| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 281, Issue 42, 31627-31637, October 20, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
via Its Interaction with Protein Kinase C*
From the Biosignal Research Center, Kobe University, Rokkodai-cho 1-1, Nada-ku, Kobe 657-8501, Japan
Received for publication, July 24, 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
PKC directly associated with DGK through its accessory domain (AD), depending on Ca2+ as well as phosphatidylserine/diolein in vitro. Mass spectrometric analysis and mutation studies revealed that PKC phosphorylated Ser-776 and Ser-779 in the AD of DGK. The phosphorylation by PKC resulted in activation of DGK because a DGK mutant in which Ser-776 and Ser-779 were substituted with glutamic acid to mimic phosphorylation exhibited significantly higher activity compared with wild type DGK and an unphosphorylatable DGK mutant. Importantly, the interaction of the two kinases and the phosphorylation of DGK by PKC could be confirmed in vivo, and overexpression of the AD of DGK inhibited re-translocation of PKC. These results demonstrate that localization and activation of the functionally correlated kinases, PKC and DGK, are spatio-temporally orchestrated by their direct association and phosphorylation, contributing to subtype-specific regulation of DGK and DAG signaling. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
DAG is phosphorylated by DAG kinase (DGK) to produce phosphatidic acid, lowering the level of DAG and consequently down-regulating PKC activation. To date, at least 10 isozymes of mammalian DGKs have been cloned and divided into five groups based on their structural motifs (8-11). All DGKs have cysteine-rich regions, homologous to the C1A and C1B motifs of PKCs, in the regulatory domain at the N terminus and possess a conserved catalytic domain in the C terminus. Type I DGKs possess an EF-hand domain at their N terminus, making these isozymes responsive to calcium with slightly different affinity (10, 12). In addition, these isozymes have a recoverin homology domain, homologous to the recoverin family of neuronal calcium sensors, at the N terminus of their EF-hand motif. This region also participates in regulating the activation and conformational changes induced by Ca2+ in concert with calcium-binding sites, EF-hands, although it is incapable of binding Ca2+ (13). Among type I isozymes, DGK
and DGK, but not DGK
, have the same core structure as the PKC C1 domains and exhibit significant binding to phorbol 12,13-bidutyrate, whereas other DGKs do not (14, 15). Consistent with this, GFP-fused DGK, but not DGK, irreversibly translocates from the cytoplasm to the plasma membrane in response to TPA (16). In contrast, DGK is located on the plasma membrane even in the resting state (17).
The PKC family also consists of at least 10 subtypes that are classified into three groups based on the structure of their regulatory domain (3, 18). Conventional PKCs (cPKC; , 1, 2, and ) have two common regions, DAG-binding C1 domain and Ca2+-binding C2 domain, in the regulatory domain. Calcium, phosphatidylserine, and DAG are required for the activation of cPKCs. Novel PKCs are activated by DAG but are Ca2+-independent. Atypical PKCs are insensitive to DAG and are Ca2+-independent. Each subtypes has shown different enzymatic properties and distinct tissue and cellular distribution, suggesting specific functions for each PKC subtype (2, 19, 20), but their individual functions have not get been fully understood.
Among these PKC and DGK subtypes, both DGK and PKC possess the functional C1 domain and Ca2+-sensitive domain and are abundantly expressed in Purkinje cells of the cerebellum (21, 22), suggesting that DGK could be coupled with PKC to form an efficient functional and subtype-specific DAG signaling pathway. In fact, we previously reported that DGK regulates re-translocation of PKC, and the sequential translocation of these enzymes is temporally well regulated; purinergic receptor stimulation induced a rapid translocation of PKC to the plasma membrane followed by membrane targeting of DGK, then PKC returned to the cytoplasm, and finally DGK was re-distributed to its initial state (16). The molecular mechanism behind the spatio-temporal localization of PKC and DGK has not yet been clarified.
In addition to our previous study, several reports have recently described interactions between PKC and DGK (23-27). These findings led us to investigate direct interactions between DGK and PKC and the phosphorylation of DGK by PKC. We found that PKC directly binds to DGK dependent on Ca2+ and PS/DO, resulting in its activation via phosphorylation of Ser-776 and Ser-779. This is the first report to show that DGK activity is up-regulated by PKC-dependent phosphorylation.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-[32P]ATP and monosodium [32P]phosphate were products of MP Biomedicals, Inc (Irvine, CA). All the other chemicals used were of analytical grade. Anti-cPKC (C-19) antibody was obtained from Santa Cruz Biotechnology, Inc. The following antibodies were purchased: rabbit anti-HA (Zymed Laboratories Inc. and Upstate%20Biotechnology">Upstate Biotechnology, Inc.); anti-MBP serum (New England Biolabs); anti-His antibody (Amersham Biosciences); peroxidase-conjugated AffiniPure goat anti-mouse IgG and peroxidase-conjugated AffiniPure goat anti-rabbit IgG (Jackson ImmunoResearch).
Plasmid Construction—cDNA fragments of rat DGK were produced by a PCR with cDNA for rat DGK as the template. The sense and antisense primers used were as follows: for catalytic domain (CD) corresponding to 423-577 amino acids, 5'-AT GGA TCC ATC ATC CCA AGT AAG G (DGK-CD-F) and 5'-TA GGA TCC CTA GTT GTA TGG GAC CTG G (DGK-CD-R); for accessory domain (AD) corresponding to 578-788 amino acids, 5'-AT GGA TCC ATC ATG AAC AAC TAT TTC TCC A (DGK-AD-F) and 5'-TA GGA TCC TTA GTC CTT TGA ACG GC (DGK-AD-R); for 
AD corresponding to 1-577 amino acids, 5'-ATA CAA TTG ATG AGT GAC GGG CAA(rat-DGK-F) and DGK-CD-R, respectively. The PCR products of DGK were first subcloned into a TA cloning vector, pCRTM2.1 (Invitrogen). The regulatory domain (RD) of DGK corresponding to amino acids 1-412 was generated by SalI and BamHI digestion of plasmid BS465 (16) and blunted by Klenow fragment treatment. For the kinase domain (KD) corresponding to amino acids 377-788, plasmid BS465 was digested with SalI and SmaI and blunted with Klenow fragment. The RD and KD were then fused to FLAG tag by subcloning into pTB701-FLAG. For baculoviral expression of Histagged DGK, wild type DGK from BS465 was subcloned into pBlueBacHisA (Novagen). Expression plasmids for proteins fused to maltose-binding protein (MBP) were constructed by subcloning the corresponding cDNA fragments into pMAL-C2 (New England Biolabs). For MBP-fused wild type DGK in pBlueBacHisA, MBP with a BglII site at the 5' and 3' termini was first amplified by PCR from a pMAL-C2 vector and subcloned into the BamHI-BglII site in pBlueBacHisA. The sense primer was 5'-AT AGA TCT ATG AAA ATC GAA GAA GGT AAA C (BglII), and the antisense primer was 5'-ATA TCT AGA AGG ATC CGA ATT CTG AAA TC (BglII). The entire coding region of DGK was then added to the C terminus of MBP in pBlueBacHisA.
A plasmid containing DsRed2 cDNA was kindly donated by Dr. Miyawaki (Brain Science Institute, Riken). cDNA fragment encoding DsRed2 with a HindIII site in the 5'-terminal end and an EcoRI site in the 3'-terminal end was obtained by PCR using the above plasmid as a template. The sense and antisense primers used were 5'-TT AAG CTT ATG GCC TCC TCC GAG AAC GTC (HindIII) and 5'-TT GAA TTC CTA CAG GAA CAG GTG GTG GCG (EcoRI), respectively. As described previously, cDNA fragments for both DsRed2 and PKC were subcloned together into the EcoRI site in pTB701. All PCR products were verified by sequencing.
Cell Culture and Transfection—COS7 cells were purchased from the Riken Cell Bank (Tsukuba, Japan). The CHO-K1 cell strain was a gift from Dr. Nishijima (National Institute of Health, Tokyo, Japan). COS7 cells were cultured in Dulbecco's modified Eagle's medium, and CHO-K1 cells were cultured in Ham's F-12 medium (Nacalai Tesque, Kyoto, Japan) at 37 °C in a humidified atmosphere containing 5% CO2. Both media contained 25 mM glucose and were buffered with 44 mM NaHCO3 and supplemented with 10% fetal bovine serum (FBS; ICN Biomedicals, Inc), penicillin (100 units/ml), and streptomycin (100 µg/ml) (Invitrogen). Transient expression in COS7 cells was performed by electroporation (28), and CHO-K1 cells were transfected by lipofection using FuGENETM 6 Transfection Reagent (Roche Applied Science) as described previously (29).
Sf9 cells were cultured in EX-CELLTM 420 (JRH Biosciences), which was buffered with 44 mM NaHCO3 and supplemented with 10% FBS and 0.5% antibiotic/antimycotic (Invitrogen) in a temperature-controlled orbital shaker operating at 150 rpm and 27.5 °C. The FBS used was not heat-inactivated.
Measurement of Phosphatidic Acid (PA) and DAG Content—We measured PA based on the methods by Aragonés et al. (30). Briefly, cells were preincubated in phosphate-free medium for 90 min and then incubated with monosodium [32P]phosphate (100 µCi/ml) for an additional 90 min. Thereafter, the cells were stimulated by ATP, and the reaction was stopped by icecold PBS at appropriate point. Lipids including PA were extracted and separated by TLC. Spot of [32P]PA was measured by BAS2500 (FUJI Film, Tokyo, Japan).
To measure DAG, CHO-K1 cells normally cultured were used. After stimulation with ATP, lipids were extracted as described above, and the amount of DAG was determined by its conversion into [32P]PA by Escherichia coli DGK in the presence of [-32P]ATP. DGK assay was performed as described below.
Recombinant Protein Expression and Purification—For bacterial expression of MBP- and glutathione S-transferase (GST) fusion proteins, BL21 (DE3) pLys cells were transformed with various expression plasmids for recombinant proteins. MBP fusion proteins were expressed and purified according to the manufacturer's instructions (New England Biolabs). In this case, expression of recombinant proteins was induced by 0.3 mM isopropyl 1-thio--D-galactoside at 25 °C for 4 h. The cells were then harvested and lysed in column buffer (20 mM Tris-HCl (pH 7.4), 1 mM EDTA, 1 mM dithiothreitol, 200 mM NaCl, 1% Triton X-100, 20 µg/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride (PMSF)) with a handy sonic instrument (Tomy Seiko Co., Ltd.). After centrifugation at 10,000 x g for 1 h, fusion proteins were purified using an amylose resin affinity column. Similarly, expression of GST fusion proteins was induced by 0.1 mM isopropyl 1-thio--D-galactoside at 25 °C for 4 h. The cells were then harvested and lysed in buffer containing 20 mM Tris-HCl (pH 8.0), 1 mM EDTA, 1 mM dithiothreitol, 5 mM MgCl2, 250 mM sucrose, 1% Triton X-100, 20 µg/ml leupeptin, 1 mM PMSF. GST fusion proteins were purified on glutathione-Sepharose 4B resin (Amersham Biosciences).
Recombinant baculoviruses encoding His-DGK and MBP-DGK were prepared according to the manufacturer's instruction (Novagen) and were purified as described above. For purification of His-DGK, collected cells were homogenized in buffer containing 20 mM NaHPO4, 300 mM NaCl, 10 mM imidazole, 1%Triton X-100, 20 µg/ml leupeptin, 1 mM PMSF. After ultracentrifugation, the supernatant was incubated with nickel-nitrilotriacetic acid-agarose (Qiagen) at 4 °C overnight. After washing with 20 mM NaHPO4, 300 mM NaCl, 20 mM imidazole, the bound proteins were eluted with 20 mM NaHPO4, 300 mM NaCl, 250 mM imidazole. Recombinant pure proteins were dialyzed against PBS(-) three times and stored at -80 °C until use.
Confocal Microscopy Analysis—CHO-K1 cells expressing PKC-DsRed2 and GFP-DGK were plated onto a glass-bottomed culture dish (MatTek Corp., Ashland, MA) and incubated for 48 h before observation. The culture medium was replaced with Ringer's solution (135 mM NaCl, 5.4 mM KCl, 1 mM MgCl2, 1.8 mM CaCl2, 5 mM HEPES, 10 mM glucose (pH 7.3)). Translocation of PKC-DsRed2 and GFP-DGK were triggered by direct addition of 10 mM ATP into the Ringer's solution to give a final concentration of 1 mM. The fluorescence of the GFP was monitored with a confocal laser scanning fluorescent microscope (LSM 510 invert, Carl Zeiss, Jena, Germany) by 488-nm argon excitation using a 515-535-nm band pass barrier filter, whereas that of the DsRed2 was monitored by 543-nm HeNe excitation using a 560-nm long pass barrier filter. All experiments were performed at 37 °C.
In Vitro Binding and Phosphorylation Assay—FLAG-DGK was transfected into COS7 cells by electroporation. After 48 h, cells were harvested and lysed in TBS-T buffer (50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1% Triton X-100, 1 mM PMSF, 20 µg/ml leupeptin). The lysates were incubated with GST or GST-PKC bound to glutathione-Sepharose at 30 °C for 30 min in the presence of 1 mM EGTA or CaCl2. Then the beads were washed with TBS-T three times and analyzed for FLAG-DGK binding by immunoblotting. For the direct binding assay using purified recombinant proteins, association of GST-PKC with MBP-DGK was performed in either PBS containing EGTA, CaCl2, PS/DO, or both CaCl2 and PS/DO at 30 °C for 30 min. After addition of glutathione-Sepharose, the reaction was continued at 4 °C for an additional 2 h. Then the beads were washed with PBS three times. The bound proteins were eluted with 10 mM glutathione, resolved by SDS-PAGE, followed by immunoblotting using anti-MBP antibody.
In vitro phosphorylation assays were performed using MBP fusion proteins as a substrate. Purified MBP fusion proteins (1.2 µg) were incubated with purified GST-PKC (0.2 µg) in the presence of 8 mg/ml PS, 0.8 mg/ml DO, or 0.5 mM Ca2+ as the control with 0.5 mM EGTA. The reaction was terminated by adding sample buffer (186 mM Tris-HCl (pH 6.7), 15% glycerol, 9% SDS, 6% 2-mercaptoethanol, bromphenol blue) and boiling at 95 °C for 5 min. The phosphorylated proteins were detected by autoradiography using BAS2500 (FUJI Film). The bound proteins to the resin were analyzed for Western blotting using anti-GST and MBP antibody.
Immunoprecipitation and Immunoblotting—Rat cerebella extracts were homogenized in TBS-T by sonication on ice and then centrifuged at 10,000 x g for 10 min. Cleared lysates were incubated in the presence of 1 mM EGTA or CaCl2 at 30 °C for 30 min. After addition of anti-DGK immobilized to protein-A-Sepharose, the reaction was continued at 4 °C for an additional 2 h. Then the beads were extensively washed with TBS-T three times and boiled in SDS sample buffer at 95 °C for 5 min. The lysate and immunoprecipitates were separated on a 7.5% SDS-PAGE. The separated proteins were transferred to a polyvinylidene difluoride membrane and blocked with 5% skim milk in 0.01 M PBS containing 0.03% Triton X-100(PBS-T). The membrane was immunostained with anti-MBP or anti-GST polyclonal antibodies for 1 h at room temperature. After three rinses with PBS-T, the membrane was incubated peroxidase-labeled anti-rabbit IgG for 1 h at room temperature. After extensively washing with PBS-T, the immunoreactive bands were visualized using an enhanced chemiluminescence detection kit (Amersham Biosciences).
In Vivo Binding and Phosphorylation Assay—COS7 cells were co-transfected with PKC and FLAG-DGK and cultured for 24 h. The transfected COS7 cells were washed three times with serum-free DMEM without sodium phosphate and sodium pyruvate (Invitrogen) containing 0.1% BSA and then labeled for 1 h with [32P]phosphoric acid (0.33 mCi/ml) in phosphate-free DMEM, 0.1% BSA. After extensively washing, the cells were subsequently incubated in phosphate-free DMEM, 0.1% BSA with or without 1 µM TPA for 10 min. Cells were then harvested in PBS and lysed in homogenate buffer (50 mM Tris-HCl (pH 7.4), 250 mM sucrose, 10 mM EGTA, 2 mM EDTA, 20 µg/ml leupeptin, 1 mM PMSF, 1% Triton X-100) containing 1 mM Na3VO4 and 1 mM NaF by sonication. After centrifugation, FLAG fusion proteins were immunoprecipitated with anti-FLAG-M2-agarose at 4 °C for 1 h. The precipitants were subjected to SDS-PAGE followed by immunoblot analysis, and the phosphorylated proteins were visualized by autoradiography using BAS2500.
|
DGK Assay—COS7 cells were transfected with GFP, GFP-DGKS/A, GFP-DGK S/E, and GFP-wild type DGK. After 48 h, cells were harvested in homogenate buffer. DGK activity in the cell lysates was examined by octyl glucoside mixed micelle assay (31). 1-Steroyl-2-arachidonoyl-sn-glycerol was used as a substrate. The reaction was performed for 10 min in the presence of [-32P]ATP, and the radioactivity of phosphatidyl acid was detected using scintillation counter. The amount of the lysates subjected to kinase assay was determined by immunoblot analysis.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
to PKC—We previously demonstrated that DGK and PKC showed spatially similar but temporally different translocations to the plasma membrane after purinergic receptor stimulation in CHO-K1 cells (16), suggesting that the delay in translocation of DGK might determine the duration of PKC activation. This experiment, however, did not compare the translocations of the kinases in the same cell. Therefore, we examined whether this phenomenon could also be observed in a single CHO-K1 cell, using GFP-DGK and PKC-DsRed2 (Fig. 1A). The addition of 1 mM ATP elicited a rapid translocation of PKC-DsRed2 from the cytoplasm to the plasma membrane after 20 s of stimulation and was re-translocated to the cytoplasm within 2 min, returning to the resting state. In contrast, the translocation of GFP-DGK was initiated around 1 min and then re-translocated within 5 min.
|
PKC translocation. On the other hand, PA production was not detected just after the stimulation but was detected at 1 min, corresponding to the translocation of GFP-DGK to the plasma membrane (Fig. 1A). Afterward, the PA level was mostly reduced to the basal level. The production of DAG and PA was temporally coincident with movement of the kinases in the cells, suggesting that the translocation reflects the activation of the two kinases.
Interestingly, the two kinases co-localized at the plasma membrane for about 1 min (Fig. 1A), suggesting a possible interaction between the kinases. We then examined whether there were interactions between DGK and PKC. Using purified GST-PKC and lysates of COS7 cells expressing FLAG-DGK, we showed that FLAG-DGK was co-precipitated with GST-PKC in the presence of Ca2+ ions but not with GST (Fig. 2A). This interaction could not be detected in the absence of Ca2+ ions.
To confirm further this interaction, we tried to inversely precipitate PKC with His-DGK. Consistently, HA-PKC was precipitated with the purified His-DGK in the presence of Ca2+ ions, although HA-PKC was faintly precipitated in the presence of 10 mM Ca2+ (Fig. 2B). We next examined the dosedependent effect of Ca2+ ions on the binding of DGK to PKC, because both kinases have Ca2+-sensitive domains, and their conformation could be changed by Ca2+ (10-12, 32). As shown in Fig. 2C, the binding of HA-PKC to His-DGK was detected at 1 µM Ca2+ and increased in a dose-dependent manner to reach maximum binding at 1-10 mM of Ca2+. These results indicated that the binding of DGK to PKC depends on the Ca2+ concentration.
|
with PKC can be seen under the physiological conditions. For this purpose, we generated polyclonal antiserum against rat-DGK (33) and used rat brain cerebellum extracts for the pull-down assay. Endogenous PKC was associated with endogenous DGK when immunoprecipitated with anti-DGK antibody, whereas no DGK was detected when control serum was used (Fig. 2D). This association was clearly increased by 1 mM Ca2+. In contrast, when immunoprecipitated with anti-PKC antibody, we could not detect the association of DGK with PKC. This may be due to the lower expression level of DGK than PKC as shown in Fig. 2D, input lane. These data demonstrate that DGK also Ca2+-dependently associated with PKC in an endogenous expression level.
However, these experiments did not rule out the possibility that DGK indirectly associates with PKC via other proteins such as anchoring proteins because we used cell extracts. Therefore, we confirmed the direct interaction between DGK and PKC using purified proteins. As shown in Fig. 2E, purified MBP-DGK directly bound to purified GST-PKC at 100 µM and 1 mM Ca2+, indicating that DGK can directly associate with PKC, although there may be accessory proteins modulating the binding.
To identify which domain of DGK is necessary for the binding to PKC, we produced DGK RD and KD for the binding assay. As shown in Fig. 3B, the KD, but not the RD, bound to PKC. Two subdomains of the KD, CD and AD, were then tested. Fig. 3C shows that MBP-AD was precipitated with GST-PKC more efficiently than the CD, demonstrating that AD of DGK is mainly involved in the association with PKC. Interestingly, in the case of interaction of the KD or AD with PKC, the Ca2+ dependence was not so significant, suggesting the importance of the Ca2+-sensitive domains of DGK for the Ca2+-dependent association to PKC. To finally confirm the importance of AD for the association, we made a DGK mutant lacking AD and tested its binding to PKC. The mutant almost lost the ability to bind to PKC, revealing that AD is important for the binding to PKC (Fig. 3D).
Phosphorylation of DGK by Activated PKC—The association between PKC and DGK suggested that PKC phosphorylates DGK. In addition, some PKCs has been reported to phosphorylate some DGKs (23, 24-26, 34). Therefore, we investigated whether DGK is phosphorylated by PKC using purified recombinant enzymes. Recombinant GST-PKC purified from insect cells still possessed an intrinsic kinase activity when stimulated; the enzyme activity of GST-PKC was enhanced by about 2-fold in the presence of Ca2+ and PS/DO compared with its basal state in the presence of EGTA (data not shown). When MBP-DGK was incubated with GST-PKC, it was phosphorylated by activated PKC but not by the enzyme in its basal state (Fig. 4A). Importantly, phosphorylated MBP was not detected under these conditions. The phosphorylation was detected at 1 min, reaching a maximum at 10 min after stimulation in vitro (Fig. 4B).
|
by PKC could be seen in vivo. For this purpose, we first stimulated the CHO-K1 cells co-expressing PKC and DGK by ATP. However, we failed to detect significant phosphorylation on DGK by PKC under the conditions, suggesting that the association and phosphorylation induced by ATP are very transient. Indeed, translocations of PKC and DGK elicited by ATP occur transiently (Fig. 1). Therefore, TPA was used as a potent stimulator, because TPA induced irreversible translocation of both PKC and DGK to the plasma membrane (16). When COS7 cells expressing FLAG-DGK and a control vector were stimulated by 1 µM TPA, FLAG-DGK was slightly phosphorylated; the phosphorylation level of FLAG-DGK, which is the ratio of intensity of the autoradiography to that of the Western blot, was changed 2.31-3.41 (Fig. 4C). On the other hand, in COS7 cells co-expressing FLAG-DGK and PKC, phosphorylation of FLAG-DGK was clearly enhanced upon stimulation; the phosphorylation levels were 3.72 and 10.8 in the absence and presence of TPA, respectively. Additionally, TPA-induced autophosphorylation of PKC could be detected, implying that TPA works as a trigger of PKC activation in this system. These results indicated that DGK was phosphorylated in vivo and suggested that DGK associated preferably with activated PKC. Indeed, in vitro binding assay showed the association between DGK and PKC was significantly enhanced in the presence of both Ca2+ and PS/DO, compared with the presence of Ca2+ or PS/DO (Fig. 4D).
Furthermore, to determine which domain of DGK was phosphorylated by PKC, the phosphorylation of each domain was investigated (Fig. 5A). In the presence of PKC activators, phosphorylations of the RD and KD were enhanced by about 2.86- and 6.03-fold, respectively. Focusing on the KD, the AD was enhanced by about 11.3-fold and the CD by 6.03-fold, suggesting that the AD of DGK is the most prominent region phosphorylated by PKC. Consistent with this finding, there was no detectable phosphorylation of MBP-DGK lacking the AD (AD) in vitro (Fig. 5A), and in vivo phosphorylation of the mutant was reduced in COS7 cells co-expressing FLAG-DGK AD and PKC; the TPA-induced phosphorylation level of the mutant was 5.50, whereas that of full length was 10.8 (Fig. 4C).
To identify serine/threonine residues in the DGK AD phosphorylated by PKC, mass spectrometric analysis was performed. Comparison of mass spectra in the absence and presence of ATP showed that one serine (m/z 604.30) or two serines (m/z 644.29) were phosphorylated in a peptide fragment, SSFF-SLRRK, corresponding to 775-783 amino acids of DGK (Fig. 5B). In addition, it was confirmed that incorporation of one or two phosphates into the AD, by observation of product ions, was by neutral loss of phosphate in MS/MS (Fig. 5C). Ser-776 and Ser-779 are similar to the consensus phosphorylation sequence recognized by PKC. Interestingly, these serines are unique to DGK among type I DGKs and are well conserved in mouse and human DGK but not present in rat DGK and (Table 1). These facts suggest the importance of their phosphorylation to subtype-specific functions.
|
PKC on DGK Activity—We hypothesized that DGK activity might be regulated by PKC phosphorylation. Therefore, we constructed mutants of DGK and compared their activity. As shown in Fig. 6A, the S/A mutants, in which Ser-776 and -779 were substituted with alanines, showed almost the same activity as wild type. In contrast, the S/E mutant that mimics the properties of phosphoserine by substitution of Ser-776 and -779 to glutamic acids showed significantly higher activities compared with wild type DGK. These data indicate that phosphorylation of Ser-776 and -779 in DGK by PKC enhances DGK activity. Quantification analysis showed further clear differences (Fig. 6B).
|
PKC by Overexpression of DGK Accessory Domain—To clarify the physiological significance of association of the DGK AD with PKC and its subsequent phosphorylation, we investigated the effect of overexpressed GFP-DGK AD on re-translocation of PKC-DsRed2. PKC-DsRed2 possessed kinase activity similar to that of PKC (data not shown). Treatment of the cells expressing PKC-DsRed2 with ATP induced rapid translocation of PKC-DsRed2 to the plasma membrane at 20 s as seen in Fig. 1 (Fig. 7B). However, when PKC-DsRed2 and GFP-DGKAD were simultaneously expressed in the same CHO-K1 cell, the membrane localization of PKC-DsRed2 lasted for at least 4 min (Fig. 7A). It is noteworthy that GFP-DGK AD also showed a slight translocation to the plasma membrane. Quantification of the data showed that co-expression of GFP-DGK AD inhibited the re-translocation of PKC-DsRed2 (Fig. 7B). This inhibitory effect of GFP-DGK AD was similar to that of DGK inhibitor treatment (16). These results indicate that the association of DGK with PKC and its phosphorylation leading to up-regulation of DGK activity contribute to the spatio-temporal regulation of PKC activity in vivo.
Interaction of PKC with DGK—Finally, to examine subtype specificity of the interaction of PKC with DGK, we performed a pulldown assay using purified His-DGK and lysates of COS7 cells expressing HA-PKC, another subtype of cPKC (Fig. 8A). As a result, PKC also interacted with DGK Ca2+-dependently in vitro. Furthermore, DGK was also phosphorylated by PKC when COS7 cells co-expressing FLAG-DGK and PKC were treated with 1 µM TPA (Fig. 8B). These results suggest that Ca2+- and DAG-regulated PKC also regulate DGK by association and phosphorylation.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
PKC directly bound to DGK in a calcium- and PS/DO-dependent manner, and the AD of DGK was necessary for the binding (Figs. 2 and 3). The AD was also a major region phosphorylated by PKC, and Ser-776 and Ser-779 were identified to be phosphorylated by PKC (Figs. 4 and 5). However, the two serines are not the only phosphorylation sites by PKC because the RD and CD were also phosphorylated to some extent (Fig. 5A). Specifically in the case of TPA stimulation, there seemed to be some phosphorylation sites in addition to Ser-776 and Ser-779 (Fig. 4C). We also demonstrated that phosphorylation of the two serines resulted in the activation of the lipid kinase (Fig. 6). These findings fit a model for regulation of the spatio-temporal localization and activation of PKC and DGK proposed in Fig. 9. In the basal state, where calcium and DAG levels are low, both PKC and DGK are localized in the cytoplasm. Upon receptor stimulation, PKC is first targeted to the plasma membrane by increased calcium ions and fully activated by DAG produced on the plasma membrane. DGK then moves to the plasma membrane. One of the reasons for the differences in onset of the translocation of the two kinases is their different sensitivities to calcium concentrations; PKC can translocate at 10 nM calcium ionophore A23187
[GenBank]
, whereas 1 µM A23187
[GenBank]
is necessary to trigger DGK translocation (data not shown). In addition to the increase in calcium concentration, the trapping of DGK by activated PKC appears to be an important mechanism for targeting DGK to the membrane. Indeed, AD of DGK itself, a major binding domain to PKC, showed accumulation on the membrane concomitant with PKC translocation (Fig. 7A), and a DGK mutant lacking the AD domain showed no translocation (data not shown). Consequently, the direct binding allows PKC to phosphorylate DGK, resulting in the full activation of DGK. Thus, DAG on the membrane is efficiently metabolized by the activated DGK, leading to attenuation of PKC activity and its re-distribution. Collectively, the process is a well organized feedback pathway to negatively regulate PKC activity.
|
|
is retained longer than PKC on the membrane after PKC is redistributed back to the cytoplasm. Phosphorylated DGK may bind to other proteins or lipids on the membrane. In fact, we previously showed DGK binds to lipids such as phorbol 12,13-dibutyrate (15), and Mérida and co-workers (35) reported DGK is activated by phosphatidylinositol 4,5-bisphosphate, suggesting the interaction of DGK with phospholipids. These interactions between DGK and lipids, including phospholipids or phosphatidic acid on the membrane, may account for the delay in re-translocation of DGK.
The physiological importance of the direct association and phosphorylation between PKC and DGK was supported by the following three findings. First, endogenous PKC and DGK were co-immunoprecipitated from cerebellum extracts inaCa2+-dependent manner (Fig. 2D). Second, when COS7 cells overexpressing both DGK and PKC were stimulated by TPA, binding and phosphorylation of DGK were confirmed (Fig. 4C). Finally, inhibition of binding and phosphorylation by the overexpression of the AD of DGK disturbed the normal re-translocation of PKC (Fig. 7). Thus, regulation of the localization and activity of PKC and DGK as described in Fig. 9 appears to function under physiological conditions. Specifically, the negative regulation of PKC by DGK may play important roles in long term depression and/or nonepisodic autosomal dominant spinocerebellar ataxia in the cerebellum because both PKC and DGK are abundantly expressed in the Purkinje cells, and PKC is involved in the induction of long term depression (36, 37) and spinocerebellar ataxia (38). However, some cells do not possess PKC. In this case, another cPKC such as PKC can regulate DGK by association and phosphorylation.
|
|
and PKC or PKC, but associations of other DGKs to PKCs have also been reported. For example, PKC binds to DGK
and phosphorylates the myristoylated alanine-rich C kinase substrate homology domain of DGK, resulting in its nuclear export (23).
PKC binds to and phosphorylates DGK
, contributing to the membrane translocation of DGK (34)). cPKC-dependent phosphorylation of serines in the pleckstrin homology domain negatively regulates the membrane translocation of DGK
(27). These results suggest that direct interactions between PKC and DGK with subsequent phosphorylation is a common mechanism to regulate the localization of DGK, contributing to subtype-specific regulation of the enzyme. Indeed, the phosphorylation sites in DGK, DGK, and DGK by the respective PKCs are unique among DGK subtypes. However, no other report has shown upregulation of DGK activity by PKC phosphorylation. In the case of DGK, unlike DGK, the phosphorylation of DGK by PKC inhibits the DAG kinase activity, resulting in potentiation of the PKC activity (24, 25). In terms of the function of DGK to terminate PKC activation, the up-regulation of DGK by PKC as shown in this study seems to be consistent with functional and effective coupling of DGK and PKC to regulate their localization and activity.
PKC is not the only binding partner for DGK. To date, interactions of DGK with Ras-GRP (39), 1-syntrophin (40), and Src (41) have been reported. In addition, Rho A binds to DGK (42), and DGK also interacts with Src (43, 44). Interestingly, Ras-GRP also has the C1 domain responsible for DAG binding and shows membrane translocation in response to signals (45). Furthermore, Ras-GRP has been shown to be phosphorylated by PKC (46). Thus, the formation of protein complexes of these enzymes, including DGK, PKC, and Ras-GRP, contributes to the regulation of DAG signaling on the membrane. Accordingly, regulation of spatio-temporal activity of these signaling molecules, especially if they have opposing functions such as PKC and DGK, is a key step to effectively transduce signals and localize second messengers to restricted areas. In the regulation of spatio-temporal activity of many enzymes, the direct association and phosphorylation (or dephosphorylation) that we show in this study may be important.
In conclusion, PKC activator-dependently binds to DGK and phosphorylates the subtype-specific serines, resulting in the activation of DGK and attenuation of PKC activity. The direct interaction and phosphorylation are mechanisms regulating spatio-temporal localization and activation of PKC and DGK, contributing to subtype-specific coupling of PKC and DGK and the regulation of DAG signaling on the membrane.
|
FOOTNOTES |
|---|
1 To whom correspondence should be addressed. Tel.: 81-78-803-5961; Fax: 81-78-803-5971; E-mail: naosaito{at}kobe-u.ac.jp.
2 The abbreviations used are: DAG, diacylglycerol; AD, accessory domain; BSA, bovine serum albumin; CHO, Chinese hamster overly; DGK, diacylglycerol kinase; DO, 1,2-diolein; KD, kinase domain; GST, glutathione S-transferase; MBP, maltose-binding protein; PBS, phosphate-buffered saline; PKC, protein kinase C; PMSF, phenylmethylsulfonyl fluoride; RD, regulatory domain; PS, phosphatidylserine; TPA, 12-O-tetradecanoylphorbol 13-acetate; GFP, green fluorescent protein; PA, phosphatidic acid; FBS, fetal bovine serum; DMEM, Dulbecco's modified Eagle's medium; HA, hemagglutinin; CD, catalytic domain; MS/MS, tandem mass spectrometry; ESI, nanoelectrospray ionization.
|
ACKNOWLEDGMENTS |
|---|
and for helpful discussions and Dr. Fumio Sakane and Dr. Hideo Kanoh for helpful discussions of our work. |
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
98.0) between 1207.59 of the precursor ion and 1109.63 of the product ion in the left spectrum, 1287.63 of the precursor ion and 1189.61 of product ion in the right spectrum, and between 1189.61 and 1091.64 of product ions in the right spectrum show that the neutral loss of phosphate(s) was observed in MS/MS. These indicate that the candidates include one or two phosphoserine residues, respectively. However, the rest of product ions was not enough to determine phosphorylation sites.
