| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 282, Issue 39, 28627-28638, September 28, 2007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
V5 Domain That Contribute to Its Insensitivity to Diacylglycerol*From the Department of Laboratory Medicine, Center for Molecular Pathology, Malmö University Hospital, Lund University, SE-205 02 Malmö, Sweden
Received for publication, March 15, 2007 , and in revised form, August 1, 2007.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
insensitive to the activator diacylglycerol, which presumably is caused by a masking of the diacylglycerol-binding C1a domain. Here we demonstrate that the diacylglycerol sensitivity of the PKC
isoforms also is suppressed by autophosphorylation of the V5 sites. To analyze conformational differences, a fusion protein ECFP-PKC-EYFP was expressed in cells and the FRET signal was analyzed. The analogous mutant with autophosphorylation sites exchanged for alanine gave rise to a substantially lower FRET signal than wild-type PKC indicating a conformational difference elicited by the mutations. Expression of the isolated PKC V5 domain led to increased diacylglycerol sensitivity of PKC. We identified acidic residues in the V5 domain that, when mutated to alanines or lysines, rendered PKC sensitive to diacylglycerol. Furthermore, mutation to glutamate of four lysines in a lysine-rich cluster in the C2 domain gave a similar effect. Simultaneous reversal of the charges of the acidic residues in the V5 and the lysines in the C2 domain gave rise to a PKC that was insensitive to diacylglycerol. We propose that these structures participate in an intramolecular interaction that maintains PKC in a closed conformation. The disruption of this interaction leads to an unmasking of the C1a domain and thereby increased diacylglycerol sensitivity of PKC. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
, I, II, and 
), novel (PKC
, 
, 
, and 
) and atypical (PKC
and 
) PKC isoforms. The classification is based on the sensitivity of the isoforms to the PKC-activating second messengers Ca2+ and 1,2-diacylglycerol. The classical isoforms respond to both activators whereas novel PKCs are insensitive to Ca2+ but activated by diacylglycerol. The atypical isoforms are not regulated by any of the activators (1-3).
In its inactive state, the PKC molecule is kept in a closed conformation by intramolecular interactions. This involves the binding of a pseudosubstrate motif (4), localized in the regulatory domain, to the substrate-binding site in the catalytic domain, thereby preventing the interaction of PKC with its substrates. To activate PKC second messengers interact with specific domains within the regulatory domain of the PKC molecule. Ca2+ binds the C2 domain and diacylglycerol interacts with one or both of the C1 domains. This generally leads to a translocation of the enzyme to cellular membranes and a concomitant conformational change resulting in the release of the pseudosubstrate from the catalytic domain and thereby an activation of the enzyme (5, 6). It is also clear that an isolated interaction of PKC with second messengers and membranes is not the sole pathway to activate PKC. There are numerous examples of interacting proteins that can either target the activated PKC to specific sites or can contribute to the direct activation of the enzyme (7-9).
The sensitivity to diacylglycerol varies considerably between the isoforms. Whereas PKC rapidly translocates to the plasma membrane, particularly the classical isoforms are less responsive to an isolated increase in diacylglycerol (10-17). This is at least partially caused by the different sensitivity of the C1 domains to diacylglycerol. For classical isoforms many studies both with purified protein in vitro and using live cells demonstrate that the C1a domain is the diacylglycerol-sensitive C1 domain while the C1b domain is either insensitive or displays a weaker binding (11, 12, 14, 18).
However, the difference in sensitivity of the C1 domains cannot entirely explain why the holo-enzymes respond differentially to diacylglycerol. One model has been suggested for the activation of PKC, which proposes that the diacylglycerol-binding site in the C1 domains is structurally masked in PKC (13). The site is putatively unmasked upon an interaction of the C2 domain with Ca2+ and cellular membranes. We have recently seen that either inhibition of the catalytic activity or mutation of the C-terminal autophosphorylation residues to alanine makes PKC respond to DAG (19). This has led us to hypothesize that the C-terminal V5 domain takes part in intramolecular interactions that contribute to the masking of the C1a domain. This study was designed to investigate what specific residues of the V5 domain that may mediate such an interaction.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
, PKCI, and PKCII, kinase-dead PKC (PKC K368R) and PKCEDM (PKC T638E/S657E) and PKCADM (PKC T638A/S657A) have been described previously (19-21). An expression vector encoding PKC lacking the V5 domain (PKC
V5) fused to EGFP was generated by amplification of the desired DNA sequence with PCR using PKCWT as a template. The PCR fragments were digested with BglII/SalI and ligated into these sites in the pEGFP-N1 vector. The QuikChange site-directed mutagenesis kit (Stratagene) was used to introduce the K197E/K199E/K209E/K211E (PKC(C2:4K/E)), D649A/D652A/E654A (PKC(V5:3D/A)), and D649K/D652K/E654K (PKC(V5:3D/K)) mutations in the PKC-EGFP sequence using the expression vector encoding PKCWT as a template. Using the same primers, (PKC(C2: 4K/E,V5:3D/K))-EGFP was generated. To generate a fusion protein with PKCWT fused to ECFP in the N-terminal and EYFP in the C-terminal a thymidine was introduced in the ECFP-C1 vector using the QuikChange site-directed mutagenesis kit to generate a suitable reading frame. The PKC-EYFP cDNA was digested with BglII/XbaI, and the fragment was inserted in the modified pECFP-C1 vector. To generate fusions with ECFP in the N terminus and EYFP in the C terminus, PKC cDNAs were excised from the EGFP vector by using BglII/SalI and inserted into the ECFP-EYFP vector. Using the QuikChange site-directed mutagenesis kit expression vectors encoding PKCI and PKCII were mutated in the autophosphorylation sites to alter threonine in the turn motif (PKCI Thr-661 and PKCII Thr-660) and serine in the hydrophobic site (PKCI Ser-642 and PKCII Ser-641) to alanine (PKCIADM and PKCIIADM). Expression vectors encoding V5EDM and V5ADM fused to ECFP were generated by amplification of the desired DNA sequence with PCR using PKCEDM and PKCADM vectors as templates. The PCR fragments were digested with BglII/SalI and ligated into these sites in the pEGFP-C1 vector. The primers are listed in Table 1. All constructs were sequenced to confirm that they contained the right mutations.
|
COS cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 100 international units/ml penicillin and 100 µg/ml streptomycin. For immunofluorescence 300,000 cells/35-mm cell culture dish were seeded on glass cover slips. For the in vitro kinase assay 1,000,000 cells were seeded on 60-mm cell culture dishes.
Cells were transfected 24 h after seeding in OptiMEM (Invitrogen) with 2 µl of Lipofectamine 2000 (Invitrogen) and 2 µg of DNA/ml OptiMEM according to the supplier's protocol.
Western Blot—Transfected SK-N-BE(2)C cells were washed twice in phosphate-buffered saline and lysed in radioimmune precipitation assay buffer (10 mM Tris-HCl, pH 7.2, 160 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% sodium dodecyl sulfate, 1 mM EDTA, 1 mM EGTA) containing 40 µl/ml protease inhibitors (Roche Applied Science). Lysates were centrifuged for 10 min at 14,000 x g at 4 °C. Proteins were electrophoretically separated on a 10% NuPAGE Novex Bis-Tris gel (Invitrogen) and transferred to a polyvinylidene difluoride membrane (Millipore). For detection, membranes were incubated with monoclonal anti-GFP (1:500) (Zymed Laboratories Inc.), polyclonal anti-phospho-PKC/II (1:1000) or anti-phospho-PKC (1:1000) (both Cell Signaling) followed by incubation with a horseradish peroxidase-labeled secondary antibody (1:5000) (Amersham Biosciences). Horseradish peroxidase was thereafter visualized using the SuperSignal system (Pierce) as substrate. The chemoluminescence was detected with a CCD camera (Fujifilm).
Immunoprecipitation—Cells were treated according to the protocol supplied with µMACS Epitope-Tagged Protein Isolation kit (Miltenyi Biotec). Transfected SK-N-BE(2)C neuroblastoma cells were washed twice in phosphate-buffered saline and lysed in lysis buffer, supplied with the kit, containing 40 µl/ml protease inhibitor (Roche Applied Science). Lysates were centrifuged for 10 min at 14,000 x g at 4 °C and incubated with anti-GFP-conjugated microbeads for 30 min. The immune complexes were recovered by applying the cell lysates on µColumns in the magnetic field of the µMACS Separator and then washed and eluted with buffers included in the kit.
Immunofluorescence—Cells were fixed with 4% paraformaldehyde in PBS for 4 min followed by permeabilization with 5% normal goat serum and 0.3% Triton X-100 in phosphate-buffered saline for 30 min. Thereafter cells were incubated with primary polyclonal anti-phospho-PKC/II (1:200) or anti-phospho-PKC (1:200) (both Cell Signaling) for 1 h followed by incubation with the secondary antibody Alexa Fluor 546-conjugated goat anti-rabbit (Molecular probes) diluted 1:800 in PBS for 1 h. Extensive washing was carried out between all steps and the cover slips were mounted on object slides.
Confocal Microscopy—Live Cells were examined by confocal microscopy on the day after transfection. The glass coverslips were washed twice with buffer H (20 mM Hepes, 137 mM NaCl, 3.7 mM KCl, 1.2 mM MgSO4, 2.2 mM KH2PO4, 1.6 mM CaCl2, 10 mM glucose, pH 7.4) and mounted on a heated stage of a Nikon microscope. The cells were examined with a Bio-Rad Radiance 2000 confocal system using a 60x lens (numerical aperture 1.4) with excitation wavelengths of 457, 488, and 543 nm and the emission filters HQ485/30, 515HQ30, and HQ545/40 for ECFP, EGFP, and EYFP, respectively. Images were captured every 5 or 10 s, and five images were taken before the addition of 1 mM carbachol or 100 µM DOG.
Fixed cells were examined using excitation wavelengths 488 nm for EGFP and 543 nm for Alexa 546. The emission filters were HQ485/30 for EGFP and HQ590/60 for Alexa 546.
FRET Analysis—SK-N-BE(2)C cells were seeded on 35-mm glass bottom culture plates (MatTek). Cells were transfected with vectors encoding PKC fused N-terminally to ECFP and C-terminally to EYFP. On the day after transfection, cells were washed and incubated in buffer H. The dish was put on the stage of an inverted fluorescence microscope (Olympus) in a 37 °C chamber. The fluorochromes were excited with a halogen lamp using the excitation filters (Chroma Technology) ET430/24x for ECFP and ET500/20x for EYFP. The fluorescence was captured with a CCD camera (Hamamatsu) using emission filters (Chroma Technology) ET470/24m for the ECFP signal and ET535/30m for EYFP. The FRET signal was detected using the ECFP excitation and EYFP emission settings. The bleed through of ECFP and EYFP into the FRET channel was calculated using cells expressing ECFP or EYFP alone and found to be 79% of the ECFP and 15% of the EYFP signal. After subtraction of background the FRET was calculated with the formula (FRETem-0.79xECFPem-0.15xEYFPem)/ECFPem using the Volocity image analysis software. The fluorescence signals were determined in an area encompassing 
30% of the extranuclear area of a cell.
Determination of PKC Activity—Transfected COS cells were washed twice in phosphate-buffered saline and lysed in buffer (20 mM Hepes, pH 7.4, 0.1% Triton X-100, and 40 µl/ml protease inhibitors), and lysates were centrifuged for 10 min at 14,000 x g at 4 °C. PKC activity was assayed in 100 µl containing 20 mM Hepes (pH 7.4), 100 µM [32P]ATP (Amersham Biosciences), 5 mM MgCl2, 50 µM substrate (synthetic peptide derived from myelin basic protein, Genescript), and 25 µl of cell lysate. Phosphatidylserine (140 µM), diacylglycerol (3.8 µM), and CaCl2 (100 µM) were included as indicated. EGTA (500 µM) was used to assay the activity in the absence of calcium. Reactions were started by addition of substrate, [32P]ATP, and MgCl2, carried out at 20 °C and interrupted after 20 min by pipetting 80 µl of the mixture onto a Whatman P-81 ion-exchange paper (1.5 x 1.5 cm). Basal kinase activity was measured by excluding lipids and calcium and adding 500 µM EGTA to the reaction mixture. The filter papers were washed 3 x 5 min in 5 ml of 0.4% phosphoric acid and transferred into scintillation vials containing 10 ml of scintillation fluid (PerkinElmer) and radioactivity was measured by scintillation counting. To estimate the activity caused by the exogenous PKC variants, the activity obtained with lysates from untransfected cells were deducted from the activity obtained with lysates from transfected cells, assayed under identical conditions.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Sensitive to Diacylglycerol—We have previously seen that stimulation of SK-N-BE(2)C neuroblastoma cells with the diacylglycerol analogue DOG induces a sustained translocation of catalytically inactive PKC and of a PKC mutant with the autophosphorylation sites exchanged for alanine. This contrasts the wild-type and autophosphorylated PKC which does not translocate upon application of DOG. These findings led us to hypothesize that the autophosphorylated V5 domain is involved in an intramolecular interaction hiding the diacylglycerol-binding C1a domain. To further establish the importance of the C-terminal V5 domain in maintaining PKC insensitive to diacylglycerol the translocation of an EGFP fusion of PKC with deleted V5 domain (PKCV5) was investigated. Cells expressing PKCV5-EGFP were stimulated with 100 µM DOG and the localization of the protein was examined by confocal microscopy. The PKC mutant responded with a sustained translocation upon stimulation with DOG (Fig. 1, D-F) contrasting PKCWT, which is insensitive to DOG (Fig. 1, A-C). Thus, the presence of the V5 domain is required to maintain PKC in a diacylglycerol-insensitive state.
|
I and PKCII with Autophosphorylation Sites Mutated to Alanine Have an Increased Sensitivity to Diacylglycerol—The phosphorylation sites in the V5 domain are considered to be essentially constitutively autophosphorylated in classical PKC isoforms. Given the homology among these isoforms we aimed at investigating whether autophosphorylation influences the diacylglycerol sensitivity of PKCI and PKCII as well. Expression vectors encoding EGFP fusions of PKCI and PKCII with the threonine of the turn motif and the serine of the hydrophobic site mutated to alanine (PKCIADM and PKCIIADM) were therefore constructed (Fig. 2A). Cells expressing wild-type PKCI or PKCII or the non-phosphorylated mutants, all fused to EGFP were stimulated with DOG and the localization of the protein was examined by confocal microscopy. PKCIADM and PKCIIADM started to translocate within 1 min and remained by the membrane throughout the entire experiment (Fig. 2, E-G and K-M). While PKCWT is almost completely insensitive to diacylglycerol PKCIWT and PKCIIWT in a few cells translocated to the plasma membrane upon stimulation with DOG (Fig. 2, B-D and H-J). However, the response is substantially weaker and more delayed than what was seen with the non-phosphorylated mutants. Taken together, the results demonstrate that autophosphorylation of the turn motif and the hydrophobic site function as a determinant for the diacylglycerol sensitivity for several classical PKC isoforms.
FRET Analysis Indicates That Diacylglycerol Sensitivity Correlates to a Conformational Change of PKC—Our previous data show that PKC can be rendered diacylglycerol-sensitive by several alterations including deletion of the pseudosubstrate or the V5 domain, alanine mutations of the autophosphorylation sites or by inhibition of the kinase activity. The pseudosubstrate presumably binds the substrate binding site in the catalytic domain maintaining the inactive PKC in a closed conformation. We have hypothesized that the diacylglycerol-binding site, the C1a domain, is masked in the inactive conformation. Modifications that more or less destabilize the closed conformation may consequently lead to an unmasking of the C1a domain and an increased sensitivity to diacylglycerol.
To test this hypothesis we took an approach with the aim to estimate conformational changes in PKC. This was done by constructing an expression vector encoding PKC with a cyan fluorescent protein (ECFP) fused to the N terminus and a yellow fluorescent protein (EYFP) fused to the C terminus (Fig. 3A). If these fluorophores are in the absolute vicinity of each other a FRET signal is generated. The signal is dependent on the distance between the fluorophores and a different conformation of PKC may lead to a changed distance between the N and C terminus of the enzyme and consequently a difference in the FRET signal.
First we investigated if the different ECFP-PKC-EYFP constructs behaved as expected when stimulated with carbachol and DOG. Carbachol stimulation leads to activation of phospholipase C resulting in a transient Ca2+ increase and a more sustained elevation of diacylglycerol levels. This results in a transient plasma membrane translocation of PKCWT and PKC with both autophosphorylation sites mutated to glutamate (PKCEDM). On the other hand, kinase inactive PKC (PKCKD) and PKC with the autophosphorylation sites mutated to alanine (PKCADM) respond with a sustained translocation. Upon carbachol addition both ECFP-PKCWT-EYFP and ECFP-PKCEDM-EYFP rapidly but transiently translocated to the plasma membrane while ECFP-PKCKD-EYFP and ECFP-PKCADM-EYFP responded with a sustained translocation (Fig. 3B). Addition of DOG did not affect the localization of ECFP-PKCWT-EYFP and ECFP-PKCEDM-EYFP whereas ECFP-PKCKD-EYFP and ECFP-PKCADM-EYFP both responded with a translocation to the plasma membrane (Fig. 3C). Thus, simultaneous fusion to variants of EGFP to both the N and C terminus of PKC does not influence the translocation pattern of PKC.
Having confirmed that the ECFP-PKC-EYFP fusion proteins respond as expected we set out to investigate whether the difference in sensitivity to diacylglycerol is reflected by an altered FRET signal. SK-N-BE(2)C cells expressing the ECFP/EYFP fusions of PKCWT, PKCKD, PKCEDM, and PKCADM were analyzed with fluorescent microscopy using a FRET filter setup. With this approach we were unfortunately not able to follow the signal during translocation. Epifluorescence did not allow for spatial discrimination between the plasma membrane and cytosol and the signal was too weak for analysis with confocal microscopy.
|
and non-phosphorylated PKC give rise to a substantially lower FRET signal than wild-type PKC and autophosphorylated PKC. This indicates that the PKC mutants that are sensitive to diacylglycerol also have a different conformation, which may explain the different sensitivity to diacylglycerol. To rule out the possibility that the change in FRET signal was due to degradation of the fusion proteins the expression of the proteins was analyzed by Western blot. All constructs were expressed at the expected size and no sign of increased degradation products could be detected for any of the constructs (Fig. 3E).
Coexpression of the V5 Domain Increases the Diacylglycerol Sensitivity of PKC—Taken together our data are congruent with a model in which the diacylglycerol sensitivity of PKC is limited by conformational restrictions, which depend on the V5 domain and the phosphorylation of the autophosphorylation sites. This could presumably be mediated by intramolecular interactions involving the V5 domain. If this is the case, breaking this interaction could lead to a more destabilized conformation and hence an increased diacylglycerol sensitivity.
To test this hypothesis constructs encoding two variants of the isolated V5 domain fused to ECFP were generated and the effect of the expression of these on the diacylglycerol sensitivity of PKC was studied. To investigate the role of phosphorylated autophosphorylation sites V5 constructs with these residues mutated to glutamate (V5EDM) or alanine (V5ADM) were generated. SK-N-BE(2)C cells were cotransfected with vectors encoding V5EDM-ECFP, V5ADM-ECFP or ECFP alone together with a vector encoding PKCWT-EYFP. Cells expressing both EYFP and ECFP were thereafter examined by confocal microscopy. The localization of PKCWT-EYFP was followed after addition of DOG (Fig. 4). In 32% of the cells coexpressing V5EDM-ECFP, PKCWT-EYFP responded to DOG with an immediate translocation and in 36% of the cells there was a more gradual translocation (Fig. 4, D-F and J). This contrasted the response in cells cotransfected with empty ECFP vector where PKCWT-EYFP was unresponsive in more than 70% of the cells (Fig. 4, A-C and J). In the remaining cells the translocation of PKCWT-EYFP was gradual. In only a minute fraction of the cells could an immediate translocation be observed. Thus, coexpression of the phosphorylated isolated PKC V5 domain enhances the diacylglycerol sensitivity of PKC.
|
V5ADM-ECFP) was examined. However, in cells coexpressing V5ADM-ECFP, PKCWT-EYFP responded in a similar manner to DOG as in cells coexpressing V5EDM-ECFP (Fig. 4, G-J). Both V5 constructs were properly expressed as determined by Western blot (Fig. 4K). Thus a negative charge on the autophosphorylation sites is not a prerequisite for an isolated V5 domain to destabilize the more closed conformation of PKC and making it sensitive to DAG. Rather it is other structures in the V5 domain that are important for disrupting a possible intramolecular interaction.
Mutation of Acidic Amino Acids in the V5 Domain Makes PKC Sensitive to Diacylglycerol—Syndecan-4 is a transmembrane heparan sulfate proteoglycan that can act as a coreceptor with integrins in the formation of focal adhesions. It has a lysine-rich variable region that directly binds to the V5 domain in PKC and thereby partly activates the enzyme (22). A putative explanation is that the interaction between syndecan-4 and the PKC V5 domain destabilizes the enzyme. It is conceivable that the lysine-rich cluster interacts with acidic residues in the V5 domain and there are three negatively charged residues (Asp-649, Asp-652, and Glu-654) positioned fairly close to each other in the PKC V5 domain. We therefore speculated that these residues take part in an intramolecular interaction that contributes to the closed conformation of the inactive PKC.
To test this hypothesis we first mutated the negatively charged amino acids Asp-652 and Glu-654 to alanines generating PKC(V5: 2D/A) (Fig. 5A). Cells expressing PKC(V5:2D/A)-EGFP were stimulated with DOG and the localization of the protein was examined by confocal microscopy (Fig. 5, B-D). In a majority of cells PKC(V5:2D/A) translocated to the plasma membrane within 5 min and remained there throughout the entire experiment (Fig. 5H). We also mutated the acidic amino acid Asp-649 to alanine, generating PKC(V5: 3D/A) to examine if this potentiated the response to DOG (Fig. 5A). PKC(V5:3D/A) translocated even more frequently within 1 min than PKC(V5:2D/A) even though there were also more cells that did not respond at all (Fig. 5, E-H).
Because we have previously seen that non-phosphorylated PKC is sensitive to DOG we wanted to rule out the possibility that the increased sensitivity to DOG was due to absent autophosphorylation caused by the mutation in the V5 domain. Cells transfected with the vector encoding PKC(V5:3D/A)-EGFP were lysed and proteins expressing EGFP were immunoprecipitated and analyzed by Western blotting using antibodies specific for phosphorylated turn motif (Thr-638) and hydrophobic site (Ser-657). Both the turn motif (Fig. 5I) and the hydrophobic site (Fig. 5J) were phosphorylated. The results demonstrate that the acidic amino acids in the V5 domain contribute to keeping PKC in a DAG-insensitive conformation.
|
—The previous results indicated that acidic residues in the V5 domain are involved in maintaining PKC in a closed conformation. This would presumably involve an interaction with basic residues in another part of the molecule. In this context it is of interest to note that phosphatidylinositol 4,5-bisphosphate (PI(4,5)P2) enhances the effect of syndecan-4 on PKC (23) and that PI(4,5)P2 binds a lysine-rich cluster in the C2 domain consisting of the residues Lys-197, Lys-199, Lys-209, and Lys-211 (24, 25).
We therefore hypothesized that the negatively charged residues in the V5 domain take part in an intramolecular interaction with the lysine-rich PI(4,5)P2-binding cluster in the C2 domain and generated vectors encoding PKC with the charge of these residues reversed (PKC(V5:3D/K) and PKC(C2:4K/E)) to study the importance of a proper charge at these sites (Fig. 6A).
SK-N-BE(2)C cells were transfected with the vectors and the localization of the mutants fused to EGFP following stimulation with DOG was examined by confocal microscopy. As for the corresponding alanine mutants, PKC with the reversed charge of acidic residues in the V5 domain responded with a translocation upon addition of DOG (Fig. 6, B-D and H). In all cells in the experiments, PKC(V5: 3D/K) was sensitive to DOG with translocation within 5 min. In nearly 80% of the cells there was a clear translocation within 1 min after stimulation.
PKC with the lysine-rich cluster in the C2 domain mutated, PKC(C2:4K/E), also responded with a translocation after addition of DOG (Fig. 6, E-H). In more than 50% of the cells PKC(C2:4K/E) translocated within 5 min. Neither mutation of the lysines in the C2 domain nor the acidic amino acids in the V5 domain abolished the phosphorylation of the turn motif (Fig. 6I) or the hydrophobic site (Fig. 6J), indicating that the protein is properly processed.
To further establish that the translocated mutants are phosphorylated an immunofluorescence analysis was performed. Because the antibodies directed toward phosphorylated turn motif and hydrophobic site will recognize endogenous PKCs we turned to COS cells to increase the signal of the transfected PKCs in relation to the signal from endogenous proteins. After treatment with 100 µM DOG a substantial amount of PKCADM, PKC(V5:3D/K), and PKC(C2: 4K/E), but not of wild-type PKC localized to the plasma membrane (Fig. 7) as was previously seen in SK-N-BE (2)C neuroblastoma cells. Antibodies directed toward the phosphorylated turn motif or hydrophobic site both recognized the wild-type PKC and the V5 and C2 mutants but not the nonphosphorylated PKCADM. The V5 and C2 mutants that localized to the plasma membrane were phosphorylated at both sites (Fig. 7, F, H, N, and P) demonstrating that the increased translocation of these isoforms is not due to a lack of phosphorylation.
|
displayed substantial activity in the presence of Ca2+ and phosphatidylserine. Diacylglycerol did not potentiate this activity nor did it activate PKC in the absence of Ca2+. Neither PKCKD nor PKCADM displayed kinase activity under any conditions. The V5:3D/K and C2:4K/E mutants were both active, albeit at a lower level than wild-type PKC. Both these mutants were active in the presence of diacylglycerol and phosphatidylserine even in the absence of Ca2+, contrasting wild-type PKC, further demonstrating that these variants have an increased sensitivity to diacylglycerol.
Simultaneous Mutations Reversing the Charge of the Acidic Residues in the V5 Domain and the Lysine-rich Cluster in the C2 Domain make PKC Insensitive to Diacylglycerol—If the charged residues investigated participate through interaction with each other in maintaining PKC in a closed conformation simultaneous introduction of the reversing mutations might restore this interaction. We therefore generated a vector encoding PKC-EGFP with both the acidic residues in the V5 domain and the lysine-rich cluster mutated (PKC(C2:4K/E,V5:3D/K)). SK-N-BE (2)C cells were thereafter transfected with the vectors and the localization of the mutant was followed with confocal microscopy (Fig. 8, A-C). In more than 85% of the PKC(C2:4K/E),V5:3D/K)-EGFP was insensitive to stimulation with DOG, which is a response similar to the response of PKCWT (Fig. 8D).
To confirm that the mutant had not completely lost the capacity to translocate, cells expressing PKC (C2:4K/E,V5:3D/K)-EGFP were stimulated with carbachol, either in the absence or presence of GF109203X (Fig. 8, E-J). In a fraction of the cells, the mutant responded with a transient translocation upon addition of carbachol, which was prolonged and enhanced in the presence of GF109203X. This resembles the response of wild-type PKC. However, addition of GF109203X did not make the mutant translocate upon application of DOG (not shown). Thus, PKC(C2:4K/E,V5:3D/K) seems to be even less prone than wild-type PKC to translocate upon stimulation with diacylglycerol.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
that conceivably participate in an intramolecular interaction maintaining the isoform in a closed conformation.
|
respond to diacylglycerol (12). The addition of PKC inhibitors have also been shown to enhance the translocation response of classical isoforms (17, 19, 28) further suggesting that modifications of the catalytic site will destabilize the closed inactive conformation of PKC.
Another putative intramolecular interaction takes place between the C1a and the C2 domains (29). For PKC, it is only the C1a domain and not the C1b domain that has the capacity to penetrate lipid bilayers and bind diacylglycerol (30). The C1a domain is presumably tethered to the C2 domain in the closed conformation, an interaction that is dependent on D55 in the C1a domain. The interaction is disrupted by Ca2+ and phosphatidylserine binding to the C2 domain, which thereby makes the C1a domain accessible for interactions with diacylglycerol in the membrane (29).
In addition, we have seen that autophosphorylation of the hydrophobic site and turn motif serves as a switch controlling the diacylglycerol sensitivity of PKC. The data in this study indicate that this is due to an altered conformation of PKC. Firstly, both the kinase-inactive variant and the mutant with the autophosphorylation sites substituted for alanine displayed a lower FRET signal when this was measured utilizing a fusion with ECFP at the N terminus and EYFP at the C terminus. This approach to study conformational changes of PKC has been used for PKC (31). Following addition of C1 domain-binding ligands there was an increase in the FRET signal, concomitant with a translocation of the PKC which presumably results in a more open conformation of the protein. Because of technical limitations, we were not able to analyze FRET signals for translocated proteins but it is conceivable that the diacylglycerol-sensitive mutants, which had a lower FRET signal, also have a more open conformation. With the constructs used, the strength in the FRET signal is dependent on the proximity of the N to the C terminus of the protein. Thus, when PKC obtains a more open conformation it seems as if the distance between the N and C termini of PKC increases while it decreases for PKC. This is conceivably related to the different setups of the regulatory domains of classical and novel isoforms with the pseudosubstrate being placed in the immediate N terminus of classical isoforms whereas it is between the C2 and the C1 domains in novel isoforms.
|
. This is congruent with a model in which disruption of an intramolecular interaction involving the V5 domain leads to a destabilization of the conformation.
It is becoming increasingly clear that the V5 domain serves an important regulatory function of PKC. If the C-terminal ten amino acids are removed from PKC its catalytic activity is lost (32). This may reflect the fact that the V5 domain, when the autophosphorylation sites are phosphorylated, positions itself on the top of the N-lobe of the catalytic domain (33). This is of importance for the activity of several members of the AGC family of protein kinases (34). Autophosphorylation is also a determinant for whether the V5 domain of PKCII interacts with PDK-1 (35) or Hsp70 (36), for the binding of PKC V5 to Hsp25 (37) and for the sensitivity of PKC to diacylglycerol (19). The PKCII V5 domain also participates both in the interaction with RACK1 (38) and for the inactive enzyme, in an intramolecular interaction with the C2 domain (39) but the role of autophosphorylation in these interactions is not settled. Thus, the V5 domain seems to have an important role in conferring the proper conformation and function to PKC.
In addition, the V5 domain is one of the regions of the PKC molecule with a high degree of variability between the isoforms and is thereby an interesting structure in terms of isoform-specific modulation. Chimeras between PKC and PKC with the V5 domain exchanged have shown that the C-terminal region is an important determinant of isoform-specific functions (40). Another example is the unique PDZ-binding domain in the C terminus of PKC (41), which mediates its interaction with PICK1 and confers to PKC its specific effects in cerebellar long-term synaptic depression (42). The isoform-specific interaction of PKC with syndecan-4 is also mediated via the V5 domain (22).
|
that is further enhanced in the presence of PI(4,5)P2 (23). It is in this context of interest to note that there is a lysine-rich cluster in the C2 domain presumably which both is involved in intramolecular interactions (43) and is a binding site for PI(4,5)P2 (24). These facts raised a hypothesis that perhaps the lysine-rich cluster in the C2 domain interacts with the acidic region in the V5 domain in the inactive PKC.
Our experiments with the acidic residues mutated to alanines or lysines are in line with such a model. Both mutations led to a diacylglycerol-responding PKC, resembling the effects of the autophosphorylation site mutants. Mutation of the lysines in the C2 cluster to glutamate also led to enhanced diacylglycerol sensitivity but not as marked as the V5 domain mutations. This may be related to the fact that the lysine-rich cluster has been shown to strengthen the binding of PKC to certain membranes (44, 45). Despite this role for the lysine-rich cluster the C2 mutation still resulted in a PKC variant with an enhanced tendency to translocate in response to diacylglycerol.
Reversing the charges of both the acidic V5 residues and the lysine-rich C2 cluster could be expected to cancel out each other if these sites interact with each other. The mutant with both sites mutated was indeed, like wild-type PKC, insensitive to diacylglycerol. In fact it did not respond to DOG even in the presence of GF109203X, which wild-type PKC does, although it translocated upon stimulation with carbachol indicating that it has not completely lost the capacity to relocate to the membrane. Nevertheless, the data show that reversing the charges of both the V5 and the C2 sites does not render a mutant functionally equivalent to wild-type PKC. However, the fact that the mutations cancel out each other in terms of diacylglycerol sensitivity is congruent with a model in which the two sites interact with each other and thereby contribute to the masking of the C1a domain.
The results highlight previously unrecognized structures in the V5 region of PKC that may be targets for factors that destabilize the closed conformation of PKC thereby promoting the binding of diacylglycerol to PKC. It can be speculated that proteins, or other molecules, that bind the V5 structure may together with diacylglycerol activate PKC. Because this part of the PKC molecule is highly variable between the isoforms this provides possibilities for isoform-specific regulation of PKC.
|
FOOTNOTES |
|---|
1 To whom correspondence should be addressed: Lund University, Center for Molecular Pathology, Entrance 78, 3rd floor, UMAS, SE-205 02 Malmö, Sweden. Tel.: 46-40-337404; E-mail: Christer.Larsson{at}med.lu.se.
2 The abbreviations used are: PKC, protein kinase C; DOG, 1,2-dioctanoylglycerol; ECFP, enhanced cyan fluorescent protein; EGFP, enhanced green fluorescent protein; EYFP, enhanced yellow fluorescent protein; FRET, fluorescence resonance energy transfer; PI(4,5)P2, phosphatidylinositol 4,5-bisphosphate; DAG, diacylglycerol; WT, wild type.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
