* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Novel p53 mutants selected in BRCA
Survey
Document related concepts
Microevolution wikipedia , lookup
History of genetic engineering wikipedia , lookup
BRCA mutation wikipedia , lookup
Nutriepigenomics wikipedia , lookup
Gene therapy of the human retina wikipedia , lookup
Vectors in gene therapy wikipedia , lookup
Genome (book) wikipedia , lookup
Point mutation wikipedia , lookup
Site-specific recombinase technology wikipedia , lookup
Cancer epigenetics wikipedia , lookup
Polycomb Group Proteins and Cancer wikipedia , lookup
No-SCAR (Scarless Cas9 Assisted Recombineering) Genome Editing wikipedia , lookup
Oncogenomics wikipedia , lookup
Strontium Dog wikipedia , lookup
List of Mutants in The Hills Have Eyes wikipedia , lookup
Transcript
Oncogene (1999) 18, 2451 ± 2459 ã 1999 Stockton Press All rights reserved 0950 ± 9232/99 $12.00 http://www.stockton-press.co.uk/onc Novel p53 mutants selected in BRCA-associated tumours which dissociate transformation suppression from other wild-type p53 functions Paul D Smith1, Susan Crossland1, Gillian Parker2, Peter Osin1, Louise Brooks3, Joanne Waller4, Elizabeth Philp1, Mark R Crompton1, Barry A Gusterson4, Martin J Allday2 and Tim Crook*,4 1 Institute of Cancer Research, Haddow Laboratories, 15 Cotswold Road, Sutton, Surrey, SM2 5NG, UK; 2Ludwig Institute for Cancer Research and Section of Virology and Cell Biology, Imperial College School of Medicine, St Mary's Campus, Norfolk Place, London W2 1PG, UK; 3Division of Clinical Sciences, London School of Hygiene and Tropical Medicine, Keppel Street, London W1, UK; 4Institute of Cancer Research, The Toby Robins Breast Cancer Research Centre, 237 Fulham Road, London SW3 6JB, UK Inheritance of germ-line mutant alleles of BRCA1 and BRCA2 confers a markedly increased risk of breast cancer and we have previously reported a higher incidence of p53 mutations in these tumours than in grade matched sporadic tumours. We have now characterized these p53 mutants. The results of these studies identify a novel class of p53 mutants previously undescribed in human cancer yet with multiple occurrences in BRCA-associated tumours which retain a pro®le of p53-dependent activities in terms of transactivation, growth suppression and apoptosis induction which is close or equal to wild-type. However, these mutants fail to suppress transformation and exhibit gain of function transforming activity in rat embryo ®broblasts. These mutants therefore fall into a novel category of p53 mutants which dissociate transformation suppression from other wild-type functions. The rarity of these mutants in human cancer and their multiple occurrence in BRCA-associated breast tumours suggests that these novel p53 mutants are selected during malignant progression in the unique genetic background of BRCA1and BRCA2-associated tumours. Keywords: p53; novel; mutation; breast; BRCA-1; BRCA-2 Introduction The tumour suppressor function of p53 is widely held to be exerted either through induction of G1 arrest via transcriptional upregulation of the cyclin dependent kinase inhibitor (CDK1) p21Waf1, or by activation of apoptosis, following DNA damage (El-Deiry et al., 1994; Ko and Prives, 1996). Induction of apoptosis appears to be associated, at least in some cell types, with expression of another p53-regulated gene Bax (Miyashita and Reed, 1995). A third p53-inducible gene, Mdm2, physically associates with p53 and via this interaction promotes the ubiquitin-dependent proteolysis of p53, thereby exerting a key role in determining the stability of p53 protein (Haupt et al., 1997; Kubbutat et al., 1997). Yet another p53-induced gene, PIG3, was recently described (Polyak et al., *Correspondence: T Crook Received 2 September 1998; revised 22 October 1998; accepted 3 November 1998 1997). PIG3 is related to a plant gene TED2, implicated in the apoptotic process associated with the formation of meristems, and the mammalian NADPH-quinone oxidoreductase which is a potent inducer of reactive oxygen species (ROS) which are powerful inducers of apoptosis. The importance of transactivation in mediating the biological eects of p53 has been illustrated by studies of human tumour associated p53 mutations the vast majority of which occur in evolutionarily conserved regions in the DNA binding domain and also by the study of rare, human tumour-derived mutants which selectively transactivate target p53 promoters. For example p53175P, originally described in metastatic cervical carcinoma (Crook et al., 1994), retains essentially wild-type ability to suppress proliferation, consistent with transactivation of p21Waf1. However, the same mutant not only fails to suppress transformation of primary rat embryo ®broblasts (REFs), perhaps by virtue of compromised ability to transactivate Bax or other p53 target genes, but also demonstrates a modest gain of function transforming phenotype (Crook et al., 1994; Rowan et al., 1996). Interestingly, mice null for Bax retain a normal p53-dependent apoptotic programme (Knudson et al., 1995). In the present study we have investigated the properties of a series of p53 mutants identi®ed in BRCA1- and BRCA2-associated carcinomas. Numerous lines of evidence implicate a role for p53 in the aetiology of