Oncogene (1999) 18, 4699 ± 4709 ã 1999 Stockton Press All rights reserved 0950 ± 9232/99 $15.00 http://www.stockton-press.co.uk/onc
Anti-apoptotic activity of p53 maps to the COOH-terminal domain and is retained in a highly oncogenic natural mutant Patrice Lassus1, Christelle Bertrand1,3, Olivier Zugasti1, Jean-Philippe Chambon1, Thierry Soussi2, DanieÁle Mathieu-Mahul1 and Urszula Hibner*,1 1
Institut de GeÂneÂtique MoleÂculaire, CNRS UMR 5535, 1919 route de Mende, 34293 Montpellier Cedex 5, France; 2CNRS UMR 218, Institut Curie, 26 rue d'Ulm, 75248 Paris Cedex 05, France
The tumour suppressor p53 plays a complex role in the regulation of apoptosis. High levels of wild type p53 potentiate the apoptotic response, while physiological range, low levels of the protein have an anti-apoptotic activity in serum starved immortalized ®broblasts. Here we report that primary ®broblast-like cells that show normal growth control are also eciently protected from apoptosis by the endogenous p53 activity. The capacity to inhibit apoptosis is not restricted to the wild type protein: the R?H175 p53 mutant fully retains the anti-apoptotic activity of the wild type p53, providing a possible explanation for its high oncogenicity. Using a series of point and deletion mutants of p53 under the control of tetracycline-regulated promoter we show that certain mutants, like the wild type, protect cells at low levels but lead to apoptosis when overexpressed. This latter eect is lost upon deletion of a proline-rich domain in the NH2 part of the protein. The antiapoptotic activity can be mapped to the extreme carboxy-terminal part of the protein and is therefore independent of other well characterized p53 activities. Our results add a new level of complexity to the network of interactions mediated by p53 in normal physiology and pathology. Keywords: apoptosis; oncogenesis; p53
Introduction Expression of the tumour suppressor p53 is tightly controlled during embryonic development and in the adult (Gottlieb et al., 1997; Komarova et al., 1997; Montes de Oca Luna et al., 1995; Schmid et al., 1991). In most adult tissues there is a low, basal level of p53 expression (MacCallum et al., 1996). Genotoxic insults and other stress conditions lead to accumulation of p53, mainly through its increased stability due to posttranslational modi®cations. The p53 protein displays multiple biochemical activities, including sequence speci®c transactivation and transcriptional repression (reviewed in Bates and Vousden, 1996; Haner and Oren, 1995; Hansen and Oren, 1997; Ko and Prives, 1996; Levine, 1997). Furthermore, p53 interacts with many viral and cellular proteins and there is convincing evidence that some of its activities are not dependent *Correspondence: U Hibner 3 Current address: Laboratoire de DieÂrenciation et Croissance, INRA, 2, place Viala, 34060 Montpellier Cedex 1 Received 24 September 1998; revised 15 March 1999; accepted 16 March 1999
on direct transcriptional control but presumably rely on speci®c protein ± protein interactions (Caelles et al., 1994; Haupt et al., 1995; Ruaro et al., 1997; Walker and Levine, 1996). The dierent biochemical activities of wild type p53 translate into several biological functions. The best documented of these are the arrest of the cell cycle, either at G1 or G2/M, and the induction of apoptosis. Cell cycle arrest depends on the transactivation of target genes by wild type p53 (Deng et al., 1995; Hermeking et al., 1997). In contrast, the involvement of the transcriptional activity of p53 in the induction of apoptosis is far from clear and seems to depend on the experimental system used (Caelles et al., 1994; Haupt et al., 1995; Polyak et al., 1997; Yonish-Rouach et al., 1996). It has been suggested that the transcriptional repression of speci®c genes, rather than activation, correlates with apoptosis induction by p53 (Murphy et al., 1996). However, both transcriptional activation and repression require the intact NH2 terminal domain of p53 (Subler et al., 1994), while p53 dependent apoptosis induction has been reported with an inactivating double point mutation in this domain (Haupt et al., 1995). We have recently described another biological function of the wild type p53 protein, namely the inhibition of apoptosis induced in immortalized ®broblasts subjected to serum deprivation (Lassus et al., 1996). We report here that this anti-apoptotic activity of p53 is also present in REF52 rat ®broblasts, an established cell line which strongly resembles nonimmortalized, primary cells in two respects: stringent growth control (Franza et al., 1986) and the induction of senescence rather than transformation in response to Ras activation (Serrano et al., 1997). Furthermore, the inhibition of apoptosis by p53 requires the carboxyl terminal domain of the protein but can be uncoupled from its transcriptional regulation activity. The majority of naturally occurring p53 mutants isolated from human tumours are single amino acid substitutions localized in the central domain of the protein (Hainaut et al., 1997). Most of these mutants have lost the ability to bind to the consensus p53 binding sites on DNA and are thus incapable of directly transactivating genes targeted by wild type p53 (Ory et al., 1994) Moreover, some of the natural mutants acquire a gain of function phenotype (Dittmer et al., 1993), conceivably related to their altered speci®city of DNA sequence recognition. Interestingly, a naturally occurring hot spot mutant of p53 (the 175R?H substitution), retains the antiapoptotic activity even though it has lost the activities
Anti-apoptotic activity of p53 P Lassus et al
4700
of transactivation of consensus target genes, the transcriptional repression and the induction of apoptosis (Friedlander et al., 1996; Ory et al., 1994). Our results oer a possible explanation for the `gain of function' phenotype of the H175 mutant.
Results Anti-apoptotic activity of p53 is present in REF52 cells Our previous study established that low levels of wild type p53 protect immortalized murine ®broblasts from apoptosis caused by serum withdrawal (Lassus et al., 1996). We set out to extend our investigation to cells whose growth control mimicks the behaviour of primary ®broblasts (Franza et al., 1986; Serrano et al., 1997). Two such cell lines, a primary human ®broblast line AS 198 and rat ®broblastic REF52 cells express low levels of wild type p53 (data not shown) and are highly resistant to apoptosis induced by culture in serum-free medium (Figure 1). Over 95% of these cells survive 48 h of serum withdrawal, similarily to Balb/c 3T3 immortalized murine ®broblasts (p53+). In contrast, Balb/c 10.1 (p537) cells undergo signi®cant cell death in the same time interval. In order to see if the resistance to apoptosis in REF52 cells was due to low levels of p53 expression, we ®rst performed transient transfection experiments aimed at either increasing the p53wt content of cells or, on the contrary, inhibiting the activity of the
Figure 1 Primary ®broblasts are resistant to apoptosis induction by serum withdrawal. Subcon¯uent cultures were rinsed in serumfree medium and cultured in the absence of serum. At indicated times ¯oating and adherent cells were harvested, pooled and counted. Viability was estimated by Trypan Blue exclusion. Over 95% of cells with endogenous wild type p53 survive 48 h of serum withdrawal, while the same treatment induces apoptosis in p537 cells. The results represent the mean+s.d. derived from triplicate counts. *: AS 198, primary human ®broblasts (p53+), &: REF52, rat ®broblasts (p53+), ^: Balb/C 3T3, immortalized mouse ®broblasts (p53+), ~: 10.1, immortalized mouse fibroblasts (p537)
endogenous p53. Two strategies were adopted for inhibiting endogenous p53 in REF52 cells: the forced expression of the HPV16 E6 protein and the expression of a series of dominant negative mutants of p53. These treatments should, respectively, direct p53 towards the degradation pathway (Schener et al., 1990) or inhibit wild type p53 activity by blocking it in inactive hetero-tetramers with mutant form of the protein. We veri®ed the expression of the E6 cDNA by RT ± PCR (data not shown) and the expression and the subcellular localization of wild type and mutant p53 in transfected cells by indirect immunofluorescence (Figure 2). Figure 3 shows the eect of perturbing endogenous p53 activity on the sensitivity of REF52 cells to undergo apoptosis in response to serum withdrawal. Exponentially growing cells were co-transfected with the plasmid encoding a truncated form of rat CD2 antigen lacking the intra cytoplasmic domain and the various constructs described previously. Staining with anti-CD2 antibody allowed us to identify the transfected population, usually 5 ± 10% of cells. The cells were then grown in the presence or absence of foetal calf serum for 14 h, collected and analysed by ¯ow cytometry.
Figure 2 Modulation of endogenous p53 expression in REF52 cells. Subcon¯uent REF52 cells were transiently transfected with a control, empty vector or expression vectors encoding dominant negative mutants of p53 (H273 and H175) or p53wt. Expression of p53 was veri®ed by indirect immuno¯uorescence. Nuclei are visualized by Hoechst 33258 dye (left panels), p53 is detected by PAb 240 monoclonal antibody. Endogenous p53 is below the limit of detection, transfected p53wt localizes in the nucleus, p53H175 mutant gives rise to both nuclear and cytoplasmic staining, p53H273 is mainly cytoplasmic but there is no exclusion from the nuclei. The same experiment repeated with the DO1 monoclonal antibody, speci®c for the human p53 gave identical results
Anti-apoptotic activity of p53 P Lassus et al
4701
Figure 3 Perturbation of endogenous p53 activity in REF52 cells aects the apoptotic phenotype upon serum starvation. Subcon¯uent cells were co-transfected with CD2 expression vector and either control, empty, vector, p53wt, HPV16 E6, p53H273 or p53H175 mutant encoding expression constructs. After 24 h cells were washed with serum-free medium and cultured for additional 14 h under serum-free conditions. Floating and adherent cells were then harvested and pooled. The transfected cell population, which typically represented about 10% of total cells, was detected by staining with anti-CD2 antibody while dead cells were labelled by Annexin V, as described in Materials and methods. The percentage of apoptotic cells in the transfected population is shown. Grey bars: control cultures (+FCS), Open bars: experimental cultures (7FCS). Means+s.d. of triplicate cultures are shown
The percentage of apoptotic cells detected by annexin V labelling in the transfected populations is shown. Under our experimental conditions there was little toxicity due to transfections, as judged by the low number (1 ± 3.5%) of annexin V positive cells in cultures maintained in the presence of serum. Serum withdrawal had a very slight eect on apoptosis of the cells transfected with the control, empty vector. In agreement with our previous results, introduction of excess wild type p53 rendered the cells signi®cantly more sensitive to apoptosis caused by serum withdrawal. Similarly, inhibition of the endogenous p53 activity either by the HPV16 E6 protein or dominant negative mutant of p53 (R?H273 point mutant) lead to increased apoptotic response of serum starved REF52 cells. Forced expression of several other p53 mutant proteins (G?P248, R?P273 and R?P175) gave rise to identical phenotype (data not shown). From these data it would appear that both inhibition of the endogenous p53 and the increase of the wild type protein expression give rise to the pro-apoptotic phenotype, suggesting that the endogenous, low levels of p53 protect REF52 cells from serum withdrawalinduced apoptosis. However, in contrast to the p53H273, transfection of the p53H175 mutant into REF52 cells had little, if any, eect on the anti-apoptotic activity of the endogenous wild type p53 (Figure 3). This result is unexpected, since p53H175 has a strong dominant negative eect on other activities of wild type p53, including its DNA binding and transactivation (Farmer et al., 1992; Kern et al., 1992). The R?H175 p53 mutant has no dominant negative activity on the anti-apoptotic function of p53wt in REF52 In order to further analyse the phenotype of REF52 cells transfected with mutant p53 we have established stable clones expressing either the H273 or the H175 mutant proteins. We chose to analyse in detail two
independent clones from each transfection, expressing comparable levels of the mutant proteins (Figure 4a). The expression levels observed in the stable clones are 30 ± 100-fold lower than those obtained by transient transfection (since less than 10% of cells are transfected in the latter case). The subcellular distribution of p53 in the stably transfected clones was analysed by immuno¯uorescence using PAb 240 and DO1 monoclonal antibodies (data not shown). Similarly to the distribution observed in transient transfectants, p53H175 was localized both in the nucleus and the cytoplasm, while p53H273 was mainly cytoplasmic, but some nuclear staining was also observed. Next we veri®ed the viability of the clones subjected to serum withdrawal. As shown in Figure 4b, the clones transfected with the control, empty vector displayed 9% loss of viability after 48 h of serum withdrawal. Consistent with the results of transient transfection assays, cells expressing the H273 mutant protein died at the signi®cantly higher rate, with about 30% of cells dead at the end of the experiment. On the other hand, the H175 expressing cells were resistant to serum withdrawal with 8 ± 11% cells dying after 48 h of starvation. These results fully con®rm the data obtained in transient transfection experiments, showing that the p53H175 mutant has no dominant negative eect on the anti-apoptotic activity of the endogenous wild type p53 in the REF52 cells. While human and murine p53 form functional hetero-tetramers, such complex formation between rat and human proteins has not been tested. We consider the lack of interaction between human and rat proteins unlikely, since all other human p53 mutant proteins tested appear to exert a dominant negative eect on the endogenous rat p53. However, we cannot formally exclude that the human H175 mutant might not interact eciently with its wild type rat counterpart. Therefore, we tested the intrinsic anti-apoptotic activity of the H175 mutant.
Anti-apoptotic activity of p53 P Lassus et al
4702
Figure 4 Inhibition of endogenous p53 by mutant proteins in stably transfected REF52 cells. Clonal cell lines expressing either the p53H175 or p53H273 proteins have been isolated. Cells transfected with the empty expression vector were used as control. (a) Western blot analysis of p53 expression in two H175 and two H273 expressing lines. Equal amounts of protein (12 mg) were loaded in each lane, including the samples from transiently transfected cells. p53 was revealed by DO-1 monoclonal antibody and GAPDH by a polyclonal antibody (a kind gift of Dr Blanchard). (b) Viability of stably transfected cell lines was assayed by Trypan Blue exclusion. Sub-con¯uent cells were subjected to serum withdrawal for 48 h, ¯oating and adherent cells were then pooled and analysed. Results represent means+s.d. of three independent cultures
The p53H175 mutant protein retains the anti-apoptotic activity We investigated the capacity of p53H175 mutant to prevent apoptosis in the absence of wild type p53. We have used an assay based on transient transfections into murine p53- ®broblasts (10.1 cells, Harvey and Levine, 1991) of wild type and mutant p53 cDNAs cloned under the tetracycline repressible promoter (Gossen and Bujard, 1992). The regulation of p53 expression was veri®ed by Western blot analysis (Figure 5a). p53 is easily detectable in cells grown in the absence but not the presence of tetracycline. Even though the sensitivity of the Western analysis is insucient for detection of very low expression levels, we reproducibly ®nd that in our transient transfection assays, even the highest concentrations of tetracycline greatly reduce gene expression but do not totally switch it o (Lassus et al., 1996). This point is illustrated in Figure 5b showing the results of immunoprecipitation of wild type p53 from transiently transfected 10.1 cells grown in the presence or absence of 2 mg/ml of tetracycline. The low, barely detectable level of p53 observed in cells grown in tetracycline are comparable to that seen for endogenous p53 in the REF52 ®broblasts (results not shown).
We have next tested the eect of high and low levels of expression of dierent p53 mutants in 10.1 cells on apoptosis induced by serum withdrawal. Cells were transfected by appropriate inducible p53 encoding vectors and maintained in the presence of 2 mg/ml of tetracycline and 10% FCS for 24 h. After washing, the cells were subjected to serum withdrawal for 14 h either in the presence or the absence of tetracycline. The percentage of apoptotic cells in the transfected populations was assayed by annexin V binding. The slight toxicity, due to the transfection, was similar in all the samples tested (Figure 5c). Results shown in Figure 5d indicate that the cells transfected with the control, empty pUHD-10.3 vector were sensitive to apoptosis induction due to serum withdrawal, as expected for the p537 cells. Forced expression of low levels of wild type p53 (+Tc) gave rise to signi®cant protection from apoptosis, while neither high nor low levels of p53H273 mutant protein protected these cells from serum starvation induced apoptosis. On the other hand, the p53H175 protein expressed at low levels gave as much protection from apoptosis as the wild type protein. Similarly to the wild type p53, the H175 mutant loses its antiapoptotic activity when strongly over-expressed in this cell line.
Anti-apoptotic activity of p53 P Lassus et al
4703
Figure 5 p53H175 mutant protein has an intrinsic anti-apoptotic activity. The 10.1 (p537) cell line was stably transfected with the pUHD 15-1 vector (Gossen and Bujard, 1992) giving rise to a 10.10 line constitutively expressing the chimaeric tetR-VP16 protein (Lassus et al., 1996). These cells were transiently transfected with wild type and mutant forms of p53 (H273 and H175) cloned in the pUHD 10-3 vector (Gossen and Bujard, 1992) under the control of a tetracycline repressible promoter. Following transfection, cells were cultured in medium supplemented with 10% foetal calf serum without tetracycline or with 2 mg/ml tetracycline for 16 h at 378C. (a) Western blot analysis of tetracycline regulated p53 expression. 10.10 cells were transfected with the empty pUHD 10-3 vector (control), or the vectors containing the wild type or mutant forms of p53, split into two plates and cultured in absence or presence of 2 mg/ml tetracycline for 24 h. Equal amount of protein from whole cell extracts were loaded in each lane. p53 was revealed with NH2 speci®c monoclonal antibody DO1. (b) 10.1 cells were transfected with the control pUHD 10-3 or the same vector encoding wt p53, cultured without tetracycline or in the presence of 2 mg/ml tetracycline and labelled with 35 S-methionine for 45 min. The radiolabelled proteins were immunoprecipitated with PAb 421 monoclonal antibody and analysed by SDS ± PAGE. The position of p53 is indicated. (c) 10.10 cells were co-transfected with vectors constitutively expressing CD2 antigen and p53 cloned under the control of tetracycline repressible promoter, as described in Materials and methods. Cells were harvested 30 h after transfection and apoptosis was measured by Annexin V labelling in the transfected (CD2+) population. Under the conditions of culture in the presence of 10% of FCS there is no signi®cant toxicity in any of the transfectants either in the absence or in the presence of 2 mg/ml of tetracycline. (d) The experiment was performed as in (c), except that the cells were cultured without serum for the last 14 h before harvesting. Solid bars: no tetracycline, open bars: 2 mg/ml tetracycline added to the culture medium
Anti-apoptotic activity of p53 P Lassus et al
4704
Mapping of the domains responsible for the anti-apoptotic activity of p53 We have analysed several deletion mutants of p53, schematically represented in Figure 6a, for their capacity to inhibit apoptosis of serum starved ®broblasts. The same experimental protocol was used as in the experiments described above. The expression of the dierent proteins was eciently regulated by tetracycline (Figure 6b). The results in Figure 6c show that the anti-apoptotic activity is unaected by the deletion of the entire transactivation domain (p53D11-69), demonstrating unambiguously that inhibition of apoptosis is independent of transcriptional regulation by p53. Removal of the COOH-terminal 38 amino acids (D355-393) resulted in the loss of the anti-apoptotic activity. The deletion of the proline-rich region (D62-91) gave rise to an interesting phenotype: the mutant displayed a full protective activity both at low and high expression levels. Finally, a mini-protein retaining only the ®rst 14 amino acids and the oligomerization as well as the COOH-terminal domains of p53 (p53DD; Shaulian et al., 1992) was as ecient as the wild type p53 in protecting cells against apoptosis, again independently of its expression level. Taken together these results indicate that the antiapoptotic activity of p53 can be localized the COOHterminal domain of the protein, while the loss of this activity upon p53 over-expression is dependent on the proline-rich domain. Discussion The wild type p53 tumour suppressor protein has multiple biological activities; the ®nal eect on cell fate of its activation is strongly dependent on the cellular context (reviewed in Haner and Oren, 1995; Hansen and Oren, 1997; Ko and Prives, 1996; Levine, 1997). Some aspects of cellular physiology which determine the overall response to p53 activation have been identi®ed. It has been reported that the level of p53 accumulation (Chen et al., 1996), as well as the induction of p21WAF1, can tip the balance of the decision between cell cycle arrest and apoptosis (Polyak et al., 1996). Furthermore, the presence or absence of survival factors can also be decisive for the type of response ellicited in a cell (Canman et al., 1995). It is clear, nevertheless, that the eect of p53 activation on cellular destiny depends on more, as yet unidenti®ed, parameters. Anti-apoptotic activity of p53 in REF52 cells We reasoned that the anti-apoptotic function of basal levels of this tumor suppressor is likely to depend on the cellular context. In particular, it was important to verify that this activity is not a peculiarity of the immortalized cell lines we have used in our initial work (Lassus et al., 1996). We ®nd that both primary human ®broblasts and rat REF52 cells are resistant to serum withdrawal-induced apoptosis. Both types of cells express basal levels of p53 and, at least in the REF52, the protection against apoptosis is abrogated
when p53 expression or activity is inhibited. The inhibition was achieved by two dierent approaches: accelerated degradation of the endogenous protein by expressing the E6 protein of HPV16 and inhibition of the endogenous p53 activity by the expression of dominant negative p53 mutants, the latter performed either by transient or stable expression of mutant proteins. Both experimental approaches gave rise to the same phenotype, namely induction of apoptosis upon serum withdrawal in the transfected cell population. Moreover, for the experiments performed in transiently transfected cells, the results re¯ect a population of cells and thus cannot be due to an artefactual selection of cells with variant responses. We therefore conclude that the physiological context of a cell line closely resembling primary ®broblasts (Franza et al., 1986; Serrano et al., 1997) is compatible with the antiapoptotic activity of low levels of wild type p53. The p53H175 mutant protein retains the anti-apoptotic activity One surprising ®nding is that the p53H175 mutant protein does not inhibit the anti-apoptotic activity of endogenous p53 in the REF52 cells. Furthermore, expression of p53H175 protein in p537 ®broblasts protects them from apoptosis. In fact, the eects of the wild type or the H175 mutant of p53 on the survival of these cells are very similar (Figure 5). The R?H substitution at the position 175 is frequent in cancer and correlates with poor prognosis for the patients with colorectal carcinoma (Goh et al., 1995). The tertiary structure of the mutant protein is predicted to be profoundly altered by this substitution (Cho et al., 1994) and, indeed, this compromises many activities of wild type p53 (Friedlander et al., 1996; Ory et al., 1994). Moreover, this point mutant has a strong dominant negative eect on the wild type p53 (Farmer et al., 1992; Kern et al., 1992), implying that its capacity to form hetero-oligomers with the wild type protein is intact. Despite the reported loss of all previously tested functions of p53wt, the H175 substitution is a `gain of function' mutation (Dittmer et al., 1993). It follows that this protein has a biological activity, which is not shared by all p53 mutants. Our discovery that p53H175 retains the anti-apoptotic activity characteristic of the wild type protein suggests a possible mechanism for the `gain of function' phenotype of this mutation. The p53H175 mutant can inhibit apoptosis of hematopoietic cells (Kremenetskaya et al., 1997) and, as reported after the submission of this manuscript, it protects colorectal cancer cells from etoposide induced apoptosis, strengthening the argument of implication of the anti-apoptotic activity of mutant p53 in cancerogenesis (Blandino et al., 1999). Anti-apoptotic activity of 53 is independent of transcriptional regulation The p53H175 mutant protein neither transactivates nor represses any known target genes of wild type p53. Since p53 mutants with altered target speci®cities have been described (Chen et al., 1993; Ludwig et al., 1996), it remained possible that p53H175 could transcriptionally regulate other, as yet undiscovered, target genes.
Anti-apoptotic activity of p53 P Lassus et al
4705
B
Figure 6 Mapping of the anti-apoptotic activity of p53wt. 10.10 cells were co-transfected with vectors constitutively expressing CD2 antigen and wild type p53 or p53 deletion mutants cloned under the control of tetracycline repressible promoter. Cells were grown in the medium supplemented with 10% FCS in the absence or presence of 2 mg/ml of tetracycline for 16 h. Cells where then rinsed in serum-free medium and the culture was continued for the following 14 h in the medium without serum. The initial tetracycline concentrations were maintained as indicated. Floating and adherent cells were harvested, pooled and analysed by ¯ow cytometry for CD2 expression (transfected cells) and Annexin V labelling (apoptotic cells). (a) Schematic representation of functional domains of a p53 monomer. The deletion mutants used in the present study are indicated. (b) Western blot analysis of tetracycline regulated p53 expression. 10.10 cells were transfected with indicated constructs, split into two plates and cultured in absence or presence of 2 mg/ml tetracycline for 24 h. Equal amount of protein from whole cell extracts were loaded in each lane. The ®gure is a composite of two blots revealed with NH2 speci®c monoclonal antibody DO1 (control, wt, DPro and DCOOH) and COOH speci®c mAb HR231 (DNH2 and DD). (c) Relative numbers of apoptotic cells in populations transfected with indicated forms of p53 expressed at high levels in the absence of tetracycline (hatched bars) and low levels in the presence of 2 mg/ml of tetracycline (open bars). The percentage of apoptotic cells in the populations transfected with the empty control vector varied between 19 and 25% and was not in¯uenced by tetracycline. In order to facilitate representation of the data, this value was arbitrarily set at one for each experiment. The results represent means+s.d. of triplicate cultures
Anti-apoptotic activity of p53 P Lassus et al
4706
However, the p53 D11-69, which protects cells from apoptosis as eciently as the wild type protein, deletes the entire transactivation domain, required for transcriptional activation and repression (Pietenpol et al., 1994; Subler et al., 1994; Unger et al., 1993), demonstrating that the inhibition of apoptosis is independent of transcriptional regulation by p53. Removal of the proline-rich region (aa 61 ± 92) preserves the anti-apoptotic activity of highly expressed p53 We previously found that the anti-apoptotic activity of p53wt is only revealed when the protein is expressed at basal, close to physiological, levels (Lassus et al., 1996). Since highly expressed p53wt potentiates the apoptotic response in our experimental system, we reasoned that the anti-apoptotic activity is masked under such conditions. However, p53H175 displays no proapoptotic activity (Friedlander et al., 1996 and this work) but nevertheless, similarly to the wild type protein, loses its protective eect when over-expressed in the 10.1 cells. Furthermore, p53H175 is inert in respect to transcriptional control of p53 target genes. Therefore, since the dependence of the anti-apoptotic activity on the level of expression of this mutant cannot be due either to apoptosis induction or transactivation by the highly expressed protein, we presume it results from protein ± protein interactions. We speculate that dierent protein partners of p53 target it to its alternative functions. In this scenario, the partner responsible for the anti-apoptotic activity would bind p53 expressed at basal level, while the
partner responsible for the inhibition of this activity would require higher levels of p53 for ecient interaction. This hypothesis, schematically represented in Figure 7, is consistent with the behaviour of p53DPro mutant. Indeed, the deletion of the proline rich domain uncouples p53 anti-apoptotic activity from its level of expression. These results suggest that an inhibitory partner interacts with p53 via the prolinerich domain. This domain is dispensable for the transactivation, but necessary for the ecient tumour growth suppression by p53 (Walker and Levine, 1996) and has recently been shown to be essential in at least one case of p53 dependent apoptosis induction (Sakamuro et al., 1997). These results, together with the data reported here, designate the proline rich region as an important regulatory domain of p53. The anti-apoptotic activity of p53 maps to the COOH-terminal domain of the protein p53DD mini-protein is a dominant negative transforming mutant both in vitro (Shaulian et al., 1992) and in vivo (Bowman et al., 1996) assays. It harbours a deletion of 288 amino acids (D15-301) which leaves the initial 14 aa and the oligomerization and the COOHterminal domains. The transforming activity of the DD protein depends on its ability to form inactive heterooligomers with the endogenous p53 (Shaulian et al., 1995). We show (Figure 6) that this mini-protein eciently protects ®broblasts from serum withdrawal induced apoptosis. The anti-apoptotic phenotype is observed in the p537 background and therefore does
Figure 7 A speculative model for apoptosis regulation as a function of p53 expression level. Apoptosis inhibition is independent of p53 transactivation activity and resides in the COOH-terminal domain of the protein. It is thus presumably due to an interaction with a speci®c partner (black circle), which would occur independently of p53 expression level. The anity of interaction with another potential partner (diamond) might be lower, so that the binding could only occur if p53 were present at elevated concentrations. This tertiary complex would no longer target p53 towards its anti-apoptotic function, but would allow the expression of activities giving rise to the tumour suppressor phenotype of p53. Similar types of interactions could occur with mutant p53 proteins, explaining the loss of the anti-apoptotic activity at high expression levels for the mutants inert for other activities of the wild type protein
Anti-apoptotic activity of p53 P Lassus et al
not require heterodimerization. Consistent with the interpretation of the role of proline-rich region in the anti-apoptotic activity of p53, p53DD, which is devoid of the proline-rich domain, protects cells against apoptosis independently of its expression level. Conversely, removal of the last 38 amino acids renders p53 totally inactive in our assay for protection against apoptosis. These results map the anti-apoptotic activity to the extreme C-terminal region of p53. The extreme carboxy-terminal region of p53 contains two major biochemical activities of p53: single stranded DNA binding and an auto-regulatory function (Hupp and Lane, 1994; Wang et al., 1993). Removal of this domain improves p53 stability (Hansen et al., 1996) and activates its binding to dsDNA target sequences (Hupp et al., 1992). The p53DCOOH retains the transactivation (Pietenpol et al., 1994) and the apoptosis induction activities (Chen et al., 1993). The mutant is active both in growth suppression (Pietenpol et al., 1994) and transformation suppression (Shaulian et al., 1995) assays. Several arguments suggest that the inhibition of speci®c DNA binding occurs by a direct interaction of the COOH domain with the central region of the protein (Hupp et al., 1993, 1995). This negative regulation appears to be enhanced in several p53 mutants, including H273, since their DNA binding activity can be restored by molecules which displace the COOH terminus (Abarzua et al., 1996; Hupp et al., 1993; Wieczorek et al., 1996). The situation is dierent in the case of H175 mutant due to its altered conformation and, indeed, the H175 mutant is unstable and totally refractory to re-activation (Abarzua et al., 1996; Friedlander et al., 1996; Wieczorek et al., 1996). However, the unfolded conformation of H175 might allow protein ± protein interactions with domains inaccessible in mutants such as H273. Several protein partners which bind to the C-terminal portion of p53 have been identi®ed. They include components of the basic transcription machinery, such as TBP (Horikoshi et al., 1995) and the DNA helicase subunits of the TFIIH (Wang et al., 1996) as well as proteins involved in DNA repair, for example CSB (Wang et al., 1996), PARP (Vaziri et al., 1997) and Ref1, a redox/repair protein (Jayaraman et al., 1997). Most interestingly, members of the heat shock protein family Hsp70, Hsc70 and their bacterial homologue dnaK interact with several domains of p53, including its carboxy-terminus (Fourie et al., 1997; Hansen et al., 1996). Hsp 70 plays a role in protection of cells against stress-induced apoptosis (Mosser et al., 1997; Polla et al., 1996), making it an attractive candidate for mediating of the anti-apoptotic activity of p53. The question of the physiological role of the inhibition of apoptosis by p53 has not been addressed directly. A study of multi-stage chemical carcinogenesis of the skin in p53+/+ and p537/7 mice shows that absence of p53 inhibits the appearance of the initial benign lesions, while it accelerates their progression towards fully malignant tumours (Kemp et al., 1993). These results suggest that cells containing p53wt survive the initial stages of oncogenesis better than their p537 counterparts, possibly re¯ecting the anti-apoptotic activity of p53 in vivo. The ®nding reported here of the preservation of the anti-apoptotic activity by an
oncogenic form of p53 raises the possibility that it could also be involved in some cases of tumour progression. Materials and methods Plasmids Wild type p53 is a full cDNA clone of human p53 (aa 1 ± 393), H175 and H273 are point mutants of human p53 in which arginine residue are replaced by histidine at the indicated positions. The deletion mutants of human p53 are: D-Proline rich domain (amino acids 61 ± 92), a kind gift of Dr Laurent Debussche, DNH2 (amino acids 11 ± 69), COOH (amino acids 355 ± 393), both kindly provided by Dr John Jenkins, p53DD (amino acids 15 ± 315), a kind gift of Dr Moshe Oren. The cDNAs were cloned in the BamHI site of pCDNA3 vector (Invitrogen) and in the EcoRI site of the pUHD 10-3 tetracycline regulated expression vector (Gossen and Bujard, 1992). The coding sequence of E6 protein of human papilloma virus 16 is cloned under the control of Harvey murine sarcoma virus LTR (Sedman et al., 1992) and is a kind gift of Dr Nicole Basset-Seguin. Truncated rat CD2 antigen, lacking the intracytoplasmic domain is expressed from the CMV promoter in the pKV461 vector, kindly provided by Dr Chris Norbury. Cell culture and transfections Adherent cells were cultured in DMEM medium suplemented with 10% foetal calf serum in 5% CO2 at 378C. For transient transfections, sub-con¯uent cells were co-transfected with the pKV461 encoding the truncated version of rat CD2 antigen and an appropriate construction in the pCDNA3 or pUHD 10-3 vector. Following transfection, cells were cultured in medium supplemented with 10% foetal calf serum and indicated tetracycline concentrations for 16 h at 378C, then rinsed once and incubated in serum-free medium with the indicated concentrations of tetracycline for additional 14 h, after which ¯oating and adherent cells were harvested, pooled and analysed. The toxicity due to the transfection typically did not exceed 5% of the cells. The transfected cells were identi®ed by labelling with anti-rat CD2 mAb OX-34 (Cedralane Laboratories). Stable transfections were performed with appropriate mutants cloned in the pEF-BOSNeo vector (Mizushima and Nagata, 1990) or the empty vector as control. Twenty-four hours following transfection, cells were selected for growth in the presence of 1 mg/ml of G418. Resistant colonies were isolated and ampli®ed. Apoptosis assay The ¯ow cytometry assay for apoptosis in the transiently transfected ®broblasts was previously described (Lassus and Hibner, 1998). Brie¯y, ¯oating and adherent cells were collected, pooled and centrifuged. The pellets were taken up in Annexin V binding buer and incubated with Annexin-VFLUOS (Boehringer Mannheim). The cells were then ®xed in 3.7% formaldehyde (Sigma), rinsed and incubated at room temperature for 2 h with biotinyled anti-rat CD2 mAB OX34 (Cedarlane), rinsed with PBS and incubated for 30 min with Streptavidin-Tri Color (Caltag). The quanti®cation of apoptosis on transfected cells population was determined by ¯ow cytometry (Becton and Dickinson FacsScan) with FL-1 channel for Annexin-V-FLUOS and FL-3 channel for CD2Tri Color. Western blot analysis The Western analysis was performed as previously described (Leroy-Viard et al., 1995).The antibodies used for detection
4707
Anti-apoptotic activity of p53 P Lassus et al
4708
of the p53 protein were mAb HR231 and DO1 recognizing the COOH and NH2 epitopes, respectively. Immunoprecipitations 106 cells were pre-incubated for 45 min in methionine and cysteine-free DMEM medium supplemented with 5% foetal calf serum. Proteins were labelled by addition of 200 mCi of 35 S labelling mix (EXPRE35S35S, NEN) for 45 min, following which the cells were collected and lysed in TENN buer (50 mM Tris, pH 7.4; 5 mM EDTA; 0,5% NP-40; 150 mM NaCl; 1 mM PMSF). The cell lysate was pre-adsorbed with protein A-Sepharose (Pharmacia) for 30 min at 48C and incubated overnight at 48C with protein A-Sepharose and 1 mg of anti p53 monoclonal antibody PAb 421 (Oncogene Research). The beads were washed four times with TENN buer and the sample analysed by SDS ± PAGE. Immuno¯uorescence Cells were grown and transfected on LabTek glass slides (Nunc). Twenty four hours later the cells were rinsed with PBS, ®xed in 3.7% formaldehyde for 10 min at room
temperature, rinsed again with PBS and permeabilized with acetone at 7208C for 30 s. Cells were then incubated ®rst with either PAb 240 or DO1 monoclonal anti-p53 antibodies, then with biotinylated goat anti-mouse Ig antibodies and ®nally with Texas Red conjugated streptavidin (Amersham) and Hoechst 33258 dye (Sigma). The ¯uorescence was observed at the 4006 magni®cation with the Zeiss Axiophot ¯uorescence microscope.
Acknowledgements We are very grateful to Drs John Jenkins, Moshe Oren and Laurent Debussche for providing plasmids with mutant p53, Dr Nicole Basset-Seguin for the E6, Dr Chris Norbury for the CD2 expression vectors, Dr Alain Sarasin for the AS 198 cells and Dr Jean Marie Blanchard for the anti-GAPDH antibody. We thank Dr Bob Hipskind for comments on the manuscript. Supported by Association pour la Recherche contre le Cancer, Ligue Nationale contre le Cancer, INSERM and CNRS. P Lassus is a recipient of a fellowship from ARC.
References Abarzua P, LoSardo JE, Gubler ML, Spathis R, Lu Y-A, Felix A and Neri A. (1996). Oncogene, 13, 2477 ± 2482. Bates S and Vousden K. (1996). Curr. Opin. Genet. Dev., 6, 12 ± 18. Blandino G, Levine AJ and Oren M. (1999). Oncogene, 18, 477 ± 485. Bowman T, Symonds H, Gu L, Yin C, Oren M and Van Dyke T. (1996). Genes Dev., 10, 826 ± 835. Caelles C, Helmberg A and Karin M. (1994). Nature, 370, 220 ± 223. Canman C, Gilmer TA, Coutts SB and Kastan MB. (1995). Genes Dev., 9, 600 ± 611. Chen J-Y, Funk WD, Wright WE, Shay JW and Minna J D. (1993). Oncogene, 8, 2159 ± 2166. Chen X, Ko LJ, Jayaraman L and Prives C. (1996). Genes Dev., 10, 2438 ± 2451. Cho Y, Gorina S, Jerey PD and Pavletich NP. (1994). Science, 265, 346 ± 355. Deng C, Zhang P, Harper JW, Elledge SJ and Leder P. (1995). Cell, 82, 675 ± 684. Dittmer D, Pati S, Zambetti G, Chu S, Teresky AK, Moore M, Finlay C and Levine AJ. (1993). Nature Genetics, 4, 42 ± 46. Farmer G, Bargonetti J, Zhu H, Friedman P, Prywes R and Prives C. (1992). Nature, 358, 83 ± 86. Fourie AM, Hupp TR, Lane DP, Sang B-C, Barbosa MS, Sambrook JF and Gething M-JH. (1997). J. Biol. Chem., 272, 19471 ± 19479. Franza BRJ, Maruyama K, Garrels JI and Ruley HE. (1986). Cell, 44, 409 ± 418. Friedlander P, Haupt Y, Prives C and Oren M. (1996). Mol. Cell. Biol., 16, 4961 ± 4971. Friedlander P, Legros Y, Soussi T and Prives C. (1996). J. Biol. Chem., 271, 25468 ± 25478. Goh HS, Yao J and Smiyth DR. (1995). Cancer Res., 55, 5217 ± 5221. Gossen M and Bujard H. (1992). Proc. Natl. Acad. Sci. USA, 89, 5547 ± 5551. Gottlieb E, Haner R, King A, Asher G, Gruss P, Lonai P and Oren M. (1997). EMBO J., 16, 1381 ± 1390. Haner R and Oren M. (1995). Curr. Opin. Genetics Dev., 5, 84 ± 90. Hainaut P, Soussi T, Shomer B, Hollstein M, Greenblatt M, Hovig E, Harris C and Montesano R. (1997). Nucleic Acids Res., 25, 151 ± 157.
Hansen R and Oren M. (1997). Curr. Opin. Cell Biol., 7, 46 ± 51. Hansen S, Hupp TR and Lane DP. (1996). J. Biol. Chem., 271, 3917 ± 3924. Harvey D and Levine AJ. (1991). Genes Dev., 5, 2375 ± 2385. Haupt Y, Rowan S, Shaulian E, Vousden KH and MO. (1995). Genes Dev., 9, 2170 ± 2183. Hermeking H, Lengauer C, Polyak K, He T-C, Zhang L, Thiagalingram S, Kinzler KW and Vogelstein B. (1997). Mole. Cell, 1, 3 ± 11. Horikoshi N, Usheva A, Chen J, Levine AJ, Weinmann R and Shenk T. (1995). Moll. Cell. Biol., 15, 227 ± 234. Hupp TR and Lane DP. (1994). Curr. Biol., 4, 865 ± 875. Hupp TR, Meek DW, Midgley CA and Lane DP. (1992). Cell, 71, 875 ± 886. Hupp TR, Meek DW, Midgley CA and PLD. (1993). Nucleic Acids Res., 21, 3167 ± 3174. Hupp TR, Sparks A and Lane DP. (1995). Cell, 83, 237 ± 245. Jayaraman L, Murthy KGK, Zhu C, Curran T, Xanthoudakis S and Prives C. (1997). Genes Dev., 11, 558 ± 570. Kemp CJ, Donehower LA, Bradley A and Balmain A. (1993). Cell, 74, 813 ± 822. Kern SE, Pietenpol JA, Thiagalingam S, Seymour A, Kinzler KW and Vogelstein B. (1992). Science, 256, 827 ± 830. Ko LJ and Prives C. (1996). Genes Dev., 10, 1054 ± 1072. Komarova EA, Chernov MV, Franks R, Wang K, Armin G, Zelnick CR, Chin DM, Bacus SB, Stark GR and Gudkov AV. (1997). EMBO J., 16, 1391 ± 1400. Kremenetskaya OS, Logacheva NP, Baryshnikov AY, Chumakov PM and Kopnin BP. (1997). Oncol. Res., 9, 155 ± 166. Lassus P, Ferlin M, Piette J and Hibner U. (1996). EMBO J., 15, 4566 ± 4573. Lassus P and Hibner U. (1998). Nuc. Acids Res., 26, 5233 ± 5234. Leroy-Viard K, Vinit M-A, Lecointe N, Jouault H, Hibner U, RomeÂo P.-H and Mathieu-Mahul D. (1995). EMBO J., 14, 2341 ± 2349. Levine AJ. (1997). Cell, 88, 323 ± 331. Ludwig RL, Bates S and Vousden KH. (1996). Mol. Cell. Biol., 16, 4952 ± 4960. MacCallum DE, Hupp TR, Midgley CA, Stuart D, Campbell SJ, Harper A, Walsh FS, Wright EG, Balmain A, Lane DP and Hall PA. (1996). Oncogene, 13, 2575 ± 2587.
Anti-apoptotic activity of p53 P Lassus et al
Mizushima S and Nagata S. (1990). Nuc. Acids Res., 18, 5322. Montes de Oca Luna R, Wagner DS and Lozano G. (1995). Nature, 378, 203 ± 206. Mosser DD, Caron AW, Bourget L, Denis-Larose C and Massie B. (1997). Mol. Cell. Biol., 17, 5317 ± 5327. Murphy M, Hinman A and J LA. (1996). Genes Dev., 10, 2971 ± 2980. Ory K, Legros Y, Auguin C and Soussi T. (1994). EMBO J., 13, 3496 ± 3504. Pietenpol JA, Tokino T, Thiagaligam S, El-Deiry W, Kinzler KW and Vogelstein B. (1994). Proc. Natl. Acad. Sci. USA, 91, 1998 ± 2002. Polla BS, Kantengwa S, FrancËis D, Salvioli S, Franceschi C, Marsac C and Cossarizza A. Proc. Natl. Acad. Sci. USA, 93, 6458 ± 6463. Polyak K, Waldman T, He T-C, Kinzler K and Vogelstein B. (1996). Genes Dev., 10, 1945 ± 1952. Polyak K, Xia Y, Zweler JL, Kinzler KW and Vogelstein B. (1997). Nature, 389, 300 ± 305. Ruaro E, Collavin L, Del Sal G, Haner R, Oren M, Levine AJ and Schneider C. (1997). Proc. Natl. Acad. Sci. USA, 94, 4675 ± 4680. Sakamuro D, Sabbatini P, White E and Prendergast GC. (1997). Oncogene, 15, 887 ± 898. Schener M, Werness BA, Huibregtse JM, Levine AJ and Howley PM. (1990). Cell, 63, 1129 ± 1136. Schmid P, Lorenz A, Hameister H and Montenarh M. (1991). Development, 113, 857 ± 865.
Sedman SA, Hubbert NL, Vass WC, Lowy DR and Schiller JT. (1992). J. Virol., 66, 4201 ± 4208. Serrano M, Lin AW, McCurrach ME, Beach D and Lowe SW. (1997). Cell, 88, 593 ± 602. Shaulian E, Haviv I, Shaul Y and Oren M. (1995). Oncogene, 10, 671 ± 680. Shaulian E, Zauberman A, Ginsberg D and Oren M. (1992). Mol. Cell. Biol., 12, 5581 ± 5592. Subler MA, Martin DW and Deb S. (1994). Oncogene, 9, 1351 ± 1359. Unger T, Mietz JA, Schener M, Yee CL and Howley PM. (1993). Mol. Cell. Biol., 13, 5186 ± 5194. Vaziri H, West MD, Allsopp RC, Davison TS, Wu Y-S, Arrowsmith CH, Poirier GG and Benchimol S. (1997). EMBO J., 16, 6018 ± 6033. Walker K and Levine AJ. (1996). Proc. Natl. Acad. Sci. USA, 93, 15335 ± 15340. Wang XW, Vermeulen W, Coursen JD, Gibson M, Lupold SE, Forrester K, Xu G, Elmore L, Yeh H, Hoeijmakers JHJ and Harris CC. (1996). Genes Dev., 10, 1219 ± 1232. Wang Y, Reed M, Wang P, Stenger JE, Mayr G Anderson ME, Schwedes JF and Tegtmeyer P. (1993). Denes Dev., 7, 2575 ± 2586. Wieczorek AM, Waterman JLF, Waterman MJF and Halazonetis TD. (1996). Nature Med., 2, 1143 ± 1146. Yonish-Rouach E, Deguin V, Zaitchouk T, Breugnot C, Mishal Z, Jenkins JR and May E. (1996). Oncogene, 12, 2197 ± 2205.
4709
