HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 26, 19567-19576, June 30, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/26/19567    most recent M001357200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Vertommen, D. Articles by Van Lint, J. Search for Related Content PubMed PubMed Citation Articles by Vertommen, D. Articles by Van Lint, J. Social Bookmarking                      What's this?

Regulation of Protein Kinase D by Multisite Phosphorylation

IDENTIFICATION OF PHOSPHORYLATION SITES BY MASS SPECTROMETRY AND CHARACTERIZATION BY SITE-DIRECTED MUTAGENESIS^*

Didier Vertommen^§, Mark Rider^¶, Youping Ni, Etienne Waelkens, Wilfried Merlevede, Jackie R. Vandenheede^**, and Johan Van Lint

From the  Afdeling Biochemie, Faculteit Geneeskunde, Campus Gasthuisberg, Katholieke Universiteit Leuven, Herestraat 49, B-3000 Leuven, Belgium and the ^¶ Institute of Cellular Pathology, Hormone and Metabolic Research Unit, Université Catholique de Louvain, B-1200 Brussels, Belgium

Received for publication, February 18, 2000, and in revised form, April 5, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Activation of the serine/threonine kinase, protein kinase D (PKD/PKCµ) via a phorbol ester/PKC-dependent pathway involves phosphorylation events. The present study identifies five in vivo phosphorylation sites by mass spectrometry, and the role of four of them was investigated by site-directed mutagenesis. Four sites are autophosphorylation sites, the first of which (Ser⁹¹⁶) is located in the C terminus; its phosphorylation modifies the conformation of the kinase and influences duration of kinase activation but is not required for phorbol ester-mediated activation of PKD. The second autophosphorylation site (Ser²⁰³) lies in that region of the regulatory domain, which in PKCµ interacts with 14-3-3. The last two autophosphorylation sites (Ser⁷⁴⁴ and Ser⁷⁴⁸) are located in the activation loop but are only phosphorylated in the isolated PKD-catalytic domain and not in the full-length PKD; they may affect enzyme catalysis but are not involved in the activation of wild-type PKD by phorbol ester. We also present evidence for proteolytic activation of PKD. The fifth site (Ser²⁵⁵) is transphosphorylated downstream of a PKC-dependent pathway after in vivo stimulation with phorbol ester. In vivo phorbol ester stimulation of an S255E mutant no longer requires PKC-mediated events. In conclusion, our results show that PKD is a multisite phosphorylated enzyme and suggest that its phosphorylation may be an intricate process that regulates its biological functions in very distinct ways.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Protein kinase D (PKD)¹ is a serine/threonine protein kinase, also called PKCµ, that was first described as a member of the novel protein kinase C (PKC) subgroup (, , , and ) (1, 2). PKD contains two cysteine-rich domains that bind diacylglycerol or phorbol ester, but it lacks the calcium binding domain seen in the classical PKCs. However, PKD also contains a pleckstrin homology domain that regulates its kinase activity (3) but does not harbor the typical PKC autoinhibitory pseudosubstrate motif. Moreover, the PKD catalytic domain is only distantly related to the kinase subdomains of the PKC family but shows homology to that of the Ca²⁺-regulated kinases, such as myosin light chain kinase and calcium/calmodulin kinase I. Finally, the substrate specificity of PKD is probably different from that of other PKCs, because it is specific for a unique peptide sequence (4). These characteristics have rendered it difficult to classify PKD in the scheme of protein kinases (5), and PKD might be the first member of a new protein kinase family and/or subgroup.

Regulation of protein kinases is achieved through a variety of mechanisms that include auto- and transphosphorylation² events and control by regulatory domains or subunits. Mutagenesis studies have highlighted the regulatory domain of PKD/PKCµ in the negative control of its activity (3, 6). Despite the fact that most protein kinases share a largely conserved catalytic domain structure, their regulation by phosphorylation is very diverse (7, 8). Phosphorylation of specific threonine, serine, or tyrosine residues can occur at a number of sites, some of which are located at the N- or C-terminal ends of the enzyme (e.g. in calmodulin-dependent kinase II and PKCII) or on other subunits (e.g. on phosphorylase kinase). A key feature for regulation is the phosphorylation of residues in the so-called kinase "activation loop" located between subdomains VII and VIII of the kinase core. Here again, this general mechanism holds for most but not for all protein kinases. Phosphorylation of residue(s) in the activation loop may be due to autophosphorylation (e.g. cAMP-dependent protein kinase, c-Src, and insulin receptor kinase) or transphosphorylation catalyzed by another protein kinase (e.g. protein kinase B, p70^S6K, extracellular signal-regulated kinase, PKC). Finally, the phosphorylation of protein kinases, either by autophosphorylation or transphosphorylation, can be the cause of activation or its consequence. For example, three phosphorylation events regulate PKCII activation, the first is catalyzed by an upstream kinase (probably 3-phosphoinositide-dependent protein kinase-1), which phosphorylates Thr-500 in the activation loop, thereby leading to kinase activation and autophosphorylation of two other sites in the C terminus (9).

PKD/PKCµ has been shown to be activated by pharmalogical agents such as phorbol ester and bryostatin 1 (10-12) and by physiological stimuli such as platelet-derived growth factor, tumor-necrosis-factor, angiotensin II, and neuropeptide agonists (13-16). Recent data have shown that PKD plays a role in the regulation of Golgi structure and function (17). Interestingly, PKD may also serve as a molecular switch to promote cell proliferation while inhibiting apoptosis (16, 18, 19). PKD activation was first described as a phorbol ester or diacylglycerol/phospholipid-dependent process. In vitro and in vivo experiments have shown that immunopurified PKD is markedly stimulated by either biologically active phorbol ester or diacylglycerol, in the presence of phosphatidylserine (10, 11, 20). More recently, attention was focussed on phosphorylation events that control the PKD activity (21). These observations were based on the fact that PKD activation was maintained during cell disruption and immunoprecipitation. Additional data, including the use of PKC inhibitors and cotransfection of PKD with constitutively active mutants of PKC and PKC, indicated that PKD was activated by phosphorylation in vivo through a PKC-dependent signal transduction pathway (11, 12, 13). Recent results have demonstrated that PKC interacts with the PH domain of PKD, suggesting a direct link between PKC and PKD (22).

Little is known about how phosphorylation regulates PKD activity, and the phosphorylation sites that mediate its biological functions have not been identified. The group of Rozengurt (23) proposed that the in vivo activation of PKD by phorbol ester results from the phosphorylation of two activation loop serine residues, namely Ser⁷⁴⁴ and Ser⁷⁴⁸, via a novel PKC-dependent signal transduction pathway. However, no sequence studies were undertaken to unambiguously determine that these two serines were actually being phosphorylated in vivo. The C-terminal Ser⁹¹⁶ was suggested to be autophosphorylated in PKCµ/PKD, because it was recognized by a phosphospecific peptide antibody (24). Phosphorylation of Ser⁹¹⁶ was also reported to be induced by phorbol ester treatment of cells in vivo.

The present study identifies five phosphorylation sites in PKD by mass spectrometry, and several of these sites were individually mutated to alanine or glutamate to study their functional role.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- DMEM and phosphate-free DMEM were from Life Technologies. Protein A-TSK gel was from Affiland (Sart-Tilman, Belgium). Glutathione-Sepharose 4B was from Amersham Pharmacia Biotech. Sequencing grade trypsin and chymotrypsin were from Roche Molecular Biochemicals. HPLC solvents were from Lab Scan. Bisindolylmaleimide I (Gö 6850) was from Calbiochem. Shrimp alkaline phosphatase, [-³²P]ATP, and [³²P]orthophosphate were from Amersham Pharmacia Biotech. All other materials were from Sigma or Roche Molecular Biochemicals.

Site-directed Mutagenesis, Expression, and Purification-- The phagemid, called pBluescript (SK)II+/PKD (pBS/PKD), containing the full-length PKD cDNA (10), was used as a template to create eight single mutations (S916A, S916E, S744A, S748A, S744E, S748E, S255A, and S255E) using the QuickChange kit (Stratagene) following instructions provided by the manufacturer. The different mutations were verified by restriction analysis and DNA sequencing. A kinase-dead mutant of PKD (K628N) was also generated by the same strategy.

The DNA sequence encoding wild-type or mutated PKD was subcloned into the eukaryotic expression vector pGMEX-T3 that has been used to overexpress gluthatione S-transferase (GST) fusion proteins in eukaryotic cells under an EF1a promoter. The pBS/PKD phagemids were cleaved with NotI to release the cDNA for PKD, which was then inserted into the compatible ends of pGMEX-T3 to create pGMEX-T3-PKD. To overexpress untagged PKD constructs, wild-type and mutant proteins were also cloned in pcDNA3 vector as described (10)

To prepare the PKD catalytic domain fusion protein (GST-catPKD), a 1014-base pair fragment comprising the entire catalytic domain of PKD was generated by polymerase chain reaction and inserted into pBluescript (SK)II+. The assembled fragment was then subcloned into pGMEXT-3 between the SalI and NotI restriction sites.

To prepare purified GST-PKD or GST-catPKD, 10-cm diameter dishes of human embryonic kidney 293 T cells (HEK 293T), expressing the SV40 large T antigen, were cultured, and each dish was transfected with 7 µg of pGMEX-T3-PKD plasmid DNA using the modified calcium phosphate method (25). Briefly, 2 × 10⁶ HEK 293T cells/dish were grown for 24 h before transfection at 37 °C and 5% CO₂ in DMEM supplemented with 10% (v/v) fetal bovine serum, 100 units/ml penicillin, and 100 µg/ml streptomycin. The DNA was mixed with equal volumes of 0.25 M CaCl₂ and BES-buffered solution and incubated for 30 min at room temperature. The calcium phosphate-DNA solution was added onto medium-containing plates and incubated for 16 h at 37 °C, 3% CO₂. The medium was then replaced with fresh DMEM containing 10% fetal bovine serum, 100 units/ml penicillin, and 100 µg/ml streptomycin. The cells were used for experimental purposes 48 h later. Phorbol ester treatment was with 1 µM PDBu for 15 min. The cells were then washed once with ice-cold phosphate-buffered saline, and each dish was lysed in 1 ml of ice-cold buffer A, pH 7.5 (50 mM Tris, 1 mM EDTA, 1 mM EGTA, 1 mM Na₃VO₄, 50 mM NaF, 5 mM NaPPi, 0.2 µM microcystin, 0.27 M sucrose, 1 mM 4-(2-aminoethyl)-benzenesulfonyl fluoride hydrochloride, 1 µg/ml leupeptin, 15 mM -mercaptoethanol, and 1% (v/v) Triton X-100). Lysates were briefly vortexed and centrifuged at 10,000 × g for 15 min. The supernatants were pooled and incubated for 1 h at 4 °C on a rotating platform with 25 µl (gel bed volume)/dish of glutathione-Sepharose previously equilibrated in one bed volume of Buffer A. The suspension was centrifuged for 10 min at 3,000 × g, and the beads were washed once with 10 bed volumes of Buffer A containing 0.5 M NaCl and once with 10 bed volumes of Buffer B, pH 8.0 (50 mM Tris, 0.1 mM EGTA, 0.27 M sucrose, 15 mM -mercaptoethanol, 10% glycerol (v/v), and 50 mM NaCl). GST-PKD was eluted from the gel with 2× one bed volume of Buffer B containing 20 mM reduced glutathione. The combined eluates were divided into aliquots and stored at 80 °C.

PKD Kinase Assay and Immunoprecipitation-- PKD activity was measured with syntide-2 peptide as substrate (1) under the conditions described in the legends to the figures and tables. One unit of PKD activity corresponds to the amount of enzyme catalyzing the formation of 1 nmol of product/min under the assay conditions. Immunoprecipitation of PKD from cell lysates with a polyclonal antibody and kinase assay was as described (15).

In Vitro Autophosphorylation Assay-- Purified GST-PKD (30 µg) was incubated in buffer containing 15 mM Tris, pH 8.0, 5 mM MgCl₂, and 0.1 mM [-³²P] MgATP (specific radioactivity of 200 cpm/pmol) for up to 60 min at 30 °C. The reaction was stopped by adding SDS-PAGE sample buffer (1% SDS, 10% glycerol, 50 mM dithiotreitol, and 12 mM Tris-HCl, pH 6.8), and the samples were boiled for 5 min for SDS-PAGE in 7.5% acrylamide gels. To determine ³²P incorporation, gels were stained with Coomassie Blue. The bands corresponding to GST-PKD were counted in a Hewlett-Packard Instant Imager together with spotted dried aliquots of the diluted stock solution of [-³²P]MgATP used in the phosphorylation experiments. Stoichiometries of ³²P incorporation (mol/mol of enzyme) were calculated from the amount of protein loaded onto the gel as quantified by the ninhydrin method (see below), and the molecular masses of the GST-PKD and GST-catPKD, taken as 127,575 and 66,975 Da, respectively.

In Vivo Labeling-- HEK 293T cells were cultured and transfected as described above. The dishes were washed five times with phosphate free DMEM containing 100 units/ml penicillin and 100 µg/ml streptomycin and labeled for 4 h with 3 ml/dish of phosphate-free DMEM containing 150 µCi/ml of [³²P]orthophosphate. Cells were stimulated with PBDu, and the PKD-GST proteins were purified as described above.