BRCA-associated breast cancer. Inheritance of mutant alleles of the breast cancer susceptibility genes BRCA1 and BRCA2 is associated with a substantial risk of developing breast, ovarian and some other cancers (Miki et al., 1994; Wooster et al., 1996). Tumorigenesis in carriers of germ-line mutations in BRCA1 and BRCA2 is frequently accompanied by loss of the wild-type allele (Smith et al., 1992; Neuhausen and Marshall, 1994; Collins et al., 1995) implying that these proteins are tumour suppressors, and they have been proposed to have a `caretaker' role in the maintenance of genomic stability, rather than a `gatekeeper' function which operates in pathways regulating cellular proliferation (Kinzler and Vogelstein, 1997; Brugarolas and Jacks, 1997). BRCA1 and BRCA2 are essential for embryonic cell proliferation, embryos homozygously deleted for either gene being inviable, undergoing a p21Waf1-mediated growth arrest at around day 5 ± 6 (Hakem et al., 1996; Suzuki et al., 1997). This lethality is partially reversed by breeding into a p53 or p21Waf1 null background (Hakem et al., 1997; Ludwig et al., 1997) suggesting that loss of Novel p53 mutations in familial breast cancer PD Smith et al 2452 function in either gene can partially reverse the cell cycle arrest consequent to loss of BRCA-encoded functions. Interestingly, analysis of ®broblasts derived from mice homozygous for a truncating mutation in BRCA2 has revealed a proliferation defect associated with induction of p21Waf1 expression, and a DNA repair defect (Connor et al., 1997; Patel et al., 1998). However, it is not known whether either or both of these defects occur in tumours arising in carriers of germline mutant alleles of BRCA1 or BRCA2. Nevertheless, these observations taken together imply that loss of p53 function may be an important genetic event in tumorigenesis in a mutant BRCA1 or 2 background. Consistent with this hypothesis, analysis of a series of high grade breast and ovarian tumours arising in carriers of germ-line BRCA1 and BRCA-2 mutations revealed frequent p53 mutation (Crook et al., 1997, 1998b). We report here the thorough characterization of these mutants identi®ed in BRCAassociated tumours and discuss their implications for p53 biology and the mechanisms of BRCA-associated carcinogenesis. promoter: 199R shows wild-type activity, 181C and 202S are close to wild-type, 150I and 219H are approximately 40% of wild-type and 158H and 240R exhibit activities about 25% of wild-type. To con®rm that the p53 mutants could also transactivate the endogenous p21Waf1 allele, Saos-2 cells were transiently transfected with the p53 mutants and processed for immunohistochemical detection of p21Waf1 protein. as predicted by the reporter assays, 181C, 202S, 199R and 150I showed similar numbers of positively staining cells as the wild-type p53. 215C was approximately 40% of the wild-type whilst 219H and 158H showed very low numbers of p21Waf1 expressing cells. Data from a typical experiment are shown in Table 1. No positive cells were detected following transfection of 240R, empty vector pCB6+ or a transactivation dead mutant, 248W. These data indicate that where a mutant shows close to wild-type activity in the reporter assay, this re¯ects capacity to transactivate the endogenous p21Waf1 allele. The two assays dier with mutants which have relatively weak activity in the reporter assay e.g. 240R, 158H and Results Novel p53 mutations in BRCA1 and BRCA2-associated tumours p53 mutations were identi®ed in genomic DNA from a high proportion of BRCA1- and BRCA2-associated tumours (Crook et al., 1998b). Having completed sequence analysis of the BRCA1- and BRCA2associated tumours, the incidence of the mutations in human cancer was determined from the IARC human p53 mutation database. Strikingly, three mutants, 150I, 199R and 202S, have not been previously described in sporadic human cancers and many of the other mutants were extremely rare in human cancer (Crook et al., 1998b). Most BRCA-tumour associated p53 mutants transactivate the p21Waf1 promoter Since studies on mice homozygously deleted for BRCA1 and BRCA2 indicate that early embryonic lethality is, at least in part, accounted for by a p53dependent induction of p21Waf1, we tested the ability of p53 mutants associated with BRCA-tumours to transactivate the human p21Waf1 promoter. Each mutant was therefore constructed by in vitro mutagenesis, subcloned into a eukaryotic expression vector and transactivating activity determined against the p21Waf1 promoter in transient transfection assays. Of the 13 BRCA-associated mutants tested, ®ve (144P, 163N, 168Y, 234C and 248W) had completely lost the ability to activate transcription over a range of input DNA concentrations up to 500 ng (data not shown). Those showing evidence of ability to transactivate the p21Waf1 promoter were then subjected to more extensive analysis. Each of these mutants were assayed in triplicate at 1, 10, 50 and 500 ng of input DNA. Results of analyses performed with 50 ng of input DNA which is optimal for wild-type DNA are shown in Figure 1. The eight transactivating mutants demonstrate a range of activities against the p21Waf1 Figure 1 Transactivating activity of p53 mutants from BRCA1and BRCA2-associated tumours against the p21Waf1 promoter. Saos-2 