-

--

I bnwokgy,

r

-

The Weiizmann

ABSTRACT Alterations in the gene encoding the cellular pSS protein are perhaps the most frequent type of genetic lesions in human cancer. At the heart of these alterations is the abrogation of the tumor suppressor activity of the normal p53. In many cases this is achieved through point mutations in p53, which often result in pronounced conformational changes. Such mutant polypeptides, which tend to accumulate to high levels in cancer cells, are believed to exert a dominant negative effect over coexpressed normal p53. Extensive research on p.53, especially in the course of the last 3 years, has already provided much insight into the biological and biochemical mechanisms that underlie its capacity to act as a potent tumor suppressor. There are now many indications that p53 may play a central role in the control of cell proliferation, cell survival, and differentiation. Nevertheless, despite the purported importance of p53 for such crucial processes, mice can develop apparently without any defect in the total absence of p53. This raises the possibility that p53 may become critically limiting only when normal growth control is lost.-Oren, M. pSS: the ultimate tumor suppressor gene? FASEBJ. 6: 3169-3176; 1992. Key Words: p53 tumor mutants tumorigenesis

suppressor

genes

dominant

negative

Inst1tut

been resolved even to this day (for historical perspectives, see ref 7). This review will present some of the more recent developments in the study of p53, with special emphasis on molecular mechanisms.

BIOLOGICAL

ACTIVITIES

OF wt p53

When trying to understand why wt p53 expression is so often abrogated during tumor development, one first has to elucidate the activities exerted by this protein when it is still present in the cell. The main approach taken to that end was the reconstitution of wt p53 expression in transformed cells that had previously lost it. So far, all these experiments have reintroduced wt p53 under various heterologous promoters; in most cases these promoters were much more potent than the relatively weak p53 gene promoter. Consequently, the protein levels attained were often well above those seen in the corresponding nontransformed cells, although in some cases (8-10) the extent of overexpression was rather modest. Thus, one must still be cautious when extrapolating from these systems to the normal functions of p53. Nevertheless, these studies have yielded interesting findings, which have played a pivotal role in formulating our current ideas on what p53 may be doing. Often investigators found it practically impossible to express wt p53 stably in transformed cells. Conceivably, the

continuous

presence

of the protein

in these cells was vigor-

IN 1989, VOGELSTEIN AND CO-WORKERS (1) reported that the p53 gene is often altered in colorectal cancer. Since that seminal report, there has been a surge of studies addressing the status of p53 in human tumors. A large body of evidence suggests that the p53 gene may well be the most frequent target for genetic alterations in human cancer (for recent reviews, see refs 2-6). These alterations range from complete deletion of the gene, resulting in no p53 synthesis at all, to a variety of different point mutations that retain the synthesis of full-length, albeit mutant, p53 protein. In fact, missense point mutations are by far the most prevalent p53 gene alterations. Such mutations occur in almost every type of human cancer, often affecting a large proportion of individual tumors (2-6). In most cases, the mutations involve various domains of the protein that have been highly conserved through evolution. This is consistent with the idea that such mutations interfere with a basic feature of the protein that

ously selected against. This obstacle was overcome through the use of regulatable p53 expression vectors. In one set of experiments (10, 11), authentic wt p53 was placed under the dexamethazone-inducible MMTV promoter. An alternative approach took advantage of a temperature-sensitive (ts),,p53 mutant (12). The protein encoded by this mutant, p53” 135, possesses a wt p53 conformation at low temperatures but switches to a typical mutant conformation when the temperature is elevated (13). Hence, cells harboring this ts mutant exhibit wt p53 activity at 32.5#{176}C, but not at 37.5#{176}C. Stable cell lines carrying this ts mutant can be established at 37.5#{176}C, where the protein does not confer any growth disadvantage, and then shifted to 32.5#{176}C in order to activate wt p53 functions and monitor the resultant cellular consequences. More recently, it has also been possible to generate stable lines expressing wt p53 under constitutive promoters

is essential

low to be tolerated by the recipient cells (8, 9, 14). On the basis of these experimental systems, a number of p53-mediated activities have been identified. The most frequent observation was that wt p53 can induce a growth arrest, which occurs primarily in the Gi phase of the cell cycle

for its proper

biochemical

functioning.

Thus,

the

common denominator of all p53-related alterations is the loss of expression of the wild-type (wt)i p53. It is now widely accepted that wt p53 is a major tumor suppressor whose ongoing activity somehow serves to maintain normal growth control. The loss of this activity will then provide cells with a selective growth advantage and contribute toward the development of cancer. The present state of understanding about p53 was not reached easily. In fact, the history of p53 is fraught with riddles and apparent inconsistencies, some of which have not

0892-6638/92/0006-3169/$01

.50.

©

FASEB

as long as the resultant

levels of expression

were sufficiently

iAbbrevjations: wt, wild-type; ts, temperature-sensitive; MCK, muscle creatine kinase; PCNA, proliferating cell nuclear antigen; IMP-DH, IMP- dehydrogenase; mAbs, monoclonal antibodies; LFS, Li-Fraumeni syndrome.

3169

www.fasebj.org by Kaohsiung Medical University Library (163.15.154.53) on November 13, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber

(10, 12, 15-18). Hence, p53 appears capable of acting as a negative regulator of cell cycle progression. It is tempting to speculate that p53 may be involved in the signaling initiated

by growth-restricting

stimuli,

such as depletion

of growth

factors, exposure to growth inhibitory cytokines, etc. However, this still has to be formally demonstrated. More recently, wt p53 has been shown to induce two additional biological processes that may account for its tumor suppressor properties. In a murine p53-negative myeloid leukemic cell line, the activation of wt p53 leads to rapid cell death (19), with distinct characteristics of apoptosis (programmed cell death). In myeloid progenitor cells, apoptosis usually follows the withdrawal of pertinent survival factors, such as IL-3 (20). It is conceivable that p53 may participate in the chain of events initiated in the absence of limiting survival factors that eventually leads to the elimination of unnecessary cells. Alternatively, p53 may be required in order to make the cells dependent on such survival factors. In either case, the loss of p53 function may allow cells to survive illegitimately. This, in turn, could provide the cells with a strong selective advantage and facilitate the establishment of neoplastic cell populations in vivo. It is noteworthy that p53-mediated apoptosis is inhibited by IL-6 (19). Hence, in the presence of activated wt p53, IL-6 may be providing the necessary survival factor function. The observation that p53 can cause apoptosis raises the possibility that other tumor suppressors may also be involved in the control of cell death and cell survival. Another important manifestation of wt p53 activity is the induction of differentiation. This was demonstrated through the introduction of stable wt p53 expression into a transformed pre-B cell line. Unlike the parental p53-negative cells, the wt p53 transfectants could undergo partial differentiation in culture (9). Moreover, such transfectants also exhibited accelerated differentiation in vivo when injected into syngeneic mice (14). When the same p53-negative parental cells were transfected with a mutant p53, their ability to undergo differentiation in vivo was practically abolished (14). In all these cases, there was a clear inverse correlation between the ability of the cells to undergo differentiation and their tumorigenic potential. That these observations are physiologically relevant is suggested by the increase in p53 levels during the course of normal hematopoietic cell maturation (21). These findings support the notion that, at least in certain cell types, the loss of p53 may contribute to tumor progression by arresting the cells in a relatively immature, continuously self-renewing state. Finally, another clue to the normal role of p53 may be the demonstration that the emergence of immortalized fibroblastic cell lines in culture is strongly correlated with the acquisition of p53 mutations (22). Hence, wt p53 may also be involved in the control of cell senescence, its loss thereby promoting the establishment of continuously proliferating tumors.

BIOCHEMICAL The

biochemical

FUNCTIONS basis

OF

for the tumor

wt p53 has still not been established

p53 suppressor

activities

unequivocally.

of

Studies

indicate that wt p53 has to be present in the cell nucleus in order to exert its antiproliferative functions (16, 17, 23, 24). Hence, the pertinent biochemical activities of p53 are most probably associated with nuclear processes. Other studies, most using the replication of DNA viruses as a model system (25-29), are consistent with a role for wt p53 in the control

of DNA replication.

One recent

study proposes

that p53 may

be essential for orchestrating the block of DNA replication that follows DNA damage (30). It is tempting to speculate that p53 binds to specific sequence elements that control the initiation of cellular DNA replication and directly represses initiation. Although very attractive, this idea is based only on correlative evidence; it still remains to be proved that p53 binding can indeed have a direct effect on the replication of bound DNA.

An alternative, although not mutually exclusive, possibility is that pS3 may be a transcriptional regulator. In the last couple of years there has been significant tion of the effects of p53 on transcription.

progress in elucidaThis in no way im-

plies that transcriptional modulation by p53 is more important or more relevant than its putative role in the control of DNA replication. Rather, the plethora of reports concerning p53 as a transcriptional modulator simply reflect the fact that, at present, there are better tools to study mammalian transcription than there are for the study of mammalian DNA replication. The p53 protein possesses a potent transactivation domain that can function very efficiently when fused to a heterologous DNA binding domain; this property is abolished by many p53 mutations associated with neoplastic processes (31-35). Moreover, wt p53 can directly activate the transcription of the muscle creatine kinase (MCK) gene promoter (36). The ability of a protein to act as a promoterspecific transcriptional activator usually requires the selective binding of this protein to defined DNA elements. Indeed, wt p53 was found capable of sequence-specific binding to DNA (3 7-39). Once again, tumor-derived p53 mutants fail to do so. More recently, it could be demonstrated that transcriptional activation of the MCK promoter is indeed mediated through sequence-specific binding of wt p53 to a DNA element located upstream to the basal promoter (C. Prives, personal communication). Furthermore, wt p53 transactivated a synthetic promoter generated by placing a p53-binding site in front of a short segment of the c-fos promoter; most important, transactivation occurred also in

an in vitro

transcription

system.

SV4O large

T antigen,

which forms noncovalent complexes with p53, effectively blocked both DNA binding and transactivation (C. Prives, personal communication). Hence, apparently p53 can act as a bona fide sequence-specific transcription factor. In addition to its ability to positively regulate specific promoters, p53 can also act as a transcriptional repressor. Thus, wt p53 has been shown to down-regulate proliferating cell nuclear antigen (PCNA) mRNA (11) and to interfere with the induction of c-fos mRNA during serum stimulation

(40). Moreover,

excess wt but not mutant

p53 was found

to

repress efficiently transcription from many promoters, including several whose activity is usually correlated with cell proliferation or malignant progression (40-42). In recent experiments, wt p53 could repress transcription from several promoters, including that of the c-myc gene, in an in vitro transcription system (N. Ragimov, Y. Aloni, and M. Oren, unpublished results). Hence, transcriptional repression by p53 could represent a direct inhibitory effect of the protein on the transcription machinery rather than merely a secondary consequence of p53-mediated growth arrest. In most experiments demonstrating transcriptional repreSsion by wt p53, actual p53 concentrations were well above physiological. Such excess may result in broader effects than those exerted by p53 in nontransfected cells. Conceivably, some promoters may respond to much lower p53 concentrations than others; such promoters may represent physiological targets of wt p53. If one or more of these genes is essential for ongoing cell proliferation, shutoff may eventually elicit a growth arrest. Furthermore, it is possible

3170 Vol. 6 October 1992 OREN The FASEB Journal www.fasebj.org by Kaohsiung Medical University Library (163.15.154.53) on November 13, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber

that as long as the cell retains its normal properties, the amount of p53 present in it at any given time is always below that required to shut off even the most responsive promoter. However, once normal regulatory networks are disrupted, particular promoters may become more sensitive to the effects of wt p53. Such a state could, for instance, be brought about by the activation of certain oncogenes. This conjecture could potentially explain the fact that transfected wt p53 often exerts more pronounced effects on highly transformed cells than on their normal or less transformed counterparts (18, 43-45).

Of the p53-repressed

promoters

c-myc gene is of particular ts p53, a marked reduction

as 1-3 h after activation

studies

so far, that of the

interest. In cells harboring the in c-myc mRNA is seen as early

of wt p53 (D. Michael,

E. Yonish-

Rouach, and M. Oren, unpublished results). Recent findings pertaining to the relationship between c-myc and p53 expression are especially provocative. The p53 gene promoter contains a DNA element capable of binding proteins of the helixloop-helix family of transcription factors (46). This element can bind protein complexes containing the c-myc product; moreover, the transcriptional activity of the p53 promoter can be stimulated by c-myc overexpression (D. Reisman and V. Rotter, personal communication). Thus, whereas normal p53 levels are very low, events such as c-myc gene amplification or deregulation may augment these levels significantly. This could effectively establish a state of constitutive wt pS3 overexpression, which in turn may repress c-myc promoter activity. When regulation of the cellular c-myc gene promoter remains essentially normal (e.g., when the gene is amplified), this could eventually restore c-myc mRNA and protein levels to more physiological values. In this manner, mutual effects of the c-myc and p53 proteins could activate a negative feedback loop whereby a cell can monitor the inappropriate overexpression of c-myc and restrain it. On the other hand, once the c-myc promoter is freed from normal regulatory constrains (e.g., in cases of chromosomal rearrangements or retroviral insertion next to the c-myc gene), the levels of p53 and c-myc will both remain elevated. In such cases, other p53-responsive genes may now be turned off, eventually leading to a growth arrest. In either case, the inactivation of wt pS3 will fix the cells in a proliferative state driven by constitutive c-myc overexpression. Such a model predicts that loss of pS3 expression will frequently be observed in tumors involving c-myc gene activation. At least in the case of Burkitt’s lymphoma, the data support this conjecture (47). Another way to approach the molecular basis of p53 action is through the characterization of other proteins with which p53 interacts specifically. The existence of such proteins has recently been demonstrated in fibroblasts carrying the ts p53 mutant. At 32.5#{176}C, when such cells are growth-arrested as a result of the induction of wt-like p53 activity (12), at least three polypeptides can be found coprecipitating with p53 (48). Of those, the most abundant is one of 95 kDa (p95).

When

the same cells are maintained

at 37.5#{176}C, where

the

ts p53 is in a mutant conformation that fails to suppress cell proliferation (12), much less p95 is associated with pS3; the same holds true for non-ts p53 mutants at either tempera-

ture (48). The binding

of p53 to p95 is therefore

augmented

under conditions where p53 is involved in imposing a growth arrest, suggesting that this interaction may play a role in the normal biochemical activities of wt p53. Further elucidation of the nature of p95, as well as of other p53-associated polypeptides, could prove most informative. An important clue to possible biochemical roles of p53 is offered by a recent study (49). Using cells carrying a temperature-inducible wt p53, it was found that increased

p53 activity correlated with a marked decrease in the activity of the enzyme IMP-dehydrogenase (IMP-DH). Furthermore, the antiproliferative effects of wt p53 could be effectively counteracted by providing the cells with xanthosine, a metabolite whose addition bypasses the biochemical block imposed by the deficiency in IMP-DH activity. These observations raise the intriguing possibility that wt p53 may specifically target guanine nucleotide biosynthesis in which IMP-DH is a rate-limiting factor. Given the central role played by guanine nucleotides in many signal transduction pathways, this could make p53 a master regulator of growthrelated signaling pathways. One attractive possibility is that the IMP-DH promoter may be exquisitely responsive to wt p53, thus representing a true in vivo target for transcriptional repression by p53. Full appreciation of the significance of this findings must await further experimental work.

CONFORMATIONAL p53 FUNCTION

REGULATION

OF

The introduction of conformation-specific monoclonal antibodies (mAbs) to p53 (50) has been a major blessing in the study of this protein. With the aid of one such mAb, PAb246, wt p53 was shown to be typically maintained in a particular conformation that has been accordingly designated the wildtype conformation. Many tumor-derived mutants of p53 exhibit a loss of reactivity with PAb246. Instead, they now react with another conformation-specific mAb, PAb24O (50). These findings imply that the mutations, many of which occur at a considerable distance away from the epitopes of the two mAbs, induce a global conformational switch in the p53 protein. This switch is correlated with the loss of the ability to exert antiproliferative effects (51). On the other hand, proteins with such “mutant” conformation (PAb240, PAb246) exhibit a pronounced oncogenic potential in vitro (52). A number of findings suggest that wt p53 may not always be locked in the wild-type conformation. One hour after serum stimulation of mouse fibroblasts carrying apparently wt p53, the majority of the protein can be found in the mutant conformation (51). Furthermore, the relative preponderance of ts mutations of p53 (12, 53) Suggests that the protein may possess an inherent tendency to switch between wildtype and mutant conformations. This has led Milner (51) to propose a conformational hypothesis for the regulation of wt p53 activity. The hypothesis assumes that wt p53 is in fact naturally poised to switch between the two alternative conformations. Signals such as exposure to growth factors induce it into the mutant (“promoter”) conformation, inactivating the antiproliferative capacity of wt p53 and allowing cell proliferation to proceed. On the other hand, under growth restrictive conditions, the protein will maintain its wild-type (“suppressor”) conformation, and block cell cycle progression. The conformational hypothesis is also consistent with studies of the subcellular localization of p53. The ability of pS3 to localize to the nucleus is a prerequisite for its capacity to exert antiproliferative effects (24). Yet, at least in serumstimulated fibroblasts, p53 spends a major part of the cell cycle outside the nucleus, entering it only in the S phase (54). It is tempting to speculate that in proliferating cells p53 is actively kept out of the nucleus during the rest of the cell cycle. This could be achieved, for instance, by masking its nuclear localization signal through interactions with other proteins or by increasing its affinity for a putative cytoplasmic anchor (17). Both these mechanisms could potentially be mediated through the conformational switch from a suppressor to a

p53 3171 www.fasebj.org by Kaohsiung Medical University Library (163.15.154.53) on November 13, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber

Ichor

C

#{149}l’

Growth-restrictive

rowth-sthnulatory

LO

conditions conditions

Positive signal(?)

Figure 1. Simplified version of the conformational switch model and its implications for cell cycle control. Depicted schematically is a normal cell, expressing wt p53. When the cell receives a signal for proliferation, p53 is proposed to switch temporarily into a mutant conformation. This may be coupled with the induction of a translocation into the cytoplasm (or inhibition of entry into the nucleus). At this stage, p53 may interact specifically with a putative cytoplasmic “anchor” Although this may merely prevent the nuclear activity of p53, it is formally possible that a positive growth-stimulatory signal is generated by this cytoplasmic p53; so far, however, there is no experimental evidence to support this possibility. When the cell is subjected to growth-restrictive conditions, p53 may adopt the active wild-type conformation. This, in turn, will be reflected in a tight interaction with putative nuclear targets, leading to the generation of an inhibitory signal. Signals of the latter type are believed to underlie the ability of wt p53 to act as a tumor suppressor. C, cytoplasm; N, nucleus. This picture is based, in part, on the model proposed by Milner (48).

promoter mode (Fig. 1). Confirmation of this model will require additional experimental data, particularly on the distribution of wt p53 in resting and continuously cycling (rather than serum-stimulated) nontransformed cells.

ONCOGENIC THE

PROPERTIES

DOMINANT

NEGATIVE

OF p53: MODEL

In experimental systems using cultured cells, transfected mutant p53 transforms efficiently, with no apparent need for the loss of the endogenous wt p53 alleles. Moreover, where the status of the endogenous alleles has been documented, they were found to remain nonmutated (16). These observations can easily be explained by invoking a dominant negative mechanism for the action of mutant p53. According to this hypothesis the mutant protein, although devoid of any biochemical activity of its own, will interfere with the function of coexpressed wt p53 and render it practically ineffective. The net effect of the mutant p53 will thus be a major reduction in cellular wt p53 activity. There are many ways in which a dominant negative effect may be exerted, including competition for targets and a direct interaction between the mutant and wt molecules (52). The latter possibility is particularly relevant in the case of p53, as mutant p53 can form tight complexes with the endogenous wt p53 in transformed and immortalized cells (see ref 2). Similar complexes can also form when both polypeptides are made in an in vitro translation system (55). In such a cell-free system, it was shown that the wt p53 polypeptide is forced to adopt the aberrant conformation of the mutant partner (55). This model is schematically illustrated in Fig. 2. The figure fur-

ther

indicates

that when

the final association

product

is a

dimer, an equimolar concentration of wt and mutant p53 should already decrease total wt p53 activity by 75%. If, as suggested recently, p53 actually assembles into tetrainers and even larger forms (56; C. Prives, personal communication), then the residual wt pSS activity should be even lower. The dominant negative model implies that the emergence of a mutation in one p53 allele, with the other allele still giving rise to wt protein, will confer a strong selective growth advantage upon the cell. Such population of p53-heterozygous cells is expected to expand rather rapidly, thereby generating

a large target pool for a second hit which will inactivate the remaining wt allele. One may further speculate that the partial deficiency in p53 function, conferred by the mutation in the first allele, predisposes the cell towards increased genomic instability and hence greatly increases the chance of the remaining allele to be affected. The relevance of the dominant negative model to “reallife” situations still is not fully established. Mutant p53 expression plasmids exert oncogenic activity in transfection experiments. Hence, a sufficient excess of mutant p53 can effectively abrogate the activity of the small amounts of endogenous wt p53. In theory, such an excess of mutant over wt p53 should also prevail in many tumor cells carrying p53 mutations, as mutant p53 is often much more stable than wt p53 (57) and accumulates to very high levels (50). One could therefore expect that the mere appearance of a stable mutant p53 protein in the cell will render it effectively p53-negative, without any further changes. Yet, even though tumors coexpressing wt along with mutant p53 can indeed

be found

(58-60),

many

others

that carry one mutated

p53

allele exhibit a loss of the remaining wt allele (1-6). Why is there a need for the loss of the remaining wt allele in tumor cells expressing large amounts of mutant p53? Presumably,

even though the presence greatly reduce endogenous enough

uncomplexed

growth-inhibitory keeps

in mind

wt p53 molecules

effects. that

of mutant p53 in these cells may wt p53 activity, there may still be

complex

This

to exert

is particularly

formation

between

at least

some

likely if one mutant

and

wt p53 may possibly occur cotranslationally (55). In this case, the proportion of wt p53 complexed with mutant p53 will be determined only by the relative translation rates of the two; the subsequent accumulation of stable mutant p53 will be immaterial to this process. On the other hand, further deletion or mutation of the remaining allele will render the cells completely devoid of p53 activity and will provide an additional selective growth advantage. One potential argument against the biological the dominant negative mode of action of mutant

relevance of p53 comes

from indications that, in living cells, a 1:1 ratio between mutant and wt p53 may be insufficient for the effective abrogation of wt p53 activity. In fact, at least at first approximation, the induced expression of comparable amounts of mutant and wt p53 does not appear to confer protection against

3172 Vol. 6 October 1992 The FASEB Journal OREN www.fasebj.org by Kaohsiung Medical University Library (163.15.154.53) on November 13, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber

wtwt

p53

p53

ill,’

genes

polypeptides

Assembled

p53

dimers

%

of total dimers

Activity

wIM

11 11 RU [f 100

25

50

25

+

+

-

-

Figure

2. Theoretical model for the dominant negative action of p53. wt = wild type p53 gene and protein; M = mutant p53 gene and protein. The filled diamond represents a single-point mutation. The bent shape of the mutant polypeptide represents the conformational change associated with many of the tumor-derived p53 mutations. As suggested by Milner (48), this conformational change is also imposed on the wt polypeptide once it associates with a mutant counterpart. The model assumes that assembly is into dimers and that any dimer containing mutant subunits will be completely inactive. In reality, p53 probably assembles into tetramers and higher forms, rather than into dimers (see text). mutant

the growth inhibitory activities of the latter (8, 43, 44, 61). Rather, in transfection experiments, clear transforming effects of mutant p53 are elicited only when the mutant is in vast excess (52). This difficulty may be resolved by examining the total levels of wt p53 activity rather than the relative proportions of mutant versus wt protein. In cells in which similar amounts of wt and mutant p53 are produced simultaneously, it is plausible that even if most wt p53 molecules are functionally blocked by the mutant, a minor fraction of the former still remains active. When wt and mutant are both produced at physiological levels, the residual amount of active wt p53 may be very low, rendering the cells practically devoid of wt p53 activity. Such a situation, expected to occur in cells carrying a mutation in one p53 allele, will presumably contribute to oncogenesis. The situation is different when both mutant and wt p53 are expressed at exceedingly high levels because of the introduction of transfected DNA. In this case, even if only a small percentage of wt p53 molecules remains active, the total amount of p53 activity in the cell could be high enough to exert measurable antiproliferative and tumor inhibitory effects.

In conclusion, it would appear that the presence of one mutant allele may already contribute directly to tumor progression. First, it will clearly cause a partial reduction in p53 activity, even though it may be less than that predicted from Fig. 2. This may go unnoticed in short-term assays with cultured cells, but may be sufficient to allow for positive selection during long-term processes in the intact animal. Alternatively, some mutants may be capable of promoting neoplastic processes independently of their interaction with the endogenous wt p53 (52). Such apparently dominant positive mutants will also be selected for whether or not the wt allele is still retained. In either case, even though the emergence of one mutated allele may already contribute to tumorigenicity, the subsequent loss of the remaining wt allele will often still provide an additional advantage and will therefore be strongly selected for.

WHAT

IS p53

ESSENTIAL

FOR?

The early description of a cell line devoid of any detectable p53 expression (62) made it clear that cells could survive without pS3. However, the many indications that p53 plays a central role in the control of cell cycle progression (10, 12, 15, 16), and perhaps other key processes such as differentiation (9, 14) and programmed cell death (19), argued very strongly that p53 is indispensable for maintaining normal growth homeostasis. Implicit in this line of thinking was that proper p53 function should be essential for normal development. It was therefore most surprising to find that mice could go through complete, apparently unimpaired, development with no p53 whatsoever (63). This was achieved by homologous recombination, rendering both p53 alleles completely nonfunctional. Such “knockout” mice developed perfectly normally, with no detectable histopathological defects. Hence, a functional p53 gene is actually dispensable for embryonic development and apparently also for a substantial portion of postnatal life. Such unexpected findings could be explained by the existence of compensatory mechanisms, which allow development to proceed normally in the absence of p53. In other words, it is assumed that p53 does play a major role in the control of normal development. However, as with other central processes, there are probably alternative pathways that operate in parallel or that can be activated once p53 function becomes deficient. While this is a reasonable possibility, it is nevertheless surprising that not even minor quantitative defects could be observed in the development of the p53-negative mice (63). Therefore one cannot dismiss the other alternative, namely, that in vivo p53 is perhaps not so important for the control of the normal cell cycle and of normal development. Whereas the p53-deficient mice develop apparently normally at first, they do tend to succumb to spontaneous tumors at a relatively very young age (63). Hence, wt p53 expression is important in ensuring tumor-free survival, its loss making the animal prone to tumor development. One could argue that p53 is the ultimate tumor suppressor, as the only meaningful consequence of its elimination is the efficient induction of cancer. Is there any molecular mechanism for p53 function that will be consistent with this “pure tumor suppressor” conjecture? In fact, recent work has pointed out such a potential mechanism. Studies of cells exposed to radiation have demonstrated that an excellent correlation exists between the response of individual cell lines to this exposure and their p53 genotype. Whereas cells expressing wt p53 uniformly responded by arresting in G1, those lacking wt p53 expres-

3173 p53 www.fasebj.org by Kaohsiung Medical University Library (163.15.154.53) on November 13, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber

sion failed to do so (30). Furthermore, after exposure to radiation, a several-fold increase occurred in steady-state p53 protein levels in the wt p53 producers; no such effect was seen in cells expressing mutant p53 (30). A proper Gi arrest may be required in order to allow the excision and repair of damaged DNA. With it, when the cell subsequently resumes active proliferation and enters the S phase, it does so with damage-free DNA. Without it, S phase follows unabated and the damage to the genome is perpetuated. The data (30) suggest that wt pS3 may somehow be involved in sensing the DNA damage and imposing a transient G1 growth arrest. This hypothesis may also explain the excessive sensitivity of many transformed cells to the introduction of wt pS3 (18, 43-45, 61). It could be argued that these cells already carry a large number of DNA lesions, fixed in the genome in the absence of p53. Once wt p53 is reexpressed, ‘the lesions are properly “sensed” and a Gl arrest ensues. In nontransformed cells, on the other hand, presumably there are no such excessive DNA lesions. Consequently, the introduction of extra wt pS3, even in large excess, will fail to impose a growth arrest. This model predicts that so long as the DNA is intact, p53 expression may not be of great importance to the cell. Once DNA damage is incurred, normal pS3 function will be required for this damage to be eliminated. Thus, p53deficient mice may develop well because their DNA is initially intact. In the absence of p53, however, they start accumulating genetic lesions. Once a sufficient number of such lesions has been generated, overt tumorigenesis ensues. As genomic instability is believed to be a key factor in bringing about cancer, a role for p53 in maintaining genomic stability is highly compatible with its capacity as a tumor suppressor. At present this model is highly speculative and rests on little experimental data. Nevertheless, it does provide a novel and provocative way of looking at the relationship between pS3 and cancer.

CAN LOSS OF TUMORIGENIC

p53 FUNCTION PROCESS?

INITIATE

A

A major conceptual issue with any tumor-related genetic lesion is whether such lesion can serve as the primary event that triggers the tumorigenic process, or whether it only provides an accelerating factor once this process has already been launched. In the case of pS3, the suggestion that loss of p53 function may be an initiating event is mainly supported by studies with Li-Fraumeni patients and with the knockout mice. The human Li-Fraumeni syndrome (LFS) is a familial predisposition to cancer characterized by multiple earlyonset malignancies. It has been found that many, though not all, LFS families carry germ-line mutations in one allele of the p53 gene (64, 65). In the tumor, the expression of the remaining wt pS3 allele is also completely abrogated. Hence, loss of pS3 function may be sufficient to initiate tumorigenesis, as is suggested also by the very high tumor rates in the knockout mice (63). Nevertheless, one would expect that the development of tumors in LFS patients and knockout mice would both occur with a much earlier onset than is actually observed. One could therefore argue that the mere presence of a mutant pS3 allele is not sufficient to set off a neoplastic process, even when the remaining wt allele is deleted or mutated. Rather, the abrogation of p53 function may predispose the cell to further genetic damage. Once a sufficient number of additional hits is reached, overt tumor development will follow. A distinction between an initiating and a predisposing role for p53 in carcinogenesis could be important when one considers the potential use of pS3 alterations

3174

Vol. 6

October 1992

as markers of preneoplastic surprising if the answers tumors, depending on the starting cell population.

states. However, it may not be vary greatly among different precise nature of the particular

CAN pS3 BE EFFECTIVE IN REVERSING TUMORIGENIC PROCESSES? An inherent hope in the study of any tumor suppressor gene is that the reintroduction of a functional copy of the gene into tumor cells, deficient in the pertinent activity, will contribute to the reversal of the malignant phenotype. Success in achieving this goal may, of course, have important clinical implications. Many reports describe the successful reconstitution of wt p53 expression in transformed cells. The targets for these experiments were cells carrying pS3 mutations as well as those lacking p53 expression altogether. Often, the induction of wt p53 activity resulted in a complete cessation of cell proliferation, with cells typically accumulating in the GI phase of the cell cycle (10, 12, 15-18). In one case, excess wt p53 activity had a more dramatic outcome, leading to programmed cell death (19). In some instances, however, the effect of wt p53 on in vitro growth properties was rather mild, and cells continued to proliferate albeit often at a reduced rate (8, 9, 66). The precise reason for these differences in the response of cultured cells to wt p53 activity is still unclear. One probability is that it may reflect differences in the amount of introduced wt p53 activity, with larger excesses giving rise to more dramatic in vitro effects. More important, however, such induced wt p53 expressors exhibit major changes in their in vivo behavior. Thus, unlike their nonmanipulated tumorigenic predecessors, they fail to elicit malignant tumors (8, 9, 66). This loss of tumorigenicity may be mediated through the induction of differentiation (14) as well as through changes in the interactions with the host immune system. It is noteworthy that transformed cells carrying p53 lesions are often far more sensitive to the effects of extra wt p53 than are cells expressing endogenous wt p53 (18, 43-45). This gives rise to the hope that, with appropriate wt p53 expression vectors, one may be able to find conditions under which the in vivo introduction of such vectors will selectively affect only tumor cells. Although this is still a long way from the actual establishment of gene therapy through wt p53, it does make this option worth further serious investigation. CONCLUSION Despite the important progress made in the study of pS3 over the last few years, many major questions are still left open. On the one hand, there are now many indications that pS3 is involved in the control of central processes such as cell proliferation, differentiation, and cell survival. On the other hand, however, apparently normal mouse development can proceed perfectly well in the total absence of any p53. Does this imply that all the earlier functional attributes of p53 are based on experimental artifacts? Most probably, this is not the case. Rather, one would like to propose that each of the above processes is too crucial for multicellular organisms, let alone highly developed species, to be left under the control of a single gene. Each step in such processes is probably regulated by an intricate network of interacting genes, assuring that it will have a very low chance of going wrong. Thus, whereas p53 may be tightly linked to the control of cell cycle progression, differentiation, and apoptosis, there probably

The FASEB Journal

OREN

www.fasebj.org by Kaohsiung Medical University Library (163.15.154.53) on November 13, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber

are a number of other gene products that can substitute for it once it becomes defunct. Such compensatory mechanisms will then allow apparently normal development to take place. One could further speculate that in order for a cell to become fully malignant, several of these backup genes have to be sequentially inactivated. The accumulated genetic damage will then eventually reach a critical mass where no compensatory pathway is available anymore, and only then will cell proliferation, differentiation, or survival become truly deregulated. This is, of course, a greatly overgeneralized model, and it may well pertain only to a subset of cell types and cancer-related situations. Nevertheless, it is highly consistent with the frequent observation that in cells carrying a series of discrete genetic lesions, it is enough to regenerate the activity of only one of the defective genes in order to restore much more normal growth control. Despite these comforting explanations for the surprising results with the p53 knockout mice, one still cannot rule out the possibility that the major role of p53 indeed is to serve as a safeguard against abnormal situations rather than to regulate normal processes. The suggestion that pS3 prevents the perpetuation of DNA damage will be in line with such a conjecture. If p53 activity becomes crucial only under circumstances where genetic damage has already been induced, it will make this cellular protein qualify as the ultimate tumor suppressor. This issue, however, will probably remain unresolved until much better understanding is gained on the precise biochemical functions of p53 and on the mechanisms responsible for regulating each of those functions. Work in the author’s laboratory is supported in part by grant ROI CA 40099 from the National Cancer Institute, by the Minerva Foundation (Munich), and by The Forchheimer Center for Molecular Genetics. The author wishes to thank Dr. E. Stanbridge for stimulating discussions, and Drs. V. Rotter, C. Prives, and L. Donehower for sharing unpublished results. REFERENCES 1. Baker, S. J., Fearon, E. R., Nigro, J. M., Hamilton, S. R., Preisinger, A. C., Jessup, J. M., vanTuinen, P., Ledbetter, D. H., Barker, D. F., Nakamura, Y., White, R., and Vogelstein, B. (1989) Chromosome 17 deletions and p53 gene mutations in colorectal carcinomas. Science 244, 217-221 2. Levine, A. J., Momand, J., and Finlay, C. A. (1991) The p53 tumour suppressor gene. Nature (London) 351, 453-456 3. Hollstein, M., Sidransky, D., Vogelstein, B., and Harris, C. C. (1991) p53 mutations in human cancers. Science 253, 49-53 4. Oren, M. (1991) The role of p53 in neoplasia. In Bioche,nical and Molecular Aspects of Selected Cancers (Pretlow, T P., and Pretlow, T. G., eds) pp. 373-391, Academic, New York 5. Weinberg, R. A. (1991) Tumor suppressor genes. Science 254, 1138-1146 6. Caron de Fromentel, C., and Soussi, T. (1992) TP53 tumor suppressor gene: a model for investigating human mutagenesis. Genes, Chromosomes and Cancer 4, 1-15 7. Jenkins, J. R., and Sturzbecher, H. W. (1988) The p53 oncogene. In The Oncogene Handbook (Reddy, E. P., Skalka, A. M., and Curran, T., eds) pp. 403-423, Elsevier, Amsterdam 8. Chen, P-L., Chen, Y., Bookstein, R., and Lee, W. -H. (1990) Genetic mechanisms of tumor suppression by the human p53 gene. Science 250, 1576-1580 9. Shaulsky, G., Goldfinger, N., Peled, A., and Rotter, V. (1991) Involvement of wild-type p53 in pre-B cell differentiation in vitro. Proc. NaIL Acad. Sci. USA 88, 8982-8986 10. Mercer, W. E., Shields, M. T, Amin, M., Sauve, G. J., Apella, E., Romano, J. W., and Ullrich, S. J. (1990) Negative growth regulation in a glioblastoma cell line that conditionally expresses human wild-type p53. Proc. NaIL Acad. &i. USA 87, 6166-6170

11. Mercer, W. E., Shields, M. T., Lin, D., Apella, E., and 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. Nail. Acad. Sci. USA 88, 1958-1962 12. Michalovitz, D., Halevy, 0., and Oren, M. (1990) Conditional inhibition of transformation and of cell proliferation by a temperature-sensitive mutant of p53. Cell 62, 671-680 13. Milner, J., and Medcalf, E. A. (1990) Temperature dependent switching between wild type and mutant forms of p53-vall35. j Mol. Biol. 216, 481-484 14. Shaulsky, G., Goldfinger, N., and Rotter, V. (1991) Alterations in tumor development in vivo mediated by expression of wild type and mutant p53 proteins. Cancer Res. 51, 5232-5237 15. Diller, L., Kassel, J., Nelson, C. E., Gryka, M. A., Litwak, G., Gebhardt, M., Bressac, B., Ozturk, M., Baker, S. J., Vogelstein, B., and Friend, S. H. (1990) p53 functions as a cell cycle control protein in osteosarcomas. MoL Cell. Biol. 10, 5772-5781 16. Martinez, J., Georgoff, I., Martinez, J., and Levine, A. J. (1991) Cellular localization and cell cycle regulation by a temperaturesensitive p53 protein. Genes Dcv. 5, 151-159 17. Gannon,J. V., and Lane, D. P. (1991) Protein synthesis required to anchor a mutant p53 protein which is temperature-sensitive for nuclear transport. Nature (London) 349, 802-806 18. Baker, S. J., Markowitz, S., Fearon, E. R., Willson, J. K. V., and Vogelstein, B. (1990) Suppression of human colorectal carcinoma cell growth by wild-type p53. Science 249, 912-915 19. Yonish-Rouach, E., Resnitzky, D., Lotem, J., Sachs, L., Kimchi, A., and Oren, M. (1991) Wild-type p53 induces apoptosis of myeloid leukaemic cells that is inhibited by interleukin-6. Nature (London) 352, 345-347 20. Williams, G. T., Smith, C. A., Spooncer, E., Dexter, T. M., and Taylor, D. R. (1990) Haemopoietic colony stimulating factors promote cell survival by suppressing apoptosis. Nature (London) 343, 76-79 21. Kastan, M. B., Radin, A. I., Kuerbitz, S. J., Onyekwere, 0., Wolkow, C. A., Civin, C. I., Stone, K. D., Woo, T., Ravindranath, Y., and Craig, R. W. (1991) Levels of p53 protein increase with maturation in human hematopoietic cells. Cancer Res. 51, 4279-4286 22. Harvey, D. M., and Levine, A. J. (1991) p53 alteration is a common event in the spontaneous immortalization of primary BALB/C murine embryo fibroblasts. Genes Dcv. 5, 2375-2385 23. Ginsberg, D., Michalovitz, D., Ginsberg, D., and Oren, M. (1991) Induction of growth arrest by temperature-sensitive p53 mutant is correlated with increased nuclear localization and decreased stability of the protein. Mol. Cell. Biol. 11, 582-585 24. Shaulsky, G., Goldfinger, N., Peled, A., and Rotter, V. (1991) Involvement of wild-type p53 protein in the cell cycle requires nuclear localization. Cell Growth Dff 2, 661-667 25. Braithwaite, A. W., Sturzbecher, H.-W., Addison, C., Palmer, C., Rudge, K., and Jenkins, J. R. (1987) Mouse p53 inhibits SV4O origin-dependent DNA replication. Nature (London) 329, 458-460 26. Gannon, J. V., and Lane, D. P. (1987). p53 and DNA polymerase a compete for binding to SV4O T antigen. Nature (London) 329, 456-458 27. Sturzbecher, H. W., Brain, R., Maimets, T., Addison, C., Rudge, K., and Jenkins, J. R. (1988) Mouse p53 blocks SV4O DNA replication in vitro and downregulates T-antigen DNA helicase activity. Oncogene 3, 405-413 28. Wang, E. H., Friedman, P. N., and Prives, C. (1989) The murifle p53 protein blocks replication of SV4O DNA in vitro by inhibiting the initiation functions of SV4O large T-antigen. Cell 57, 379-392 29. Wilcock, D., and Lane, D. P. (1991) Localization of p53, retinoblastoma and host replication proteins at sites of viral replication in herpes-infected cells. Nature (London) 349, 429-431 30. Kastan, M. B., Onyekwere, 0., Sidransky, D., Vogelstein, B., and Craig, R. W. (1991) Participation of p53 protein in the cellular response to DNA damage. Cancer Res. 51, 6304-6311 31. Raycroft, L., Wu, H., and Lozano, G. (1990) Transcriptional activation by wild-type but not transforming mutants of the p53

p53 3175 www.fasebj.org by Kaohsiung Medical University Library (163.15.154.53) on November 13, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber

anti-oncogene. Science 249, 1049-1051 32. Fields, S., and Jang, S. K. (1990) Presence of a potent transcription activating sequence in the p53 protein. Science 249, 1046-1049 33. O’Rourke, R. W., Miller, C. W., Kato, G. J., Simon, K. J., Chen, D. L., Dang, C. V., and Koeffler, H. P. (1990) A potential transcriptional activation element in the p53 protein. Oncogene 5, 1829-1832 34. Raycroft, L., Schmidt, J. R., Yoas, K., Hao, M., and Lozano, G. (1991) Analysis of p53 mutants for transcriptional activation. MoL Cell. Biol. 11, 6067-6074 35. Shohat Foord, 0., Bhattacharya, P., Reich, Z., and Rotter, V. (1991)

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

A DNA

binding

domain

is contained

in the C-terminus

of wild type p53 protein. NucL Acids Res. 19, 5191-5198 Weintraub, H., Hauschka, S., and Tapscott, S. J. (1991) The MCK enhancer contains a p53 responsive element. Proc. NaIL Acad. Sci. USA 88, 4570-4571 Bargonetti, J., Friedman, P. N., Kern, S. E., Vogelstein, B., and Prives, C. (1991) Wild-type but not mutant p53 immunopurified proteins bind to sequences adjacent to the SV4O origin of replication. Cell 65, 1083-1091 Kern, S. E., Kinzler, K. W., Bruskin, A., Jarosz, D., Friedman, P., Prives, C., and Vogelstein, B. (1991) Identification of p53 as a sequence-specific DNA-binding protein. Science 252, 1708-1711 El-Deiry, W. S., Kern, S. E., Pietenpol, J. A., Kinzler, K. W., and Vogelstein, B. (1992) Definition of a consensus binding site for p53. Nature Genet. 1, 4S-49 Ginsberg, D., Mechta, F., Yaniv, M., and Oren, M. (1991) Wildtype p53 can down-modulate the activity of various promoters. Proc. NaIL Acad. Sci. USA 88, 9979-9983 Santhanam, U., Ray, A., and Sehgal, P. B. (1991) Repression of the interleukin 6 gene promoter by p53 and the retinoblastoma susceptibility gene products. Proc. NaIl. Acad. Sci. USA 88, 760S-7609 Chin, K. V., Ueda, K., Pastan, I., and Gottesman, M. M. (1992) Modulation of activity of the promoter of the human MDR1 gene by ras and p53. Science 255, 459-462 Eliyahu, D., Michalovitz, D., Eliyahu, S., Pinhasi- Kimhi, 0., and Oren, M. (1989) Wild-type p53 can inhibit oncogenemediated focus formation. Proc. Nail. Acad. Sci. USA 86, 8763-8767 Finlay, C. A., Hinds, P. W., and Levine, A. J. (1989) The p53 proto-oncogene can act as a suppressor of transformation. Cell 57, 1083-1093 Casey, G., Lo-Hsueh, M., Lopez, M. E., Vogelstein, B., and Stanbridge, E. (1991) Growth suppression of human breast cancer cells by the introduction of a wild-type p53 gene. Oncogene, 6, 1791-1798 Ronen, D., Rotter, V., and Reisman, D. (1991) Expression from the murine p53 promoter is mediated by factor binding to a downstream helix-loop-helix recognition motif. Proc. NaiL Acad. Sci. USA 88, 4128-4132 Wiman, K. G., Magnusson, K. P., Ramqvist, G., and Klein, G. (1991) Mutant p53 detected in a majority of Burkitt lymphoma cell lines by monoclonal antibody PAb24O. Oncogene 6, 1633-1640 Barak, Y., and Oren, M. (1992) Enhanced binding of a 95 kDa protein to p53 in cells undergoing p53-mediated growth arrest. EMBOJ 11, 2115-2121 Sherley, J. L. (1991) Guanine nucleotide biosynthesis is regis-

3176

Vol. 6

October 1992

50.

51. 52. S3.

lated by the cellular p53 concentration. j BioL Chem. 266, 24815-24828 Gannon, J. V., Greaves, R., Iggo, R., and Lane, D. P. (1990) Activating mutations in p53 produce a common conformational effect. A monoclonal antibody specific for the mutant form. EMBOJ 9, 1595-1602 Milner, J. (1991) p53 in the normal control of cell proliferation. Curr. Opin. Cell BioL 3, 282-286 Michalovitz, D., Halevy, 0., and Oren, M. (1991) p53 mutations: gains or losses? j Cell. Biochem. 45, 22-29 Unger, T., Nau, M. N., Segal, S., and Minna, J. D. (1992) p53 - a transdominant regulator of transcription whose function is ablated by mutations occurring in human cancer. EMBOJ. 10, 1383-1390

54. Shaulsky, G., Ben-Zeev, A., and Rotter, V. (1990) Subcellular distribution of the p53 protein during the cell cycle of Balb/c 3T3 cells. Oncogene 5, 1707-1711. 5S. Milner, J., and Medcalf, E. A. (1991) Cotranslation of activated mutant p53 with wild type drives the wild-type p53 protein into the mutant conformation. Cell 65, 765-774 S6. Stenger, J. E., Mayr, G. A., Mann, K., and Tegtmeyer, P. (1992) Formation of stable p53 homotetramers and multiples of tetramers. Molec. Carcinogen. 5, 102-106 57. Jenkins, J. R., Rudge, K., Chumakov, P., and Currie, G. A. (1985) The cellular oncogene p53 can be activated by mutagenesis. Nature (London) 317, 816-818 58. Halevy, 0., Rodel, J., Peled, A., and Oren, M. (1991) Frequent p53 mutations in chemically-induced murine fibrosarcoma. Oncogene 6, 1593-1600 59. Chiba, I., Takahashi, T., Nau, M., D’Amico, D., Curiel, D., Mitsudomi, T, Carbone, D., Piantadosi, S., Koga, H., Reissmann, P., Slamon, D., Holmes, E., and Minna, J. (1990) MutatiOns in the p53 gene are frequent in primary, resected nonsmall cell lung cancer. Oncogene 5, 1603-1610 60. Davidoff, A. M., Kerns, B. J. M., Iglehart, J. D., and Marks, J. R. (1991) Maintenance of p53 alterations throughout breast cancer progression. Cancer R#{128}s. 51, 260S-2610 61. Johnson, P., Gray, D., Mowat, M., and Benchimol, S. (1991) Expression of wild type p53 is not compatible with continued growth of p53-negative tumor cells. Mol. CelL BioL 11, 1-11 62. Wolf, D., Admon, S., Oren, M., and Rotter, V. (1984) Abelson murine leukemia virus-transformed cell that lack p53 protein synthesis express aberrant p53 mRNA species. Mol. CelL Biol. 4, 552-558 63. Donehower, L. A., Harvey, M., Slagle, B. L., McArthur, M. J., Montgomery, C. A., Butel, J. S., and Bradley, A. (1992) p53-deficient mice are developmentally normal but susceptible to spontaneous tumours. Nature (London) 356, 215-221 64. Malkin, D., Li, F. P., Strong, L. C., Fraumeni,J. F.,Jr., Nelson, C. E., Kim, D. H., Kassel, J., Gryka, M. A., Bischoff, F. Z., Tainsky, M. A., and Friend, S. H. (1990) Germ line p53 mutations in a familial syndrome of breast cancer, sarcomas, and other neoplasms. Science 250, 1233-1238 65. Srivastava, S., Zou, Z. Q, Pirollo, K., Blattner, W., and Chang, E. H. (1990) Germ-line transmission of a mutated p53 gene in a cancer-prone family with Li-Fraumeni syndrome. Nalure (London) 348, 747-749 66. Cheng, J., Yee, J.-K., Yeargin, J., Friedmann, T, and Haas, M. (1992) Suppression of acute lymphoblastic leukemia by the human wild-type p53 gene. Cancer Res. 52, 222-226

The FASEB Journal

OREN

www.fasebj.org by Kaohsiung Medical University Library (163.15.154.53) on November 13, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber