HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 20, 18151-18160, May 17, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/20/18151    most recent M104252200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Huang, D. Articles by Bryer-Ash, M. Search for Related Content PubMed PubMed Citation Articles by Huang, D. Articles by Bryer-Ash, M. Social Bookmarking                      What's this?

Focal Adhesion Kinase (FAK) Regulates Insulin-stimulated Glycogen Synthesis in Hepatocytes^*

Danshan Huang, Anthony T. Cheung^§, J. Thomas Parsons^¶, and Michael Bryer-Ash^§

From  UCLA Gonda (Goldschmied) Diabetes Center and the Research Service, West Los Angeles Veterans Administration Medical Center, Los Angeles, California 90095, ^§ University of Tennessee Health Science Center and the Research Service, Veterans Administration Medical Center, Memphis, Tennessee 38103, and the ^¶ University of Virginia, Charlottesville, Virginia 22908

Received for publication, May 10, 2001, and in revised form, January 16, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Experimental data support a role for FAK, an important component of the integrin signaling pathway, in insulin action. To test the hypothesis that FAK plays a regulatory role in hepatic insulin action, we overexpressed wild type (WT), a kinase inactive (KR), or a COOH-terminal focal adhesion targeting (FAT) sequence-truncated mutant of FAK in HepG2 hepatoma cells. In control untransfected (NON) and vector (CMV2)- and WT-transfected cells, insulin stimulated an expected 54 ± 13, 37 ± 4, and 47 ± 12 increase in [U-¹⁴C]glucose incorporation into glycogen, respectively. This was entirely abolished in the presence of either KR (1 ± 7%) or FAT mutants (0 ± 8%, n = 5, p < 0.05 for KR or FAT versus other groups), and this was associated with a significant attenuation of incremental insulin-stimulated glycogen synthase (GS) activity. Insulin-stimulated serine phosphorylation of Akt/protein kinase B was significantly impaired in mutant-transfected cells. Moreover, the ability of insulin to inactivate GS kinase-3 (GSK-3), the regulatory enzyme immediately upstream of GS, by serine phosphorylation (308 ± 16, 321 ± 41, and 458 ± 34 optical densitometric units (odu) in NON, CMV2, and WT, respectively, p < 0.02 for WT versus CMV2) was attenuated in the presence of either FAT (205 ± 14, p < 0.01) or KR (189 ± 4, p < 0.005) mutants. FAK co-immunoprecipitated with GSK-3, but only in cells overexpressing the KR (374 ± 254 odu) and FAT (555 ± 308) mutants was this association stimulated by insulin compared with NON (209 ± 92), CMV2 (47 ± 70), and WT (39 ± 31 odu). This suggests that FAK and GSK-3 form both a constitutive association and a transient complex upon insulin stimulation, the dissociation of which requires normal function and localization of FAK. We conclude that FAK regulates the activity of Akt/protein kinase B and GSK-3 and the association of GSK-3 with FAK to influence insulin-stimulated glycogen synthesis in hepatocytes. Insulin action may be subject to regulation by the integrin signaling pathway, ensuring that these growth and differentiation-promoting pathways act in a coordinated and/or complementary manner.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Integrins constitute a cell-surface transmembrane receptor family that mediates growth and differentiation-promoting functions upon activation by engagement with basement membrane or other extracellular ligands (1). Clustering of a number of intracellular signaling proteins at the site of focal contact (focal adhesion complex formation) is a central feature of integrin activation (2, 3). FAK¹ is a 125-kDa cytosolic protein-tyrosine kinase named for its predominant cellular localization within the focal adhesion complex, which is tyrosine-phosphorylated and activated upon integrin engagement (4) and which interacts with PI3K, Shc, and Grb2 to promote integrin signaling (5-7).

It has also been shown that the tyrosine phosphorylation state of FAK is regulated in a complex manner by growth factor stimulation and that FAK associates with both the IR and IRS proteins upon insulin stimulation (8-11). In the case of platelet-derived growth factor and insulin-like growth factor-I, FAK undergoes tyrosine phosphorylation at low ligand concentrations but is dephosphorylated at higher concentrations (7). In addition, the regulation of the tyrosine phosphorylation state of FAK upon insulin stimulation is a function of the level of IR expression (4, 12), cell type (13), and the adhesion state of the cell (9, 14). It has also previously been reported that integrin engagement stimulates both IRS-1-associated PI3K activity and Akt/protein kinase B (Akt/PKB) activity (15), which are important steps in insulin action to stimulate glucose transport and glycogen synthesis. Integrin engagement appears to play a stimulatory role in insulin signaling (8) and specifically in insulin-stimulated mitogenesis (16), suggesting that the state of cell contact is an important regulator of insulin action.

In light of the above data, it is important to establish the role of FAK in the regulation of insulin action. Hitherto, the tyrosine phosphorylation state of FAK in response to insulin administration in vivo was unknown. However, we recently showed that in the liver of healthy Sprague-Dawley rats, FAK rapidly undergoes tyrosine phosphorylation upon insulin stimulation under euglycemic conditions in vivo (14). Moreover, we observed that when insulin resistance was ameliorated in obese Zucker rats by administration of an adenoviral construct containing cDNA encoding for a soluble inhibitor of tumor necrosis factor-, this was associated with a consistent increase (>4-fold) in the tyrosine phosphorylation state of a 125-kDa protein, the identity of which was subsequently confirmed to be FAK (14). Using HepG2 cells, a human hepatoma cell line, we further showed that pretreatment with tumor necrosis factor- abolished insulin-stimulated tyrosine phosphorylation of FAK without altering IR abundance or IR tyrosine phosphorylation.

In light of the above data, we undertook the present study to test the hypothesis that FAK plays an important role in the regulation of hepatic insulin signaling and insulin action. In the present work, we employ mutant FAK constructs overexpressed in HepG2 hepatoma cells and show that both the focal adhesion-targeting property and tyrosine kinase activity of FAK are essential for insulin-mediated stimulation of glycogen synthesis and that FAK acts to promote insulin action downstream of PI3K by a specific interaction with GSK-3 and Akt/PKB.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials and Supplies-- Antibodies to FAK and GSK-3 were obtained from Transduction Laboratories (Lexington, KY). Anti-IR -subunit, anti-IRS-1, and anti-p85 polyclonal antibodies were obtained from Upstate Biotechnology Inc. (Lake Placid, NY). Antibody to Akt and anti-phospho-specific Akt polyclonal antibody, which recognizes serine 473-phosphorylated Akt as well as anti-phospho-specific GSK-3 polyclonal antibody, which recognizes serine 9-phosphorylated GSK-3, were from Cell Signaling Technology (Beverly, MA). Antibody against phosphotyrosine (pY99) was purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA). Chemiluminescence kit, protein A/G-conjugated agarose beads, horseradish peroxidase-conjugated anti-mouse and anti-rabbit secondary antibodies were all purchased from Pierce. All cell culture reagents were obtained from Invitrogen. Nitrocellulose membrane was from Schleicher & Schuell. D-[U-¹⁴C]Glucose was from Amersham Biosciences, Inc. [-³²P]ATP was obtained from PerkinElmer Life Sciences. UDP-[U-¹⁴C]glucose (303 mCi/mM) was from ICN (Irvine, CA). The pFLAG-CMV2 vector, anti-FLAG antibodies, and all other chemicals were purchased from Sigma unless otherwise stated.

Synthesis of pFLAG-CMV2 Constructs with Mutant FAK cDNAs-- Mutant FAK cDNAs were FAK wild-type (WT), K454R mutant of FAK, in which Lys-454 was replaced by Arg (KR), or the COOH-terminal deletion variant dl^686-1011 of FAK (FAT) (17). cDNA was subcloned into the EcoRI site of pBluescript/KS (pBS, Stratagene). The cDNAs encoding WT and mutant forms of FAK were then excised from the pBS plasmid by SmaI and SalI digestion and directionally subcloned into EcoRV and SalI cut site of NH₂-terminal FLAG-tagged pFLAG-CMV2 expression vector, since the COOH-terminal of FAK is required for location of FAK to the focal adhesion complex (17). Insert orientations were confirmed by direct sequencing and restriction fragment analysis (data not shown).

Cell Culture, Transfection, and Insulin Stimulation-- HepG2 is a cell line derived from a human hepatocyte carcinoma (Invitrogen). Cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% (v/v) fetal bovine serum, 10 mM glutamine, and 15 mM Hepes, pH 7.4. Cells were grown to 70-80% confluence in 100-mm dishes (Costar, Cambridge, MA) and were then transiently transfected using Tfx-20 reagent (Promega, Madison, WI) with pFLAG-CMV-expressing various FAK constructs as described above. Control cells were transfected with the pFLAG-CMV2 empty vector only. Transfection efficiency exceeded 50% (data not shown). Forty-eight hours post-transfection, cells were starved overnight in fetal bovine serum-free DMEM before performing experiments. For insulin stimulation, recombinant human insulin was added to a final concentration of 100 nM. After the indicated treatment times, cells were washed twice with ice-cold phosphate-buffered saline (PBS), and liquid N₂ was added. 1 ml of radioimmune precipitation lysis buffer (50 mM Tris-HCL, pH 7.4, 150 mM NaCl, 0.5% deoxycholate, and 1% Nonidet P-40) containing aprotinin (20 µg/ml), leupeptin (20 µg/ml), 1 mM sodium orthovanadate, 10 mM NaF, 10 mM sodium pyrophosphate, 1 mM phenylmethylsulfonyl fluoride) was then added, and the cells were scraped off and transferred to 1.5-ml microcentrifugation tubes. They were incubated on ice for 40 min and vortexed frequently. After centrifugation at 100,000 × g for 30 min at 4 °C, the clarified supernatants were separated into aliquots of equal protein content with radioimmune precipitation lysis buffer (Bio-Rad).

Determination of the Rate of Cell Growth and Cell Morphology-- Cells were suspended at 0.5 × 10⁵ cells/ml in complete medium, plated on 6-well plates in triplicate samples (0.5 × 10⁵ cells/well), and transfected with FAK mutants. At 24, 48, and 72 h after transfection, cells were harvested in 0.05% trypsin, EDTA and washed twice with PBS. Cells were then resuspended in 100 µl of PBS and mixed with an equal volume of 0.4% trypan blue (Sigma). 10 µl of suspended cells were then loaded onto a hemocytometer, and viable cells were counted under a light microscope. Cell growth was expressed as the number of live cells per well.

Cell morphology was quantified by determination of the cellular aspect ratio according to the method of Ridyard and Sanders (18), after capturing the cell image on Image 1.61 software (NIH). Measurements were taken on 13 different cells in each culture. The aspect ratio was calculated as the long axis of the cell divided by its short axis.

Immunoprecipitation and Western Blotting-- 500 µg of cell lysate protein was incubated with 10 µg of the respective antibody and agarose-conjugated Protein G at 4 °C overnight. The immunoprecipitates were then washed three times in a buffer containing 50 mM Hepes, 100 mM Na₄P₂O₇, 100 mM NaF, 10 mM EDTA, 1 mM Na₃VO₄, 1 mM phenylmethylsulfonyl fluoride, and 0.1 mg/ml aprotinin. The washed immunoprecipitates were boiled in 2× Laemmli buffer (19) and separated by 7.5% SDS-PAGE. Proteins were electroblotted onto a nitrocellulose membrane, and the membrane was blocked in 5% nonfat dry milk in Tris-buffered saline containing 0.1% Tween 20. Membranes were blotted with various primary antibodies followed by the appropriate horseradish peroxidase-conjugated secondary antibody. Results were visualized by Supersignal chemiluminescence kit (Pierce) according to the manufacturer's instructions. In cases where reblotting with a different primary antibody was required, the membrane was stripped by incubation in 50 mM Tris-HCL, pH 6.8, 100 mM 2-mercaptoethanol, and 2% SDS for 30 min at 50 °C with constant agitation. The membrane was then washed thoroughly in Tris-buffered saline/Tween 20 and reprobed with the appropriate antibody.

Co-immunoprecipitation Studies-- Cell lysates were prepared as described above. Samples containing 500 µg of protein were precleared for 2 h at 4 °C with protein G-Sepharose that had previously been washed 3 times with lysis buffer and were immunoprecipitated overnight at 4 °C with 1 µg of monoclonal anti-GSK-3 or anti-Akt/PKB antibody. Protein G-Sepharose (40 µl of beads in suspension) was added, and the incubation was continued for 3 h. After washing the precipitates 3 times with lysis buffer, 20 µl of Laemmli stopping buffer was added to each pellet, and the samples were placed in boiling water for 10 min before SDS-PAGE and immunoblotting. When the blots were probed with the monoclonal anti-FAK antibody, horseradish peroxidase-conjugated mouse antibody was used as secondary antibody. The blots were then stripped by incubation in 100 mM -mercaptoethanol, 2% SDS, 62.5 mM Tris-HCl, pH 6.7, at 50 °C for 30 min with agitation, followed by thorough rinsing, blocking, and reprobing with monoclonal GSK-3 or polyclonal Akt/PKB antibody.

Measurement of Glycogen Synthesis in HepG2 Cells-- The determination of glycogen synthesis was made by measurement of the incorporation of D-[U-¹⁴C]glucose into glycogen (20). Confluent transfected HepG2 cells in 6-well plates were incubated overnight in fetal bovine serum-free DMEM and then for 3 h in fetal bovine serum-free DMEM containing 5.5 mM glucose and 0.33 µCi/ml D-[U-¹⁴C]glucose in the absence or presence of 100 nM insulin. The incubation was terminated by removal of the medium and rinsing the cells five times in ice-cold PBS. Cells were solubilized with 20% KOH for 2 h, and an aliquot (100 µl) of the resulting lysate was removed for protein analysis. Lysates were extracted with 8% (w/v) tricarboxylic acid, neutralized with 2.0 M HCl, then boiled for 5 min, and 1 mg of glycogen was added as a carrier to each sample. Total glycogen was precipitated by the addition of 80% ethanol (final concentration) for 2 h at 20 °C followed by centrifugation at 1100 × g for 10 min. This step was then repeated. After the pellets had been re-dissolved in distilled water, the samples were precipitated again as described above. The amount of incorporated radioactivity was determined by liquid scintillation counting.

Measurement of GS Activity in HepG2 Cells-- GS activity was determined by a modification of previously described methods (21, 22). HepG2 cells transfected with FAK mutants were incubated in serum-free medium overnight followed by the addition of 100 nM insulin or saline for 10 min at 37 °C. After washing with PBS, cells were scraped into 500 µl of GS assay buffer (50 mM Tris-HCl, pH 7.8, 100 mM NaF, 20 mM EDTA, 0.5% glycogen, 1 mM phenylmethylsulfonyl fluoride, 1 mM benzamidine, 10 µg/ml leupeptin, 10 µg/ml aprotinin) and incubated on ice for 40 min with frequent mixing. After centrifugation at 100,000 × g for 30 min at 4 °C, the clarified supernatants were separated into aliquots with GS assay buffer to equal protein content (Bio-Rad). To measure GS activity, 50 µl of supernatant (4 µg of protein/µl) were added to an equal volume of GS buffer containing 2 µCi/ml UDP-[¹⁴C]glucose and 15 mg/ml glycogen in the presence or absence of 10 mM glucose 6-phosphate. After a 15-min incubation at 37 °C, assay tubes were chilled for a further 15 min on ice. Contents were then spotted onto labeled Whatman filter papers (GF/A; 2.4 cm), which were immediately immersed in 70% ethanol at 4 °C, mixed for 40 min, then washed twice more in 70% ethanol (for 15 and 60 min) to remove unincorporated substrate from precipitated glycogen. Filters were air-dried, and radioactivity was counted with 10 ml of liquid scintillation fluid.

Measurement of PI3K Activity-- PI3K activity was measured by the technique of Backer et al. (23). After incubation with insulin, cells were lysed at 4 °C for 20 min in buffer containing 20 mM Tris-HCl, pH 7.4, 137 mM NaCl, 1 mM MgCl₂, 1 mM CaCl₂, 10% (v/v) glycerol, 1% Nonidet P-40; 1 mM Na₃VO₄, and 1 mM phenylmethylsulfonyl fluoride. The resulting lysates were incubated with either anti-p85 or anti IRS-1 antibody for 2 h and then for 1 h with protein A-Sepharose (4 µg/ml of lysate). Immunoprecipitates were washed twice with buffer A, 1% Nonidet P-40, twice with 0.5 M LiCl/100 mM Tris-HCl, pH 7.5, and twice with 10 mM Tris-HCL, pH 7.4, 100 mM NaCl, 1 mM EDTA, and pellets were directly incubated with phosphatidylinositol (0.1 mg/ml) in medium containing 880 µM ATP (30 µCi of [-³²P]ATP), 20 mM MgCl₂ at room temperature for 10 min with constant shaking in a final volume of 80 µl. After the addition of 20 µl of 8 M HCl, lipids were extracted with 160 µl CHCl₃/CH₃OH (1:1, by volume). The samples were centrifuged, and the lower organic phase was removed and applied to a silica gel plate (Merck) and coated with 1% potassium oxalate. TLC plates were developed in CHCl₃/CH₃OH/H₂O/NH₄OH (60:47:11.3:2), dried, and visualized by autoradiography.

Statistical Analysis-- All statistical comparisons were conducted using Student's 2-tailed t test for paired or unpaired samples as appropriate, with appropriate correction for multiple comparisons (SigmaStat statistical software version 2.0, SPSS Inc, Chicago, IL). Data are reported as the means ± S.E.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Overexpression of WT and Mutant FAK Constructs in HepG2 Cells-- Recent studies have suggested several roles for integrin signaling via FAK in the regulation of cell survival (24, 25), proliferation (26, 27), spreading (28), and migration (26, 29, 30). Overexpression of FAK stimulated cell migration, and mutation of Tyr-397 to Phe abolished this property (31). To investigate the effect of various constructs of FAK on cell proliferation and morphology, we first measured the growth rate of cells transfected with various constructs of FAK: FAK WT, K454R mutant of FAK, in which Lys-454 is replaced by Arg (KR), or the COOH-terminal deletion variant dl^686-1011 of FAK (FAT). As shown in Fig. 1, the cell growth curves were similar among the 5 groups (Fig. 1C), and 48 h after transfection all 3 of constructs were overexpressed by ~4-5-fold. Analysis of cell shape showed that there were no significant differences in the morphology of the transfected and non-transfected cells. The mean values for the aspect ratio were 1.23 ± 0.04, 1.27 ± 0.06, and 1.20 ± 0.03 in WT and FAT- and KR mutant-transfected cells, respectively. The mean value in non-transfected cells was 1.24 ± 0.05. In HepG2 cells transfected with the FAT mutant, the expression level of endogenous FAK was similar to that of non-transfected cells (Fig. 1A, left), suggesting that transfection did not alter the cellular phenotype. Fig. 1A, right, confirms the specificity of the anti-FLAG antibody for detecting exogenous FAK tagged with FLAG, since endogenous FAK is not visible.

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Expression of FAK and epitope-tagged FAK variants in HepG2 cells. A, cells were suspended at 1.0 × 105 cells/ml in complete medium, then transfected with FAK mutants. At the times shown, cells were resuspended and counted as described under "Experimental Procedures." Whole cell lysates were prepared, and 30 µg of total cell protein was resolved by SDS-PAGE and Western blotting with antibody to FLAG epitope to show exogenous FAK (A, right) or antibody to FAK and to show both endogenous and exogenous FAK (A, left). B, total FAK abundance was quantitated by densitometry. NON, untransfected cells; WT, cells transfected with wild-type FAK; FAT, cells transfected with COOH-terminal focal adhesion targeting region deficient mutant FAK; KR, cells transfected with kinase-incompetent mutant FAK; CMV2, cells transfected with empty vector. Cell growth was expressed as the number of live cells per dish ± S.E. for four separate experiments (C).

Insulin Induces a Rapid and Transient Tyrosine Dephosphorylation of FAK followed by Phosphorylation in Adherent HepG2 Cells-- To determine the effects of insulin stimulation on FAK tyrosine phosphorylation, HepG2 cells, either untransfected or transfected with WT FAK, were treated with 100 nM insulin for 0-60 min. Extracts were then immunoprecipitated with anti-FAK antibody and immunoblotted with anti-phosphotyrosine antibody then stripped and reblotted with anti-FAK antibody to quantify the amount of FAK immunoprecipitated. After 5 min of insulin treatment, a significant decrease in FAK phosphotyrosine immunoreactivity was observed in both groups versus the non-insulin-stimulated state (Fig. 2A). The decrease in FAK phosphorylation in both groups was transient, because anti-phosphotyrosine immunoreactivity gradually increased with longer exposure to insulin, indicating that FAK was either subsequently re-phosphorylated or phosphorylated on tyrosine residues that are not phosphorylated in the non-insulin-stimulated state. In the basal state, as reported by Hildebrand et al. (17), the kinase-defective variant KR FAK showed significantly reduced tyrosine phosphorylation. After exposure to insulin for 10 min, tyrosine phosphorylation of KR FAK was unchanged compared with basal levels (Fig. 2B). This is an interesting observation indicating that the phosphorylation state of FAK is dependent upon its kinase activity. It has been reported that this protein may function in dominant negative fashion by occupying all of the potential binding sites for enzymatically active FAK (17). FAT-FAK showed an intermediate level of tyrosine phosphorylation, both in the presence and absence of insulin stimulation. The tyrosine phosphorylation state of FAK in the FAT and WT FAK groups was not significantly altered by insulin treatment (Fig. 2B).

View larger version (69K): [in this window] [in a new window]   Fig. 2.   Insulin-induced tyrosine phosphorylation of exogenous FAK. HepG2 cells that were not transfected or transfected with FAK WT were exposed to 100 nM insulin for different periods. Exogenous WT FAK was immunoprecipitated (IP) with a monoclonal anti-FLAG antibody (A, top right), and untransfected endogenous FAK was immunoprecipitated with a monoclonal anti-FAK antibody (A, top left) and both were immunoblotted (IB) with PY99 monoclonal antiphosphotyrosine antibody. The membrane was then stripped and reblotted with monoclonal anti-FAK antibody (A, lower panel). HepG2 cells that were not transfected or transfected with FAK constructs were exposed to 0 nM insulin () or 100 nM insulin (+) for 10 min. Exogenous FAK was immunoprecipitated with a monoclonal anti-FLAG antibody, and the precipitates were immunoblotted with PY99 antibody (B, upper panel). The membrane was then stripped and reblotted with monoclonal anti-FAK antibody (B, lower panel). Representative immunoblots are shown.

Insulin Fails to Stimulate Glycogen Synthesis in HepG2 Cells Transfected with FAT and KR FAK Mutants-- As shown in Fig. 3, overexpression of FAT and KR FAK effectively abolished incremental insulin-mediated glucose incorporation into glycogen in HepG2 cells, indicating that correct cellular localization and intact kinase activity of FAK are required for insulin-stimulated glycogen synthesis. The mutants appear to act in a dominant negative manner, since mutant expression did not markedly diminish expression of endogenous FAK (Fig. 1). However, overexpression of WT FAK did not increase basal or incremental insulin-mediated glucose incorporation into glycogen compared with that seen in untransfected or vector-transfected cells, suggesting that parental endogenous expression levels of FAK were sufficient for maximum insulin action. The basal mean values were similar in all groups at 128 ± 22, 127 ± 21, 147 ± 25, 140 ± 26, and 142 ± 24 pmol/1 h/mg of protein in NON, CMV2, WT, FAT, and KR, respectively.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Insulin-stimulated glucose incorporation into glycogen in HepG2 cells that were not transfected or transfected with FAK constructs. HepG2 cells were incubated for 16 h in DMEM containing 5 mM glucose and then for 3 h in fresh medium containing 15 mM [U-14C] glucose with or without 100 nM insulin. Values are the means ± S.E. for five different experiments and represent percent change in glycogen synthesis versus non-insulin stimulated cells. *, p < 0.05 relative to NON, CMV2, or WT. Labels are as for Fig. 1.