cells were transfected with 1 mg of each reporter plasmid, together with the indicated amounts of each expression plasmid and 250 ng of pCMV b-gal. Lysates were assayed for luciferase and b-galactosidase activity as described in Materials and methods. The data are expressed as relative to wild-type p53 transactivating capacity Table 1 Induction of expression of endogenous p21waf1 in Saos-2 cells by transfection of p53 mutants p53 plasmid 1 2 3 pCB6+ Wild-type 181C 202S 150I 219H 158H 240R 248W 0 26 21 17 21 1 1 0 0 0 92 64 89 88 24 6 4 0 0 94 39 68 67 13 1 1 0 Data shown are numbers of nuclei positive for expression of p21Waf1. Results from three typical experiments are shown Novel p53 mutations in familial breast cancer PD Smith et al 219H, in this case, the ability to induce the p21Waf1 protein is low or zero. Mutants retain p53-dependent suppression of proliferation The proliferative defect observed in BRCA7/7 embryos and in ®broblasts homozygous for a truncating mutation in BRCA2 (Hakem et al., 1997; Ludwig et al., 1997; Connor et al., 1997) is hypothesized to be p53-dependent. As such, it was surprising to observe retention of p21Waf1 transactivating activity in numerous p53 mutants from BRCA1and BRCA2-associated tumours, and it was clearly important to directly assess whether the mutants were compromised for p53-dependent inhibition of proliferation. We therefore utilized the well-validated stable transfection assay of p53-dependent growth arrest in Saos-2 (p537/7) cells. Of the 13 mutants tested, only four (163N, 168Y, 234C and 248W) were completely negative for growth arrest whereas the remaining nine mutants exhibited reproducible growth suppressor activity (Figure 2). Mutants with wild-type or close to wild-type ability to transactivate the p21Waf1 promoter (181C, 199R, 202S and 215C) suppressed Saos-2 growth with close to wild-type eciency. There is a good correlation between p21Waf1 transactivating activity and Saos-2 growth suppression. Interestingly however, 144P suppressed growth by 40 ± 50% despite being null for p21Waf1 transactivation. Dierential transactivating activity of p53 mutants The retention of transactivating activity against the p21Waf1 promoter by many of the BRCA-associated p53 mutants raised the question of whether the basis for their selection in BRCA-associated tumours re¯ected loss of activity against other p53 target promoters, rather than p21Waf1. Each mutant was therefore assayed against four other p53 responsive promoters: mdm2, Bax, IGF BP-3 and the recently described human PIG3 (Figure 3). Experiments were performed as described previously for the p21Waf1 promoter. As for the p21Waf1 promoter, 144P, 163N, 168Y, 234C and 248W failed to transactivate any of the promoters over a range of input DNA concentrations (data not shown). Analyses of the remaining eight mutants, which transactivated p21Waf1, are shown in Figure 3. These segregated broadly into two groups. Group 1 comprises those mutants (181C, 219H, 158H and 240R) which transactivate p21Waf1 and mdm2 (181C and 219H also transactivate PIG3) but not Bax or IGF BP-3. Group 2 consists of those mutants (150I, 199R, 202S and 215C) which transactivate all ®ve tested promoters. Since 202S, 199R and 150I mutants transactivate all ®ve promoters tested at close to or more than wild-type eciency and do so over a range of DNA concentrations (data not shown), we conclude that they are fully transcriptionally competent. Some striking individual variations are worthy of note. For example, 202S reproducibly showed activity against IGF BP-3 which was at least twofold higher than wild-type. Similarly, 150I transactivated PIG3 signi®cantly more eciently than wild-type. Apoptosis-inducing activity of p53 mutants A number of studies have led to the proposal that the capacity to induce apoptosis is a critical determinant of the ability of p53 proteins to mediate transformation suppression (Lowe et al., 1994). In order to investigate whether the p53 mutants from BRCA1- and BRCA2associated tumours retained apoptosis-inducing activity, each was analysed in an in vitro assay of p53dependent apoptosis (Crook et al., 1998a). Plasmids encoding each mutant were transfected into Saos-2 (p537/7) cells. Sixty hours after transfection, cells were ®xed and stained for p53 with mAb DO-1 and for DNA content with propidium iodide. p53-positive and -negative fractions were identi®fed by ¯ow cytometry and subjected to cell cycle analysis (Figure 4). Mutants 144P and 168Y were completely negative for the induction of apoptosis. 163N, 219H, 234C and 240R displayed about 10% of wild-type activity, comparable with a previously characterized mutant 234H (Parker et al., 1996) (Figure 4). In contrast, a number of the BRCA-associated mutants retained signi®cant capacity to induce apoptosis (Figure 4). 150I, 199R and 202S displayed close to wild-type apoptosis-inducing activity, 181C and 215C were 30 ± 40% of wild-type, whilst 158H exhibited 15% of wild-type activity. Apoptosis-competent p53 mutants are undescribed in sporadic tumours Figure 2 Analysis of growth suppressor functions of p53 mutants. Suppression of Saos-2 growth. Each dish of Saos-2 cells received 2 mg of pCB6+ p53 plasmid or vector alone. Data shown are the mean of at least four separate experiments and are expressed as survival relative to cells receiving pCB6+ vector alone The demonstration that some of the BRCA-associated p53 mutants were competent not only for transactivation and growth suppression, but also, most strikingly, for induction of apoptosis, was unexpected since the vast majority of human tumour-derived p53 mutants have lost these properties (Ko and Prives, 1996). It was of interest, therefore, to correlate this activity with the incidence of the mutants in sporadic human tumours. 2453 Novel p53 mutations in familial breast cancer PD Smith et al 2454 Figure 3 Transactivating activity of p53 mutants from BRCA1- and BRCA2-associated tumours against Mdm2, Bax, PIG3 and IGF BP-3 promoters. Saos-2 cells were transfected with 1 mg of each reporter plasmid, together with the indicated amounts of each expression plasmid and 250 ng of pCMV b-gal. Lysates were assayed for luciferase and b-galactosidase activity as described in Materials and methods. The data are expressed as relative to wild-type p53 transactivating capacity Those mutants which were totally compromised for p53-mediated transactivation, growth suppression and apoptosis (144P, 163N, 168Y, 234C and 248W) have been commonly described in the IARC database, whereas mutants 150I, 199R and 202S, which retained close to wild-type levels of transactivating, growth suppression and apoptosis-inducing activity, are previously undescribed in sporadic tumours. Each p53 mutant fails to suppress transformation of REFs Expression of human wild-type p53 inhibits the oncogenic transformation of early passage primary rat embryo ®broblasts (REFs) by nuclear acting oncoproteins such as that encoded by human papillomavirus type 16 E7 (HPV16E7) and EJ-ras (Crook et al., 1991). The majority of p53 mutants derived from human tumours do not have this property (Ko and Prives, 1996; Levine et al., 1994). Moreover, some mutant p53 proteins exhibit gain of function allowing them to cooperate with ras to transform REFs. To determine the phenotype of the p53 proteins from BRCA1- and BRCA2-associated breast tumours, each mutant was introduced into REFs either in combination with HPV16E7 and an activated ras gene or with ras alone. The mutants 144P, 158H, 163N, 168Y, 234C and 248W not only failed to suppress transformation by HPV16E7 and ras, but markedly enhanced transformation (Figure 5a). Moreover, each clone was clearly capable of co-operating alone with ras to eciently transform REFs. Indeed, comparison with 173L (a p53 mutant with potent transforming function (Crook et al., 1994)) revealed that each clone was at least as potent as 173L (Figure 5b). Analysis of mutants 150I, 199R and 202S which possessed an otherwise wild-type phenotype revealed, surprisingly, that each was compromised in its ability to suppress the transforming activity of HPV16E7 and ras (Figure 5a). Furthermore, the same clones, when transfected with ras alone, were capable of generating transformed colonies above the background of vector Novel p53 mutations in familial breast cancer PD Smith et al controls, although colony numbers were clearly below those of the highly transforming complete loss of function mutants (Figure 5b). Other mutants, 181C and 215C, which are substantially wild-type also exhibited both loss of transformation suppression and independent gain of transforming function. To verify expression of the transfected mutant p53 sequences, transformed colonies were propagated as cell lines and Figure 4 Flow cytometric analysis of p53-induced apoptosis in Saos-2 cells. In the graphs showing the total cell population, the Xaxis (FL2A) represents DNA content and the Y-axis FITC. The p53+ population is gated in the top right-hand box. In the graphs showing p53+ve and 7ve populations separately, the X-axis (FL2A) represents DNA content and the Y-axis the number of cells. Apoptosis is measured as the accumulation of cells with a sub-G1 DNA content (M1) which appears to the left of the G1/G0 peaks (M2). (a) Mutants with apoptotic activity are compared with wild-type p53 and a previously characterized mutant which retains the ability to induce apoptosis, 283H (Crook et al., 1988a); (b) Mutants negative for apoptosis are compared with a previously described highly transforming mutant 234H (Parker et al., 1996); (c) Summary of at least three separate experiments; the average level of apoptosis induced by each mutant is expressed as a percentage of wild-type 2455 Novel p53 mutations in familial breast cancer PD Smith et al 2456 Figure 6 Expression of mutant p53 proteins in transformed REF cell lines. Lysates of cell lines transformed by mutant p53 and ras were subjected to Western blotting and p53 protein was detected by the DO-1 antibody. The identity of the p53 mutants in each cell line are: Lane 1: untransfected primary REFs; Lane 2: 240R; Lane 3: 181C; Lane 4: 248W; Lane 5: 163N; Lane 6: 234C; Lane 7: 158H; Lane 8: 168Y; Lane 9: 144P; Lanes 10 and 13: Wild-type p53; Lane 11: 150I; Lane 12: 202S transfected, wild-type p53 was not expresed (Figure 6). These data are clearly consistent with suppression of transformation by wild-type p53 and loss of this property by mutants 150I, 199R and 202S, despite retention of ecient apoptosis-inducing activity. Discussion Figure 