Protein Kinase C-ζ as a Downstream Effector of Phosphatidylinositol 3-Kinase during Insulin Stimulation in Rat Adipocytes

Insulin provoked rapid increases in enzyme activity of immunoprecipitable protein kinase C-ζ (PKC-ζ) in rat adipocytes. Concomitantly, insulin provoked increases in32P labeling of PKC-ζ both in intact adipocytes and during in vitro assay of immunoprecipitated PKC-ζ; the latter probably reflected autophosphorylation, as it was inhibited by the PKC-ζ pseudosubstrate. Insulin-induced activation of immunoprecipitable PKC-ζ was inhibited by LY294002 and wortmannin; this suggested dependence upon phosphatidylinositol (PI) 3-kinase. Accordingly, activation of PI 3-kinase by a pYXXM-containing peptide in vitro resulted in a wortmannin-inhibitable increase in immunoprecipitable PKC-ζ enzyme activity. Also, PI-3,4-(PO4)2, PI-3,4,5-(PO4)3, and PI-4,5-(PO4)2 directly stimulated enzyme activity and autophosphoralytion in control PKC-ζ immunoprecipitates to levels observed in insulin-treated PKC-ζ immunoprecipitates. In studies of glucose transport, inhibition of immunoprecipitated PKC-ζ enzyme activity in vitro by both the PKC-ζ pseudosubstrate and RO 31-8220 correlated well with inhibition of insulin-stimulated glucose transport in intact adipocytes. Also, in adipocytes transiently expressing hemagglutinin antigen-tagged GLUT4, co-transfection of wild-type or constitutive PKC-ζ stimulated hemagglutinin antigen-GLUT4 translocation, whereas dominant-negative PKC-ζ partially inhibited it. Our findings suggest that insulin activates PKC-ζ through PI 3-kinase, and PKC-ζ may act as a downstream effector of PI 3-kinase and contribute to the activation of GLUT4 translocation.


Insulin provoked rapid increases in enzyme activity of immunoprecipitable protein kinase C-(PKC-) in rat adipocytes. Concomitantly, insulin provoked increases in 32 P labeling of PKC-both in intact adipocytes and during in vitro assay of immunoprecipitated PKC-; the latter probably reflected autophosphorylation, as it was inhibited by the PKC-pseudosubstrate. Insulin-induced activation of immunoprecipitable PKC-was inhibited by LY294002 and wortmannin; this suggested dependence upon phosphatidylinositol (PI) 3-kinase. Accordingly, activation of PI 3-kinase by a pYXXM-containing peptide in vitro resulted in a wortmannin-inhibitable increase in immunoprecipitable PKC-enzyme activity. Also, PI-3,4-(PO 4 ) 2 , PI-3,4,5-(PO 4 ) 3 , and PI-4,5-(PO 4 ) 2 directly stimulated enzyme activity and autophosphoralytion in control PKC-immunoprecipitates to levels observed in insulin-treated PKC-immu-
noprecipitates. In studies of glucose transport, inhibition of immunoprecipitated PKC-enzyme activity in vitro by both the PKC-pseudosubstrate and RO 31-8220 correlated well with inhibition of insulin-stimulated glucose transport in intact adipocytes. Also, in adipocytes transiently expressing hemagglutinin antigentagged GLUT4, co-transfection of wild-type or constitutive PKC-stimulated hemagglutinin antigen-GLUT4 translocation, whereas dominant-negative PKC-partially inhibited it. Our findings suggest that insulin activates PKC-through PI 3-kinase, and PKC-may act as a downstream effector of PI 3-kinase and contribute to the activation of GLUT4 translocation.
The atypical protein kinase C (PKC), 1 PKC-, is ubiquitously expressed, but little is known about its activation or actions. This ignorance partly derives from the fact that PKC-is not activated by membrane-associated diacylglycerol (DAG) or phorbol esters, generally does not translocate appreciably from cytosol to membrane when activated, and is not depleted by prolonged phorbol ester treatment. Consequently, methods used to evaluate DAG-sensitive conventional (␣, ␤, and ␥) and novel (␦, ⑀, , and ) PKCs are not relevant to PKC-and other DAG-insensitive, atypical PKCs. Although not activated by DAG, PKC-is activated in vitro by phosphatidylserine and polyphosphoinositides, including D3-PO 4 derivatives of phosphatidylinositol (PI) (1,2). Because of its activation by polyphosphoinositides, PKC-has been suspected to operate downstream of PI 3-kinase; however, direct experimental evidence for this suspicion is lacking, particularly in intact cells. Since insulin increases total polyphosphoinositide levels (3)(4)(5)(6), probably largely through PI 3-kinase activation (7), we examined the possibility that insulin activates PKC-by a PI 3-kinasedependent mechanism. To this end, we assayed immunoprecipitable PKC-(a) following treatment of intact adipocytes with insulin in the presence and absence of PI 3-kinase inhibitors; (b) following PI 3-kinase activation in vitro by a pYXXMcontaining peptide; and (c) in response to polyphosphoinositides added directly to the assay of PKC-in vitro. Also, since PI 3-kinase appears to be required for insulin-stimulated GLUT4 translocation and subsequent glucose transport, we questioned whether PKC-, as an effector of PI 3-kinase, may play a role in this process. To this end, we used PKC-inhibitors and examined the effects of transiently transfected wildtype, constitutive, and dominant-negative PKC-in adipocytes co-transfected with hemagglutinin antigen (HA)-tagged GLUT4.

MATERIALS AND METHODS
As described (8), rat adipocytes were prepared from epididymal fat pads of 200 -250-g male Sprague-Dawley rats. For acute incubations, the cells were suspended and incubated in glucose-free Krebs-Ringer phosphate (KRP) buffer containing 1% bovine serum albumin with or without insulin (Elanco), tetradecanoylphorbol-13-acetate (TPA; Sigma), wortmannin (Sigma), LY294002 (Biomol), RO 31-8820 (Alexis), and/or myristoylated PKC-or PKA pseudosubstrate peptides (Quality Controlled Biochemicals) as indicated in the text. For overnight incubations, the cells were suspended in Dulbecco's modified Eagle's medium (Life Technologies, Inc.) containing 1% bovine serum albumin, 5 mM glucose, and other indicated treatments; after overnight incubations, the cells were transferred to glucose-free KRP for acute treatments. Following acute treatments, glucose transport was assayed by measurement of [ 3 H]2-deoxyglucose (2-DOG; 0.1 mM; NEN Life Science Products) uptake over 1 min as described (8), or PKC-enzyme activity was assayed as described below.
In some experiments, adipocytes were co-transfected by electroporation (as described by Quon et al. (9,10)) in the presence of (a) pCIS2 eukaryotic expression vector containing cDNA encoding HA-tagged GLUT4 (kindly supplied by Drs. Michael Quon  and dominant-negative forms of PKC-were described previously (11), and the constitutive form of PKC-was generated by mutating Ala 119 in the pseudosubstrate region to Asp 119 (i.e. GCC to GAC) using a Transformer Site-Directed Mutagenesis Kit from CLONTECH; the sequence of this construct was confirmed by sequence analysis) or (i) a dominantnegative mutant form of the ⌬p85 subunit of PI 3-kinase (⌬p85; kindly supplied by Drs. Wataru Ogawa and Masato Kasuga). These cells were incubated overnight in Dulbecco's modified Eagle's medium containing 25 mM Hepes, 200 nM (ϪN)-N 6 -(2-phenylisopropyl)-adenosine, and 5% bovine serum albumin to allow time for expression of the inserts. Cells were then washed and resuspended in glucose-free KRP medium and treated with or without 10 nM insulin for 30 min prior to the addition of 2 mM KCN and assessment of the translocation of HA-tagged GLUT4 to the plasma membrane, as described (9,10). This treatment was performed using anti-HA mouse monoclonal antibody (Berkeley Antibody Co.) and 125 I-labeled sheep anti-mouse IgG (second) antibody (Amersham Corp.) to measure cell surface content of expressed GLUT4 containing the exofacial HA epitope.
To measure PKC-enzyme activity, as described previously (11), the reactions were rapidly stopped by adding ice-cold KRP medium, and cells were washed and sonicated in Buffer A (20 mM Tris/HCl (pH 7.5), 0.25 M sucrose, 1.2 mM EGTA, 20 mM ␤-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride, 20 g/ml leupeptin, 20 g/ml aprotinin, 1 mM Na 3 VO 4 , 1 mM Na 4 P 2 O 7 , and 1 mM NaF). Homogenates were centrifuged at 500 ϫ g for 10 min to remove nuclei, debris, and the fat cake. Triton X-100 (1%), Nonidet (0.5%), and 150 mM NaCl were then added to the total cell lysate, and, after 30 min equilibration at 4°C, preimmune antiserum or polyclonal anti-PKC-antiserum (obtained from Life Technologies, Inc., or Santa Cruz Biotechnologies) (both antisera were raised against a nearly identical C-terminal sequence and gave comparable results) was added to samples containing 500 g of lysate protein. After overnight equilibration at 0 -4°C, immunoprecipitates were collected on protein AG-Sepharose beads by centrifugation, washed, and suspended in Buffer B (50 mM Tris/HCl (pH 7.5), 5 mM MgCl 2 , 100 M Na 3 VO 4 , 100 M Na 4 P 2 O 7 , 1 mM NaF, and 100 M phenylmethylsulfonyl fluoride). Suspensions were then incubated for 8 min at 30°C in 100 l of Buffer B containing 3-5 Ci of [␥-32 P]ATP (NEN Life Science Products), 50 M ATP, 4 g of phosphatidylserine, and 40 M [ 159 Ser]PKC-⑀(AA153-164)-NH 2 (Upstate Biotechnology, Inc.), a preferred substrate for PKC- (12). In some experiments, PI-3,4-(PO 4 ) 2 or PI-3,4,5-(PO 4 ) 3 (Matreya) or PI-4,5-(PO 4 ) 2 (Fluka) was added as a sonicated suspension to the assay. Reactions were stopped by addition of 10 l of 5% acetic acid. Aliquots of the reaction mixtures were spotted on p81 filter paper, washed in 5% acetic acid, and counted for 32 P radioactivity. Blank values (approximately 10 -15%) were determined from precipitates obtained with preimmune serum or from incubations conducted in the presence of 100 M PKC-pseudosubstrate and were subtracted from total cpm to determine PKC--specific cpm. Reaction rates were linear with respect to time (see Fig. 1, inset) and were markedly dependent upon addition of phosphatidylserine but not Ca 2ϩ or diolein. Insulin treatment did not influence either the amount of PKC-recovered in immunoprecipitates or the blank values. Approximately 50% of total cellular PKC was recovered in the immunoprecipitates, and this recovery was not improved by the addition of a 2-fold excess of antibody. As reported previously (11), PKC-immunoprecipitates contained no detectable PKC-␣, ␤, ␦, or ⑀, which are known to be present in rat adipocytes (13). The dependence of enzyme activity on phosphatidylserine suggested that protein kinase B and protein kinase N were absent from PKC-immunoprecipitates, since immunoprecipitable PKB and PKN activities are independent of phosphatidylserine. 2 In all experiments, control and treated cells were derived from the same batch of adipocytes, and all samples were subsequently assayed in parallel; consequently, results of each treatment were expressed relative to the corresponding mean control value, arbitrarily set at 1. An example of absolute cpm typically observed in control and insulinstimulated PKC-immunoprecipitates in an individual experiment is shown in the inset of Fig. 1.
In some assays of immunoprecipitated PKC-, exogenous substrate was omitted, and 32 P labeling of PKC-itself was determined after its resolution by SDS-polyacrylamide gel electrophoresis, electrolytic transfer to nitrocellulose membranes, and analysis in the Bio-Rad Molecular Analyst phosphorimager.
PKC isoforms were immunoblotted as described (11), except that polyclonal antibodies used for PKC-␤ 1 and PKC-␤ 2 were obtained from Santa Cruz Biotechnology. Blots were quantified in a Bio-Rad Molecular Analyst Chemiluminescence Imager. PI 3-kinase activity in adipocyte homogenates was measured as described previously (8).

RESULTS
Insulin provoked rapid increases in the enzyme activity recovered in PKC-immunoprecipitates prepared from total adipocyte lysates (Fig. 1). Increases were observed throughout a 20-min treatment period with 10 nM insulin, but the increases appeared to be biphasic, with peaks at 0.5-1 and 10 min, and were maximal at insulin concentrations of 1-10 nM (see Fig. 1, inset). In general, increases in immunoprecipitable PKC-enzyme activity at 10 min of 10 nM insulin treatment were approximately 2-3-fold, although they varied from 1.5-fold to 10-fold. This variability appeared to primarily reflect differences in basal activity, but in addition, the timing of the insulin-induced stimulatory peaks may have varied slightly in individual experiments.
As shown in Fig. 2, insulin-induced increases in immunoprecipitable PKC-were observed, not only in acutely incubated adipocytes but also in adipocytes that were incubated overnight prior to acute insulin treatment. As expected, the marked down-regulation of DAG-sensitive PKCs (␣, ␤1, ␤2, ␦, and ⑀) by overnight TPA (1 M) treatment failed to influence basal or insulin-stimulated immunoprecipitable PKC-enzyme activity and interestingly, at least in some experiments (e.g. see Fig. 2 showed that PKC-␣, -␤1, -␤2, -␦, and -⑀ can be dissociated from glucose transport effects of insulin in the rat adipocyte (also see below); and (c) documented that PKC-is activated by acute insulin treatment in adipocytes that are either used directly or first placed into primary culture and incubated overnight with or without TPA (see below).
To determine whether PKC-may be activated by a PI 3kinase-dependent mechanism, we used three approaches. First, we used two relatively specific, but structurally different, inhibitors of PI 3-kinase. In concentrations that inhibit PI 3-kinase and insulin effects on glucose transport, both LY294002 and wortmannin inhibited insulin effects on PKCactivation in intact adipocytes (Fig. 3); in contrast, these inhibitors did not inhibit PKC-enzyme activity when added directly to PKC-immunoprecipitates (not shown). Second, we used a peptide (DADS(pY)ENMDNP-NH 2 ) that, by virtue of its pYXXM motif, activates the SH2 domain of PI 3-kinase (see Ref. 14). Upon incubation of this peptide with control adipocyte homogenates, there was a dose-related activation of immunoprecipitable PKC- (Fig. 4); moreover, this activation was blocked by 100 nM wortmannin, confirming that it was caused by PI 3-kinase activation (Fig. 4). Note that PI 3-kinase activity in these homogenates was increased nearly 2-fold by pYXXMcontaining peptide and was fully blocked by LY294002 (Fig. 4). Third, we added polyphosphoinositides (PI-3,4-(PO 4 ) 2 , PI-3,4,5-(PO 4 ) 3 , and PI-4,5-(PO 4 ) 2 ) directly to control and insulin-stimulated PKC-immunoprecipitates. As shown in Fig. 5, each of these polyphosphoinositides, in doses found to be optimal in other experiments, stimulated enzyme activity of control PKC-immunoprecipitates but had a relatively small or no significant effect on insulin-stimulated PKC-immunoprecipitates; accordingly, these polyphosphoinositides obliterated or narrowed the difference in enzyme activity between control and insulin-stimulated immunoprecipitates. These findings suggested that PI 3-kinase is not only required for insulin-induced activation of immunoprecipitable PKC-in rat adipocytes but, through its lipid products, can account for this activation.
Certain evidence (8,15,16) suggests that a protein kinase distal to PI 3-kinase (8, 17) may be required for insulin effects on GLUT 4 translocation and glucose transport. Insulin effects on these processes are sensitive to PKC inhibitors but, in general, the required concentrations of these inhibitors (8,15,16) have exceeded those required to inhibit conventional and novel DAG-sensitive PKCs. Along these lines, it may be noted that bisindolemaleimides are potent inhibitors of conventional PKCs but inhibit PKC-only at relatively high concentrations (18,19), and this was presently found to be the case for the bisindolemaleimide derivative RO 31-8220, which inhibited recombinant forms of PKC-␣, -␤1, -␤2, -␥, -␦, -⑀, -, and -with EC 50 values of approximately 40, 20, 15, 15, 30, 100, 20, and 1000 nM, respectively (Fig. 6). Presently, we also found that the inhibition of enzyme activity in adipocyte PKC-immunoprecipitates in vitro closely matched the inhibition of insulinstimulated glucose transport observed in intact adipocytes in response to increasing doses of RO 31-8220 (Fig. 7); EC 50 values for inhibition of both processes were approximately 4 M. (The differences between EC 50 values of RO 31-8220 for inhibiting soluble preparations of recombinant PKC-versus PKC-immobilized in immunoprecipitates (or, for that matter, contained in intact, lipid-laden adipocytes) may reflect differences in effective local concentrations of RO 31-8220 at catalytic enzyme sites under these different conditions.) In other studies, we found that RO 31-8220 did not inhibit insulin-induced activation of PI 3-kinase (8) or PI 3-kinase-dependent activation of PKB or glycogen synthase. 2 Thus, RO 31-8220 does not inhibit basic insulin signaling mechanisms that involve PI 3-kinase. We have also found that Go 6976, another bisindolemaleimide derivative that potently inhibits conventional PKCs (␣, ␤, ␥) but not novel or atypical PKCs, did not inhibit insulin-stimulated glucose transport in concentrations up to 100 M (data not shown).
In addition to RO 31-8220, the cell-permeable (see Ref. 20), myristoylated PKC-pseudosubstrate (myr-SIYRRGAR-RWRKL) but not the myristoylated PKA pseudosubstrate (myr-GRTGRRNAI) inhibited insulin-stimulated glucose transport in a time-and concentration-dependent manner in intact adipocytes (Fig. 7). Full inhibition was achieved at 90 -  )). Of particular interest, enzyme activity of immunoprecipitated PKC-in vitro was inhibited by concentrations of PKC-pseudosubstrate that were nearly identical to those required for inhibition of glucose transport in intact cells (Fig.  7). Of further interest, in other studies, 2 we have found that the PKC-pseudosubstrate does not inhibit immunoprecipitated PKB, and PKB is clearly not the PKC-pseudosubstrate-sensitive protein kinase that is required for glucose transport.
As another approach to test the possibility that PKC-may play a role in insulin stimulation of glucose transport, we co-transfected HA-tagged GLUT4 and various forms of PKCinto rat adipocytes. After overnight incubation to allow time for expression (documented by transfecting HA-tagged PKCalong with HA-tagged GLUT4 in some experiments; see Fig. 8), we found that wild-type PKCand, to a greater extent, pointmutated constitutive PKC-increased HA-tagged GLUT4 translocation ( Fig. 9 and Table I). Stimulatory effects of insulin on HA-tagged GLUT4 translocation were still observed in cells expressing wild-type and constitutive PKC-, although the percent increases, because of relatively greater increases in basal HA-tagged GLUT4 translocation, were diminished in these cells. In contrast to consistently observed stimulatory effects of wild-type and constitutive PKC-, a kinase-inactive, dominantnegative form of PKC-(see Ref. 11) failed to significantly alter basal HA-tagged GLUT4 translocation, but in about half of the experiments, it inhibited insulin-stimulated HA-tagged GLUT4 translocation ( Fig. 9 and Table I) by 53 Ϯ 4% (mean Ϯ S.E.; n ϭ 4; p Ͻ 0.005; paired t test); this inhibition was only slightly less than that observed in adipocytes transfected with dominant-negative ⌬p85, viz., 61 Ϯ 10% (n ϭ 4; p Ͻ 0.01), which was also clearly inhibitory in about half of the experiments. Results from an experiment in which all forms of PKCand ⌬p85 were used in parallel (i.e. in the same batch of adipocytes) are shown in Fig. 9; results from multiple experiments are summarized in Table I Fig. 9 for details of transfections and incubations. Note that (a) HA-tagged forms of PKCwere used in B but not in A; (b) we used both HA-tagged and non-HA-tagged forms of wild-type PKC-and dominant-negative PKC-in the functional experiments described in Fig. 9 and Table I (comparable results of HA-tagged GLUT4 translocation were obtained, regardless of the presence or absence of the HA epitope tag on PKCconstructs); and (c) transfection with PKC-constructs did not alter levels of HA-tagged GLUT4.
transfected cells, we presume (from observed stimulatory effects of transfected wild-type PKC-on co-transfected HAtagged GLUT4 translocation) that a significant fraction of expressed transfected wild-type PKC-was enzymatically active, even without insulin addition.
Since PKC-would be expected to operate downstream of PI 3-kinase in the activation of glucose transport, it was of interest to see whether wortmannin altered GLUT4 translocation. As shown in Fig. 9, 100 nM wortmannin blocked insulin effects on HA-tagged GLUT4 translocation but did not inhibit the stimulatory effects of constitutive PKC-. Thus, the activated form of PKC-appeared to operate independently or downstream of PI 3-kinase.
In addition to increases in enzyme activity, insulin provoked rapid increases in 32 P labeling of immunoprecipitable 70-kDa PKC-in cells that had been incubated in the presence of 32 PO 4 for 2 h prior to insulin addition (Fig. 10) (ATP is labeled to constant specific activity during a 2-h period (21)). This observation provided confirmatory evidence that the activation of PKC-observed in vitro truly reflected an activation of PKCin intact cells. On the other hand, it may be noted that maximal increases in 32 P labeling of PKC-occurred at 5 min, and this was different from the timing of maximal increases in PKCenzyme activity (compare Figs. 10 and 1).
In addition to increasing 32 P labeling of PKC-in intact adipocytes, insulin provoked increases in 32 P labeling of PKCthat occurred during the in vitro assay of immunoprecipitated PKC- (Fig. 11); this suggested that PKC-autophosphorylated in vitro in response to insulin treatment in intact cells, and this postulation was confirmed by the finding that the PKCpseudosubstrate markedly inhibited 32 P labeling of PKC-in vitro (Fig. 11). Of further note, PI-3,4-(PO 4 ) 2 and, to a lesser extent, PI-3,4,5-(PO 4 ) 3 increased 32 P labeling of PKC-in control but not insulin-treated PKC-immunoprecipitates (Fig.  11). These findings suggested that both insulin and D3-PO 4 polyphosphoinositides enhanced both the enzyme activity and autophosphorylation of PKC-.

DISCUSSION
The requirement for PI 3-kinase in insulin-induced activation of PKC-in intact adipocytes and the direct activating effects of both pYXXM-stimulated PI 3-kinase and D3-PO 4 polyphosphoinositides on PKC-enzyme activity in vitro suggested that PI 3-kinase is both necessary and sufficient for PKC-activation. The precise mechanism for insulin-induced PKC-activation is not entirely certain, but since added polyphosphoinositides stimulated control but not insulintreated immunoprecipitates and since added polyphosphoinositides narrowed or obliterated the difference in enzyme activity between control and insulin-stimulated PKC-immunoprecipitates, it follows that polyphosphoinositides may have largely accounted for increases in enzyme activity observed in insulin-stimulated PKC-immunoprecipitates. In this scenario, insulin-induced increases in either D3-PO 4 polyphosphoinositides themselves and/or other presently uncertain stimulatory processes or factors that are induced in response to increases in these polyphosphoinositides (e.g. autophosphoryl-  Table I   ation of PKC-) apparently were carried over into the PKCimmunoprecipitate. Studies are currently in progress to evaluate these possibilities.
With respect to the possibility that phosphorylation may have modified PKC-enzyme activity during insulin treatment, it should be noted that the peak of 32 P labeling of PKCin intact adipocytes seemed to occur at a time when enzyme activity had temporarily receded, i.e. at the 5-min nadir of the biphasic activation sequence. This raises the possibility that some phosphorylation sites on PKC-may have either no effect or an inhibitory effect on PKC-enzyme activity. Accordingly, it is possible that there are multiple insulin-sensitive phosphorylation sites on PKC-, with various effects on enzyme activity. Further studies on specific PKC-phosphorylation sites will be required to evaluate this possibility and identify the kinase(s) responsible for these phosphorylations in intact adipocytes.
The findings that the PKC-pseudosubstrate and RO 31-8220 provoked similar dose-dependent inhibitory effects on immunoprecipitated PKC-enzyme activity and insulin-stimulated glucose transport suggested that PKC-or a closely related kinase may be required for glucose transport effects of insulin. Along these lines, DAG-sensitive PKCs, despite being targets for the PKC-pseudosubstrate and RO 31-8220, can be excluded as being required for insulin stimulation of glucose transport, as suggested by the fact that phorbol ester-dependent depletion of PKC-␣, -␤1, -␤2, -␦, and -⑀ and the conventional PKC inhibitor Go 6976 did not inhibit insulin stimulation of glucose transport; this conclusion was also supported by the fact that inhibitory effects of RO 31-8220 on insulin-stimulated glucose transport more closely matched the inhibition of PKCand occurred at considerably higher concentrations than those required to inhibit PKC-␣, -␤1, -␤2, -␥, -␦, -⑀, and -.
As alluded to above, PKB does not appear to be the PKCpseudosubstrate-sensitive protein kinase that is required for insulin-stimulated glucose transport. This does not imply that PKB is not required for insulin-stimulated glucose transport, but it seems clear that at least one PKC or a closely related protein kinase is required, perhaps along with PKB. In this regard, as discussed above, DAG-sensitive PKCs appear to be ruled out. It is therefore tempting to suggest that PKC-and/or another atypical PKC is the PKC-pseudosubstrate-sensitive kinase that is required for insulin stimulation of glucose transport. However, we cannot rule out the possibility that the PKC-pseudosubstrate, as well as RO 31-8220, may bind to and thus inhibit catalytic sites of other relevant, but presently unknown, non-PKC protein kinases.
The activation of GLUT4 translocation by wild-type PKC-, and even more so by constitutive PKC-, in transiently transfected adipocytes provided further evidence that PKC-may participate in the activation of GLUT4 translocation by insulin or other agonists. Indeed, two forms of constitutive PKC-, one point-mutated in the pseudosubstrate region (see above) and one in which the regulatory domain was deleted, were nearly as effective as insulin in stimulating HA-tagged GLUT4 translocation. However, as with inhibitor studies, the finding that transient transfection of wild-type and constitutive PKC-led to increases in HA-tagged GLUT4 translocation must be interpreted cautiously, as these increases reflect slowly developing responses (i.e. over 16 -24 hours) that could be due to activation of a variety of factors and other non-PKC-signaling pathways. Similarly, the partial inhibition of insulin-stimulated glucose transport by dominant-negative PKC-suggested that PKCmay be required for this stimulation, but this inhibition may also reflect nonspecific alterations in other signaling components in cells transfected with this interfering PKC-mutant. Obviously, other experimental approaches will be needed to further evaluate the importance of PKC-in insulin-stimulated glucose transport. FIG. 10. Effects of insulin on 32 P labeling of immunoprecipitable PKCin rat adipocytes. Cells were incubated for 2 h in low phosphate (0.12 mM) KRP buffer containing 10 mM HEPES, 2 mM glucose, 1% bovine serum albumin, and 100 Ci of 32 PO 4 /ml and then treated with or without 10 nM insulin for the indicated times. PKC-was immunoprecipitated from whole cell lysates, and precipitates were resolved by SDS-polyacrylamide gel electrophoresis and analyzed for immunoreactivity and 32 P labeling of 70-kDa PKC-. Insets show a representative autoradiogram and its corresponding immunoblot (note equal loading of PKC-). The graph shows results (mean Ϯ S.E.) of 3 determinations in which 32 P labeling of PKC-was quantified with a Bio-Rad phosphorimager.
It may be noted that the present findings on insulin-induced activation of PKC-and its potential role in the activation of GLUT4 translocation and glucose transport in rat adipocytes are in keeping with findings in 3T3/L1 fibroblasts and adipocytes, in which stable transfection with pCDNA3 containing cDNA encoding wild-type and dominant-negative PKCled to increases and decreases, respectively, of glucose transport (11). Of further interest, we have found that PKC-is activated by insulin in L6 myotubes (23) and rat skeletal muscles; 2 it is therefore tempting to suggest that PKC-is activated by insulin and may participate in regulating glucose transport in a variety of insulin-sensitive cell types.
In summary, our findings suggest that insulin activates PKC-in rat adipocytes largely through PI 3-kinase-dependent increases in polyphosphoinositides. Our findings also suggest that PKC-may be required for and may participate in the translocation of GLUT4 and the activation of glucose transport in rat adipocytes. Further studies are needed to test these hypotheses.