Identification of Phosphorylation Sites by Electrospray Ionization-Tandem Mass Spectrometry (ESI-MS/MS)-- To identify autophosphorylation sites, GST-PKD (50 µg) was incubated at 30 °C, with 0.1 mM [-³²P]MgATP (specific activity, 300-500 cpm/pmol). After 60 min, the reaction was stopped by adding 10% trichloroacetic acid (v/v) and left on ice for 1 h. Precipitated protein was collected by centrifugation (12,000 × g for 15 min), washed once with ice-cold acetone, and resuspended in 50 µl of 0.1 M Tris-HCl, pH 8.5, 0.6% (w/v) n-octylglucoside for overnight digestion at 30 °C with 1 µg of sequencing grade chymotrypsin or trypsin. Peptides were separated by reversed-phase narrowbore HPLC on a Vydac C18 column (1.0 mm × 25 cm) in an acetonitrile gradient in 0.1% (v/v) trifluoroacetic acid (solvent A). Elution was performed with the following gradient program: 5-100% solvent B (70% acetonitrile in solvent A) over 100 min at a flow rate of 40 µl/min generated by a model 140B Applied Biosystems solvent delivery system. Peptides were collected by hand, and radioactive peptides were identified by Cerenkov counting. Radioactive peptides were dried under vacuum and redissolved in 4-6 µl of 60% (v/v) methanol, 1% (v/v) acetic acid for nanospray ESI-MS/MS.

To identify transphosphorylation sites, GST-PKD (50 µg) purified from ³²P-labeled cells, was precipitated, redissolved, and digested as described above. Peptides were separated by reversed-phase HPLC on a Amersham Pharmacia Biotech C2/C18 column (2.1 mm × 10 cm) connected to a Amersham Pharmacia Biotech SMART system. Column was equilibrated in 0.1% (v/v) trifluoroacetic acid (solvent A). Elution was performed with the following gradient program: 7-70% solvent B (100% acetonitrile, 0.1%(v/v) trifluoroacetic acid) over 80 min at a flow rate of 80 µl/min. Peptides absorbing at 215 nm were collected, counted, dried, and dissolved as described above for nanospray ESI-MS/MS.

Radioactive peaks were analyzed by nanospray ESI-MS/MS. Briefly, 2-3 µl of the radioactive peptides were analyzed in a LCQ (Finnigan MAT LCQ, San Jose, CA) equipped with a nano-electrospray ionization source. Spectra were taken in full MS and zoom scan mode to determine parent masses and their charge state. The source voltage was set at 0.8 kV with a scan time of 3.6 s. The collision energy was adjusted to the minimum needed for fragmentation.

Identification of Phosphorylation Sites by HPLC-ESI-MS/MS-- To identify in vivo phosphorylation sites, peptides were separated by reversed-phase HPLC on a C18 capillary column (0.3 mm × 25 cm, LC Packings) with an acetonitrile gradient in 0.05% (v/v) formic acid (solvent A). Elution was performed with the following gradient: 0-100% solvent B (95% acetonitrile (v/v) in 0.05% (v/v) formic acid) over 100 min at a flow rate of 5 µl/min generated by a 140B pump (Applied Biosystems) connected to a flow splitter (1/20, Accurate solvent splitter, LC Packings). Mass spectra were recorded on-line in the LCQ (Finnigan MAT LCQ, San Jose, CA) using the standard electrospray ionization source. Electrospray was performed at a voltage of 5.6 kV with a scan time of 1.2 s. Mass spectra were acquired in a mode that alternated single MS scans (m/z 500-2000) with MS² and MS³ scans.

Other Methods-- Protein was measured by the Bradford method (26) using -globulin as a standard or by the reaction with ninhydrin after trichloroacetic acid precipitation and complete alkaline hydrolysis (27) using bovine serum albumine as a standard. SDS-polyacrylamide gel electrophoresis analysis in 10% or 7.5% (w/v) acrylamide was as described (28). Kinetic constants were calculated by fitting data to a hyperbola by nonlinear least square regression using a computer program (Ultrafit, Biosoft, Cambridge, UK)

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Purification of Wild-type and Mutant PKD Preparations-- Engineering of a N-terminal GST tag in PKD allowed rapid purification of the protein by a one-step procedure. Cell lysates were directly mixed with glutathione-Sepharose 4B, and GST-PKD preparations were eluted with reduced gluthatione. SDS-polyacrylamide gel electrophoresis analysis of the purified preparations of wild-type PKD (GST-PKD), PKD catalytic domain (GST-catPKD), and mutant proteins showed single 132,000- or 68,000-Da bands in agreement with the calculated masses. The purified proteins were stored in elution buffer at 80 °C with no appreciable loss of activity over several months.

Characterization of Purified Wild-type and Catalytic Domain PKD-- The catalytic domain of PKD had a 12-fold higher k_cat than wild-type PKD (6 s¹ versus 0.5 s¹). The affinities for MgATP and syntide-2 were similar for the two recombinant enzymes, suggesting that the overall structure of the catalytic domain was maintained and that the GST tag had no influence on the kinetic properties of the enzyme (Table I). Likewise, addition of a N-terminal green fluorescent protein tag in PKD has previously been shown to have no influence on PDBu-induced translocation, basal catalytic activity, phorbol ester binding, and kinase activation (29). Moreover, in vitro incubation of GST-PKD with PS/PDBu micelles led to a 5-fold stimulation of PKD activity (not shown). A comparable stimulation was observed with untagged PKD (10, 23).

View this table: [in this window] [in a new window]   Table I Kinetic properties of PKD in the recombinant GST-tagged wild-type and mutant preparations under basal or PDBu-stimulated conditions HEK-293T cells were transiently transfected with vectors expressing GST-PKD, GST-catPKD, and the different mutants of GST-PKD. The cells were stimulated with PDBu, and the recombinant proteins were purified as described under "Experimental Procedures." Purified PKD activity was measured at 30 °C in buffer containing 15 mM Tris-Cl, pH 8.0, 5 mM MgCl2, 1 mg/ml bovine serum albumin, 500 µM [-32P]MgATP (100 cpm/pmol), and 500 µM syntide-2 peptide (1). For the syntide-2 and MgATP saturation curves, the concentration of substrates were varied up to 10 times the Km. The concentration of the other substrate was 500 µM, except for the S744A mutant, where the syntide-2 concentration was 1 mM. The values are the means ± S.E. for at least three determinations.

Incubation of HEK 293T cells with PDBu caused a 3-fold increase in k_cat for GST-PKD with no effect on the affinities for MgATP and syntide-2 (Table I). By contrast, PDBu treatment had no effect on the activity of GST-catPKD. Therefore, deletion of the regulatory domain of PKD leads to a constitutively active kinase, and these results suggest that the region for PKC-dependent activation is located outside the catalytic domain. These results support those obtained in experiments using partial deletions or point mutations in the regulatory domain of PKD (3, 6).

Time course experiments of in vitro autophosphorylation showed that ³²P incorporation into GST-PKD was maximal after 60 min and was maintained for up to 80 min (not shown). The initial rate of GST-PKD autophosphorylation was independent of enzyme concentration with an activity of 50 pmol/min/mg (not shown), indicating that autophosphorylation of GST-PKD occurs via an intramolecular event at a very slow rate. Indeed, for Dictyostelium myosin light chain kinase, which has a catalytic domain possessing 40% identity with the PKD kinase domain, autophosphorylation is also intramolecular but with a 15-fold faster rate (30). The stoichiometry of autophosphorylation of GST-PKD and GST-catPKD was 0.4 and 0.2 mol of phosphate incorporated per mol of enzyme, respectively, suggesting the existence of at least two autophosphorylation sites, one located in the catalytic domain and the other in the regulatory domain. The low incorporation of radioactive phosphate in vitro could reflect the fact that these sites were already largely phosphorylated in vivo (see below).

Ser⁹¹⁶ Is an Autophosphorylation Site in Vitro and in Vivo and Is Involved in the Down-regulation of PKD Activity after PDBu Stimulation-- Purified GST-PKD was autophosphorylated in vitro by incubation with [-³²P]MgATP. Following trichloroacetic acid precipitation, the protein was digested with chymotrypsin, and peptides were separated by reversed-phase HPLC, and one major radioactive peak was observed (not shown). This peak was analyzed by nanospray ESI-MS/MS to identify the phosphorylation site. The fraction contained several ions in full MS mode, only one of which (m/z = 883.4, P1 in Table II) lost 98 Da in the ion trap when subjected to a low collision energy, and its mass was decreased to m/z = 785.3 (Fig. 1). The difference of 98 Da corresponds to the loss of H₃PO₄ through -elimination, leaving dehydroalanine in place of the phosphorylated serine residue (31). An ion of m/z = 803.4 (883.4 minus 80 Da for the PO₃² group) could correspond to a theoretical chymotryptic peptide with an average mass of 803.9 Da representing the sequence ⁹¹²SERVSIL⁹¹⁸. Indeed, when the ion of m/z = 785.3 was fragmented in the ion trap, the sequence of the chymotryptic fragment was confirmed, and the phosphorylated residue was identified as Ser⁹¹⁶ (Fig. 1). To see whether this site was phosphorylated in vivo, GST-PKD was purified from unstimulated or PDBu-stimulated cells and digested with chymotrypsin. The resulting peptides were analyzed by on-line capillary HPLC-ESI-MS/MS. Ser⁹¹⁶ was found to be phosphorylated in both conditions, indicating that Ser⁹¹⁶ is phosphorylated in vivo (Table II). Ser⁹¹⁶ was also found to be autophosphorylated in GST-catPKD, indicating that autophosphorylation of Ser⁹¹⁶ does not depend on the presence of the regulatory domain. Some minor radioactive HPLC peaks contained phosphopeptides generated by missed cleavages during proteolysis (P3 and P4 in Table II).

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Identification of Ser916 as an in vitro phosphorylation site. The major 32P-labeled peak from the chymotryptic digestion of in vitro autophosphorylated GST-PKD was analyzed by nanospray ESI-MS/MS and found to contain one phosphopeptide ion (P1 in Table II). a, MS2 spectrum of this mono charged phosphopeptide ion (m/z 883.4). A loss of 98 Da is observed (H3PO4) to produce an ion with m/z 785.3. b, MS3 spectrum of the ion arising from loss of 98 Da (m/z 785.3 of the mono charged ion in a). The b4 and b5 fragments have a mass difference of 69 Da corresponding to dehydroalanine, identifying the product of phosphoserine after losing 98 Da. The B label denotes dehydroalanine, and the b and y labels refer to ions containing the N-terminal or C-terminal ends of the molecule, respectively.

To study the role of Ser⁹¹⁶ autophosphorylation, this residue was mutated to alanine (S916A) or glutamate (S916E), and the kinetic properties of PKD were studied. The two mutants had similar kinetic parameters compared with wild-type GST-PKD, with no drastic changes in k_cat or K_m (Table I). Moreover, the S916A and S916E mutants could be activated in cells treated with PDBu to the same extent as wild-type GST-PKD (Table I). Mutation of Ser⁹¹⁶ to glutamate did not overcome the need for PKC-activity in the PDBu-mediated PKD activation (see below and Fig. 6). Therefore, autophosphorylation of Ser⁹¹⁶ is not required either for activity or for in vivo activation by PDBu.

Other roles for phosphorylation sites in the C terminus in PKCs have been proposed. For example, C-terminal phosphorylation sites may increase protein stability or increase the resistance to dephosphorylation by protein phosphatases (32). C-terminal phosphorylations have also been reported to affect protein subcellular partitioning, sensitivity to proteolysis, or affinity for substrates, phosphatidylserine, or Ca²⁺ (9, 33). Experiments were therefore undertaken to see whether phosphorylation at Ser⁹¹⁶ in GST-PKD could cause similar changes. For the S916E and S916A mutants, the PS/PDBu dependence of GST-PKD substrate phosphorylation was measured in the presence of mixed micelles containing Triton X-100 (34). None of the mutations significantly affected the PS/PDBu dependence of the GST-PKD activity (not shown). We then tested the sensitivity of PKD toward proteolysis by trypsin. This technique has been used to study conformational changes in PKC, for example those induced by membrane binding (35). Incubation of the in vivo PDBu-stimulated wild-type and S916E preparations with trypsin (0.02 unit ml¹) led to extensive proteolysis of the native enzyme and the appearance of a 42,000-Da fragment, which corresponds to the mass of the catalytic domain. No intact GST-PKD was left after incubation with 0.2 unit ml¹ of trypsin. By contrast, the S916A mutant was more resistant to proteolysis, requiring higher concentrations of trypsin to obtain a similar pattern of proteolysis, and intact enzyme was still apparent after incubation with 0.2 unit ml¹ trypsin (Fig. 2). We also investigated the sensitivity of the in vivo PDBu-stimulated wild-type and Ser⁹¹⁶ mutants toward dephosphorylation by alkaline phosphatase. Following in vitro autophosphorylation with [-³²P]MgATP and incubation with alkaline phosphatase, the wild-type and S916E preparations showed a time-dependent decrease in their extent of phosphorylation, whereas the S916A mutant was resistant to dephosphorylation (Fig. 3). Finally, we investigated whether Ser⁹¹⁶ mutation could affect any in vivo properties of PKD, by examining the time-dependent down-regulation of PKD activity after PDBu stimulation. For this experiment untagged PKD constructs were cloned in pcDNA 3 vector and transiently transfected in HEK-293T cells. After PDBu treatment, cells were washed and incubated for another 6 h in DMEM without PDBu. PKD activity was measured after immunoprecipitation at different time points. The wild-type and S916E mutant showed a time-dependent decrease in activity reaching 56 or 64% of initial activity, respectively, after 6 h (Fig. 4). The slow down-regulation of the PKD activity seen after phorbol ester treatment confirms previous studies (36). In contrast, PKD activity of the S916A increased during the first hour of incubation and decreased thereafter at a similar rate compared with the wild-type reaching 85% of initial activity after 6 h (Fig. 4). We also tested whether this decrease in activity was reversible. After 6 h of down-regulation, cells were restimulated with PDBu without changing the medium. The PKD activity measured after the second PDBu stimulation was similar to the initial activity (after the first PDBu treatment), indicating that down-regulation of PKD activity occurs via a reversible mechanism, probably reflecting reversible dephosphorylation of the enzyme (Fig. 4).

View larger version (11K): [in this window] [in a new window]   Fig. 2.   Sensitivity to trypsin proteolysis of wild-type, S916A, and S916E recombinant GST-PKD preparations. Purified PKD (5 µg) from PDBu-stimulated cells was incubated for 15 min at 30 °C in buffer containing 20 mM Tris-Cl, pH 8.0, 0.3 mM CaCl2, and the indicated concentrations of trypsin. Proteins were separated by SDS-PAGE in 10% acrylamide and stained with Coomassie Brilliant Blue. Intact GST-PKD and the 42,000-Da fragment generated from the wild type are indicated by arrows.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Sensitivity of wild-type, S916A, and S916E recombinant GST-PKD preparations to alkaline phosphatase treatment. Purified wild-type (), S916A (), or S916E () (10 µg of each) from PDBu-stimulated cells were autophosphorylated in vitro as described under "Experimental Procedures." Proteins were incubated at 30 °C in buffer containing 50 mM Tris-Cl, pH 9.0, 20 mM MgCl2 with 25 units/ml of shrimp alkaline phosphatase. Aliquots were taken at the indicated times, and proteins were separated by SDS-PAGE in 10% acrylamide. The extents of phosphorylation of the GST-PKD bands were measured after Coomassie Blue staining and phosphorimaging (Molecular Dynamics). The results are the means of two separate experiments. 100% corresponds to 0.40, 0.27, and 0.25 mol of phosphate incorporated per mol of wild-type, S916A, and S916E, respectively.

View larger version (25K): [in this window] [in a new window]   Fig. 4.   Time course of down-regulation of PKD activity after phorbol ester stimulation. HEK 293T cells transiently expressing wild-type (), S916A (), or S916E () untagged PKD were stimulated with 500 nM PDBu for 15 min. After extensive washing with phosphate-buffered saline, cells were lysed (time 0) or incubated in DMEM without PDBu for the indicated times and then lysed. PKD activity was measured after immunoprecipitation with a polyclonal antibody as described (15). 100% of PKD activity corresponds to the initial activity measured right after PDBu stimulation (time 0). The results are the means ± S.E. for three separate determinations. Inset, relative activity of PKD in HEK 293T cells transiently expressing wild-type PKD. Closed bar, after PDBu stimulation; open bar, 6 h after PDBu stimulation; shaded bar, 6 h after PDBu stimulation and restimulated 15 min with PDBu. The results are the means ± S.E. for three separate determinations. *, p < 0.05 versus wild type at same time.

Identification of Ser²⁰³ as a Second in Vitro and in Vivo Autophosphorylation Site-- The S916A and S916E mutants still autophosphorylate, suggesting the existence of other autophosphorylation sites, which were not detected after chymotryptic cleavage. Therefore, autophosphorylated GST-PKD was digested with trypsin instead of chymotrypsin. One major radioactive peak was isolated by HPLC and analyzed by nanospray ESI-MS/MS. This peak did not contain a peptide corresponding to the predicted short sequence containing phosphorylated Ser⁹¹⁶ (VpSIL), which may not have been retained by the C18 column. A double charged ion with m/z = 777.2 was found in the major peak, which lost H₃PO₄ in the ion trap, and its m/z decreased to 728.3. This could correspond to the tryptic peptide ²⁰¹RLSNVSLTGLGTVR²¹⁴ (P5 in Table II). Fragmentation of the m/z = 728.3 ion confirmed the sequence and identified Ser²⁰³ as the autophosphorylation site (not shown). We also looked for this phosphorylation site in vivo by on-line capillary HPLC-ESI-MS/MS. GST-PKD purified from unstimulated or PDBu-stimulated cells was digested by trypsin or chymotrypsin. Ions corresponding to phosphopeptides containing Ser²⁰³ were found in both conditions with both proteases (P5 and P7 in Table II). Fragmentation of the ions losing H₃PO₄ confirmed that Ser²⁰³ is phosphorylated in vivo (Fig. 5).

View larger version (31K): [in this window] [in a new window]   Fig. 5.   Identification of Ser203 as an in vivo phosphorylation site. Purified GST-PKD was digested with trypsin, and the peptides were separated and analyzed on-line by capillary HPLC-ESI-MS/MS. The LC/MS data were scanned for ions with m/z predicted for the unphosphorylated, monophosphorylated, and diphosphorylated peptides containing Ser203. After identifying candidate ions from this initial analysis, the sample was run a second time, and the ions were selected for on-line CID in MS2 and MS3 mode. a, MS full scan for a phosphopeptide ion of m/z 777.3 (P5 in Table II). b, MS3 spectrum of the ion arising from loss of 98 Da (m/z 728.3 of the double charged ion in a). The y12 and y11 fragments have a mass difference of 69 Da that corresponds to the mass of dehydroalanine identifying Ser203 as phosphoserine. For other details of the labeling, see legend to Fig. 1.

Identification of Ser²⁵⁵ as an in Vivo Transphosphorylation Site in PKD-- HEK-293T cells, transiently expressing GST-PKD, were labeled with [³²P]orthophosphate and stimulated with PDBu. Following cell lysis, GST-PKD was purified and digested with chymotrypsin, and the three major radioactive peaks isolated by HPLC were analyzed by nanospray ESI-MS/MS. Two peaks contained the autophosphorylation sites Ser⁹¹⁶ and Ser²⁰³. The third peak contained a double charged ion with m/z = 753.1, which lost H₃PO₄ to give m/z = 704.0 in the ion trap of the mass spectrometer. Fragmentation of this ion identified a peptide corresponding to the sequence ²⁴⁷IGREKRSNSQSY²⁵⁸, in which Ser²⁵⁵ was the phosphorylated residue (P6 in Table II). This peak was not labeled in GST-PKD purified from unstimulated cells. Ser²⁵⁵ is located between the two cysteine-rich domains and is conserved in PKCµ. It has basic residues at positions 3 and 4, which have been shown to be important for PKC substrate recognition. The sequence lacks positive residues at position +2 and +3, but basic residues at these positions are not absolutely required for novel PKC family members (such as PKC and PKC) (4). However, purified PKC or PKC preparations did not phosphorylate GST-PKD in vitro (not shown), suggesting an indirect role of these kinases in the activation of PKD. To test the role of Ser²⁵⁵ phosphorylation in PKD activation, it was mutated to alanine or glutamate (S255A and S255E). The two mutations decreased the k_cat of the enzyme (2-4-fold) without affecting the affinities for MgATP or syntide-2 (Table I). Thus, mutation of Ser²⁵⁵ into glutamate certainly did not induce a constitutively active form of PKD. However, PDBu treatment led to a greater degree of activation of the S255E mutant than the wild type (11-fold versus 3-fold). Surprisingly, mutation of Ser²⁵⁵ to alanine did not abolish PDBu-induced activation, indicating that this site is not essential for PKD activation. To test whether this site might be phosphorylated downstream of a PKC-dependent signaling pathway, we studied the PDBu-induced activation of the S225E and S255A mutants in the presence and absence of Gö 6850. This PKC inhibitor prevents the activation of PKD in response to phorbol ester or mitogens (11, 15, 21). Treatment of HEK 293T cells with Gö 6850 significantly decreased the PDBu-induced PKD activation of wild-type and S255A GST-PKD but had no effect on the activation of the S255E mutant (Fig. 6), indicating that this site is indeed phosphorylated by a PKC-dependent pathway upon stimulation with phorbol ester.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Inhibition of PDBu-induced activation of PKD by bisindolylmaleimide I. HEK 293T cells transiently expressing wild-type and mutated GST-PKD were incubated in DMEM with (open bars) or without (closed bars) 4 µM bisindolylmaleimide I (Gö 6850) for 2 h. Cells were then treated with 1 µM PDBu for 10 min or directly lysed (control). Lysis, purification of GST-PKD and PKD activity measurements were performed as described under "Experimental Procedures." Results are expressed as fold activation of PKD versus control conditions (no PDBu). The results are the means ± S.E. for three separate determinations. *, p < 0.05 versus condition without Gö 6850.

In Vivo Activation of Full-length PKD by Phorbol Ester Does Not Encompass Phosphorylation of the Activation Loop Ser⁷⁴⁴ and Ser⁷⁴⁸-- Ser⁷⁴⁴ and Ser⁷⁴⁸ in the activation loop of PKD have been proposed to become phosphorylated in intact cells in response to PDBu stimulation (23). Rozengurt and colleagues used a combination of mutational analysis and two-dimensional peptide mapping to demonstrate the potential role of these two residues in the activation of PKD. However, the two phosphorylation sites were not unambiguously identified, and a detailed kinetic study of the serine mutants was not performed. We did not find any radioactive peptides containing phosphorylated Ser⁷⁴⁴ or Ser⁷⁴⁸ in ³²P-labeled HEK-293T cells that were transiently expressing GST-PKD, with or without PDBu stimulation. However, this alone does not rule out the in vivo phosphorylation of these sites, because they could have been already phosphorylated by PDBu-independent mechanisms involved in the maturation of the enzyme, as already described for conventional PKCs (37). We therefore looked for the Ser⁷⁴⁴ and Ser⁷⁴⁸ phosphorylation sites in the activation loop of PKD by on-line capillary HPLC-ESI-MS/MS. Unlabeled HEK 293T cells were treated with or without PDBu and GST-PKD was purified for digestion with trypsin or chymotrypsin. No phosphopeptides containing Ser⁷⁴⁴ and Ser⁷⁴⁸ were detected in PDBu-stimulated or unstimulated cells. Moreover, we were able to identify and sequence the peptides in which the two activation loop serines were nonphosphorylated. This indicates that these two sites are not phosphorylated in full-length PKD, either in PDBu-stimulated or unstimulated cells.

We also searched for Ser⁷⁴⁴ and Ser⁷⁴⁸ phosphorylation in GST-catPKD. Following in vitro autophosphorylation with [-³²P]MgATP, GST-catPKD was digested with chymotrypsin, and peptides were separated by HPLC. Radioactive peaks were analyzed by nanospray ESI-MS/MS. In addition to the previously identified phosphorylated Ser⁹¹⁶, we were able to identify and sequence another phosphopeptide in which Ser⁷⁴⁸ was phosphorylated (P2 in Table II), indicating that this is an in vitro autophosphorylation site in the expressed catalytic domain. GST-catPKD was also analyzed by on-line capillary HPLC-ESI-MS/MS directly following purification from unstimulated cells. Ser⁹¹⁶ and Ser⁷⁴⁸ were phosphorylated, and we also identified phosphorylated Ser⁷⁴⁴ (P8 in Table II). This suggests that all three sites are phosphorylated in vivo. None of these phosphorylation sites were detected by on-line capillary HPLC-ESI-MS/MS in a kinase-dead mutant of GST-catPKD (K628N). This indicates that Ser⁷⁴⁴, Ser⁷⁴⁸, and Ser⁹¹⁶ are in vivo autophosphorylation sites in GST-catPKD. Because only Ser⁷⁴⁸ and Ser⁹¹⁶ could be autophosphorylated in vitro, we conclude that Ser⁷⁴⁴ is constitutively phosphorylated in vivo under basal conditions.

We also decided to investigate by site-directed mutagenesis the roles of Ser⁷⁴⁴ and Ser⁷⁴⁸ in PKD activation by PDBu. The two serine residues were mutated to Glu or Ala in the full-length GST-PKD to generate four single points mutants (S744A, S744E, S748A, and S748E). The mutants were then expressed in HEK 293T cells and purified as described above. We studied the effects of the mutations on kinetic parameters (Table I). Mutation of Ser⁷⁴⁴ or Ser⁷⁴⁸ to Ala drastically decreased the k_cat by 22- or 8-fold, respectively, and increased the K_m for syntide-2. The S744E mutant also displayed a lower k_cat (4-fold), and there was a slight decrease in affinity for MgATP (2-fold). Mutation of Ser⁷⁴⁸ to glutamate had no effect on the kinetic properties of PKD. Interestingly, all mutants could be activated in cells treated with PDBu. The S744A and S748A mutants were activated to the same extent as the wild type (3-5-fold), whereas the S744E and S748E mutants displayed a 24- or 8-fold activation, respectively. Moreover, the S744E and S748E mutants were sensitive to Gö 6850-induced inhibition of PDBu-mediated PKD activation (Fig. 6). These results indicate that neither of the activation loop serines is involved in PDBu-induced activation but that they may be involved in catalysis or in maintaining the conformation of the enzyme protein. Phosphorylation of Ser⁷⁴⁴ and Ser⁷⁴⁸ in GST-catPKD is probably responsible for the higher catalytic activity of this construct. Indeed mutation of Ser⁷⁴⁴ to alanine in GST-catPKD decreased the k_cat 20-fold to 0.25 s¹.

Activation of PKD by Proteolysis-- PKC was originally described as a protein kinase that could be activated by limited proteolysis (38). However, it is now generally accepted that a reversible activation of PKC, rather than irreversible proteolytic activation, is the major means of regulation for this family of kinases (39). Nevertheless, an increasing number of kinases having a large regulatory domain (p21-activated protein kinase, mitogen-activated protein kinase kinase kinase-1, PKC, and protein kinase C-related kinase-1) are activated by proteolysis (40-43). We tested whether PKD could be activated by proteolysis. Purified GST-PKD was incubated with 0.02 unit ml¹ of trypsin, and a time course of PKD activation was studied (Fig. 7). Partial proteolysis resulted in the appearance of three major bands analyzed by SDS-PAGE with masses of 90,000, 42,000, and 26,000 Da (Fig. 7) and resulted in an increase in the PKD activity (4-14-fold depending on the PKD/Trypsin molar ratio). In vivo treatment with PDBu prior to proteolysis had no effect on this proteolytic activation, suggesting that the trypsin cleavage sites do not only become exposed after phorbol ester stimulation (Fig. 7).

View larger version (13K): [in this window] [in a new window]   Fig. 7.   PKD activation by limited proteolysis. a, purified GST-PKD (25 µg) was incubated with 0.02 unit/ml trypsin in a buffer containing 20 mM Tris, 0.3 mM CaCl2, pH 8.0, at 30 °C. Aliquots were taken at the indicated time and analyzed by SDS-PAGE in 10% acrylamide. The gel was stained with Coomassie Brilliant Blue. The three major proteolytic fragments are indicated by arrows (masses of 90,000, 42,000, and 26,000 Da). b, GST-PKD purified from unstimulated or PDBu-stimulated HEK 293T cells was incubated with 0.02 unit/ml of trypsin (filled symbols) in a buffer containing 20 mM Tris, 0.3 mM CaCl2, pH 8.0, at 30 °C. Aliquots were taken at the indicated time, and PKD activity was measured as described under "Experimental Procedures," except that 1 mM 4-(2-aminoethyl)-benzenesulfonyl fluoride hydrochloride was included in the assay buffer to inhibit any further proteolysis. The results are representative of three separate determinations. WT, wild type.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In this study we identified five phosphorylation sites in PKD, shown schematically in Fig. 8. Two sites are located in the regulatory domain (Ser²⁵⁵ and Ser²⁰³) in a region that not only mediates the interaction of PKD/PKCµ with other proteins but also controls the catalytic activity of the enzyme. Two other sites are present in the activation loop (Ser⁷⁴⁴ and Ser⁷⁴⁸) but are only phosphorylated by the isolated catalytic domain and not by the full-length protein. Their phosphorylation is not regulated by phorbol ester. The presence of a C-terminal phosphorylation site (Ser⁹¹⁶) links PKD to the growing number of kinases phosphorylated on their C terminus, an event that is thought to provide an electrostatic anchor that structures the kinase and/or alters its surface to promote or disrupt protein-protein interactions (44-46).

View larger version (57K): [in this window] [in a new window]   Fig. 8.   Localization of phosphorylation sites in the wild-type and the catalytic domain of PKD. Schematic outline of the structural domains of PKD containing the two cysteine-rich domains (CRD), the pleckstrin homology domain (PH), and the kinase core (CAT). Phosphorylation sites are indicated by an open circle (autophosphorylation) or a closed circle (transphosphorylation). Proteins that associate with PKD are listed under the region of PKD with which they have been proposed to interact (17, 22, 50, 54, 55).

Ser⁹¹⁶ at the C-terminal end was found to be autophosphorylated in vitro and in vivo both in GST-PKD and in GST-catPKD (Fig. 1 and Table II). Interestingly, Ser⁹¹⁶ is not in a C-terminal phosphorylation site consensus sequence FXXF(S/T)(Y/F) identified in other kinases (protein kinase B, conventional PKCs, novel PKCs, and p70^S6K) (47). Using an anti-phospho-Ser⁹¹⁶ antibody, it was shown that phosphorylation of Ser⁹¹⁶ correlated with PKD activity and was induced by phorbol ester or by antigen receptor triggering in lymphocytes (24). Replacement of Ser⁹¹⁶ by alanine or glutamate demonstrated that autophosphorylation of Ser⁹¹⁶ is not required for activity or for in vivo activation by phorbol ester (Table I). Likewise in PKC II, where two autophosphorylation sites are located at the C terminus, one site (Ser-660) does not control the kinase activity but rather plays a structural role in both the active site and the regulatory region by increasing the affinity for substrates, phosphatidylserine and Ca²⁺ (9, 33). The S916A mutant showed a reduced sensitivity to proteolysis (Fig. 2) and to dephosphorylation by alkaline phosphatase (Fig. 3), which indicates that the Ser⁹¹⁶ to alanine mutant might have a more closed conformation and that a C-terminal negative charge would favor a more open structure. Moreover, the S916A mutant exhibited a delayed time-dependent down-regulation of its activity after phorbol ester stimulation (Fig. 4). We also showed that the down-regulation of PKD activity is a reversible process, probably under the control of a protein phosphatase (Fig. 4). This is in agreement with our previous report showing that the activation of PKD can be fully reversed in vitro by protein phosphatases PP1_c and PP2A_c (15). There exist several examples of C-terminal kinase phosphorylation/dephosphorylation as a regulatory mechanism for kinase activity down-regulation. For protein kinase B, down-regulation of the kinase activity occurs via dephosphorylation of the two major regulatory phosphorylation sites (Thr³⁰⁸ in the activation loop and Ser⁴⁷³ at the C-terminal end). It is known that their mutation to aspartate leads to a constitutively active enzyme that cannot be down-regulated (48). As a variation on this theme, IKK also contains C-terminal autophosphorylation sites involved in the down-regulation of its kinase activity (49). However, replacement of these serines with alanine in IKK results in a mutant that remains active four times longer than the wild-type enzyme. Likewise, mutation of Ser⁹¹⁶ to alanine in PKD instigates a slower down-regulation of the kinase activity, whereas a negative charge at this position induced by phosphorylation or mutagenesis seems to favor the process (Fig. 4), possibly by inducing a conformational change. This could render other phosphorylation sites more accessible to protein phosphatases that are involved in the reversal of kinase activation. It should be mentioned that the down-regulation of PKD activity was postponed but not abolished in the S916A mutant, indicating that additional mechanisms are involved.

14-3-3 proteins have been proposed to associate with PKCµ and to negatively regulate PKCµ kinase activity (50). Mutational analysis suggested that this association involved two serine pairs (serines 205/208 and 219/223 in PKCµ), and both these pairs of serine residues were proposed to be autophosphorylation sites of PKCµ. However, in these studies, only combinations of double mutants were tested, which cannot pinpoint the individual residues required for interaction with 14-3-3. Moreover the phosphorylated serine residue(s) were not positively identified. These two serine pairs are conserved in PKD and correspond to serines 203/206 and 217/221. Here we identified Ser²⁰³ as an in vitro and in vivo autophosphorylation site in PKD. Ser-206 was detected as being nonphosphorylated (Table II). We have no evidence that a serine residue in the second pair is phosphorylated. Moreover, we could identify and sequence peptides containing Ser-217 and Ser-221 in their nonphosphorylated states in autophosphorylated PKD.

We identified Ser²⁵⁵ as a PDBu-induced transphosphorylation site in ³²P-labeled cells, downstream of a PKC-dependent pathway (Table II). Although the S255E mutant is not constitutively active, its activation by phorbol ester no longer requires PKC activity (Fig. 6). The demonstration that of all the mutants tested only the S255E mutant can still be activated by PDBu in the presence of PKC inhibitors indicates that the stimulation of PKD by phorbol ester also encompasses events other than PKC-mediated phosphorylation. In view of the fact that PKD can be partially stimulated in vitro by the addition of PS/PDBu micelles, one could envisage that PDBu uses a bifurcating pathway for the activation of PKD in vivo. One signaling path leads to a PKC-dependent transphosphorylation of Ser²⁵⁵, whereas the second path involves other PDBu-dependent events perhaps involving direct PDBu binding to the PKD zinc fingers. Replacement of Ser²⁵⁵ by Ala demonstrated that Ser²⁵⁵ phosphorylation is not strictly required for PKD activation by PDBu (Table I) but may be required for more efficient activation by the PKC pathway. This suggests that the activation of the S255A mutant may be the result of a phosphorylation of a neighboring serine and moreover points to the possibility that the PKC-dependent path of PKD-activation encompasses several equivalent phosphorylation sites in the vicinity of Ser²⁵⁵. Multisite phosphorylation is a characteristic of many protein kinases, e.g. in mitogen-activated protein kinase activated protein (MAPKAP) kinase-2 where any two of three sites must be phosphorylated to achieve maximal activation (51). A negative charge on the Ser²⁵⁵ site (induced by mutagenesis or by phosphorylation) may be a prerequisite for the in vivo activation of PKD via the second phorbol ester-mediated pathway.

Our results demonstrate that Ser⁷⁴⁴ and Ser⁷⁴⁸ are not phosphorylated in full-length GST-PKD in response to phorbol ester stimulation (Table II). However, these two sites are autophosphorylated in vivo in GST-catPKD, which probably explains why GST-catPKD is highly active and cannot be further stimulated by the PDBu/PKC pathway (Table I). PKD does not belong to the family of RD kinases, which are defined as kinases where the conserved catalytic aspartate is preceded by an arginine residue. Most RD kinases are regulated by phosphorylation in the activation loop (8). In the three-dimensional structures of several RD kinases, the arginine residue in the "RD motif" interacts with the phospho-amino acid in the activation loop (e.g. cAMP-dependent protein kinase, mitogen-activated protein kinase, and cyclin-dependent kinase 2) or with a corresponding acidic residue (e.g. phosphorylase kinase) to promote the correct positioning of the catalytic site residues (7). Non-RD kinases, like PKD, are proposed not to be regulated by phosphorylation in the activation loop (8). There is an interesting double substitution in twitchin, the only non-RD kinase of known structure, with the ion pair seen in RD kinases being replaced by two uncharged residues (valine and leucine) (52). This probably explains why this kinase is not regulated by phosphorylation in the activation loop. In PKD, an intermediate situation is found where the arginine of the RD motif is replaced by a cysteine, whereas Ser⁷⁴⁸, a possible phosphorylation target corresponding to the phosphorylated residue in RD kinases, is present in the activation loop. If phosphorylation occurs on this serine in PKD, interaction with the cysteine residue would be weak. It is known from crystallographic data that the activation loop plays an important role in substrate recognition and in the correct positioning of catalytic residues (7). Additional interactions mediated by phosphorylated Ser⁷⁴⁴ and Ser⁷⁴⁸ might promote structural changes in the catalytic site of GST-catPKD and induce a higher catalytic activity. Neither of these two sites lies in the highly conserved activation loop phosphorylation site consensus sequence T(F/L)CGT identified among the AGC family of kinases (47). Nevertheless, they may be important for the in vivo activation of PKD/PKCµ by diacylglycerol-independent pathways such as the G-mediated regulation of PKD in Golgi structure and function (17) or form the basis of the in vitro stimulation of PKD by dextran sulfate (23). Our results may be at variance with earlier studies (23) but emphasize the need for positive identification of phosphorylation sites rather than indirect evidence from site-directed mutagenesis studies.

One might envisage an alternative pathway for PKD activation where a protease-driven mechanism would be involved. Interestingly, one study has shown that stable PKCµ transfectants exhibit a reduced sensitivity to tumor necrosis factor-induced apoptosis (16), which reported that PKCµ is stimulated by tumor necrosis factor and promotes the activation of NF-B-dependent genes, counteracting apoptotic signals. More recently, it was shown that treatment of cells with various apoptosis-inducing agents caused a caspase-3-mediated proteolytic cleavage of PKCµ between the regulatory and catalytic domain (53). The caspase-3 cleavage site in PKCµ was determined by site-directed mutagenesis, but this site is not conserved in PKD. In the present report we show that PKD can be activated in vitro by limited proteolysis (Fig. 7). This irreversible activation could be due to the removal of the inhibitory regulatory domain and/or to the unmasking of new phosphorylation sites in the kinase domain. Indeed, autophosphorylation on Ser⁷⁴⁴ and Ser⁷⁴⁸, observed only with the isolated catalytic domain, could perhaps play a role in the activation of PKD by proteolysis that has the potential to generate free catalytic domain.

In conclusion, phosphorylation of PKD at particular sites may be an intricate mechanism for the selective control of its biological functions. More work is in progress on the hierarchy of the observed phosphorylations and their potential role for the association of PKD with other proteins. Finally, further studies will be needed to investigate whether other activating signaling pathways such as G subunits or caspase-mediated proteolysis lead to differential phosphorylation of PKD.

	ACKNOWLEDGEMENTS

We thank Sarah Vander Perre and Dominique Revets for expert technical assistance. We are especially grateful to Peter Parker (ICRF, London) for providing PKC and PKC constructs.

	FOOTNOTES

^* This work was supported by Interuniversitiare Attractiepolen Grant P4/26, Fonds voor Wetenschappelijk Onderzoek-Vlaanderen Grants G.0193.99 and 1.5.409.98, and European Biomed Program Grant BMH4-CT96-0300.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^§ Supported by Postdoctoral fellowship Grant IUAP P4/26.

Research Associate of the National Fund for Scientific Research (Belgium).

^** Research Director of the Fonds voor Wetenschappelijk Onderzoek- Vlaanderen.

Postdoctoral fellow of the Fonds voor Wetenschappelijk Onderzoek-Vlaanderen. To whom correspondence should be addressed. Tel.: 32-16-345-719; Fax: 32-16-345-995; E-mail: johan.vanlint@med.kuleuven.ac.be.

Published, JBC Papers in Press, April 12, 2000, DOI 10.1074/jbc.M001357200

² Transphosphorylation refers to a phosphorylation catalyzed by another kinase.

	ABBREVIATIONS

The abbreviations used are: PKD, protein kinase D; PKC, protein kinase C; PDBu, phorbol 12,13-dibutyrate; PS, phosphatidyl-L-serine; BES, (N,N-bis[2-hydroxyethyl]-2-aminoethanesulfonic acid); DMEM, Dulbecco's modified Eagle's medium; HPLC, high pressure liquid chromatography; GST, gluthatione S-transferase; PAGE, polyacrylamide gel electrophoresis; ESI-MS/MS, Electrospray Ionization-Tandem Mass Spectrometry.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES