A small synthetic peptide, which inhibits the p53-hdm2 interaction, stimulates the p53 pathway in tumour cell lines1

doi:10.1006/jmbi.2000.3738 available online at http://www.idealibrary.com on

J. Mol. Biol. (2000) 299, 245±253

A Small Synthetic Peptide, which Inhibits the p53-hdm2 Interaction, Stimulates the p53 Pathway in Tumour Cell Lines Patrick CheÁne*, Jean Fuchs, Jacqueline Bohn Carlos GarcõÂa-EcheverrõÂa, Pascal Furet and Doriano Fabbro Novartis, Oncology Department, CH-4002, Basel Switzerland

The hdm2 protein negatively regulates p53 tumour suppressor activity. Upon binding to p53, hdm2 stimulates p53 degradation and inhibits its transcriptional activity. Moreover, the hdm2 protein is overexpressed in various tumours inactivating p53. We report here that an octamer synthetic peptide derived from p53 inhibits the p53-hdm2 interaction in vitro. In cellular assays, this untagged peptide penetrates tumour cells and induces the accumulation of p53. The accumulation of p53 leads to its activation. Two gene products transcriptionally regulated by p53, p21Waf1/Cip1 and hdm2, are induced in the presence of the peptide. When used with tumour cells that overexpress hdm2, the peptide induces the death of these tumour cells by apoptosis. The mode of action of this peptide differs from that of DNA-damaging agents (e.g. cisplatin) in that it does not induce p53 phosphorylation on serine 15. This work validates with a low molecular mass molecule our current knowledge on the regulation of the p53 pathway by the hdm2 protein. It also shows that inhibitors of the p53-hdm2 interaction are very attractive candidates for the activation of the p53 pathway in tumours expressing wild-type p53. # 2000 Academic Press

*Corresponding author

Keywords: p53; hdm2; mdm2; protein-protein interaction; peptide

Introduction In normal cells, the wild-type (wt) p53 protein is present at a very low concentration. However, in response to various stimuli, such as DNA damage, hypoxia, or metabolite changes, it accumulates and induces cell-cycle arrest and/or apoptosis (Levine, 1997). Originally, it was shown that p53 stimulates the synthesis of hdm2 and that the newly synthesised hdm2 protein inhibits the transcriptional activity of p53 because it binds to its transactivation domain (Chen et al., 1993; Momand et al., 1992; Picksley et al., 1994). This leads to the conclusion that p53 and hdm2 form an autoregulatory feedback loop (Picksley & Lane, 1993). It has been established that hdm2 regulates the p53 protein by an additional mechanism that favours the degradation of p53 (Haupt et al., 1997; Kubbutat et al., 1997; Midgley & Lane, 1997). hdm2 acts as a shuttling protein (Tao & Levine, 1999); on interacting with p53, it induces its nuclear export into the cytoplasm, where it is degraded. The inhibition of E-mail address of the corresponding author: [email protected] 0022-2836/00/010245±9 $35.00/0

nuclear export with leptomycin B, a nuclear export inhibitor (Wolff et al., 1997), favours the accumulation of p53 in the nucleus and its activation (Freedman & Levine, 1998; Lain et al., 1999). Inhibition of the interaction between p53 and hdm2 is an attractive target for cancer therapy (Lane & Hall, 1997). The inhibition of this interaction in tumour cells should lead to the accumulation and the activation of p53, and this in turn would induce tumour cell death by apoptosis. All tumour cells expressing the wt p53 protein could therefore be targeted by this approach, assuming that they have a functional p53 pathway. Three independent groups have recently demonstrated the feasibility of such an approach. (1) Lane and collaborators (Bottger et al., 1997) have shown that the expression of a modi®ed thioredoxin protein, which expresses at its surface a peptidic inhibitor of the p53-hdm2 interaction, induces both p53 accumulation and activation. (2) Blaydes and collaborators (Blaydes et al., 1997) have revealed that the microinjection of an antibody (3G5) that binds to hdm2 prevents its association with p53, thus promoting the activity of p53. (3) Finally, Wasylyk # 2000 Academic Press

246 and co-workers (Wasylyk et al., 1999) have shown that a peptidic inhibitor fused to the gluthatione S-transferase (GST) protein induces p53 activity in tumour cells. In addition, Chen and co-workers (Chen et al., 1998) have revealed that antisense oligonucleotides raised against hdm2 can induce p53 activity. These different approaches tend to support the idea that inhibitors of the p53-hdm2 interaction could be used in cancer therapy. However, the 3G5 antibody, the modi®ed thioredoxin, and the GST-fusion proteins are large molecules, and it remains to be established whether small molecules can inhibit the p53-hdm2 interaction in a cellular environment. This is a key issue when protein-protein interactions are targeted for therapeutic applications. Because hdm2 has additional properties, e.g. ubiquitin ligase activity (Honda & Yasuda, 1999), the data observed with the antisense oligonucleotides may not only re¯ect the inhibition of the p53-hdm2 interaction, but could also be the result of the total loss of hdm2 in the cell. Here, we show that an octamer synthetic peptide, which potently inhibits the p53-hdm2 interaction in vitro, stimulates the p53 pathway in tumour cells that express wt p53. The results obtained with this untagged peptide show that it is possible to generate small molecules that inhibit the p53-hdm2 interaction and induce p53 tumour suppressor activity in cancer cell lines.

Results The AP peptide inhibits the p53-hdm2 interaction in vitro The AP peptide is an octamer oligomer that contains four non-proteinogenic amino acid residues (C.G.-E. et al., unpublished results). The sequence of the peptide is the result of an optimisation process that took into account the structural information derived from the X-ray structure of a p53 peptide co-crystallised with hdm2 (Kussie et al., 1996). The potency of the AP peptide has been evaluated in the ELISA hdm2 (Bottger et al., 1996) (Figure 1). The AP peptide showed a high potency in the assay with an IC50 of 5 nM. An analogue of the AP peptide, the IP peptide, in which the important 6-chlorotryptophan residue has been replaced by alanine, shows very low inhibitory activity under the same conditions (Figure 1). Since the AP peptide is very potent in vitro, it was used as a tool to inhibit the p53-hdm2 interaction in cells. However, because the cellular penetration of free peptides is usually low, the AP peptide was used at a concentration of 100 mM. The AP peptide induces p53 accumulation in HCT-116 cells The ®rst expected cellular effect of inhibitors of the p53-hdm2 interaction is that they should induce wt p53 accumulation, because they prevent its association with hdm2 and therefore its degra-

Inhibitor of the p53-hdm2 Interaction

Figure 1. The AP peptide is a potent inhibitor of the p53-hdm2 interaction in vitro. The inhibitory properties of both the AP (*) and the IP (*) peptides were determined in the ELISA hdm2 (BoÈttger et al., 1996). The IC50 of the AP peptide is 5(1) nM (n ˆ 4). The IP peptide did not inhibit the p53-hdm2 interaction in a concentration range 50 nM-2000 mM.

dation. Colon carcinoma HCT-116 cells that express wt p53 (O'Connor et al., 1997) were chosen to evaluate the effect of the AP peptide on p53 accumulation. To measure the accumulation of p53 in the cells, an ELISA was used (see Materials and Methods). This assay shows a linear response when the amount of p53 varies between 0.06 and 0.5 ng (data not shown). The ELISA was used in this linear range to determine the amount of p53 protein present in the cell lysates. The HCT-116 cells were incubated for different lengths of time in the presence of 100 mM AP and IP. After incubation, the cells were lysed and their p53 content determined by ELISA. The results obtained show that the AP peptide induced the accumulation of the p53 protein, whereas no increase in p53 concentration was observed in the presence of the IP peptide (Figure 2(a)). The p53 accumulation induced by the AP peptide was observed by Western blot (data not shown). Following the addition of the AP peptide, the p53 concentration increased, reached a maximum, and then decreased. This decrease might be the consequence of a progressive degradation of the peptide. A decrease in the peptide concentration should lead to a decrease in the p53 concentration. To evaluate this possibility, the medium containing the peptide was removed every 24 hours and replaced by medium containing fresh peptide. The results show that the p53 concentration remained constant when fresh AP peptide was present throughout the experiment (Figure 2(b)). This suggests that the decrease in p53 concentration observed after 24 hours treatment was due to the degradation of the AP peptide.

Inhibitor of the p53-hdm2 Interaction

247 were incubated in the presence of 100 mM AP and IP peptides and the cellular accumulation of both p21Waf1/Cip1 and hdm2 proteins was determined by Western blot at different incubation times. The experimental data show that, after nine hours of treatment, p21Waf1/Cip1 and hdm2 started to accumulate and that maximum induction was achieved after treatment for 24 hours following a single dose of AP peptide (Figure 3(a)). By contrast, the IP peptide did not induce the accumulation of either the p21Waf1/Cip1 or the hdm2 protein. The AP peptide was next studied to determine whether its induction of p53 activity is dosedependent. The accumulation of the p21Waf1/Cip1 protein in HCT-116 cells treated with different concentrations of the AP peptide was analysed by Western blot after 24 hours treatment. The experimental results show that raising the AP peptide concentration increased the expression of the p21Waf1/Cip1 protein (Figure 3(b)). The effect of the AP peptide on the induction of p53 activity was therefore dose-dependent and a dose-dependent effect was observed for p53 accumulation (data not shown). The AP peptide is active only in cell lines that express wild-type p53

Figure 2. Induction of p53 accumulation by the AP peptide in HCT-116 cells. (a) The HCT-116 cells were incubated at different times in the presence of the AP ( & ) or the IP (&) peptide. The cells were lysed and the protein concentration determined. The amount of p53 present in 50 mg of total protein extract was estimated by ELISA. The standard deviation (n ˆ 4) for each experiment is given. Both peptides were used at a concentration of 100 mM. (b) AP peptide (100 mM) was added at the beginning of the experiment and not replaced until the end of the experiment (&) or replaced every 24 hours by a fresh peptide solution ( & ). At each indicated time-point, the cells were lysed and the p53 concentration dosed in ELISA.

The AP peptide induces the accumulation of p21Waf1/Cip1 and hdm2 in HCT-116 cells Since the AP peptide induces p53 accumulation, the induction of two gene products, p21Waf1/Cip1 and hdm2, which are transcriptionally regulated by p53 (Barak et al., 1993; el-Deiry et al., 1993) was monitored to ascertain whether the peptide stimulates p53 activity. The HCT-116 cells

The speci®city of the AP peptide was evaluated in different cell lines. Its effect in HCT-116 cells that express wt p53 was compared with its activity in SKBR-3 cells, which express a mutated form of p53 (O'Connor et al., 1997), and in Saos-2 cells, which do not express p53 (Diller et al., 1990). The experimental data (Figure 4(a)) indicate that the AP peptide induces p53 accumulation only in tumour cells that contain wt p53. This effect was not restricted to the HCT-116 cells, but was observed also in the breast carcinoma MCF-7 and NCI-H460 cell lines, which express wt p53 (data not shown). The induction of p53 activity was evaluated in Western blot by determining the accumulation of the p21Waf1/Cip1 protein. Following peptide treatment, the p21Waf1/Cip1 protein accumulated only in the cells that express the wt p53 protein (Figure 4(b)). This suggests that the AP peptide activates the p53 pathway only in the presence of wt p53. The AP peptide induces apoptosis in a cell line that overexpresses hdm2 The hdm2 gene product is overexpressed in many different cancers (Momand & Zambetti, 1997). This overexpression may lead to an inhibition of p53 activity and therefore to the loss of its tumour suppressor activity. Inhibitors of the p53hdm2 interaction would be very attractive for the treatment of such tumours. The AP peptide was evaluated in a cell line that overexpresses the hdm2 protein and contains a wt p53, the OSA-CL (Bottger et al., 1997; Florenes et al., 1994; Wasylyk et al., 1999). The AP peptide induced the accu-

248

Inhibitor of the p53-hdm2 Interaction

Figure 3. The AP peptide induces p21Waf1/Cip1 and hdm2 in HCT-116 cells. (a) The cells were treated with 100 mM AP or IP peptide for different incubation times. After incubation, the cells were lysed and the protein concentration determined. Total protein extract (100 mg) was analysed by Western blot for p21Waf1/Cip1 and hdm2 accumulation with the speci®c monoclonal antibodies EA10 and IF2, respectively. Molecular masses (kDa) are given. (b) The cells were treated with increasing concentrations of the AP peptide, and 24 hours after treatment the accumulation of p21Waf1/Cip1 was analysed by Western blot.

mulation both of p53 and of p21Waf1/Cip1 in OSACL cells (Figure 5(a) and (b)). Similar to the experiments described with the HCT-116 cells, the IP peptide did not stimulate the accumulation of either p53 or p21Waf1/Cip1 in these cells. Since tumours overexpressing hdm2 are good targets for inhibitors of the p53-hdm2 interaction, the AP peptide was next studied to determine whether it could induce the death of these cells. Because of its low solubility in aqueous buffers, the stock solutions of the AP peptide for these experiments were made in dimethyl sulphoxide (DMSO). This permits the use of higher concentrations of peptide. The OSA-CL cells were incubated with different concentrations of the AP and the IP peptides. After 72 hours incubation, the number of surviving cells was determined (see Materials and Methods). The results show that the AP peptide induces the death of the OSA-CL cells and that this effect is dosedependent (Figure 5(c)): 500 mM AP peptide (the maximal dose used in the assay) induces the death of about 90 % of the OSA-CL cells present in the assay. In contrast, incubation with increasing con-

centrations of the IP peptide does not lead to a dose-dependent cell death (Figure 5(c)). Because the AP peptide induces the death of the OSA-CL, it was next investigated if this death was due to the induction of apoptosis. ICE-like proteases (ICE, interleukin-1b-converting enzyme) have been implicated in the execution phase of apoptosis (Nicholson et al., 1995) and are therefore good markers for measuring the induction of apoptosis. One of these proteins, CPP32 (caspase-3), is activated in p53-induced apoptosis (Chandler et al., 1997). The CPP32 activity in cellular lysates of OSA-CL cells was therefore measured after treatment with the AP peptide. The occurrence of CPP32 activity was time-dependent and reached a maximum after 72 hours treatment (data not shown). The CCP32 activity induced after 72 hours by the AP peptide was not detected in the presence of the IP peptide or in the untreated cells (Figure 5(d)). The activity was completely inhibited in the presence of the CPP32 inhibitor Ac-DEVD-CHO, indicating that the activity measured in the assay is speci®c for CPP32 activation (data not shown).

249

Inhibitor of the p53-hdm2 Interaction

Figure 4. The AP peptide is a speci®c inducer of the p53 pathway. (a) The induction of p53 accumulation by the AP peptide was determined by ELISA in three different cell lines. The HCT-116, SKBR-3 and Saos-2 cells, respectively, express wt p53, a mutant p53 (Arg175His) and no p53. The amount of p53 present in these cells was determined after 24 hours treatment in the absence of the peptide (PBS) or in the presence of 100 mM AP peptide. The measured A450 of the non-treated cells (PBS) was 0.83, 2.8 and 0.23 for the HCT-116, SKBR-3 and Saos-2 cells, respectively. (b) In the same experiments, the induction of the p21Waf1/Cip1 protein was estimated by Western blot. The data represent the accumulation of p21Waf1/Cip1 in the cell lysates after 24 hours treatment.

The AP peptide does not induce the phosphorylation of serine 15 on p53 Recent ®ndings have shown that DNA damage induces the phosphorylation of p53 on serine 15 and serine ÿ37 (Shieh et al., 1997). These phosphorylations reduce the ability of p53 to form a complex with hdm2 and may therefore be responsible for the stabilisation of p53 after DNA damage. Since the AP peptide inhibits the p53hdm2 interaction, experiments were conducted to determine whether it induces p53 phosphorylation on serine 15. Cisplatin, which leads to p53 accumulation and activation in HCT-116 cells (Vikhanskaya et al., 1999), was chosen as control substance for DNA damage. The cells were treated with 5 mM cisplatin or 100 mM AP peptide, and the accumulation of p53 was analysed by ELISA (Figure 6(a)). After 24 hours treatment, a similar accumulation of the p53 protein was observed in the presence both of cisplatin and of the AP peptide. HCT-116 cellular extracts were analysed by Western blot with the Phospo-p53 (Ser15) antibody for the presence of p53 phosphorylated on serine 15 (Figure 6(b)). A massive phosphorylation on serine 15 was observed after 24 hours treatment with 5 mM cisplatin, whereas only very slight phosphorylation on serine 15 was observed in some experiments in the presence of the AP peptide. The fact that very little phosphorylation was observed on serine 15 even after 72 hours treatment with the AP peptide suggests that this difference was not due to slower kinetics of phosphorylation in the presence of the AP peptide (data not shown).

Discussion The activation of p53 in tumour cells induces apoptosis (Levine, 1997). The identi®cation of drugs that speci®cally stimulate this pathway of cell death in human tumours would therefore be of interest in cancer therapy. One attractive possibility is to activate the p53 pathway by inhibiting the p53-hdm2 interaction (Lane & Hall, 1997). Because p53 bound to hdm2 is exported into the cytoplasm and degraded (Haupt et al., 1997; Kubbutat et al., 1997; Midgley & Lane, 1997; Tao & Levine, 1999), the inhibition of this interaction should induce p53 accumulation and possibly cell death by apoptosis. Two different groups have reported the use of a peptidic inhibitor of the p53-hdm2 interaction in cellular experiments (Bottger et al., 1997; Wasylyk et al., 1999). However, this peptide was inserted into the thioredoxin protein or fused to the GST. Although these reports show that such hybrid molecules activate the p53 pathway, they do not demonstrate that small molecules are able to do the same. This is a key issue when protein-protein interactions are targeted, because apart from the fact that they have inhibitory properties owing to the presence of the fused peptide, these fusion proteins can create steric clashes that enhance the effect of the peptide. This could also be the case with the antibody 3G5 (Blaydes et al., 1997). Moreover, some of the experiments carried out in these studies were done by transfection, and it has been shown that the p53 pathway can be stimulated by the transfection procedure (Renzing & Lane, 1995). Evidence is still required to show that small mol-

250

Inhibitor of the p53-hdm2 Interaction

ecules are capable of inhibiting the p53-hdm2 interaction and stimulating the p53 pathway in tumour cells. This is a key step for the development of such new drugs. In this study, we used a highly potent inhibitor of the p53-hdm2 interaction, the AP peptide. This octamer peptide has an IC50 of 5 nM, 60 times lower than that of the dodecamer peptide previously inserted in the thioredoxin protein (Bottger et al., 1997) or fused to GST (Wasylyk et al., 1999). Because of the high potency of the AP peptide in vitro, we tried to use it without further modi®cations in cellular experiments. Surprisingly, the AP peptide was active in cells, and no additional tag was required for its penetration in the cells. To determine the exact intracellular concentration of the peptide, experiments with a radiolabelled analog of the AP (IP) peptide should be carried out. Since the AP peptide is a small molecule, it can be considered the ®rst inhibitor of the p53-hdm2 interaction to have been described as being active in cells. The effects obtained with the AP peptide fully substantiate our current knowledge of the p53hdm2 interaction. In tumour cells that express wt p53, the AP peptide stimulated p53 accumulation and p53 activity as measured by the induction of two genes that are regulated by p53, p21Waf1/Cip1 and hdm2. This effect was observed only in tumour cells that express wt p53. No induction or activation of p53 occurred in cells that contain either a mutated form of p53 or no p53 at all. In all the cell lines tested, the AP peptide induced a weak (less than twofold) accumulation of p53, but a strong accumulation of p21Waf1/Cip1. The induction of p21Waf1/Cip1 was observed also at lower peptide concentration (25 mM) while, in the same conditions, very little p53 accumulation was observed. This shows that only small amounts of p53 are required to induce the accumulation both of p21Waf1/Cip1 and of hdm2. Moreover, the peptide induced very weak, if any, phosphorylation on serine 15. This is similar to the activation of p53 by the upstream oncoprotein/p19ARF pathway, which does not induce phosphorylation on this serine residue (de Stanchina et al., 1998). The AP peptide would therefore appear to activate p53 by means of a mechanism different from that used by DNAdamaging agents that induce p53 phosphorylation

Figure 5. The AP peptide is active in a cell line that overexpresses hdm2. OSA-CL cells were not treated (PBS) or were treated with 100 mM AP or IP peptide for 24 hours. The cells were lysed and the amounts of (a)

p53 and (b) p21Waf1/Cip1 proteins present in the lysates determined by ELISA and Western blot, respectively. (c) OSA-CL cells were incubated with different concentrations of the AP (*) and the IP (*) peptides. The number of viable cells was then determined with the Cell Proliferation Reagent WST-1. Standard deviations are given. (d) AP or IP peptide (100 mM) was added to the cells. After 72 hours incubation, the cells were lysed and 100 mg of total protein extract was analysed for CPP32 activity using the CaspACE Assay System, Fluorometric (Promega) according to the manufacturers.

251

Inhibitor of the p53-hdm2 Interaction

induces cell death by apoptosis as measured by the induction of CPP32 activity. Our data support the view that low molecular mass inhibitors of the p53-hdm2 interaction can stimulate the p53 pathway. Blaydes and WynfordThomas, however, have shown that the disruption of this interaction in normal cells induces p53 (Blaydes & Wynford-Thomas, 1998). It is now very important to determine whether a therapeutic window exists for such compounds. The design of more stable analogues of the AP peptide with better penetration of cells should allow us to measure this therapeutic window in animal models.

Materials and Methods Cell lines The HCT-116, OSA-CL, Saos-2 and SKBR-3 cells were from the ATCC. The cells were treated with the different compounds between 80 % con¯uence and full con¯uence. Peptide synthesis The description of the peptide synthesis will be given elsewhere (C.G.-E. et al., unpublished results). The sequences of the active peptide (AP) and the inactive peptide (IP) are: Ac-Phe-Met-Aib-Pmp-…6-Cl†Trp-Glu-Ac3 c-Leu-NH2 …AP† …CGP 84700† Ac-Phe-Met-Aib-Pmp-Ala-Glu-Ac3 c-Leu-NH2 …IP† …NVP-AAW422-NX-1†

Figure 6. The AP peptide does not induce the phosphorylation of p53 serine 15. (a) HCT-116 cells were incubated for different times in the presence of 5 mM cisplatin ( & ) or 100 mM AP peptide (&). The amount of p53 expressed in these cells was then determined by ELISA. (b) The same protein extracts were analysed by Western blot with the Phospo-p53 (Ser15) antibody. The lane marked PBS corresponds to HCT-116 cells that have not been treated.

on serine 15. These results suggest that p53 does not need to be modi®ed to become active. Genes regulated by p53 could then be activated just by a gene dosage effect. This hypothesis needs to be con®rmed by a more extended analysis of p53 modi®cations after treatment with the AP peptide. The peptide is active in tumour cells that overexpress hdm2. These cells are very good targets for such compounds, because their high hdm2 content might inhibit p53. In these cells, the AP peptide

where Aib is a-aminoisobutyric acid, Ac3c is 1-aminocyclopropanecarboxylic, Pmp is phosphonophenylalanine, and (6-Cl)Trp is 6-chlorotryptophan. To make the AP and IP peptide stock solutions, peptides were dissolved at room temperature in PBS. The solutions were then centrifuged for ten minutes at 14,000 rpm and the soluble fractions were dosed at 280 nm by spectrophotometry. In these conditions both peptides had a similar solubility. Protein purification and peptide testing by ELISA The recombinant p53 and GST-hdm2(1-188) proteins were puri®ed as described (BoÈttger et al., 1996). The AP and IP peptides were tested for their ability to inhibit the p53-hdm2 interaction in the ELISA hdm2 (BoÈttger et al., 1996). Measurement of p53 accumulation by ELISA The cells, in six-well dishes, were incubated for the indicated times in the presence or absence of 100 mM AP and IP peptides. After incubation, the cells were lysed with the lysis buffer (50 mM Tris-HCl (pH 7.5), 5 mM EGTA, 1 % (v/v) Triton X-100, 150 mM NaCl, 1 mM phenylmethylsulphonyl ¯uoride, 80 mg/ml aprotinin and 50 mg/ml leupeptin) and the protein concentration was determined with the BCA protein assay reagent (Pierce). Total protein extracts (50 mg) were used in the

252 Rapid Format Pan p53 Assay-Quantitative ELISA kit (Calbiochem) to determine their content in p53. For each measurement, 2 ng of puri®ed p53 was used to ensure that the signal observed did not reach saturation. Measurement of p21Waf/Cip1, hdm2 and p53 phosphorylated on serine 15 induction in Western blot The cells, in six-well dishes, were incubated for the indicated times in the presence or absence of 100 mM AP and IP peptides. After incubation, the cells were lysed and the protein concentration determined as described above. Total protein extracts (100 mg) were heat denatured in the presence of SDS-buffer and loaded onto a polyacrylamide gel. After migration, the proteins were transferred onto an Immobilon-P membrane (Millipore). The p21Waf/Cip1, hdm2 and p53 phosphorylated on serine 15 were detected with the EA10 antibody (Oncogene Science) diluted at 1/150, the IF2 antibody (Oncogene Science) diluted at 1/50, and the Phospo-p53 (Ser15) antibody (New England Biolabs) diluted at 1/1000. The different primary antibodies were detected with secondary antibodies conjugated to horseradish peroxidase (Biorad) and the ECL Western blotting detection reagent (Amersham Life Science). Cell proliferation assay OSA-CL cells (15,000 cells/well) were seeded in a 96well plate and incubated for 24 hours to allow them to attach to the plate. The medium was removed and replaced by fresh medium containing the AP or the IP peptides dissolved in DMSO. The cells were then incubated for 72 hours and the number of viable cells was determined with the Cell Proliferation Reagent WST-1 (Roche Diagnostics) according to the manufacturers. The presence of DMSO up to a concentration of 0.5 % (concentration used in the assay) does not modify cell survival noticeably (data not shown). Measurement of the CPP32 activity The cells were incubated for 72 hours in the presence of 100 mM AP or IP. After incubation, the cells were lysed with the lysis buffer, and the protein concentration was determined. The CPP32 activity from 100 mg of total protein extracts was determined using the CaspACE Assay System, Fluorometric (Promega) according to the manufacturer's instructions.

References Barak, Y., Juven, T., Haffner, R. & Oren, M. (1993). mdm2 expression is induced by wild-type p53 activity. EMBO J. 12, 461-468. Blaydes, J. P. & Wynford-Thomas, D. (1998). The proliferation of normal human ®broblasts is dependent upon negative regulation of p53 function by mdm2. Oncogene, 16, 3317-3322. Blaydes, J. P., Gire, V., Rowson, J. M. & WynfordThomas, D. (1997). Tolerance of high levels of wildtype p53 in transformed epithelial cells dependent on auto-regulation by mdm-2. Oncogene, 14, 18591868. Bottger, V., Bottger, A., Howard, S. F., Picksley, S. M., Chene, P., Garcia-Echeverria, C., Hochkeppel, H. K. & Lane, D. P. (1996). Identi®cation of novel mdm2

Inhibitor of the p53-hdm2 Interaction binding peptides by phage display. Oncogene, 13, 2141-2147. Bottger, A., Bottger, V., Sparks, A., Liu, W. L., Howard, S. F. & Lane, D. P. (1997). Design of a synthetic Mdm2-binding mini protein that activates the p53 response in vivo. Curr. Biol. 7, 860-869. Chandler, J. M., Alnemri, E. S., Cohen, G. M. & MacFarlane, M. (1997). Activation of CPP32 and Mch3 alpha in wild-type p53-induced apoptosis. Biochem. J. 322, 19-23. Chen, J., Marechal, V. & Levine, A. J. (1993). Mapping of the p53 and mdm-2 interaction domains. Mol. Cell. Biol. 13, 4107-4114. Chen, L., Agrawal, S., Zhou, W., Zhang, R. & Chen, J. (1998). Synergistic activation of p53 by inhibition of MDM2 expression and DNA damage. Proc. Natl Acad. Sci. USA, 95, 195-200. de Stanchina, E., McCurrach, M. E., Zindy, F., Shieh, S. Y., Ferbeyre, G., Samuelson, A. V., Prives, C., Roussel, M. F., Sherr, C. J. & Lowe, S. W. (1998). E1A signaling to p53 involves the p19(ARF) tumor suppressor. Genes Dev. 12, 2434-2442. Diller, L., Kassel, J., Nelson, C. E., Gryka, M. A., Litwak, G., Gebhardt, M., Bressac, B., Ozturk, M., Baker, S. J., Vogelstein, B., et al. (1990). p53 functions as a cell cycle control protein in osteosarcomas. Mol. Cell. Biol. 10, 5772-5781. el-Deiry, W. S., Tokino, T., Velculescu, V. E., Levy, D. B., Parsons, R., Trent, J. M., Lin, D., Mercer, W. E., Kinzler, K. W. & Vogelstein, B. (1993). WAF1, a potential mediator of p53 tumor suppression. Cell, 75, 817-825. Florenes, V. A., Maelandsmo, G. M., Forus, A., Andreassen, A., Myklebost, O. & Fodstad, O. (1994). MDM2 gene ampli®cation and transcript levels in human sarcomas: relationship to TP53 gene status. J. Natl Cancer Inst. 86, 1297-1302. Freedman, D. A. & Levine, A. J. (1998). Nuclear export is required for degradation of endogenous p53 by MDM2 and human papillomavirus E6. Mol. Cell. Biol. 18, 7288-7293. Haupt, Y., Maya, R., Kazaz, A. & Oren, M. (1997). Mdm2 promotes the rapid degradation of p53. Nature, 387, 296-299. Honda, R. & Yasuda, H. (1999). Association of p19(ARF) with Mdm2 inhibits ubiquitin ligase activity of Mdm2 for tumor suppressor p53. EMBO J. 18, 2227. Kubbutat, M. H., Jones, S. N. & Vousden, K. H. (1997). Regulation of p53 stability by Mdm2. Nature, 387, 299-303. Kussie, P. H., Gorina, S., Marechal, V., Elenbaas, B., Moreau, J., Levine, A. J. & Pavletich, N. P. (1996). Structure of the MDM2 oncoprotein bound to the p53 tumor suppressor transactivation domain. Science, 274, 948-953. Lain, S., Midgley, C., Sparks, A., Lane, E. B. & Lane, D. P. (1999). An inhibitor of nuclear export activates the p53 response and induces the localization of HDM2 and p53 to U1A-positive nuclear bodies associated with the PODs. Exp. Cell. Res. 248, 457472. Lane, D. P. & Hall, P. A. (1997). MDM2-arbiter of p53's destruction. Trends Biochem. Sci. 22, 372-374. Levine, A. J. (1997). p53, the cellular gatekeeper for growth and division. Cell, 88, 323-331. Midgley, C. A. & Lane, D. P. (1997). p53 protein stability in tumour cells is not determined by mutation but

Inhibitor of the p53-hdm2 Interaction is dependent on Mdm2 binding. Oncogene, 15, 11791189. Momand, J. & Zambetti, G. P. (1997). Mdm-2: ``big brother'' of p53. J. Cell. Biochem. 64, 343-352. Momand, J., Zambetti, G. P., Olson, D. C., George, D. & Levine, A. J. (1992). The mdm-2 oncogene product forms a complex with the p53 protein and inhibits p53-mediated transactivation. Cell, 1237, 343-352. Nicholson, D. W., Ali, A., Thornberry, N. A., Vaillancourt, J. P., Ding, C. K., Gallant, M., Gareau, Y., Grif®n, P. R., Labelle, M., Lazebnik, Y. A., et al. (1995). Identi®cation and inhibition of the ICE/ CED-3 protease necessary for mammalian apoptosis. Nature, 376, 37-43. O'Connor, P. M., Jackman, J., Bae, I., Myers, T. G., Fan, S., Mutoh, M., Scudiero, D. A., Monks, A., Sausville, E. A., Weinstein, J. N., Friend, S., Fornace, A. J., Jr & Kohn, K. W. (1997). Characterization of the p53 tumor suppressor pathway in cell lines of the National Cancer Institute anticancer drug screen and correlations with the growth-inhibitory potency of 123 anticancer agents. Cancer Res. 57, 4285-4300. Picksley, S. M. & Lane, D. P. (1993). The p53-mdm2 autoregulatory feedback loop: a paradigm for the regulation of growth control by p53? Bioessays, 15, 689-690. Picksley, S. M., Vojtesek, B., Sparks, A. & Lane, D. P. (1994). Immunochemical analysis of the interaction of p53 with MDM2; ®ne mapping of the MDM2

253 binding site on p53 using synthetic peptides. Oncogene, 9, 2523-2529. Renzing, J. & Lane, D. P. (1995). p53-dependent growth arrest following calcium phosphate-mediated transfection of murine ®broblasts. Oncogene, 10, 18651868. Shieh, S. Y., Ikeda, M., Taya, Y. & Prives, C. (1997). DNA damage-induced phosphorylation of p53 alleviates inhibition by MDM2. Cell, 91, 325-334. Tao, W. & Levine, A. J. (1999). Nucleocytoplasmic shuttling of oncoprotein Hdm2 is required for Hdm2mediated degradation of p53. Proc. Natl Acad. Sci. USA, 96, 3077-3080. Vikhanskaya, F., Colella, G., Valenti, M., Parodi, S., D'Incalci, M. & Broggini, M. (1999). Cooperation between p53 and hMLH1 in a human colocarcinoma cell line in response to DNA damage. Clin. Cancer Res. 5, 937-941. Wasylyk, C., Salvi, R., Argentini, M., Dureuil, C., Delumeau, I., Abecassis, J., Debussche, L. & Wasylyk, B. (1999). p53 mediated death of cells overexpressing MDM2 by an inhibitor of MDM2 interaction with p53. Oncogene, 18, 1921-1934. Wolff, B., Sanglier, J. J. & Wang, Y. (1997). Leptomycin B is an inhibitor of nuclear export: inhibition of nucleo- cytoplasmic translocation of the human immunode®ciency virus type 1 (HIV-1) Rev protein and Rev-dependent mRNA. Chem. Biol. 4, 139-147.

Edited by A. R. Fersht (Received 17 January 2000; received in revised form 23 March 2000; accepted 31 March 2000)