Identification of Serine 643 of Protein Kinase C-δ as an Important Autophosphorylation Site for Its Enzymatic Activity*

To investigate the role of serine/threonine autophosphorylation of protein kinase C-δ (PKC-δ), we mutated serine 643 of PKC-δ to an alanine residue (PKC-δS643A). Two different expression vectors containing PKC-δS643A mutant cDNAs were transfected and expressed in 32D myeloid progenitor cells.In vitro autophosphorylation assays demonstrated 65–83% reduction in autophosphorylation of PKC-δS643A in comparison to wild type PKC-δ (PKC-δWT). The enzymatic activity of PKC-δS643A mutant as measured by phosphorylating the PKC-δ pseudosubstrate region-derived substrate was also reduced more than 70% in comparison to that of PKC-δWT. In vivo labeling and subsequent two-dimensional phosphopeptide analysis demonstrated that at least one phosphopeptide was absent in PKC-δS643A when compared with PKC-δWT, further substantiating that serine 643 is phosphorylated in vivo. Localization and 12-O-tetradecanoylphorbol-13-acetate-dependent translocation and tyrosine phosphorylation of PKC-δS643A were not altered in comparison to PKC-δWT, indicating that mutagenesis did not affect the structural integrity of the mutant protein. 12-O-Tetradecanoylphorbol-13-acetate-mediated monocytic differentiation of 32D cells overexpressing PKC-δS643A mutant protein was impaired in comparison to that of PKC-δWT transfectant. Taken together, our results demonstrate that serine 643 of PKC-δ is a major autophosphorylation site, and phosphorylation of this site plays an important role in controlling its enzymatic activity and biological function.

Protein kinase C (PKC) 1 is composed of a family of serine/ threonine kinases. To date, 11 different PKC isoenzymes have been identified that are divided into three different subgroups, conventional PKCs (cPKCs), novel PKCs (nPKCs), and atypical PKCs (1)(2)(3). PKCs have been defined as important signaling molecules in cell growth, differentiation, secretion of hormones and neurotransmitters, and cellular transformation (2). PKC-␦ belongs to nPKC subgroup and is ubiquitously expressed in many tissues and cell lines (4).
We have focused our efforts on understanding the role of PKC-␦ in various signaling transduction pathways. Overexpression of wild type of PKC-␦ (PKC-␦WT) in 32D myeloid progenitor cells led to monocytic differentiation in response to 12-O-tetradecanoylphorbol-13-acetate (TPA) treatment (5), suggesting a causal role for PKC-␦ in hematopoietic cell differentiation. An ATP binding mutant of PKC-␦ (PKC-␦K376R) was generated by site-directed mutagenesis and was demonstrated to lack autophosphorylation capacity in vitro completely (6). Moreover, the PKC-␦K376R mutant competitively inhibited PKC-␦WT phosphorylation of an exogenous substrate in vitro. Recently, our group and several others (7)(8)(9)(10)(11)(12) observed tyrosine phosphorylation of PKC-␦ in vivo in response to its activation by various agonists. PKC-␦ was also demonstrated to be an important substrate in the platelet-derived growth factor ␤ receptor (PDGF-␤R) pathway (13). It was phosphorylated by the activated PDGF-␤R in vivo and in vitro on tyrosine residue(s) (10,13). The relevance of PKC-␦ in mediating c-sis/ PDGF-B transformation of NIH 3T3 cells was recently elucidated (14). In this study, expression of the PKC-␦K376R mutant led to dramatic inhibition of c-sis-induced NIH 3T3 cell transformation. These results demonstrate that PKC-␦ plays a physiological role in a signaling pathway leading to malignant transformation of fibroblasts induced by sis oncogene.
Serine/threonine phosphorylation of PKC in vivo was first observed approximately 10 years ago (15)(16)(17)(18)(19). Several in vivo phosphorylation sites have been mapped utilizing different methods (20 -22). Based on studies performed on cPKCs (20,(23)(24)(25)(26)(27), it is generally believed that PKC is first synthesized as an immature precursor protein that does not show any catalytic activity. Phosphorylation of PKC on the "activation loop," which corresponds to threonines 497 and 500 of PKC-␣ (23) and ␤II (26), respectively, by an unidentified PKC kinase then renders PKC catalytic domain competent. However, transphosphorylation of PKC on its activation loop does not alter the mobility of the protein as observed by SDS-polyacrylamide gel electrophoresis (PAGE). Subsequent autophosphorylation on threonine 641 of PKC-␤II results in the first upward shift of the mobility of the protein. This event is followed by a second autophosphorylation on serine 660 of PKC-␤II which further shifts the protein to the mature 80-kDa form. Generation of diacylglycerol through different mechanisms recruits PKC to the membrane where the pseudosubstrate region-mediated autoinhibition of the catalytic domain is released. The enzyme is then able to phosphorylate substrates and transmit the downstream signals. How the mature enzyme returns to the cytosol after activation remains unclear. This may be regulated by serine/threonine phosphatase activity (1).
Autophosphorylation of PKC has been observed both in vivo and in vitro (15)(16)(17)(18)(19). It is thought that autophosphorylation of * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 1 The abbreviations used are: PKC, protein kinase C; WT, wild type; TPA, 12-O-tetradecanoylphorbol-13-acetate; HA, hemagglutinin; PDGF-␤R, platelet-derived growth factor ␤ receptor; PAGE, polyacrylamide gel electrophoresis; anti-Tyr(P), anti-phosphotyrosine; mAb, monoclonal antibody. PKC enhances its binding to phorbol ester and reduces the K m for its substrates in vitro (16,18). Several in vivo autophosphorylation sites for different PKC isoenzymes have been mapped (20 -22). Recently, conserved threonine autophosphorylation sites on two cPKCs (PKC-␣ and PKC-␤I) were characterized by site-directed mutagenesis (23,28,29). Mutation of threonine 638 to alanine in the PKC-␣ molecule did not dramatically affect its enzymatic activity (23). In striking contrast, mutation of this conserved site (threonine 642 to alanine) in PKC-␤I completely abolished its enzymatic activity and in vivo phosphorylation (29). Since PKC-␦ belongs to the nPKC subfamily and a serine residue rather than a threonine residue exists at this conserved position (see Fig. 1), we have attempted to elucidate whether PKC-␦ is phosphorylated on this conserved site and, if so, whether this phosphorylation would influence PKC-␦ function. Our results indicate that serine 643 is a major PKC-␦ autophosphorylation site, and phosphorylation of this site significantly affects its enzymatic activity.

Construction of a Serine to Alanine Mutant of Murine PKC-␦, cDNA Expression Vectors, and Cell Lines-The
Bio-Rad Muta-gene Phagemid in vitro mutagenesis kit (version 2) was used for the site-directed mutagenesis. The oligonucleotide 5Ј-GAATGAGAAACCTCAGCTTG-CATTCAG-3Ј was used as a mutant primer in the in vitro mutagenesis reaction where the serine residue at amino acid 643 of murine PKC-␦ was changed to alanine (underlined in the sequence). The successful mutation of this site generated a new BsmI restriction site that was used to screen all the reaction products. The mutation was confirmed by DNA sequencing. The PKC-␦S643A mutant cDNA was subcloned into pCEV-HA (three hemagglutinin epitope repeats, neo selection) and pLTR (two HA epitope repeats, gpt selection) vectors, generating pCEV-␦S643A-HA and pLTR-␦S643A-HA, respectively. The generation of these two vectors and subcloning of PKC-␦WT cDNA into these vectors have been previously described (6,30). The 32D cells were transfected with different cDNA expression vectors using the electroporation procedure described previously (5). 32D cells and transfectants were cultured in RPMI 1640 medium with 10% fetal calf serum and 5% WEHI-3B conditioned medium as a source of murine interleukin-3.
Immunoprecipitation, Immunoblot Analysis, and Subcellular Fractionation-These procedures have been described previously (6,10,13,30). Briefly, the 32D transfectants were serum-starved for 2 h and left untreated or stimulated with 100 ng/ml TPA (Sigma) for 10 min. The cell pellets were lysed in Triton X-100 containing lysis buffer (13) and clarified by centrifugation. For immunoprecipitation, equal amounts of proteins (1-5 mg per sample) were incubated with polyclonal anti-PKC-␦ serum (5 l per sample, Calbiochem) together with 40 l of protein G-coupled Sepharose (Pharmacia Biotech, Inc.) or with anti-HA monoclonal antibody (mAb; 4 g per sample, Boehringer Mannheim) together with 25 l of protein A-Sepharose beads (Pierce). Anti-phosphotyrosine (anti-Tyr(P), 2 g/ml, Upstate Biotechnology) and anti-PKC-␦ (1:1000) were utilized for immunoblot analysis. The enhanced chemiluminescence system (Amersham Corp.) was used to visualize proteins, and the densities of the bands from SDS-PAGE and autoradiography were quantified by using a densitometer (Molecular Dynamics). The method for the subcellular fractionation has been described before (6,13).
In Vitro PKC-␦ Autophosphorylation Assay-The in vitro autophosphorylation assay utilizing anti-HA antibody for immunoprecipitation was performed by following a previously described protocol (6). Briefly, cell lysates were immunoprecipitated with anti-HA antibody as described above. Washed immunoprecipitates were incubated on ice for 30 min with 50 l of autophosphorylation buffer that contained 20 mM Tris-HCl, pH 7.5, 5 mM magnesium acetate, 50 g/ml phosphatidylserine (Sigma), 100 ng/ml TPA, 10 g/ml leupeptin, 1 mM Na 3 VO 4 , 1 M ATP (Boehringer Mannheim), and 5 Ci of [␥-32 P]ATP (3000 Ci/mmol, Amersham Corp.). The reaction was stopped by washing twice with Triton X-100 containing lysis buffer, and denatured proteins were separated by SDS-PAGE. The dried gel was autoradiographed.
In Vivo Labeling and Two-dimensional Phosphopeptide Analysis-Both in vivo labeling and subsequent two-dimensional phosphopeptide analysis have been described previously (30). Briefly, serum-starved 32D transfectants were labeled with [ 32 P]orthophosphate (1 mCi/ml; NEN Life Science Products) for 3 h and were stimulated with TPA (100 ng/ml) for 10 min. Cell lysates were immunoprecipitated with anti-HA mAb, and immunoprecipitates were resolved by SDS-PAGE. Radiolabeled PKC-␦WT-HA and PKC-␦S643A-HA bands were excised from the gel and exhaustively digested with trypsin (tosylphenylalanyl chloromethyl ketone-treated). The resulting phosphopeptides were resolved by thin layer electrophoresis, pH 8.9, followed by ascending chromatography, pH 1.9. Dried plates were autoradiographed for 1 week.
In Vitro PKC-␦ Activity Assays-DE52 ion exchange chromatography to enrich PKC from the cell lysates and the subsequent measurement of PKC activity utilizing PKC-␦ pseudosubstrate region-derived peptide as a substrate have been described previously (6,13,30). Direct measurement of PKC-␦ activity on PKC-␦ substrate utilizing anti-HA immunoprecipitates as the kinase sources was also employed. Briefly, the equal amounts of protein (6 mg per sample) from the various PKC-␦ transfectants were immunoprecipitated with anti-HA antibody (4 g per sample). Washed immunoprecipitates were incubated at room temperature with 40 l of reaction buffer that contained 10 M PKC-␦ substrate derived from PKC-␦ pseudosubstrate region (6), 20 mM Tris-HCl, pH 7.5, 1 mM CaCl 2 , 10 M magnesium acetate, 1 M TPA, 50 g/ml phosphatidylserine (Sigma), 30 M ATP, and 30 Ci of [␥-32 P]ATP for 20 min. The reaction tube was centrifuged, and 20 l of the supernatant was spotted on phosphocellulase disk sheets (Life Technologies, Inc.). The sheets were washed twice with 1% phosphoric acid and twice with distilled water, and samples were analyzed by liquid scintillation. The nonspecific catalytic activity was measured in the same reaction buffer except that TPA and phosphatidylserine were omitted from the reaction. The specific PKC-␦ activity was calculated by subtracting the nonspecific catalytic activity from the total catalytic activity and expressed as counts per min (cpm).

FIG. 1. Serine 643 of PKC-␦ is a conserved phosphorylation site in other PKC isoenzymes.
Sequence alignment is based on the previous report described by Keranen et al. (20). In vivo phosphorylation of PKC-␣, -␤I, and -␤II at the threonine sites 638, 642, and 641, respectively, has been demonstrated (20,23,29). These sites correspond to serine 643 in PKC-␦. The conserved residues are shown as bold letters.
FIG. 2. The PKC-␦S643A mutant protein is expressed in the various 32D transfectants. Equal amounts of cell lysates (4 mg per sample) from 32D cells and transfectants were immunoprecipitated (IP) with anti-PKC-␦ serum. Washed immunoprecipitates were subjected to SDS-PAGE, and proteins transferred to an Immobilon membrane were immunoblotted (Blot) with anti-HA mAb.

Mutation of PKC-␦ Serine 643 and Expression of This Mutant in 32D
Cells-In an attempt to define which amino acids within PKC-␦ are autophosphorylation sites and determine whether mutation of one of these sites would affect PKC-␦ enzymatic activity, we chose to mutate serine 643 to alanine by sitedirected mutagenesis. This putative autophosphorylation site is conserved in other PKC sequences, including PKC-␣, PKC-␤I, and PKC-␤II (Fig. 1). In vivo phosphopeptide mapping or site-directed mutagenesis of the corresponding threonine sites within PKC-␣, PKC-␤I, and PKC-␤II revealed that these residues were all phosphorylated in vivo (20 -23, 28, 29). The mutant cDNA, designated PKC-␦S643A, was inserted into the pCEV-HA (3 ϫ HA repeats) vector, generating pCEV-␦S643A-HA, or into pLTR-HA vector (2 ϫ HA repeats), generating pLTR-␦S643A-HA. PKC-␦WT cDNA was previously inserted into these same vectors and designated pCEV-␦WT-HA and pLTR-␦WT-HA, respectively (30).
32D cells were transfected with expression vectors containing the various cDNA constructs, and drug-resistant 32D transfectants were subjected to immunoprecipitation and immunoblot analysis to detect PKC-␦S643A and PKC-␦WT expression. As shown in Fig. 2, immunoprecipitation with anti-PKC-␦ serum followed by immunoblot analysis with the anti-HA mAb detected both pLTR-␦WT-HA and pCEV-␦WT-HA proteins with mobilities of 80 and 90 kDa, respectively. The mobilities of PKC-␦WT proteins expressed in these two vectors were identical to those reported in our previous study (30). Endogenous PKC-␦ expression in 32D cells was not detected, since the anti-HA mAb was utilized for immunoprecipitation. The levels of PKC-␦S643A expression in cells transfected with pLTR-HA and pCEV-HA vectors were 2.8-and 1.8-fold higher than those of PKC-␦WT in the corresponding vectors, respectively (Fig. 2).
Autophosphorylation of the PKC-␦S643A Mutant Is Reduced in Comparison to That of PKC-␦WT-We performed in vitro autophosphorylation assays utilizing the anti-HA mAb for immunoprecipitation. As shown in Fig. 3A, autophosphorylation of pLTR-␦S643A-HA protein was reduced by 54% when compared with that of the pLTR-␦WT-HA molecule. Autophosphorylation of the pCEV-␦S643A-HA protein was decreased by 37% when compared with that of pCEV-␦WT-HA (Fig. 3B). Autophosphorylation of endogenous PKC-␦ from parental 32D cells was not detected since the anti-HA mAb would not recognize endogenous PKC-␦. By normalizing protein expression levels of PKC-␦S643A in comparison to those of PKC-␦WT in the various transfectants (see Fig. 2), an 83% reduction in pLTR-␦S643A-HA autophosphorylation and a 65% reduction in pCEV-␦S643A-HA autophosphorylation were observed. These results strongly suggest that serine 643 of PKC-␦ is a major autophosphorylation site, and mutation of this site dramatically reduces autophosphorylation.
Comparison of Tryptic Phosphopeptides Generated from PKC-␦S643A and PKC-␦WT by Two-dimensional Phosphopeptide Analysis-To confirm that serine 643 is an in vivo phosphorylation site, two-dimensional tryptic phosphopeptide analysis was performed. As shown in Fig. 4A, tryptic digestion of in vivo labeled PKC-␦WT-HA from the TPA-treated transfectant resulted in the detection of approximately 20 distinct phosphopeptides. The phosphopeptide pattern generated from PKC-␦WT-HA is consistent to that generated in a previous study (30), assuring that this assay is very reproducible. Although most of PKC-␦WT-HA phosphopeptides were also detected from tryptic digestion of in vivo labeled PKC-␦S643A-HA after TPA treatment of 32D/pLTR-␦S643A-HA transfectant, two phosphopeptides (peptides 5 and 14) were absent from PKC-␦S643A-HA sample. The reduced intensity of peptide 5 in a mixture experiment, where equal amounts of PKC-␦WT-HA and PKC-␦S643A-HA samples were mixed before performing two-dimensional phosphopeptide analysis, confirmed that pep- tide 5 was missing in PKC-␦S643A-HA (compare peptide 5 in Fig. 4, A and C). Since the PKC-␦WT-HA sample migrated slightly slower than the others in chromatography, only a tail of peptide 14 can be observed (Fig. 4A). This peptide was not detected in PKC-␦S643A-HA sample (Fig. 4B). Therefore, whether the intensity of peptide 14 detected in the mixture experiment was reduced (Fig. 4C) is difficult to judge. In addition, the intensity of peptide 11 was greatly reduced in the PKC-␦S643A-HA sample when compared with PKC-␦WT-HA, and intermediate intensity was observed in the mixture experiment (Fig. 4C). On the other hand, phosphopeptide 19 may be absent in PKC-␦WT-HA. Taken together, the results of twodimensional phosphopeptide analysis clearly indicate that the absence or reduction in intensity of phosphopeptides 5, 14, and 11 may account for the reduced autophosphorylation of PKC-␦S643A in vitro (see Fig. 3).
Enzymatic Activity of PKC-␦S643A Mutant Is Greatly Decreased in Comparison to That of PKC-␦WT-The enzymatic activity of PKC-␦S643A expressed in pLTR-HA system was measured utilizing two separate procedures. In the first assay, the activities were measured utilizing anti-HA immunoprecipitates as the kinase sources. This method has been recently used in other PKC studies to measure PKC activity (8,31). As shown in Table I, the immunoprecipitates derived from pLTR-␦WT-HA and pLTR-␦S643A-HA mutant transfectants displayed similar nonspecific catalytic activities when they were incubated with the PKC-␦ pseudosubstrate region-derived peptide in the absence of TPA and phosphatidylserine, two important cofactors required for specific PKC activation in vitro. However, the specific PKC-␦ catalytic activity of pLTR-␦S643A-HA mutant was reduced by 54% when compared with that of pLTR-␦WT-HA protein.
In another PKC activity assay, DE52 ion exchange chromatography was utilized to enrich PKC-␦ proteins before performing the kinase assay (6,30). pLTR-␦WT-HA overexpression resulted in a 14-fold increase in the enzymatic activity com-pared with that of endogenous PKC-␦ (Table II). The increased activity observed in the pLTR-␦WT-HA transfectant correlated with the levels of overexpressed PKC-␦ protein (data not shown). Expression of pLTR-␦S643A-HA reduced its specific catalytic activity by 39% compared with that of pLTR-␦WT-HA. By normalizing the protein expression level of pLTR-␦S643A-HA in comparison to that of pLTR-␦WT-HA, a 78 -84% reduction in pLTR-␦S643A-HA enzymatic activity was calculated from the results of these two assays (see Fig. 2). In summary, these results indicate that PKC-␦ serine 643 is not only important for autophosphorylation but also for transphosphorylation of its in vitro substrate.
The PKC-␦S643A Mutant Protein Is Not Thermal Labile-Recent work on the PKC-␣T638A mutant suggested that mutation of threonine 638 rendered the enzyme very sensitive to heat treatment (23). Thus, we were interested in determining whether there were any changes in the heat sensitivity of PKC-␦S643A mutant in comparison to PKC-␦WT. As shown in Fig. 5, pLTR-␦WT-HA, pLTR-␦S643A-HA, and endogenous PKC activities remained very stable even after a 30-min period of preincubation at 25°C. Surprisingly, both pLTR-␦WT-HA and pLTR-␦S643A-HA mutant activities were slightly increased after the preincubation period. This result suggests that phosphorylation of PKC-␦ on serine 643 does not affect the heat stability of the enzyme, even though the enzymatic activity and autophosphorylation of PKC-␦S643A are greatly reduced in comparison to PKC-␦WT.

Localization, Translocation, and Tyrosine Phosphorylation of PKC-␦ Are Not Altered When Serine 643 Autophosphorylation
Is Abolished-PKC-␦ normally resides in the cytosol (S100) of the cell. In response to stimulation by TPA, a portion translocates to the membrane fraction (P100) (13). Our previous data demonstrated that PKC-␦ was tyrosine-phosphorylated in vivo in response to TPA stimulation, and tyrosine-phosphorylated PKC-␦ could be detected only in the membrane fraction (6,10,13,30). Thus, we investigated whether mutation of serine 643 would affect localization, translocation, or tyrosine phosphorylation of the enzyme. As shown in Fig. 6A, the pLTR-␦S643A-HA mutant as well as pLTR-␦WT-HA proteins resided in the cytosol in resting cells after cell fractionation and immunoprecipitation with anti-HA mAb followed by anti-PKC-␦ immunoblot analysis. Stimulation with TPA for 10 min caused translocation of a similar portion of both pLTR-␦WT-HA and pLTR-␦S643A-HA mutant proteins to the membrane fraction (Fig. 6A, lanes 4 and 6). Reblotting the membrane with anti-Tyr(P) mAb showed that both pLTR-␦WT-HA and pLTR-␦S643A-HA mutant proteins were tyrosine-phosphorylated in TPA-stimulated samples (Fig. 6B, lanes 4 and 6). As previously demonstrated (6,13), tyrosine phosphorylation was observed only in the membrane fraction. Taken together, the results indicate that autophosphorylation of PKC-␦ on serine 643 does not affect localization, translocation, or tyrosine phosphorylation of the enzyme. reduced enzymatic activity as measured by utilizing anti-HA immunoprecipitates as PKC-␦ kinase sources The method for measuring PKC-␦ activity by using anti-HA immunoprecipitates as PKC-␦ sources was described under "Experimental Procedures." Only one sample from each lysate was utilized for anti-HA immunoprecipitation and the subsequent activity assay. Thus, no standard deviation was available. PKC-␦ specific activity was obtained by subtracting nonspecific activity from total catalytic activity. The activity is presented as cpm.

TABLE II
The PKC-␦S643A mutant protein expressed in 32D cells possesses reduced enzymatic activity as measured by utilizing DE52 column eluates as PKC-␦ kinase sources The method for PKC enrichment by DE52 ion exchange chromatography and the subsequent activity assay has been described previously (6,30). PKC-␦ specific activity was obtained by subtracting nonspecific activity from total catalytic activity. The results represent the mean value of three individual samples. The activity is presented as cpm. ated by the PKC-␦S643A Mutant Transfectant in Comparison to the PKC-␦WT Transfectant Is Impaired-TPA treatment of 32D cells overexpressing PKC-␦WT was able to mediate monocytic differentiation, as judged by changes in morphology, cell adhesion, nonspecific esterase staining, and cell surface differentiation marker expression (5,6). Since mutation of PKC-␦ on serine 643 reduced its autophosphorylation and its enzymatic activity, we tested whether PKC-␦-mediated monocytic differentiation would be affected. Treatment of the pLTR-␦S643A-HA mutant transfectant with TPA overnight resulted in reduced cell adhesion and less morphological changes indicative of the macrophage phenotype as analyzed by Wright-Giemsa staining when compared with the pLTR-␦WT-HA transfectant (data not shown). Flow cytometric analysis was utilized to detect cell surface differentiation marker expression. As seen in Fig. 7, stimulation of pLTR-␦WT-HA transfectant with TPA overnight resulted in increased expression of Mac-1 (Fig. 7A) and Fc␥II/IIIR (Fig. 7B). TPA treatment of the pLTR-␦S643A-HA mutant transfectant resulted in reduced increases in marker expression in comparison to the pLTR-␦WT-HA transfectant (Fig. 7, A and B). However, the TPA-induced increase in marker expression observed for the pLTR-␦S643A-HA mutant transfectant was still greater than that for the parental 32D cells, indicating that the remaining kinase activity provided by the pLTR-␦S643A-HA mutant was able to partially mediate the differentiation process. These results suggest that serine autophosphorylation on amino acid 643 plays an important role in PKC-␦-mediated monocytic differentiation of 32D myeloid progenitor cells. DISCUSSION In the present study, we have demonstrated that serine 643 of PKC-␦ is a major autophosphorylation site in vitro and autophosphorylation of PKC-␦ on this site is required for its full enzymatic activity. TPA-induced monocytic differentiation of 32D cells overexpressing PKC-␦S643A is reduced in compari-son to the PKC-␦WT transfectant, suggesting that the mutant protein is less efficient at activating key substrate(s) which affect the differentiation process. The effects of site-directed mutagenesis of PKC-␣ and PKC-␤I at similarly conserved sites were recently reported (23,29). Although no in vitro autophosphorylation data were presented in either study, transphosphorylation of the histone substrate in vitro by PKC-␣T638A mutant was reduced by 26% (23). In contrast, the PKC-␤IT642A mutant completely abolished in vivo phosphorylation and enzymatic activity (29). Whether mutagenesis of PKC-␤I affected the general conformation of the protein remains to be determined. This was suggested by the inability to label in vivo the PKC-␤IT642A mutant protein with [ 32 P]orthophosphate. Although an ATP binding mutant of PKC-␦ (PKC-␦K376R) generated in our laboratory was completely devoid of autophosphorylation capacity (6), it could still be labeled in vivo by [ 32 P]orthophosphate. 2 Two-dimensional phosphopeptide mapping of the PKC-␦K376R mutant revealed that at least two autophosphorylation sites were absent when compared with PKC-␦WT, indicating that other sites in addition to serine 643 2 W. Li, unpublished observations. FIG. 5. PKC-␦S643A mutant protein is thermal stable. Equal amounts of the various cell lysates were enriched for PKC-␦ using DE52 chromatography. Eluates were incubated in a 25°C water bath for various periods and assayed for PKC-␦ activity using the PKC-␦ pseudosubstrate region-derived peptide as a substrate (6). The PKC-␦ specific activity was calculated by subtracting the nonspecific catalytic activity from the total catalytic activity as described in Table II. The variation between the three samples utilized to calculate the mean value of the total catalytic and nonspecific activity was less than 5% of the mean value. The lines with diamonds, circles, and squares represent PKC-specific activity from 32D/pLTR-␦WT-HA, 32D/pLTR-␦S643A-HA, and the parental 32D line, respectively.
FIG. 6. Localization, translocation, and tyrosine phosphorylation of the PKC-␦S643A mutant protein are not altered in comparison to PKC-␦WT. A, 32D cells and the transfectants were serumstarved for 2 h and stimulated with TPA for 10 min. The membrane fraction (P100) was separated from cytosolic fraction (S100) according to previously established methods (13). Equal amounts of proteins were immunoprecipitated (IP) with anti-HA mAb and resolved by SDS-PAGE. Transferred proteins were immunoblotted (Blot) with anti-PKC-␦ serum. B, the same Immobilon membrane utilized in A was reblotted (Blot) with anti-Tyr(P) mAb. must contribute to PKC-␦ autophosphorylation. 2 Moreover, the present results provide evidence that autophosphorylation of the PKC-␦S643A mutant is not completely abolished (see Fig.  3). PKC-␦S643A mutant could be labeled in vivo to a similar extent as PKC-␦WT (see Fig. 4). Based on recent mapping and site-directed mutagenesis results involving PKC-␣ at serine 657 (24) and PKC-␤II at serine 660 (20), we predict that the corresponding serine 662 of PKC-␦ may be an additional autophosphorylation site.
Generation of a serine 643 to alanine mutant of PKC-␦ did not affect the translocation of PKC-␦ from the cytosol to the membrane in response to TPA stimulation, nor did it affect its tyrosine phosphorylation in vivo. These data indicate that sitedirected mutagenesis did not alter the general conformation of the molecule. This is also suggested by the similar two-dimensional phosphopeptide pattern observed for both PKC-␦WT and PKC-␦S643A (see Fig. 4, A and B). Translocation of PKC from the cytosol to the membrane is dependent on the binding of phorbol ester or endogenously produced diacylglycerol to the regulatory domain of PKC (1). Tyrosine phosphorylation of PKC-␦ has also been mapped at the N terminus of PKC-␦ (30). Therefore, it was not surprising that mutation of serine 643 did not affect these events since this mutation resides in the C terminus of the molecule. Although phosphorylation has been implicated to be important for PKC localization, expression of PKC-␦S643A did not alter the localization of the molecule. This can be best explained by the finding that in vitro autophospho-rylation was diminished by only 65-83% in the mutant (see Fig. 3). Thus, alternative autophosphorylation sites may compensate and allow the mutant protein to normally regulate localization through phosphorylation and dephosphorylation dynamics.
In summary, our results demonstrate that serine 643 is a major autophosphorylation site of PKC-␦. Autophosphorylation of PKC-␦ on this site is indispensable for its full enzymatic activity but is not required or sufficient for determining the localization, translocation, or tyrosine phosphorylation of PKC-␦. Mapping the remaining autophosphorylation site(s) within PKC-␦ should make it feasible to determine the complete role of autophosphorylation and its effects on the various aspects of PKC-␦ function.