–35. . LEVINE, A.J., MOMAND, J. & FINLAY, C.A. (1991) Nature 351, 453–456. . MARTIN, D.P., ITO, A., HORIGOME, K., LAMPE, P.A. & JOHNSON, E.M. JR. (1992) J. Neurobiol. 23, 1205–1220. . MATTSON, M.P., CHENG, B., DAVIS, D., BRYANT, K., LEIBENBURG, I. & ROYDEL, R.E. (1992) J. Neurosci. 12, 376–389. . CARSON, D.A. & RIBEIRO, J.M. (1993) Lancet 341, 1251–1254. . COUGEON, M.L. & MONTAGNIER (1993) Science 260, 1269–1270. . WYLLIE, A.H. (1980) Nature 284, 555–556. . COHEN, J.J. & DUKE, R.C. (1984) J. Immunol. 132, 38–42. . WYLLIE, A.H., MORRIS, R.G., SMITH, A.L. & DUNLOP, D. (1984) J. Pathol. 142, 67–77. . MCCONKEY, D.J., HARTZELL, P. & ORRENIUS (1990) Arch. Biochem. Biophys. 278, 284–287. . ARENDS, M.J., MORRIS, R.G. & WYLLIE, A.H. (1990) Am. J. Pathol. 136, 593– 601. . COMPTON, M.M. (1992) Cancer Metastasis Rev. 11, 105–119. . PEITSH, M.C., MANNHERZ, H.G. & TSCHOPP, J. (1994) Trends Cell. Biol. 4, 37–41. . TANUMA, S. & SHIKAWA, D. (1994) Biochem. Biophys. Res. Commun. 203, 789–797. . SHIOKAWA, D., OHYAMA, H., YAMADA, T., TAKAHASHI, K. & TANUMA, S. (1994) Eur. J. Biochem. 226, 23–30. . DUVAL, E. & WYLLIE, A.H. (1986) Immunol. Today 7, 115–119. . DYSON, J.E.E. (1986) Cells Tiss. Kinet. 19, 311–324. . ALLEN, T.D. (1987) In Perspectives on Mammalian Cell Death. POTTEN, C.S. (ed). Oxford University Press, Oxford, pp. 39–65. . WALKER, P.R., SMITH, C., YOUDALE, T., LEBLANC, J., WHIFIELD, J.F. & SIKORSKA, M. (1991) Cancer Res. 51, 1078–1085. . ROY, C., BROWN, D.L., LITTLE, J.E., VALENTINE, B.K., WALKER, P.R., SIKORSKA, L.M., LEBLANC, J. & CHALY, N. (1992) Exp. Cell Res. 200, 416–424. . TOMEI, L.D., SHAPIRO, J.P. & COPE, F.O. (1993) Proc. Natl Acad. Sci. USA 90, 853–857. . BRAWN, D.G., SUN, X.-M. and CHEN, G.M. (1993) J. Biol. Chem. 3037–3039. . OBERHAMMER, F., WILSON, J.W., DIVE, C., MORRIS, I.D., HICKMAN, J.A., WAKELING, A.E., WALKER, P.R. & SIKORSKA, M. (1993) EMBO J. 12, 3679– 3684. . ELLIS, R.E. & HORVITZ, H.R. (1986) Cell 44, 817–829. . HENGARTNER, M. & HORVITZ, H.R. (1994) Nature 369, 318–320. . BURGOYNE, L.A., WAGAR, M.A. & ATKINSON, M.R. (1970) Biochem. Biophys. Res. Commun. 39, 254–259.
52
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70. 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. HEWISH, D.R. & BURGOYNE, L.A. (1973) Biochem. Biophys. Res. Commun. 56, 703–710. . ISHIDA, R., AKIYOSHI, H. & TAKAHASHI, T. (1974) Biochem. Biophys. Res. Commun. 59, 658–665. . NAKAMURA, M., SAKAKI, Y., WATANABE, N. & TAKAGI, Y. (1981) J. Biochem. 89, 143–152. . HASHIDA, T., TANAKA, Y., MATSUNAMI, N., YOSHIHARA, K., KAMIYA, T., TANIGAWA, Y. & KOIDE, S.S. (1982) J. Biol Chem. 257, 13114–13119. . GAIDO, M.L. & CIDLOWSKI, J.A. (1991) J. Biol. Chem. 266, 18580–18585. . PEITSCH, M.C., POLZAR, B., STEPHAM, H., CROMPTON, T., MACDONALD, H.R., MANNHERZ, H.G. & TSHOPP, J. (1993) EMBO J. 12, 371–377. . NIKONOVA, L.V., BELELSKY, I.P. & UMANSKY, S.R. (1993) Eur.J. Biochem. 215, 893–901. . RIBEIRO, J.M. & CARSON, D.A. (1993) Biochemistry 32, 9129–9136. . KAUFMANN, S.H. (1989) Cancer Res. 49, 5870–5878. . NICHOLSON, D.W., ALI, A., THORNBERRY, N.A., VAILLANCOURT, J.P., DING, D.K., GALLANT, M., GAREAU, Y., GRIFFIN, P.R., LABELLE, M., LAZEBNIK, Y.A. et al. (1995) Nature 376, 37–43. . CHIOCCA, E.A., DAVIES, P.J.A. & STEIN, J.P. (1988) J. Biol. Chem. 263, 11584– 11589. . FESUS, L. (1991) Immunol. Lett. 30, 277–282. . KNIGHT, C.R., REES, R.C. & GRIFFIN, M. (1991) Biochim. Biophys. Acta 1096, 312–318. . YAMADA, T. & OHYAMA, H. (1988) Int. J. Radiat. Biol. 53, 65–75. . IJIRI, K. & POTTEN, C.S. (1987) Br. J. Cancer 55, 113–123. . BAXTER, G.D., COLLINS, R.J., HARMON, B.V., KUMR, S., PRENTICE, R.L., SMITH, P.J. & LAVIN, M.F. (1989) J. Pathol. 158, 123–129. . MARTIN, S.J., LENNON, S.V., BONHAM, A.M. & COTTER, T.G. (1990) J. Immunol. 145, 1859–1867. . ITOH, N., YONEHARA, S., ISHII, A., YONEHARA, M., MISUSHIMA, S. & NAGATA, S. (1991) Cell 66, 233–243. . WATANABE-FUKUNAGA, R. (1992) J. Immunol. 148, 1274–1279. PAWLITA, M. & MAIER, G. (1992) J. Biol. Chem. 267, 10709–10715. . SUDA, T. & NAGATA, S. (1994) J. Exp. Med. 179, 873–879. . SUDA, T. & NAGATA, S. (1993) Cell 75, 1169–1178. . TAKAHASHI, T. & NAGATA, S. (1994) Cell 76, 969–976. . WATANABE-FUKINAGA, R., BRANNON, C.I., COPELAND, N.G., JENKINS, N.A. & NAGATA, S. (1992) Nature 356, 314–317. . GRANINGER, W.B. (1992) Wien Klin. Wochenschr. 104, 205–207. . PANTALEO, G., GRAZIOSI, C. & FAUCI, A.S. (1993) N. Engl. J. Med. 328, 327–335. . TERAI, C., KORNBLUTH, R., PAUZA, C.D., RICHMAN, D.D. & CARSSON, D.A. (1991) J. Clin. Invest. 87, 1710–1714. . GOLSTEIN, P., OJCIUS, D.D. & YOUNG, J.D.E. (1991) Immunol. Rev. 121, 29–65. . RAHELU, M., WILLIAMS, G.T., KUMANARATNE, D.S., EATON, G.C. & GASTON, J.S.H. (1993) J. Immunol. 150, 4856–4866. . ROUVIER, ?., LUCIANI, M.-F. & GOLSTEIN, P. (1993) J. Exp. Med. 177, 195–200. . EVAN, G.I., WYLLIE, A.H., GILBERT, C.S., LITTLEWOOD, T.D., LAND, H., WATERS, C.M., PENN, L.Z. & HANCOCK, D.C. (1992) Cell 69, 119–128. . SHI, Y.F., GLYNN, J.M., GUILBERT, L.J., COTTER, T.G., BISSONNETTE, R.P. & GREEN, D.R. (1992) Science 257, 212–214. . RAO, L., DEBBAS, M., SABBATINI, P., HOCKENBERY, D. & KOSMERYER, S. (1992) Proc. Natl Acad. Sci. USA 89, 7742–7746. . GINSBERG, D., MICHAEL-MICHALOVITZ, D. & OREN, M. (1991) Mol Cell Biol. 11, 582–585. . YONISHI-ROUACH, Z., RESNITZKY, D., LOTEM, J., SACHS, L. & KIICHI, A. (1991) Nature 352, 345–347. . CLARKE, A.R., PURDIE, C.A., HARRISON, D.J., MORRIS, R.G., BIRD, C.C., HOOPER, M.L. & WYLLIE, A.H. (1993) Nature 362, 849–852.
MOLECULAR MECHANISMS OF APOPTOSIS
87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114
53
. LEWINE, A.J., MOMAND, J. & FINLAY, C.A. (1991) Nature 351, 453–456. . VAUX, D.L., CORY, S. & ADAMS, J.M. (1988) Nature 335, 440–442. . FIMKE, J., FIRTZEN, R. & TERNES, P. (1992) Blood 80, 459–469. . BISSONETTE, R.P., ZCHEVERRI, F., MAHBOUBI, A. & GREEN, D.R. (1992) Nature 359, 552–554. . KORSMEYER, S.J. (1992) Cancer Surveys 15, 105–118. . VAUX, D.L., AGUILA, H.L. & WEISSMAN, I.L. (1992) Int. Immunol. 4, 821–824. . OLTVAI, Z.N., MILLIMAN, C.L. & KORSMEYER, S.J. (1993) Cell 74, 609–619. . MAKINO, K., IKEKITA, M., KONDO, T., TANUMA, S. & OHSHIMA, H. (1994) Colloid Polymer Sci. 272, 487–492. . SAVILL, J., DRANSFIELD, I., HOGG, N. & HASLETT, C. (1990) Nature 343, 170– 173. . SAVILL, J., FADOK, V., HENSON, P. & HASLETT, C. (1993) Immunol. Today 14, 131–136. . BUTTYAN, R., OLSSON, C.A., PINTAR, J., CHANG, C. & BANDYK, M.G. (1989) Mol. Cell. Biol. 9, 3473–3481. . BETTUZZI, S., HIIPAKKA, R.A., GILNA, P. & LIAO, S. (1989) Biochem. J. 257, 293–296. . LEGER, J., LEGUELLEC, R. & TENNISWOOD, P.R. (1988) Prostate 13, 131–142. . LASKOWSKI, M., SR. (1971) In The Enzymes, BOYER, P.D., LANDY, H. & MYRBACK, K. (eds) Vol. 4, pp. 289–311. Academic Press, New York. . LACKS, S.A. (1981) J. Biol. Chem. 256, 2644–2648. . ANAI, M., SASAKI, M. MUTA, A. & MIYAGAWA, T. (1981) Biochim. Biophys. Acta 656, 183–188. . KREUDER, V., DIECKHOFF, J., SITTIG, M. & MANNHERZ, H.G. (1984) Eur. J. Biochem. 139, 389–400. . BARRY, A. & EASTMAN, A. (1993) Arch. Biochem. Biophys. 300, 440–450. . MURAI, K., YAMANAKA, M., AKAGI, K. & ANAI, M. (1980) J. Biochem. 87, 1097– 1103. . ALNEMRI, E.S. & LITWACK, G. (1990) J. Biol. Chem. 265, 17323–17333. . BANSAL, N., HOULE, A.G. & MELNYKOVYCH, G. (1990) J. Ceil Physiol. 143, 105–109. . SUGIMURA, T. (1973) Prog. Nucleic Acid Res. Mol. Biol. 13, 127–151. . UEDA, K. & HAYAISHI, O. (1985) Ann. Rev. Biochem. 54, 73–100. . TANUMA, S. (1989) Biochem. Biophys. Res. Commun. 163, 1047–1055. . YOSHIHARA, K., TANIGAWA, Y. BURZIO, L. & KOIDE, S.S. (1975) Proc. Natl Acad. Sci. USA 72, 289–293. . NELIPOVICH, P.A., NIKONOVA, L.V. & UMANSKY, S.R. (1988) Int. J. Radiat. Biol. 53, 749–765. . TANUMA, S. & SHIOKAWA, D. (1995) Apoptosis: Programmed Molecular Subcell Biology, pp. 1–12. Springer-Verlag, Heidelberg. . TANUMA, S. (1995) Tiss. Cul. Res. Commun. 14, 138–140.
4 Identification of Genes Associated with Cell Death CARILEE A.LAMB and J.JOHN COHEN Department of Immunology, B-184, University of Colorado Medical School, Denver, Colorado 80262, USA
4.1 Introduction Apoptosis seems to be an orderly program. Most of the morphological hallmarks of apoptosis are similar from cell to cell, regardless of the type of cell or the nature of the inducer that triggered the process. In contrast, necrosis is generally thought of as random disintegration of a cell that has been so injured that it cannot recover; it is difficult to define necrosis on a biochemical basis, except to say that eventually the cell becomes unable to operate its ion pumps, or keep its plasma membrane intact and functioning. The development of our knowledge about apoptosis supports the notion of an orderly and regulated apoptotic program, which any cell can activate when necessary. In 1972 Kerr, Wyllie, and Currie described in detail the morphology of cells that die when damage is too great to repair, but less severe than that which leads to necrosis; and they suggested the term ‘apoptosis’ for this morphology [1]. Although workers had described cell death of this sort in previous years, the 1972 paper formalized a way of looking at and describing it. They also pointed out the similarity to examples of morphogenetic death, that is, the death of cells during embryonic development that is necessary for the formation of organs and limbs. Apoptosis soon came to be used synonymously with ‘programmed cell death’. This latter term was originally used to describe death of cells in invertebrates [2], for example, of the intersegmental muscles during eclosion. However, because apoptosis is also seen in cells that have been damaged, the term programmed cell death is somewhat inappropriate. Nevertheless, the evidence seems to point in the direction of a ‘cell death program’, which can be activated physiologically or pathologically. 4.2 Genetic Regulation of Apoptosis In 1984, Wyllie’s group [3] and ours [4] showed that the apoptotic death of rodent thymus cells upon exposure to glucocorticoids was inhibited if protein or RNA synthesis were blocked by agents such as emetine, cycloheximide or actinomycin D. This suggested very strongly that the steroids were not killing the cells directly; rather, they were inducing a form of suicide, in which there was new expression of a lethal gene or genes. Thus, death could be considered as another glucocorticoid response, one restricted to
IDENTIFICATION OF GENES ASSOCIATED WITH CELL DEATH
55
cortical thymocytes. In general, glucocorticoids work by binding cytoplasmic receptors which are then translocated to the nucleus where they interact with glucocorticoid-response elements; the result can be negative or, more commonly, positive regulation of transcription. The evidence supported the idea of new gene expression in apoptosis induced by dexamethasone, but what were the genes? Either they were a novel set of ‘death’ or ‘suicide’ genes, or they were conventional glucocorticoid-responsive genes which interacted in a peculiar way with other elements in thymocytes but not in other cells. This question has not been answered. It is clear that there are many glucocorticoid-responsive genes in thymocytes [5], but which, if any, of the genes that have been described so far are actually part of a death pathway, or directly activate such a pathway, has yet to be determined. It was soon shown that a characteristic pattern of apoptosis could be induced in thymocytes by a wide range of inducers, so that the response to glucocorticoids was not morphologically unique. Nevertheless, it is unique in that it requires the glucocorticoid receptor; viewed in this context, the receptor is a ‘suicide protein’ for thymocytes. But only in this context: low-dose radiation, for example, does not require the presence of a functional glucocorticoid receptor to cause apoptosis in thymocytes (in our hands, doses of the steroid antagonist RU486 that completely abrogate glucocorticoid-induced apoptosis have no effect on apoptosis caused by ionizing radiation). On the other hand, it has been shown that the growth-regulatory gene p53 is necessary for radiation-induced, but not glucocorticoid-induced, apoptosis of thymocytes [6, 7]. Thymus cells from animals in which the p53 gene has been inactivated by homologous recombination retain their sensitivity to glucocorticoids but are resistant to low-dose radiation. Thus p53 is a ‘suicide protein’ for thymocytes under certain circumstances but not others. These findings suggest that there may be a ‘final common pathway’ of apoptosis, that is, a series of biochemical steps that all cells must use to generate the morphological changes that we recognize. None of these has been definitively characterized as to their molecular basis, or the role they play in the cascade of apoptotic events. Even the hallmark of apoptosis, the cleavage of nuclear chromatin into a ladder of nucleosome-sized fragments, has been shown to be optional, in the sense that certain cells undergoing unmistakable apoptosis fail to cleave their DNA in this manner [8, 9]. Of the other signs of apoptosis, cell shrinkage is a mystery; zeiosis is probably associated with cytoskeletal rearrangements, but the data are conflicting [10–13]; and the recognition of apoptotic cells by phagocytes can be mediated by more than one mechanism [14–16]. But since all cells undergoing apoptosis by definition demonstrate at least some of these changes, we are led to conclude that the final common pathway is real. If it is, then in each cell, and with each inducer, there must be a ‘private’ pathway that leads to and activates the ‘common’ pathway. The glucocorticoid receptor is clearly part of a private apoptotic pathway that exists in thymocytes but not, for example, in liver cells, which do not die when treated with these steroids. The p53 protein is part of the ‘private’ radiation pathway, but cannot be in the common pathway. However, several private pathways may share a gene or gene product: for example, etoposide, a topoisomerase inhibitor, induces apoptosis in thymocytes via a p53-dependent route [7]. 4.3 Genes Associated with Apoptosis Four technical approaches have been used to identify mammalian genes associated with apoptosis. First, genes associated with the apoptotic phenotype have been isolated on the basis of their differential expression in normal and apoptotic cells. Second, molecules which transduce a signal to undergo apoptosis in response to external stimuli have been identified (primarily as cell surface and steroid receptors). Third, sequences from apoptotic genes found in lower organisms have been used as probes to screen cDNA
56
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
libraries and genetic databases for homologous mammalian sequences. Finally, analysis of different cancers has yielded several genes whose inactivation by mutation or deregulation by chromosomal translocation have been found to be involved in the process of cellular transformation. Over- or under-expression of several of these genes has been shown to induce or protect from apoptosis. 4.4 Identifying Apoptosis-associated Messages It is our impression that we do not yet know enough about apoptosis to distinguish sharply between private and final common mechanisms. This makes the identification of genes involved in the process of apoptosis a difficult and risky endeavor. In the first attempt to identify mammalian ‘death genes’, subtractive hybridization was used to enrich the cDNAs of messages preferentially expressed in thymocytes undergoing apoptosis after exposure to glucocorticoids [17]. This approach is valuable but is not perfect; using it, one cannot reliably obtain messages that are only slightly more abundant in apoptotic cells, as might be the case for a regulatory protein. Furthermore, because there is a general degradation of messages during apoptosis [18], any message whose degradation is somewhat slower than the average will appear to be enriched. The subtractive approach, based on gene message abundance rather than on some selectable property, is thus limited. Messages called RP-2 and RP-8 were isolated in this manner [19], and their role in apoptosis is under investigation. They have some intriguing attributes. RP-2 apparently encodes a protein with two membrane-spanning domains; both N- and C-termini are cytoplasmic, and the intervening region is extracellular and probably glycosylated. It seems to be a member of a family of genes whose products function as purine-regulated calcium channels of the PX2 type [20, 21]. This is particularly interesting as calcium has frequently been implicated in the process of apoptosis, and extracellular ATP (to which these channels are sensitive) can cause apoptosis in at least some normal and transformed types of cell [22–24]. However, the PX2 ion channel also plays an important role in synaptic transmission, and its function in other types of cell will have to be determined. RP-8 encodes a novel protein with a zinc-finger motif, which may be involved in transcriptional regulation. It is expressed in the central nervous system in areas where apoptotic cell death is prominent [25]. 4.5 Apoptosis Transduced from the Cell Surface Although cell-surface receptors are usually thought of as involved in positive signaling and adhesion, it is becoming clear that there are receptors that transduce a lethal signal. There are several situations where this sort of interaction might be important. In development, certain cells die as tissues assume their mature configurations. Most of this morphogenetic death is probably due to deprivation of essential survival factors, but it remains an interesting possibility that certain cells execute others which fail to meet some criterion: the conceptual antithesis of the nurse cell. The immune system, with its role in immune surveillance, is responsible for assuring the rapid death of abnormal somatic cells, especially those undergoing malignant transformation or carrying an internal parasite (virus, fungus, protozoan or bacterium). It has been shown that in virus-containing target cells, cytotoxic T cells induce DNA fragmentation of viral DNA at the same rate as genomic DNA [26]. This is a superb adaptation to the problem of viral control, as simply lysing an infected target might actually facilitate viral spread. Cytotoxic T cells, when activated, express a molecule called Fas ligand [27]. This molecule is a member of the rapidly-growing tumor necrosis factor family of surface ligands. Its target is a molecule expressed on the surface of many different types of cells in the
IDENTIFICATION OF GENES ASSOCIATED WITH CELL DEATH
57
body, called Fas or APO-1 or, more recently, CD95. The engagement of CD95 by its ligand can be mimicked by antibody to CD95 which cross-links the receptor. The nature of the lethal signals has not yet been established, but death is by apoptosis and it can be extremely rapid—as it is when cytotoxic T cells themselves are used [28]. Tumor necrosis factor can also cause apoptosis in target cells that bear at least one of the known TNF receptors [29]. Again, the mechanism is not fully understood, but may involve the sphingomyelinase pathway [30–32]. In lymphocytes, abnormal signaling through the normal activation transducers can cause apoptosis [33]; in this case, it would appear that these are not death signals per se but signals which, arriving at the wrong time or out of the proper sequence, ‘confuse’ the cell and cause it to activate its suicide program to avoid possible harm to the organism. This probably plays an essential role in the regulation of the lymphocyte repertoire, as receptor-mediated stimulation of immature T or B cells causes apoptosis, whereas it would cause activation in the corresponding mature cell. A gene associated with this kind of apoptosis is nur77, previously identified as an orphan steroid receptor. Its expression is increased in thymus cells and hybridomas induced to undergo apoptosis by cross-linking their receptors for antigen [34, 35]. Antisense or dominant negative versions of nur77 block apoptosis in these cells, which suggests that it plays an important role in cells that receive signals via their antigen receptors. It does not seem to be involved in thymocyte apoptosis induced by other agents such as glucocorticoids. 4.6 Cell Death Genes in Invertebrates and Mammalian Homologues The most complete genetic analysis of the cell death program has been performed in the nematode Caenorhabditis elegans. Several characteristics make C. elegans amenable to study: the knowledge of the pattern of cells that die to produce the mature animal, the availability of mutants, and the ability to examine the whole animal morphologically. Genes involved in killing, engulfment and inhibition of the death program have been discovered. The ced-3 and ced-4 genes are required for cell death to occur [36]. Recently, Ced-3 has been shown to be homologous to the human and murine interleukin-1 converting enzyme (ICE), a cysteine protease [37]. Using the predicted protein sequence of Ced-3 to compare its probable three-dimensional structure to that known by X-ray crystallography of ICE has indicated that the residues involved in substrate recognition and catalysis are conserved in Ced-3 [38]. That ICE is involved in mammalian cell death was demonstrated by the suppression of neuronal apoptosis by CrmA [39], a specific inhibitor of ICE encoded by a cowpox virus gene [40]. In addition, the overexpression of ICE in Rat-1 fibroblasts induces cell death, which is inhibited by Bcl-2 and CrmA [41]. These structural and functional data yield evidence supporting a proteolytic function for Ced-3. It has yet to be determined if ICE, like Ced-3, is involved in developmental cell death. No mammalian homologue has yet been identified for Ced-4, although from an analysis of its predicted sequence it is postulated to be a calcium-binding protein [42]. A third C. elegans gene, ced-9, has been identified as a negative regulator of apoptosis. Loss of Ced-9 function results in death of the organism [43]. Conversely, gain-of-function mutants prevent all cells from dying [43]. The mammalian protooncogene bcl-2 has been shown to be a structural and functional homologue of ced-9 [44]. The ability of bcl-2 to substitute for ced-9 in the prevention of normal cell death in C. elegans suggests the cell death program is relatively conserved between vertebrates and invertebrates [44]. Two mammalian genes, bax [45] and bcl-x [46], have been shown to be homologous to bcl-2. Bax negatively regulates the function of Bcl-2 [45], possibly through the formation of heterodimers [47]. Bcl-x is alternatively spliced to form two mRNAs, bcl-xL and bcl-xS, which appear to function independently of
58
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Bcl-2 to inhibit and promote apoptosis, respectively [46]. It remains to be determined if Ced-9 activity is analogously modulated by other C. elegans proteins. Analysis of Drosophila mutants may also yield insight into the switch or upstream effectors of the apoptotic process. White et al. [48] isolated the reaper gene from a region spanning a deletion in four independent Drosophila mutants that displayed the absence of normal cell death as detected by the lack of acridine orange staining. Introduction of the reaper gene into mutants restored cell death. In addition to having a defect in developmentally programmed cell death, the mutant embryos show decreased sensitivity to low dose radiation, only undergoing apoptosis at relatively high doses. The sequence of the reaper transcript has one open reading frame that is predicted to code for a protein of 65 amino acids. This deduced protein sequence is homologous to the intracellular ‘death domains’ of CD95 and the TNF receptor [61]. 4.7 Oncogenes Associated with Apoptosis Increasingly, contributions to our knowledge of apoptosis are coming from research into the genetic mutations associated with certain types of cancer. Several genes shown to be inactivated (p53) or dysregulated (c-myc, bcl-2) by chromosomal changes are being demonstrated to play a role in the signaling or suppression of apoptosis. It has been found that p53 is mutated or deleted in approximately 50% of human cancers, suggesting the wild-type protein is active in tumor suppression. Analysis of p53 knock-out mice demonstrated that p53 mediates growth arrest and apoptosis induced by DNA damaging agents, irradiation, and etoposide, but not glucocorticoids or calcium ionophores [6, 7]. Recently, it was shown that p53 mediates growth arrest through its role as a transcriptional activator. p53 induces the expression of a 21 kD protein termed WAF1 (wild-type p53 activated fragment; [49]) or Cip1 (cdk interacting protein; [50]) which interacts with and inhibits cyclin-kinase complexes, thereby preventing cell cycle progression. WAF1/Cip1 was found to be induced in cells undergoing p53-mediated apoptosis [51]. Interestingly, it appears that, in a somatotropic progenitor cell line, p53-dependent induction of apoptosis by UV irradiation does not require new RNA or protein synthesis [52], suggesting that other cellular events may be involved in p53-mediated signaling of apoptosis. C-myc, the cellular homologue of a gene in many transforming retroviruses, has been found to be dysregulated by chromosomal translocations in Burkitt lymphomas and mouse plasmacytomas. Recently it has been discovered that in addition to its involvement in cellular proliferation, c-myc is also implicated in some models of apoptosis, a seemingly opposing function. Dysregulation or overexpression of c-myc in Rat-1 fibroblasts or IL-3 dependent myeloid cells cultured under conditions (low serum [53] or growthfactor deprivation [54]) which normally signal quiescence results instead in apoptosis. In addition, experiments using c-myc antisense oligonucleotides have implicated c-myc in activation-induced apoptosis in T cell hybridomas [55]. The dysregulation of a gene that inhibits apoptosis, like mutation or deletion of a tumor suppressor gene, predisposes a cell to oncogenic transformation. The t(14;18) chromosomal translocation found in follicular lymphomas juxtaposes the bcl-2 gene with the immunoglobulin heavy chain locus. Studies using Bcl-2-Ig transgenic mice have shown that this inappropriate expression of Bcl-2 results in prolonged mature B-cell survival [56]. Bcl-2 inhibits apoptosis induced by various agents in many cell types, although it is unable to release the cell from growth arrest [57, 58]. This suggests that Bcl-2 blocks a component of the apoptotic pathway that is distinct from cell-cycle regulation. Interestingly, Bcl-2 does not appear to block apoptosis induced by CTL [59], and only partially inhibits apoptosis mediated by TNF or Fas [60]. This indicates that Bcl-2 probably does not function in the ‘final common pathway’ of apoptosis.
IDENTIFICATION OF GENES ASSOCIATED WITH CELL DEATH
59
4.8 Conclusions Everything in the cell is under genetic control, in that all molecules in a cell are either themselves gene products or metabolites dependent upon the actions of gene products. Therefore when we say that apoptosis is genetically regulated we are not saying anything particularly profound. The real question is whether there is a single genetically-determined program which defines apoptosis, as there seems to be with mitosis. The goal of mitosis is creation; that of apoptosis, destruction. Just as the process of mitosis is complex, and interference with it at many different levels can impair it, so it may be that apoptosis is a complex pathway, and interference at many different places may activate it. We have not yet identified a gene whose expression is the sine qua non of apoptosis. Whether such a gene exists is a question whose answer will be of enormous interest to discover. References 1 2 3
4 5
6 7 8 9
10 11 12 13 14
. KERR, J.F.R., WYLLIE, A.H. & CURRIE, A.R. (1972) Apoptosis: a basic biological phenomenon with wideranging implications in tissue kinetics. Br. J. Cancer 26, 239–257. . LOCKSHIN, R.A. & WILLIAMS, C.M. (1964) Programmed cell death. II. Endocrine potentiation of the breakdown of the intersegmental muscles of silkmoths. J. Insect. Physiol. 10, 643–649. . WYLLIE, A.H., MORRIS, R.G., SMITH, A.L. & DUNLOP, D. (1984) Chromatin cleavage in apoptosis: association with condensed chromatin morphology and dependence on macromolecular synthesis. J. Pathol. 142, 67–77. . COHEN, J.J. & DUKE, R.C. (1984) Glucocorticoid activation of a calcium-dependent endonuclease in thymocyte nuclei leads to cell death. J. Immunol. 132, 38–42. . HARRIGAN, M.T., BAUGHMAN, G., CAMPBELL, N.F. & BOURGEOIS, S. (1989) Isolation and characterization of glucocorticoid- and cyclic AMP-induced genes in T lymphocytes. Mol. Cell Biol. 9, 34383446. . LOWE, S.W., SCHMITT, E.M., SMITH, S.W., OSBORNE, B.A. & JACKS, T. (1993) p53 is required for radiation-induced apoptosis in mouse thymocytes. Nature 362, 847–849. . CLARKE, A.R., PURDIE, C.A., HARRISON, D.J., MORRIS, R.G., BIRD, C.C., HOOPER, M.L. & WYLLIE, A.H. (1993) Thymocyte apoptosis induced by p53-dependent and independent pathways. Nature 362, 849–852. . SELLINS, K.S. & COHEN, J.J. (1991) Cytotoxic T lymphocytes induce different types of DNA damage in target cells of different origins. J. Immunol. 147, 795–803. . OBERHAMMER, F., WILSON, J.W., DIVE, C., MORRIS, I.D., HICKMAN, J.A., WAKELING, A.E., WALKER, P.R. & SIKORSKA, M. (1993) Apoptotic death in epithelial cells: cleavage of DNA to 300 and/or 50 kb fragments prior to or in the absence of internucleosomal fragmentation. EMBO J. 12, 3679–3684. . KOLBER, M.A., BROSCHAT, K.O. & LANDA-GONZALEZ, B. (1990) Cytochalasin B induces cellular DNA fragmentation. FASEB J. 4, 3021–3027. . MARTIN, S.J. & COTTER, T.G. (1990) Disruption of microtubules induces an endogenous suicide pathway in human leukaemia HL-60 cells. Cell Tissue Kinet. 23, 545–559. . GODMAN, G.C., MIRANDA, A.F., DEITCH, A.D. & TANENBAUM, S.W. (1975) Action of cytochalasin D on cells of established lines. III Zeiosis and movements at the cell surface . J. Cell Biol. 64, 644–667. . COTTER, T.G., LENNON, S.V., GLYNN, J.M. & GREEN, D.R. (1992) Microfilament-disrupting agents prevent the formation of apoptotic bodies in tumor cells undergoing apoptosis. Cancer Res. 52, 997–1005. . FADOK, V.A., VOELKER, D.R., CAMPBELL, P.A., COHEN, J.J., BRATTON, D.L. & HENSON, P.M. (1992) Exposure of phosphatidylserine on the surface of apoptotic lymphocytes triggers specific recognition and removal by macrophages. J. Immunol. 148, 2207–2216.
60
15
16
17 18 19 20 21 22 23
24 25 26 27 28 29 30 31 32
33 34
35 36
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. FADOK, V.A., SAVILL, J.S., HASLETT, C., BRATTON, D.L., DOHERTY, D.E., CAMPBELL, P.A. & HENSON, P.M. (1992) Different populations of macrophages use either the vitronectin receptor or the phosphatidylserine receptor to recognize and remove apoptotic cells. J. Immunol. 149, 4029–4035. . HALL, S.E., SAVILL, J.S., HENSON, P.M. & HASLETT, C. (1994) Apoptotic neutrophils are phagocytosed by fibroblasts with participation of the fibroblast vitronectin receptor and involvement of a mannose/fucosespecific lectin. J. Immunol. 153, 3218–3227. . OWENS, G.P. & COHEN, J.J. (1992) Identification of genes involved in programmed cell death. Cancer Metastasis Rev. 11, 149–156. . CIDLOWSKI, J.A. (1982) Glucocorticoids stimulate ribonucleic acid degradation in isolated rat thymic lymphocytes in vitro. Endocrinology 111, 184–190. . OWENS, G.P., HAHN, W.E. & COHEN, J.J. (1991) Identification of mRNAs associated with programmed cell death in immature thymocytes. Mol. Cell. Biol. 11, 4177–4188. . VALERA, S., HUSSY, N., EVANS, R.J., ADAMI, N., NORTH, R.A., SUPRENANT, A. & BUELL, G. (1994) A new class of ligand-gated ion channel defined by P2x receptor for extracellular ATP. Nature 371, 516–519. . BRAKE, A.J., WAGENBACH, M.J. & JULIUS, D. (1994) New structural motif for ligand-gated ion channels defined by an ionotropic ATP receptor. Nature 371, 519–523. . MURGIA, M., PIZZO, P., STEINBERG, T.H. & DI VIRGILIO, F. (1992) Characterization of the cytotoxic effect of extracellular ATP in J774 mouse macrophages. Biochem. J. 288, 897–901. . ZOETEWEIJ, J.P., VAN DE WATER, B., DE BONT, H.J., MULDER, G.J. & NAGELKERKE, J.F. (1992) Involvement of intracellular Ca2+ and K+ in dissipation of the mitochondrial membrane potential and cell death induced by extracellular ATP in hepatocytes. Biochem. J. 288, 207–213. . FILIPPINI, A., TAFFS, R.E., AGUI, T. & SITKOVSKY, M.V. (1990) Ecto-ATPase activity in cytolytic Tlymphocytes. Protection from the cytolytic effects of extracellular ATP. J. Biol Chem. 265, 334–340. . MAHALIK, T.J., HAHN, W.E., CLAYTON, G.H. & OWENS, G.P. (1994) Programmed cell death in developing grafts of fetal substantia nigra. Exp. Neurol. 129, 27–36. . SELLINS, K.S. & COHEN, J.J. (1989) Polyoma viral DNA is damaged in target cells during cytotoxic T lymphocyte-mediated killing. J. Virol. 63, 572–578. . SUDA, T., TAKAHASHI, T., GOLSTEIN, P. & NAGATA, S. (1993) Molecular cloning and expression of the Fas ligand, a novel member of the tumor necrosis factor family. Cell 75, 1169–1178. . DUKE, R.C., CHERVENAK, R. & COHEN, J.J. (1983) Endogenous endonuclease-induced DNA fragmentation: an early event in cell-mediated cytolysis. Proc. Natl Acad. Sci. USA 80, 6361–6365. . WRIGHT, S.C., KUMAR, P., TAM, A.W., SHEN, N., VARMA, M. & LARRICK, J. W. (1992) Apoptosis and DNA fragmentation precede TNF-induced cytolysis in U937 cells. J. Cell Biochem. 48, 344–355. . OBEID, L.M., LINARDIC, C.M., KAROLAK, L.A. & HANNUN, Y.A. (1993) Programmed cell death induced by ceramide. Science 259, 1769–1771. . KOLESNICK, R. & GOLDE, D.W. (1994) The sphingomyelin pathway in tumor necrosis factor and interleukin-1 signaling. Cell 77, 325–328. . JARVIS, W.D., KOLESNICK, R.N., FORNARI, F.A., TRAYLOR, R.S., GEWIRTZ, D.A. & GRANT, S. (1994) Induction of apoptotic DNA damage and cell death by activation of the sphingomyelin pathway. Proc. Natl Acad. Sci. USA 91, 73–77. . NEWELL, M.K., HAUGHN, L.J., MAROUN, C.R. & JULIUS, M.H. (1990) Death of mature T cells by separate ligation of CD4 and the T-cell receptor for antigen. Nature 347, 286–289. . LIU, Z.G., SMITH, S.W., MCLAUGHLIN, K.A., SCHWARTZ, L.M. & OSBORNE, B.A. (1994) Apoptotic signals delivered through the T-cell receptor of a T-cell hybrid require the immediate-early gene nur77. Nature 367, 281–284. . WORONICZ, J.D., CALNAN, B., NGO, V. & WINOTO, A. (1994) Requirement for the orphan steroid receptor Nur77 in apoptosis of T-cell hybridomas . Nature 367, 277–281. . YUAN, J. & HORVITZ, H.R. (1990) The Caenorhabditis elegans genes ced-3 and ced-4 act cell autonomously to cause programmed cell death. Dev. Biol. 138, 33–41.
IDENTIFICATION OF GENES ASSOCIATED WITH CELL DEATH
37 38
39 40
41
42 43 44 45 46
47 48 49
50 51
52 53
54
55 56
61
. YUAN, J., SHAHAM, S., LEDOUX, S., ELLIS, H.M. & HORVITZ, H.R. (1993) The C. elegans cell death gene ced-3 encodes a protein similar to mammalian interleukin-1 beta-converting enzyme. Cell 75, 641–652. . WILSON, K.P., BLACK, J.F., THOMSON, J.A., KIM, E.E., GRIFFITH, J.P. , NAVIA, M.A., MURCKO, M.A., CHAMBERS, S.P., Aldape, R.A., RAYBUCK, S. A. & LIVINGSTON, D.J. (1994) Structure and mechanism of interleukin-1 beta con verting enzyme. Nature 370, 270–275. . GAGLIARDINI, V., FERNANDEZ, P.A., LEE, R.K., DREXLER, H.C., ROTELLO, R.J., FISHMAN, M.C. & YUAN, J. (1994) Prevention of vertebrate neuronal death by the crmA gene. Science 263, 826–828. . RAY, C.A., BLACK, R.A., KRONHEIM, S.R., GREENSTREET, T.A., SLEATH, P.R., SALVESEN, G.S. & PICKUP, D.J. (1992) Viral inhibition of inflammation: cowpox virus encodes an inhibitor of the interleukin-1 converting enzyme. Cell 69, 597– 604. . MIURA, M., ZHU, H., ROTELLO, R., HARTWIEG, E.A. & YUAN, J. (1993) Induction of apoptosis in fibroblasts by IL-1 beta-converting enzyme, a mammalian homolog of the C. elegans cell death gene ced-9. Cell 75, 653–660. . YUAN, J. & HORVITZ, H.R. (1992) The Caenorhabditis elegans cell death gene ced-4 encodes a novel protein and is expressed during the period of extensive programmed cell death. Development 116, 309–320. . HENGARTNER, M.O., ELLIS, R.E. & HORVITZ, H.R. (1992) Caenorhabditis elegans gene ced-9 protects cells from programmed cell death. Nature 356, 494–499. . HENGARTNER, M.O. & HORVITZ, H.R. (1994) C. elegans cell survival gene ced-9 encodes a functional homolog of the mammalian proto-oncogene bcl-2. Cell 76, 665–676. . OLTVAI, Z.N., MILLIMAN, C.L. & KORSMEYER, S.J. (1993) Bcl-2 heterodimerizes in vivo with a conserved homolog, Bax, that accelerates programmed cell death. Cell 74, 609–619. . BOISE, L.H., GONZALEZ-GARCIA, M., POSTEMA, C.E., DING, L., LINDSTEN, T., TURKA, L.A., MAO, X., NUNEZ, G. & THOMPSON, C.B. (1993) bcl-x, a bcl-2-related gene that functions as a dominant regulator of apoptotic cell death. Cell 74, 597– 608. . YIN, X.M., OLTVAL, Z.N. & KORSMEYER, S.J. (1994) BH1 and BH2 domains of Bcl-2 are required for inhibition of apoptosis and heterodimerization with Bax [see comments]. Nature 369, 321–323. . WHITE, K., GRETHER, M.E., ABRAMS, J.M., YOUNG, L., FARRELL, K. & STELLER, H. (1994) Genetic control of programmed cell death in Drosophila. Science 264, 677–683. . EL-DEIRY, W.S., TOKINO, T., VELCULESCU, V.E., LEVY, D.B., PARSONS, R., TRENT, J.M., LIN, D., MERCER, W.E., KINZLER, K.W. & VOGELSTEIN, B. (1993) WAF1, a potential mediator of p53 tumor suppression. Cell 75, 817–825. . HARPER, J.W., ADAMI, G.R., WEI, N., KEYOMARSI, K. & ELLEDGE, S.J. (1993) The p21 Cdk-interacting protein Cipl is a potent inhibitor of Gl cyclin-dependent kinase. Cell 75, 805–816. . EL-DEIRY, W.S., HARPER, J.W., O’CONNOR, P.M., VELCULESCU, V.E., CANMAN, C.E., JACKMAN, J., PIETENPOL, J.A., BURRELL, M., HILL, D.E. & WANG, Y. (1994) WAF1/CIP1 is induced in p53-mediated G1 arrest and apoptosis. Cancer Res. 54, 1169–1174. . CAELLES, C., HELMBERG, A. & KARIN, M. (1994) p53-dependent apoptosis in the absence of transcriptional activation of p53-target genes [see comments]. Nature 370, 220–223. . EVAN, G.I., WYLLIE, A.H., GILBERT, C.S., LITTLEWOOD, T.D., LAND, H., BROOKS, M., WATERS, C.M., PENN, L.Z. & HANCOCK, D.C. (1992) Induction of apoptosis in fibroblasts by c-myc protein. Cell 69, 119–128. . ASKEW, D.S., ASHMUN, R.A., SIMMONS, B.C. & CLEVELAND, J.L. (1991) Constitutive c-myc expression in an IL-3-dependent myeloid cell line suppresses cell cycle arrest and accelerates apoptosis. Oncogene 6, 1915–1922. . WADHWA, R., KAUL, S.C., IKAWA, Y. & SUGIMOTO, Y. (1991) Protein markers for cellular mortality and immortality. Mutat. Res. 256, 243–254. . MCDONNELL, T.J., DEANE, N., PLATT, F.M., NUNEZ, G., JAEGER, U., MCKEARN, J.P. & KORSMEYER, S.J. (1989) bcl-2-immunoglobulin transgenic mice demonstrate extended B cell survival and follicular lymphoproliferation. Cell 57, 79–88.
62
57 58 59 60 61
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. VAUX, D.L., CORY, S. & ADAMS, J.M. (1988) Bcl-2 gene promotes haemopoietic cell survival and cooperates with c-myc to immortalize pre-B cells. Nature 335, 440–442. . CHIOU, S.K., RAO, L. & WHITE, E. (1994) Bcl-2 blocks p53-dependent apoptosis. Mol Cell Biol. 14, 2556–2563. . VAUX, D.L., AGUILA, H.L. & WEISSMAN, I.L. (1992) Bcl-2 prevents death of factor-deprived cells but fails to prevent apoptosis in targets of cell mediated killing. Int. Immunol. 4, 821–824. . ITOH, N., TSUJIMOTO, Y. & NAGATA, S. (1993) Effect of bcl-2 on Fas antigen-mediated cell death. J. Immunol. 151, 621–627. . GOLSTEIN, P., MARGUET, D. & DEPRAETERE, V. (1995) Homology between Reaper and the cell death domains of Fas and TNFR1. Cell 1, 185–186.
5 Genes Involved in Apoptosis NICOLA J.MCCARTHY, ELIZABETH A.HARRINGTON and GERARD I.EVAN Biochemistry of the Cell Nucleus Laboratory, Imperial Cancer Research Fund Laboratories, 44 Lincoln’s Inn Fields, London WC2A 3PX, UK
5.1 Introduction Apoptosis is a genetically controlled cell death pathway that is triggered by diverse stimuli (reviewed in [1]). Although the phenomenon of apoptosis has been known for a long time, awareness of its critical role in metazoan biology has been slow to evolve, principally because cell death is commonly viewed as a pathological phenomenon restricted to tissue subject to damage or trauma. Recently, however, it has become increasingly clear that apoptosis is a fundamental component of tissue homeostasis. The hunt for genes that control apoptosis is at present an intense area of research, mainly because genetic lesions that restrict cell death also grossly affect cell population number, thereby leading to a number of diverse pathologies. For example, transgenic mice in which over-expression of the apoptosis-suppressing protooncogene bcl-2 is targeted to the B cell lineage exhibit a substantially expanded B-cell population. Although this hyperplasia is not at first clinically pathological, from these hyperplastic cells secondary tumourigenic lesions arise with high frequency. Although there is as yet no overall defined molecular pathway for apoptosis, several gene products have been shown to either induce (c-Myc, p53, Bax, Fas) or suppress (Bcl-2, Abl, p19EIB) it. In this chapter, these key genetic players and their effects on cell death will be reviewed, with particular attention paid to the possible interactions that may occur between these genes that lead to malignancy. 5.2 Models of Programmed Cell Death Many physiological stimuli have been identified that trigger apoptosis, including withdrawal of survival factors [2–4], stimulation of the T-cell receptor in immature thymocytes [5, 6] and the action of TNF or Fas ligand on many cell types [7, 8]. The response of cells to such stimuli is, in turn, dependent upon the intracellular state of the cell. Thus, changes in gene expression can alter the response of a cell to a given apoptotic stimulus (Figure 5.1). The regulation of apoptosis in mammalian cells appears to be complex, reflecting the organisational complexity of the mammalian form. This complexity makes difficult the identification of the basic
64
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 5.1 External and internal stimuli regulating cell fate. An external stimulus received by a cell can effect several cellular responses. The response of the stimulated cell is in turn dependent on its internal information such as its genotype or its developmental history. For example deregulated c-Myc expression causes apoptosis in response to serum withdrawal in Rat-1 fibroblast cells. The same signal in fibroblasts with regulated c-Myc results in Myc downregulation and cellular quiescence, not cell death.
machinery that regulates mammalian apoptosis. However, investigations into the control of cell population growth in less complex and genetically tractable invertebrate organisms, such as the nematode worm Caenorhabditis elegans, have facilitated the identification of cell death regulatory genes. Such strategies have been invaluable in (a) demonstrating that programmed cell death is an essential component of tissue homeostasis in widely diverse organisms and (b) identifying, by homology, new gene families involved in the control of programmed cell death in mammalian cells. 5.2.1 Invertebrate Models of Programmed Cell Death Much of our knowledge of the molecular control of programmed cell death or apoptosis has emerged from study of the nematode C. elegans. C. elegans development is invariant: each adult hermaphrodite worm has exactly the same body plan comprising the same number of cells. The entire developmental lineage of the worm has been precisely mapped [9–11] and hence it has been established that exactly 131 specific cells undergo programmed death during the animal’s development. Using appropriate mutants, key genes involved in these programmed cell deaths have been identified. Of particular interest are two genes, cell death gene (ced)-3 and ced-4. Loss-of-function of either ced-3 or ced-4 results in an animal in which no developmental deaths occur. Instead, the ‘undead’ cells take up similar differentiated forms to their sister cells [9]. Construction of appropriate mosaic animals demonstrates that ced-3 and -4 gene products act
GENES INVOLVED IN APOPTOSIS
65
within the cells that die rather than as killer signals produced by neighbouring or surrounding cells [12]. Thus, cell death mediated by ced-3 and ced-4 is ‘suicide’ and not ‘murder’. Clearly, if genes exist that trigger cell death, mechanisms must exist that control and antagonize it. Suppression of cell death in C. elegans is mediated by the ced-9 gene product. Ced-9 gain-of-function mutants exhibit no developmental cell death. Nonetheless, the resulting animals are viable despite their excess cells. In contrast, ced-9 loss-of-function mutations exhibit massive cell death, even in cells that do not normally die during development [13]. Thus, ced-9 appears to be a general suppressor of programmed cell death whose expression is necessary to inhibit a ‘default’ suicide function in many cell types in the nematode. Functional suppression of cell death in ced-9 loss-of-function mutants can be partially restored by the transgenic expression of bcl-2 [14, 15], a proto-oncogene that suppresses apoptotic cell death in mammalian cells (see later, under the heading, ‘Bcl-2: a Suppressor of Apoptosis’). Thus, ced-9 and bcl-2 are functional homologues of each other, implying evolutionary conservation of the cell death machinery. In turn, this suggests that other genes involved in nematode cell death, such as ced-3 and ced-4, might also exist in mammalian cells. To date, sequence analysis of ced-4 has identified no known mammalian homologues. However, several mammalian homologues of ced-3 have recently been found. The first to be characterized was the interleukin 1 -converting enzyme (ICE) [16]. ICE is a 45 kD cysteine protease that cleaves specific substrate poly peptides at aspartate residues [17, 18]. The active form of ICE is produced by auto-catalytic cleavage into subunits of 10 kD and 20 kD: active ICE comprises two of each subunit. ICE has been shown to induce apoptosis when ectopically expressed in Rat-1 fibroblasts [19]. ICE expression also induces apoptosis in dorsal root ganglion cells and this cell death is suppressed by the poxvirus protein Crm A, a member of the serpin family of protease inhibitors with specific inhibitory activity for ICE [20, 21]. Many, perhaps most, viruses express gene products that suppress suicidal death of their host cells, thereby promoting host cell survival until viral induced cell lysis occurs and releases mature virus (see section 5.3). Further evidence indicating a key role for proteases in triggering apoptosis is provided by the recent identification of two other homologues of ICE/Ced-3. One is a putative ICE-like protein, prICE. prICE does not cleave IL-1 but instead cleaves the enzyme poly (ADP-ribose) polymerase (PARP) [22]. The precise significance of this discovery is not yet clear since the prICE protein has not been purified and any involvement of PARP in apoptosis is unsubstantiated (for review see [23]). However, prICE has been functionally implicated in an in vitro model system of nuclear fragmentation with similarities to apoptosis in vivo. A second ICE homologue is the human nedd-2 gene, aka ich-1 [24]. The nedd-2/ich-1 gene is transcribed as two spliced variant mRNAs. One encodes ICH-1L, which is 435 amino acids in length and comprises both the 10 kDa and 20 kDa active subunits comparable to active ICE. The other is ICH-1S, which is 312 amino acids in length and terminates 21 amino acids after the active pentapeptide found in ICH-1L. These two different forms Ich-1 also have opposing effects on cell viability. Expression of ICH-1L kills Rat-1 cells in the presence of serum, whereas ICH-1s prevents cell death upon serum withdrawal. Interestingly Crm A is not able to fully delay cell death induced by ICH-1L suggesting that Crm A does not efficiently inhibit ICH as it does ICE. The diversity of mammalian homologues of the nematode ced-3 gene illustrates the general theme that corresponding systems controlled by one gene in C. elegans may be controlled by several genes in more complex mammalian systems.
66
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
5.2.2 Control of Apoptosis in Mammalian Cells One of the more surprising features of cell death in mammalian cells is that apoptosis is controlled by genes that are already described as key regulators of cell growth and differentiation. These include a number of oncogenes and tumour suppressor genes. c-Myc and Apoptosis The proto-oncogene c-myc has long been implicated in the control of cell proliferation [25]. Its elevated or deregulated expression in virtually all tumours suggests that lesions affecting c-myc expression play a key part in the development of neoplasia. c-myc encodes a short-lived, nuclear phosphoprotein, c-Myc, whose expression is tightly regulated by mitogen availability. In quiescent fibroblasts, mitogen addition stimulates c-myc expression which peaks after 2–3 hours [26], suggesting a role for c-myc at the G0/G1 transition. Thereafter, however, c-myc expression is maintained throughout the cell cycle, albeit at reduced level, suggesting a continuous role in the maintenance of cell proliferation. Evidence for the involvement of c-Myc in the progression from quiescence to proliferation is provided by the observation that the ectopic induction of c-Myc in the absence of other immediate early gene expression is sufficient to drive quiescent fibroblasts into cycle [27, 28]. Evidence favours the notion that the involvement of c-myc in proliferation is due to its action as a transcription factor. c-Myc is a sequence-specific DNA-binding protein which possesses an amino-terminal domain with transcriptional modulatory activity and dimerizes with a stable protein, Max, via a C-terminal HLH-LZ dimerization domain [29– 32]. To date, however, the target genes for c-myc action are poorly defined. Given its well established role in the regulation of cell proliferation, the observation that c-Myc also induces programmed cell death (apoptosis) was something of a surprise. Transgenic mice that constitutively express c-Myc within the lymphoid compartment show a substantially elevated level of apoptosis [33–35]. Moreover, c-Myc induces apoptosis in fibroblasts subjected to serum deprivation or cytotoxic agents [28, 36] and in factor-deprived IL-3-dependent myeloid cells [37]. c-Myc is also implicated in activation-induced apoptosis of T-cell hybridomas [38] and in TNF-induced cell death [39, 40]. Mutagenesis analysis indicates that the regions of the c-Myc protein that mark it as a transcription factor are all required for both the mitogenic and apoptotic action of c-Myc [28, 41]. Specifically, these regions are the amino-terminal transactivational domain and the carboxy-terminal DNA-binding and dimerisation basic-helix-loop-helixleucine zipper domain required for interaction with Max. This implies that c-Myc induces both of the ‘opposing’ functions, proliferation and cell death, via a transcriptional mechanism and, moreover, that these functions are obligatorily interlocked. Two simple models have been proposed to explain c-Myc induced apoptosis in cells deprived of serum or treated with cytostatic agents. In the first model, death arises because of a conflict between the growth promoting activity of c-Myc and the growth suppressive actions of mitogen removal or cytostatic agents. In this model, apoptosis is not a physiological function of c-Myc: rather it is a pathological consequence of ‘inappropriate’ c-Myc expression coupled with impeded growth. The second model proposes that induction of the apoptotic programme is a bona fide physiological function of c-Myc that is obligatorily coupled to cMyc induced proliferation. In order to proliferate, therefore, a cell requires two independent sets of signals, one to trigger mitogenesis and the other to suppress the concomitant apoptotic programme. This ‘dual key’ model (see Figure 5.2) proposes the testable hypothesis that c-Myc induces death in low serum due to the absence of survival factors that serve to inhibit the apoptotic programme, rather than due to absence of
GENES INVOLVED IN APOPTOSIS
67
Figure 5.2 The dual key model. In this model induction of apoptosis is a normal and obligate function of c-Myc that is regulated by availability of anti-apoptotic survival factors or Bcl-2. For a proliferating cell apoptosis represents a default pathway occurring in the absence of survival signals. In this scenario a cell that acquires a potential tumourigenic lesion such as deregulated c-Myc expression will only proliferate while survival factors are present. Since survival factors are limiting in vivo deregulated c-Myc expression results in a non-viable phenotype once the cell outgrows the paracrine supply of survival factors. Thus the suppression of neoplasia is ‘hardwired’ into cellular growth control.
serum mitogens required for cell cycle progression [3]. Substantial evidence now favours the ‘dual key’ model. Apoptosis is rapidly induced in c-Myc expressing cells proliferating in high serum when cycloheximide is used to block protein synthesis, strongly suggesting that the apoptotic programme that cMyc implements requires no de novo protein synthesis to be manifest and must therefore pre-exist in otherwise viable cells even in the absence of a conflict of growth signals [3, 28]; for example, cells expressing c-Myc and growing in high serum. This in turn implies the existence of serum factors that serve to suppress the potential apoptotic programme activated by c-Myc. Recently, two such factors have been defined that suppress c-Myc-induced apoptosis in mesenchymal cells: the insulin-like growth factors (IGFs) and platelet derived growth factor (PDGF) (see Figure 5.3). No other tested fibroblasts’ mitogen (e.g. epidermal growth factor, fibroblast growth factor, bombesin or IL1) exerts any anti-apoptotic effect. Moreover, the antiapoptotic activities of IGF and PDGF are not dependent upon their mitogenic activities—both suppress apoptosis under conditions in which cell proliferation is profoundly blocked with cytostatic agents and during the post-commitment S/G2 phases of the cell cycle [3] when fibroblasts no longer require mitogens for completion of the cell cycle. The notion that the anti-apoptotic action of certain cytokines is unlinked to their mitogenic action fits well with the known survival-potentiating activities of IGF-I and PDGF in nonproliferating post-mitotic cells such as neurones [42, 43]. Presumably, individual cell lineages are responsive to their own unique sets of survival factors. For example, interleukin-3 appears to be a major anti-apoptotic cytokine for many haematopoietic lineages [44–46] whereas PDGF is not. In summary, it appears that c-Myc continually drives both a proliferative and an apoptotic programme. cMyc-expressing cells die more rapidly when treated with cytostatic and cytotoxic drugs because they have a primed apoptotic pathway and are therefore poised to commit suicide in response to such agents. Thus, proliferating cells expressing c-Myc continuously require the presence of survival factors such as IGFs and PDGF to suppress apoptosis. Such a coupling between proliferation and apoptosis provides an excellent homeostatic mechanism for controlling the numbers of proliferating cells within a tissue and suppressing neoplasia—any proliferating clone is programmed to die if it outgrows the paracrine supply of survival factors. This idea has dramatic implications for our understanding of the genesis of tumours because it
68
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 5.3 The effect of different growth factors upon c-Myc apoptosis in serum-starved Rat-1/c-MycER fibroblasts. Subconfluent Rat-1/c-MycER cells were deprived of serum for 48 hours. C-Myc expression was the activated by the addition of 2 µM -oestradiol in the presence (black circles) or absence (white squares) of cytokine; IGF-1 (100 ng ml −1), PDGF (10 ng ml−1), IGF-II (10 ng ml−1), bFGF (100 ng ml−1), EGF (10 ng ml−1) and bombesin (10 ng ml−1). Approximately 100 cells were monitored by time lapse video microscopy for 20 hours. Apoptotic cell deaths were scored from the last time the cells appeared normal and cumulative cell deaths were plotted against time. Images were acquired at the rate of 12 frames per hour. c-Myc-induced apoptosis is suppressed by insulin-like growth factors (IGFs) and PDGF. Potent fibroblast mitogens such as EGF, bFGF and bombesin do not exert an anti-apoptotic effect.
GENES INVOLVED IN APOPTOSIS
69
implies that all neoplasms must have evolved some mechanism to suppress apoptosis. Indeed, the more rapidly the tumour grows, the more suppression of apoptosis is required. Thus, lesions within the apoptotic pathway are certain to be important components of carcinogenesis that may well provide novel targets for pharmacological intervention. Bcl-2: a Suppressor of Apoptosis Bcl-2 was first identified on human chromosome 18 as the site of reciprocal translocation in follicular Bcell lymphoma [47]. It encodes a membrane-associated protein, Bcl-2, present in the endoplasmic reticulum, nuclear and outer mitochondrial membranes. The anti-apoptotic activity of bcl-2 was first noted when its expression was observed to prolong the survival of an interleukin-3-dependent myeloid cell line upon removal of the cytokine without inducing proliferation [4, 48], Bcl-2 protects cells from apoptosis induced by survival factor removal in many cell types, including mesenchymal cells and sympathetic and sensory neurones [49]. Bcl-2 is widely expressed during embryonic development. However, in the adult it is restricted to the immature and stem cell populations in epithelia such as skin and intestine, long-lived cells such as resting memory B-lymphocytes, peripheral sensory neurones and glandular epithelial tissues that undergo repeated cycles of hyperplasia and involution [50]. In transgenic mice, constitutive expression of bcl-2 in the lymphoid compartment leads to an increase in the numbers of mature resting B-cells and potentiates their longevity. Affected T-cells are markedly resistant to the cytotoxic effects of radiation, glucocorticoids and anti-CD3 although thymic self-censorship appears normal [51, 52]. In addition, such transgenic mice exhibit a low, but significant, incidence of malignant lymphoma. Intriguingly, bcl-2 synergizes with c-myc in tumour progression—co-expression of c-Myc with Bcl-2 leads to a much greater tumour incidence than with either gene alone. Such synergy is also observed in vitro where Bcl-2 cooperates with c-Myc to immortalize pre-B-cells [48, 53]. Studies in fibroblasts have demonstrated that bcl-2 and c-myc synergize because Bcl-2 suppresses c-Myc induced apoptosis whilst leaving its proliferating activities unaffected [36, 54, 55]. This type of synergy between bcl-2 and c-myc oncogenes provides the first example of a novel mechanism of oncogene co-operation which differs from ‘classical’ oncogene cooperation of the type observed between, for example, c-myc and activated ras. Activated ras does not suppress c-Myc-induced apoptosis. In addition to its role in the development of neoplasia, Bcl-2 may also be important in the resistance of tumours to chemotherapeutic drugs. Bcl-2 protects fibroblasts from death following treatment with the topoisomerase II inhibitor etoposide, a drug frequently used in chemotherapy [36]. Its expression also protects cells from the cytocidal actions of a vast array of cytotoxic agents [56–58]. Bcl-2 does not appear to mediate its effects through enhanced capacity to exclude or metabolize cytotoxic drugs, nor does it confer resistance to DNA damage by genotoxic agents. Rather, it acts by suppressing the tendency of damaged cells to commit suicide [59, 60]. This is a novel mechanism of drug resistance with important implications for the design of future cancer therapies. Tumour cell lines in vitro transfected with a bcl-2-expressing plasmid are significantly more resistant to treatment with several different chemotherapeutic drugs. Not only do they survive the period of drug exposure but they go on to form colonies in semi-solid media after treatment [57]. This implies that tumour cells that inappropriately express bcl-2 may survive higher doses of chemotherapeutic agents in vivo despite sustaining significant genetic damage. Such mutated cells would then be free to proliferate after treatment is finished and their mutated phenotypes might well then form the basis of further and more aggressive tumour growth. Although Bcl-2 protects against apoptosis in an variety of biological systems, its molecular functions remain elusive. Bcl-2 knockout mice exhibit stunted growth and postnatal mortality [61]. Initially, their
70
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
haematopoietic development proceeds normally but is then followed by complete apoptotic involution of primary lymphoid organs. Animals also develop polycystic kidneys with concomitant renal failure, and this is the major cause of their mortality. Finally, affected animals fail to develop pigment during second hair follicle cycle and their hair consequently turns grey at around five weeks of age. One current idea is that the bcl-2 protein may regulate the levels of reactive oxygen species (ROS) within cells [62], an idea not inconsistent with the bcl-2 knockout phenotype—polycystic kidneys also occur in animals subjected to various metabolic poisons, and the hair pigmentation cycle requires formation of melanin which also involves ROS. Another intriguing observation is the interaction between Bcl-2 and the 23 kDa R-Ras protein believed to be involved in aspects of signal transduction [63]. It is now clear that the Bcl-2 protein is but one of several related proteins whose structures and function have been conserved throughout metazoan evolution. The family currently consists of ced-9 from C. elegans, the mammalian proteins Bcl-Xs and Bcl-XL (splice variants derived from the same gene), Mcl-1, A1 and Bax, and the viral proteins p35 (baculovirus), BHRF1 (Epstein-Barr virus), VG16 (herpes saimiri), LMW5 HL (African swine fever virus) and p19E1B (adenovirus) [64–71]. Functionally, members fall into two groups. Bcl-2, Bcl-XL, BHRF1, p19E1B and Ced-9 all suppress apoptosis whereas Bax and Bcl-Xs promote it. The striking functional evolutionary conservation of bcl-2 is underscored by its ability to protect ced-9 loss-of-function mutants of C. elegans [14, 15]. Both Bax and the smaller splice variant of Bcl-X, Bcl-XS antagonize the antiapoptotic effect of bcl-2. Bax heterodimerizes with bcl-2, but it is unclear whether bcl-2 blocks a default Bax-dependent suicide function, whether Bax blocks a Bcl-2-mediated survival function, or whether the relative levels of Bcl-2 and Bax simply determine the propensity of an individual cell to undergo programmed cell death, as has been proposed [72]. However, recent elegant experiments suggest that Bcl-2 functions to suppress cell death via heterodimerization with Bax. Site-specific mutagenesis of bcl-2 indicates that the mutations that affect bcl-2 function also disrupt its heterodimerization with Bax yet still permit bcl-2 homodimerization [73]. In contrast to Bax, Bcl-XS has not been shown to heterodimerize with bcl-2 and its mode of action remains unclear. Intriguingly, double positive (CD4+ CD8+) thymocytes, most of which are programmed to die, express XS whereas long-lived neurones exclusively express Bcl-XL. Clearly it will be of interest to determine the mechanisms that regulate the splicing of Bcl-X [64]. Expression of Immediately Early Genes in Apoptosis Increased expression of immediate early genes such as c-fos and c-jun is observed in certain cells undergoing apoptosis in response to varying stimuli [28, 74]. Transient [75] or prolonged [74] increases in c-fos expression are seen in certain apoptotic cells. For example, the embryonic nervous system undergoes extensive apoptosis throughout its development and using transgenic mice expressing Lac-Z driven from a c-fos promoter this has been shown to correlate with continuous c-fos expression [74]. Both c-Fos and c-Jun are induced in IL-6 and IL-2 dependent cells upon withdrawal of growth factor. Apoptosis induced by withdrawal of either IL-2 or IL-6 can be partially inhibited with antisense oligonucleotides to either c-fos or c-jun or to both [76]. Increases in c-fos, junB, junD and c-jun mRNAs are also observed in involuting mouse mammary glands, along with a transient increase in AP-1 activity [77]. However, not all cells undergoing apoptosis express Fos or Jun. Indeed, in some cases, apoptosis is associated with a decrease in AP-1 activity [78]. Thus, expression of immediate early genes during cell death may well be cell typespecific.
GENES INVOLVED IN APOPTOSIS
71
Tumour Suppressor Genes and their Role in Apoptosis Tumour suppressor genes suppress unrestrained cell growth through their specific inhibitory effects on the cell cycle and they are implicated in the development of neoplasia following their loss from genomes of transformed and cancerous cells. One tumour suppressor gene, that encoding the protein p53, is functionally inactivated in some 70% of all human tumours. p53 has a well-described role in cell cycle control, through its transcriptional induction of the cyclin-Cdk inhibitor p21WAF-1/Cip1 inresponse to DNA damage [79, 80]. Recently, however, it has also emerged that p53 is a potent trigger of apoptosis. Ectopic expression of p53 in the p53-negative leukaemic cell line M1 results in the induction of apoptosis [81]. Why p53 induces spontaneous apoptosis in this cell line is unclear but it may be due to intrinsic damage accumulated within the genome during tumourigenesis in the absence of functional p53. The role of p53 in sensing genomic damage and triggering concomitant apoptosis appears to be of fundamental importance in its action as a tumour suppressor protein [82, 83]. The normally short-lived p53 protein becomes stabilized following genotoxic stress and this triggers cell cycle arrest in G1 and, in cells that sustain irreparable levels of DNA damage, apoptosis. Thymocytes derived from mice in which p53 has been genetically deleted (p53null mice) are resistant to induction of apoptosis by either -radiation or the drug etoposide. Furthermore, the ability of p53null cells to proliferate despite extensive genomic damage probably explains the marked genomic instability of p53-negative cells. p53 may also be involved in non-DNA damage associated triggers of apoptosis, such as cytokine withdrawal [84, 85]. A second tumour suppressor gene implicated in control of apoptosis is the retinoblastoma (rb) gene. Originally identified as the gene lost in human retinoblastoma, sporadic loss of rb is associated with many diverse human cancers. Children who inherit one mutant allele of rb have a 90% chance of acquiring loss of the second allele during retinal development resulting in bilateral retinoblastoma by the age of 3. Creation of rb knock-out mice has demonstrated that the rb gene is critical for normal embryonic development, since death occurs in developing rbnull embryos at around 14–16 days post coitus [86–88]. Death results from defects in both neuronal and haematopoietic development that become apparent around 10–12 days p.c. Prior to this, rbnull embryos are indistinguishable from their normal counterparts. Increased mitotic and apoptotic cells are evident in the spinal cord, hind and intermediate brain in rbnull embryos and, although some mature neurones are present, mitosis appears to outweigh the normal process of terminal differentiation observed in wild-type mice at the same developmental point [86, 87]. In rbnull mice, early haematopoiesis occurs normally in the yolk sac of the embryo. Stem cells from the yolk sac then seed in the liver and would normally give rise to the eight mature lineages of the haematopoietic system, including enucleated red blood cells (RBC). However, in rbnull embryos, RBC largely remain in an immature nucleated state [86, 88]. In vitro culture of Rbnull haematopoietic stem cells in the presence of cytokines required for erythropoiesis indicates that the RBC precursors are developmentally blocked and cannot attain a mature phenotype [88]. It would appear that the erythrocyte progenitor cells are unable to proceed through end-stage differentiation, resulting in levels of haemoglobin that are too low to support the developing embryo. This indicates that the rb gene product, p105rb plays a critical role in the terminal differentiation of immature red blood cells. Whether the loss of neurones in rbnull mice is due to ananalogous developmental block in neuronal development is not clear. It is possible, however, that in the absence of normal neuronal differentiation, essential interactions between neuronal cells required to prevent apoptosis fail to occur so leading to massive cell death [43, 89]. Unlike rb-negative humans, neither rbnull nor rb heterozygous mice develop retinoblastoma. However, p105rb function has been specifically ablated in mouse retinal photoreceptor cells using a vector directing photoreceptor-specific expression of the human papilloma virus gene E7 which sequesters and functionally inactivates p105rb. In such animals, substantial apoptosis occurs at the stage when photoreceptor cells are
72
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
normally undergoing terminal differentiation. Moreover, if the same experiment is conducted in p53null mice, retinoblastoma results [90]. Thus, both Rb and p53 need to be functionally inactivated in order to produce ocular tumours in mice. Perhaps loss of p105rb serves to maintain cells in an immature cycling state whilst p53 loss prevents apoptosis. Analogous results have been obtained in a different tumour model using an SV40 Large T antigen mutant (T121) that perturbs p105rb function but does not interact with p53 [91]. These mice develop low grade brain tumours, a function dependent on ablation of p105rb activity. In mice transgenic for wt T antigen much more aggressive tumours occur, resulting in death of the mice within 6 weeks. Expression of the T121 fragment in p53null mice also results in highly malignant tumours. However, when expressed in the background of p53+/null heterozygotes, low grade tumours predominate at first, although nodules of more malignant cells appear eventually. These malignant cells have all lost their functional p53 allele, demonstrating conclusively that in the absence of functional p105rb loss of p53 is required for highly metastatic tumours to evolve, whereas in the presence of functional p53 tumours still arise but tumour growth is much inhibited by extensive apoptosis. 5.3 Viral Gene Suppression of Apoptosis p53 is the target of many DNA tumour virus genes that sequester or functionally inactivate the protein in order to restrict the cell’s capacity to arrest growth and commit suicide. Presumably, this is a useful strategy for the virus because it promotes proliferation of the infected cells and, at the same time, suppresses the ability of the cell to kill both itself and the infecting virus. Indeed, many different viruses have evolved a variety of mechanisms to suppress apoptosis suggesting that suppression of host cell suicide may be a common and important feature of virus infection. Adenovirus p55E1B and human papilloma virus E6 both bind p53 [92–94]. p55E1B inactivates the transcriptional activity of p53 by converting its transactivation domain into a trans-repressor [95] whilst HPV E6 targets p53 for rapid destruction [96]. Hence, both genes serve to suppress normal p53 activity and its induction of G1 arrest and apoptosis in response to DNA damage [83, 97–99]. Suppression of apoptosis by viral genes is not only prevalent in mammalian viral models but is also important in promoting efficient infection of insect hosts by viruses such as Baculovirus [100]. 5.3.1 Anti-apoptotic Insect Virus Genes In 1991, Clem and colleagues isolated a baculovirus gene product, p35, that delays apoptosis in infected insect cells. A second gene, iap, (inhibitor of apoptosis) has been isolated from a related baculovirus strain [100]. Homologues of p35 [101] and Iap [100, 102] have also been identified in related baculovirus virus strains. One known virus homologue of Iap shares 60% amino acid homology and suppresses apoptosis in viral complementation assays when co-transfected with an iap-deficient mutant p35 baculovirus. A second IAP homologue has a lower (30%) amino acid homology and does not complement mutant p35 function. All three virus iap homologues encode polypeptides that contain a C3HC4 zinc finger together with a characteristic series of amino acid repeats—these regions may be important for the anti-apoptotic function of these proteins [103]. The C3HC4 type of zinc finger is also found in some 30 other known polypeptides, including those encoded by the mammalian genes ret, bmi-1, pml-RARA, mel-18 and c-abl, [100]. Of these, c-abl at least possesses anti-apoptotic activity [104] and bmi-1 co-operates with c-Myc and may also suppress apoptosis.
GENES INVOLVED IN APOPTOSIS
73
Infection of SF-2 insect cells by baculovirus normally produces progeny virions by about 24 hours post infection. Infection of cells with virus lacking either p35 or iap results in apoptosis of infected cells between 9 and 12 hours post infection, severely compromising production of progeny virions. Normally, expression of p35 occurs at around 6–9 hours post infection and this temporally correlates with an apparent suppression of apoptosis. In fact, limited blebbing of the host cell cytoplasm is often evident at this time in cells infected with wild type virus but in cells infected by virus with defective p35, the blebbing increases and apoptosis proceeds. Transcription of host cell genes normally declines as a late function of virus infection, around the same time as apoptosis is induced by p35-negative mutants. Precisely how p35 and iap are able to inhibit host cell apoptosis is at present unclear but both genes also suppress apoptosis induced by the RNA synthesis inhibitor actinomycin D. Thus, both iap and p35 must interact with as yet unknown pre-existing cellular component(s) of the apoptotic pathway(s) [105]. Infection of TN-368 cells (derived from a different lepidopteran species from that from which SF-2 cells were derived) with p35 deficient virus does not result in apoptosis and viral production appears essentially normal [66]. Unlike SF-2 cells, TN-368 do not undergo apoptosis upon treatment with actinomycin D. Several inferences may be drawn from these data. The cessation of host cell gene expression that leads to apoptosis may comprise part of a host cell defence mechanism which p35 and iap have evolved to overcome [103]. However, in cell lines that are not triggered to undergo cell death by RNA synthesis inhibition the presence of such genes is presumably not required. Therefore, p35 and iap represent typical host-range genes that have presumably evolved to widen the number of host cell types permissive for baculovirus replication. If so, such virus genes may not, like bcl-2, be universal inhibitors of apoptosis but may instead interact very specifically with components of apoptotic pathway(s) peculiar to specific cell types. Consistent with this, neither p35 nor iap shares any homology with bcl-2 and, moreover, neither bcl-2 nor adenovirus p19E1B is able to prevent apoptosis during baculovirus infection in the absence of either baculovirus gene [103]. 5.3.2 Anti-apoptotic Genes in Mammalian Viruses Genes with apparent anti-apoptotic activity are present in many viruses that infect mammalian cells. The herpes simplex virus (HSV) gene 1 34.5 encodes a protein, infected cell protein (ICP-34.5), which is required for neurovirulence during HSV infection [106]. Unlike wild-type HSV, viruses lacking 1 34.5 are unable to effectively infect neuronal cells and are no longer lethal when injected into mice. One of the actions of ICP-34.5 is to prevent the shut-down of host cell protein synthesis in infected neurones which normally triggers apoptosis [107]. Analogous effects are also mediated by host range genes in other viruses, for example the CHOhr gene of cow pox virus (CPV). Vaccinia virus infection of non-permissive CHO cells results in early shut-off both host cell and viral protein synthesis and subsequent apoptotic cell death. Death can be prevented in these cells by expression of the CPV host range gene (CHOhr) which results in CHO cells becoming permissive for VV replication [108]. Expression of CHOhr suppresses CHO cell apoptosis during early infection and so permits late viral protein synthesis to occur. This results in release of mature virus upon the lytic death of the cell 15–20 hours after infection. Interestingly, expression of the adenovirus gene p19E1B also prevents CHO cell apoptosis in response to VV infection. However, p19E1B expression is not sufficient to permit expression of late virus genes and productive infection [109], a result comparable to p35 replacement by p19E1B in baculovirus. This implies that host range genes such as CHOhr do more than merely suppress apoptosis of host cells. Another CPV host range gene is crm A, which encodes a specific inhibitor of the ICE protease, a mammalian homologue of the nematode ced-3 killer gene product (see above). Expression of the CrmA protein prolongs survival of dorsal root ganglion neurones deprived of
74
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
NGF or when micro-injected with a plasmid expressing ICE [21] (see section 2.1). CrmA is required for effective virus propagation in vivo. In addition to suppressing apoptosis, CrmA may also block processing and release of IL-1 . This may suppress the recruitment of other immune cells into the infected area, enabling the release of viral particles to progress unchecked for a longer time and so increase virulence. Transformation of primary rodent cells by adeno virus requires both E1A and E1B protein products [110]. Although E1A alone is capable of inducing cell cycle progression and low level cellular immortalization, coexpression of E1B is required for high efficiency transformation. The E1B gene encodes two distinct proteins of 19 kDa and 55 kDa, either of which markedly enhances the transforming ability of E1A. Transformation of cells by E1A alone is inefficient because E1A alone is a potent inducer of apoptosis in infected cells. Such apoptosis is suppressed by expression of the p19E1B protein, a potent anti-apoptotic polypeptide of the same functional class as Bcl-2 [111–113]. As with c-myc, bcl-2 also co-operates with E1A to allow formation of foci that would otherwise regress by apoptosis [114]. The larger 55 kDa E1B protein binds p53 and also acts to suppress E1A-induced apoptosis which is, in part, p53-dependent [115, 116]. Why should a virus employ two independent strategies for suppressing apoptosis, one analogous to bcl-2 (19 kDa) and the other to inactivate p53 (55 kDa)? It may be that p19E1B and p55E1B, by interacting with different parts of the host cell apoptotic machinery provide a synergistic antiapoptotic function [114]. In this context, it is interesting to note that p19E1B has been shown to inhibit apoptosis in response to diverse stimuli, including anti-Fas antibodies or TNF [112, 113], neither of which operates through p53. The Epstein-Barr virus is a -herpesvirus that has evolved two independent mechanisms for suppressing host cell apoptosis. Infection of mature resting peripheral human or primate B-cells by EBV produces latently infected lymphoblastoid cell lines (LCLs) that exhibit continuous proliferation in vitro [117]. Once established within a host cell, EBV can embark upon either a latent or a lytic life-cycle. In a latent cycle, no progeny virus is produced: instead, only a few (1–8) ‘latent’ genes are expressed. In contrast, the lytic cycle leads to production of progeny virus and involves expression of nearly all of the ~100 viral genes. In general, continuously proliferating LCLs in culture express the restricted pattern of eight EBV latent genes (EBNAs 1, 2, 3A, 3B, 3C, LP, LMP1 and 2(TP)). In contrast, EBV-positive Burkitt’s lymphoma cells derived from biopsies of tumours express one latent gene, EBNA 1 [118]. These different ‘forms’ of latency are referred to as group I or group III cell lines respectively and each has differing sensitivities to the induction of apoptosis in response to various stimuli [119]. Group I cells readily enter apoptosis in response to serum deprivation or the presence of calcium ionophores whereas group III cells are insensitive to such stimuli. The insensitivity of Group III cells to apoptosis appears in part to be due to high levels of expression of bcl-2 that is induced by expression of the EBV membrane protein LMP1 [120]. The underlying mechanism by which LMP1 induces bcl-2 expression is unknown and may be restricted to cells of the B lineage [121]. A second EBV gene involved in suppression of host cell apoptosis is BHRF1 (BamH1 fragment H Rightward open reading Frame 1) [69]. BHRF1 encodes a putative transmembrane protein, [69, 122] with ~25% amino acid sequence homology to Bcl-2 [123, 124] and similar anti-apoptotic activity [125, 126] (N.McCarthy, unpublished results; A.Fanidi, personal communication). Unlike p19E1B, which is required for cellular transformation by adenovirus, BHRF1 is not required for EBV infection or transformation in vitro [127]. Thus although BHRF1 is anti-apoptotic, its functional role within EBV biology is unclear.
GENES INVOLVED IN APOPTOSIS
75
5.4 Fas/Apo-1-induced Apoptosis The Fas/Apo-1 receptor is expressed on several different cell types including B and T cells and belongs to a protein receptor superfamily comprising the tumour necrosis factor receptor, the low affinity nerve growth factor receptor and CD40 [128]. Binding of anti-Fas/Apo-1 antibodies to the Fas/Apo-1 receptor induces apoptosis within the target cell [8, 129] and this response is important for several of the kill mechanisms involved in the immune response [130, 131]. Fas/Apo-1-induced apoptosis is one component of the kill pathway activated during cytotoxic T-cell mediated killing. This was initially shown in T-cell hybridomas specifically selected for a reduced number of cell surface ligands, so allowing the ligand-receptor interactions between effector and target cells to be identified. Interactions between the Fas ligand and the Fas receptor represent a calcium-independent component of T-cell-mediated killing illustrating that this pathway is distinct from the calcium-requiring perforin death pathway [132, 133]. Although different, the Fas/Apo-1 and perforin pathways are complementary: both pathways together are sufficient for 100% of detectable cell-mediated cytotoxicity in vitro. Perforin induced death occurs in ~66% of effector/target cell interactions, with Fas/ Apo-1 inducing the remaining 33% of target cell death. This ratio may differ with respect to either Fas receptor expression on target cells, and may be further modulated by the presence of different pathogens, or by variations in perforin content and Fas ligand expression in CTLs [134]. As well as its role in killing immunocyte target cells, Fas/Apo-1-triggered apoptosis is also involved in the removal of the surplus cytotoxic T- and B-cells that persist after elimination of an invading pathogen. Activated CTLs in vitro express Fas/Apo-1 in an inactive state, i.e. triggering of the receptor with monoclonal antibodies does not induce apoptosis. Such cells do not become responsive to Fas/Apo-1induced death until a lag period of ~7 days has elapsed [130, 132]. Thus, in these cells, coupling of the apoptotic machinery to the Fas/Apo-1 receptor is delayed. This allows for CTL killing of the invading pathogen prior to initiating the clearance of active T-cells to terminate the immune response. A similar response is observed in activated B-cells [130]. In the absence of either of the Fas/Apo-1 receptor or ligand, not only is the termination of the immune response compromised but the elimination of self-reactive thymocytes is also compromised. Mice lacking either Fas receptor expression (lpr mice) or mice lacking Fas ligand expression (gld mice) both have profound lymphadenopathy as well as autoimmune disease similar to human systemic lupus erythromatosis. Thus the identification of the gene products involved in the Fas/ Apo-1 pathway could reveal novel targets for pharmacological intervention in the treatment of autoimmune diseases. Further investigation into the Fas/Apo-1 signalling pathway should also facilitate identification of new proteins involved in the regulation of programmed cell death. 5.5 ‘Death’ Genes Inhibitors of protein synthesis delay apoptosis in many cell systems. This implies that the affected cell is somehow active in its own death—that is, apoptosis requires either the continued or the de novo synthesis of particular gene products before apoptosis can occur. Some, but not all, such gene products might be specific to the cell death programme and would therefore be particularly interesting as targets for therapy in diseases in which apoptosis plays a part. Apoptosis during regression of the rat ventral prostate was one of the first mammalian models in which genes expressed during the cell death period were analysed [135–140]. Prostate epithelial cells re-enter cell
76
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
cycle upon androgen withdrawal and express c-fos (transiently) and c-myc (continuously). However, once in cycle few cells reach G2/M phase due to the rapid onset of apoptosis [141]. p53 is stabilized in apoptotic prostate cells but it is unclear whether p53 stabilization is due to DNA damage incurred when cells re-enter cell cycle in the absence of androgen or whether p53 is induced via a non DNA-damage pathway upon removal of the hormone [141]. The search for bona fide ‘death’ genes—i.e. genes that are exclusively expressed prior to cell death—has not proven fruitful. Subtractive hybridization experiments in thymocytes identified a number of cDNAs that are expressed in cells undergoing apoptosis in response to dexamethasone [142–146] but none appears to encode a protein that is specifically required for a thymocyte to die. Arguably, genes that are directly required for apoptosis may be more readily identified by dissection of triggering pathways, such as those induced by growth factor withdrawal or upon Fas/ Apo-1 stimulation. 5.6 Summary—Apoptosis and the Possibilities for Novel Pharmacological Intervention in Disease We have reviewed several important aspects of programmed cell death (apoptosis). First, we have argued that programmed cell death is a highly conserved process essential for maintaining tissue homeostasis. Cells that outgrow their paracrine supply of survival factors are programmed to die [43]. This ‘default’ death programme is activated concomitantly with the cell’s proliferative programme. Thus, every cell that enters cycle, and thereby presents a neoplastic risk to the organism, can survive only so long as its death is suppressed by survival factors [3, 28]. Thus, suppression of neoplasia is ‘hardwired’ into the growth of metazoan cells. Second, we have indicated that apoptosis is regulated by several known genes, amongst them oncogenes, tumour suppressor genes and proteases. In this summary section, we take two examples of genes that regulate apoptosis, the bcl-2 and ICE/Ced-3/ Nedd2 families, and use them to illustrate ways in which it may be possible to intervene in apoptosis pharmacologically. Genes that regulate cell death are present in both invertebrates and vertebrates, indicating that programmed cell death has been conserved throughout multicellular evolution. Suppression of cell death during development of the nematode worm C. elegans is mediated by the gene ced-9, a member of the bcl-2 gene family which comprises a number of related genes involved in regulating apoptosis. Bcl-2 itself, and the functional homologues Bcl-XL, BHRF1 and p19E1B, all inhibit apoptosis whereas Bax and Bcl-XS antagonize the anti-apoptotic action of Bcl-2-like proteins and so promote programmed cell death. It is known that Bcl-2 and Bax homodimerize and heterodimerize with each other and it has been proposed that the relative intracellular concentrations of Bax homodimers versus Bax/Bcl-2 heterodimers sets the innate tendency of a cell to survive or die under particular conditions [73]. The full extent of possible interactions between individual Bcl-2 family members is not yet clear but it is likely that dimeric interactions are possible between many members of the Bcl-2 and Bax family. Evidence is accumulating to support the notion that the maintenance of a functional cell death pathway is essential for preventing aberrant cell growth such as neoplasia. Thus, the antagonistic roles played by Bcl-2 and Bax, and their homologues, in determining cell survival are likely to be important factors in carcinogenesis. A second class of gene involved in the cell death pathway is the ced-3/nedd-2/ICE gene family. Three mammalian homologues of the C. elegans ced-3 death gene product have been identified—ICE, prICE and Ich-1. Like CED-3, these proteins are all cysteine proteases [16, 22, 24]. A fundamental question is where these proteases act within the death pathway—are they triggers that commit a cell to apoptosis or merely factors that influence the tendency of cell to undergo apoptosis? Both CED-3 and active ICE induce
GENES INVOLVED IN APOPTOSIS
77
apoptosis when expressed ectopically in various cell types [19, 21, 147]. However, the action of both CED-3 and active ICE in triggering apoptosis can apparently be modulated by either Bcl-2 or survival factors, suggesting either that both of these anti-apoptotic agents serve to directly modulate activity of the protease or that they act downstream of the ICE protease function. As regulatable triggers of apoptosis, the ICE proteases present attractive candidate targets for therapies aimed at modulating apoptosis. This is because the cleavage of a substrate by cysteine proteases is a well understood molecular mechanism that had been extensively characterized prior to the discovery of the role of ICE proteases in apoptosis [17, 18]. Several peptide and polypeptide inhibitors of ICE are already known, including the product of the CrmA gene of poxvirus—a serpin specific for ICE which can block ICE-induced apoptosis when expressed [21]. Moreover, the co-crystal structure of ICE and its substrate has recently been resolved [148]. This will greatly facilitate the rational design of inhibitors of the ICE family proteases. Moreover, identifying substrates that are cleaved by this family of proteases may provide a step forward in understanding how these particular genes regulate apoptosis. Interactions Between Genes that Regulate Apoptosis and their Relevance to Tumourigenesis As discussed, p53, Bcl-2, Bax, pRB and c-Myc all act to control apoptosis in varying circumstances. However, it is not yet clear whether any one of these genes is required to regulate apoptosis in response to every apoptotic stimulus that any given cell may encounter. For example, p53 is clearly involved in apoptosis induced by DNA damage but it may also be involved in pathways of apoptosis where damage to the DNA is not a trigger, such as androgen withdrawal from rat prostate cells [141]. Lesions in genes that suppress apoptosis are often identified in tumours because inhibition of cell death appears to be an important component of carcinogenesis. Therefore, it is important to establish whether specific genes are involved in one or many pathway(s) controlling apoptosis because any one gene that is required for many instances of cell death is presumably a prime candidate for mutation during tumourigenesis. Bcl-2 is one such gene product whose expression suppresses apoptosis in response to several stimuli. Bcl-2 co-operates with c-Myc in tumourigenesis by providing an anti-apoptotic signal that blocks Myc induced death favouring Myc induced proliferation. Bcl-2 in this instance effectively replaces the survival signal normally provided by specific cytokines [3, 28]. However, unlike cytokines, Bcl-2 is not a restricted survival source in vivo, hence its expression significantly increases a cells likelihood of surviving in micro-environments that are not optimal for cell survival, as illustrated in mice transgenic for Bcl-2. Transgenic expression of Bcl-2 increases the numbers of pre-B- and B-cells and although this is not clinically detrimental to the animal at first, after 12 months clonal tumours appear. In this instance the increase in B-cell numbers has raised the chance of a cell undergoing further genetic mutations leading to malignant outgrowth [149]. However, expression of Bcl-2 does not completely block a cell from entering apoptosis. Rather, it limits the propensity for an individual cell to undergo apoptosis in response to a given trigger. Thus, cells expressing Bcl-2 are more likely to survive the effects of cytotoxic drugs but are not immune to their effects and can still be killed by extended or elevated levels of the same drug. Of course, the problem then becomes one of specificity—how to kill a resistant tumour cells under conditions where normal somatic cells survive. The greatest risk must be that tumour cells with suppressed apoptosis may well survive genotoxic damage that kills other cells, and then go on to propagate despite the DNA damage they have incurred. In such a circumstance, the chemotherapeutic drug would act to drive mutation and concomitant evolution of the surviving tumour cells—arguably the worst conceivable outcome of therapy because progressively more aggressive clonal variants may result, possibly through lesions in yet other
78
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
genes that suppress differing apoptotic pathways [57]. By identifying genes that are important in the regulation of apoptosis, and the lesions that affect such genes that arise during carcinogenesis, it may be possible to develop drugs that specifically target and correct apoptotic lesions. The result of such drugs would be the suicidal elimination of tumour cells, surely the ideal result of any anti-neoplastic therapy. References 1 2 3 4
5
6
7 8 9 10 11 12 13 14 15 16 17
18
19
. WILLIAMS, G.T. (1991) Programmed cell death: apoptosis and oncogenesis. Cell 65, 1097–1098. . WILLIAMS, G.T., SMITH, C.A., SPOONCER, E., DEXTER, T.M. & TAYLOR, D.R. (1990) Haemopoietic colony stimulating factors promote cell survival by suppressing apoptosis. Nature 343, 76–79. . HARRINGTON, E., FANIDI, A., BENNETT M. & EVAN, G. (1994) Modulation of Myc-induced apoptosis by specific cytokines. EMBO J. 13, 3286–3295. . NUNEZ,G., LONDON, L., HOCKENBERY, D., ALEXANDER, M., MCKEARN, J.P. & KORSMEYER, S.J. (1990) Deregulated Bcl-2 gene expression selectively prolongs survival of growth factor-deprived hemopoietic cell lines. J. Immunol. 144, 3602–3610. . SMITH, C.A., WILLIAMS, G.T., KINGSTON, R., JENKINSON, E.J. & OWEN, J.J.T. (1989) Antibodies to CD3/T-cell receptor complex induce death by apoptosis in immature T cells in thymic cultures. Nature 337, 181–184. . MCCONKEY, D.J., HARTZELL, P., AMADOR, P.J., ORRENIUS, S. & JONDAL, M. (1989) Calciumdependent killing of immature thymocytes by stimulation via the CD3/T cell receptor complex. J. Immunol. 143, 1801–1806. . LASTER, S.M., WOOD, J.G. & GOODING, L.R. (1988) Tumor necrosis factor can induce both apoptic and necrotic forms of cell lysis. J. Immunol. 141, 2629–2634. . YONEHARA, S., ISHII, A. & YONEHARA, M. (1989) A cell-killing monoclonal antibody (anti-Fas) to a cell surface antigen co-downregulated with the receptor of tumor necrosis factor. J. Exp. Med. 169, 1747–1756. . ELLIS, H.M. & HOROVITZ, H.R. (1986) Genetic control of programmed cell death in the nematode C. elegans. Cell 44, 817–829. . HEDGECOCK, E.M., SULSTON, J.E. & THOMSON, J.N. (1983) Mutations affecting programmed cell deaths in the nematode Caenorhabditis elegans. Science 220, 1277– 1279. . SULSTON, J.E., SCHIERENBERG, E., WHITE, J.G. & THOMSON, J.N. (1983) The embryonic cell lineage of the nematode Caenorhabditis elegans. Dev. Biol. 100, 64–119. . YUAN, J.Y. & HORVITZ, H.R. (1990) The Caenorhabditis elegans genes ced-3 and ced-4 act cell autonomously to cause programmed cell death. Dev. Biol. 138, 33–41. . HENGARTNER, M.O., ELLIS, R.E. & HORVITZ, H.R. (1992) Caenorhabditis elegans gene ced-9 protects cells from programmed cell death. Nature 356, 494–499. . VAUX, D.L., WEISSMAN, I.L. & KIM, S.K. (1992) Prevention of programmed cell-death in Caenorhabditiselegans by human bcl-2. Science 258, 1955–1957. . HENGARTNER, M. & HORVITZ, H. (1994) C. elegans cell survival gene ced-9 encodes a functional homolog of the mammalian proto-oncogene bcl-2. Cell 76, 665–676. . YUAN, J.Y., SHAHAM, S., LEDOUX, S., ELLIS. H.M. & HORVITZ, H.R. (1993) The C-elegans cell-death gene ced-3 encodes a protein similar to mammalian interleukin-1 beta converting enzyme. Cell 75, 641–652. . THORNBERRY, N.A., BULL, H.G., CALAYCAY, J.R., CHAPMAN, K.T., HOWARD, A.D., KOSTURA, M.J., MILLER, D.K., MOLINEAUX, S.M., WEIDNER, J.R., AUNINS, J. et al. (1992) A novel heterodimeric cysteine protease is required for interleukin-1 beta processing in monocytes. Nature 356, 768–774. . CERRETTI, D.P., KOZLOSKY, C.J., MOSLEY, B., NELSON, N., VAN, N.K., GREENSTREET, T.A., MARCH, C.J., KRONHEIM, S.R., DRUCK, T., CANNIZZARO, L.A. et al. (1992) Molecular cloning of the interleukin-1 beta converting enzyme. Science 256, 97–100. . MIURA, M., ZHU, H., ROTELLO, R., HARTWIEG, E.A. & YUAN, J.Y. (1993) Induction of apoptosis in fibroblasts by IL-1-beta-converting enzyme. Cell 75, 653–660.
GENES INVOLVED IN APOPTOSIS
20
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
41 42 43
79
. RAY, C.A., BLACK, R.A., KRONHEIM, S.R., GREENSTREET, T.A., SLEATH, P.R., SALVESEN. G.S. & PICKUP, D.J. (1992) Viral inhibition of inflammation: cowpox virus encodes an inhibitor of the interleukin-1 beta converting enzyme. Cell 69, 597–604. . GAGLIARDINI, V., FERNANDEZ, P.A., LEE, R.K., DREXLER, H.C., ROTELLO, R.J., FISHMAN, M.C. & YUAN, J. (1994). Prevention of vertebrate neuronal death by the crmA gene. Science 263, 826–828. . LAZEBNIK, Y., KAUFMANN, S. DESNOYERS, S. POIRIER, G. & EARNSHAW, W. (1994) Cleavage of poly (ADP-ribose) polymerase by a proteinase with properties like ICE. Nature 371, 346–347. . SMITH, C.A., MCCARTHY, N.J. & WILLIAMS, G.T. (1993). Cell recognition of apoptotic cells. In: Blood Cell Biochemistry. 5: Macrophages and related cells. M.A. Horton (ed). Plenum Press, New York, pp. 393–421. . WANG, L., MIURA, M., BERGERON, L., ZHU. H. & YUAN, J. (1994) Ich-1, an Ice/ced-3-related gene, encodes both positive and negative regulators of programmed cell-death. Cell 78, 739–750. . SPENCER, C.A. & GROUDINE, M. (1991) Control of c-myc regulation in normal and neoplastic cells. Adv. Cancer Res. 56, 1–48. . KELLY, K., COCHRAN, B.H., STILES, C.D. & LEDER, P. (1983) Cell specific regulation of the c-myc gene by lymphocyte mitogens and platelet-derived growth factor. Cell 35, 603–610. . EILERS, M., SCHIRM, S. & BISHOP, J.M. (1991) The MYC protein activates transcription of the alphaprothymosin gene. EMBO J. 10, 133–141. . EVAN, G., WYLLIE, A., GILBERT, C., LITTLEWOOD, T., LAND, H., BROOKS, M., WATERS, C., PENN. L. & HANCOCK, D. (1992) Induction of apoptosis in fibroblasts by c-myc protein. Cell 63, 119–125. . KATO, G.J., BARRETT, J., VILLA, G.M. & DANG, C.V. (1990) An aminoterminal c-myc domain required for neoplastic transformation activates transcription. Mol Cell Biol. 10, 5914–5920. . KRETZNER, L., BLACKWOOD, E. & EISENMAN, R. (1992) Myc and Max possess distinct transcriptional activities. Nature 359, 426–429. . AMATI, B., DALTON, S., BROOKS, M., LITTLEWOOD, T., EVAN, G. & LAND, H. (1992) Transcriptional activation by c-Myc oncoprotein in yeast requires interaction with Max. Nature 359, 423–426. . AMATI, B., BROOKS, M., LEVY, N., LITTLEWOOD, T., EVAN, G. & LAND. H. (1993) Oncogenic activity of the c-Myc protein requires dimerisation with Max. Cell 72, 233–245. . NEIMAN, P.E., THOMAS, S.J. & LORING, G. (1991) Induction of apoptosis during normal and neoplastic Bcell development in the bursa of Fabricius. Proc. Natl Acad. Sci. USA 88, 5857–5861. . LANGDON, W.Y., HARRIS. A.W. & CORY, S. (1988) Growth of Eµ-myc transgenic B-lymphoid cells in vitro and their evolution toward autonomy. Oncogene Res. 3, 271– 279. . DYALL, S.D. & CORY, S. (1988) Transformation of bone marrow cells from Eµ-myc transgenic mice by Abelson murine leukemia virus and Harvey murine sarcoma virus. Oncogene Res. 2, 403–409. . FANIDI, A., HARRINGTON, E. & EVAN, G. (1992) Cooperative interaction between c-myc and bcl-2 protooncogenes. Nature 359, 554–556. . ASKEW, D., ASHMUN, R., SIMMONS, B. & CLEVELAND, J. (1991) Constitutive c-myc expression in IL-3dependent myeloid cell line suppresses cycle arrest and accelerates apoptosis. Oncogene 6, 1915–1912. . SHI, Y., GLYNN, J., GUILBERT, L., COTTER, T., BISSONETTE, R. & GREEN, D. (1992) Role for c-myc in activation-induced apoptotic cell death in T cell hybridomas. Science 257, 212–214. . JANICKE, R., LEE, F. & PORTER, A. (1994) Nuclear c-myc plays an important role in the cytotoxicity of tumor-necrosis-factor-alpha in tumor-cells. Mol. Cell. Biol. 14, 5661– 5670. . FERNANDEZ, A., MARIN, M.C., MCDONNELL, T., & ANANTHASWAMY, H.N. (1994) Differential sensitivity of normal and Ha-ras-transformed C3H mouse embryo fibroblasts to tumor necrosis factor: induction of bcl-2, c-myc, and manganese superoxide dismutase in resistant cells. Oncogene 9, 2009–2017. . AMATI, B., LITTLEWOOD, T., EVAN. G. & LAND, H. (1994) The c-Myc protein induces cell cycle progression and apoptosis through dimerisation with Max. EMBO J. 12, 5083–5087. . BARRES, B.A., HART, I.K., COLES, H.S., BURNE, J.F., VOYVODIC, J.T., RICHARDSON, W.D. & RAFF, M.C. (1992) Cell death in the oligodendrocyte lineage. J. Neurobiol. 23, 1221–1230. . RAFF, M.C. (1992) Social controls on cell survival and cell death. Nature 356, 397– 400.
80
44 45 46 47 48 49 50
51 52 53 54 55 56 57
58
59
60
61 62 63 64
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. RODRIGUEZ, T.G., COLLINS, M.K., GARCIA, I. & LOPEZ, R.A. (1992) Insulin-like growth factor-I inhibits apoptosis in IL-3-dependent hemopoietic cells. J. Immunol. 149, 535–540. . COLLINS, M.K., MARVEL, J., MALDE, P. & LOPEZ-RIVAS, A. (1992) Interleukin 3 protects murine bone marrow cells from apoptosis induced by DNA damaging agents. J. Exp. Med. 176, 1043–1051. . CROSS, M. & DEXTER, T.M. (1991) Growth factors in development, transformation, and tumorigenesis. Cell 64, 271–280. . TSUJIMOTO, Y., COSSMAN, J., JAFFE, E. & CROCE, C.M. (1985) Involvement of the bcl-2 gene in human follicular lymphoma. Science 228, 1440–1443. . VAUX, D.L., CORY, S. & ADAMS, J.M. (1988) Bcl-2 gene promotes haemopoietic cell survival and cooperates with c-myc to immortalize pre-B cells. Nature 335, 440–442. . ALLSOPP, T.E., WYATT, S., PATERSON, H.F. & DAVIES, A.M. (1993) The proto-oncogene bcl-2 can selectively rescue neurotrophic factor-dependent neurons from apoptosis. Cell 73, 295–307. . HOCKENBERY, D.M., ZUTTER, M., HICKEY, W., NAHM. M. & KORSMEYER, S.J. (1991) BCL2 protein is topographically restricted in tissues characterized by apoptotic cell death. Proc. Natl Acad. Sci. USA 88, 6961–6965. . SENTMAN, C.L., SHUTTER, J.R., HOCKENBERY, D., KANAGAWA, O. & KORSMEYER, S.J. (1991) bcl-2 inhibits multiple forms of apoptosis but not negative selection in thymocytes. Cell 67, 879–888. . STRASSER, A., HARRIS. A.W. & CORY, S. (1991) bcl-2 transgene inhibits T cell death and perturbs thymic self-censorship. Cell 67, 889–899. . STRASSER, A., HARRIS, A.W., BATH, M.L. & CORY, S. (1990) Novel primitive lymphoid tumours induced in transgenic mice by cooperation between myc and bcl-2. Nature 348, 331–333. . BISSONNETTE, R., ECHEVERRI, F., MAHBOUBI, A. & GREEN, D. (1992) Apoptotic cell death induced by c-myc is inhibited by bcl-2. Nature 359, 552–554. . WAGNER, A.J., SMALL, M.B. & HAY, N. (1993) Myc-mediated apoptosis is blocked by ectopic expression of bcl-2. Mol. Cell Biol. 13, 2432–2440. . TSUJIMOTO, Y. (1989) Stress-resistance conferred by high level of bcl-2 alpha protein in human B lymphoblastoid cell. Oncogene 4, 1331–1336. . MIYASHITA, T. & REED, J.C. (1992) bcl-2 gene transfer increases relative resistance of S49.1 and WEHI7.2 lymphoid cells to cell death and DNA fragmentation induced by glucocorticoids and multiple chemotherapeutic drugs. Cancer Res. 52, 5407–5411. . WALTON, M.I., WHYSONG, D., O’CONNOR, P.M., HOCKENBERY, D., KORSMEYER, S.J. & KOHN, K.W. (1993) Constitutive expression of human Bcl-2 modulates nitrogen mustard and camptothecin induced apoptosis. Cancer Res. 53, 1853–1861. . FISHER, T.C., MILNER, A.E., GREGORY, C.D., JACKMAN, A.L., AHERNE, G. W., HARTLEY, J.A., DIVE, C. & HICKMAN, J.A. (1993) bcl-2 modulation of apoptosis induced by anticancer drugs: resistance to thymidylate stress is independent of classical resistance pathways. Cancer Res. 53, 3321–3326. . KAMESAKI, S., KAMESAKI, H., JORGENSEN, T., TANIZAWA, A., POMMIER, Y. & COSSMAN, J. (1993) Bcl-2 protein inhibits etoposide-induced apoptosis through its effects on events subsequent to topoisomerase II-induced DNA strand breaks and their repair. Cancer Res. 53, 4251–4256. . VEIS, D., SORENSON, C., SHUTTER, J. & KORSMEYER, S. (1993) Bcl-2-deficient mice demonstrate fulminant lymphoid apoptosis, polycystic kidneys and hypopigmented hair. Cell 75, 229–240. . HOCKENBERY, D., OLTVAI, Z., YIN, X., MILLIMAN, C. & KORSMEYER, S.J. (1993) Bcl-2 functions in an antioxidant pathway to prevent apoptosis. Cell 75, 241–251. . FERNADEZ-SARABIA, M. & BISCHOFF, J. (1993) Bcl-2 associates with the ras-related protein R-ras p23. Nature 366, 274–275. . BIOSE, L., GONZÁLEZ-GARCIA, M., POSTEMA, C., DING, L., LINDSTEN, T., TURKA, L., MAO, X., NUÑEZ, G. & THOMPSON, C. (1993) bcl-x, a bcl-2-related gene that functions as a dominant regulator of apoptotic cell death. Cell 74, 597–608.
GENES INVOLVED IN APOPTOSIS
65
66 67 68 69
70
71
72 73 74
75
76
77
78 79
80 81 82 83 84
81
. KOZOPAS, K.M., YANG, T., BUCHAN, H.L., ZHOU, P. & CRAIG, R.W. (1993) MCL1, a gene expressed in programmed myeloid cell differentiation, has sequence similarity to BCL2. Proc. Natl Acad. Sci. USA. 90, 3516–3520. . LIN, E., ORLOFSKY, A., BERGER, M. & PRYSTOWSKY, M. (1993) Characterization of A1, a novel hemopoietic-specific early-response gene with sequence similarity to bcl-2. J. Immunol. 151, 1979–1988. . OLTVAI, Z., MILLIMAN, C. & KORSMEYER, S. (1993) Bcl-2 heterodimerizes in vivo with a conserved homolog, Bax, that accelerates programed cell death. Cell 74, 609–619. . CLEM, R.J., FECHHEIMER, M. & MILLER, L.K. (1991) Prevention of apoptosis by a baculovirus gene during infection of insect cells. Science 254, 1388–1390. . PEARSON, G.R., LUKA, J., PETTI, L., SAMPLE, J., BIRKENBACH, M., BRAUN, D. & KIEFF, E. (1987) Identification of an Epstein-Barr virus early gene encoding a second component of the restricted early antigen complex. Virology 160, 151–161. . ALBRECHT, J.C., NICHOLAS, J., CAMERON, K.R., NEWMAN, C., FLECK-ENSTEIN, B. & HONESS, R.W. (1992) Herpesvirus saimiri has a gene specifying a homologue of the cellular membrane glycoprotein CD59. Virology 190, 527–530. . NEILAN, J., LU, Z., AFONSO, C., KUTISH, G., SUSSMAN. M. & ROCK, D. (1993) An African Swine Fever virus gene with similarity to the proto-oncogene bcl-2 and the Epstein-Barr virus gene. BHRF1 J. Virol 67, 4391–4394. . KORSMEYER, S.J., SHUTTER, J.R., VEIS, D.J., MERRY, D.E. & OLTVAI, Z.N. (1993) Bcl-2/Bax: a rheostat that regulates an anti-oxidant pathway and cell death. Semin. Cancer Biol. 4, 327–332. . YIN, X.-M., OLTVAI. Z. & KORSEMEYER, S. (1994) BH1 and BH2 domains of Bcl-2 are required for inhibition of apoptosis and heterodimerization with bax. Nature 369, 321–323. . SMEYNE, R., VENDRELL, M., HAYWARD, M., BAKER, S., MIAO, G., SCHILLING, K., ROBERTSON, L., CURRAN, T. & MORGAN, J. (1993) Continuous c-fos expression precedes programmed cell-death in vivo. Nature 363, 166–169. . WU, F.Y., CHANG, N.T., CHEN, W.J. & JUAN, C.C. (1993) Vitamin K3-induced cell cycle arrest and apoptotic cell death are accompanied by altered expression of c-fos and c-myc in nasopharyngeal carcinoma cells. Oncogene 8, 2237–2244. . COLOTTA, F., POLENTARUTTI, N., SIRONI, M. & MANTOVANI, A. (1992) Expression and involvement of c-fos and c-jun protooncogenes in programmed cell death induced by growth factor deprivation in lymphoid cell lines. J. Biol. Chem. 267, 18278– 18283. . MARTI, A., JEHN, B., COSTELLO, E., KEON, N., KE, G., MARTIN. F. & JAGGI, R. (1994) Protein kinase A and AP-1 (c-Fos/JunD) are induced during apoptosis of mouse mammary epithelial cells. Oncogene 9, 1213–1223. . GOLDSTONE, S.D. & LAVIN, M.F. (1994) Prolonged expression of c-jun and associated activity of the transcription factor AP-1, during apoptosis in a human leukaemic cell line. Oncogene 9, 2305–2311. . EL-DEIRY, W., TOKINO, T, VELCULESCU, V., LEVY, D., PARSONS, R., TRENT, J, LIN, D., MERCER, W., KINZLER. K.& VOGELSTEIN, B. (1993) WAF1, a potential mediator of p53 tumor suppression. Cell 76, 817–825. . HARPER, J., ADAMI, G., WEI, N., KEYOMARSI. K. & ELLEDGE, S. (1993) The p21 Cdk-interacting protein Cip1 is a potent inhibitor of G1 cyclin-dependent kinases. Cell 76, 805–816. . YONISH-ROUACH, E., RESNITZKY, D., LOTEM, J., SACHS, L., KIMCHI, A. & OREN, M. (1991) Wildtype p53 induces apoptosis of myeloid leukaemic cells that is inhibited by interleukin-6. Nature 352, 345–347. . LANE, D.P. (1992) Cancer. p53, guardian of the genome. Nature 358, 15–16. . LANE, D. (1993) A death in the life of p53. Nature 362, 786. . LOTEM, J. & SACHS, L. (1993) Regulation by bcl-2, c-myc, and p53 of susceptibility to induction of apoptosis by heat shock and cancer chemotherapy compounds in differentiation-competent and -defective myeloid leukemic cells. Cell Growth Differ. 4, 41–47.
82
85
86
87
88 89 90
91 92 93 94
95 96
97 98 99
100 101 102 103 104 105
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. GOTTLIEB, E., HAFFNER, R., VON RUDEN, T., WAGNER, E.F. & OREN, M. (1994) Down-regulation of wild-type p53 activity interferes with apoptosis of IL-3-dependent hematopoietic cells following IL-3 withdrawal. EMBO J. 13, 1368–1374. . LEE, E.-H., CHANG, C.-Y., HU, N., WANG, Y.-C., LAI, C.-C., HERRUP, K., LEE. W.-H., & BRADLEY, A. (1992) Mice deficient for Rb are nonviable and show defects in neurogenesis and haematopoiesis. Nature 359, 288–294. . CLARKE, A., MAANDAG, E., VAN ROON, M., VAN DER LUGT, N., VAN DER VALK, M., HOOPER, M., BERNS. A. & TE RIELE, H. (1992) Requirement for a functional Rb-1 gene in murine development. Nature 359, 328–330. . JACKS, T., FAZELI, A., SCHMITT, E., BRONSON, R., GOODELL, M. & WEINBERG, R. (1992) Effects of an Rb mutation in the mouse. Nature 359, 295–300. . RAFF, M., BARRES, B., BURNE, J., COLES, H., ISHIZAKI, Y. & JACOBSON, M. (1993) Programmed cell death and the control of cell survival: lessons from the nervous system. Science 262, 695–700. . HOWES, K., RANSOM, L., PAPERMASTER, D., LASUDRY, J., ALBERT. D. & WINDLE, J. (1994) Apoptosis or retinoblastoma—alternative fates of photoreceptors expressing the HPV-16 E7 gene in the presence or absence of p53. Genes Dev. 8, 1300– 1310. . SYMONDS, H., KRALL, L., REMINGTON, L., SAENZ-ROBLES, M., LOWE, S., JACKS. T. & VAN DYKE, T. (1994) P53-dependent apoptosis suppresses tumour growth and progression in vivo. Cell 76, 703–711. . BRAITHWAITE, A.W. & JENKINS, J.R. (1989) Ability of p53 and the adenovirus E1b 58-kilodalton protein to form a complex is determined by p53. J. Virol. 63, 1792– 1799. . GANNON, J.V. & LANE, D.P. (1987) p53 and DNA polymerase alpha compete for binding to SV40 T antigen. Nature 329, 456–458. . SARNOW, P., HO, Y.S., WILLIAMS. J. & LEVINE, A.J. (1982) Adenovirus E1b-58kd tumor antigen and SV40 large tumor antigen are physically associated with the same 54 kd cellular protein in transformed cells. Cell 28, 387–394. . YEW, P.R., LIU, X. & BERK, A.J. (1994) Adenovirus E1B oncoprotein tethers a transcriptional repression domain to p53. Genes Dev. 8, 190–202. . CIECHANOVER, A., SHKEDY, D., OREN, M. & BERCOVICH, B. (1994) Degradation of the tumor suppressor protein p53 by the ubiquitin-mediated proteolytic system requires a novel species of ubiquitin-carrier protein, E2. J. Biol. Chem. 269, 9582– 9589. . CLARKE, A.R., PURDIE, C.A., HARRISON, D.J., MORRIS, R.G., BIRD, C.C., HOOPER, M.L., & WYLLIE, A.H. (1993) Thymocyte apoptosis induced by p53-dependent and independent pathways. Nature 362, 849–852. . MOMAND, J., ZAMBETTI, G.P., OLSON, D.C. GEORGE, D. & LEVINE, A.J. (1992) The mdm-2 oncogene product forms a complex with the p53 protein and inhibits p53-mediated transactivation. Cell 69, 1237–1245. . KESSIS, T.D., SLEBOS, R.J., NELSON, W.G., KASTAN, M.B., PLUNKETT, B. S., HAN, S.M., LORINCZ, A.T., HEDRICK, L. & CHO, K.R. (1993) Human papillomavirus 16 E6 expression disrupts the p53-mediated cellular response to DNA damage. Proc. Natl Acad. Sci. USA 90, 3988–3992. . CROOK, N.E., CLEM. R.J. & MILLER, L.K. (1993) An apoptosis-inhibiting baculovirus gene with a zinc finger-like motif. J. Virol. 67, 2168–2174. . KAMITA, S.G., MAJIMA, K. & MAEDA, S. (1993) Identification and characterization of the p35 gene of Bombyx mori nuclear polyhedrosis virus that prevents virus-induced apoptosis. J. Virol. 67, 455–463. . BIRNBAUM, M.J., CLEM, R.J. & MILLER, L.K. (1994) An apoptosis-inhibiting gene from a nuclear polyhedrosis virus encoding a polypeptide with Cys/His sequence motifs. J. Virol. 68, 2521–2528. . CLEM, R.J. & MILLER, L.K. (1994) Control of programmed cell death by the baculovirus genes p35 and iap. Mol. Cell Biol. 14, 5212–5222. . EVANS, C.A., OWEN, L.P., WHETTON, A.D., & DIVE, C. (1993) Activation of the Abelson tyrosine kinase activity is associated with suppression of apoptosis in hemopoietic cells. Cancer Res. 53, 1735–1738. . CLEM, R.J. & MILLER, L.K. (1993) Apoptosis reduces both the in vitro replication and the in vivo infectivity of a baculovirus. J. Virol. 67, 3730–3738.
GENES INVOLVED IN APOPTOSIS
106 107
108 109
110 111 112
113 114
115 116 117 118 119
120
121 122
123 124 125
83
. CHOU, J., KERN, E.R., WHITLEY, R.J. & ROIZMAN, B. (1990) Mapping of herpes simplex virus-1 neurovirulence to gamma 134.5, a gene nonessential for growth in culture. Science 250, 1262–1266. . CHOU, J. & ROIZMAN, B. (1992) The gamma 1(34.5) gene of herpes simplex virus 1 precludes neuroblastoma cells from triggering total shutoff of protein synthesis characteristic of programmed cell death in neuronal cells. Proc. Natl Acad. Sci. USA 89, 3266– 3270. . SPEHNER, D., GILLARD, S., DRILLIEN, R. & KIRN, A. (1988) A cowpox virus gene required for multiplication in Chinese hamster ovary cells. J. Virol. 62, 1297–1304. . INK, B.S., GILBERT, C.S. & EVAN, G.I. (1995) Prevention of vaccinia virus-induced apoptosis in nonpermissive Chinese hampster ovary (CHO) cells by the cowpox virus CHOhr and adenovirus E1Bp19 genes. J. Virol. 69 (2), 661–668. . BRANTON, P.E., BAYLEY, S.T. & GRAHAM, F.L. (1985) Transformation by human adenoviruses. Biochim. Biophys. Acta 780, 67–94. . WHITE, E., CIPRIANI, R., SABBATINI, P. & DENTON, A. (1991) Adenovirus E1B 19-kilodalton protein overcomes the cytotoxicity of E1A proteins. J. Virol. 65, 2968– 2978. . WHITE, E., SABBATINI, P., DEBBAS, M., WOLD, W., KUSHER, D. & GOODING, L. (1992) The 19kilodalton Adenovirus E1B transforming protein inhibits programmed cell death and prevents cytolysis by tumour necrosis factor . Mol Cell Biol. 12, 2570–2580. . HASHIMOTO, S., ISHII. A. & YONEHARA, S. (1991) The E1b oncogene of adenovirus confers cellular resistance to cytotoxicity of tumor necrosis factor and monoclonal anti-Fas antibody. Int. Immunol. 3, 343–351. . RAO, L., DEBBAS, M., SABBATINI, P., HOCKENBERY, D., KORSMEYER. S. & WHITE, E. (1992) The adenovirus E1A proteins induce apoptosis, which is inhibited by the E1B 19-kDa and Bcl-2 proteins. Proc. Natl Acad. Sci. USA 89, 7742–7746. . DEBBAS, M. & WHITE, E. (1993) Wild-type p53 mediates apoptosis by E1A, which is inhibited by E1B. Genes Dev. 7, 546–554. . LOWE, S. & RULEY, H.E. (1993) Stabilization of the p53 tumor suppressor is induced by adenovirus 5 E1A and accompanies apoptosis. Genes Dev. 7, 535–45. . ROGERS, R.P., STROMINGER, J.L. & SPECK, S.H. (1992) Epstein-Barr virus in B lymphocytes: viral gene expression and function in latency. Adv. Cancer Res. 58, 1–26. . GREGORY, C.D., ROWE, M. & RICKINSON, A.B. (1990) Different Epstein-Barr virus-B cell interactions in phenotypically distinct clones of a Burkitt’s lymphoma cell line. J. Gen. Virol. 71:1481–1495. . GREGORY, C.D., DIVE, C., HENDERSON, S., SMITH, C.A., WILLIAMS, G.T., GORDON. J. & RICKINSON, A.B. (1991) Activation of Epstein-Barr virus latent genes protects human B cells from death by apoptosis. Nature 349, 612–614. . HENDERSON, S., ROWE, M., GREGORY, C., CROOM, C.D., WANG, F., LONGNECKER, R., KIEFF. E. & RICKINSON, A. (1991) Induction of bcl-2 expression by Epstein-Barr virus latent membrane protein 1 protects infected B cells from programmed cell death. Cell 65, 1107–1115. . MARTIN, J.M., VEIS, D., KORSMEYER. S.J. & SUGDEN, B. (1993) Latent membrane protein of EpsteinBarr virus induces cellular phenotypes independently of expression of Bcl-2. J. Virol. 67, 5269–5278. . PFITZNER, A.J., STROMINGER, J.L., & SPECK, S.H. (1987) Characterization of a cDNA clone corresponding to a transcript from the Epstein-Barr virus BamHI M fragment: evidence for overlapping mRNAs. J. Virol. 61, 2943–2946. . BECKER, Y., TABOR, E. & ASHER, Y. (1991) Epstein-Barr virus BHRF1 gene but not the cellular protooncogene bcl-2 is expressed in ataxia-telangiectasia lymphoblastoid lines. Virus Genes 5, 33–45. . CLEARY, M.L., SMITH, S.D. & SKLAR, J. (1986) Cloning and structural analysis of cDNAs for bcl-2 and a hybrid bcl-2/immunoglobulin transcript resulting from the t(14;18) translocation. Cell 47, 19–28. . TARODI, B., SUBRAMANIAN, T. & CHINNADURAI, G. (1994) Epstein-Barr virus BHRF1 protein protects against cell death induced by DNA-damaging agents and heterologous viral infection. Virology 201, 404–407.
84
126
127
128
129
130 131 132 133 134 135 136 137 138 139 140
141 142
143 144 145 146
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. HENDERSON, S., HUEN, D., ROWE, M., DAWSON, C., JOHNSON. G. & RICKINSON, A. (1993) EpsteinBarr virus-coded BHRF1 protein, a viral homolog of Bcl-2, protects human B-cells from programmed cell-death. Proc. Natl Acad. Sci. USA 90, 8479–8483. . MARCHINI, A., TOMKINSON, B., COHEN, J.I. & KIEFF, E. (1991) BHRF1, the Epstein-Barr virus gene with homology to Bc 12, is dispensable for B-lymphocyte transformation and virus replication. J. Virol. 65, 5991–6000. . OEHM, A., BEHRMANN, I., FALK, W., PAWLITA, M., MAIER, G., KLAS, C., LI, W.M., RICHARDS, S., DHEIN, J., TRAUTH, B.C., PONSTINGL, H. & KRAMMER, P.H. (1992) Purification and molecular cloning of the APO-1 cell surface antigen, a member of the tumor necrosis factor/nerve growth factor receptor superfamily. Sequence identity with the Fas antigen. J. Biol Chem. 267, 10709–10715. . TRAUTH, B.C., KLAS, C., PETERS, A.M., MATZKU, S., MOLLER, P., FALK, W., DEBATIN, K.M. & KRAMMER, P.H. (1989) Monoclonal antibody-mediated tumor regression by induction of apoptosis. Science 245, 301–305. . DANIEL, P.T. & KRAMMER, P.H. (1994) Activation induces sensitivity toward APO-1 (CD95)-mediated apoptosis in human B cells. J. Immunol. 152, 5624–5632. . KLAS, C., DEBATIN, K.M., JONKER, R.R. & KRAMMER, P.H. (1993) Activation interferes with the APO-1 pathway in mature human T cells. Int. Immunol. 5, 625–630. . VIGNAUX, F. & GOLSTEIN, P. (1994) Fas-based lymphocyte-mediated cytotoxicity against syngeneic activated lymphocytes: a regulatory pathway? Eur. J. Immunol. 24, 923–927. . ROUVIER, E., LUCIANI, M.F. & GOLSTEIN, P. (1993) Fas involvement in Ca(2+)-independent T cellmediated cytotoxicity. J. Exp. Med. 177, 195–200. . LOWIN, B., HAHNE, M., MATTMAN. C. & TSCHOPP, J. (1994) Cytolytic T-cell cytotoxicity is mediated through perforin and Fas lytic pathways. Nature 370, 650–652. . BUTTYAN, R., ZAKERI, Z., LOCKSHIN. R. & WOLGEMUTH, D. (1988) Cascade induction of c-fos, c-myc, and heat shock 70 K transcripts during regression of the rat ventral prostate gland. Mol. Endocrinol. 2, 650–657. . BUTTYAN, R., OLSSON, C.A., PINTAR, J., CHANG, C., BANDYK, M., NG, P.Y. & SAWCZUK, I.S. (1989) Induction of the TRPM-2 gene in cells undergoing programmed death. Mol. Cell Biol. 9, 3473–3481. . LEGER, J.G., MONTPETIT, M.L. & TENNISWOOD, M.P. (1987) Characterization and cloning of androgenrepressed mRNAs from rat ventral prostate. Biochem. Biophys. Res. Commun. 147, 196–203. . MONTPETIT, M.L., LAWLESS, K.R. & TENNISWOOD, M. (1986) Androgen-repressed messages in the rat ventral prostate. Prostate, 8, 25–36. . COLLARD, M.W. & GRISWOLD, M.D. (1987) Biosynthesis and molecular cloning of sulfated glycoprotein 2 secreted by rat Sertoli cells. Biochemistry, 26, 3297–3303. . CHENG, C.Y., CHEN, C.L., FENG, Z.M. MARSHALL, A. & BARDIN, C.W. (1988) Rat clusterin isolated from primary Sertoli cell-enriched culture medium is sulfated glycoprotein-2 (SGP-2). Biochem. Biophys. Res. Commun. 155, 398–404. . COLOMBEL, M., OLSSON, C.A., NG, P.Y. & BUTTYAN, R. (1992) Hormoneregulated apoptosis results from reentry of differentiated prostate cells onto a defective cell cycle. Cancer Res. 52, 4313–4319. . HARRIGAN, M.T., BAUGHMAN, G., CAMPBELL, N.F. & BOURGEOIS, S. (1989) Isolation and characterization of glucocorticoid- and cyclic AMP-induced genes in T lymphocytes. Mol. Cell Biol. 9, 3438–3446. . OWENS, G., HAHN. W. & COHEN, J. (1991) Identification of mRNAs associated with programmed cells death in immature thymocytes. Mol. Cell Biol. 11, 4177–4188. . BAUGHMAN, G., LESLEY, J., TROTTER, J., HYMAN, R. & BOURGEOIS, S. (1992) Tcl-30, a new T cellspecific gene expressed in immature glucocorticoid-sensitive thymocytes. J. Immunol. 149, 1488–1496. . BAUGHMAN, G., HARRIGAN, M.T., CAMPBELL, N.F., NURRISH, S.J. & BOURGEOIS, S. (1991) Genes newly identified as regulated by glucocorticoids in murine thymocytes. Mol. Endocrinol. 5, 637–644. . HARRIGAN, M.T., CAMPBELL, N.F. & BOURGEOIS, S. (1991) Identification of a gene induced by glucocorticoids in murine T-cells: a potential G protein-coupled receptor. Mol. Endocrinol. 5, 1331–1338.
GENES INVOLVED IN APOPTOSIS
147 148
149
85
. ELLIS, R.E. & HORVITZ, H.R. (1991) Two C. elegans genes control the programmed deaths of specific cells in the pharynx. Development 112, 591–603. . WALKER, N.P.C., TALANIAN, R.V., BRADY, K.D., DANG, L.C., BUMP, N.J., FERENZ, C.R., FRANKLIN, S., GHAYUR, T., HACKET, M.C., HAMMILL, L.D., HERZOG, L., HUGUNIN, M. et al. (1994) Crystal structure of the cystein protease interleukin-1 - converting enzyme: a (p20/p10)2 homodimer. Cell 78, 343–352. . MCDONNELL, T.J. & KORSMEYER, S.J. (1991) Progression from lymphoid hyperplasia to high-grade malignant lymphoma in mice transgenic for the t(14; 18). Nature 349, 254–256.
6 The Role of p53 in Apoptosis SCOTT W.LOWE Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 11724, USA
6.1 Introduction Over the last two decades, molecular and genetic approaches have identified numerous molecules linked to carcinogenesis by virtue of their mutation in human tumors. These molecules have been classified as either oncogenes or tumor suppressor genes [1, 2]. Oncogenes are molecules that, when constitutively activated by mutation, promote neoplastic growth in a dominant manner. By contrast, tumor suppressor genes are genetically recessive—loss of function of these molecules enhances tumorigenesis. Since cancer results from net tissue expansion, much of cancer research has focused on how oncogenes and tumor suppressor genes regulate cell proliferation. Thus, considerable evidence indicates that oncogenes such as ras and cmyc normally function in signal transduction pathways that promote cell growth, whereas the products of several tumor suppressor genes (e.g. the retinoblastoma protein) play essential roles in negative growth control. Only recently has it become widely appreciated that imbalances in cell survival also contribute to neoplastic disease. This new direction in cancer research has emerged from increases in our understanding of the process of apoptosis, or programmed cell death. Apoptosis is a form of cell death that is characterized by distinct morphological and physiological features, but is most remarkable in that it requires a genetic program for its execution. In principle, mutations that attenuate the cell death program would allow inappropriate cell survival and contribute to net tissue expansion—the primary feature of the malignant phenotype. Molecular evidence is emerging to support this view. For example, activation of the bcl-2 oncogene, while having no effect on proliferation, promotes lymphogenesis by inhibiting apoptosis [3]. Moreover, the p53 tumor suppressor gene may inhibit tumor progression by promoting apoptosis. It is the latter topic that will be the focus of this chapter.
THE ROLE OF P53 IN APOPTOSIS
87
6.2 The p53 Tumor Suppressor 6.2.1 p53 Mutation in Human Malignancy p53 mutations are among the most frequent genetic lesions detected in human cancer and occur in a wide range of tumor types [4]. Germ-line mutations in p53 are associated with Li-Fraumeni syndrome, a familial cancer syndrome which predisposes individuals to a variety of tumors [5, 6]. In many cancers, however, p53 mutations are a late event in tumorigenesis (see for example [7]). Moreover, p53 mutations have been associated with clinically aggressive cancers and poor patient prognosis (reviewed in [8]). The majority of p53 mutations occurring in human tumors are single nucleotide changes that produce altered or truncated gene products [8]. Missense mutations in one allele of p53 often are accompanied by loss of the remaining normal allele. Nevertheless, even one mutant allele is often sufficient for promoting tumorigenesis. This mutational pattern may be explained by in vitro studies indicating that many mutant p53 proteins are capable of inhibiting wild-type p53 function in a dominant-negative manner (reviewed in [9, 10]). In general, the genetic data from human cancer support the notion that p53 functions as a tumor suppressor gene. Moreover, the involvement of p53 mutations in the etiology of diverse tumor types indicates that p53 plays a central role in carcinogenesis, normally affecting processes common to many tissues. Of great importance, therefore, is a better understanding of the biochemical and biological functions of p53 that limit tumor growth. 6.2.2 Biochemical and Biological Properties of p53 p53 is a sequence-specific DNA binding protein (see for example [11]) that can activate [12–14] or inhibit [15, 16] transcription. p53 may also directly regulate DNA synthesis [17] or DNA repair [18]. In recent years, p53 has been shown to interact physically with an ever-growing list of proteins [19], but this approach has yet to provide a definitive understanding of how p53 suppresses tumor growth. In this regard, the DNA binding activity of p53 has provided the best biochemical link with the genetic evidence from human tumors: all tumor-derived p53 mutants produce proteins with reduced affinity for DNA [10]. These data strongly suggest that DNA binding is essential for tumor suppression by p53. The co-crystal structure of p53 with its consensus DNA sequence supports this view, since many of the most frequent p53 mutations occur at residues with essential structural roles for the DNA-protein interaction [20]. Overexpression of either wild-type or dominant-negative mutant p53 alleles has been used to enhance or inhibit the biological activities of p53. These types of experiments have identified several biological processes in which p53 might participate, including negative growth control [21, 23], apoptosis [24, 25], senescence [26, 27] and differentiation [28]. However, forced overexpression of p53 cannot reveal the physiological circumstances in which p53 is activated to participate in these processes. Moreover, studies utilizing mutant p53 alleles to block endogenous p53 function are complicated by the observation that these mutants may possess gain-of-function activities [29]. Gene targeting of p53 has allowed investigators to assess the biological activities of endogenous p53 using a classical genetic approach. By comparison of mouse strains that differ only in the presence or absence of functional p53 genes, phenotypic differences can be directly or indirectly attributed to p53 function. Given the impact of p53 mutation on tumorigenesis, it is surprising that mice homozygous for p53
88
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
deletions develop normally. Nevertheless, mutant animals are highly predisposed to the development of spontaneous and carcinogen-induced tumors [30–33]. p53 therefore truly functions as a tumor suppressor— whose apparent function is to inhibit neoplastic growth. These mutant mouse strains also provide a source of cells which can be used to investigate endogenous p53 function in vitro. By comparing cells derived from normal and p53-deficient animals, p53 has been implicated in the cellular response to DNA damage and in the maintenance of genomic integrity [34, 35]. Normal cells possess cell cycle checkpoints that limit proliferation following situations of cellular stress. For example, treatment of normal fibroblasts with DNA-damaging agents produces G1 or G2 arrest, which is preceded by increases in p53 levels and protein stability [36]. p53 is required for the G1 arrest, since fibroblasts derived from p53-deficient mice fail to arrest in G1 following DNA damage [34]. These p53deficient cells are more prone to aneuploidy and DNA amplification events, suggesting that the loss of the G1 checkpoint promotes genomic instability [35]. 6.2.3 The `Guardian of the Genome' The involvement of p53 in the cellular response to DNA damage has suggested a mechanism whereby p53 suppresses neoplastic growth [37]. According to this view, p53 is an essential component of a DNA damage control system which, when operating normally, facilitates DNA repair and thereby reduces the likelihood that cells will sustain oncogenic mutations. The role of p53 in tumor suppression is therefore indirect, since cells defective for p53 function are more prone to additional mutations, some of which are oncogenic. It is these secondary mutations that actually provide the growth or survival advantage to cells acquiring p53 mutations. This model is attractive, since it accounts for the observation that p53 is not required for normal growth and development. If p53 is activated to suppress cell growth following DNA damage, then it functions under circumstances that might be uncommon during embryogenesis. Moreover, the effects of DNA damage on the incidence of experimental tumors are consistent with this mechanism. For example, mice that are deficient for p53 or harbor mutant p53 transgenes are extremely prone to radiation and carcinogeninduced tumors [31, 38, 39]. The view that p53 inhibits tumorigenesis by limiting the occurrence of oncogenic mutations predicts that p53 mutation should be a good initiator of tumor growth. In many cancers, however, p53 mutations occur late in tumor progression (see for examples [7, 40–45]). Furthermore, loss of p53 does not influence the initiation, but rather the progression of carcinogen-induced skin papillomas to malignant carcinomas in mice [46]. Indeed, this indirect mechanism of tumor suppression is not the only means in which p53 mutation contributes to cancer. 6.3 p53 can Promote Apoptosis Insight into additional mechanisms by which p53 suppresses oncogenic transformation and inhibits tumor growth has emerged from the observation that p53 promotes apoptosis. These observations suggest that loss of p53 function can contribute to tumorigenesis by allowing inappropriate cell survival. The potential involvement of p53 in apoptosis was first demonstrated using a myeloid leukemia cell line expressing only a temperature-sensitive p53 allele [24]. These cells remained viable at the restrictive temperature (i.e. p53 inactivated) but rapidly underwent apoptosis at the permissive temperature for wild-
THE ROLE OF P53 IN APOPTOSIS
89
type p53 expression. Apoptosis apparently was not due to toxic side-effects of p53 overexpression, because interleukin-6 prevented cell death without influencing p53 levels. Subsequent studies revealed that p53 overexpression induces apoptosis in many tumor lines. For example, p53 induces apoptosis in p53-deficient cells derived from a human colon carcinoma [25], a non-small cell lung carcinoma [47], a osteogenic sarcoma [48], a Burkitt’s lymphoma [49], a v-myc induced murine T cell lymphoma [50], murine erythroleukemias [51], and certain oncogenically-transformed primary rodent cells containing endogenous p53 [52, 53]. Although these studies are informative and may have implications for gene therapy in cancer, they tell us little about physiological circumstances in which p53 participates in apoptosis. In particular, the levels of wild-type p53 in these experiments are often well above levels ever achieved by endogenous p53. The initial enthusiasm over the potential role of p53 in apoptosis was diminished by the observation that p53deficient mice develop normally—p53 therefore has no essential role in the physiological cell deaths that occur during embryogenesis. Nevertheless, it is now apparent that endogenous p53 participates in apoptosis under physiological conditions, with important implications for cancer. 6.4 Dependent Apoptosis in Normal Tissues p53-deficient animals and cells derived from these animals have been instrumental in identifying the circumstances under which endogenous p53 participates in apoptosis. At present, p53 has been implicated in apoptosis in the thymus, bone marrow, and intestine. 6.4.1 Apoptosis in Immature Thymocytes Apoptosis of cells in the developing thymus is particularly well-characterized, and provides an excellent system for investigating the role of endogenous p53 in the process. Thymocytes have a remarkably low threshold for apoptosis, presumably because of the requirement to eliminate most of these cells during positive and negative selection. Apoptosis contributes to the deletion of autoreactive thymocytes [54], and is induced by several agents that may mimic physiological conditions present during thymocyte development. These include the combination of phorbol esters and calcium ionophores [55], glucocorticoids such as dexamethasone [56], and -radiation [57]. Treatment with low doses of -radiation induces apoptosis in normal thymocytes but not thymocytes derived from p53-deficient animals, indicating that endogenous p53 is required for the process (Figure 6.1). It is important to note that not all apoptotic programs require p53 function, since p53-deficiency has no effect on thymocyte apoptosis induced by several other stimuli [58–60]. Similar results have been achieved in vivo. In particular, irradiation of whole animals induces apoptosis and depletion of CD4+ CD8+ thymocytes in normal but not p53-deficient animals, whereas dexamethasone treatment induces apoptosis and thymus ablation regardless of p53 status [60]. Interestingly, thymocytes derived from mice heterozygous for p53 deletions display an intermediate resistance to these agents, indicating that even a partial reduction in p53 dosage imparts a substantial survival advantage to these cells. These results established the involvement of endogenous p53 in a cell death pathway, but also demonstrated that apoptosis can occur in the absence of p53 function. Thus, cell death in the thymus can be subdivided into at least two distinct pathways, one requiring p53 and one that is p53 independent.
90
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 6.1 p53-dependent apoptosis of thymocytes induced by -irradiation but not dexamethasone. Thymocytes derived from p53−/− (squares), p53+/− (triangles), and p53+/+ (circles) animals were treated with (A) 1 µM dexamethasone or (B) 5 Gy -radiation, and viability was assessed by dye exclusion at various times thereafter [60]. Cell death was accompanied by internucleosomal DNA degradation, a physiological feature of apoptosis (not shown). Reproduced from [60] with permission.
6.4.2 Apoptosis in Hematopoietic Lineages Myeloid progenitor cells derived from p53-deficient mice are also resistant to apoptosis induced by irradiation, and cells derived from heterozygous mice display an intermediate level of resistance [59]. Other treatments that induce apoptosis in progenitor cells also display some level of p53-dependence. For example, p53-deficient cells are more resistant to heat shock and more readily survive sub-optimal concentrations of growth factors when compared to their normal counterparts. Similar results have been obtained using blast cells from patients with acute myeloblastic leukemia (AML) [61]. Survival of lines derived from AML blasts is either growth factor dependent or independent. AML blasts that proliferate autonomously (i.e. growth factor independent) express mutant forms of p53, whereas lines that require exogenous growth factors contain p53 with a wild-type conformation. Upon growth factor depletion, non-autocrine lines rapidly undergo apoptosis. Apoptosis can be inhibited by the addition of p53 antisense oligonucleotides, implying that p53 is required for this process. 6.4.3 Apoptosis in Epithelial Stem Cells Epithelial stem cells of the large and small intestine are particularly sensitive to apoptosis following treatment with a variety of genotoxic agents [62]. The stem cells are located near the base of the intestinal crypts and migrate outward during differentiation. In mice expressing wild-type p53 genes, -irradiation rapidly induces apoptosis of these stem cells, although their differentiated counterparts remain largely
THE ROLE OF P53 IN APOPTOSIS
91
unaffected. When p53-deficient animals are irradiated with identical doses, virtually no apoptosis is observed [63, 64]. It seems likely that p53-dependent apoptosis may be a typical outcome of DNA damage in many stem cell types. 6.4.4 Physiological Role of p53-Dependent Apoptosis in Normal tissues Most of the studies described above have used exogenous sources of radiation to trigger p53-dependent apoptosis. While these studies implicate endogenous p53 in apoptosis, radiation is not a physiological stimulus. What, therefore, is the physiological role of p53 in apoptosis? -Radiation produces various forms of cellular damage, most significantly double stranded breaks in genomic DNA. It is widely believed, though not proven, that DNA damage is the stimulus that recruits p53 to participate in apoptosis. Although tissues rarely receive the level of damage required to measure apoptosis in whole cell populations, -irradiation may mimic physiological forms of DNA damage encountered by individual cells in vivo. For example, errors in DNA replication or mitosis, oxidative damage, or exposure to natural or synthetic carcinogens may all produce various forms of DNA damage. It is plausible that p53-dependent apoptosis occurs only under these limited circumstances, a suggestion consistent with the observation that p53 has no essential role during development. By contrast, most tissues undergo p53-dependent growth arrest rather than apoptosis following DNA damage (most cells have mechanisms that arrest cells in both G1 and G2 following -irradiation, but the G2 arrest does not require p53 function [34]). Since entry into S phase could fix deleterious mutations, the p53regulated G1 checkpoint may allow sufficient time for DNA repair. It is intriguing that tissues that are particularly prone to radiation-induced apoptosis have in common the potential for rapid expansion. Perhaps apoptosis functions to eliminate these potentially dangerous cells following genomic damage to avoid the increased possibility of mutations that would accompany DNA repair. In the case of thymocytes, p53-dependent apoptosis may have another physiological role. During thymocyte development, T cell receptor genes undergo somatic recombination to produce clones expressing unique T cell receptors. These clones undergo both positive and negative selection, with the survivors eventually becoming peripheral T cells (reviewed in [65]). Since the generation of functional T cell receptor genes requires DNA cleavage and rejoining, p53 may provide protection against defective rearrangement by promoting apoptosis when T cell receptor genes are cleaved but are not rejoined. Such circumstances would cause DNA damage similar to that produced by -radiation, allowing the defective cells to be efficiently eliminated by p53-dependent apoptosis. The availability of mouse strains defective in T cell receptor gene rearrangement will allow this hypothesis to be tested directly. 6.4.5 Mechanism of Apoptosis The molecular mechanism underlying p53-dependent apoptosis in normal tissues is poorly understood. Increases in p53 levels occur following -irradiation of both thymocytes and epithelial stem cells [64, 66]. Actinomycin D prevents apoptosis of thymocytes but has no effect on p53 induction (S.L., unpublished), suggesting that new gene expression is not required for increased p53 levels. This response is reminiscent of the p53 stabilization induced following -irradiation of fibroblasts [34], a cell type which undergoes p53dependent growth arrest. It is therefore possible that upstream events that occur in both p53-dependent apoptosis and growth arrest are similar in both cell types. p53-mediated transcriptional transactivation of
92
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
p21Cip1/WAF1 is thought to play an essential role in p53-dependent growth arrest (see for discussion, [67]), but the downstream effectors of apoptosis remain unknown. Intriguing candidates for regulation by p53 during apoptosis are the products of the bcl-2 gene family. Constitutive activation of bcl-2 contributes to the etiology of human follicular lymphomas and promotes B and T cell lymphoma in transgenic mice (reviewed in [3]). Over-expression of bcl-2 inhibits multiple forms of apoptosis, including radiation-induced apoptosis of mouse thymocytes expressing normal p53 [68, 69]. Thus, bcl-2 apparently acts downstream of p53 in the apoptotic program. bax is structurally related to bcl-2, but functionally promotes apoptosis [70]. bcl-2 and bax can form homodimeric complexes or heterodimerize with each other, suggesting that the ratio of bcl-2 to bax expression may ultimately determine cell survival following an apoptotic stimulus [70]. Both decreases in bcl-2 and increases in bax message and protein levels are associated with apoptosis induced by p53 overexpression in M1 leukemia cells, suggesting that p53 regulates apoptosis by influencing the Bcl-2/Bax ratio [71–73]. However, p53-deficient thymocytes do not overexpress bcl-2 and the bcl-2 message does not decrease during radiation-induced apoptosis (E.Schmitt and T.Jacks, personal communication). This raises the possibility that the p53-induced changes in bcl-2 and bax expression are a consequence of non-physiological levels of p53. 6.4.6 p53-Dependent Apoptosis and Tumor Suppression The data define another mechanism by which p53 can act as a tumor suppressor gene (Figure 6.2). In some tissues, mutational inactivation of p53 during tumorigenesis may allow the accumulation of oncogenic mutations due to the removal of an important G1 checkpoint (section 6.2.3). In thymocytes, hematopoietic progenitor cells, epithelial stem cells, and perhaps other cell types, the absence of p53 function provides an immediate selective advantage, allowing inappropriate cell survival. The failure to eliminate cells that have acquired genetic damage could lead to an increased incidence of mutation and ultimately to the selection of cells that have undergone neoplastic transformation. It is noteworthy that among the various tumor types that develop in p53 homozygous mutant mice, thymic lymphoma is by far the most common [30, 32, 33]. However, tumors arising from other tissues which readily undergo p53-dependent apoptosis are rarely observed in p53-deficient animals (e.g. myeloid leukemia and colon carcinoma). Perhaps these mice succumb to other cancers prior to the onset of these tumors. p53 also is implicated in apoptosis following growth factor withdrawal of certain hematopoietic cells in culture (section 6.4.2). Although it remains possible that growth factor depletion produces apoptosis by indirectly generating DNA damage, these observations raise the possibility that other stimuli can promote p53-dependent apoptosis. Nevertheless, cells acquiring p53 mutations may have a significant survival advantage compared to those expressing only wild-type p53, particularly under growth limiting conditions. Such an advantage would provide additional pressure for p53 inactivation during tumor progression.
THE ROLE OF P53 IN APOPTOSIS
93
Figure 6.2 The role of p53 in the cellular response to DNA damage. In fibroblasts (I and II), loss of a p53-dependent cell cycle checkpoint may contribute to tumor growth by increasing the likelihood that cells fix oncogenic mutations following genomic damage. p53 mutation provides an indirect advantage to these cells, since subsequent mutations actually drive tumor growth. In normal tissues prone to p53-dependent apoptosis (III and IV), cells lacking functional p53 are also prone to acquiring oncogenic mutations, since they more readily survive DNA damage. However, p53 loss provides an immediate survival advantage to these cells, thereby increasing the target size for additional mutations. The shaded nuclei represent cells containing damaged DNA.
6.5 p53-Dependent Apoptosis and Aberrant Proliferation 6.5.1 Apoptosis is a Common Feature of Malignant Tumors Tumor biologists have long observed a discrepancy between observed increases in tumor volume and the expected increase in tumor volume based on the percentage of proliferating cells. In many instances, the observed tumor doubling time is less than 5% of the potential growth rate, implying that there is a significant amount of cell death occurring in most tumors (reviewed in [74]). Consequently, factors that decrease the ‘cell loss factor’ in tumors can dramatically enhance tumor progression. Cell loss in tumors occurs primarily by necrosis or apoptosis. Tumor necrosis results from hypoxia and nutrient deprivation, so necrotic cells typically are observed in zones at a distance from blood vessels [74]. Although necrosis contributes significantly to tumor cell loss, it cannot explain the magnitude of cell death observed in many tumors [75]. Until recently, the contribution of apoptosis to cell death in tumors has been largely overlooked, perhaps because dying cells shrink and are rapidly phagocytosed by macrophages or neighboring cells. However, apoptosis accounts for a large proportion of cell loss in tumors, particularly during periods of tumor regression [76, 77].
94
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
The susceptibility of normal and neoplastic cells to apoptosis has several implications with regard to cancer (reviewed in [77, 78]). First, suppression of apoptosis in normal cells could cause hyperplasia, creating an expanded cell population from which cells acquiring oncogenic mutations could arise. Second, since apoptosis contributes significantly to the cell death in tumors, mutations that suppress apoptosis could promote tumor progression. Finally, the cytotoxicity of many anticancer agents may result from their ability to activate apoptosis. Consequently, mutations in apoptotic programs may reduce the effectiveness of cancer therapy. 6.5.2 Oncogenes can Promote both Proliferation and Apoptosis Recent advances have suggested a molecular basis for the enhanced rate of apoptosis occurring in malignant tumors. In particular, it has become apparent that many oncogenic alterations that increase proliferation also promote apoptosis. The balance between these processes is determined by additional genetic or environmental factors. In principle, circumstances which attenuate apoptosis while leaving proliferation unaffected could have a dramatic effect on tumor growth. Emerging evidence indicates that p53 modulates apoptosis in oncogene-expressing cells, a process that has been most extensively characterized for the adenovirus early region 1A (E1A) oncogene. The Role of Adenovirus E1A and E1B in Oncogenic Transformation Transforming interactions between E1A and other viral or cellular oncogenes provide a well-characterized model of multistep carcinogenesis [79]. The E1A oncogene, while unable to transform primary cells alone, collaborates with the adenovirus early region 1B (E1B) gene or activated ras oncogenes to transform primary cells to a tumorigenic state [80]. E1A promotes proliferation and S phase entry, probably by associating with cellular proteins involved in negative growth control [81, 82]. The E1B-encoded proteins, p19E1B and p55E1B, have no obvious effect on cell proliferation in the absence of E1A [83] but allow sustained proliferation of E1A-expressing cells [84]. The differential activities of E1A and E1B on apoptosis may explain their collaborative interactions in adenovirus transformation. Although E1 A initiates cell proliferation, it also increases susceptibility to apoptosis, which is particularly pronounced under conditions of serum deprivation or high cell density [84]. Consequently, proliferation cannot be sustained. Proliferation and apoptosis are tightly linked, since mutational analysis of E1A has been unable to separate regions of E1A which promote S phase entry from those that induce apoptosis [85]. E1B counters apoptosis in E1A-expressing cells, thereby allowing proliferation to continue without directly influencing cell growth [84]. While these studies do not rule out the possibility that E1B has additional activities, they suggest that its primary function is to inhibit apoptosis, thereby promoting oncogenic transformation. p53 is Required for E1A-induced Apoptosis Several observations suggested that p53 participates in E1A-associated apoptosis. First, the adenovirus p55E1B protein physically interacts with p53 and blocks p53-mediated transactivation [86] (although p19E1B is more effective at blocking E1A-induced apoptosis [84]). Second, cells expressing E1A contain metabolically stabilized p53 protein, leading to a 5–10-fold increase in p53 levels [87]. Increases in p53
THE ROLE OF P53 IN APOPTOSIS
95
levels and protein stability occur shortly after the introduction of E1A and are independent of E1Bexpression. Third, p53 overexpression induces apoptosis in cells expressing E1A [52]. The role of endogenous p53 in this process has been examined directly by introducing E1A into normal and p53-deficient embryonic fibroblasts [88]. These studies demonstrate that p53 is required for efficient apoptosis induced by E1A, since p53-deficient cells are resistant to E1A-induced apoptosis (Figure 6.3). Moreover, inac of p53 appears to be the primary function of E1B, since p53-deficiency substitutes for E1B in promoting the growth, survival, and transformation of E1A-expressing cells [88]. These results suggest that p53-dependent apoptosis is a cellular response to forced proliferation, which limits the oncogenic potential of E1A-expressing cells. Many Oncogenic Alterations Modulate p53-dependent Apoptosis Recent studies demonstrate that other viral and cellular genes modulate p53dependent apoptosis. Both the human papilloma virus E7 oncoprotein and a N-terminal mutant of simian virus 40 (SV40) large T antigen (which lacks the p53 binding domain) can promote apoptosis in a p53-dependent manner [89–91]. These viral oncoproteins are related to E1A in that they physically associate with a similar set of cellular proteins. For both large T antigen and E7, loss of p53-dependent apoptosis is associated with enhanced tumorigenicity (section 6.5.4). Cellular alterations that promote proliferation can also promote p53-dependent apoptosis. For example, one cellular target inactivated by the E1A, E7, and large T oncoproteins is the retinoblastoma gene product (pRB) [81, 92, 93]. Inactivation of Rb by gene deletion also can promote both aberrant proliferation and apoptosis (for example, see [94]). In at least some instances apoptosis associated with Rb deletion requires endogenous p53 function ([95]). One consequence of Rb inactivation (by deletion or by viral oncoproteins) is constitutive activation of the E2F transcription factors (reviewed in [96]), which can promote S phase entry [97]. Therefore, it is perhaps not surprising that constitutive activation of E2F synergizes with wildtype p53 to promote apoptosis in fibroblasts [98]. Forced expression of ras oncogenes promotes apoptosis in mouse embryo fibroblasts cultured in low serum or following DNA damage, but leads to transformation without apoptosis in p53-deficient fibroblasts [99]. c-myc overexpression also promotes apoptosis under growth-limiting conditions or following treatment with DNA-damaging agents [100, 101]. In mouse embryonic fibroblasts, apoptosis is attenuated in the absence of functional p53 (S.L., unpublished). Inactivation of p53-dependent apoptosis can occur by several mechanisms. First, homozygous deletion of p53 or expression of dominant negative p53 mutants can inactivate wild-type p53 function and apoptosis. Second, p53 can be inactivated by physical interactions with viral or cellular proteins, although cellular proteins that physically associate with p53 (e.g. Mdm-2 [102]) have not been tested for their ability to inhibit p53-dependent apoptosis. Finally, factors that act downstream of p53 can inhibit apoptosis, including p19E1B and bcl-2 [52, 103]. In several instances, oncogenes that promote apoptosis collaborate with those that repress apoptosis in oncogenic transformation (e.g. adenovirus E1A and E1B [104], papilloma virus E7 and E6 [105], c-myc and bcl-2 [101, 106]). It is important to note, however, that oncogenic transformation does not require escape from apoptosis, since activated ras oncogenes cooperate with E1A to transform cells to a highly malignant state but do not block p53-dependent apoptosis [88].
96
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 6.3 E1A-expressing in 0.1% FBS. A plasmid vector co-expressing E1A and hygromycin phosphotransferase was introduced into p53+/+(+/+), p53+/−(+/−), and p53−/−(−/−), MEFs and colonies were isolated by selection by selection in hygromycin B. After 3 weeks, E1A-expressing colonies were marked and photographed at the indicated times following transfer to medium containing 0.1% FBS. Both p53+/+ and p53−/− untransfected MEFs remained viable (~90% viability) for at least 6 days after transfer to 0.1% FBS (data not shown). Reproduced from [88] with permission.
6.5.3 p53 can Directly Suppress Oncogenic Transformation These observations have suggested a fundamentally different means by which p53 suppresses oncogenic transformation and limits tumor growth [88]. In this view, certain oncogenic alterations induce a cellular response which compensates, at least in part, for increased proliferation by enhancing apoptosis. Tumor growth can continue, but is limited by environmental conditions such as inadequate concentrations of growth or survival factors (Figure 6.4). Only then does p53 mutation provide a selective advantage— thereby releasing cells from susceptibility to apoptosis and allowing proliferation to continue unabated. In
THE ROLE OF P53 IN APOPTOSIS
97
Figure 6.4 Model for effects of p53 mutation during tumor progression. Early oncogenic alterations promote proliferation but also apoptosis. Tumor growth can occur, but is limited by environmental conditions that activate apoptosis. p53 mutation dramatically reduces the propensity of cells to engage the apoptotic program, allowing proliferation to continue unchecked.
this manner, p53 directly limits tumor progression by destroying aberrantly growing cells. Indeed, this view provides a plausible explanation for situations where p53 mutations occur late in tumor progression, after other oncogenic alterations have already occurred (for examples, see [7, 43, 107–109]). 6.5.4 p53-dependent Apoptosis Suppresses Tumor Progression in vivo The study of tumor progression in transgenic and ‘knock-out’ mice has provided strong evidence to support the notion that p53-dependent apoptosis occurs in response to aberrant proliferation and limits the progression of developing tumors. Several recent examples are discussed below. Proliferation and Apoptosis in the Mouse Embryonic Lens Mice homozygous for deletion of the Rb tumor suppressor contain several tissues that display aberrant proliferation and increased apoptosis, including the embryonic lens (Figure 6.5). The increased proliferation may be a direct consequence of Rb inactivation, which prevents cell cycle exit and normal differentiation [95]. In double-mutant mice in which both Rb and p53 are inactivated, cell death is dramatically reduced while proliferation remains unaffected. These results therefore link p53-dependent apoptosis with aberrant proliferation in vivo. While one might predict that this would lead to an increased incidence of lens tumors, mice homozygous for Rb deletion die during embryogenesis from other defects [94]. However, similar results have been obtained using transgenic mice in which pRB or p53 are inactivated by lens-specific expression of the papilloma virus E7 or E6 oncoproteins,
98
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 6.5 Association between aberrant proliferation, apoptosis, and p53 in the embryonic lens. Lens sections derived from Rb+/+ p53+/+ (A and D), Rb−/− p53+/+ (B and E) and Rb−/− p53−/− (C and F) embroys at 13.5 days of gestation. (A, B and C) Lens cells containing degraded DNA, a physiological feature of apoptosis, were identified using the TUNEL assay [133]. Representative apoptotic cells are marked with arrows. (E and F) Animals were exposed to bromodeoxyuridine (BrdU) in utero prior to preparation of embryonic lens sections. Proliferating cells, which incorporate BrdU, were identified using standard histological methods. Representative positive cells are marked with arrows, e=anterior epithelial layer; f=lens fiber cells; r=retina. Contributed by T.Jacks, Massachusetts Institute of Technology.
respectively [90]. In this model, inhibition of p53-dependent apoptosis correlates with the appearance of lens cell tumors. These observations may account for the synergistic action of Rb and p53 deletions in the etiology of certain murine tumors [110] and suggest a biological basis for the frequent inactivation of both Rb and p53 in human cancer. Progression of T-antigen-induced Tumors of the Choroid Plexus The role of p53 in experimental tumors of the choroid plexus provides a compelling example of how p53 mutation contributes to tumor progression by inactivation of apoptosis. In this tissue, a weakly oncogenic form of T antigen (which inactivates the Rb family of proteins but not p53) disrupts proliferation, but also increases apoptosis [91]. Thus, in the presence of functional p53, growth deregulation produces only hyperplasia. Inactivation of p53—by viral oncoprotein binding, by deletion, or by spontaneous mutation— dramatically increases the occurrence of aggressive tumors. Importantly, the increased malignancy of tumors with inactivated p53 is not a function of more efficient tumor initiation or increased proliferation, but instead correlates with a decreased incidence of apoptosis. Thus, inhibition of apoptosis enhances the growth of tumors initiated by other oncogenic alterations, allowing progression to a more malignant state. p53 Mutations in Anaplastic Wilm’s Tumor In human cancer, p53 mutations are frequently associated with progression of tumors to more aggressive forms, but the role of apoptosis in this process has been examined only in Wilm’s tumor. p53 mutations are a rare event in Wilm’s tumor, and are observed primarily in an aggressive subtype characterized by its anaplastic morphology [111]. Nevertheless, p53 mutations are apparently a critical event in progression to anaplasia. This view stems from the observation that some Wilm’s tumors contain focal regions of anaplasia surrounded by non-anaplastic tumor, and that p53 mutations occur specifically within these early focal
THE ROLE OF P53 IN APOPTOSIS
99
growths. The anaplastic foci display a striking reduction in spontaneous apoptosis compared to the relatively high incidence in the surrounding tumor. These results tightly link p53 mutation, reduced apoptosis, and tumor progression in a human cancer and suggest that, in some tumors, inactivation of p53 provides a selective advantage to abnormally proliferating cells by reducing their propensity for apoptosis. 6.6 Anticancer Therapies and p53-dependent Apoptosis The tumor-specific action of most anticancer agents has been attributed to their cytotoxic effects on actively proliferating cells. However, a body of evidence suggests that anticancer agents actually kill tumor cells by triggering apoptosis (reviewed in [77]). This emerging view has profound implications for our understanding of the therapeutic response in human tumors. First, the role of apoptosis in the cytotoxicity of anticancer agents suggests that events subsequent to the interaction of agents with their primary intracellular target play a substantial role in determining tumor response. Second, factors that increase tumor cell propensity for apoptosis may influence the therapeutic index whereby anticancer agents selectively kill tumor cells. Finally, since the execution of apoptosis requires a genetic program, mutations in apoptotic pathways could lead to cross-resistance to a wide range of anticancer therapies. One of the primary cellular lesions produced by many anticancer agents is DNA damage. Thus, the potential involvement of p53 in the therapeutic response was originally suggested by the observation that p53 acts to limit proliferation following genomic damage. Since failure to pause and repair damaged DNA should more readily produce debilitating cellular mutations, it was predicted that cells harboring p53 mutations would display decreased survival following treatment with anticancer agents (i.e. cancer therapy should be more effective in cells with p53 mutations). Although some reports have supported this view [112], this is typically not the case ([113, 114]; see section 6.6.2). Some cell types respond to DNA damage by inducing apoptosis rather than growth arrest (section 6.4). In these tissues, inactivation of p53 increases cell survival following anticancer therapy. However, in other tissues, DNA damage produces growth arrest without apoptosis. It is intriguing that tumors derived from the latter cell types respond to the same damage by apoptosis, implying that the threshold for apoptosis changes during tumorigenesis. Recent studies using model systems have suggested a molecular explanation for this phenomenon. In particular, oncogenes that promote apoptosis in response to physiological stimuli also enhance the cytotoxicity of many anticancer agents. Moreover, p53 function is required for this enhanced sensitivity to apoptosis, such that inactivation of p53 produces cross-resistance to numerous agents. 6.6.1 Model Systems Cells transformed by the adenovirus E1A and activated ras oncogenes are highly susceptible to apoptosis when cultured in the absence of growth factors (section 6.5.2). As shown in Figure 6.6, treatment with radiation and several compounds used in chemotherapy produce a similar decrease in viability in cells expressing wild-type p53. The dying cells have both morphological and physiological features of apoptosis, including membrane blebbing, DNA fragmentation, chromatin condensation and nuclear fragmentation [115]. Importantly, both transformed cells lacking p53 and the untransformed fibroblasts are largely resistant to these agents. These oncogenes therefore increase the susceptibility of fibroblasts to apoptosis, providing the ‘therapeutic index’ in which anticancer agents specifically target tumor cells. The ability of anticancer agents to specifically target oncogenically transformed cells is unrelated to their proliferation rate, since
100
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 6.6 Viability of the untransfected MEFs and p53+/+ and p53−/− clones transformed by E1A and ras was estimated 36 hours after treatment with the indicated dose of -radiation (A), or 24 hours after incubating with the indicated concentration of adriamycin (B), 5-fluorouracil (C), or etoposide (D). Dying cells contained condensed chromatin, fragmented nuclei, and degraded DNA, apparently resulting from internucleosomal cleavage (not shown). Closed circles=untransfected p53+/+ MEFs; open circles=untransfected p53−/− MEFs; closed squares=p53+/+ cells oncogenically transformed by E1A and T24 H-ras; open squares=p53−/− cells oncogenically transformed by E1A and T24 H-ras. Reproduced from [115] with permission.
p53-deficient fibroblasts transformed by E1A and ras grow rapidly but are resistant to apoptosis. In these cells, loss of p53 produces a ‘multidrug resistant’ phenotype. Similar results have been obtained using normal and p53-deficient embryonic fibroblasts expressing either E1A or ras alone [99, 115]. These observations have been extended to include the therapeutic response of bona fide tumors. Cells transformed by E1A and ras form tumors independently of p53 expression, allowing the generation of tumors differing primarily in their p53 status [116]. The response of these tumors to anticancer therapy displays a striking dependence on p53: tumors derived from p53-expressing cells regress following treatment with radiation or adriamycin while p53-deficient tumors are resistant to the same regimens (Figure 6.7). The difference between tumor response appears to be the ability of these agents to trigger apoptosis. Thus,
THE ROLE OF P53 IN APOPTOSIS
101
Figure 6.7 Effect of p53 on tumor response to -irradiation. Tumor volumes were estimated at various times after injection of p53+/+ (closed circles) or p53−/− (open circles) fibroblasts transformed by E1A and ras into nude mice. Upon reaching an appropriate volume (indicated by the arrows), the mice were irradiated with 7 Gy and tumors monitored for growth or regression. Reproduced from [134] with permission.
treatment induces massive apoptosis in tumors derived from p53-expressing cells but little apoptosis in p53deficient tumors (Figure 6.8). Interestingly, de novo p53 mutations are associated with treatment resistance and relapse of tumors derived from cells originally expressing only wild-type p53 [116]. These studies establish that defects in apoptosis can produce a mode of ‘multidrug’ resistance to antitumor therapy. Moreover, they suggest that oncogenic changes that accompany malignant transformation lowers the threshold at which anticancer agents trigger apoptosis. Other mutations may have the opposite effect, promoting both tumor progression and leading to treatment resistance. Since p53 mutations are extremely prevalent in human cancer, these studies suggest that p53 status may be an important determinant of tumor response to therapy. 6.6.2 Does p53 Mutation Enhance Drug Resistance in Human Cancer? At present, clinical studies have not been designed to properly test whether p53 status influences the therapeutic response of human tumors. Nevertheless, several observations are consistent with this view. For example, the studies from model systems predict that cancer therapy would be less effective in patients harboring tumors with p53 mutations. Indeed, the presence of p53 mutation has been associated with poor patient prognosis in a variety of cancers (for examples, see [109, 117–120]). Moreover, several highly curable tumor types rarely harbor p53 mutations. These include certain leukemias and lymphomas [121], testicular cancer [122], and Wilm’s tumor [109]. In acute lymphoblastic leukemia [123] and multiple myeloma [108], p53 mutations have been identified in relapse phase tumors that were not detectable prior to chemotherapy. These results suggest that p53 mutations confer a survival advantage to cells undergoing anticancer therapy. Although the associations described above are striking, they neither implicate p53 mutation in drug resistance nor demonstrate that wild-type p53 promotes apoptosis in response to anticancer agents.
102
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 6.8 Massive apoptosis induced by -radiation in p53+/+ but not p53−/− tumors. Tumors derived from either p53+/+ (A) or p53−/− (B) transformed fibroblasts co-expressing E1A and ras were recovered 48 hours after irradiation with 7 Gy and stained with hematoxylin and eosin.
However, studies using human tumor lines have linked p53 mutation with radioresistance [124, 125], and reintroduction of wild-type p53 into p53-deficient lung carcinoma cells restores apoptosis following cisplatin treatment [126]. Moreover, p53 mutations dramatically reduce the probability that patients with B-cell chronic lymphocytic leukemia will undergo tumor regression and enter remission following chemotherapy, providing a direct clinical association between p53 mutation and drug resistance [117]. Finally, in Wilm’s tumor, p53 mutation is linked to reduced apoptosis in a subtype that typically responds poorly to chemotherapy (N.Bardeesy and J.Pelletier, personal communication). Considerably more investigation will be required to determine the circumstances and the extent to which p53-dependent apoptosis contributes to the therapeutic response of human tumors. The situation is likely to be complicated by a variety of factors, including the type of p53 mutation, expression of modifying genes, tissue of tumor origin, and type of agent(s) used in therapy. Nevertheless, additional investigation should increase our understanding of both apoptosis and the therapeutic response, and may ultimately produce improvements in the treatment of human malignancy. 6.7 Mechanism of p53-dependent Apoptosis in Oncogene-expressing Cells The mechanism in which p53 participates in apoptosis in oncogene-expressing cells (or tumors) is unknown. However, re-introduction of p53 into p53-deficient tumor cells restores apoptosis directly or in synergy with anticancer agents (for examples, see [24, 25, 47]). This indicates that resistance to apoptosis is a direct consequence of p53 inactivation, not a result of secondary mutations occurring from the genomic instability that also accompanies p53 loss.
THE ROLE OF P53 IN APOPTOSIS
103
It has been suggested that apoptosis in oncogenically-transformed cells involves three distinct phases: (i) priming, (ii) triggering, and (iii) execution [127]. As described below, p53 apparently functions primarily in the triggering step. 6.7.1 Oncogenes Alter the Threshold of Apoptosis Immature thymocytes are capable of immediately initiating apoptosis given an appropriate stimulus (i.e. they are primed [127]), while normal fibroblasts are not. Oncogenes such as c-myc, E1A, and ras (among others) dramatically increase the susceptibility of fibroblasts to apoptosis. The expression of these oncogenes is not sufficient for apoptosis; rather, these cells possess (or have access to) the machinery required for the execution of apoptosis. Cells can continue to proliferate under favorable conditions, but initiate apoptosis following DNA damage or other conditions that only arrest growth in normal cells. Priming of cells for apoptosis appears to be linked to forced proliferation, c-myc overexpression, activated ras oncogenes, and E1A each deregulate proliferation, and the regions of E1A and c-myc that promote apoptosis are indistinguishable from those that promote proliferation [85, 100]. One mechanism of priming may involve disruption of Rb function leading to constitutive activation of the E2F transcription factors, which normally regulate S phase entry (section entitled ‘Many Oncogenic Alterations Modulate p53-dependent Apoptosis’). While this model may account for the enhanced apoptosis occurring in cells expressing viral oncogenes or lacking Rb it cannot readily explain how ras oncogenes, which have no direct effect on Rb function, also prime cells for apoptosis [99]. This raises the possibility that priming is a general cellular response to aberrant proliferation. It is not clear whether p53 participates in priming cells for apoptosis. p53 levels are elevated in cells expressing E1A even under circumstances where cells proliferate without significant apoptosis [87]. In these cells, p53 levels do not increase further following some apoptotic triggers, suggesting the high p53 levels are not sufficient for apoptosis (see discussion in [88]). Nevertheless, high p53 levels may be necessary for apoptosis, since normal cells express low levels of p53 without adversely effecting growth or viability. Although these observations suggest that p53 stabilization participates in the priming process, it remains to be determined whether other oncogenes that promote p53-dependent apoptosis also stabilize p53. 6.7.2 Physiological Stimuli and Genotoxic Agents Trigger Apoptosis Apoptosis is triggered by stimuli that cause susceptible cells to execute the apoptotic program. In the absence of priming, the trigger may produce only growth arrest. For instance, low serum concentrations, high cell density, and ionizing radiation induce growth arrest in normal cells but apoptosis in cells expressing E1A, ras, or both E1A and ras [99, 115]. Oncogenes that prime cells for apoptosis can also interfere with cell cycle checkpoints that are necessary for antiproliferative stimuli to arrest cell growth [88, 115 , 128], suggesting that apoptosis is triggered when further proliferation is perceived as aberrant [88]. This view is controversial, however, since several lines of evidence suggest that apoptosis is part of the normal spectrum of c-myc activities [129, 130]. In the latter view, agents that trigger apoptosis alter c-myc activities from proliferation to apoptosis. In oncogene-expressing cells susceptible to p53-dependent apoptosis, p53 appears primarily concerned with the triggering step. This is most apparent in p53-deficient tumor lines, where re-introduction of high
104
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
levels of wild-type p53 directly induces, apoptosis. These lines were presumably primed for apoptosis, since most normal cells simply arrest following p53 overexpression. A controversial area of investigation involves the nature of the molecular signal(s) required to initiate p53-dependent apoptosis. Clearly, DNA damage can elicit a p53 response in normal and oncogenically transformed cells, but is this the only event capable of initiating the p53-dependent apoptotic program? Several instances have been discussed in which p53-dependent apoptosis is associated with circumstances not known to damage DNA. These include apoptosis induced by serum deprivation in myeloid progenitor cells, in fibroblasts expressing certain oncogenes, and in hyperplastic tumors of the choriod plexus. It remains possible that forced S phase entry produces aberrant DNA synthesis and damage, thereby engaging the p53 pathway. Alternatively, other stimuli may also feed into the p53 pathway. Although the DNAdamage induced pathway has been partially well-characterized at the molecular level (reviewed in [67]), the molecular details of other routes to p53 activation remain unknown. 6.7.3 Execution of Apoptosis does not Require p53 Execution of apoptosis involves the molecular events causing cell shrinkage, chromatin condensation, DNA fragmentation and ultimately cell death. Most of the molecules involved in these processes are unknown. It does not appear that p53 functions in the execution of apoptosis in either normal or transformed cells, primarily since apoptosis occurs in many circumstances without the presence of functional p53. For example, glucocorticoids readily induce apoptosis in p53-deficient thymocytes [58–60] and serum depletion eventually induces apoptosis in p53-deficient cells transformed by E1A and ras (although much more slowly than if p53 is present). Chemotherapeutic agents also induce apoptosis in p53-deficient cells, although apoptosis typically requires considerably higher doses than in p53-expressing cells [115]. Thus, the machinery required for execution of apoptosis seems unaffected by p53 loss, but is less likely to be engaged in the absence of biologically active p53. 6.7.4 Biochemical Activities of p53 Involved in Apoptosis It is not known whether p53-dependent apoptosis in preneoplastic and malignant cells involves similar mechanisms as those occurring in normal tissues. Of fundamental importance may be the observation that oncogenic changes can alter the function of p53 from facilitating growth arrest to promoting apoptosis. Thus, untransformed fibroblasts respond to low levels of -radiation by undergoing p53-dependent growth arrest while their oncogenically-transformed derivatives undergo p53-dependent apoptosis following the same dose (section 6.6.1). The ability of p53 to arrest cell growth appears to be via transcriptional activation of genes such as GADD45 and p21Cip1/WAF1 [34, 131, 132], but transcriptional activation by p53 is not necessary for apoptosis induced by p53 overexpression [53]. While this suggests that apoptosis and growth arrest involve biochemically distinct activities of p53, apoptosis induced by p53 overexpression may be fundamentally different from that involving endogenous p53. Considerably more effort will be required to elucidate the biochemical mechanism underlying p53-dependent apoptosis, but the molecular details of this process stand to provide important insight into p53 function and tumorigenesis.
THE ROLE OF P53 IN APOPTOSIS
105
6.8 Summary: Implications of p53-dependent Apoptosis in Human Cancer The role of p53 in apoptosis has several implications for the development of neoplastic disease. In tissues that are normally prone to p53-dependent apoptosis, inactivation of p53 may promote tumor growth by allowing inappropriate cell survival, increasing the probability of additional mutation. In other tissues, p53 mutation may enhance the progression of neoplastic or preneoplastic lesions by eliminating an apoptotic program that keeps tumor growth in check. Since p53 also participates in a cell cycle checkpoint that arrests cell growth in response to DNA damage [34], several mechanisms may account for the high frequency of p53 mutation in human tumors. A more detailed understanding of these processes will be essential for unraveling the complexities of human cancer. The potential involvement of p53 in the therapeutic response to anticancer agents has important ramifications for cancer therapy. The experimental systems described in this chapter provide provocative models for how anticancer agents elicit their tumor-specific action, as well as establish that defects in apoptosis can produce treatment-resistant tumors. These systems also suggest a basis for the side-effects of cancer therapy, since many of the tissues that are adversely affected by radiation or chemotherapy normally undergo apoptosis following DNA damage (e.g. thymus, myeloid progenitor cells, epithelial stem cells). Ironically, the presence of functional p53 in these tissues may be deleterious to patients receiving cancer therapy. Assuming that further investigation confirms the existence of similar processes in human cancer, the role of p53 in apoptosis may have a substantial impact on the treatment of human malignancy. First, the identification of p53 mutations may become important in treatment decisions. Second, a better understanding of why therapies work should allow known drugs to be used more effectively and facilitate the rational design of better agents. In particular, the use of both p53-dependent and independent drugs may prove effective in combination. Finally, the use of drugs or genes that restore p53 activity may increase the efficacy of classical anticancer agents. Acknowledgments The author wishes to thank H.E.Ruley, T.Jacks, and D.Housman for excellent guidance and support and J.S.Smith, M.E.McCurrach, and T.Jacks for editorial assistance. References 1 2 3 4 5
6 7
. WEINBERG, R.A. (1991) Tumor suppressor genes. Science 254, 1138–1146. . WEINBERG, R.A. (1992) The integration of molecular genetics into cancer management. Cancer 70, 1653–1658. . KORSMEYER, S.J. (1992) Bcl-2 initiates a new category of oncogenes: regulators of cell death. Blood 80, 879–886. . HOLLSTEIN, M., SIDRANSKY, D., VOGELSTEIN, B. & HARRIS, C.C. (1991) p53 mutations in human cancers. Science 253, 49–53. . MALKIN, D., LI, F.P., STRONG, L.C., FRAUMENI, J.J.F., NELSON, C.E., KIM, D.H., KASSEL, J., GRYKA, M.A., BISCHOFF, F.Z., TAINSKY, M.A. & FRIEND, S.H. (1990) Germ line p53 mutations in a familial syndrome of breast cancer, sarcomas, and other neoplasms. Science 250, 1233–1238. . SRIVASTAVA, S., ZOU, Z., PIROLLO, K., BLATTNER, W. & CHANG, E. (1990) Germ-line transmission of a mutated p53 gene in a cancer-prone family with Li-Fraumeni syndrome. Nature 348, 747–749. . FEARON, E.R. & VOGELSTEIN, B. (1990) A genetic model for colorectal tumorigenesis. Cell 61, 759–767.
106
8 9 10 11 12 13 14
15
16 17 18
19 20 21 22 23
24 25 26 27
28 29 30
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. GREENBLATT, M.S., BENNETT, W.P., HOLLSTEIN, M. & HARRIS, C.C. (1994) Mutations in the p53 tumor suppressor gene: clues to cancer etiology and molecular pathogenesis. Cancer Res. 54, 4855–4878. . LEVINE, A.J., MOMAND, J. & FINLAY, C.A. (1991) The p53 tumour suppressor gene. Nature 351, 453–456. . VOGELSTEIN, B. & KINZLER, K.W. (1992) p53 function and dysfunction. Cell 70, 523–526. . EL-DEIRY, W.S., KERN, S.E., PIETENPOL, J.A., KINZLER, K.W. & VOGELSTEIN, B. (1992) Definition of a consensus binding site for p53. Nature Genet. 1, 45–49. . FARMER, G., BARGONETTI, J., ZHU, H., FRIEDMAN, P., PRYWES, R. & PRIVES, C. (1992) Wild-type p53 activates transcription in vitro. Nature 358, 83–86. . KERN, S.E., PIETENPOL, J.A., THIAGALINGAM, S., SEYMOUR, A., KINZLER, K.W. & VOGELSTEIN, B. (1992) Oncogenic forms of p53 inhibit p53-regulated gene expression. Science 256, 827–830. . ZAMBETTI, G.P., BARGONNETI, J., WALKER, K., PRIVES, C. & LEVINE, A.J. (1992) Wild-type p53 mediates positive regulation of gene expression through a specific DNA sequence element. Genes Dev. 6, 1443–1152. . SETO, E., USHEVA, A., ZAMBETTI, G.P., MOMAND, J., HORIKOSHI, N., WEINMANN, R., LEVINE, A.J. & SHENK, T. (1992) Wild-type p53 binds to the TATA-binding protein and represses transcription. Proc. Natl Acad. Sci. USA 89, 12028–12032. 12028–12032. . MACK, D.H., VARTIKAR, J., PIPAS, J.M. & LAIMINS, L.A. (1993) Specific repression of TATA-mediated but not initiator-mediated transcription by wild-type p53. Nature 363, 281–283. . DUTTA, A., RUPPERT, J.M., ASTER, J.C. & WINCHESTER, E. (1993) Inhibition of DNA replication factor RPA by p53 . Nature 365, 79–82. . WANG, X.W., FORRESTER, K., YEH, H., FEITELSON, M.A., GU, J.R. & HARRIS, C.C. (1994) Hepatitis B virus X protein inhibits p53 sequence-specific DNA binding, transcriptional activity, and association with transcription factor ERCC3. Proc. Natl Acad. Sci. USA 91, 2230–2234. . PIETENPOL, J.A., & VOGELSTEIN, B. (1993) No room at the p53 inn. Nature 365, 17–18. . CHO, Y., GORINA, S., JEFFREY, P.D. & PAVLETICH, N.P. (1994) Crystal structure of a p53 tumor suppressor-DNA complex: understanding tumorigenic mutations. Science 265, 346–355. . MICHALOVITZ, D., HALEVY, O. & OREN, M. (1990) Conditional inhibition of transformation and of cell proliferation by a temperature-sensitive mutant of p53. Cell 62, 671–680. . MARTINEZ, J., GEORGOFF, I., MARTINEZ, J. & LEVINE, A.J. (1991) Cellular localization and cell cycle regulation by a temperature-sensitive p53 protein. Genes Dev. 5, 151–159. . MERCER, W.E., SHIELDS, M.T., LIN, D., APPELLA, E. & ULLRICH, S.J. (1991) Growth suppression induced by wild-type p53 protein is accompanied by selective down-regulation of proliferating-cell nuclear antigen expression. Proc. Natl Acad. Sci. USA 88, 1958–1962. . YONISH-ROUACH, E., RESNITZKY, D., LOTEM, J., SACHS, L., KIMCHI, A. & OREN, M. (1991) Wildtype p53 induces apoptosis of myeloid leukaemic cells that is inhibited by interleukin-6. Nature 352, 345–347. . SHAW, P., BOVEY, R., TARDY, S., SAHLI, R., SORDAT, B. & COSTA, J. (1992) Induction of apoptosis by wild-type p53 in a human colon tumor-derived cell line. Proc. Natl Acad. Sci. USA 89, 4495–4499. . HARVEY, D.M. & LEVINE, A.J. (1991) p53 alteration is a common event in the spontaneous immortalization of primary BALB/c murine embryo fibroblasts. Genes Dev. 5, 2375–2385. . HARVEY, M., SANDS, A.T., WEISS, R.S., HEGI, M.E., WISEMAN, R.W., PANTIAZIS, P., GIOVANELLA, B.C., TAINSKY, M.A., BRADLEY, A. & DONEHOWER, L.A. (1993) In vitro growth characteristics of embryo fibroblasts isolated from p53-deficient mice. Oncogene 8, 2457–2467. . SHAULSKY, G., GOLDFINGER, N., PELED, A. & ROTTER, V. (1991) Involvement of wild-type p53 in preB-cell differentiation in vitro. Proc. Natl Acad. Sci. USA 88, 8982– 8986. . DITTMER, D., PATI, S., ZAMBETTI, G., CHU, S., TERESKY, A.K., MOORE, M., FINLAY, C. & LEVINE, A.J. (1993) Gain of function mutations in p53. Nature Genet. 4, 42–46. . DONEHOWER, L.A., HARVEY, M., SLAGLE, B.L., MCARTHUR, M.J., MONTGOMERY, C.A., BUTEL, J.A. & BRADLEY, A. (1992) Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumours. Nature 356, 215–220.
THE ROLE OF P53 IN APOPTOSIS
31
32 33
34
35 36 37 38 39 40
41
42 43 44 45
46
47
48
49 50
107
. HARVEY, M., MCARTHUR, M.J., MONTGOMERY, C.A., BUTEL, J.S., BRADLEY, A. & DONEHOWER, L.A. (1993) Spontaneous and carcinogen-induced tumorigenesis in p53-deficient mice. Nature Genet. 5, 225–229. . JACKS, T., REMINGTON, L., WILLIAMS, B.O., SCHMITT, E.M., HALACHMI, S., BRONSON, R.T. & WEINBERG, R.A. (1994) Tumor spectrum analysis in p53-mutant mice. Curr. Biol. 4, 1–7. . PURDIE, C.A., HARRISON, D.J., PETER, A., DOBBIE, L., WHITE, S., HOWIE, S.E.M., SALTER, D.M., BIRD, C.C., WYLLIE, A.H., HOOPER, M.L. & CLARKE, A.R. (1994) Tumour incidence, spectrum and ploidy in mice with a large deletion in the p53. Oncogene 9, 603–609. . KASTAN, M.B., ZHAN, Q., EL-DEIRY, W.S., CARRIER, F., JACKS, T., WALSH, W.V., PLUNKETT, B.S., VOGELSTEIN, B. & FORNACE, A., JR. (1992) Amammalian cell cycle checkpoint pathway utilizing p53 and GADD45 is defective in ataxia-telangiectasia. Cell 71, 587–597. . LIVINGSTONE, L.R., WHITE, A., SPROUSE, J., LIVANOS, E., JACKS, T. & TLSTY, T.D. (1992) Altered cell cycle arrest and gene amplification potential accompany loss of wild-type p53. Cell 70, 923–935. . MALTZMAN, W. & CZYZYK, L. (1984) UV irradiation stimulates levels of p53 cellular tumor antigen in nontransformed mouse cells. Mol. Cell Biol. 4, 1689–1694. . LANE, D.P. (1992) p53, guardian of the genome. Nature 358, 15–16. . LEE, J.M. & BERNSTEIN, A. (1993) p53 mutations increase resistance to ionizing radiation. Proc. Natl Acad. Sci. USA 90, 5742–5746. . KEMP, C.J., WHELDON, T. & BALMAIN, A. (1994) p53-deficient mice are extremely susceptible to radiation-induced tumorigenesis. Nature Genet. 8, 66–69. . YAMADA, Y., YOSHIDA, T., HAYASHI, K., SEKIYA, T., YOKOTA, J., HIROHASHI, S., NAKATANI, K., NAKANO, H., SUGIMURA, T. & TERADA, M. (1991) p53 gene mutations in gastric cancer metastases and in gastric cancer cell lines derived from metastases. Cancer Res. 51, 5800–5805. . ICHIKAWA, A., HOTTA, T., TAKAGI, N., TSUSHITA, K., KINOSHITA, T., NAGAI, H., MURAKAMI, Y., HAYASHI, K. & SAITO, H. (1992) Mutations of p53 gene and their relation to disease progression in B-cell lymphoma. Blood 79, 2701–2707. . TSUDA, H. & HIROHASHI, S. (1992) Frequent occurrence of p53 gene mutations in uterine cancers at advanced clinical stage and with aggressive histological phenotypes. Jpn. J. Cancer Res. 83, 1184–1191. . BOOKSTEIN, R., MAC GROGAN, D., HILSENBECK, S.G., SHARKEY, F. & ALLRED, D.C. (1993) p53 is mutated in a subset of advanced-stage prostate cancers. Cancer Res. 53, 3369–3373. . ODA, T., TSUDA, H., SCARPA, A., SAKAMOTO, M. & HIROHASHI, S. (1992) Mutation pattern of the p53 gene as a diagnostic marker for multiple hepatocellular carcinoma. Cancer Res. 52, 3674–3678. . OHUE, M., TOMITA, N., MONDEN, T., FUJITA, M M., FUKUNAGA, M., TAKAMI, K., YANA, I., OHNISHI, T., ENOMOTO, T., INOUE, M., SHIMANO, T. & MORI, T. (1994) A frequent alteration of p53 gene in carcinoma in adenoma of colon. Cancer Res. 54, 4798–4804. . KEMP, C.J., DONEHOWER, L.A., BRADLEY, A. & BALMAIN, A. (1993) Reduction of p53 gene dosage does not increase initiation or promotion but enhances malignant progression of chemically induced skin tumors. Cell 74, 813–822. . FUJIWARA, T., GRIMM, E.A., MUKHOPADHYAY, T., ZHANG, W.W., OWEN-SCHAUB, L.B. & ROTH, J.A. (1994) Induction of chemosensitivity in human lung cancer cells in vivo by adenovirus-mediated transfer of the wild-type p53 gene. Cancer Res. 54, 2287–2291. . RADINSKY, R., FIDLER, I.J., PRICE, J.E., ESUMI, N., TSAN, R., PETTY, M.C., BUCANA, C.D. & BARELI, M. (1994) Terminal differentiation and apoptosis in experimental lung metastases of human osteogenic sarcoma cells by wild type p53. Oncogene 9, 1877–1883. . RAMQVIST, T., MAGNUSSON, K.P., WANG, Y., SZEKELY, L., KLEIN, G. & WIMAN, K.G. (1993) Wildtype p53 induces apoptosis in a Burkitt lymphoma (BL) line that carries mutant p53. Oncogene 8, 1495–1500. . WANG, Y., RAMQVIST, T., SZEKELY, L., AXELSON, H., KLEIN, G. & WIMAN, K.G. (1993) Reconstitution of wild-type p53 expression triggers apoptosis in a p53-negative v-myc retrovirus-induced T-cell lymphoma line. Cell Growth Diff. 4, 467–473.
108
51 52 53 54 55 56 57 58 59 60 61 62 63
64
65 66 67 68 69 70 71 72
73
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. RYAN, J.J., DANISH, R., GOTTLIEB, C.A. & CLARKE, M.F. (1993) Cell cycle analysis of p53-induced cell death in murine erythroleukemia cells. Mol. Cell Biol. 13, 711–719. . DEBBAS, M. & WHITE, E. (1993) Wild-type p53 mediates apoptosis by E1A, which is inhibited by E1B. Genes Dev. 7, 546–554. . CAELLES, C., HEIMBERG, A. & KARIN, M. (1994) p53-dependent apoptosis in the absence of transcriptional activation of p53-target genes. Nature 370 220–223. . MACDONALD, H.R. & LEES, R.K. (1990) Programmed death of autoreactive thymocytes. Nature 343, 4495–4499. . CRABTREE, G.R. (1989) Contingent genetic regulatory events in T lymphocyte activation. Science 243, 335–362. . WYLLIE, A.H. (1980) Glucocorticoid-induced thymocyte apoptosis is associated with endogenous endonuclease activation. Nature 284, 555–556. . SELLINS, K.S. & COHEN, J.J. (1987) Gene induction by gamma-irradiation leads to DNA fragmentation in lymphocytes. J. Immunol. 139, 3199–3206. . CLARKE, A.R., PURDIE, C.A., HARRISON, D.J., MORRIS, R.G., BIRD, C.C., HOOPER, M.L. & WYLLIE, A.H. (1993) Thymocyte apoptosis induced by p53-dependent and independent pathways. Nature 362, 849–852. . LOTEM, J. & SACHS, L. (1993) Hematopoietic cells from mice deficient in wild-type p53 are more resistant to induction of apoptosis by some agents. Blood 82, 1092–1096. . LOWE, S.W., SCHMITT, E.M., SMITH, S.W., OSBORNE, B.A. & JACKS, T. (1993) p53 is required for radiation-induced apoptosis in mouse thymocytes. Nature 362 847–849. . ZHU, Y.M., BRADBURY, D.A. & RUSSELL, N.H. (1994) Wild-type p53 is required for apoptosis induced by growth factor deprivation in factor-dependent leukaemic cells. Br. J. Cancer 69, 468–472. . POTTEN, C.S., MERRITT, A., HICKMAN, J., HALL, P. & FARANDA, A. (1994) Characterization of radiation-induced apoptosis in the small intestine and its biological implications. Int. J. Rad. Biol. 65, 71–78. . CLARKE, A.R., GLEDHILL, S., HOOPER, M.L., BIRD, C.C. & WYLLIE, A.H. (1994) p53 dependence or early apoptotic and proliferative responses within the mouse intestinal epithelium following gamma-irradiation. Oncogene 9, 1767–1773. . MERRITT, A.J., POTTEN, C.S., KEMP, C.J., HICKMAN, J.A., BALMAIN, A., LANE, D.P. & HALL, P.A. (1994) The role of p53 in spontaneous and radiation-induced apoptosis in the gastrointestinal tract of normal and p53-deficient mice. Cancer Res. 54, 614–617. . KRAMMER, P.H., BEHRMANN, I., DANIEL, P., DHEIN, J. & DEBATIN, K.M. (1994) Regulation of apoptosis in the immune system. Curr. Opin. Immunol. 6, 279–289. . LOWE, S.W., CHECOVICH, W.J., RAPACZ, J. & ATTIE, A.D. (1988) Defective receptor binding of low density lipoprotein from pigs possessing mutant apoplipoprotein B alleles. J. Biol. Chem. 263, 15467–15473. . MARX, J. (1993) How p53 suppresses cell growth. Science 262, 1644–1645. . SENTMAN, C.L., SHUTTER, J.R., HOCKENBERY, D., KANAGAWA, O. & KORSMEYER, S. (1991) bcl-2 inhibits multiple forms of apoptosis but not negative selection in thymocytes. Cell 67, 879–888. . STRASSER, A., HARRIS, A.W. & CORY, S. (1991) bcl-2 transgene inhibits T cell death and perturbs thymic self-censorship. Cell 67, 889–899. . OLTVAI, Z.N., MILLIMAN, C.L. & KORSMEYER, S.J. (1993) bcl-2 Heterodimerizes in vivo with a conserved homolog, bax, that accelerates programmed cell death. Cell 74, 609–619. . MIYASHITA, T., HARIGAI, M., HANADA, M. & REED, J.C. (1994) Identification of a p53-dependent negative response element in the bcl-2 gene. Cancer Res. 54, 3131–3135. . MIYASHITA, T., KRAJEWSKI, S., KRAJEWSKA, M., WANG, H.G., LIN, H.K., LIEBERMANN, D.A., HOFFMAN, B. & REED, J.C. (1994) Tumor suppressor p53 is a regulator of bcl-2 and bax gene expression in vitro and in vivo. Oncogene 9, 1799–1805. . SELVAKUMARAN, M., LIN, H.K., MIYASHITA, T., WANG, H.G., KRAJEWSKI, S., REED, J.C., HOFFMAN, B. & LIEBERMANN, D. (1994) Immediate early upregulation of bax expression by p53 but not TGF beta 1: A paradigm for distinct apoptotic pathways. Oncogene 9, 1791–1798.
THE ROLE OF P53 IN APOPTOSIS
74 75 76
77 78 79 80 81
82 83 84
85 86
87 88 89
90 91 92
93 94 95
109
. WYLLIE, A.H. (1985) The biology of cell death in tumours. Anticancer Res. 5, 131–136. . KERR, J.F., WYLLIE, A.H. & CURRIE, A.R. (1972) Apoptosis: a basic biological phenomenon with wideranging implications in tissue kinetics. Br. J. Cancer 26, 239– 257. . SEARLE, J., LAWSON, T.A., ABBOTT, P.J., HARMON, B. & KERR, J.F.R. (1975) An electron-microscope study of the mode of cell death induced by cancer-chemotherapeutic agents in populations of proliferating normal and neoplastic cells. J. Path. 116, 129–138. . KERR, J.F.R., WINTERFORD, C.M. & HARMON, B.V. (1994) Apoptosis—Its significance in cancer and cancer therapy. Cancer 73, 2013–2026. . DIVE, C. & HICKMAN, J.A. (1991) Drug-target interactions: only the first step in the commitment to a programmed cell death? Br. J. Cancer 64, 192–196. . RULEY, H.E. (1990) Transforming collaborations between ras and nuclear oncogenes. Cancer Cells 2, 258–268. . RULEY, H.E. (1983) Adenovirus early region 1A enables viral and cellular trnasforming genes to transform primary cells in culture. Nature 304, 602–606. . WHYTE, P., BUCHKOVICH, K.J., HOROWITZ, J.M., FRIEND, S.H., RAYBUCK, M., WEINBERG, R.A. & HARLOW, E. (1988) Association between an oncogene and an anti-oncogene: the adeno virus E1A proteins bind to the retinoblastoma gene product. Nature 334, 124–129. . WHYTE, P., WILLIAMSON, N.M. & HARLOW, E. (1989) Cellular targets for transformation by the adeno virus E1A proteins. Cell 56, 67–75. . VAN DEN ELSEN, P., HOUWELING, A. & VAN DER EB, A. (1983) Expression of region E1B of human adeno viruses in the absence of region E1A is not sufficient for complete transformation. Virol. 128, 377–390. . RAO, L., DEBBAS, M., SABBATINI, P., HOCKENBERY, D., KORSMEYER, S. & WHITE, E. (1992) The adeno virus E1A proteins induce apoptosis which is inhibited by the E1B 19K and Bcl-2 proteins. Proc. Natl Acad. Sci. USA 89, 7742–7746. . WHITE, E., CIPRIANI, R., SABBATINI, P. & DENTON, A. (1991) Adenovirus E1B 19-kilodalton protein overcomes the cytoxicity of E1A proteins. J. Virol. 65, 2968–2978. . SARNOW, P., HO, Y.S., WILLIAMS, J. & LEVINE, A.J. (1982) Adenovirus E1b-58 kd tumor antigen and SV40 large tumor antigen are physically associated with the same 54 kd cellular protein in transformed cells. Cell 28, 387–394. . LOWE, S.W. & RULEY, H.E. (1993) Stabilization of the p53 tumor suppressor is induced by adenovirus E1A and accompanies apoptosis. Genes Dev. 7, 535–545. . LOWE, S.W., JACKS, T., HOUSMAN, D.E. & RULEY, H.E. (1994) Abrogation of oncogene-associated apoptosis allows transformation of p53-deficient cells. Proc. Natl Acad. Sci. USA 91, 2026–2030. . HOWES, K.A., RANSOM, L.N., PAPERMASTER, D.S., LASUDRY, J.G.H., ALBERT, D.M. & WINDLE, J.J. (1994) Apoptosis or retinoblastoma: alternative fates of photoreceptors expressing the HPV-16 E7 gene in the presence or absence of p53. Gene Dev. 8, 1300–1310. . PAN, H.C. & GRIEP, A.E. (1994) Altered cell cycle regulation in the lens of HPV-16 E6 or E7 transgenic mice: Implications for tumor suppressor gene function in development. Gene Dev. 8, 1285–1299. . SYMONDS, H., KRALL, L., REMINGTON, L., SAENZ-ROBLES, M., LOWE, S.W., JACKS, T. & VAN DYKE, T. (1994) p53-dependent apoptosis suppresses tumor growth and progression in vivo. Cell 78, 703–711. . DECAPRIO, J.A., LUDLOW, J.W., LYNCH, D., FURUKAWA, Y., GRIFFIN, J., PIWNICA-WORMS, H., HUANG, C.M. & LIVINGSTONE, D.M. (1988) SV40 large T antigen forms a specific complex with the product of the retinoblastoma susceptibility gene. Cell 54, 275–283. . DYSON, N., HOWLEY, P.M., MUNGER, K. & HARLOW, E. (1989) The human papilloma virus-16 E7 oncoprotein is able to bind the retinoblastoma gene product. Science 243, 934–936. . JACKS, T., FAZELI, A., SCHMITT, E.M., BRONSON, R.T., GOODELL, M.A. & WEINBERG, R.A. (1992) Effects on an Rb mutation in the mouse. Nature 359, 295– 300. . MORGENBESSER, S.D., WILLIAMS, B.O., JACKS, T. & DEPINHO, R.A. (1994) p53-dependent apoptosis produced by Rb-deficiency in the developing mouse lens. Nature 371, 72–74.
110
96 97 98 99
100
101 102 103 104 105
106 107 108 109
110 111 112
113
114 115 116
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. NEVINS, J.R. (1992) E2F: a link between the Rb tumor suppressor protein and viral oncoproteins. Science 258, 424–429. . JOHNSON, D.G., SCHWARZ, J.K., CRESS, W.D. & NEVINS, J.R. (1993) Expression of transcription factor E2F1 induces quiescent cells to enter S phase. Nature 365, 349–352. . WU, X.W. & LEVINE, A.J. (1994) p53 and E2F-1 cooperate to mediate apoptosis. Proc. Natl Acad. Sci USA 91, 3602–3606. . TANAKA, N., ISHIHARA, M., KITAGAWA, M., HISASHI, H., KIMURA, T., MATSUYAMA, T., LAMPHIER, M.S., ALZAWA, S., MAK, T.W. & TANIGUCHI, T. (1994) Cellular commitment to oncogeneinduced transformation or apoptosis is dependent on the transcription factor IRF-1. Cell 77, 829–839. . EVAN, G.I., WYLLIE, A.H., GILBERT, C.S., LITTLEWOOD, T.D., LAND, H., BROOKS, M., WATERS, C., PENN, L.Z. & HANCOCK, D.C. (1992) Induction of apoptosis in fibroblasts by c-myc protein. Cell 69, 119–128. . FANIDI, A., HARRINGTON, E.A. & EVAN, G.I. (1992) Cooperative interaction between c-myc and bcl-2 protooncogenes. Nature 359, 554–556. . MOMAND, J., ZAMBETTI, G.P., OLSON, D.C., GEORGE, D. & LEVINE, A.J. (1992) The mdm-2 oncogene product forms a complex with the p53 protein and inhibits p53-mediated transactivation. Cell 69, 1237–1245. . CHIOU, S.K., RAO, L. & WHITE, E. (1994) Bcl-2 blocks p53-dependent apoptosis. Mol. Cell Biol. 14, 2556–2563. . BRANTON, P.E., BAYLEY, S.T. & GRAHAM, F.L. (1985) Transformation by human adenoviruses. Biochim. Biophys. Acta 780, 67–94. . MUNGER, K., WERNESS, B.A., DYSON, N., PHELPS, W.C. & HOWLEY, P.M. (1989) The E6 and E7 genes of the papillomavirus type 16 together are necessary and sufficient for transformation of primary human keratinocytes. J. Virol. 63, 4417–4421. . BISSONNETTE, R.P., ECHEVERRI, F., MAHBOUBI, A. & GREEN, D.R. (1992) Apoptotic cell death induced by c-myc is inhibited by bcl-2. Nature 359, 552–554. . MAZARS, R., PUJOL, P., MAUDELONDE, T., JEANTEUR, P. & THEILLET, C. (1991) p53 mutations in ovarian cancer: a late event? Oncogene 6, 1685–1690. . NERI, A., BALDINI, L., TRECCA, D., CRO, L., POLLI, E. & MAIOLO, A.T. (1993) p53 gene mutations in multiple myeloma are associated with advanced forms of malignancy. Blood 81, 128–135. . BARDEESY, N., FALKOFF, D., PETRUZZI, M.J., NOWAK, N., ZABEL, B., ADAM, M., AGUIAR, M.C., GRUNDY, P., SHOWS, T. & PELLETIER, J. (1994) Anaplastic Wilm’s tumour, a subtype displaying poor prognosis, harbours p53 gene mutations. Nature Genet. 7, 91–97. . WILLIAMS, B.O., REMINGTON, L., ALBERT, D.M., MUKAI, S., BRONSON, R. T., & JACKS T. (1994) Cooperative tumorigenic effects of germline mutations in Rb and p53. Nature Genet. 7, 480–484. . ZUPPAN, C.W., BECKWITH, J.B. & LUCKEY, D.W. (1988) Anaplasia in unilateral Wilm’s tumor: a report from the national Wilm’s tumor study pathology center. Hum. Path. 19, 1199–1209. . PETTY, R.D., CREE, I.A., SUTHERLAND, L.A., HUNTER, E.M., LANE, D.P., PREECE, P.E. & ANDREOTTI, P.E. (1994) Expression of the p53 tumour suppressor gene product is a determinant of chemosensitivity. Biochem. Biophys. Res. Comm. 199, 264–270. . BRACHMAN, D.G., BECKETT, M., GRAVES, D., HARAF, D., VOKES, E. & WEICHSELBAUM, R.R. (1993) p53 mutation does not correlate with radiosensitivity in 24 head and neck cancer cell lines. Cancer Res. 53, 3667–3369. . SLICHENMYER, W.J., NELSON, W.G., SLEBOS, R.J. & KASTAN, M.B. (1993) Loss of a p53-associated G1 checkpoint does not decrease cell survival following DNA damage. Cancer Res. 53, 4164–4168. . LOWE, S.W., RULEY, H.E., JACKS, T. & HOUSMAN, D.E. (1993) p53-dependent apoptosis modulates the cytotoxicity of anticancer agents. Cell 74, 954–967. . LOWE, S.W., BODIS, S., MCCLATCHEY, A., REMINGTON, L., RULEY, H.E., FISHER, D., HOUSMAN, D.E. & JACKS, T. (1994) p53 status and the efficacy of cancer therapy in vivo. Science 266, 807–810.
THE ROLE OF P53 IN APOPTOSIS
117
118
119 120
121
122 123 124
125 126
127 128
129 130 131
132 133 134
111
. ELROUBY, S., THOMAS, A., COSTIN, D., ROSENBERG, C.R., POTMESIL, M., SILBER, R. & NEWCOMB, E.W. (1993) p53 gene mutation in B-cell chronic lymphocytic leukemia is associated with drug resistance and is independent of MDR1/MDR3 gene expression. Blood 82, 3452–3459. . HORIO, Y., TAKAHASHI, T., KUROISHI, T., HIBI, K., SUYAMA, M., NIIMI, T., SHIMOKATA, K., YAMAKAWA, K., NAKAMURA, Y. & UEDA, R. (1993) Prognostic significance of p53 mutations and 3p deletions in primary resected non-small cell lung cancer. Cancer Res. 53, 1–4. . THORLACIUS, S., BORRESEN, A.L. & EYFJORD, J.E. (1993) Somatic p53 mutations in human breast carcinomas in an Icelandic population: a prognostic factor. Cancer Res. 53, 1637–1641. . NAKAI, H., MISAWA, S., TANIWAKI, M., HORIIKE, S., TAKASHIMA, T., SERIU, T., NAKAGAWA, H., FUJII, H., SHIMAZAKI, C., MARUO, N., AKAOGI, T., UIKE, N., ABE, T. & KASHIMA, K. (1994) Prognostic significance of loss of a chromosome 17p and p53 gene mutations in blast crisis of chronic myelogenous leukaemia . Br. J. Haematol. 87, 425–427. . GAIDANO, G., BALLERINI, P., GONG, J.Z., INGHIRAMI, G., NERI, A., NEWCOMB, E.W., MAGRATH, I.T., KNOWLES, D.M. & DALLA-FAVERA, R. (1991) p53 mutations in human lymphoid malignancies: association with Burkitt lymphoma and chronic lymphocytic leukemia. Proc. Natl Acad. Sci. USA 88, 5413–5417. . HEIMDAL, K., LOTHE, R.A., LYSTAD, S., HOLM, R., FOSSA, S.D. & BORRESEN, A.L. (1993) No germline TP53 mutations detected in familial and bilateral testicular cancer. Genes Chrom. Cancer 6, 92–97. . HSIAO, M.H., YU, A.L., YEARGIN, J., KU, D. & HAAS, M. (1994) Nonhereditary p53 mutations in T-cell acute lymphoblastic leukemia are associated with the relapse phase. Blood 83, 2922–2930. . O’CONNOR, P.M., JACKMAN, J., JONDLE, D., BHATIA, K., MAGRATH, I. & KOHN, K.W. (1993) Role of the p53 tumor suppressor gene in cell cycle arrest and radiosensitivity of Burkitt’s lymphoma cell lines. Cancer Res. 53, 4776–4780. . MCILWRATH, A.J., VASEY, P.A., ROSS, G.M. & BROWN, R. (1994) Cell cycle arrests and radiosensitivity of human tumor cell lines: Dependence on wild-type p53 for radiosensitivity. Cancer Res. 54, 3718–3722. . FUJIWARA, T., GRIMM, E.A., MUKHOPADHYAY, T., CAI, D.W., OWEN-SCHAUB, L.B. & ROTH, J.A. (1993) A retroviral wild-type p53 expression vector penetrates human lung cancer spheroids and inhibits growth by inducing apoptosis. Cancer Res. 53, 4129–4133. . WYLLIE, A.H. (1993) Apoptosis (The 1992 Frank Rose Memorial Lecture). Br. J. Cancer 67, 205–208. . DEMERS, G.W., FOSTER, S.A., HALBERT, C.L. & GALLOWAY, D.A. (1994) Growth arrest by induction of p53 in DNA damaged keratinocytes in bypassed by human papillomavirus 16 E7. Proc. Natl Acad. Sci. USA 91, 4382–4386. . SHI, Y., GLYNN, J.M., GUILBERT, L.J., COTTER, T.G., BISSONNETTE, R.P. & GREEN, D.R. (1992) Role for c-myc in activation-induced apoptotic cell death in T cell hybridomas. Science 257, 212–214. . HARRINGTON, E.A., BENNETT, M.R., FANIDI, A. & EVAN, G.I. (1994) c-myc-induced apoptosis in fibroblasts is inhibited by specific cytokines. EMBO J. 13, 3286– 3295. . EL-DEIRY, W.S., TORINO, T., VELCULESCU, V.E., LEVY, D.B., PARSONS, R., TRENT, J.M., LIN, D., MERCER, W.E., KINZLER, K.W. & VOGELSTEIN, B. (1993) WAF1, a potential mediator of p53 tumor suppression. Cell 75, 817–825. . HARPER, J.W., ADAMI, G.R., WEI, N., KHANDAN, K. & ELLEDGE, S.J. (1993) The p21 cdk-interacting protein Cip1 is a potent inhibitor of G1 cyclin-dependent kinases. Cell 75, 805–816. . GAVRIELI, Y., SHERMAN, Y. & BENSASSON, S.A. (1992) Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J. Cell Biol. 119, 493–501. . LOWE, S.W., BODIS, S., BARDEESY, N., MCCLATCHEY, A., REMINGTON, L., RULEY, H.E., FISHER, D.E., JACKS, T., PELLETIER, J. & HOUSMAN, D.E. (1994) Apoptosis and the prognostic significance of p53 mutation. Cold Spring Harbor Symp. Quant. Biol. 59, 419–426.
7 Bcl-2 and the Regulation of Programmed Cell Death in Cancer JOHN C.REED La Jolla Cancer Research Foundation, Oncogene & Tumor Suppressor Gene Program, La Jolla, CA 92037, USA
7.1 Introduction Members of the bcl-2 gene family play a central role in regulating the relative sensitivity and resistance of cells to a wide variety of apoptotic stimuli. The first member of this multigene family, bcl-2, was discovered by virtue of its involvement in the t(14;18) chromosomal translocations commonly found in lymphomas [1–4]. Deregulation of the bcl-2 gene either by translocations in B-cell lymphomas or by other mechanisms in several other types of cancer contributes to neoplastic cell expansion by prolonging cell survival rather than by accelerating rates of cell division [5–8]. The Bcl-2 protein also can protect tumor cells from apoptosis induced by radiation and nearly all cytotoxic anticancer drugs [9–12], thus contributing to treatment failures in patients with cancer [13–16]. In addition to cytogenetic and molecular studies of human cancers which have suggested that bcl-2 represents a critical point of regulation of apoptotic processes in cells, hints of the central importance of bcl-2 as a regulator of programmed cell death have come from investigations of the genetics of developmental cell death in lower organisms as well as from viral genetics, where homologs of bcl-2 have been discovered such as the ced-9 gene in the nematode C. elegans [17] and the BHRF-1 and E1b-19 kDa proteins in Epstein-Barr virus and adenovirus, respectively [18–20]. Several additional homologs of bcl-2 have recently been discovered in mammals, including humans, revealing the presence of a multigene family [21–24]. Interestingly, some members of the bcl-2 gene family function as inhibitors of cell death, similar to bcl-2, whereas others are enhancers of apoptosis that oppose the actions of the Bcl-2 protein. Many of these Bcl-2 family proteins have the capacity to interact with each other through formation of homo- and heterotypic dimers [21, 25], revealing an important role for protein-protein interactions in the orchestration of Bcl-2 family protein function and suggesting approaches to pharmacologically manipulating the physiological cell death pathway. 7.2 Discovery of Bcl-2 at the Breakpoints of t(14;18) Translocations in Lymphomas Chromosomal translocations represent a general mechanism of proto-oncogene activation in human lymphomas and leukemias [reviewed in 26]. In ~80% of neoplasms of B-cell origin, translocations
BCL-2 IN CANCER
113
involving one of the immunoglobulin (Ig) gene loci can be detected by routine cytogenetic techniques. In these translocations, typically a cellular proto-oncogene located on a different chromosome becomes fused in a cis-configuration with the Ig heavy-chain (IgH) locus on chromosome 14 or one of the light-chain (IgL) loci on chromosomes 2 or 22. The result is that powerful transcriptional enhancer elements associated with the Ig gene loci exert their influence on the juxtaposed cellular proto-oncogene, thus deregulating its expression and causing continuously high levels of transcription of the involved oncogene. The first discovered example of chromosomal translocations as a mechanism for activating a cellular proto-oncogene came from studies of Burkitt lymphomas, were the c-myc gene becomes involved in translocations with the IgH or IgL genes in nearly 100% of cases [26]. Using this as a paradigm, it was speculated by Croce and colleagues that other novel proto-oncogenes might similarly become activated by a mechanism involving chromosomal translocations. Among the more common translocations seen in B-cell malignancies, are the t(11;14), typical of mantle cell lymphoma (also called intermediate differentiated lymphoma [IDL]), and the t(14;18) seen in most follicular lymphomas (also termed nodular poorly differentiated lymphoma [NPDL]). Thus the names bcl-1 and bcl-2 (for B-cell lymphoma-1 and 2) were coined for the genes that at that time were speculated to exist on chromosomes 11 and 18, respectively. Using probes derived from the IgH locus as a starting point, several independent groups of investigators then ‘walked’ across the t(14;18) breakpoints of non-Hodgkin’s lymphomas and into the adjacent bcl-2 gene sequences [1–4]. 7.2.1 Structure and Consequences of t(14;18) Translocations in Follicular Lymphomas Over 85% of follicular non-Hodgkin’s lymphomas contain t(14;18) translocations, suggesting that this genetic alteration represents an early event in the pathogenesis of these malignancies that arise from germinal center B-cells [27, 28]. DNA sequence analysis of the breakpoints of t(14;18) chromosomes has demonstrated that no two are identical [29]. Nevertheless, the breakpoints cluster into two regions on chromosome 18, with the major breakpoint cluster region (mbr) located within the 3 -untranslated region of the bcl-2 gene and a less commonly involved minor cluster region (mcr) residing 3 - and completely downstream of the bcl-2 transcriptional unit [30]. Thus, the breakpoints of t(14;18) chromosomes do not involve the coding regions of the bcl-2 gene. Though speculative, it has been suggested that t(14;18) translocations may arise due to errors in the normal DNA recombination mechanisms involved in the cutting and splicing of gene segments in the IgH locus, where heavy-chain proteins are encoded in separate V, D and J gene segments [1, 31]. This speculation is based in part of the finding of sequences within or near bcl-2 on chromosome 18 that resemble the classical heptamer/nonamer motifs flanking the V, D and J gene segments, which presumably represent recognition elements for cellular recombinases. Other theories however have also been advanced in an effort to explain the origins of t(14;18) and related translocations in human cancers, particularly the idea that Chi-like motifs located within or near bcl-2 may serves as DNA substrates for mediating illicit recombination events [32]. In this regard, lymphomas that contain t(14;18) translocations typically are widely disseminated and involve the bone marrow at the time of diagnosis [33], which lends indirect support to the notion of an origin for these translocations due to errors in V, D, J gene recombination, in as much as these gene rearrangements normally take place in the bone marrow at the preB-cell stage of B-cell differentiation. Regardless of the actual mechanisms responsible, the t(14;18) breakpoint can serve as a clonal marker for diagnosis and monitoring of patients with lymphomas, using the polymerase chain reaction (PCR) for detection of t(14;18)-containing cells at frequencies as low as 10−6 [33]. In this regard, PCR-based detection of malignant cells that harbor a t(14;18) has shown suggestions of
114
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
prognostic utility, particularly for monitoring patients after receiving ablative therapy followed by autologous bone marrow transplantation [34]. Based on nuclear run-on transcription assays, t(14;18) translocations appear to chiefly deregulate bcl-2 gene expression at the transcriptional level [36, 37]. Though some t(14;18) breakpoints fall within the 3’untranslated region of bcl-2 and result in the production of bcl-2/IgH fusion transcripts, measurements of the turnover of normal and fusion transcripts suggest similar half-lives of ~2.5 to 3 h under most circumstances [36–38]. In experiments where bcl-2 ‘minigenes’ have been linked with IgH enhancer elements in plasmid constructs, high levels of bcl-2 expression were obtained in B-cell lines, consistent with the notion that as-acting regulatory elements located in the IgH locus are fundamentally responsible for the alterations in bcl-2 gene expression seen in t(14;18)-containing B-cell lymphomas [39–41]. Hypomethylation of the promoter region of the bcl-2 gene also occurs in the translocated but not the unrearranged bcl-2 alleles of t(14;18)-containing lymphoma cell lines [39, 42]. In addition, changes in the DNase-I hypersensitivity of specific sites in this region of the bcl-2 gene have been reported, again consistent with the idea that transcriptional mechanisms play a major role in the deregulation of bcl-2 gene expression caused by t(14; 18) translocations [43]. The in vivo consequences of t(14;18) chromosomal translocations, where deregulation of bcl-2 gene expression is concerned, can perhaps best be appreciated by immunohistochemical comparisons of the patterns of Bcl-2 protein production in normal and neoplastic lymph nodes (Figure 7.1). In normal lymph nodes, Bcl-2 protein is found at high levels in the mantle zone region, a cuff of small dense lymphoid cells that surrounds the germinal center regions of secondary follicles. The mantle zone region comprises functionally a population of mostly long-lived ‘memory’ B-cells with recirculating capacity. Little Bcl-2 immunoreactivity is found within the centers of the follicles, where the germinal center B-cells reside, a population of cells that have recently encountered specific antigen and that are highly prone to apoptotic cell death [44–46]. These patterns of Bcl-2 immunoreactivity in normal nodes suggest that bcl-2 gene expression is normally shut-off as recirculating B-cells enter germinal centers and encounter antigens. Based on in vitro investigations using isolated germinal center B-cells, it has been suggested that those B-cells that successfully compete for antigen and that receive appropriate co-stimuli from helper T-cells are induced to re-express bcl-2 and thus are spared from apoptosis and allowed to exist the germinal center and join either the pool of recirculating memory cells seen in the mantle zone or to differentiate into antibodyproducing plasma cells [47]. In contrast to this distinct pattern of Bcl-2 immunoreactivity seen in normal nodes, immunostaining of follicular lymphoma specimens reveals strong Bcl-2 immunoreactivity in the germinal center compartment [48]. Presumably, the t(14;18)-containing lymphoma cells that take up residence in the germinal centers find the follicular regions of nodes a conducive environment for clonal expansion, where they enjoy a selective survival advantage relative to their normal B-cell counterparts. The natural history of follicular lymphomas is consistent with the function of bcl-2 as a regulator of cell lifespan, as opposed to cell division. Follicular non-Hodgkin’s lymphomas, for example, are considered lowgrade tumors, with patients experiencing median survivals of 5 to 8 years, even if untreated, compared to only 1 to 2 years for high-grade lymphomas [49]. In these patients, there occurs as gradual accumulation of malignant cells, culminating inevitably in patient death, despite various attempts at therapeutic intervention. Typically, >99% of the malignant cells are resting G0/G1-phase cells, again in keeping with notion that the primary defect in these indolent B-cell tumors represents a selective survival advantage instead of an increased proliferative rate. Probably the best evidence that deregulation of bcl-2 is indeed directly responsible for the characteristics of these tumors comes from transgenic mouse experiments were the transgene consisted of a human bcl-2 minigene linked to the IgH enhancer, thus emulating the t(14;18) structure and resulting in high levels of Bcl-2 protein production specifically in B-cells in these animals. In
BCL-2 IN CANCER
115
Figure 7.1 Normal and abnormal patterns of Bcl-2 protein production in lymph nodes The figure shows typical immunostaining results where Bcl-2 protein was localized using an specific antibody followed by detection using a diaminobenzidine-based colorimetric method that produces a brown color. Nuclei are counterstained with hematoxylin. In (A) [top panel], a benign reactive node is shown. Note that germinal centers do not stain for Bcl-2 but the surrounding cuff of mantle zone lymphocytes are strongly immunostained for Bcl-2. Scattered lymphocytes between the follicles are also immunostained. In (B) [bottom panel], a case of t(14;18)-containing follicular lymphoma is shown. Note the strong Bcl-2 immunoreactivity within the follicular center lymphocytes.
116
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
these mice, follicular expansions of mature B-cells were observed in nodes and spleen, which were initially polyclonal in origin. The preponderance of these bcl-2-expressing B-cells were small resting G0/G1-phase cells, which exhibited markedly prolonged survival in vitro when explanted into cultures, compared to Bcells derived from transgene-negative littermate controls [7, 8, 50]. Thus, the characteristics of the lymphoproliferative disorder seen in bcl-2/Ig transgenic mice is highly reminiscent of follicular lymphoma as it occurs in patients. Interestingly, in both patients with follicular lymphomas and in mice with bcl-2/Ig transgenes, transformation of the low-grade lesions to aggressive rapidly fatal tumors occurs frequently, and is often accompanied by the activation of additional cellular proto-oncogenes such as c-myc [51–53]. 7.3 Bcl-2 Gene Activation in Other Types of Cancer In addition to its activation because of chromosomal translocations in B-cell lymphomas, high levels and aberrant patterns of bcl-2 gene expression have been reported in several types of human cancer. Though the data are based largely on qualitative comparisons using immunohistochemical assays, it nevertheless appears that alterations in bcl-2 expression may occur in as much as about half of all cancers, including ~90% of colorectal adenocarcinomas, 30–60% of prostate cancers, 70% of breast adenocarcinomas, ~20% of squamous cell non-small cell lung cancers, ~60% of gastric cancers, ~80% of undifferentiated nasopharnygeal cancers, ~70% of chronic lymphocytic leukemias (CLLs), and various percentages of acute lymphocytic leukemias (ALLs), acute myelogenous leukemias (AMLs), neuroblastomas, renal cancers, small cell lung cancers, and melanomas [15, 16, 42, 56–63]. The details of how the patterns or levels of Bcl-2 protein production differ from normal cells vary among tumor types. Immunohistochemical analysis of Bcl-2 protein in colorectal lesions provides a particularly striking example of an alteration in the normal patterns of bcl-2 gene expression that can be seen in solid tumors. In the normal colonic mucosa, strong Bcl-2 immunoreactivity is present in the stem cell population that lies in the base of the crypts. Bcl-2 immunoreactivity dissipates in intensity as the colonic epithelial cells migrate up the crypts to the lumenal surface where they die by programmed cell death [56, 64]. In contrast, the normal gradient of bcl-2 expression is lost in pre-neoplastic adenomatous and malignant lesions of the colon, and instead, Bcl-2 protein is present at high levels throughout the length of the colonic crypts. Thus, deregulation of bcl-2 gene expression appears to occur relatively early in the progression of colorectal cancers, though the mechanisms involved remain unclear at present. In contrast to colorectal adenocarcinomas where deregulation of bcl-2 gene expression appears to represent an early event, activation of bcl-2 may often occur as a relatively late event in adenocarcinomas of the prostate. In an immunohistochemical analysis of primary tumor specimens by one group, for example, Bcl-2 positivity was seen in only about 10% of cases and was associated with aggressive histology (Gleason grades 9 and 10) [70]. Furthermore, the Bcl-2 immunostaining was heterogenious in these tumors, with some populations of tumor cells containing no Bcl-2 immunoreactivity and other more anaplastic cells displaying strong positivity, consistent with the idea that bcl-2 gene activation occurred as a relative late event in the pathogenesis of these malignancies. In contrast to primary tumors, bone metastases from heavily treated patients with hormone-refractory disease were strongly Bcl-2 positive in over 50% of cases, again suggesting that deregulation of bcl-2 occurred late in the course of progression of these tumors from androgen-sensitive localized disease to androgen-independent metastatic cancer. In prostate, colorectal, and other solid tumors examined to date, no evidence of gross structural alterations in the bcl-2 gene has been discovered by Southern blot analysis. Unlike the t(14;18)
BCL-2 IN CANCER
117
translocations seen in lymphomas, therefore, trans-rather than cis-regulatory mechanisms presumably are responsible for the alterations in bcl-2 expression that occur in most types of non-lymphomatous cancer. 7.3.1 Loss of p53 Tumor Suppressor as a Potential Mechanism of bcl-2 Deregulation One of the cis-regulatory mechanism that potentially may contribute to bcl-2 deregulation in cancers is loss of the p53 tumor suppressor gene. The p53 gene becomes inactivated in over half of all human cancers. The protein encoded by this gene has at least two important actions with regards to its ability to function as a tumor suppressor. First, p53 induces cell cycle arrest at the G1/S-border [97]. Second, p53 can induce apoptosis in some types of cells [98, 99]. In many cases however induction of p53 alone is insufficient to spontaneously trigger apoptotic cell death but can markedly increase the sensitivity of tumor cells to apoptosis induced by radiation or DNA-damaging drugs [91, 100]. The protein encoded by the p53 gene binds DNA and functions at least in part as a transcriptional regulator, acting as either an inhibitor or an inducer of gene expression depending on the particular target gene. Though the mechanisms by which p53 down-regulates the expression of particular target genes remains undetermined, its ability to upregulate gene expression has been associated with binding directly to specific DNA sequences having the consensus 5 -PuPuPuC(T/A)(A/T)GPyPyPy-3 [65]. Using a myeloblastic leukemia cell line that had lost p53 and a temperature-sensitive mutant of p53, Miyashita, et al. showed that conditional restoration of p53 function resulted in rapid down-regulation of bcl-2 mRNA levels, followed subsequently by a decline in Bcl-2 protein levels and apoptotic cell death [66]. In transient co-transfection assays, wild-type p53 was shown to be capable of down-regulating in a p53deficient human lung cancer line H358 the expression of reporter gene plasmids that contained a 195 bp DNA fragment derived from the 5 -untranslated region (5 -UTR) of the bcl-2 gene [67]. This p53-dependent negative response element (PNRE) functioned regardless of orientation and position, suggesting it has the characteristics of a transcriptional silencer. Immunoblot and immunohistochemical analysis of Bcl-2 protein levels in p53-deficient transgenic mice (‘knock-outs’) revealed elevated levels of Bcl-2 protein in some tissues, including spleen, thymus and prostatic epithelium, compared to normal littermate control animals that retained both copies of their p53 genes [66]. However, loss of p53 did not detectably affect bcl-2 expression in many tissues, implying that the extent to which basal levels of p53 influence bcl-2 is highly tissuespecific. For example, bcl-2 is not normally expressed in the liver and in the absence of p53, it was still not expressed, implying the existence of p53-independent mechanisms for repression of bcl-2 [66]. Indeed, a p53-independent negative regulatory element (NRE) has been described in the bcl-2 gene [43, 66]. The prediction of these observations is that in some but not all types of cancer, depending on the tissue of origin, loss of p53 will be associated with deregulation of bcl-2 gene expression. The data available thus far, however, clearly indicate that regulation of bcl-2 gene expression is complex, with multiple factors potentially providing input into the bcl-2 gene promoter gene. Thus, loss of p53 as a single parameter may not necessarily correlate with elevations in bcl-2 gene expression. It is unknown however what the in vivo influence of p53 is on bcl-2 expression in the setting of chemotherapeutic drug- or radiation-induced DNA damage, which is known to upregulate p53 protein levels and p53 transcriptional activity [68, 69]. Thus, while basal levels of p53 activity may be insufficient to significantly impact bcl-2 gene expression in some types of cells, the elevated levels of p53 activity associated with genotoxic stress conceivably could be important as an in vivo mechanism for downregulating bcl-2 and inducing apoptosis.
118
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
7.4 Bcl-2 and Chemoresistance in Cancer Though bcl-2 plays an important role in the origins of cancer where it contributes to neoplastic cell expansion by delaying or preventing normal cell turnover due to programmed cell death, perhaps more important are the potential effects of overexpression of this gene on responses to therapy. Using gene transfer methods to over-express bcl-2 in leukemia and solid tumor cell lines that contained low levels of Bcl-2 protein, as well as antisense approaches to reduce the levels of Bcl-2 protein in t(14;18)-containing lymphoma cell lines that contained high levels of this protein, it has been shown that the levels of Bcl-2 protein correlate with relative sensitivity or resistance to a wide spectrum of chemotherapeutic drugs as well as -irradiation [9–11, 16, 22, 73–83]. Included among the drugs that Bcl-2 has been experimentally shown to render cells more resistant to killing by are: dexamethasone, cytosine arabinoside (Ara-C), methotrexate, cyclophosphamide, adriamycin, daunomycin, 5-fluoro-deoxyuridine, 2-chlorodeoxyadenosine, fludarabine, taxol, etoposide (VP-16), camptothecin, nitrogen mustards, mitoxantrone, cisplatin, vincristine and some retinoids. The extent to which gene transfer-mediated elevations in Bcl-2 protein levels provide protection from the cytotoxic effects of these drugs varies, depending on the particular drug and the cell line, but can be as much as 4 or more logs (10000×) or as little as half a log (5×). When translated to clinical situations, however, even a 5-fold increase in resistance may be highly significant, given that most attempts to employ so-called ‘high-dose’ aggressive chemotherapy involve a mere doubling of the concentrations of drugs. The observation that Bcl-2 provides protection against such a wide variety of drugs which have markedly diverse mechanisms of action suggests that they all utilize the same final common pathway for ultimately inducing cell death and that Bcl-2 is a regulator of this pathway. Indeed, several studies have provided evidence that chemotherapeutic drugs, as well as -radiation, when administered in vitro to tumor cell lines induce cell death through mechanisms consistent with apoptosis as opposed to necrosis [84, 85]. Furthermore, the data argue that despite the diversity of their biochemical mechanisms of action, all of these drugs have in common the ability to activate the programmed cell death pathway at some point that lies upstream of Bcl-2. The drug resistance imparted to cancer cells by elevated levels of Bcl-2 protein differs from all other previously described forms of chemoresistance. Traditionally, pharmacologist have thought of the chemoresistance problem in cancer in terms of four major issues: (i) problems with delivery of drug to the target, such as when a drug is metabolized to an inactive product or when the mdr-1 gene product, Pglycoprotein, is over-produced in the plasma membrane of cancer cells and pumps drugs out of the cell; (ii) modification of the drug target, an example of which is amplification of the gene for dihydrofolate reductase which often occurs following exposure to methotrexate or loss of estrogen receptors in response to treatment with antiestrogens; (iii) increased rates of repair of damage to DNA or other structures; and (iv) diminished rates of drug-induced damage to DNA or other macromolecules, as can occur for some drugs when glutathione levels are elevated in tumors. Bcl-2, in contrast, appears to act through a different mechanism. Studies from several laboratories [74, 77, 83], for example, have shown that Bcl-2 does not prevent entry of drugs into cells. Bcl-2 also does not alter the extent to which drugs induce damage to DNA or the rate at which cells repair damaged DNA. Furthermore, no effects have been found of Bcl-2 on nucleotide pools or rates of cell cycling, which represent additional variables which can influence the relative sensitivity of cell to anticancer drugs. Similarly, though Bcl-2 was reported to produce elevations in intracellular glutathione levels in one neural cell line [86], this has not been observed in several other tumor and leukemia lines, indicating that no consistent relation of Bcl-2 to this intracellular antioxidant exists [87 and unpublished data]. It appears therefore that in the setting of Bcl-2 over-production, drugs still enter cells and induce damage, but this damage is somehow ineffectively translated into signals for cell death. In fact,
BCL-2 IN CANCER
119
it has been shown that anticancer drugs can still induce cell cycle arrest when Bcl-2 is present at high levels, but the cells typically fail to die or do so at markedly slower rates compared to control transfected cells [9, 10, 74, 77]. Thus, Bcl-2 can convert anticancer drugs from cytotoxic to cytostatic. Furthermore, when drugs are removed from cultures, a scenario that is analogous to the cessation of drugs that occurs clinically between cycles of chemotherapy, bcl-2 expressing cells can often reinitiate cell growth at higher rates than their control counterparts, in clonogenic cell assays [10, 77]. Similar effects have been reported for -irradiation, where again clonogenic assays indicate that bcl-2 can be highly radioprotective [73]. Presumably, therefore, because they do not die as easily when exposed to drugs and radiation, cells with elevated levels of Bcl-2 protein are able to survive through the period of drug treatment or radiation and then repair damaged DNA and resume their proliferation when drugs are withdrawn or after radiation. Taken together, these observations suggest that Bcl-2 defines a new category of drug- and radio-resistance gene, i.e., those that regulate the physiological cell death pathway. In addition to in vitro experiments, clinicocorrelative studies of bcl-2 expression in cancer patients have suggested that bcl-2 gene activation and high levels of Bcl-2 protein production may be important determinants of prognosis in at least some subgroups of patients. For example, in two studies of patients with non-Hodgkin’s lymphomas (NHLs) having diffuse histology with a large cell component (DLCL), an association was found between bcl-2 gene rearrangements and shorter survival, shorter disease-free survival (DFF), or failure to achieve a complete remission (CR) [13, 14, 87]. The data approached statistical significance (P=0.07) in a third study of DLCL but the median survival in this case was short (2 years), suggesting a need for longer follow-up [105]. Though bcl-2 status was not of prognostic significance in five other reports involving patients with aggressive histology NHL, in one study the combination of p53 and Bcl-2 immunostaining data defined a subgroup of patients at high risk for death [106]. Thus, as discussed above, the interplay between p53 and bcl-2 gene regulation may have been a contributor to the particularly poor prognosis observed for these patients. Furthermore, in several of the studies where the correlation between bcl-2 and survival did not reach statistical significance, there was a tendency of patients with evidence of bcl-2 gene activation to relapse or die sooner. For example, the 3 year survival for patients with bcl-2-positive tumors was only 45% compared to 75% in a report by Romaguera et al. [107] and the time to treatment failure was shorter for patients with Bcl-2-positive tumors (48% vs. 11%) in a study by Jacobson et al. [108]. Similarly, survival at 5 years was shorter for patients with Bcl-2-positive DLCL (35% vs. 46%) in a report by Piris et al. [106] as well as in a study by Offit et al. [14]. In patients with follicular lymphomas, Yunis et al. reported a significant association between bcl-2 gene rearrangements and both failure to achieve CR and reduced survival in cases where the histology included a large-cell component (FLCL) [13]. Conversely, in an analysis of patients with low-grade NHLs (follicular small-cleaved cell and follicular mixed cell), bcl-2 status was not of prognostic significance [109]. One limitation of this study, however, was that the size of the bcl-2-negative group was small, since >85% of low-grade NHLs contain a t(14;18). Taken together, these data suggest a trend towards a clinically significant role for bcl-2 gene activation and poor outcome in patients with lymphomas, particularly those that present a nodal (as opposed to extranodal) disease and where the histology includes a large cell component (DLCL; FLCL). Further studies involving larger groups of patients that received uniform treatment, however, are clearly required before any firm conclusions can be drawn as to the usefulness of Bcl-2 as a prognostic indicator when used either alone or in combination with other laboratory tests such as p53 immunostaining. In addition to lymphomas, suggestions of an association between bcl-2 and poor responses to therapy have been found in patients with acute myelogenous leukemia (AML), where the presence of 20% of Bcl-2positive cells correlated with failure to achieve CR and shorter survival [16], as well as in men with adenocarcinoma of the prostate where Bcl-2-positive immunostaining was correlated with failure to respond
120
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
to anti-androgen therapy [15]. Interestingly, in another report where a cross-sectional analysis of prostate cancers was performed, positive Bcl-2 immunostaining was found in 100% of hormone-independent cancers [57], again suggesting an association between elevations in Bcl-2 protein production and poor response to hormonal therapy. Though no survival data were available, Bcl-2 immunostaining was also positively correlated with unfavorable histology and N-myc gene amplification in one study of children and infants with neuroblastoma [110]. In addition, while no correlation between Bcl-2 immunostaining and histology or N-myc was noted by Krajewski et al. [111], elevated Bcl-2 immunostaining was seen in residual nests of viable tumor cells in 4 of 5 patients after therapy suggesting that Bcl-2 may have been cytoprotective. Correlations of Bcl-2 immunostaining with survival in a study of women with lymph nodenegative breast cancer and of patients with squamous cell carcinoma of the lung paradoxically suggested an inverse correlation between Bcl-2 and poor outcome [58, 60]. In these studies, however, the treatment was primarily or even exclusively surgical, with only some patients receiving local regional radiotherapy and none treated with systemic chemotherapy. Thus, the relevance of these findings in breast and lung cancers may be limited where the issue of bcl-2 as a modulator of chemosensitivity is concerned. Taken together, therefore, on balance the data suggest but fall short of proving that Bcl-2 status can be an important determinant of prognosis for patients with at least some types of cancer. Presumably, this association can be attributed to the ability of Bcl-2 to render tumor cells relatively more resistant to induction of apoptosis by a wide range of anticancer drugs as well as radiation. Further uni- and multivariate analyses of larger data sets from well controlled studies in which patients receive uniform therapy are required however before the relative importance of Bcl-2 as a determinant of clinical outcome is fully known. 7.4.1 Bcl-2 Blocks both p53-dependent and p53-independent Pathways for Drug-induced Apoptosis As mentioned above, p53 can be a regulator of bcl-2 gene expression. In addition, gene transfer studies have demonstrated that enforced production of Bcl-2 protein at high levels can partially or completely block apoptosis induced by p53 [101–103], suggesting a direct functional connection between p53 and its ability to both induce apoptosis and to down-regulate bcl-2 gene expression. Recently, a central role has emerged for p53 as a regulator of chemo- and radioresistance in tumors. For example in vitro gene transfer studies have shown that cultured cell lines which lack functional p53 exhibit increased resistance to induction of apoptosis by multiple anticancer drugs and radiation [91, 96]. Furthermore, p53 ‘knock-out’ mice, experience less radiation-induced apoptosis in the small intestine compared to normal littermate control animals [92, 93]. In addition, thymocytes isolated from p53 knock-out mice have impaired apoptotic responses to -irradiation and topoisomerase inhibitors, relative to p53-expressing control animals [94, 95]. Loss of p53 has also been associated with worse prognosis for patients with several types of cancers [for examples, see 88–90]. It is therefore tempting to speculate that p53 and bcl-2 may be functionally linked in a pathway that controls drug- and radiosensitivity. However, the situation is likely to be more complex, in that some types of anticancer drugs are able to induce apoptosis through mechanisms that are p53-independent and yet suppressible by Bcl-2. Studies with p53 knock-out mice, for example, have shown that while apoptosis induced in thymocytes by -radiation and DNA-damaging drugs is dependent on p53, apoptosis stimulated by glucocorticoids and calcium-ionophores is not [94, 95]. In contrast, both of these pathways for cell death are blocked in thymocytes derived from transgenic mice that produce high levels of Bcl-2 protein in the thymus [80–82]. Moreover, even DNA-damaging drugs and radiation can induce apoptosis through p53-
BCL-2 IN CANCER
121
independent mechanisms in some types of cells. In contrast to immature thymocytes, for example, apoptosis can be triggered in mature T-cells derived from p53 knock-out mice by radiation and chemotherapeutic drugs, despite the absence of p53 [104]. Again, apoptosis in these cells is also suppressible by Bcl-2. Thus, Bcl-2 appears to function at a point distal to the convergence of p53-independent and p53-dependent limbs of a final common pathway for drug-and radiation-induced apoptotic cell death. 7.5 Bcl-2 Regulates a Distal Event in an Evolutionarily Conserved Pathway for Cell Death In addition to rendering tumor cells relatively more resistant to induction of apoptosis by chemotherapeutic drugs and radiation, the Bcl-2 protein can also provide protection against a broad range of stimuli and insults that trigger the physiological cell death pathway. Most of these data are derived from gene transfer studies where cells were stably transfected with a bcl-2 expression vector, versus a control vector, and then challenged in various ways that are known to result in apoptotic cell death. For example, in hemopoietic and lymphoid cells, gene transfer-mediated elevations in Bcl-2 protein levels have been shown to markedly prolong survival when cells are placed into cultures without growth factors [5, 6, 112–114]. These Bcl-2 transfected cells still undergo cell cycle arrest in G0/G1-phase in the absence of growth factors, indicating that Bcl-2 does not render cells factor-independent for growth but rather specifically prolongs survival without simultaneously simulating mitogenesis. Similarly, microinjection of bcl-2 expression plasmids into sympathetic neurons strikingly delays apoptotic cell death caused by Nerve Growth Factor (NGF) deprivation [115]. Interestingly, Bcl-2 also can protect sensory neurons from death induced by withdrawal of NGF, Brain-Derived Neutrophic Factor (BNDF), or neurotrophin-3 (NT-3) but not ciliary neurons from Ciliary Neurotrophic Factor (CNTF) deprivation [116]. Bcl-2 also provides protection against apoptosis induced in neuronal cell lines by L-glutamate, an excitotoxic neurotransmitter thought to play an important role in stroke [117, 118]. Cell death induced by free radicals, drugs that generate free radicals in cells, and agents that interfere with glutathione synthesis in cells is also opposed by Bcl-2, though if very high concentrations of these agents are employed the cytoprotective effects of Bcl-2 can be overwhelmed [86, 8 7, 119]. In this case, however, the cell death typically is necrotic rather than apoptotic. TGF- (Transforming Growth Factor- ) induces apoptosis in some types of cells, and functions essentially as a tumor suppressor in epithelial and hemopoietic tissues. In a myeloblastic leukemia line, TGF- downregulated bcl-2 gene expression and induced apoptosis through a mechanism that was completely suppressible by transfection with a bcl-2 expression vector [136]. Interestingly, TGF- -mediated cell cycle arrest was not blocked by Bcl-2. In addition, cytokines that induce cell death such as Tumor Necrosis Factor (TNF) and Fas-ligand have also been shown to utilize Bcl-2-suppressible pathways to mediate their cytotoxic actions [123–125], though in some types of cells Bcl-2 provides little or no protection for reasons that will be discussed below [126, 127]. Similarly, cell death induced by cytolytic Tcells (CTLs) can be partially blocked by over-production of Bcl-2 protein in target cells [128, 129]. CTLs however have at their disposal a variety of mechanisms for killing target cells, some of which involve apoptosis (TNF, Fas-Ligand, proteases, ATP) and others necrosis (perforin), and thus Bcl-2 does not protect in all instances [82, 130]. Moreover, gene transfer studies have documented that Bcl-2 can protect cells against apoptosis induced by serine proteases derived from the cytotoxic granules of CTLs [131], as well as cysteine proteases of the ICE (interleukin-1- Converting Enzyme) family [132–134]. Bcl-2 also increases resistance to cell death induced by heat shock [120], as well as by calcium ionophores [9, 80–82, 119] and even some types of viruses [121, 122]. With regards to viruses, for example, the Tax protein of HIV has
122
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
been shown to induce apoptosis via a Bcl-2-suppressible mechanism [122]. In addition, Bcl-2 has been demonstrated to suppress the pro-apoptotic effects of the adenovirus E1a protein, which renders cells more sensitive to induction of apoptosis by serum withdrawal, DNA-damaging drugs and radiation [135]. In addition to viral oncogenes such as E1a, apoptosis induced by certain cellular oncogenes including cMyc, c-Myb, and R-Ras can be blocked by Bcl-2 [136–139]. In the case of c-Myc, for example, it has been shown that Myc simultaneously stimulates cellular pathways for both cell proliferation and apoptosis [140, 141] , The apoptotic effects of Myc can be suppressed by supplying cells with appropriate growth factors which generate survival signals. In that absence of growth factors, however, Myc-transfected cells undergo rapid apoptotic cells death via a mechanisms that is completely suppressible by Bcl-2 [137, 138]. This cooperation between Myc and Bcl-2 may explain why low-grade lymphomas that contain a t(14;18) involving bcl-2 take on an aggressive rapidly fatal phenotype when a subsequent t(8;14) translocation occurs that activates the c-myc gene [51, 52]. It has been argued that the dual role of Myc as both an inducer of mitogenesis and apoptosis helps to build additional controls into cell growth regulation, thus cordinating extracellular stimuli with intracellular gene expression. In addition, however, these observations imply that tumor cells may be more dependent on genes such as bcl-2 for their survival, thus offering hope that if the means of pharmacologically inhibiting bcl-2 function were developed, tumor cells would be rendered relatively more vulnerable to apoptosis compared with normal cells. In this regard, enforced production of high levels of Bcl-2 protein through gene transfer manipulations has also been shown to protect cell from apoptosis induced by loss of attachment to excellular matrix proteins mediated by certain integrins [124]. It is conceivable therefore that deregulated expression of bcl-2 contributes in at least some types of cancers to the acquisition of anchorage-independent growth, local invasiveness, and metastatic properties. This dependence on bcl-2 for survival in the absence of appropriate cell attachment signals may therefore again render tumor cells more dependent on bcl-2 than normal cells which retain their appropriate attachments to other cells and excellular matrix. The broad range of stimuli against which Bcl-2 can protect suggests that the Bcl-2 protein functions at a distal point in what may represent a final common pathway for apoptotic cell death. Thus, despite the various upstream ‘signals’ that are generated by these stimuli, eventually they must utilize the same mechanisms to ultimately kill cells, since Bcl-2 can provide protection from all of them. Nearly all of the cell death-inducing stimuli mentioned above have been shown to trigger apoptosis, as opposed to necrosis. In addition, however, Bcl-2 has also been reported to provide protection even in one model of necrotic cell death [86]. Furthermore, elements of the cell death pathway regulated by Bcl-2 appear to be well conserved throughout evolution, in that the human Bcl-2 protein has been shown to block cell death when expressed in insect cells, nematodes, and even yeast under some circumstances [17, 25, 86, 142, 143]. Though Bcl-2 clearly can have profound effects on the relative sensitivity of cells to apoptosis induction by a wide variety of insults and stimuli, most data argue that Bcl-2 is not absolutely required for cell survival. In experiments where antisense techniques were used to achieve reductions in Bcl-2 protein levels, for example, spontaneous cell death did not result, though the cells were markedly more sensitive to induction of death by growth factor deprivation and chemotherapeutic drugs [78, 144]. Similarly, in thymocytes derived from bcl-2 knock-out mice, rates of spontaneous cell death were not appreciably different for bcl-2-deficient and normal cells, but absence of bcl-2 was correlated with greater sensitivity to apoptosis induced by glucocorticoids and radiation [145, 146]. In fact, the relative normalcy of bcl-2 knock-out mice argues persuasively that Bcl-2 is not necessarily required for cell survival. Thus the effects of Bcl-2 on cell death pathways may be more analogous to the volume knob on a radio or record player than the on-off button. Bcl-2 does not turn-on a cell survival pathway or turn-off a cell death pathway, but rather adjusts the
BCL-2 IN CANCER
123
magnitude of cell death ‘signals’ so that either cell survival signals are amplified or cell death signals squelched. Despite the broad significance of Bcl-2 for regulation of cell death, there have been reported some scenarios where gene transfer-mediated elevations in Bcl-2 protein levels have failed to protect against cell death [80, 81, 126, 127, 130]. Furthermore, in some cases, the cell death process was clearly consistent with apoptosis as opposed to necrosis, such as with antigen receptor-induced apoptosis in some B-and Tlymphocyte cell lines [148, 150]. Though these data have often been used to argue for the existence of bcl-2independent pathways that regulate apoptosis, it is also possible that the mechanisms involved in cell death induction did indeed involve the bcl-2 pathway but that the mere over-production of the Bcl-2 protein was insufficient to provide protection for a variety of reasons, including absence of partner proteins that Bcl-2 may require to fulfill its mission as a cell death blocker, presence of high levels of proteins that inhibit Bcl-2, or stimulation of posttranslational modifications of the Bcl-2 protein that impair its function. 7.6 The Bcl-2 Protein: Possible Mechanisms of Action The predicted amino-acid sequence of the Bcl-2 protein has failed to provide any clues about the biochemical mechanism by which this protein blocks cell death. In humans, mice, rats and chickens, the protein has a molecular mass of ~25 to 26kDa and contains a stretch of hydrophobic amino-acids near its Cterminus that constitutes a transmembrane domain [151]. The intracellular membranes into which Bcl-2 inserts are strikingly unusual compared to other known proteins. A combination of subcellular fractionation, immunofluorescence confocal, laser-scanning, and electron microscopic methods have provided conclusive evidence that Bcl-2 is associated with mitochondria, specifically the outer mitochondrial membrane, as opposed to the inner membrane where many of the steps of oxidative phosphorylation occur [152–156]. Consistent with the absence of Bcl-2 from the inner membrane, it has been shown by use of mutant cells that lack mitochondrial DNA that absence of a complete respiratory chain does not interfere with ability of Bcl-2 to block apoptosis [157]. Bcl-2 immunoreactivity in the outer membrane of mitochondria is not uniformly distributed, but rather is patchy in its distribution—a property which is suggestive of proteins that associate with the mitochondrial junctional complexes where the inner and outer membranes come into contact and where various transport phenomenon occur. In addition to the mitochondrial outer membrane, much of the Bcl-2 protein is found in the nuclear envelope. Similar to the situation with mitochondria, electron microscopic data suggest that the Bcl-2 protein is non-uniformly distributed in the nuclear envelope, but occurs in a punctate pattern that is reminiscent of nuclear pore complexes where the inner and outer nuclear membranes come into contact and where transport between nucleus and cytosol of proteins, RNA, and possibly ions occurs. Bcl-2 is also found in at least parts of the endoplasmic reticulum. Though the functional significance of the unusual intracellular distribution of the Bcl-2 protein remains unclear, the possible association of Bcl-2 with mitochondrial junctional complexes (MJCs) and nuclear pore complexes (NPCs) is of particular interest. The nucleus and mitochondria have several features in common, including the fact that both contain DNA. Both the nucleus and mitochondria are also the only intracellular organelles that have a two membrane system, an outer and an inner membrane. The MJCs and NPCs where these membranes come into contact are the sites of transport of macromolecules and possibly some ions into and out of these organelles. Undoubtedly, the MJCs and NPCs are also critical structures for maintaining the integrity of mitochondria and nuclei, and disruptions of these multiprotein complexes would presumably compromise the structure and function of these essential organelles. In support of a role for Bcl-2 in regulating some aspect of protein transport in the nuclear envelope, reduced ratios of nuclear to cytosolic
124
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
cdc-2 and cdk-2 kinase have been detected in HeLa cells transfected with bcl-2 expression plasmids [158]. In addition, translocation of the p53 protein from cytosol into the nucleus was reported to be prevented by co-transfection of a mouse erythroleukemia line with the combination of bcl-2 and c-myc expression vectors [102]. Conversely, Bcl-2 was able to block p53-induced apoptosis in a v-myc-transformed T-cell lymphoma with out disturbing p53 translocation into the nucleus [101], suggesting that interference with transport of p53 is not a consistent observation among different tumor lines that simultaneously over-produce Bcl-2 and Myc oncoproteins. Gene transfer-mediated elevations in Bcl-2 protein also do not interfere with nuclear accumulation of p53 in other cell lines where c-myc is not overexpressed [103, 160, 198]. Moreover, Bcl-2 has been shown to protect the cytoplasm of enucleated cells from ‘apoptosis’, suggesting that the presence of a nucleus is not essential for Bcl-2 action [159]. A role for Bcl-2 in regulating protein transport in mitochondria however has not been explored. Another possible functional implication of the intracellular locations of the Bcl-2 protein is suggested by data showing that Bcl-2 can influence intracellular Ca2+ homeostasis [161, 162]. For example, in an IL-3dependent hemopoietic cell line 32 D a striking loss of Ca2+ from the endoplasmic reticulum (ER) was seen in control cells prior to apoptosis induction by growth factor withdrawal, whereas ER pools of Ca2+ were maintained in the normal range in cells over-producing Bcl-2. Conversely, estimates of mitochondrial Ca2+ pools suggested that elevations occur in the amounts of releasable Ca2+ in mitochondria and that Bcl-2 prevents the accumulation of Ca2+ in this organelle [161]. A functional connection between dysregulation of intracellular Ca2+ and apoptosis has been well-established by experimentation involving use of Ca2+ionophores and other agents, including the observation that apoptosis is induced by thapsigargin—a drug that poisons the Ca2+-ATPase of the ER and results in massive loss of Ca2+ from this organelle [163]. Similarly, gene transfer-mediated elevations in calbindin-D, a Ca2+ binding protein that resides in the lumen of the ER, have been shown to delay the onset of apoptosis in a glucocorticoid-treated lymphoid cell line, arguing that increasing the ability of the ER to sequester Ca2+ protects against apoptosis [164]. In this regard, the rate of efflux of Ca2+ from the ER was shown to be substantially reduced in Bcl-2-transfected WEHI7.1 T-cell lymphoma cells compared to controls when treated with thapsigargin [162]. The presence of Bcl-2 in nuclear and ER membranes, therefore, may have some relevance to the fact that most of the Ca2+ in cells is sequestered in the lumen of the ER and, by extension, the space between the inner and outer nuclear membranes. Furthermore, in most types of cells, the mitochondria represent the next largest intracellular storage site for Ca2+, again suggesting that Bcl-2 is at least located in the right places to function either directly or indirectly as a regulator of intracellular Ca2+ homeostasis. Perhaps relevant to a role for mitochondria in Ca2+ sequestration during apoptosis, mitochondria were reported to be absolutely required for the apoptosis-like nuclear disintegration seen in a cell-free assay in which ‘apoptotic’ cytosolic extracts were mixed with nuclei; and addition of Ca2+ ionophores blocked nuclear destruction in this system [165]. It has also been suggested that Bcl-2 may function in an antioxidant pathway, based on the findings that: (i) Bcl-2 prevents induction of apoptotic and (in some cases) necrotic cell death induced by agents that either result in oxygen free radical production or that deplete intracellular glutathione; (ii) over-expression of certain antioxidant enzymes such as forms of superoxide dismutase (SOD) or glutathione peroxidase can also render cells more resistant to induction of cell death analogous to Bcl-2; and (iii) Bcl-2 prevents the accumulation of lipid peroxides, suggesting that Bcl-2 somehow nullifies damage to membranes by reactive oxygen species [86, 87]. The relevance of these findings to the intracellular locations of the Bcl-2 protein could be that mitochondrial, ER and plasma membranes are the major sites of free-radical generation in cells. Additional evidence supporting a possible role for a redox mechanism for Bcl-2 comes from studies of SOD-deficient yeast, where expression of the human Bcl-2 protein was shown to restore growth under
BCL-2 IN CANCER
125
aerobic conditions [86]. Also, in bcl-2 knock-out mice, hypopigmentation of coat hairs occurs during the second hair follicle cycle—a finding which has been speculated to reflect a defect in one of the redoxdependent steps of melanin synthesis in melanocytes [146, 166]. However, in a cell-free system for ‘apoptosis’ in which Bcl-2 can function to prevent nuclear breakdown and DNA degradation, chemicals that modulate redox conditions had no significant effects on either induction of apoptotic-like changes in nuclei or the ability of Bcl-2 to function [165]. Also, Bcl-2 is able to block apoptosis induced by staurosporine and anti-Fas antibodies in fibroblasts grown under anaerobic conditions, arguing against a requirement for reactive oxygen specifies [167], though these observations do not exclude a role for redoxsensitive, thiol-based chemical reactions. At present, however, no data have been obtained that directly link Bcl-2 to the regulation of antioxidant pathways or any other particular mechanism such as Ca2+ or protein transport. 7.7 Bcl-2 Homologs and Interacting Proteins In the absence of a clear biochemical function for the Bcl-2 protein, a number of groups have searched for proteins that interact with Bcl-2 in the hopes that the predicted amino-acid sequences of these Bcl-2interacting proteins would provide insights into the mechanism of action of Bcl-2. Using a variety of interaction cloning techniques, as well as protein purification and sequencing, several proteins have now been identified which are capable of specifically binding to or at least co-immunoprecipitating with Bcl-2. One class of Bcl-2-interaction proteins represents homologs of Bcl-2, which can form heterotypic dimers with Bcl-2 as well as homotypic dimers with themselves in some cases. A second class of Bcl-2-binding proteins can be defined as proteins that do not share homology with Bcl-2. Knowledge about these proteinprotein interactions is beginning to provide insights into the molecular details of how the Bcl-2 protein functions. 7.7.1 Bcl-2 Homologs At present, six mammalian homologs of Bcl-2 have been reported, including Bax, Bcl-X, Mcl-1, A1, Bad and Bcl-y [21–24, 168, 169] (Table 7.1). Some of these proteins have additional forms that arise through alternative splicing mechanisms, the most interesting to date of which are the long and short forms of Bcl-X. The Bcl-X-L and Bcl-X-S proteins have opposing functions, with Bcl-X-L functioning as a blocker of cell death analogous to Bcl-2 and the Bcl-X-S protein acting as an antagonist of Bcl-2 which accelerates apoptotic cell death [3]. In addition to Bcl-X-S, some of the other Bcl-2-like proteins have been shown to function as an inducers of rather than protectors from cell death, including Bax, Bad, and Bak [21, 168, 169]. Conversely, the Mcl-1 and A1 proteins function as cell death blockers, though perhaps less efficiently than Bcl-2 [our unpublished observations]. In addition to mammalian homologs, several homologs of Bcl-2 have been described in viruses, including the E1b-19 kDa protein of adenovirus, the BHRF-1 protein of EpsteinBarr virus (EBV) and the LMWS-HL open reading frame found in the African Swine Fever Virus [18–20, 170]. Both E1b and BHRF-1 function as blockers of cell death, whereas the properties of the LMWS-HL protein have yet to be reported. A homolog of bcl-2 has also been discovered in the nematode, C. elegans, which functions as a blocker of cell death and has been termed ced-9 [17]. Figure 7.2 depicts the structures of the mammalian homologs of Bcl-2. Sequence alignments have identified the presence of three conserved domains, termed Bcl-2 domains (BD) a, b and c [171]. As shown,
126
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
most members of this family contain a stretch of hydrophobic amino-acids at their C-terminus that presumably allows for post-translational insertion into membranes. The Bad and A1 proteins however lack any obvious transmembrane domains. In addition, alternatively spliced versions of Bcl-2, Bcl-X-L and Bax have been described that do not contain membrane anchoring sequences [2, 21, 172]. For the most part, however, these splicing variants are Table 7.1 Characteristics of known Bcl-2 interacting proteins Protein
Description
Binding to other Bcl-2 homologs*
Function
Bax
Bcl-2 homolog
Death
Bcl-X-L
Bcl-2 homolog
Bcl-X-L, Mcl-1, A1, Bad [not Bcl-X-S or Bad] Bcl-X-L, Mcl-1, Bcl-X-S, Bad Bcl-X-L
Bcl-X-S
Survival
Bcl-2 homolog Death [missing BD(b) and BD(c)] Mcl-1 Bcl-2 homolog Bcl-X-L, Bax Survival [PEST sequences] [weak] # A1 Bcl-2 homolog Bcl-2, Bcl-X-L, Bax Survival [no TM domain] [weak] # Bak Bcl-2 homolog Bcl-2, Bcl-X-L, BHRF-1, Death [also called Cdn-1 and 19. E1b-19 kDa 1] [Not Bcl-X-S] Bad Bcl-2 homolog Bcl-X-L Death [no TM domain; missing [not Bax] BD(a)] BAG-1 Ubiquitin-like domain; Bcl-2, Bcl-X-L, Bcl-X-S Survival acidic; no TM domain R-Ras GTPase N.T. Death Raf-1 serine/threonine-protein N.T. Survival kinase Nip-1 Phosphodiester homology; E1b-19 kDa Unknown TM domain; PEST sequences Nip-2 Ca2+ -binding motif E1b-19 kDa Unknown homology; no TM domain; PEST sequences Nip-3 Calbindin-D domain; E1b-19 kDa Unknown TM domain; PEST sequences * Only those interactions with other Bcl-2 family protein that have been experimentally documented are indicated. In many cases, binding to other members of the Bcl-2 protein family has not been tested. # Anti-apoptotic activity may be weaker than for Bcl-2, based on gene transfer studies. TM=transmembrane domain. N.T.=not tested.
relatively less abundant than the forms shown in Figure 7.2 and their functions not well studied to date. In studies where truncation mutants of Bcl-2 have been prepared that lack the transmembrane domain, function
BCL-2 IN CANCER
127
as a blocker of apoptosis in lymphokine-dependent hemopoietic cells was shown to be impaired relative to the wild-type Bcl-2 protein, but not completely absent [87, 173]. Similar results were obtained for an epithelial cancer cell line, where apoptosis was induced by E1b-19-kDa-deficient adenovirus [174]. Moreover, replacement of the transmembrane domain of Bcl-2 with heterologous membrane targeting sequences derived from either the IL-2 receptor of a mitochondrial outer membrane protein Mas70p restored the anti-
128
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 7.2 Structure of Bcl-2 family proteins The structures of the human Bcl-2 protein and some of the known mammalian homologs of Bcl-2 are depicted. The domains with a high degree of conservation of sequence homology are indicated as A, B, and C, as well as the locations of the transmembrane domains (TM). The NR-13 protein is derived from chicken [189] and Ced-9 is from C. Elegans.
apoptotic function of these truncation mutants to nearly normal levels [173, 174]. Conversely, in TNFtreated L929 fibroblasts and NGF-deprived sympathetic neurons, inhibition of cell death by transmembranedeficient versions of Bcl-2 was essentially comparable to the wild-type Bcl-2 protein, suggesting that Bcl-2 need not necessarily target membranes [175]. Given that Bcl-2 can form homotypic dimers with itself as well as heterotypic dimers with several of its homologs [25], it remains possible that transmembranedeficient versions of Bcl-2 were nevertheless able to localize at least in part to the usual membrane sites
BCL-2 IN CANCER
129
Figure 7.3 Model for functional and physical interactions among Bcl-2 family proteins. The evidence available to date supports a model wherein Bax functions possibly as a homodimer to increase the sensitivity of cells to apoptotic stimuli. Binding of Bax by either Bcl-2, Bcl-X-L, or Mcl-1 disrupts Bax/Bax homodimerization and protects cells from apoptosis. Proteins such as Bcl-X-S and Bad probably exert their influence indirectly by forming heterodimers with Bcl-2, Bcl-X-L, and possibly other anti-apoptotic Bcl-2 homologs. This then prevents Bcl-2 or Bcl-XL from forming heterodimers with Bax, thus leaving Bax unopposed and consequently increasing sensitivity to cell death stimuli.
through protein-protein interactions. In this regard, subcellular localization of the Bcl-X-L and Bcl-X-S proteins suggest that these proteins reside at the same or similar membrane sites as Bcl-2 [172]. Also, though only examined at the level of conventional light microscope the intracellular immunostaining patterns of antibodies specific for Bax and Mcl-1 are very similar to Bcl-2 in that punctate immunostaining of cytosolic structures resembling mitochondria is seen, as well as nuclear and perinuclear membranes in some cells [176, 177]. Likewise, electron microscopic analysis of the subcellular localization of the BHRF-1 protein from EBV suggests association with the outer mitochondrial membrane and other sites typically occupied by Bcl-2 [178]. The 19-kDa E1b protein also has a similar localization, but appears to reside more so in nuclear and perinuclear membranes than mitochondrial membranes [179]. Though some members of the Bcl-2 protein family were discovered by virtue of their ability to bind to Bcl-2, this is not the case for the majority of these proteins. Thus, to date, it is unknown whether all of the homologs of Bcl-2 can form heterotypic dimers with the Bcl-2 protein. Nevertheless, several homologs have been investigated in this regard, mostly by use of yeast two-hybrid assays, though in some cases in vitro bindings studies using recombinant fusion proteins or coimmunoprecipitation experiments involving mammalian cells have also been performed. These studies suggest that Bcl-2 can form heterodimers with the Bax, Bcl-X-L, Bcl-X-S, Mcl-1 Bad and Bak proteins [21, 25, 168, 169]. In those cases tested, the Bcl-XL protein appears to have similar binding characteristic, and has been shown to interact specifically with Bax, Bcl-X-S, Mcl-1, Bad, and itself, in addition to Bcl-2 [25, 168]. This result is not entirely surprising, since the Bcl-2 and Bcl-X-L proteins are 47% identical in their amino-acids sequences [22]. Essentially nothing is known at present about the affinities of these protein-protein interactions. Furthermore, though it is convenient to think of these protein interactions as homo- and heterotypic dimers, their stoichiometry remains undetermined to date.
130
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Investigations of the domains within Bcl-2 family proteins required for homoand heterotypic dimer formation have thus far been consistent with an antiparallel or head-to-tail arrangement, wherein structures present within the first ~ 80 aminoacids of Bcl-2 appear to interact with structures present within the carboxyl portions of the protein [25]. For example, it was shown by use of yeast two-hybrid assays that amino-acids 1–81 of the human Bcl-2 protein can mediate interactions with amino-acids 83–218, whereas the 83–218 fragment cannot homodimerize with itself. Either the 1–81 or the 83–218 fragment of Bcl-2 appears to be sufficient for interactions with Bcl-X-L and Mcl-1, but other Bcl-2 family proteins have not been tested in this regard. The BD(a), BD(b) and BD(c) domains of Bcl-2 all appear to be required for Bcl-2 to retain cell death blocking function [124, 180, 181]. Of particular note, it has been shown in a lymphokinedependent hemopoietic cell line that mutations in the BD(b) and BD(c) domains of Bcl-2 (also termed BH1 and BH2 domains) are required for Bcl-2 to co-immunoprecipitate with Bax but do not impair associations with wild-type Bcl-2 protein [180]. These mutant forms of Bcl-2 that fail to bind to Bax are also deficient in function where blocking of cell death caused by lymphokine deprivation is concerned. These results have been interpreted as evidence that for Bcl-2 to function, it must be able to bind to Bax. Since gene transfermediated elevations in Bax protein levels accelerate the rate of cell death caused by growth factor withdrawal, whereas Bcl-2 has the opposite effect, it has been argued that the ratio of Bax and Bcl-2 proteins determines the relative sensitivity of cells to apoptosis [21]. Furthermore, the mutagenesis studies suggest that the Bcl-2/Bax interaction defines a critical aspect of this regulation of susceptibility to apoptotic cell death. When expressed in yeast (S. cerevisiae), the Bax protein confers a lethal phenotype that can be specifically neutralized by co-expression of Bcl-2, Bcl-X-L, or Mcl-1 [25]. For suppression of Bax-induced death in yeast, it is not necessary that the transmembrane domains of the Bcl-2, Bcl-X-L and Mcl-1 protein be included [25]. Mutant versions of Bcl-2 that fail to bind to Bax also fail to suppress Bax-mediated cytotoxicity in yeast [25, 181]. These observations suggest that elements of the Bax/ Bcl-2 pathway may be conserved even in single-cell eukaryotic organisms and raise the possibility of applying yeast genetics approaches to delineation of some of the downstream effectors or even upstream activators involved in the physiological cell death pathway. It is not immediately obvious why single cell organisms such as yeast might have mechanisms for committing suicide, unlike multicellular organisms where altruistic cell death can be easily reconciled with the greater goal of protection of the whole organism. One idea however is that yeast may use suicide as a means of minimizing the deleterious effets of viruses, so that viral production would be limited and the likelihood of infection of all progeny of a given yeast cell reduced. Elements of this cell death pathway could then have been transferred to multicellular eukaryotes, where again it would have served the organism when confronted with viruses. Clearly, however, the finding of cell deathblocking homologs of Bcl-2 in viruses suggests that some viruses have ‘learned’ to subvert the suicide defense mechanism, a finding that may also be relevant to the issue of viral latency whereby viral genomes can ensure their longevity by piggybacking on the chromosomes of host cells and then resume their replication at opportune times. It may also be relevant in this regard that genes encoded within the EBV virus also appear to be able to induce expression of the endogenous bcl-2 gene, though it is uncertain whether this is a direct effect of viral proteins on the expression of host cell genes versus a selection phenomenon where only those cells that happen to have higher levels of Bcl-2 protein survive [182–184]. In contrast to Bax, the cell death enhancing protein Bcl-X-S is not lethal when expressed in yeast, at least when lacking its transmembrane domain, which is the only way that the function of this protein in yeast has been tested thus far [25]. This observation has suggested that Bcl-X-S may enhance cell death by a different mechanism than Bax. Because of an alternative splicing event, the Bcl-X-S protein is missing a wellconserved region that includes the BD(b) and BD(c) domains [22]. Its ability to interact with Bcl-2 and Bcl-
BCL-2 IN CANCER
131
X-L thus appears to be dependent on preservation of the NH2-terminal portion of the molecule where the BD (a) region resides. Interestingly, the interaction of Bcl-X-S with either Bcl-2 or Bcl-X-L in two-hybrid assays has been reported to be significantly stronger than dimerization of Bcl-2 and Bcl-X-L with themselves or each other [25], though certainly such assays are far from quantitative. When taken together with the evidence that Bcl-2/Bax interactions may be particularly important for Bcl-2 to function as a cell death inhibitor, these observations suggest that Bcl-X-S may antagonize Bcl-2 by binding to it and thus preventing Bcl-2 from forming heterodimers with Bax. This would then leave Bax unopposed to increase the sensitivity of cells to apoptotic stimuli. A similar story appears to apply for the Bad protein. This homolog of Bcl-2 is composed essentially of only the carboxyl-portions of Bcl-2 including domains with homology to the BD(b) and BD(c) regions, but lacks the NH2-terminal sequences where BD(a) resides and also lacks a transmembrane domain. Nevertheless, Bad can bind to Bcl-2 and Bcl-X-L and neutralize their anti-cell death activities in mammalian cells. Interestingly, Bad appears to be relatively specific for Bcl-X-L in that it coimmunoprecipitates with Bcl-X-L much more efficiently than with Bcl-2 and also is considerably more effective at negating Bcl-X-L function than Bcl-2 [168]. 7.7.2 Other Bcl-2-Binding Proteins At least six other proteins have been reported that can bind either directly or indirectly to Bcl-2. These include BAG-1, R-Ras, Raf-1, Nip-1 Nip-2 and Nip-3 [129, 185–187]. At present, it remains unknown whether the interaction of Bcl-2 with any of these proteins is essential for its function as a blocker of apoptosis. BAG-1 The BAG-1 protein was discovered using an interaction cloning technique where recombinant Bcl-2 protein was overlaid onto -gt11 expression cDNA libraries [185]. The BAG-1 protein is 219 amino-acids in length in mice and is acidic in nature (pI 4.18). A domain within BAG-1 has as much as 50% amino-acid sequence identity with some ubiquitin and ubiquitin-like proteins, raising the possibility of a connection to protease pathways for protein degradation. Downstream of this ubiquitin-like domain is a region that, based on computer predictions, may assume an mostly -helical conformation with some of the helices being amphipathic and thus good candidates for participation in binding with other protein via coiled-coil type interactions. Otherwise, however, the predicted primary sequence of the BAG-1 protein reveals no clues as to its potential biochemical activities. In gene transfer studies, BAG-1 was shown to have anti-cell death activity, thus its name, ‘Bcl-2associated AthanoGene-T (BAG-1). In Balb/c-3T3 fibroblasts, gene transfer-mediated elevations in BAG-1 protein levels were associated with prolonged survival in the setting of treatment with staurosporine, a general kinase inhibitor that is a potent inducer of apoptosis in fibroblasts and many other types of cells. More important, co-transfection of BAG-1 and Bcl-2 into Jurkat T-cells provided strong protection from cell death induced by anti-Fas antibody, cytolytic T-cells (CTLs), as well as staurosporine. In contrast, expression of Bcl-2 alone was only partially effective at blocking cell death and BAG-1 by itself afforded little protection. Thus, BAG-1 can functionally cooperate with Bcl-2, resulting in markedly more efficient suppression of cell death induced by anti-Fas antibodies and CTLs than either Bcl-2 or BAG-1 alone. The significance of these observations lies in the controversy over whether Fas and CTLs induce cell death through Bcl-2-dependent vs. -independent mechanisms [82, 125, 127, 128, 130]. Previously, some investigators had speculated that Fas and CTLs may activate the cell death pathway through a Bcl-2-
132
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
independent mechanism, because gene transfer-mediated elevations in Bcl-2 protein levels were by themselves insufficient to provide protection [82, 127, 130]. The BAG-1 findings demonstrate that these stimuli do indeed kill cells through a Bcl-2-dependent pathway, but indicate that for Bcl-2 to block cell death under these circumstances adequate levels of an additional partner protein (BAG-1) must be maintained. It remains to be determined whether the combination of elevated levels of BAG-1 and Bcl-2 can also abrogate the apoptotic effects of other types of stimuli that have been reported to involve Bcl-2independent mechanisms, such as crosslinking of surface receptors for antigen on lymphoid precursors [148–150]. In this regard, the observation that Bcl-X-L protects WEHI231 B-cell lymphoma cells from antiIg induced apoptosis, whereas Bcl-2 fails to do so [149], begs the question of whether there may exist homologs of BAG-1 that preferentially interact with Bcl-X-L as opposed to Bcl-2. Alternatively, such results could potentially be explained by high levels of a Bcl-2-specific inhibitor, possibly a protein that functions in a converse manner to the Bad protein which is relatively specific for Bcl-X-L. R-Ras and Raf-1 Two potential signal transducing proteins have been identified that interact directly or indirectly with Bcl-2: the GTPase R-Ras and the serine/threonine-kinase Raf-1 [185, 186]. R-Ras is a 23-kDa member of the Ras family of small molecular weight GTPases. Mutant versions of R-Ras that constitutively bind GTP and that therefore are chronically in an active conformation (such as a mutant with glycine to valine substitution at position 38) induce anchorage-independent growth in NIH-3T3 cells and render these cells highly tumorigenic in nude mice, but unlike their p21-Ha-Ras counterparts fail to induce morphological transformation [188]. Like Ha-Ras, the R-Ras protein binds to Raf-1 kinase in a GTP-dependent fashion via its effector domain [190, 191]. R-Ras was identified as a Bcl-2 interacting protein during yeast two-hybrid screening of cDNA libraries [185]. This protein was also reported to coimmunoprecipitate with Bcl-2 from bcl-2-transfected HeLa cells. However, Bcl-2 and R-Ras could not be co-immunoprecipitated from 32D hemopoietic cells in which both the Bcl-2 and R-Ras proteins were over-expressed by gene cotransfection, nor from Sf9 cells co-infected with recombinant Bcl-2 and R-Ras baculoviruses [189]. Under these same conditions, however, Bcl-2 could be readily co-immunoprecipitated with Bax, and R-Ras co-immunoprecipitated with Raf-1 kinase. Thus, if R-Ras does associate with Bcl-2, it presumably does so with lower affinity, lower stoichiometry, or more transiently than some other proteins. In both 32D cells and NIH-3T3 cells, R-Ras(V38) significantly accelerates the rate of apoptotic cell death caused by growth factor withdrawal [189]. Furthermore, co-expression of Bcl-2 completely nullifies this effect of R-Ras. Bcl-2 protein however had no effect on R-Ras GTPase activity in vitro, suggesting that it does not function as a GTPase-activating protein (GAP) for R-Ras. Because no GDP exchange proteins for R-Ras have been discovered thus far, it has not been possible to test the idea that Bcl-2 might oppose R-Ras by blocking its loading with GTP. Thus, it remains possible that Bcl-2 acts as an inhibitor of R-Ras exchange proteins. However, examination of the ratio of GTP/GDP associated with R-Ras in 32D cells demonstrated that Bcl-2 does not appear to alter guanine nucleotide binding by R-Ras in cells. Furthermore, the observation that Bcl-2 blocks the effects on cell death of even the constitutively active R-Ras (V38) protein (GTP/GDP+GTP was >85% in 32D cells) argues that Bcl-2 functions downstream of R-Ras. One idea then is that Bcl-2 some how interferes with the interaction of R-Ras with an effector protein, such as Raf-1. In this regard, it was shown that Bcl-2 does not block the ability of R-Ras(V38) to bind to and induce activation of Raf-1 kinase in Sf9 cells [189]. However, at least three other effector proteins, in addition to Raf-1, have been described for Ha-Ras. Consequently, the notion of Bcl-2 preventing R-Ras from
BCL-2 IN CANCER
133
interacting with a downstream effector remains possible, but Raf-1 kinase would appear not to represent that effector protein. It is also unlikely that R-Ras accelerates rates of cell death in the setting of growth factor withdrawal by using Bax as a downstream effector, since Bax could not be co-immunoprecipitated with R-Ras or R-Ras(V38) in transfected 32D cells that contained high levels of these proteins [189]. Interestingly, though it has been difficult in some types of cells to coimmunoprecipitate R-Ras and Bcl-2, this is not the case for Raf-1 and Bcl-2. In both transfected 32D cells that co-expressed Bcl-2 and a truncated version of Raf-1 that retained essentially only the catalytic domain and in Sf9 insect cells coinfected with Bcl-2 and Raf-1 baculoviruses, Bcl-2 and Raf-1 were reported to coimmunoprecipitate with reasonably high stoichiometry (5–30%) under the same conditions where no detectable association of R-Ras and Bcl-2 could be found. In addition to a potential physical interaction, Bcl-2 and Raf-1 have been shown to functionally interact in that co-expression of Bcl-2 and a constitutively active version of Raf-1 in 32D cells resulted in synergistic prolongation of survival in the absence of lymphokines, compared to cells transfected with either Bcl-2 or Raf-1 alone [186]. The ability of Raf-1 and Bcl-2 to co-immunoprecipitate however does not necessarily imply direct binding of these proteins, since no interaction was detected in yeast two-hybrid experiments. Mapping studies indicate that the C-terminal half of the Raf-1 kinase where the catalytic domain resides is sufficient for co-immunoprecipitation with Bcl-2. In contrast, sequences located in the NH2-terminal end of Raf-1 are directly involved in binding to Ras proteins [186, 190, 191]. Though mapping to the catalytic domain, kinase activity appears not to be necessary for Raf-1 coimmunoprecipitation with Bcl-2, based on experiments performed using a point-mutant form of Raf-1 with a disrupted ATP-binding site [168]. The observations showing that Raf-1 can cooperate with Bcl-2 to protect cells from apoptosis seem paradoxical when one considers that R-Ras accelerates cell death and yet can activate Raf-1. The apparent solution to this dilemma however lies in experimental evidence which suggests that Raf-1 enters into separate independent complexes with R-Ras and Bcl-2, such that very little (<1%) of the Bcl-2 or R-Ras in cells can be found in a 3-way complex that simultaneously involves Bcl-2, R-Ras, and Raf-1. Thus, Bcl-2/ Raf-1 and R-Ras/Raf-1 represent largely independent protein complexes and presumably therefore have distinct functions. Though the intracellular locations of the R-Ras protein have yet to be defined, one further potential implication is that Bcl-2 and R-Ras may also target Raf-1 to different membranes in cells and thus influence the protein substrates with which it in comes into contact. Further work is required however before the significance of these protein-protein interactions for modulation of Bcl-2 function are understood. In this regard, Bcl-2 appears not to be a substrate of Raf-1, but the possibility of phosphorylation of Bax, other Bcl-2 family proteins, or other Bcl-2-interacting proteins has not been explored. Interestingly, antibodies directed against the EBV homolog of Bcl-2, BHRF-1, have been reported to co-immunoprecipitate a poorly characterized serine/threonine-kinase activity [192], but it remains to be determined whether this is Raf-1. Nip-1, Nip-2, Nip-3 Using the E1b-19kDa protein as a bait, cDNAs were cloned that encoded three separate proteins which can bind to both E1b and Bcl-2 [187]. All of the 19-kDa-interacting proteins (Nips) were shown to coimmunoprecipitate with E1b and Bcl-2, but not with mutants of E1b-19 kDa that lack anti-apoptotic function. Immunofluorescence microscopy also suggests that the Nip proteins reside in similar locations in cells as E1b-19 kDa and Bcl-2. The Nip-1 protein is 228 amino-acids in length and has a transmembrane domain located at its C-terminus, similar to the topology of the Bcl-2 and E1b-19 kDa proteins. A region within the Nip-1 protein has homology to the catalytic domain of some phosphodiesterases. Nip-2 is
134
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
predicted to be 315 amino-acids in length and contains potential Ca2+-binding sites as well as a region with homology to non-catalytic portions of the Rho-GAP protein. Though it lacks a hydrophobic anchor, Nip-2 appears to localize mostly to nuclear envelope and ER, similar to E1b-19kDa. Nip-3 is a 194 amino-acid protein with homology to calbindin-D, but some splice variants may lack the calbindin-D domain. Nip-3 contains a transmembrane domain near its C-terminus. Of the Nip proteins, the intracellular distribution of Nip-3 is most reminiscent of the mitochondrial pattern seen for Bcl-2 but changes to a predominantly nuclear membrane and perinuclear membrane pattern upon transfection of cells with E1b-19 kDa. All of the Nip proteins contain PEST sequences, implying that their levels are regulated by protein degradation mechanisms. The function of the Nip proteins has yet to be reported.
7.8 Expression of Bcl-2 Family Proteins in Normal Tissues and Cancers The in vivo patterns of Bcl-2, Bax, Bcl-X, and Mcl-1 protein production have been determined by immunohistochemical means in most normal adult tissues in either humans, mice, or both [64, 176, 177, 20 3, 204]. In addition, the developmental patterns of bcl-2 expression have been examined in fetal tissues from humans and mice, revealing that bcl-2 expression is more widespread in the embryo than in adult tissues [205–207]. From these results, a few general conclusions can be reached. First, expression of all of these bcl-2 family genes occurs in a wide variety of adult tissues in vivo. Second, the regulation of the expression of bcl-2, bax, bcl-X, and mcl-1 is highly tissue-specific, and moreover, varies depending on the stage of differentiation or activation of the cells in many cases. Third, in some instances the patterns of expression of these genes correlate with the in vivo regulation of programmed cell death in a way that coincides with the known function of the proteins. In other cases, however, there is no obvious correlation or even an inverse correlation with the known function of some Bcl-2-like proteins and cell death regulation in vivo. In these cases, however, the issue of protein-protein interactions must be taken into consideration, since various homoand heterotypic interactions among these proteins modulate their functions. Fourth, the highly tissuespecific regulation of these genes clearly dictates which of the various members of this multigene family will be expressed in any particular cell at any given time, thus influencing the repertoire of Bcl-2-like proteins that are available to participate in homo- and heterotypic dimer formation. A few examples of the tissue-specific regulation of bcl-2 family gene expression are particularly noteworthy. As mentioned above, for instance, bcl-2 expression in the colonic epithelium occurs in a gradient with high levels of Bcl-2 protein residing in the stem cells located in the base of the crypts with progressively lower levels of Bcl-2 protein present as the cells advance up the crypts to the lumenal surface where they die by programmed cell death [64, 176]. In contrast, Bax protein is present in a gradient with an opposite orientation, such that higher levels of Bax immunoreactivity are seen at the tops of the crypts than in the cells located in the bases [174]. Interestingly, despite having at least modest anti-cell death activity, the pattern of Mcl-1 immunoreactivity in the colonic mucosa is similar to Bax [204]. In fact, the gradients of Bcl-2 and Mcl-1 in several epithelia run in opposing directions, and in general, Bcl-2 tends to be expressed more in self-renewing, long-lived stem cell populations whereas Mcl-1 tends to be highest in more differentiated epithelial cells. It may be relevant therefore that Mcl-1 appears to be less potent than Bcl-2 as a cell death blocker [193, 194], given that the differentiated cells of these complex epithelia undergo programmed cell death as part of the normal homeostatic mechanisms that counterbalance cell production with cell loss in essentially all adult tissues with self-renewal capacity.
BCL-2 IN CANCER
135
Another potentially important observation from the immunostaining results comes from the analysis of Bax in the central nervous system. Despite their longlived nature, most neurons in the adult brain immunostain strongly for Bax. Given the role of Bax as a promoter of cell death, one implication of this finding is that Bax is presumably actively opposed in neurons by other Bcl-2 family proteins, perhaps Bcl-XL in particular which is expressed at high levels in the adult brain [22, 172]. The high levels of Bax in neurons may also explain why these cells are particularly vulnerable to a variety of cell death stimuli. In this regard, it was reported that Bax immunoreactivity was elevated in neurons that had early degenerative changes consistent with ischemia, raising the possibility of a dynamic regulation of bax gene expression in CNS neurons [176]. Finally, an additional striking example of differential tissue-specific regulation of expression of bcl-2 family genes conies from immunohistochemical analysis of lymph nodes. As discussed above, Bcl-2 protein is normally found at high levels in the long-lived mantle zone lymphocytes that surround the germinal centers of secondary lymphoid follicles. In contrast, very little Bcl-2 immunostaining is typically seen in germinal center B-cells. In contrast, follicular center lymphocytes are strongly Mcl-1 positive, while mantle zone lymphocyte are completely Mcl-1 negative [177]. The implication therefore is that while Bcl-2 and Mcl-1 have been shown to be physically capable of interacting [25], the differential regulation of these members of the bcl-2 gene family precludes such interactions at least in most normal peripheral B-cells. In contrast to Bcl-2 and Mcl-1, Bax immunostaining is found in most lymphocytes in nodes, whereas Bcl-X immunoreactivity was limited to lymphoblastic cells in the interfollicular regions of node (probably activated T-cells) and plasma cells [176, 203]. These immunostaining results derived from lymphoid tissues, therefore, provide another example of how the strikingly different regulation of the expression of bcl-2 related genes ultimately determines which of the various Bcl-2 family proteins are available within any given cell to participate in homo- and heterotypic dimer formation. Though little has been published thus far about the regulation of bax, bcl-x, and other bcl-2 homologs in human cancers, preliminary observations suggest that alterations in the expression of some of these genes may occur in tumors. For example, elevated levels of Bcl-X-L protein have been detected in several cases of acute leukemia in association with transformation to a drug-resistant phenotype and reductions in Bax protein levels have been observed in breast and ovarian cancers as well as in chronic lymphocytic leukemia specimens [our unpublished data]. Further analysis of the expression of these and other bcl-2 family genes however is required to delineate the mechanisms responsible for and the prognostic significance of alterations in the expression of other bcl-2-like genes in human cancers. 7.8.1 Molecular Mechanisms of Bcl-2 Gene Regulation At present, understanding of the mechanisms responsible for regulation of bcl-2 gene expression remains limited. The major promoter of the bcl-2 gene is located ~1.7 kbp upstream of the open reading frame and contains several GC boxes (Spl binding sites) but no TATAA box, typical of ‘housekeeping’ gene promoters [37]. This finding suggests that regulation of bcl-2 gene expression may be modulated in large part by events that occur after transcription initiation. In this regard, a p53-independent represser element has been mapped to the 5 -UTR of the bcl-2 gene which inhibits gene expression only when positioned downstream of promoters, suggesting it may regulate a block to transcript elongation [43, 67]. A p53dependent represser element was also mapped to this same general region, but displayed the ability to inhibit regardless of whether positioned upstream or downstream of heterologous promoters [67]. Discordance between bcl-2 mRNA and Bcl-2 protein levels has also been described, suggesting the potential
136
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
for regulation at the level of mRNA translation or protein degradation [208–210]. For example, bcl-2 mRNA was detected in germinal center lymphocytes by in situ hybridization, whereas Bcl-2 protein was not. In this regard, the long 5 -UTR of the bcl-2 mRNAs contains 11 upstream AUGs that represent good candidates for participation in the modulation of translational efficiency. A wide variety of receptors for cytokines and lymphokines, as well as phorbol esters and some retinoids have been reported to either positively or negatively regulate signal transduction pathways that control bcl-2 gene expression, including IL-2, IL-3, IL-4, IL-6, IL-10, TGF- TNF, kit-ligand and CD40 [38, 47, 76, 136, 211–215]. For the most part, however, few molecular details are available as to how these signal transducing proteins influence the expression of the bcl-2 gene. 7.8.2 Tumor Suppressor p53 is a Direct Transcriptional Regulator of bax Gene Expression Besides bcl-2, little if anything is known about the mechanisms that control the expression of most other bcl-2 family genes, including bcl-X, mcl-1, bad, and bak. Recently, however, the promoter region of the human bax gene has been cloned, revealing the presence of four sites with homology to the consensus sequence for p53-binding located upstream of a TATAA box and the transcription start site [195]. In reporter gene assays, p53 was reported to strongly trans-activate the bax promoter. The p53 responsive element was mapped to a 39 bp region that included the four p53-binding site consensus sequences. Furthermore, p53 protein was demonstrated by gel-retardation assays to be capable of binding in a specific fashion to double-strand oligonucleotides containing this 39 bp sequence derived from the bax promoter. In addition, in a p53-deficient myeloblastic leukemia line, a temperature-sensitive version of p53 was shown to induce marked elevations in bax mRNA levels and Bax protein in a conditional fashion [66]. In these same experiments, p53 inhibited expression of bcl-2, as discussed above [66]. Finally, in p53 knock-out mice, reductions in steady-state levels of Bax protein were detected in some tissues by immunoblotting or immunohistochemical assays [66]. Taken together, these findings suggest a potential mechanism by which p53 may control the sensitivity of cells to apoptosis induced by radiation, DNA-damaging chemotherapeutic drugs, and even growth factor deprivation [91, 96, 196]. It has been shown for example that DNA damage induced by drugs or radiation results in elevations in p53 protein and p53 transcriptional activity [68, 69]. As depicted in Figure 7.4, therefore, these elevations in p53 would be expected to induce an increase in Bax protein production and simultaneously a reduction in Bcl-2 protein synthesis. In doing so, p53 would shift the Bcl-2/Bax ratio into a condition of Bax excess and thus place the cell at increased risk of apoptosis. Several caveats about the model proposed in Figure 7.4 for p53-mediated control of sensitivity to apoptosis however deserve discussion. First, the transcriptional and post-transcriptional inputs into the bax and bcl-2 gene are likely to be complex. Therefore, while p53 is one potential regulator, clearly other transacting factors also exist that make contributions to the final net output from the bax and bcl-2 genes. One question that must be pursued for the future therefore is the determination of in which tissues p53 represents a dominant player as a regulator of bcl-2 and bax, and in which types of cancer does loss of p53 lead to elevations in Bcl-2, decreases in Bax, or both. A second caveat is that these observed effects of p53 on bcl-2 and bax gene expression, while correlating with the apoptotic behavior of cells, do not necessarily explain the mechanism by which p53 increases the sensitivity of cells to apoptotic stimuli. Certainly, the gene transfer experiments showing that enforced production of high levels of Bcl-2 protein can partially or completely block p53-induced apoptosis in leukemia and lymphoma cell lines argue in favor of this hypothesis [101–104, 160] but do not prove it. In this regard, it has also been reported that p53 can induce
BCL-2 IN CANCER
137
Figure 7.4 Regulation of bcl-2 and bax gene expression by p53 may contribute to mechanisms of p53-induced apoptosis. In the model depcited, DNA damage caused for example by chemotherapeutic drugs or radiation induces increases in p53 protein and p53 transcriptional activity. The p53 protein then directly upregulates bax gene expression and also downregulates bcl-2 gene expression. Consequently, the ratio of Bcl-2/Bax protein levels decreases and cells are therefore rendered more sensitive to cell death stimuli. The model does not depict the many other effects that p53 may have on cell and apoptotic pathways that could modify the schema shown, nor does it imply that p53 is the predominant regulator of bcl-2 and bax gene in all types of cells.
138
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
apoptosis even in the presence of cycloheximide or actinomycin-D, at least in a large-T antigen-induced pituitary tumor line when subjected to UV-radiation [100]. In some cells, therefore, it would appear that p53-mediated elevations in bax gene expression are not necessarily required for p53-induced apoptosis. These findings however do not exclude a potentially important functional role for p53-induced decreases in bcl-2 gene expression. Furthermore, in some types of cells, the steady-state levels of Bax protein may already be sufficiently high that a further p53-mediated increase is not required. In this regard, it has been reported that enforced production of high levels of Bcl-2 does not block the ability of p53 to trans-activate reporter gene constructs that contain typical p53 binding sites, but does interfere with trans-repression mediated by p53 [198]. In fact, when co-transfected with expression plasmids producing either Bcl-2 or E1b-19 kDa, p53 paradoxically acted as an upregulator of the expression of some reporter gene constructs that are normally repressed by p53. Finally, p53 does not induce apoptosis in all types of cells in which it can nevertheless induce cell cycle arrest. This is particularly true when comparing normal and transformed cells, where the latter are often markedly more sensitive to p53-dependent apoptosis induced by radiation, DNA-damaging drugs and growth factor deprivation [94, 96, 197]. It will be of interest in these circumstances to compare the effects of p53 on the expression of bax and bcl-2 in search of correlations with susceptibility to apoptosis. In this regard, -irradiation has been shown to induce rapid increases in Bax protein in radiosensitive tissues but not in radioresistant tissues [194]. Though speculative, the connection between p53 and regulation of bax gene expression may provide insights into the clinical behavior of some types of cancer. For example, the ability of p53 to induce expression of bax may explain the somewhat paradoxical observation that patients with follicular lymphomas typically respond well to therapy at least initially, despite the high levels of Bcl-2 in these neoplasms caused by t(14;18) translocations. In this regard, though relapse occurs almost invariably, many patients can be induced into partial or complete clinical remissions a few times, often by use of the same drug regimen. Eventually, however, most patients experience transformation of their disease to an unresponsive state. The ability of follicular lymphoma patients to respond initially could theoretically be explained by induction of increases in p53 protein levels and transcriptional activity as a result of druginduced damage to DNA. Eventually, this p53-Bax pathway for induction of apoptosis may fail due to loss of p53 expression or function, mutations in the bax gene, or other mechanisms resulting in an nonresponsive state where the ratio of Bcl-2 to Bax remains high and the tumor cells thus fail to undergo apoptotic cell death. Consistent with this idea, p53 gene mutations have previously been associated with histological transformation of follicular lymphomas [199, 200]. The finding that p53 can be a direct transcriptional activator of bax gene expression, as well as an inhibitor of bcl-2, may have important implications for diseases beyond cancer. For example, induction of p53 expression in association with apoptosis of neurons has been reported in an animal model that was designed to emulate some of the biochemical events that occur during strokes [201]. In this regard, hypoxia has been reported to be among the inducers of increases in p53 activity in cells [202]. Marked increases in Bax protein have also been reported in ischemia damaged neurons in a model of transient global ischemia in the rat [193]. Thus, it will be of interest to explore whether direct functional connections exist between p53, bax, bcl-2 and the regulation of apoptosis in the nervous system, and perhaps in other scenarios such as ischemic myocardium.
BCL-2 IN CANCER
139
7.9 Conclusions The discovery of bcl-2 and its homologs has revealed a critical point in the regulation of the physiological cell death pathway. Alterations in the expression of bcl-2 have been found in the majority of human cancers. Moreover, Bcl-2 and its homologs contribute significantly to the regulation of the relative resistance of tumor cells to apoptosis induced by radiation and essentially all chemotherapeutic drugs. The hope is that with improved understanding of the mechanisms that regulate the expression of bcl-2 family genes in normal and malignant cells, as well as with advances in our knowledge of the molecular details of how Bcl-2 and its various interacting proteins biochemically function, it will eventually become possible to develop novel approaches to the treatment of cancer and other diseases that involve the programmed cell death pathway. Acknowledgements I thank all the members of my laboratory whose untiring dedication led to much of the data described in this review, Cecilia Stephens for manuscript preparation, Christine Aimé-Sempé for Figure 2 and the National Cancer Institute, American Cancer Society, and Leukemia Society of America for support. References 1 2 3 4
5 6 7
8
9
10 11
. TSUJIMOTO, Y., COSSMAN, J., JAFFE, E. & CROCE, C. (1985) Involvement of the bcl-2 gene in human follicular lymphoma. Science 228, 1440–1443. . TSUJIMOTO, Y., COSSMAN, J.A.-J & CROCE, C. (1985) Involvement of the bcl-2 gene in human follicular lymphoma. Science 228, 1440–1443. . CLEARY, M.L., SMITH, S.D. & SKLAR, J. (1986) Cloning and structural analysis of cDNAs for bcl-2 and a hybrid bcl-2/immunoglobulin transcript resulting from the t(14;18) translocation. Cell 47, 19–28. . BAKHSHI, A., JENSEN, J.P., GOLDMAN, P., WRIGHT, J.J., MCBRIDE, O.W., EPSTEIN, A.L. & KORSMEYER, S.J. (1985) Cloning the chromosomal breakpoint of t(14;18) human lymphomas: clustering around JH on chromosome 14 and near a transcriptional unit on 18. Cell 41, 899–906. . VAUX, D.L., CORY, S. & ADAMS, J.M. (1988) Bcl-2 gene promotes haemopoietic cell survival and cooperates with c-myc to immortalize pre-B-cells. Nature 335, 440–442. . HOCKENBERY, D., NUNEZ,G., MlLLIMAN, C., SCHREIBER, R.D. & KORSMEYER, S.J. (1990) Bcl-2 is an inner mitochondrial membrane protein that blocks programmed cell death. Nature 348, 334–336. . MCDONNELL, T.J., DEANE, N., PLATT, F.M., NUMEZ, G., JAEGER, ., MCKEARN, J.P. & KORSMEYER, S.J. (1989) Bcl-2-immunoglobulin transgenic mice demonstrate extended B-cell survival and follicular lymphoproliferation. Cell 57, 79–88. . KATSUMATA, M., SIEGEL, R., LOUIE, D., MIYASHITA, T., TANAKA, S., GREENE, M. & REED, J.C. (1992) Differential effects of bcl-2 on B- and T-lymphocytes in transgenic mice Proc. Natl Acad. Sci. USA 89, 11376–11380. . MIYASHITA, T., & REED, J.C. (1992) BCL-2 gene transfer increases relative resistance of S49.1 and WEH17. 2 lymphoid cells to cell death and DNA fragmentation induced by glucocorticoids and multiple chemotherapeutic drugs. Cancer Res. 52, 5407– 5411. . MIYASHITA, T. & REED, J.C. (1993) Bcl-2 oncoprotein blocks chemotherapy-induced apoptosis in a human leukemia cell line. Blood 81, 151–157. . COLLINS, M.K.L., MARVEL, J., MALDE, P. & LOPEZ-RIVAS, A. (1992) Interleukin-3 protects murine bone marrow cells from apoptosis induced by DNA damaging agents. J. Exp. Med. 176, 1043–1051.
140
12
13 14
15
16
17 18
19
20
21 22
23
24 25
26 27 28 29 30
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. GRIFFITHS, S.D., GOODHEAD, D.T., MARSDEN, S.J., WRIGHT, E.G., KRAJEWSKI, S., REED, J.C., KORSMEYER, S.J. & GREAVES, M. (1994) Interleukin 7-dependent B lymphocyte precursor cells are ultrasensitive to apoptosis. J. Exp. Med. 179, 1789–1797. . YUNIS, J.J., MAYER, M.G., ARENSEN, M.A., AEPPLI, D.P., OKEN, M.M. & FRIZZERA, G. (1991) Bcl-2 and other genomic alterations in the prognosis of large-cell lymphomas. N. Engl. J. Med. 320, 1047–1054. . OFFIT, K., KODURU, P.R.K., HOLLIS, R., FILIPPA, D., JHANWAR, S.C., CLARKSON, B.C. & CHAGANTI, R.S.K. (1989) 18q21 rearrangement in diffuse large cell lymphoma: incidence and clinical significance. Br. J. Haematol. 72, 178–183. . MCDONNELL, T.J., TRONCOSO, P., BRISBAY, S.M., LOGOTHETIS, C., CHUNG, L.W.K., HSIEH, J.-T., TU, S.-M. & CAMPBELL, M.L. (1992) Expression of the proto-oncogene bcl-2 in the prostate and its association with emergence of androgen-independent prostate cancer. Cancer Res. 52, 6940–6944. . CAMPOS, L., ROUALULT, J.-P., SABIDO, O., ROUBI, N., VASSELON, C., ARCHIMBAUD, E., MAGAUD, J.-P. & GUYOTAT, D. (1993) High expression of bcl-2 protein in acute myeloid leukemia cells in associated with poor response to chemotherapy. Blood 81, 3091–3096. . HENGARTNER, M.O. & HORVITZ, H.R. (1994) C. elegans cell survival gene ced-9 encodes a functional homolog of the mammlian proto-oncogene bcl-2. Cell 76, 665–676. . HENDERSON, S., HUEN, D., ROWE, M., DAWSON, C., JOHNSON, G. & RICKINSON, A. (1993) EpsteinBarr virus-coded BHRF1 protein, a viral homologue of bcl-2, protects human B-cells from programmed cell death . Proc. Natl. Acad. Sci. USA 90, 8479–8483. . TAKAYAMA, S., CAZALS-HATEM, D.L., KITADA, S., TANAKA, S., MIYASHITA, T., HOVEY, L.R., HUEN, D., RICKINSON, A., VEERAPANDIAN, P., KRAJEWSKI, S., SAITO, K. & REED, J.C. (1994) Evolutinary conservation of function among mammalian, avian, and viral homologs of the bcl-2 oncoprotein: structure-function implications. DNA Cell. Biol. 13, 679–692. . WHITE, E., SABBATINI, P., DEBBAS, M., WOLD, W.S.M., KUSHER, D.I. & GOODING, L.R. (1992) The 19-kilodalton adenovirus E1B transforming protein inhibits programmed cell death and prevents cytolysis by tumor necrosis factor- . Mol. Cell. Biol. 12, 2570–2580. . OLTVAI, Z., MILLIMAN, C. & KORSMEYER, S.J. (1993) Bcl-2 heterodimerizes in vivo with a conserved homology, Bax, that accelerates programmed cell death. Cell 74, 609–619. . BOISE, L.H., GONZALEZ-GARCIA, M., POSTEMA, C.E., DING, L., LINDSTEN, T., TURKA, L.A., MAO, X., NUNEZ, G. & THOMPSON, C.B. (1993) bcl-x, a bcl-2-related gene that functions as a dominant regulator of apoptotic cell death. Cell 74, 597–608. . KOZOPAS, K.M., YANG, T., BUCHAN, H.L., ZHOU, P. & CRAIG, R. (1993) Mcl-1, a gene expressed in programmed myeloid cell differentiation, has sequence similarity to bcl-2. Proc. Natl Acad. Sci. USA 90, 3516–3520. . LIN, E.Y., ORLOFSKY, A., BERGER, M.S. & PRYSTOWSKY, M.B. (1993) Characterization of A1, a novel hemopoietic-specific early-response gene with sequence similarity to bcl-2. J. Immunol. 151 1979–1988. . SATO, T., HANADA, M., BODRUG, S., IRIE, S., IWAMA, N., BOISE, L.H., THOMPSON, C.B., GOLEMIS, E., FONG, L., WANG, H.-G. & REED, J.C. (1994) Interactions among members of the bcl-2 protein family analyzed with a yeast two-hybrid system. Proc. Natl Acad. Sci. USA 91, 9238–9242. . CROCE, C.M., TSUJIMOTO, Y., ERIKSON, J. & NOWELL, P. (1984) Chromosome translocations and B-cell neoplasia. Lab. Invest. 51, 258–267. . YUNIS, J.J., FRIZZERA, G., OKEN, M.M., MCKENNA, J., THEOLOGIDES, A. & ARNESON, M. (1987) Multiple recurrent genomic defects in follicular lymphoma. New Engl. J. Med. 316, 79–84. . FUKUHARA, S., ROWLEY, J.D., VARIAKOJIS, D. & GOLOMB, H.M. (1979) Chromosome abnormalities in poorly differentiated lymphocytic lymphoma. Cancer Res. 39, 3119–3128. . EICK, S., KRIEGER, G., BOLZ, I. & KNEBA, M. (1990) Sequence analysis of amplified t(14;18) chromosomal breakpoints in B-cell lymphomas. J. Pathology. 162, 127–133. . WEISS, L.M., WARNKE, R.A., SKLAR, J. & CLEARY, M.L. (1987) Molecular analysis of the t(14;18) chromosomal translocation in malignant lymphomas. N. Engl. J. Med. 317, 1185–1189.
BCL-2 IN CANCER
31
32
33
34 35
36
37
38 39 40 41
42 43 44
45 46 47
48
141
. TSUJIMOTO, Y., LOUIE, E., BASHIR, M.M. & CROCE, C.M. (1988) The reciprocal partners of both the t(14; 18) and t(11;14) translocations involved in B-cell neoplasms are rearranged by the same mechanism. Oncogene 2, 347–351. . JAEGER, U., PUTSCHER, B., KARTH, G.D., KNAPP, S., MANNHALTER, C. & LECHNER, K. (1993) Mechanism of the chromosomal translocation t(14;18) in lymphoma: detection of a 45-Kd breakpoint binding protein. Blood 81, 1833–1840. . GRIBBEN, J.G., FREEDMAN, A.S., WOO, S.D., BLAKE, K., SHU, R.S., FREEMAN, G., LONGTINE, J.A., PINKUS, G.S. & NADLER, L.M. (1991) All advanced stage non-Hodgkin’s lymphomas with a polymerase chain reaction amplifiable breakpoint of bcl-2 have residual cells containing the bcl-2 rearrangement at evaluation and after treatment. Blood 78, 3275–3280. . LEE, M.-S., CHANG, K.-S., CABANILLAS, F., FREIREICH, E.J., TRUJILLO, J.M. & STASS, S.A. (1987) Detection of minimal residual cells carrying the t(14;18) by DNA sequence amplification. Science 237, 175–178. . GRIBBEN, J.G., NEUBERG, D., FREEDMAN, A.S., GIMMI, C.D., PESEK, K.W., BARBER, M., SAPORITO, L., WOO, S.D., CORAL, F., SPECTOR, N., RABINOWE, S.N., GROSSBARD, M.L., RITZ, J. & NADLER, L.M. (1993) Detection by polymerase chain reaction of residual cells with the bcl-2 translocation is associated with increased risk of relapse after autologous bone marrow transplantation for B-cell lymphoma. Blood 81, 3449–3457. . REED, J., TSUJIMOTO, Y., EPSTEIN, S., CUDDY, M., SLABIAK, T., NOWELL, P. & CROCE, C. (1989) Regulation of bcl-2 gene expression in lymphoid cell lines containing t(14;18) or normal #18 chromosomes. Oncogene Res. 4, 271–282. . SETO, M., JAEGER, U., HOCKETT, R.D., GRANINGER, W., BENNETT, S., GOLDMAN, P. & KORSMEYER, S.J. (1988) Alternative promoters and exons, somatic mutation and deregulation of the bcl-2-Ig fusion gene in lymphoma. EMBO J. 7, 123– 131. . REED, J.C., TSUJIMOTO, Y., ALPERS, J.D., CROCE, C.M. & NOWELL, P.C. (1988) Regulation of bcl-2 proto-oncogene expression during normal human lymphocyte proliferation. Science 236, 1295–1299. . REED, J., HALDAR, S., CUDDY, M., CROCE, C. & MAKOVER, D. (1988) Deregulated bcl-2 expression enhances growth of a human B-cell line. Oncogene 4, 1123–1127. . TSUJIMOTO, Y. (1989) Overexpression of the human bcl-2 gene product results in growth enhancement of Epstein-Barr virus-immortalized B-cells. Proc. Natl Acad. Sci. USA 86, 1958–1962. . NUÑEZ, G., SETO, M., SEREMETIS, S., FERRERO, D., GRIGNANI, F., KORSMEYER, S.J. & DALLAFAVERA, R. (1989) Growth- and tumor-promoting effects of deregulated Bcl-2 in human B-lymphoblastoid cells. Proc. Natl Acad. Sci. USA 86, 4589–4593. . HANADA, M., DELIA, D., AIELLO, A., STADTMAUER, E. & REED, J.C. (1993) bcl-2 gene hypomethylation and high-level expression in B-cell chronic lymphocytic leukemia. Blood 82, 1820–1828. . YOUNG, R.L. & KORSMEYER, S.J. (1993) A negative regulatory element in the bcl-2 5 -untranslated region inhibits expression from an upstream promoter. Mol. Cell Biol. 13, 3686–3697. . LIU, Y.-J., CAIRNS, J.A., HOLDER, M.J., ABBOT, S.D., JANSEN, K.U., BONNEFOY, J.-Y., GORDON, J. & MACLENNAN, I.C.M. (1991) Recombinant 25-kDa CD23 and interleukin 1a promote the survival of germinal center B-cells: evidence for bifurcation in the development of centrocytes resuced from apoptosis. Eur. J. Immunol. 21, 1107–1114. . MACLENNAN, I.C.M. & GRAY, D. (1986) Antigen-driven selection of virgin and memory B-cells. Immunol. Rev. 91, 61–66. . SPRENT, J. (1994) T and B memory cells. Cell 76, 315–322. . LIU, Y.-J., MASON, D.Y., JOHNSON, G.D., ABBOT, S., GREGORY, C.D., HARDIE, D.L., GORDON, J. & MACLENNAN, I.C.M. (1991) Germinal center cells express bcl-2 protein after activation by signals which prevent their entry into apoptosis. Eur. J. Immunol. 21, 1905–1910. . PEZZELLA, F., TSE, A.G.D., CORDELL, J.L., PULFORD, K.A.F., GATTER, K.C. & MASON, D.Y. (1990) Expression of the bcl-2 oncogene protein is not specific for the 14;18 chromosomal translocation. Amer J. Pathology 137, 225–232.
142
49 50
51 51b 52
53 54
55 56 57
58
59
60
61
62 63 64 65 66
67
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. HORNING, S.J.& ROSENBERG, S.A. (1984) The natural history of initially untreated low-grade non-Hodgkin’s lymphomas. N. Engl J. Med. 311, 1471–1475. . MCDONNELL, T.J., NUÑEZ, G., PLATT, F.M., HOCKENBERY, D., LONDON, L., MCKEARN, J.P. & KORSMEYER, S.J. (1990) Deregulatedbcl-2-immunoglobulin transgene expands a resting but responsive immunoglobulin M and D-expressing B-cell population. Mol. Cell. Biol. 10, 1901–1907. . GAUWERKY, C.E., HOXIE, J., NOWELL, P.C. & CROCE, C.M. (1988) Pre-B-cell leukemia with a t(8;14) and a t(14;18) translocation is preceded by follicular lymphoma. Oncogene 2, 431–435. . LEE, J.T., INNES JR., D.J. & WILLIAMS, M.E. (1989) Sequential bcl-2 and c-myc oncogene rearrangements associated with the clinical transformation of non-Hodgkin’s lymphoma. J. Clin. Invest. 84, 1454–1459. . DE JONG, D., VOETDIJK, B.M.H., BEVERSTOCK G.C., VAN OMMEN, G.J.B., WILLEMZE, R. & KLUIN, P.M. (1988) Activation of the c-myc oncogene in a precursor-B-cell blast crisis of follicular lymphoma, presenting as composite lymphoma. N. Engl. J. Med. 318, 1373–1378. . MCDONNELL, T.J. & KORSMEYER, S.J. (1991) Progression from lymphoid hyperplasia to high-grade malignant lymphoma in mice transgenic for the t(14;18). Nature 349, 254–256. . REED, J.C., MEISTER, L., TANAKA, S., CUDDY, M., YUM, S., GEYER, C. & PLEASURE, D. (1991) Differential expression of bcl-2 proto-oncogene in neuroblastoma and other human tumor cell lines of neural origin. Cancer Res. 51, 6529. . LAUWERS, G.Y., SCOTT, G.V. & HENDRICKS, J. (1994) Immunohistochemical evidence of aberrant bcl-2 protein expression in gastric epithelial dysplasia. Cancer 73, 2900–2904. . BRONNER, M., CULIN, C., REED, J.C. & FURTH, E.E. (1994) Bcl-2 protooncogene and the gastrointestinal epithelial tumor progression model. Amer. J. Pathol. 146, 20–26. . COLOMBEL, M., SYMMANS, F., GIL, S., O’TOOLE, K.M., CHOPIN, D., BENSON, M., OLSSON, C.A., KORSMEYER, S. & BUTTYAN, R. (1993) Detection of the apoptosis-suppressing oncoprotein bcl-2 in hormonerefractory human prostate cancers. Amer. J. Pathol. 143, 390–400. . SILVESTRINI, R., VENERONI, S., DAIDONE, M.G., BENINI, E., BORACCHI, P., MEZZETTI, M., DIFRONZO, G., RILKE, F. & VERONESI, U. (1994) The bcl-2 protein: a prognostic indicator strongly related to p53 protein in lymph node-negative breast cancer patients. J. Natl Cancer. Inst. 86, 499–504. . LEEK, R.D., KAKLAMANIS, L., PEZZELLA, F., GATTER, K.C. & HARRIS, A.L. (1994) bcl-2 in normal human breast and carcinoma, association with oestrogen receptor-positive, epidermal growth factor receptornegative tumours and in situ cancer. Br. J. Cancer 69, 135–139. . PEZZELLA, F., TURLEY, H., KUZU, I., TUNGEKAR, M.F., DUNHILL, M.S., PIERCE, C.B., HARRIS, A., GATTER, K.C. & MASON, D.Y. (1993) bcl-2 protein in non-small-cell lung carcinoma. N. Engl. J. Med. 329, 690–694. . DELIA, D., AIELLO, A., SOLIGO, D., FONTANELLA, E., MELANI, C., PEZZELLA, F., PIEROTTI, M.A. & DELLA PORTA, G. (1992) bcl-2 proto-oncogene expression in normal and neoplastic human myeloid cells. Blood 79, 1291–1298. . LU, Z.-L., ELIA, G., LUCAS, S. & THOMAS, J.A. (1993) Bcl-2 proto-oncogene expression in Epstein-BarrVirus associated nasopharyngeal carcinoma. Int. J. Cancer. 53, 29–35. . IKEGAKI, N., KATSUMATA, M., MINNA, J. & TSUJIMOTO, Y. (1994) Expression of bcl-2 in small cell lung carcinoma cells. Cancer Res. 54, 6–8. . HOCKENBERY, D.M., ZUTTER, M., HICKER, W., NAHM, M. & KORSMEYER, S. J. (1991) Bcl-2 protein is topographically restricted in tissues characterized by apoptotic cell death. Proc. Natl Acad. Sci. USA 88, 6961. . EL-DEIRY, W.S., KERN, S.E., PIETENPOL, J.A., KINZLER, K.W. & VOGELSTEIN, B. (1992) Definition of a consensus binding site for p53. Nature Genetics 1, 45–49. . MIYASHITA, T., KRAJEWSKI, S., KRAJEWSKA, M., WANG, H.G., LIN, H.K., HOFFMAN, B., LIEBERMAN, D. & REED, J.C. (1994) Tumor suppressor p53 is a regulator of bcl-2 and bax in gene expression in vitro and in vivo. Oncogene 9, 1799– 1805. . MIYASHITA, T., HARIGAI, M., HANADA, M. & REED, J.C. (1994) Identification of a p53-dependent negative response element in the bcl-2 gene. Cancer Res. 54, 3131–3135.
BCL-2 IN CANCER
68 69 70 71 72 73
74
75
76
77
78 79
80 81
82 83
84 85
86
143
. ZHAN, Q., CARRIER, F. & FORNANCE, JR., A.J. (1993) Induction of cellular p53 activity by DNAdamaging agents and growth arrest. Mol. Cell. Biol. 13, 4242–4250. . LU, X. & LANE, D.P. (1993) Differential induction of transcriptionally active p53 following UV or ionizing radiation: defects in chromosome instability syndromes? Cell 75, 765–778. . SHABAIK, A.S., KRAJEWSKI, S., BURGAN, A., KRAJEWSKA, M. & REED, J.C. bcl-2 proto-oncogene expression in normal, hyperplastic, and neoplastic prostate tissue. J. Urologic Pathology, in press. . FURUYA, Y., KRAJEWSKI, S., EPSTEIN, J.I., REED, J.C. & ISAACS, J.T. Enhanced expression of Bcl-2 and the progression of human and rodent prostatic cancers. Clinical Cancer Res. (in press). . DOLE, M., NUÑEZ, G., MERCHANT, A.K., MAYBAUM, J., RODE, C.K., BLOCK, C.A. & CASTLE, V.P. (1994) Bcl-2 inhibits chemotherapy-induced apoptosis in neuroblastoma. Cancer Res. 54, 256–260. . EPPERLY, M.W., SANTUCCI, M.A., REED, J.C., SHIELDS, D., HALLORAN, A. & GREENBERGER, J.S. (1994) Expression of the human bcl-2 transgene increases the radiation resistance of a hematopoietic progenitor cell line. Rad. Oncology Invest. 2, 77–83. . FISHER, T.C., MILNER, A.E., GREGORY, C.D., JACKMAN, A.L., AHERNE, G.W. , HARLEY, J.A., DIVE, C. & HICKMAN, J.A. (1993) Bcl-2 modulation of apoptosis induced by anticancer drugs: resistance to thymidylate stress is independent of classical resistance pathways. Cancer Res. 53, 3321–3326. . GRIFFITHS, S.D., GOODHEAD, D.T., MARSDEN, S.J., WRIGHT, E.G., KRAJEWSKI, S., REED, J.C., KORSMEYER, S.J. & GREAVES, M. (1994) Interleukin 7-dependent B lymphocyte precursor cells are ultrasensitive to apoptosis. J. Exp. Med. 179, 1789–1797. . HANADA, M., KRAJEWSKI, S., TANAKA, S., CAZALS-HATEM, D., SPENGLER, B.A., ROSS, R.A., BIEDLER, J.L. & REED, J.C. (1993) Regulation of bcl-2 oncoprotein levels with differentiation of human neuroblastoma cells. Cancer Res. 53, 4978–4986. . KAMESAKI, S., KAMESAKI, H., JORGENSEN, T.J., TANIZAWA, A., POMMIER, Y. & COSSMAN, J. (1993) Bcl-2 protein inhibits etoposide-induced apoptosis through its effects on events subsequent to topoisomerase II-induced DNA strand breaks and their repair . Cancer Res. 53, 4251–4256. . KITADA, S., TAKAYAMA, S., DERIEL, K., TANAKA, S. & REED, J.C. (1994) Reversal of chemoresistance of lymphoma cells by antisense-mediated reduction of bcl-2 gene expression. Antisense Res. Dev. 4, 71–79. . REED, J.C., KITADA, S., TAKAYAMA, S. & MIYASHITA, T. (1994) Regulation of chemoresistance by the bcl-2 oncoprotein in non-Hodgkin’s lymphoma and lymphocytic leukemia cell lines. Annals of Oncology 5 (Suppl. 1) S61–S65. . SENTMAN, C.L., SHUTTER, J.R., HOCKENBERY, D., KANAGAWA, O. & KORSMEYER, S.J. (1991) bcl-2 inhibits multiple forms of apoptosis but not negative selection in thymocytes. Cell 67, 879–888. . SIEGEL, R.M., KATSUMATA, M., MIYASHITA, T., LOUIE, D.C., GREENE, M.I. & REED, J.C. (1992) Inhibition of thymocyte apoptosis and negative antigenic selection in bcl-2 transgenic mice. Proc. Natl Acad. Sci. USA 89, 7003–7007. . STRASSER, A., HARRIS, A.W. & CORY, S. (1991) Bcl-2 transgene inhibits T cell death and perturbs thymic self-censorship. Cell 67, 889–899. . WALTON, W.I., WHYSONG, D., O’CONNOR, P.M., HOCKENBERY, D., KORSMEYER, S.J. & KOHN, K.W. (1993) Constitutive expression of human bcl-2 modulates mitrogen mustard and camptothecin induced apoptosis. Cancer Res. 53 1853–1861. . EASTMAN, A. (1990) Activation of programmed cell death by anticancer agents: cisplatin as a model system. Cancer Cells 2, 275–280. . KAUFMANN, S.H. (1989) Induction of endonucleolytic DNA cleavage in human acute myelogenous leukemia cells by etoposide, campthothecin, and other cytotoxic anticancer drugs: A cautionary note. Cancer Res. 49, 5870–5878. . KANE, D.J., SARAFIN, T.A., AUTON, S., HAHN, H., GRALLA, F.B., VALENTINE, J.C., ORD, T. & BREDESEN, D.E. (1993) Bcl-2 inhibition of neural cell death: decreased generation of reactive oxygen species . Science 262, 1274–1276.
144
87 88
89
90
91 92
93
94 95 96
97
98 99 100 101 102
103
104 105 106
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. HOCKENBERY, D., OLTVAI, Z., YIN, X.-M., MILLIMAN, C. & KORSMEYER, S.J. (1993) Bcl-2 functions in an antioxidant pathway to prevent apoptosis. Cell 75, 241–251. . SUN, X., CARSTENEN, J., ZHANG, M., STAL, O., WINGREN, S., HATSCHEK, T. & NORDENSKJOLD, B. (1992) Prognostic significance of cytoplasmic p53 oncoprotein in colorectal adenocarcinoma. Lancet 340, 1369–1373. . THOR, A.D., MOORE, D.H. II, EDGERTON, S.M., KAWASAKI, E.S., REIHSAUS, E., LYNCH, H.T., MARCUS, J.N., SCHWARTZ, L., CHEN, L.C. & MAYALL, B.H. (1992) Accumulation of p53 tumor suppressor gene protein: an independent marker of prognosis in breast cancers. J. Natl Cancer Inst. 84, 845–855. . VISAKORPI, T., KALLIONIEMI, O., HEIKKINEN, A., KOIVOULA, T. & ISOLA, J. (1992) Small subgroup of aggressive, highly proliferative prostatic carcinomas defined by p53 accumulation. J. Natl Cancer Inst. 84, 883–887. . LOWE, S.W., RULEY, H.E., JACKS, T. & HOUSEMAN, D.E. (1993) p53-dependent apoptosis modulates the cytotoxicity of anticancer agents. Cell 74, 957–967. . MERRITT, A.J., POTTEN, C.S., KEMP, C.J., HICKMAN, J.A., BALMAIN, A., LANE, D.P. & HALL, P.A. (1994) The role of p53 in spontaneous and radiation-induced apoptosis in the gastrointestinal tract of normal and p53-deficient mice. Cancer Res. 54, 614–617. . CLARKE, A.R., GLEDHILL, S., HOOPER, M.L., BIRD, C.C. & WYLLIE, A.H. (1994) p53 dependence of early apoptotic and proiferative responses within the mouse intestinal epithelium following -irradiation. Oncogene 9, 1767–1773. . LOWE, S.W., SCHMITT, E.M., SMITH, S.W., OSBORNE, B.A. & JACKS, T. (1993) p53 is required for radiation-induced apoptosis in mouse thymoctyes. Nature 362, 847–849. . CLARKE, A.R., PURDIE, C.A., HARISON, D.J., MORRIS, R.G., BIRD, C.C., HOOPER, M.L. & WYLLIE, A.H. (1993) Thymocyte apoptosis induced by p53-dependent and independent pathways. Nature 362, 849–852. . BRISTOW, R.G., JANG, A., PEACOCK, J., CHUNG, S., BENCHIMOL, S. & HILL, R.P. (1994) Mutant p53 increases radioresistance in rat embryo fibroblasts simultaneously transfected with HPV 16-E7 and/or activated H-ras. Oncogene 9, 1527–1536. . MERCER, W.E., SHIELDS, M.T., AMIN, M., SAUVE, G.J., APPELLA, E., ROMANO, J.W. & ULLRICH, S.J. (1990) Negative growth regulation in a glioblastoma tumor cell line that conditionally expresses human wildtype p53. Proc. Natl Acad. Sci. USA 87, 6166–6170. . YONISH-ROUACH, E., RESNITZKY, D., LOTEM, J., SACHS, L., KIMCHI, A. & OREN, M. (1991) Wildtype p53 induces apoptosis of myeloid leukaemic cells that is inhibited by interleukin-6. Nature 352, 345–347. . SHAW, P., BOVEY, R., TARDY, S., SAHLI, R., SORDAT, B. & COSTA, J. (1992) Induction of apoptosis by wild-type p53 in a human colon tumor-derived cell line. Proc. Natl Acad. Sci. USA 89, 4495–4499. . CAELLES, C., HEIMBERG, A. & KARIN, M. (1994) p53-dependent apoptosis in the absence of transcriptional activation of p53-target genes. Nature 370, 220–224. . WANG, Y., SZEKELY, L., OKAN, I., KLEIN, G. & WIMAN, K.G. (1993) Wild-type p53 triggered apoptosis is inhibited by bcl-2 in a v-myc-induced T-cell lymphoma line. Oncogene 8, 3427–3431. . RYAN, J.J., PROCHOWNIK, E., GOTTLIEB, C.A., APEL, I.J., MERINO, R., NUÑEZ, G. & CLARKE, M.F. (1994) c-myc and bcl-2 modulates p53 function by altering p53 subcellular trafficking during the cell cycle. Proc. Natl Acad. Sci. USA 91, 5878–5882. . SELVAKUMARAN, M., LIN, H.-K., MIYASHITA, T., WANG, H.-G., KRAJEWSKI, S., REED, J.C., HOFFMAN, B. & LIEBERMANN, D. (1994) Immediate early upregulation of bax expression by p53 but not TGFbl: a paradigm for distinct apoptotic pathways. Oncogene 9, 1791–1798. . STRASSER, A., HARRIS, A.W., JACKS, T. & CORY, S. (1994) DNA damage can induce apoptosis in proliferating lymphoid cells via p53-independent mechanisms inhibitable by bcl-2. Cell 79, 329–339. . OFFIT, K., WONG, G., FILIPPA, D.A., TAO, Y. & CHAGANTI, R.S.K. (1991) Cytogenetic analysis of 434 consecutively ascertained specimens of non-Hodgkin’s lymphoma: clinical correlations. Blood 74, 1508–1515. . PIRIS, M.A., PEZELLA, F., MARTINEZ-MONTERO, J.C., ORRADRE, J.L., VILLUENDAS, R., SANCHEZBEATO, M., CUENA, R., CRUZ, M.A., MARTINEZ, B., GARRIDO, M.C., GATTER, K., AIELLO, A.,
BCL-2 IN CANCER
107
108 109
110
111
112
113
114
115 116 117 118 119
120 121 122
123
124
145
DELIA, D., GIARDINI, R. & RILKE, F. (1994) p53 and bcl-2 expression in high-grade B-cell lymphomas: correlation with survival time. Br. J. Cancer 69, 337–341. . ROMAGUERA, J.E., PUGH, W., LUTHRA, R., GOODACRE, A. & CABANILLAS, F. (1993) The clinical relevance of t(14;18)/Bcl-2 rearrangement and Del 6q in diffuse large cell lymphoma and immunoblastic lymphoma. Ann. Oncol. 4, 51–54. . JACOBSON, J.O., WILKES, B.M., KWIATKOWSKI, D.J., MEDEIROS, L.J., AISENBERG, A.C. & HARRIS, N.L. (1993) Bcl-2 rearrangements in de novo diffuse large cell lymphoma. Cancer 72, 231–236. . PEZZELLA, F., JONES, M., RALFKIAER, E., ERSBOLL, J., GATTER, K.C. & MASON, D.Y. (1992) Evaluation of bcl-2 protein expression and 14;18 translocation as prognostic markers in follicular lymphoma. Br. J. Cancer 65, 87–89. . CASTLE, V.P., HEIDELBERGER, K.P., BROMBERG, J., OU, X., DOLE, M. & NUÑEZ, G. (1993) Expression of the apoptosis-suppressing protein bcl-2, in neuroblastoma is associated with unfavourable histology and N-myc amplification. Amer. J. Pathol. 143, 1543–1550. . KRAJEWSKI, S., CHATTEN, J., HANADA, M., WOMER, R. & REED, J.C. (1995) Immunohistochemical analysis of Bcl-2 oncoprotein in human neuroblastoma: Correlations with tumor cell differentiation and N-myc protein. Lab Invest. 71, 42–54. . REED, J.C., CUDDY, M., HALDAR, S., CROCE, C., NOWELL, P., MAKOVER, D. & BRADLEY, K. (1990) BCL2-mediated tumorigenicity in a T-lymphoid cell line: synergy with C-MYC and inhibition by BCL2 antisense. Proc. Natl Acad. Sci. USA 87, 3660. . NUÑEZ, G., LONDON, L., HOCKENBERY, D., ALEXANDER, M., MCKEARN, J.P. & KORSMEYER, S.J. (1990) Deregulated bcl-2 gene expression selectively prolongs survival of growth factor-deprived hemopoietic cell lines. J. Immunol. 144, 3602–3610. . BORZILLO, G.V., ENDO, K. & TSUJIMOTO, Y. (1992) Bcl-2 confers growth and survival advantage to interleukin 7-dependent early pre-B cells which become factor independent by a multistep process in culture. Oncogene 7, 869–876. . GARCIA, I., MARTINOU, I., TSUJIMOTO, Y. & MARTINOU, J.-C. (1992) Prevention of programmed cell death of sympathetic neurons by the bcl-2 proto-oncogene. Science 258, 302–304. . ALLSOPP, T.E., WYATT, S., PATERSON, H.F. & DAVIES, A.M. (1993) The proto-oncogene bcl-2 can selectively rescue neurotrophic factor-dependent neurons from apoptosis. Cell 73, 295–307. . BEHL, C., HOVEY III, L., KRAJEWSKI, S., SCHUBERT, D. & REED, J.C. (1993) Bcl-2 prevents killing of neuronal cell lines by L-glutamate but not by amyloid- peptide. BBRC 197, 949–956. . ZHONG, L.-T., KANE, D.J. & BREDESEN, D.E. (1993) Bcl-2 blocks glutamate toxicity in neural cell lines. Mol. Brain Res. 19, 353–355. . ZHONG, L.T., SARAFIAN, T., KANE, D.J., CHARLES, A.C., MAH, S.P., EDWARDS, R.H. & BREDESEN, D.E. (1993) bcl-2 inhibits death of central neural cells induced by multiple agents. Proc. Natl Acad. Sci. USA 90, 4533–4537. . TSUJIMOTO, Y. (1989) Stress-resistance conferred by high level of bcl-2 protein in human B lymphoblastoid cell. Oncogene 4, 1331–1336. . LEVINE, B., HUANG, Q., ISAACS, J.T., REED, J.C., GRIFFIN, D.E. & HARDWICK, J.M. (1993) Conversion of lytic to persistent alphavirus infection by the bcl-2 cellular oncogene. Nature 361, 739–742. . YAMADA, T., YAMAOKA, S., GOTO, T., NAKAI, M., TSUJIMOTO, Y. & HATANAKA, M. (1994) The human T-cell leukemia virus type I Tax protein induces apoptosis which is blocked by the bcl-2 protein. J. Virol. 68, 3374–3379. . HENNET, T., BERTONI, G., RICHTER, C. & PETERHANS, E. (1993) Expression of bcl-2 protein enhances and survival of mouse fibrosarcoid cells in tumor necrosis factor-mediated cytotoxicity. Cancer Res. 53, 1456–1460. . BORNER, C., MARTINOU, I., MATTMANN, C., IRMLER, M., SCHARER, E., MARTINOU, J.-C. & TSCHOPP, J. (1994) The protein bcl-2 does not require membrane attachment, but two conserved domains to suppress apoptosis. J. Cell Biol. 126, 1059– 1068.
146
125 126
127
128 129
130 131 132
133 134
135
136
137 138 139
140
141
142
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. ITOH, N., TSUJIMOTO, Y. & NAGATA, S. (1993) Effect of bcl-2 on fas antigen-mediated cell death. J. Immunol. 151, 621–627. . VANHAESEBROECK, B., REED, J.C., DEVALCK, D., MIYASHITA, Y., TANAKA, S. , BEYAERT, R., VAN ROY, F. & FIERS, W. (1993) Effect of Bcl-2 proto oncogene expression on tumor necrosis factormediated cytotoxicity. Oncogene 8, 1075–1081. . OWEN-SCHAUB, L.B., RADINSKY, R., KRUZEL, E., BERRY, K. & YONEHARA, S. (1994) Anti-Fas on nonhematopoietic tumors: levels of Fas/APO-1 and bcl-2 are not predictive of biological responsiveness. Cancer Res. 54, 1580–1586. . TORIGOE, T., MILLAN, J.A., TAKAYAMA, S., TAICHMAN, R., MIYASHITA, T. & REED, J.C. (1994) Bcl-2 inhibits T-cell-mediated cytolysis of leukemia cell line. Cancer Res. 54, 4851–4854. . TAKAYAMA, S., SATO, T., KRAJEWSKI, S., KOCHEL, K., IRIE, S., MILLAN, J. & REED, J.C. (1995) Cloning and functional analysis of BAG-1: a novel Bcl-2 binding protein with anti-cell death activity. Cell 80, 279–284. . VAUX, D.L., AGUILA, H.L. & WEISSMAN, I.L. (1992) Bcl-2 prevents death of factor-deprived cells but fails to prevent apoptosis in targets of cell mediated killing. Int. Immunol. 4, 821–824. . CHIU, V.K., WALSH, C.M., LIU, C.-C., REED, J.C. & CLARK, W.R. (1995) bcl-2 blocks degranulation-but not fas-based cell-mediated cytotoxicity. J. Immunol. 154, 2023–2032. . MIURA, M., ZHU, H., ROTELLO, R., HARTWIEG, E.A. & YUAN, J. (1993) Induction of apoptosis in fibroblasts by II–1 -converting enzyme, a mammalian homolog of the C. elegans cell death gene ced-3. Cell 75, 653–660. . WANG, L., MIURA, M., BERGERON, L., ZHU, H. & YUAN, J (1994) Ich-1, and ice/ced-3-related gene, encodes both positive and negative regulators of programmed cell death. Cell 78, 739–750. . KUMAR, S., KINOSHITA, M., NODA, M., COPELAND, N.G. & JENKINS, N.A. (1994) Induction of apoptosis by the mouse Nedd2 gene, which encodes a protein similar to the product of the Caenorhabditis elegans cell death gene ced-3 and the mammalian IL-1 beta-converting enzyme. Genes Dev. 8, 1613–1626. . RAO, L., DEBBAS, M., SABBATINI, P., HOCKENBERY, D. & KORSMEYER, S. (1992) The adeno virus E1A proteins induce apoptosis, which is inhibited by the E1B 19-kda and bcl-2 proteins. Proc. Natl Acad. Sci. USA 89, 7742–7746. . SELVAKUMARAN, M., LIN, H.K., REED, J.C., TANAKA, S., LIEBERMANN, D.A. & HOFFMAN, B. (1994) The novel differentiation-associated gene MyD 118, proto-oncogenes myb, myc, and bcl-2, and the differentiated state modulate TGFb1-induced apoptosis of myeloid leukemia cells. Mol. Cell. Biol. 14, 2352–2360. . BISSONNETTE, R.P., ECHEVERRI, F., MAHBOUBI, A. & GREEN, D.R. (1992) Apoptotic cell death induced by c-myc is inhibited by bcl-2. Nature 359, 552–554. . FANIDI, A., HARRINGTON, E.A. & EVAN, G.I. (1992) Cooperative interaction between c-myc and bcl-2 protooncogenes. Nature 359, 554–556. . WANG, H.-G., MILLAN, J.A., COX, A.D., DER, C.J., RAPP, U.R., BECK, T. & REED, J.C. (1995) R-Ras accelerates apoptosis caused by growth factor deprivation via a bcl-2 suppressible mechanism. J. Cell. Biol. 129, 1103–1114. . ASKEN, D.S., ASHMEN, R.A., SIMMONS, B.C. & CLEVELAND, J.L. (1991) Constitutive c-myc expression in an IL-3-dependent myeloid cell line suppresses cell cycle arrest and accelerates apoptosis. Oncogene 6, 1915–1922. . EVAN, G.I., WYLLIE, A.H., GILBERT C.S., LITTLEWOOD, T.D., LAND, H., BROOKS, M., WATERS, C.M., PENN, L.Z. & HANCOCK, D.C. (1992) Induction of apoptosis in fibroblasts by c-myc protein. Cell 69, 119–128. . ALNEMRI, E.S., ROBERTSON, N.M., FERNANDES, T.F., CROCE, C.M. & LITWACK, G. (1992) Overexpressed full-length human Bcl-2 extends the survival of baculovirus-infected Sf9 insect cells . Proc. Natl Acad. Sci. USA 89, 7295–7299.
BCL-2 IN CANCER
143 144 145
146 147
148
149
150
151 152 153
154 155
156
157 158 159 160 161
162
147
. VAUX, D.L., WEISSMAN, I.L. & KIM, S.K. (1992) Prevention of programmed cell death in Caenorhabditis elegans by human bcl-2. Science 258, 1955–1957. . KITADA, S., MIYASHITA, T., TANAKA, S. & REED, J.C. (1993) Investigations of antisense oligonucleotides targeted against bcl-2 RNAs. Antisense Res. Develop. 3, 157– 169. . NAKAYAMA, K.-I., NAKAYAMA, K., NEGISHI, I., KUIDA, K., SHINKAI, Y., LOUIE, M.C., FIELDS, L.E., LUCAS, P.J., STEWART, V., ALT, F.W. & LOH, D.Y. (1993) Disappearance of the lymphoid system in bcl-2 homozygous mutant chimeric mice. Science 261, 1584–1588. . VEIS, D.J., SORENSON, C.M., SHUTTER, J.R. & KORSMEYER, S.J. (1993) Bcl-2-deficient mice demonstrate fulminant lymphoid apoptosis, polycystic kidneys, and hypopigmented hair. Cell 75, 229–240. . DENG, G. & PODACK, E.R. (1993) Suppression of apoptosis in a cytotoxic T-cell line by interleukin 2mediated gene transcription and deregulated expression of the protooncogene bcl-2. Proc. Natl Acad. Sci. USA 90, 2189–2193. . CUENDE, E., ALES-MARTINEZ, J.E., DING, L., GONZALEZ-GARCIA, M., MARTÍNEZ-A, C. & NUÑEZ, G. (1993) Programmed cell death by bcl-2-dependent and independent mechanisms in B lymphoma cells. EMBO J. 12, 1555–1560. . GOTTSCHALK, A.R., BOISE, L.H., THOMPSON, C.B. & QUINTÀNS, J. (1994) Indentification of immunosuppressant-induced apoptosis in a marine B-cell line and its prevention by bcl-X but not bcl-2. Proc. Natl Acad. Sci. USA 91, 7350–7354. . MAKOVER, D., CUDDY, M., BRADLEY, K., ALPERS, J., SUKHATME, V. & REED, J. (1999) Phorbol ester-mediated inhibition of growth and regulation of proto-oncogene expression in the human T cell leukemia line JURKAT. Oncogene 6, 455–460. . CHEN-LEVY, Z. & CLEARY, M.L. (1990) Membrane topology of the Bcl-2 protooncogenic protein demonstrated in vitro. J. Biol. Chem. 265, 4929–4933. . MONAGHAN, P., ROBERTSON, D., ANDREW, T., AMOS, S., DYER, M.J.S., MASON, D.Y. & GREAVES, M.F. (1992) Ultrastructural localization of bcl-2 protein. J. Histochem. Cytochem. 40, 1819–1825. . KRAJEWSKI, S., TANAKA, S., TAKAYAMA, S., SCHIBLER, M.J., FENTON, W. & REED, J.C. (1993) Investigations of the subcellular distribution of the bcl-2 oncoprotein: residence in the nuclear envelope, endoplasmic reticulum, and outer mitochrondrial membranes. Cancer Res. 53, 4701–4714. . DE JONG, D., PRINS, F.A., MASON, D.Y., REED, J.C., VAN OMMEN, G.B. & KLUIN, P.M. (1994) Subcellular localization of the bcl-2 protein in malignant and normal lymphoid cells. Cancer Res. 54, 256–260. . LITHGOW, T., VAN DRIEL, R., BERTRAM, J.F. & STRASSER, A. (1994) The protein product on the oncogene bcl-2 is a component of the nuclear envelope, the endoplasmic reticulum, and the other mitochondrial membrane. Cell. Growth Differ. 3, 411–417. . AKAO, Y., OTSUKI, Y., KATAOKA, S., ITO, Y. & TSUJIMOTO, Y. (1994) Multiple subcellular localization of bcl-2: detection in nuclear outer membrane, endoplasmic reticulum membrane, and mitochondrial membranes. Cancer Res. 54, 2468–2471. . JACOBSON, M.D., BURNE, J.F., KING, M.P., MIYASHITA, T., REED, J.C. & RAFF, M.C. (1993) Apoptosis and bcl-2 protein in cells without mitochondrial DNA. Nature 361, 365–368. . MEIKRANTZ, W., GISSELBRECHT, S., TAM, S.W. & SCHLEGEL, R. (1994) Activation of cyclin Adependent protein kinases during apoptosis. Proc. Natl Acad. Sci. USA 91, 3754–3758. . JACOBSON, M.D., BURNE, J.F. & RAFF, M.C. (1994) Programmed cell death and Bcl-2 protection in the absence of a nucleus. EMBO J. 13, 1899–1910. . CHIOU, S.K., RAO, L. & WHITE, E. (1994) Bcl-2 blocks p53-dependent apoptosis. Mol. Cell Biol. 14, 2556–2563. . BAFFY, G., MIYASHITA, T., WILLIAMSON, J.R. & REED, J.C. (1993) Apoptosis induced by withdrawal of Interleukin-3 [IL-3] from an IL-3-dependent hematopoietic cell line associated with repartitioning of intracellular calcium and is blocked by enforced Bcl-2 oncoprotein production. J. Biol. Chem. 268, 6511–6519. . LAM, M., DUBYAK, G., CHEN, L., NUÑEZ, G., MIESFELD, R.L. & DisTELHORST, C.W. (1994) Evidence that bcl-2 represses apoptosis by regulating endoplasmic reticulum-associated Ca2+ fluxes. PNAS 91, 6569–6573.
148
163 164
165 166
167 168 169
170
171 172
173
174
175 176
177
178
179 180
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. LAM, M., DUBYAK, G. & DISTELHORST, C.W. (1993) Effect of glucocorticosteroid treatment on intracellular calcium homeostasis in mouse lymphoma cells. Mol. Endocrinol. 7, 686–693. . DOWD, D.R., MACDONALD, P.N., KOMM, B.S., HAUSSLER, M.R. & MIESFELD, R.L. (1992) Stable expression of the calbindin-D28K complementary DNA interferes with the apoptotic pathway in lymphocytes. Mol. Endocrinol. 6, 1843–1848. . NEWMEYER, D., FARSCHON, D.M. & REED, J.C. (1994) Cell-free apoptosis in Xenopus egg extracts: by Bcl-2 inhibits a latant cytoplasmic phase. Cell 79, 353–364. . KAMADA, S., SHIMONO, A., SHINTO, Y., TSUJIMURA, T., TAKAHASHI, T., NODA, T., KITAMURA, Y., KONDOH, H. & TSUJIMOTO, Y. (1995) Bcl-2 deficiency in mice leads to pleiotropic abnormalities: accelerated lymphoid cell death in thymus and spleen, polycystic kidney, distorted intestine and hair hypopigmentation. Cancer Res. (in press). . CHITTENDEN, T., HARRINGTON, E.A., O’CONNOR, R., FLEMINGTON, C., LUTZ, R.J., EVAN, G.I. & GUILD, B.C. (1995) Induction of apoptosis by the Bcl-2 homologue Bak. Nature 374, 733–736. . KIEFER, M.C., BRAUER, M.J., POWERS, V.C., WU, J.J., UBANSKY, S.R., TOMEI, L.D. & BARR, P.J. (1995) Modulation of apoptosis by the widely distributed Bcl-2 homologue Bak. Nature 374, 736–739. . FARROW, S.N., WHITE, J.H.M., MARTINOU, I., RAVEN, T., PUN, K.-T., GRINHAM, C.J., MARTINOU, J.-C. & BROWN, R. (1995) Cloning of a bcl-2 homologue by interaction with adenovirus E1B 19K. Nature 374, 731–733. . NEILAN, J.G., LU, Z., AFONSON, C.L., KUTISH, G.F., SUSSMAN, M.D. & ROCK, D.L. (1993) An African swine fever virus gene with similarity to the protooncogene bcl-2 and the Epstein-Barr virus gene BHRF1. J. Virol. 67, 4391–4394. . SATO, T., IRIE, S., KRAJEWSKI, S. & REED, J.C. (1994) Cloning and sequencing of cDNA encoding rat bcl-2 protein. Gene 140, 291–292. . GONZÁLEZ-GARCIA, M., PÉREZ-BALLESTERO, R., DING, L., DUAN, L., BOISE, L.H., THOMPSON, C.B. & NUÑEZ, G. (1994) bcl-XL is the major bcl-X mRNA form expressed during murine development and its product localizes to mitochondria. Development 120, 3033–3042. . TANAKA, S., SAITO, K. & REED, J.C. (1993) Structure-function analysis of the apoptosis-suppressing bcl-2 oncoprotein: Substitution of a heterologous transmembrane domain restores function to truncated Bcl-2 proteins. J. Biol. Chem. 268, 10920–10926. . NGUYEN, M., BRANTON, P.E., WALTON, P.A., OLTVAI, Z.N., KORSMEYER, S.J. & SHORE, G.C. (1994) Role of membrane anchor domain of bcl-2 in suppression of apoptosis caused by E1B-defective adenovirus. J. Biol. Chem. 269, 16521–16524. . FRISCH, S.M. & FRANCIS, H. (1994) Disruption of epithelial cell-matrix interactions induces apoptosis. J. Cell. Biol. 124, 619–626. . KRAJEWSKI, S., KRAJEWSKA, M., SHABAIK, A., MIYASHITA, T., WANG, H.-G. & REED, J.C. (1994) Immunohistochemical determination of in vivo distribution of Bax, a dominant inhibitor of Bcl-2. Amer. J. Pathol. 145, 1323–1336. . KRAJEWSKI, S., BODRUG, S., GASCOYN, R., BEREAN, K., KRAJEWSKA, M. & REED, J.C. (1994) Immunohistochemical analysis of mcl-1 and bcl-2 proteins in normal and neoplastic lymph nodes. Amer. J. Pathol. 145, 515–525. . HICKISH, T., ROBERTSON, D., CLARKE, P., HILL, M., DI STEFANO, F., CLARKE, C. & CUNNINGHAM, D. (1994) Ultrastructural localization of BHRF1: An Epstein-Barr virus gene product which has homology with bcl-2. Cancer Res. 54, 2808– 2811. . WHITE, E. & CIPRIANI, R. (1989) Specific disruption of intermediate filaments and the nuclear lamina by the 19-kDa product of the adenovirus E1B oncogene. Proc. Natl Acad. Sci. USA 86, 9886–9890. . YIN, X.-M., OLTVAI, Z.N. & KORSMEYER, S.J. (1994) BH1 and BH2 domains of Bcl-2 are required for inhibition of apoptosis and heterodimerization with Bax. Nature 369, 321–323.
BCL-2 IN CANCER
181
182
183
184 185 186
187
188 189 190 191
192 193 194 195 196 197
198 199 200 201
149
. HANADA, M., AIMÉ-SEMPÉ, C., SATO, T. & REED, J.C. (1995) Structure-function analysis of bcl-2 protein: identification of conserved domains important for homodimerization with bcl-2 and heterodimerization with bax. J. Biol. Chem. 270, 11962– 11968. . HENDERSON, S., ROWE, M., GREGORY, C., CROOM-CARTER, D., WANG, F., LONGNECKER, R., KIEFF, E. & RICKINSON, A. (1991) Induction of bcl-2 expression by Epstein-Barr virus latent membrane protein 1 protects infected B-cells from programmed cell death. Cell 65, 1107–1115. . FINKE, J., FRITZEN, R., TERNES, P., TRIVEDI, P., BROSS, K.J., LANGE, W., MERTELSMANN, R. & DOLKEN, G. (1992) Expression of bcl-2 in Burkitt’s lymphoma cell lines: Induction by latent Epstein-Barr virus genes. Blood 80, 459–469. . MARTIN, J.M., VEIS, D., KORSMEYER, S.J. & SUGDEN, B. (1993) Latent membrane protein of EpsteinBarr virus induces cellular phenotypes independently of expression of bcl-2. J. Virol. 67, 5269–5278. . FERNANDEZ, M.J. & BISCHOFF, J.R. (1993) Bcl-2 associates with the ras-related protein R-ras p23. Nature 366, 274–275. . WANG, H.-G., MIYASHITA, T., TAKAYAMA, S., SATO, T., TORIGOE, T., KRAJEWSKI, S., TANAKA, S., HOVEY, L. III, TROPPMAIR, J., RAPP, U.R. & REED, J.C. (1994) Apoptosis regulation by interaction of Bcl-2 protein and Raf-1 kinase. Oncogene 9, 2751–2756. . BOYD, J.M., MALSTROM, S., SUBRAMANIAN, T., VENKATESH, L.K., SCHAEPER, U., ELANGOVAN, B., D’SA-EIPPER, C. & CHINNADURAI, G. Adenovirus E1B 19 kDA and bcl-2 proteins interact with a common set of cellular proteins. Cell 79, 341–351. . Cox, A.D., BRTVA, T.R., LOWE, D.G. & DER, C.J. (1994) R-ras induces malignant, but not morphologic, transformation of NIH 3T3 cells. Oncogene 9, 3281–3288. . GILLET, G., GUERIN, M., TREMBLEAU, A. & BRUN, G. (1995) A Bcl-2-related gene is activated in avian cells transformed by the Rous sarcoma virus. EMBO J. 14, 1372– 1381. . REY, I., TAYLOR-HARRIS, P., VAN ERP, H. & HALL, A. (1994) R-ras interacts with rasGAP, neurofibromin and c-raf but does not regulate cell growth or differentiation. Oncogene 9, 685–692. . SPAARGAREN, M., MARTIN, G.A., MCCORMICK, F., FERNANDEZ-SARABIA, M.J. & BISCHOFF, J.R. (1994) The Ras-related protein R-ras interacts directly with Raf-1 in a GTP-dependent manner. Biochem. J. 300, 303–307. . KOCACHE, M.M. & PEARSON, G.R. (1990) Protein kinase activity associated with a cell cycle regulated, membrane-bound Epstein-Barr virus induced early antigen. Intervirology 31, 1–13. . KRAJEWSKI, S., MAI, J.K., KRAJEWSKA, M., SIKORSKA, M., MOSSAKOWSKI, M.J. & REED, J.C. (1995) Upregulation of bax protein levels in neurons following cerebral ischemia. J. Neurosci. 15, 6364–6376. . KITADA, S., KRAJEWSKI, S., MIYASHITA, T., KRAJEWSKA, M. & REED, J.C. (1995) -radiation induces upregulation of Bax protein and apoptosis in radiosensitive cells in vivo. Oncogene, in press. . MIYASHITA, T. & REED, J.C. (1995) Tumor suppressor p53 is a direct transcriptional activator of human bax gene. Cell 80, 293–299. . GOTTLIEB, E., HAFFNER, R., VON RÜDEN, T., WAGNER, E.F. & OREN, M. (1994) Down-regulation of wild-type p53 activity interferes with apoptosis of IL-3-withdrawal. EMBO J. 13, 1368–1374. . TANAKA, N., ISHIHARA, M., KITAGAWA, M., HARADA, H., KIMURA, T., MATSUYAMA, T., LAMPHIER, M.S., AIZAWA, S., MAK, T.W. & TANIGUCHI, T. (1994) Cellular commitment to oncogeneinduced transformation or apoptosis is dependent on the transcription factor IRF-1. Cell 77, 829–839. . SHEN, Y. & SHENK, T. (1994) Relief of p53-mediated transcriptional repression by the adenovirus E1B 19kDa protein or the cellular Bcl-2 protein. Proc. Natl Acad. Sci. USA 91, 8940–8944. . LO COCO, F., GAIDANO, G., LOUIE, D.C., OFFIT, K., CHAGANTI, R.S.K. & DALLA-FAVERA, R. (1993) p53 mutations are associated with histologic transformation of follicular lymphoma. Blood 82, 2289–2295. . SANDER, C.A., YANO, T., CLARK, H.M., HARRIS, C., LONGO, D.L., JAFFE, E.S. & RAFFELD, M. (1993) p53 mutation is associated with progression in follicular lymphomas. Blood 82, 1994–2004. . SAKHI, S., BRUCE, A., SUN, N., TOCCO, G., BAUDRY, M. & SCHREIBER, S.S. (1994) p53 induction is associated with neuronal damage in the central nervous system. Proc. Natl Acad. Sci. USA 91, 7525–7529.
150
202
203 204
205 206 207 208
209
210
211 212 213
214
215
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. GRAEBER, T.G., PETERSON, J.F., TSAI, M., MONICA, K., FORNACE, A.J. JR. & GIACCIA, A.J. (1994) Hypoxia induces accumulation of p53 protein, but activation of a G1-phase checkpoint by low-oxygen conditions is independent of p53 status. Mol. Cell. Biol. 14, 6264–6277. . KRAJEWSKI, S., KRAJEWSKA, M., SHABAIK, A., WANG, H.-G., IRIE, S., FONG, L. & REED, J.C. (1994) Immunohistochemical analysis of in vivo patterns of Bcl-X expression. Cancer Res. 54, 5501–5507. . KRAJEWSKI, S., BODRUG, S., KRAJEWSKA, M., SHABAIK, A., GASCOYNE, R., BEREAN, K. & REED, J.C. (1995) Immunohistochemical analysis of Mcl-1 protein in human tissues: differential regulation of Mcl-1 and Bcl-2 protein production suggests a unique role for Mcl-1 in control of programmed cell death in vivo. Amer. J. Pathol. 146, 1309–1319. . LEBRUN, D.P., WARNKE, R.A. & CLEARY, M.L. (1993) Expression of bcl-2 in fetal tissues suggests a role in morphogenesis. Amer. J. Pathol. 142, 743–753. . LU, Q.-L., POUSOM, R., WONG, L. & HANBY, A.M. (1993) bcl-2 expression in adult and embryonic nonhaematopoietic tissues. J. Pathol. 169, 431–437. . NOVACK, D.V. & KORSMEYER, S.J. (1994) Bcl-2 protein expression during murine development. Am. J. Pathol. 145, 61–73. . KONDO, E., NAKAMURA, S., ONOUE, H., MATSUO, Y., YOSHINO, T., AOKI, H., HAYASHI, K., TAKAHASHI, K., MINOWADA, J., NOMURA, S. & AKAGI, T. (1992) Detection of bcl-2 protein and bcl-2 messenger RNA in normal and neoplastic lymphoic tissues by immunohistochemistry and in situ hybridization. Blood 80, 2044–2051. . CHLEQ-DESCHAMPS, C.M., LEBRUN, D.P., HUIE, PH., BESNIER, D.P., WARNKE, R.A., SIBLEY, R.K. & CLEARY, M.L. (1993) Topographical dissociation of bcl-2 messenger RNA and protein expression in human lymphoid tissues. Blood 81, 293–298. . SAKURAGI, N., OOUKOS, G., FURTH, E.E., BRONNER, M.P., VANARSDALE, C.M., KRAJEWSKI, S., REED, J.C. & STRAUSS, J.F. III (1994) Differentiation dependent expression of the bcl-2 gene in the trophoblast lineage. J. Soc. Gyn. Invest. 1, 164–172. . OTANI, H., ERDOS, M. & LEONARD, W.J. (1993) Tyrosine kinase(s) regulate apoptosis and bcl-2 expression in a growth factor-dependent cell line. J. Biol. Chem. 268, 22733–22736. . LEVY, Y. & BROUET, J.-C. (1994) Interleukin-10 prevents spontaneous death of germinal center B-cells by induction of the bcl-2 protein. J. Clin. Invest. 93, 424–428. . FERNANDEZ, A., MARIN, M.C., MCDONNEL, T. & ANANTHASWAMY, H.N. (1994) Differential sensitivity of normal and Ha-ras-transformed C3H mouse embryo fibroblasts to tumor necrosis factor: induction of bcl-2, c-myc, and manganese superoxide dismutase in resistant cells . Oncogene 9, 2009–2017. . CARSON, W.E., HALDER, S., BAIOCCHI, R.A., CROCE, C.M. & CALIGIURI, M.A. (1994) The c-kit ligand suppresses apoptosis of human natural killer cells through the upregulation of bcl-2. Proc. Natl Acad. Sci. USA 91, 7553–7557. . PREHN, J., BINDOKAS, V., MARACILLI, C., KRAJEWSKI, S., REED, J.C. & MILLAN, R. Protective effect of transforming growth factor- 1 (TGF- ) on -amyloid neurotoxicity in rat hippocampal neurones. Mol. Pharm. (in press).
8 Molecular and Genetic Control of Apoptosis in Drosophila JOHN M.ABRAMS Department of Cell Biology and Neuroscience, The University of Texas Southwestern Medical Center, 5323 Harry Hinds Boulevard, Dallas, Texas 75235–9039, USA
8.1 Introduction As animals develop and age, a large number of cells are eliminated by a genedirected process referred to as programmed cell death (PCD) or apoptosis. This phenomenon is a critical determinant of embryonic pattern formation which, later in life, also serves to regulate the type and number of cells within a given tissue. The extrinsic signals and cues which elicit PCD have been extensively investigated and well characterized in a variety of experimental contexts [1–7] and numerous studies have begun to elucidate the importance of apoptosis in the etiology and treatment of human disease [reviewed in 8]. The cytological features of apoptotic cell deaths are strikingly similar in all metazoans and several lines of recent evidence suggest that some molecular components required for this process are highly conserved [9, 10]. Yet despite the prevalence of apoptosis in biological systems, and its obvious clinical importance, the central biochemical mechanisms that implement this suicidal process are poorly understood. How do otherwise healthy cells become committed to die? How do cells execute this destructive decision once the apoptotic fate determination has been made? Experiments on model systems such as the nematode and the fruit fly have begun to answer these questions at the molecular level. By applying classical genetic tools in combination with modern molecular techniques, studies of PCD in these organisms have provided remarkable insights into the nature of cellular suicide and offered the first molecular clues regarding genetic components required for apoptosis. The apparent conservation of at least some apoptosis genes (see below) suggests that this information could eventually permit us to develop treatments for human diseases that are caused, or exacerbated by, the misregulation of apoptosis. Cell death defective mutations in the nematode Caenorhabditis elegans provided the first substantial genetic evidence that ‘naturally-occurring’ cell death was indeed an active cellular process. One of the earliest PCD mutants isolated was found to lack a nuclease responsible for the characteristic destruction of apoptotic nuclei and yet, in these animals, normal frequencies of cell deaths were still observed [11]. Other single-gene mutations were found that could block all cell deaths [12] suggesting that a common mechanism of killing is utilized among all the cells which die in this animal. Cell death defective mutants have been grouped according to whether they affect the specification of PCD, the execution of PCD or the final engulfment of dead cell corpses [reviewed in 4] and genes corresponding to several of these mutations have
152
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
been cloned [9, 10, 13]. One gene which normally acts to suppress PCD in nematodes, ced-9, [14] shows a significant degree of structural similarity to bcl-2, a proto-oncogene which suppresses apoptosis in mammals [9]. More intriguing, perhaps, is the observation that human bcl-2 can partially rescue ced-9 lossof-function mutations in transgenic nematodes [9, 15]. Another important cell death gene in nematodes, ced-3, encodes a cysteine protease homologous to mammalian IL-1/ converting enzyme [10] which can induce apoptosis in cultured fibroblasts [16]. These observations suggest that at least some components involved in PCD may have conserved functions. Compared to the depth of knowledge available for C. elegans, the genetics of PCD in Drosophila is far less well characterized. Yet the fruit fly also affords a very powerful experimental context to study PCD and, recently, a number of loci have been shown to specifically influence or regulate cell death in this organism [17–21]. In contrast to the nematode C. elegans, many cell deaths in the fly are not strictly predetermined by lineage [17, 19]. Moreover, this organism displays remarkable developmental plasticity, including the ability to eliminate cells (by apoptosis) that do not successfully complete their developmental program. Drosophila is thus uniquely suited for studying how cell interactions or environmental stresses can specify or trigger the cell death fate. This chapter provides a broad review of PCD in Drosophila and attempts to emphasize the utility of this model system for dissecting the molecular basis of apoptosis during normal and abnormal development. 8.2 Detection and Occurrence of PCD During Development Cell death has been described in many tissues at virtually all developmental stages of Drosophila. In general, investigators have utilized histological stains (such as toluidine blue) on fixed preparations and/or vital stains on live preparations to selectively identify dying cells. The vital dye acridine orange (AO) can be used to visualize cell death in dissected imaginal discs [19, 22, 23] and, with some adaptation, AO can be used to visualize PCD in the embryo [17]. AO specifically detects apoptotic cell deaths (as opposed to necrotic deaths) and this label persists even after cell corpses have been entirely engulfed within a macrophage [17]. Live preparations can also be stained with Nile blue to detect apoptosis [17]. In fixed preparations, PCD can be visualized by histochemical methods which exploit the characteristic degradation of chromatin in apoptotic nuclei [18]. One technique [24], called TUNEL labeling, detects free DNA ends in situ with terminal deoxytransferase. Another method exploits nick translation labeling in situ to similarly detect cleaved DNA [25]. The application of these methods to Drosophila tissue suggests that an evolutionarily conserved mechanism is responsible for DNA degradation during apoptosis. Although many of descriptions of PCD in Drosophila occurred prior to common usage of the term ‘apoptosis’, an examination of published ultrastructure often shows that the mode of death is clearly apoptotic (nuclear material condenses, cells become shrunken and cellular fragments are engulfed). For instance, an examination of PCD in ovarian chambers reported that the nuclei of some nurse cells were severely condensed and the resulting cellular debris was engulfed by neighboring follicle cells [26]. Some of the earliest descriptions of cell death in Drosophila were reported by Fristrom [27, 28] studying imaginal development of the eye and wing disc. These reports established that ectopic or inappropriate cell death during development accounted for structural defects in adults from a variety of mutant strains (see below). They also demonstrated that genotypes could have remarkable and predictable effects upon the incidence of cell death. PCD is a very prominent feature during embryonic development [29]. Ultrastructural examination of cell deaths at this stage show a striking resemblance to apoptotic cell deaths observed in vertebrate systems
CONTROL OF APOPTOSIS IN DROSOPHILA
153
[17]. Patterns of cell death in the embryo are dynamic and widespread among many different organs and tissues, yet a fairly stereotypical distribution and number are observed for each stage. After embryogenesis, conspicuous numbers of dying cells are not observed until metamorphic changes are initiated. In fact, the widespread histolysis of larval tissue during metamorphosis is often cited as a classic example of PCD. In young pupa, larval muscles separate from their attachment sites and gradually degenerate into fragments that are consumed by phagocytes [30]. Most head and thoracic muscles die earlier than abdominal muscles. This predictable and orderly destruction lent early support to the notion that cell death is a ‘programmed’ event subject to regulation by genetic information. At this stage, scattered cell death has also been observed in the developing wing [31] and eyes of wild type flies [19]. Wolff and Ready [19, 23] have undertaken an extensive analysis of the patterns of apoptotic cell death which occur during normal eye development. In this context, they have found that as the cell collapses, prominent microtubules are lost and that the remaining fragments are engulfed by neighboring cells. Examples of cell death also occur after metamorphosis is complete and continues in the adult stage. The death of larval adipose cells, for instance, initiates in the pupal stage and continues through the third day of adult life [32]. Studies reported by Kimura and Truman [33] have described cell death in the nervous and musculature system within the first day after eclosion. Their analyses (see below) suggest that signals governing death of muscle cells are different from those regulating neuronal death. 8.3 Treatments that Induce Apoptosis Ionizing radiation is the most commonly used agent to induce apoptosis in Drosophila. Irradiation has historically been used to generate mosaic animals during early larval development [34] and studies to characterize the biological consequences of these treatments cited clear effects upon the incidence of cell death. During these treatments ~30% of the cells in a developing wing are killed, probably as the result of defective cell divisions [35]. In wing discs, treatment with ionizing radiation showed evidence of both apoptosis and necrotic death [36] whereas only apoptotic deaths were observed in embryos exposed to similar treatments [17, 18]. Not surprisingly, the stages and cell types that are most vulnerable to radiation correlate with replicative potential [37–39]. Moreover, notable differences in the kinetics and onset of apoptosis may be influenced by exposure doses [17]. Induction of cell death by ionizing radiation may reflect a process of ‘altruistic suicide’ or ‘cell-replacement repair’ whereby the elimination of cells that may harbor damaging mutations stimulates the proliferation of healthy cells to replace them [40]. 8.4 Signals that Govern Cell Death Among several potential endocrine regulators of PCD in Drosophila, the ecdysteroid hormones are the most well understood. Regulation of cellular physiology and gene expression cascades by steroids are extremely well characterized at the molecular and genetic levels [e.g. 41, 42]. Much of our knowledge regarding the regulation of PCD by ecdysteroids derives from analogy to studies of the moth Manduca sexta and the giant silkworm Antheraea polyphemus [3, 43] which are particularly amenable to detailed physiological studies because of their large size. In these insects, metamorphic cell death is largely governed by falling titers of ecdysteroids. Withdrawal of this hormone can either directly trigger cell death [44] or, alternatively, lower ecdysone levels can ‘prime’ cells for this fate [45]. Because these deaths can be suppressed by cyclohexamide [46] or actinomycin D [43] they probably require macromolecular synthesis. Other
154
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
endocrine factors, including juvenile hormone [47] and eclosion hormone [45] have been implicated as potential PCD cues in larger insects and may serve similar functions in Drosophila. Mutations in genes that encode receptors for these endocrine signals should provide exciting avenues for exploring hormonal mechanisms that regulate PCD. In fact, probes for the ecdysone receptor [48] have already provided some tantalizing clues. This receptor encodes three protein isoforms [49] that share DNA and hormone binding domains yet differ in their N-terminal regions. Expression patterns of each isoform show distinct tissue distributions that correlate with differential responses to ecdysone during metamorphosis [49]. High expression levels of one isoform, EcR-A, specifically anticipates PCD in a heterogeneous group of cells in the central nervous system [50]. Apparently, the predetermined fate of these doomed cells is reflected by preferential expression of this receptor isoform. The importance of this observation was substantiated by showing that treatment with ecdysone could block the death of these cells if provided 3 h prior to degeneration [50]. Withdrawal of ecdysone may trigger the onset of death in neurons expressing high levels of EcR-A by direct derepression of death-related genes. In this context, the functional relevance of the differing N terminal regions in these receptor isoforms should provide important clues regarding the molecular basis of differential cellular responses to a common hormonal signal. 8.5 Mutations that Cause Ectopic Cell Death Many mutations in Drosophila are associated with excessive cell deaths in specific regions or structures during development. The first rigorous analysis of this issue was reported by Fristrom [27, 28] who showed that characteristic patterns of ectopic cell death were associated with six different mutations. Ectopic cell death mutations now include a large number of loci, many of which appear to be required for critical developmental events [17, 20, 51–54]. Where analyzed, these cell deaths appear to be apoptotic in character [17, 51]. Table 8.1 lists genes that cause excessive or ectopic cell death when mutated. Table 8.1 A representative listing of Drosophila genes that are associated with excessive cell death when mutated. For each locus, tissue(s) where ectopic cell death has been described are indicated along with the relevant citation. If available, molecular information regarding identified protein motifs or homologues is cited in parentheses and also referenced Gene/mutation (protein or motifs)
Tissue affected
References
Bar (homeo domain)
Eye disc
eyeless vestigial apterous (homeo, LIM domains)
Eye disc Wing disc Wing disc
Beadex abnormal wing disc (metastasis suppressor nm23)
Wing disc Wing disc
cut (homeo domain)
Wing disc
decapentaplegic (TGF-B)
All imaginal discs
27 55 27 28 27, 56 57 27 58 59 27 60 51 61
CONTROL OF APOPTOSIS IN DROSOPHILA
Gene/mutation (protein or motifs)
Tissue affected
References
fushi-tarazu (homeo domain)
Embryo
achaete-scute (helix-loop-helix domain)
Embryonic ectoderm
sine oculis (homeo domain)
Optic lobes
l(3)c43hs1 rotund n-chimaerin (GAP) l(1) giant small-optic lobes eyes absent torpedo/Ellipse (EGF receptor family) stardust Star fizzy polyhomeotic (DNA binding) fused rough deal crumbs (EGF repeats) 13 X-linked lethals
Wing disc Imaginal discs Embryo Optic lobes Eye disc Embryo/eye disc Embryonic epithelium Eye, wing disc Embryo Embryo Embryo Eye, abdomen, wing Embryonic epithelium Various discs
52 62, 63 64 65 66 53 67 68 69 66 20 70, 71 72 73 74 75 76 77 54 78
155
Table 8.1 shows that a variety of loci are associated with characteristic patterns of ectopic cell death in many regions of the developing fly. Even a cursory examination suggests that most of these genes encode functions that are involved in the specification of cell fate decisions which occur either during embryonic development or during imaginal development. As is the case with vertebrates, this correlation suggests that the inability to properly differentiate frequently leads to the onset cell death [17]. The fact that so many different mutations can specifically influence the pattern of cell death argues that apoptosis is the outcome of a default program triggered by conflicting developmental signals [6, 79]. This property may be essential for allowing a degree of plasticity during cell fate determination (see below). Estimates from one screen for cell death defective mutations suggest that nearly 20% of the genes in the fly genome can cause excessive PCD when mutated [18]. Because excessive PCD is frequently observed in association with defective development, it is difficult to establish whether these phenotypes reflect the secondary consequences of aberrant development or specific apoptosis functions. Mutations displaying the opposite phenotype (see below) are therefore more likely to reveal specific components of the apoptosis machinery. 8.6 Mutations and Genes that Reduce the Incidence of Apoptosis Some mutations in Drosophila cause a reduction of the incidence of PCD in specific tissues. These loci could affect components necessary to specify cell death, and may be similar to a class of genes in C. elegans that either determine or trigger PCD in the nematode [4]. Mutations at apterous were perhaps the first phenotypes associated with failures in PCD [32, 80]. In wild type adults, cells of the larval fat body
156
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
disappear entirely within the first 4 days after emergence. Animals bearing lesions at apterous initially show delayed rates of adipose cell death and, eventually, these animals show a complete block to PCD in these cells. More than 20% of the larval cells that would otherwise have died were found to persist in apterous adults. Several experiments, including cell transplantations, suggested that the PCD phenotype was nonautonomous and possibly associated with endocrine, neuronal or cytolytic factors [32]. Ironically, mutations at this same locus may also cause excessive PCD in other cell types (see Table 8.1). Because apterous apparently encodes a transcription factor [57] reduced adipose cell death could reflect inappropriate regulation of an extracellular agent which normally triggers the PCD response. Three mutations affecting eye development are associated with reduced cell death phenotypes [19, 81]. At least one allele of Notch reduces the number of cell deaths in the developing retina [81]. This gene is required for a variety of local cell interactions that specify differentiation fates. PCD phenotypes associated with mutations at this locus could therefore result from misrouting of cell fate decisions rather than a direct failure to die [81]. Two other mutations that disrupt retinal development, roughest and echinus, also appear to reduce the incidence of PCD. In the case of the former, mosaic analyses suggested that gene action in this context is cell autonomous and may reflect failures in one of many signaling pathways that provoke apoptosis [19]. Sequence analysis shows that roughest is a large transmembrane protein containing several immunoglobulin-like motifs and that the allele which affects PCD disrupts the intracellular domain [82]. Other genes appear to affect the rate at which cell death occurs [21]. After an adult fly emerges from the puparium, a group of muscles in the head degenerate within 12 hours of eclosion. Two mcd (muscle cell death) mutations delay this process showing blocks to fragmentation and/or absorption of the muscle cell corpses. PCD in the head muscles of mcd mutants is clearly distinct from wild type, suggesting that these genes might function as part of the machinery which dismantles dying cells [21]. Apoptosis is a common response to viral infection that may impede additional viral spread in metazoan hosts [83] and examples of this strategy to eliminate infected host cells occurs in both vertebrates and invertebrates. Evolutionary pressures have apparently led to the inclusion of genes in some viral genomes that restrain this apoptotic response in infected hosts [79]. The p35 gene from the insect baculovirus Autographa californica, for instance, blocks a virally-induced apoptotic response in its natural host [84]. What is most intriguing about this protein is that its function can be observed across the phyletic spectrum. Reports have documented p35 mediated anti-apoptotic effects in the nematode [85], in Drosophila [25] and in mammalian neurons [86]. The cross-species effects of this molecule argue that p35 targets elements of a highly conserved cell death pathway. The biochemical function of this protein is unknown but its cytosolic localization and a lack of sequence similarity to other anti-apoptotic proteins, suggest that p35 may prevent cell death in a fashion that is distinct from members of the bcl-2 family [87]. In Drosophila, when p35 expression is directed either to the developing eye or throughout the embryo, substantially reduced PCD is observed in these tissues [25]. The adult phenotype associated with reduced PCD in the developing retina appears to resemble phenotypes described for mutations at echinus [19]. Moreover, X-irradiation-induced apoptosis in the eye disc is also reduced by p35 expression, suggesting that similar target components are utilized during induced cell deaths (see below). 8.7 Reaper, a Gene Required for Apoptosis The reaper (rpr) region is required for all cell deaths which occur during embryonic development in Drosophila [17, 18]. This locus was recently identified from a screen which sampled ~50% of the fly genome for cell death defective mutations [18]. Deletions of the rpr region prevent all embryonic deaths
CONTROL OF APOPTOSIS IN DROSOPHILA
157
[17, 18] and, in this regard, resemble C. elegans mutations at ced-3 and ced-4. Early embryogenesis is fairly normal in rpr mutants and, although segmentation and cuticular structures are normal, these embryos ultimately show failures in head involution and nerve cord condensation and they do not hatch as larvae. These defects probably reflect the anatomic consequences of failures in cell death commencing at earlier developmental stages. A number of independent criteria establish that deletions for the rpr region are, in fact, cell death defective [18]. First, direct examination by multiple histological criteria, including electron microscopy and TUNEL labeling, shows a conspicuous absence of apoptotic cells. Exposure of mutant embryos to Xirradiation, however, can induce some apoptotic deaths (see below) thus suggesting that the genetic lesion blocks either the commitment to PCD or an early step in the execution of PCD. Second, the macrophages in rpr mutants are unusually small and devoid of internalized cell corpses with which they normally are engorged. Third, significant numbers of supernumerary cells persist in the nervous system of these mutants. Antibodies to either to the elf-1 protein [88] or the Krupple (Kr) gene product [89] detect a 2–3 fold excess of immunoreactive cells in the nervous systems of embryos that are deleted for the 75C1, 2 interval. Apparently, cells that would otherwise die may adopt differentiated fates that are typical of their surrounding tissue. This collection of mutant phenotypes identified the first genetic function that is globally required for all PCD in the Drosophila embryo. When exposed to ionizing radiation, embryos homozygous for rpr deletions show a significant amount of induced apoptotic deaths (~50). However, the actual numbers of induced cell deaths are far less than those observed for wild type siblings exposed to the same treatment [18]. This radiation resistant phenotype is reminiscent of cells that lack p53 [90, 91] and demonstrates that: 1) loss rpr confers protection against the induction of ectopic death by at least one extrinsic agent and 2) some apoptotic deaths can occur in the absence of rpr function. Apparently, ionizing radiation can induce apoptosis through at least two genetic pathways, one of which depends on functions that are utilized during normal embryonic development. Deletions at rpr can also prevent excessive cell deaths associated with at least one ectopic cell death mutation, crumbs (see section 8.5: Mutations that cause ectopic cell death). These observations generally imply that genes which are necessary for PCD may also be utilized during apoptotic deaths that are induced in abnormal contexts. The cell death phenotype was originally defined by a deletion interval (DfH99) that maps to genomic region 75C1, 2 on the third chromosome [18]. By establishing a detailed molecular examination of the obligate interval, we were able to identify the rpr transcript. This gene is specifically expressed in cells that will later die and can partially restore PCD in mutant embryos. The hypothetical translation product for rpr encodes a 65 amino acid open reading frame bearing no obvious similarities to known proteins. Comparison of this sequence to the rpr gene derived from Drosophila simulans shows that five out of five nucleotide differences between the sampled interval are silent changes in the third codon position. This observation confirms the biological relevance of the predicted reading frame and also serves to establish that the relevant gene product is, in fact, a protein rather than RNA. Despite extensive efforts, no point mutations in rpr have, as yet, been isolated [18]. The gene could either be poorly mutable (its small size could account for this) or there may be redundant cell death functions in the rpr region. The lack of single gene mutations in rpr have thus far hampered studies of the potential role of this locus during postembryonic cell deaths. Current evidence suggests that rpr function can be activated by very disparate cues to trigger a common apoptosis pathway and indications from early experiments with germline transformants suggest that expression of a rpr cDNA driven by the heat inducible promoter is sufficient to trigger extensive apoptotic deaths in transgenic DfH99 embryos [18]. Figure 8.1 shows a schematized version of a working model for how the rpr gene might be regulated during normal PCD and during ectopic cell death. For the sake of
158
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 8.1 Model for regulation of the rpr gene. For purposes of simplicity, this schematic diagram presumes that the observed in situ patterns of rpr RNA are mediated at the level of transcriptional control and envisions cis elements at the 5 end of the gene. The diagram illustrates the hypothetical assignment of a molecular convergence point for disparate cell death cues at the rpr gene. Signals for embryonic apoptosis and a substantial portion of induced or ectopic deaths operate through the rpr locus (thick arrows). A modest number of induced or ectopic deaths (e.g. after irradiation) can die via a rpr-independent pathway (thin arrow). Grouping of all forms of naturally-occurring and ectopic cell deaths under single headings is for convenience only—there are clearly many diverse signals which can provoke these pathways. A major question posed by this model is whether signals for embryonic deaths and those utilized during induced or ectopic apoptosis converge at, or prior to, the immediate regulation of rpr.
simplicity, this diagram presumes that a major factor governing the distribution of rpr RNA operates through transcription control elements. In the embryo, all naturally-occurring cell deaths appear to require activation of rpr function. Because individuals deleted for rpr are also substantially resistant to apoptosis induced by X-irradiation and developmental defects, a majority of ectopic deaths also appear to operate through a rpr-dependent pathway. We propose that this gene occupies a central switch position for converging information derived from distinct cell death cues. Some apoptotic deaths can still be induced in the absence of rpr (albeit at a much lower numbers) suggesting that a rpr-independent pathway can also be utilized to trigger programmed cell death in some contexts. The model proposes that signals associated with normal development provide relevant cues to activate rpr expression and trigger apoptosis in the embryo. Signals associated with cell injury (e.g. provoked by X-irradiation) and/or defective development can also activate rpr expression and may, in addition, activate an alternate cell death pathway which does not require rpr. The few cell deaths which can be induced in rpr− mutants after X-irradiation clearly display normal apoptotic features. We therefore suspect that rpr may not be part of the ‘apoptosis machinery’ itself, but instead may trigger a set of commonly used effector molecules [6]. One important question raised by this model is whether the signals utilized during ‘natural’ deaths and those utilized during ‘induced’ deaths converge at, or prior to, the immediate regulation of rpr. It is plausible to consider, for instance, that the rpr promoter region may define a molecular integration site for a variety of cell death signals. Alternatively, it is also possible that the integration of cues governing cell death occurs upstream of the rpr locus. 8.8 Other Candidate Cell Death Genes Drosophila The Drosophila rox8 gene (DmTIAR) is a homologue of human TIA-1-type nucleolysins that was identified a screen for molecules containing an RNA-recognition motif [92]. The nucleolysins are components of
CONTROL OF APOPTOSIS IN DROSOPHILA
159
cytolytic granules natural killer cells and cytotoxic T lymphocytes that contain RNA binding domains. These human proteins have been implicated apoptosis physiology by virtue of their ability to induce DNA fragmentation permeabilized cells [93]. It is not yet known whether the Drosophila nucleolysin-related protein shares similar apoptotic functions but a genetic analysis of this locus should resolve the issue. Polyubiquitin is another gene that may be involved during PCD Drosophila. Induction of this gene is associated with the death of some muscles and neurons Manduca and Drosophila suggesting that ubiquitindependent proteolysis may play a role during PCD [47, 94]. Genetic studies of this locus and the Drosophila ubiquitin-conjugating enzyme UbcD1 [95] could establish ubiquitin-mediated protein degradation as a potential effector pathway during apoptosis. 8.9 Engulfment of Apoptotic Cells Engulfment of dead cell corpses is the final stage of apoptotic death throughout the animal kingdom. This process may protect tissues from potential damage associated with exposure to the contents of dying cells [96] yet the mechanism(s) by which dying cells are recognized and ingested are not clear. In C. elegans, there are seven loci, falling into two mutant classes, that can influence the efficiency of engulfment [97]. Double mutants in representatives from both classes show the most severe impairments, suggesting that distinct, parallel pathways for the clearance of apoptotic cells may be utilized. Engulfment in the nematode does not rely on ‘professional phagocytes’ and, although ingestion is dispensable for cell killing, it is required for activation of an endonuclease function associated with the nuc1 gene [4]. In most animals, apoptotic debris is typically removed by ‘professional phagocytes’ or macrophages, although examples of engulfment by neighboring cells also occur [19, 27]. In some instances, macrophages appear to function in the killing process itself [98]. Studies from mammalian systems have implicated several cell surface proteins in this process, including the CD36 membrane protein and a phosphatidylserine receptor [96]. At least one Drosophila gene which shows significant homology to CD36 has been described [99]. The ‘professional phagocytes’ in Drosophila are referred to as macrophages or hemocytes. Although these cells are most frequently observed to engulf late stage corpses, they display a remarkable capacity to recognize and discriminate dying cells, even at very early stages of apoptosis [17]. Macrophages appear to be a homogeneous group of cells which derive from the mesoderm of the head prior to, and independently of, the first signs of PCD [100]. Because abundant cell death can be observed in mutations that lack macrophages, these cells are apparently not required for cell killing in the embryo [100]. There are no mutations yet described that show specific failures in the engulfment process. The predicted phenotype of such mutants would retain normal numbers of hemocytes which fail to phagocytose apoptotic cells. There are presently at least two plausible candidate genes that could be involved in this process. The class C scavenger receptor [101] and peroxidasin [102] are both specifically expressed in Drosophila macrophages and have been suggested to play roles in the recognition and processing of dying cells. Scavenger receptors endocytose a wide variety of polyanionic macromolecules [103]. These membrane proteins have been implicated in the onset of atherosclerosis because of their affinity for modified lipoproteins (LDL) such as oxidized or acetylated LDL (AcLDL) . In mammals, there are at least two classes of scavenger receptors (class A and class B/CD36) which exhibit distinct ligand binding profiles [104] and both have been implicated in the recognition of apoptotic cells [96, 101, 103]. One potential mechanism to account for the recognition of apoptotic cells involves phophatidylserine, a ligand for macrophage scavenger receptors [105] which is restricted to the inner monolayer in healthy cells. This
160
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
asymmetric distribution is lost during apoptotic death, and may be recognized on the surface by one or more receptors [106], including the macrophage scavenger receptors. Scavenger receptor activity was detected in Drosophila embryos that had been injected with a fluorescenated ligand (AcLDL) after gastrulation had occurred [101]. The modified lipoprotein was shown to accumulate specifically inside macrophages and can be used as a marker for these cells in live preparations. In cell culture studies, the Schneider L2 Drosophila cell line also tested positive for scavenger receptors while the Kc cell line did not [101]. The L2 cell receptor activity showed kinetic behavior and binding specificities that were remarkably similar to the mammalian class A receptors and, so far, one scavenger receptor has been cloned from these cells [107]. Mutations in this gene could permit rigorous analysis of the role of this receptor in the recognition and ingestion of apoptotic cells by macrophages. Another gene that has been proposed to play a role in the recognition of apoptosis by macrophages is peroxidasin, a unique heme peroxidase that appears to be deposited in the extracellular matrix [102]. This protein, originally referred to as protein X [17, 108], is synthesized by macrophages and may have both intracellular and extracellular oxidation activity [102]. These properties have led to the intriguing proposal that peroxidasin is anchored to basement membranes and could thus function to ‘mark’ damaged or dying cells that might secrete H2O2 or other peroxidase substrates [102]. Cells ‘marked’ in this way could be recognized for disposal by scavenger receptors [102]. This proposal is attractive because it relies upon distinct and multiple step-wise activities for the detection of damaged self which could impart both high selectivity and fidelity to the recognition process. 8.10 Conclusion Over the past several years, apoptosis research has witnessed spectacular growth. This expanding level of interest is also reflected among researchers who use Drosophila to dissect the genetic and molecular components that underlie biological processes. As has consistently turned out to be the case with other facets of developmental biology, the parallels between cell death in Drosophila and mammals are striking. The cytomorphological changes associated with death in both systems are indistinguishable from each other. In both Drosophila and mammals, apoptosis is regulated by hormones, is associated with DNA degradation, and occurs in reproducible patterns during development. Also, in both systems, excessive apoptosis occurs in association with defective development and is readily induced by exposure to X-irradiation. Studies in this model system permitted a description of the first gene product known (rpr) to specifically anticipate the death of a cell prior to any morphological manifestation of apoptosis [18] and, although our knowledge of apoptosis in Drosophila is still rudimentary, the prospects for continued progress are outstanding. Strong evidence argues for a high degree of evolutionary conservation among components required for PCD, and it is therefore likely that understanding apoptosis in Drosophila will have profound relevance to the understanding and treatment of various human pathologies. Genes homologous to rpr, for instance, could provide excellent targets for the discovery of therapeutic drugs that are intended to block or induce apoptosis in humans. Studies in Drosophila may also permit us to uncover potential connections between mechanisms of apoptosis and mechanisms of aging. The remarkable demonstration of extended life-span in transgenic flies engineered to overexpress antioxidant genes [109] is very intriguing in light of the proposed antioxidant effects of bcl-2 [110, 111], an apoptosis regulator in mammals.
CONTROL OF APOPTOSIS IN DROSOPHILA
161
References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
17 18 19 20 21 22 23
24
. SAUNDERS, J.W., JR. & FALLON, J.F. (1967) Cell death in morphogenesis. In: Major problems in developmental biology, LOCKE, M. (ed). Academic Press, New York, pp. 289–314. . HINCHCLIFFE, J.R. (1981) Cell death during embryogenesis. In: Cell death in biology and pathology, BOWEN, I.D. & LOCKSHIN, R.A. (eds). Chapman and Hall, London, pp. 35–78. . TRUMAN, J. (1984) Cell death in invertebrate nervous systems. Ann. Rev. Neurosci. 7, 171–188. . ELLIS, R.E., YUAN, J. & HORVITZ, H.R. (1991) Mechanisms and functions of cell death. Annu. Rev. Cell. Biol. 7, 663–698. . HURLE, J.M. (1988) Cell death in developing systems. Meth. Achiev. exp. Pathol. 13, 55–86. . RAFF, M.C. (1992) Social controls on survival and cell death. Nature 356, 397–400. . OPPENHEIM, R.W. (1991) Cell death during development of the nervous system. Ann. Rev. Neurosci. 14, 453–501. . TOMEI, L.D. & COPE, F.O. (1994) Apoptosis II: The molecular basis of apoptosis in disease. Cold Spring Harbor Laboratory Press, Cold Spring Harbor. . HENGARTNER, M.O. & HORVITZ, H.R. (1994) C. elegans cell survival gene ced-9 encodes a functional homolog of the mammalian proto-oncogene bcl-2. Cell 76, 665– 676. . YUAN, J., SHAHAM, S., LEDOUX, S., ELLIS, H.M. & HORVITZ, E.R. (1993) The C. elegans cell death gene ced-3 encodes a protein similar to mammalian Interleukin-1B-converting enzyme. Cell 75, 641–652. . SULSTON, J.E. (1976) Post-embryonic development in the ventral nerve cord of Caenorhabditis elegans. Philos. Trans. R. Soc. London Ser. 275, 287–298. . ELLIS, H.M. & HORVITZ, H.R. (1986) Genetic control of programmed cell death in the nematode C. elegans. Cell 44, 817–829. . YUAN, J. & HORVITZ, H.R. (1992) The Caenorhabditis elegans cell death gene ced-4 encodes a novel protein and is expressed during the period of extensive programmed cell death. Development 116, 309–320. . HENGARTNER, M.O., ELLIS, R.E. & HORVITZ, H.R. (1992) Caenorhabditis elegans gene ced-9 protects cells from programmed cell death . Nature 356, 494–499. . VAUX, D.L., WEISSMAN, I.L. & KIM, S.K. (1992) Prevention of programmed cell death in Ceanorhabditis elegans by human bcl-2. Science 258, 1955–1957. . MIURA, M., ZHU, H., ROTELLO, R., HARTWIG, E.A. & YUAN, J. (1993) Induction of apoptosis in fibroblasts by IL-1B-converting enzyme, a mammalian homolog of the C. elegans cell death gene ced-3. Cell 75, 653–650. . ABRAMS, J.M., WHITE, K., FESSLER, L. & STELLER, H. (1993) Programmed cell death during Drosophila embryogenesis. Development 111, 29–44. . WHITE, K., GRETHER, M., ABRAMS, J.M., YOUNG, L., FARRELL, K. & STELLER, H. (1994). Genetic control of programmed cell death in Drosophila. Science 264, 677–683. . WOLFF, T. & READY, D.F. (1991) Cell death in normal and rough eye mutants of Drosophila. Dev. Biol. 113, 825–839. . BONINI, N.M., LEISERSON, W.M. & BENZER, S. (1993) The eyes absent gene: genetic control of cell survival and differentiation in the developing Drosophila eye. Cell 72, 379–395. . KIMURA, K. & TANIMURA, T. (1992) Mutants with delayed cell death of the ptilinal head muscles in Drosophila. J. Neurogenet. 8, 57–69. . SPREIJ, T.E. (1971) Cell death during the development of the imaginal discs of Calliphora erythrocephala. Netherlands J. Zool. 21, 221–264. . WOLFF, T. & READY, D.F. (1993) Pattern formation in the Drosophila retina. In: The Development of Drosophila melanogaster, BATE, M. & ARIAS, A.M. (eds). Cold Spring Harbor Laboratory Press, Cold Spring Harbor, pp. 1277–1326. . GAVRIELI, Y., SHERMAN, Y. & BEN-SASSON, S.A. (1992) Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J. Cell. Biol. 119, 493–501.
162
25 26 27 28 29 30 31 32 33 34 35 36. 37
38 39 40 41 42 43 44 45 46 47 48
49
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. HAY, B., WOLFF, T. & RUBIN, G.M. (1994) Expression of the baculovirus p35 prevents cell death in Drosophila. Development 120, 2121–2129. . GIORGI, F. & DERI, P. (1976) Cell death in ovarian chambers of Drosophila melanogaster. J. Embryol. exp. Morph. 35, 521–533. . FRISTROM, D. (1969) Cellular degeneration in the production of some mutant phenotypes in Drosophila melanogaster. Molec. Gen. Genetics 103, 363–379. . FRISTROM, D. (1968) Cellular degeneration in wing development of the mutant vestigial of Drosophila melanogaster. J. Cell. Biol. 39, 488–491. . CAMPOS-ORTEGA, J.A. & HARTENSTEIN, V. (1985) The embryonic development of Drosophila melanogaster. Springer-Verlag; Berlin. . BODENSTEIN, D. (1950) The postembryonic development of Drosophila. In: Biology of Drosophila , DEMEREC (ed). John Wiley & Sons, Inc, New York, pp. 275–364. . JAMES, A.A. & BRYANT, P.J. (1981) Mutations causing pattern deficiencies and duplications in the imaginal wing disk of Drosophila melanogaster. Dev. Biol. 85, 39–54. . BUTTERWORTH, F.M. (1972) Adipose tissue of Drosophila melanogaster. V. Genetic and experimental studies of an extrinsic influence on the rate of cell death in the larval fat body. Dev. Biol. 28, 311–325. . KIMURA, K. & TRUMAN, J.W. (1990) Postmetamorphic cell death in the nervous and muscular systems of Drosophila melanogaster. J. Neurosci. 10, 403–411. . ASHBURNER, M. (1989) Drosophila: a laboratory manual. Cold Spring Harbor Laboratory Press, New York, pp. 161–167. . JAMES, A.A. & BRYANT, P.J. (1981) A quantitative study of cell death and mitotic inhibition in -irradiated imaginal wing discs of Drosophila melanogaster. Radiat. Res. 87, 552–564. ABBOT, L.A. (1983) Ultrastructure of cell death in Gamma- or X-irradiated imaginal wing discs of Drosophila. Radiat. Res. 96, 611–627. . WURGLER, F.E. & ULRICH, H. (1976) Radiosensitivity of embryonic stages. In: The genetics and biology of Drosophila, ASHBURNER, M. & NOVITSKI, E. (eds). Academic Press, London and New York, pp. 1269–1298. . ABRAMS, J.M. unpublished observations. . FRYXELL, K.J. & KUMAR, J.P. (1993) Characterization of the radiation sensitive stage in the development of the compound eye of Drosophila. Mutat. Res. 285, 181–189. . KONDO, S. (1988) Altruistic cell suicide in relation to radiation hormesis. Int. J. Radiat. Biol. 53, 95–102. . ASBURNER, M., CHIHARA, C., MELTZER, P. & RICHARDS, G. (1974) Temporal control of puffing activity in polytene chromosomes. Cold Spring Harbor Symp. Quant. Biol. 38, 655–662. . SEGRAVES, W.A. & HOGNESS, D.S. (1990). The E75 ecdysone-inducible gene responsible for the 75B early puff in Drosophila encodes two new members of the steroid receptor superfamily. Genes Dev. 4, 204–219. . TRUMAN, J. (1992) Insect systems for the study of programmed neuronal death. Experimental Gerontology 27, 17–28. . SCHWARTZ, L.M. & TRUMAN, J.W. (1983) Hormonal control of rates of metamorphic development in the tobacco horn worm, Manduca sexta. Dev. Biol. 99, 103–144. . SCHWARTZ, L.M. & TRUMAN, J.W. (1982) Peptide and steroid regulation of muscle degeneration in an insect. Science 1420–1421. . FARBACH, S.E. & TRUMAN, J.W. (1988) Cycloheximide inhibits ecdysteroidregulated neuronal death in the moth Manduca sexta. Society Neurosci. Abst. 14, 368. . SCHWARTZ, L.M. (1992) Insect muscle as a model for programmed cell death. J. Neurobiol. 23, 1312–1326. . KOELLE, M.R., TALBOT, W.S., SEGRAVES, W.A., BENDER, M.T., CHERBAS, P. & HOGNESS, D.S. (1991) The Drosophila EcR gene encodes an ecdysone receptor, a new member of the steroid receptor superfamily. Cell 67, 59–77. . TALBOT, W.S., SWYRYD, E.A. & HOGNESS, D.S. (1993) Drosophila tissues with different metamorphic responses to ecdysone express different ecdysone receptor isoforms. Cell 73, 1323–1337.
CONTROL OF APOPTOSIS IN DROSOPHILA
50
51 52 53
54 55
56 57
58 59 60 61 62 63
64 65 66 67 68
69 70
163
. RABINOW, S., TALBOT, W.S., HOGNESS, D.S. & TRUMAN, J.W. (1993) Programmed cell death in the Drosophila CNS is ecdysone-regulated and coupled with a specific ecdysone receptor isoform. Development 119, 1251–1259. . BRYANT, P. (1988) Localized cell death caused by mutations in a Drosophila gene coding for a transforming growth factor homolog. Dev. Biol. 128, 386–395. . MAGRASSI, L. & LAWRENCE, P.A. (1988) The pattern of cell death in fushi tarazu, a segmentation gene of Drosophila. Development 104, 447–451. . CHEYETTE, B.N.R., GREEN, P.J., MARTIN, K., GARREN, H., HARTENSTEIN, V. & ZIPURSKY, S.L. (1994) The Drosophila sine oculis locus encodes a homeodomain-containing protein required for the development of the entire visual system. Neuron 12, 977–996. . TEPASS, U., THERES, C. & KNUST, E. (1990) crumbs encodes an EGF-like protein expressed on apical membranes of Drosophila epithelial cells and required for organization of epithelia. Cell 61, 787–799. . HIGASHIJIMA, S., KOJIMA,T., MICHIUE, T., ISHIMARU, S., EMORI, Y. & SAIGO, K. (1992) Dual Bar homeo box genes of Drosophila required in two photoreceptor cells, R1 and R6, and primary pigment cells for normal eye development. Genes Dev. 6, 50–60. . SEDLAK, B.J. & MANZO, R. (1984) Localized cell death in Drosophila imaginal wing disc epithelium caused by the mutation apterous-blot. Dev. Biol. 104, 489–496. . COHEN, B., MCGUFFIN, M.E., PFEIFLE, C., SEGAL, D. & COHEN, S.M. (1992) apterous, a gene required for imaginal disc development in Drosophila encodes a member of the LIM family of developmental regulatory proteins. Genes Dev. 6, 715–729. . DEAROLF, C.R., HERPSPERGER, E. & SHEARN, A. (1988) Developmental consequences of awdb3, a cell autnomous lethal mutation in Drosophila induced by hybrid dysgenesis. Dev. Biol. 129, 159–168. . ZINYK, D.L., MCGONNIGAL, B.G. & DEAROLF, C.R. (1993) Drosophila awdkpn, a homologue of the metastasis suppressor gene nm23, suppresses the Tum-1 haematopoietic oncogene. Nature Genetics 4, 195–201. . BLOCHLINGER, K., JAN, L.Y. & JAN, Y.N. (1991) Transformation of sensory organ identity by ecoptic expression of Cut in Drosophila. Genes Dev. 5, 1124–1135. . PADGETT, R.W., ST. JOHNSTON, R.D. & GELBART, W.M. (1987) A transcript from a Drosophila pattern gene predicts a protein homologous to the transforming growth factor-beta family. Nature 325, 81–84. . KUROIWA, A., HAFEN, E. & GEHRING, W. (1984) Cloning and transcriptional analysis of the segmentation gene fushi tarazu of Drosophila. Cell 37, 825–831. . WEINER, A.J., SCOTT, M. & KAUFMAN, T. (1984) A molecular analysis of fushi tarazu, a gene in Drosophila melanogaster that encodes a product affecting embryonic segment number and cell fate. Cell 37, 843–851. . FERNANDO, J. & CAMPOS-ORTEGA, J.A. (1990) Defective neuroblast commitment in mutants of the achaetescute complex and adjacent genes of D. melanogaster. Neuron 5, 81–89. . VILLARES, R. & CABRERA, C.V. (1987) The achaete-scute gene complex of D. melanogaster: conserved domains in a subset of genes required for neurogenesis and their homolgy to myc. Cell 50, 415–424. . FISCHBACH, K.F. & TECHNAU, G. (1984) Cell Degeneration in the developing optic lobes of the sine oculis and small optic lobes mutants of Drosophila melanogaster. Dev. Biol. 104, 219–239. . SEDLAK, B.J. (1986) Spatial and temporal relationships between cell death and tissue overgrowth in imaginal wing discs of the Drosophila mutant l(3)c43hs1. Dev. Genet. 6, 199–212. . AGNEL, M., RODER, L., VOLA, C. & GRIFFEN-SHAE, R. (1992) A Drosophila rotund transcript expressed during spermatogenesis and imaginal disc morphogenesis encodes a protein which is similar to human Rac GTPaseactivating (racGAP) proteins. Mol. Cell. Biol. 12, 5111–5122. . PETSCHEK, J.P. & MAHOWALD, A.P. (1990) Different requirements for 1(1) giant in two embryonic domains of Drosophila melanogaster. Dev. Genet. 11, 88–96. . CLIFFORD, R. & SCHUPBACH, T. (1992) The torpedo (DER) receptor tyrosine kinase is required at multiple times during Drosophila embryogenesis. Development 115, 853– 872.
164
71 72 73 74 75
76 77 78 79 80 81 82
83 84 85
86 87 88 89 90 91 92
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. BAKER, N.E. & RUBIN, G.M. (1992) Ellipse mutations in the Drosophila homologue of the EGF receptor affect pattern formation, cell division and cell death in eye imaginal discs. Dev. Biol. 150, 381–396. . TEPASS, U. & KNUST, E. (1993) Crumbs and stardust act in a genetic pathway that controls the organization of the epithelia in Drosophila melanogaster. Dev. Biol. 159, 311–326. . HEBERLEIN, U., HARIHARAN, I.K. & RUBIN, G.M. (1993) Star is required for neuronal differentiation in the Drosophila retina and displays dosage-sensitive interactions with Ras1 . Dev. Biol. 160, 51–63. . DAWSON, I.A., ROTH, S., AKAM, M. & ARTAVANIS-TSAKONAS, S. (1993) Mutations of the fizzy locus cause metaphase arrest in Drosophila melanogaster embryos. Development 117, 359–376. . DURA, J., RANDSHOLT, N.B., DEATRICK, J., ERK, L, SANTAMARARIA, P., FREMAN, J.D., FREEMAN, S.J., WEDDELL, D. & BROCK, H.W. (1987) A complex genetic locus, polyhomeotic, is required for segmental specification and epidermal development in D. melanogaster. Cell 51, 829–839. . MARTINEZ-ARIAS, A. (198 5) The development of fused embryos of Drosophila melanogaster. J. Embryol Exp. Morphol. 87, 99–114. . KARESS, R.E. & GLOVER, D.M. (1989) rough deal: a gene required for proper mitotic segregation in Drosophila. J. Cell Biol. 109, 2951–2961. . MURPHY, C. (1974) Cell death and autonomous gene action in lethals affecting imaginal discs in Drosophila melanogaster. Dev. Biol. 39, 23–36. . WHITE, E. (1993) Death defying acts: a meeting review on apoptosis. Genes Dev. 7, 2277–2284. . BUTTERWORTH, F.M. & KING, R.C. (1965) The developmental genetics of apterous mutants in Drosophila melanogaster. Genetics 52, 1153–1174. . CAGAN, R.L. & READY, D.F. (1989) Notch is required for successive cell decisions in the developing Drosophila retina. Genes Dev. 3, 1099–1112. . RAMOS, R.G., IGLOI, G.L., LICHTE, B., BAUMANN, U., MAIER, D., SCHNEIDER, T., BRANDSTATTER, J.H., FROHLICH, A. and FISCHBACH, K.F. (1993) The irregular chiasm C-roughest locus of Drosophila, which affects axonal projections and programmed cell death, encodes a novel immunoglobulin-like protein. Genes Dev. 7, 2533–2547. . CLAUSTON, W.M. & KERR, J.F.R. (1985) Apoptosis, lymphocytotoxicity and the containment of viral infections . Med. Hypoth. 18, 399–404. . CLEM, R.J., FECHHEIMER, M. & MILLER, L.K. (1991) Prevention of apoptosis by a baculovirus gene during infection of insect cells. Science 254, 1388–1390. . SUGIMOTO, A., FRIESEN, P.D. & ROTHMAN, J.H. (1994) Baculovirus p35 prevents developmentally programmed cell death and rescues a ced-9 mutant in the nematode Caenorhabditis elegans. EMBO J. 13, 2023–2028. . RABIZADEH, S., LACOUNT, D.J., FRIESEN, P.D. & BREDESEN, D.E. (1993) Expression of the baculovirus p35 gene inhibits mammalian neuronal cell death. J. Neurochem. 61, 2318–2321. . HERSHBERGER, P.A., LACOUNT, D.J. & FRIESEN, P.D. (1994) The apoptotic suppressor p35 is required early during baculovirus replication and is targeted to the cytosol of infected cells. J. Virol. 68, 3467–3477. . BRAY, S.J., BURKE, B., BROWN, N.H. & HIRSH, J. (1989) Embryonic expression pattern of family of Drosophila proteins that interact with a central nervous system regulatory element . Genes Dev. 3, 1130–1145. . GAUL, U., SEIFERT, E., SCHUH, R. & JACKLE, H. (1987) Analysis of Krupple protein distribution during early Drosophila development reveals posttranscriptional regulation. Cell 50, 639. . CLARKE, A.R., PURDIE, C.A., HARRISON, D.J., MORRIS, R.J., BIRD, C.C., HOOPER, M.L. & WYLIE, A.H. (1993) Thymocyte apoptosis induced by p53-dependent and independent pathways. Nature 362, 849–852. . LOWE, S.W., SCHMITT, E.M., SMITH, S.W., OSBORNE, B.A. & JACKS, T. (1993) p53 is required for radiation-induced apoptosis in mouse thymocytes. Nature 362, 847–849. . BRAND, S. & BOURBON, H. (1993) The developmentally-regulated Drosophila gene rox8 encodes an RRMtype RNA binding protein structurally related to human T1A-1-type nucleolysins. Nucleic Acids Res. 21, 3699–3704.
CONTROL OF APOPTOSIS IN DROSOPHILA
93
165
. TIAN, Q., STREULI, M., SAITO, H., SCHLOSSMAN, S.F. & ANDERSON, P. (1991) A polyadenylate binding protein localized to the granules of cytolytic lymphocytes induced DNA fragmentation in target cells. Cell 67, 629–639. 94 . SCHWARTZ, L.M., MYER, A., KOSZ, L., ENGELSTEIN, M. & MAIER, C. (1990) Activation of polyubiquitin gene expression during developmentally programmed cell death. Neuron 5, 411–419. 95 . TREIER, M., SEUFERT, W. & JENTSCH, S. (1992) Drosophila UbcD1 encodes a highly conserved ubiquitinconjugating enzyme involved in selective protein degradation. EMBO J. 11, 367–372. 96 . SAVILL, J., FADOK, V., HENSON, P. & HASLETT, C. (1993) Phagocyte recognition of cells undergoing apoptosis. Immunology Today 14, 131–136. 97 . ELLIS, R.E., JACOBSON, D.M. & HORVITZ, H.R. (1991) Genes required for the engulfment of cell corpses during programmed cell death in Caenorhabditis elegans. Genetics 129, 79–94. 98 . LANG, R.A. & BISHOP, J.M. (1993) Macrophages are required for cell death and tissue remodeling in the developing mouse eye. Cell 74, 453–462. 99 . HART, K. & WILCOX, M. (1993) A Drosophila gene encoding an epithelial membrane protein with homolgy to CD36/LIMP II. J. Mol Biol. 234, 249–253. 100 . TEPASS, U., FESSLER, L.L, AZIZ, A. & HARTENSTEIN, V. (1994) Embryonic origin of hemocytes and their relationship to cell death in Drosophila. Development 120, 1829–1837. 101 . ABRAMS, J.M., LUX, A., STELLER, H. & KRIEGER, M. (1992) Macrophages in Drosophila embryos and L2 cells exhibit scavenger receptor-mediated endocytosis. Proc. Natl Acad. Sci. USA 89, 10375–10379. 102 . NELSON, R.E., FESSLER, L.I., TAKGI, Y., BLUMBERG, B., KEENE, D.R., OLSON, P.F., PARKER, C.G. & FESSLER, J.H. (1994) Peroxidasin: a novel enzyme-matrix protein of Drosophila development. EMBO J. 13, 3438–3447. 103 . KRIEGER, M. & HERZ, J. (1994) Structures and functions of multiligand lipoprotein receptors: Macrophage scavenger receptors and LDL receptor-related protein (LRP). Annu. Rev. Biochem. 63, 601–637. 104 . ACTON, S.L., SCHERER, P., LODISH, H. & KRIEGER, M. (1994) Expression cloning of SR-B1, a CD36related class B scavenger receptor. J. Biol. Chem. 269, 21003– 21009. 105 . NISHIKAWA, K., ARAI, H. & INOUE, K. (1990) Scavenger receptor-mediated uptake and metabolism of lipid vesicles containing acidic phospholipids by mouse peritoneal macrophages. J. Biol Chem. 265, 5226–5231. 106 . FADOK, V.A., VOELKER, D.R., CAMPBELL, P.A., COHEN, J.J., BRATTON, D.L. & HENSON, P.M. (1992) Exposure of phosphatidyl serine on the surface of lymphocytes triggers specific recognition and removal by macrophages. J. Immunol 148, 2207–2216. 107 . PEARSON, A., LUX, A. & KRIEGER, M. (1995) Expression cloning of dSR-CI, a class C macrophagespecific scavenger receptor from Drosophila melanogaster. Proc. Natl Acad. Sci. USA 92, 4056–4060. 108. FESSLER, J.H. & FESSLER, L.I. (1989) Drosophila extracellular matrix. Ann Rev. Cell Biol. 5, 309–339. 109 . ORR, W.C. & SOHAL, R.S. (1994) Extension of life-span by overexpression of superoxide dismutase and catalase in Drosophila melanogaster. Science 263, 1128–1130. 110 . KANE, D.J., SARAFIAN, T.A., ANTON, R., HAHN, H., GRALLA, E., VALENTINE, J., ORD, T. & BREDESEN, D.E. (1993) Bcl-2 inhibition of neural death: decreased generation of reactive oxygen species. Science 262, 1274–1276. 111 . HOCKENBERY, D.M., OLTVAI, Z.N., YIN, X., MILLIMAN, C.L. & KORSMEYER, S. (1993) Bcl-2 functions in an anitoxidant pathway to prevent apoptosis. Cell 75, 241–251.
9 Evolution of Apoptosis DAVID L.VAUX The Walter and Eliza Hall Institute of Medical Research, Post Office Royal Melbourne Hospital 3050, Australia
9.1 Introduction Cell death occurring as a normal part of physiology has been observed in every multicellular organism in which investigators have bothered to look. This raises the question of how it evolved, and what it evolved to do. This chapter will attempt to answer these questions, but the answers are incomplete, and are largely speculative. However, while they can’t explain all the observed cell death phenomena, they are consistent with most of it, and will hopefully encourage the experiments required to support or disprove them. Evolution is driven by natural selection; any process that has persisted, including cell death, must have either aided the organism directly, helped it compete with others, or promoted the replication of its genes. Physiological cell death, the killing of a cell by itself or by another of the organism’s cells, will be examined from this perspective. 9.2 Observations of Cell Death in Different Organisms Physiological cell death, in which there is no evidence of physical injury or the presence of biological pathogens, has been recorded in many different forms of life [1, 2]. During animal development cell death is used as an aid to morphogenesis. For example, the hands and feet of humans develop as round, paddle-like structures, and the cells that make up the webbing between the fingers die, allowing them to separate. Cell death is seen in other vertebrates, such as during metamorphosis in amphibia. The tadpole’s tail is removed by programmed cell death [3]. In the earliest protochordates, such as the tunicate Botryllus, cycles of both sexual and asexual reproduction involve cell death [4]. After the asexual budding of daughter colonies, the parents’ generation synchronously die. Physiological cell death has been well described in insects [5, 6] and nematodes [1]. Cell death is also seen in many different types of plants. In trees the water conducting xylem is composed of the walls of dead cells. In flowering plants cell death is seen in the ovary and amongst pollen [7]. In Volvox cateri, an aquatic alga that in some ways resembles a spherical colony of flagellated single-celled
EVOLUTION OF APOPTOSIS
167
organisms, rather than a multicellular plant, somatic cells of one generation die to allow release of daughter colonies [8]. Whether physiological cell death occurs in single-celled organisms is less certain. There are, however, a few anecdotal reports of cell death, or degradation of genetic material, in a variety of protozoa, yeast and prokaryotes. Physiological cell death is seen in the stalk of the slime mold, Dictyostelium, but whether this is really a single celled organism is debatable. Dictyostelium exists in two forms, as free living individual amoeba capable of vegetative growth, or, when food is scarce, as a multicellular slug that gives rise to a stalk supporting a mass of spores. The stalk is composed of dead, highly vacuolated cells [9]. Whether the stalk cells ‘age’ or apoptose has not been determined, but their act is altruistic, as they enable the spores, which are the only cells that can produce progeny, to disperse more efficiently. In ciliates the DNA exists in two forms of nucleus. While the micronucleus serves as the germline nucleus, it does not express its genes. These are copied and rearranged and expressed from the macronucleus to allow for vegetative growth. Mating cells exchange haploid micronuclei, and a new macronucleus develops from a new diploid micronucleus [10]. The old macronucleus is destroyed by endonucleases, in a manner somewhat reminiscent of the DNA degradation commonly seen in apoptosis. It has been suggested that even bacteria may sometimes commit suicide. Escherichia coli bearing a genetic element designated e14 appear to activate a protease in response to infection with T4 phage that cleaves the essential translational elongation factor EF-Tu, causing death of the bacterium [11]. This may represent altruistic cell suicide. 9.3 Natural Selection of Genes for Cell Death The study of cell death has been neglected compared to study of cell division. Perhaps one reason for this is that the idea of cells killing themselves seems counter-intuitive, because most of us are brought up with a Darwinian mindset: that everything has been selected to maximize survival and proliferation. However, this narrow view of selfish cells and selfish organisms is not consistent with the occurrence of altruism. Any form of self-inflicted death, whether it is cell death or death of an organism, is the ultimate in altruistic behavior. Richard Dawkins has explained that if the unit of selection that is given greatest priority is the gene, rather than the individual, it is easy to justify and explain altruistic behavior. His theory of the ‘selfish gene’ suggests that as long as some copies of a gene can persist and replicate, it does not matter if other copies are lost [12]. Any behavior of a bird that promotes the survival of copies of its genes carried by its chicks will be positively selected, even if it includes risky behavior such as acting as a decoy to lure away the fox. The hypothetical genes encoding this altruistic behavior would thus be able to ‘selfishly’ persist in the chicks. If genes are really what drives the altruistic suicide of an organism in order to protect other copies of themselves, then altruistic suicide ought only occur for the benefit of genetically related individuals. On one hand, there ought to be only minimal altruism for members of another species, as different species have no way of sharing their genes. The more closely related the individuals, the more exact copies of genes would be shared, and the greater the gain for the selfish gene. Among communal insects, the hymenoptera, which include ants, bees and wasps, the haploid-diploid pattern of inheritance means that the only way genes borne by a non-reproductive worker can be propagated is if she directs her efforts to the good of the queen and drones. Altruistic behavior, even to the point of suicide after delivery of a bee sting, for example, is very common amongst these insects.
168
APOSTOSIS IN NORMAL DEVELOPMENT AND CANCER
A multicellular organism can be thought of as a collection of ultra-communal individuals, each of whom share exactly the same DNA. In this case there is absolutely no genetic penalty for cell suicide, as long as the genes can be passed on by the germ cells. From a genetic point of view we should not be surprised by cell death; we should expect it. In the case of single-celled organisms, in which there is no distinction between the germ line and somatic line, the genetic penalties for cell death are very real and substantive. Unicellular organisms are immortal in the sense that they are able to multiply indefinitely through successive generations, and they may not age and senesce the way somatic cells do in metazoans, but although immortal and ageless, they may at times be suicidal. It is interesting to speculate that death due to ageing may have evolved in early metazoans, but altruistic cell death evolved earlier, in single-celled organisms. Whether the benefits of altruistic suicide by a single-celled organism for its relatives out-weigh the genetic penalties may need to be settled by observation. Perhaps the evolutionary step to multicellularity with concomitant cellular specialization into immortal germ and mortal somatic cell lines also allowed altruistic cell death to become a fully utilized strategy. 9.4 The Mechanism of Cell Death Genetic studies in Caenorhabditis elegans have told us a lot about the molecular participants in apoptosis [1]. The ability of mammalian and insect cell death genes to function in the nematode [13, 14] and insect genes to function in nematode and mammalian cells [15] shows that the mechanism for cell death is strongly conserved, and findings in different phyla of animals are likely to be generally applicable. A fertilized C. elegans ovum takes about 14 h to hatch, and is able to bear its own young at about 2 days of age. During development of the hermaphrodite exactly 1090 nuclei are formed, but the physiological fate of 131 of the cells is to die [1]. Exquisite knowledge of the pattern of development allowed the examination of mutant worms in which the process of cell death was abnormal. The cell death abnormal (ced) genes thus identified fell into several groups. Some ced genes encoded products that were not strictly necessary for cell death itself, but were important for disposal of the cellular corpses. Cell corpses tend to accumulate in worms with mutations to these genes, rather than being quickly engulfed and degraded by neighboring cells [16]. Three genes appeared to be key players in the process of cell death. Mutation of ced-3 or ced-4 prevents death of the 131 cells, so they presumably encode ‘killer’ proteins [16]. Another gene, ced-9, encodes a protein with the opposite effect: worms bearing overactive ced-9 alleles also have the extra cells, so ced-9 protects against cell death [17, 18]. Interestingly, preventing cell death in C. elegans by mutating ced-3 or ced-4 does not alter the lifespan of the worm. It lives to its normal age, but dies with an extra 131 cells. This suggests that the mechanisms regulating senescence and programmed cell death in metazoans are different, and may have evolved separately. There is abundant evidence that the effector mechanisms of apoptosis have been highly conserved, as several cell death genes function in organisms as different as nematodes, insects and mammals. The cell death blocking protein ced-9 has structural similarity to the product of the mammalian oncogene bcl-2 [18], and bcl-2 is able to prevent cell death in C. elegans [13], so the two genes are related, and their function is highly conserved. For programmed cell death to occur in worms two genes are required, ced-3 and ced-4 [16]. Presumably ced-9 (and to a lesser extent bcl-2) can either prevent activation of ced-3 and ced-4 or prevent them from interacting with their targets. Exactly how they achieve this, and whether it is direct or indirect, has not been determined. A product of an insect virus designated p35 that inhibits apoptosis in insect cells can also rescue ced-9 mutant worms, and can prevent apoptosis of mammalian cells [15].
EVOLUTION OF APOPTOSIS
169
The ‘killer’ protein ced-3 resembles products of at least two mammalian genes, interleukin-1converting enzyme (ICE) [19] and nedd2 [20], a gene first isolated from a day 10 mouse neuro-epithelial library [21]. Other homologs are likely to be identified over the next few years. ICE is a cysteine protease that cleaves its substrates at aspartate residues [22]. The motif surrounding the active cysteine residue is shared by ced-3 and nedd2, suggesting they act in the same manner [19, 20]. Sequence comparisons between ICE and serine proteases suggest that they may be related, as ICE bears a serine residue at a site that corresponds to the active serine in serine proteases [23]. Crystallography has revealed that ICE probably functions as a tetramer consisting of two p20 and two p10 subunits [24]. The quaternary structure of ICE is distinct from other cysteine proteases, and its pattern of sensitivity to protease inhibitors is also unique [23]. As expression of ICE can cause apoptosis, and this can be inhibited by specific ICE protease inhibitors, one step in the process of apoptosis must be cleavage by ICE of some intracellular substrate [25]. The identity of the substrate(s) of these cell death cysteine proteases that lead to cellular collapse are not yet known. While ICE can cleave pro-IL-1 to yield the mature form, pro-IL-1 is unlikely to be the only substrate of ICE, and it is unlikely that cell death is mediated via production of mature IL-1 . Despite the power of genetic approaches in C. elegans, the gene(s) encoding the substrate(s) of ced-3 have not been found. One possible explanation for this is that the substrate may be an essential component of every cell. If both copies of this gene were mutated, no development at all would be possible, and the animal would not appear as a cell death abnormal (ced) mutant. The seminal discovery that ICE was a homolog of ced-3 and could also mediate apoptosis was surprising for several reasons, and raised many questions [19]. First of all, is ICE an apoptosis gene, and is cell death mediated through mature IL-1 ? Yuan and coworkers showed that transient transfection of ICE into rat fibroblasts caused apoptosis, but overexpression of almost everything has been shown to cause apoptosis in one cell or another [25]. Importantly, they showed that cell death could be blocked by specific inhibitors of ICE, and that mutations to the reactive cysteine blocked its activity. These experiments proved that ICE is an apoptosis gene in mammalian cells. It is unlikely that cell death is mediated by IL-1 for several reasons. Firstly, cells have been transfected with truncated constructs that encode the mature form of IL-1 , and these cells secrete bioactive IL-1 but are not killed by it [26]. Secondly, the known activities of IL-1 include acting as an ‘alarm’ cytokine that activates lymphocytes and acts as a chemo-attractant. IL-1 functions as endogenous pyrogen during infections to increase temperature and blood flow. It does not commonly cause cells to undergo apoptosis [27]. It is more likely that there is another substrate of ICE whose cleavage results in cellular collapse. So what is the role of IL-1 , and why is it activated by ICE, a cell death protein? One possibility is that pro-IL-1 sits in the cell until the cell activates ICE to undergo apoptosis, in response to a virus, for example. Pro-IL-1 will then be cleaved to release bioactive IL-1 which diffuses from the dying cell. The ability of pro-IL-1 to act as a substrate for ICE may have occurred as an evolutionary afterthought, to alert neighboring cells and immunocytes with IL-1 receptors that a cell is undergoing apoptosis, and the organism may be under viral attack. The idea that a cell undergoing apoptosis may activate the immune system, promoting inflammation, is contrary to the usual view of apoptosis [28], in which scarring rarely occurs and phagocytosis of apoptotic bodies is performed by neighboring cells, rather than professional phagocytes [29]. Perhaps ICE is used to mediate apoptosis in situations of viral infection, but other homologs of ced-3 such as nedd2 being used more for apoptosis during development and normal homeostasis, when inflammation is not required. The mechanisms that implement cell death in plants are largely unknown, although the identification of genes that regulate the hypersensitivity response may give a lead [30] (see below). Whether plant cells degrade their DNA during cell death, or whether bcl-2-like genes can function in plants has not been
170
APOSTOSIS IN NORMAL DEVELOPMENT AND CANCER
determined. If bcl-2 can influence the death of plant cells it would indicate that the mechanism bcl-2 controls evolve in a presumably single-celled common ancestor of plants and animals. 9.5 Uses for Cell Death To understand the evolution of a characteristic, it is necessary to know how it is useful to the organism; how it allows it to survive and reproduce better. Of course, this is not to exclude the possibility that some characteristic may evolve for one reason, but then get used for another. In mammals, cell death is commonly used for morphogenesis during development, to maintain appropriate cell numbers, and as a defence mechanism. During embryogenesis cell death is used to sculpt the body. In the brain it is used to remove neurons that have not made contact with target cells that can provide factors to give a survival signal. In animals where the sexes develop differently, cell death is used to remove tissues that are not needed in that particular sex. For example, breast tissue that starts to develop in male mice regresses by cell death under control of testosterone. In amphibia and insects, cell death is used during metamorphosis. The set of tissues and organs required for the larval existence are different to those required as an adult, and death of the cells in the earlier tissues allows nutrients to be recycled for later use. The giant muscle cells in the larvae of the tobacco hawkmoth, Manduca sexta, for example, are programmed to die under control of the molting hormone ecdysone, and the muscle proteins are degraded for later use in the adult moth by ubiquitin-mediated proteolysis [6, 31]. In animals with relatively long lives, in which the life of the organism is considerably longer than the lifespan of most of its cells, cells that have performed their tasks are not left to linger or senesce but kill themselves. In the skin and the gut, for example, the epithelial cells are at constant risk from environmental factors such as mutagens in the diet and UV irradiation. New cells are constantly produced, and the old ones apoptose [32]. In the mammalian immune system cells are produced with randomly assorted antigen receptors. To be prepared for foreign antigens cells with novel specificities must be produced, but the old cells must be removed to prevent overpopulation. Similarly, cells with unreactive antigen receptors can be disposed of, as can the majority of immune cells after the infection is dealt with. The liver is an example of an organ that can hypertrophy in response to increasing demand, but regresses when the demand decreases. Although some growth is cellular hypertrophy, most is due to the production of new cells. Regression occurs by apoptosis [33, 34]. Plants are under constant threat from micro-organisms. Because they cannot flee their attackers, their defence mechanisms are more limited than those of animals. It is well known that one disease resistance mechanisms is production of toxins, but cell death is also used by plants [35]. In what is known as the ‘hypersensitivity response’ a plant of a particular genotype recognizes a pathogen bearing certain markers, and the cells at the site of contact die and dessicate. Furthermore, cells nearby that may not be in direct contact with the pathogen are given a signal from neighboring cells so that they too succumb to form a necrotic lesion. The pathogen is left on dead cells that provide a poorer environment than viable cells. Although the hypersensitivity response requires death of some plant tissue, this loss is outweighed by the protection gained by the plant against spread of the pathogen. A mutation to a negative regulator of this cell death response in Arabidopsis called acd2 results in ‘hyper-hyper-sensitive’ plants that inherit the tendency to spontaneously form lesions in the absence of pathogens [30]. The mechanism of the response may involve many more genes, as mutations at at least four other loci also result in this phenotype of
EVOLUTION OF APOPTOSIS
171
spontaneous lesions, but increased resistance [36]. Significantly, these mutants show even greater resistance to pathogens than the wild type. Not only plants use cell death in responding to pathogens. Cells from insects and vertebrates undergo apoptosis when exposed to certain viruses. For example, insect cells undergo apoptosis when infected with baculovirus. That apoptosis is a defensive response to the virus, rather than an effect triggered in the cell by the virus to help it replicate, is shown by the ability of apoptosis to reduce viral replication and infectivity [37, 38]. Chicken cells apoptose when infected with chicken anemia virus [39], as do rodent neuronal cells when infected with Sindbis virus [40, 41]. Cell death may also be used as protection against cells bearing mutated or abnormal DNA in a process controlled by p53 [42]. p53 is a tumor suppressor gene that can activate apoptosis [43]. The importance of p53 is shown by the early and frequent onset of tumors in mice or humans lacking one or both p53 alleles [44]. The role of p53 protein is to detect abnormal or unreplicated DNA, and to halt the cell cycle, allowing the DNA to be repaired. p53 may alternatively send the cell to its death by apoptosis. Rapid interphase apoptosis of thymocytes due to irradiation is mediated by p53, as thymocytes from p53 null mice are insensitive to irradiation, but retain their ability to undergo apoptosis triggered in other ways, such as by dexamethasone [45, 46]. While the activity of p53 may protect against our own mutated cells, it may also protect against virus infection, as p53 may be activated by the unusual structure of viral nucleic acid. 9.6 Apoptosis and Necrosis Although it is clear that apoptosis has frequently been used as a defence mechanism, it is not known how cells detect the presence of an infecting virus particle. In some cases abnormal nucleic acid may be the sign. Perhaps single stranded DNA or double stranded RNA results in the activation of p53. Sometimes viruses need to use the host cell’s DNA polymerase to copy their own genome. Inappropriate activation of replication machinery may also alert cells to the presence of a virus. Cell may monitor several metabolic parameters, and may be pre-programmed to undergo apoptosis if they detect a change, under the assumption that they have been infected. Clues as to the mechanism by which a cell becomes aware of an infecting virion may come from studying cases in which a single viral gene has been shown to be necessary for the apoptotic response. For example, for apoptosis in response to Chicken Anemia Virus, the VP3 protein must be produced [47]. Inappropriate implementation of the cell death program in response to a stimulus wrongly interpreted by the cell to be due to a virus may explain why so many physical and pharmacological agents have been found to trigger apoptosis [48]. It may also explain why treatment of cells with cytotoxic agents that would be expected to cause necrosis instead induce changes associated with cell suicide. The most classical model of necrosis is that seen after interruption to the blood supply, for example following ligature of an artery. Large numbers of adjoining cells are affected; resolution is by inflammatory infiltrate of professional phagocytic cells and there is scarring and disruption to the normal architecture [49]. Loss of oxygen is clearly not a physiological occurrence and mammalian cells have an absolute requirement for it to survive. Despite this, in some cases apoptosis, rather than necrosis is observed in some instances of anoxia [50], and in other circumstances that lead to reduction of ATP such as when cells are treated with sodium azide [51, 52]. The fact that forced expression of bcl-2 in some cells can delay their death, and prevent changes characteristic of apoptosis, even though it may not be able to prevent the eventual death of the cell, suggests that some cells may respond to the early changes caused by oxygen deprivation
172
APOSTOSIS IN NORMAL DEVELOPMENT AND CANCER
by activating their cell suicide machinery. These azide-treated cells may commit suicide before they are killed. The importance of these observations is that more cells may die in response to anoxia than need die. After a thrombosis leading to a heart attack or stroke, for example, heart muscle or neuronal cells may kill themselves even after receiving a sub-lethal amount of anoxia. A pharmaceutical agent that could block cell suicide may thus limit the damage following such vascular accidents. 9.7 Virus Defence Against Apoptosis As activation of proteases and endonuclease by a host cell is likely to be effective in limiting virus production, viruses are under selective pressure to counter defensive apoptosis. Many viruses carry genes whose products interact with parts of the apoptosis mechanism. Anti-apoptosis genes have been found in viruses that infect mammalian and insect cells. Epstein-Barr Virus and African Swine Fever Virus carry genes that resemble bcl-2, both structurally and functionally [53, 54]. Human Papilloma Virus, Adenovirus, SV40 virus and Polyoma Virus carry genes that can degrade, sequester or otherwise inactivate p53 [55, 56]. Cowpox virus carries a gene, crmA, that can specifically bind to and inhibit ICE [57]. Several insect viruses carry genes that can block apoptosis [58, 59]. These have a variety of different structures and may act in several different ways [60], but they are likely to interact with components of the same apoptotic pathway, for example, p35 can rescue C. elegans ced-9 mutants [15]. 9.8 The Role of Cytotoxic Cells The fact that many viruses carry genes that allow them to block defensive apoptosis of the host cell, has left the organism with a problem. How can it defend itself against an intracellular pathogen if the autonomous cell death mechanisms are blocked? The solution adopted by vertebrates is to use cell mediated cytotoxicity to murder the infected cell, since it may not be able to kill itself. In mammals the main defence against viruses is provided by cytotoxic T lymphocytes (CTL) and natural killer (NK) cells. These cells can identify infected cells and kill them, thus preventing the virus using the cell’s synthetic machinery to replicate. CTL recognize processed viral peptides presented by MHC molecules on the surface of the infected cell. They attach to these target cells and kill them without killing themselves, so they can continue to look for other targets. Cells killed by CTL die exhibiting all the characteristic features of apoptosis [61, 62]. DNA degradation occurs rapidly, usually prior to any significant leakage of cellular contents into the media [63]. CTL kill by transferring the contents of enzyme-bearing granules through pores in the surface of the target cell formed with perforin. One of the most important contents of the granules is a serine pro tease called granzyme B [64], as cytotoxic function is severely diminished in both granzyme B or perforin knockout mice [65, 66]. Granzyme B, which is also known as fragmentin because of its ability to cause rapid DNA degradation in cells when added together with perforin [67], is a protease that resembles ICE as it cleaves at aspartate residues, and can be blocked by modified peptides that closely resemble the inhibitors of ICE [23, 64]. The identical appearance of apoptosis mediated by CTL and cell autonomously by homologs of ced-3, and the fact that both ICE and granzyme B are Asp-ases, suggests that they may cleave the same intracellular substrates. Significantly, unlike apoptosis mediated by ICE, apoptosis caused by granzyme B or CTL cannot be blocked by bcl-2 [68] or, presumably, by viruses carrying bcl-2-like genes. Thus CTL are likely to be effective against viruses that carry anti-apoptosis genes.
EVOLUTION OF APOPTOSIS
173
9.9 Conclusions Physiological cell death is an ancient process whose mechanism is probably shared by all metazoans. Plants also commonly use cell death. Whether they use the same mechanism has not been determined, but if this proves to be the case, then the mechanism for cell death must have evolved from a presumably single celled common ancestor. Altruistic cell death has not been convincingly described among single celled organisms, but as long as surviving relatives bore copies of the cell death genes, it is not precluded. Understanding the evolution of cell death is helped by considering that natural selection operates on genes rather than whole organisms. The use of apoptosis as a defence mechanism against intracellular pathogens has pressured viruses to develop anti-apoptosis genes. Studying these is likely to provide more clues to the components of this fascinating process. Acknowledgments This work was supported by the Cancer Research Institute of New York and the Anti-Cancer Council of Victoria. References 1 2 3
4 5 6 7 8 9 10 11 12 13 14
15
. ELLIS, R.E., YUAN, J.Y. & HORVITZ, H.R. (1991) Mechanisms and functions of cell death. Annu. Rev. Cell Biol. 7, 663–698. . VAUX, D.L. (1993) Toward an understanding of the molecular mechanisms of physiological cell death. Proc. Natl Acad. Sci USA 90, 786–789. . KERR, J.F., HARMON, B. & SEARLE, J. (1974) An electron-microscope study of cell deletion in the anuran tadpole tail during spontaneous metamorphosis with special reference to apoptosis of striated muscle fibers. J. Cell Sci. 14, 571–585. . LAUZON, R., PATTON, C.W. & WEISSMAN, I.L. (1993) A morphological and immunohistochemical study of programmed cell death in botryllus schlosseri (tunicata, ascidiacea). Cell Tissue Res. 272, 115–127. . TRUMAN, J.W. (1992) Insect systems for the study of programmed neuronal death. Exp. Gerontol. 27, 17–28. . SCHWARTZ, L.M. (1992) Insect muscle as a model for programmed cell death. J. Neurobiol . 23, 1312–1326. . NEWBIGIN, E., ANDERSON, M.A. & CLARKE, A. (1993) Gametophytic self incompatibility systems. Plant Cell 5, 1315–1324. . KIRK, D.L. (1988) The ontogeny and phylogeny of cellular differentiation in Volvox. Trends Genet. 4, 32–36. . WILLIAMS, J.G. (1991) Regulation of cellular differentiation during Dictyostelium morphogenesis. Curr. Op. Gen. Dev. 1, 358–362. . PRESCOTT, D. (1994) The DNA of ciliated protozoa. Microbiol. Rev. 58, 233–267. . SHUB, D.A. (1994) Bacterial altruism? Curr. Biol. 4, 555–556. . DAWKINS, R. (1978) The selfish gene. OUP, Oxford. . VAUX, D.L., WEISSMAN, I.L. & KIM, S.K. (1992) Prevention of programmed cell death in Caenorhabditis elegans by human bcl-2. Science 258, 1955–1957. . ALNEMRI, E.S., ROBERTSON, N.M., FERNANDES, T.F., CROCE, C.M. & LITWACK, G. (1992) Overexpressed full-length human BCL2 extends the survival of baculovirus-infected Sf9 insect cells. Proc. Natl Acad. Sci. USA 89, 7295–7299. . SUGIMOTO, A., FRIESEN, P.D. & ROTHMAN, J.H. (1994) Baculovirus p35 prevents developmentally programmed cell death and rescues a ced 9 mutant in the nematode caenorhabditis elegans. EMBO J. 13, 2023–2028.
174
16 17 18 19 20
21 22
23
24 25
26 27 28 29 30 31 32 33 34
35 36 37 38
APOSTOSIS IN NORMAL DEVELOPMENT AND CANCER
. ELLIS, H.M. & HORVITZ, H.R. (1986) Genetic control of programmed cell death in the nematode C. elegans. Cell 44, 817–829. . HENGARTNER, M.O., ELLIS, R.E. & HORVITZ, H.R. (1992) Caenorhabditis elegans gene ced-9 protects cells from programmed cell death. Nature 356, 494–499. . HENGARTNER, M.O. & HORVITZ, H.R. (1994) C. elegans cell survival gene ced-9 encodes a functional homolog of the mammalian proto-oncogene bcl-2. Cell 76, 665–676. . YUAN, J.Y., SHAHAM, S., LEDOUX, S., ELLIS, H.M. & HORVITZ, H.R. (1993) The C. elegans cell death gene ced 3 encodes a protein similar to mammalian interleukin 1 beta converting enzyme. Cell 75, 641–652. . KUMAR, S., KINOSHITA, M., NODA, M., COPELAND, N.G. & JENKINS, N.A. (1994) Induction of apoptosis by mouse Nedd2 gene encoding a protein similar to the product of C. elegans cell death gene ced-3 and mammalian IL-1 -converting enzyme. Function of nedd2. Genes Dev. (in press). . KUMAR, S., TOMOOKA, Y. & NODA, M. (1992) Identification of a set of genes with developmentally downregulated expression in the mouse brain. Biochem. Biophys. Res. Commun. 185, 1155–1161. . HOWARD, A.D., KOSTURA, M.J., THORNBERRY, N., DING, G.J., LIMJUCO, G., WEIDNER, J., SALLEY, J.P., HOGQUIST, K.A., CHAPLIN, D.D. & MUMFORD, R.A. et al. (1991) IL-1-converting enzyme requires aspartic acid residues for processing of the IL-1 beta precursor at two distinct sites and does not cleave 31-kDa IL-1 alpha. J. Immunol. 147, 2964–2969. . THORNBERRY, N.A., BULL, H.G., CALAYCAY, J.R., CHAPMAN, K.T., HOWARD, A.D., KOSTURA, M.J., MILLER, D.K., MOLINEAUX, S.M., WEIDNER, J.R., AUNINS, J. et al. (1992) A novel heterodimeric cysteine protease is required for interleukin-1 beta processing in monocytes. Nature 356, 768–774. . WILSON, K.P. (1994) Structure and mechanism of interleukin-1 converting enzyme. Nature 370, 270–275. . MIURA, M., ZHU, H., ROTELLO, R., HARTWEIG, E.A. & YUAN, J. (1993) Induction of apoptosis in fibroblasts by IL-1 -converting enzyme, a mammalian homolog of the C. elegans cell death gene ced-3. Cell 75, 653–660. . SIDERS, W.M., KLIMOVITZ, J.C. & MIZEL, S.B. (1993) characterization of the structural requirements and cell type specificity of IL-1 alpha and IL-1 beta secretion. J. Biol. Chem. 268, 22170–22174. . DI GIOVINE, F.S. & DUFF, G.W. (1990) Interleukin 1: the first interleukin. Immunol. Today 11, 13–21. . RAFF, M.C. (1992) Social controls on cell survival and cell death. Nature 356, 397–400. . KERR, J.F., WYLLIE, A.H. & CURRIE, A.R. (1972) Apoptosis: a basic biological phenomenon with wideranging implications in tissue kinetics. Br. J. Cancer 26, 239–257. . GREENBERG, J.T. & AUSUBEL, F.M. (1993) Arabidopsis mutants compromised for the control of cellular damage during pathogenesis and aging. Plant J. 4, 327–341. . SCHWARTZ, L.M., JONES, M.E., Kosz, L. & KUAH, K. (1993) Selective repression of actin and myosin heavy chain expression during the programmed death of insect skeletal muscle. Dev. Biol. 158, 448–455. . POTTEN, C.S., AL BARWARI, S., HUME, W.J. & SEARLE, J. (1977) Circadian rhythms of presumptive stem cells in three different epithelia of the mouse. Cell Tissue Kinet. 10, 557–568. . COLUMBANO, A., LEDDA, C.G., CONI, P.P., FAA, G., LIGUORI, C., SANTA, C.G. & PANI, P. (1985) Occurrence of cell death (apoptosis) during the involution of liver hyperplasia. Lab. Invest. 52, 670–675. . OBERHAMMER, F., BURSCH, W., TIEFENBACHER, R., FROSCHL, G., PAVELKA, M., PURCHIO, T. & SCHULTE, H.R. (1993) Apoptosis is induced by transforming growth factor-beta 1 within 5 hours in regressing liver without significant fragmentation of the DNA. Hepatology 18, 1238–1246. . SHIELDS, R. & STRATFORD, R. (1993) Debugging tomatoes. Nature 366, 508–509. . DIETRICH, R.A., DELANEY, T.P., UKNES, S.J., WARD, E.R., RYALS, J.A. & DANGL, J.L. (1994) Arabidopsis mutants simulating disease resistance response. Cell 77, 565–577. . CLEM, R.J. & MILLER, L.K. (1993) Apoptosis reduces both the in vitro replication and the in vivo infectivity of a baculovirus. J. Virol. 67, 3730–3738. . LEE, J.C., CHEN, H.H., WEI, H.L. & CHAO, Y.C. (1993) Superinfection induced apoptosis and its correlation with the reduction of viral progeny in cells persistently infected with hz 1 baculovirus. J. Virol. 67, 6989–6994.
EVOLUTION OF APOPTOSIS
39
40 41 42. 43 44
45 46 47
48 49 50
51 52. 53
54
55 56 57
58 59
175
. JEURISSEN, S.H., WAGENAAR, F., POL, J.M. VAN DER & NOTEBORN, M.H. (1992) Chicken anemia virus causes apoptosis of thymocytes after in vivo infection and of cell lines after in vitro infection. J. Virol. 66, 7383–7388. . LEVINE, B., HUANG, Q., ISAACS, J.T., REED, J.C., GRIFFIN, D.E. & HARDWICK, J.M. (1993) Conversion of lytic to persistent alphavirus infection by the bcl-2 cellular oncogene. Nature 361, 739–742. . GRIFFIN, D.E., LEVINE, B., UBOL, S. & HARDWICK, J.M. (1994) The effects of alphavirus infection on neurons. Ann. Neurol. 35, S23-S27. LANE, D.P. (1992) Worrying about p53. Curr. Biol. 2, 581–583. . YONISH, R.E., RESNITZKY, D., LOTEM, J., SACHS, L., KIMCHI, A. & OREN, M. (1991) Wild-type p53 induces apoptosis of myeloid leukaemic cells that is inhibited by interleukin-6. Nature 353, 345–347. . DONEHOWER, L.A., HARVEY, M., SLAGLE, B.L., MCARTHUR, M.J., MONTGOMERY, C.J., BUTEL, J.S. & BRADLEY, A. (1992) Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumours. Nature 356, 215–221. . LOWE, S.W., SCHMITT, E.M., SMITH, S.W., OSBORNE, B.A. & JACKS, T. (1993) p53 is required for radiation-induced apoptosis in mouse thymocytes. Nature 362, 847–849. . CLARKE, A.R., PURDIE, C.A., HARRISON, D.J., MORRIS, R.G., BIRD, C.C., HOOPER, M.L. & WYLLIE, A.H. (1993) Thymocyte apoptosis induced by p53-dependent and independent pathways. Nature 362, 849–852. . NOTEBORN, M.H.M., TODD, D., VERSCHUEREN, C.A.J., DEGAUW, H.W.F.M., CURRAN, W.L., VELDKAMP, S., DOUGLAS, A.J., MCNULTY, M. S., VANDEREB, A.J. & KOCH, G. (1994) A single chicken anemia virus protein induces apoptosis. J. Virol. 68, 346–351. . VAUX, D.L., HAECKER, G. & STRASSER, A. (1994) An evolutionary perspective on apoptosis. Cell 76, 777–779. . SEARLE, J., KERR, J.F. & BISHOP, C.J. (1982) Necrosis and apoptosis: distinct modes of cell death with fundamentally different significance. Pathol Annu. 2, 229–259. . SHERIDAN, J.W., BISHOP, C.J. & SIMMONS, R.J. (1984) Effects of hypoxia on the kinetic and morphological characteristics of human melanoma cells grown as colonies in semi-solid agar medium. Br. J. Exp. Pathol. 65, 171–180. . STRASSER, A., HARRIS, A.W. & CORY, S. (1991) bcl-2 transgene inhibits T cell death and perturbs thymic self-censorship. Cell 67, 889–899. VAUX, D.L., WHITNEY, D. & WEISSMAN, I.L. (1994) Activation of physiological cell death mechanisms by a necrosis-causing agent. Microsc. Res. Tech. (in press). . NEILAN, J.G., LU, Z., AFONSO, C.L., KUTISH, G.F., SUSSMAN, M.D. & ROCK, D.L. (1993) An African swine fever virus gene with similarity to the protooncogene bcl-2 and the Epstein-Barr virus gene BHRF1. J. Virol. 67, 4391–4394. . PEARSON, G.R., LUKA, J., PETTI, L., SAMPLE, J., BIRKENBACH, M., BRAUN, D. & KIEFF, E. (1987) Identification of an Epstein-Barr virus early gene encoding a second component of the restricted early antigen complex. Virology 160, 151–161. . SCHEFFNER, M., HUIBREGTSE, J.M., VIERSTRA, R.D. & HOWLEY, P.M. (1993) The HPV-16 E6 and E6AP complex functions as a ubiquitin-protein ligase in the ubiquitination of p53. Cell 75, 495–505. . DEBBAS, M. & WHITE, E. (1993) Wild-type p53 mediates apoptosis by E1A, which is inhibited by E1B. Genes. Dev. 7, 546–554. . RAY, C.A., BLACK, R.A., KRONHEIM, S.R., GREENSTREET, T.A., SLEATH, P.R., SALVESEN, G.S. & PICKUP, D.J. (1992) Viral inhibition of inflammation: cowpox virus encodes an inhibitor of the interleukin-1 beta converting enzyme. Cell 69, 597–604. . CLEM, R.J., FECHHEIMER, M. & MILLER, L.K. (1991) Prevention of apoptosis by a baculovirus gene during infection of insect cells. Science 254, 1388–1390. . HAY, B.A., WOLFF, T. & RUBIN, G.M. (1994) Expression of baculovirus p35 prevents cell death in Drosophila. Development 120, 2121–2129.
176
60 61 62
63
64
65
66
67 68
APOSTOSIS IN NORMAL DEVELOPMENT AND CANCER
. BIRNBAUM, M.J., CLEM, R.J. & MILLER, L.K. (1994) An apoptosis inhibiting gene from a nuclear polyhedrosis virus encoding a polypeptide with cys/his sequence motif. J. Virol. 68, 2521–2528. . CLOUSTON, W.M. & KERR, J.F. (1985) Apoptosis, lymphocytotoxicity and the containment of viral infections. Med. Hypoth. 18, 399–404. . DON, M.M., ABLETT, G., BISHOP, C.J., BUNDESEN, P.G., DONALD, K.J., SEARLE, J. & KERR, J.F. (1977) Death of cells by apoptosis following attachment of specifically allergized lymphocytes in vitro. Aust. J. Exp. Biol. Med. Sci. 55, 407–417. . RUSSELL, J.H., MASAKOVSKI, V.R. & DUBOIS, C.B. (1980) Mechanisms of immune lysis. I. Physiological distinction between target cell death mediated by cytotoxic T lymphocytes and antibody plus complement. J. Immunol. 124, 1100–1105. . SHI, L., KAM, C.M., POWERS, J.C., AEBERSOLD, R. & GREENBERG, A.H. (1992) Purification of three cytotoxic lymphocyte granule serine proteases that induce apoptosis through distinct substrate and target cell interactions. J. Exp. Med. 176, 1521–1529. . HEUSEL, J.W., WESSELSCHMIDT, R.L., SHRESTA, S., RUSSELL, J.H. & LEY, T.J. (1994) cytotoxic lymphocytes require granzyme b for the rapid induction of DNA fragmentation and apoptosis in allogeneic target cells. Cell 76, 977–987. . KAGI, D., LEDERMANN, B., BURKI, K., SEILER, P., ODERMATTOLSEN, K.J., PODACK, E.R., ZINKERNAGEL, R.M. & HENGARTNER, H. (1994) Cytotoxicity mediated by T cells and natural killer cells is greatly impaired in perforin deficient mice. Nature 369, 31–37. . SHI, L., KRAUT, R.P., AEBERSOLD, R. & GREENBERG, A.H. (1992) A natural killer cell granule protein that induces DNA fragmentation and apoptosis. J. Exp. Med. 175, 553–566. . VAUX, D.L., AGUILA, H.L. & WEISSMAN, I.L. (1992) Bcl-2 prevents death of factor-deprived cells but fails to prevent apoptosis in targets of cell mediated killing. Int. Immunol. 4, 821–824.
10 Hormonal Control of Apoptosis in Gonadal Tissues HÅKAN BILLIG1,2 and AARON J.W.HSUEH1
1
Division of Reproductive Biology, Department of Gynecology and Obstetrics, Stanford University School of Medicine, Stanford, CA 94305–5317, USA 2 Department of Physiology, University of Göteborg, Medicinargatan 11, 413 90 Göteborg, Sweden 10.1 Introduction
Normal tissue development includes cell proliferation, differentiation and ‘controlled’ cell death [1]. During the fertile life span of mammals, the function of the gonads are hormonally regulated. Even though proliferation of somatic cells in the ovary and germ cells in the testis plays important roles in normal gonadal development and function, degeneration of gonadal cells results in the depletion of a majority of the potential number of germ cells in both sexes. In the human ovary, two million oocytes are found at birth and 400000 follicles are present at the onset of puberty. However, only 400 follicles could possibly be ovulated during the female reproductive life [2]. At the time of menopause, no follicles or oocytes can be found in the ovary. Therefore, more than 99.9% of follicles, including oocytes, granulosa and theca cells are deleted and this process is an integral part of the normal ovarian function. Indeed, it is the ‘norm’ for a follicle to die rather than to ovulate. The degenerative process by which 99.9% of follicles are irrevocably committed to undergo cell death is termed atresia. Despite its critical role during the recruitment of follicles for ovulation, the mechanisms underlying the onset and progression of atresia remain poorly understood [3]. In the testis, germ cell deletion during normal spermatogenesis has been estimated to result in the loss of up to 75% of potential numbers of mature sperm cells during normal spermatogenesis in the adult testis [4–6]. Morphological analysis indicated that germ cell demise is most commonly found in spermatogonia and is present throughout testis development [5]. During neonatal and pubertal development, the incidence of testis germ cell degeneration varies [7]. Although individual seminiferous tubules in the testis do not undergo an ‘all-or-none’ degenerative process like atresia in the ovarian follicle, it is clear that germ cell death plays an important role in testis physiology and pathophysiology. In addition, the seminiferous tubules are highly sensitive to damage by ionizing radiation, chemotherapeutic agents, and hyperthermia, highlighting the role of testis cell death in pathological conditions.
178
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
10.2 Morphological Studies on the Degeneration of Testicular and Ovarian Cells In the ovary, the underlying mechanisms for the massive death of both somatic and germ cells are not clear. In mammals, there are at least four degenerative stages during ovarian development to account for the massive loss of ovarian cells. During migration of the primordial germ cells from the yolk sac to the genital ridge, these cells undergo degeneration unless rescued by stem cell factor (SCF) and related factors [8, 9]. Coincident with their entry into meiosis, germ cells undergo attrition before formation of the follicle [10]. At penultimate stage of development, developing early antral follicles either differentiate or undergo atresia. If ovulatory signals are absent, the mature follicles undergo degeneration. After ovulation, the corpus lutea have finite life span and undergo luteolysis. The morphological features of follicular atresia have been described as progressive changes in all follicular cell types. Early signs of atresia are the presence of scattered pyknotic nuclei in the granulosa cell layer among cells that appear morphologically normal [11]. In follicles of more advanced stages of atresia further follicular deterioration could be seen including detachment of the granulosa cell layer from the basement membrane [12], fragmentation of the basal lamina [13], and the presence of cell debris in follicular antrum [14]. A reduction of protein [15] and DNA [11, 16] synthesis within granulosa cells of atretic follicles has also been reported. During follicular atresia, meiosis-like changes occur in the oocyte which later is fragmented. These changes are accompanied by disruption of oocytecumulus connections [17]. In ovine atretic follicles [18], theca cells undergo degenerative changes. In contrast, human, rat and rabbit theca cells undergo extensive hypertrophy during follicle atresia [19, 20]. Regardless of species differences, however, it is generally accepted that the earlier stages of follicular demise are correlated with disorganization and degeneration of granulosa cells. In the testis, morphological signs of germ cell degeneration during spermatogenesis were recognized almost a century ago [21]. It has been estimated that the deletion of spermatogonia in adult testis results in 25 to 75% decreases of the theoretical optimal sperm production [4–6]. In the testis, the germ cells are the major cell type undergoing degeneration [22]. Peaks of germ cell loss have been shown in three distinct stages of spermatogenesis: during mitotic divisions of type A spermatogonia, during meiotic divisions of spermatocytes and during spermiogenesis [5, and references therein]. The germ cell degeneration shows distinct morphological signs such as margination and condensation of nuclear chromatin and phagocytosis by Sertoli cells without signs of inflammatory response [22], consistent with morphology found during apoptotic cell death [23, 24]. Although the morphological changes seen in apototic cell death resemble the changes seen during follicle atresia and testis cell degeneration, no biochemical analysis has been performed until recently. 10.3 Occurrence and Localization of Apoptosis in the Gonads The most striking feature of apoptosis is the activation of calcium/magnesiumdependent endonuclease activity which specifically cleaves cellular DNA between regularly-spaced nucleosomal units that can be visualized as a distinct ladder of DNA bands following agarose gel electrophoresis and ethidium bromide staining [25, 26]. The Ca2+/Mg2+-dependent endonuclease is present in diverse cell types [26, 27], including the ovary [28]. Once activated, the apoptotic degradation of DNA is irreversible. Indeed, recent studies indicated that rat testis [29–31] and atretic rat, avian and porcine follicles contain DNA fragments resembling those found in of oligonucleosomes of different sizes [32–34].
APOPTOSIS IN GONADAL TISSUES
179
Figure 10.1 Quantitative analysis of DNA fractionation and in situ DNA 3 -end-labeling assay. Isolated DNA is 3 -endlabeled with 32P-ddATP and fractionated using agarose gels. After autoradiography, portions of each lane corresponding to DNA less than 15 kB are isolated, and radioactivity is estimated for quantification of the degree of internucleosomal DNA fragmentation [36]. Alternatively, cellular DNA is 3 -end-labeled in situ with digoxigenin-ddUTP. Incorporated digoxigenin-ddUTP is detected by anti-digoxigenin antibodies conjugated to alkaline phosphatase and visualized after the addition of color substrates. Reprinted from [130] with permission.
Early studies on follicle atresia relied completely on morphological criteria, such as the occurrence of pyknotic nuclei in small populations of cells in a given atretic follicle. The evaluation of the morphological signs is subjective, time-consuming and, more importantly, only semi-quantitative. In addition, different labs set different criteria for atresia based on the presence of degenerating oocyte, varying numbers of pyknotic granulosa cells in a given histological section, and germinal vesicle breakdown of the oocyte [35]. The study of atresia and apoptosis in ovarian follicles has become more accessible with the development of sensitive, quantitative DNA fractionation and in situ DNA end-labeling assays. In the quantitative DNA fractionation assay, isolated DNA is labeled in its 3 -ends with 32P-ddATP. Electrophoretic separation of the labeled DNA enables both quantitative estimation of 32P-ddATP incorporation and autoradiographic visualization of apoptotic DNA fragmentation (Figure 10.1) [36]. In contrast, specific cell types undergoing apoptosis can be studied on histological sections after in situ 3 -end labeling of DNA with a modified nucleotide, digoxigenin-ddUTP. The digoxigenin-ddUTP is visualized on the sections by alkaline
180
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 10.2 Testicular germ cell apoptosis. In situ 3 -end labeling of testis sections from adult rats. 3 -end labeling of DNA with digoxigenin-ddUTP was done on testis sections from adult rats. In many, but not all tubules, increased labeling is seen in some meiotic spermatocytes.
phosphatase conjugated anti-digoxigenin antibodies after addition of color substrates (Figure 10.1) [37]. With the recent application of these methods to the analysis of ovarian cell death, it is clear that, in the ovary, granulosa cells are the cellular sites of apoptosis during follicle atresia [32, 33, 37]. Similar DNA 3 -end labeling and in situ analyses have been used to analyze testis cell apoptosis. In the normal developing testis, increased levels of apoptotic DNA fragmentation were detected in rats between 16 to 32 days of age, while lower levels were found in younger and in adult animals [31]. Consistent with morphological signs of degeneration, in situ analysis indicated that apoptosis in rats at 20 and 32 days of age were most frequently found in spermatocytes [7]. In adult rats, apoptotic DNA fragmentation in seminiferous tubules showed stage-dependent differences. The lowest levels are found in stage VIII while 2fold higher levels of apoptotic DNA fragmentation are found in stages I and XII–XIV. Morphological signs of cell degeneration in type A2, A3 and A4 spermatogonia and spermatocytes have been found in these stages of the tubules [7, 22, 38–40]. In situ analysis demonstrated spermatocytes to be the major cell type with increased DNA fragmentation whereas apoptosis in somatic cells (Leydig and Sertoli cells) could not be detected (Figure 10.2) [29–31]. Thus, in contrast to ovarian studies showing massive apoptosis of somatic granulosa cells, germ cells (mainly spermatocytes) are the major cells in the testis affected [29–31]. 10.4 Hormonal Control of Gonadal Cell Apoptosis The factors which trigger apoptosis in diverse tissues appear to be tissue-specific, although it is believed that these various early apoptotic signals in different cell types may ultimately lead to a common pathway
APOPTOSIS IN GONADAL TISSUES
181
which locks all cells into the irrevocable progression of apoptotic cell death [1]. Steroid hormones regulate apoptosis in several organs either as suppressive or stimulatory signals. These include the thymus gland (glucocorticoids: increases apoptosis [25, 41]), prostate (androgens: decreases apoptosis [42, 43]) and mammary gland (estrogens: decreases apoptosis [44]). Recent studies have also ascribed a role for diverse ‘growth factors’ in the control of apoptosis during embryogenesis [45], hematopoietic stem cell development [46], B-cell selection [47] and neuronal cell survival [48]. In each case, the removal of growth factors triggers an irreversible progression of apoptotic cell death. Examples for cell-cell communication to determine cell survival or death have also been found. For instance, in the developing nervous system the neurons are dependent on neurotrophic paracrine factors, like NGF, produced by target cells to survive [49]. In the gonads, diverse hormones and growth factors can act as survival factors to inhibit apoptosis or as apoptotic factors to induce cell demise through endocrine, paracrine and autocrine mechanisms. Furthermore, the action of these factors are dependent on the stage of differentiation of the target cells. 10.4.1 Role of Gonadotropins In the ovary, decrease of circulating gonadotropins through hypophysectomy [50] or blockade of the LH/ FSH surge [51] on the day of proestrus leads to massive atresia of preovulatory follicles. However, the general process of atresia does not appear to be temporally related to marked changes in serum levels of pituitary gonadotropins because atresia could be found in all stages of the reproductive cycle [15, 52]. Furthermore, early atretic follicles can be ‘rescued’ by administration of exogenous gonadotropins [14, 53]. Thus, gonadotropins are likely to be survival factors in preventing follicle apoptosis. Indeed, cells collected from atretic rat [54, 55], sheep [56] or hamster [57] follicles display decreased responsiveness to gonadotropins. Also the expression of mRNA for LH and FSH receptors are decreased in atretic porcine follicles showing increased apoptotic DNA degradation [34]. In addition to the morphological analysis of atresia discussed earlier, measurement of ovarian DNA fragmentation and in situ analysis also indicated that treatment of hypophysectomized immature rats with FSH decreases follicle apoptosis in the granulosa cells in vivo [58]. Furthermore, preovulatory follicles isolated from equine chorionic gonadotropin-treated rats contain predominantly intact high MW DNA. However, a time-dependent, spontaneous onset of internucleosomal DNA fragmentation characteristic of apoptotic cell death occurred during culture of the follicles. When cultures were treated with either FSH or hCG, the spontaneous onset of apoptotic DNA fragmentation of preovulatory follicles was prevented, underscoring the role of gonadotropins as follicle survival factors [59]. In the testis, it is established that spermatogenesis is gonadotropin-dependent and the optimal function of testis cells is supported by LH and FSH [60, 61]. Hypophysectomy or neutralization of circulating gonadotropins increases degeneration of spermatogenic cells [7, 38], [62] while treatment with gonadotropins prevents germ cell degeneration at specific stages of spermatogenesis [7, 38]. Recently, biochemical evidence was presented that links testicular cell death to increased internucleosomal DNA fragmentation [29]. Hypophysectomy increases internucleosomal DNA fragmentation demonstrating the pituitary dependence of testicular apoptosis. In vivo treatment of hypophysectomized rats with LH and FSH inhibits testicular apoptosis [29]. Even though Ley dig and Sertoli cells are dependent on gonadotropins for their development [63, 64], hypophysectomy of adult animals does not induce morphological signs of apoptosis in these cells [65, 66]. Although gonadotropin deprivation was shown to increase apoptotic DNA fragmentation in the enriched interstitial cell preparation from immature rats [29], in situ data indicate that only germ cells and not Leydig or Sertoli cells undergo apoptosis [30, 31].
182
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 10.3 Gonadotropin regulation of testicular apoptosis: age-related changes in gonadotropin dependence. Pretreatment of rats with a potent GnRH antagonist that suppresses circulating levels of FSH and LH causes a marked increase in apoptotic DNA fragmentation in 16- to 32-day-old animals. Testicular DNA was 3 -end labeled with [32P]ddATP and fractionated through 2% agarose gels followed by exposure on films to visualize the incorporation of [32P]ddATP into DNA 3 -ends (V=vehicle-treated, A=GnRH antagonist-treated). Reprinted from [31] with permission.
Furthermore, our recent studies indicate the gonadotropin-dependence of testis germ cell apoptosis is agerelated. A marked increase in apoptotic DNA fragmentation was seen in 16- to 32-day-old animals pretreated with a potent GnRH antagonist, that suppresses circulating levels of FSH and LH (Figure 10.3) [31]. Furthermore, in situ labeling of testis DNA fragments demonstrated that an increase in the number of apoptotic spermatocytes was evident in 16- to 32-day-old rats pretreated with GnRH antagonist. These findings are consistent with the cell stages previously shown to have increased morphological signs of degeneration after hypophysectomy in rats at similar age [7]. However, in animals younger than 16 days of age or adults, the GnRH antagonist treatment did not affect the level of apoptotic DNA fragmentation in spite of decreased levels of FSH, suggesting the role of gonadotropins as survival factors is age-dependent. 10.4.2 Role of Sex Steroids In the ovary, sex steroids are important intraovarian regulators of follicle atresia and when estrogen is given to hypophysectomized rats, fewer follicles become atretic [67]. Furthermore, the profile of sex steroid production in healthy follicles differs from that in atretic ones. In rat and hamster, the production of both estrogens and androgens decreases in atretic follicles [51, 54, 55, 68]. In human, ovine and porcine species the production of estradiol production by atretic follicles is also decreased but the production of androgens is increased [69–71]. The common denominator for all these species is a decreased estrogen production in atretic follicles. In general, changes in steroidogenesis can be observed prior to morphological signs of atresia [51, 54]. We have analyzed ovarian DNA fragmentation in rats treated with estrogens and androgens [37]. Immature rats were hypophysectomized and implanted with estrogen (DES, diethylstilbestrol) capsules. Two days later, estrogen implants were removed from some animals followed by treatment with estrogens
APOPTOSIS IN GONADAL TISSUES
183
Figure 10.4 Sex steroid regulation of ovarian apoptosis. Time-dependent effect of estrogen withdrawal on total ovarian DNA fragmentation. Hypophysectomized immature rats at 22 days of age were implanted with estrogen capsules (DES). Two days after operation, the estrogen implants were removed (−DES groups) and animals were sacrificed at 12 h intervals up to 48 h. Animals without removal of the DES implants served as controls (+DES groups). Total ovarian DNA was 3 -end labeled with [32P]-ddATP and fractionated through 2% agarose gels followed by autoradiography. Time 0 represents the time of removal of the DES implants. Reprinted from [37] with permission.
with or without androgens. After estrogen withdrawal, apoptotic DNA fragmentation increased in the ovary (Figure 10.4). In situ analysis of DNA fragmentation on histological sections of ovaries demonstrated that apoptosis induced by estrogen withdrawal is confined to the granulosa cells in early antral and pre-antral follicles (Figure 10.5). No increase in DNA breakdown was detected in theca and interstitial cells or granulosa cells of primordial and primary follicles. In contrast, replacement with DES or estradiol benzoate completely prevented the observed increases in granulosa cell apoptosis [37]. In contrast to estrogens, treatment with androgens have been shown to be atretogenic to ovarian follicles. In vivo treatment with androgens causes a dose- and time-dependent decrease of ovarian weight [72, 73] and an increase in morphological signs of atresia in estrogen-treated hypophysectomized rats [72]. Endogenously produced androgens are also atretogenic because atresia induced by low doses of hCG [74] was inhibited by androgen receptor blockers and testosterone antibodies [75]. Furthermore, the antiatretogenic effect of estrogen was blocked by treatment with testosterone (Figure 10.5). The specificity of testosterone action was further demonstrated by the lack of effect of progesterone and cortisol on ovarian apoptosis. These data suggest that sex steroids play an important role in the regulation of ovarian apoptotic cell death with estrogens preventing apoptosis whereas androgens antagonize the effect of estrogens [37]. The survival of male germ cells is probably dependent on gonadotropins as well as intratesticular androgens induced by LH. In contrast to the ovary [37], intratesticular androgens can act as survival factors and treatment with androgens prevents apoptotic DNA fragmentation [29]. Indeed, androgens alone can maintain spermatogenesis in adult animals [76]. It is believed that intratesticular androgens, secreted by Leydig cells in response to LH stimulation, play important paracrine roles in preventing germ cell degeneration.
184
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
10.4.3 Role of Growth Factors Granulosa cells isolated from preovulatory follicles of PMSG-treated rat contained predominantly intact high MW DNA. However, a time-dependent, spontaneous onset of internucleosomal DNA fragmentation occurred in these granulosa cells during serum-free culture suggesting a withdrawal or lack of survival factors. Treatment of these cultured granulosa cells with EGF, TGF , or bFGF inhibited the spontaneous onset of apoptotic DNA cleavage whereas treatment with IGF-I or insulin had no effect. Furthermore, inclusion of a tyrosine kinase inhibitor (genistein) completely blocked the ability of EGF, TGF , and bFGF to suppress apoptosis in these cells [77]. The presence of EGF/TGF and basic FGF and their receptors in the ovary have been demonstrated at both the protein and mRNA levels [78–82]. In addition, the levels of ovarian TGF message are up-regulated following FSH stimulation in vivo [81]. Although immunocytochemical analysis has localized TGF protein to the theca-interstitial layer of follicles in several species [81, 83], high-affinity receptors for EGF/TGF [84–86] and bFGF [87] have been localized to granulosa cells of ovarian follicles. Therefore, it is possible that apoptotic cell death in granulosa cells of follicles selected for ovulation is prevented by the paracrine actions of EGF/TGF produced and secreted by theca-interstitial cells, or the autocrine actions of bFGF synthesized by granulosa cells. In the ovary, IGFI is produced locally and plays an important intra-ovarian role in the augmentation of gonadotropin stimulation of follicle differentiation [88]. In rats, the ovarian mRNA for IGFI is abundant (fourth only to liver, oviduct and uterus: [89, 90]), underscoring the importance of local control mechanisms. Early studies have also demonstrated the presence of IGF type I receptors in the granulosa cells [91] and the stimulation of ovarian IGFI levels by both FSH [92, 93]and GH [94], although gonadotropins are probably the prime stimulator of granulosa cell IGFI production [95]. IGFs in body fluids are bound to IGF binding proteins (IGFBPs) and six IGFBPs have been cloned and sequenced [96]. Although circulating IGFBPs prolong the half life of IGFs, these proteins mostly inhibit IGF actions [97–99]. IGFBP 4 and 5 are produced by rat granulosa cells [100] and FSH treatment decreases the secretion of these proteins in rat ovaries [101]. In atretic human follicles of both normal and polycystic ovarian syndrome (PCO) ovaries, high levels of IGFBPs were detected [102, 103]. In situ mRNA analysis further demonstrated the presence of IGFBPs in atretic but not healthy follicles [104]. Furthermore, the binding of endogenously produced IGFI by exogenously added IGFBP-3 results in the suppression of gonadotropin stimulation of follicle growth and subsequent ovulation [97, 98], suggesting that part of the physiological effects of FSH are mediated by endogenously produced IGFI. Using the preovulatory in vitro follicle model, we recently demonstrated that treatment with IGFI, like FSH and hCG, prevents spontaneous onset of apoptosis [59]. Furthermore, the suppressive effect of both IGFI and hCG on follicle DNA fragmentation was prevented by co-treatment with IGFBP-3 (Figure 10.6). The finding that IGFBP3 is capable of neutralizing the action of hCG to suppress spontaneous apoptosis further indicates that endogenous IGFI is secreted outside the granulosa cells to exert its action [59]. These data suggest that IGFI may also be a survival factor for ovarian follicles and the ability of gonadotropins to suppress apoptosis is partially mediated by endogenously produced IGFI. The apoptosis-suppression action of gonadotropins is manifested by decreasing the production of IGFBPs [101] as well as the stimulation of IGFI production [92, 93]. In contrast to the ovary, regulators of testis cell apoptosis by growth factors has not been studied.
APOPTOSIS IN GONADAL TISSUES
185
Figure 10.5 Sex steroid regulation of ovarian apoptosis. In situ 3 -end labeling of ovarian DNA from rats with or without estrogen and testosterone treatment. Immature hypophysectomized rats were treated with estrogen as in Figure 10.4 followed by in situ 3 -end labeling with digoxigenin-ddUTP [37]. Panels A and B: Ovary after 48 h of estrogen treatment. Panels C and D: Ovary 48 h after termination of estrogen treatment (corresponding to −DES 48 h in Figure 10.4) showing labeling of granulosa cells (gc) but not theca cells (tc) in preantral follicles (arrow). Also indicated is lack of labeling in primordial follicles (arrowhead) and interstitial tissue. Panels E and F: Ovary after 48 h treatment with testosterone and estrogen showing labeling of granulosa cells (gc) but not theca cells (tc). Panel G: Ovary with the same treatment as in panel C but 3 -end labeling was performed with ddATP instead of digoxigenin-ddUTP demonstrating the level of non-specific labeling. Bar lengths equal 250 µm. Reprinted from [37] with permission.
10.4.4 Role of Other Factors In rat ovaries, the regulatory peptide GnRH and its agonists have been shown to exert direct inhibitory effects
186
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 10.6 Gonadotropin and growth factor regulation of ovarian apoptosis. Inhibition of apoptosis in cultured ovarian follicles by hCG and IGF-I, reversal, by IGF bindings proteins. DNA from preovulatory follicles incubated for 24 h with hCG or IGF-I with or without IGF binding protein (IGFBP-3) was 3 -end labeled and gel fractionated as in Figure 10.1. Incorporation of [32P]-ddATP was determined in portions of the gels corresponding to low-molecularweight DNA (<15 kB) and expressed as percentage change compared with DNA from follicles cultured for 24 h without hormone treatment. Time zero represents DNA isolated from follicles prior to culture. Reprinted from [59] with permission.
on follicle differentiation by acting through specific ovarian receptors in granulosa and theca cells [105, 106]. Because treatment with GnRH antagonists in vivo augments gonadotropin stimulation of follicle growth, inhibin production and the content of IGF type I receptor in the ovary, the presence of an ovarian GnRHlike substance has been postulated [107–109]. Although low levels of GnRH mRNA have been demonstrated in rat ovaries using reverse-transcription polymerase chain reaction [110], the presence of the peptide remains elusive. In hypophysectomized, estrogen-treated rats, treatment with a GnRH agonist has been shown to directly induce ovarian apoptotic DNA fragmentation with or without co-treatment with FSH [58]. In situ analysis further indicated that the apoptotic cell death is confined to the granulosa cells. Because GnRH is known to increase intracellular Ca2+ and phosphatidyl inositol turnover in rat granulosa cells [111, 112], these studies may provide a model to analyze the activation of the Ca2+/ Mg2+-dependent endonuclease through the protein kinase C pathway during hormonally induced apoptosis. Recent studies have also indicated that the pleiotropic cytokine interleukin-6 (IL-6) is produced by granulosa cells [113] under the control of FSH, IL-1 , IL-1 , lipopolysaccharide but not TNF- . Using cultured granulosa cells from PMSG-primed rats, IL-6 was shown to stimulate granulosa cell apoptotic DNA fragmentation, suggesting that the cytokine may be an intra-follicular atretogenic factor [114]. Studies using diverse animal models have shown that proteins of the bcl-2 gene family are likely to be the direct regulators of apoptosis onset in many cell types. Only limited data are available regarding the regulation of cellular bcl-2 levels. In some cells, extracellular signals that promote cell survival stimulate the expression of bcl-2 [115]. Although bcl-2 expression is topographically restricted in tissues characterized by apoptotic cell death (e.g. hemato-lymphoid tissue and brain), Bcl-2 protein was not found in the human ovary [116, 117]. However, the exact status of the autopsy material used in these limited studies is unclear and a recent report found bcl-2 expression in chicken ovaries [118]. The expression of the bcl-2 gene in ovarian follicles of mammals and the effect of its over-expression on follicle atresia remain to be studied.
APOPTOSIS IN GONADAL TISSUES
187
In the testis, experimentally induced bilateral cryptorchidism (re-position of the testis inside the abdomen) induces germ cell apoptosis prior to decreases in the level of LH receptor expression [30]. In this case, apoptosis is not due to lack of gonadotropin stimulation since the procedure causes significant increases in serum levels of LH and FSH [119–122]. Furthermore, in unilaterally cryptorchid rats, the abdominal testis showed apoptotic DNA degradation while the contralateral sham oper-ated testis in the same animal remains viable in spite of exposure to the same circulating hormone levels [30]. These studies demonstrate that elevation of testis temperature or other changes associated with cryptorchidism may regulate male germ cell apoptosis. Indeed, morphological signs of apoptosis in spermatocytes have been demonstrated in testis exposed to higher temperature [123]. During apoptotic cell death in ventral prostate the mRNA level of a 70 kDa heat shock protein is increased [124]. In contrast, inactivation of heat shock proteins in fibroblasts increases their sensitivity to heat-induced apoptosis [125]. Because, the mRNA and gene products of several heat shock proteins have been detected in both normal and heat-treated testis [126–129], both activation and inactivation of different heat shock proteins might play a role in apoptotic germ cell death of the testis [30]. 10.5 Comparison of Ovarian and Testicular Cell Apoptosis It is clear that different peptide and steroid hormones play important roles as survival and apoptotic factors in the regulation of follicular and testicular apoptosis (Table 10.1). The main products of gonadal cells, the sex steroids, are regulators of apoptosis. In addition, the less studied IL-6 and GnRH-like peptides may also play important intra-ovarian regulatory roles. However, it is also clear that the specific stage of differentiation of gonadal cells determines their susceptibility to survival or apoptotic factors. For instance, granulosa cells in primordial follicles do not Table 10.1 Comparison of ovarian and testicular apoptosis Reduction of potential germ cell pool through apoptosis Primary cell type affected Survival factors gonadotropins steroids growth factors Cell types responsive to survival factors Apoptotic factors
Cell types responsive to apoptotic factors
Ovary
Reference Testis
Reference
99.9%
2
50–75%
5
Granulosa cells
37
Germ cells
31
LH, FSH Estrogens bFGF, TGF , EGF IGF-I Granulosa cells Theca cells
58, 59 37 77 59 133
LH, FSH Androgens ?
29, 31 29
134
Androgens GnRH IGFBPs Granulosa cells
37 58 59 133
Leydig cells Sertoli cells Germ cells Temperature (mediated by local factors?) ?
30
188
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
undergo apoptosis after estrogen withdrawal or androgen stimulation, while granulosa cells in more differentiated follicles do [37]. Furthermore, testis germ cells in young or adult rats do not undergo apoptosis after suppression of circulating FSH levels while 16–32-day-old animals do [31]. In addition, different developmental phases in the spermatogenesis demonstrate varying, but stage-specific, degeneration and apoptosis [40, 31]. Based mainly on studies using preovulatory follicles of the rat, one can postulate a model for the regulation of follicle apoptosis by intra-ovarian hormonal mechanisms involving several growth factors [130]. Because granulosa cells are the exclusive site of apoptotic cell death in rats, endocrine and paracrine signals converge on this cell type to regulate apoptosis. However, studies using cultured granulosa cells indicated that treatments with FSH, LH/hCG and IGFI are ineffective in the prevention of spontaneous apoptosis, despite of their apoptosis-suppressing action in cultured follicles. Because FSH receptors are present exclusively in the granulosa cells, these findings suggest an important role of the neighboring theca cells. FSH or LH may act at the granulosa cells to produce a theca cell stimulator which, in turn, increases the secretion of EGF or TGF by the theca cells. Subsequently, these growth factors could diffuse back to granulosa cells to inhibit apoptosis. Because bFGF is produced by granulosa cells [82], this peptide could also play an autocrine role in the regulation of follicle apoptosis. IGFI and insulin are also ineffective in preventing spontaneous apoptosis in cultured granulosa cells [77] but capable of suppressing it in cultured follicles [59]. Because granulosa cell is the main site of IGFI synthesis in rats, one can further hypothesize that gonadotropins stimulate the production of IGFI production by granulosa cells and the secreted IGFI acts on theca cells to stimulate the production of EGF/ TGF . Again, theca cell TGF /EGF or other factors may diffuse back to granulosa cells to inhibit apoptosis. Increases in germ cell apoptosis after GnRH antagonist treatment in rats between 16 and 32 days of age are correlated with decreases in serum FSH but not LH levels. It is, therefore, postulated that FSH is a survival factor for these cells whereas LH, and intra-testicular androgens induced by LH, play less important roles in germ cell survival during these developmental stages. Because FSH receptors are found exclusively in the Sertoli cells of the seminiferous tubules but not in germ cells themselves [131], it is likely that the FSH stimulation of Sertoli cells results in the stimulation of an intratubular factor essential for the survival of the male germ cells [132]. However, in adult animals, androgens alone can maintain spermatogenesis making it likely to be a survival factor as well [76]. 10.6 Conclusion Although early morphological analysis has clearly documented that a majority of ovarian follicles and testicular germ cells undergo degeneration during reproductive life, only recent studies have demonstrated the involvement of gonadal apoptotic cell death. Availability of a quantitative method to analyze apoptotic DNA fragmentation and in situ methods to localize the specific cell types involved in DNA degradation open new experimental possibilities to increase the understanding of the hormonal control of gonadal cell apoptosis. Analysis of follicle cell apoptosis confirms and extends earlier findings showing the role of gonadotropins and estrogens as follicle survival factors as well as the role of androgens as an atretogenic factor. In addition, several ovarian growth factors and regulatory peptides have recently been identified as either survival factors (EGF/TGF , basic FGF and IGFI) or atretogenic factors (interleukin-6 and GnRH). Likewise, analysis of testicular cell apoptosis confirms and extends earlier findings showing the role of gonadotropins and androgens as testicular survival factors. In addition, it is also clear in both gonads that
APOPTOSIS IN GONADAL TISSUES
189
the susceptibility to both apoptotic and survival factors is dependent on the stage of differentiation of the target cells. Even though the primary target cell types undergoing apoptosis in the gonads differ (germ cells in the testis and granulosa cells in the ovary), apoptosis is a physiological process in the reduction of the potential germ cell pool in both sexes. Future analysis of the mechanism of gonadal cell apoptosis should provide new understanding on the molecular process underlying the hormonal action as well as to provide better treatment protocols for the management of pathological conditions involving abnormal gonadal cell demise such as premature ovarian failure and polycystic ovary syndrome in the female and oligospermia and cryptorchidism in the male. Acknowledgements This work was supported by NIH Grant HD-13527 and Swedish MRC Grant B93– 10380. References 1 2 3 4 5 6 7 8 9
10 11 12 13
14 15 16 17
. RAFF, M.C. (1992) Social controls on cell survival and cell death. Nature 356, 397– 400. . BAKER, T.G. (1963) A quantitative and cytological study of germ cells in human ovaries. Proc. Royal Soc. London Ser. B 158, 417–433. . HIRSHFIELD, A.N. (1991) Development of follicles in the mammalian ovary. Int. Rev. Cytol. 124, 43–101. . OAKLAND, E. (1956) A description of spermatogenesis in the mouse and its use in analysis of the cycle of seminiferous epithelium and germ cell renewal. Am. J. Anat. 99, 391–413. . HUCKINS, C. (1978) The morphology and kinetics of spermatogonial degeneration in normal adult rats. An analysis using a simplified classification of germinal epithelium. Anat. Rec. 190, 905–926. . DE ROOIJ, D.G. & LOK, D. (1987) Regulation of the density of spermatogonia in the seminiferous epithelium of the Chinese hamster: II Differentiating spermatogonia. Anat. Rec. 217, 131–136. . RUSSEL, L.D., ALGER, L.E. & NEQUIN, L.G. (1987) Hormonal control of pubertal spermatogenesis. Endocrinology, 120 1615–1632. . GODIN, I., DEED, R., COOKE, J., ZSEBO, K., DEXTER, M. & WYLIE, C.C. (1991) Effects of the Steel gene product on mouse primordial germ cell culture. Nature 352, 807–809. . DOLCI, S., WILLIAMS, D.E., ERNST, M.K., RESNICK, J.L., BRANNAN, C.I., LOCK, L.F., LYMAN, S.D., BOSWELL, S. & DONOVAN, P.J. (1991) Requirement for mast cell growth factor for primordial germ cell survival in culture. Nature 352, 809–811. . BEAUMONT, H.M. & MANDL, A.M. (1962) A quantitative and cytological study of oogonia and oocytes in foetal and neonatal rat. Proc. Royal Soc. London Ser. B 155, 557–579. . HIRSHFIELD, A.N. (1989) Rescue of atretic follicles in vitro and in vivo. Biol. Reprod. 40, 181–190. . JUNQUIERA, L.C., CARNEIRO, J. & KELLY, R.O. (1989) Basic histology. Appleton and Lange, Norwalk, pp. 443–444. . BAGAVANDOSS, P., MIDGLEY JR, A.R. & WICHA, M. (1983) Developmental changes in the ovarian follicular basal lamina detected by immunofluorescence and electron microscopy. J. Histochem. Cytochem. 31, 633–640. . HAY, M.F., CRAN, D.G. & MOOR, R.M. (1976) Structural changes occurring during atresia in sheep ovarian follicles. Cell Tissue Res. 169, 515–529. . BYSKOV, A.G.S. (1979) Atresia. In: Ovarian follicular development and function, MIDGLEY, A.R. & SADLER, W.A. (eds). Raven Press, New York, pp. 41–57. . GREENWALD, G.S. (1989) Temporal and topographic changes in DNA synthesis after induced follicular atresia. Biol. Reprod. 40, 175–181. . TSAFRIRI, A. & BRAW, R.H. (1984) Experimental approaches to atresia in mammals. Oxford Rev. Reprod. Biol. 6, 226–26.
190
18 19 20 21 22
23 24 25 26 27 28 29 30 31
32
33 34
35 36 37 38 39 40
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. O’SHEA, J.D., HAY, M.F. & CRAN, D.G. (1978) Ultrastructural changes in the theca interna during follicular atresia in sheep. J. Reprod. Fertil. 54, 183–187. . BRAW, R.H., BYSKOV, A.G., PETERS, H. & FABER, M. (1976) Follicular atresia in the human infant ovary. J. Reprod. Fertil. 46, 55–59. . ERICKSON, G.F., MAGOFFIN, D.A., DYER, C.A. & HOFEDITZ, C. (1985) The ovarian androgen producing cells: a review of structure/function relationships. Endocr. Rev. 6, 371–399. . REGAUD, C.P. (1900) Dégénérescence des cellules seminales chez mammifères en l’abscence de tout état pathologique. C.R. Soc. Biol. 52, 268–270. . ALLAN, D.J., HARMON, B.V. & ROBERTS, S.A. (1992) Spermatogonial apoptosis has three morphologically recognizable phases and shows no circadian rhythm during normal spermatogenesis in the rat. Cell Prolif. 25, 241–250. . COHEN, J.J. (1993) Apoptosis. Immunol. Today 14, 126–130. . SCHWARTZ, L.M. & OSBORNE, B.A. (1993) Programmed cell death, apoptosis and killer genes. Immunol. Today 14, 582–590. . WYLLIE, A.H., KERR, J.F.R. & CURRIE, A.R. (1980) Cell death: the significance of apoptosis. Int. Rev. Cytol. 68, 251–306. . ARENDS, M.J., MORRIS, R.G. & WYLLIE, A.H. (1990) Apoptosis: the role of the endonuclease. Am. J. Pathol. 136, 593–608. . GAIDO, M.L. & CIDLOWSKI, J.A. (1991) Identification, purification, and characterization of calciumdependent endonuclease (NUC18) from apoptotic rat thymocytes. J. Biol. Chem. 266, 18580–18585. . ZELEZNIK, A.J., IHRIG, L. & BASSETT, S.G. (1989) Developmental expression of Ca++/Mg++-dependent endonuclease activity in rat granulosa and luteal cells. Endocrinology 125, 2218–2220. . TAPANAINEN, J.S., TILLY, J.L., VIHKO, K.K. & HSUEH, A.J.W. (1993) Hormonal control of apoptotic cell death in the testis: gonadotropins and androgens as testicular cell survival factors. Mol. Endocrinol. 7, 643–650. . SHIKONE, T., BILLIG, H. & HSUEH, A.J.W. (1994) Experimentally-induced cryptorchidism increases apoptosis in rat testis. Biol. Reprod. (in press). . BILLIG, H., FURUTA, I., RIVIER, C., TAPANAINEN, J., PARVINEN, M. & HSUEH, A.J.W. (1994) Apoptosis in testis germ cells: developmental changes in gonadotropin dependence and localization to selective tubule stages (manuscript). . HUGHES JR, F.M. & GOROSPE, W.C. (1991) Biochemical identification of apoptosis (programmed cell death) in granulosa cells: evidence for a potential mechanism underlying follicular atresia. Endocrinology 129, 2415–2422. . TILLY, J.L., KOWALSKI, K.I., JOHNSON, A.L. & HSUEH, A.J.W. (1991) Involvement of apoptosis in ovarian follicular atresia and postovulatory regression. Endocrinology 129, 2799–2801. . TILLY, J.L., KOWALSKI, K.I., SCHOMBERG, D.W. & HSUEH, A.J.W. (1992) Apoptosis in atretic ovarian follicles is associated with selective decreases in messenger RNA transcripts for gonadotropin receptors and cytochrome P450 aromatase. Endocrinology 131, 1670–1676. . GREENWALD, G.S. & ROY, S.K. (1994) Follicular development and its control. In: The physiology of reproduction, KNOBIL, E. & NEILL, J.D. (eds). Raven Press, New York, vol. 2, pp. 629–724. . TILLY, J.L. & HSUEH, A.J.W. (1992) A microscale autoradiographic analysis of apoptotic DNA fragmentation. J. Cellular Physiol. 131, 1063–1068. . BILLIG, H., FURUTA, I. & HSUEH, A.J.W. (1993) Estrogens inhibit and androgens enhance ovarian granulosa cell apoptosis. Endocrinology 133, 2204–2212. . RUSSEL, L.D. & CLERMONT, Y. (1977) Degeneration of germ cells in normal, hypophysectomized and hormone treated hypophysectomized rats. Anat. Rec. 187, 347– 366. . CLERMONT, Y. (1962) Quantitative analysis of spermatogenesis of the rat: a revised model for the renewal of spermatogonia. Am. J. Anat. 177, 111–129. . KERR, J.B. (1992) Spontaneous degeneration of germ cells in the normal rat testis: assessment of cell types and frequency during the spermatogenic cycle. J. Reprod. Fertil. 95, 825–830.
APOPTOSIS IN GONADAL TISSUES
41 42 43 44 45 46 47 48
49 50 51 52 53 54 55 56 57 58
59
60 61 62 63 64
191
. COMPTON, M.M. & CIDLOWSKI, J.A. (1986) Rapid in vivo effects of glucocorticoids on the integrity of rat lymphocyte genomic deoxyribonucleic acid. Endocrinology 118, 38–45. . SANFORD, N.L., SEARLE, J.W. & KERR, J.F.R. (1984) Successive waves of apoptosis in the rat prostate after repeated withdrawal of testosterone stimulation. Pathology 16, 406–410. . KYPRIANOU, N. & ISAACS, J.T. (1988) Activation of programmed cell death in the rat ventral prostate after castration. Endocrinology 122, 552–562. . CHO-CHUNG, Y.S. (1978) Interaction of cyclic AMP and estrogen in tumor growth control. In: Endocrine control of neoplasia, SHARMA, R. & CRISS, W.E. (eds). Raven Press, New York, pp. 335–346. . HASSELL, J.R. & PRATT, R.M. (1977) Elevated levels of cAMP alters the effect of epidermal growth factor in vitro on programmed cell death in the secondary palatial epithelium. Exp. Cell. Res. 106, 55–62. . WILLIAMS, G.T., SMITH, C.A., SPOONCER, E., DEXTER, T.M. & TAYLOR, D.R. (1990) Hematopoietic colony stimulating factors promote cell survival by suppressing apoptosis. Nature 343, 76–79. . HOCKENBERY, D., NUNEZ, G., MlLLIMAN, C., SCHREIBER, R.D. & KORSMEYER, S.J. (1990) Bcl-2 is an inner mitochondrial membrane protein that blocks programmed cell death. Nature 348, 334–336. . MARTIN, D.P., SCHMIDT, R.E., DISTEFANO, P.S., LOWRY, O.H., CARTER, J.G. & JOHNSON JR, E.M. (1988) Inhibitors of protein synthesis and RNA synthesis prevent neuronal cell death caused by nerve growth factor deprivation. J. Cell Biol. 106, 829–844. . RAFF, M.C., BARRES, B.A., BURNE, J.F., COLES, H.S., ISHIZAKI, Y. & JACOBSON, M.D. (1993) Programmed cell death and the control of cell survival: Lessons from the nervous system. Science 262, 695–700. . INGRAM, D.L. (1953) The effect of hypophysectomy on the number of oocytes in the adult albino rat. J. Endocr. 9, 307–311. . BRAW, R.H. & TSAFRIRI, A. (1980) Follicles explanted from pentobarbitone-treated rats provide a model for atresia. J. Reprod. Fertil. 59, 259–265. . BRAND, A. &, DE JONG, W.H.R. (1973) Qualitative and quantitative micromorphological investigations of the tertiary follicle population during the oestrous cycle of the sheep. J. Reprod. Fertil. 33, 431–439. . BRAW, R.H. & TSAFRIRI, A. (1980) Effect of PMSG on follicular atresia in the rat ovary. J. Reprod. Fertil. 59, 267–272. . UILENBROEK, J.T.J., WOUTERSEN, P.J.A. & VAN DER SCHOOT, P. (1980) Atresia of preovulatory follicles: gonadotropin binding and steroidogenic activity. Biol. Reprod. 23, 219–229. . BRAW, R.H., BAR-AMI, S. & TSAFRIRI, A. (1981) Effect of hypophysectomy on atresia of rat preovulatory follicles. Biol. Reprod. 25, 989–996. . CARSON, R.S., FINDLAY, J.K., BURGER, H.G. & TROUNSON, A.O. (1979) Gonadotropin receptors of the ovine ovarian follicle during follicular growth and atresia. Biol. Reprod. 21, 75–87. . SHAHA, C. & GREENWALD, G.S. (1984) Autoradiographic analysis of changes in ovarian binding of FSH and hCG during induced follicular atresia in the hamster. J. Reprod. Fertil. 66, 197–201. . BILLIG, H., FURUTA, I. & HSUEH, A.J.W. (1994) Gonadotropin releasing hormone (GnRH) directly induces apoptotic cell death in the rat ovary: biochemical and in situ detection of DNA fragmentation in granulosa cells. Endocrinology 134, 245– 252. . CHUN, S.Y., BILLIG, H., TILLY, J.L., FURUTA, I., TSAFRIRI, A. & HSUEH, A. J.W. (1994) Gonadotropin suppression of apoptosis in cultured preovulatory follicles: mediatory role of endogenous IGF-I. Endocrinology (in press). . PARVINEN, M. (1982) Regulation of the seminiferous epithelium. Endocr. Rev. 3, 404– 417. . STEINBERGER, E. (1991) Hormonal control of mammalian spermatogenesis. Physiol. Rev. 51, 1–22. . RAJ, H.G.M. & DYM, M. (1976) The effects of selective withdrawal of FSH and LH on spermiogenesis in the immature rat. Biol. Reprod. 14, 489–494. . KERR, J.B. & SHARPE, R.M. (1985) Follicle-stimulating hormone induction of Leydig cell maturation. Endocrinology 116, 2592–2604. . ALMIRON, I. & CHEMES, H. (1988) Spermatogenic onset. II FSH modulates mitotic activity of germ and Sertoli cells in immature rats. Int. J. Androl. 11, 235–246.
192
65
66
67 68 69
70 71 72
73 74
75 76 77
78 79 80 81
82 83 84 85
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. GOSH, S., BARTKE, A., GRASSO, P., REICHERT, L.E. & RUSSEL, L.D. (1992) Structural manifestations of the rat Sertoli cell to hypophysectomy: a correlative morphometric and endocrine study. Endocrinology 131, 485–497. . RUSSEL, L.D., CORBIN, T.J., REN, H.P., ARADOR, A., BARTKE, A. & GOSH, S. (1992) Structural changes in rat Leydig cells posthypophysectomy: a morphometric and endocrine study. Endocrinology 131, 498–508. . INGRAM, D.L. (1959) The effect of oestrogen on the atresia of ovarian follicles. J. Endocr. 19, 123–125. . TERRANOVA, P.F. (1981) Steroidogenesis in experimentally induced atretic follicles of the hamster: a shift from estradiol to progesterone synthesis. Endocrinology 108, 1855– 1890. . CARSON, R.S., FINDLAY, J.K., CLARKE, I.J. & BURGER, H.F. (1981) Estradiol, testosterone, and androstenedione in ovine follicular fluid during growth and atresia of ovarian follicles. Biol. Reprod. 24, 105–113. . MAXSON, W.S., HANEY, A.F. & SCHOMBERG, D.W. (1985) Steroidogenesis in porcine atretic follicles: loss of aromatase activity in isolated granulosa and theca. Biol. Reprod. 33, 495–501. . MOOR, R.M., DOTT, H.M. & CRAN, D.G. (1978) Macroscopic identification and steroidogenic function of atretic follicles in sheep. J. Endocr. 77, 309–318. . HILLIER, S.G. & ROSS, G.T. (1979) Effects of testosterone on ovarian weight, follicular morphology and intraovarian progesterone concentration in estrogen-primed hypophysectomized immature female rats. Biol. Reprod. 20, 261–268. . PAYNE, R.W., HELLBAUM, A.A. & OWENS, J.N. (1956) The effect ofandrogen on the ovaries and uterus of the estrogen treated hypophysectomized immature rat. Endocrinology 59, 306–316. . LOUVET, J.P., HARMAN, S.M. & ROSS, G.T. (1975) Effects of human chorionic gonadotropin, human intestinal cell stimulating hormone and human follicle-stimulating hormone on ovarian weights in estrogenprimed hypophysectomized immature female rats. Endocrinology 96, 1179–1186. . LOUVET, J.P., HARMAN, S.M., SCHREIBER, J.R. & ROSS, G.T. (1975) Evidence for a role of androgens in follicular maturation. Endocrinology 97, 366–372. . ROBERTS, K.P. & ZIRKIN, B.R. (1991) Androgen regulation of spermatogenesis in the rat. Ann. NY. Acad. Sci. 637, 90–106. . TILLY, J.L., BILLIG, H., KOWALSKI, K.I. & HSUEH, A.J.W. (1992) Epidermal growth factor and basic fibroblast growth factor suppress the spontaneous onset of apoptosis in cultured rat ovarian granulosa cells and follicles by a tyrosine kinase-dependent mechanism. Mol. Endocrinol. 6, 1942–1950. . HSU, C., HOLMES, C.D. & HAMMOND, J. (1987) Ovarian epidermal growth factor-like activity. Concentrations in porcine follicular fluid during follicular enlargement. Biochem. Biophys. 147, 242–247. . WESTERGAARD, L.G. & ANDERSEN, C.Y. (1989) Epidermal growth factor (EGF) in human preovulatory follicles. Hum. Reprod. 4, 257–260. . KHAN-DAWOOD, F.S. (1987) Human corpus luteum: immunocytochemical localization of epidermal growth factor. Fertil. Steril. 47, 916–919. . KUDLOW, J.E., KOBRIN, M.S., PURCHIO, A.F., TWARDZIK, D.R., HERNANDEZ, E.R., ASA, S.L. & ADASHI, E.Y. (1987) Ovarian transforming growth factor-alpha gene expression: immunohistochemical localization to the theca-interstitial cells. Endocrinology. 121, 1577–1579. . NEUFELD, G., FERRARA, N., SCHWEIGERER, L., MITCHELL, R. & GosPODAROWICZ, D. (1987) Bovine granulosa cells produce basic fibroblast growth factor. Endocrinology 121, 597–603. . LOBB, D.K., KOBRIN, M.S., KUDLOW, J.E. & DORRINGTON, J.H. (1989) Transforming growth factoralpha in the adult bovine ovary: identification in growing ovarian follicles. Biol. Reprod. 40, 1087–1093. . JONES, P.B.C., WELCH, JR. & HSUEH, A.J.W. (1982) Regulation of ovarian progestin production by epidermal growth factor in cultured rat granulosa cells. J. Biol. Chem. 257, 11268–11273. . ST-ARNAUD, R., WALKER, P., KELLY, P.A. & LABRIE, F. (1983) Rat ovarian epidermal growth factor receptors: characterization and hormonal regulation. Mol. Cell Endocrinol. 31, 43–52.
APOPTOSIS IN GONADAL TISSUES
86 87 88 89 90
91 92 93
94 95
96 97
98
99
100
101
102 103 104 105
193
. FENG, P., KNECHT, M. & CATT, K.J. (1987) Hormonal control of epidermal growth factor receptors by gonadotropins during granulosa cell differentiation. Endocrinology 120, 1121–1126. . SHIKONE, T., YAMOTO, M. & NAKANO, R. (1992) Follicle-stimulating hormone induces functional receptors for basic fibroblast growth factor in rat granulosa cells. Endocrinology 131, 1063–1068. . ADASHI, E.Y., RESNICK, C.E., D’ERCOLE, A.J., SVOBODA, M.E. & VAN WYK, J.J. (1985) Insulin-like growth factors as intraovarian regulators of granulosa cell growth and function. Endocr. Rev. 6, 400–420. . MURPHY, L.J., BELL, G.I. & FRIESEN, H.G. (1987) Tissue distribution of insulin-like growth factor I and II messenger ribonucleic acid in the adult rat. Endocrinology 120, 1279–1282. . CARLSSON, B., HILLENSJÖ, T., NILSSON, A., TÖRNELL, J. & BILLIG, H. (1993) Expression of insulinlike growth factor-I (IGF-I) in the rat fallopian tube: possible autocrine and paracrine action of Fallopian tubederived IGF-I on the Fallopian tube and on the preimplantation embryo. Endocrinology 133, 2031–2039. . DAVOREN, J.B., KASSON, B.G., LI, C.H. & HSUEH, A.J.W. (1986) Specific insulin-like growth factor I and II binding sites on rat granulosa cells: relation to IGF action. Endocrinology 119, 2155–2162. . HSU, C.J. & HAMMOND, J.M. (1987) Gonadotropins and estradiol stimulate immunoreactive insulin-like growth factor-I production by porcine granulosa cells in vitro. Endocrinology 120, 198–207. . HERNANDEZ, E.R., ROBERTS, C.T.J., LEROITH, D. & ADASHI, E.Y. (1989) Rat ovarian insulin-like growth factor I (IGF-I) gene expression is granulosa cell-selective : 5 -untranslated mRNA variant representation and hormonal regulation. Endocrinology 125, 572–574. . DAVOREN, J.B. & HSUEH, A.J.W. (1986) Growth hormone increases ovarian levels of immunoreactive somatomedin C/insulin-like growth factor I in vivo. Endocrinology 118, 888–890. . CARLSSON, B., CARLSSON, L. & BILLIG, H. (1989) Estrus cycle-dependent covariations of the level of insulin-like growth factor-I (IGF-I) messenger ribonucleic acid and protein in the rat ovary. Mol. Cell. Endocrinol. 64, 271–275. . SHIMASAKI, S. & LING, N. (1991) Identification and molecular characterization of insulin-like growth factor binding proteins (IGFBP-1, -2, -3, -4, -5 and -6). Progr. Growth Factor Res. 3, 243–266. . UI, M., SHIMONAKA, M., SHIMASAKI, S. & LING, N. (1989) An insulin-like growth factor-binding protein in ovarian follicular fluid blocks follicle-stimulating hormonestimulated steroid production by ovarian granulosa cells. Endocrinology 125, 912–916. . BICSAK, T.A., SHIMONAKA, M., MALKOWSKI, M. & LING, N. (1990) Insulin-like growth factor binding protein inhibition of granulosa cell function: effect on cAMP, DNA synthesis and comparison with the effect of an IGF-I antibody. Endocrinology 126, 2184–2189. . ADASHI, E.Y., RESNICK, C.E., RICCIARELLI, E., HURWITZ, A., KOKIA, E., TEDESCHI, C., BOTERO, L., HERNANDEZ, E.R., ROSENFELD, R.G. & CARLSSON-SKWIRUT, C. (1992) Granulosa cell-derived insulin-like growth factor (IGF) binding proteins. J. Clin. Invest. 90, 1593–1599. . ERICKSON, G.F., NAKATANI, A., LING, N. & SHIMASAKI, S. (1992) Localization of insulin-like growth factor-binding protein-5 messenger ribonucleic acid in rat ovaries during the estrous cycle. Endocrinology 130, 1867–1878. . ADASHI, E.Y., RESNICK, C.E., HERNANDEZ, E.R., HURWITZ, A. & ROSENFELD, R.G. (1990) Folliclestimulating hormone inhibits the constitutive release of insulin-like growth factor binding proteins by cultured rat ovarian granulosa cells. Endocrinology 126, 1305–1307. . CATALDO, N.A. & GIUDICE, L.C. (1992) Follicular fluid IGF binding protein profiles in polycystic ovary syndrome. J. Clin. Endo. Metab. 74, 695–697. . CATALDO, N.A. & GIUDICE, L.C. (1992) IGF binding protein profiles in human ovarian follicular fluid correlate with follicular functional status. J. Clin. Endo. Metab. 74, 821–829. . ERICKSON, G.F., NAKATANI, A., LING, N. & SHIMASAKI, S. (1992) Cyclic changes in insulin-like growth factor-binding protein-4 messenger ribonucleic acid in the rat ovary. Endocrinology 130, 625–636. . JONES, P.B. & HSUEH, A.J.W. (1981) Regulation of progesterone-metabolizing enzyme by adrenergic agents, prolactin, and prostaglandins in cultured rat ovarian granulosa cells. Endocrinology 109, 1347–1354.
194
106 107 108
109
110 111
112 113 114 115
116
117
118
119
120 121 122 123 124 125
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. JONES, P.B., CONN, P.M., MARIAN, J. & HSUEH, A.J.W. (1980) Binding of gonadotropin releasing hormone agonist to rat ovarian granulosa cells. Life Sci. 27, 2125–2132. . BIRNBAUMER, L., SHAHABI, N., RIVIER, J. & VALE, W. (1985) Evidence for a physiological role of gonadotropin-releasing hormone (GnRH) or GnRH-like material in the ovary. Endocrinology 116, 1367–1370. . RIVIER, C. & VALE, W. (1989) Immunoreactive inhibin secretion by the hypophysectomized female rat: demonstration of the modulating effect of gonadotropin-releasing hormone and estrogen through a direct ovarian site of action. Endocrinology 124, 195– 198. . ADASHI, E.Y., RESNICK, C.E., VERA, A. & HERNANDEZ, E.R. (1991) In vivo regulation of granulosa cell type IIGF receptors: evidence for an inhibitory role for the putative endogenous ligand(s) of the ovarian GnRH receptor. Endocrinology 128, 3130– 3137. . OIKAWA, M., DARGAN, C.M., NY, T. & HSUEH, A.J.W. (1990) Expression of gonadotropin-releasing hormone and prothymosin-alpha messenger ribonucleic acid in the ovary. Endocrinology 127, 2350–2356. . WANG, J., STEELE, G.L., BAIMBRIDGE, K.G., RODWAY, M.R. & LEUNG, P.C. (1992) Intracellular calcium and the signaling mechanism of luteinizing hormone-releasing hormone in rat granulosa cells. Am. J. Obstet. Gynecol. 167, 541–547. . NAOR, Z. & YAVIN, E. (1982) Gonadotropin-releasing hormone stimulates phospholipid labeling in cultured granulosa cells. Endocrinology 111, 1615–1619. . ROY, S.K. & GREENWALD, G.S. (1991) Mediation of follicle-stimulating hormone action on follicular deoxyribonucleic acid synthesis by epidermal growth factor. Endocrinology 129, 1903–1908. . GOROSPE, W.C. & SPANGELO, B.L. (1993) Interleukin-6; potential roles in neuroendocrine and ovarian function. Endocrine J. 1, 3–9. . LIU, Y.J., MASON, D.Y., JOHNSON, G.D., ABBOT, S., GREGORY, C.D., HARDIE, D.L., GORDON, J. & MACLENNAN, I.C. (1991) Germinal center cells express bcl-2 protein after activation by signals which prevent their entry into apoptosis. Eur. J. Immunol. 21, 1905–1910. . HOCKENBERY, D.M., ZUTTER, M., HICKEY, W., NAHM, M. & KORSMEYER, S.J. (1991) Bcl-2 protein is topographically restricted in tissues characterized by apoptotic cell death. Proc. Natl Acad. Sci. USA 88, 6961–6965. . REED, J.C., MEISTER, L., TANAKA, S., CUDDY, M., YUM, S., GEYER, C. & PLEASURE, D. (1991) Differential expression of bcl2 protooncogene in neuroblastoma and other human tumor cell lines of neural origin. Cancer Res. 51, 6529–6538. . EGUCHI, T., EWERT, D.L. & TSUJIMOTO, Y. (1992) Isolation and characterization of the chicken bcl-2 gene: expression in a variety of tissues including lymphoid and neuronal organs in adult and embryo. Nucleic Acids Res. 20, 4187–4192. . KERR, J.B., RISBRIDGER, G.P., MURRAY, P.J. & KNELL, C.M. (1978) Effect of unilateral cryptorchidism on the intertubular tissue of the adult testis: evidence for intracellular changes within the Leydig cells. Int. J. Androl. 11, 209–223. . AU, C.L., ROBERTSON, D.M. & DE KRETSER, D.M. (1983) In vitro bioassay of inhibin in testes of normal and cryptorchid rats. Endocrinology 112, 239–244. . SWEDLOFF, R.S., WALSH, P.C., JACOBS, H.S. & ODELL, W.D. (1971) Serum LH and FSH during sexual maturation in the male rat: effect of castration and cryptorchidism. Endocrinology 88, 120–128. . AMATAYAKUL, K., RYAN, R., UOZUMI, T. & ALBERT, A. (1971) A reinvestigation of testicular-anterior pituitary relationships in the rat: I. Effects of castration and cryptorchidism. Endocrinology 88, 872–880. . ALLAN, D.J., HARMON, B.V. & KERR, J.F.R. (1987) Cell death in spermatogenesis. In: Perspective on mammalian cell death. Oxford University Press, London, pp. 229–258. . BUTTYAN, R., ZAKERI, Z., LOCKSHIN, R. & WOLGEMUTH, D. (1988) Cascade induction of c-fos, c-myc, and heat shock 70 K transcripts during regression of the rat ventral prostate gland. Mol. Endocrinol. 2, 650–657. . RIABOWOL, K.T., MIZZEN, L.A. & WELCH, W.J. (1987) Heat shock is lethal to fibroblast microinjected with antibodies against hsp 70. Science 242, 433–436.
APOPTOSIS IN GONADAL TISSUES
126 127 128 129
130 131 132 133 134
195
. LEMAIRE, L. & HEINLEIN, U.A.O. (1990) Detection of secreted and temporarily inducible heat shock responsive proteins in mouse testicular tissue. Life Sci. 48, 365–372. . ZAKAERI, S.F., WELCH, W.J. & WOLGEMUTH, D.J. (1990) Characterization and inducibility of hsp 70 proteins in the male mouse germ line. J. Cell Biol. 111, 1785–1702. . ITOH, H. & TASHIMA, Y. (1991) Different expression time of the 105/kDa protein and 90-kDa heat-shock protein in rat testis. FEBS 289, 110–112. . ROSARIO, M.O., PERKINS, S.L., O’BRIEN, D.A., ALLEN, R.L. & EDDY, E.M. (1992) Identification of the gene for developmentally expressed 70 kDA heatshock protein (P70) of mouse spermatocytogenic cells. Dev. Biol. 150, 1–11. . HSUEH, A.J.W., BILLIG, H. & TSAFRIRI, A. (1994) Ovarian follicle atresia: hormonally controlled apoptotic process. Endocr. Rev. 15, 707–724. . HECKERT, L.L. & GRISWALD, M.D. (1991) Expression of follicle-stimulating hormone receptor mRNA in rat testes and Sertoli cells. Mol. Endocrinol. 5, 670–677. . BARDIN, C.W., CHENG, C.Y., MUSTO, N.A. & GUNSALUS, G.L. (1988) The Sertoli cell. In: The physiology of reproduction, KNOBIL, E. & NEIL, J. (eds). Raven Press, New York, pp. 933–974. . HSUEH, A.J.W., ADASHI, E.Y., JONES, P.B.C. & WELSH JR., T.H. (1984) Hormonal regulation of the differentiation of cultured ovarian granulosa cells. Endocr. Rev. 5, 76–127. . VORNBERGER, W., PRINS, G., MUSTO, N.A. & SUAREZ-QUIAN, C.A. (1994) Androgen receptor distribution in the rat testis: new implications for androgen regulation of spermatogenesis. Endocrinology 134, 2307–2316.
11 Apoptosis in the Immune System CHRISTOPHER D.GREGORY Department of Immunology, University of Birmingham Medical School, Birmingham, B15 2TT, UK
11.1 Introduction In recent years, we have witnessed an enormous upsurge of interest in cell death as a physiological process, and the immune system has provided one of the most important arenas in which the field has developed. By the same token, the steadily increasing body of knowledge of the molecular mechanisms which control survival, death and disposal of cells has led to a better understanding of the immune system. This chapter will attempt to provide a broad overview of apoptosis as it relates to the immune system, taking into account elements of both innate and acquired immunity. Particular attention will be paid to the selection of lymphocytes prior to the initiation of immune responses, and also to the control of cell numbers during (and after) response to antigen. The cytotoxic effector mechanisms of lymphocytes will also be considered. Discussion of the innate immune system will focus on the survival of neutrophils and the clearance of apoptotic cells by macrophages. Finally, failure of the normal controls of cell survival, death and clearance will be considered as contributory factors in the pathogenesis of diseases of the immune system: autoimmunity, immunodeficiency and neoplasia. Historical perspective will not be a major aim of this chapter; where appropriate the reader will be referred to additional reviews, a number of which have been published recently (for example see reference [1] for a historical overview of apoptosis in the immune system). 11.2 Regulation of Apoptosis in Lymphocyte Selection During Development 11.2.1 Development of T Lymphocytes in the Thymus The majority of cells of the T lineage mature in the thymus and express on their surface a heterodimeric receptor for antigen (TCR) composed of disulphide-linked and chains in combination with CD3 [2]. A minor subset of T cells which express heterodimeric TCR also develops in the thymus, but its
APOPTOSIS IN THE IMMUNE SYSTEM
197
Figure 11.1 Highly simplified scheme of T cell development in the thymus showing points of entry into apoptosis (†). TCR- and - gene rearrangements are indicated by labelling of the nucleus (0=germline, R=rearranged). Surface expression of TCR- , - , X (gp33 [8]), CD4 and CDS polypeptides is also indicated. †1=default or glucocorticoidinduced death; †2=antigen receptor-induced death.
maturational pathway(s) and the nature of the antigens recognised are less well defined than those of TCR-expressing T cells [3]. Mature T cells fall into two functional categories depending upon which of the two co-receptors, CD4 or CDS, they express on their surface. Thus, TCR of CD4+ T cells recognise protein antigens presented as peptides in the context of class II major histocompatibility complex (MHC) antigens and the majority of these cells function as helper/inducer T cells. CD8+ T cells recognise peptides held by class I MHC molecules and generally function as cytotoxic cells. In the thymus, large numbers of immature T cells are generated from relatively few precursors and during their development undergo a series of selection steps resulting in the production of mature cells that display a diverse repertoire of TCR specificities. The selection pressures favour maturation of thymic lymphocytes (thymocytes) displaying TCR that interact with self MHC antigens, but militate against successful maturation of cells expressing TCR that recognise self peptides [4]. The majority of the vast number of immature cells produced in the thymic cortex fail to survive the rigours of selection and die in situ. The signals which determine thymocyte fate are generated through the clonotypic TCR molecules in conjunction with a variety of additional thymocyte surface receptors (including CD3, CD4, CDS, IL-2R, CD2, CD28, CD44, CD45, LFA-1 and Fas) that interact with ligands produced in the thymic microenvironment. The precise details of the nature and outcome of these interactions are under intense investigation (reviewed in [5, 6]), but it is widely accepted that thymocyte development proceeds in an ordered sequence delineated both by the histological architecture of the thymus and by the pattern of molecules, including TCR polypeptides, displayed on the thymocyte surface (Figure 11.1). Details of the sequential genetic and phenotypic changes that occur during thymocyte maturation have been extensively reviewed [4, 5, 7]. However, in order to appreciate the role of apoptosis in regulating the massive numbers of cells produced in the thymus, the developmental stages defined so far will be considered here in broad outline (Figure 11.1). For the T cell lineage, maturation in the thymus can by
198
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
divided into three phases according to the expression on the thymocyte surface of the TCR chains: (1) the early phase, which is not regulated by the TCR, (2) the intermediate phase, which is dependent upon expression of the chain of the TCR, and (3) the late phase, dependent upon expression of both and TCR chains [5]. (1) In the early phase, the intrathymic T cell precursors become committed to the T lineage and undergo rearrangement of TCR genes. Almost nothing is yet known about the mechanisms which underlie commitment of the earliest T cell precursors, or of the environmental factors which regulate rearrangement of TCR genes and expression of TCR polypeptide chains. Rearrangement and expression of TCR- and genes precede that of and . Cells which initiate TCR- gene rearrangement, a process that occurs before gene rearrangement, are CD2+IL-2R +CD4−CD8− in surface phenotype and may also express surface CD3 . Once a TCR- chain locus has been productively rearranged, allelic exclusion occurs and further TCR- gene rearrangements are inhibited. (2) A strong body of evidence from studies of transgenic SCID (severe combined immunodeficiency, i.e. rearrangement-deficient) and RAG (recombinase activating gene) -deficient mice into which TCRtransgenes were introduced, indicates that successful expression of the TCR- chain is critical for the continued development of thymocytes along the lineage [8]. Transit to the next developmental stage, the so-called ‘double-positive’ stage is characterised by the simultaneous surface expression of each of the coreceptors, CD4 and CD8 (although CD8 appears before CD4) and an increase in cell number. The TCRpolypeptide has been found to be expressed on the thymocyte surface as a disulphide-linked heterodimer in combination with a novel 33 kDa glycoprotein that functions as a ‘surrogate’ TCR- chain (cf. the VpreB and 5 products which act as surrogate light chains during B cell development; see section 11.2.2). It has been suggested that TCR- /gp33, which also associates with CD3 chains, constitutes a signal transduction complex that may be important in interacting with a thymic environmental ligand to generate signals responsible for allelic exclusion and transition to the next developmental stage [8]. (3) The TCR-dependent phase of thymocyte maturation follows productive rearrangement of a TCRlocus in a cell which expresses TCR- chains. TCR- chains are expressed at the cell surface after heterodimerisation with TCR- polypeptides, although the point at which the switch from TCR- /gp33 to TCRoccurs at the thymocyte surface has yet to be determined. The CD4+CD8+TCR- + blast population which, in the adult thymus, accounts for only a small percentage (<5%) of the cells, gives rise to the major population of thymocytes (representing >80% of all thymocytes) [5]. These small, resting doublepositive cells (CD4+CD8+) are located in the cortex of the thymus and express low levels of TCR on their surface. High-level expression of surface TCR follows the process of positive selection, a phase of thymocyte differentiation that is dependent upon interaction of the TCR with MHC molecules, and results in the emergence of single-positive (CD4+or CD8+) cells. Since CD4+cells display a propensity to interact with MHC class II molecules and CD8+cells interact with MHC class I molecules, it has been proposed that downregulation of either CD4 (from MHC class I-interactive thymocytes) or CD8 (from MHC class IIinteractive cells) occurs as a result of an instructive signal which follows the MHC interaction. An alternative proposal (for which evidence has also been provided) is a stochastic/selective model which predicts that MHC class I-interactive or class II-interactive cells will only survive if they express CD8 or CD4, respectively, the requirement for survival being co-receptor crosslinking [9]. What controls are imposed on the apoptotic program of a thymocyte during its differentiation? An attractive theory of control of cell death and survival, which may be true of all cells in all tissues, is that death is truly programmed and occurs by default unless factors capable of generating survival signals are present in the cellular microenvironment [10]. The thymus has been described as an organ which ‘selects the useful, neglects the useless and destroys the harmful’ [4]. Useless cells, for example those which are
APOPTOSIS IN THE IMMUNE SYSTEM
199
unsuccessful in making productive TCR gene rearrangements, or which express TCR that fail to interact with MHC molecules, do not receive the requisite survival signals and undergo death by default. Thus progression to the next stage of differentiation is the key to thymocyte survival. The long-standing observation that cortical thymocytes rapidly undergo apoptosis in response to glucocorticoid hormones, coupled with the finding that the thymus is greatly enlarged in adrenalectomised mice may indicate that the massive death of cortical thymocytes observed in situ represents glucocorticoid-induced death of ‘neglected’ cells [5]. Activation of the apoptotic programme of thymocytes, in addition to its role in disposing of useless cells, is also critically important in the induction of self-tolerance (‘destruction of the harmful’). It has been clearly demonstrated that signals generated at TCR of immature (double- and single-positive) cells rapidly induce the apoptotic self-destruction of thymocytes. Antibodies to CD3 have been observed to activate apoptosis in double-positive cells in vitro in mouse foetal thymic organ cultures [11]. In vivo administration of CD3 antibody was also found to induce apoptosis in virtually all CD4+CD8+ thymocytes and in approximately 50% of CD4+CD8−cells [12]. Direct induction of apoptosis in CD4+CD8+cells by antigen was shown in a transgenic mouse system in which virtually all thymocytes and peripheral T cells express a TCR that recognises a chicken ovalbumin peptide antigen [13]. In these experiments, intraperitoneal administration of the peptide caused rapid deletion of double-positive thymocytes. In another transgenic model, in which double-positive thymocytes expressing a TCR specific for the male (H-Y) antigen were used, Kisielow and colleagues showed that apoptosis of double-positive cells from transgenic female mice occurred rapidly after culture with thymic or splenic antigen-presenting cells from MHC-matched male animals [14]. These workers subsequently demonstrated that all subsets of TCR-expressing thymocytes, whether double- or single-positive, are sensitive to TCR-triggered apoptosis (reviewed in [5]). The apoptotic response of immature thymocytes to signals generated by TCR-antigen interaction contrasts starkly with the mitotic response of mature T cells to TCR-transduced signals. This paradox will be considered later (see section 11.3.1). In the thymus, a further paradox of TCR signalling is that of positive and negative selection. Positive selection, as already described, is the process which favours survival of thymocytes expressing TCR that interact with self MHC molecules. Negative selection, by contrast, is apoptotic deletion of thymocytes which express TCR recognising self antigens—i.e. self peptides in the context of self MHC molecules. Three main hypotheses have been tested in order to understand how the thymus provides the education signals, which must involve newly-expressed TCR, to promote survival of cells interacting with self MHC molecules yet at the same time to induce apoptosis in cells which recognise self antigen in the context of the same MHC molecules [15]. Firstly, outcome (positive versus negative selection) may be dependent upon the lineage of the thymic stromal cells (epithelial or bone marrowderived) with which the thymocyte interacts. Secondly, sensitivity to selection signals may be developmentally linked, with stages susceptible to positive selection preceding those responsive to negative selection. Thirdly, the generation of positive and negative selection signals may be dependent upon the affinity/avidity of the interaction between TCR and MHC (plus peptide), with high affinity/avidity interactions leading to induction of apoptosis and low affinity/avidity interactions favouring survival. Most of the recent experimental evidence supports the third model (reviewed in [5, 16]) and indicates that both positive and negative selection of thymocytes can occur at the same developmental stage and that the response pathway is determined by the quality of the interaction of the TCR, CD4 and CD8 co-receptors as well as additional ancillary molecules, with available thymic ligands. This argument therefore implies that the basis for selection of thymocytes is the divergence of intracellular signalling pathways initiated by the newly-generated TCR. There is evidence that the coupling of second messenger pathways to TCR in maturing thymocytes may change as a result of positive selection [17]. Furthermore, expression of nur77, the
200
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 11.2 Simplified scheme of progression or death during initial maturation in the B lineage. Heavy (H) and light (L) chain gene rearrangements are indicated in the nucleus (0=germline; DJ, VDJ, VJ rearrangements as indicated). Antigen receptor components SL (surrogate light chains, VpreB, 5), µ heavy chains and and Ig polypeptides are also shown as well as the X chains (p130/p55 [28] which are associated with the SL at the surface during the early stages of development) and IL7 receptors (IL7R). †1=default death; †2=antigen receptor-induced death.
orphan steroid receptor which functions as a transcription factor, appears to provide a hallmark of the apoptotic response of thymocytes to TCR cross-linking. The differential expression of this protein in TCRtriggered apoptotic and proliferating T cells suggests that nur77 marks a point of divergence in TCRgenerated intracellular signalling [18, 19]. Detailed understanding of the biochemistry of positive and negative selection of thymocytes requires further experimental investigation. What is known of the genetic basis of thymocyte survival? Although there is some evidence that c-myc and p53 play roles in driving apoptosis during T cell development, such evidence is so far only indirect [20–22]. Investigations into the repression of apoptosis by the proto-oncogene, bcl-2, the product of which is known to act as an ‘apoptosis-antidote’ in a variety of cellular systems (see Reed, this volume) suggest that bcl-2 regulates thymocyte survival during maturation. Thus, whereas mature T cells of the thymic medulla, like peripheral T cells, mainly express relatively high levels of bcl-2, double-positive cortical thymocytes which express low levels of TCR on their surface only express low levels of bcl-2 [23]. Interestingly, immature double-negative thymocytes show high-level bcl-2 expression, suggesting that the apoptosis-repressor is required for the survival of immature cells prior to TCR gene rearrangement and surface TCR expression. Experiments with bcl-2 knock-out mice, however, have shown that bcl-2 is not essential for selection of T cells in the thymus, but rather has a much more important role to play in promoting survival of mature T cells in the periphery [24, 25]. Indeed, expression of a bcl-2 transgene in SCID mice failed to remove the block on T cell development (i.e. inability to make productive TCR gene rearrangements) exhibited by these mice. Furthermore, the finding that production of normal numbers of double-positive thymocytes is restored by a TCR transgene in both SCID and bcl-2/SCID mice strongly suggests that T cell sensitivity to control of apoptosis by Bcl-2 does not occur until after cell surface expression of the TCR [26]. In summary, lymphocyte development in the thymus involves the generation of large numbers of cells, the majority of which are harmlessly removed by apoptosis either by default because they fail to express a receptor for antigen or because their TCR are ‘inappropriate’ (i.e. do not interact with self MHC molecules or interact with too high an affinity). The capacity to survive is critically dependent upon signals generated at TCR and is also likely to involve a host of additional molecules (e.g. CD2, CD4, CD8, CD45, CD28, LFA-1, IL-2R, Fas), the ligands for which are expressed on or by thymic stromal elements. The
APOPTOSIS IN THE IMMUNE SYSTEM
201
mechanisms underlying thymocyte survival are complex and the integration of signals originating at TCR and auxiliary molecules cause the repression of apoptosis largely independently of Bcl-2. 11.2.2 Development of B Lymphocytes in the Bone Marrow The enormously diverse receptor repertoire of B lymphocytes, like that of T cells, is initially generated in the absence of exogenous antigen by recombination of a relatively small number of antigen receptor gene segments [27]. The immunoglobulin (Ig) antigen receptors of B cells, unlike TCR, can interact with antigens in their native forms and, under appropriate conditions, B cells differentiate to secrete antibodies, soluble Ig molecules of identical specificity to the surface Ig receptors. In mammals, B cell development is initiated in foetal life at numerous locations, but in adults is maintained in the bone marrow whence, in a mouse, around 3 million cells exit each day to recirculate along defined migratory pathways through the blood, lymphoid tissues and lymphatics [28]. Distinct stages in the maturation of B cells in the bone marrow of mice have been identified according to their commitment to the B lineage, the recombination status of their Ig gene segments, and their expression of antigen receptor components (Figure 11.2) (reviewed in [28, 29]). Taking the nomenclature of Melchers and colleagues [28, 29], in the earliest cells that are committed to the B lineage, the ‘pro-B' cells, the Ig gene segments (VH, DH and JH for the heavy chains and VL and JL for the light chains) remain in germline configuration. These cells probably divide asymmetrically to produce one daughter that retains the features of a pro-B cell (and maintains the size of the pro-B cell pool—around 0.5 to 1 million cells) while the other undergoes DH and JH rearrangements to become a ‘pre-B-F cell. The pool of pre-B-I cells (around 2 million cells) expands under the influence of IL-7 and other stroma-derived factors and differentiates further (i.e. undergoes VH rearrangements) to produce the ‘pre-B-II population (about 50 million cells). The ‘immature B’ cell population is derived from pre-B-II cells that have undergone rearrangements at light chain loci and comprises approximately 20 million cells. Of these, only a fraction reach full maturity: around 3 million per day leave the bone marrow to enter the circulating mature B cell pool [28, 29]. In B cells, expression of antigen receptors requires the synthesis and association of the Ig heavy chains (usually of the µ isotype) and light chains (K or ) after productive recombination of VH, DH, JH with Cµ as well as VL, JL with CK or C gene segments. B cell receptors composed of µ heavy chains in combination with K or light chains do not appear at the cell surface until the immature B cell stage. At the pre-B cell stages, light chain genes are in germline configuration and recombination of heavy chain gene segments is under way or, in the later phase, complete. Such cells express µ heavy chains not only in their cytoplasm, but also, as a complex with the surrogate light chain components VpreB and 5, at the cell surface. Indeed, surrogate light chains are expressed at the cell surface during all progenitor and precursor B cell developmental stages. At the pro-B cell stage they are associated with ‘surrogate’ heavy chains p55 and p130, while on pre-B cells they are associated with DHJHCµ polypeptides (formed after certain DJ recombinations which fall into a reading frame that allows transcription of a heavy chain with an incomplete variable region) as well as with complete (VHDHCµ ,) chains [28, 29]. It is clear from the population dynamics of the B cell developmental pathway of the average mouse that large numbers of cells are produced and lost at the pre-B cell stages and that cell deletion continues at high frequency during the immature B cell stage prior to release from the bone marrow. Two mechanisms have been proposed to account for loss of B lineage cells during development in the bone marrow: classical apoptosis, and rapid internalisation by complex resident macrophages [30]. The relationship, if any, between macrophage deletion and the rapid phagocytic clearance that follows programmed cell death (see
202
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
section 11.5.2) has not been reported. Just as for T cells, successful B cell maturation is critically dependent upon sequential success in rearrangement and expression on the cell surface of antigen receptor molecules. Death by default is therefore expected of cells that fail to synthesise firstly heavy chain, and secondly light chain, polypeptides. It is not yet known how the surrogate light chains, complexed with their various partners on the surface of pro- and pre-B cells, interact with the bone-marrow microenvironment or how signals generated through such interactions affect B cell survival. However, it has been proposed [28] that, since pre-B-II cells appear to be positively selected for the expression of µ heavy chains, the expression of such chains in combination with surrogate light chains at the cell surface may allow the cells to interact with microenvironmental ligands (possibly including self and foreign antigens) and thereby receive positive or negative selection signals. Early B cells that express antigen receptors, like immature T cells, do not proliferate in response to antigen. Rather, they are either induced to die or undergo functional deletion. Studies of antigen receptortransgenic mice have elegantly shown that exposure of B cells to certain antigens during their initial development in the bone marrow results in their elimination. Nemazee and colleagues demonstrated that B cells expressing a receptor that interacts with membrane-bound MHC class I molecules were clonally deleted, either in the bone marrow or soon after exit to the periphery [31, 32]. Other membrane-bound autoantigens, as well as double-stranded DNA, have also been found to cause clonal deletion of B cells [33–36]. Elegant experiments by Goodnow and co-workers showed that the response of early B cells to autoantigen was dependent on the form in which the antigen was presented : membrane-bound antigen induced deletion whereas soluble antigen led to clonal anergy (i.e. functional unresponsiveness) [33, 37]. Elimination of autoreactive cells from the bone marrow appears to operate in two stages that occur in sequence [38]. Firstly, autoreactive cells are developmentally blocked such that they fail to make the transition from the immature to the mature B cell stage and consequently are restricted in their expression of surface molecules involved in migration and activation in the periphery. This developmental block, which was found to be reversible if antigen was removed, leads to cell death, presumably by apoptosis. That cell deletion occurs by apoptosis is strongly implied by the observation that expression of the apoptosisrepressing oncogene, bcl-2, extended survival of autoreactive, developmentally blocked cells [38]. A recently-reported mechanism by which autoreactive B cells can avoid elimination is through functional deletion by antigen receptor editing [39–41]. This process involves the modification of antigen receptors of surface IgM+ early B cells by further recombination of their light chain variable region genes. A change in receptor specificity may therefore prevent autoantigen-induced death. As mentioned previously, recognition of self antigen by newly-produced B cells may induce functional silencing, also known as clonal ‘anergy’, rather than death. In their original experiments, Goodnow and colleagues demonstrated, using a transgenic mouse system in which all B cells expressed a receptor for hen egg lysozyme (HEL), that in doubletransgenic animals which in addition secrete HEL, all the B cells became anergic. These cells were apparently able to recirculate normally through secondary lymphoid tissues and home, as normal B cells, to the B cell areas (primary follicles) but were unable to clonally expand and produce antibody [33, 42, 43]. More recently, the same group has modified its transgenic systems to emulate normal B cell physiology more closely. Thus, when autoreactive, anergic cells recirculate with normal B cells, they are selectively excluded from entry into the follicles of the lymphoid tissues and, as a result, undergo apoptosis prematurely [44]. Expression of bcl-2 inhibited death of the autoreactive cells, but did not affect entry into follicles. These results have important implications for understanding the paradoxical link between autoimmunity and immunodeficiency and also indicate that follicles provide an important microenvironment for B cell survival (see section 11.2.3 below).
APOPTOSIS IN THE IMMUNE SYSTEM
203
Apoptosis is therefore likely to play an important role in selection of B cells during and soon after their generation in the bone marrow. Maturation blocks due to failure of expression of antigen receptor components may lead to apoptosis by default. Failure to express receptors necessary to interact with bone marrow stromal factors such as IL-7 or inability to compete for space in order to access surface-bound stromal molecules may also be important in the control of B cell survival in the bone marrow [30, 45, 46]. The CD40 molecule, a member of the nerve growth factor receptor family, is expressed by B lineage cells from the pro-B cell stage onwards [47]. Engagement of this surface receptor by a ligand displayed on activated T cells is known to generate survival signals for mature B cells (see section 11.2.3). Whether signals derived from CD40 interacting with a ligand expressed in the bone marrow microenvironment affect the survival of developing B cells is not yet known. Bcl-2, however, which can be expressed as a result of CD40-induced survival of mature B cells [48, 49], may promote survival during B cell maturation. Thus, recent data indicate that levels of Bcl-2 protein are high at the pro-B cell stage, downregulated at the pre-B and immature B cell stages and return to high levels in mature B cells [50]. These results imply that downregulation of Bcl-2 may be important to permit antigen receptor-based selection to proceed, although experiments with murine cell lines which are thought to emulate immature B cells have shown that Bcl-2 is not sufficient to prevent antigen receptor-triggered apoptosis [51]. Bcl-2-transgenic mice, however, do display enhanced survival of both pre- and immature B cell compartments [26, 50, 52]. In bcl-2 knock-out mice, B cell maturation initially proceeds, but is not sustained [24, 25]. 11.2.3 B-1 B Cells B-1 B cells are distinguished from conventional B cells, not only by expression of the surface marker Ly-1 (CD5+) but also by differences in distribution and function (see [53] and [54] for reviews). B-l cells are the predominant subset of B cells present in foetal life, but in adults represent only a minor component of the B cells of the peripheral blood, lymph nodes and spleen, although they predominate in the pleural and peritoneal cavities. B-l B cells have been found to produce low affinity IgM antibodies of broad specificity encoded by germline V-region genes and their clear involvement in mucosal immunity has led to the suggestion that their antibodies provide a first line of defence at mucosal sites [53]. Whilst conventional B cells are continually produced in the bone marrow throughout adult life, B-1 B cells do not appear to arise, in adult animals, from bone-marrow precursors and it has been suggested that they comprise a B-lineage distinct from that of conventional B cells. B-1 cells also exhibit self-replenishing activity. As described previously, conventional B cells that display self-reactive antigen receptors are likely to be eliminated by activation of their apoptotic programme either during their maturation in the bone marrow or soon after entry into the circulating pool. Recent data have indicated that autoreactive B-1 B cells, by contrast, are capable of escaping such elimination by residing in peripheral niches, away from certain self antigens. Thus, in an anti-RBC autoantibody transgenic mouse model, it was shown that B-1 B cells displaying autoreactivity, unlike the conventional B cells, escaped deletion and expanded in the peritoneal cavity [36, 54, 55]. Although by its very nature this model is an artificial one, particularly since virtually all the B cells are RBC autoantibody-specific and all conventional B cells are deleted, it does serve to illustrate a number of properties regarding B-1 B cell survival. The autoreactive B-1 cells in these mice are not inherently insensitive to apoptotic signals resulting from exposure to antigen, since apoptosis is induced rapidly after i.p. injection of RBC. This suggests that the survival of these cells is dependent upon their ability to occupy sites in which they are effectively ‘hidden’ from the autoantigen. The mechanisms by which autoreactive B-1 cells escape deletion and migrate to appropriate niches from which particular self
204
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
antigens are excluded, are not known. Furthermore, no information is yet available on cellular or microenvironmental factors that contribute to the survival of B-l cells at these sites. However, the ability of such cells to escape from autoantigen-induced apoptosis has important implications for the development of autoimmune disease [54]. 11.3. Apoptosis and the Modulation of Immune Responses 11.3.1 Antigen-induced Apoptosis of Mature T Cells It has become increasingly clear over the past 5 years that clonal deletion of antigen-reactive T cells by apoptosis plays an important role in the limitation and termination of ongoing immune responses and in the acquisition of peripheral tolerance. Although apoptosis of T cells occurs by default as a result of, for example, lymphokine withdrawal [56], and can be activated by non-TCR molecules such as Fas (see section 11.3.3), evidence that TCR-antigen interactions are involved in inducing apoptosis in mature T cells has been obtained in a number of systems. Thus, it has been shown that asynchronous ligation of CD4 and TCR by antibodies induces apoptosis of murine splenic T cells [57]. Chronic exposure to low doses of exogenous superantigen results in deletion of mature T cells in vivo in the absence of proliferation [58]. Cycling T cells, under appropriate conditions, also undergo apoptosis as a result of TCR engagement [59]. The observation that activated mature T cells are primed to undergo apoptosis upon re-encounter with antigen has been made in several studies (reviewed in [60]). Lenardo and colleagues have introduced the term ‘propriocidal’ regulation to describe the physiological processes that underlie the phenomenon. In the original studies, it was shown, using both in vitro and in vivo models, that activated T cells were rapidly induced into apoptosis by TCR engagement and that the apoptotic response of the cells was critically dependent upon prior exposure to IL-2 [61]. Subsequent work showed that IL-4 could be substituted for IL-2 in this system and indicated that the apoptotic response to TCR activation strictly required cycling T cells. Indeed, results of cell-cycle blocking experiments suggested that susceptibility to TCR-triggered apoptosis was associated with the capacity to enter S phase [61, 62]. Strong additional evidence has led to the proposal that propriocidal regulation provides a mechanistic basis for the long-established phenomenon of high-dose antigen tolerance [63] and implies that suppression of immune responses by high antigen doses is caused by an intrinsic property of antigen-reactive T cells (i.e. the capacity to undergo apoptosis in response to antigen), obviating the need to invoke additional regulatory elements such as suppressive networks or veto cells [60]. Whilst high doses of antigen might induce a state of tolerance through propriocidal apoptosis of the reactive T cells, the response of mature T cells to lower levels of antigen would be predicted to be qualitatively different. Thus, in the absence of excess antigen the proliferating T cell would be expected to leave the cell cycle and re-enter the resting state. To maintain T cell homeostasis, much of the initially reactive population must die away, but some survive to become memory T cells [64]. Recent results suggest that downregulation of Bcl-2 could provide a critical component of human T cell activation which allows for the physiological removal of reactive cells by apoptosis and for the selection of memory cells [64, 65]. To explain the paradox that T cells of memory phenotype (cells that have changed cell surface expression of leucocyte common antigen (CD45) isoform from CD45RA to CD45RO) are low in Bcl-2 and susceptible to apoptosis, it has been suggested that the memory T cell pool (CD45RO+) is maintained by continuous restimulation with consequent upregulation of Bcl-2. In support of this notion, IL-2 was found both to
APOPTOSIS IN THE IMMUNE SYSTEM
205
Figure 11.3 Highly schematic representation of a secondary follicle (adapted from [67]). D=dark zone of the germinal centre containing centroblasts (CB); L=light zone of the germinal centre containing a dense network of follicular dendritic cells (FDC), centrocytes (CO undergoing selection and apoptotic centrocytes (†); M=mantle zone containing small, recriculating slgM+slgD+B cells (B).
repress apoptosis and to induce upregulation of Bcl-2 in CD45RO+ cells in vitro. In addition, the observation that CD45RO+ T cells can be maintained on fibroblasts in the absence of Bcl-2-upregulation has led to the proposal that stromal factors may be important for the survival of CD45RO+ Bcl-2lo T cells in vivo [64, 65]. 11.3.2 Affinity Maturation of T-dependent B Cell Responses
B cell responses to T-dependent antigens, i.e. antibody responses that require T cell help, are characterised by a further process of selection that increases productivity of higher affinity antibodies. The process occurs in the germinal centres of secondary lymphoid follicles and involves the hypermutation of immunoglobulin V-region genes (reviewed in [66, 67]) (Figures 11.3,11.4). A proportion of B cells responding to antigen in the T cell-rich extrafollicular areas of secondary lymphoid tissues enter the follicles (for reasons which have yet to be established) which, in unimmunised animals, comprise a meshwork of antigen-presenting cells known as follicular dendritic cells (FDC) together with small recirculating B cells expressing IgM and IgD on their surface. The newly-arrived antigen-reactive B blast cell from the T zone, then proliferates so rapidly (with a doubling time of only about 6–7 h) that the follicle is soon filled with daughter progeny which push the population of small resting IgM+IgD+B cells to the periphery of the follicle, creating the mantle zone. The zone of proliferating B blasts, which are IgM+ IgD−, generate the germinal centre of the follicle, a histologically distinct and complex structure in which the subsequent stages of hypermutation, selection and differentiation take place. Histologically, with the follicle orientated such that the mantle zone lies at the top, the germinal centre can be divided (Figure 11.3) into a dark zone at the bottom, above which is the light zone composed of basal and apical regions [67]. The dark zone contains the progeny of the B blasts, now known as centroblasts, which have downregulated their antigen receptor expression to display little or no surface Ig. These cells proliferate asymmetrically to produce daughter cells which either continue to proliferate in the
206
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 11.4 Schematic representation of the antibody response to a T cell-dependent (i.e. protein) antigen (Ag) occurring in the extrafollicular T cell zones and in the germinal centres of the follicles of secondary lymphoid tissue. IDC=interdigitating cell; FDC=follicular dendritic cell; CB=centroblast; CC=centrocyte. See text for details.
centroblast pool, or exit from the cell cycle to produce the non-dividing centrocytes that re-express surface Ig and populate the light zone. The process of somatic hypermutation of V-region genes begins in the centroblasts. The random nature of this process necessitates a mechanism of tight selection of cells expressing appropriate mutant receptors. Thus, it would be predicted that cells expressing mutant receptors with high affinity for the antigen which had elicited the response would be positively selected, while those failing to re-express receptors as a result of missense mutations, or expressing low affinity or self-reactive mutant receptors would be eliminated. A large body of evidence derived from germinal centre reactions in vivo and also the behaviour of isolated germinal centre cells in vitro indicates that germinal centre B cells are primed for apoptosis and that the positive selection of cells bearing appropriate mutant receptors results in the generation of signals that block cell death [48, 49, 66–69]. Germinal centre B cells isolated from human tonsils according to their low buoyant density and their lack of expression of IgD and CD39 rapidly undergo apoptosis when placed in culture at 37ºC; after about 18 h all cells have succumbed. Death can be delayed for about 24–36 h if the cells are exposed to immobilised antiIg antibodies [68]. The finding that immobilised, but not soluble, antibody promoted survival of germinal centre B cells has been taken as reasonably strong evidence that this in vitro system, in which germinal centre B cells are exposed to anti-Ig antibody coated onto sheep red blood cells, effectively mimics the in vivo situation in which the B cells of the germinal centre interact with antigen held in the form of immune complexes by the resident FDC [67]. The histological picture of the germinal centre shows evidence of highrate apoptosis occurring in the basal light zone where cells displaying the classical morphological features of apoptosis are encountered relatively frequently, both free and within tingible body macrophages. The FDC network is particularly dense in the light zone and it has been proposed that centrocytes, after
APOPTOSIS IN THE IMMUNE SYSTEM
207
expressing their newly mutated antigen receptors, compete in the basal light zone for antigen held on FDC processes in order to receive a survival signal; in the absence of such a signal, apoptosis ensues [67]. The provision, by antigen, of a positive selection signal that blocks apoptosis may be sufficient to permit the germinal centre B cell to enter a further stage in its differentiation, which is almost undoubtedly influenced by the microenvironment of the apical light zone [67]. However, whilst positive selection by antigen would be expected to militate against the survival of centrocytes lacking antigen receptors or expressing low affinity receptors, it might be predicted that such a mechanism, at the same time as selecting cells displaying receptors of high affinity for the foreign antigen in question, might also promote the survival of cells that, unfortuitously, have become autoreactive as a result of the hypermutation process. Recent studies have suggested that the CD40-CD40L interaction could provide an important discriminatory and long-term survival signal for germinal centre B cells [70]. CD40, as mentioned previously, is constitutively expressed on the surface of B cells from the pro-B cell stage onwards. The molecule belongs to a family of cell surface receptors, including CD27, CD30, Fas, the low affinity NGF receptor and TNF receptors type I and type II, that share a characteristic series of three or four cysteine-rich extracellular domains [71]. The CD40-ligand is a type II integral membrane protein homologous to TNF that is expressed transiently on the surface of activated T cells [72]. Of importance to the present discussion is the finding that CD40L provides isolated germinal centre B cells with a more potent survival signal than that provided by immobilised anti-Ig antibodies [68–70]. It has been proposed that CD40L might be available to germinal centre B cells that process and present antigen to specific T helper cells in the light zone [70] (Figure 11.4). Indeed, it has been demonstrated that appreciable numbers of T helper cells are present in the light zones of germinal centres and antigen-specific T helper cells have been found to enter the germinal centres of immunised mice. The absence of self-reactive T cells would be expected to restrict the CD40L-induced survival signal to germinal centre B cells reactive with the immunising antigen. Therefore, long-term survival of autoreactive cells would be avoided. The mechanism underlying selection of germinal centre B cells may therefore involve repression of apoptosis at two stages: (1) through a short-term survival signal generated by the antigen receptor and (2) via a subsequent long-term survival signal resulting from the interaction of CD40 with its counterstructure. It is interesting to note that germinal centre B cells downregulate expression of Bcl-2, an event which is likely to be important in the priming of these cells for apoptosis. Extended CD40L-induced survival of germinal centre B cells in vitro is associated with re-expression of Bcl-2 [49]. In situ, Bcl-2 protein is found at high level in the cells of the follicular mantle and in only occasional cells of the germinal centre — perhaps these represent cells which have received all the signals for long-term survival, allowing them to exit the germinal centre and enter the long-lived memory B cell pool [48, 67]. 11.3.3 Fas and its Counterstructure Fas, also known as APO-1 and, more recently, CD95, was identified independently by two groups, one in Germany and one in Japan, who described monoclonal antibodies that rapidly induced apoptosis after crosslinking a cell surface protein of approximately 48kDa [73, 74]. It has subsequently been shown that this molecule together with its physiological ligand, are likely to play important roles in regulating immune responses. The Fas/Fas-ligand interaction also constitutes a pathway by which cytotoxic T lymphocytes cause death of the target cells with which they interact (see section 11.4.2). Cloning of Fas using a cDNA library from a human T cell lymphoma identified the molecule as a transmembrane protein which showed homology to a family of structures including TNF-receptors types I and II, low affinity NGF-receptor,
208
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
CD27, CD30, CD40 and OX40, grouped according to the presence of a series of cysteine-rich repeats in their extracellular domains [75]. Cloning of APO-1 demonstrated that it was, as suspected, identical to Fas [76]. It has been found that the intracytoplasmic domain of Fas contains a sequence of 68 amino acids, shared with the type I TNFR, that is critical for delivery of its apoptotic signal (although Fas and TNFR, at least type I TNFR, appear to signal via distinct biochemical pathways [77]). The intracytoplasmic region of the molecule, notably the C-terminal 15 amino acids, also appears to be involved in negative regulation of the apoptotic signal, since removal of this sequence promotes apoptosis-signalling by Fas [78]. Construction of a fusion protein between murine Fas and human Ig-Fc permitted the identification of a physiological ligand for Fas [79]. The sequence of this structure showed that it was a type II transmembrane protein and, not surprisingly, proved to be homologous to TNF. Fas-ligand has been purified as a 40-kD glycoprotein from the surface of the cytotoxic T cell hybridoma from which the original cDNA was isolated [80]. Recent work has shown that Fas transcripts in human cells encode variant Fas lacking the transmembrane domain [81]. Such a soluble Fas molecule may be involved in regulating the function of membrane-bound Fas in apoptosis-signalling by competing for Fas-ligand. Fas is expressed on a variety of cell types. Lymphocytes of both T and B lineages express the protein which is observed at higher levels in activated, as compared to resting, cells [82]. Although its function appears primarily to be that of an apoptosis-signalling molecule, its surface expression per se, either on T cells or B cells, does not guarantee that it will generate an apoptotic response after ligation. Indeed, like other members of the TNFR/NGFR family to which it belongs, Fas may be a multifunctional molecule; it has been shown, for example, that it is capable of transducing proliferative signals in human T cells [83]. It has been suggested that lymphocyte apoptosis could be partially regulated throught the inverse expression of Fas and Bcl-2 [82]. Thus, germinal centre B cells are high in Fas and low in Bcl-2; the reciprocal is true of the cells of the follicular mantle. There is strong evidence from a number of sources that Fas is also important in the physiological regulation of T cells, especially in the periphery. Of particular significance is the finding that mice which carry a homozygous recessive mutation known as lpr (for lymphoproliferation) develop lymphadenopathy and systemic autoimmunity as a consequence of T cell dysfunction brought about by impaired or mutant Fas expression (reviewed in [84]). The lpr mutation was recently shown to be caused by a retroviral insert into the second intron of the Fas gene that affected transcriptional readthrough and splicing resulting in a marked reduction of Fas protein translation. An allelic form of lpr, lprcg, that results in an identical syndrome of lymphadenopathy and autoimmunity, is caused by a point mutation affecting the intracytoplasmic region of Fas and consequently such mice produce apoptosis-signalling-defective Fas molecules. Both the T cells and B cells of lpr mice appear to be defective in their ability to undergo apoptosis. The lymphadenopathy which these animals develop does not occur in thymectomised or anti-CD4 antibody-treated mice and is characterised by the accumulation of large numbers of abnormal double-negative T cells in the peripheral lymphoid tissues. These cells display a distinct pattern of cell surface markers from that of the double-negative cells of the thymus [84]. An additional syndrome, gld (generalised lymphoproliferative disease) that is non-allelic to lpr, produces near-identical symptoms to lpr. Recent results indicating that this condition is caused by a point mutation in a C-terminal sequence of Fas-ligand that is highly conserved among members of the TNF family [85], are consistent with the original suggestion that lpr and gld syndromes are manifestations of defects in separate arms of a single receptor-ligand system [86]. Thus, mutant Fas-ligand from gld mice expressed in COS cells failed, unlike its normal counterpart, to induce apoptosis in cells expressing Fas. The consequences of Fas/Fas-ligand defects for development of autoimmune disease will be discussed in section 11.6 below.
APOPTOSIS IN THE IMMUNE SYSTEM
209
Figure 11.5 Two molecular pathways of cytotoxic T lymphocyte (CTL)-mediated cytotoxicity (modified from [92]).
11.4 Cell-mediated Cytotoxicity The destruction of target cells by cytotoxic T lymphocytes (CTL) or natural killer (NK) cells has been the subject of numerous reviews in recent years (for details, see [87–92]) and therefore will only be considered briefly here. It has been known for about two decades that killing of target cells by CTL can be divided into discrete phases: (1) target cell recognition and binding by the CTL, (2) delivery of the lethal hit (‘programming for lysis’), and (3) target cell disintegration, this final stage not requiring continual presence of the cytotoxic cell. The molecular mechanisms that underlie the second stage have been the subject of intense investigation and debate, the field being divided into two schools of thought: one supporting the contention that target cell death is induced by the exocytosis of molecules produced by cytotoxic lymphocytes, the other following the premise that target cell destruction is mediated by signals resulting from the docking of T cell surface-bound receptors with ligands on the target. Both factions have been proved correct with the formal demonstration of two, mutually exclusive, destructive pathways (Figure 11.5). 11.4.1 Mode of Target Cell Destruction: the Role of Perforin Cytolytic granules isolated from CTL and NK cells have been found to contain proteins (‘perforins’) structurally homologous to components of the membrane attack complex of the complement system (reviewed in [87, 89, 91]). It has been suggested that the molecular basis of the lethal hit is the directed exocytosis of perforins towards the zone of contact between lymphocyte and target cell, and the subsequent polymerisation of perforin proteins to form pore-like structures in the target cell membrane [87, 93]. This turns out to be an oversimplified view for one important reason: death of nucleated cells by purified perforin displays all the characteristics of necrosis, whereas much evidence documents CTL-mediated target cell death as apoptosis (see [88]). Recent studies of perforin knock-out mice indicate, however, that the vast majority of anti-viral CTL activity (as demonstrated in vitro and in vivo) is absent from such animals [94]. Taken together, these results suggest that, whilst perforins are undoubtedly important for cytotoxicity by CTL, the perforin-based mechanism is not limited to perforin secretion and polymerisation with subsequent passive lysis of the target cell. The identification of a target cell mutant that resists CTL-triggered death yet is still recognised by CTL lends support to the idea that the target cell plays an active, genetically programmed role in its own destruction [95].
210
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Serine esterases are likely to provide key components of the perforin-based mechanism, their contribution being to ensure that target cells undergo apoptosis. Physiologically, target cell death by apoptosis would be preferred over necrosis, since an apoptotic mechanism would lead both to rapid clearance of dying target cells and, in concert with endonucleolytic cleavage of genomic DNA, provide a means to destroy viral genomes present in infected cells [96]. Serine esterases (seven in mouse and three in man), termed ‘granzymes’, have been found in the granules of CTL and several lines of evidence support the contention that they operate, in the presence of perforin components, to direct target cell death towards apoptosis. Notably, granzyme A (CTLA-3), in combination with purified perforin, directed target cell death towards apoptosis in two experimental model systems [97, 98]. Furthermore, serine esterases from rat cytotoxic lymphocytes were found to induce apoptosis in target cells in the presence of perforin [99]. Granzyme Bknockout mice have been used to show that this serine protease is critical for the induction of target cell apoptosis by alloreactive CTL [100]. An additional constituent of the granules of cytotoxic lymphocytes which may be important in inducing apoptosis upon access to the interior of target cells is TIA-1 [101]. This is likely to be an RNA-binding protein which, along with a related molecule TIA-1R, can induce DNA fragmentation in permeabilised cells, is present in numerous cell types and, as such, may form part of a central apoptotic mechanism [102]. 11.4.2 Cell-mediated Cytotoxicity by Fas/Fas-ligand Interaction It has been argued for some years that more than one mechanism is operational in lymphocyte-mediated cytotoxicity. The major proportion of the cytotoxicity displayed by most CTL preparations is dependent upon extracellular Ca2+, and physiological perforin-based cytotoxicity may be expected to be largely Ca2+dependent, both in terms of insertion of perforin in the target cell membrane (although this can also occur when Ca2+is chelated) and in terms of the requirement for Ca2+ in granule secretion. However, a number of investigators have observed that CTL can display cytotoxic activity in the absence of Ca2+(reviewed in [92]). Recent studies have demonstrated that a Ca2+-independent pathway of CTL-mediated cytotoxicity that appears to function quite independently of perforin (though both mechanisms can be simultaneously displayed by a single clone of effector cells) operates through Fas (on the target cell) interacting with its ligand (on the CTL) [92, 103]. Indeed, it has been found that expression of Fas-ligand in non-lymphoid (COS) cells endows them with cytotoxic capability [79], indicating that no association of Fas-ligand with a lymphoid, or even leucocyte, lineage-restricted molecule is required in this cytotoxic pathway. Comparison of cytotoxic T cell activity from perforin knock-out and normal animals indicates that all T cell-mediated cytotoxicity can be accounted for by the perforin- and Fas-based mechanisms [104]. 11.5 The Innate Immune System In the immune system, the relevance of apoptosis is not limited to lymphocytes, the specific agents of adaptive or acquired immunity. The apoptotic process also plays a central role in the innate immune system. Thus, not only are the cellular populations of the innate immune system subjected to control by apoptosis, but also they serve to illustrate a critically important feature of the apoptotic mechanism: the rapid clearance of the dying cells by phagocytosis, a process which itself has implications for physiological control of the immune system. In this section, survival and death of the single most important cell type of the innate
APOPTOSIS IN THE IMMUNE SYSTEM
211
immune system, the neutrophil, will be considered and the molecular mechanisms underlying phagocytic clearance of apoptotic cells will be discussed. 11.5.1 Apoptosis in Neutrophils Neutrophil granulocytes form an essential first line of defence against infection and are important mediators of inflammation. Throughout life, they are produced in the bone marrow at a phenomenal rate: in adult humans the peripheral blood neutrophil population is maintained by the steady-state supply of 1–3×1010 newly-formed cells per day [105] balanced by the efficient removal of senescent cells. In vitro, isolated neutrophils spontaneously undergo apoptosis at 37°C: human peripheral blood neutrophil populations routinely display the characteristic morphological features of apoptosis after only a few hours [106]. HL60 myeloblastic leukemia cells undergo apoptosis soon after differentiation toward neutrophils [107]. It is probable that, under normal circumstances in vivo, too, aged neutrophils undergo apoptosis and, as a result, are rapidly cleared by phagocytes (see below), although apoptosis may not be the only route by which aged neutrophils are recognised and phagocytosed ([108] and see below). A body of evidence also strongly supports the proposal that phagocytosis of apoptotic neutrophils represents a crucial facet of the inflammatory process necessary for resolution [109]. Although nothing is yet known of the nature of the short-term ‘clock’ that operates in determining neutrophil survival time, one attractive possibility is that it involves the gradual loss of a factor which functions to inhibit the cell’s inherent tendency to activate its apoptotic programme [1] (indeed, as already discussed, apoptosis by default may be a feature of almost any cell type [10]). Resetting the neutrophil death clock, at least to a limited degree, can be achieved in a number of ways: bcl-2 expression and exposure to NGF or LPS delay neutrophil apoptosis, while TNF has been reported to speed up the death programme [108, 110–112]. It is therefore possible that environmental factors, notably soluble mediators of inflammation, could influence the rate of neutrophil apoptosis at different sites; a cell at an inflammatory site might therefore display a different apoptotic rate as compared to a circulating cell. Neutrophil apoptosis has been found to be associated with reduction or loss in ability to change shape, to respond to chemotactic stimuli, to degranulate, undergo respiratory burst, and to phagocytose opsonized zymosan [113]. In addition, apoptotic neutrophils appear capable of specifically downregulating certain surface glycoproteins as has been described for CD 16 (Fc RIII) [114]. Therefore it seems that, once initiated, the apoptotic mechanism of the neutrophil causes the effective ‘paralysis’ of the cell which becomes functionally isolated from its environment. 11.5.2 Phagocytosis of Apoptotic Cells A crucial component of the apoptotic process, perhaps the most important, is the marking of dying cells for disposal by phagocytes. So efficient is the recognition mechanism that cells undergoing apoptosis are consumed by phagocytes prior to plasma membrane breakdown. The rapidity with which apoptotic cells (or their membrane-bound fragments, apoptotic bodies) are removed is such that free cells exhibiting the morphological features of apoptosis are not often encountered in vivo. Thus, apoptosis is characterised by an efficient disposal mechanism that militates against the potentially catastrophic consequences of leakage of the intracellular contents from dying cells, a feature of necrosis. The histotoxic and inflammatory effects of the contents of neutrophil granules are well documented (see [115]) and the potentially injurious specialist
212
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 11.6 Molecules associated with the clearance of apoptotic cells by macrophages (modified from [117] and also using data from [124, 125]). PS=phosphatidylserine; TSP=thrombospondin; VNR= v 3 integrin vitronectin receptor; CHO=altered carbohydrate. The 61D3 antigen remains poorly characterised [124]. Note that the counterstructures for PS, TSP, CHO, 61D3 antigen and ICAM-3 operating in these cell/cell interactions have not yet been identified.
molecules of certain other cell types (the granzymes and perforins of cytotoxic lymphocytes, for example) provide clear cases in support of a mechanism of clearance which pre-empts the consequences of plasma membrane disruption in dying cells. However, it seems likely that almost any cell type can inflict damage upon neighbouring tissue if its lysosomal contents are released from their safe intracellular packaging. An additional, indirect, danger of plasma membrane leakage prior to clearance may result from the exposure of normally intracellular components to the extracellular milieu. Thus, stimulation of B cell clones bearing receptors for such antigens could induce the production of disruptive or destructive autoantibodies as occurs in the autoimmune condition systemic lupus erythematosus (SLE). In the absence of an efficient clearance mechanism, therefore, apoptosis would be ineffectual as a physiological mode of cell death. As already mentioned, free apoptotic cells are only rarely encountered in vivo: very often the only evidence of an apoptotic cell in a histological section is provided by its ‘pyknotic’ remnants within a tissue macrophage (which has been known to be renamed as a consequence: the ‘tingible body’ macrophage of the lymphoid follicle is so called because its cytoplasm is laden with the remains of apoptotic centrocytes). In addition, neighbouring viable tissue cells appear to play a highly effective role in phagocytosing their apoptotic counterparts and it is possible that virtually any cell type has the potential to function in this way [115]. What molecules mark apoptotic cells for phagocytosis and with what counterstructures do they interact on phagocytes? A clue to the complexity of this question is provided by studies of the nematode, Caenorhabditis elegans (see also Yuan and Drexler, this volume), an animal which lacks professional phagocytes, but in which seven genes are associated with the phagocytic clearance of cells that undergo programmed death in the normal course of worm development [116]. In mammalian systems little is yet known of the mechanisms underlying such ‘non-professional’ phagocytosis. However, the complexity of the molecular interactions that occur between macrophages and apoptotic cells is beginning to be appreciated, largely as a result of the work of Savill and colleagues, who have recently provided two excellent reviews on this subject [115, 117]. Based on work performed in murine and human systems, Figure 11.6 summarises the molecular structures implicated in the recognition and phagocytosis of apoptotic cells by macrophages. To date, all published investigations have relied upon either lymphocytes or granulocytes (in most cases neutrophils) to provide a source of apoptotic cells. The earliest investigations into the problem were perfomed by Wyllie and colleagues [118, 119] who studied the binding, at low temperature, of apoptotic murine thymocytes to peritoneal macrophages. Their experiments, in which it was shown that cell/cell binding could be inhibited
APOPTOSIS IN THE IMMUNE SYSTEM
213
by simple sugars that interacted with the macrophage surface, strongly suggested that the intercellular adhesion pathway involved altered carbohydrate on the apoptotic cells and a putative lectin-like moiety on the macrophages. Subsequently, Savill and co-workers reported that the removal of senescent (apoptotic) human neutrophils by monocyte-derived macrophages was mediated by the v 3 (vitronectin receptor) integrin on the macrophage surface [120]. This molecule appears to interact with the apoptotic cell surface indirectly: thus, v 3 has been shown to bind to thrombospondin (TSP) which functions as a molecular bridge to interact with an as yet unknown counter-structure on apoptotic cells. Macrophage CD36, a further TSP-ligand, was also shown to play a role in apoptotic cell recognition [121]. Evidence in support of an additional pathway, operating in humans and mice (and indeed across these species), that mediates recognition and phagocytosis of apoptotic lymphocytes and neutrophils by macrophages has been offered recently [122, 123]. The experiments demonstrated that phosphatidylserine (PS), which flips out during apoptosis to be expressed on the external surface of the plasma membrane of apoptotic cells, acts as a ligand for a putative PS-receptor on macrophages. Furthermore it was found that, not only were the v 3 and PS recognition pathways separate, but also that different populations of macrophages preferentially utilise one or other. Finally, two further molecules have been implicated in macrophage/apoptotic cell interactions: one, the 61D3-antigen, a poorly-characterised surface structure of human macrophages, the other, ICAM-3, on the surface of human B cells [124, 125]. It is clear from Figure 11.6 that, although several molecules contributing to the mechanism of apoptotic cell clearance have now been identified, definitive pathways, which link apoptotic cell to macrophage, have yet to be formulated. This is an important area for future investigation. Other important questions also remain. For example, are apoptotic cells chemotactic for monocytes or macrophages? How are phagocytosed and degraded apoptotic cells perceived by the immune system? The potentially detrimental immunological consequences of release of internal cellular components as would be expected to occur in necrosis have already been considered above. Would not the constituent proteins of apoptotic cells, after degradation to peptides by macrophages (which are after all efficient in the art of antigen-presentation) be available for recognition by T cells? Perhaps such recognition of apoptotic-self has already been selected against in the thymus. Alternatively, it is conceivable that apoptotic components enter an endogenous processing pathway. These possibilities have yet to be investigated. The molecular basis of innate immunity to apoptotic-self, the efficient and swift phagocytosis of cells dying by apoptosis, is thus a relatively neglected area of cell death biology that may be anticipated to develop significantly in the near future. Important new clues to the structure of novel components involved in mammalian systems are likely to be gained from identification of the key recognition molecules in C. elegans. Studies of how surface changes of apoptotic cells that are critical for recognition and clearance are related to other biochemical events characteristic of the apoptotic process are also clearly warranted. In this regard it is interesting to note that, in a recent study of the clearance of senescent murine neutrophils by macrophages, it was shown that, whilst the bcl-2 product effectively inhibited neutrophil apoptosis as assessed by conventional criteria, it failed to block the recognition mechanism that triggered macrophage clearance [108].
11.6 Conclusions: Apoptosis and the Immune System in Health and Disease In concluding this chapter, it is perhaps most informative to leave normality aside and consider dysregulation of the physiological death process and its consequences for the immune system. In broad terms,
214
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
in view of the clear role played by apoptosis in the selection of lymphocytes, the simplest predictions would be that inappropriate activation of the apoptotic mechanism in these cells would have implications for immunodeficiency whilst inappropriate inhibition of the process would contribute to autoimmunity and neoplasia. But what evidence is there to support such a causeeffect relationship? Immunodeficiency may be caused by lymphocyte developmental blocks which lead to apoptosis by default, cell death in these cases being the normal, programmed response occurring in the absence of an essential survival signal (failure to rearrange antigen receptor genes, for example). However, active induction of the death programme can also elicit immunodeficiency, the strongest case in point being that of acquired immunodeficiency syndrome, AIDS. Much evidence supports the idea that HIV activates T cell apoptosis, resulting in severe depletion of the CD4+ population as occurs in AIDS patients. It has been shown that apoptosis in T cells can be induced directly as a result of the envelope glycoprotein, gp120, of HIV interacting with CD4 prior to the interaction of the TCR with its specified antigen. Additionally it has been proposed that the virus may induce the progressive depletion of CD4+ T cells by encoding a superantigen (reviewed in [126, 127]). Several lines of evidence support the view that defective regulation of apoptosis plays a causative role in the development of autoimmunity. Apart from the potential of autoreactive CD5+B cells to escape deletion, as already discussed, defective Fas/ Fas-ligand interactions as occur in the lpr and gld mutations lead to the development of autoimmune symptoms that are dependent, at least in part, upon aberrant T cell survival. lpr and gld animals develop autoantibodies to an array of antigens, including nuclear components, which, in extreme cases cause death from vasculitis and glomerulonephritis [84]. Failure to eliminate autoreactive T cells has been implicated in the spontaneous development of autoimmune conditions (such as autoimmune hemolytic anemia, and SLE-like disease) in other autoimmunity-prone mouse strains [128]. In addition, in certain of these strains, it has been found that triggering of mature B cell apoptosis by anti-Ig antibodies is also defective [128]. Interestingly, constitutive expression of a bcl-2 transgene in B-lineage cells was found to cause an SLE-like syndrome with autoantibody production and glomerulonephritis, suggesting that Bcl-2, through its ability to block apoptosis, inhibited normal deletion of self-reactive B cells [52]. Finally, it is tempting to speculate that defective clearance mechanisms may contribute to autoimmune conditions by exposing lymphocytes to intracellular components in an immunogenic form. Of possible relevance here is the recent finding that SLE autoantigens are clustered in blebs at the surface of apoptotic keratinocytes [129]. It seems rational to argue that a cell’s capacity to escape from apoptosis would predispose to malignancy. The strongest support for this idea, at least as far as the immune system is concerned, is that the follicular lymphoma t(14;18) translocation-associated gene, bcl-2, whose only known function is to regulate apoptosis, is oncogenic in transgenic mice, with the majority of tumours also carrying a mutation of a further potent oncogene, c-myc [130–132]. On the other hand, the product of c-myc can itself function to drive apoptosis in several cell types [20, 133, 134] including the monoclonal human B cell tumour, Burkitt’s lymphoma (BL) [134], all cases of which carry a c-myc/Ig translocation. Indeed, BL, which, like follicular lymphoma, is a tumour of germinal centre B cells but one in which Bcl-2 is downregulated [136], serves to illustrate that a malignant cell can retain the apoptotic characteristics of its normal counterpart. Thus BL cells, retain a phenotypic vestige of the normal germinal centre B cells from which they are derived, and remain ‘primed’ to undergo apoptosis [137]. Therefore, although a constitutive block on apoptosis may provide an important contributory factor to the evolution of a malignant clone, it is clearly not a prerequisite for lymphocyte neoplasia that the apoptotic mechanism is shut down. On the contrary, induction of apoptosis is an important mode of anti-lymphoma/leukemia therapy, not only with regard to anticancer drugs,
APOPTOSIS IN THE IMMUNE SYSTEM
215
but also to specific agents such as antigen receptor peptide ligands and anti-idiotypic antibodies, which can induce apoptosis of follicular lymphoma cells in vitro and tumor regression in vivo [138, 139]. 11.7 Summary With the realisation that cell population sizes can be physiologically regulated, not only by proliferation and differentiation, but also by death, it is logical that a controllable cell death programme would play a critical role in the effective construction and normal functioning of the immune system. In the case of lymphocytes, establishment of a diverse antigen receptor repertoire from a limited number of receptor gene segments requires that antigen receptor gene rearrangements are initiated in vast numbers of T and B cells; in the B lineage, further genetic changes during affinity maturation of antibody responses are similarly initiated in a greatly expanded population of antigen-responsive cells. In each case, the inherent death programme that results in apoptosis either engages by default in useless lymphocytes (for example those unable to produce a functional receptor), or is actively engaged as a result of an environmental signal (for example a self antigen). Thus, the apoptotic process of the lymphocyte is regulated to select for immunologically useful cells. Antigen-reactive lymphocytes which are ‘inappropriately’ activated (without co-stimulation for example) may also engage their apoptotic programme, and the termination of on-going responses, as well as the establishment of immunological memory are also likely to require modulation of the lymphocyte death programme. High-turnover lymphocyte populations in which genetic changes are in progress are inherently primed for activation of their apoptotic programme which provides a failsafe mechanism for the swift removal of aberrant or dangerous cells. References 1 2
3 4 5 6 7 8 9 10 11
. COHEN, J.J. (1991) Programmed cell death in the immune system. Adv. Immunol. 50, 55–85. . MEUER, S.C., FITZGERALD, K.A., HUSSEY, R.E., HODGDON, J.C., SCHLOSSMAN, S.F. & REINHERZ, E.L. (1983) Clonotypic structures involved in antigen-specific human T-cell function: relationship to the T3 molecular complex. J. Exp. Med. 157, 705–719. . HAAS, W., PEREIRA, P. & TONEGAWA, S. (1993) Gamma/delta cells. Ann. Rev. Immunol. 11, 637–685. . VON BOEHMER, H., TEH, H.S. & KISIELOW, P. (1989) The thymus selects the useful, neglects the useless and destroys the harmful. Immunol. Today 10, 57–61. . KISIELOW, P. (1995) Apoptosis in intrathymic T-cell development. In: Apoptosis And The Immune Response, GREGORY, C.D. (ed). Wiley-Liss Inc, New York, pp. 13–53. . WALLACE, V.A., PENNINGER, J. & MAK, T. (1994) CD4, CD8 and tyrosine kinases in thymic selection. Curr. Opin. Immunol. 5, 235–240. . KRUISBEEK, A.M. (1993) Development of T cells. Curr. Opin. Immunol. 5, 227–234. . GROETTRUP, M. & VON BOEHMER, H. (1993) A role for a pre-T-cell receptor in T-cell development. Immunol Today. 14, 610–614. . DAVIS, C.B. & LITTMAN, D.R. (1994) Thymocyte lineage commitment: is it instructed or stochastic? Curr. Opin. Immunol. 6, 266–272. . RAFF, M.C., BARRES, B.A., BURNE, J.F., COLES, H.S., ISHIZAKI, Y. & JACOBSON, M.D. (1993) Programmed cell death and the control of cell survival: lessons from the nervous system. Science 262, 695–700. . SMITH, C.A., WILLIAMS, G.T., KINGSTON, R., JENKINSON, E.J. & OWEN, J.J.T. (1989). Antibodies to CD3/T receptor complex induce cell death by apoptosis in immature thymic cultures. Nature 337, 181–184.
216
12
13 14 15 16 17 18 19
20 21 22 23
24
25 26 27 28
29 30 31 32 33
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. SHI, Y., BISSONNETTE, R.P., PARFREY, N., SZALAY, M., KUBO, R.T. & GREEN, D.R. (1991) In vivo administration of antibodies to the CD3-T cell receptor complex induces cell death (apoptosis) in immature thymocytes. J. Immunol. 146, 3340– 3346. . MURPHY, K., HEIMBERGER, A. & LOH, D.Y. (1990) Induction by antigen of intrathymic apoptosis of CD4 +8+TCRlo thymocytes in vivo. Science 250, 1720–1722. . SWAT, W., IGNATOWICZ, L., VON BOEHMER, H. & KISIELOW, P. (1991) Clonal deletion of CD4+CD8+ thymocytes in suspension culture by extrathymic antigen-presenting cells. Nature 351, 150–152. . ROBEY, E.A., RAMSDELL, F., KIOUSSIS, D., SHA, W.C., LOH, D.Y., AXEL, R. & FOWLKES, B.J. (1992) The level of CD8 expression can determine the outcome of thymic selection. Cell 69, 1089–1096. . ROTHENBERG, E.V. (1994) Signalling mechanisms in thymocyte selection. Curr. Opin. Immunol. 6, 257–265. . MALISSEN, B. & SCHMITT-VERHULST, A.M. (1993) Transmembrane signalling through the T-cell-receptorCD3 complex. Curr. Opin. Immunol. 5, 324–333. . WORONICZ, J.D., CALNAN, B., NGO, V. & WINOTO, A. (1994) Requirement for the orphan steroid receptor Nur77 in apoptosis of T cell hybridomas. Nature 367, 277– 281. . LIU, Z.G., SMITH, S.W., MCLAUGHLIN, K.A., SCHWARTZ, L.M. & OSBORNE, B.A. (1994) Apoptotic signals delivered through the T cell receptor of a T cell hybrid require the immediate early gene nur77. Nature 367, 281–284. . SHI, Y., GLYNN, J.M., GUILBERT, L.J., COTTER, T.G., BISSONNETTE, R.P. & GREEN, D.R. (1992) Role for c-myc in activation-induced apoptotic cell death in T cell hybridomas. Science 257, 212–214. . LOWE, S.W., SCHMITT, E.M., SMITH, S.W., OSBORNE, B.A. & JACKS, T. (1993) p53 is required for radiation induced apoptosis in mouse thymocytes. Nature 362, 847–849. . CLARKE, A.R., PURDIE, C.A., HARRISON, D.J., MORRIS, R.G., BIRD, C.C., HOOPER, M.L. & WYLLIE, A.H. (1993) Thymocyte apoptosis induced by p53 dependent and independent pathways. Nature 362, 849–852. . TAO, W., TEH, S.-J., MELHADO, I., JIRIK, F., KORSMEYER, S.J. & TEH, H.-S. (1994) The T cell receptor repertoire of CD4–8+ thymocytes is altered by overexpression of the BCL-2 protooncogene in the thymus. J. Exp. Med. 179, 145–153. . NAKAYAMA, K.-L, NAKAYAMA, K., NEGISHI, I., KUIDA, K., SHINKAI, Y., LOUIE, M.C., FIELDS, L.E., LUCAS, P.J., STEWART, V., ALT, F.W. & LOH, D.Y. (1993) Disappearance of the lymphoid system in Bcl-2 homozygous mutant chimeric mice. Science 261, 1584–1588. . VEIS, D.J., SORENSON, C.M., SHUTTER, J.R. & KORSMEYER, S.J. (1993) Bcl-2-deficient mice demonstrate fulminant lymphoid apoptosis, polycystic kidneys, and hypopigmented hair. Cell 75, 229–240. . STRASSER, A., HARRIS, A.W., CORCORAN, L.M. & CORY, S. (1994) Bcl-2 expression promotes B- but not T-lymphoid development in scid mice. Nature 368, 457–460. . TONEGAWA, S. (1983) Somatic generation of antibody diversity. Nature 302, 575–581. . MELCHERS, F., KARASUYAMA, H., HAASNER, D., BAUER, S., KUDO, A., SAKAGUCHI, N., JAMESON, B. & ROLINK, A. (1993) The surrogate light chain in B-cell development. Immunol. Today. 14, 60–68. . ROLINK, A. & MELCHERS, F. (1993) Generation and regeneration of cells of the B-lymphocyte lineage. Curr. Opin. Immunol. 5, 207–217. . OSMOND, D.G. (1993) Early B cell development. In: Mechanisms of B Cell Neoplasia 1993, MELCHERS, F. & POTTER, M. (eds). Editiones Roche, Basel, pp. 31–38. . NEMAZEE, D.A. & BÜRKI, K. (1989) Clonal deletion of B lymphocytes in a transgenic mouse bearing anti-MHC class I antibody genes. Nature 337, 562–566. . NEMAZEE, D.A. & BÜRKI, K. (1989) Clonal deletion of autoreactive B lymphocytes in bone marrow chimeras. Proc. Natl Acad. Sci. USA 86, 8039–8043. . HARTLEY, S.B., CROSBIE, J., BRINK, R., KANTOR, A.A., BASTEN, A. & GOODNOW, C.C. (1991) Elimination from peripheral lymphoid tissues of self-reactive B lymphocytes recognizing membrane-bound antigens. Nature 353, 765–769.
APOPTOSIS IN THE IMMUNE SYSTEM
34 35 36 37
38
39 40 41 42 43 44 45 46 47 48
49
50 51
52
53 54
217
. BROMBACHER, F., KOHLER, G. & EIBEL, H. (1991) B cell tolerance in mice transgenic for anti-CD8 immunoglobulin mu chain. J. Exp. Med. 174, 1335–1346. . OKAMOTO, M., MURAKAMI, M., SHIMIZU, A., OZAKI, S., TSUBATA, T., KUMAGAI, S. & HONJO, T. (1992) A transgenic model of autoimmune hemolytic anemia. J. Exp. Med. 175, 71–79. . ERIKSON, J., RADIC, M.Z., CAMPER, S.A., HARDY, R.R., CARMACK, C. & WEIGERT, M. (1991) Expression of anti-DNA immunoglobulin transgenes in nonautoimmune mice. Nature 349, 331–334. . GOODNOW, C.C., CROSBIE, J., ADELSTEIN, S., LAVOIE, T.B., SMITH-GILL, S.J., BRINK, R.A., PRITCHARD-BRISCOE, H., WOTHERSPOON, J.S., LOBLAY, R.H., RAPHAEL, K., TRENT, R.J. & BASTEN, A. (1988) Altered immunoglobulin expression and functional silencing of self-reactive B lymphocytes in transgenic mice. Nature 334, 676–682. . HARTLEY, S.B., COOKE, M.P., FULCHER, D.A., HARRIS, A.W., CORY, S., BASTEN, A. & GOODNOW, C.C. (1993) Elimination of self-reactive B lymphocytes proceeds in two stages: arrested development and cell death. Cell 72, 325–335. . GAY, D., SAUNDERS, T., CAMPER, S. & WEIGERT, M. (1993) Receptor editing: an approach by autoreactive B cells to escape self-tolerance. J. Exp. Med. 177, 999–1008. . TIEGS, S.L., RUSSELL, D.M. & NEMAZEE, D. (1993) Receptor editing in self-reactive bone marrow B-cells. J. Exp. Med. 177, 1009–1020. . RADIC, M.Z., ERIKSON, J., LITWIN, S. & WEIGERT, M. (1993) B lymphocytes may escape tolerance by revising their antigen receptors. J. Exp. Med. 177, 1165–1173. . BASTEN, A., BRINK, R., PEAKE, P., ADAMS, E., CROSBIE, J., HARTLEY, S. & GOODNOW, C.C. (1991) Self tolerance in the B-cell repertoire. Immunol. Rev. 122, 5–19. . GOODNOW, C.C. (1992) Transgenic mice and analysis of B-cell tolerance. Ann. Rev. Immunol. 10, 489–518. . CYSTER, J.G., HARTLEY, S.B., GOODNOW, C.C. (1994) Competition for follicular niches excludes selfreactive cells from the recirculating B-cell repertoire. Nature 371, 389–395. . JACOBSEN, K. & OSMOND, D.G. (1990) Microenvironmental organization and stromal cell associations of B lymphocyte precursor cells in mouse bone marrow. Eur. J. Immunol . 20, 2395–2404. . JACOBSEN, K., MIYAKE, K., KINCADE, P.W. & OSMOND, D.G. (1992) Highly restricted expression of a stromal cell determinant in mouse bone marrow in vivo. J. Exp. Med. 176, 927–935. . LAW, C.-L. & CLARK, E.A. (1994) Cell-cell interactions that regulate the development of B-lineage cells. Curr. Opin. Immunol. 6, 238–247. . LIU, Y.-J., MASON, D.Y., JOHNSON, G.D., ABBOT, S., GREGORY, C.D., HARDIE, D.L., GORDON, J. & MACLENNAN, I.C.M. (1991) Germinal center cells express bcl-2 protein after activation by signals which prevent their entry into apoptosis. Eur. J. Immunol. 21, 1905–1910. . HOLDER, M.J., WANG, H., MILNER, A.E., CASAMAYOR, M., ARMITAGE, R., SPRIGGS, M.K., FANSLOW, W.C., MACLENNAN, I.C.M., GREGORY, C.D. & GORDON, J. (1993) Suppression of apoptosis in normal and neoplastic human B lymphocytes by CD40 ligand is independent of Bcl-2 induction. Eur. J. Immunol. 23, 2368– 2371. . MERINO, R., DING, L., VEIS, D.J., KORSMEYER, S.J. & NUNEZ, G. (1994) Developmental regulation of the Bcl-2 protein and susceptibility to cell death in B lymphocytes. EMBO J. 13, 683–691. . CUENDE, E., ALES-MARTINEZ, J.E., DING, L., GONZALEZ-GARCIA, M., MARTINEZ-A, C. & NUNEZ, G. (1993) Programmed cell death by bcl-2-dependent and independent mechanisms in B lymphoma cells. EMBO J. 12, 1555–1560. . STRASSER, A., WHITTINGHAM, S., VAUX, D.L., BATH, M.L., ADAMS, J.M., CORY, S. & HARRIS, A. (1991) Enforced Bcl-2 expression in B-lymphoid cells prolongs antibody responses and elicits autoimmune disease. Proc. Natl Acad. Sci. USA 88, 8661–8665. . KANTOR, A.B. & HERZENBERG, L.A. (1993) Origin of murine B cell lineages. Ann. Rev. Immunol. 11, 501–538. . TSUBATA, T. (1995) B-1 cells (Ly-1 or CD5+ B cells), apoptosis and autoimmunity. In: Apoptosis And The Immune Response, GREGORY, C.D. (ed). Wiley-Liss Inc, New York, pp. 217–256.
218
55
56 57 58 59 60
61 62 63
64 65
66 67 68 69
70 71 72
73
74
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. MURAKAMI, M., TSUBATA, T., OKAMOTO, M., SHIMIZU, A., KUMAGAI, S., IMURA, H. & HONJO, T. (1992) Antigen-induced apoptotic death of Ly-1 B cells responsible for autoimmune disease in transgenic mice. Nature 357, 77–80. . DUKE, R.C. & COHEN, J.J. (1986) IL-2 addiction: withdrawal of growth factor activates a suicide program in dependent T cells. Lymphokine Res. 5, 289–299. . NEWELL, M.K., HAUGHN, L.J., MAROUN, C.R. & JULIUS, M.H. (1990) Death of mature T cells by separate ligation of CD4 and the T-cell receptor for antigen. Nature 347, 286–289. . MCCORMACK, J.E., CALLAHAN, J.E., KAPPLER, J. & MARRACK, P.C. (1993) Profound deletion of mature T cells in vivo by chronic exposure to exogenous superantigen. J. Immunol. 150, 3785–3792. . LENARDO, M.J. (1991) Interleukin-2 programs mouse T lymphocytes for apoptosis. Nature 353, 858–861. . CRITCHFIELD, J.M., BOEHME, S.A. & LENARDO, M.J. (1995) The regulation of antigen-induced apoptosis in mature T lymphocytes. In: Apoptosis And The Immune Response, GREGORY, C.D. (ed). Wiley-Liss Inc, New York, pp. 55–114. . BOEHME, S.A. & LENARDO, M.J. (1993) Propriocidal apoptosis of mature T lymphocytes occurs at S phase of the cell cycle. Eur. J. Immunol. 23, 1552–1560. . BOEHME, S.A. & LENARDO, M.J. (1993) Ligand-induced apoptosis of mature T lymphocytes (propriocidal regulation) occurs at distinct stages of the cell cycle. Leukemia 7, S45–S49. . CRITCHFIELD, J.M., RACKE, M.K., ZUNIGA-PFLUCKER, J.C., CANNELLA, B., RAINE, C.S., GOVERMAN, J. & LENARDO, M.J. (1994) T cell deletion in high antigen dose therapy of autoimmune encephalomyelitis. Science 263, 355–361. . AKBAR, A.N., SALMON, M., SAVILL, J. & JANOSSY, G. (1993) A possible role for bcl-2 in regulating Tcell memory—a ‘balancing act’ between death and survival. Immunol. Today 14, 526–532. . AKBAR, A.N., BORTHWICK, N., SALMON, M., GOMBERT, W., BOFILL, M., SHAMSADEEN, N., PILLING, D., PETT, S., GRUNDY, J.E. & JANOSSY, G. (1993) The significance of low bcl-2 expression by CD45RO T cells in normal individuals and patients with acute viral infections. J. Exp. Med. 178, 427–438. . MACLENNAN, I.C.M. & GRAY, D. (1986) Antigen-driven selection of virgin and memory B cells. Immunol. Rev. 91, 61–85. . LIU, Y.-J., JOHNSON, G.D., GORDON, J. & MACLENNAN, I.C.M. (1992) Germinal centres in T celldependent antibody responses. Immunol Today 13, 17–21. . LIU, Y.-J., JOSHUA, D.E., WILLIAMS, G.T., SMITH, C.A., GORDON, J. & MACLENNAN, I.C.M. (1989) Mechanism of antigen-driven selection in germinal centres. Nature 342, 929–931. . LIU, Y.-J., CAIRNS, J.A., HOLDER, M., ABBOT, S.D., JANSEN, K.U., BONNEFOY, J.-Y., GORDON, J. & MACLENNAN, I.C.M. (1991) Recombinant 25-kDa CD23 and interleukin 1 promote the survival of germinal center B cells: evidence for bifurcation in the development of centrocytes rescued from apoptosis. Eur. J. Immunol. 21, 1107–1114. . GORDON, J., KATIRA, A., HOLDER, M., MACDONALD, I. & POUND, J. (1994) Central role of CD40 and its ligand in B lymphocyte responses to T-dependent antigens. Cell. Mol. Biol. 40 (Suppl. I), 1–13. . MALLETT, S. & BARCLAY, A.N. (1991) A new superfamily of cell surface proteins related to the nerve growth factor receptor. Immunol. Today 12, 220–223. . ARMITAGE, R.J., FANSLOW, W.C., STROCKBINE, L., SATO, T.A., CLIFFORD, K.N., MACDUFF, B.M., ANDERSON, D.M., GIMPEL, S.D., DAVIS-SMITH, T., MALISZEWSKI, C.R., CLARK, E.A., SMITH, C.A., GRABSTEIN, K.H., COSMAN, D. & SPRIGGS, M.K. (1992) Molecular and biological characterization of a murine ligand for CD40. Nature 357, 80–82. . TRAUTH, B.C., KLAS, C., PETERS, A.M.J., MATZKU, S., MOLLER, P., FALK, W., DEBATIN, K.-M. & KRAMMER, P.H. (1989) Monoclonal antibody-mediated tumor regression by induction of apoptosis. Science 245, 301–305. . YONEHARA, S., ISHII, A. & YONEHARA, M. (1989) A cell-killing monoclonal antibody (anti-Fas) to a cell surface antigen co-downregulated with the receptor of tumor necrosis factor. J. Exp. Med. 169, 1747–1756.
APOPTOSIS IN THE IMMUNE SYSTEM
75
76
77 78 79 80 81
82
83
84 85
86
87 88 89 90 91 92
93 94
219
. ITOH, N., YONEHARA, S., ISHII, A., YONEHARA, M., MIZUSHIMA, S.-I., SAMESHIMA, M., HASE, A., SETO, Y. & NAGATA, S. (1991) The polypeptide encoded by the cDNA for human cell surface antigen Fas can mediate apoptosis. Cell 66, 233–243. . OEHM, A., BEHRMANN, I., FALK, W., PAWLITA, M., MAIER, G., KLAS, C., LI-WEBER, M., RICHARDS, S., DHEIN, J., TRAUTH, B.C., PONSTINGL, H. & KRAMMER, P.H. (1992) Purification and molecular cloning of the APO-1 cell surface antigen, a member of the tumor necrosis factor/nerve growth factor receptor superfamily. Sequence identity with the Fas antigen. J. Biol. Chem. 276, 10709–10715. . WONG, G.H. & GOEDDEL, D.V. (1994) Fas antigen and p55 TNF receptor signal apoptosis through distinct pathways. J. Immunol. 152, 1751–1755. . ITOH, N. & NAGATA, S. (1993) A novel protein domain required for apoptosis. Mutational analysis of human Fas antigen. J. Biol. Chem. 268, 10932–10937. . SUDA, T., TAKAHASHI, T., GOLSTEIN, P. & NAGATA, S. (1993) Molecular cloning and expression of the Fas ligand, a novel member of the tumor necrosis factor family. Cell 75, 1169–1178. . SUDA, T. & NAGATA, S. (1994) Purification and characterization of the Fas ligand that induces apoptosis. J. Exp. Med. 179, 873–879. . CHENG, J., ZHOU, T., LIU, C., SHAPIRO, J.P., BRAUER, M.J., KIEFER, M.C., BARR, P.J. & MOUNTZ, J.D. (1994) Protection from Fas mediated apoptosis by a soluble form of the Fas molecule. Science 263, 1759–1762. . YOSHINO, T., KONDO, E., CAO, L., TAKAHASHI, K., NOMURA, S. & AKAGI, T. (1994) Inverse expression of bcl-2 protein and Fas antigen in lymphoblasts in peripheral blood T and B lymphocytes. Blood 83, 1856–1861. . ALDERSON, M.R.,ARMITAGE,R.J., MARASKOVSKY, E., TOUGH, T.W., ROUX, E., SCHOOLEY, K., RAMSDELL, F. & LYNCH, D.H. (1993) Fas transduces activation signals in normal human T lymphocytes. J. Exp. Med. 178, 2231–2235. . COHEN, P.J. & EISENBERG, R.A. (1995) Fas/Apo-1: A cell surface receptor that signals apoptosis. In: Apoptosis And The Immune Response, GREGORY, C.D. (ed). Wiley-Liss Inc, New York, pp. 169–186. . TAKAHASHI, T., TANAKA, M, BRANNAN, C.I., JENKINS, N.A., COPELAND, N.G., SUDA, T. & NAGATA, S. (1994) Generalized lymphoproliferative disease in mice, caused by a point mutation in the Fas ligand. Cell 76, 969–976. . ALLEN, R.D., MARSHALL, J.D., ROTHS, J.B. & SIDMAN, C.L. (1990) Differences defined by bone marrow transplantation suggest that Ipr and gld are mutations of genes encoding an interacting pair of molecules. J. Exp. Med. 172, 1367–1375. . TSCHOPP, J. & NABHOLZ, M. (1990) Perforin-mediated target cell lysis by cytolytic T lymphocytes. Ann. Rev. Immunol. 8, 279–302. . GOLSTEIN, P., OJCIUS, D.M. & YOUNG, J.D.E. (1991) Cell death mechanisms and the immune system. Immunol. Rev. 121, 29–65. . PODACK, E.R., HENGARTNER, H. & LICHTENHELD, M.G. (1991) A central role of perform in cytolysis? Ann. Rev. Immunol. 9, 129–157. . DUKE, R.C. (1991) Apoptosis in cell-mediated immunity. In: Apoptosis: The Molecular Basis of Cell Death, TOMEI, L.D. & COPE, F.O. (eds). Cold Spring Harbour Laboratory Press, pp. 209–226. . YAGITA, H., NAKATA, M., KAWASAKI, A., SHINKAI, Y. & OKUMURA (1992) Role of perform in lymphocyte-mediated cytolysis. Adv. Immunol. 51, 215–242. . GOLSTEIN, P., LUCIANI, M.-F., MATTÉI, M.-G. & FOA, C. (1995) Molecular mechanisms of T lymphocyte cytotoxicity, with emphasis on the Fas pathway. In: Apoptosis And The Immune Response, GREGORY, C.D. (ed). Wiley-Liss Inc, New York, pp. 143–168. . HENKART, P.A. (1985) Mechanism of lymphocyte-mediated cytotoxicity. Ann. Rev. Immunol. 3, 31–58. . KÄGI, D., LEDERMAN, B., BURKI, K., SEILER, P., ODERMATT, B., OLSEN, K.J., PODACK, E.R., ZINKERNAGEL, R.M. & HENGARTNER, H. (1994) Cytotoxicity mediated by T cells and natural killer cells is greatly impaired in perforin deficient mice. Nature 369, 31–37.
220
95
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. UCKER, D.S., WILSON, J.D. & HEBSHI, L.D. (1994) Target cell death triggered by cytotoxic T lymphocytes: a target cell mutant distinguishes passive pore formation and active cell suicide mechanisms. Mol. Cell. Biol. 14, 427–436. 96 . MARTZ, E. & HOWELL, D.M. (1989) CTL: virus control cells first and cytolytic cells second? Immunol. Today 10, 79–86. 97 . HAYES, M.P., BERREBI, G.A., HENKART, P.A. (1989) Induction of target cell DNA release by the cytotoxic T lymphocyte granule protease granzyme A. J. Exp. Med. 170, 933–946. 98 . SHIVER, J.W., SU, L. & HENKART, P.A. (1992) Cytotoxicity with target DNA breakdown by rat basophilic leukemia cells expressing both cytolysin and granzyme A. Cell 71, 315–322. 99 . SHI, L. KAM, C.-M., POWERS, J.C., AEBERSOLD, R. & GREENBERG, A.H. (1992) Purification of three cytotoxic lymphocyte granule serine proteases that induce apoptosis through distinct substrate and target cell interactions. J. Exp. Med. 176, 1521– 1529. 100 . HEUSEL, J.W., WESSELSCHMIDT, R.L., SHRESTA, S., RUSSELL, J.H. & LEY, T.J. (1994) Cytotoxic lymphocytes require granzyme B for the rapid induction of DNA fragmentation and apoptosis in allogeneic target cells. Cell 76, 977–987. 101 . TIAN, Q., STREULI, M., SAITO, H., SCHLOSSMAN, S.F. & ANDERSON, P. (1991) A polyadenylate binding protein localized to the granules of cytolytic lymphocytes induces DNA fragmentation in target cells. Cell 67, 629–639. 102 . KAWAKAMI, A., TIAN, Q., DUAN, X., STREULI, M., SCHLOSSMAN, S.F. & ANDERSON, P. (1992) Identification and functional characterization of a TIA-1-related nucleolysin. Proc. Natl Acad. Sci. USA 89, 8681–8685. 103 . ROUVIER, E., LUCIANI, M.-F., GOLSTEIN, P. (1993) Fas involvement in Ca2+-independent T cell-mediated cytotoxicity. J. Exp. Med. 177, 195–200. 104 . KÄGI, D., VIGNAUX, F., LEDERMAN, B., BURKI, K., DEPRAETERE, V., NAGATA, S., HENGARTNER, H., & GOLSTEIN, P. (1994) Fas and perform pathways as major mechanisms of T cell mediated cytotoxicity. Science 265, 528–530. 105 . HELLEWELL, P.G. & HENSON, P.M. (1993) Neutrophils and their mediators. In: Clinical Aspects of Immunology, LACHMANN, P.J. et al. (eds), Black well Scientific Publications , pp. 505–522. 106 . SAVILL, J.S., WYLLIE, A.H., HENSON, J.E., WALPORT, M.J., HENSON, P.M. & HASLETT, C. (1989) Macrophage phagocytosis of aging neutrophils in inflammation. Programmed cell death in the neutrophil leads to its recognition by macrophages. J. Clin. Invest. 83, 865–875. 107 . MARTIN, S.J., BRADLEY, J.G. & COTTER, T.G. (1990) HL60 cells induced to differentiate towards neutrophils subsequently die via apoptosis. Clin. Exp. Immunol. 79, 448–453. 108 . LAGASSE, E. & WEISSMAN, I.L. (1994) bcl 2 inhibits apoptosis of neutrophils but not their engulfment by macrophages. J. Exp. Med. 179, 1047–1052. 109. HASLETT, C. (1992) Resolution of acute inflammation and the role of apoptosis in the tissue fate of granulocytes. Clin. Sci. Colch. 83, 639–648. 110 . KANNAN, Y., USAMI, K., OKADA, M., SHIMIZU, S. & MATSUDA, H. (1992) Nerve growth factor suppresses apoptosis of murine neutrophils. Biochem. Biophys. Res. Commun. 186, 1050–1056. 111 . TAKEDA, Y., WATANABE, H., YONEHARA, S., YAMASHITA, T., SAITO, S. & SENDO, F. (1993) Rapid acceleration of neutrophil apoptosis by tumor necrosis factor alpha. Int. Immunol. 5, 691–694. 112 . YAMAMOTO, C., YOSHIDA, S.-I., TANIGUCHI, H., QIU, M.H., MIYAMOTO, H. & MIZUGUCHI, Y. (1993). Lipolysaccharide and granulocyte colony stimulating factor delay neutrophil apoptosis and ingestion by guinea pig macrophages. Infect. Immun. 61, 1972–1979. 113 . WHYTE, M.K., MEAGHER, L.C., MACDERMOT, J. & HASLETT, C. (1993) Impairment of function in aging neutrophils is associated with apoptosis. J. Immunol. 150, 5124–5134. 114 . DRANSFIELD, I., MCDOWALL, A., SAVILL, J.S., BUCKLE, A.-M., HASLETT, C. & HOGG, N. (1994) Neutrophil apoptosis is associated with a reduction in CD 16 (Fc RIII) expression. J. Immunol. 153, 1254–1263.
APOPTOSIS IN THE IMMUNE SYSTEM
115 116 117 118 119 120 121
122
123
124 125
126
127 128 129
130 131 132 133
134
221
. SAVILL, J. (1995) The innate immune system: recognition of apoptotic cells. In: Apoptosis And The Immune Response, GREGORY, C.D. (ed). Wiley-Liss Inc, New York, pp. 341–369. . ELLIS, R.E., JACOBSON, D.M. & HORVITZ, H.R. (1991) Genes required for the engulfment of cell corpses during programmed cell death in Caenorrhabditis elegans. Genetics 129, 7–94. . SAVILL, J., FADOK, V., HENSON, P. & HASLETT, C. (1993) Phagocyte recognition of cells undergoing apoptosis. Immunol. Today 14, 131–136. . DUVALL, E., WYLLIE, A.H. & MORRIS, R.G. (1985) Macrophage recognition of cells undergoing programmed cell death. Immunology 56, 351–358. . MORRIS, R.G., HARGREAVES, A.D., DUVALL, E. & WYLLIE, A.H. (1984) Hormone-induced cell death 2. Surface changes in thymocytes undergoing apoptosis. Am. J. Pathol. 115, 426–436. . SAVILL, J., DRANSFIELD, I., HOGG, N. & HASLETT, C. (1990) Vitronectinreceptor mediated phagocytosis of cells undergoing apoptosis. Nature 343, 170–173. . SAVILL, J., HOGG, N., REN, Y. & HASLETT, C. (1992) Thrombospondin cooperates with CD36 and the vitronectin receptor in macrophage recognition of neutrophils undergoing apoptosis. J. Clin. Invest. 90, 1513–1522. . FADOK, V.A., VOELKER, D.R., CAMPBELL, P.A., COHEN, J.J., BRATTON, D.L. & HENSON, P.M. (1992) Exposure of phosphatidylserine on the surface of apoptotic lymphocytes triggers specific recognition and removal by macrophages. J. Immunol. 148, 2207–2216. . FADOK, V.A., SAVILL, J.S., HASLETT, C., BRATTON, D.L., DOHERTY, D.E., CAMPBELL, P.A. & HENSON, P.M. (1992) Different populations of macrophages use either the vitronectin receptor or the phosphatidylserine receptor to recognize and remove apoptotic cells. J. Immunol. 149, 4029–4035. . FLORA, P.K. & GREGORY, C.D. (1994) Recognition of apoptotic cells by human macrophages: inhibition by a monocyte/macrophage-specific monoclonal antibody. Eur. J. Immunol. 24, 2625–2632. . FLORA, P.K. & GREGORY, C.D. (1994) Recognition pathways in the interaction of macrophages with apoptotic cells. In: SCHLOSSMAN et al. (eds), Leucocyte Typing V, Oxford University Press, Oxford, pp. 1675–1677. . GOUGEON, M.-L. & MONTAGNIER, L. (1994) Apoptosis in peripheral T lymphocytes during HIV infection: influence of superantigens and correlation with AIDS pathogenesis. In: Apoptosis II: The Molecular Basis of Apoptosis in Disease, TOMEI. L.D. & COPE, F.O. (eds). Cold Spring Harbour Laboratory Press, pp. 5–19. . AMEISEN, J.C. (1995) HIV infection and T-cell death. In: Apoptosis And The Immune Response, GREGORY, C.D. (ed). Wiley-Liss Inc, New York, pp. 115–142. . TSUBATA, T., MURAKAMI, M. & HONJO, T. (1994) Antigen-receptor cross-linking induces peritoneal B-cell apoptosis in normal but not autoimmunity-prone mice. Curr. Biol. 4, 8–17. . CASCIOLA-ROSEN, L.A., ANHALT, G. & ROSEN, A. (1994) Autoantigens targeted in systemic lupus erythematosus are clustered in two populations of surface structures on apoptotic keratinocytes. J. Exp. Med. 179, 1317–1330. . VAUX, D.L., CORY, S. & ADAMS, J.M. (1988) Bcl-2 gene promotes haemopoietic cell survival and cooperates with c-myc to immortalize pre-B cells. Nature 335, 440–442. . STRASSER, A., HARRIS, A.W., BATH, M.L. & CORY, S. (1990) Novel primitive lymphoid tumours induced in transgenic mice by cooperation between myc and bcl-2. Nature 348, 331–333. . MCDONNELL, T.J. & KORSMEYER, S.J. (1991) Progression from lymphoid hyperplasia to high-grade malignant lymphoma in mice transgenic for the t(14;18). Nature 349, 254–256. . ASKEW, D.S., ASHMUN, R.A., SIMMONS, B.C. & CLEVELAND, J.L. (1991) Constitutive c-myc expression in an IL-3-dependent myeloid cell line suppresses cell cycle arrest and accelerates apoptosis. Oncogene 6, 1915–1922. . EVAN, G.I., WYLLIE, A.H., GILBERT, C.S., LITTLEWOOD, T.D., LAND, H., BROOKS, M., WATERS, C.M., PENN, L.Z. & HANCOCK, D.C. (1992) Induction of apoptosis in fibroblasts by c-myc protein. Cell 69, 119–128.
222
135 136
137
138
139
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. MILNER, A.E., GRAND, R.J.A., WATERS, C.M. & GREGORY, C.D. (1993) Apoptosis in Burkitt lymphoma cells is driven by c-myc. Oncogene 8, 3385–3391. . HENDERSON, S., ROWE, M., GREGORY, C.D., CROOM-CARTER, D., WANG, F., LONGNECKER, R., KIEFF, E. & RICKINSON, A.B. (1991) Induction of bcl-2 expression by Epstein-Barr virus latent membrane protein 1 protects infected B cells from programmed cell death. Cell 65, 1107–1115. . GREGORY, C.D., DIVE, C., HENDERSON, S., SMITH, C.A., WILLIAMS, G.T., GORDON, J. & RICKINSON, A.B. (1991) Activation of Epstein-Barr virus latent genes protects human B cells from death by apoptosis. Nature 349, 612–614. . RENSCHLER, M.F., BHATT, R.R., DOWER, W.J. & LEVY, R. (1994) Synthetic peptide ligands of the antigen binding receptor induce programmed cell death in a human B cell lymphoma. Proc. Natl Acad. Sci. USA 91, 3623–3627. . VUIST, W.M., LEVY, R. & MALONEY, D.G. (1994) Lymphoma regression induced by monoclonal antiidiotypic antibodies correlates with their ability to induce Ig signal transduction and is not prevented by tumor expression of high levels of bcl-2 protein. Blood 83, 899–906.
12 Suppression of Apoptosis by Monoclonal Antibodies in Mouse Malignant T-lymphoma Cells NAOYA FUJITA, SHIRO KATAOKA and TAKASHI TSURUO Institute of Molecular and Cellular Biosciences, University of Tokyo, Yayoi 1–1–1, Bunkyo-ku, Tokyo 113, Japan
12.1 Introduction Lymphatic vessels are the most common passageways for cancer metastasis, and the regional lymph nodes draining the site of primary tumors are usually the first sites of metastasis in cancer patients. In order to establish metastasis in lymph nodes, it is necessary for tumor cells to adhere preferentially to lymph nodes using cell-surface adhesion molecules. It has been shown that the expression of CD44 variants facilitates metastasis of rat pancreatic and mammary carcinoma to lung and lymph nodes [1]. CD44 [2–4] and other cellsurface adhesion molecules, including LFA-1 [5] and MEL-14 antigen [6], have been implicated in metastasis of lymphoma cells. The precise mechanisms underlying lymphoma metastasis, however, have not been elucidated. We have previously established a mouse malignant T-lymphoma CS-21 (CMS) cell line from a lymphoma that developed spontaneously in a BALB/c mouse. CS-21 lymphoma cells produce a high incidence of lymph node metastasis following subcutaneous (s.c.) injection [7]. CS-21 lymphoma cells grow continuously in vitro when they are cocultured under the cell-cell attached conditions with CA-12 stromal cells prepared from lymph nodes. CS-21 lymphoma cells, however, are unable to proliferate by themselves and undergo apoptosis when separated from CA-12 stromal cells [8]. As a result of our study on the growth properties of CS-21 lymphoma cells, we found that CS-21 lymphoma cells required for cell growth at least two types of molecule provided by CA-12 stromal cells—cell adhesion molecules and soluble factors. Soluble factors by themselves, however, could not prevent CS-21 lymphoma cells from undergoing apoptotic cell death. Thus, we have previously postulated that the cell-adhesion molecules, not the soluble factors, play a crucial role in CS-21 cell survival [8]. To study further the possible involvement of cell adhesion molecules in CS-21 lymphoma cell growth in lymph nodes, we prepared monoclonal antibodies (mAbs) which recognized CS-21 cell-surface molecules and partially inhibited adhesion of lymphoma cells to the stromal cells. Then we examined the effects of the antibodies on growth and survival of CS-21 lymphoma cells. Some mAbs against cell adhesion molecules, such as MCS-5 directed against a 168-kDa cell-surface protein and MCS-19 against a 23-kDa protein, suppressed apoptosis and stimulated growth of CS-21 lymphoma cells [9].
224
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 12.1 Morphological analysis of CS-21 lymphoma cells. CS-21 lymphoma cells were examined and photographed using a Nikon Diaphot light microscope before separation (A), immediately after the separation (B), and 24 h after separation (C) from CA-12 stromal cells. Bars=100µm.
Our results indicate that the paracrine factors secreted from CA-12 lymph node stromal cells support CS-21 cell proliferation and that CS-21 lymphoma cell adhesion by 168-kDa and 23-kDa cell-surface proteins to the stromal cell prevents apoptosis of CS-21 lymphoma cells. We assume that apoptosisinhibitory signals transmitted through 168-kDa and 23-kDa proteins on CS-21 cell surfaces may be involved in CS-21 cell survival and growth in lymph nodes [9]. 12.2 Apoptosis of Malignant T-lymphoma Cells 12.2.1 Apoptosis of CS-21 Cells in vitro We have previously established two cell lines, CMS and CMA, from a malignant lymphoma that developed spontaneously in the axillary lymph node of a BALB/c mouse [7]. CS-21 lymphoma cells and CA-12 stromal cells were cloned from the CMS and CMA cells, respectively [8]. CA-12 stromal cells injected s.c. did not form a tumor mass in mice, but the cells have properties similar to the stromal cells in lymph nodes. CS-21 lymphoma cells, however, possessed tumorigenecity as high as that of parental CMS cells. CS-21 cells were Thy-1.2 and L3T4 (CD4)-positive and Lyt 2.1 (CD8)-negative malignant T-lymphoma cells. When CS-21 lymphoma cells were cultured with CA-12 lymph node stromal cells, the lymphoma cells grew exponentially in a time-dependent manner. Most CS-21 lymphoma cells adhered to CA-12 stromal cells but some were found beneath the stromal cells (Figure 12.1 A). When CS-21 lymphoma cells were separated from the CA-12 stromal cells and cultured alone, their viability rapidly declined (Figure 12.2). The morphological features of CS-21 lymphoma cells were examined following the separation. Almost all of the cells were round and appeared immediately to be viable (Figure 12.1B). Viable CS-21 lymphoma cells have a large round nucleus and a relatively small volume of cytosol (Figure 12.3A). After 24 h in isolated culture, however, many of the lymphoma cells shrank and appeared to be dead (Figure 12.1C). The nuclei of some cells were condensed and fragmented (Figure 12.3B). The cellular membranes were convoluted, and the cells separated into ‘apoptotic bodies’, which contained morphologically intact
SUPPRESSION OF APOPTOSIS BY MONOCLONAL ANTIBODIES
225
Figure 12.2 Effects of CA-12 stromal cells on growth of CS-21 lymphoma cells. CS-21 lymphoma cells were freshly isolated from CA-12 stromal cells. The lymphoma cells (5×104 cells/ml) were cultured with (open circles) or without (closed circles) CA-12 stromal cells (1×104 cells/ml) that had been plated on the previous day. The number of viable cells was determined by trypan blue exclusion. The vertical bars represent SD values of triplicate determinations.
mitochondria and other organelles. These electron microscopic observations clearly indicated that CS-21 lymphoma cells died by apoptosis (programmed cell death) [10] when separated from CA-12 stromal cells [8]. To determine the type of cell death more precisely, DNA fragmentation analysis was carried out (Figure 12.4). Fragmented DNA was not detected in CS-21 lymphoma cells immediately after separation from CA-12 stromal cells. However, after only 3 h of incubation, the degradation of DNA into oligonucleosomal fragments was observed. This type of degradation is characteristic of apoptotic cell death [11]. These results confirm that CS-21 lymphoma cells underwent apoptosis with DNA fragmentation when they were separated from CA-12 stromal cells [8]. 12.2.2 Suppression of Apoptosis by Adhesion to Stromal Cells Because proliferation and suppression of apoptosis in CS-21 lymphoma cells was observed only when they were cocultured with CA-12 stromal cells, some signals for growth and suppression of apoptosis could be provided by CA-12 stromal cells. We first examined the possible involvement of a soluble factor (or factors) secreted by CA-12 stromal cells in cell growth and apoptosis inhibition, because many studies have reported their importance in these processes. For example, interleukin-2 (IL-2)-dependent cytotoxic T-cell line, CTLL-2, underwent apoptosis without IL-2 in the cell culture [12] but could survive and grow when IL-2 was added [13]. Other cytokines or growth factors, such as IL-3 [14, 15], IL-5 [16], IL-6 [17], G-CSF [15], GM-CSF [15], and erythropoietin [18], had similar effects on various cell lines. The conditioned medium from the cocultures of CA-12 stromal cells and CS-21 lymphoma cells (CS-21/ CA-12 CM) increased DNA synthesis of CS-21 lymphoma cells. The total number of viable cells, however, were kept at the initial levels in 40% cocultured conditioned medium. The DNA fragmentation analysis
226
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 12.3 Transmission electron micrograph of CS-21 lymphoma cells. CS-21 lymphoma cells were separated from CA-12 stromal cells and cultured for 0 h (A) and 24 h (B). Micrographs were taken with an Hitachi H7000 electron microscope. N indicates the normal nucleus. Arrows point to condensed and fragmented nucleus in an apoptotic CS-21 cell. Bars=5µm.
showed that adding the conditioned medium did not suppress the degradation of nuclear DNA into oligonucleosomal fragments [8]. To identify the soluble factors, CS-21 lymphoma cells were cultured in the presence of IL-1, IL-2, IL-3, IL-4, IL-5, IL-6, bFGF, NGF. These cytokines and growth factors, however, could not stimulate growth of CS-21 lymphoma cells. Other cytokines or growth factors might be involved in the growth promotion. Interestingly, we found that adding CA-12 stromal cells to the CS-21 lymphoma cell culture suppressed DNA degradation and significantly enhanced DNA synthesis and CS-21 cell growth. CA-12 stromal cells at
SUPPRESSION OF APOPTOSIS BY MONOCLONAL ANTIBODIES
227
Figure 12.4 Agarose gel electrophoresis of DNA of CS-21 lymphoma cells after separation from CA-12 stromal cells. CS-21 lymphoma cells were freshly isolated from the cultures with CA-12 stromal cells and then incubated for 0, 3, 6, 12, 18, and 24 h (lanes 2–7, respectively) without CA-1 stromal cells. DNA was isolated from the cells and applied to 2% agarose gel [41]. Molecular size markers are the 100-bp DNA ladder (lane 1).
105 cells/10-cm dish especially promoted growth of the lymphoma cells and completely suppressed the degradation of nuclear DNA. The results indicate that some soluble factors, those that stimulate the proliferation of CS-21 lymphoma cells but do not prevent apoptosis, are secreted by CA-12 stromal cells and that a signal that promotes the survival of CS-21 lymphoma cells may be transmitted by CA-12 stromal cells through direct cell-cell contact [8]. Table 12.1 Anti-CS-21 Monoclonal antibodies No. of the clone
Ig Class
Antigen MW (kDa)
MCS-1 MCS-5, 17, 55, 72 MCS-12, 19, 24, 34, 41, 65, 81, 87, 93
IgG IgG IgG
15 168 23
228
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
12.3 Suppression of Apoptosis by Monoclonal Antibodies Against Cell Adhesion Molecules 12.3.1 Production and Characterization of anti-CS-21 mAbs To identify which cell adhesion molecules mediated cell adhesion to CA-12 stromal cells, we immunized five SD rats with intrasplenic injections [19] of viable CS-21 lymphoma cells. We successfully raised 14 clones of the hybridomas, which produced mAbs and partially inhibited CS-21 cell binding to a monolayer of CA-12 stromal cells [9]. The molecular sizes of the antigens recognized by these mAbs were determined. One of the 14 mAbs recognized a 15-kDa cell surface protein, nine bound 23-kDa proteins, and four bound 168-kDa proteins under both reducing and nonreducing conditions (Table 12.1). As shown in Figure 12.5, MCS-1, MCS-5, and MCS-19 immunoprecipitated CS-21 cell surface molecules of 15 kDa, 168 kDa, and 23 kDa, respectively. When we added CS-21 lymphoma cells to a CA-12 stromal cell monolayer, the lymphoma cells adhered to the stromal cells in about 1 h and grew. Some lymphoma cells, grew beneath the stromal cells, called emperipolesis [20, 21]. The purified antibodies (MCS-1, 5, and 19) reduced CS-21id cell (a CS-21 variant that can grow without CA-12 cells) adhesion to a CA-12 cell monolayer by 20–30% at a concentration of 10 µg/ ml, but they did not reduce emperipolesis. As each antibody recognized different antigens with different molecular weights, CS-21 lymphoma cells seemed to posses at least three groups of cell adhesion molecules. The antigens of these mAbs may be the adhesion molecules that mediate lymphocyte binding to stromal and endothelial cells, such as the immunoglobulin-like superfamily [22], integrin superfamily [23], and LEC-CAM family [24]. In fact, such cell adhesion molecules as LFA-1 [25, 26] and VLA-4 [27, 28] were expressed on the surface of CS-21 lymphoma cells. However, the molecular sizes of the cell surface antigens of MCS-1, 5, and 19 were 15 kDa, 168 kDa, and 23 kDa, respectively, which suggests that these antigens were neither LFA-1 nor VLA-4. 12.3.2 Inhibition of Apoptosis by mAbs Because soluble growth factors secreted from CA-12 lymph node stromal cells failed to suppress apoptosis of CS-21 lymphoma cells, cell-cell contact might be important for apoptosis inhibition. Reports indicate that many cell adhesion molecules (e.g. LFA-1 [29], CD28 [30]) not only participate in cell-cell adhesion but also in signal transduction. Therefore, we examined the effect on apoptosis suppression of the mAbs capable of inhibiting cell-cell adhesion. We found that MCS-5 and MCS-19 at 10 mg/ml suppressed DNA fragmentation, but MCS-1, normal rat IgG, or conditioned medium from CA-12 stromal cells did not (Figure 12.6). The inhibition of apoptosis by MCS-5 or MCS-19 was further confirmed by quantitative estimation of DNA fragmentation. In the absence of each mAb, approximately 40% of DNA was fragmented after a 24-h culture. While in the presence of MCS-5 or MCS-19, DNA fragmentation was 17% or 5%, respectively. The fragmentation of DNA proceeded prominently in the absence of mAbs for 48 h in culture. In contrast, the inhibition of DNA fragmentation by the mAbs became more evident during this period. The proliferative effects of MCS-5 and MCS-19 were weaker than those of soluble factors from CA-12 stromal cells. However, the effects of MCS-5 and MCS-19 on CS-21 cell growth were enhanced in the
SUPPRESSION OF APOPTOSIS BY MONOCLONAL ANTIBODIES
229
Figure 12.5 Immunoprecipitation of the antigens recognized by MCS-1, MCS-5 and MCS-19. The cell surface proteins of CS-21 lymphoma cells were biotin-labeled with NHS-LC-biotin [42, 43] and then solubilized with buffer containing 1% Triton X-100 [9]. The proteins bound to mAbs were immunoprecipitated using protein G-sepharose 4B gel beads coated with mAbs. After electrophoresis with -mercaptoethanol using a 4–20% gradient polyacrylamide gel [44], the proteins were transblotted onto a nitrocellulose membrane and detected by incubating with ECL mixture using Kodak XOmat AR film.
presence of the soluble factors. Actually, the numbers of viable CS-21 lymphoma cells cultured in the presence of both soluble factors and mAbs were almost equal to those in the coculture with CA-12 stromal cells [9]. These results indicate that 23-kDa and 168-kDa cell adhesion molecules, recognized by MCS-5 and MCS-19, respectively, transmit apoptosis-inhibitory signals from CA-12 stromal cells and enhance CS-21 cell growth in the presence of soluble factors (Figure 12.7). 12.4 Summary Apoptosis is a cellular suicide functionally opposite to mitosis. It may play an important role in tissue growth control and removal of damaged and premalignant cells. Apoptosis induction and inhibition are regulated not only by soluble factors but also by cell-cell contact. There have been reports on Fas antigen in
230
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 12.6 Suppression of CS-21 cell apoptosis by MCS-5 and MCS-19. CS-21 cells were freshly isolated from the cultures with CA-12 stromal cells (lane 2) and then incubated with medium alone (lane 3), 10 µg/ml of MCS-19 (lane 4), MCS-5 (lane 5), MCS-19 (lane 6), normal rat IgG (lane 7), or conditioned medium of CA-12 stromal cells (lane 8) for 24 h. DNA was isolated from each sample and analyzed for fragmentation by electrophoresis in 2% agarose gel plates [41]. Molecular size markers are the 100-bp DNA ladder (lane 1).
Figure 12.7 Possible mechanisms of CS-21 lymphoma cell growth.
various tissues [31], CD3/TCR in T cells [32], surface immunoglobulin (sIg) in B cells [33], and other molecules that participate in apoptosis induction through cell-cell contact. Similarly, some molecules have been reported which involved with apoptosis inhibition [9, 34–36], except growth factors that suppress apoptosis of growth factor-dependent cells [13–18]. The study of apoptosis induction and suppression is important because disorder in either process has resulted in various diseases, including autoimmune disease or tumor development [37–40]. The mechanisms of apoptosis inhibition through cell-cell interaction in B cells have recently been reported [34, 35]. In B cells, apoptosis was induced by crosslinking sIg [33]. Such signals of apoptosis induction was overcome by adding anti-CD40 mAb [34, 35] or by adding helper T cell clone that express the ligand of CD40 (CD40L) [35]. These findings indicate that the interactions between B cells and helper T cells through CD40-CD40L is important in B cell survival. Likewise, CS-21 cell adhesion to CA-12 stromal
SUPPRESSION OF APOPTOSIS BY MONOCLONAL ANTIBODIES
231
cells either through the 23-kDa or the 168-kDa proteins participates in apoptosis inhibition of CS-21 cells [9]. Further studies are required to characterize the 23-kDa and 168-kDa cell adhesion molecules and also to identify their ligands expressed on CA-12 stromal cells. Acknowledgement This study was supported in part by grants from the Ministry of Education, Science and Culture, Japan, and Special Coordination Fund of the Science and Technology Agency, Japan. References 1
2
3 4
5 6 7
8
9
10 11 12 13 14 15
. GÜNTHERT, U., HOFMANN M., RUDY, W., REBER, S., ZVÖLLER, M., HAUSSMANN, I., MATZKU, S., WENZEL, A., PONTA H. & HERRLICH, P. (1991) A new variant of glycoprotein CD44 confers metastatic potential to rat carcinoma cells. Cell 65, 13–24. . HORST, E., MEIJER, C.J.L.M., RADASZKIEWICZ, T., OSSEKOPPELE, G.J., VAN KRIEKEN, J.H.J.M. & PALS, S.T. (1990) Adhesion molecules in the prognosis of diffuse large-cell lymphoma: expression of a lymphocyte homing receptor (CD44), LFA-1 (CD11a/18), and ICAM-1 (CD54). Leukemia 4, 595–599. . JALKANEN, S., JOENSUU, H. & KLEMI, P. (1990) Prognostic value of lymphocyte homing receptor and S phase fraction in non-Hodgkin’s lymphoma. Blood 75, 1549–1556. . PALS, S.T., HORST, E., OSSEKOPPELE, G.J. FIGDOR, C.G., SCHEPER, R.J. & MEIJER, C.J.L.M. (1989) Expression of lymphocyte homing receptor as a mechanism of dissemination in non-Hodgkin’s lymphoma. Blood 173, 885–888. . ROOS, E. (1991) Adhesion molecules in lymphoma metastasis. Cancer Metastasis Rev. 10, 33–48. . BARGATZE, R.F., WU, N.W., WEISSMAN, I.L. & BUTCHER, E.C. (1987) High endothelial venule binding as a predictor of the dissemination of passaged murine lymphomas. J. Exp. Med. 166, 1125–1131. . TSURUO, T., OH-HARA, T., YAMORI, T., TSUKAGOSHI S., ISHIKAWA, T. & SUGANO, H. (1988) Development of lymphosarcoma lines with high metastatic ability to lymph nodes and visceral organs in BALB/c mice. Clin. Expl. Metastasis 6, 141–152. . KATAOKA, S., NAITO, M., FUJITA, N., ISHII, H., ISHII, S., YAMOTI, T., NAKAJIMA, M. & TSURUO, T. (1993) Control of apoptosis and growth of malignant T lymphoma cells by lymph node stromal cells. Exp. Cell Res. 207, 271–276. . FUJITA, N., KATAOKA, S., NAITO, M., HEIKE, Y., BOKU, N., NAKAJIMA, M. & TSURUO, T. (1993) Suppression of T-lymphoma cell apoptosis by monoclonal antibodies raised against cell surface adhesion molecules. Cancer Res. 53, 5022–5026. . KERR, J.F.R., WYLLIE, A.H. & CURRIE, A.R. (1972) Apoptosis: a basic biological phenomenon with wideranging implications in tissue kinetics. Br. J. Cancer 26, 239–257. . WYLLIE, A.H. (1980) Glucocorticoid-induced thymocyte apoptosis is associated with endogenous endonuclease activation. Nature 284, 555–556. . DUKE, R.C. & COHEN, J.J. (1986) IL-2 addiction: withdrawal of growth factor activates a suicide program in dependent T cells. Lymphokine Res. 5, 289–299. . NIETO, M.A. & LÓPEZ-RIVAS, A. (1989) IL-2 protects T lymphocytes from glucocorticoid-induced DNA fragmentation and cell death. J. Immunol. 143, 4166–4170. . RODRIGUEZ-TARDUCHY, G., COLLINS, M. & LÓPEZ-RIVAS, A. (1990) Regulation of apoptosis in interleukin-3-dependent hemopoietic cells by interleukin-3 and calcium ionophores. EMBO J. 9, 2997–3002. . WILLIAMS, G.T., SMITH, C.A., SPOONCER, E., DEXTER, T.M. & TAYLOR, D. R. (1990) Haemopoietic colony stimulating factors promote cell survival by suppressing apoptosis. Nature 343, 76–79.
232
16
17
18 19 20
21 22 23 24 25
26
27 28 29
30 31
32
33
34 35
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. YAMAGUCHI, Y., SUDA, T., OHTA, S., TOMINAGA, K., MIURA, Y. & KASAHARA, T. (1991) Analysis of the survival of mature human eosinophils: interleukin-5 prevents apoptosis in mature human eosinophils: Blood 78, 2542–2547. . COLOTTA, F., POLENTARUTTI, N., SIRONI, M. & MANTOVANI, A. (1992) Expression and involvement of c-fos and c-jun protooncogene in programmed cell death induced by growth factor deprivation in lymphoid cell lines. J. Biol. Chem. 267, 18278–18283. . KOURY, M.J. & BONDURANT, M.C. (1990) Erythropoietin retards DNA breakdown and prevents programmed death in erythroid progenitor cells. Science 248, 378–381. . SPITZ, M., SPITZ, L., THORPE, R. & EUGUI, E. (1984) Intrasplenic primary immunization for the production of monoclonal antibodies. J. Immunol. Meth. 70, 39–43. . DREYER, D.A., SHULLENBERGER, C.C. & DMOCHOWSKI, L. (1964) A study on intracellular lymphocytes (‘emperipolesis’) in tissue culture of lymph nodes from patients with malignant lymphoma. Texas Rep. Biol. Med. 22, 61–69. . IOACHIM, H.L. (1965) Emperipolesis of lymphoid cells in mixed cultures. Lab. Invest. 14, 1784–1794. . WILLIAMS, A.F. & BARCLAY, A.N. (1988) The immunoglobulin superfamily-domains for cell surface recognition. Annu. Rev. Immunol. 6, 381–405. . HYNES, R.O. (1987) Integrins: A family of cell surface receptors. Cell 48, 549–554. . BRANDLEY, B.K., SWIEDLER, S.J. & ROBBINS, P.W. (1990) Carbohydrateligands of the LEG cell adhesion molecules. Cell 63, 861–863. . SANCHEZ-MADRID, F., NAGY, J.A., ROBBINS, E., SIMON, P & SPRINGER, T.A. (1983) A human leukocyte differentiation antigen family with distinct -subunits and a common -subunit: the lymphocyte function associated antigen (LFA-1), the C3bi complement receptor (OKM1/Mac-1), and the p150, 95 molecule. J. Exp. Med. 158, 1785– 1803. . WAWRYK, S.O., NOVOTNY, J.R., WICKS, I.P., WILKINSON, D., MAHER, D., SALVARIS, E., WELCH, K., FECONDO, J & BOYD, A.W. (1989) The role of the LFA-l/ICAM-1 interaction in human leukocyte homing and adhesion. Immunol. Rev. 108, 135–161. . HEMLER, M.E., ELICES, M.J., PARKER, C & TAKADA, Y. (1990) Structure of the integrin VLA-4 and its cell-cell and cell-matrix adhesion functions. Immunol. Rev. 114, 45–65. . TAKADA, Y., ELICES, M.J., GROUSE, C. & HEMLER, M.E. (1989) The primary structure of the 4 subunit of VLA-4: homology to other integrins and a possible cell-cell adhesion function. EMBO J. 8, 1361–1368. . VAN SEVENTER, G.A., SHIMIZU, Y., MORGAN, K.J. & SHAW, S. (1990) The LFA-1 ligand ICAM-1 provides an imporant costimulatory signal for T cell receptor-mediated activation of resting T cells. J. Immunol. 144, 4579–4586. . HARDING, F.A., MCARTHUR, J.G., GROSS, J.A., RAULET, D.A. & ALLISON, J.P. (1992) CD28-mediated signalling co-stimulates murine T cells and prevents induction of anergy in T-cell clones. Nature 356, 607–609. . WATANABE-FUKUNAGA, R., BRANNAN, C.I., ITOH, N., YONEHARA, S., COPELAND, N.G., JENKINS, N.A. & NAGATA, S. (1992) The cDNA structure, expression and chromosomal assignment of the mouse Fas antigen. J. Immunol. 148, 1274–1279. . SMITH, C.A., WILLIAMS, G.T., KINGSTON, R. JENKINSON, E.J. & OWEN, J.J.T. (1989) Antibodies to CD3/T-cell receptor complex induce death by apoptosis in immature T cells in thymic cultures. Nature 337, 181–184. . MURAKAMI, M., TSUBATA, T., OKAMOTO, M., SHIMIZU, A., KUMAGAI, S., IMURA, H. & HONJO, T. (1992) Antigen-induced apoptotic death of Ly-1 B cells responsible for autoimmune disease in transgenic mice. Nature 357, 77–80. . VALENTINE, M.A. & LICCIARDI, K.A. (1992) Rescue from anti-IgM-induced programmed cell death by the B cell surface proteins CD20 and CD40. Eur. J. Immunol. 22, 3141–3148. . TSUBATA, T., WU, J. & HONJO, T. (1993) B-cell apoptosis induced by antigen receptor crosslinking is blocked by a T-cell signal through CD40. Nature 364, 645–648.
SUPPRESSION OF APOPTOSIS BY MONOCLONAL ANTIBODIES
36
37
38
39 40 41 42
43
44
233
. KOOPMAN, G., KEEHNEN, R.M.J., LINDHOUT, E., NEWMAN, W., SHIMIZU, Y., VAN SEVENTER, G.A., DE GROOT, C. & PALS, S.T. (1994) Adhesion through the LFA-1 (CD18/CD18)-ICAM-l (CD54) and the VLA-4 (CD49d)-VCAM-l (CD106) pathways prevents apoptosis of germinal center B cells. J. Immunol. 152, 3760–3767. . WATANABE-FUKUNAGA, R., BRANNAN, C.I., COPE LAND, N.G., JENKINS, N.A. & NAGATA, S. (1992) Lymphoproliferation disorder in mice explained by defects in Fas antigen that mediates apoptosis. Nature 356, 314–317. . ADACHI M., WATANABE-FUKUNAGA, R & NAGATA, S. (1993) Aberrant transcription caused by the insertion of an early transposable element in an intron of the Fas antigen gene of lpr mice. Proc. Natl Acad. Sci. USA 90, 1756–1760. . TSUJIMOTO, Y., COSSMAN, J., JAFFE, E. & CROCE, C.M. (1985) Involvement of the bcl-2 gene in human follicular lymphoma. Science 228, 1440–1443. . MEYAARD, L., OTTO, S.A., JONKER, R.R., MIJINSTER, M.J., KEET, R.P.M. & MIEDEMA, F. (1992) Programmed death of T cells in HIV-1 infection. Science 257, 217–219. . GUNJI, H., KHARBANDA, S. & KUFE, D. (1991) Induction of internucleosomal DNA fragmentation in human myeloid leukemia cells by l- -arabinofuranosylcytosine. Cancer Res. 51, 741–743. . HSU, S.M., RAINE, L. & FANGER, H. (1981) A comparative study of the peroxidase-antiperoxidase method and avidin-biotin complex method for studying polypeptide hormones with radioimmunossay antibodies. Am. J. Clin. Pathol. 75, 734–738. . HSU, S.M., RAINE, L & FANGER, H. (1981) Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabeled antibody (PAP) procedures. J. Histochem. Cytochem. 29, 577–580. . LAEMMLI, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685.
13 Apoptosis as a Predictive Factor for Cancer Therapy FRANCESCO PEZZELLA,1 GIAMPIETRO GASPARINI,2 ADRIAN L.HARRIS3 and KEVIN C.GATTER4 1
Department of Pathology, European Institute of Oncology, Via Ripamonti 435, 20141 Milano, Italy
2
Department of Radiotherapy and Oncology, St Bortolo Hospital, 36100 Vicenza, Italy 3
4
ICRF Molecular Oncology Laboratories, Institute of Molecular Medicine, John Radcliffe Hospital, Headington, Oxford, OX3 9DU, UK
University Department of Cellular Science, John Radcliffe Hospital, Headington, Oxford, OX3 13.1 9DU, UK Introduction
The discovery of bcl-2 gene [1, 2] and the demonstration of its role in inhibiting apoptosis [3] has led to the definition of a new category of oncogenes which are involved in the regulation of programmed cell death (apoptosis) [4]. Alterations in these genes promote growth by preventing cell death, whereas the oncogenes previously known promoted cellular growth. Since the discovery of bcl-2, several genes have been described that are involved in apoptosis regulation [5–12]. The other discovery that turned out to be of particular interest was that the anti-oncogene p53 induces apoptosis [13–14] and either its deletion or inactivation are believed to prevent apoptosis from occurring, providing a growth advantage. Different agents used in cancer treatment induce apoptosis. Therefore it is possible that apoptotic pathways could indicate the outcome of chemotherapy. However, in practical terms, the subject is complex and, in approaching the problem of apoptosis as a predictive factor in cancer therapy we should keep in mind three points. The first point is the concept of predictiveness. While prognosis refers to the identification of likelihood of survival, and for how long, on the grounds of clinical and pathological findings, predictiveness is concerned with the prediction of the response to a certain treatment according to the biological characteristics of the disease. The best possible example of use of predictiveness is in infective diseases. The isolation of the aetiological agent of an infection and its testing ‘in vitro’ against a range of drugs leads us to identify the one that will be able to eradicate the infection ‘in vivo’. In tumours the situation is much more complex. A tumour can have many different subclones with different genetic alterations and metabolic pathways activated. Furthermore once we have detected an onco-protein in a neoplastic cell, we cannot always be sure that it is active: e.g. detection of wild-type p53 does not automatically mean that it is active. We need to be sure that no mdm-2 protein is co-expressed [15] binding and inactivating p53. We must therefore keep in mind the complexity of a tumour when trying to predict whether it will respond to a treatment on the grounds of the genes expressed. The second point is that assessment of apoptosis is difficult. It may not be adequate just to quantify apoptotic cells, which are just a final event. Rather it my be necessary to identify which genes are involved in its regulation and their aberrant function.
A PREDICTIVE FACTOR FOR CANCER THERAPY
235
The third point to keep in mind is that eventually, to predict whether a tumour will respond or not to a certain treatment, the expression of proteins related to other processes such as proliferation and drug resistence pathways will have to be considered. We will focus our attention on bcl-2 and p53 genes because they are the two most widely studied and because they belong to two opposite pathways respectively preventing and inducing apoptosis in the same group of human tissues. Furthermore they are presently the only two such genes that one can investigate by immunohistochemistry in both normal and neoplastic tissues. 13.2 bcl-2 and p53 Protein Expression in Normal Tissue 13.2.1 Immunohistochemical Detection of bcl-2 and p53 Proteins Immunohistochemical detection of bcl-2 was first described as most likely to occur in non-Hodgkin’s lymphomas with the 14;18 translocation and bcl-2 was believed not to be detectable in normal tissues [16]. Successive studies have demonstrated that this is not the case as bcl-2 is expressed in normal lymphoid tissue and lymphomas without the 14;18 translocation [17, 18]. Subsequently, Hockenbery et al. [19] showed that bcl-2 is present also in other tissues and is associated with cells that are protected from apoptosis. Therefore in each type of tissue a peculiar pattern of expression is detectable and this must be kept in mind when tumours originating from each tissue are stained for bcl-2. The story of the p53 gene and its protein is even more intriguing. Originally believed to be detectable only in tumours with a mutated gene [20], we have now established that p53 protein, like bcl-2, is detectable in a variety of normal tissues [21]. p53 is present in two localizations: cytoplasmic, widely detectable, and nuclear, in a few cells. These two localizations reflect at least two immunological forms [22]. p53 is, to the best of our knowledge, at present the only known protein for which the histopathologist needs two or more different antibodies in order to detect it. 13.2.2 bcl-2 Protein Expression in Normal Tissue Lymphoid System Small lymphocytes show intense cytoplasmic staining for bcl-2 both in peripheral blood and in secondary lymphatic organs. In these latter both the interfollicular areas and the mantle zone are positive. Characteristic is the lack of bcl-2 expression in most of the germinal centre cells. In the thymus cortical thymocytes are mostly negative whereas the more mature thymocytes present in the medulla are positive. In normal myeloid cells bcl-2 is expressed in bone marrow precursors but is absent from mature granulocytes [19, 23]. Epithelia In complex epithelia the expression of bcl-2 is limited to long-lived cells: basal cells and melanocytes in skin and respiratory epithelium, cryptic proliferating cells in the small and large intestine. In pancreas both
236
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
exocrine and endocrine glands appear to be positive [19, 24]. In epithelia regulated by hormones bcl-2 expression differs depending on the organ. In prostate, bcl-2 is expressed in the basal androgen-independent cells whereas the columnar superficial cells are negative [19, 25, 26] while in breast bcl-2 is expressed in epithelial cells of lobules and small and large ducts. The staining is heterogeneous and confined to the luminal cells [19, 27, 28]. However, Leek et al. have described a few cases in which normal epithelium appears negative. Central Nervous System Adult neurones are bcl-2 positive with staining present in the cytoplasm surrounding the nuclei whereas the axonal tract is negative [19]. 13.2.3 p53 Expression in Normal Tissues Understanding p53 protein function in normal and pathological conditions is influenced by the common observation that it cannot be demonstrated by immunostaining in normal tissues [20]. We have recently demonstrated [21] that wild-type p53 can be detected in most lymphoid and epithelial human cells in a cytoplasmic/ perinuclear localization. In the same tissues a few scattered cells shown nuclear staining. One can distinguish between these forms by using two anti-p53 antibodies: PAb248 for the cytoplasmic/ perinuclear and PAb240 for the nuclear form. Lymphoid Tissue p53 is not detectable both in its cytoplasmic and nuclear forms in peripheral blood small lymphocytes by immunostaining, but all the granulocytes show cytoplasmic/ perinuclear staining and a few of them also have nuclear positivity. Reactive tonsil, lymph node and spleen show cytoplasmic positivity in germinal centre, mantle zone and interfollicular areas. Only a few cells with nuclear staining are present and they are mostly in the germinal centres and, more rarely, in the interfollicular area. All the thymocytes show cytoplasmic staining whereas nuclear staining is present only in a few cells, mostly in the cortical area [21]. Epithelia We have demonstrated that p53 is detectable in prostatic and respiratory epithelia. Basal proliferating cells are negative whereas the upper layers of well differentiated cells show cytoplasmic/perinuclear p53. Gastric, duodenal and colonic mucosa cells are mostly positive [21]. Table 13.1 bcl-2 and p53 expression in normal human tissues Lymphoid tissue Peripheral blood lymphocytes Lymph node, tonsil and spleen paracortical area mantle zone
bcl-2 p53 cytoplasmic staining (PAb 248)
p53 nuclear staining (PAb 240)
pos
neg
pos
pos pos
pos pos
pos (a few) neg
A PREDICTIVE FACTOR FOR CANCER THERAPY
germinal centres Thymus cortex medulla Myeloid cells Neutrophils Epithelia Colon cryptic others Respiratory epithelium basal columnar Prostate basal columnar Liver (hepatocytes)
bcl-2 p53 cytoplasmic staining (PAb 248)
p53 nuclear staining (PAb 240)
neg
pos
pos (a few)
neg pos
pos pos
pos (a few) neg
neg
pos (all)
pos (a few)
pos neg
* pos
* neg
pos neg
neg pos
neg neg
237
pos neg neg neg pos neg neg neg neg * Relationship between bcl-2 and p53 expression in cryptic cells has not yet been established.
Figure 13.1 Tonsil, reactive follicular hyperplasia: mantle zone lymphocytes (lower right corner) are positive for bcl-2 and cytoplasmic p53 but do not show nuclear p53. In most of the germinal centre cells bcl-2 is no longer expressed; they remain positive for cytoplasmic p53 and a few now show nuclear p53. The pictures of p53 staining are reproduced, with the permission of the BMJ Publishing Group, from Figures 3d and 3e in reference 21.
13.2.4 Relationship Between bcl-2 and p53 Expression in Normal Tissues Comparison of staining in a limited number of normal tissues [19, 21, 24] has demonstrated that p53 and bcl-2 expression occurs with a striking degree of reciprocity (Table 13.1). In general, bcl-2 is associated with long-lived cells while p53 We originally found an inverse relationship between bcl-2 and p53 expression in non-Hodgkin’s lymphomas [31]. At that time the reported ability of wild-type p53 to induce apoptosis [13] led us to suggest that accumulation of inactivated nuclear p53 in bcl-2-negative follicular and diffuse lymphomas might lead to the same final effect, i.e. inhibition of apoptosis. Such an inverse relationship between p53
238
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Figure 13.2 Bronchial mucosa: basal cells are positive for bcl-2 and negative for p53, the columnar cells are bcl-2negative but positive for cytoplasmic p53. The picture of p53 staining is reproduced, with the permission of the BMJ Publishing Group, from Figure 4b in reference 21.
and bcl-2 has been subsequently reported also in breast carcinomas [27, 18, 19] and thyroid carcinomas [32]. Now it is possible to understand these immunohistochemical observations since recent work in vitro has established that p53 is a regulator of bcl-2 and its associated protein, bax [33, 34]. 13.3 bcl-2 and p53 Expression in Tumours and Prediction of Survival or Response to Therapy 13.3.1 Haematopoietic Malignancies A number of studies looking at the relationship between bcl-2 expression and resistance to drugs has been performed mainly on cell lines. Myshita et al. [35] investigated the pre-B-cell leukaemia cell line 697, which is highly sensitive to dexamethasone and to a variety of drugs commonly used in the treatment of NHLs. After transfection with bcl-2 resulting in a 20 fold increase of protein level, the line became highly resistant to killing by dexamethasone, which was nevertheless still able to suppress the proliferation. A similar effect was also found with a variety of drugs (VP 16, vincristine, hydroperoxycyclophosphamide, methotrexate). Additional investigations by the same authors [36] have confirmed the same data on murine lymphoma cell lines. Furthermore they have shown that the opposite it is also true: i.e. inhibition of bcl-2 synthesis restores sensitivity to a drug. The resistance to methotrexate of the cell line SU-DHL-4, bearing the 14;18 translocation and hence high levels of bcl-2, was almost completely abolished by treatment with bcl-2
A PREDICTIVE FACTOR FOR CANCER THERAPY
239
antisense oligomers. The reduction of viability obtained is greater than the sum of the reduction with methotrexate or antisense alone, indicating a synergistic effect. Also, in five myeloid cell lines an association between bcl-2 expression and resistance to a series of drugs (adriamycin, cytosin arabinoside and methotrexate) [37] has been reported. The authors noticed that mutant p53 genes cooperate with c-myc in the suppression of apoptosis. However, B chronic lymphocytic leukaemic cells from eight patients appear to be more sensitive to apoptosis than lymphocytes from normal controls when treated with cyclohexamide, a protein-synthesis inhibitor [38]. The authors have not tested the cases investigated for bcl-2 expression, but others [39] have recently demonstrated that 95% of this lymphoproliferative disease has bcl-2 levels higher than PBL, and 70% has levels equal or greater than those found in a t(14;18)-positive lymphoma cell line. These data indicate that bcl-2-positive malignancies are not necessarily resistant to apoptosis-induced therapy. B-CLL cells seem to become even more sensitive than normal small lymphocytes to induction of apoptosis, at least with some drugs. This finding underlines the need to investigate drug resistance or sensitivity by looking at more than just one protein involved in apoptosis regulation. Different results have been obtained by Campos et al. [40] who found in a series of 82 patients with AML results comparable to those obtained ‘in vitro’ [37]. Treatment was variable according to age but all the patients received an anthracycline drug or mitoxantrone and cytosine-arabinoside. The number of patients achieving complete remission was significantly higher among bcl-2-negative patients. Also, the overall survival was better for patients with leukaemias with less than 20% of bcl-2-positive cells. On the other hand, cells from leukaemias with more than 20% of bcl-2-positive cells showed greater viability in liquid culture. Retrospective studies on series of non-Hodgkin’s lymphomas have failed to produce a result comparable to that suggested by ‘in vitro’ studies [35]. It has been controversial whether the presence of the 14;18 translocation and/or bcl-2 protein expression has any prognostic significance in follicular non-Hodgkin’s lymphomas [41, 42]. In a study, on a series of 70 patients with follicular lymphomas, we have exploited the possibility of detecting both the 14;18 translocation, by the polymerase chain reaction, and protein expression, using our anti-bcl-2 monoclonal antibody, in paraffin embedded material. We used archival lymph node biopsies up to 30 years old and found no evidence of correlation between patient survival and either the translocation or the expression of the protein in the tumour [43]. We obtained similar results in a series of 119 B-cell lymphomas, where bcl-2 expression alone is not related to overall survival. Poorer prognosis is associated with p53 nuclear expression and, interestingly enough, the subgroup expressing both nuclear p53 and bcl-2 appears to have the worst outcome [44]. Therefore, on clinical grounds, bcl-2 expression is apparently not predictive of the outcome of currently used protocols in low and high-grade lymphoma treatment. However, these studies were not designed to investigate the role of bcl-2 as a predictive factor as the patients were treated with standard protocols in different centres and often with variations depending on individual needs. Furthermore, on biological grounds, ‘in vivo’ studies cannot be easily compared with ‘in vitro’ data. In the latter a single cell line is tested against a single agent. Instead, a combination of drugs and radiotherapy is given to patients. Furthermore, a tumour is composed of different subclones which can have different biological properties and can also acquire mutations during treatment.
240
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
13.3.2 Breast Carcinoma and bcl-2 In breast carcinoma a clear-cut association has been demonstrated between bcl-2 and estrogen receptor positivity [27, 29] while estrogen receptor negative cases are usually p53 positive. Therefore, two major groups of breast carcinomas emerge: tamoxifen-sensitive, ER, PR, bcl-2-positive and estrogen-independent, EGFR and p53-positive. However, two smaller groups are present, respectively, ER-positive bcl-2-negative and vice versa. Study of these cases should provide more information on the relationship between hormone stimulation and bcl-2 expression. To date, two preliminary studies have been performed to assess the potential predictive value of bcl-2 expression in human node-positive breast cancer patients treated with adjuvant treatments. Nathan et al. [45] determined bcl-2 expression with immunocytochemistry in a series of 107 breast cancer specimens of node positive patients enrolled in the Ludwig Breast Cancer Studies I–IV. bcl-2 was associated with both estrogen and progesterone receptors. Overall, 70% of the tumours expressed bcl-2 protein and patients with high levels of bcl-2 tended to have a better relapse-free and overall survival in univariate analysis. Since these patients were enrolled in trials of adjuvant therapy, these results raise the possibility that bcl-2 expression may be predictive of the efficacy of some adjuvant treatments. A second study by Gasparini et al. [28] analysed bcl-2 protein expression in two series of node-positive breast cancer patients treated either with adjuvant tamoxifen (81 cases) or adjuvant chemotherapy (CMF schedule) (99 cases), bcl-2 protein was expressed in 65% of the cases and was significantly associated with both estrogen and progesterone receptors and inversely correlated with p53 in both groups of patients. In univariate analysis for relapse free survival after a median follow-up of 5 years, patients with bcl-2-positive tumours had a significantly better outcome than the bcl-2-negative in both groups. As far as overall survival is concerned, patients with bcl-2-positive tumours had a better outcome in the group treated with adjuvant chemotherapy. In multivariate analysis in the group treated with tamoxifen those patients with lack of expression of estrogen receptor and bcl-2-negative tumours and those with more than three nodes involved had a poorer relapse free survival. In the CMF treated group, bcl-2 negativity was significant for poor overall survival as well as the number of involved nodes and tumour size. Thus, in contrast with in vitro data on drug resistance in haematopoietic lines [36] bcl-2 expression in human node-positive breast cancer patients was associated with a better outcome in patients treated either with adjuvant chemotherapy or hormone therapy. Overall, the results of the study by Gasparini et al. [28] suggest that expression of bcl-2 protein and the number of metastatic axillary nodes are independent significant indicators in patients with node-positive breast cancer, irrespective of the type of adjuvant treatment administered. These findings need to be confirmed in larger series of cases to establish whether or not bcl-2 expression can be used to select those node-positive patients who are more likely to respond to adjuvant treatments. 13.3.3 Prostate and bcl-2 Studies in animals have demonstrated that human androgen-sensitive prostate carcinoma cell lines undergo apoptosis following androgen ablation [46]. To investigate the involvement of bcl-2 in prostate cancer McDonnell et al. [25] have carried out a study looking at the expression of bcl-2 in human tumours and variations of bcl-2 mRNA synthesis in rat in relationship to androgen presence. In the first part of their study, the authors showed that in most of the cases of androgen-dependent prostatic carcinoma bcl-2 protein is absent and in the remaining is weakly expressed, sometimes just in a few cells. On the contrary all the
A PREDICTIVE FACTOR FOR CANCER THERAPY
241
androgen independent carcinomas were strongly positive for bcl-2 with the exception of some bone marrow metastases. To establish whether bcl-2 expression is due to lack of androgen stimulation and whether androgen can abrogate bcl-2 synthesis, two cohorts of rats were studied. In the first group an increase in prostate bcl-2 mRNA was shown after castration. In the second this effect was prevented by administration of testosterone following the castration. These data have been confirmed by Colombel et al. [26]. In primary localized cancers, staining for bcl-2 was heterogeneous, ranging from negative to strongly positive cases, while in a small series of hormone-resistant tumours all cases were positive. Residual neoplastic cells, after luteinizing-hormone releasing-hormone treatment protocol, were all bcl-2-positive. Metastatic tumours, also in bone, have been found to have a cell population mostly positive for bcl-2. With the exception of the discordance about the bone metastasis, found bcl-2-negative in the former study and positive in the latter, all these studies clearly indicate that bcl-2 is associated with more aggressive, androgen-independent prostate carcinoma. 13.3.4 p53 and Drugs Sensitivity in a Murine Cell Line Lowe et al. [47] have demonstrated that in opposition to the effect of bcl-2, transfection of wild-type p53 into a murine fibroblastic cell line restores sensitivity to treatment with both radiation and drugs (5fluoruracil, etoposide and adriamycin) inducing apoptosis. Transfection with either mutant or inactive p53 fails to achieve the same result. The presence of wild-type p53 turned out to be crucial to obtain a dramatic loss of viability. These data further point out the need to assess more than a single gene in order to predict sensitivity or loss of response to induction of apoptosis. 13.4 Conclusion Investigation of the expression of proteins regulating apoptosis as a predictive factor for cancer treatment is at its beginning. However three main points are already clear: 1. The expression of each of the proteins regulating apoptosis is likely to have a different significance in different organs, as recently pointed out for bcl-2 [48]. Indeed it has already been demonstrated that this protein has a predictive meaning in myeloid leukaemias [40] and breast cancer [28]. 2. It will be necessary to investigate not only the primary tumour, but, when possible, also the posttreatment residual tumour and the metastasis. This approach will allow us to understand how neoplastic cell populations with different biological markers, respond to treatment. This type of information could possibly lead us to choose both the best treatment for resistant clones and to identify in the primary tumours potentially resistant cells already present and treat them immediately in an appropriate way. 3. It is likely that, in most types of tumour, to predict the sensitivity to treatment it will be necessary to assess the expression of more than one protein, possibly belonging to different pathways involved in both suppression and induction of apoptosis. The bcl-2-p53 system is the first, even if incomplete, model available for such studies. Acknowledgements The authors thank Beppe Viale for critically reviewing the manuscript.
242
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
References 1
2
3 4 5 6
7
8 9 10
11
12
13 14 15 16
17
18
. TSUJIMOTO, Y., FINGER, L.R., YUNIS, J., NOWELL, P.C. & CROCE, C.M. (1984) Cloning of the chromosome breakpoint of neoplastic B cells with the t(14;18) chromosome translocation. Science 226, 1097–1099. . CLEARY, M.L. & SKLAR, J. (1985) Nucleotide sequence of a t(14;18) chromosomal breakpoint in follicular lymphoma and demonstration of a breakpoint cluster region near a transcriptionally active locus on chromosome 18. Proc. Natl Acad. Sci. USA 82, 7439– 7443. . HOCKENBERY, D., NUNEZ, G., MlLLIMAN, C., SCHREIBER, R.D. & KORSMEYER, S.J. (1990) Bcl-2 is an inner mitochondrial membrane protein that blocks programmed cell death. Nature 348, 334–336. . KORSMEYER, S. (1992) Bcl-2 initiates a new category of oncogenes: regulators of cell death. Blood 80, 879–886. . BUTTY AN, R., OLSSON, C.A., PIN TAR, J. CHANG,C., BANDYK, M., NG, P-Y. & SAWCZUK, I.S. (1989) Induction of the TRPM-2 gene in cells undergoing programmed cell death. Mol. Cell. Biol. 9, 3473–3481. . WATNABE-FIKUNAGA, R., BRANNAN, C. L, COPELAND, N.G., JENKINS, N.A. & NAGATA, S. (1992) Lymphoprolifertive disorders in mice explained by defects in Fas antigen that mediates apoptosis. Nature 356, 314–317. . EVAN, G.I., WYLLIE, A.H., GILBERT, C.S., LITTLEWOOD, T.D., LAND, H., BROOKS, M., WATERS, C.M., PEN, L.Z. & HANCOCK, D.C. (1992) Induction of apoptosis in fibroblasts by -c-myc protein. Cell 69, 119–128. . HENGARTNER, M.O., ELLIS, R.E. & HORVITZ, H.R. (1992) Caenorhabditis elegans gene ced-9 protects cells from programmed cell death. Nature 356, 494–499. . OLTVAI, Z.N., MILLIMAN, C.L. & KORSMEYER, S. (1993) Bcl-2 heterodimerizes in vivo with a conserved homolog, Bax, that accelerates programmed cell death. Cell 74, 609–619. . KOZOPAS, K.M., YANG, T., BUCHAN, H.L., ZHOU, P. & CRAIG, R.W. (1993) MCL1, a gene expressed in programmed myeloid cell differentiation, has sequence similarity to bcl-2. Proc. Natl Acad. Sci USA 90, 3516–3520. . BOISE, L.H., GONZALES-GARCIA, M., POSTEMA, C.E., DING, L., LINDSTEN, T., TURKA, L.A., MAO, X., NUNEZ, G. & THOMPSON, C.B. (1993) bcl-x, a bcl-2-related gene that functions as a dominant regulator of apoptotic cell death. Cell 74, 597– 608. . EL-DEIRY, W.S., HARPER, W., O’CONNOR, P.M., VELCILESCU, V.E., CANMAN, C.E., JACKMAN, J., PIETENPOL, J.A., BURRELL, M., HILL, D.E., WANG, Y., WIEMAN, K.G., MERCER, W.E., KASTAN, M.B., KOHN, K.W., ELLEDGE, S.J., KINZLER, K.W. & VOGELSTEIN, B. (1994) WAF1/CIP1 is induced in p52mediated G1 arrest and apoptosis. Cancer Res. 54, 1169–1174. . YONIS-ROUACH, E., RESNITZKY, D., LOTEM, J., SACHS, L., KIMCHI, A. & OREN, M. (1991) Wildtype p53 induces apoptosis of myeloid leukaemic cells that is inhibited by interleukin-6. Nature 352, 345–347. . SHAW, P., BOVEY, R., TARDY, S., SAHALI, R., SORDAT, B. & COSTA, J. (1992) Induction of apoptosis by wild-type p53 in a human colon tumor-derived cell line. Proc. Natl Acad. Sci. USA 89, 4495–4499. . OLINER, J.D., KINZLER, K.W., MELTZER, P.S., GEORGE, D.L. & VOGELSTEIN, B. (1992) Amplification of a gene encoding a p53-associated protein in human sarcomas. Nature 358, 80–83. . NGAN, B.Y., CHEN-LEVY, Z., WEISS, L.M., WARNKE, R.A. & CLEARY, M.L. (1988) Expression in nonHodgkin’s lymphoma of the Bcl-2 protein associated with the t(14;18) translocation. New Engl. J. Med. 318, 1638–1644. . PEZZELLA, F., TSE, A.G.D., CORDELL, J.L., PULFORD, K.A.F., GATTER, K.C. & MASON, D.Y. (1990) Expression of the bcl-2 protein is not specific for the 14;18 chromosomal translocation. Am. J. Path. 137, 225–232. . PEZZELLA, F., GATTER, K.C., MASON, D.Y., BASTARD, C., DUVAL, C., KRAJEVSKI, A., TURNER, G.E., ROSS, F.M., CLARK, H., JONES, D.B., LEROUX, D. & LE MARC’HADOUR, F. (1990) Bcl-2 protein is expressed in follicular lymphomas in the absence of the 14;18 translocation. Lancet 336, 1510–1511.
A PREDICTIVE FACTOR FOR CANCER THERAPY
19
20 21
22 23
24
25
26
27
28
29
30 31
32 33
34 35 36
243
. HOCKENBERY, D.M., ZUTTER, M., HICKEY, W., NAHM, M. & KORSMEYER, S. (1991) Bcl-2 protein is topographically restricted in tissues characterized by apoptotic cell death. Proc. Natl Acad. Sci. USA 88, 6961–6965. . RODRIGUES, N.R., ROWAN, A., SMITH, M.E. F., KERR, I.B., BODMER, W.F., GANNON J.V. & LANE, D.P. (1990) p53 mutations in colorectal cancer. Proc. Natl Acad. Sci. USA 87, 7555–7559. . PEZZELLA, F., MICKLEM, K., TURLEY, H., PULFORD, K., JONES, M., KOCIALKOWSKI, S., DELIA, D., AIELLO, A., BICKNELL, R., SMITH, K., HARRIS, A., GATTER K.C. & MASON, D.Y. (1994) An antibody for detecting p53 protein by immunohistochemistry in normal tissues. J. Clin. Path. 47, 592–596. . MILNER, J. (1984) Different forms of p53 detected by monoclonal antibodies in nondividing lymphoctes. Nature 310, 143–145. . DELIA, D., AIELLO, A., SOLIGO, D., FONTANELLA, E., MELANI, C., PEZZELLA, F., PIEROTTI, M.A. & DELLA PORTA, G. (1992) Bcl-2 proto-oncogene expression in normal and neoplastic human myeloid cells. Blood 79, 1291–1298. . PEZZELLA, F., TURLEY, H., KUZU, I., TUNGEKAR, M.F., DUNNILL, M.S., PIERCE, C.B., HARRIS, A., GATTER, K.C. & MASON, D.Y. (1993) Bcl-2 protein in non-small-cell lung carcinoma. Immunohistochemical evidence for abnormal expression and correlation with survival. N. Engl. J. Med. 29, 690–694. . MCDONNELL,T. J., TRONCOSO, P., BRISBAY, S.M., LONGOTHETIS, C., CHUNG, L.W., HSIEH, J.T., TU, S.M. & CAMPBELL, M.L. (1992) Expression of the protooncogene bcl-2 in the prostate and its association with emergence of an androgen-independent prostate cancer. Cancer Res. 52, 6940–6944. . COLOMBEL, M., SYMMANS, F., GIL, S., O’TOOLE, K.M., CHOPIN, D., BENSON, M., OLSSON, C.A., KORSMEYER, S. & BUTTYAN, R. (1993) Detection of the apoptosis-suppressing oncoprotein bcl-2 in hormonerefractory human prostate cancers. Am. J. Pathol. 143, 390–400. . DOGLIONI, C., DEI TOS, B., LAURINO, L., CHIARELLI, C., BARBARESCHI, M. & VIALE, G. (1994) The prevalence of bcl-2 immunoreactivity in breast carcinomas and its clinicopathological correlates, with particular reference to oestrogen receptor status. Virch. Arch. 424, 47–51. . GASPERINI, G., BARBARESCHI, M., DOGLIONI, C., DALLA PALMA, P., MAURI, F.A., MALUTA, S., BORACCHI, P., BEVILACQUA, P., CAFFO, O., MORELLI, L., VERDERIO, P., PEZZELLA, F. & HARRIS, A.L. (1995) Expression of bcl-2 protein predicts efficacy of adjuvant treatments in operable node-positive breast cancer. Clinical Cancer Res. 1, 189–198. . LEEK, R.D., KAKLAMANIS, L., PEZZELLA, F., GATTER, K.C. & HARRIS, A.L. (1994) bcl-2 in normal human breast and carcinoma, association with oestrogen receptor positive, epidermal growth factor receptornegative tumours and in situ cancer. Br. J. Cancer. 6, 135–139. . IGGO, R., GATTER, K., BARTEK, J., LANE, D. & HARRIS, A.L. (1990) Increased expression of mutant forms of p53 oncogene in primary lung cancer. Lancet 335, 675– 679. . PEZZELLA, F., MORRISON, H., JONES, M., GATTER, K.C., LANE, D., HARRIS, A.L. & MASON, D.Y. (1993) Immunohistochemical detection of p53 and bcl-2 protein in non-Hodgkin’s lymphoma. Histopathology 22, 39–44. . PILOTTI, S., COLLINI, P., RILKE, F., CATTORETTI, G., DEL BO, R. & PIEROTTI, M. (1994) Bcl-2 protein expression in carcinomas originating from the follicular epithelium of the thyroid gland. J. Pathol. 172, 337–342. . MIYASHITA, T., KRAJEWSKI, S., KRAJEWSKA, M., WANG, H.G., LIN, H.K., LIEBERMANN, D.A., HOFFMAN, B., REED & J.C. (1994) Tumour suppressor p53 is a regulator of bcl-2 and bax gene expression in vitro and in vivo. Oncogene 9, 1799–1805. . HALDAR, S., NEGRINI, M., MONNE, M., SABBIONI, S. & CROCE, C.M. (1994) Down-regulation of bcl-2 by p53 in breast cancer cells. Cancer. Res. 54, 2095–2097. . MYASHITA, T. & REED, J. (1993) Bcl-2 oncoprotein blocks chemotherapy-induced apoptosis in a human leukaemia cell line. Blood 81, 151–157. REED, J.C., KITADA, S., TAKAYAMA, S. & MIYASHITA, T. (1994) Regulation of chemoresistance by the bcl-2 oncoprotein in non-Hodgkin’s lymphoma and lymphocytic cell lines. Ann. Oncol. 5(Suppl. 1), S61–S65.
244
37
38.
39 40
41 42 43
44
45
46 47 48
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. LOTEM, J. & SACHS, L. (1993) Regulation by bcl-2, c-myc and p53 of susceptibility to induction of apoptosis by heat schok and cancer chemotherapy compounds in differentiation competent and defective myeloid leukemic cells. Cell. Growth. Differ. 4, 41–47. COLLINS, R.J., HARMON, B.V., SOUVLIS, T., POPE, J.H., KERR, J.F.R. (1991) Effects of cyclohexamide on B-chronic lymphocytic leukaemic and normal lymphocytes in vitro: induction of apoptosis. Br. J. Cancer. 64, 518–522. . HANADA, M., DELIA, D., AIELLO, A., STADMAUER, E. & REED, J. (1993) bcl-2 gene hypomethilation and high-level expression in B-cell Chronic Lymphocytic leukemia. Blood 82, 1820–1828. . CAMPOS, L., ROUAULT, J-P., SABIDO, O., ORIOL, P., ROUBI, N., VASSELON, C. ARCHIMBAUD, E., MAGAUD, J-P. & GUYOTAT, D. (1993) High expression of bcl-2 protein in acute myeloid leukaemia cells is associated with poor response to chemotherapy. Blood 81, 3091–3096. . YUNIS, J.J., MAYER, M.G., ARNESEN, M.A., AEPPLI, D.P., OKEN. M.M. & FRIZZERA, G. (1989) Bcl-2 and other genomic alterations in the prognosis of large cell lymphomas. N. Engl. J. Med. 320, 1047–1054. . LEVINE, E.G., ARTHUR, D.C., FRIZZERA, G., PETERSON, B.A., HURD, D.D. & BLOOMFIELD, C.D. (1988) Cytogenetic abnormalities predict outcome in non-Hodgkin’s lymphoma. Ann. Int. Med. 108, 14–20. . PEZZELLA, F., JONES, M., RALFKIAER, E., ERSBOLL, J. , GATTER, K.C. & MASON, D.Y. (1992) Evaluation of bcl-2 protein expression and 14;18 translocation prognostic markers in follicular lymphoma. Br. J. Cancer 65, 87–89. . PIRIS, M.A., PEZZELLA, F., MARTINEZ-MONTERO, J.C., ORRADRE, J.L., VILLUENDAS, R., SANCHEZ-BEATO, M., CUENA, R., CRUZ, M.A., MARTINEZ, B., GARRIDO, M.C., GATTER, K., AIELLO, A., DELIA, D., GIARDINI, R. & RILKE, F. (1994) P53 and bcl-2 expression in high grade B-cell lymphomas, correlation with survival time. Br. J. Cancer 69, 337–341. . NATHAN, B., GUSTERSON, B., JADAYEL, D., O’HARE, M., ANBAZHAGAN, R., JAYATILAKE, H., EBBS, S., MICKLEM, K., PRICE, K., GELBER, R., REED, R., SENN, H-J., GOLDHIRSCH, A. & DYER, M.J.S. (1994) Expression of bcl-2 in primary breast cancer and its correlation with tumour phenotype. Ann. Oncol. 5, 409–414. . KYPRIANOU, N., ENGLISH, H.F. & ISAACS, J.T. (1990) Programmed cell death during regression of PC-82 human prostate cancer following androgen ablation. Cancer Res. 50, 3748–3753. . LOWE, S.W., RULEY, H.E., JACKS, T. & HOUSMAN, D.E. (1993) p53-dependent apoptosis modulates the cytotoxicity of anticancer agents. Cell 74, 957–967. . HARRIS, A.L. (1994) What does Bcl-2 mean in solid tumours—friend or foe? Ann. Oncol. 5, 388–390.
14 Control of Cell Death in the NervousSystem HANNES C.A.DREXLER1 and JUNYING YUAN2 1
2
Max Planck Institut für klinische und physiologische Forschung, Abt Molekulare Zellbiologie, Parkstr. 1, D-61231 Bad Nauheim, Germany
Cardiovascular Research Center, Massachusetts General Hospital-East, 149 13th St., Charlestown, MA 02129, USA; Department of Medicine, Harvard Medical School, Boston, MA 02115, USA
14.1 Introduction The phenomenon of cell death during development was first described systematically in a review by Glucksmann et al. over 40 years ago (1950) but it was largely ignored with the exception of a small group of neuronal embryologists (for review see [1]). Cell death during neuronal development of both vertebrate and invertebrate animals is dramatic both in its extensive distribution and its proportions. It has been described in detail for various locations in the CNS as well as the PNS of the vertebrate embryo. (For an extensive review on the types of neural cell undergoing developmentally regulated cell death, as well as a discussion of the role of neurotrophins, see Jacobson M., Neuronal death and neurotrophic factors, in: Developmental Neurobiology, 3rd edition, Plenum Press, New York.) In some areas, up to 50% or more of the initially generated neurons die. The time of death usually coincides with synaptogenesis and this is generally believed to be a consequence of a limited amount of neurotrophin supply, for which a surplus number of neurons compete. This competition between supernumerary neurons mainly serves two purposes: (1) it ensures that all target cells are innervated, and (2) it matches the number of neurons with the target size. The death of neurons not only plays a pivotal role during embryogenesis when a functional nervous system is to be established but also occurs in the adult, preferentially under various pathological conditions which almost always are accompanied by devastating consequences for the patients. Examples include Alzheimer’s disease, amyotrophic lateral sclerosis and other types of neuronal injury inflicted by ischemia, hypoglycemia or excitotoxic agents. While it is not yet clear whether these latter types of neurodegenerative events also trigger the execution of an intrinsic genetic program leading to the death of certain neuron populations, there is mounting evidence for the hypothesis that, although cell death is no uniform process, and may phenotypically vary between different cell types and cell death inducers, parts of the same celluar machinery as in developmental cell death may be implicated in neurodegenerative events as well. The phenomenon of neuronal cell death has already been extensively reviewed (e.g. by [1, 2]). Hence, the aim of this work is to focus on more recent studies that have been carried out to elucidate the molecular mechanisms that are underlying the neuronal cell death process. Direct experimental evidence for the involvement of particular molecular components are scarce in many instances of neuronal cell death, but
246
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
observations made for certain molecules in other cell death model systems may be helpful to reveal at least the direction of future investigations on cell death mechanisms in the nervous system. Due to the exponential growth currently observed in apoptosis research and to the limited scope of this review however, omissions had to be made and we apologize for not including many of the already ‘classical’ research articles on apoptosis. 14.2 Neural Trophic Factors and their Receptors Numerous studies of cell death in the nervous system have been carried out to address the role targetderived trophic factors play in neuronal survival. The neurotrophic factor theory has provided us during the past few decades with the conceptual framework to understand better the survival and death of neurons. It is based on studies which demonstrated that the survival of developing neurons is dependent on the target tissue size, on the trophic factors these target cells produce and the extent of synaptic contact formation of a neuron with a particular target cell (reviewed by [3]). Surplus neurons are eliminated as there is limited supply of neurotrophins and neurons that fail to establish contact with their proper target do not survive as they do not receive the trophic support required for their survival. Recent studies however have shed some doubts on this simplistic form of the neurotrophic factor concept. There is substantial evidence that neurons may derive trophic support not only from the target cell they innervate, but may also depend on certain afferent input signals for survival [4]. In addition, it has become quite clear that a particular neurotrophin is capable of reacting not only with its high-affinity receptor but also with other neurotrophin receptors as well (reviewed by [5]) and that its effect is dependent on the integration of a multitude of variables to which a particular neuron is exposed. These caveats should be kept in mind during the following discussion and should make us aware of the possibility that ‘nature has opted for a fuzzy strategy’ [3] when it comes to decide which cell should live and which cells should die. The family of neurotrophins consists of four closely related polypeptides: the archetype neurotrophin, nerve growth factor (NGF), which was discovered some 40 years ago on the basis of its capacity to increase survival and neurite outgrowth of sensory and sympathetic neurons, and three homologous polypeptides that were identified only very much later: brain-derived neurotrophic factor (BDNF), neurotrophin-3 (NT-3) and neurotrophin-4/5 (NT-4/5). Depending on the pattern of neurotrophin receptor expression, all of these trophic factors promote the survival of particular subsets of neurons. A great deal of knowledge about the physiological role neurotrophins and their receptors play stems from germline-targeted knockout mice. NGF knockout mice have been generated which display complete loss of sympathetic neurons as well as of sensory neurons responsible for thermo- and noci-reception [6]. NGF (−/−) mice are viable at birth but die soon thereafter within a four week period. Interestingly, the number of cholinergic neurons in the basal forebrain does not seem to be reduced, although there is a decrease in cholinacetyltransferase level in these neurons. This supports earlier experimental data suggesting that NGF may not be essential for the survival of these CNS neurons but rather mediate the maintenance of their cholinergic phenotype [7]. The mice that are homozygous mutant for the high-affinity NGF receptor trkA display equivalent defects in the peripheral nervous system: neuron numbers in the trigeminal ganglia as well as in the DRG and the superior cervical ganglion are markedly reduced. However, in contrast to NGF (−/−) mice the number of projections from basal forebrain cholinergic neurons to the hippocampus is significantly reduced in trkA (−/ −) mice. It should be noted that the observed reduction in the number of cholinergic fibers may not reflect a true reduction in cholinergic cell numbers but may as well be due to an incompetence in axon formation or maintenance of the cholinergic phenotype.
CONTROL OF CELL DEATH IN THE NERVOUS SYSTEM
247
Thus, it seems that in the CNS only part of the effects mediated by signal transduction via the TrkA receptor are due to the action of NGF and that cholinergic neurons may be able to derive their trophic support from other sources as well. In other words, the mechanisms of survival and fiber outgrowth may diverge and this functional separation occurs already at the level of the high-affinity Trk receptors. In the peripheral nervous system NT-3 or NT-4/5 at least do not seem to be able to function as NGF-substitutes: NT-3 and NT-4/ NT-5 are capable of activating the NGF receptor trkA as well, albeit with a considerably lower affinity (see below). Consequently, in NGF (−/−) mice which still have a functional trkA receptor, the number of sensory or sympathetic neurons should have been increased compared to trkA ( −/−) mice where signal transduction through the trkA receptor is completely abrogated. BDNF and NT4/5 are the two ligands that bind with high affinity to the TrkB receptor. Gene knock-out studies for the receptor as well as for BDNF have shown that neuron populations in the peripheral nervous system are preferentially affected. TrkB (−/−) mice that do not express a catalytically active receptor show defects in motoneurons especially in the facial motor nucleus and in lumbar segments of the spinal cord motorcolumn. There are also lesions in DRG neurons which are probably responsible for mechanoreception and in the trigeminal ganglion. Conversely, sympathetic neurons do not seem to be affected in trkB ( −/−) mice. This is also true for the neurons in the central nervous system although there is widespread expression of the trkB gene in the CNS [8], again pointing at the involvement of a certain functional redundancy of trophic support in the CNS. As for the mice that are homozygous mutant for BDNF, although BDNF does support the survival of facial motoneuron and spinal montoneurons after axotomy [9, 10] as well as the survival of cultured spinal cord motoneurons [11], there is surprisingly no loss of motoneurons observed in these locations. BDNF knock-out mice, however, display severe defects in sensory ganglia and DRG which is in good correlation with the trkB deficient mice. Because NT4/5 is also able to bind to and activate trkB receptor it is conceivable that NT4/5 has its primary function in motoneuron development and survival. NT4/5 knock-out mice have not been reported so far but certainly will provide further insight into the interaction of BDNF and NT4/5 with TrkB receptor. NT3 is the most promiscuous trophic factor of this neurotrophin family. Apart from binding to its primary high-affinity receptor trkC, it also binds to and activates trkA and trkB receptors, though with a significantly reduced affinity. Mice that are homozygously mutant for either NT3 [12] or the tyrosine kinase domain of trkC have been generated [13]. Both types of knock-out mice demonstrate a corresponding array of proprioception deficiencies. Histological analyses revealed that the number of DRG neurons involved in proprioception are reduced and IA afferent fibers projecting to the dorsal horn of the spinal cord are completely missing. Little to nothing is known about the physiological function of trkA or trkB receptor occupied by NT3. While the role of the trk receptors regarding the mediation of survival promoting signals in certain subset of neurons could be pinpointed by the knock-out mutants, the part of the p75 low affinity NGF receptor (p75LNGFR) which binds all neurotrophins equally well, is much less well defined. p75LNGFR (−/−) mutant mice develop quite normal with a decreased thermoreception as the most striking phenotype. Two major functions have been postulated for the p75LNGFR. (1) There is growing evidence that p75LNGFR possesses an intrinsic signalling capacity that is independent of the trk receptors. For instance, cell death of a neuronal cell line expressing p75LNGFR but not trkA could be blocked by NGF. It is interesting to note that p75LNGFR expression by itself was sufficient to increase the apoptosis rate in the absence of NGF or monoclonal antibody in the transfected cells [14]. Experiments carried out by Von Bartheld et al. demonstrate that p75LNGFR may be mediating the death of neurons in the isthmo-optic nucleus of the chick embryo [15]. A recent publication demonstrates that p75LNGFR mediated signal transduction can induce either cell death or survival, depending on the developmental stage of the receptor expressing cell. The authors show by using
248
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
antisense oligonucleotides that repression of p75LNGFR expression reduced the NGF-mediated survival in sensory neurons from 12- and 15-day-old mice but increased survival among sensory neurons from older embryonic or newborn animals [16]. (2) Alternatively, in analogy to the TGF-beta type III receptor or to cell surface bound proteoglycans for the fibroblast growth factors, p75LNGFR may represent an assistant receptor without signalling ability that is capturing ligands for the signal transducing master receptor. p75LNGFR may also be an integral part of a signal transducing receptor complex in which the activity of trk receptor is supported by the action of p75LNGFR [17, 18]. CNTF is a neurotrophic factor different from the neurotrophins mentioned above in so far as it lacks a classical signal peptide and does not appear to be released in significant amounts from producer cells via the ER and Golgi pathway [19]. In vitro CNTF acts as a pleiotrophin supporting the survival of a variety of neuronal cell types from the central and peripheral nervous system like sympathetic and parasympathtic neurons, spinal cord motor neurons and sensory neurons [20–23]. Furthermore, it enhances survival of oligodendrocytes and influences the differentiation of glial cells [24, 25]. The initial emphasis of CNTF as a classical neurotrophin has gradually shifted to the understanding that CNTF is a factor that exerts its trophic effect during nerve injury or lesion. Consistent with this notion is the lack of a classical signal peptide in the molecule. Axotomy experiments have shown that administration of CNTF in vivo is able to prevent the degeneration of the lesioned nerves [26]. In the wobbler mutant mice, which have a phenotype of progressive motoneuron dysfunction, CNTF synergizes with BDNF in blocking the degeneration process [27] and it could also slow down the degenerative process in pmn mutant mice (pmn=progressive motoneuronopathy). Disruption of the CNTF gene does not affect the number of motoneurons generated during embryonic development of the knock-out mice but rather caused a progressive atrophy and a functional loss of spinal motoneurons with increasing age of the CNTF (−/−) mice [20]. This is in contrast to earlier experiments carried out in chick embryos which have demonstrated that CNTF could rescue motoneurons in the lumbar motorcolumn from ontogenetic cell death [28, 29, 30]. Support for the notion that CNTF actually could play a role during embryonic development and, for example, accounts for the survival of a subset of spinal motoneurons, comes from a recent study indicating that CNTF is expressed in the head region of E11 rat embryos [31], although others have claimed that CNTF expression during embryogenesis is not detectable [19, 32, 26]. The potential for CNTF to act as a neurotrophin during embryogenesis is further complicated by the lack of a hydrophobic signal sequence. It is unclear at the moment how CNTF could be released from the astrocytes or the myelinating Schwann cells, which are the major producer cells. In vitro data using astrocyte conditioned medium or cells transfected with a CNTF expression vector at least do not show a detectable release of biologically active CNTF. Alternative secretion mechanisms as suggested for other cytokines like aFGF, bFGF or IL-1 which also lack a hydrophobic secretion signal, could be involved but further studies are necessary to settle this issue. The activation of the CNTF receptor occurs in two steps: (1) CNTF binds to the alpha subunit of the CNTF receptor; (2) the CNTF/alpha subunit complex then associates with two beta components, gp130 and leukemia inhibitor factor receptor (LIFR which dimerize and represent the signal transducing unit. As one of the substrates that is phosphorylated subsequently, p91 has been identified. Interestingly, p91 is rapidly translocated upon CNTF-induced phosphorylation into the nucleus of SK-N-MC neuroepithelioma cells and can activate transcription of a reporter gene which contains a CNTF responsive element (CNTF-RE) from cfos. Thus, neuronal genes that carry the CNTF-RE may be important mediators of the survival-promoting effect of CNTF [33]. There are also indications that the promotor of the rat superoxide dismutase (SOD), an enzyme implicated in the etiology of amyotrophic lateral sclerosis where spinal motoneurons are degenerating, contains a sequence similar to the CNTF-RE. Thus, the identification of genes that carry such
CONTROL OF CELL DEATH IN THE NERVOUS SYSTEM
249
CNTF-REs could help to obtain more detailed information about CNTF’s mode of action as well as a better general understanding of the cellular components that are part of the intracellular machinery of neuronal cell death suppression. The brain is one of the richest sources for fibroblast growth factor-1 and -2 (FGF-1 and FGF-2). Soon after they had been discovered, primarily because of their strong mitogenic activity on endothelial cells and their ability to stimulate angiogenesis, it became clear that the FGFs are, in addition, potent survival factors for primary cultures of hippocampal neurons and other types of neuron (reviewed by [34]). Although extensive studies have been carried out especially with FGF-1 and FGF-2 and although their survival enhancing activity in vitro is undoubted, the question whether these pleiotrophic growth factors act as ‘real’ neurotrophin under physiological conditions in vivo still remains a matter of controversy. Almost all models which describe an in vivo function of FGF are based on afflicting various forms of injuries or trauma followed by activation of endogenous FGF or subsequent application of exogenous FGF [34]. In this respect FGF resembles more CNTF which also appears to act primarily as an emergency neurotropohin or repair factor. The picture starts to become even more dilute by two other facts. (1) There are four receptor genes encoding high affintiy receptors for FGFs (FGFR-1 to FGFR-4) as well as multiple isoforms of these receptors. In addition, there are various low affinity receptors for FGF which are able not only to sequester ligand at the cell surface but also to present ligand to the high-affinity receptors. Our current knowledge about the distribution and specificity of these receptors/receptor complexes in the nervous system is incremental at best. (2) Virtually nothing is known about the relevance of the other members of the FGF family (FGF-3 to FGF-9) to neuronal survival. One exception is FGF-5 which has been claimed to function as a muscle-derived survival factor for spinal cord motoneurons in culture [35]. Another similarity between CNTF and FGF-l/FGF-2 is represented by the absence of a hydrophobic signal sequence in all three molecules. Consequently, none of these potent factors is released via the ERGolgi pathway. However, FGF can readily be found associated with the extracellular matrix and it therefore has been claimed that FGF 1/2 may be released through a non-classical secretion pathway [36]. This would be consistent again with the aforementioned model systems, in which the neurodegenerative effects induced by nerve lesion/chemical or physical trauma etc. can be amended by FGF, because in these situations FGF could readily be released from the intracellular stores. 14.3 The Role of the ICE Family of Genes in Controlling Neuronal Cell Death Since the beginning of the 1980s one particularly fruitful field of programmed cell death research has been a small nematode, Caenorhabditis elegans, which has been adopted by many developmental biologists to identify developmentally important genes. Of specific interest was the finding that, of the 1090 somatic cells of an adult hermaphrodite, exactly 131 cells are eliminated in a precisely defined and invariable spatial and temporal manner. Genetic analyses of mutations that effect various aspects of programmed cell death in C. elegans resulted in the identification of 14 genes that act in this process. Among these 14 genes, there are 3 genes for which a universal requirement in the control of all 131 cell deaths has been demonstrated [140, 141]. The product of the ced-9 gene acts as a suppressor of the cell death program and is structurally and functionally similar to that of the human protooncogene bcl-2 (see below). For the product of the ced-4 gene which, together with that of the ced-3 gene, is involved in the actual killing process, no functional homologue in vertebrates has been described yet [37]. Recent genetic evidence suggested that ced-4 acts upstream from ced-3 and may be a positive regulator of ced-3 (S. Shaham and H.R.Horvitz, personal communication).
250
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Like ced-4, the C. elegans gene ced-3 is also required in dying cells for programmed cell death to occur, suggesting that ced-3 is a component of the cell death machinery [37]. Molecular analysis of the ced-3 gene revealed that the product of ced-3 is homologous to the mammalian cysteine pro tease interleukin-1 converting enzyme (ICE) [38, 39]. Ced-3 and ICE share an overall identity of 28% on the amino acid level with a nearly identical pentapeptide covering the active site of ICE. ICE is involved in the processing of the 33 kDa IL-1 percursor molecule into the biologically active 17.5 kDa form. Overexpression of ICE in rat-1 fibroblasts induced death of the ICE expressing cells which displayed morphological features of apoptosis [40]. This type of cell death could be completely abrogated by overexpression of the crmA gene, a cowpox virus immediate early gene that has been described as specific inhibitor of ICE [41] as well as by overexpressing the human proto-oncogene bcl-2 which is known to suppress apoptosis in a variety of cell death model systems (see also discussion below). This result suggests that ICE has the ability to induce apoptosis, a necessary but not sufficient condition for a cell death gene. The following experiment provided stronger evidence that the genes in the ICE family are essential for cells to undergo programmed cell death in vertebrates. Dorsal root ganglion neurons from chick embryos subjected to withdrawal of nerve growth factor and serum from the culture medium rapidly undergo cell death within 3–4 days. Microinjection of an expression vector into these neurons directing crmA gene expression under the control of the chicken -actin promotor resulted in the survival of 80% of the neurons after 3 days (compared to less than 20% in the controls) and of 60% after 6 days of NGF removal (compared to less than 10% of the control population) [42]. Expression of bcl-2 in these neurons showed equivalent survival rates, and coexpression of crmA and bcl-2 did not reveal any synergism between both gene products. Overexpression of an ICE-lacZ fusion protein on the contrary resulted in the death of all fusion protein expressing cells whereas expression of an ICE mutant carrying a point mutation in the active site (Cys Gly) which drastically reduced the enzymatic activity resulted in the survival of at least 2–2.7 times as many neurons. These results imply (1) that ICE or a closely related molecule is necessary to mediate neuronal cell death upon neurotrophin withdrawal, (2) that the actual killing process requires the same enzymatic activity of ICE needed for the pre IL-1 processing, and (3) that ICE and bcl-2 are compounds acting within the same molecular pathway(s) to cell death as no synergistic or additive effect in the coexpression experiments could be observed. It is also rather questionable whether the mature biologically active 17 kDa IL-1 polypeptide per se has a direct cell death inducing activity [43]. The cysteine protease ICE could act on other as yet unidentified substrates besides pre-IL-1 . The association of these proteins with ICE would then lead to the further execution of the cell death program downstream of ICE. The potent ICE inhibitor CrmA exhibits some features of the serpin type class of serine protease inhibitors. It has been argued therefore by Komiyama et al. that it is the ‘substrate binding geometry, not the catalytic mechanism of a proteinase, that dictates its reactivity with protein inhibitors’ [44], and one might add, with other substates as well. In a recent publication, the Ich-1/nedd2 gene has been described as a molecule with extensive structural homology to ICE that, in addition, parallels ICE in its biological activity [45, 46]. Originally, nedd2 had been identified as one of a set of genes with unknown function that are downregulated during mouse brain development [47]. The gross amount of neuronal cell death occurs during embryonic development and it is therefore tempting to speculate that the Ich-1/nedd2 gene may be the one involved in these cell death events. The Ich-1/nedd2 gene encodes two alternatively spliced mRNAS: Ich-1L and Ich-1s [45]. The protein product of Ich-1L is 435 amino acids long and shares 29% identity on the amino acid level with both CED-3 and ICE. Overexpression of Ich-1L induced programmed cell death which can be inhibited effectively by bcl-2 but only weakly by crmA. On the contrary, expression of Ich-1s, which is a truncated version of Ich-1L did not induce cell death; instead, it can protect against cell death induced by serum
CONTROL OF CELL DEATH IN THE NERVOUS SYSTEM
251
deprivation [45]. These results suggest that alternative splicing may represent a switch by which programmed cell death could be regulated. 14.4 The Role of bcl-2 in Neuronal Cell Death Overexpression of the bcl-2 proto-oncogene can rescue a variety of cell types treated by trophic factor deprivation or exposed to noxious stimuli from undergoing cell death (see the contribution of J.Reed, this volume, for a detailed discussion and [48]). bcl-2 and the C. elegans gene ced-9 are functional homologues [49] which have been demonstrated by the ability of bcl-2 to suppress the normally occurring cell death in wild-type C. elegans as well as to substitute for the ced-9 function in mutant animals lacking a functional ced-9 gene [49, 50]. bcl-2 can suppress the cell death in neuronal cell lines, DRG neurons or sympathetic neurons induced by neurotrophin withdrawal [42, 51–53]. Transgenic mice have recently been generated which are overexpressing bcl-2 under the control of the neuron-specific enolase and the phosphoglycerate kinase promotor, resulting in massive suppression of neuronal loss during embryonic development [54]. Various neuroblastoma cells express elevated levels of bcl-2 protein which correlates with an unfavorable histology and poor stage disease as well as an increased resistance to chemotherapeutic drugs due to the survival enhancing activity of bcl-2 [55–57]. It has also been demonstrated that bcl-2 is expressed in postmitotic neurons [58–60], suggesting that bcl-2 is contributing to the survival of these cells which are completely dependent on trophic support and/or synaptic activity. However, the survival enhancing effect of bcl-2 does not extend to all types of neurons. Microinjection of a bcl-2 expression vector into chick neurons demonstrated that, although bcl-2 can suppress cell death in a variety of trophic factor-dependent primary neurons, there is no effect on the survival of CNTF-dependent ciliary neurons [58]. It has been speculated that neurotrophins might increase the expression levels of bcl-2 or promote its activity, thereby preventing cell death, whereas in the absence of the trophic support bcl-2 would become inactivated and cell death could occur [42]. It is unclear whether such a mechanism does exist. BDNF, for example, which supports survival of cultured cerebellar granule neurons, could not influence the levels of bcl-2 messenger RNA in these cultures [61]. It is intriguing though to consider the cyclic changes of bcl-2 expression during the endometrial cycle [62, 63] oscillating between a peak of Bcl-2 staining towards the end of the follicular phase and the absence of Bcl-2 expression during the secretory activity. The authors also describe a similarity between the bcl-2 expression pattern and that of oestrogen and progesterone receptors, indicative of a role for ovarian hormones in the regulation of bcl-2 expression. It was shown in addition that disappearance of bcl-2 expression in glandular cells at the secretory phase coincided with the appearance of apoptotic cells. Further support for the possible control of bcl-2 expression on a transcriptional level comes from a series of recent publications [64–66] describing in a myeloid leukemia cell line the downregulation of bcl-2 expression via a negative response element in the bcl-2 gene 5 untranslated region by the p53 tumor suppressor gene. Interestingly there is a concomitant increase in the levels of bax expression resulting in an overall enhanced cell death kinetic, bcl-x is a member of the bcl-2/ced-9 family which have been shown to exist in long- and short-splice variants [67, 68]. The short forms of bcl-x act as a dominant negative suppressor of bcl-2, whereas the long form has similar cell-death-suppressing activity like bcl-2. The long form of bcl-x also seems to be predominantly expressed in tissues containing long-lived postmitotic cells like the brain. Neuronal damage elicited by systemic administration of the glutamate analogue kainic acid led to increased p53 mRNA levels in sensitive brain regions and apoptotic cell death in the corresponding areas [69]. Others have shown that mice deficient in p53 exhibit elevated Bcl-2 and reduced Bax protein levels in
252
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
several tissues [66]. It is therefore intriguing to speculate that the relative levels of the various bcl-2 family members and consequently the net increase or decrease in cell survival might be transcriptionally regulated by p53. How does bcl-2 increase the survival of neuronal cells at least in vitro? Recent findings proposed a role of bcl-2 in preventing cells from the effects of oxidative stress and placed bcl-2 at a central position in an antioxidant pathway of cell rescue [70, 71]. Observations made in a hypothalamic neural cell line support this notion by demonstrating that bcl-2 overexpression could be causally linked to a decreased generation of reactive oxygen species [72]. In L929 lymphoma cells, for example, pro-oxidant compounds induce apoptosis by releasing Ca2+ from mitochondria, resulting in increased cytosolic Ca2+ levels which subsequently activate the cell death program. Bcl-2 expression in these cells can prevent Ca2+ release from mitochondria by stabilizing and maintaining the membrane potential across the mitochondrial membrane ( psi) [73]. Treatment of the same cells with the ionophore nigericin, which increases psi, protected the cells against TNF mediated cytotoxicity [74]. Expression of bcl-2 in WEHI7.2 cells has been reported to regulate the flux of Ca2+ across the ER membrane inhibiting also the Ca2+ -dependent activation of the cell death program [75]. These potential functions of bcl-2 are consistent with the described intracellular localization of Bcl-2 protein to the mitochondrial membrane, the endoplasmic reticulum and the perinuclear membrane. Considering the above mentioned results, it is perhaps surprising that germline-targeted bcl-2 (−/−) mice showed no detectable deficiencies in the nervous system [76]. One possible explanation is the existence of a redundant system of genes with similar functions, e.g. bcl-x1 may be sufficient by itself to compensate for the loss of bcl-2. The creation of double knock-out mice (e.g. bcl-2 (−/−) and bcl-x1 (−/−)) should be helpful in shedding some light onto the role of distinct members of the bcl-2/ced-9 family of cell-death-suppressor genes. 14.5 Pathology One interesting hypothesis yet to be proven is that the genes that control developmental programmed cell death may also contribute to neuronal degenerative diseases. Thus, it is of interest to consider if there is any evidence to suggest the involvement of programmed cell death genes in such diseases as Alzheimer’s disease, amytrophic lateral sclerosis, Parkinson’s disease or stroke. 14.5.1 Alzheimer's Disease Alzheimer’s disease is characterized by the formation of abnormal structures in the brain, particularly in the neocortex and the hippocampus. These depositions are referred to as amyloid plaques in the intercellular space and neurofibrillary tangles, which are located within the cytoplasm. Formation of these structures is associated with cerebral degeneration due to massive loss of neurons and synapses with the consequence of severe dementia as well as behavioural and cognitive havoc in affected patients. The deposition of amyloid peptide (A ) in the core structure of the amyloid fibrils of senile plaques is thought to play a central role in amyloid plaque formation (reviewed by [77, 78]). A is a 39 to 43 amino acid long fragment from APP, the -amyloid precursor protein, and is released from this precursor protein by proteolytic cleavage involving an as yet unidentified protease, termed ‘ -secretase’. A thorough discussion of the
CONTROL OF CELL DEATH IN THE NERVOUS SYSTEM
253
molecular pathways of A release from APP expressing cells is clearly outside the scope of this article. We would like to focus instead on the neurotoxic effects of A peptide. The finding that extracellular A can be identified in human CSF and the plasma of healthy individuals in concentrations similar to those in AD patients, and that it is also released under normal culture conditions from various primary cells and cell lines, has shed considerable doubt on the initial hypothesis that A production per se and A deposition is sufficient to elicit Alzheimer’s disease. It also implies that A has a function outside any pathological situation. In vitro, A has been shown to exert trophic effects on neurite integrity or even neurite outgrowth in a series of studies [79–81]. In a recent article, Lofiler and colleagues demonstrated that transiently high concentrations of soluble APP could be detected in the rat brain during early postnatal development as well as in the adult olfactory bulb, which is characterized by persistant synaptogenesis. The authors correlated the high levels of APP with cell differentiation and synaptogenesis [82]. Disruption of the APP gene in mice by inserting a neo cassette resulted in the expression of a shortened APP protein without axon 2. The mutant mice were severely impaired in spatial learning and exploratory behavior, and they showed an increased incidence of agenesis of the corpus callosum, suggesting that APP may have a role in developmental or synaptic plasticity. Expression of the mouse homeobox gene Hox-3.1 in vitro resulted in a repression of APP transcription, suggesting a potentially important function of APP during embryonic development of the nervous system [83]. Other proteins like phosphorylated tau protein isoforms, which have previously been associated with the paired helical filaments of neurofibrillary tangles in AD brains only, show an expression pattern during embryogenesis that also suggests a functional role during differentiation and axon formation rather than for neuronal degeneration [84]. Low concentrations of A 1–40 peptide (EC50=0.06 nM) can induce neuronal differentiation in primary cultures of hippocampal neurons, whereas high concentrations (EC50=100 nM) were neurotoxic [85]. Interestingly, the authors also determined that the trophic and the toxic activity of A 1–40 resides in a 10 amino acid sequence (amino acids 23–35) and that this sequence has a high structural and functional homology with tachykinin peptides. A physiological role for the A peptide as tachykinin antagonist is therefore conceivable, although others have doubted this function of A peptides [86, 87]. There is increasing evidence that A -mediated neurotoxicity follows an apoptotic pathway. Neurons exposed to A peptides display characteristic features of apoptosis such as nuclear condensation, cell shrinkage, membrane blebbing and DNA fragmentation into oligonucleosomes [88]. Chronic application of rather high concentrations (25–100 µM) of A 25–35 to primary hippocampal neurons in culture resulted in neuronal cell death by apoptosis [89]. It could also be demonstrated that A 25–35 peptide could polymerize under the culture conditions to amyloid-like fibrils. It is possible that it is the aggregate form of A which is responsible for neurite dystrophy and neuron death [90–92]. Overexpression of APP in a postmitotic neuron model reveals aberrant processing of APP and also cleavage fragments that potentially could give rise to amyloid-like fibrils [93]. Aberrant processing of APP in a neuronal cell line could be due to the fact that the protein is produced in abundant amounts within the cells and this might simply exceed the cell’s capacity to process APP correctly [94]. To summarize these few examples, it seems plausible that in vitro A peptides, depending on the proper concentration and conformation, are able directly to induce neuronal cell death by utilizing most likely an apoptotic pathway. This may have important implications for the conceptual design of a therapy against Alzheimer’s disease. It has been shown that a variety of non-neuronal and neuronal cell types whose survival is dependent on the presence of growth or survival factors undergo apoptosis upon growth factor withdrawal [95–100]. A role for growth factors in the control of APP expression levels has been postulated. BDNF messenger RNA, for example, appears to be downregulated in the hippocampus of AD patients [101]. Removal of NGF from the culture of rat sympathetic ganglion neurons results in neuronal cell death.
254
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
Smith et al. could show that the death of these neurons is associated with a shift of the relative amounts for APP695 and APP751/770 mRNA towards the high-molecular-weight forms of APP [102]. Regulation of APP expression levels has also been reported for Transforming Growth Factor-beta 1 (TGF- 1) and IL-1 in astrocytes and human glioma cells [103, 104]. A direct involvement of the genes that control programmed cell death in the pathogenesis of Alzheimer’s disease has not been demonstrated. It is interesting to note, though, that the cerebrospinal fluid of Alzheimer’s patients contains significantly higher levels of IL-1 than that in age-matched controls [105]. It is not yet clear which tissue or cells are the source of IL-1 in these Alzheimer’s patients. However, the finding that IL-1 can stimulate the expression of APP on the transcriptional level implicates the possibility of a link between IL-1 , APP and neuronal death [106–108]. 14.5.2 Excitotoxicity L-Glutamate is the primary neurotransmitter in the mammalian CNS responsible for excitatory synaptic transmission. Usually, L-glutamate concentrations in the extracellular space of neurons are low and rise only transiently in spacially restricted areas during synaptic transmission. Sustained exposure of excitable neurons to elevated levels of L-glutamate may be an important cause for various dysfunctions of the brain, e.g. stroke and ischemia, Huntingdon’s disease and Alzheimer’s disease. Three classes of receptor have been identified by virtue of binding characteristic pharmacologic agonists: (1) N-methyl-D-aspartate, (2) quisqualate, and (3) kainate. All three classes of receptor are associated with cation channels, the best characterized still being the NMDA receptor type. Glutamate activation of this receptor regulates influx of Ca2+ into neurons, which is the effector for long-term potentiation as well as excitotoxicity. A number of growth factors are capable of ameliorating the effects of NMDA mediated excitotoxicity. For instance it has been demonstrated that FGF and excitotoxic amino acids have opposing effects on hippocampal pyramidal neurons: excitotoxic amino acids induced dendritic regression and cell death, whereas FGF (FGF) promoted neurite outgrowth and cell survival by raising the threshold of killing for exci to toxic amino acids [109]. FGF may have achieved this effect by weakening the CA2+ response to glutamate mainly by enhancing the Na +—dependent Ca2+ extrusion. Another study demonstrated that bFGF can significantly inhibit neurotoxicity mediated by NMDA receptor in striatal neurons [110]. NGF, bFGF and the insulin-like growth factors (IGFs) protected rat hippocampal and septal neurons as well a human cortical neurons from NMDA mediated excitotoxicity induced by hypoglycemia [111]. By using the calcium indicator fura-2, the authors show that the growth factors could stabilize intraneuronal Ca2+concentrations which, as in the other examples mentioned before, appears to be the primary intracellular mediator of excitotoxicity. Glutamate can induce cell death of PC 12 phaeochromocytoma cells. When bcl-2 is overexpressed in these cells they become resistant to the treatment with the excitotoxic amino acid [112] exemplifying again the potential role of bcl-2 in maintaining intracellular Ca2+ homeostasis (compare also with [75, 73]). 14.5.3 Ischemia Trauma, stroke, excitotoxicity and hypoxia are conditions of ischemia which is associated with extensive neuronal cell death in sensitive brain regions. Recent studies have indicated that common biochemical mechanisms may exist between these forms of neuronal cell death and apoptosis in vivo. It has always been
CONTROL OF CELL DEATH IN THE NERVOUS SYSTEM
255
postulated that apoptosis is an active process requiring macromolecular synthesis by the cells undergoing cell death and that inhibition of transcription (by actinomycin D) or translation (by cycloheximide) could then block the execution of the intrinsic death program. Necrotic cell death, on the other hand, cannot be blocked by inhibitors of macromolecular synthesis and occurs upon gross perturbation of cellular homeostasis. As a hallmark of apoptosis degradation of DNA into oligonucleosomal size fragments has been described which results in a ‘ladder’ pattern of fragments with 180–200 bp and multiples thereof, when analyzed by gel electrophoresis. It should be noted though that it becomes increasingly common sense to evalute morphological criteria in parallel to DNA fragmentation to describe the cell death process in a more holistic way. Recent studies demonstrated that, after experimentally induced ischemia, characteristic apoptotic DNA fragmentation occurs in affected brain regions [113–115]. Interestingly, the size of the infarction could be reduced by continuous infusion of cycloheximide, indicating that ongoing protein synthesis is a prerequisite for ischemic neuronal cell death [114]. These results are substantiated by numerous in vitro experiments which demonstrate that the protein synthesis inhibitor cycloheximide can block neuronal cell death in various experimental paradigms [116–124]. However, other reports claim that ischemic brain injury does not uniformly induce apoptotic DNA fragmentation, but that random DNA fragmentation could be observed as well suggesting locally confined differences in the affected brain regions [115, 125]. This variability may be due to experimental conditions. Dragunow et al. could for example demonstrate that hypoxia-induced neuronal cell death induced transcription of immediate early genes like c-Fos, Fos B, Jun B, Jun D, and cJun proteins only under mild hypoxia (15 min) whereas severe hypoxia (60 min) induced infarction and necrotic cell death. Under severe hypoxia expression of these immediate early genes was observed only in the non-ligated part of the rat brain [126]. The mode of neuronal cell death during ischemic brain injury (‘apoptotic’ versus ‘necrotic’) has potentially important implications: the identification of components of the cell death machinery would allow a specific targeting in therapeutic approaches yet to be developed to preserve as many neurons as possible. Recent publications suggested that apoptosis does indeed contribute to neuronal damage after ischemic injury. Martinou et al. [55] have shown that the infarct size is 50% smaller in the brain of transgenic mice expressing bcl-2 compared to wild-type mice after middle cerebral artery occlusion, suggesting that bcl-2 can offer some protection against ischemic injury. The recent identification of interleukin-1 converting enzyme (ICE) as the human ced-3 homologue is also of particular interest: after focal ischemia, and especially in hypertensive rats, messenger RNA levels for interleukin-1 are upregulated [127]. This is by no means final proof for an engagement of ICE in ischemiainduced neuronal cell death, but it is intriguing to speculate about a potential role which this enzyme or a closely related molecule might have under the given circumstances. 14.5.4 Superoxide Dismutase and Motoneuron Disease (ALS) An increasing number of recent findings focused on the correlation of neurodegenerative events with an increased generation of reactive oxygen species. In these studies a central role is ascribed to the superoxide anion and the Cu/Zn superoxide dismutase, a metalloenzyme which catalyzes the dismutation of the superoxide radical to O2 and H2O2. The growing interest in reactive oxygen species is due partly to reports that link familiar cases of amyotrophic lateral sclerosis to defects in the superoxide dismutase enzyme, where a number of mutations have been identified which result in a decreased SOD activity and hence in an increased susceptibility to situations were cytotoxic superoxide radicals are generated [128–132]. The mutations in SOD have been localized to amino acids which are involved in the formation of the
256
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
catalytically active homodimer: patients carrying heterozygote mutations of SOD displayed a 50% reduction of SOD activity in red blood cells [131]. It should be noted, however, that SOD activity does not seem to be affected in sporadic cases of amyotrophic lateral sclerosis [129]. Furthermore, expression of high levels of human SOD containing a substitution of Gly to Ala at position 93, a change that has little effect on enzymatic activity of SOD, caused motor neuron degeneration in transgenic mice [130]. These results suggest that dominant gain-of-function mutations in SOD are sufficient for the pathogenic manifestation of familial ALS but do not clarify the potential contribution of reactive oxygen species in the pathogenesis of this disease. However, support for a participation of oxygen radicals in neurotoxicity comes from various in vitro experiments. Cultured cortical neurons from transgenic mice overexpressing human copper-zinc-superoxide dismutase have been used to study the role of SOD in glutamate-induced neural death and it was shown that neurons overexpressing SOD are protected against this type of cell death. This indicates that oxidative stress mediated by superoxide radicals is part of the NMDA receptor mediated neurotoxicity [133, 132]. Neurotoxicity induced in mouse spinal cord neurons by the generation of superoxide and hydrogen peroxide through the xanthin oxidase/hypoxanthine system could be blocked by scavengers of ROS as well as by antagonists of the NMDA receptor [134]. Treatment of PCT2 cells or spinal cord organotypic cultures using antisense oligonucleotides against superoxide dismutase could demonstrate reduced cell survival in the presence or absence of NGF and induction of apoptosis in spinal neurons, respectively [135, 136]. NGF has been shown to exert its neuroprotective effect against ROS mediated cytotoxicity, e.g. by induction of a free radical detoxifying mechanism, such as catalase activity in PC 12 cells, thereby reducing the sensitivity against H2O2 and hydroxyl radicals [137, 138]. Also, microglia-mediated PC12 cell death has been shown to be due to superoxide radical generation in the activated microglia [139]. The potential association of increased levels of ROS with mutations of the SOD gene in familial ALS is particularly exciting in respect to the capability of bcl-2 to participate in protecting cells against reactive oxygen species (ROS) from undergoing apoptotic cell death. It remains to be examined whether in vivo ROS are directly responsible for inducing the apoptotic elimination of motoneurons in ALS or whether another, as yet unidentified function of SOD is the primary cause for the massive motoneuron degeneration. Since accumulation of ROS has in addition been proposed to be an important contribution to aging, it also would be interesting to investigate the involvement of a ROS-dependent apoptosis mechanism in neurodegenerative events during aging. Acknowledgments This work is supported in part by grants to J.Y. from Bristol Myer-Squibb and from the National Institute of Aging. References 1
2 3
. GLÜCKSMANN, A. (1950) Cell deaths in normal vertebrate ontogeny. Biol. Rev. Cambridge Philos. Soc. 26, 59–86. 1*. OPPENHEIM, R.W. (1991) Cell death during development of the nervous system. Annu. Rev. Neurosci. 14, 453–501. . JOHNSON, E.M., & DECKWERTH, T.L. (1993) Molecular mechanisms of developmental neuronal death. Annu. Rev. Neurosci. 16, 31–46. . KORSCHING, S. (1993) The neurotrophic factor concept: a reexamination. J. Neurosci. 13, 2739–2748.
CONTROL OF CELL DEATH IN THE NERVOUS SYSTEM
4
5 6
7 8 9 10 11
12 13
14 15
16 17
18
19
20 21 22 23
257
. FUBER, S., OPPENHEIM, R.W. & PREVETTE, D. (1987) Naturally-occurring neuron death in the ciliary ganglion of the chick embryo following removal of preganglionic input: evidence for the role of afferents in ganglion cell survival. J. Neurosci. 7, 1816– 1832. . KLEIN, R. (1994) Role of neurotrophins in mouse neuronal development. FASEB J. 8, 738–744. . CROWLEY, C., SPENCER, S.D., NISHIMURA, M.C., CHEN, K.S., PITTS-MEEK, S., ARAMANINI, M.P., LING, L.H., MCMAHON, S.B., SHELTON, D.L., LEVINSON, A.D. et al. (1994) Mice lacking nerve growth factor display perinatal loss of sensory and sympathetic neurons yet develop basal forebrain cholinergic neurons. Cell 76, 1001–1011. . SVENDSEN, C.N., COOPER, J.D. & SOFRONIEW, M.V. (1991) Trophic factor effects on septal cholinergic neurons. Ann. N.Y. Acad. Sci. 640, 91–94. . HOFER, M., PAGLIUSI, S.R., HOHN, A., LEIBROCK, J. & BARDE, Y.A. (1990) Regional distribution of brain-derived neurotrophic factor mRNA in the adult mouse brain. EMBO J. 9, 2459–2464. . SENDTNER, M., HOLTMANN, B., KOLBECK, R., THOENEN, H. & BARDE, Y.A. (1992) Brain-derived neurotrophic factor prevents the death of motoneurons in newborn rats after nerve section Nature 360, 757–759. .OPPENHEIM, R.W., YIN, Q.W., PREVETTE, D. & YAN, Q. (1992) Brain-derived neurotrophic factor rescues developing avian motoneurons from cell death. Nature 360, 755–757. . HENDERSON, C.E., CAMU, W., METTLING, C., GOUIN, A., POULSEN, K., KARIHALOO, M., RULLAMAS, J., EVANS, T., MCMAHON, S.B., ARMANINI, M.P. et al. (1993) Neurotrophins promote motoneuron survival and are present in embryonic limb bud. Nature 360, 266–270. . ERNFORS, P., LEE, K.F., KUCERA, J. & JAENISCH, R. (1994) Lack of neurotrophin-3 leads to deficiencies in the peripheral nervous system and loss of limb proprioceptive afferents. Cell 77, 503–512. . KLEIN, R., SILOS-SANTIAGO, I., SMEYNE, R.J., LIRA, S.A., BRAMBILLA, R., BRYANT, S., ZHANG, L., SNIDER, W.D. & BARABACID, M. (1994) Disruption of the neurotrophin-3 receptor gene trkC eliminates Ia muscle afferents and results in abnormal movement. Nature 368, 249–251. . RABIZADEH, S., OH, J, ZHONG, L.T., YANG, J., BITLER, C.M., BUTCHER, L.L. & BREDESEN, D.E. (1993) Induction of apoptosis by the low-affinity NGF receptor. Science 261, 345–348. . VON BARTHELD, C.S., KINOSHITA, Y., PREVETTE, D., YIN, Q.-W., OPPENHEIM, R.W. & BOTHWELL, M. (1994) Positive and negative effects of neurotrophins on the isthmo-optic nucleus in chick embryos. Neuron. 12, 639–654. . BARRETT, G.L. & BARLETT, P.F. (1994) The p75 nerve growth factor receptor mediates survival or death depending on the stage of sensory neuron development. Proc. Natl Acad. Sci. USA 91, 6501–6505. . HANTZOPOULOS, P.A., SURI, C., GLASS, D.J., GOLDFARB, M.P. & YANCOPOULOS, G.D. (1994) The low affinity NGF receptor, P75, can collaborate with each of the Trks to potentiate functional responses to the neurotrophins. Neuron. 13, 187– 201. . VERDI, J.M., BIRREN, S.J., IBANEZ, C.F., PERSSON, H., KAPLAN, D.R., BENEDETTI, M., CHAO, M.V. & ANDERSON, D.J. (1994) p75LNGFR regulates Trk signal transduction and NGF-induced neuronal differentiation in MAH cells. Neuron. 12, 733–745. . STÖCKLI, K.A., LILLJEN, L.E., NAHERNOE, M., BREITFELD, G. HUGHES, R.A., RAFF, M.C., THOENEN, H. & SENDTNER, M. (1991) Regional distribution, development changes, and cellular localization CNTF-messenger RNA and protein in the rat brain. J. Cell Biol. 115, 447–459. . MASU, Y., WOLF, E., HOLTMANN, B., SENDTNER, M., BREM, G. & THOENEN, H. (1993) Disruption of the CNTF gene regults in motor neuron degeneration. Nature 365, 27–32. . LIN, L.F., ARMES, L.G., SOMMER, A., SMITH, D.J. & COLLINS, F. (1990) Isolation and characterization of ciliary neurotrophic factor from rabbit sciatic nerves. J. Biol. Chem. 265, 8942–8947. . BOTTNER, D., BRUGGEMANN, W. & UNSICKER, K. (1989) Ciliary neurotrophic factor supports targetdeprived preganglionic sympathetic spinal cord neurons. Neurosci. Lett. 105, 316–320. . HAGG, T. & VARON, S. (1993) Ciliary neurotrophic factor (CNTF) prevents degeneration of adult rat substantia nigra dopaminergic neurons in vivo Proc. Natl Acad. Sci. USA . 90, 6315–6319.
258
24 25 26 27 28
29
30 31
32 33 34 35
36
37 38 39
40
41
42 43
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. ANDERSON, D.J. (1989) New roles for PDGF and CNTF in controlling the timing of glial cell differentiation in the optic nerve. Trends Neurosci. 12, 83–85. . BARRES, B.A., SCHMID, R., SENDNTER, M. & RAFF, M.C. (1993) Multiple extra-cellular signals are required for long-term oligodendrocyte survival. Development 118, 283–295. . SENDTNER, M., ARAKAWA, Y., STÖCKLI, K.A., KREUTZBERG, G.W. & THOENEN, H. (1991) Effect of ciliary neurotrophic factor (CNTF) on motoneuron survival. J. Cell. Sci. Suppl. 15, 103–109. . MITSUMOTO, H., IKEDA, K., KLINKOSZ, B., CEDARBAUM, J.M., WONG, V. & LINDSAY, R.M. (1994) Arrest of motor neuron disease in wobbler mice cotreated with CNTF and BDNF. Science 265, 1107–1110. . NURCOMBE, V., WREFORD, N.G. & BERTRAM, J.F. (1991) The use of the optical disector to estimate the total number of neurons in the developing chick lateral motor column: effects of purified growth factors. Anat. Rec. 231, 416–424. . OPPENHEIM, R.W., PREVETTE, D., HAVERKAMP, L.J., HOUENOU, L. YIN, Q.W. & MCMANAMAN, J. (1993) Biological studies of a putative avian muscle-derived neurotrophic factor that prevents naturally occurring motoneuron death in vivo. J. Neurobiol.24, 1065–1079. . WEWETZER, K.J., MACDONALD, J.R., COLLINS, F. & UNSICKER, K. (1990) CNTF rescues motoneurons from ontogenetic cell death in-vivo but not in-vitro. N euro-report. 1, 203–206. . IP, N.Y., MCCLAIN, J., BARREZUETA, N.X., ALDRICH, T.H., PAN, L. LI, Y.P., WIEGAND, S.J., FRIEDMAN, B., DAVIS, S. & YNACOPOULOS, G.D. (1993) The alpha component of the CNTF receptor is required for signalling and defines potential CNTF targets in the adult and during development. Neuron 10, 89–102. . DOBREA, G.M., UNNERSTALL, J.R. & RAO, M.S. (1992) The expression of CNTF message and immunoreactivity in the central and peripheral nervous system of the rat. Dev. Brain Res. 66 209–219. . BONNI, A., FRANK, D.A., SCHINDLER, C. & GREENBERG, M.E. (1993) Characterization of a pathway for ciliary neurotrophic factor signaling to the nucleus. Science 262, 1575–1579. . BAIRD, A. (1994) Fibroblast growth factors: activities and significance of non-neurotrophin neurotrophic growth factors. Curr. Opp. Neurobiol. 4, 78–86. . HUGHES, R.A., SENDTNER, M., GOLDFARB, M., LINDHOLM, D. & THOENEN, H. (1993) Evidence that fibroblast growth factor-5 is a major muscle-derived survival factor for cultured spinal motoneurons. Neuron. 10, 369–377. . JACKSON, A., FRIEDMAN, S., ZHAN, X., ENGLEKA, K.A., FOROUGH, R. & MACIAG, T. (1992) Heat shock induces the release of fibroblast growth factor 1 from NIH 3T3 cells. Proc. Natl Acad. Sci. USA 89, 10691–10695. . YUAN, J. & HORVITZ, H.R. (1990) The Caenorhabditis elegans genes ced-3 and ced-4 act autonomously to cause programmed cell death. Dev. Biol. 138, 33–41. . YUAN, J.Y., SHAHAM, S., LEDOUX, S., ELLIS, H.M. & HORVITZ, H.R. (1993) The C. elegans cell death gene ced-3 encodes a protein similar to mammalian interleukin-1 beta-converting enzyme. Cell 75, 641–652. . THORNBERRY, N.A., BULL, H.G., CALACAY, J.R., CHAPMAN, K.T., HOWARD, A.D., KOSTURA, M.J., MILLER, D.K., MOLINEAUX, S.M., WEIDNER, J.R., AUNINS, J. et al. (1992) A novel heterodimeric cysteine protease is required for interleukin 1-beta processing in monocytes. Nature 356, 768–774. . MIURA, M., ZHU, H., ROTELLO, R., HARTWIEG, E.A. & YUAN, J.Y. (1993) Induction of apoptosis in fibroblasts by IL-1 beta-converting enzyme, a mammalian homolog of the C. elegans cell death gene ced-3. Cell 75, 653–660. . RAY, C.A., BLACK, R.A., KRONHEIM, S.R., GREENSTREET, T.A., SLEATH, P. R., SALVESEN, G.S. & PICKUP, D.J. (1992) Viral inhibition of inflammation: cowpox virus encodes an inhibitor of the interleukin converting enzyme. Cell 69, 697– 604. . GAGLIARDINI, V., FERNANDEZ, P.-A., LEE, R.K.K., DREXLER, H.C.A., ROTELLO, R.J., FISHMAN, M.C. & YUAN, J. (1994) Prevention of vertebrate neuronal death by the crmA gene. Science 263, 826–828. . DINARELLO, C.A. (1980) Biology of interleukin 1. FASEB J. 2, 108–115.
CONTROL OF CELL DEATH IN THE NERVOUS SYSTEM
44
45 46
47 48 49 50 51 52 53 54
55
56
57
58 59
60 61
62 63
259
. KOMIYAMA, T., RAY, C.A., PICKUP, D.J., HOWARD, A.D., THORNBERRY, N.A., PETERSON, E.P. & SALVESEN, G. (1994) Inhibition of interleukin-1 beta converting enzyme by the cowpox virus serpin crma—an example of cross-class inhibition. J. Biol Chem. 269, 19331–19337. . WANG, L., MIURA, M., BERGERON, L., ZHU, H. & YUAN, J.Y. (1994) Ich-1, an ice/ced-3 related gene, encodes both positive and negative regulators of programmed cell death. Cell 78, 739–750. . KUMAR, S., KINOSHITA, M., NODA, M., COPELAND, N.G. & JENKINS, N.A. (1994) Induction of apoptosis by the mouse nedd2 gene, which encodes a protein similar to the product of the Caenorhabditis elegans cell death gene ced-3 and the mammalian IL-1 beta-converting enzyme. Genes. Dev. 8, 1613–1626. . KUMAR, S., TOMOOKA, Y. & NODA, M. (1992) Identification of a set of genes with developmentally downregulated expression in the mouse brain. Biochem. Biophys. Res. Comm. 185, 1155–1161. . KORSMEYER, S.J. (1992) Bcl-92—An antidote to programmed cell death. Cancer Surv. 15, 105–118. . HENGARTER, M.O. & HORVITZ, H.R. (1994) C. elegans cell survival gene ced-9 encodes a functional homology of the mammalian proto-oncogene bcl-2. Cell 76, 665– 676. . VAUX, D.L., WEISSMAN, I.L. & KIM, S.K. (1992) Prevention of programmed cell death in Caenorhabditis elegans by human bcl-2. Science 258, 1955–1957. . BATISTATOU, A., MERRY, D.E., KORSMEYER, S.J. & GREENE, L.A. (1993) Bcl-2 affects survival but not neuronal differentiation of PC12 cells. J. Neurosci. 13, 4422–4428. . GARCIA, I., MARTINOU, I., TSUJIMOTO, Y. & MARTINOU, J.C. (1992)Prevention of programmed cell death of sympathetic neurons by the bcl-2 proto-oncogene. Science 258, 302–304. . MAH, S.P., ZHONG, L.T., LIU, Y., ROGHANI, A., EDWARDS, R.H. & BREDESEN, D.E. (1993) The protooncogene bcl-2 inhibits apoptosis in PC12 cells. J. Neuro-chem. 60, 1183–1186. . MARTINOU, J.C., DUBOISDAUPHIN, M., STAPLE, J.K., RODRIGUEZ, I., FRANKOWSKI, H., MISSOTTEN, M., ALBERTINI, P., TALABOT, D., CATSICAS, S., PIETRA, C. et al. (1994) Overexpression of BCL-2 in trangenic mice protects neurons from naturally occurring cell death and experimental ischemia. Neuron. 13, 1017–1030. . CASTLE, V.P., HEIDELBERGER, K.P., BROMBERG, J., OU, X.G., DOLE, M. & NUNEZ, G. (1993) Expression of the apoptosis suppressing protein bcl-2, in neuroblastoma is associated with unfavorable histology and N-myc amplification. Am. J. Pathol. 143, 1543–1550. . HANADA, M., KRAJEWSKI, S., TANAKA, S., CAZALSHATEM, D., SPENGLER, B.A., ROSS, R.A., BIEDLER, J.L. & REED, J.C. (1993) Regulation of Bcl-2 oncoprotein levels with differentiation of human neuroblastoma cells. Cancer. Res. 53, 4978–4986. . REED, J., MEISTER, L., TANAKA, S., CUDDY, M., YUM, S., GEYER, C. & PLEASURE, D. (1991) Differential expression of bcl2 protooncogene in neuroblastoma and other human tumor cell lines of neural origin. Cancer. Res. 51, 6529–6538. . ALLSOPP, T.E., WYATT, S., PATERSON, H.F. & DAVIES, A.M. (1993) The proto-oncogene bcl-2 can selectively rescue neurotrophin factor-dependent neurons from apoptosis. Cell 73, 295–307. . HOCKENBERY, D.M., ZUTTER, M., HICKEY, W., NAHM, M. & KORSMEYER, S.J. (1991) BCL2 protein is topographically restricted in tissues characterized by apoptotic cell death. Proc. Natl Acad. Sci. USA 88, 6961–6965. . MIZUGUCHI, M., IKEDA, K., ASADA, M., MIZUTANI, S. & KAMOSHITA, S. (1994) Expression of bcl-2 protein in murine neural cells in culture. Brain Res. 649, 197–202. . CASTREN, E., OHGA, Y., BERZAGHI, M.P., TZIMAGIORGIS, G., THOENEN, H. & LINDOHOLM, D. (1994) bcl-2 messenger RNA is localized in neurons of the developing and adult rat brain. Neuroscience. 61, 165–177. . GOMPEL, A., SABOURIN, J.C., MARTIN, A., YANEVA, H., AUDOUIN, J., DECROIX, Y. & POITOUT, P. (1994) Bcl-2 expression in normal endometrium during the menstrual cycle. Am. J. Pathol. 144, 1195–1202. . OTSUKI, Y., MISAKI, O., SUGIMOTO, O., ITO, Y., TSUJIMOTO, Y. & AKAO, Y. (1994) Cyclic bcl-2 gene expression in human uterine endometrium during menstrual cycle. Lancet. 344, 28–29.
260
64
65 66
67
68 69 70 71 72
73 74
75
76 77 78 79 80 81 82 83
84
85
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. SELVAKUMARAN, M., LIN, H.K., MIYASHITA, T., WANG, H.G., KRAJEWSKI, S., REED, J.C., HOFFMAN, B. & LIEBERMANN, D. (1994) Immediate early upregulation of bax expression by p53 but not TGF beta 1: a paradigm for distinct apoptotic pathways. Oncogene. 9, 1791–1798. . MIYASHITA, T., HARIGAI, M., HANADA, M. & REED,!. C. (1994) Identification of a p53-dependent negative response element in the bcl-2 gene. Cancer Res. 54, 3131–3135. . MIYASHITA, T., KRAJEWSKI, S., KRAJEWSKA, M., WANG, H.G., LIN, H.K., LIEBERMANN, D.A., HOFFMAN, B. & REED, J.C. (1994) Tumor suppressor p53 is a regulator of bcl-2 and bax gene expression in vitro and in vivo. Oncogene. 9, 1799–1805. . BOISE, L.H., GONZALEZGARCIA, M., POSTEMA, C.E., DING, L.Y., LINDSTEN, T., TURKA, L.A., MAO, X.H., NUNEZ, G. & THOMPSON, C.B. (1993) bcl-x, a bcl-2-related gene that functions as a dominant regulator of apoptotic cell death. Cell 74, 597–608. . OLTVAI, Z.N., MILLIMAN, C.L. & KORSMEYER, S.J. (1993) bcl-2 Heterodimerizes in vivo with a conserved homolog, bax, that accelerates programmed cell death. Cell 74, 609–619. . SAKHI, S., BRUCE, A., SUN, N., Tocco, G., BAUDRY, M. & SCHREIBER, S.S. (1994) p53 induction is associated with neuronal damage in the central nervous system. Proc. Natl Acad. Sci. USA 91, 7525–7529. . HOCKENBERY, D.M., OLTVAI, Z.N., YIN, X.M., MILLIMAN, C.L. & KORSMEYER, S.J. (1993) Bcl-2 functions in an antioxidant pathway to prevent apoptosis. Cell 75, 241–251. . SARAFIAN, T.A. & BREDESEN, D.E. (1994) Invited commentary: is apoptosis mediated by reactive oxygen species? Free Radical Res. 21, 1–8. . KANE, D.J., SARAFIAN, T.A., ANTON, R., HAHN, H., BUTLER GRALLA, E. SELVERSTONE VALENTINE, J. ORD, T. & BREDESEN, D.E. (1993) Bcl-2 inhibition of neural cell death: decreased of reactive oxygen species. Science 262, 1274–1277. . RICHTER, C. (1993) Pro-oxidants and mitochondrial Ca2+—their relationship to apoptosis and oncogenesis. FEBS Lett 325, 104–107. . HENNET, T., BERTONI, G. RICHTER, C. & PETERHANS, E. (1993) Expression of BCL-2 protein enhances the survival of mouse fibrosarcoid cells in tumor necrosis factor-mediated cytotoxicity. Cancer Res. 53, 1456–1460. . LAM, M., DUBYAK, G., CHEN, L., NUNEZ, G. MIESFELD, R.L. & DIS TELHORST, C.W. (1994) Evidence that BCL-2 represses apoptosis by regulating endoplasmic reticulum-associated Ca2+ fluxes. Proc. Natl Acad. Sci. USA 91, 6569–6573. . VEIS, D.J., SORENSON, C.M., SHUTTER, J.R. & KORSMEYER, S.J. (1993)Bcl-2-deficient mice demonstrate fulminant lymphoid apoptosis, polycystic kidneys, and hypopigmented hair. Cell 75, 229–240. . KATZMAN, R. & SAITOH, T. (1991) Advances in Alzheimer’s disease. FASEB J. 5, 278–286. . SELKOE, D.J. (1991) The molecular pathology of Alzheimer’s disease. Neuron 6, 487– 498. . KOO, E.H., PARK, L. & SELKOE, D.J. (1993) Proc. Natl Acad. Sci USA 90, 4748–4752. . WHITSON, J.S. (1990) Beta-amyloid protein promotes neuritic branching in hippocampal cultures. Neurosci. Lett. 110, 319–324. . WHITSON, J.S., SELKOE, D.J. & COTMAN, C.W. (1989) Amyloid beta protein enhances the survival of hippocampal neurons in vitro. Science 243, 279–282. . LOFFLER, J. & HUBER, G. (1992) Beta-amyloid precursor protein isoforms in various rat brain regions and during brain development. J. Neurochem. 59, 1316–1324. . VIOLETTE, S.M., SHASHIKANT, C.S., SALBAUM, J.M., BELTING, H.G., WANG, J.C.H. & RUDDLE, F.H. (1992) Repression of the beta-amyloid gene in a Hox-3.1-producing cell line. Proc. Natl Acad. Sci. USA 89, 3805–3809. . POPE, W., ENAM, S.A., BAWA, N., MILLER, B.E., GHANBARI, H.A. & KLEIN, W.L. (1993) Phosphorylated tau epitope of Alzheimer’s disease is coupled to axon development in the avian central nervous system. Exp. Neurol. 120, 106–113. . YANKER, B.A., DUFFY, L.K. & KIRSCHNER, D.A. (1990) Neurotrophic and neurotoxic effects of amyloid protein: reversal by tachkyinin neuropeptides. Science 250, 279–250.
CONTROL OF CELL DEATH IN THE NERVOUS SYSTEM
86
87
88
89
90 91 92 93 94
95 96 97
98 99 100 101
102
103 104
261
. KOWALL, N.W., BEAL, M.F., BUSCIGLIO, J. DUFFY, L.K. & YANKNER, B.A. (1991) An in vivo model for the neurodegenerative effects of beta amyloid and protection by substance P. Proc. Natl Acad. Sci. USA 88, 7247–7251. . KOZLOWSKI, M.R., SPANOYANNIS, A. MANLY, S.P., FIDEL, S.A. & NEVE, R.L. (1992) The neurotoxic carboxy-terminal fragment of the Alzheimer amyloid precursor binds specifically to a neuronal cell surface molecule: pH dependence of the neurotoxicity and the binding. J Neurosci. 12, 1679–1687. . LOO, D.T., COPANI, A., PIKE, C.J., WHITTEMORE, E.R., WALENCEWICZ, A. J. & COTMAN, C.W. (1993) Apoptosis is induced by beta-amyloid in cultured central nervous system neurons. Proc Natl Acad Sci USA 90, 7951–7955. . FORLONI, G., CHIESA, R., SMIROLDO, S., VERGA, L., SALMONA, M. TAGLIAVINI, F. & ANGERETTI, N. (1993) Apoptosis mediated neurotoxicity induced by chronic application of beta amyloid fragment 25–35. Neuroreport 4, 523–526. . FRASER, P.E., LEVESQUE, L. & MCLACHLAN, D.R. (1994) Alzheimer A beta amyloid forms an inhibitory neuronal substrate. J. Neurochem. 62, 1227–1230. . ROHER, A.E., BALL, M.J., BHAVE, S.V. & WAKADE, A.R. (1991) Beta-amyloid from Alzheimer disease brains inhibits sprouting and survival of sympathetic neurons. Biochem. Biophys. Res. Commun. 174, 572–579. . PIKE, C.J., CUMMINGS, B.J. & COTMAN, C.W. (1992) beta-Amyloid induces neuritic dystrophy in vitro: similarities with Alzheimer pathology. Neuroreport 3, 769– 772. . YOSHIKAWA, K., AIZAWA, T. & HAYASHI, Y. (1992) Degeneration in vitro of postmitotic neurons overexpressing the Alzheimer amyloid protein precursor [see comments]. Nature 359, 64–67. . FUKUCHI, K., KAMINO, K., DEEB, S.S., FURLONG, C.E., SUNDSTROM, J.A., SMITH, A.C. & MARTIN, G.M. (1992) Expression of carboxy-terminal region of the beta-amyloid precursor protein in a heterogeneous culture of neuroblastoma cells—evidence for altered processing and selective neurotoxicity. Mol. Brain Res. 16, 37–46. . WILLIAMS, G.T., SMITH, C.A., SPOONCER, E., DEXTER, T.M. & TAYLOR, D.R. (1990) Haemopoietic colony stimulating factors promote cell survival by suppressing apoptosis. Nature 343, 76–79. . BATISTATOU, A. & GREENE, L.A. (1991) Aurintricarboxylic acid rescues PC12 cells and sympathetic neurons from cell death caused by nerve growth factor deprivation. J. Cell Biol 115, 461–471. . BAFFY, G., MIYASHITA, T., WILLIAMSON, J.R. & REED, J.C. (1993) Apoptosis induced by withdrawal of interleukin-3 (IL-3) from and IL-3-dependent hematopoietic cell line is associated with repartitioning of intercellular calcium and is blocked by enforced Bcl-2 oncoprotein production. J. Biol. Chem. 268, 6511–6519. . RODRIGUEZ-TARDUCHY, G., COLLINS, M.K. L., GARCIA, I. & LOPEZRIVAS, A. (1992) Insulin-like growth factor-I inhibits apoptosis in IL-3-dependent hemopoietic cells. J. Immunol. 149, 535–540. . MESNER, P.W., WINTERS, T.R. & GREEN, S.H. (1992) Nerve growth factor withdrawal-induced cell death in neuronal PC 12 cells resembles that in sympathetic neurons. J. Cell Biol. 119, 1669–1680. . VAUX, D.L. (1993) Toward an understanding of the molecular mechanisms of physiological cell death. Proc. Natl Acad. Sci. USA 90, 786–789. . PHILLIPS, H.S., HAINS, J.M., ARAMANINI, M., LARAMEE, G.R., JOHNSON, S.A. & WINSLOW, J.W. (1991) BDNF mRNA is decreased in the hippocampus of individuals with Alzheimer’s disease. Neuron 7, 695–702. . SMITH, C.J., JOHNSON, E.M., OSBORNE, P., FREEMAN, R.S., NEVEU, I. & BRACKET, P. (1993) NGF deprivation and neuronal degeneration trigger altered beta-amyloid precursor protein gene expression in the rat superior cervical ganglia in vivo and in vitro. Mol. Brain Res. 17, 328–334. . GRAY, C.W. & PATEL, A.J. (1993) Regulation of beta-amyloid precursor protein isoform messenger RNAs by transforming growth factor-beta-1 and interleukin-1-beta in astrocytes. Mol. Brain Res. 19, 251–256. . YANG, F.M., JANSEN, L., FRIEDRICHS, W.E., BUCHANAN, J.M. & BOWMAN, B.H. (1993) IL-1-beta decreases expression of amyloid precursor protein gene in human glioma cells. Biochem. Biophys. Res. Commun. 191, 1014–1019.
262
105
106
107 108
109 110 111 112 113 114 115 116
117 118
119 120 121 122
123
124
APOPTOSIS IN NORMAL DEVELOPMENT AND CANCER
. CACABELOS, R., BARQUERO, M., GARCIA, P., ALVAREZ, X.A. & VARELA DE SEIJAS, E. (1991) Cerebrospinal fluid interleukin-1 beta (IL-1 beta) in Alzheimer’s disease and neurological disorders. Meth. Find Exp. Clin. Pharmacol. 13, 455–458. . FORLONI, G., DEMICHELI, F., GIORGI, S., BENDOTTI, C. & ANGERETTI, N. (1992) Expression of amyloid precursor protein mRNAs in endothelial, neuronal and glial cells: modulation by interleukin-1. Brain. Res. Mol. Brain Res. 16, 128–134. . DONNELLY, R.J., FRIEDHOFF, A.J., BEER, B., BLUME, A.J. & VITEK, M.P. (1990) Interleukin-1 stimulates the beta-amyloid precursor protein promoter. Cell Mol. Neurobiol. 10, 485–495. . GOLDGABER, D., HARRIS, H.W., HLA, T., MACIAG, T., DONNELLY, R.J., JACOBSEN, J.S., VITEK, M.P. & GAJDUSEK, D.C. (1989) Interleukin 1 regulates synthesis of amyloid beta-protein precursor mRNA in human endothelial cells. Proc. Natl Acad. Sci. USA 86, 7606–7610. . MATTSON, M.P. (1990) Excitatory amino acids, growth factors, and calcium: a teetertotter model for neural plasticity and degeneration. Adv. Exp. Med. Biol. 268, 211–220. . FREESE, A., FINKLESTEIN, S.P. & DIFIGLIA, M. (1992) Basic fibroblast growth factor protects striatal neurons in vitro from NMDA-receptor mediated excitotoxicity. Brain Res. 575, 351–355. . MATTSON, M.P. & CHENG, B. (1993) Growth factors protect neurons against excitotoxic ischemic damage by stabilizing calcium homeostasis. Stroke 24, 1136–1140. . ZHONG, L.T., KANE D.J. & BREDESEN, D.E. (1993) BCL-2 Blocks Glutamate Toxicity in Neural Cell Lines. Mol. Brain Res. 19, 353–355. . KIHARA, S., SHIRAISHI, T., NAKAGAWA, S., TODA, K. & TABUCHI, K. (1994) Visualization of DNA double strand breaks in the gerbil hippocampal CA1 following transient ischemia. Neurosci. Lett. 175, 133–136. . LINNIK, M.D., ZOBRIST, R.H. & HATFIELD, M.D. (1993) Evidence supporting a role for programmed cell death in focal cerebral ischemia in rats. Stroke 24, 2002–2008. . TOMINAGA, T., KURE, S., NARISAWA, K. & YOSHIMOTO, T. (1993) Endonuclease activation following focal ischemic injury in the rat brain. Brain Res 608, 21–26. . MARTIN, D.P., SCHMIDT, R.E., DISTEFANO, P.S., LOWRY, O.H., CARTER, J. G. & JOHNSON, E.M. (1988) Inhibitors of protein synthesis and RNA synthesis prevent neuronal cell death caused by nerve growth factor deprivation. J. Cell Biol. 106, 829–844. . DIPASQUALE, B., MARINI, A.M. & YOULE, R.J. (1991) Apoptosis and DNA degradation induced by lmethyl-4-phenylpyridinium in neurons. Biochem. Biophys. Res. Commun. 181, 1442–1448. . EDWARDS, S.N., BUCKMASTER, A.E. & TOLKOVSKY, A.M. (1991) The death programme in cultured sympathetic neurones can be suppressed at the posttranslational level by nerve growth factor, cyclic AMP, and depolarization. J. Neurochem. 57, 2140– 2143. . FAHRBACH, S.E., CHOI, M.K. & TRUMAN, J.W. (1994) Inhibitory effects of actinomycin D and cycloheximide on neuronal death in adult Manduca sexta. J. Neurobiol. 25, 59–69. . LINNIK, M.D., HATFIELD, M.D., SWOPE, M.D. & AHMED, N.K. (1993) Induction of programmed cell death in a dorsal root ganglia×neuroblastoma cell line. J. Neurobiol. 24, 433–446. . MARTIN, D.P., ITO, A., HORIGOME, K., LAMPE, P.A. & JOHNSON, E.J. (1992) Biochemical characterization of programmed cell death in NGF-deprived sympathetic neurons. J. Neurobiol. 23, 1205–1220. . SVENDSEN, C.N., KEW, J.N., STALEY, K. & SOFRONIEW, M.V. (1994) Death of developing septal cholinergic neurons following NGF withdrawal in vitro: protection by protein synthesis inhibition. J. Neurosci. 14, 75–87. . WEEKS, J.C., DAVIDSON, S.K. & DEBU, B.H. (1993) Effects of a protein synthesis inhibitor on the hormonally mediated regression and death of motoneurons in the tobacco hornworm, Manduca sexta. J. Neurobiol. 24, 125–140. . RATAN, R.R., MURPHY, T.H. & BARABAN, J.M. (1994) Macromolecular synthesis inhibitors prevent oxidative stress-induced apoptosis in embryonic cortical neurons by shunting cysteine from protein synthesis to glutathione. J. Neurosci. 14, 4385–4392.
CONTROL OF CELL DEATH IN THE NERVOUS SYSTEM
125
126
127
128
129
130
131
132 133
134 135 136 137 138
139
140 141
263
. HERON, A., POLLARD, H., DESSI, F., MOREAU, J., LASBENNES, F., BENARI, Y. & CHARRIAUTMARLANGUE, C. (1993) Regional variability in DNA fragmentation after global ischemia evidenced be combined histological and gel electrophoresis observations in the rat brain. J. Neurochem. 61, 1973–1976. . DRAGUNOW, M., BEILHARZ, E. SIRIMANNE, E., LAWLOR, P., WILLIAMS, C., BRAVO, R. & GLUCKMAN, P. (1994) Immediate-early gene protein expression in neurons undergoing delayed death, but not necrosis, following hypoxic-ischaemic injury to the young rat brain Mol. Brain Res. 25, 19–33. . LIU, T., MCDONNELL, P.C., YOUNG, P.R., WHITE, R.F., SIREN, A.L., HALLENBECK, J.M., BARONE, F.C. & FEURERSTEIN, G.Z. (1993) Interleukin-1 beta messenger RNA expression in ischematic rat cortex. Stroke 24, 1746–1751. . ROSEN, D.R., SIDDIQUE, T., PATTERSON, D. FIGLEWICZ, D.A. SAPP, P. HENTATI, A., DONALDSON, D., GOTO, J., O’REGAN, J.P., DENG, H.X. et al. (1993) Mutations in Cu/Zn superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis [published erratum in Nature 1993 Jul 22; 364(6435): 362] [see comments]. Nature 362, 59–62. . ROBBERECHT, W., SAPP, P., VIAENE, M.K., ROSEN, D., MCKENNA, Y.D., HAINES, J., HORVITZ, R., THEYS, P. & BROWN, R.J. (1994) Cu/Zn superoxide dismutase activity in familial and sporadic amyotrophic lateral sclerosis. J. Neurochem. 62, 384–387. . GURNEY, M.E., PU, H., CHIU, A.Y., DAL, C.M., POLCHOW, C.Y., ALEXANDER, D.D., CALIENDO, J., HENTATI, A., KWON, Y.W., DENG, H.X. et al. (1994) Motor neuron degeneration in mice that express a human Cu,Zn superoxide dismutase mutation [see comments]. Science 264, 1772–1775. . DENG, H.X., HENTATI, A., TAINER, J.A., IQBAL, Z., CAYABYAB, A., HUNG, W.Y., GETZOFF, E.D., HU, P., HERZFELDT, B., ROOS, R.P. et al. (1993) Amyotrophic lateral sclerosis and structural defects in Cu,Zn superoxide dismutase [see comments]. Science 261, 1047–1051. . CHAN, P.H., CHU, L., CHEN, S.F., CARLSON, E.J. & EPSTEIN, C.J. (1990) Reduced neurotoxicity in transgenic mice overexpressing human copper-zinc-superoxide dismutase. Stroke 21, suppl. III, 80–82. . CHAN, P.H., CHU, L., CHEN, S.F., CARLSON, E.J. & EPSTEIN, C.J. (1990) Attenuation of glutamateinduced neuronal swelling and toxicity in transgenic mice overexpressing human CuZn-superoxide dismutase. Acta Neurochir. Suppl. Wien 51, 245–247. . MICHIKAWA, M., LIM, K.T., MCLARNON, J.G. & KIM, S.U. (1994) Oxygen radical-induced neurotoxicity in spinal cord neuron cultures. J. Neurosci. Res. 37, 62–70. . TROY, C.M., & SHELANSKI, M.L. (1994) Down-regulation of copper/zinc superoxide dismutase causes apoptotic death in PC 12 neuronal cells. Proc. Natl Acad. Sci. USA 91, 6384–6387. . ROTHSTEIN, J.D. BRISTOL, L.A., HOSLER, B., BROWN, R.J. & KUNCL, R.W. (1994) Chronic inhibition of superoxide dimutase produces apoptotic death of spinal neurons. Proc. Natl Acad. Sci. USA 91, 4155–4159. . PAN, Z. & PEREZ-POLO, R. (1993) Role of nerve growth factor in oxidant homeostasis: glutathione metabolism. J. Neurochem. 61, 17134–1721. . JACKSON, G.R., APFFEL, L., WERRBACH, P.K. & PEREZ, P.J. (1990) Role of nerve growth factor in oxidant-antioxidant balance and neuronal injury. I. Stimulation of hydrogen peroxide resistance. J. Neurosci. Res. 25, 360–368. . TANAKA, M., SOTOMATSU, A., YOSHIDA, T., HIRAI, S. & NISHIDA, A. (1994) Detection of superoxide production by activated microglia using a sensitive and specific chemiluminescence assay and microgliamediated PC12h cell death. J. Neurochem. 63, 266–270. . ELLIS, H.M. & HORVITZ, H.R. (1986) Genetic control of programmed cell death in the nematode C. elegans. Cell 44, 817–829. . HENGARTER, M.O., ELLIS, R.E. & HORVITZ, H.R. (1992) C. elegans gene ced-9 protects cells from programmed cell death. Nature 356, 494–499.
Index
A1 78, 142, 152 abl gene 71 81 acd2 194 acidophilic bodies 2 ACTH 4 actinomycin D 103, 174 adenylate cyclase 43, 44 ADPRT (ADP-ribosyltransferase) 24, 29, 56 adrenal cortex 4, 10 adriamycin 134, 270 ageing 14, 21, 55 AIDS 14, 39, 45, 55 AKA ICH-1 gene 73 albicotin 3 alkylating agents 47 allograft rejection 10 ALS 289, 290 Alzheimer’s disease 55, 285–7 aminobenzamide 29 AMP (cyclic) 43, 47 androgen deprivation 54 androgens 205, 207–9, 212 anoxia 195 Antheraea polyphemus 174 anti-CD3 77 antioxidant pathway 141, 142 antipain 23 AP1 79 APO-1 13, 43 apoptotic bodies 2, 6–10, 12, 13, 21, 22, 30, 31, 41, 43, 52, 54 apterous gene 176 Ara-C (cytosine arabinoside) 134, 270 Arabidopsis 194
ascialic glycoprotein receptor 53 asp-ase 23 ATA (aurintricarboxylic acid) 26, 27, 50 ATP 30, 43, 47, 63, 138, 149, 195 autoimmune (effects, disease) 14, 30, 39, 45, 55, 56, 84 Bad 142, 145, 152 BAG-1 146, 147, 152 bax gene 47, 65, 71, 78, 79, 86, 103, 142, 144–6, 150–5 B-cells 13, 64, 71, 77, 83, 84, 87, 129, 139, 205, 231, 232, 261 bcl-1 gene 128 bcl-2 gene 14, 43, 45–7, 54, 65, 66, 71, 73, 77–9, 82–7, 97, 103, 108, 127–56, 172, 177, 182, 192, 193, 195, 196, 211, 230 bcl-2 237, 240, 265–73, 283, 284, 288, 289 bcl-x genes 65, 78, 79, 86, 142, 144–7, 150–3 Bcl-Y 142, 145, 152, 153 BDNF (brain derived neurotrophic factor) 278, 280, 287 BHRF1 viral protein 78, 84, 86, 127, 142, 144, 149 BLT-esterase 23 BM 13674 cells 9, 12 bmi-1 gene 81 BNDF (brain derived neutrophic factor) 137 bombesin 75, 76 Botryllus 189 breast cancer 271 budding 9, 21 Burkitt’s lymphoma 9, 12, 66 Caenorhabditis elegans 13, 42, 46, 64, 65, 72, 74, 78, 86, 127, 142, 171, 172, 176, 177, 180, 191, 196, 241, 242, 282 calbindin-D 150 264
INDEX
calcium (ions, ionophore) 13, 23, 47, 48, 54, 55, 63–5, 83, 100, 137, 138, 141, 285, 288 calpain 23, 24 camptothecin 134 carbohydrates 13, 30 carcinogene 13 castration 8, 10 cathepsin-D 23 CD95 64 cdk-2 kinase 140 ced genes 42, 46, 64, 65, 73, 74, 78, 83, 85, 86, 127, 142, 172, 177, 191–3, 196, 282, 285 cell cycle 45, 47 cellular condensation 13 cellular turnover 21 CG (chorionic gonadotropin) 208 chaotropic agents 31, 52 chemotherapy 1, 10, 39, 54, 78 chlorodeoxyadenosine 134 CHOhr gene 82, 83 chromatin (see nuclear chromatin) Cip1 (cdk-interacting protein) 65 cisplatin 134 Civatte bodies 2 CL granule protease 23 CL-mediated cytotoxicity 22, 23 CMF treatment 271 C-nitroso compounds 29 CNTF (ciliary neurotrophic factor) 137, 280–2 complement-mediated lysis 21 cortical cells 4 Councilman bodies 2 CrmA 65, 73, 74, 83, 86, 283 crumbs 178 crypt lumen 7 cyclin-cdk (cdc) 47 cycloheximide 23, 174 cyclophilin 26, 27 cyclophosphamide 134 cyclosporin 26 cysteine-protease 24 cytokine 39, 45, 77, 80 cytosine arabinoside 7 cytosine arabinoside (see Ara-C) cytotoxic cells 22, 64 cytotoxic T lymphocytes (CTL) 196, 237, 238 degenerative neural disease 39 Dexamethasone (DEX) see glucocorticoid
265
DfH99 embryos 178 DG (diacyl glycerol) 43, 44, 47 Dictyostelium 190 DmTIAR (see rox 8) DNA damaging agents 47, 65, 78, 87, 99, 102, 108, 111, 112, 117, 134, 137, 138, 141, 153 DNA fragmentation 1, 11, 22–5, 27, 28, 40, 42, 47, 48, 55, 56, 62, 181, 190, 196, 206, 208, 288 death genes 14 endonuclease 11, 14, 23–5, 28, 40, 42, 43, 48, 50, 51, 54, 55, 180, 202, 211 ladder patterns 11, 12, 48, 50, 202 linker regions 11, 48 DNA repair 98 DNA replication 47 DNase 26, 27, 48–51, 55 Drosophila mutants 65, 171–82 drug resistance 114, 127, 133, 134, 271 duanomycin 134 e14 locus 190 E1A gene 46, 83, 105–8, 113, 115, 116, 138 E1B gene 54, 83, 105, 106, 108, 127, 142–4, 149, 155 E2F 106, 115 E6 108, 110 E7 106, 108, 110 EBV (Epstein-Barr virus) 47, 54, 83, 127, 144, 196 ecdysteroid hormone (ecdysone) 174, 194 echinus gene 176, 177 EGF (epidermal growth factor) 29, 75, 76, 208, 210, 214, 271 elf-1 gene 177 embryogenesis 5, 8, 10, 173, 177, 180, 193 endomytrium 8 endonuclease (see DNA endonuclease) endoplasmic reticulum 41, 44, 77, 141 Escherichia coli 190 estrogen receptor 271 estrogens 205, 207–9, 211–13 etoposide (VP-16) 23, 62, 65, 78, 80, 134 excitotoxicity 287, 288 5FU (5-fluoro-deoxyuridine) 134 Fas/APO-1 (CD95) 43, 45, 47, 64, 66, 71, 83–5, 138, 141, 147, 235–7 FGF (fibroblast growth factor) 75, 76, 208, 214, 281, 287, 288 fludarabine 134 fos 43, 79, 85, 288
266
INDEX
fragmentin (Granzyme B) 23, 196, 238 FSH (follicle stimulating hormone) 208, 210–13 1 34.5 gene 82 G-actin 50 GADD45 gene 116 gene expression 13 gld mice 84 glucocorticoids 11, 23, 24, 28, 29, 45–8, 55, 61–65, 77, 100, 101, 116, 133, 137, 139, 141, 205 glutamyl-bis-spermidine 52 glutamylysine isopeptide 31, 52 glutathione 134 glutathione peroxidase 141 glycan 52 glycoprotein 52 GMP (cyclic) 43 GnRH peptide 210, 214 gonadotropins 205–7, 211, 213 Granzyme B (fragmentin) 23, 196, 238 granzymes (CL granule proteases) 23 growth factor 13 growth factors 208–10 GTP-binding protein 44 guanylate cyclase 43, 44 heat shock 138, 212 HeLa cells 7 heliotrine 3 histones 24 HIV (human immunodeficiency virus) 45 HL-60 cells 23, 29, 45 hormones 1, 10, 21, 24, 39, 42, 174, 201–14 hyperthermia 41 hypoxia 41, 54, 105 iap gene 81, 82 ICE (interleukin-1 converting enzyme) 64, 65, 73, 74, 83, 85, 86, 138, 192, 196, 282, 283, 289 Ich genes 73, 74, 86, 283 IGF (insulin-like growth factor) 75–7, 210–14 IL (interleukin) 75, 77, 79, 83, 100, 140, 144, 153, 172, 192, 193, 211, 214, 256, 257, 281–3 immune system 11–13, 21, 22, 223–44 inflammation 2, 3, 10, 22, 41, 54, 193 influenza virus 45 intestinal crypt 7, 11 iodoacetamide 23 ionic channels 43
IP3 (inositol-1.4.5-triphosphate) 43, 44, 47 irradiation 1, 11, 23, 39, 45, 47–9, 54, 55, 62, 65, 66, 77, 80, 100, 101, 102, 113, 116, 137–9, 173, 177, 178, 181, 194 ischemia 3, 41, 288 jun 43, 79, 288 juvenile hormone 174 K cells 13 karyolytic bodies 2 Kr (Krupple) 177 Kupffer cells 3 large T 106, 110, 155 leupeptin 24 LFA-1 228 LH (luteotropic hormone) 208, 210, 212, 213 lipids 30 LMP1 protein 83 LMW5 HL viral protein 78, 142 lp2 gene 45, 237 lpr mice 84 lysis 22 lysozomes 3, 6, 8, 24, 27 macrophage 7, 9, 21, 30, 41, 43, 180, 181 malformation 39 Manduca sexta 174, 179, 194 MARs (matrix attachment regions) 28 Mas 70p 144 Max 74 MC5–5 253 mcd gene 176 mcl-1 gene 142, 145, 150–3 Mcl-1 78, 144 Mdm-2 108 mdr gene 134 met-ase 23 metamorphosis 10 metamylocytes 23 methotrexate 134, 270 methylnicotinomide 29 MHC 224–7 mitochondria 21, 41, 47, 77, 140, 141, 150, 285 mitosis 5, 14, 54, 66 mitoxantrone 134, 270 myb gene 138
INDEX
myc 14, 43, 46, 47, 54, 65, 66, 71, 72, 74, 75, 77, 78, 81, 83, 85–7, 97, 108, 115, 116, 131, 138, 140, 227 necrosis 2, 3, 5, 11, 21–3, 40, 41, 47, 55, 61, 105, 139, 172, 195 nedd-2 gene 73, 85, 86, 192 N-ethylmaleimide 23 neurodegenerative disease 14 neurotrophic factors 45 NGF (nerve growth factor) 45, 83, 84, 137, 205, 235, 236, 240, 278, 280 Nip 146, 149, 150, 152 nitrogen mustard 134 NK (natural killer) cells 13, 22, 196, 237 NMDA receptor 290 Notch gene 176 NT (neurotrophin) 278 nuc-1 gene 180 Nuc18 26, 27, 48, 51 nuclear envelope 6 nuclear fragmentation 6 nuclear chromatin 6, 7, 21, 26, 41, 47, 48, 62, 172 nuclease (see DNA endonuclease) nucleosomes 11, 40, 62 nur 77 gene 64 oncogenes 13, 15, 74, 85, 105, 116, 128 ovary 202, 203, 211 p-chloromercuribenzenesulfate 23 P-glycoprotein 134 p105rb 80, 81 p19E1B viral protein 71, 78, 82–4, 86, 105, 106, 108 p21WAF-1/Cip1 79, 103, 116 p35 viral protein 78, 81–3, 177, 196 p53 14, 15, 43, 46, 47, 54, 62, 65, 66, 71, 79, 80, 81, 83, 85, 86, 97–117, 132, 136, 137, 140, 153–5, 194, 195, 227, 266–73, 285 p55E1B viral protein 81, 105, 106 parenchymal cells 10 PARP (poly (ADP-ribose) polymerase) 73 PDGF (platelet derived growth factor) 75–7 perform 22, 23, 84, 138, 196 peroxidasin 180, 181 phagocytosis 3, 6–10, 12, 13, 30, 46, 52, 53, 62, 105, 173, 193, 195, 240–2 phagolysosomes 6, 9 phenylmethylsulfonyl fluoride 23 phorbol ester 100
267
phosphatidylethanolamine 30 phosphatidylinositol-4, 5-diphosphate 44 phosphatidylserine (receptor) 30, 53 phospholipase 43, 44 phospholipids 30 phosphoramidone 23 pml gene 81 PMSG (pregnant mare serum gonadotropin) 208 polyamine 25 polyubiquitin 179 pRB 86, 106 prednisone 4 preneoplastic foci 13 prostate 8, 10, 54, 85, 87, 205, 212, 272 protease 22, 23, 43, 54, 85, 138, 172 protein kinase 44 protein synthesis 11, 13, 43, 45, 47, 61, 66, 75 protein tyrosine kinase 44 protein x (see peroxidasin) proteolysis 22, 23, 29, 31 Px2 ion channel 63 radiation (see irradiation) Raf-1 146–9, 152 RAG (recombinase activating gene) mice 225 ras (R-Ras) 78, 97, 105, 108, 113, 116, 138, 146–9, 152 rb gene (see also pRB) 43, 47, 80, 106, 109, 110, 115 reaper gene (rpr) 65, 177–9, 181 ret gene 81 retinoids 134 Rho-GAP protein 149 RNA (mRNA, RNA synthesis) 11, 13, 24, 31, 43, 45, 47, 61, 66, 82 RNase 24, 29 ROS (reactive oxygen species) 78 ROS 290 rotamase 26 roughest gene 176 rox 8 (DmTIAR) 179 RP-2 63 RP-8 63 RU486 62 sarcoma 7 scavenger receptors 180, 181 SCID (severe combined immunodeficiency) mice 225, 227 secondary necrosis 9 serine protease 23, 24
268
INDEX
SGP-2 (sulphated glycoprotein) 54 shrinkage necrosis 3–5 SLE (systematic lupus erythematosus) 45, 241 SOD (superoxide dismutase) 141, 281, 289 spermine 25 sphingomyelin 30, 64 staurosporine 147 steroid receptor 63 Sunburn cells 2 T-cell hybridoma 23, 66, 74, 84 T-cell (receptor) 10, 13, 22–4, 45, 48, 64, 71, 77, 84, 137, 139, 223–44 T121 (SV40 Large T antigen mutant) 80 tadpole tail 10, 189 Tax viral protein 138 taxol 134 TCR 24, 223–8, 232, 243 testis 202, 204–6, 212–14 testosterone (see androgens) TGF (transforming growth factor) 29, 137, 138, 208, 210, 214, 287 thapsigargin 141 thrombospondin (receptor) 30, 53 thy-1 24 tingible bodies 2 TLCK (N-tosyl-L lysyl-chloromethylketone) 22 TNF (tumour necrosis factor) 23, 43, 45, 64, 66, 71, 74, 83, 84, 138, 144, 235, 236, 285 topoisomerase inhibitor 23, 62, 78 TPCK (N-tosyl-L-phenylalanine-chloromethylketone) 23 transglutaminase 13, 30–32, 43, 52 translocase 30 Trk receptor 279, 280 TRPM-2 (testosterone-repressed prostate message) 54 tumour regression 8 tumour suppression genes 13–15, 66, 74, 79, 80, 85, 98, 99, 103, 132 tumour promoters 13, 54 TUNEL labeling 172, 177 UbcD1 (ubiquitin-conjugating enzyme) 180 VG16 viral protein vincristine 134 viral infection 13, 41, 47, 54, 81–3, 138, 194, 196 vitronectin (receptor) 30, 53 Volvox cateri 189 VP3 protein 195
WAF1 (wild-type p53 activated fragment) 65 zeiosis 62 zinc ions 48, 50, 55