Rottlerin Inhibits Insulin-Stimulated Glucose Transport in 3T3-L1 ...

0013-7227/02/$15.00/0 Printed in U.S.A.

Endocrinology 143(10):3884 –3896 Copyright © 2002 by The Endocrine Society doi: 10.1210/en.2002-220259

Rottlerin Inhibits Insulin-Stimulated Glucose Transport in 3T3-L1 Adipocytes by Uncoupling Mitochondrial Oxidative Phosphorylation AYSE G. KAYALI, DARRELL A. AUSTIN,

AND

NICHOLAS J. G. WEBSTER

Medical Research Service (A.G.K., D.A.A., N.J.G.W.), San Diego Veterans Affairs Healthcare System, San Diego, California 92161; University of California San Diego/Whittier Diabetes Program (A.G.K., N.J.G.W.) and University of California San Diego Cancer Center (N.J.G.W.), Department of Medicine, University of California, San Diego, California 92093 There is increasing evidence that protein kinase C (PKC) isoforms modulate insulin-signaling pathways in both positive and negative ways. Recent reports have indicated that the novel PKC␦ mediates some of insulin’s actions in muscle and liver cells. Many studies use the specific inhibitor rottlerin to demonstrate the involvement of PKC␦. In this study, we investigated whether PKC␦ might play a role in 3T3-L1 adipocytes. We found that PKC␦ is highly expressed in mouse adipose tissue and increased on 3T3-L1 adipocyte differentiation, and insulin-stimulated glucose transport is blocked by rot-

T

HE PROTEIN KINASE C (PKC) family mediates downstream signaling from many growth factors and G protein-coupled receptors. This family of serine/threonine kinases comprises at least 10 isoforms that can be divided into three general categories: conventional PKCs (␣, ␤, and ␥) are dependent on calcium and diacylglycerol (DAG); the novel PKCs (␦, ⑀, ␩, and ␪) are dependent on DAG but not calcium; the atypical PKCs (␫, ␭, and ␨) do not depend on either calcium or DAG but are activated by inositol-phospholipids or phosphatidic acid (1– 4). The regulatory region of the conventional and novel PKCs contain C1 and C2 domains that bind DAG and calcium, respectively. There is also a related class of DAG-dependent kinases, including PKC␮/ PKD and PKC␯, which have somewhat larger regulatory regions that contain dual C1 domains (5–7). All of the DAGdependent PKC isoforms can be artificially activated by phorbol esters. Whether activation of PKC is involved in insulin’s metabolic effects has long been a matter of controversy. It was not until the identification, cloning, and characterization of the different PKC isoforms described above that definitive studies could be performed. There is now good evidence that PKC isoforms do indeed impinge upon insulin-signaling pathways in both positive and negative ways (8). Overexpression of the conventional PKCs has been shown to directly inhibit the kinase activity of the insulin receptor in vitro Abbreviations: AICAR, 5-Aminoimidazole-4-carboxamide ribonucleoside; AMPK, AMP-dependent kinase; DAG, diacylglycerol; DNP, dinitrophenol; FCCP, carbonyl cyanide p-trifluoromethoxyphenylhydrazone; GLUT4, glucose transporter 4; HDM, high-density microsome; IR, insulin receptor; IRS-1, insulin receptor substrate 1; LDM, low-density microsome; PI3K, phosphatidylinositol-3-kinase; PKC, protein kinase C; PMA, phorbol 12-myristate 13-acetate; PMSF, phenylmethylsulfonylfluoride.

tlerin. The phosphorylation state and activity of PKC␦ are not altered by insulin, but the protein translocates to membranes following insulin treatment. In contrast to the results with rottlerin, inhibition of PKC␦ activity or expression has no effect on glucose transport in adipocytes, unlike muscle cells. Lastly, we found that rottlerin lowers adenosine triphosphate levels in 3T3-L1 cells by acting as a mitochondrial uncoupler, and this is responsible for the observed inhibition of glucose transport. (Endocrinology 143: 3884 –3896, 2002)

through serine phosphorylation (9). Hyperglycemia activates the PKC␤ isoform selectively, thus leading to insulin resistance (10). On the other hand, ectopic expression of ␤II enhances insulin-stimulated glucose transport in the NIH3T3 cell, and insulin has been shown to modulate the alternative splicing of the PKC␤ isoform, favoring the ␤II splice variant, in L6 muscle cells (11, 12). Preventing the change in splicing blocks insulin-stimulated glucose transport and artificially causing the ␤II splice by transfection of the splicing factor SRp40 mimic the effect of insulin (13). Thus, PKC␤ has both negative and positive effects on insulin-signaling depending on the specific splice variant expressed. More recently, the atypical PKCs ␭ and ␨ have been shown to be activated by the phosphoinositide-dependent kinase 1 leading to glucose transporter-4 (GLUT4) translocation and stimulation of glucose transport in 3T3-L1, rat adipocytes, and L6 muscle cells (14 –18). Overexpression of wild-type PKC␨/␭ stimulates both basal and insulin-stimulated glucose transport, whereas a dominant-negative PKC␨/␭ blocks transport. More importantly, adenoviral overexpression of PKC␨ in the tibialis anterior muscle in rats elevated basal glucose transport in the perfused hind limb (19). Interestingly, PKC␨ and ␭ have identical effects on glucose transport, the ␭ isoform having a predominant role in 3T3-L1 adipocytes, in which ␨ is not expressed, and the ␨ isoform having a major role in primary adipocytes and muscle (15–18). The interpretation of these studies is often complicated by the fact that many of the reagents used to inhibit ␨ or ␭ are not isoform specific but will inhibit both. Once again, overexpression of an atypical PKC can also have negative effects because PKC␨ phosphorylates insulin receptor substrate (IRS)-1, impairing its ability to activate phosphatidylinositol 3-kinase (PI3K) in response to insulin in NIH3T3 and Fao cells (20, 21). These

3884

Kayali et al. • Rottlerin Uncouples 3T3-L1 Cells

isoforms are also activated by phosphatidic acid, generated by non-PI3K-dependent mechanisms, and may be the point of convergence in signaling to GLUT4 translocation (22, 23). A number of previous papers have proposed a role for novel PKCs in insulin action in muscle and liver, but there are no reports implicating PKC␦ in adipocytes (24, 25). Overexpression of PKCs ␦ and ␪ inhibits the kinase activity of the insulin receptor, and elevated free fatty acids specifically activate the PKC␪ isoform in muscle, causing insulin resistance (26 –29). On the positive side, expression of PKC␦ or ⑀ enhances insulin-stimulated glucose transport in NIH3T3 cells (11). Consistent with this, insulin increases PKC␦ activity in L6 muscle cells, HepG2 hepatoma cells, primary rat skeletal muscle cells, and mouse skin keratinocytes. Blockade of PKC␦ signaling via expression of a dominant-negative protein or transfection of antisense oligonucleotides prevents the stimulation of glucose transport in rat skeletal muscle cells, proliferation of skin keratinocytes, and stimulation of pyruvate dehydrogenase in HepG2 and L6 cells (24, 25, 30). Many studies of PKC␦ signaling have used rottlerin as a specific inhibitor, in conjunction with other methods, as evidence for the involvement of PKC␦ in a variety of biological processes (31). In this report, we investigated whether insulin signals via PKC␦ in the 3T3-L1 adipocyte cell model as has been shown for liver and muscle. We show that insulin causes a relocation of PKC␦ to different membrane fractions but does not change its phosphorylation state or activity. Inhibition of PKC␦ activity or expression has no effect on insulin-stimulated glucose transport. However, rottlerin inhibits glucose transport in a PKC␦-independent manner by acting as a mitochondrial uncoupler. Materials and Methods Materials and cell culture Porcine insulin was kindly provided by Eli Lilly & Co. (Indianapolis, IN). The PKC inhibitors were obtained from Calbiochem (San Diego, CA). Carbonyl cyanide p-trifluoromethoxyphenylhydrazone (FCCP) was from Sigma (St. Louis, MO). Protein A/G-agarose, src, and fyn antibodies were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). ERK1/2 antibodies (M12320), antibodies to individual PKC isoforms, and antiphosphotyrosine antibodies (pY20) were from BD-Transduction Laboratories, Inc. (Lexington, KY). Phospho-ERK antibodies (anti-ACTIVE MAPK) were from Promega Corp. (Madison, WI). Phospho-Akt (Ser473), phospho-PKC␦ (Thr505) antibodies were from Cell Signaling Inc. (Beverly, MA). PKC␦ antibodies for immunoprecipitation were from Zymed Laboratories, Inc. (South San Francisco, CA) and Santa Cruz Biotechnology, Inc.. Enhanced chemiluminescence reagents were from Amersham (Piscataway, NJ). Myristoylated cell-permeable inhibitory peptides to PKC␨ (SIYRRGARRWRKL), PKC␦ (AVRDMRQTVAVGVIKAV), and PKC␮ (AALVRQMAVAFFFK) were synthesized by Biosource Technologies, Inc. (Camarillo, CA) or United Biochemical Research Inc. (Seattle, WA). [3H]-2-Deoxyglucose (50 ␮Ci/mmol) was from Perkin-Elmer/NEN Life Science Products (Boston, MA). Unless noted, all other reagents were supplied by Sigma or Fisher Scientific Co. (Springfield, NJ). Mice were housed and handled under a protocol that had been approved by the University of California San Diego Animal Subjects Committee in accordance with Institutional Animal Care and Use Committee guidelines.

Cell culture The 3T3-L1 cells were maintained in DMEM with 25 mm glucose with 50 U/ml penicillin, 50 ␮g/ml streptomycin, and 10% fetal calf serum in

Endocrinology, October 2002, 143(10):3884 –3896 3885

a 10% CO2 environment. Cells were then differentiated 2 d after confluence by the addition of the media containing 500 ␮m isobutylmethylxanthine, 25 ␮m dexamethasone, and 4 ␮g/ml insulin. After 2 d, cells were grown in media containing only insulin for another 3 d. Differentiated adipocytes were replated into appropriate dishes. Subsequently, medium was changed to DMEM regular glucose with 10% fetal calf serum and replaced every 2 d until the cells were well differentiated (d 10) with many lipid droplets. For the antisense experiments, cells in 12-well plates were transfected with the PKC␦ antisense oligonucleotide (GAAGGAGATGCGCTGGAA) or the corresponding sense oligonucleotide (TTCCAGCGCATCTCCTTC) using Fugene 6 (Roche, Indianapolis, IN) following the manufacturer’s instructions. Phosphorothioate oligonucleotides were synthesized by GENSET Corp. (La Jolla, CA). Cells were transfected with oligos every 2 d until use.

Glucose transport Cells in 12-well plates were rinsed three times with PBS at 23 C and incubated with 0.5 ml Krebs-Ringer phosphate-HEPES buffer containing 0.5% fatty acid-free BSA for 2 h. Cells were stimulated with insulin (50 ng/ml) at 37 C for 30 min. In some experiments, cells were pretreated with various pharmacological inhibitors or cell-permeable peptides for 30 min. The transport reaction was initiated by adding 50 ␮l Krebs-Ringer phosphate-HEPES buffer containing 1.1 mm 2deoxyglucose (50 ␮m final) with [3H]2-deoxyglucose (0.2 ␮Ci/well). After a 10-min incubation, cells were washed three times with ice-cold PBS containing 100 mm phloretin, solubilized in 1% SDS, 0.1% NaOH, neutralized, and counted. Nonspecific uptake was measured in the presence of 25 ␮m cytochalasin B.

PKC␦ activity PKC␦ activity was measured with the Signatect assay system (Promega Corp.) following a protocol similar to Braiman et al. (24). Briefly, 3T3-L1 adipocytes were resuspended in cold lysis buffer [25 mm Tris (pH 7.4), 0.05% Triton X-100, 0.5 mm EDTA, 0.5 mm EGTA, 0.5 mm phenylmethylsulfonylfluoride (PMSF), 1 ␮g/ml leupeptin, and 1 ␮g/ml aprotinin] and lysed with a cold Dounce homogenizer. The clarified lysate was immunoprecipitated with anti-PKC␦ antibodies (Zymed Laboratories, Inc. or Santa Cruz Biotechnology, Inc.). Immune complexes were washed five times with ice-cold lysis buffer with 0.2 m NaCl and two times with kinase buffer [20 mm Tris (pH 7.5), 5 mm MgCl2, 0.25 mm EGTA, 0.4 mm CaCl2, 0.1 mg/ml BSA, 0.32 mg/ml phosphatidylserine]. Some reactions also contained 32 ng/ml diacylglycerol. The kinase reaction was performed in the same buffer containing 0.5 ␮Ci ␥32P-ATP, 100 ␮m ATP, and 100 ␮m biotinylated Neurogranin (28 – 43) peptide (AAKIQASFRGHMARKK) or a biotinylated EF-1␣ peptide (AVRDMRQTVAVGVIKAV) as a substrate for 5 min at 30 C and was terminated by the addition of half the volume of 7.5 m guanidine hydrochloride. Samples were spotted onto SAM capture membranes, washed extensively, and counted.

Immunoblotting The 3T3-L1 cells were serum starved for 16 h in 12-well plates and then stimulated with insulin for 5 or 10 min at 37 C. The cells were washed with ice-cold PBS and lysed in radioimmunoprecipitation assay buffer containing 20 mm Tris-HCl (pH 7.5), 1 mm EDTA, 140 mm NaCl, 1% Nonidet P-40, 1 mm sodium orthovanadate, 1 mm PMSF, and 10 ␮g/ml aprotinin. Samples were centrifuged at 2500 rpm for 10 min, fat aspirated, and pellet discarded. Equal amounts of protein from the control and insulin-treated lysates were solubilized in 2⫻ SDS-sample buffer. The proteins were denatured by boiling for 5 min, separated by electrophoresis on 7.5% SDS-PAGE, and transferred to polyvinylidenedifluoride membranes. The filter was blocked with 3% BSA in Trisbuffered saline-0.2% Tween-20 for 30 min and incubated with the antiphosphotyrosine, anti-phospho-Akt (Ser473), anti-phospho-ERK (Thr202/Tyr204), anti-phospho PKC␦ (Thr505), or anti-PKC␦ antibodies. For immunoblotting of individual PKC isoforms, whole-cell extracts from mouse white adipose tissue, 3T3-L1 fibroblasts, or 3T3-L1 adipocytes were made directly with 2⫻ SDS-sample buffer containing 2 mm sodium orthovanadate and 200 mm sodium fluoride. Samples were boiled and fractionated by SDS-PAGE, transferred to polyvinylidenedi-

3886

Endocrinology, October 2002, 143(10):3884 –3896


fluoride, and immunoblotted in a multichannel blotter (Hoefer, San Francisco, CA) using antibodies at a dilution of 1:1000. The filters were washed with Tris-buffered saline-0.2% Tween-20 for 30 min and incubated with horseradish-peroxidase conjugated secondary antibodies, and tyrosine-phosphorylated proteins were visualized by enhanced chemiluminescence.

Subcellular fractionation Cells were stimulated with 100 ng/ml insulin for 20 min and then washed three times with ice-cold PBS. Cells were scraped into ice-cold HES buffer [255 mm sucrose, 20 mm HEPES (pH 7.4), and 1 mm EDTA] supplemented with 100 mm sodium fluoride, 10 mm sodium pyrophosphate, 1 mm sodium vanadate, and protease inhibitors (1 ␮g/ml leupeptin, 1 ␮g/ml pepstatin, 1 mg/ml benzamidine, 0.5 mm PMSF). Cells were then homogenized using a Teflon/glass homogenizer. Subcellular fractionation was carried out as described previously (32). Internal and plasma membrane fractions were kept at ⫺70 C before protein quantification and immunoblotting with PKC␦ antibodies.

Immunostaining The 3T3-L1 adipocytes were plated on 10-mm acid-washed glass coverslips and stimulated with insulin for 30 min. Cells were washed with PBS and fixed with 3.7% formaldehyde in PBS for 20 min at room temperature. Following two washes in PBS, the cells were permeabilized and blocked in PBS containing 5% BSA and 0.5% Nonidet P-40 for 10 min. Coverslips were incubated with the anti-PKC␦ antibody (1:400 dilution) for 60 min at room temperature, washed once in PBS, and then incubated with fluorescein isothiocyanate-conjugated antirabbit IgG antibody (1:100 dilution) in PBS with 5% BSA and 0.5% Nonidet P-40 for 30 min at room temperature. Finally, the coverslips were extensively washed with PBS, rinsed with water, and mounted in PBS containing 15% gelvatol (polyvinyl alcohol), 33% glycerol, and 0.1% sodium azide.

14 12 10 8 6

*

4 2

*

Measurement of intracellular ATP The 3T3-L1 cells in 12-well plates were treated with increasing concentrations of rottlerin or FCCP for 30 min. Cells were washed twice with cold PBS and lysed in 0.1 m NaOH/0.5 mm EDTA. Extracts were incubated at 60 C for 20 min and frozen at ⫺20 C. Extracts were diluted 1:200 into 0.2 m NaOH/0.5 mm EDTA, and ATP was measured on 20 ␮l of the dilution by adding 100 ␮l luciferase/luciferin (Sigma) and measuring light output for 15 sec in a Turner Designs luminometer. Serial dilutions of ATP from 30 ␮m to 30 nm were used to generate a standard curve.

Results PKC isoform expression in fat cells

Initially we examined expression of PKC isoforms in mouse epididymal fat pads, and 3T3-L1 adipocytes before and after differentiation. Whole-cell extracts were immunoblotted for the conventional PKC isoforms ␣, ␤, and ␥; the novel PKC isoforms ␦, ⑀, and ␪; and the atypical PKC isoforms ␫, ␭, and ␨. We also blotted for the less related PKC␮ isoform. Epididymal fat pads express the novel isoforms ␦ and ⑀, the atypical isoforms ␫ and ␭, and a lower level of the conventional isoforms ␣ and ␤ (Fig. 1A). The antibodies to ␫ and ␭ cross-react, so we cannot distinguish between these two isoforms. The 3T3-L1 adipocytes express high levels of the novel isoforms ␦ and ⑀, and the atypical isoforms ␫ and ␭, both before and after differentiation. The conventional PKC isoforms ␣, ␤, and ␥ are expressed highly in the preadipocyte cells, but expression drops with differentiation (Fig. 1A). PKC␮ is expressed at similar levels in all three extracts, but PKC␪ and ␨ are not detectable in any of the three. The lack of expression of PKC␨ confirms previous findings (18). These results confirmed that the novel PKC isoform ␦ is expressed

FIG. 1. PKC isoform expression in 3T3-L1 adipocyte. A, Whole-cell extracts from mouse adipose tissue, 3T3-L1 preadipocytes, and 3T3-L1 adipocytes were immunoblotted with antibodies for individual PKC isoforms. B, The 3T3-L1 adipocytes were treated with increasing doses of rottlerin (0.3, 1, 3, 10 ␮M) for 30 min before stimulation with insulin (50 ng/ml) and measurement of 2-deoxyglucose uptake. Asterisks indicate statistical significance vs. insulin alone (P ⬍ 0.05).

at a high level in fat and is one of the major isoforms expressed in 3T3-L1 adipocytes. We then sought to determine whether PKC␦ is involved in insulin-stimulated glucose transport as has been shown for primary neonatal rat muscle cells and L6 muscle cells (Fig. 1B). Increasing concentrations of the inhibitor rottlerin caused a dose-dependent decrease in insulin-stimulated glucose transport (half-maximal effective dose, 2–3 ␮m) but had no effect on basal transport (data not shown). This finding suggested that PKC␦ might be involved in insulin-stimulated transport in adipocytes as has been shown for muscle cells. Insulin does not activate PKC␦

If insulin signals to glucose transport via PKC␦, then it is reasonable that insulin might activate the kinase. PKC␦ activation was monitored using an antibody against phosphoThr505. This phosphorylation site lies on the activation loop of PKC␦ and is phosphorylated by phosphoinositide-dependent kinase 1 (33). Whole-cell extracts from 3T3-L1 adipocytes stimulated with insulin were immunoblotted with the



antibody to pThr505 PKC␦ (Fig. 2A). This site was phosphorylated in the basal, unstimulated cells and phosphorylation did not increase with insulin stimulation for 2–30 min. The blots were stripped and reblotted for phospho-ERK to verify that the cells had been adequately starved of serum and that insulin caused an increase in ERK phosphorylation. Thus, phosphorylation of the activation loop residue is constitutive and not stimulated by insulin. We confirmed the immunoblotting data by measuring PKC␦ kinase activity. PKC␦ was immunoprecipitated from 3T3-L1 cells that had been stimulated with insulin for increasing times. The immunoprecipitates were subjected to an in vitro kinase assay using a biotinylated peptide substrate. Two different antibodies and two substrate peptides were used in independent experiments. No increase in total PKC␦ kinase activity was observed with insulin treatment in agreement with the immunoblotting data (Fig. 2B). This is in contrast to reports on muscle and liver cells in which insulin increases PKC␦ phosphorylation and activity. Insulin causes a redistribution of PKC␦

The novel PKC isoforms are calcium independent but are activated by DAG generated by phospholipase C. The DAG also serves to localize the kinase to membrane compartments. We have previously shown that insulin causes an increase in DAG levels in 3T3-L1 adipocytes (32). Although insulin does not activate PKC␦, the intracellular localization of the kinase may be altered by insulin treatment. Consequently, we monitored the cellular localization of PKC␦ in different membrane fractions of insulin-stimulated 3T3-L1 adipocytes. Insulin causes a biphasic increase in DAG in

A

Time: 0

2

5

10

15 30 min

pThr505 PKCδ

PKC␦ is associated with PKC␭ in 3T3-L1 adipocytes

PKC activity (% basal)

pERK

B

140 120 100 80 60 40 20 Time: 0

these cells, an initial transient peak at 30 sec, followed by a slower more sustained rise at 10 –30 min. Starved cells were stimulated for increasing times with insulin and membrane fractions prepared by sucrose gradient centrifugation. Initially we prepared plasma membrane and mitochondrial/ nuclear fractions because PKC␦ has been shown to translocate to these membranes in other cells. Insulin caused an increase in PKC␦ localization in both plasma membrane and mitochondrial/nuclear fractions by immunoblotting with either the pThr505 antibody or an antibody against native PKC␦ (Fig. 3A). Interestingly, the redistribution of PKC␦ coincides with the sustained increase in DAG at 15–30 min, rather than the acute peak of DAG at 30 sec. We subsequently investigated whether PKC␦ might relocate to other microsomal fractions. Cells were stimulated with insulin for increasing times and then fractionated into plasma membrane, mitochondrial/ nuclear, and low (LDM)- and high-density microsomal (HDM) fractions by sucrose gradient centrifugation. A longer time course of insulin treatment was used because the translocation occurred at 15–30 min in the initial experiments. Insulin caused an increase in pThr505-PKC␦ in both the mitochondrial/nuclear and plasma membrane fractions as before (Fig. 3A). Insulin also caused an increase in pThr505PKC␦ in the LDM fraction and a concomitant decrease in PKC␦ in the HDM fraction (Fig. 3B). The redistribution was confirmed by immunofluorescence. Cells were stimulated with insulin, fixed, and stained with an antibody to PKC␦. In the unstimulated state, PKC␦ staining is homogenous throughout the cell. Stimulation with insulin causes a redistribution of PKC␦ staining to membrane structures (Fig. 3C). Thus, insulin alters the intracellular localization but not activity of PKC␦ in 3T3-L1 adipocytes. This would be consistent with insulin signaling via PKC␦ and agrees with the relocalization of PKC␦ observed in muscle and liver cells.

0.5

1

5

10

30 min

FIG. 2. PKC␦ phosphorylation and activity, A, 3T3-L1 adipocytes were stimulated with insulin (50 ng/ml) for the indicated times. Whole-cell extracts were immunoblotted for phospho-PKC␦ (Thr505) or phospho-ERK (Thr202/Tyr204). B, 3T3-L1 adipocytes were stimulated with insulin (50 ng/ml) for the indicated times. PKC␦ was immunoprecipitated from the cell lysates and subjected to an immune-kinase assay using two different substrate peptides. Results from two different PKC␦ antibodies using two substrates were averaged (⫾ SEM).

PKC␦ has been shown to associate with a number of signaling proteins including the insulin receptor in muscle cells and PI3K in TF-1 erythroleukemia cells (26, 34). PKC␦ is tyrosine phosphorylated in response to insulin, phorbol 12myristate 13-acetate (PMA), platelet-derived growth factor, epithelial growth factor, ligation of the IgE receptor and various apoptotic stimuli, and serine phosphorylated by PKC␨ (24, 35–39). Therefore, we investigated whether PKC␦ could be found in association with components of these signaling pathways in 3T3-L1 adipocytes. Cell lysates from unstimulated and insulin-stimulated adipocytes were immunoprecipitated with antibodies to PKC␦. The pellets and supernatants were then immunoblotted for phosphotyrosine (Fig. 4A). Insulin causes an increase in tyrosine phosphorylation of the insulin receptor (IR) and IRS-1 in the supernatants, but no tyrosine-phosphorylated proteins were found in the pellets. Reblotting for PKC␦ demonstrated that the antibodies precipitated more than 70% of the protein. This result demonstrated that PKC␦ does not associate with the IR in 3T3-L1 cells, unlike muscle cells, and furthermore showed that PKC␦ is not tyrosine phosphorylated either in the basal or insulin-stimulated state. We subsequently in-

3888



FIG. 3. Subcellular fractionation and immunofluorescence. A, 3T3-L1 adipocytes were stimulated with insulin (50 ng/ml) for the indicated times. Cell lysates were fractionated into plasma membrane and mitochondrial/nuclear fractions by sucrose gradient centrifugation. Equal amounts of protein were immunoblotted with antibodies to phospho-PKC␦ (Thr505) or native PKC␦. B, Cells were stimulated for the indicated times and cell lysates fractionated into plasma membrane, HDMs, LDMs, and mitochondrial/nuclear fraction by sucrose density gradient centrifugation. Equal amounts of protein were immunoblotted for phospho-PKC␦ (Thr505). C, 3T3-L1 adipocytes were plated on acid-washed cover slips and stimulated with insulin (50 ng/ml) for 30 min. Cells were fixed, permeabilized, and stained with an antibody to PKC␦, followed by a rhodamine-conjugated secondary antibody. Cellular localization was determined by immunofluorescence microscopy.

A FIG. 4. Association of PKC␦ with PKC␭, A, 3T3-L1 adipocytes were stimulated with insulin for 15 min. Cell lysates were immunoprecipitated with antibodies to PKC␦ and the resulting pellets and supernatants immunoblotted with antibodies to phosphotyrosine. The positions of the tyrosinephosphorylated ␤-subunit of the IR and IRS-1 are indicated. The blots were stripped and reblotted for PKC␦. B, Identical cell lysates were immunoprecipitated for PKC␭, PKC␮, src, or fyn. The pellets and supernatants were immunoblotted for PKC␦.

B Insulin:

-

+

-

Insulin:

+ IRS1

I.P. PKCδ I.B. pY

IRβ

I.P. PKCδ I.B. PKCδ

PKCδ Pellets

vestigated whether PKC␦ is associated with PKC␭, PKC␮, src, or fyn. Three of these kinases, PKC␭, src, and fyn, have been shown to phosphorylate PKC␦ so they might be expected to interact. Cell lysates from unstimulated or insulinstimulated cells were immunoprecipitated with antibodies for PKC␭, PKC␮, src, or fyn, and then the pellets and su-

Sups

-

Sups. +

Pellets +

I.P. PKCλ I.B. PKCδ I.P. PKCµ I.B. PKCδ I.P. src I.B. PKCδ I.P. fyn I.B. PKCδ

pernatants were immunoblotted for PKC␦. Only antibodies to PKC␭ were able to precipitate PKC␦ (Fig. 4B). The finding of a complex between PKC␦ and PKC␭ is intriguing because PKC␭ has been shown to be critical for insulin signaling to GLUT4 translocation and stimulation of glucose transport in these cells.



Inhibition of PKC␦ activity does not inhibit glucose transport

so we used cell-permeable peptide inhibitors to inhibit individual PKC isoforms. Cells were pretreated with each cellpermeable peptide (100 ␮m) for 30 min before stimulation with insulin. The PKC␨ pseudosubstrate inhibitor, which also inhibits the other atypical PKCs, blocked insulin-stimulated glucose transport, as has been published previously, but equivalent peptides for PKC␦ and PKC␮ were without effect (Fig. 5B). We verified that the PKC␦ peptide was able to inhibit glucose transport in L6 muscle cells. Differentiated L6 myotubes were preincubated with rottlerin (10 ␮m) or the PKC␦ inhibitory peptide (100 ␮m) for 30 min before stimulation with insulin for 30 min. Insulin causes an approximate 2-fold increase in 2-deoxyglucose transport in these cells. Both rottlerin and the PKC␦ peptide reduced insulin-stimulated transport to basal levels (Fig. 5C), neither agent altered basal transport rates (data not shown). This result again argues against a role for PKC␦ in insulin-stimulated glucose transport in 3T3-L1 adipocytes unlike L6 muscle cells. The novel and conventional PKC isoforms are all DAG dependent. Consequently, these isoforms can be downregulated by chronic treatment with phorbol-esters. Cells were treated overnight with 100 nm PMA and whole-cell extracts immunoblotted for PKC isoform expression as before. The conventional isoforms ␣, ␤, and ␥ and the novel

The alteration in cellular localization of PKC␦ is consistent with a role in insulin signaling. The earlier results with rottlerin suggested a requirement for PKC␦ activity in insulinstimulated glucose transport. To confirm this, we performed a direct comparison of the effect of rottlerin on glucose transport with other well-characterized PKC inhibitors. The 3T3-L1 adipocytes were pretreated with increasing doses of rottlerin, calphostin C (a DAG antagonist), bisindolylmaleimide I (an ATP-binding site inhibitor selective for the ␣, ␤, ␥, ␦, and ⑀ isoforms of PKC), Go6976 (a PKC␣, ␤, and ␮ inhibitor), or Ro 31– 8220 (an inhibitor of conventional and novel PKCs) for 30 min, stimulated with insulin for 30 min, and then 2-deoxyglucose transport measured. As before, rottlerin caused a dose-dependent decrease in transport. Calphostin C, BIM, or Go6976 did not inhibit transport at concentrations that would inhibit PKC. Only Ro 31– 8220 inhibited transport at 10 –20 ␮m, a concentration that inhibits the atypical PKC isoforms ␨, ␭, and ␫ (Fig. 5A). Therefore, these results are inconsistent with insulin signaling via PKC␦ to transport but are consistent with the published data on PKC␭. All the chemical inhibitors used have nonspecific effects,

80 70 60 FIG. 5. Inhibition of conventional and novel PKCs does not inhibit glucose transport. A, 3T3-L1 adipocytes were treated with increasing doses of rottlerin, calphostin C, bisindolylmaleimide I (BIM), Go6976, or Ro 31-8220 for 30 min and then stimulated with insulin (50 ng/ml) and 2-deoxyglucose transport measured. Results are the mean and SEM of three experiments in triplicate. B, 3T3-L1 adipocytes were treated with cell-permeable inhibitory peptides to PKC␨, PKC␦, or PKC␮ (100 ␮M) for 30 min and then stimulated with insulin (50 ng/ml) for 30 min and 2-deoxyglucose transport measured. Results are the mean and SEM of three to five experiments in triplicate. Asterisks indicate statistical significance vs. insulin alone (P ⬍ 0.05). C, L6 myotubes were treated with the cell-permeable PKC␦ inhibitory peptide (100 ␮M) or rottlerin (10 ␮M) for 30 min and then stimulated with insulin (50 ng/ml) for 30 min and 2-deoxyglucose transport measured. Results are the mean and SEM of two experiments. Asterisks indicate statistical significance vs. insulin alone (P ⬍ 0.05).

50 40 30

∗

∗

∗

20 10

∗

∗

0

* * ∗

3890



PKC isoform ␦ were not detected showing complete downregulation (Fig. 6A). Expression of the novel PKC isoform ⑀ was reduced greater than 90%, but the atypical PKC isoforms ␫ and ␭ and PKC␮ were not altered. Basal and insulin-stimulated glucose transport were measured on the PMA downregulated cells. PMA treatment raised basal transport 2.5fold but had no effect on insulin-stimulated transport (Fig. 6B). Therefore, insulin is still able to stimulate glucose transport in the absence of PKC␦. Interestingly, pretreatment of down-regulated cells with rottlerin still blocked insulin-stimulated transport, strongly suggesting that the effect of rottlerin is not due to inhibition of PKC␦. Transport is still sensitive to wortmannin, so the mechanism of insulin stimulation is likely unchanged by the PMA down-regulation. Inhibition of PKC␦ expression does not inhibit glucose transport

The pharmacological methods used above to inhibit PKC␦ activity could be criticized for nonspecific effects. To obtain evidence for a role for PKC␦ in insulin-stimulated glucose transport using an independent approach, we reduced PKC␦ expression using antisense oligonucleotides. We initially determined the stability of the PKC␦ protein by inhibiting protein synthesis with cycloheximide. The half-life of PKC␦ in 3T3-L1 adipocytes is 24 h. Complete elimination of the protein can be seen by 96 h (Fig. 7A). Consequently, we transfected cells with two doses of antisense oligonucleotides to PKC␦ for 4 and 5 d. The oligonucleotide sequence was pub-

A

B

2-DOG Transport (fold over basal)

PKC

α

β

δ

ε

γ

ι

λ

µ

8 7

θ

ζ

DMSO

6 5 4 3

PMA

∗

∗

∗ ∗

2 1 Basal

Insulin

Ins + Rottlerin

Ins + Wortmannin

FIG. 6. Down-regulation of PKCs with PMA does not inhibit glucose transport. A, 3T3-L1 adipocytes were treated overnight with 100 nM PMA. Whole-cell extracts were immunoblotted with antibodies to individual PKC isoforms as before. B, PMA-down-regulated cells were treated with rottlerin (10 ␮M) or wortmannin (100 nM) for 30 min and then stimulated with insulin (50 ng/ml) and 2-deoxyglucose transport measured. Results are the mean and SEM of two experiments in triplicate. Asterisks indicate statistical significance vs. insulin alone (P ⬍ 0.05).

lished previously (40). Whole-cell extracts were immunoblotted for PKC␦. Transfection of cells with 5 ␮g antisense oligonucleotide for 5 d causes an 80% reduction in PKC␦ (Fig. 7B). These conditions were then used for transport studies. Adipocytes were transfected with antisense or sense oligonucleotides to PKC␦ for 5 d. Basal and insulin-stimulated transport were measured. The antisense oligonucleotide to PKC␦ had no effect on basal or insulin-stimulated glucose transport (Fig. 7C). Therefore, neither inhibition of PKC␦ activity or reduction of PKC␦ expression reduces glucose transport. Rottlerin is a mitochondrial uncoupler in 3T3-L1 adipocytes

The results from the previous sections indicate that PKC␦ is not involved in glucose transport, but rottlerin, a supposedly specific PKC␦ inhibitor, reduces insulin-stimulated transport. Obviously, rottlerin must have another target within the cell. Rottlerin does not inhibit glucose transport directly when added to a transport assay but requires pretreatment (data not shown). It was recently shown that rottlerin functions as a mitochondrial uncoupler to increase oxygen consumption and reduce ATP levels in rat parotid acinar cells (41). To test whether rottlerin uncouples 3T3-L1 adipocytes, cells were treated with increasing doses of rottlerin (1, 3, and 10 ␮m) or the chemical uncoupler FCCP (1, 3, and 10 ␮m) for 30 min and intracellular ATP levels were measured on cell extracts using luciferase/luciferin. Both rottlerin and FCCP cause a dose-dependent decrease in cellular ATP levels. At the maximal concentration (10 ␮m), both agents reduce ATP levels by 80% (Fig. 8A). In parallel, we determined the effects of increasing doses of rottlerin and FCCP on glucose transport. Both agents cause a similar dosedependent decrease in insulin-stimulated glucose transport (Fig. 8B). This is the first demonstration that FCCP inhibits glucose transport in adipocytes and suggests that the ability of rottlerin to inhibit transport is related to its ability to act as a mitochondrial uncoupler. Dinitrophenol (DNP) was used as an antiobesity drug for many years because it increases energy expenditure by uncoupling oxidative phosphorylation. We tested whether reduction of ATP levels using DNP would also inhibit glucose transport. Cells were treated with 0.5 mm DNP for 30 min before stimulation with insulin for 30 min and measurement of 2-deoxyglucose transport. In parallel, cellular ATP levels were determined using luciferase/luciferin. Treatment of 3T3-L1 cells with DNP caused an 80% reduction in ATP and a more than 90% reduction in glucose transport (Fig. 8C), confirming the correlation between ATP levels and transport that was seen with rottlerin and FCCP. Rottlerin reduces insulin-stimulated activation of Akt and ERK

Given the ability of rottlerin to reduce ATP levels, one conceivable mechanism for the inhibition of transport is due to reduced kinase activity. To investigate this possibility, cells were treated with increasing doses of rottlerin for 30 min and then stimulated with insulin. Whole-cell extracts were immunoblotted with antibodies to phosphotyrosine. Tyrosine phosphorylation of the insulin receptor and IRS-1 are



FIG. 7. Antisense elimination of PKC␦ does not inhibit glucose transport. A, 3T3-L1 adipocytes were treated with cycloheximide (100 ␮g/ml) and harvested at varying times thereafter. Equal amounts of protein were immunoblotted for PKC␦. B, 3T3-L1 adipocytes were transfected with two doses of antisense oligonucleotide to PKC␦ every 2 d. Cells were harvested after 4 and 5 d and equal amounts of protein immunoblotted for PKC␦. C, 3T3-L1 adipocytes were transfected with 5 ␮g antisense or sense oligonucleotide to PKC␦ every 2 d for 5 d. Cells were stimulated with insulin (50 ng/ml) and 2-deoxyglucose transport measured. Results are mean and SEM of two experiments in triplicate.

preserved, even at the highest concentration of rottlerin that reduces ATP levels by 80% (Fig. 9A). The effects on Akt activation are more pronounced however. Phosphorylation of Akt on Ser473 is reduced in a dose-dependent manner with increasing concentrations of rottlerin (Fig. 9B). Rottlerin has been shown to shorten activation of ERK in PC12 cells (40). Therefore, cells were treated with 10 ␮m rottlerin for 30 min and stimulated with insulin for increasing times. Wholecell extracts were immunoblotted with antibodies to the dually phosphorylated form of ERK (Thr204/Tyr204). Insulin can still activate ERK at early times points (5 and 10 min) in the presence of rottlerin, but the kinase is rapidly dephosphorylated by 30 min (Fig. 9C). Thus, the prolonged activation of ERK by insulin in 3T3-L1 adipocytes is prevented by rottlerin. These studies suggested that rottlerin may inhibit insulin signaling because phosphorylation of a number of key proteins is reduced. To determine whether this defect is responsible for the reduced glucose transport, we performed a time course of rottlerin treatment in 3T3-L1 adipocytes. Cells were treated with insulin (50 ng/ml) for 30 min before measuring 2-deoxyglucose transport. Rottlerin (10 ␮m) was added at different times relative to the insulin: 30 and 15 min before insulin; at the same time as insulin; or 5, 10, and 15 min after insulin. Addition of rottlerin before or at the same time as insulin blocked glucose transport (Fig. 9D). More importantly, addition of rottlerin 5 min after insulin was still able to block glucose transport, but the inhibitory effect was progressively lost at later times. This is significant because activation of signaling via the insulin receptor occurs within the first 5 min after insulin stimulation. The reduced glucose transport is therefore not simply a reflection of impaired signaling. Rottlerin causes irreversible uncoupling of mitochondria

Reductions in ATP levels with rottlerin and FCCP are observed rapidly, within 5 min of addition. To determine whether rottlerin is acting as a chemical uncoupler like FCCP or DNP, we performed a wash-out experiment. Cells were

treated with 10 ␮m rottlerin or FCCP for 30 min, and then the cells were washed and the medium was changed to medium lacking the uncouplers. Cells were allowed to recover for 0, 30, 60, or 90 min, and then ATP levels were measured using luciferase/luciferin. The reduced ATP levels because of treatment with FCCP returned to normal after a 30-min wash-out (Fig. 10A). Unexpectedly, ATP levels following rottlerin treatment did not recover, even after a 90-min wash-out. In parallel cells, we measured insulin-stimulated glucose transport. Following the wash-out of rottlerin and FCCP for different times, cells were stimulated with insulin for 30 min and 2-deoxygluocse transport measured. The impaired glucose transport in cells treated with rottlerin did not recover, even after a 90-min wash-out consistent with the continued reduction in ATP levels (Fig. 10B). The results with FCCP were more striking. Glucose transport was normalized in cells that were stimulated with insulin immediately after removal of the FCCP. Allowing cells to recover for 30 min or longer before insulin stimulation led to a superinduction of transport. Therefore, rottlerin is not acting as a chemical uncoupler but must irreversibly inhibit a component of the electron transport chain. Discussion

Because of the evidence for PKC␦ involvement in insulin action in other cell types, we decided to investigate whether this PKC isoform might play a role in 3T3-L1 adipocytes. Initial experiments demonstrated that PKC␦ is highly expressed in 3T3-L1 adipocytes and mouse fat, and rottlerin inhibited glucose transport in a dose-dependent manner. However, we were not able to document an increase in either phosphorylation of PKC␦ or activity with insulin. The insulin receptor has been shown to associate with PKC␦ in muscle and NIH3T3 cells, but we did not observe any association in 3T3-L1 cells, which may explain the lack of phosphorylation. We did observe, however, the translocation of PKC␦ to the plasma membrane, LDM, and nuclear/mitochondrial fractions by sucrose gradient centrifugation and immunofluorescence. This would be consistent with insulin causing a relocation of PKC␦ without changing its inherent activity.

3892


5000 4500 4000 3500 3000 2500 2000 1500 1000 500


Rottlerin FCCP

∗ ∗

∗ ∗

∗

0

5

10

30 25 ∗ ∗

20 15

Rottlerin FCCP ∗

10 ∗

∗ ∗

5

0

5

10

100

60 ATP 2-DOG

80

40

60 40 20

*

20 *

0

FIG. 8. Rottlerin decreases ATP levels. A, 3T3-L1 adipocytes were treated with increasing doses of rottlerin or the chemical uncoupler FCCP for 30 min, and cellular ATP levels were measured using a luciferase/luciferin assay. B, Cells were treated with increasing doses of rottlerin and FCCP as before and stimulated with insulin (50 ng/ml) and 2-deoxyglucose transport measured. Results are the mean and SEM of two experiments in triplicate. Asterisks indicate statistical significance vs. cells treated with vehicle alone (P ⬍ 0.05). C, Parallel cultures of 3T3-L1 cells were treated with 0.5 mM DNP for 30 min. Cellular ATP levels were measured on one set of cells using the luciferase/luciferin assay and are expressed as percent of basal ATP level (mean ⫹ SEM). The other set of cells was stimulated with insulin for 30 min and 2-deoxyglucose transport measured. Results are the mean and SEM of two experiments in triplicate. Asterisks indicate statistical significance vs. cell without DNP (P ⬍ 0.05).

PKC␦ activity was inhibited using a variety of pharmacological agents or cell-permeable inhibitor peptides. We were able to document the importance of PKC␭, as has been published, but inhibition of PKC␦ had no effect on glucose transport. Similarly, elimination of PKC␦ using antisense oligo-

nucleotides or down-regulation with chronic PMA also had no effect. Therefore, we were left with the conundrum that rottlerin inhibited glucose transport in a PKC␦-independent manner. The ability of rottlerin to inhibit PKC␦ has been questioned recently. Davies et al. (42) tested a series of kinase inhibitors, including rottlerin, against a panel of recombinant kinases. Rottlerin was unable to inhibit PKC␦ in this assay but was able to inhibit MAPK-activated kinase 2 (IC50 5.4 ␮m) and p38-regulated/activated kinase (IC50 1.9 ␮m). It has also been reported to inhibit calmodulin-dependent kinase III (31). Similarly, in another study rottlerin was unable to inhibit immunoprecipitated PKC␦ activity in the presence of DAG and phosphatidylserine and unexpectedly stimulated kinase activity in their absence (41). In this latter study, they showed that rottlerin reduces cellular ATP levels by uncoupling mitochondrial oxidative phosphorylation. The drop in ATP levels was sufficient to prevent the tyrosine phosphorylation of PKC␦ in response to carbachol or PMA. This effect was seen only in primary parotid acinar cells but not in the PC12 and RPG1 cell lines. It was proposed that primary cells rely on oxidative phosphorylation for ATP production, whereas cultured cell lines rely on both glycolysis and oxidative phosphorylation. Hence, rottlerin did not function as an uncoupler in either PC12 or RPG1 cells, even though rottlerin increased oxygen consumption in these cells. We show here that rottlerin is able to uncouple 3T3-L1 adipocytes similar to the chemical uncouplers FCCP and DNP. This would suggest that ATP production in these adipocytes is primarily driven by oxidative phosphorylation rather than glycolysis. This is consistent with a switch in energy metabolism during differentiation. We also show that the drop in ATP levels is sufficient to inhibit insulin signaling because Akt and ERK phosphorylation are impaired. Both FCCP, DNP, and rottlerin decreased insulin-stimulated glucose transport in parallel with the decrease in ATP. Basal transport was not altered by these agents, so the ability of transporters already resident in the plasma membrane to transport glucose is not dependent on ATP. There are two possible explanations for the decreased transport, either an insulin-signaling defect because of the impaired kinase phosphorylation or a failure of GLUT4-containing vesicles to translocate to the plasma membrane. The trafficking of GLUT4 vesicles is mediated by ATP-dependent microtubule motors, so the lowered ATP level may directly inhibit vesicle motion. The impairment in signaling is not the sole explanation for the reduced transport because addition of rottlerin 5 min after insulin was still able to inhibit transport. Activation of PI3K signaling is maximal within 5 min of insulin stimulation, so the inhibition by rottlerin is unrelated to signaling in this experiment. We also found that the uncoupling effect of rottlerin is irreversible. It is unlikely, therefore, that rottlerin is a chemical uncoupler, but it may directly inhibit a component of the electron transport chain. This notion is supported by the observation that the mitochondrial toxin rotenone, which is a competitive inhibitor of complex I (nicotinamide-adenine dinucleotidecoenzyme Q) of the electron transport chain, also inhibits


FIG. 9. Rottlerin impairs insulin activation of Akt and ERK. The 3T3-L1 cells were treated with increasing doses of rottlerin for 30 min and stimulated with insulin for 5 min. Whole-cell extracts were immunoblotted for phosphotyrosine. B, Cells were treated with increasing doses of rottlerin and stimulated with insulin. Whole-cell extracts were immunoblotted for phospho-Akt (Ser473). C, Cells were treated with rottlerin (10 ␮M) for 30 min and stimulated with insulin for the indicated times. Whole-cell extracts were immunoblotted for phospho-ERK (Thr202/Tyr204). D, Cells were stimulated with insulin (50 ng/ml) for 30 min before measurement of 2-deoxyglucose transport. Rottlerin (10 ␮M) was added to the cultures 30 and 15 min before insulin stimulation (⫺30 and ⫺15); at the same time as insulin stimulation (0); or 5, 10, or 15 min after insulin stimulation (⫹5, ⫹10, and ⫹15). Results are the mean and SEM of three determinations. Asterisks indicate statistical significance vs. insulin alone (P ⬍ 0.05).


16.00 14.00 12.00 10.00

*

8.00

*

6.00 4.00

*

*

2.00

glucose transport in a manner similar to the chemical uncouplers (data not shown). A number of papers have investigated the effects of lowering ATP levels on glucose transport. Glucosamine causes insulin resistance and reduces cellular ATP levels. Hresko et al. (43) found that treatment of 3T3-L1 adipocytes with glucosamine for 2.5 h in glucose-free medium blocked insulinstimulated glucose transport and caused an 80% reduction in ATP levels. Severe depletion of ATP with azide mimicked the effect of glucosamine and impaired insulin signaling and GLUT4 translocation. The authors reported a tight correlation between cellular ATP levels and the magnitude of the insulin-induced increase in glucose transport. Kang et al. (44) reported that treatment of 3T3-L1 adipocytes with glucosamine for 5 h in medium containing glucose lowered ATP levels by a more modest 15% but still impaired insulinstimulated glucose transport by 50%. In the second article, glucosamine treatment was performed in the presence of glucose, which is preferentially transported over glucosamine, resulting in a milder reduction in ATP. Treatment of 3T3-L1 cells with 0.05 mm dinitrophenol or 3 mm azide to obtain a similar reduction in ATP was insufficient to inhibit insulin-stimulated glucose transport but, unlike glucosamine, increased basal glucose transport 4-fold. Higher

doses of DNP or azide to reduce ATP levels by 40% were able to inhibit insulin signaling, but the effects on transport were identical with the lower doses leading the authors to conclude that a mild reduction in ATP levels per se does not cause insulin resistance. The results presented here are consistent with the report of Hresko et al. The acute reduction in ATP levels with rottlerin and FCCP is very dramatic (⬎80%). This reduction is sufficient to impair insulin signaling and reduce glucose transport without activating basal transport. The induction of basal transport by a prolonged mild reduction of ATP reported by Kang et al. (44) is similar to effects seen in muscle. Treatment of muscle strips or L6 muscle cells with mitochondrial uncouplers leads to an increase in glucose transport (45– 48). One possible explanation for this increase is the activation of PKC signaling because of the release of calcium from the mitochondria upon depletion of ATP (46). Elevation of intracellular calcium to levels insufficient to cause contraction increases glucose transport (49). Alternatively, reduction of ATP levels and concomitant increase in the ADP/ATP ratio leads to the activation of the AMP-dependent kinase (AMPK) (50). This kinase can directly activate glucose transport by causing translocation of GLUT4 vesicles (51). The kinase can be activated artificially by the compound 5-aminoimidazole-4-carbozamide ribonu-

3894



*

*

*

*

* FIG. 10. Rottlerin is an irreversible inhibitor of mitochondrial function. The 3T3-L1 cells were treated with rottlerin (10 ␮M) or FCCP (10 ␮M) for 30 min. A, Rottlerin and FCCP were washed off for 0, 30, 60, or 90 min and cellular ATP levels were determined using the luciferase/luciferin assay. Results are the mean and SEM of three determinations and are expressed as relative light units (RLUs). Asterisks indicate statistical significance vs. basal untreated cells (P ⬍ 0.05). B, Rottlerin and FCCP were washed off for 0, 30, 60, or 90 min and cells were stimulated with insulin for 30 min and 2-deoxyglucose transport measured. Results are the mean and SEM of three determinations. Asterisks indicate statistical significance vs. insulin-stimulated cells (P ⬍ 0.05).

*

*

0

cleoside (AICAR), which is phosphorylated inside the cell to 5-aminoimidazole-4-carboxamide ribonucleotide, an AMP mimic. Treatment of muscle cells or strips with AICAR stimulates glucose transport to the same extent as insulin or contractions (52–54). Indeed, contractions activate AMPK␣2, and the effects of AICAR and contractions are not additive, which has lead to the suggestion that AMPK mediates exercise-induced transport (55, 56). Treatment of 3T3-L1 adipocytes with AICAR stimulates basal transport only 2-fold and impairs insulin-stimulated transport by 50% (57). The increase in basal transport is similar to the 4-fold induction due to chronic 5-h treatment with low doses of DNP or azide as reported by Kang et al. (44). So the stimulatory effect observed with mild ATP depletion may be linked to activation of AMPK. We did not observe an increase in basal transport with acute reductions in ATP, but it may take longer to manifest an effect on AMPK. In summary, we have shown that PKC␦ is highly expressed in 3T3-L1 adipocytes. Its phosphorylation state and activity are not altered by insulin, but the protein translocates to membranes following insulin treatment. Inhibition of PKC␦ activity or expression has no effect on glucose trans-

*

30

*

60

*

*

*

90

0

30

60

90 min

port in adipocytes, unlike muscle cells. Lastly, the effects of rottlerin to inhibit glucose transport are most likely related to its ability to act as a mitochondrial uncoupler and reduce ATP levels. Acknowledgments Received March 4, 2002. Accepted June 14, 2002. Address all correspondence and requests for reprints to: Nicholas J. G. Webster, Department of Medicine (0673), University of California, San Diego, 9500 Gilman Drive, La Jolla, California 92093-0673. E-mail: [email protected]. This work was supported by a Diabetes Research Center Grant from the VA and the Juvenile Diabetes Foundation and a fellowship from the Drown Foundation (Los Angeles, CA) (to A.G.K.).

References 1. Dempsey EC, Newton AC, Mochly-Rosen D, Fields AP, Reyland ME, Insel PA, Messing RO 2000 Protein kinase C isozymes and the regulation of diverse cell responses. Am J Physiol Lung Cell Mol Physiol 279:L429 –L438 2. Newton AC 2001 Protein kinase C: structural and spatial regulation by phosphorylation, cofactors, and macromolecular interactions. Chem Rev 101:2353–2364 3. Ron D, Kazanietz MG 1999 New insights into the regulation of protein kinase C and novel phorbol ester receptors. FASEB J 13:1658 –1676


4. Toker A 1998 Signaling through protein kinase C. Front Biosci 3:D1134 –D1147 5. Valverde AM, Sinnett-Smith J, Van Lint J, Rozengurt E 1994 Molecular cloning and characterization of protein kinase D: a target for diacylglycerol and phorbol esters with a distinctive catalytic domain. Proc Natl Acad Sci USA 91:8572– 8576 6. Hayashi A, Seki N, Hattori A, Kozuma S, Saito T 1999 PKCnu, a new member of the protein kinase C family, composes a fourth subfamily with PKCmu. Biochim Biophys Acta 1450:99 –106 7. Sturany S, Van Lint J, Muller F, Wilda M, Hameister H, Hocker M, Brey A, Gern U, Vandenheede J, Gress T, Adler G, Seufferlein T 2001 Molecular cloning and characterization of the human protein kinase D2. A novel member of the protein kinase D family of serine threonine kinases. J Biol Chem 276: 3310 –3318 8. Formisano P, Beguinot F 2001 The role of protein kinase C isoforms in insulin action. J Endocrinol Invest 24:460 – 467 9. Bollag GE, Roth RA, Beaudoin J, Mochly-Rosen D, Koshland Jr DE 1986 Protein kinase C directly phosphorylates the insulin receptor in vitro and reduces its protein-tyrosine kinase activity. Proc Natl Acad Sci USA 83:5822– 5824 10. King GL 1996 The role of hyperglycaemia and hyperinsulinaemia in causing vascular dysfunction in diabetes. Ann Med 28:427– 432 11. Cooper DR, Watson JE, Patel N, Illingworth P, Acevedo-Duncan M, Goodnight J, Chalfant CE, Mischak H 1999 Ectopic expression of protein kinase C␤II, -␦, and -⑀, but not -␤I or -␨, provide for insulin stimulation of glucose uptake in NIH-3T3 cells. Arch Biochem Biophys 372:69 –79 12. Chalfant CE, Mischak H, Watson JE, Winkler BC, Goodnight J, Farese RV, Cooper DR 1995 Regulation of alternative splicing of protein kinase C␤ by insulin. J Biol Chem 270:13326 –13332 13. Patel NA, Chalfant CE, Watson JE, Wyatt JR, Dean NM, Eichler DC, Cooper DR 2001 Insulin regulates alternative splicing of protein kinase C␤II through a phosphatidylinositol 3-kinase-dependent pathway involving the nuclear serine/arginine-rich splicing factor, SRp40, in skeletal muscle cells. J Biol Chem 276:22648 –22654 14. 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 138:4721– 4731 15. 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 337:461– 470 16. 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 274:25308 –25316 17. 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 272:30075–30082 18. 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 18:6971– 6982 19. 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 274:22139 –22142 20. Ravichandran LV, Esposito DL, Chen J, Quon MJ 2001 Protein kinase C-␨ phosphorylates insulin receptor substrate-1 and impairs its ability to activate phosphatidylinositol 3-kinase in response to insulin. J Biol Chem 276:3543– 3549 21. Liu YF, Paz K, Herschkovitz A, Alt A, Tennenbaum T, Sampson SR, Ohba M, Kuroki T, LeRoith D, Zick Y 2001 Insulin stimulates PKC␨-mediated phosphorylation of insulin receptor substrate-1 (IRS-1). A self-attenuated mechanism to negatively regulate the function of IRS proteins. J Biol Chem 276:14459 –14465 22. Bandyopadhyay G, Sajan MP, Kanoh Y, Standaert ML, Quon MJ, Reed BC, Dikic I, Farese RV 2001 Glucose activates protein kinase C-zeta/lambda through proline-rich tyrosine kinase-2, extracellular signal-regulated kinase, and phospholipase D. J Biol Chem 276:35537–35545 23. Limatola C, Schaap D, Moolenaar WH, van Blitterswijk WJ 1994 Phosphatidic acid activation of protein kinase C-␨ overexpressed in COS cells: comparison with other protein kinase C isotypes and other acidic lipids. Biochem J 304(Pt 3):1001–1008 24. Braiman L, Alt A, Kuroki T, Ohba M, Bak A, Tennenbaum T, Sampson SR 1999 Protein kinase C␦ mediates insulin-induced glucose transport in primary cultures of rat skeletal muscle. Mol Endocrinol 13:2002–2012 25. Caruso M, Maitan MA, Bifulco G, Miele C, Vigliotta G, Oriente F, Formisano P, Beguinot F 2001 Activation and mitochondrial translocation of protein kinase C␦ are necessary for insulin stimulation of pyruvate dehydrogenase complex activity in muscle and liver cells. J Biol Chem 276:45088 – 45097


26. Braiman L, Alt A, Kuroki T, Ohba M, Bak A, Tennenbaum T, Sampson SR 2001 Insulin induces specific interaction between insulin receptor and protein kinase C␦ in primary cultured skeletal muscle. Mol Endocrinol 15:565–574 27. Chalfant CE, Ciaraldi TP, Watson JE, Nikoulina S, Henry RR, Cooper DR 2000 Protein kinase C␪ expression is increased upon differentiation of human skeletal muscle cells: dysregulation in type 2 diabetic patients and a possible role for protein kinase C␪ in insulin-stimulated glycogen synthase activity. Endocrinology 141:2773–2778 28. Schmitz-Peiffer C, Browne CL, Oakes ND, Watkinson A, Chisholm DJ, Kraegen EW, Biden TJ 1997 Alterations in the expression and cellular localization of protein kinase C isozymes ⑀ and ␪ are associated with insulin resistance in skeletal muscle of the high-fat-fed rat. Diabetes 46:169 –178 29. Kellerer M, Mushack J, Seffer E, Mischak H, Ullrich A, Haring HU 1998 Protein kinase C isoforms ␣, ␦ and ␪ require insulin receptor substrate-1 to inhibit the tyrosine kinase activity of the insulin receptor in human kidney embryonic cells (HEK 293 cells). Diabetologia 41:833– 838 30. Shen S, Alt A, Wertheimer E, Gartsbein M, Kuroki T, Ohba M, Braiman L, Sampson SR, Tennenbaum T 2001 PKC␦ activation: a divergence point in the signaling of insulin and IGF-1-induced proliferation of skin keratinocytes. Diabetes 50:255–264 31. Gschwendt M, Muller HJ, Kielbassa K, Zang R, Kittstein W, Rincke G, Marks F 1994 Rottlerin, a novel protein kinase inhibitor. Biochem Biophys Res Commun 199:93–98 32. Eichhorn J, Kayali AG, Austin DA, Webster NJ 2001 Insulin activates phospholipase C-␥1 via a PI-3 kinase dependent mechanism in 3T3–L1 adipocytes. Biochem Biophys Res Commun 282:615– 620 33. Le Good JA, Ziegler WH, Parekh DB, Alessi DR, Cohen P, Parker PJ 1998 Protein kinase C isotypes controlled by phosphoinositide 3-kinase through the protein kinase PDK1. Science 281:2042–2045 34. Ettinger SL, Lauener RW, Duronio V 1996 Protein kinase C␦ specifically associates with phosphatidylinositol 3-kinase following cytokine stimulation. J Biol Chem 271:14514 –14518 35. Soltoff SP, Toker A 1995 Carbachol, substance P, and phorbol ester promote the tyrosine phosphorylation of protein kinase C␦ in salivary gland epithelial cells. J Biol Chem 270:13490 –13495 36. Benes C, Zheng Y, Soltoff SP 2000 Activation of PKC␦ by tyrosine phosphorylation in rat parotid acinar cells. J Korean Med Sci 15(Suppl):S40 –S41 37. Denning MF, Dlugosz AA, Threadgill DW, Magnuson T, Yuspa SH 1996 Activation of the epidermal growth factor receptor signal transduction pathway stimulates tyrosine phosphorylation of protein kinase C␦. J Biol Chem 271:5325–5331 38. Zang Q, Lu ZM, Curto M, Barile N, Shalloway D, Foster DA 1997 Association between v-Src and protein kinase C␦ in v-Src-transformed fibroblasts. J Biol Chem 272:13275–13280 39. Blass M, Kronfeld I, Kazimirsky G, Blumberg PM, Brodie C 2002 Tyrosine phosphorylation of protein kinase C␦ is essential for its apoptotic effect in response to etoposide. Mol Cell Biol 22:182–195 40. Corbit KC, Foster DA, Rosner MR 1999 Protein kinase Cdelta mediates neurogenic but not mitogenic activation of mitogen-activated protein kinase in neuronal cells. Mol Cell Biol 19:4209 – 4218 41. Soltoff SP 2001 Rottlerin is a mitochondrial uncoupler that decreases cellular ATP levels and indirectly blocks protein kinase C␦ tyrosine phosphorylation. J Biol Chem 276:37986 –37992 42. Davies SP, Reddy H, Caivano M, Cohen P 2000 Specificity and mechanism of action of some commonly used protein kinase inhibitors. Biochem J 351: 95–105 43. Hresko RC, Heimberg H, Chi MM, Mueckler M 1998 Glucosamine-induced insulin resistance in 3T3-L1 adipocytes is caused by depletion of intracellular ATP. J Biol Chem 273:20658 –20668 44. Kang J, Heart E, Sung CK 2001 Effects of cellular ATP depletion on glucose transport and insulin signaling in 3T3-L1 adipocytes. Am J Physiol Endocrinol Metab 280:E428 –E435 45. Hayashi T, Hirshman MF, Fujii N, Habinowski SA, Witters LA, Goodyear LJ 2000 Metabolic stress and altered glucose transport: activation of AMPactivated protein kinase as a unifying coupling mechanism. Diabetes 49: 527–531 46. Khayat ZA, Tsakiridis T, Ueyama A, Somwar R, Ebina Y, Klip A 1998 Rapid stimulation of glucose transport by mitochondrial uncoupling depends in part on cytosolic Ca2⫹ and cPKC. Am J Physiol 275:C1487–C1497 47. Bashan N, Burdett E, Guma A, Sargeant R, Tumiati L, Liu Z, Klip A 1993 Mechanisms of adaptation of glucose transporters to changes in the oxidative chain of muscle and fat cells. Am J Physiol 264:C430 –C440 48. Tsakiridis T, Vranic M, Klip A 1995 Phosphatidylinositol 3-kinase and the actin network are not required for the stimulation of glucose transport caused by mitochondrial uncoupling: comparison with insulin action. Biochem J 309:1–5 49. Youn JH, Gulve EA, Holloszy JO 1991 Calcium stimulates glucose transport in skeletal muscle by a pathway independent of contraction. Am J Physiol 260:C555–C561 50. Stapleton D, Mitchelhill KI, Gao G, Widmer J, Michell BJ, Teh T, House CM, Fernandez CS, Cox T, Witters LA, Kemp BE 1996 Mammalian AMP-activated protein kinase subfamily. J Biol Chem 271:611– 614

3896


51. Kurth-Kraczek EJ, Hirshman MF, Goodyear LJ, Winder WW 1999 5⬘ AMPactivated protein kinase activation causes GLUT4 translocation in skeletal muscle. Diabetes 48:1667–1671 52. Kaushik VK, Young ME, Dean DJ, Kurowski TG, Saha AK, Ruderman NB 2001 Regulation of fatty acid oxidation and glucose metabolism in rat soleus muscle: effects of AICAR. Am J Physiol Endocrinol Metab 281:E335–E340 53. Buhl ES, Jessen N, Schmitz O, Pedersen SB, Pedersen O, Holman GD, Lund S 2001 Chronic treatment with 5-aminoimidazole-4-carboxamide-1-␤-dribofuranoside increases insulin-stimulated glucose uptake and GLUT4 translocation in rat skeletal muscles in a fiber type-specific manner. Diabetes 50:12–17 54. Derave W, Ai H, Ihlemann J, Witters LA, Kristiansen S, Richter EA, Ploug


T 2000 Dissociation of AMP-activated protein kinase activation and glucose transport in contracting slow-twitch muscle. Diabetes 49:1281–1287 55. Hayashi T, Hirshman MF, Kurth EJ, Winder WW, Goodyear LJ 1998 Evidence for 5⬘ AMP-activated protein kinase mediation of the effect of muscle contraction on glucose transport. Diabetes 47:1369 –1373 56. Musi N, Hayashi T, Fujii N, Hirshman MF, Witters LA, Goodyear LJ 2001 AMP-activated protein kinase activity and glucose uptake in rat skeletal muscle. Am J Physiol Endocrinol Metab 280:E677–E684 57. Salt IP, Connell JM, Gould GW 2000 5-Aminoimidazole-4-carboxamide ribonucleoside (AICAR) inhibits insulin-stimulated glucose transport in 3T3–L1 adipocytes. Diabetes 49:1649 –1656

ATTENTION AUTHORS: All new manuscripts submitted to Endocrinology after November 1, 2002 should be sent to the new editorial office: Jeffrey E. Pessin, Editor-in-Chief Endocrinology 4350 East West Highway, Suite 500 Bethesda, MD 20814-4426