THE MECHANISMS USED by insulin to regulate the

Transcription

1 /00/$03.00/0 Vol. 141, No. 11 Endocrinology Printed in U.S.A. Copyright 2000 by The Endocrine Society Effects of Adenoviral Gene Transfer of Wild-Type, Constitutively Active, and Kinase-Defective Protein Kinase C- on Insulin-Stimulated Glucose Transport in L6 Myotubes* GAUTAM BANDYOPADHYAY, YOSHINORI KANOH, MINI P. SAJAN, MARY L. STANDAERT, AND ROBERT V. FARESE J. A. Haley Veterans Hospital Research Service and Department of Internal Medicine, University of South Florida College of Medicine, Tampa, Florida THE MECHANISMS USED by insulin to regulate the translocation of the GLUT4 glucose transporter to the plasma membrane and subsequent glucose transport in skeletal muscle cells and adipocytes are particularly important, but poorly understood. It is now generally accepted that phosphatidylinositol (PI) 3-kinase is required, but factors that operate distal to PI 3-kinase remain uncertain. Inhibitor and transfection studies have provided the initial evidence that two protein kinases that are activated by PI 3-kinase, viz. atypical protein kinases C (PKCs) (1 4) and/or protein kinase B (PKB) (5 8), may be required for insulin effects on GLUT4 translocation and glucose transport in skeletal muscle cells and adipocytes. However, these experimental approaches have inherent caveats and do not provide entirely convincing answers. Further evidence that supports a potential role for atypical PKCs and PKB in insulin-stimulated glucose transport has been obtained from use of adenoviral gene transfer methods (9) and microinjection of antibodies (10, 11) in 3T3/L1 adipocytes. To date, these perhaps more convincing experimental approaches have not been used to gain insight into the question of whether atypical PKCs and PKB are required for insulin-stimulated GLUT4 translocation and glucose transport in skeletal muscle-type cells. Received May 1, Address all correspondence and requests for reprints to: Robert V. Farese, M.D., Research Service (VAR 151), J. A. Haley Veterans Hospital, Bruce B. Downs Boulevard, Tampa, Florida rfarese@com1.med.usf.edu. * This work was supported by funds from the Department of Veterans Affairs Merit Review Program and NIH Research Grant 2R01-DK A1. ABSTRACT We used adenoviral gene transfer methods to evaluate the role of atypical protein kinase Cs (PKCs) during insulin stimulation of glucose transport in L6 myotubes. Expression of wild-type PKC- potentiated maximal and half-maximal effects of insulin on 2-deoxyglucose uptake, but did not alter basal uptake. Expression of constitutively active PKC- enhanced basal 2-deoxyglucose uptake to virtually the same extent as that observed during insulin treatment. In contrast, expression of kinase-defective PKC- completely blocked insulin-stimulated, but not basal, 2-deoxyglucose uptake. Similar to alterations in glucose transport, constitutively active PKC- mimicked, and kinase-defective PKC- completely inhibited, insulin effects on GLUT4 glucose transporter translocation to the plasma membrane. Expression of kinase-defective PKC-, in addition to inhibition of atypical PKC enzyme activity, was attended by paradoxical increases in GLUT4 and GLUT1 glucose transporter levels and insulinstimulated protein kinase B enzyme activity. Our findings suggest that in L6 myotubes, 1) atypical PKCs are required and sufficient for insulin-stimulated GLUT4 translocation and glucose transport; and 2) activation of protein kinase B in the absence of activation of atypical PKCs is insufficient for insulin-induced activation of glucose transport. (Endocrinology 141: , 2000) Here we used adenoviral gene transfer methodology to introduce various wild-type and mutant forms of the atypical PKC, PKC-, into L6 myotubes, a continuous cell line derived from rat skeletal muscle. This approach allowed us to introduce measured, reasonable amounts of these PKCs into a large fraction of L6 myotubes, and, accordingly, more critically test the hypothesis that atypical PKCs play an important role during insulin stimulation of GLUT4 translocation and glucose transport in this important cell type. Materials and Methods Cell culture, viral infections, and incubation conditions L6 cells were grown to confluence and fully differentiated to myotubes as described previously (2). Myotubes were cultured in 24-well plates for 2-deoxyglucose uptake studies and in 100-mm plates for studies of GLUT4 translocation and enzyme activation. Myotubes were cultured over a 48-h period with the indicated concentrations [as multiplicity of infection (MOI), i.e. number of viral particles per cell] of either adenovirus alone or adenovirus encoding wild-type, constitutively active, or kinase-defective PKC- (provided by Dr. Masato Kasuga, Kobe University, Kobe, Japan) as described previously in studies of 3T3/L1 adipocytes (11). At the end of the 48-h period, which allowed sufficient time for expression of encoded PKCs, as described previously (2), cells were equilibrated at 37 C in glucose-free Krebs-Ringer phosphate buffer (KRP) containing 1 mg/ml BSA and then treated with or without insulin as described in the text. Studies of glucose transport After insulin treatment for 30 min, the uptake of [ 3 H]2-deoxyglucose over 5 min and the recovery of immunoreactive GLUT4 in the plasma membrane (purified by ultracentrifugation) were measured as described previously (2). [ 3 H]2-deoxyglucose uptake results are expressed as counts per min/well. Note that infection of myotubes with adenovirus 4120

2 PKC- AND INSULIN-STIMULATED GLUCOSE TRANSPORT 4121 alone or adenovirus encoding wild-type, constitutively active, or kinasedefective PKC- did not alter the recovery of cellular protein per well or per plate and additionally, except for adenovirus encoding constitutively active PKC- (which provoked increases in uptake; see below), did not alter the level of basal [ 3 H]2-deoxyglucose uptake. PKC- / assays Activation of PKC- / was assessed as described previously (1 4). In brief, after activation of cells with or without insulin for 10 min, PKC- / was immunoprecipitated from postnuclear (centrifuged at 700 g for 10 min to remove nuclei, cellular debris, and floating fat) cell lysates with a rabbit polyclonal antiserum (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) that targets the nearly identical C-termini of both PKC- and PKC-. Immunoprecipitates were collected on protein A-Sepharose G beads, washed, and incubated for 8 min at 30 C in 100 l buffer containing 50 mm Tris-HCl (ph 7.5), 5 mm MgCl 2, 100 m Na 3 VO 4, 100 m Na 4 P 2 O 7,1mm NaF, 100 m phenylmethylsulfonylfluoride, 4 g phosphatidylserine, 3 5 Ci [ - 32 P]ATP (NEN Life Science Products, Boston, MA), 50 m ATP, and, as substrate, 40 m PKC- pseudosubstrate serine 159 analog (Biosource Technologies, Inc., Camarillo, CA). After incubation, 32 P-labeled substrate was trapped on p81 filter paper and counted for radioactivity. PKB assays Activation of immunoprecipitable PKB was measured as described previously (12), using a kit obtained from Upstate Biotechnology, Inc. (Lake Placid, NY). Western analyses As previously described (1 4), cell lysates or subcellular fractions were boiled and stored in Laemmli buffer, subjected to SDS-PAGE, transferred to nitrocellulose membranes, and immunoblotted with the following: 1) rabbit polyclonal antiserum that targets the C-termini of both PKC- and PKC- (Santa Cruz Biotechnology, Inc.), 2) rabbit polyclonal anti-pkb antiserum (Upstate Biotechnology, Inc.), 3) rabbit polyclonal anti-glut4 antiserum (Biogenesis, Bournemouth, UK), 4) rabbit polyclonal anti-glut1 antiserum (provided by Dr. Ian Simpson), 5) rabbit polyclonal antiserum that targets the surrounding peptide sequence that includes phosphoserine 473 in PKB (New England Biolabs, Inc.), 6) goat polyclonal antiserum that recognizes a specific N-terminal sequence of PKC- (Santa Cruz Biotechnology, Inc.), and 7) mouse monoclonal antibody that recognizes a specific internal sequence of PKC- (Transduction Laboratories, Inc., Lexington, KY). Immunoblots were quantitated by measurement of chemiluminescence in a Bio-Rad Laboratories, Inc., Molecular Analyst Chemiluminescence/Phosphorescence Imaging System (Richmond, CA). FIG. 1. Expression of PKC- in myotubes infected with adenovirus containing complementary DNA (cdna) inserts encoding wild-type (WT), constitutively active (CONST), and kinase-defective (KD) PKC-. Myotubes were infected with increasing amounts (as MOI) of adenoviruses containing no cdna insert or cdna inserts encoding various forms of PKC- as indicated. After incubation for 48 h, cell lysates were examined for levels of total combined, endogenous plus virus-derived exogenous, atypical PKCs, i.e. and, as determined by immunoblotting with an antiserum (Santa Cruz Biotechnology, Inc.) that targets the C-termini of both PKC- and PKC-. Note that constitutively active PKC- lacks amino acids of the N-terminus and migrates below the level of endogenous 75-kDa PKC- /, i.e. at about 63 kda. The numerical values shown here indicate the ratio of total endogenous plus virus-derived exogenous PKC- / to the endogenous PKC- / level found in cells infected with virus alone (set at unity). See the text for the ratios observed in multiple experiments. Immunoblots were quantitated by measurement of extended chemiluminescence in a Bio-Rad Laboratories, Inc., Molecular Analyst Chemiluminescence/Phosphorescence Imaging System as described in Materials and Methods. Results Expression of PKC- in adenovirus-infected myotubes In preliminary studies we found that most ( 90%) cells stained positively for expression of -galactosidase enzyme activity after infection with 5 10 MOI adenovirus-encoding -galactosidase. As shown in Fig. 1, there were no significant changes in the level of endogenous 75-kDa PKC- / in cells infected with adenovirus alone. In cells infected with adenovirus encoding wild-type, constitutively active, and kinase-defective PKC-, there were dose-related increases in the expression of these PKCs (Fig. 1). In cells infected with adenovirus encoding wild-type and constitutively active PKC-, total cellular combined PKC- / (endogenous) plus PKC- (virus-derived exogenous) was increased approximately 2-fold (i.e. a 100% increase relative to the endogenous PKC- / ) at 5 10 MOI adenovirus in the experiment depicted in Fig. 1. Note that constitutively active PKC- lacks N-terminal amino acids (which contain the autoinhibitory pseudosubstrate sequence) and therefore migrates faster than endogenous or wild-type 75-kDa PKC- /, i.e. at approximately 63 kda. Also note that in other studies (see Ref. 12 and unpublished observations), we found that truncated forms of atypical PKCs are partially activated, but nevertheless can be further activated by insulin, most likely via increases in PI 3-kinase/3-phosphoinositide-dependent protein kinase-1-dependent phosphorylation of the threonine 410 loop site and subsequent autophosphorylation of threonine 560. In cells infected with adenovirus encoding kinase-defective PKC-, increases in total cellular PKC- / were approximately 4.5- to 5.3-fold at 5 and 10 MOI in the experiment depicted in Fig. 1. In multiple experiments, ratios (mean se) of total endogenous plus exogenous PKC- / to endogenous PKC- / in cells infected with 10 MOI adenovirus encoding wild-type, constitutively acting, and kinase-

3 4122 PKC- AND INSULIN-STIMULATED GLUCOSE TRANSPORT Endo 2000 Vol. 141 No. 11 defective PKC- were (n 7), (n 16), and (n 16), respectively. As shown in Fig. 2, rat-derived L6 myotubes, like rat vastus lateralis skeletal muscle, were found to contain primarily PKC- and a relatively small amount of PKC- compared with mouse skeletal muscle, which contained primarily PKC- and a relatively small amount of PKC-. As reported previously (12), PKC- and PKC- are closely (72%) homologous (13), contain the same pseudosubstrate sequence, are activated similarly by insulin through PI 3-kinase- dependent increases in PI-3,4,5- (PO 4 ) 3 (12), and, moreover, function interchangeably in supporting insulin-dependent GLUT4 translocation (4). The latter interchangeability allowed us to use PKC- constructs despite the fact that PKC- is the predominant atypical PKC in L6 myotubes. In confirmation of this assumption of interchangeability, as described below, mutant forms of PKC- were indeed found to markedly alter PKC- enzyme activity and insulin effects on GLUT4 translocation and glucose transport in L6 myotubes. FIG. 2. Levels of immunoreactive PKC- and PKC- in L6 myotubes relative to levels in rat and mouse vastus lateralis muscles. Equal amounts of lysate protein were resolved by SDS-PAGE and blotted with anti( )-PKC- / antiserum (Santa Cruz Biotechnology, Inc.; rabbit polyclonal antiserum that recognizes the nearly identical C- terminal sequences of PKC- and PKC- ), anti-pkc- antiserum (Santa Cruz Biotechnology, Inc.; goat polyclonal antiserum that recognizes a specific N-terminal sequence of PKC- ), and anti-pkc- antibodies (Transduction Laboratories, Inc.; mouse monoclonal antibody that recognizes a specific internal sequence of PKC- ). FIG. 3. Effects of adenoviral gene transfer of various forms of PKC- on basal and insulin-stimulated glucose transport in L6 myotubes. Myotubes (in 24-well plates) were infected with the indicated doses (as MOI) of adenovirus alone (A) or adenovirus containing complementary DNA (cdna) encoding wild-type (C), constitutively active (B), or kinase-defective (D) PKC-. After incubation for 48 h to allow time for expression, cells were incubated for 30 min in glucose-free KRP medium with (o) or without ( ) 100 nm insulin, after which uptake of [ 3 H]2-deoxyglucose over 5 min was measured. Note that adenoviral constructs did not alter cellular protein content per well. Shown here are the mean SE for three or more determinations (n 3 12) or the mean range for two determinations (n 2). Asterisks indicate P 0.005, as determined by t test comparisons of peak alterations: a, insulin-stimulated value at 10 MOI virus encoding wild-type PKC- vs. insulin-stimulated value at 10 MOI virus alone; b, insulin-stimulated value at 10 MOI virus encoding kinase-defective PKC- vs. insulin-stimulated value at 10 MOI virus alone; and c, control value at 25 MOI for virus encoding constitutively active PKC- vs. control value at 25 MOI virus alone. Effects of expression of various forms of PKC- on glucose transport Infection of myotubes with adenovirus alone had little or no effect on basal or insulin-stimulated glucose transport (Fig. 3). Expression of wild-type PKC- in virus-infected myotubes had little or no effect on basal glucose transport (i.e. 2-deoxyglucose uptake), but at viral doses of 5 10 MOI it consistently enhanced both maximal and half-maximal effects of insulin on glucose transport (Figs. 3 and 4). Expression of increasing amounts of truncated, constitutively active PKC- provoked dose-related increases in basal glucose transport, and at adenoviral doses of 10 MOI and higher, basal transport activity approached that observed with insulin treatment (Fig. 3). Perhaps most importantly, expression of kinase-defective PKC- at adenoviral doses of 5 10 MOI completely inhibited the effects of insulin on glucose transport, but had little or no effect on basal glucose transport (Figs. 3 and 4). The presently observed effects of wild-type, constitutively active, and kinase-defective PKC- could not be explained by alterations in cell recovery (see Materials and Methods) or levels of GLUT1 or GLUT4 glucose transporters. Indeed, as shown in Fig. 5, the levels of both transporters were increased significantly by expression of kinase-defective PKC- and, if anything, were decreased, albeit not statistically significantly, by expression of constitutively active PKC-. Effects of expression of various forms of PKC- on PKC- / enzyme activity In myotubes infected with virus alone, insulin provoked a 2.5-fold increase in immunoprecipitable PKC- / enzyme activity (Fig. 6). The expression of kinase-defective PKC- in virus-infected myotubes led to a decrease in overall intrinsic

4 PKC- AND INSULIN-STIMULATED GLUCOSE TRANSPORT 4123 FIG. 4. Effects of adenoviral gene transfer of wild-type (WT) and kinase-defective (KD) PKC- on dose-dependent effects of insulin on glucose transport in L6 myotubes. Myotubes (in 24-well plates) were infected without (control) or with the indicated doses (MOI) of adenovirus encoding WT or KD PKC- (shown at right), and 48 h later myotubes were incubated for 30 min in glucose-free KRP medium with the indicated doses of insulin, after which the uptake of [ 3 H]2- deoxyglucose over 5 min was measured. Note that adenoviral constructs did not alter the cellular protein content per well. Shown here are mean values of four closely agreeing values (for simplicity of presentation, SEs, which did not exceed 10% of the mean, are not shown). Asterisks indicate P 0.05, as determined by t test comparison of insulin-stimulated uptake in cells expressing WT or KD PKC- vs. uptake in cells treated with same doses of insulin and the same MOI of adenovirus alone (f). enzyme activity of basal PKC- / (Fig. 6A); this probably reflects the fact that these immunoprecipitates contained both enzymatically active endogenous wild-type PKC- / and virus-derived kinase-defective PKC-, so the data in Fig. 6A were normalized to reflect the enzyme activity of equal amounts of immunoprecipitated PKC- /. Moreover, in cells expressing kinase-defective PKC-, there was not only a loss of the ability of insulin to provoke increases in overall PKC- / activity, but, for uncertain reasons, PKC- / enzyme activity actually decreased after insulin treatment (Fig. 6). We thought it important to assay total combined endogenous plus exogenous virus-derived atypical PKC enzyme activity, rather than endogenous PKC- / activity, in these studies, as this mixture of virus-derived exogenous PKC- plus endogenous PKC- / is what the intact virus-infected cell contains and uses to regulate biological processes. In this context, the virus-derived kinase-defective PKC- would be expected to compete with endogenous wild-type atypical PKCs for activating factors and substrates in both intact cells and in the assay in vitro. Indeed, we previously reported that in rat adipocytes (which, like rat skeletal muscles and rat-derived L6 myotubes, contain primarily PKC- ), 1) plasmid-mediated expression of kinase-inactive forms of either PKC- or PKC- leads to inhibition of total cellular, combined (endogenous plus exogenous), insulin-stimulated, immunoprecipitable PKC- / (4); and 2) plasmid-mediated expression of both kinase-inactive and activation-resistant (threonine 410 mutated to alanine) forms of PKC- inhibits insulin-induced activation of epitope-tagged wild-type PKC- (14). In addition, assuming that total atypical PKC enzyme activity in cells expressing kinase-defective PKC- largely reflects that FIG. 5. Alterations in levels of immunoreactive GLUT1 and GLUT4 glucose transporters in L6 myotubes infected with adenoviruses (adeno) encoding wild-type (WT), constitutively active (CST), and kinasedefective (KD) PKC-. Myotubes were infected with 10 MOI adenovirus encoding the indicated forms of PKC-, and 48 h later cell lysates were analyzed for levels of immunoreactive GLUT1 and GLUT4 glucose transporters. Shown here are representative immunoblots and relative levels of transporters, as determined by measurement of extended chemiluminescence in a Bio-Rad Laboratories, Inc., Molecular Analyst Phosphorescence/Chemiluminescence Imaging System. Numerical values are the mean SE of the number of determinations given in parentheses, i.e. comparisons of transporter levels in cells infected with adenovirus encoding the indicated PKCs vs. transporter levels in corresponding control