Insulin-stimulated GS Activity Is Impaired in HepG2 Cells Transfected with FAT and KR FAK Mutants-- To determine the mechanism of the inhibition of insulin-stimulated glycogen synthesis in FAT and KR mutant cells, GS activity was assayed. An in vitro assay in which UDP-[¹⁴C]glucose incorporation into glycogen was determined in the presence of GS precipitated from HepG2 cells overexpressing the WT and mutant forms of FAK was employed. The data are shown in Fig. 4 and demonstrate that incremental insulin-stimulated GS activity was significantly impaired in the presence of the two mutant FAK constructs. In NON, CMV2, and WT preparations, insulin stimulation led to a 2.3-, 2.1-, and 2.1-fold increase in GS activity, respectively. In contrast, in the preparations overexpressing FAT and KR, GS activity increased by 1.7- and 1.3-fold, respectively (p < 0.05 versus NON, CMV2, and WT). There was a significant difference between basal and insulin-stimulated GS activity in FAT (p < 0.05) but not in KR cells. Basal non-insulin-stimulated GS activity was similar in all preparations (not shown). Interestingly, the increment in insulin-stimulated GS activity in WT cells was similar to that seen in control preparations, which paralleled the results for glycogen synthesis seen in Fig. 3.

View larger version (33K): [in this window] [in a new window]   Fig. 4.   GS activity in HepG2 cells transfected with FAK constructs. HepG2 cells were incubated for 10 min with or without 100 nM insulin. GS activity was assayed in the supernatants of cell lysates as described under "Experimental Procedures." GS activity is expressed as a percent change from the basal non-insulin-stimulated value. Values shown are means ± S.E. for three different experiments. *, p < 0.05 relative to NON, CMV2, or WT, Labels are as for Fig. 1.

Overexpressing FAT and KR FAK Mutants Impairs Insulin-mediated GSK-3 Ser-9 Phosphorylation in HepG2 Cells-- Activation of GS is a rate-limiting step in glycogen synthesis, and GS is regulated by both allosteric and phosphorylation-dephosphorylation mechanisms (32). There is evidence that inhibition of GSK-3 activity by insulin-mediated serine phosphorylation is a key regulatory mechanism of activation of GS upon insulin stimulation (33). Insulin treatment results in phosphorylation on Ser-9 of GSK-3, leading to enzyme inactivation, thus releasing GS from its inhibitory influence (34-36). To determine whether the observed effect of FAT or KR FAK on glycogen synthesis was accompanied by changes in the activation state of GSK-3, the effect of insulin on Ser-9 phosphorylation of GSK-3 was examined. Non-transfected HepG2 cells and cells transfected with FAK constructs were exposed to insulin for 10 min. GSK-3 was then immunoprecipitated, and Ser-9-phosphorylation was analyzed using an anti-GSK-3 Ser-9 antibody. As shown in Fig. 5A, insulin stimulation led to a marked increase in GSK-3 Ser-9 phosphorylation in HepG2 cells transfected with WT FAK (458 ± 34 optical densitometry units (odu)) compared with non-transfected cells (308 ± 16) or those transfected with vector alone (321 ± 41). By contrast, insulin-stimulated phosphorylation of GSK- 3 Ser-9 was significantly attenuated in HepG2 cells transfected with FAT (205 ± 14 odu) or KR (189 ± 4) FAK. These data are shown quantitatively in Fig. 5B and demonstrate that the effect of FAT or KR FAK on glycogen synthesis is accompanied by a decreased capacity for insulin to serine phosphorylate and inactivate GSK-3. Interestingly, although overexpression of WT FAK potentiated insulin-mediated serine phosphorylation of GSK-3, insulin-induced glycogen synthesis in these cells was unchanged (Fig. 3).

View larger version (28K): [in this window] [in a new window]   Fig. 5.   Insulin-induced Ser-9 phosphorylation of GSK-3. HepG2 cells that were not transfected or transfected with FAK constructs were exposed to 0 nM insulin () or 100 nM insulin (+) for 10 min. GSK-3 was immunoprecipitated (IP) with a monoclonal anti-GSK-3 antibody, and the Ser-9 phosphorylation of GSK-3 was analyzed with anti-phospho-Ser-9-GSK-3 antibody. A, a representative immunoblot is shown. B, levels of GSK-3 Ser-9 phosphorylation were quantitated by densitometry. Results are expressed as means ± S.E. of three different experiments. *, p < 0.05 versus NON or CMV2.

Association of FAK with GSK-3 in Response to Insulin Is Increased in HepG2 Cells Overexpressing FAT and KR Mutants-- The association of FAK with GSK-3 was then examined in HepG2 cells transfected or untransfected with FAK constructs. GSK-3 was immunoprecipitated from cell lysates with or without prior insulin stimulation, and immunoprecipitates were analyzed for the presence of FAK. As shown in Fig. 6, in HepG2 cells transfected with FAT or KR FAK, treatment with insulin for 10 min increased the amount of FAK that co-precipitated with GSK-3. In the case of FAT mutant, association of endogenous FAK, rather than the truncated mutant itself, with GSK-3 appears to be stimulated by insulin treatment. In contrast, insulin treatment was associated with no significant increase in association between FAK and GSK-3 in HepG2 cells transfected with WT FAK, empty vector, or untransfected cells (Fig. 6A). These data are shown quantitatively in Fig. 6B and demonstrate that the effect of FAT or KR FAK on insulin-mediated GSK-3 Ser-9 phosphorylation is also accompanied by increased insulin-stimulated association between FAK and GSK-3, These findings suggest that a transient association between FAK and GSK-3 is necessary for insulin action to promote glycogen synthesis and that failure to terminate this association appropriately may lead to impaired insulin action, at least on glycogen synthesis.

View larger version (43K): [in this window] [in a new window]   Fig. 6.   Association between GSK-3 and FAK in response to insulin treatment in cells transfected or untransfected with FAK constructs. Lysates were prepared from cells incubated without () or with (+) 100 nM insulin for 10 min. GSK-3 was immunoprecipitated (IP) from each sample, and the precipitates were immunoblotted (IB) with anti-FAK antibody (A, upper panel) then stripped and reblotted with anti-GSK-3 antibody (A, lower panel). B, the association of GSK-3 with FAK was measured by densitometry. Results are expressed as means ± S.E. for three different experiments. *, p < 0.05 versus NON or CMV2.

Insulin-mediated Akt/PKB-Ser-473 Phosphorylation Is Impaired in HepG2 Cells Overexpressing FAT and KR Mutants-- Because Akt/PKB is known to be the upstream regulator of GSK-3 in insulin signaling to glycogen synthesis (37), we analyzed the phosphorylation state of one key residue in Akt/PKB, namely Ser-473, using a phosphospecific antibody. Phosphorylation of Ser-473 of Akt/PKB occurs as a result of insulin stimulation (38). As shown in Fig. 7A, there was a marked increase in phosphorylation of Ser-473 in response to exposure to insulin in HepG2 cells transfected with either WT FAK or vector alone as well as in untransfected cells. However, similar to the effect on Ser-9 phosphorylation of GSK-3 presented above, the insulin-stimulated increase in Akt/PKB Ser-473 phosphorylation was significantly impaired in HepG2 cells overexpressing FAT or KR FAK mutants compared with that seen with WT FAK, vector alone, or in untransfected cells (Fig. 7A). These data are shown quantitatively in Fig. 7B. There was no immunoprecipitable association between FAK and Akt/PKB either in the basal or insulin-stimulated state (data not shown).

View larger version (24K): [in this window] [in a new window]   Fig. 7.   Insulin-induced Ser-473 phosphorylation of Akt/PKB. HepG2 cells that either were or were not transfected with FAK constructs were exposed to 0 nM insulin () or 100 nM insulin (+) for 10 min. Panel A, upper band, Akt/PKB was immunoprecipitated (IP) with a monoclonal anti-Akt/PKB antibody, and the Ser-473 phosphorylation of Akt/PKB was analyzed with an anti-pSer-473-Akt/PKB antibody. IB, immunoblot. Panel A, lower band, the membrane was then stripped and reblotted with anti-GSK-3 antibody. Panel B, levels of Akt/PKB Ser-473 phosphorylation were quantitated by densitometry. Results are expressed as means ± S.E. for three different experiments. *, p < 0.05 versus non-transfected, CMV2, or WT-transfected cells.

IR- and IRS-1 Phosphorylation and PI3K Activity Are Unaffected in Cells Overexpressing FAT or KR Mutants-- Akt/PKB links insulin signaling to metabolic function through insulin-mediated activation of the IR and engagement of IRS proteins with PI3K (39). Because PI3K activation is central to realizing insulin metabolic effects (40), we also studied the regulation of IR-/IRS-1 tyrosine phosphorylation or protein abundance and both total and IRS-1-associated PI3K activity by FAK mutant constructs in HepG2 cells in response to insulin. As shown in Figs. 8 and 9, the level of IR- or IRS-1 tyrosine phosphorylation and protein abundance was similar among five experimental conditions after insulin stimulation. In addition, as shown in Fig. 10, after insulin stimulation the expected increase in docking of the p85 subunit of PI3K with IRS-1 occurred; however, this was similar among the five groups. Total PI3K activity in cell lysates was also similar both with and without insulin stimulation in the five experimental groups (data not shown). Moreover, Fig. 11 shows that insulin treatment resulted in the expected stimulation of specific IRS-1-associated PI3K activity, but that this was unaffected by the expression of WT FAK or either of the mutant FAK species. These data suggest that the effect of FAK on insulin-stimulated glycogen synthesis is localized downstream of IR, IRS-1, and PI3K.

View larger version (22K): [in this window] [in a new window]   Fig. 8.   The effect of FAK mutants on insulin-induced tyrosine phosphorylation of IR- subunit. HepG2 cells that were not transfected or transfected with FAK constructs were exposed to 0 nM insulin () or 100 nM insulin (+) for 10 min. The IR- subunit was immunoprecipitated (IP) with a polyclonal anti-IR- antibody, and the precipitates were immunoblotted with the monoclonal anti-phosphotyrosine antibody (panel A). The membrane was then stripped and reblotted with polyclonal anti-IR- antibody (panel B). Representative immunoblots (IB) are shown. The levels of IR- tyrosine phosphorylation were similar among the five groups after insulin stimulation.

View larger version (45K): [in this window] [in a new window]   Fig. 9.   Effect of transfection of FAK mutants on insulin-induced tyrosine phosphorylation of IRS-1. HepG2 cells that were not transfected or transfected with FAK constructs were exposed to 0 nM insulin (-) or 100 nM insulin (+) for 10 min. IRS-1 was immunoprecipitated with a polyclonal anti-IRS-1 antibody, and the precipitates were immunoblotted with the monoclonal anti-phosphotyrosine antibody (panel A) then stripped and reblotted with polyclonal anti-IRS-1 antibody (panel B). Representative immunoblots (IB) are shown. The levels of IRS-1 tyrosine phosphorylation were similar among the five groups after insulin stimulation.

View larger version (27K): [in this window] [in a new window]   Fig. 10.   Effect of FAK mutants on association of IRS-1 with p85 of PI3K in HepG2 cells. After overnight serum starvation, cells were stimulated with 0 nM insulin () or 100 nM insulin (+) for 5 min. IRS-1 was immunoprecipitated (IP) with a polyclonal anti-IRS-1 antibody, and the precipitates were immunoblotted (IB) with an antibody specific for the p85 subunit of PI3K (upper panel) then stripped and reblotted with polyclonal anti-IRS-1 antibody (lower panel). Representative immunoblots are shown. There was no difference in the amount of IRS-1-associated with p85 of PI3K among the five groups.

View larger version (50K): [in this window] [in a new window]   Fig. 11.   Effect of FAK mutants on IRS-1-associated PI3K activity. After overnight serum starvation, cells were stimulated with 0 nM insulin () or 100 nM insulin (+) for 5 min. Lysates were prepared, and PI3K activity was measured in IRS-1 immunoprecipitates. Activity was quantitated by densitometry. Panel A, a representative autoradiogram of thin layer chromatography together with quantitation of the results of three different experiments is shown. There was no significant difference in the values among the five groups either with or without insulin stimulation.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Integrins are transmembrane receptors involved in interactions between cells and extracellular matrix protein that mediate both outside-in and inside-out signaling (41). FAK, a cytosolic tyrosine kinase, is a key component of the integrin signaling pathway (42, 43). After integrin engagement by cell contact, FAK is activated and phosphorylated on tyrosine 397, which serves as a docking site for c-Src (44). Src may regulate FAK activity by phosphorylating it on additional tyrosine residues and facilitating its interaction with other signaling proteins, including the p85 subunit of PI3K (6). Tyrosine phosphorylation of FAK requires that cells are adherent to the extracellular matrix, and it has been reported that the level of tyrosine phosphorylation is reduced in suspended cells (1). When suspended cells are replated on extracellular matrix substrates, FAK becomes highly tyrosine-phosphorylated and shows enhanced kinase activity (45).

Insulin is a pleiotropic hormone that both effects and affects a broad range of biologic functions including regulation of glucose and fat metabolism, protein synthesis, cytoskeletal rearrangement, and cell growth and differentiation (46). It is now accepted that insulin-mediated glycogen synthesis utilizes a signaling cascade involving PI3K (47, 48) and the downstream effector Akt/PKB (49-51). The ability of insulin to regulate the activation state of FAK was, however, a recent observation (9, 12). Most available information in the literature focuses on the role of FAK in insulin-mediated effects on the cytoskeleton (13, 42). As a result, the precise role of FAK in the other metabolic actions of insulin remains poorly understood. Cross-talk between integrin and insulin signaling is suggested by the finding that insulin-stimulated DNA synthesis is modulated by engagement of _V3 integrin (16, 52, 53). It has been shown that insulin promotes association of IRS-1 with integrin (16) and also that tyrosine-phosphorylated IRs rapidly associate with integrin at focal adhesion contacts (53). Furthermore, cell adhesion and integrin engagement promote insulin's ability to activate PI3K (7). Integrin also modulates the insulin and insulin-like growth factor-1-signaling pathways by regulating cellular IRS-1 mRNA expression, and the pathway of FAK-mediated signaling to JNK appears to be involved in this process (11).

Insulin and insulin-like growth factor-I induce either phosphorylation or dephosphorylation of FAK, depending upon whether cells are in the attached or suspended state (8-10, 15). As also reported by others (9, 17), we find that in attached cell preparations, insulin-induced dephosphorylation of FAK occurs in the presence of endogenous FAK and both WT and FAT mutant FAK constructs. Whether or not FAK itself directly interacts with insulin signaling proteins and the site at which such an interaction occurs has not hitherto been fully addressed. The studies described here, which were carried out in an insulin-responsive cell line, demonstrate that FAK is likely to regulate insulin action on glycogen synthesis at a level downstream of IR, IRS-1, and PI3K, apparently via direct interaction with Akt/PKB and GSK-3.

Evidence in the present work supporting a role for an interaction between FAK and GSK-3 in the regulation of glycogen synthesis by insulin includes the following. First, overexpression of FAT or KR mutant FAK abolished insulin-stimulated glucose incorporation into glycogen, was accompanied by significantly reduced serine phosphorylation of both Akt/PKB and GSK-3, and significantly increased the insulin-stimulated co-immunoprecipitation of FAK with GSK-3. The regulation of the serine phosphorylation state of GSK-3 by FAK may thus be secondary to its ability to regulate Akt/PKB activation. Insulin-mediated serine phosphorylation of GSK-3 was up-regulated in WT overexpression, whereas both Akt/PKB phosphorylation and glycogen synthesis were unchanged. Possible explanations for this finding include additional regulation of GS by Akt/PKB through a mechanism other than via inhibition of GSK-3 alone or that endogenous levels of FAK permit activation of Akt/PKB and inhibition of GSK-3 sufficient for maximum insulin-stimulated glycogen synthesis in this system. It has previously been reported that Akt/PKB and GSK-3 can be activated by PI3K-independent mechanisms (54, 55).

Initially, the observed association between FAK and GSK-3 in the basal state was unexpected. However, Ding et al. (56) report that -catenin is a substrate of GSK-3 in the Wnt-signaling pathway (56). -Catenin is involved in mediation of cell adhesion and transcription and associates with focal adhesion proteins, including FAK (57, 58). However, activation of the Wnt-signaling pathway does not regulate GS activity, and does not lead to the phosphorylation of serine 9 of GSK-3 nor activation of Akt/PKB, which occur after insulin stimulation. Thus, the association between FAK and GSK-3 may involve intermediary proteins in more than one signaling pathway.

Although the basal association between FAK and GSK-3 was increased, this was not further disproportionately stimulated by exposure to insulin in WT cells. By contrast, the abolition of insulin-stimulated glycogen synthesis seen in cells overexpressing FAT and KR mutants was associated in both cases with a clear insulin-mediated increase in association between FAK and GSK-3. Interestingly, in the case of FAT, the increased association involved only endogenous FAK and not FAT itself. This suggests that either localization to the focal adhesion complex is required for the subsequent insulin-stimulated association-dissociation to occur or that FAT is acting as a dominant negative at a step outside the focal adhesion complex, thereby inhibiting the required dissociation of even normally localized FAK from GSK-3.

Because the phosphate groups on GS turn over rapidly (59), it has been proposed that inhibition of GSK-3 serves as a physiologically relevant mechanism for regulating GS activity. The regulation of GSK-3 by insulin has been shown to be mediated by Akt/PKB (60, 61). Upon insulin stimulation, the serine 473 residue of Akt/PKB is phosphorylated, and Akt/PKB is activated (62). Akt/PKB has been implicated as a mediator of metabolic responses to insulin, including GS activation (63). Subsequently, GSK-3 is phosphorylated on serine 9, which leads to a decrease in GSK-3 activity (61, 64). Although this has usually been detected as a 50-70% drop, it is apparently sufficient to relieve the inhibition of GS and allow glycogen synthesis to proceed in an insulin-responsive manner (56). Therefore, further potentiation in Akt/PKB serine phosphorylation might not have led to increased glycogen synthesis. Apparently, GSK-3 inhibition alone is not sufficient for the dephosphorylation and activation of GS (32). Our data suggest that FAK mutants may potentially inhibit insulin-stimulated phosphatase activity and that FAK plays a regulatory role in insulin-stimulated glycogen synthesis and that this is achieved through an interaction between FAK, Akt/PKB, and GSK-3. The fact that incremental insulin-stimulated GS activity was partially, but significantly, attenuated in the presence of FAT and KR mutants indicates that impaired regulation of GS activity is at least partly responsible for the reduction in insulin-stimulated glycogen synthesis seen in the presence of these mutants.

The mechanism whereby these FAK mutants inhibit the ability of insulin to induce phosphorylation of Akt/PKB remains unclear. It is possible that endogenous FAK interacts with an unidentified substrate that is necessary for the interaction between PI3K and Akt/PKB and that mutant FAK exerts a dominant negative effect on endogenous FAK. A similar dissociation between PI3K activity and activation of Akt/PKB has previously been reported (65, 66). Lin et al. (65) find that the PKC activator phorbol dibutyrate decreased the ability of insulin to phosphorylate both Akt/PKB and GSK-3 by 90 and 35%, respectively, in rat skeletal muscle, whereas PI3K activation was unaffected (65). Also, Matsumoto et al. (66) report that in Chinese hamster ovary cells and L6 myocytes, expression of kinase-defective PKC inhibited insulin-mediated phosphorylation and activation of Akt/PKB while having no inhibitory effect on phosphorylation or activity of PI3K (66). Whether PKC is involved in the downstream regulation of insulin signaling by FAK remains to be investigated.

We also found that the effect of overexpression of these FAK constructs on glycogen synthesis was unaccompanied by changes in the tyrosine phosphorylation state of either the IR- subunit or IRS-1, nor was the level of expression of either of these upstream proteins affected. Importantly, IRS-1-associated PI3K activity was also unaltered. In this respect, our findings differ from those of Lebrun et al. (10), who reported that FAK interacts directly with IRS-1 and regulates IRS-1-associated PI3K activity. The following may be offered to explain this discrepancy. First, the studies were performed in different cell systems, FAK^/ cells and DA2 cells, of fibroblast lineage, whereas HepG2 cells are insulin-responsive human hepatoma cells (9). Second, Lebrun et al. (11) studied cells that were in the suspended state for 2 h (11), whereas we utilized attached cells. Because the nature of the cross-talk between insulin/insulin-like growth factor-I receptors and FAK is dependent upon cell architecture, the interaction of the insulin/insulin-like growth factor-I-signaling system with integrin will vary accordingly (9). Finally, Goode et al. (67) already reported that GSK-3 can be phosphorylated by certain isoforms of PKC in vitro. Bogdanovic et al. (68) also reported that under conditions of insulin resistance PKC down-regulates Akt/PKB and that this is unaccompanied by a change in PI3K activity. We hypothesize that the FAK mutants used in this study may also down-regulate Akt/PKB activity. Moreover, the increased association of FAK mutants with GSK-3 upon insulin stimulation emphasizes this down-regulation. Possibly, this down-regulation is PKC-mediated, leading to reduced phosphorylation of GSK-3. However, further studies will be required to address this issue.

The effect of these FAK mutants to inhibit insulin-mediated glycogen synthesis in this adherent cell system appeared to parallel their inhibition of the insulin-mediated dephosphorylation of FAK. It is unclear whether this dephosphorylation is necessary for insulin-mediated glycogen synthesis to occur in hepatocytes. In studies using a mutant insulin receptor Y1210F overexpressed in Chinese hamster ovary cells, Van der Zon et al. (69) found that insulin-mediated dephosphorylation of FAK was almost abolished but that insulin-mediated glycogen synthesis was maintained, although somewhat diminished (69), suggesting that FAK dephosphorylation is not an essential component of insulin-mediated glycogen synthesis. Future studies with PTP inhibitors may also address this question.

We conclude that FAK regulates the activity of Akt/PKB, GSK-3 and GS to influence insulin-mediated glycogen synthesis in hepatocytes and, thus, may play an important role in regulating hepatic insulin action. Insulin action may be subject to secondary regulation by the integrin-signaling pathway, ensuring that these growth and differentiation-promoting pathways act in a coordinated and/or complimentary manner.

	ACKNOWLEDGEMENT

The advice of Dr. Joan Guinovart on the assay for glycogen synthase activity is gratefully acknowledged. This work is based in part upon support provided by the Office of Research and Development of the Department of Veterans Affairs.

	FOOTNOTES

^* This work was supported in part by a Merit Review Award from the Veterans Administration and by a research grant from the American Diabetes Association (to M. B.-A.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Supported in part by National Institutes of Health General Clinical Research Center Grant RR00211 and the UCLA Gonda (Goldschmied) Diabetes Center Endowment. To whom correspondence should be addressed: Division of Endocrinology, Diabetes, and Hypertension, School of Medicine, UCLA, 900 Veteran Ave, Warren Hall 24-130, Los Angeles, CA 90095. Tel.: 310-267-4639; Fax: 310-794-7654; E-mail: mbryerash@ mednet.ucla.edu.

Published, JBC Papers in Press, January 23, 2002, DOI 10.1074/jbc.M104252200

	ABBREVIATIONS

The abbreviations used are: FAK, focal adhesion kinase; PI3K, phosphatidylinositol 3-kinase; IR, insulin receptor; IRS-1, insulin receptor substrate-1; PKB and PKC, protein kinase B and C, respectively; GS, glycogen synthase; GSK-3, GS kinase-3; WT, wild type; DMEM, Dulbecco's modified Eagle's medium; PBS, phosphate-buffered saline; KR, kinase inactive mutant; FAT, focal adhesion targeting sequence-deleted mutant; NON, untransfected; CMV2, vector transfected.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES