Autophosphorylation of Dictyostelium myosin II heavy chain-specific protein kinase C is required for its activation and membrane dissociation.

Myosin II heavy chain (MHC)-specific protein kinase C (MHC-PKC) isolated from the ameba, Dictyostelium discoideum, regulates myosin II assembly and localization in response to the chemoattractant cAMP. cAMP stimulation of Dictyostelium cells leads to translocation of MHC-PKC from the cytosol to the membrane fraction, as well as causing an increase in both MHC-PKC phosphorylation and its kinase activity. MHC-PKC undergoes autophosphorylation with each mole of kinase incorporating about 20 mol of phosphate. The MHC-PKC autophosphorylation sites are thought to be located within a domain at the COOH-terminal region of MHC-PKC that contains a cluster of 21 serine and threonine residues. Here we report that deletion of this domain abolished the ability of the enzyme to undergo autophosphorylation in vitro. Furthermore, after this deletion, cAMP-dependent autophosphorylation of MHC-PKC as well as cAMP-dependent increases in kinase activity and subcellular localization were also abolished. These results provide evidence for the role of autophosphorylation in the regulation of MHC-PKC and indicate that this MHC-PKC autophosphorylation is required for the kinase activation in response to cAMP and for subcellular localization.

We have previously reported the isolation of a MHC 1 -specific PKC (MHC-PKC) from the ameba, Dictyostelium that phosphorylates Dictyostelium MHC specifically and is homologous to ␣, ␤, and ␥ subtypes of mammalian PKC (1,2). In vitro phosphorylation of MHC by MHC-PKC results in inhibition of myosin II thick filament formation (1) by inducing the formation of a bent monomer of myosin II, whose assembly domain is tied up in an intramolecular interaction that precludes the intermolecular interaction necessary for thick filament formation (3).
The MHC-PKC which is expressed during Dictyostelium development has been implicated in the increase in MHC phosphorylation observed in response to cAMP stimulation (1). We have recently found that elimination of MHC-PKC abolishes this cAMP-induced MHC phosphorylation, indicating that MHC-PKC is the enzyme which phosphorylates MHC in response to cAMP stimulation (4). MHC-PKC null cells exhibit a substantial myosin II overassembly in vivo, as well as aberrant cell polarization, chemotaxis, and morphological differentiation. Cells that overexpress MHC-PKC contain highly phosphorylated MHC. They show no apparent cell polarization and chemotaxis, and exhibit impaired myosin II localization (4). These findings establish that, in Dictyostelium, the MHC-PKC plays an important role in regulating the cAMP-induced myosin II localization required for cell polarization and, consequently, for efficient chemotaxis.
When cells of Dictyostelium are starved, they acquire the ability to bind cAMP to specific cell surface receptors and to respond to this signal by chemotaxis, which requires phosphorylation and reorganization of myosin II (5)(6)(7)(8)(9). That is, the myosin II, which exists as thick filaments, translocates to the cortex (9) in response to cAMP stimulation. This translocation is correlated with a transient increase in the rate of MHC as well as light chain phosphorylation (5,6,10). We have recently shown that cAMP exerts its effects on myosin II via the regulation of MHC-PKC (11). cAMP stimulation of Dictyostelium cells results in translocation of MHC-PKC from the cytosol to the membrane fraction, as well as increasing MHC-PKC phosphorylation and its kinase activity (11). We could also show that MHC is phosphorylated by MHC-PKC in the cell cortex, and this leads to myosin II dissociation from the cytoskeleton (11).
Members of the PKC family are composed of a single polypeptide consisting of an NH 2 -terminal lipid binding regulatory domain and an ATP-binding catalytic domain located in the COOH-terminal region of the protein (12,13). PKC is phosphorylated on multiple serine and threonine residues located in both domains (14 -17). Three clusters of autophosphorylation sites have been mapped in vitro in the PKC ␤II isoenzyme at the NH 2 terminus, the COOH-terminal tail, and the hinge region between the regulatory and catalytic domains (14). However, in vivo studies have indicated that the autophosphorylation sites are all localized at the COOH-terminal tail of PKC (18,19). Bond and co-workers (20) have replaced the three clusters of the autophosphorylation sites found in vitro by alanine residues. Altering the autophosphorylation sites in this way at the NH 2 -terminal region or at the hinge region did not affect the activity or subcellular localization of the kinase. However, replacing the autophosphorylation sites at the COOH-terminal tail with alanine residues resulted in inactive, Triton X-100-insoluble enzyme (20). Similar results have been reported by Su et al. (21), who found that deletion of 23 amino acids from the COOH terminus of PKC␣, which includes the autophosphorylation sites, caused total loss of catalytic activity. These results indicate that the PKC autophosphorylation sites located at the COOH-terminal tail play an important role in the regulation of the kinase in vivo.
MHC-PKC also undergoes phosphorylation in vivo, which consists of both autophosphorylation and phosphorylation by another kinase, presumably cGMP-dependent protein kinase.
These two kinds of phosphorylation involve different phosphorylation sites (11). Here we report that deletion of 33 amino acids from the COOH-terminal tail of MHC-PKC, which includes a cluster of 21 serine and threonine residues, abolished both MHC-PKC autophosphorylation in vitro and the cAMPdependent MHC-PKC autophosphorylation. This indicates that this portion of the protein contains the in vitro and in vivo MHC-PKC autophosphorylation sites. We further present data showing that MHC-PKC autophosphorylation plays an important role in the kinase activation and subcellular localization.

EXPERIMENTAL PROCEDURES
Cell Culture and Development-Dictyostelium discoideum strain amebas were grown in HL-5 medium (22), harvested at a density of 2 ϫ 10 6 cells/ml, washed twice in MES buffer (20 mM MES, pH 6.8, 0.2 mM CaCl 2 , 2 mM MgSO 4 ), and resuspended at a density of 2 ϫ 10 7 cells/ml to initiate development. Cells were shaken at 100 rpm at 22°C for 3.5 h. 5 mM caffeine was added to the suspension 30 min prior to the addition of cAMP.
Expression of MHC-PKC⌬ST-All DNA manipulations were carried out using standard methods (23). We used the expression vector pDXA-HY which contains the actin-15 promoter and allows the expression of proteins carrying a NH 2 -terminal His tag (24). pDXA-MHC-PKC⌬ST was constructed as follows. The vector pBS-MHCK (2), containing a 2.6-kilobase pair MHC-PKC cDNA clone, was digested with SmaI and SwaI, which deleted a fragment of 309 base pairs from the 3Ј of the MHC-PKC clone. The deleted fragment contains 99 base pairs of coding region and 210 base pairs of noncoding region. The deleted coding region is 33 amino acids in length and contains a cluster of 21 serine and threonine (ST) residues which are thought to be the MHC-PKC autophosphorylation domain (Fig. 1, ST domain). The resulting MHC-PKC fragment was named MHC-PKC⌬ST. This MHC-PKC⌬ST fragment was sequenced to confirm the deletion of the ST domain. It was cloned into pDXA-HY digested with SmaI. pDXA-MHC-PKC⌬ST was used for the transformation of MHC-PKC Ϫ cells (4) using calcium phosphate precipitate (25). Clones were selected on the basis of their resistance to G418 (Boehringer Mannheim) and screened using Western blot analysis (see below).
Purification of His-tagged MHC-PKC⌬ST-50 ml of 2 ϫ 10 6 cells/ml expressing MHC-PKC⌬ST were washed twice in 20 mM phosphate buffer (pH 6.5), and the cells were lysed in 1 ml of lysis buffer containing 20 mM HEPES (pH 7.5), 1% Triton X-100, 0.2% Nonidet P-40, 200 mM KCl, 5 mM ␤-mercaptoethanol, and protease inhibitor mix (2 mM phenylmethylsulfonyl fluoride, 200 M leupeptin, and 200 M pepstatin). The extracts were centrifuged in a microcentrifuge for 15 min at 4°C, and the supernatant was incubated with 50 l of a slurry of Ni ϩ -agarose beads (Qiagen) in 20 mM phosphate buffer, pH 6.5, and 200 mM KCl for 1 h at 4°C. The bead-protein complex was washed three times with lysis buffer, twice with lysis buffer containing 20 mM imidazole, and twice with lysis buffer containing 50 mM imidazole. The MHC-PKC⌬ST was eluted with 100 l of lysis buffer containing 150 mM imidazole and then eluted with 100 l of lysis buffer containing 250 mM imidazole.
Western Blot Analysis-Cells were developed for 4 h in shaking flasks as described above. Cells were washed in 10 mM Tris-HCl, pH 8.0, and 150 mM KCl and lysed in 50 mM Tris-HCl, pH 8.0, 20 mM sodium pyrophosphate, pH 6.8, 5 mM EDTA, 5 mM EGTA, 0.5% Triton X-100, and protease inhibitor mix. Protein was determined by the method of Peterson (26), and lysates were electrophoresed on sodium dodecyl sulfate-polyacrylamide gels (SDS-PAGE) (27). Western blots were probed with MHC-PKC polyclonal antibody (2), and the blots were developed using a horseradish peroxidase-coupled secondary antibody (Bio-Rad Laboratories). ECL was performed using a kit from Amersham Corp.
Autophosphorylation-Developed Ax2 cell suspensions containing 5 ϫ 10 6 cells/ml were added to an equal volume of ice-cold 2 ϫ lysis buffer (40 mM Tris-Cl, pH 7.5, 0.2% Nonidet P-40, 2 mM dithiothreitol, 10 mM EDTA, and protease inhibitor mix) and centrifuged for 5 min in a microcentrifuge at 4°C. The supernatant was precleared by incubation with 30 l of rabbit preimmune serum at 4°C for 1 h followed by incubation with Staphylococcus A cells at 4°C for 30 min. Staphylococcus A cells were centrifuged, and 10 l of MHC-PKC antibody were added to the supernatant and incubated at 4°C for 1 h, followed by incubation with 50 l of protein A-conjugated agarose (100 mg/ml) at 4°C for 1 h and then centrifuged for 1 min. Pellets were washed twice in 1 ϫ lysis buffer containing 1 mg/ml bovine serum albumin and once with 1 ϫ lysis buffer only. The protein A-MHC-PKC complex was resuspended in 200 l of phosphatase mix (20 mM Tris-HCl, pH 7.5, 200 M MnCl 2 , 1 mM dithiothreitol, 0.5 mM CaCl 2 , 0.04 mM EDTA, 300 mM KCl, and 5 units of alkaline phosphatase (Boehringer Mannheim)) and incubated at 37°C for 30 min. The phosphatase-treated protein A-MHC-PKC complex was washed twice in 1 ϫ lysis buffer and incubated in 200 l of phosphatase inhibitor mix (100 mM NaF, 200 M Na 3 VO 4 , 10 mM KH 2 PO 4 and protease inhibitor mix) for 30 min at room temperature, followed by two washes in 1 ϫ lysis buffer. The phosphatasetreated protein A-MHC-PKC complex was resuspended in 100 l of 20 mM Tris-HCl, pH 7.5, 8 mM MgCl 2 , 0.2 mM [␥-32 P]ATP and incubated at 22°C for 20 min and then centrifuged for 1 min. The pellets were washed in 1 ϫ lysis buffer and resuspended in SDS sample buffer and boiled for 5 min. The supernatants from a microcentrifuge spin were loaded on SDS-PAGE and analyzed using autoradiography and Phos-phorImaging. Autophosphorylation of MHC-PKC⌬ST was performed as described above using MHC-PKC⌬ST purified with Ni ϩ -agarose beads as described above.
Phosphorylation-Dictyostelium Ax2 and MHC-PKC⌬ST cells were developed and treated with caffeine as described above. Before and after the application of cAMP stimulus, 100 l of developed cells were added to 200 l of reaction mixture containing 0.2% Triton X-100, 8 mM MgCl 2 , 10 mM Tris-HCl (pH 7.5), and 0.2 mM [␥-32 P]ATP and incubated for 30 s at 22°C. The MHC-PKC and MHC-PKC⌬ST were immunoprecipitated and analyzed using SDS-PAGE. Densitometric scanning of the Coomassie Blue-stained gels was used to determine the relative amounts of immunoprecipitated MHC-PKC and MHC-PKC⌬ST. The amounts of 32 P incorporated into the proteins were determined using the PhosphorImager. Relative phosphorylation of MHC-PKC and MHC-PKC⌬ST was determined by dividing the values obtained with the PhosphorImager by the values obtained by scanning of the Coomassie Blue-stained gels. In vivo phosphorylation of MHC was carried out as described previously (5,6).
MHC-PKC Activity-This was assayed directly using the kinase extracted from the insoluble cell fraction. Following resuspension of 1 ϫ 10 7 developed Ax2 and MHC-PKC⌬ST cells in 1 ml of sonication buffer (10 mM Tris-HCl, pH 7.5, 50 mM KCl, and protease inhibitor mix), they were stimulated with 1 M cAMP and lysed by sonication using an ultrasonic cell disruptor (Microson) model XL with a small sized tip at 50% output power. The extract was spun in a microcentrifuge for 20 min at 4°C. MHC-PKC and MHC-PKC⌬ST were extracted from the insoluble fraction using sonication buffer containing 1% Triton X-100 and 0.5 M KCl. For kinase assay, the solubilized MHC-PKC or MHC-PKC⌬ST were incubated with LMM58 (0.5-1 mg/ml), 6 mM MgCl 2 , 0.2 mM [␥-32 P]ATP (500 cpm/pmol), 1 mM DTT for 10 min at 22°C on a rotator. Reaction was initiated by the addition of ATP and stopped by the addition of 5% trichloroacetic acid. The precipitated LMM58 were pelleted in a microcentrifuge after incubation for 15 min on ice, washed twice with 5% trichloroacetic acid, resuspended in 20 l of SDS-PAGE sample buffer, and electrophoresed on 7% SDS-PAGE gels. To determine incorporation of 32 P into LMM58, the gels were stained and destained, and the bands were cut out of the gels and counted in a scintillation counter in 5 ml of scintillation fluid. The amounts of MHC-PKC and MHC-PKC⌬ST in the cell extracts were determined using densitometric scanning of Western blots and normalized to the total amount of protein determined as described previously (26).
Biochemical Analysis of MHC-PKC and MHC-PKC⌬ST Distribution-Following resuspension of 1 ϫ 10 7 developed Ax2 and MHC-PKC⌬ST cells in 1 ml of sonication buffer (10 mM Tris, pH 7.5, 50 mM KCl, and protease inhibitor mix), they were lysed by sonication as described above, and the extract was spun in a microcentrifuge for 20 min at 4°C. The soluble fraction was immunoprecipitated with MHC-PKC antibody as described above. MHC-PKC and MHC-PKC⌬ST were extracted from the insoluble fraction using sonication buffer containing 1% Triton X-100 and 0.5 M KCl. The extract was spun in a microcentrifuge for 10 min at 4°C, and the solubilized MHC-PKC and MHC-PKC⌬ST were immunoprecipitated as described above. To quantify the amounts of MHC-PKC and MHC-PKC⌬ST in the soluble and insoluble fractions, the immunoprecipitated MHC-PKC and MHC-PKC⌬ST from both fractions was electrophoresed on 7% SDS-PAGE gels and the Coomassie Blue-stained gels were analyzed as described above.
Triton-resistant Cytoskeleton Analysis-Triton-insoluble cytoskeleton analysis was performed as described previously (28). Supernatant and cytoskeletal pellet fractions were resuspended in SDS-PAGE sample buffer, boiled for 5 min, and electrophoresed on 7% SDS-PAGE gels. The relative amounts of myosin II were determined by SDS-PAGE gel analysis as described above.

Expression of MHC-PKC⌬ST Protein in MHC-PKC Cell
Line-Phosphorylation of sites located at the COOH-terminal tail of several PKC is important for the regulation of the kinase activity and subcellular localization (18,20,21,29). MHC-PKC undergoes autophosphorylation in vitro and phosphorylation in vivo in response to cAMP stimulation, and this phosphorylation coincides with the activation of the kinase (11). The MHC-PKC autophosphorylation sites are thought to be located within the COOH-terminal tail of MHC-PKC, which contains a cluster of 21 serine and threonine residues (Fig. 1, ST domain) (1,2). To study the in vivo role of MHC-PKC autophosphorylation, we engineered an MHC-PKC in which the putative autophosphorylation domain was deleted (MHC-PKC⌬ST). The MHC-PKC⌬ST was expressed in Dictyostelium cells that were previously engineered to lack expression of MHC-PKC (so-called MHC-PKC Ϫ cells) (4).
MHC-PKC Ϫ cells transformed with the pDXA-MHC-PKC⌬ST construct (see "Experimental Procedures") expressed MHC-PKC⌬ST at 125-150% of the level of MHC-PKC in Ax2 cells (Fig. 2). The expressed MHC-PKC⌬ST migrated on SDS-PAGE with an apparent molecular mass of about 80 kDa (Fig.  2), fitting well with the predicted molecular mass of 80 kDa and indicating that the protein was not phosphorylated. In contrast, MHC-PKC migrated on SDS-PAGE of cell extracts of Ax2 with an apparent molecular mass of about 94 kDa (Fig. 2), although the predicted molecular mass of MHC-PKC is 84 kDa (2). These results are consistent with a migration of the autophosphorylated form of MHC-PKC on SDS-PAGE (1), indicating that the MHC-PKC was in its phosphorylated form, under these experimental conditions. In addition to the 94-kDa band of MHC-PKC, another band with an apparent molecular mass of 90 kDa was found (Fig. 2). This protein could be either partially phosphorylated form of MHC-PKC or a degradation product of MHC-PKC.
MHC-PKC but Not MHC-PKC⌬ST Underwent Autophosphorylation in Vitro-To find out whether the ST domain of MHC-PKC is indeed the kinase autophosphorylation domain, we studied the ability of MHC-PKC⌬ST to undergo autophosphorylation in vitro. To do this, we developed Ax2 and MHC-PKC⌬ST cells. The MHC-PKC was immunoprecipitated and the MHC-PKC⌬ST purified using Ni ϩ -agarose beads as described under "Experimental Procedures." To increase autophosphorylation, the proteins were treated with phosphatase prior to the autophosphorylation reaction as described under "Experimental Procedures." The phosphatase-treated MHC-PKC immunocomplexed to protein A-Sepharose and the MHC-PKC⌬ST-Ni ϩ -agarose complex were incubated with [␥-32 P]ATP and MgCl 2 , and the extent of the autophosphorylation was analyzed using autoradiography as described under "Experimental Procedures." On addition of ATP and MgCl 2 to MHC-PKC, it migrated with a molecular mass of 94 kDa (Fig. 3A) which is consistent with a molecular mass of the autophosphorylated form of MHC-PKC (1). Autoradiography revealed that MHC-PKC indeed underwent autophosphorylation (Fig. 3B), and similar results have been previously reported (1). MHC-PKC appeared as a doublet on SDS-PAGE (Fig. 3A), which may represent different extents of autophosphorylation. In contrast, addition of ATP and MgCl 2 to MHC-PKC⌬ST did not alter the migration pattern of the protein, and it migrated with a molecular mass predicted for nonphosphorylated truncated MHC-PKC protein (Fig. 3C). Autoradiography of MHC-PKC⌬ST revealed that the protein was unable to undergo autophosphorylation (Fig. 3D). These results indicate that the ST domain is the in vitro MHC-PKC autophosphorylation domain.
MHC-PKC and Not MHC-PKC⌬ST Is Phosphorylated in Response to cAMP Stimulation-We have previously reported that, in response to cAMP stimulation, MHC-PKC undergoes a phosphorylation which is composed of both autophosphorylation and phosphorylation in a cGMP-dependent manner (11). To address whether MHC-PKC⌬ST is able to undergo in vivo phosphorylation in response to cAMP stimulation, we stimulated Ax2 and MHC-PKC⌬ST cells with cAMP and a total lysate of amebas labeled with [␥-32 P]ATP. MHC-PKC and MHC-PKC⌬ST were immunoprecipitated and the levels of their phosphorylation were determined as described under "Experimental Procedures." MHC-PKC was transiently phosphorylated in response to cAMP stimulation (Fig. 4, A and B), with peak phosphorylation at about 40 s (Fig. 4B) (see also Dembinsky et al. (11)). In contrast, addition of cAMP to cells expressing the MHC-PKC⌬ST protein resulted in very low phosphorylation levels (Fig. 4, A and B), whose magnitude was similar to the basal level of MHC-PKC phosphorylation determined 120 s after cAMP stimulation (Fig. 4B). This low level of phosphorylation in the MHC-PKC⌬ST may be due to the phosphorylation of the protein by another kinase possibly cGMP-dependent protein kinase, as we recently suggested (11). The inability of MHC-PKC⌬ST to undergo both autophosphorylation in vitro (Fig. 3) and cAMP-dependent phosphorylation (Fig. 4), indicate that the in vitro and in vivo autophosphorylation sites of MHC-PKC are the same and are located within the ST domain. The finding of relatively high phosphorylation levels of MHC-PKC prior to cAMP stimulation (Fig. 4, A and B) is consistent with the results presented in Fig. 2, in which MHC-PKC from Ax2 extract migrated on SDS-PAGE with an apparent molecular mass consistent with migration of the kinase in its autophosphorylated form (1).

MHC-PKC⌬ST Protein Remains in the Cell Membrane
Regardless of cAMP Stimulation-We reported previously that, on cAMP stimulation, MHC-PKC translocates to the membrane, presumably as part of the kinase activation mechanism. This translocation coincides with the kinase phosphorylation (11). It was therefore of interest to study the localization properties of MHC-PKC⌬ST on cAMP stimulation.
Ax2 and MHC-PKC⌬ST cells were developed and treated with caffeine, stimulated with cAMP, and lysed using sonication, and the MHC-PKC and MHC-PKC⌬ST were immunoprecipitated from the soluble and the insoluble fractions using specific MHC-PKC polyclonal antibody (see "Experimental Procedures"). Fig. 5 shows that, prior to cAMP stimulation, about 30% of the MHC-PKC resided in the insoluble fraction, whereas cAMP stimulation was followed by a rapid transient association of up to about 60% of the MHC-PKC with the membrane fraction, as reported previously (11). In contrast, about 70% of MHC-PKC⌬ST remained in the membrane regardless of cAMP stimulation (Fig. 5). These results indicate that the MHC-PKC autophosphorylation mechanism is involved in the dissociation of the kinase from the membrane.
MHC-PKC, but Not MHC-PKC⌬ST, Responds to cAMP Stimulation by Increasing Its Kinase Activity-To examine the possible role of MHC-PKC autophosphorylation in the activation of the kinase, we studied the specific activity of MHC-PKC and MHC-PKC⌬ST in response to cAMP stimulation. Ax2 and MHC-PKC⌬ST cells were stimulated with cAMP, and the kinase was solubilized from cell membranes and assayed for kinase activity as described under "Experimental Procedures." Although Dictyostelium contains several myosin heavy chain kinases, all except MHC-PKC are located in the cytosol (for review, see Tan et al. (30)). Furthermore, cells in which the MHC-PKC was eliminated do not show MHC phosphorylation activity in their membranes (4). Accordingly, all subsequent kinase assays were performed on MHC-PKC or MHC-PKC⌬ST that were solubilized from the cell membrane fraction. Fig. 6 shows that cAMP stimulation of Ax2 cells resulted in a transient increase in membrane-associated MHC-PKC kinase activity as reported previously (11). These cAMP-stimulated increases in MHC-PKC activity coincided with the cAMPstimulated membrane-association (Fig. 5) and phosphorylation (Fig. 4) of MHC-PKC, suggesting that these processes are linked to each other and may be required for the activation of MHC-PKC, as was proposed previously (11). In contrast, cAMP stimulation of MHC-PKC⌬ST did not increase the kinase activity (Fig. 6). MHC-PKC⌬ST showed a basal level of MHC kinase activity with magnitude similar to that shown by MHC-PKC isolated from nonstimulated Ax2 cells (Fig. 6). These results indicate that MHC-PKC autophosphorylation plays a role in the cAMP-dependent activation of the kinase. Interestingly, MHC-PKC⌬ST activity decreased 60 s after cAMP stim- ulation, reaching a level of activity similar to that of MHC-PKC determined 120 s after cAMP stimulation. The decrease in MHC-PKC⌬ST activity indicates that, in addition to autophosphorylation, there is another mechanism(s) that regulate the activity of MHC-PKC. One such possible mechanism is a cGMPdependent phosphorylation (11).

MHC-PKC⌬ST Cells Exhibit Nonphosphorylated Highly Triton-insoluble Myosin II-We reported previously that MHC-
PKC is the kinase that phosphorylates MHC in response to cAMP stimulation (4). We now tested how the cAMP-induced MHC phosphorylation is affected by the inability of MHC-PKC⌬ST to undergo autophosphorylation and to increase its activity in a cAMP-dependent manner. As shown in Fig. 7A Ax2 cells, but not MHC-PKC⌬ST cells, responded to cAMP stimulation by an increase in MHC phosphorylation. These results indicate that deletion of the autophosphorylation domain of MHC-PKC resulted in elimination of the cAMP-dependent MHC phosphorylation. These results are consistent with the finding that the MHC-PKC autophosphorylation plays an important role in the regulation of MHC-PKC and, consequently in the regulation of myosin II. Unstimulated developed Ax2 and MHC-PKC⌬ST cells exhibited the same basal levels of MHC phosphorylation (Fig. 7A).
MHC phosphorylation regulates the distribution of myosin II within the cell (4 -6, 11). In order to study what happens to the localization properties of myosin II in the MHC-PKC⌬ST mutant cells when there is no cAMP-dependent increase in the MHC-PKC⌬ST activity, we isolated cAMP-stimulated actinenriched Triton-insoluble cytoskeletons (see "Experimental Procedures"). 33% of the myosin II was insoluble in unstimulated developed Ax2 cells (Fig. 7B). Addition of cAMP resulted in a rapid accumulation of myosin II associated with the Triton-insoluble cytoskeleton (up to 65%), followed by an increase in myosin II solubility (Fig. 7B), as found previously (5,6,11). However, about 50% of the myosin II was already insoluble in unstimulated developed MHC-PKC⌬ST cells. Addition of cAMP resulted in a gradual increase in myosin II insolubility (up to 70%). In contrast to wild type cells, the increase in myosin II insolubility was not followed by a decrease in its insolubility. These findings are similar to those obtained for MHC-PKC Ϫ cells (4). DISCUSSION Previous studies have suggested a correlation between PKC activation and its autophosphorylation (20,(31)(32)(33)(34). The autophosphorylation sites located at the COOH-terminal tail of several PKC are important in the regulation of the kinase activity and subcellular localization (18,20,21,29). Here we have examined the functional role of autophosphorylation of MHC-PKC using deletion and biochemical analyses. Our results indicate that (i) it is the ST domain which is autophosphorylated in vitro and in vivo in response to cAMP stimulation and (ii) that MHC-PKC autophosphorylation plays a role in its partition into the cytosol and its activation in vivo.
We have previously shown that cAMP stimulation of Dictyostelium cells resulted in several changes in MHC-PKC behavior (11). First, cAMP stimulation causes MHC-PKC to translocate to the membrane. Second, translocation coincides with increases in MHC-PKC phosphorylation. A third change is an increase in MHC-PKC activity. The association of MHC-PKC with the membrane is necessary for MHC-PKC activation since the cytosolic MHC-PKC has a very low kinase activity. There are two types of MHC-PKC phosphorylation, which are different in their extent and their sites; autophosphorylation, which may occur at the membrane and accounts for most of the MHC-PKC phosphorylation, and cGMP-dependent phosphorylation, possibly via cGMP-dependent protein kinase which may take place in the cytosol. The two different phosphorylations occur on different serine and/or threonine residues in MHC- PKC (11).
Here, deletion of the ST domain abolished the in vitro MHC-PKC autophosphorylation. However, in vivo MHC-PKC⌬ST still exhibits a low level of phosphorylation which may represent cGMP-dependent phosphorylation. These results are consistent with our previous report that MHC-PKC and MHC-PKC⌬ST are phosphorylated in vitro by cGMP-dependent protein kinase (11). A similar phosphorylation of PKC by a heterologous kinase and its involvement in the regulation of PKC has also been reported for PKC␣ (35), however the identity of this kinase is unknown. We suggested previously that a plausible candidate for a PKC kinase is a cGMP-dependent protein kinase (11). Phosphorylation of MHC-PKC by another kinase may be required for the kinase translocation to the membrane which is required for the kinase activation, whereas the MHC-PKC autophosphorylation may be required for its activation and membrane dissociation.
Previous studies with mammalian PKC have shown that elimination of the COOH-terminal autophosphorylation sites fully inactivates the kinase (21,29). In contrast, upon deletion of the ST domain the MHC-PKC retained its basal kinase activity and lost the cAMP-dependent increases in its activity. These findings may indicate that the expressed protein is folded properly and the absence of cAMP-dependent increases in the kinase activity was a result of the ST domain deletion, it further suggests that this domain plays a key role in the cAMPdependent kinase activation. It is noteworthy that the basal level of MHC-PKC⌬ST activity decreased 60 s after cAMP stimulation. This may result from phosphatases removing the phosphates that were incorporated into the MHC-PKC⌬ST by cGMP mechanism thereby decreasing the kinase activity.
Deletion of the ST domain results in kinase that is unable to partition into the cytosol, indicating that the autophosphorylation of the ST domain plays a role in the dissociation of the kinase from the membrane. The seemingly contradictory results that autophosphorylation is required for both kinase activation and membrane dissociation may be explained as follow: the cluster of the 21 serine/threonine residues is autophosphorylated in two steps. In the first step only a fraction of the sites is autophosphorylated and this autophosphorylation is required for the activation of the kinase in response to cAMP. In the second step the remaining sites are autophosphorylated and this phosphorylation is required for dissociation of the kinase from the membrane. We are currently attempting to express MHC-PKC proteins in which the autophosphorylation sites are randomly converted to alanine residues. Experiments with these altered MHC-PKC proteins will enable us to explore the in vivo role of the different autophosphorylation sites.
The myosin II Triton-solubility and phosphorylation characterization in MHC-PKC⌬ST are similar to that in the mutant from which the MHC-PKC was eliminated (4). The inability of MHC-PKC⌬ST to increase its activity in response to cAMP stimulation is also reflected in the state of MHC phosphorylation and its Triton-solubility; MHC in MHC-PKC⌬ST cells is not phosphorylated in response to cAMP stimulation. These results are consistent with the findings that the cAMP-dependent MHC phosphorylation is carried out by MHC-PKC (4). If the enzyme does not respond catalytically to cAMP stimulation, the cell cannot respond to cAMP stimulation by phosphorylating the MHC.
The aberrant cAMP-dependent myosin II Triton-solubility in MHC-PKC⌬ST cells is consistent with the finding that it is through MHC phosphorylation that myosin II distributed throughout the cell in response to cAMP stimulation. MHC-PKC⌬ST and MHC-PKC cells (4) contain highly Triton-insolu-ble myosin II. This contrasts with wild type cells in which cAMP stimulation resulted in translocation of myosin II to the cortex, followed by dissociation of myosin II from the membrane fraction (5,6,11). In the two mutant cell lines, cAMP stimulation resulted in a gradual increase in myosin II association with the cytoskeleton with no apparent dissociation. The simplest explanation for these results is that, in unstimulated MHC-PKC⌬ST cells, the absence of MHC-PKC drives myosin II molecules into filaments in vivo and that these filaments have high affinity for the cortical cytoskeleton. cAMP stimulation, additional myosin II translocates to the cytoskeleton and, subsequently, cannot dissociate from it because of the presence of catalytically inactive MHC-PKC⌬ST. The observed phenotype of MHC-PKC⌬ST cells described here is not a result of the expression of MHC-PKC⌬ST per se but rather results from the absence of the ST domain; this is indicated by previous experiments in which we engineered cells expressing MHC-PKC under the same actin promoter, and the resulting cells had a wild type phenotype.
The detailed mechanism by which autophosphorylation could control MHC-PKC activity remains to be elucidated. One possibility is that the autophosphorylation triggers a conformational change that is important in relieving inhibition of MHC-PKC pseudosubstrate prototope similar to PKC (12). Supporting this hypothesis, the putative substrate (and pseudosubstrate) binding site of PKC is thought to be located near the COOH-terminal autophosphorylation sites (36,37). This model is further supported by the observation that the catalytic fragment of PKC retains histone kinase activity, even though it can no longer autophosphorylate (16). This suggests that removal of the regulatory domain eliminates the requirement for COOH-terminal autophosphorylation. Alternatively, this autophosphorylation may be critical for enzyme-substrate recognition or for interaction of MHC-PKC with regulatory factors.
A model consistent with both our data and previous data is that MHC-PKC is first synthesized as an inactive precursor that is cytosolic (11). The kinase is then recognized by a cGMPdependent protein kinase, which phosphorylates it, possibly at the C-2 and C-4 domains. This phosphorylation may be required for the kinase translocation to the cell membrane. The membrane-bound enzyme is then stimulated by autophosphorylation at the ST domain. The first step of autophosphorylation may activate the MHC-PKC causing MHC phosphorylation, followed by a second step of autophosphorylation resulting in a decrease in the enzyme's membrane affinity so that it partitions into the cytosol. This returns the enzyme to its basal state.