5 Transformation suppression and gain of function transformation properties p53 mutants from BRCA1- and BRCA2-associated tumours. (a) Transformation of primary REFs by p53 mutants co-transfected with pEJ6.6. (b) Eect of p53 mutants on transformation of primary REFs by HPV16E7 (pCB6+16E7) co-transfected with pEJ6.6. Data are the total number of transformed colonies per plate. The error bars represent standard error of the mean p53 expression was then analysed by Western blotting using the human ± speci®c DO-1 antibody. Expression of each mutant was readily detectable in transformed cell lines (Figure 6), although the level of expression of mutants with close to wild-type transactivating phenotypes was lower than that of mutants negative for transactivation. Interestingly, analysis of the occasional transformed colonies arising on plates which received wild-type p53 revealed that the The ®ndings in this paper have important implications for our understanding of both BRCA-associated breast cancer and p53 biology. Our recent ®ndings that p53 mutation is more frequent in BRCA-associated breast tumours than in a grade matched set of sporadic breast tumours (Crook et al., 1997, 1998b) was initially consistent with the hypothesis that p53 mutation might occur in such tumours in order to abrogate p53-dependent induction of p21Waf1 expression. This model was suggested by studies of embryos homozygously deleted for BRCA1 or BRCA2 which revealed a proliferation defect that was partially relieved by crossing into a p53 null or p21Waf1 null background (Hakem et al., 1997; Ludwig et al., 1997). Furthermore, analysis of murine ®broblasts homozygous for a truncated BRCA2 which revealed a proliferation defect associated with a modest increase in steadystate levels of p53 protein, suggested the possibility that the proliferation defect might be, at least in part, p53dependent (Connor et al., 1997). Taken together, these studies implied that abrogation of p21Waf1 induction via loss of function in p53 might, at least in part, relieve the proliferation defect, and provided a molecular rationale for the selection of p53 mutations. As a corollary of this, we had anticipated that loss of p21Waf1 transactivation would be a property common to mutants from BRCA-associated tumours. However, we now show the converse, namely that the majority of the p53 mutants we have identi®ed and analysed retain signi®cant ability to transactivate the p21Waf1 promoter. Since the large majority of tumour-derived p53 mutants have lost this activity, this was a most unexpected result and prompted additional studies to determine whether other properties of wild-type p53 are intact in these p53 mutants. Analysis of the ability of each mutant to suppress the proliferation of the Saos-2 cell line revealed that this property too was retained by many of the mutants. Previous studies have demonstrated that ecient suppression of Saos-2 proliferation is often a function of the ability of p53 to transactivate expression of p21Waf1 (Crook et al., Novel p53 mutations in familial breast cancer PD Smith et al 1994). The retention of this property by many of the mutants in the present series is consistent with their ability to transactivate p21Waf1. Taken together with the analysis of p21Waf1 transactivation, these observations demonstrate that ability to activate a p21Waf1 dependent G1 checkpoint is substantially retained by these p53 mutants. We conclude, therefore, that the higher frequency of p53 mutation in BRCA-associated tumours does not re¯ect an absolute requirement either for loss of p53-dependent p21Waf1 induction or for abrogation of a predicted p53-mediated proliferation block. Aside from their utility in addressing the role of p53 mutation in BRCA-associated tumorigenesis, study of these mutants has aorded signi®cant insight into general mechanisms of tumour suppression by p53. We demonstrate a close correlation between apoptosis-inducing activity of the p53 mutants in this series and ability to transactivate PIG3. To our knowledge, this is the ®rst detailed analysis of transactivation of PIG3 by human tumour-derived p53 mutants. The correlation between apoptosis and transactivation of PIG3 is intriguing since Polyak and colleagues, in describing the cloning and preliminary functional characterization of the cDNA, showed that expression of PIG3 was not per se sucient to induce apoptosis (Polyak et al., 1997). The properties of 181C and 215C clearly support this observation, since each transactivates PIG3 more eciently than wild-type p53, yet has lost about 2/3 of the apoptotic activity of the wild-type protein, implying that transactivation of additional genes is required. The close relationship we observe between apoptosis and induction of PIG3 suggests, therefore, that PIG3 may be necessary, although clearly not sucient, for implementation of p53-dependent apoptosis. These observations then raise the question of which additional gene(s) are required for p53-mediated apoptosis. An attractive possibility is Bax, since previous analysis of p53 mutants 175P and 181L demonstrated that both mutants, although competent to suppress proliferation through transactivation of p21Waf1 (Crook et al., 1994) lacked apoptosis-inducing activity, a defect associated