Protein Kinase C δ Regulates Ser46 Phosphorylation of p53 Tumor Suppressor in the Apoptotic Response to DNA Damage*

The p53 tumor suppressor is activated in the cellular response to genotoxic stress. Transactivation of p53 target genes dictates cell cycle arrest and DNA repair or induction of apoptosis; however, a molecular mechanism responsible for these distinct functions remains unclear. Recent studies revealed that phosphorylation of p53 on Ser46 was associated with induction of p53AIP1 expression, resulting in the commitment of the cell fate into apoptotic cell death. Moreover, upon exposure to genotoxic stress, p53DINP1 was expressed and recruited a kinase(s) to p53 that specifically phosphorylated Ser46. Here, we show that the pro-apoptotic kinase, protein kinase C δ (PKCδ), is involved in phosphorylation of p53 on Ser46. PKCδ-mediated phosphorylation is required for the interaction of PKCδ with p53. The results also demonstrate that p53DINP1 associates with PKCδ upon exposure to genotoxic agents. Consistent with these results, PKCδ potentiates p53-dependent apoptosis by Ser46 phosphorylation in response to genotoxic stress. These findings indicate that PKCδ regulates p53 to induce apoptotic cell death in the cellular response to DNA damage.

The cellular response to genotoxic stress includes cell cycle arrest, activation of DNA repair and, in the event of irreparable damage, induction of apoptosis. The signaling mechanisms responsible for regulation of the DNA damage response are largely unclear. Certain insights have been derived from the finding that the ␦ isoform of protein kinase C (PKC) 2 is activated in response to DNA damage. Notably, PKC␦ is cleaved to a 40-kDa catalytically active fragment by caspase-3 in cells treated with DNA-damaging agents (1,2). The finding that overexpression of the PKC␦ catalytic fragment (PKC␦CF) induces chromatin condensation and DNA fragmentation supports a role for PKC␦ cleavage in the induction of apoptotic cell death (3). Interaction of PKC␦CF with the nuclear DNA-dependent protein kinase catalytic subunit (DNA-PKcs) inhibits the function of DNA-PKcs to form complexes with DNA and to phosphorylate its downstream target, p53 (4). In addition, cells deficient in DNA-PK are resistant to apoptosis induced by PKC␦CF overexpression (4). Other studies have shown that PKC␦ interacts with the c-Abl tyrosine kinase upon exposure to genotoxic stress (5). c-Abl is a pro-apoptotic tyrosine kinase that targets to the nucleus following genotoxic stress (6 -8). Importantly, c-Abl-mediated phosphorylation activates PKC␦ and induces translocation of PKC␦ to the nucleus (5). In concert with these findings, tyrosine phosphorylation of PKC␦ is necessary for its nuclear translocation and subsequent caspase-dependent cleavage in the apoptotic response to DNA damage (9). A recent study demonstrated that nuclear-targeted PKC␦ interacts with and phosphorylates a pro-apoptotic molecule, Rad9 (10). PKC␦ regulates the interaction of Rad9 with Bcl-2 and the hRad9-mediated apoptotic response to DNA damage (10). Furthermore, previous studies showed that cells derivedfromPKC␦-nulltransgenicmiceweredefectiveinmitochondriadependent apoptosis (11). These findings collectively support a pivotal role for PKC␦ in the induction of apoptosis in response to DNA damage.
The p53 tumor suppressor functions in the cellular response to stress by inducing cell cycle arrest, DNA repair, senescence, differentiation, or apoptosis (12). Genotoxic stress is associated with stabilization of p53 and induction of p53-mediated transcription. Selective transactivation of p53 target genes dictates cell cycle arrest and DNA repair, or induction of apoptosis (13)(14)(15). However, the mechanism by which p53 determines the choice of cell fate is largely unknown. Available evidence suggests that promoter selectivity of p53 is regulated by its phosphorylation. For instance, Ser 15 and Ser 20 phosphorylation influences binding of p53 to promoters for cell cycle arrest and DNA repair genes. Further phosphorylation of Ser 46 following severe DNA damage increases the affinity of p53 for promoters of pro-apoptotic genes, such as p53AIP1 (16). In this regard, Ser 46 kinase(s) could function in p53-dependent apoptosis.
The present study demonstrates that PKC␦ associates with p53. This association is required for PKC␦-mediated phosphorylation of p53 on Ser 46 upon exposure to genotoxic stress. Furthermore, PKC␦ promotes p53-dependent apoptosis in response to DNA damage.
Plasmids-PKC␦ expression plasmids were described previously (17,18). p53 cDNA was amplified by PCR from human fetal brain cDNA library, then cloned into the pcDNA3-FLAG vector. The N-terminal region of p53 (amino acids 1-92) was cloned into the pGEX4T-1 vector (Amersham Biosciences). Various mutations were introduced by sitedirected mutagenesis and were confirmed by sequencing.
Cell Transfections-Cell transfections were performed as described (19). The total DNA concentration was kept constant by including an empty vector.
In Vitro Binding Assays-Cell lysates were incubated with purified GST, GST-PKC␦ regulatory domain (RD), or GST-PKC␦CF (17) in lysis buffer for 2 h at 4°C. The adsorbates were resolved by SDS-PAGE and analyzed by immunoblotting with anti-FLAG or anti-GST.
Assessment of Apoptosis-Apoptotic cells were detected by TUNEL assays using the DeadEnd Colorimetric TUNEL System (Promega).

RESULTS
PKC␦ Phosphorylates p53 on Ser 46 -To investigate whether PKC␦ is involved in phosphorylation of p53, MCF-7 cells were treated with the DNA-damaging agent adriamycin (ADR) in the presence or absence of the PKC␦-specific inhibitor, rottlerin (21). p53 was phosphorylated on Ser 15 and Ser 20 in response to ADR treatment regardless of PKC␦ activity (Fig. 1A). By contrast, phosphorylation on Ser 46 was diminished by pretreatment with rottlerin ( Fig. 1B). Moreover, consistent with previous results (22), inhibition of PKC␦ attenuated the expression level of p53 in relatively later periods following DNA damage (Fig. 1B). Similar results were obtained with U2-OS cells (data not shown). Comparable results were also observed when cells were treated with other DNAdamaging agents, such as etoposide and cisplatin (data not shown). To extend these findings using ectopically expressed p53, 293T cells were transfected with FLAG-tagged p53. Similar to endogenous p53, exogenous p53 was phosphorylated on Ser 46 following ADR treatment (Fig.  1C). In contrast, there was little if any phosphorylation of overexpressed p53 on Ser 46 in rottlerin-pretreated cells (Fig. 1C). These results indicate that PKC␦ is involved in phosphorylation of p53 on Ser 46 .
To further define the role for PKC␦ in Ser 46 phosphorylation, FLAG-p53 was co-transfected into 293T cells together with the Myc vector, Myc-PKC␦CF, or the Myc-PKC␦CF(K3 R) mutant, which is catalytically inactive (18). Expression of catalytically active PKC␦ was associ-ated with prominent phosphorylation of p53 on Ser 46 ( Fig. 2A, upper). Conversely, Ser 46 phosphorylation was completely abrogated by expression of the dominant negative PKC␦CF(K3 R) mutant. Moreover, the level of p53 expression paralleled that of Ser 46 phosphorylation ( Fig. 2A, FIGURE 1. PKC␦ is involved in phosphorylation of p53 on Ser 46 in response to DNA damage. A, MCF-7 cells were pretreated with or without rottlerin for 1 h followed by the treatment with ADR for the indicated times. Cell lysates were subjected to immunoblot analysis (IB) with the indicated antibodies. B, MCF-7 cells were treated as described above. Cell lysates were subjected to immunoprecipitation (IP) with anti-p53 followed by immunoblot analysis with anti-phospho-p53 (Ser 46 ) (upper panel) or anti-p53 (middle panel). Lysates were also subjected to immunoblotting with anti-tubulin (lower panel). C, 293T cells were transfected with FLAG-p53 and then treated with ADR in the presence or absence of rottlerin. Cell lysates were analyzed by immunoprecipitation (IP) with anti-FLAG followed by immunoblotting with anti-phospho-p53 (Ser 46 ) (upper panel) or anti-FLAG (middle panel). Lysates were also analyzed by immunoblotting with anti-tubulin (lower panel). lower). To determine if PKC␦ functions in DNA damage-induced Ser 46 phosphorylation, 293T cells were co-transfected with FLAG-p53 and GFP vector or GFP-PKC␦CF(K3 R). In GFP vector-transfected cells, Ser 46 phosphorylation was induced following ADR treatment (Fig. 2B). By contrast, co-expression with the GFP-PKC␦CF(K3 R) mutant completely abrogated Ser 46 phosphorylation upon exposure to ADR (Fig.  2B). To confirm whether PKC␦ is responsible for Ser 46 phosphorylation in response to DNA damage, we knocked down PKC␦ by transfection of cells with PKC␦ siRNAs. Down-regulation of PKC␦ was associated with attenuation of Ser 46 phosphorylation following genotoxic stress (Fig.  2C). To establish a direct role for PKC␦ in Ser 46 phosphorylation, kinase-active recombinant PKC␦ was incubated with ATP and GST, GST-p53-(1-92) wild type or the GST-p53-(1-92) mutant in which Ser 46 is substituted with Ala. The finding that purified PKC␦ phosphorylated GST-p53-(1-92), and not GST, indicates that PKC␦ is the Ser 46 kinase in vitro (Fig. 2D). Notably, co-incubation of PKC␦ with the GST-p53-(1-92) S46A mutant abrogated reactivity with anti-pSer 46 (Fig.  2D). These results collectively support the direct role for PKC␦ in phosphorylation of p53 on Ser 46 in response to DNA damage.
PKC␦ Interacts with p53-To examine whether PKC␦ associates with p53, MOLT-4 cells, which highly express p53, were treated with ADR. Lysates were immunoprecipitated with anti-PKC␦ or, as a control, IgG. Immunoblot analysis of the precipitates with anti-p53 revealed that PKC␦ associates with p53 (Fig. 3A). Moreover, treatment with ADR was associated with increased formation of PKC␦-p53 complexes (Fig. 3A). In the reciprocal experiment, immunoblot analysis of anti-p53 immunoprecipitates with anti-PKC␦ demonstrated a low, but substantial level of interaction between PKC␦ and p53 (Fig. 3B). Importantly, the finding that PKC␦CF was detectable in anti-p53 immunoprecipitates from ADR-treated MOLT-4 cells indicated that the catalytic fragment of PKC␦ is responsible for binding to p53 (Fig. 3B, upper). Similar results were obtained with U2-OS cells (data not shown). To identify the region of PKC␦ that associates with p53, lysates from 293T cells transfected with FLAG-p53 were incubated with purified GST, GST-PKC␦CF or GST-PKC␦ regulatory domain (RD). Analysis of precipitates with anti-FLAG showed the binding of p53 to PKC␦CF, but not to PKC␦RD (Fig.  3C). These findings indicate that the catalytic domain of PKC␦ is required for binding to p53.

PKC␦ Forms Complexes with p53DINP1 in Response to Genotoxic
Stress-A previous study demonstrated that expression of the p53-inducible gene, p53DINP1, by DNA damage enhances Ser 46 phosphorylation and leads to apoptotic cell death (23). Notably, p53DINP1 recruits an unknown kinase(s) responsible for Ser 46 phosphorylation to p53 (23). These data indicate that p53DINP1 interacts with a Ser 46 kinase(s) upon exposure to genotoxic stress. To examine the possibility that PKC␦ associates with p53DINP1, 293T cells were co-transfected with GFP-p53DINP1 and FLAG vector, FLAG-PKC␦ full-length (FL), FLAG-PKC␦CF, or FLAG-PKC␦RD. Immunoblot analysis of anti-FLAG immunoprecipitates with anti-GFP demonstrated that FLAG-PKC␦FL, and not CF or RD, was associated with GFP-p53DINP1 (Fig. 4A). To extend these findings to endogenous proteins, MOLT-4 cells were left untreated or treated with ADR. Cell lysates were immunoprecipitated with anti-p53DINP1 followed by immunoblotting with anti-PKC␦. The results showed the interaction of p53DINP1 with PKC␦ (Fig. 4B). Moreover, treatment with ADR was associated with increased formation of  p53DINP1-PKC␦ complexes (Fig. 4B). Similar results were obtained with U2-OS cells (data not shown). Comparable results were also observed when cells were treated with etoposide (data not shown). These findings demonstrate an inducible binding of PKC␦ with p53DINP1 following genotoxic stress.
Whereas PKC␦ induces p53 and p53DINP1 expression is regulated by p53, it is conceivable that PKC␦ modulates the expression level of p53DINP1 in response to DNA damage. To examine this hypothesis, MCF-7 cells were treated with ADR in the presence or absence of rottlerin. The expression level of p53DINP1 was enhanced by ADR treatment (Fig. 4C). By contrast, induction of p53DINP1 expression was attenuated by pretreatment of cells with rottlerin (Fig. 4C). Similar results were obtained with U2-OS cells (data not shown). To confirm PKC␦-mediated regulation of p53DINP1, 293T cells were co-transfected with FLAG-p53 and GFP vector or GFP-PKC␦CF(K3 R). Immunoblot analysis with anti-p53DINP1 revealed that expression of the dominant negative PKC␦ mutant inhibited ADR-induced up-regulation of p53DINP1 (Fig. 2B). These findings support the mechanism by which PKC␦ regulates p53DINP1 expression through p53 activation.

PKC␦ Potentiates p53-dependent Apoptosis in Response to DNA
Damage-Previous studies have shown that activation of PKC␦ following genotoxic stress is associated with the execution of apoptosis (3,4,10); however, this mechanism is largely unknown. Importantly, the present study demonstrates that PKC␦ phosphorylates p53 on Ser 46 , which is responsible for the induction of apoptosis. In this regard, PKC␦ could be involved in p53-dependent apoptosis following genotoxic stress. To address this issue, U2-OS cells were pretreated with rottlerin followed by the treatment with etoposide for 24 h. Treatment of cells with etoposide increased induction of apoptosis (Fig. 5A). In contrast, pretreatment with rottlerin substantially attenuated etoposide-induced apoptosis (Fig. 5A). Similar findings were obtained with MCF-7 cells (data not shown). Comparable results were also observed when cells were treated with ADR (data not shown). To extend these findings, PKC␦ was knocked down in U2-OS cells by transfection with PKC␦ siRNAs. Knocking down PKC␦ attenuated the induction of apoptosis elicited by etoposide treatment (Fig. 5B). These results suggest that etoposide-induced apoptosis is, at least in part, a PKC␦-dependent mechanism. To further define the role for PKC␦ in p53-dependent apo- ptosis, p53-deficient HCT116 cells (HCT116/p53 Ϫ/Ϫ ) (24) were transfected with FLAG vector, FLAG-p53 wild type (wt), or the FLAG-p53 S46A mutant, in which Ser 46 was replaced with Ala. Cells were then treated with etoposide for 24 h in the presence or absence of rottlerin. TUNEL assays demonstrated that treatment of vector-expressing cells with etoposide induced apoptotic cell death (Fig. 5, C and D). Furthermore, etoposide-induced apoptosis was enhanced by ectopic expression of p53 wt, and not the p53 S46A mutant (Fig. 5, C and D). Importantly, pretreatment with rottlerin was substantially attenuated etoposide-induced apoptosis regardless of p53 expression (Fig. 5, C and D). These findings provide support for the involvement of PKC␦ phosphorylation of p53 on Ser 46 in the apoptotic response to DNA damage.

DISCUSSION
The mechanisms by which genotoxic stress is converted into intracellular signals that control cellular fate are largely unknown. Certain insights have been derived from the findings that PKC␦ is activated in the response of cells to agents that arrest DNA replication or induce DNA lesions (1,2). The available evidence indicates that full-length PKC␦ is activated as an early event within 1 h of exposure to genotoxic agents (5,17). Phosphorylation of PKC␦ on tyrosine is a mechanism for PKC␦ activation by DNA-damaging agents (5,9). In this context, activation of PKC␦ is induced, at least in part, by c-Abl-dependent phosphorylation (5). Previous studies have also shown that treatment of cells with DNA-damaging agents is associated with translocation of PKC␦ to the nucleus (10,25). Inhibition of PKC␦ kinase activity attenuates nuclear targeting of PKC␦. Whereas nuclear PKC␦ associates with DNA-PKcs and hRad9 (4,10), the nuclear targets of PKC␦ are otherwise largely unknown. The present findings demonstrate that nuclear PKC␦ also interacts with p53. Binding of PKC␦ to p53 was detectable constitutively and increased in response to DNA damage. In this context, nuclear targeting of PKC␦ and induction of p53 expression following genotoxic stress were both associated with increases in the formation of PKC␦-p53 complexes. Significantly, the present studies demonstrate that the catalytic fragment of PKC␦ is responsible for binding to p53. Previous studies have shown that PKC␦ is activated as a later event in the genotoxic stress by caspase-3-mediated cleavage (1,2,26). The cleaved C-terminal 40-kDa fragment contains the ATP binding and kinase domains, which is constitutively active. Interestingly, a recent study demonstrated that PKC␦ activates caspase-3 by its phosphorylation (27). Thus, caspase-3 could be activated by PKC␦-dependent phosphorylation, then cleaved PKC␦CF fragment by activated caspase-3 might potentiate interaction of PKC␦ with p53. In addition to the interaction, we found that PKC␦ was associated with induction of p53 expression in later periods following DNA damage. In concert with these results, a previous study showed that inhibition of PKC␦ activity attenuated basal level of p53 expression (22). By contrast, another study demonstrated that treatment of HeLa cells with rottlerin increased cisplatin-mediated p53 level (28). Obviously, further studies will be needed to clarify a mechanism by which PKC␦ regulates p53 expression.
In contrast to many p53 phosphorylation sites, Ser 46 is phosphorylated in a later period following genotoxic stress (16). Moreover, this phosphorylation is required for expression of p53AIP1 that functions in induction of apoptosis (16). These findings thus suggest that Ser 46 phosphorylation is essential for p53-dependent apoptosis in response to DNA damage. However, little is known about Ser 46 kinase(s). Available lines of evidence revealed that HIPK2 phosphorylated Ser 46 upon exposure to UV, but not ionizing radiation (29,30). p38 MAPK was also reported as a Ser 46 kinase; however, this phenomenon is controversial (16,(31)(32)(33). In this context, the present studies demonstrate that PKC␦ is a novel candidate for Ser 46 kinase upon exposure to genotoxic stress. The evidence that PKC␦ is fully activated after cleavage by caspase-3 might be associated with a delayed phosphorylation of p53 on Ser 46 . Moreover, the present findings that the catalytic fragment of PKC␦ preferably associates with p53 further support a role for PKC␦ in Ser 46 phosphorylation. We also show that PKC␦ associates with p53DINP1 following DNA damage. A previous study demonstrated that p53DINP1 recruits kinase(s) responsible for Ser 46 phosphorylation to p53. In this regard, inducible association of PKC␦ with p53DINP1 provided a further support for the involvement of PKC␦ in Ser 46 phosphorylation. Importantly, coimmunoprecipitation studies indicate that fragments of PKC␦ such as PKC␦CF or RD was insufficient for binding to p53DINP1. A potential explanation is that the binding domain extends to both regions of PKC␦. Another possibility is that specific tertiary structure is required for the binding. Whereas PKC␦CF can interact with p53, this interaction might be independent of the recruitment by p53DINP1. However, given the recent finding that active caspase-3 translocates to the nucleus after induction of apoptosis (34), it is conceivable that PKC␦ is cleaved in the nucleus after recruitment of PKC␦ by p53DINP1 to p53.
Upon exposure to genotoxic stress, p53 functions in both cell cycle arrest and induction of apoptosis. However, the mechanism by which p53 determines these distinct outcomes is largely unknown. A previous study suggested that, for repairable DNA damage, p53 is phosphorylated on Ser 15 and Ser 20 and induces G 1 arrest genes, such as p21. If DNA damage is severe and irreparable, Ser 46 is phosphorylated, which then triggers induction of pro-apoptotic genes, such as p53AIP1 (16). Thus, Ser 46 kinase(s) are activated in response to DNA damage and induce apoptosis, at least in a p53-dependent mechanism. In this context, the present results demonstrate that PKC␦ phosphorylates p53 on Ser 46 and induces apoptosis, at least in part, in a Ser 46 phosphorylation-dependent manner. Moreover, PKC␦ induces p53 expression in later periods following DNA damage. These findings thus support a model in which activation of PKC␦ by genotoxic stress induces phosphorylation of p53 on Ser 46 , resulting in the commitment of cell fate to apoptosis.