Introduction

The p53 protein is a tightly regulated nuclear phosphoprotein that plays a critical role in the cellular response to environmental stress. Upon a cell's exposure to stress, such as hypoxia or genomic damage, the p53 protein is expressed at high levels and is post-translationally modified. These modifications, which include phospho-rylation, acetylation and glycosylation occur rapidly and lead to the activation of p53 (Lees-Miller et al., 1990; Ko and Prives, 1996), resulting in either G1 growth arrest or programmed cell death (Levine, 1997; Agarwal et al., 1998). In this activated state, p53 exists as a tetramer that binds DNA in a sequence-specific manner (El-Deiry et al., 1992; Funk et al., 1992). Target genes of this transcription factor include, but are not limited to, MDM2, p21waf, bax and gadd45 genes (Kastan et al., 1992; Harper et al., 1993; Juven et al., 1993; Wu et al., 1993; Miyashita and Reed, 1995).

Following DNA damage, p53 is phosphorylated on serine residues 15, 20, 33 and 37 within the amino-terminal domain. These phosphorylation events are thought to play a key role in regulating p53 stability and activity (Lambert et al., 1998; Ashcroft and Vousden, 1999). Various studies have highlighted the importance of several of these residues in regulating p53 function; however, the significance of phosphorylated S37 has been more difficult to ascertain. p53 is phosphorylated at S37 in response to both ultraviolet (UV)- and -irradiation (Sakaguchi et al., 1998) and may have different functions in response to these two different DNA-damaging treatments. Bean and Stark (2001) have suggested that phosphorylation at this site following UV-irradiation is needed to maintain p53 stability, since mutation of S37 to alanine impaired the accumulation of p53 after exposing cells to UV. In addition, Sakaguchi et al. (1998) suggested that the phosphorylation of S37 and/or S33 in response to UV- or -irradiation increased the affinity of p53 for coactivators PCAF and p300 and promoted the acetylation of p53 carboxy-terminal residues Lys-382 and Lys-320. Acetylation at either of these sites was thought to inactivate the nonsequence-specific binding activity of the p53 carboxyl-terminus and enhance sequence-specific DNA binding.

It has been shown that multiple protein serine/threonine phosphatases can dephosphorylate p53 in vitro. These phosphatases include protein phosphatase 1 (PP1), protein phosphatase 2A (PP2A), protein phosphatase 5 (PP5), Wip1 and Cdc14 (Scheidtmann et al., 1991; Takenaka et al., 1995; Fiscella et al., 1997; Zuo et al., 1998; Li et al., 2000). In vivo, cells treated with the PP1 and PP2A inhibitor okadaic acid (Cohen, 1991) accumulated hyperphosphorylated p53 (Bialojan and Takai, 1988). Unfortunately, the specific hyperphosphorylated amino-acid residues were not determined (Yan et al., 1997). Okadaic acid induces programmed cell death in many cell types, suggesting that PP1 or PP2A are components of signaling pathways that regulate apoptosis (Yan et al., 1997). PP2A has been shown to act at different points in the cell cycle in various systems, has been implicated as a positive and a negative stimulus for cell proliferation (Cohen and Cohen, 1989; Kawabe et al., 1997; Wilson et al., 1999) and appears to be essential for cell viability (Ruediger et al., 1991; Baharians and Schonthal, 1998; Gotz et al., 1998). Moreover, inhibition of PP2A by okadaic acid can negatively regulate the immediate early gene expression induced by various extracellular stimuli, such as growth factors and hormones (Guy et al., 1992).

PP2A is one of the four major classes of serine/threonine phosphatases (Cohen et al., 1989). The core of the PP2A holoenzyme is the catalytic C subunit (PP2Ac). This subunit predominately associates with the regulatory A subunit to form an AC dimer that can exist alone or form a trimeric ABC complex with a B subunit family member (Mumby and Walter, 1993; Wera and Hemmings, 1995), which regulates the substrate specificity and subcellular localization of the holoenzyme. Subunit composition and post-translational modifications, such as phosphorylation and methylation, determine the activity of PP2Ac (Chen et al., 1992; Bryant et al., 1999).

In the present study, we show that phosphorylation at S37 is important for the transcriptional activity of p53. Furthermore, we demonstrate that p53 and PP2Ac associate in vivo following ionizing radiation. The increase in association correlates with an increase in PP2A protein and activity, suggesting that PP2A is involved in the rapid turnover of the phosphoserine 37 residue on p53 and regulates the transcriptional activation of the p53 protein after ionizing radiation.

Results

Transient phosphorylation of p53 at serines 15 and 37 following -irradiation

Following DNA damage by UV- or -irradiation, p53 is stabilized and phosphorylated at a number of residues, including serines 15 and 37. A Western blot time course of Molt-4 nuclear extracts (NEs) harvested at 0, 4, 8, 12 and 24 h after -irradiation using specific antibodies is shown in Figure 1. p53 protein rapidly increased within the first 4 h following -irradiation and was phosphorylated at serines 15 and 37. Both S15 and S37 remained phosphorylated through the 8 h time point. Interestingly, the level of S15 and S37 phosphorylation decreased at later time points. The decrease in phosphorylation cannot be entirely attributed to a decrease in protein levels, since the level of total p53 protein did not change as significantly as the drop in the levels of phosphorylated p53 over the same time course.

Figure 1.

p53 is transiently phosphorylated at S15 and S37 in vivo following -irradiation. NEs were prepared at the times indicated from Molt-4 cells exposed to 8 Gy -rays from a 68 Mark I ¹³⁷Cs irradiator. The samples were analysed by Western blot for p53 using DO-1, anti-p53-phosphoserine 15 and anti-p53-phospho-serine 37 as noted. The figure is representative of three independent time courses

Full figure and legend (40K)

Serine 37 is required for p53 transcriptional activity

The transient nature of p53 phosphorylation at serine 15 and 37 in vivo suggests their potential importance in the regulation of p53. Previous studies, which used the N-terminal domain of p53 fused to the DNA-binding domain of the yeast GAL4 protein to measure the transactivation of p53 and p53 phosphorylation site mutants, showed that phosphorylation of S15, but not S37, is critical for p53-dependent transactivation (Dumaz and Meek, 1999). To determine the effect of these phosphorylation site mutants in the context of the whole protein and without the added complexity of p53 degradation, we cotransfected p53/MDM2 knockout mouse embryonic fibroblast (MEF) cells, which have no endogenous p53 or MDM2 activity, with wild-type p53 or mutants p53-S15A and p53-S37A. Western blot analysis verified that the amount of transiently expressed p53 was similar to the endogenous level of p53 expressed in -irradiated Molt-4 cells (Figure 2a). In addition, phosphorylation of the p53 constructs at S15 or S37 indicated that the transfection alone triggered an effective p53 DNA damage response (Dumaz and Meek, 1999). In light of this, it was not necessary to further -irradiate the cells after transfection. Consistent with previous reports, the p53-S15A mutant caused a decrease in the transcriptional activity of p53 in the reporter assays (Figure 2b). Interestingly, we found that the p53-S37A mutant also caused a marked decrease in p53 transactivation activity in the context of the whole protein. Such a decrease was not seen when the Gal4-p53-S37A mutant was transfected into the p53/MDM2 knockout MEFs (data not shown), suggesting that the GAL4 construct behaves differently than the full-length protein.

Figure 2.

Full-length phosphorylation site mutant p53-S37A has reduced transcriptional activity compared to wild type. p53- and MDM2-null MEFS (p53^-/-, MDM2^-/-) were cotransfected with a MDM2-luciferase reporter and full-length wild-type p53 or p53 phosphorylation site mutants p53-S15A and p53-S37A. At 24 h post-transfection, WCEs were analysed by Western blot (a) or assayed for transcriptional activity of p53 (b). WCEs from Molt-4 cells harvested 4 h after mock treatment (Molt4) or -irradiation (Molt4 ) were used to compare the transient p53 protein expression to endogenous p53 expression. WT p53 was detected using DO-1 and p53 phosphorylated at S15 and S37 was detected using anti-p53-phosphoserine antibodies specific for these residues. The data displayed are representative of at least two separate transfection experiments assayed in triplicate

Full figure and legend (73K)

The decrease in p53 transcriptional activity seen using the full-length p53-S37A construct cannot be explained by the protein expression since Western blots of the extracts indicate that the protein levels were comparable in all samples (Figure 2a). This result might be expected since the MEF cells were deficient in MDM2. Furthermore, nonspecific squelching by the mutant was also not responsible for the low transcriptional activity given that experiments were performed within the linear response range such that decreasing or increasing the amount of p53-S37A transfected into the cells led to a reduction or elevation, respectively, in the observed transcriptional activity (data not shown). Thus, these data suggest that transactivation of p53 is regulated by phosphorylation at S37.

Okadaic acid inhibits the phosphatase activity seen in vivo and in vitro

Upon observing the importance of S37 for the transcriptional activity of p53, we sought to determine how phosphorylation at this site was regulated. We reasoned that PP1 and/or PP2A protein serine/threonine phosphatases may be responsible for the phosphatase activity directed against S37 contained within the -irradiated extracts. Whole-cell extracts (WCEs) were prepared from Molt-4 cells 4 h after -irradiation in the presence or absence of 50 nM okadaic acid. Okadaic acid is a strong phosphatase inhibitor, capable of inhibiting both PP1 and PP2A at high concentrations (Bialojan and Takai, 1988). At lower concentrations, okadaic acid specifically inhibits PP2A (Cohen, 1991). p53 was either immunoprecipitated with anti-p53 antibody DO-1 (Figure 3a) and probed with anti-p53-phosphoserine 37 antibody (Figure 3b) or immunoprecipitated with anti-p53-phosphoserine 37 antibody and probed with anti-p53 antibody AB-7 (Figure 3c). As shown in Figure 3a, the anti-p53 antibody immunoprecipitated the same amount of p53 protein from the -irradiated extracts in the presence or absence of okadaic acid. However, Western blot analysis with anti-p53-phosphoserine 37 antibody indicated that WCE from cells -irradiated following the addition of okadaic acid had approximately a 3–7-fold increase in S37 phosphorylation (Figure 3b, c). p53 was not detected in nonirradiated cells (Figure 3a–c).

Figure 3.

Okadaic acid inhibits phosphatase activity in vivo. WCEs were prepared from Molt-4 cells 4 h after 8 Gy of -irradiation in the presence or absence of 50 nM okadaic acid. p53 was immunoprecipitated with anti-p53 antibody DO-1 and probed with either DO-1 (a) or with anti-p53-phosphoserine 37 (b). (c) WCEs were immunoprecipitated with anti-p53-phosphoserine 37 and then probed with anti-p53 antibody AB-7. The data presented are representative of at least two independent experiments each

Full figure and legend (91K)

Purified PP2A dephosphorylates p53 at serine 37 in vitro

We next used purified PP2A enzyme to treat GST-p53 proteins in phosphatase assays in vitro. After phospho-rylating GST-p53 at S15 and S37 and GST-p53-S15A at S37 with DNA-PK, the proteins were incubated with PP2A or PP2A in the presence of 10 nM okadaic acid. The reaction products were analysed by Western blot using anti-p53-phosphoserine 37 antibody. Our results indicate that purified PP2A dephosphorylated p53 at S37 and that this dephosphorylation was inhibited by okadaic acid (Figure 4).

Figure 4.

Purified PP2A dephosphorylates p53 at S37 in vitro. Following treatment with DNA-PK and cold ATP, GST-p53 (a) and the phosphorylation site mutant GST-p53-S15A (b) were incubated with purified PP2A or PP2A and 10 nM okadaic acid. The products were separated on a gel, transferred to a PVDF membrane and probed with anti-p53-phosphoserine 37 antibody

Full figure and legend (91K)

Protein level of the catalytic subunit of PP2A increases after -irradiation

Although PP2A protein expression is tightly regulated (Baharians and Schonthal, 1998), Wilson et al. (1999) have reported an increase in the activity of PP2A associated with an increase in the level of protein. Western blots of MCF-7 WCE were probed with antibodies directed against the different subunits of the PP2A complex. Although the rabbit antibody used to detect the catalytic subunit of PP2A (PP2Ac) recognizes the carboxy-terminal domain of the protein in a region where the protein can be methylated, the antibody was not affected by methylation or demethylation of PP2Ac (data not shown). Figure 5 shows a time-dependent, 1.7-fold, increase in the protein level of PP2Ac following irradiation, as well as a 13-fold increase in the level of p53. Interestingly, the expression level of the regulatory subunits, A and B, did not vary significantly over time. This latter result may be due to the fact that the antibodies used to detect A and B recognize a global pool of the regulatory subunits and therefore may not be able to detect an increase of one specific isoform.

Figure 5.

-Irradiation increases PP2Ac protein levels in MCF-7 cells. WCEs were harvested at the times indicated from MCF-7 cells exposed to 8 Gy -rays. Extracts were loaded onto a gel and then transferred to PVDF membranes. Blots were probed with an anti-p53 antibody DO-1, anti-PP2Ac antibody, anti-PP2A/A antibody or anti-PP2A/B antibody as indicated. The membranes were subsequently reprobed for -tubulin to control for loading. This gel is representative of three independent experiments

Full figure and legend (96K)

p53 associates with PP2Ac in MCF-7 cells

The experiments presented thus far demonstrate that PP2A can dephosphorylate p53 in vitro and in vivo and that the level of PP2Ac and p53 is increased following -irradiation. It was of interest to determine whether these results could be correlated with the intracellular distribution pattern of these two proteins. To address this point, we analysed the endogenous p53 and PP2Ac distribution in MCF-7 cells by indirect immunofluorescence microscopy.

In nonirradiated cells, p53 staining was not detectable above background (Figure 6, panel a). Upon -irradiation, p53 staining localized to the nucleus in numerous small foci and the intensity of p53 showed a diffuse punctuate fluorescence signal concentrating in the nucleoplasm that increased with time (panels b, c, g and j). This increase in p53 signal was consistent with the increased protein levels seen in Figure 5. PP2Ac also displayed similar punctuate staining throughout the nucleoplasm, the intensity of which also increased with time after -irradiation (panels d–g, j) and was consistent with Western blot analysis of PP2Ac (Figure 5).

Figure 6.

p53 associates with PP2Ac in MCF-7 cells. Endogenous p53 (red) and PP2Ac (green) were detected by indirect immunofluorescence microscopy using specific antibodies, 2 h (panels b and e) and 4 h (panels c and f) after 8 Gy of -irradiation. In panels a and d, MCF-7 cells were not -irradiated. Colocalization of p53 with PP2Ac in the nucleus is shown in panels g and j corresponding to 2 and 4 h after -irradiation, respectively. In both panels g and j, an example of colocalization area in the nucleoplasm is marked by a rectangle and enlarged as an inset. Left insets are red signal only, representing p53, middle insets are green signal only, representing PP2Ac and right insets are the overlay images of both. In the overlays, yellow–orange indicates colocalizations. Two sets of arrows in panels g and j point to the planes of linescans displayed in panels h and i (2 h after -irradiation) and in panels k and l (4 h after -irradiation). In panels h, i, k, l, red and green curves represent the fluorescent intensity distributions of p53 and PP2Ac signals at the different positions along the linescan. Scale bars, 10 m in panels a and d, 2 m in panels g and j. Each of the images is representative of 60 images collected from three independent experiments

Full figure and legend (585K)

The punctate nucleoplasmic distribution of PP2Ac and p53 prompted us to investigate the possibility of intranuclear association between these two proteins. We compared the nucleoplasmic distribution of p53 to that of the PP2Ac by high-resolution dual-staining indirect immunofluorescence microscopy (Van Steensel et al., 1996; Elbi et al., 2002). We observed significant colocalizations (yellow–orange) between p53 (red) and PP2Ac (green) 2 and 4 h after -irradiation (panels g and j). Association of p53 with PP2Ac was verified by linescan analysis. The representative linescans in Figure 5 (panels h, i, k and l) indicate the overlap of some but not all fluorescence intensity peaks from both signals. Quantitation of colocalization percentages showed that 203% (means.d.) of p53 distributions colocalized with PP2Ac distributions. From these experiments, we concluded that a subpopulation of p53 associates with PP2Ac in the nucleoplasm of -irradiated MCF-7 cells.

p53 and the catalytic subunit of PP2Ac coimmunoprecipi-tate in -irradiated cells

To further confirm the interaction between PP2Ac and p53 in the nucleus of the MCF-7 cells following -irradiation, we performed a coimmunoprecipitation (co-IP) time-course experiment using NEs from -irradiated MCF-7 cells. The p53 protein was immunoprecipitated from the NEs using an antibody directed toward the amino-terminus of p53, anti-p53 pAb1801. The protein complexes were resolved on a denaturing gel and Western blots performed using an antibody directed against PP2Ac. As seen in Figure 7a, there was a substantial amount of PP2Ac present in immunoprecipitates prepared with a monoclonal antibody directed against p53, but not with the mouse IgG control. The co-IP was minimal at the zero time point; however, after 4 h the ratio of p53 to PP2Ac increased. The amount of PP2A associated at the 4 h time point was three times more than that calculated using the ratio of PP2A to p53 in the untreated cells, further confirming that -irradiation increased the specific association of p53 and PP2Ac.

Figure 7.

p53 associated PP2Ac activity. (a) NEs were harvested at times indicated from MCF-7 cells exposed to 8 Gy -rays and immunoprecipitations were performed with an IgG control or p53 1801 antibody. The immunoprecipitates were separated on a gel, transferred to a PVDF membrane and probed for PP2Ac. Protein bands were detected by ECL. p53 was detected with DO-1. (b) p53 1801 immunoprecipitates were collected from a time-course experiment using WCE from -irradiated MCF-7 cells. Immunoprecipitates were analysed for phosphatase activity as described in 'Materials and methods'. Each of these results is representative of three independent experiments

Full figure and legend (128K)

To assess whether the p53-associated PP2Ac was active, we performed in vitro phosphatase assays using the immunoprecipitated p53-PP2Ac complex. Phosphorylation site mutants GST-p53-S15A and GST-p53-S37A were labeled by DNA-PK at serine 37 and 15, respectively. The labeled GST-p53 proteins were incubated with immunoprecipitates collected from a time-course experiment using WCEs from -irradiated cells. Consistent with the results of the co-IP experiment, there was a time-dependent reduction in the phosphorylation level at S37. This decrease was not seen in the IgG control (Figure 7b). Interestingly, the level of ³²P-labeled S15 (GST-p53-S37A) remained relatively constant for all time points. Together, these data indicate that under ionizing radiation conditions, PP2A associates with p53 and that, following this interaction, the phosphatase activity of PP2A is directed specifically toward the S37 residue on p53.

Discussion

The p53 tumor suppressor plays a critical role in mechanisms that respond to a wide range of cellular stresses. In response to DNA damage, p53 is phosphorylated at serine residues within its N-terminal domain (Shieh et al., 1997, 1999; Siliciano et al., 1997). These post-translational modifications are important for p53 regulatory events ranging from activation to degradation (Ashcroft and Vousden, 1999). The functional significance of phosphorylated S37 has remained somewhat elusive. We have now shown using full-length p53 that phosphorylation at S37 is important for the transcriptional activity of p53. In addition, we have shown that this activity is regulated by the dephosphorylation of S37 by PP2A.

The GAL4-p53 constructs used in studies by Dumaz and Meek (1999) did not contain the carboxy-terminus of p53. The absence of this domain may explain why only S15 and not S37 was found to be critical for p53-dependent transactivation. We confirmed the relative roles of S15 and S37 phosphorylation for p53-dependent transactivation using reporter assays to compare the N-terminal activation domains of GAL4-p53 constructs. Moreover, our results showed for the first time that the single mutation of S37 to alanine dramatically impairs p53-dependent transactivation in the context of the whole protein. These results suggest that the carboxy-terminus of p53 induces an inactive conformation that is overcome by phosphorylation at S37. In addition, these results are consistent with the phosphorylation–acetylation cascade proposed by Sakaguchi et al. (1998).

Phosphorylation of p53 at S37 is a transient event evident 4 h post-irradiation and absent at later time points (Sakaguchi et al., 1998 and this work). Interestingly, we could detect phosphatase activity toward p53 at S37 as early as 2 h after -irradiation (Figure 7b). During this time frame, p53 protein was still accumulating and being phosphorylated in response to DNA damage (Figure 1). Therefore, regulation of p53 at this site appears to be continuous in response to DNA damage and does not occur in waves. This dynamic 'on–off' switch affords tight regulation of this critical protein.

A role for PP2A in p53 regulation would be consistent with reports that PP2A may be involved in the genomic stability of a cell (Schonthal, 1998). Of particular interest is the observation that the increase in PP2A activity and expression correlates with its association with p53 in the nucleus. This association was observed by co-IP and indirect immunofluorescence microscopy experiments. Both techniques indicated that p53 was associated with PP2Ac following -irradiation (Figures 6, 7). Further investigation is needed to determine the mechanism by which PP2Ac expression is increased following -irradiation.

Although it has been reported that cyclin G recruits PP2A to bind to and dephosphorylate MDM2 thereby regulating both MDM2 and p53 (Okamoto et al., 1996, 2002; Kimura and Nojima, 2002), we have been unable to show the presence of cyclin G in the p53-PP2A complexes immunoprecipitated in this work (unpublished data). The absence of such a complex suggests that PP2A regulates p53 using a pathway distinct from the one that utilizes cyclin G. PP2A might regulate p53 transcriptional activity through interaction with a deacetylase. A similar scenario, involving the dephosphorylation of cAMP-responsive element-binding protein (CREB), was recently demonstrated by Canettieri et al. (2003). Phosphorylation of CREB at S133 promotes recruitment of CREB-binding protein (CBP) and p300, which stimulate acetylation of promoter-bound histones during transcription. Histone deacetylase 1 promoted S133 dephosphorylation by way of a stable interaction with PP1. We are currently investigating this possibility.

In conclusion, we show for the first time that phosphorylation of p53 at S37 is important for the transcriptional activity of p53. Furthermore, we provide novel evidence that p53 molecules associate with PP2A in vivo following -irradiation and that this association results in a rapid turnover of phosphoserine 37 on p53. These observations correlate well with the previous work proposing that phosphorylation of p53 at S37 in response to DNA damage leads to acetylation of the carboxy-terminus and sequence-specific DNA binding. It is also possible that the transcriptional activity stimulated by the transient phosphorylation at S37 in the response to DNA damage may aid in the recruitment of other transcription factors to the promoter by inhibiting MDM2 binding to p53. This latter possibility does not preclude a direct role for S37 in the accumulation of p53. Whether or not UV-irradiation, which also causes phosphorylation of p53 at S37, elicits the same functional response of p53 remains to be seen.

Materials and methods

Cell culture growth conditions

Molt-4 cells were maintained in RPMI 1640 supplemented with 2 mM glutamine, 50 U/ml penicillin, 0.1 mg/ml streptomycin and 10% fetal bovine serum (FBS) at 37°C in a 5% CO₂ atmosphere. MCF-7, a human cell line with a wild-type p53 and MEFs (p53^-/- MDM2^-/-) were each cultured in Dulbecco's modified Eagle's medium supplemented with 2 mM glutamine, 50 U/ml penicillin, 0.1 mg/ml streptomycin and 10% FBS at 37°C in a 5% CO₂ atmosphere.

Nuclear extracts

Cells (1 10⁷) were harvested with trypsin, washed and gently resuspended in 400 l of buffer A (10 mM HEPES (pH 7.8), 10 mM KCl, 2 mM MgCl₂, 0.1 mM EDTA, 1 mM DTT, 1 mM PMSF, 4 g/ml leupeptin). After 15 min incubation on ice, cells were lysed with 0.5 % NP40. Nuclear pellets were obtained by centrifugation, resuspended in 50 l of buffer B (20 mM HEPES (pH 7.9), 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 1 mM PMSF) and incubated at 4°C for 15 min. The NE was collected by centrifugation.

Transient transfections and reporter assays

At 1 day after plating, 60–80% confluent MEFs (p53^-/-, MDM2^-/-) were cotransfected with 0.025–0.5 g p53 construct to obtain equal protein expression levels as determined by Western blot, 0.5 g MDM2-luciferase reporter or GAL4-TK-luciferase reporter and 0.4 g CMV--galactosidase control plasmid using Effectene Transfection Reagent (Qiagen Inc.). Wild-type and phosphorylation site mutant p53 constructs (S15A and S37A) have been previously reported (Pise-Masison et al., 2000) and GAL4-p53 constructs were a kind gift from D Meek (Dumaz and Meek, 1999). Cells were harvested 24 h after transfection for Western blot or luciferase/-galactosidase analysis (Tropix).

Western blot analysis

Molt-4 NEs (50 g) from cells -irradiated with 8 Gy from a 68 Mark I Irradiator (JL Shepherd & Assoc.) were separated by electrophoresis on 10% Tris-glycine gels (Novex). The proteins were transferred to PVDF membranes (Millipore) and analysed for the presence of p53 using anti-p53 DO-1 (Oncogene Research), anti-p53-phosphoserine 15 (Pise-Masison et al., 2000) and anti-p53-phosphoserine 37 (Cell Signaling) antibodies. Protein bands were visualized using enhanced chemiluminescence (ECL) (Pierce). To analyse proteins from transfected MEFs (p53^-/-, MDM2^-/-), WCEs were prepared using Eucaryotic Lysis Buffer (50 mM Tris-HCL pH 7.4, 120 mM NaCl, 5 mM EDTA, 0.5% NP40, 50 mM NaF, 0.2 mM Na₃VO₄, and one complete protease inhibitor tablet (Roche Applied Science)/50 ml). The lysates were incubated on ice for 20 min, cleared by centrifugation at 4°C for 10 min and 25 g was analysed as above. The individual PP2A subunits were detected in 20 g samples of WCEs from -irradiated MCF-7 cells using antibodies directed against the catalytic subunit PP2Ac (Upstate USA, Inc.) and the two regulatory subunits A and B (Calbiochem). Band intensities were analysed by densitometry using ImageQuant (Molecular Dynamics).

Immunoprecipitation

Anti-p53 DO-1-conjugated agarose (Oncogene Research) and anti-p53-phosphoserine 37 (Cell Signaling) were used to immunoprecipitate p53 from Molt-4 NEs. The NE was prepared from 5 10⁷ cells irradiated with approximately 8 grays in the presence or absence of 50 nM okadaic acid. Proteins were immunoprecipitated overnight at 4°C. Protein G/plus protein A agarose (Oncogene Research) was added to the phosphoserine 37 mixture and incubated for another 2 h to precipitate the p53 phosphorylated at S37. All the beads were collected by centrifugation then washed three times with ice-cold NE buffer B. After the final wash, proteins were released with the addition of SDS loading buffer. Western blot analysis was performed as described using anti-p53 AB-7 (Oncogene Research) and anti-p53-phosphoserine 37 antibodies. Co-IP of PP2A and p53 was performed by incubating 300 g of MCF-7 NE with 400 ng of mouse monoclonal anti-p53 1801 (Oncogene Research) for 1 h at 4°C. Protein G–Sepharose (Fast Flow 4, Pharmacia Biotech) was added for 1 h at 4°C to precipitate the 1801 antibody. The beads were collected by centrifugation and then washed five times with ice-cold TNN buffer (50 mM Tris-HCl (pH 7.4), 120 mM NaCl, 5 mM EDTA, 0.5% NP40, 0.2 mM Na₃VO₄, 1 mM DTT, 1 mM PMSF, 20 g/ml aprotinin). After the addition of SDS loading buffer, Western blot analysis was performed as described above.

Expression and purification of GST-p53 proteins

GST-p53-S15A and GST-p53-S37A were prepared from Escherichia coli HB101 cells grown to a 600_A reading of approximately 0.5 and induced for 4 h at 37°C with 0.1 mM isopropyl--thiogalactopyranoside (Lambert et al., 1998). The bacteria were pelleted, resuspended in ice-cold phosphate-buffered saline (PBS) with 1% Triton X-100 and then lysed by sonication. After the removal of debris by centrifugation, the fusion protein was adsorbed onto glutathione–Sepharose 4B beads (Pharmacia Biotech). The beads were washed three times with cold lysis buffer and resuspended in the same buffer.

Phosphorylation of p53 by DNA-PK

DNA-PK, which specifically phosphorylates p53 on serines 15 and 37, was used to introduce phosphate groups at these amino acids (Lambert et al., 1998; Ashcroft and Vousden, 1999). Phosphorylation of p53 by DNA-PK (Promega) was carried out according to the manufacturer's instructions using 50 l of the GST-p53 fusion protein as a substrate, 5 Ci of [-³²P]ATP (unless otherwise indicated) and 10 l of 0.1 mM cold ATP. Following phosphorylation, the beads were washed three times with phosphatase buffer (20 mM Tris pH 7.5, 50 mM NaCl, 6 mM MgCl₂, 5 mM EGTA).

Immunofluorescence microscopy

MCF-7 cells were grown on glass coverslips and kept at subconfluent level. Indirect immunofluorescence was performed as previously described (Elbi et al., 2002). Briefly, after exposure to 8 Gy -irradiation, cells were fixed with 2% paraformaldehyde in PBS. The cells were permeabilized in 0.5% Triton X-100 in PBS for 5 min on ice and incubated with monoclonal anti-p53 1801 (Oncogene Research) or polyclonal anti-PP2Ac (Upstate USA, Inc.). Cells were rinsed in PBS and incubated with Texas Red-conjugated goat-anti-mouse and FITC-conjugated goat-anti-rabbit antibodies (Jackson Immunolabs). Cells were mounted using prolong (Molecular Probes) and observed with the NIKON E800 microscope equipped with a cooled CCD camera (MicroMax, 5 MHz). The colocalization percentage was calculated using 60 randomly selected cells from three independent experiments. Single optical sections from the middle of the cells were used for quantitations. Red and green fluorescent signals with fluorescent intensity values ranging within the top 35% for each fluorescent channel were identified by thresholding using Metamorph software (Universal Imaging Corp.) and the percentage of pixels having the same position in both threshold images was calculated.

In vitro phosphatase assays

The GST-p53 proteins used for the purified PP2A phosphatase assays were phosphorylated by DNA-PK at serines 15 and 37 using cold ATP. Phosphatase reactions were carried out in a 20 l volume containing 1 l of phosphorylated GST-p53 protein (50% slurry), 0.025 U purified PP2A (Upstate USA, Inc.), 50 mM Tris pH 7, 0.1 mM CaCl₂, and 1.0 mM MgCl₂. Where indicated, reactions also contained 10 nM okadaic acid. Reactions were allowed to proceed for 30 min at 30°C and were subsequently stopped with the addition of loading buffer. Samples were analysed by Western blot as described. The GST-p53 substrates used for the immunoprecipitated PP2A phosphatase assay were phosphorylated using [-³²P]ATP, loaded onto Tris-glycine gels (Novex), visualized using a PhosphorImager (Molecular Dynamics) and analysed by densitometry using ImageQuant.

References

1. Agarwal ML, Taylor MV, Chernov MV, Chernova OB and Stark GR. (1998). J. Biol. Chem., 273, 1–4. | Article | PubMed | ISI | ChemPort |
2. Ashcroft M and Vousden KH. (1999). Oncogene, 18, 7637–7643. | Article | PubMed | ISI | ChemPort |
3. Baharians Z and Schonthal AH. (1998). J. Biol. Chem., 273, 19019–19024. | Article | PubMed | ISI | ChemPort |
4. Bean LJ and Stark GR. (2001). Oncogene, 20, 1076–1084. | Article | PubMed | ISI | ChemPort |
5. Bialojan C and Takai A. (1988). Biochem. J., 256, 283–290. | PubMed | ISI | ChemPort |
6. Bryant JC, Westphal RS and Wadzinsky BE. (1999). Biochem. J., 339, 241–246. | Article | PubMed | ISI | ChemPort |
7. Canettieri G, Morantte I, Guzman E, Asahara H, Herzig S, Anderson SD, Yates JR and Montminy M. (2003). Nat. Struct. Biol., 10, 175–181. | Article | PubMed | ISI | ChemPort |
8. Chen J, Martin Bl and Brautignan DL. (1992). Science, 257, 1261–1264. | Article | PubMed | ISI | ChemPort |
9. Cohen P. (1991). Methods Enzymol., 201, 389–397. | PubMed |
10. Cohen P and Cohen PTW. (1989). J. Biol. Chem., 264, 21435–21438. | PubMed | ISI | ChemPort |
11. Cohen P, Klumpp S and Schelling DL. (1989). FEBS Lett., 254, 596–600.
12. Dumaz N and Meek DW. (1999). EMBO J., 18, 7002–7010. | Article | PubMed | ISI | ChemPort |
13. El-Deiry WS, Kern SE, Pietenpol JA, Kinzler KW and Vogelstein B. (1992). Nat. Genet., 1, 45–49. | Article | PubMed | ChemPort |
14. Elbi C, Misteli T and Hager GL. (2002). Mol. Biol. Cell, 13, 2001–2015. | PubMed |
15. Fiscella M, Zhang H, Fan S, Sakaguchi K, Shen S, Mercer WE, Vande Woude GF, O'Connor PM and Appella E. (1997). Proc. Natl. Acad. Sci. USA, 94, 6048–6058. | Article | PubMed | ChemPort |
16. Funk WD, Pak DT, Karas RH, Wright WE and Shay JW. (1992). Mol. Cell. Biol., 12, 2866–2871. | PubMed | ISI | ChemPort |
17. Gotz J, Probst A, Ehler E, Hemmings B and Kues W. (1998). Proc. Natl. Acad. Sci. USA, 95, 12370–12375. | Article | PubMed | ChemPort |
18. Guy GR, Cao X, Chua SP and Tan YH. (1992). J. Biol. Chem., 267, 1846–1852. | PubMed |
19. Harper JW, Adami GR, Wei N, Keyomarsi K and Elledge SJ. (1993). Cell, 75, 805–816. | Article | PubMed | ISI | ChemPort |
20. Juven T, Barak Y, Zauberman A, George DL and Oren M. (1993). Oncogene, 8, 3411–3416. | PubMed | ISI | ChemPort |
21. Kastan MB, Zhan Q, El-Deiry WS, Carrier F, Jacks T, Walsh W, V, Plunkett BS, Vogelstein B and Fornace AJJ. (1992). Cell, 71, 587–597. | Article | PubMed | ISI | ChemPort |
22. Kawabe T, Muslin AJ and Korsmeyer SJ. (1997). Nature, 385, 454–458. | Article | PubMed | ISI | ChemPort |
23. Kimura SH and Nojima H. (2002). Genes Cells, 7, 869–880. | PubMed |
24. Ko LJ and Prives C. (1996). Genes Dev., 10, 1054–1072. | Article | PubMed | ISI | ChemPort |
25. Lambert FP, Kashanchi F, Radonovitch MF, Shiekhattar R and Brady JN. (1998). J. Biol. Chem., 273, 33048–33053. | Article | PubMed | ISI | ChemPort |
26. Lees-Miller SP, Chen Y-R and Anderson CW. (1990). Mol. Cell. Biol., 10, 6472–6481. | PubMed | ChemPort |
27. Levine AJ. (1997). Cell, 88, 323–331. | Article | PubMed | ISI | ChemPort |
28. Li L, Ljungman M and Dixon JE. (2000). J. Biol. Chem., 275, 2410–2414. | Article | PubMed | ISI | ChemPort |
29. Miyashita T and Reed JC. (1995). Cell, 80, 293–299. | Article | PubMed | ISI | ChemPort |
30. Mumby M and Walter G. (1993). Physiol. Rev., 73, 673–699. | PubMed | ISI | ChemPort |
31. Okamoto K, Kamibayashi C, Serrano M, Prives C, Mumby MC and Beach D. (1996). Mol. Cell. Biol., 16, 6593–6602. | PubMed |
32. Okamoto K, Li H, Jensen MR, Zhang T, Taya Y, Thorgeirsson SS and Prives C. (2002). Mol. Cell, 9, 761–771. | Article | PubMed | ISI | ChemPort |
33. Pise-Masison CA, Mahieux R, Jiang H, Ashcroft M, Radonovich M, Duvall J, Guillerm C and Brady JN. (2000). Mol. Cell. Biol., 20, 3377–3386. | Article | PubMed | ISI | ChemPort |
34. Ruediger R, Van Wart Hood JE, Mumby M and Walter G. (1991). Mol. Cell. Biol., 11, 4282–4285. | PubMed |
35. Sakaguchi K, Herrera JE, Saito S, Miki T, Bustin M, Vassilev A, Anderson CW and Appella E. (1998). Genes Dev., 12, 2831–2841. | PubMed | ISI | ChemPort |
36. Scheidtmann KH, Mumby MC, Rundell K and Walter G. (1991). Mol. Cell. Biol., 11, 1996–2003. | PubMed | ISI | ChemPort |
37. Schonthal AH. (1998). Front. Biosci., 3, 1262–1273.
38. Shieh S-Y, Ikeda M, Taya Y and Prives C. (1997). Cell, 91, 325–334. | Article | PubMed | ISI | ChemPort |
39. Shieh S-Y, Taya Y and Prives C. (1999). EMBO J., 18, 1815–1823. | Article | PubMed | ISI | ChemPort |
40. Siliciano JD, Canman CE, Taya Y, Sakaguchi K, Appella E and Kastan MB. (1997). Genes Dev., 11, 3471–3481. | PubMed | ISI | ChemPort |
41. Takenaka I, Morin F, Seizinger BR and Kley N. (1995). J. Biol. Chem., 270, 5405–5411. | Article | PubMed | ISI | ChemPort |
42. Van Steensel B, van Binnendijk EP, Hornsby CD, Van der Voort HT, Krozowski Z, De Kloet ER and Van Driel R. (1996). J. Cell Sci., 109, 787–792. | PubMed | ChemPort |
43. Wera S and Hemmings BA. (1995). Biochem. J., 311, 17–29. | PubMed | ISI | ChemPort |
44. Wilson NJ, Moss ST, Csar XF, Ward AC and Hamilton JA. (1999). Biochem. J., 339, 517–524. | Article | PubMed |
45. Wu X, Bayle JH, Olson D and Levine AJ. (1993). Genes Dev., 7, 1126–1132. | Article | PubMed | ISI | ChemPort |
46. Yan Y, Shay JW, Wright WE and Mumby MC. (1997). J. Biol. Chem., 272, 15220–15226. | Article | PubMed | ISI | ChemPort |
47. Zuo Z, Dean N and Honkanen RE. (1998). J. Biol. Chem., 273, 12250–12258. | Article | PubMed | ISI | ChemPort |

Acknowledgements

We thank Michael Radonovich for preparation of -irradiated cells, Dr Tom Misteli for technical suggestions on image processing and Dr Cynthia Pise-Masison for helpful discussions. Microscopy images in this article were generated in the imaging facility of Tom Misteli and processed in the NIH, NCI Core Imaging Facility.