with inability to transactivate Bax (Rowan et al., 1996). Comparison of the transactivating abilities of 199R and 215C suggests, however, that Bax is unlikely to represent a strong candidate for a co-operating pro-apoptotic gene. Although both 199R and 215C transactivate PIG3 as well as wild-type and both mutants exhibit comparable activity against the Bax promoter, 199R is essentially wild-type for apoptosis induction whereas 215C retains only 30% activity. Furthermore, 181C which possesses 100% wild-type activity against PIG3 retains 30% wild-type apoptotic activity yet is negative for transactivation against Bax. The observation that mice homozygously deleted for Bax remain competent to eect the p53dependent apoptotic programme further argues against a critical role for transactivatioan of Bax expression (Knudson et al., 1995). These results strongly suggest, but do not prove, that activation of expression of genes other than Bax must occur for full implementation of the p53-dependent apoptotic programme. One conclusion which can be drawn from these studies is that the in vitro pro-apoptotic functions of p53 correlate with comprehensive sequence speci®c transactivation of all the promoters tested particularly Bax, IGF-BP3 and PIG3. It has been reported for some p53 mutants that both the growth suppression and apoptotic functions of p53 have components which do not depend entirely upon sequence speci®c transcription (Walker and Levine, 1996; Haupt et al., 1995) although these observations were made with in vitro generated mutations. It has also been reported that p53-dependent apoptosis can occur in the presence of inhibitors of transcription (Caelles et al., 1994; Wagner et al., 1994). Of note therefore, in the present study is 144P which is null for transactivation but retains some capacity to suppress the growth of Saos-2 cells, the same mutant is however null for in vitro apoptosis and transformation suppression. Therefore if 144P retains nontransactivating functions of p53 they are clearly insucient to induce in vitro apoptosis or suppress transformation. The most striking conclusion from our studies is the identi®cation of a group of p53 mutants occurring uniquely in BRCA1- and BRCA2-associated tumours, which are null for transformation suppression and have acquired transforming ability despite retaining a phenotype close to that of the wild-type protein in terms of transactivation, growth suppression and apoptosis-inducing functions. In previous studies, the transforming ability of mutants 175P and 181L was shown to correlate with absent or severely compromised apoptosis-inducing activity. Accompanying the diminished apoptotic activity was loss of ability to transactivate Bax. As such, possession of close to wildtype apoptosis-inducing activity and transactivating activity against Bax by mutants 150I and 202S, and of signi®cant activity by 199R and 215C, clearly dierentiates these mutants from 175P and 181L. Similarly, whereas 175P lacks detectable activity against IGF-BP3, 150I, 199R, 202S and 215C all possess signi®cant transactivating activity towards this promoter further distinguishing them from 175P and 181L. To summarize, since 202S, 199R and 150I mutants transactivate all 5 promoters tested at close to or more than wild-type eciency and do so over a range of DNA concentrations (data not shown), we can ®nd no evidence that they are transcriptionally defective. However, we cannot formally exclude the possibility that these mutants fail to transactivate other p53 responsive genes and that this underlies their failure to suppress transformation. It is a tantalising possibility therefore that these mutants represent the minimal loss of function sucient to lose p53 dependent tumour suppression, as such they would make an invaluable tool with which to uncover the key p53 dependent events which mediate tumour suppression. Taken together, the results of these analyses place these p53 mutants into a novel category. Their multiple occurrence in BRCA-associated tumours, loss of transformation suppression and gain of transforming function indicate that they are genuine oncogenic mutants despite the retention of an otherwise close to wild-type pro®le of biological properties. Retention of such wild-type properties, particularly ecient p53dependent apoptotic activity, is exceedingly unusual in sporadic human tumour-derived p53 mutants (Ko and Prives, 1996) and it is therefore remarkable to observe 2457 Novel p53 mutations in familial breast cancer PD Smith et al 2458 this property in several of the analysed mutants in the present series. The only described p53 mutant with a similar phenotype is 283H which also retains wild-type activities such as transactivation and apoptosis inducing capacity (Crook et al., 1988a). However, the mutants described here dier from 283H because they are retained in the nucleus (Paul Smith, unpublished observations) whereas 283H is constitutively activated for nuclear export (Crook et al., 1988a). Consistent with the rarity of retention of wild-type p53-dependent activities, analysis of the IARC database reveals that 150I, 199R and 202S which exhibited activity most comparable to wild-type have never been described previously in sporadic human cancer. Whilst the compromised transformation suppression functions of these mutants provides an obvious biological explanation for their presence in tumours, the basis for the selection of these mutants rather than the full loss of function mutants commonly seen in sporadic breast cancer such as 175H and 273C is of interest. The occurrence of these mutants uniquely in BRCAassociated tumours implies that the genetic background provided by speci®c tumour-associated BRCA1 and BRCA2 mutations underlies the selection of these mutants. The genetic background of the tumours may confer tolerance to or abrogate the transactivating and/or apoptotic functions of p53 which are retained by 150I, 199R and 202S and in so doing allows expression of the transforming properties. The demonstration (Zhang et al., 1998) that mutant BRCA1 proteins inhibit the ability of p53 to transactivate the p21Waf1 and Bax promoters provides strong evidence to support this hypothesis. A second possibility is that the transactivating ability of these mutants is tolerated as a result of loss of function in other components of the p53 pathway. The recent observation that p19ARF is involved in activation of p53 function is clearly of interest in this respect (de Stanchina et al., 1998; Zindy et al., 1998). Moreover, an inverse correlation between expression of the human p14ARF and p53 function has been described (Stott et al., 1998). Alternatively, it is possible that survival of cells within certain genetic backgrounds in conjunction with the mutant BRCA-allele, is dependent upon retention of some wild-type p53 functions. Further investigation into the role of p53 mutation in the development and progression of BRCA1 and BRCA2 associated tumours will have to include the eects of BRCA-mutant alleles on p53 dependent activities and a thorough characterization of the molecular pathology and immuno-phenotype of the tumours. Materials and methods Tumours Paran-embedded specimens were removed from pathology archives and the diagnosis of each tumour con®rmed by examination of haematoxylin and eosin-stained tissue sections by two pathologists (PO and BAG). The p53 mutations described in this series were identi®ed by SingleStrand Conformation Polymorphism analysis (SSCP) and DNA sequencing in tumours arising in carriers of germ-line mutant alleles of BRCA1 and BRCA2 as described previously (Crook et al., 1997, 1998b). In vitro mutagenesis and plasmids Mutant p53 clones were constructed from pALTER1p53 using the `Altered Sites' system (Promega) as previously described (Crook et al., 1994). Mutants were identi®ed by sequencing, then subcloned into pCB6+ for eukaryotic expression. Construction of pCB6+16E7 was described previously (Parker et al., 1996). Transactivation assays The ability of the p53 mutants to transactivate p53 responsive promoters was assayed in Saos-2 (p537/7) human osteosarcoma cells essentially as previously described (Crook et al., 1998a) with minor modi®cations. Brie¯y, 1.256105 cells were transfected by the calcium phosphate co-precipitation technique, with 1 mg of luciferase reporter vector and 1 ± 500 ng of the pCB6 + p53 expression plasmid. Each assay included 250 ng of pCMVb-gal as an internal transfection control. Twentyfour hours post transfection, cells were lysed and processed for luciferase activity measurements according to the manufacturers' instructions (Promega). b-galactosidase assays were performed using b-D-galactopyranoside (Boehringer Mannheim) as a substrate and activity was measured using a micro-titre plate reader (Dynex Technologies). Activation of expression of endogenous p53 target genes was con®rmed by immuno¯uorescence staining of aliquots of transfected cells with the anti-p21 Waf1 antibody SX21 (Fredersdorf et al., 1996). Saos-2 proliferation assay Assays for the p53-dependent suppression of proliferation were performed essentially as previously described (Crook et al., 1998a). Brie¯y, 56105 Saos-2 cells were transfected with ®xed amounts of pCB6+ p53 expression plasmids and grown in the presence of 500 mg/ml G418 for 3 ± 4 weeks, ®xed in formol saline and stained in Giemsa. Rat embryo ®broblast transformation assays REF transformation assays were performed as described previously (Crook et al., 1994, 1998a; Parker et al., 1996). Following transfection, cells were grown in the presence of 500 mg/ml G418 sulphate. Transformed colonies were counted after 3 ± 4 weeks. Expression of p53 in transformed cell lines was con®rmed by Western blotting of cell lysates using the DO-1 antibody. Analysis of apoptosis by ¯ow cytometry Sixty hours after transfection, the assay for apoptosis was performed essentially as described previously (Crook et al., 1998a). After ®xing and staining for p53 with mAb Do1 and DNA with propidium iodide (PI), analysis was performed with a FACSort (Becton Dickinson). Cells were measured for their FITC ¯uorescence intensity (green channel) and PI intensity (red channel). Total populations were gated to remove doublets and very small debris and the p53+ve population was gated for high FITC compard to control transfected cells. M1 represents the sub-G1 apoptotic population, M2 the G1/G0 peak, M3 represents S phase and M4 is G2/M. Acknowledgements We thank K Vousden for pCB6+173L, R Weinberg for pEJ6.6, B Vogelstein for Waf1 luciferase and PIG3 luciferase, J Reid for Bax luciferase, L Buckbinder for IGF BP3 Box A luciferase, Moshe Oren for MDM2 luciferase and Xin Lu for the anti-p21Waf1 antibody. Novel p53 mutations in familial breast cancer PD Smith et al T Crook is a Leopold Muller Fellow. J Waller is funded by US Army grant number DAMD17-94-4066, L Brooks is supported by the Medical Research Council, P Osin by the Gilbert Fellowship and G Parker by the Wellcome Trust. Various parts of the work were supported by the Cancer Research Campaign and Breakthrough Breast Cancer. References Brugarolas J and Jacks T. (1997). Nature Med., 3, 721 ± 722. Caelles C, Helmberg A and Karin M. (1994). Nature, 370, 220 ± 223. Collins N, McManus R, Wooster R, Mangion J, Seal S, Lakhani SR, Ormiston W, Daly PA, Ford D, Easton D and Stratton MR. (1995). Oncogene, 10, 1673 ± 1675. Connor F, Bertwistle D, Mee PJ, Ross GM, Swift S, Grigorieva E, Tybulewicz VL, Ashworth A. (1997). Nature Genetics, 17, 423 ± 430. Crook T, Fisher C and Vousden KH. (1991). J. Virol., 65, 505 ± 510. Crook T, Marston N, Sara E and Vousden KH. (1994). Cell, 79, 817 ± 827. Crook T, Crossland S, Crompton MR, Osin P and Gusterson BA. (1997). Lancet, 350, 638 ± 639. Crook T, Parker GA, Rozycka M, Crossland S and Allday MJ. (1998a). Oncogene, 16, 1429 ± 1441. Crook TR, Brooks LA, Crossland S, Osin P, Barker KT, Waller J, Philp E, Smith PD, Yulug I, Peto J, Parker G, Allday MJ, Crompton MR and Gusterson BA. (1998b). Oncogene, 17, 1681 ± 1689. El-Deiry WS, Harper JW, O'Connor PM, Velculescu VE, Canman CE, Jackman J, Pientenpol JA, Burrell M, Hill DE, Wang Y, Wilman KG, Mercer WE, Kastan MB, Kohn KW, Elledge SJ, Kinzler KW and Vogelstein B. (1994). Cancer Res., 54, 1169 ± 1174. Fredersdorf S, Milne AW, Hall PA and Lu X. (1996). Am. J. Path., 148, 825 ± 835. Hakem R, de la Pompa JL, Sirard C, Mo R, Woo M, Hakem A et al. (1996). Cell, 11, 1009 ± 1023. Hakem R, de la Pompa JL, Elia A, Potter J and Mak TW. (1997). Nature Genetics, 16, 298 ± 301. Haupt Y, Rowan S, Shaulian E, Vousden KH and Oren M. (1995). Genes Dev., 9, 2170 ± 2183. Haupt Y, Maya R, Kazaz A and Oren M. (1997). Nature, 387, 296 ± 299. Kinzler KW and Vogelstein B. (1997). Nature, 386, 761 ± 763. Knudson CM, Tung KS, Tourtellotte WG, Brown GA and Korsmeyer SJ. (1995). Science, 270, 96 ± 99. Ko LJ and Prives C. (1996). Genes Dev., 10, 1054 ± 1072. Kubbutat MHG, Jones SN and Vousden KH. (1997). Nature, 387, 299 ± 303. Levine AJ, Perry ME, Chang A, Silver A, Dittmer D, Wu M and Welsh D. (1994). Br. J. Cancer, 69, 409 ± 416. Lowe SW, Jacks T, Housman DE and Ruley HE. (1994). Proc. Natl. Acad. Sci. USA, 91, 2026 ± 2030. Ludwig T, Chapman DL, Papaioannou VE and Efstratiadis A. (1997). Genes Dev., 11, 1126 ± 1241. Miki Y, Swensen J, Shattuck-Eidens D, Futreal PA, Harshman K, Tavtigian S, Liu Q, Cochran C, Bennet LM, Wei D, Bell R, Rosenthal J, Hussey C, Thanh T, McClure M, Frye C, Hattier T, Phelps R, Haugen-Strano A, Katcher H, Kazulo Y, Gholami Z, Shaer D, Stone S, Bayer S, Wray C, Bogden R, Dayananth P, Ward J, Tonin P, Narod S, Bristow PK, Lai M, Barrett C, Lewis C, Neuhausen S, Cannon-Albright L, Goldger D, Wiseman R, Kamb A and Skolnick MH. (1994). Science, 266, 66 ± 71. Miyashita T and Reed JC. (1995). Cell, 80, 293 ± 299. Neuhausen SL and Marshall CJ. (1994). Cancer Res., 54, 6069 ± 6072. Parker GA, Crook T, Bain M, Sara E, Farrell PJ and Allday MJ. (1996). Oncogene, 13, 2541 ± 2549. Patel KJ, Yu VPCC, Lee H, Corcoran A, Thistlethwaite FC, Evans MJ, College WH, Friedman LS, Ponder BAJ and Venkitaraman AR. (1998). Mol. Cell, 1, 347 ± 357. Polyak K, Xia Y, Zweier JL, Kinzler KW and Vogelstein B. (1997). Nature, 389, 300 ± 305. Rowan S, Ludwig RL, Haupt Y, Bates S, Lu X, Oren M and Vousden KH. (1996). EMBO J., 15, 827 ± 838. Smith SA, Easton DF, Evans DGR and Ponder BAJ. (1992). Nature Genetics, 2, 128 ± 131. de Stanchina E, McCurrach ME, Zindy F, Shieh S-Y, Ferbeyre G, Samuelson AV, Prives C, Roussel MF, Sherr CJ and Lowe SW. (1998). Genes Dev., 12, 2434 ± 2442. Stott FJ, Bates S, James MC, McConnell BB, Starborg M, Brookes S, Palmero I, Ryan K, Hara E, Vousden KH and Peters G. (1998). EMBO J., 17, 5001 ± 5014. Suzuki A, de la Pompa JL, Hakem R, Elia A, Yoshida R, Mo R, Nishina H, Chuang T, Wakeham A, Itie A, Koo W, Billia P, Ho A, Fukumoto M, Hui CC and Mak TW. (1997). Genes Dev., 11, 1242 ± 1252. The Breast Cancer Linkage Consortium. (1997). Lancet, 349, 1505 ± 1510. Wagner AJ, Kokontis JM and Hay N. (1994). Genes Dev., 8, 2817 ± 2830. Walker KK and Levine AJ. (1996). Proc. Natl. Acad. Sci. USA, 93, 15335 ± 15340. Wooster R, Bignell G, Lancaster J, Swift S, Seal S, Mangion J, Collins N, Gregory S, Gumbs C, Micklem G, Barfoot R, Hamoudi R, Patel S, Rice C, Biggs P, Hashim Y, Smith A, Connor F, Arason A, Gudmundsson J, Ficenec D, Kelsell D, Ford D, Tonin P, Bishop DT, Spurr N, Ponder BAJ, Eeles R, Peto J, Devilee P, Cornelisse C, Lynch H, Narod S, Lenoir G, Egilsson V, Barkadottir RB, Easton DF, Bentley DR, Futreal PA, Ashworth A and Stratton MR. (1996). Nature, 378, 789 ± 792. Zhang H, Somasundaram K, Peng Y, Tian H, Zhang H, Bi D, Weber BL and El-Deiry WS. (1998). Oncogene, 16, 1713 ± 1721. Zindy F, Eischen CM, Randle DH, Kamijo T, Cleveland JL, Sherr CJ and Roussel MF. (1998). Genes Dev., 12, 2424 ± 2433. 2459