cells infected with adenovirus alone (set at unity). Asterisks indicate P 0.05, by paired t test. of the enzymatically active, endogenous PKC- (see above), from the data shown in Fig. 6, it seems clear that kinasedefective PKC- served as a very effective inhibitor of insulin-induced activation of this endogenous PKC- in L6 myotubes. In contrast to kinase-defective PKC-, expression of constitutively active PKC- was attended by the anticipated increases in the intrinsic enzyme activity of PKC- / immunoprecipitated from lysates of control myotubes (Fig. 6A). In addition to increasing basal intrinsic PKC- / enzyme activity, the total cellular content of enzymatically active PKC- /, i.e. endogenous wild-type PKC- / plus virus-derived, constitutively active PKC-, was increased substantially (see Fig. 6C), and total cellular basal PKC- / enzyme activity was increased approximately 4-fold in cells infected with adenovirus encoding constitutively active PKC- (Fig. 6B). Nevertheless, insulin continued to provoke increases in PKC- / enzyme activity in cells infected with constituitively active PKC- (Fig. 6), and this probably reflects 1) the activation of endogenous wild-type PKC- /, and/or 2) the fact that N-terminally truncated forms of atypical PKCs are only partially activated and can be further activated by insulin, most likely via 3-phosphoinositide-dependent kinase-1 increases in phosphorylation (12) (our unpublished observations). Effects of expression of kinase-defective PKC- on PKB activity As PKB has been reported to be required for insulininduced translocation of epitope-tagged GLUT4 glucose

5 4124 PKC- AND INSULIN-STIMULATED GLUCOSE TRANSPORT Endo 2000 Vol. 141 No. 11 FIG. 6. Effects of adenoviral gene transfer of kinase-defective and constitutively active PKC- on enzyme activity of immunoprecipitable PKC- / in L6 myotubes. Myotubes (in 100-mm plates) were infected with 10 MOI of adenoviruses (adeno) encoding constitutively active (CONSTIT) or kinase-defective (KD) PKC-, and 48 h later myotubes were incubated for 30 min in glucose-free KRP medium with or without 100 nm insulin, after which equal amounts (250 g protein) of postnuclear cell lysates were examined for enzyme activity of immunoprecipitable combined PKC- /. Shown here are the mean SE of six determinations. Values for PKC- / enzyme activity in A were normalized to reflect an amount of immunoprecipitable PKC- / equal to that in cells infected with adenovirus alone (see C for a representative immunoblot). Values in B reflect the total cellular PKC- / activity. transporter to the plasma membrane and presumably for subsequent glucose transport in L6 myotubes (8), it was important to determine whether kinase-defective PKC- interfered with insulin-induced activation of PKB. On the contrary, expression of kinase-defective PKC- was attended by increases in insulin-stimulated PKB enzyme activity (Fig. 7) and PKB/serine 473 phosphorylation (Figs. 7 and 8). This FIG. 7. Effects of adenoviral gene transfer of kinase-defective (KD) PKC- on PKB enzyme activity (A), total immunoreactive PKB (B), and phosphorylation of serine 473 in PKB (C) in L6 myotubes. Myotubes (in 100-mm plates) were infected with 10 MOI adenovirus alone or adenovirus encoding KD-PKC-, and 48 h later myotubes were incubated in glucose-free KRP medium for 10 min with or without 100 nm insulin as indicated. After incubation postnuclear cell lysates were analyzed for immunoprecipitable PKB enzyme activity (A) or immunoreactivity of total PKB (B) or phosphoserine 473-PKB (C) as described in Materials and Methods. B and C, Values are expressed relative to the mean control level for each immunoblot. Values are the mean SE of three determinations in A and four determinations in B and C. Increases in insulin-stimulated PKB activity (A) and levels of PKB (B) and phosphoserine 473-PKB (C) in cells infected with KD-PKC- were all statistically significant: P (A), P (B), and P (C). finding of increased PKB activation/phosphorylation may be due in part to increases in immunoreactive PKB content (Figs. 7 and 8), but may also reflect the fact that atypical PKCs

6 PKC- AND INSULIN-STIMULATED GLUCOSE TRANSPORT 4125 may be both functional and nonfunctional pools of plasma membrane-associated GLUT4. Nevertheless, despite increased basal levels, insulin provoked further increases in plasma membrane GLUT4 levels in cells expressing wildtype PKC-. Moreover, in myotubes expressing constitutively active PKC-, basal plasma membrane GLUT4 levels were increased to virtually the same extent as that observed with insulin (Fig. 8). And, perhaps most importantly, in myotubes expressing kinase-defective PKC-, insulin failed to provoke increases in plasma membrane GLUT4 content (Fig. 8). FIG. 8. Effects of adenoviral gene transfer of various forms of PKC- on insulin-induced translocation of GLUT4 to the plasma membrane in L6 myotubes. Myotubes (in 100-mm plates) were infected with 10 MOI of adenovirus (adeno) alone or adenovirus encoding wild-type (WT), constitutively active (CST), or kinase-defective (KD) PKC-, and 48 h later myotubes were incubated for 30 min in glucose-free KRP medium with or without 100 nm insulin, after which plasma membranes were isolated and examined for immunoreactive GLUT4 and PKC- / (top two panels). Lysates were also examined for levels of immunoreactive PKC- /, phosphoserine 473 (ps473) PKB, and total PKB. Note the increases in immunoreactive PKC- / in both total cell lysates and plasma membranes in cells expressing wild-type, constitutively active, and kinase-defective PKC-. Also note the increases in total PKB and phosphoserine 473-PKB in cells expressing kinase-defective PKC-. Immunoblots shown here are representative of four determinations. See Figs. 5 and 7 for quantitative comparisons of changes in total GLUT4, PKB, and phosphoserine 473-PKB in cells expressing kinase-defective PKC-. can act as negative modulators of PKB (15). Whatever the correct explanation, the presently observed inhibitory effects of kinase-defective PKC- on glucose transport cannot be explained by inhibition of PKB. Effects of expression of various forms of PKC- on GLUT4 translocation to the plasma membrane In conjunction with alterations in insulin-stimulated glucose transport, the expression of wild-type, constitutively active, and kinase-defective PKC- in adenovirus-infected myotubes provoked qualitatively similar changes in the recovery of immunoreactive GLUT4 in the plasma membrane. As shown in Fig. 8, in cells infected with virus alone, there was little or no GLUT4 recovered in the plasma membrane fraction in the basal state, and insulin provoked a marked increase in plasma membrane GLUT4 content. In myotubes expressing wild-type PKC-, the basal level of plasma membrane GLUT4 was surprisingly high (Fig. 8) despite the fact that there was no concomitant increase in basal glucose transport (see Figs. 3 and 4). This apparent disparity may reflect the fact that actual glucose transport may be dependent not only on the plasma membrane level of GLUT4, but also on other insulin-induced factors that appear to be required to increase the transport activity of plasma membrane-associated GLUT4 (16, 17). Alternatively, it is possible that there Discussion The present findings provided strong evidence that atypical PKCs are required for the effects of insulin on translocation of endogenous GLUT4 to the plasma membrane and subsequent glucose transport in L6 myotubes. In this regard, it is important to note that the level of expression of kinasedefective PKC- that led to essentially complete inhibition of the effects of insulin on both GLUT4 translocation and glucose transport, viz. at a viral titer of 5 10 MOI, was approximately 3- to 4-fold greater than the level of endogenous PKC- /. Assuming that expression of a given amount of kinase-defective (i.e. inactive) atypical PKC would effectively dilute the enzymatic and biological activity of an equal amount of endogenous atypical PKC by 50%, a 4-fold excess of kinase-defective atypical PKC would be expected to inhibit endogenous atypical PKC by approximately 80%, i.e. slightly less than the virtually complete inhibition of insulinstimulated GLUT4 translocation and glucose transport presently observed. This disparity may reflect the fact that total cellular PKC- / enzyme activity was paradoxically diminished by insulin treatment in cells expressing kinase-defective PKC-. Dependency of endogenous GLUT4 translocation and overall glucose transport has also been reported in adenoviral gene transfer studies in 3T3/L1 adipocytes (11). However, in 3T3/L1 cells, the maximal level of inhibition of GLUT4 translocation and glucose transport caused by expression of kinase-defective PKC- was approximately 50 60% compared with the virtually complete inhibition we observed in L6 myotubes. As the 3T3/L1 adipocytes and L6 myotubes used in these studies were transfected with saturating amounts of the same adenoviral PKC- constructs, it is possible that despite partial similarities, the signaling factors used by insulin to activate GLUT4/glucose transport systems may be partly different in these two cell types. However, another possibility is that lesser inhibitory effects of kinase-defective PKC- on insulin-stimulated glucose transport in 3T3/L1 adipocytes may reflect lesser rates of adenoviral infection in these cells. In addition to the apparent requirement for atypical PKCs during insulin-stimulated GLUT4 translocation and glucose transport in L6 myotubes (as suggested by studies with kinase-defective PKC- ), the fact that expression of wild-type PKC- potentiated insulin effects suggested that atypical PKCs actively contribute to insulin-stimulated GLUT4 translocation and glucose transport. Interestingly, similar potentiating effects of overexpressed wild-type PKC- on insulin-

7 4126 PKC- AND INSULIN-STIMULATED GLUCOSE TRANSPORT Endo 2000 Vol. 141 No. 11 stimulated epitope-tagged GLUT4 translocation or glucose transport have been observed in transiently transfected rat adipocytes (4) and in rat skeletal muscles injected in vivo with adenovirus encoding wild-type PKC- (18). In keeping with the possibility that PKC- / contributes to insulin-stimulated glucose transport, it may be noted that even in the absence of insulin treatment, constitutively active PKC- was capable of provoking increases in GLUT4 translocation and glucose transport comparable to those of insulin. On the other hand, it is questionable if the expression of a constitutively active atypical PKC truly mimics the signaling system(s) used by insulin in intact cells. The present findings are in agreement with those of our previous studies that suggested a requirement for atypical PKCs during insulin-stimulated GLUT4 translocation and glucose transport in L6 myotubes (2), 3T3/L1 adipocytes (1), and rat adipocytes (3, 4). However, the presently used adenoviral gene transfer methodology provided more convincing evidence for this hypothesis than the stable and transient transfection methodology used in our previous studies. In this regard, stable transfection approaches are open to the criticisms that they do not necessarily provide a homogeneous population of uniformly transfected cells and, moreover, may select cells that employ aberrant signaling circuits. In both L6 myotubes (2) and 3T3/L1 cells (1), for example, the stable transfection approach yielded cell populations in which expression of kinase-defective PKC- inhibited insulin-stimulated glucose transport and GLUT4 translocation to the plasma membrane by only 50%, and it was uncertain whether this reflected a failure to obtain cells that were uniformly transfected, the possibility that there was only a 2-fold increase in total PKC- / levels in uniformly transfected cells, or the existence of multiple parallel mechanisms used by insulin to activate the glucose transport system. Similarly, in transient transfections, an even smaller percentage of cells was successfully transfected, and it was generally necessary to use an exogenously cotransfected epitopetagged GLUT4 or another extraneous marker to focus on successfully transfected cells. Obviously, there are many assumptions in the transient cotransfection approach, e.g. that cotransfections have occurred in the same cell population, and even if all assumptions are correct, this approach does not provide direct information on the endogenous glucose transport system. Moreover, in the transient transfection approach, particularly at low transfection rates, it is generally necessary to use relatively large amounts of plasmid to be certain that significant protein expression has occurred, and this not only leads to excessive amounts of wild-type or mutant protein in successfully transfected cells, but also is attended by considerable uncertainty as to the ratio of transfected protein to the endogenous protein under study. Obviously, we were able to avoid many of these caveats in the present adenoviral gene transfer studies. It was of interest that there appeared to be an inverse relationship between PKC- / enzyme activity and levels of GLUT4 and GLUT1 glucose transporters. Thus, expression of constitutively active PKC- was attended by mild, but statistically insignificant, decreases in the levels of these transporters, and expression of kinase-defective PKC- was attended by more substantial, statistically significant, increases in levels of these transporters. A similar inverse relationship between PKC- enzyme activity and glucose transporter levels was also observed in previous stable transfection studies in 3T3/L1 adipocytes (1) and L6 myotubes (2), and this relationship may reflect a homeostatic mechanism that attempts to maintain a sufficient, but not excessive, level of glucose transport. It was of interest to find that despite enhanced effects of insulin on PKB phosphorylation and activation in cells expressing kinase-defective PKC-, insulin was unable to stimulate GLUT4 translocation or glucose transport. This finding provided clear evidence that the activation of PKB in the absence of atypical PKC activation is not sufficient for activation of the insulin-dependent glucose transport system, and activation of an atypical PKC may be indispensable for this action of insulin in L6 myotubes. This does not necessarily imply that PKB is not required for insulin-stimulated glucose transport in these cells. Indeed, the findings of Akimoto et al. (13) suggest that PKB is required for insulinstimulated GLUT4 translocation in L6 myotubes. Accordingly, at this point, it appears that both PKB and atypical PKCs are required for insulin-stimulated glucose transport in L6 myotubes. In summary, we used adenoviral gene transfer methods to introduce various forms of PKC- into L6 myotubes to examine the role of atypical PKCs during insulin stimulation of endogenous GLUT4 translocation and glucose transport. We found that expression of kinase-defective PKC- completely inhibited, wild-type PKC- potentiated, and constitutively active PKC- fully mimicked the effects of insulin on GLUT4 translocation and glucose transport. Our findings provided strong support for the hypothesis that atypical PKCs are required for and contribute directly to the effects of insulin on GLUT4 translocation and glucose transport in the L6 skeletal muscle cell line. Acknowledgment We thank Sara M. Busquets for her invaluable secretarial assistance. References 1. Bandyopadhyay G, Standaert ML, Zhao L, Yu B, Avignon A, Galloway L, Karnam P, Moscat J, Farese RV 1997 Activation of protein kinase C (,, and ) by insulin in 3T3/L1 cells. Transfection studies suggest a role for PKC- in glucose transport. J Biol Chem 272: Bandyopadhyay G, Standaert ML, Galloway L, Moscat J, Farese RV 1997 Evidence for involvement of protein kinase C (PKC)- and noninvolvement of diacylglycerol-sensitive PKCs in insulin-stimulated glucose transport in L6 myotubes. Endocrinology 138: Standaert ML, Galloway L, Karnam P, Bandyopadhyay G, Moscat J, Farese RV 1997 Protein kinase C- as a downstream effector of phosphatidylinositol 3-kinase during insulin stimulation in rat adipocytes. Potential role in glucose transport. J Biol Chem 272: Bandyopadhyay G, Standaert ML, Kikkawa U, Ono Y, Moscat J, Farese RV 1999 Effects of transiently expressed atypical (, ), conventional (, ) and novel (, ) protein kinase C isoforms on insulin-stimulated translocation of epitope-tagged GlUT4 glucose transporters in rat adipocytes: specific interchangeable effects of protein kinases C- and C-. Biochem J 337: Kohn AD, Summers SA, Birnbaum MJ, Roth RA 1996 Expression of a constitutively active Akt Ser/Thr kinase in 3T3/L1 adipocytes stimulates glucose uptake and glucose transporter 4 translocation. J Biol Chem 271: Tanti J, Grillo S, Grémeaux T, Coffer PJ, Van Obberghen E, Le Marchand- Brustel Y 1997 Potential role of protein kinase B in glucose transporter 4 translocation in adipocytes. Endocrinology 138: Cong L-N, Chen H, Li Y, Zhau L, McGibbon MA, Taylor SI, Quon MJ 1997

8 PKC- AND INSULIN-STIMULATED GLUCOSE TRANSPORT 4127 Physiologic role of Akt in insulin-stimulated translocation of GLUT4 in transfected rat adipose cells. Mol Endocrinol 11: Wang Q, Somwar R, Bilan PJ, Liu Z, Jin J, Woodgett JR, Klip A 1999 Protein kinase B/Akt participates in GLUT4 translocation in L6 myoblasts. Mol Cell Biol 19: Hill MM, Clark SF, Tucker DF, Birnbaum MJ, James DE, Macaulay SL 1999 A role for protein kinase B /Akt2 in insulin-stimulated GLUT4 translocation in adipocytes. Mol Cell Biol 19: Imamura T, Vollenweider P, Egawa K, Clodi M, Ishibashi K, Nakashima N, Ugi S, Adams JW, Brown JH, Olefsky JM 1999 G -q/11 protein plays a key role in insulin-induced glucose transport in 3T3/L1 adipocytes. J Biol Chem 274: Kotani K, Ogawa W, Matsumoto M, Kitamura T, Sakaue H, Hino Y, Miyake K, Sano W, Akimoto K, Ohno S, Kasuga M 1998 Requirement of atypical protein kinase C for insulin stimulation of glucose uptake but not for Akt activation in 3T3 L1 adipocytes. Mol Cell Biol 18: Standaert ML, Bandyopadhyay G, Perez L, Price D, Galloway L, Poklepovic A, Sajan MP, Cenni V, Sirri A, Moscat J, Toker A, Farese RV 1999 Insulin activates protein kinases C- and C- by an autophosphorylation-dependent mechanism and stimulates their translocation to GLUT4 vesicles and other membrane fractions in rat adipocytes. J Biol Chem 274: Akimoto K, Mizuno K, Osada S, Hirai S, Tanuma S, Suzuki K, Ohno S 1994 A new member of the third class in the protein kinase C family, PKC, expressed dominantly in an undifferentiated mouse embryonal carcinoma cell line and also in many tissues and cells. J Biol Chem 269: Bandyopadhyay G, Standaert ML, Sajan MP, Karnitz LM, Cong L, Quon MJ, Farese RV 1999 Dependence of insulin-stimulated glucose transporter 4 translocation on 3-phosphoinositide-dependent protein kinase-1 and its target threonine-410 in the activation loop of protein kinase C-. Mol Endocrinol 13: Doornbos RP, Theelen M, vanderhoeven PCJ, van Blitterswijk WJ, Verkleij AJ, van Bergen en Henegouwen PMP 1999 Protein kinase C is a negative regulator of protein kinase B activity. J Biol Chem 274: Sweeney G, Somwar R, Ramlal T, Volchuk A, Ueyama A, Klip A 1999 An inhibitor of p38 mitogen-activated protein kinase prevents insulin-stimulated glucose transport but not glucose transporter translocation in 3T3 L1 adipocytes and L6 myotubes. J Biol Chem 274: Hausdorff SF, Fingar DC, Morioka K, Garza LA, Whiteman EL, Summers SA, Birnbaum MJ 1999 Identification of wortmannin-sensitive targets in 3T3 L1 adipocytes. Dissociation of insulin-stimulated glucose uptake and GLUT4 translocation. J Biol Chem 274: Etgen GJ, Valasek KM, Broderick CL, Miller AR 1999 In vivo adenoviral delivery of recombinant human protein kinase C- stimulates glucose transport activity in rat skeletal muscle. J Biol Chem 274: