T.Unger, T.Juven-Gershon and E.Moallem contributed equally to this work
Critical role for Ser20 of human p53 in the negative regulation of p53 by Mdm2
Article first published online: 1 APR 1999
Copyright © 1999 European Molecular Biology Organization
The EMBO Journal
Volume 18, Issue 7, pages 1805–1814, April 1, 1999
How to Cite
Unger, T., Juven-Gershon, T., Moallem, E., Berger, M., Vogt Sionov, R., Lozano, G., Oren, M. and Haupt, Y. (1999), Critical role for Ser20 of human p53 in the negative regulation of p53 by Mdm2. The EMBO Journal, 18: 1805–1814. doi: 10.1093/emboj/18.7.1805
- Issue published online: 1 APR 1999
- Article first published online: 1 APR 1999
- Manuscript Accepted: 2 FEB 1999
- Manuscript Revised: 16 DEC 1998
- Manuscript Received: 10 AUG 1998
In response to environmental stress, the p53 phosphoprotein is stabilized and activated to inhibit cell growth. p53 stability and activity are negatively regulated by the murine double minute (Mdm2) oncoprotein in an autoregulatory feedback loop. The inhibitory effect of Mdm2 on p53 has to be tightly regulated for proper p53 activity. Phosphorylation is an important level of p53 regulation. In response to DNA damage, p53 is phosphorylated at several N-terminal serines. In this study we examined the role of Ser20, a potential phosphorylation site in human p53, in the regulation of p53 stability and function. Substitution of Ser20 by Ala (p53-Ala20) significantly increases the susceptibility of human p53 to negative regulation by Mdm2 in vivo, as measured by apoptosis and transcription activation assays. Mutation of Ser20 to Ala renders p53 less stable and more prone to Mdm2-mediated degradation. While the in vitro binding of p53 to Mdm2 is not increased by the Ala20 mutation, the same mutation results in a markedly enhanced binding in vivo. This is consistent with the conclusion that phosphorylation of Ser20 in vivo attenuates the binding of wild-type p53 to Mdm2. Peptides bearing non-phosphorylated Ser20 or Ala20 compete with p53 for Mdm2 binding, while a similar peptide with phosphorylated Ser20 does not. This implies a critical role for Ser20 in modulating the negative regulation of p53 by Mdm2, probably through phosphorylation-dependent inhibition of p53–Mdm2 interaction.
The p53 phosphoprotein plays a key role in tumor suppression and in the cellular response to stress. The p53 protein is normally expressed at low levels in a latent inactive form. However, upon exposure to stress conditions such as hypoxia, DNA damage or reduction in ribonucleotide triphosphate levels, p53 is modified post-translationally with subsequent accumulation of active protein (Gottlieb and Oren, 1996; Levine, 1997; Agarwal et al., 1998; Giaccia and Kastan, 1998; Prives, 1998). Activation of p53 by stress leads to growth inhibition. This includes the induction of growth arrest, presumably to allow sufficient time for DNA repair, or apoptosis which eliminates abnormal cells. These activities are abrogated by mutations in p53, which constitute the most common type of genetic alterations in many different human malignancies (Velculescu and El-Deiry, 1996). Most of the mutations are localized to the central DNA-binding domain of p53, and these mutations prevent p53 from binding to its target DNA sequences (e.g. Unger et al., 1993).
The biological effects of p53 are largely due to sequence-specific binding to DNA and transcriptional activation of target genes. The transactivation domain of p53, located at the N-terminus of the protein (Unger et al., 1992), interacts with components of the basal transcription machinery (Lu and Levine, 1995) and promotes the transcription of genes containing p53-binding sequences. Different sets of p53 responsive genes mediate different functions of p53, including growth arrest by p21waf1/cip1 and 14-3-3σ (El-Deiry et al., 1993; Hermeking et al., 1997) and apoptosis by bax, CD95/Fas, IGF-BP3 and PIG3 (Buckbinder et al., 1995; Miyashita and Reed, 1995; Owen-Schaub et al., 1995; Polyak et al., 1997).
The protein stability and biochemical activities of p53 are tightly regulated by a variety of mechanisms. One of the mechanisms most likely to mediate p53 regulation is phosphorylation. The p53 protein is phosphorylated on several serine residues within the N- and C-terminal regions by several cellular kinases (reviewed in Steegenga et al., 1996; Milczarek et al., 1997; Fuchs et al., 1998a; Giaccia and Kastan, 1998; Meek, 1998; Prives, 1998). There are at least three phosphorylation sites near the C-terminus of human p53 (hp53) at amino acids 315, 378 and 392. Ser315 has been shown to be a target of p34cdc2 kinase and of cyclin-dependent kinase 2 (cdk2) (Addison et al., 1990; Bischoff et al., 1990; Price et al., 1995). Ser392 is phosphorylated in vitro by purified casein kinase II (Meek et al., 1990), and Ser378 is a site for phosphorylation by protein kinase C (PKC) (Baudier et al., 1992). Phosphorylation of these C-terminal serines can enhance the in vitro specific DNA-binding activity of p53 (Hupp et al., 1992; Takenaka et al., 1995; Wang and Prives, 1995; Lu et al., 1997). Similarly, dephosphorylation of human p53 Ser376 enhances binding to 14-3-3 protein and increases specific DNA binding (Waterman et al., 1998). The substitution of mouse p53 Ser389 (equivalent to Ser392 in human p53) to glutamate activates p53 in vivo (Hao et al., 1996), but substitution to Ala did not affect the ability of p53 to bind DNA or to suppress cell growth (Fiscella et al., 1994). This discrepancy can be explained by the recent observation that phosphorylation of Ser392 occurs in response to some environmental stresses but not others, suggesting independent mechanisms of activating p53 (Kapoor and Lozano, 1998; Lu et al., 1998).
At the N-terminus of p53, several potential phosphorylation sites have been identified including Ser6, 9, 15, 20, 33, 37 and Thr18 of human p53 (hp53) (reviewed in Steegenga et al., 1996; Milczarek et al., 1997; Fuchs et al., 1998a; Giaccia and Kastan, 1998; Meek, 1998; Prives, 1998). Ser15 and 37 of hp53 can be phosphorylated in vitro by the DNA-dependent protein kinase (DNA-PK) (Lees Miller et al., 1992). Ser7, 9 and 12 of mouse p53 can be phosphorylated in vitro by casein kinase I (CKI) and CKI-like PK270 (Milne et al., 1992). In vitro phosphorylation of Ser34 of mouse p53 by the UV-induced Jun kinase 1 (JNK) has been shown (Milne et al., 1995). Further, the Raf-1 serine/threonine kinase phosphorylates p53 in vitro on one or more sites within the first 27 residues (Jamal and Ziff, 1995). Although multiple sites of phosphorylation have been identified in p53, relatively little is known about their role in the regulation of p53 activity in vivo. The phosphorylation of Ser15, which partially alleviates the inhibitory effect of Mdm2 on p53 (Shieh et al., 1997), was shown to be induced by DNA damage (Siliciano et al., 1997). Recent evidence strongly implicates the ATM (ataxia telengiectasia mutated) kinase in Ser15 phosphorylation (Banin et al., 1998; Canman et al., 1998). Phosphorylation of Ser33 and Ser37 also enhances C-terminal acetylation of p53 with a consequent increase in specific DNA binding (Sakaguchi et al., 1998).
The mdm2 oncogene was found to be amplified in sarcomas and several other types of cancers (reviewed by Momand and Zambetti, 1997; Lozano and Montes de Oca Luna, 1998). Mdm2 binds p53 within its transactivation domain and blocks its transcriptional activity (Oliner et al., 1993; Chen et al., 1995). Moreover, Mdm2 abrogates the ability of p53 to induce growth arrest and apoptosis (Chen et al., 1996; Haupt et al., 1996). Since the mdm2 gene itself is a target gene for activation by p53, a negative autoregulatory loop exists between the two proteins (reviewed in Piette et al., 1997). A lag is sometimes found between activation of p53 and the consequent induction of mdm2, defining a time window within which p53 is allowed to exert its inhibitory activities. Alternatively, p53 may continue to operate even after the induction of mdm2 has occurred. In this case, however, p53 must be rendered immune to the inhibitory effects of Mdm2, possibly through covalent modification of either or both proteins, and perhaps also through the action of additional partner proteins. In addition, Mdm2 has more recently been shown to promote p53 for degradation in a ubiquitin-dependent manner (Haupt et al., 1997; Kubbutat et al., 1997), by enhancing the ubiquitination of p53 (Honda et al., 1997). The interaction between the two proteins is obligatory for the inhibitory activities of Mdm2 (Chen et al., 1996; Haupt et al., 1996, 1997; Kubbutat et al., 1997), and facilitates nuclear–cytoplasmic shuttling of p53 (Roth et al., 1998). Blocking the interaction between p53 and Mdm2 by specific mutations, or by peptides and antibodies directed towards the binding site, all prevent the degradation and inactivation of p53 by Mdm2 (Böttger et al., 1997; Haupt et al., 1997; Kubbutat et al., 1997; Midgley and Lane, 1997). Thus, through parallel mechanisms, Mdm2 shuts off p53 activity and terminates the propagation of its growth inhibitory signal. This tight regulation of p53 by Mdm2 is essential for normal embryonic development (Jones et al., 1995; Montes de Oca Luna et al., 1995).
In this study we investigated the role of serine 20 in the regulation of p53 activity and stability. We have recently shown that substitution of serine to alanine at residue 20 (p53-Ala20) partially impaired the apoptotic activity of p53 (Unger et al., 1999). We now report that this reduced apoptosis is due to enhanced susceptibility of p53-Ala20 to negative regulation by Mdm2. Similarly, the transcriptional activity of p53-Ala20 is more sensitive than that of wild-type (wt) p53 to inhibition by Mdm2. Notably, substitution of Ser20 to alanine renders p53 more susceptible to Mdm2-directed proteolytic degradation. All of these observations indicate that mutation of Ser20 to alanine enhances the interaction of p53 with Mdm2. In support of this conclusion, analysis of p53–Mdm2 complexes within transfected cells revealed a much more prominent coprecipitation of Mdm2 with p53-Ala20 than with wt p53. However, this increased interaction is not due to the Ala mutation itself, as p53-Ala20 did not exhibit any enhanced association with Mdm2 when the binding was performed in vitro. Hence, the differential binding in vivo must be due to a modification which occurs specifically on residue 20 of p53 within living cells. That this modification is the phosphorylation of Ser20 is strongly supported by the finding that p53-derived peptides with a non-phosphorylated Ser20 or with Ala20 were able to compete efficiently with wt p53 for Mdm2 binding, while a similar peptide with phosphorylated Ser20 failed to compete. Our results provide the first in vivo demonstration for an important role played by Ser20 in the modulation of p53 activity and stability, and imply that phosphorylation of p53 on this residue interferes with its ability to associate with Mdm2 and be negatively regulated by it. This, in turn, might allow p53 to retain its activity under conditions of extended stress, even after the mdm2 gene has become maximally induced.
Mutation of Ser20 to Ala renders p53-mediated apoptosis excessively sensitive to inhibition by Mdm2
If Ser20 of hp53 is a site of phosphorylation in vivo, mutation to Ala should prevent this phosphorylation. Thus, whereas wt p53 may be phosphorylated to a variable extent on Ser20, the p53-Ala20 mutant remains completely non-phosphorylated at this position at all times. To assess the possible consequences of p53 phosphorylation on Ser20 and of its abrogation, we compared the behavior of wt p53 and p53-Ala20 in a variety of biological and biochemical assays. These analyses revealed that p53-Ala20 possesses a partially impaired apoptotic activity, as measured in transient transfection assays (Unger et al., 1999). Further support for this conclusion is provided in Figure 1, where H1299 cells were transfected with expression plasmids encoding either wt p53 or p53-Ala20. Seventy-two hours post-transfection cells were harvested, stained for p53 and subjected to flow-cytometric analysis (Haupt et al., 1996). The non-transfected sub-population (NT in Figure 1A) exhibits low fluorescent intensity, while the transfected sub-population (T) exhibits high levels of p53 fluorescence (Figure 1A). The cell-cycle distribution of each sub-population was analyzed separately. Cells with DNA content of less than 2N (Sub-G1 in Figure 1B) are considered apoptotic (Darzynkiewicz et al., 1992). As seen in Figure 1C and D, while wt p53 induced 21% apoptosis (after subtraction of the background obtained in the non-transfected cells), p53-Ala20 induced only 4%. A comparison of cumulative data, representing a total of seven independent experiments (Figure 1E), clearly supports the conclusion that p53-Ala20 has impaired apoptotic activity.
The apoptotic activity of p53 is efficiently inhibited by Mdm2 in H1299 cells (Haupt et al., 1996). Ser20 resides within the Mdm2-binding site of p53 (residues 19–26; Picksley et al., 1994; Kussie et al., 1996), which engages in tight interactions with a hydrophobic cleft within Mdm2. It is conceivable that phosphorylation of Ser20 may interfere with this binding. In that case, substitution of Ser20 by Ala, by preventing phosphorylation, is expected to augment the inhibitory effect of Mdm2. This might render the mutant protein less competent for induction of apoptosis in cells such as H1299, in which endogenous human Mdm2 (Hdm2) is induced in response to p53 activation (Haupt et al., 1996). This notion was tested by comparing the extent by which limiting amounts of transfected Mdm2 inhibit the apoptotic activity of p53-Ala20 relative to wt p53. H1299 cells were transfected with 3 μg of wt p53 or p53-Ala20 DNA either alone or together with 1 μg of mdm2 DNA, and the extent of apoptosis within the transfected sub-population was determined as above. Notably, expression levels of p53 were slightly lower in cells cotransfected with p53-Ala20 than in those with wt p53 (Figure 2A; see below). As expected (Haupt et al., 1996), co-transfection with mdm2 blocked the apoptotic activity of wt p53 by 45% (Figure 2B). Remarkably, the apoptotic activity of p53-Ala20 was inhibited to a significantly higher extent (70%). A similar difference was observed when different ratios of p53 to mdm2 were used (data not shown). Presumably, the exogenous Mdm2 acts in concert with the endogenously induced Hdm2 to block p53-mediated apoptosis. Our data indicate that this block is significantly more pronounced in the case of p53-Ala20, implying that mutation of Ser20 to Ala indeed causes the apoptotic activity of p53 to become more sensitive to inhibition by Mdm2.
Transcriptional activity of p53-Ala20 is excessively sensitive to inhibition by Mdm2
The inhibition of p53-mediated apoptosis by Mdm2 in H1299 cells was shown to be largely due to inhibition of the transcriptional activity of p53 (Haupt et al., 1996). Therefore, the extent by which Mdm2 blocks the transcriptional activity of p53 mutated at residue 20 was evaluated using the luciferase assay. The effect of Mdm2 on wt p53 was compared with that seen with p53-Ala20 or with p53 carrying a substitution to Asp at position 20 (p53-Asp20). Asp is negatively charged and in some instances may mimic a phosphorylated state. H1299 cells were transiently transfected with 5 ng of either wt p53, p53-Ala20 or p53-Asp20 expression plasmid DNA, together with increasing amounts of mdm2 expression plasmid. The transcriptional activity of p53 was deduced from the activity of the luciferase reporter gene driven by the p53-responsive cyclin G promoter (Zauberman et al., 1995). Low concentrations of mdm2 expression plasmid were chosen such that the amounts of encoded Mdm2 protein will be limiting, resulting in only partial inhibition of p53 activity. In the absence of exogenous Mdm2, the transcriptional activity of p53-Ala20 was comparable with that of wt p53 (Figure 3, columns 2, 5 and 8). However, marked differences in the transcriptional activities of the three forms of p53 were revealed in the presence of exogenous Mdm2. Co-transfection of either 5 or 10 ng of mdm2 expression plasmid inhibited p53-Ala20 activity by as much as 74% (Figure 3, column 6) and 89% (column 7), respectively. On the other hand, the activities of wt p53 and p53-Asp20 were inhibited more mildly by either 5 ng (Figure 3, wt p53 by 31% and p53-Asp20 by 18%, columns 3 and 9) or 10 ng (Figure 3, wt p53 by 55% and p53-Asp20 by 53%, columns 4 and 10) of mdm2 DNA. Hence, as in the case of apoptosis, the transcriptional activity of p53-Ala20 is also selectively sensitive to the inhibitory effect of Mdm2.
p53-Ala20 is more susceptible than wt p53 to Mdm2-mediated degradation
Mdm2 promotes the degradation of p53 through the ubiquitin–proteasome system (Haupt et al., 1997; Honda et al., 1997; Kubbutat et al., 1997). The observations described above demonstrated that substitution of Ser20 by Ala renders the protein more sensitive to the effects of Mdm2. It was therefore reasonable to expect that p53-Ala20 may also be more susceptible to the destabilizing activity of Mdm2. Evidence supporting this prediction was obtained in the FACS analysis (Figure 2A) which suggested that the average cellular content of p53-Ala20 is lower than that of wt p53 in the presence of co-transfected mdm2, and that a distinct sub-population with higher p53 protein content existed in the wt p53-transfected cells under those conditions.
To establish more directly that p53-Ala20 is indeed more susceptible than wt p53 to Mdm2-promoted degradation, the steady-state level of the two proteins was compared by Western blot analysis following transfection with relatively low amounts of expression plasmid DNA. H1299 cells were transfected with wt p53 or p53-Ala20, either alone or together with co-transfected mdm2. Using 50 ng of each p53 expression plasmid without any exogenous mdm2, the level of p53-Ala20 was only about half that of wt p53 (Figure 4A, lane 1). This difference became much more dramatic when low amounts of exogenous mdm2 were also included. Inclusion of as little as 50 ng of mdm2 plasmid was already sufficient to achieve maximum reduction of p53-Ala20 steady-state levels, while the level of wt p53 was reduced only 2-fold (Figure 4A, lane 2, and 4B). Inclusion of 100 ng mdm2 expression plasmid reduced wt p53 levels by a further 2-fold (Figure 4A, lane 3, and 4B). Similar effects were observed when 10 or 100 ng of the p53 expression plasmids were used (data not shown). These results demonstrate that substitution of Ser20 by Ala destabilizes p53 by rendering it much more sensitive to Mdm2-mediated degradation. Presumably, the relatively lower levels of p53-Ala20 obtained even without exogenous Mdm2 (Figure 4A, lane 1) may also be due to this increased sensitivity, in this case mediated exclusively by the p53-induced endogenous Hdm2 of H1299 cells.
If the above notion is correct, it is expected that in the absence of endogenous Hdm2 the expression levels of wt p53 and p53-Ala20 should be comparable. This prediction was tested in embryonic fibroblasts derived from p53−/−/mdm2−/− double knock-out mice (McMasters et al., 1996). When these cells were transfected with low amounts of the two p53 expression plasmids, p53-Ala20 indeed did not accumulate to lower levels than wt p53 (Figure 5). In fact, the level of p53-Ala20 (Figure 5, lane 2) was even higher than that of wt p53 (lane 1); the reason for this is presently unclear. Thus, the steady-state levels of the Ala20 mutant are compromised only when the cells are capable of expressing endogenous Mdm2, implicating this endogenous protein in the selective down-regulation of p53-Ala20. When the p53 expression plasmids were transfected into the double knock-out fibroblasts together with exogenous Mdm2, the picture was essentially similar to that observed in H1299 cells: p53-Ala20 expression was reduced far more severely than that of wt p53 (data not shown). Collectively, these results demonstrate that Ser20 plays an important role in regulating p53 stability, through dictating the susceptibility of the protein to Mdm2-mediated degradation.
Mdm2 binds preferentially to p53-Ala20 in vivo but not in vitro
The increased responsiveness of p53-Ala20 to negative regulation by Mdm2 raised the possibility that the Ser20 to Ala substitution altered the binding affinity of p53 to Mdm2 within cells. This conjecture is attractive because Ser20 resides within the Mdm2-binding domain of p53 (Picksley et al., 1994; Kussie, et al., 1996). A co-immunoprecipitation assay was therefore performed in order to compare the two types of p53 with regard to Mdm2 binding. To increase the sensitivity of the assay a mutant Hdm2, Hdm2Δ RING, lacking the C-terminal RING finger domain was used. While this mutant binds p53, it can not promote its degradation (Figure 6AII; R.Vogt Sionov and Y.Haupt, unpublished results; see also Kubbutat et al., 1997). This allows the accumulation of substantial amounts of p53, thereby facilitating detection of stable p53–Mdm2 complexes. Twenty-four hours post-transfection, cell extracts were prepared and Hdm2/p53 complexes were monitored by immunoprecipitation with the anti-p53 monoclonal antibody PAb421, followed by Western blot analysis with the anti-Hdm2 monoclonal antibody SMP14. This revealed that the amount of Hdm2Δ RING co-immunoprecipitated with p53-Ala20 was higher (14-fold) than that precipitated with wt p53 (Figure 6AI, compare lanes 1 and 2), despite lower expression level of p53-Ala20 as compared with wt p53 (Figure 6AII, lanes 1 and 2). A similar experiment was also carried out with intact wt Hdm2; this time the effect of Hdm2 on p53 degradation was partially blocked by treating the transfected cells for 2 h with the proteasome inhibitor, ALLN. Again, a higher amount of Hdm2 was co-immunoprecipitated with p53-Ala20 than with wt p53 (Figure 6AI, lanes 3 and 4), although the former was expressed at a substantially lower level (Figure 6AII, lanes 3 and 4). Thus, substitution of Ser20 by Ala increases markedly the association of p53 with Hdm2 within cells.
The lower association of wt p53 with Hdm2 relative to that of p53-Ala20 could be due to the fact that the in vivo phosphorylation of the former but not the latter on Ser20 interferes with binding to Hdm2. Alternatively, it could be argued that the Ala substitution itself augments the affinity of p53 for Hdm2, as compared with the wild-type protein with serine at position 20. To distinguish between these two possibilities, the interaction of different p53 forms with Mdm2 was monitored in vitro. If reduced Mdm2 binding requires phosphorylation on Ser20, it is likely that such modification will not occur in vitro and therefore both p53-Ala20 and wt p53 will bind equally well to Mdm2. p53-Ala20, wt p53 and p53-Asp20 were each translated in a reticulocyte lysate in the presence of [35S]methionine. Equal amounts of radiolabeled p53 of each type were reacted with either a glutathione S-transferase (GST)–Mdm2 fusion protein or GST alone. As seen in Figure 6B, all three forms of p53 bound preferentially to GST–Mdm2. However, unlike in the in vivo situation, the amount of bound p53-Ala20 was not higher but actually even lower than that of wt p53, while the binding capacity of p53-Asp20 was similar to that of wt p53. Taken together, these results suggest that a specific modification of Ser20 in vivo, presumably phosphorylation, compromises the binding capacity of p53 to Mdm2.
To explore this suggestion more directly, we employed a peptide competition assay. Three peptides encompassing the Mdm2-binding domain of human p53 (residues 14–29) were synthesized: one with the wild-type sequence (Ser20), one with Ala at residue 20 (Ala20), and one with a phosphorylated Ser20 [Ser(P)20]. The ability of these three peptides to block binding between human p53 and mouse Mdm2 was tested by ELISA at different peptide concentrations. Figure 6C shows that both Ser20 and Ala20 peptides, at 100 μM as well as 250 μM, competed for the binding of wt hp53 to Mdm2 by 43 and 63%, respectively. In contrast, at the same concentrations Ser(P)20 had no significant effect on binding. Similarily, at a peptide concentration of 500 μM, p53 binding to Mdm2 was competed markedly by both Ser20 (72%) and Ala20 (79%), while competition by Ser(P)20 was only 24% (Figure 6C). This argues that phosphorylation of p53 on Ser20 significantly reduces the affinity for Mdm2. Overall, our data support the conjecture that the in vivo phosphorylation of p53 on Ser20 by a cellular kinase(s) can strongly reduce the tightness of its interaction with Mdm2.
p53 is a key player in the cellular response to environmental stress, including various types of genomic damage. Activation of p53 in response to stress involves post-translational modifications, such as specific phosphorylation at serine residues within the N-terminal transactivation domain (reviewed in Steegenga et al., 1996; Milczarek et al., 1997; Giaccia and Kastan, 1998; Fuchs et al., 1998a; Meek, 1998). A recent study showed that two serines within the first 24 amino acids of p53 are phosphorylated following exposure to ionizing and UV irradiation (Siliciano et al., 1997). This phosphorylation induces accumulation of the p53 protein and enhances its transcriptional activity (Siliciano et al., 1997). One of the two phosphorylation sites has been identified as Ser15 (Siliciano et al., 1997), a potential phosphorylation site for ATM and DNA-PK (Lees-Miller et al., 1992; Shieh et al., 1997; Banin et al., 1998; Canman et al., 1998; Woo et al., 1998). The identity of the second serine is not yet known, leaving Ser6, 9 and 20 as the potential sites (Siliciano et al., 1997). We have recently shown that substitution of serine to alanine at residues 6 and 9 does not alter the biological effects of p53, while mutations in either Ser15 or Ser20 impair the apoptotic activity of p53 (Unger et al., 1999). These findings suggest that, along with Ser15, Ser20 plays a role in regulating p53 activity; regulation is most probably achieved through phosphorylation of Ser20 in response to stress. Indeed, recent work shows clearly that Ser20 becomes rapidly phosphoylated upon exposure of cells to DNA damage (Shieh et al., 1999).
The ability of Mdm2 to inhibit p53 activities and promote its degradation (Haupt et al., 1997; Kubbutat et al., 1997; Momand and Zambetti, 1997), together with the fact that Ser20 resides within the Mdm2-binding domain of p53 (Kussie et al., 1996), strongly suggest a role for Mdm2 in mediating the reduced activity of p53-Ala20. This conjecture was confirmed experimentally by the finding that p53-Ala20 is markedly more susceptible than wt p53 to inhibition by Mdm2, as revealed when the apoptotic and transcriptional activities of p53-Ala20 were compared with those of wt p53 in the presence of limiting amounts of Mdm2. Significantly, Mdm2-promoted destabilization of p53 was enhanced by the Ser20 to Ala substitution. This differential susceptibility to Mdm2-promoted proteolytic degradation also explains the observation that p53-Ala20 accumulates less efficiently than wt p53 in cells like H1299 which express endogenous Hdm2, but not in cells lacking Mdm2.
The generality of these findings is supported by several observations. First, the increased sensitivity of p53-Ala20 to inhibition by Mdm2 can be demonstrated in multiple cell types. Secondly, the enhanced inhibition of p53-Ala20 by Mdm2 is seen not only with exogenous overexpressed protein, but also with the endogenously-induced Mdm2. Thirdly, elimination of Mdm2, such as in the mdm2 null cells, is sufficient to stabilize p53-Ala20. These results support a physiological role for Ser20 in the regulation of p53 stability and activity.
It is rather difficult to assess to what extent the promotion of p53 degradation accounts for the ability of Mdm2 to block p53 activities. In cells expressing relatively low levels of p53, this may well be the major determinant. However, when larger amounts of p53 plasmid were used, resulting in levels of p53-Ala20 more comparable with those of wt p53, the former still induced 4-fold less apoptosis (Figure 1E). Presumably, a lower ratio of Mdm2 to p53 may suffice for inhibiting the apoptotic activity of p53, while a high ratio may be required to promote its degradation as well. Therefore, the promotion of p53 for degradation and the inhibition of its activities through other mechanisms may have additive effects in terminating the p53 signal once the stress has been dealt with properly.
The increased susceptibility of p53-Ala20 to negative regulation by Mdm2 can be explained by the enhanced binding between the two proteins in vivo but not in vitro (Figure 6A and B). Presumably, phosphorylation of Ser20 reduces the binding of p53 to Mdm2, thereby alleviating the inhibitory effects of Mdm2. Indeed, a peptide with phosphorylated Ser20 was unable to compete for p53/Mdm2 binding while similar peptides with non-phosphorylated Ser20 or with Ala20 efficiently competed with wt p53 for Mdm2 binding (Figure 6C). Since the association between p53 and Mdm2 is mediated primarily through hydrophobic interactions (Kussie et al., 1996), the extra charges contributed by the phosphate group might directly interfere with this association. In addition, phosphorylation on Ser20 may also exert long-range effects, by augmenting inhibitory interactions between the N-terminal domain of p53 and other parts of the molecule, or by imposing conformational changes on the p53 protein, as suggested for phosphorylation on the adjacent Ser15 (Shieh et al., 1997). Of note, transient transfection of DNA represents an effective DNA damage signal, which induces a strong p53 activation response (Renzing and Lane, 1995; Huang et al., 1996). It is thus conceivable that under the conditions of our transient transfection assays, a significant fraction of the exogenous wt p53 protein might have undergone phosphorylation on N-terminal residues, notably Ser20. Consequently, this phosphorylated sub-population is expected to have become a less favorable Mdm2 target, accounting for the pronounced apoptotic and transcriptional activities of the transfected p53 despite the induction of endogenous Mdm2. On the other hand, mutation of Ser20 to Ala prevents p53 phosphorylation on this residue and retains the protein in a continuously Mdm2-sensitive state. This is reflected in less efficient accumulation of p53-Ala20, coupled with impaired biological activity.
Our findings complement those of Shieh et al. (1997), who showed that phosphorylation of Ser15 by DNA-PK attenuates p53–Mdm2 interaction and consequently activates p53. Overall, blocking the Mdm2-p53 interaction can provide an efficient physiological mechanism for activating and stabilizing p53, as also shown by targeting the binding site of p53 and/or Mdm2 using specific mutations, antibodies or short peptides (Böttger et al., 1997; Haupt et al., 1997; Kubbutat et al., 1997; Midgley and Lane, 1997). Covalent modifications in Mdm2 may also reduce its interaction with p53, and are expected to have a similar effect on p53 activity. In fact, it has been reported that phosphorylation of Mdm2 by DNA-PK within its p53-binding domain interferes significantly with p53 binding and with the ability of Mdm2 to abrogate the transcriptional function of p53 (Mayo et al., 1997). Other mechanisms may also be utilized towards overcoming the negative regulation of p53 by Mdm2. For instance, through direct interaction with Mdm2, the p19ARF tumor suppressor protein neutralizes Mdm2 and thereby collaborates with p53 in growth suppression (de Stanchina et al., 1998; Pomerantz et al., 1998; Zhang et al., 1998; Zindy et al., 1998). In addition, it has recently been reported that the adenovirus E1A protein promotes p53 accumulation by blocking the induction of mdm2 gene expression by p53 (Thomas and White, 1998), as well as by interfering with the association of Mdm2 with p300, which is important for p53 degradation (Grossman et al., 1998). Thus, the interaction between p53 and Mdm2 is a major target for regulation, through which p53 activity and stability are modulated.
Overall, our results strongly implicate Ser20 in the regulation of p53 activity in response to stress. Together with the findings of Shieh et al. (1999) who demonstrated that p53 is rapidly phosphorylated on Ser20 in vivo in response to ionizing radiation, our results support the following scenario: upon exposure to DNA damage and other types of stress Ser20 is phosphorylated, probably along with Ser15. The newly phosphorylated p53 protein, which is now released from negative regulation by Mdm2, accumulates and becomes activated to elicit its biological effects. Thus, through modification of Ser20, environmental cues are converted into a p53-dependent cellular response. Identifying the kinase(s) that phosphorylates Ser20 will help to delineate the signaling cascade leading to p53 activation.
Materials and methods
Cells and transfection
Mouse embryo fibroblasts were grown in Dulbecco's modified Eagle's medium (DMEM), while H1299 cells were maintained in Roswell Park Memorial Institute (RPMI). All cells were grown at 37.5°C in the presence of 10% fetal calf serum. The H1299 cell line is derived from a human lung adenocarcinoma and is devoid of any p53 expression. Cells were transfected by the calcium phosphate procedure, which includes a shock with 10% glycerol for 1 min at the end of the transfection (6–16 h from the start). The exact amount of plasmid used in each experiment is indicated in the corresponding figure legends. Whenever needed, an empty vector was used to maintain a constant amount of DNA in each transfection mix.
Luciferase assays, Western blot and immunoprecipitation analyses were carried out as previously described (Haupt et al., 1996). The antibodies used were: anti-human p53 monoclonal antibodies PAb1801, PAb421 and DO-1, anti-Hdm2 SMP14 (Picksley et al., 1994) and anti-α-tubulin antibody (DM 1A, Sigma).
In vitro binding and peptide competition assay
GST alone and GST–Mdm2 (encoding full-length mouse Mdm2) were purified from bacteria using standard procedures. Human wt p53, p53-Ala20 and p53-Asp20 were transcribed and translated in vitro in the presence of [35S]methionine using the TNT T7-coupled reticulocyte lysate system (Promega). In vitro translated p53 proteins were incubated with GST fusion proteins immobilized on glutathione agarose beads (Sigma) for 2 h at 4°C in 250 μl of binding buffer [25 mM Tris–HCl pH 7.2, 50 mM NaCl, 0.2% Nonidet P-40 (NP-40)], with continuous shaking. Following incubation the beads were washed three times with a large excess of washing buffer (100 mM Tris–HCl pH 8.0, 100 mM NaCl, 1% NP-40) and boiled in protein sample buffer. The separated proteins were resolved by SDS–PAGE and visualized by exposure to X-ray film after fluorography.
For peptide competition assays, GST–Mdm2 (residue 6–121) was purified from bacteria by glutathione agarose beads and human p53 fused to six histidines (p53-His6) was purified using nickel column chromatography (Qiagen). Peptides were purified by HPLC (100%; Alpha Diagnostic). The peptides used were:
ELISA plates were coated with 20 ng of GST–Mdm2 in phosphate-buffered saline (PBS) overnight at 4°C. Plates were blocked in 2% BSA in PBS. One hundred microliters of peptides in PBST (PBS/0.05 Tween 20)/2% bovine serum albumin (BSA), at the indicated concentrations, were added to the blocked plates for 1 h at room temperature. One hundred microliters of purified p53-His6 in PBST/1% BSA were added to the plates and incubated for an additional 15 min at room temperature. Plates were washed three times with PBST and incubated with anti-p53 PAb421 for 1 h at 4°C. Plates were washed again and incubated with an horseradish peroxidase (HRP)-coupled goat-anti-mouse IgG for 30 min at room temperature and then washed again. A color reaction was established with 100 μl of ABTS (2,2′ Azinobis[3-ethylbenzthiazoline-sulfonic acid] (1 mg/ml)/H2O2 (0.01%) in 0.1 M phosphate-citrate buffer pH 4.2 for 15 min at room temperature. The reaction was stopped with 100 μl 0.2 M citric acid. Absorbance was measured at 405 nm using Victor2 multilabel counter (Wallac).
The human p53 cDNA Ser20 to Ala or Asp mutants were generated by PCR-based oligonucleotide-directed mutagenesis. The resultant PCR products were cloned into the XbaI–human HindIII sites of pRC/CMV vector (Invitrogen). The mutant plasmids were subjected to DNA sequencing to confirm the desired mutation and to ensure that no misincorporations occurred during PCR amplification. Expression plasmids for human wt p53 were pCMV-Neo-Bam-p53 and PRC/CMV wt p53; for mouse wt mdm2, pCOC-mdm2 X2 (Barak et al., 1994; Haupt et al., 1996), and for mutant human mdm2, pCMV-Neo-Bam-Hdm2 RING. All of these cDNAs were placed under the CMV (cytomegalovirus) immediate early enhancer/promoter. The reporter plasmid used was cyclin G-luciferase (Zauberman et al., 1995). The bacterial expression plasmids for in vitro binding were GST alone and GST–Mdm2 (full-length mouse mdm2), and for peptide competition assay GST–Mdm2 (6–121) and His6-tagged human p53, kindly provided by Z.Ronai (Fuchs et al., 1998b).
At the indicated time post-transfection, adherent and floating cells were collected, fixed in methanol and stained for p53 using a mixture of anti-p53 antibodies DO-1 and PAb1801, followed by an FITC-conjugated secondary antibody. Samples were analyzed in a cell sorter (FACSCalibur) using the CellQuest software (Becton Dickinson). The FITC fluorescent intensity was plotted on a logarithmic scale. Cells were collected separately according to their fluorescent intensity. Cells with high fluorescent intensity represent the successfully transfected sub-population, while cells with low fluorescent intensity represent the non-transfected sub-population. The cell cycle distribution of each sub-population was determined by measuring the DNA content through propidium iodide staining. The apoptotic fraction was determined by quantitating the number of cells possessing a sub-G1 DNA content (Haupt et al., 1996).
We thank S.Y.Shieh and C.Prives for sharing unpublished results, A.Zauberman for the cyclin G luc reporter plasmid, R.Maya for pCMV-Neo-Bam-Hdm2Δ RING, Z.Ronai for the His-p53, B.Vogelstein for pCMV-Neo-Bam-p53 and pCMV-Neo-Bam-Hdm2, Y.Barak for pCOCmdm2X2 plasmid, D.Lane and S.Picksley for the gift of DO-1, PAb1801 and SMP14, and Y.Ben-Neriah for IκB peptide. This work was supported by the Basil O'Connor Starter Scholar Research Award grant No. 5-FY97-0700 (Y.H.), by the Concern Foundation (Y.H.), by the Center for Excellence program of the Israel Science Foundation (M.O. and Y.H.), and by NIH grant RO1 CA40099 (M.O.). E.M. is a recipient of the Eshkol Research Foundation.
- 1990) The p53 nuclear localisation signal is structurally linked to a p34cdc2 kinase motif. Oncogene, 5, 423–426. , and (
- 1998) The p53 network. J Biol Chem, 273, 1–4. , , , and (
- 1998) Enhanced phosphorylation of p53 by ATM in response to DNA damage. Science, 281, 1674–1677. et al. (
- 1994) Regulation of mdm2 expression by p53: alternative promoters produce transcripts with nonidentical translation potential. Genes Dev, 8, 1739–1749. , , and (
- 1992) Characterization of the tumor suppressor protein p53 as a protein kinase C substrate and a S100b-binding protein. Proc Natl Acad Sci USA, 89, 11627–11631. , , , and (
- 1990) Human p53 is phosphorylated by p60-cdc2 and cyclin B-cdc2. Proc Natl Acad Sci USA, 87, 4766–4770. , , , and (
- 1997) Design of a synthetic Mdm2-binding mini protein that activates the p53 response in vivo. Curr Biol, 7, 860–869. , , , , and (
- 1995) Induction of the growth inhibitor IGF-binding protein 3 by p53. Nature, 377, 646–649. , , , , , and (
- 1998) Activation of the ATM kinase by ionizing radiation and phosphorylation of p53. Science, 281, 1677–1679. , , , , , , and (
- 1995) Regulation of transcription functions of the p53 tumor suppressor by the mdm-2 oncogene. Mol Med, 1, 142–152. , and (
- 1996) Mdm-2 inhibits the G1 arrest and apoptosis functions of the p53 tumor suppressor protein. Mol Cell Biol, 16, 2445–2452. , , and (
- 1992) Features of apoptotic cells measured by flow cytometry. Cytometry, 13, 795–808. , , , , , and (
- 1998) E1A signaling to p53 involves the p19ARF tumor suppressor. Genes Dev, 12, 2434–2442. et al. (
- 1993) WAF1, a potential mediator of p53 tumor suppression. Cell, 75, 817–825. et al. (
- 1994) The carboxy-terminal serine-392 phosphorylation site of human p53 is not required for wild-type activities. Oncogene, 9, 3249–3257. , , , , , , , and (
- 1998a) Stress-activated kinases regulate protein stability. Oncogene, 17, 1483–1490. , and (
- 1998b) JNK targets p53 ubiquitination and degradation in nonstressed cells. Genes Dev, 12, 2658–2663. , , , , , and (
- 1998) The complexity of p53 modulation: emerging patterns from divergent signals. Genes Dev, 12, 2973–2983. and (
- 1996) p53 in growth control and neoplasia. Biochim Biophys Acta, 1287, 77–102. and (
- 1998) p300/MDM2 complexes participate in MDM2-mediated p53 degradation. Mol Cell, 2, 405–415. , , , , , , , and (
- 1996) Mutation of phosphoserine 389 affects p53 function in vivo. J Biol Chem, 271, 29380–29385. , , , , and (
- 1996) Cell type-specific inhibition of p53-mediated apoptosis by Mdm2. EMBO J, 15, 1596–1606. , and (
- 1997) Mdm2 promotes the rapid degradation of p53. Nature, 387, 296–299. , , and (
- 1997) 14-3-3σ is a p53-regulated inhibitor of G2/M progression. Mol Cell, 1, 3–11. , , , , , , and (
- 1997) Oncoprotein MDM2 is a ubiquitin ligase E3 for tumor suppressor p53. FEBS Lett, 420, 25–27. , and (
- 1996) Sensitivity and selectivity of the DNA damage sensor responsible for activating p53-dependent G1 arrest. Proc Natl Acad Sci USA, 93, 4827–4832. , and (
- 1992) Regulation of the specific DNA binding function of p53. Cell, 71, 875–886. , , and (
- 1995) Raf phosphorylates p53 in vitro and potentiates p53-dependent transcriptional transactivation in vivo. Oncogene, 10, 2095–2101. and (
- 1995) Rescue of embryonic lethality in Mdm2-deficient mice by absence of p53. Nature, 378, 206–208. , , and (
- 1998) Functional activation of p53 via phosphorylation following DNA damage by UV but not γ radiation. Proc Natl Acad Sci USA, 95, 2834–2837. and (
- 1997) Regulation of p53 stability by Mdm2. Nature, 387, 299–303. , and (
- 1996) Structure of the MDM2 oncoprotein bound to the p53 tumor suppressor transactivation domain. Science, 274, 948–953. , , , , , and (
- 1992) Human DNA-activated protein kinase phosphorylates serines 15 and 37 in the amino-terminal transactivation domain of human p53. Mol Cell Biol, 12, 5041–5049. , , , and (
- 1997) p53, the cellular gatekeeper for growth and division. Cell, 88, 323–331. (
- 1998) MDM2 function. Biochim Biophys Acta, 1377, M55–M59. and (
- 1995) Human TAF (II)31 protein is a transcriptional coactivator of the p53 protein. Proc Natl Acad Sci USA, 92, 5154–5158. and (
- 1997) The CDK7–cycH–p36 complex of transcription factor IIH phosphorylates p53, enhancing its sequence-specific DNA binding activity in vitro. Mol Cell Biol, 17, 5923–5934. , , and (
- 1998) Ultraviolet radiation, but not gamma radiation or etoposide-induced DNA damage, results in the phosphorylation of the murine p53 protein at serine-389. Proc Natl Acad Sci USA, 95, 6399–6402. , , and (
- 1997) Mdm-2 phosphorylation by DNA-dependent protein kinase prevents interaction with p53. Cancer Res, 57, 5013–5016. , and (
- 1996) mdm2 deletion does not alter growth characteristics of p53-deficient embryo fibroblasts. Oncogene, 13, 1731–1736. , , and (
- 1998) Multisite phosphorylation and the integration of stress signals at p53. Cell Signal, 10, 159–166. (
- 1990) The p53 tumour suppressor protein is phosphorylated at serine 389 by casein kinase II. EMBO J, 9, 3253–3260. , , and (
- 1997) p53 protein stability in tumour cells is not determined by mutation but is dependent on Mdm2 binding. Oncogene, 15, 1179–1189. and (
- 1997) p53 phosphoryl- ation: biochemical and functional consequences. Life Sci, 60, 1–11. , and (
- 1992) Phosphorylation of the p53 tumour-suppressor protein at three N-terminal sites by a novel casein kinase I-like enzyme. Oncogene, 7, 1361–1369. , , and (
- 1995) p53 is phosphorylated in vitro and in vivo by an ultraviolet radiation-induced protein kinase characteristic of the c-Jun kinase, JNK1. J Biol Chem, 270, 5511–5518. , , and (
- 1995) Tumor suppressor p53 is a direct transcriptional activator of the human bax gene. Cell, 80, 293–299. and (
- 1997) Mdm-2: ‘big brother’ of p53. J Cell Biochem, 64, 343–352. and (
- 1995) Rescue of early embryonic lethality in mdm2-deficient mice by deletion of p53. Nature, 378, 203–206. , and (
- 1993) Oncoprotein MDM2 conceals the activation domain of tumour suppressor p53. Nature, 362, 857–860. , , , , and (
- 1995) Wild-type human p53 and a temperature-sensitive mutant induce Fas/APO-1 expression. Mol Cell Biol, 15, 3032–3040. et al. (
- 1994) Immunochemical analysis of the interaction of p53 with MDM2; fine mapping of the MDM2 binding site on p53 using synthetic peptides. Oncogene, 9, 2523–2529. , , and (
- 1997) Mdm2: keeping p53 under control. Oncogene, 15, 1001–1010. , and (
- 1997) A model for p53-induced apoptosis. Nature, 389, 300–305. , , , and (
- 1998) The Ink4a tumor suppressor gene product, p19Arf, interacts with MDM2 and neutralizes MDM2's inhibition of p53. Cell, 92, 713–723. et al. (
- 1995) Cdk2 kinase phosphorylates serine 315 of human p53 in vitro. Oncogene, 11, 73–80. , and (
- 1998) Signaling to p53: breaking the Mdm2-p53 circuit. Cell, 95, 5–8. (
- 1995) p53-dependent growth arrest following calcium phosphate-mediated transfection of murine fibroblasts. Oncogene, 10, 1865–1868. and (
- 1998) Nucleo-cytoplasmic shuttling of the hdm2 oncoprotein regulates the levels of the p53 protein via a pathway used by the human immunodeficiency virus rev protein. EMBO J, 17, 554–564. , , , and (
- 1998) DNA damage activates p53 through a phosphorylation–acetylation cascade. Genes Dev, 12, 2831–2841. , , , , , , and (
- 1997) DNA damage-induced phosphorylation of p53 alleviates inhibition by MDM2. Cell, 91, 325–334. , , and (
- 1999) DNA damage-inducible phosphorylation of p53 at N-terminal sites including a novel site, Ser20, requires tetramerization. EMBO J, 18, 1815–1823. , and (
- 1997) DNA damage induces phosphorylation of the amino terminus of p53. Genes Dev, 11, 3471–3481. , , , , and (
- 1996) How phosphorylation regulates the activity of p53. J Mol Biol, 263, 103–113. , and (
- 1995) Regulation of the sequence-specific DNA binding function of p53 by protein kinase C and protein phosphatases. J Biol Chem, 270, 5405–5411. , , and (
- 1998) Suppression of the p300-dependent mdm2 negative-feedback loop induces the p53 apoptotic function. Genes Dev, 12, 1975–1985. and (
- 1992) p53: a transdominant regulator of transcription whose function is ablated by mutations occurring in human cancer. EMBO J, 11, 1383–1390. , , and (
- 1993) Functional domains of wild-type and mutant p53 proteins involved in transcriptional regulation, transdominant inhibition and transformation suppression. Mol Cell Biol, 13, 5186–5194. , , , and (
- 1999) Mutations in serines 15 and 20 of human p53 impair its apoptotic activity. Oncogene (in press). , , , , , and (
- 1996) Biological and clinical importance of the p53 tumor suppressor gene. Clin Chem, 42, 858–868. and (
- 1995) Increased and altered DNA binding of human p53 by S and G2/M but not G1 cyclin-dependent kinases. Nature, 376, 88–91. and (
- 1998) ATM-dependent activation of p53 involves dephophorylation and association with 14-3-3 proteins. Nature Genet, 19, 175–178. , , and (
- 1998) DNA-dependent protein kinase acts upstream of p53 in response to DNA damage. Nature, 394, 700–704. , , , and (
- 1995) A functional p53-responsive intronic promoter is contained within the human mdm2 gene. Nucleic Acids Res, 23, 2584–2592. , , , and (
- 1998) ARF promotes MDM2 degradation and stabilizes p53: ARF-INK4a locus deletion impairs both the Rb and p53 tumor suppression pathways. Cell, 92, 725–734. , and (
- 1998) Myc signaling via the ARF tumor suppressor regulates p53-dependent apoptosis and immortalization. Genes Dev, 12, 2424–2433. , , , , , and (