Identification of p122RhoGAP (Deleted in Liver Cancer-1) Serine 322 as a Substrate for Protein Kinase B and Ribosomal S6 Kinase in Insulin-stimulated Cells*

Protein kinase B (PKB or Akt) plays an essential role in the actions of insulin, cytokines, and growth factors, although the substrates for PKB that are relevant to many of its actions require identification. In this study, we have reported the identification of p122RhoGAP, a GTPase-activating protein selective for RhoA and rodent homologue of the tumor suppressor deleted in liver cancer (DLC1) as a novel insulin-stimulated phosphoprotein in primary rat adipocytes. We have demonstrated that Ser-322 is phosphorylated upon insulin stimulation of intact cells and that this site is directly phosphorylated in vitro by PKB and ribosomal S6 kinase, members of the AGC (protein kinases A, G, and C) family of insulin-stimulated protein kinases. Furthermore, expression of constitutively active mutants of PKB or mitogen-activated protein kinase/extracellular signal-regulated kinase kinase (MEK) stimulates Ser-322 phosphorylation in intact cells, demonstrating that activation of the PKB or MEK pathway is sufficient for Ser-322 phosphorylation in vivo. Indeed, in primary adipocytes, insulin-stimulated Ser-322 phosphorylation was almost exclusively regulated by the phosphatidylinositol 3-kinase/PKB pathway, whereas in immortalized cells, insulin-stimulated phosphorylation was predominantly regulated by the MEK/extracellular signal-regulated kinase/ribosomal S6 kinase pathway, with the phosphatidylinositol 3-kinase/PKB pathway playing a minor role. These results demonstrate that p122RhoGAP Ser-322 acts as an integrator of signal transduction in a manner dependent on the cellular context.

Protein kinase B (PKB 2 or Akt) is a protein serine/threonine kinase that plays a key role in intracellular signaling and cellular homeostasis. The kinase is activated by dual phosphorylation on Thr-308 by 3-phosphoinositide-dependent kinase-1 (1) and on Ser-473 in a hydrophobic motif by a kinase that remains to be identified, although recent evidence suggests that ataxia telangiectasia-mutated, DNA-protein kinase, integrin-linked kinase and/or mTOR/Rictor may be involved (2)(3)(4)(5). Phosphorylation of these residues requires the production of PtdIns-3,4,5-P 3 in response to the activation of Class I PI 3-kinases (reviewed in Refs. 6 and 7). Insulin, numerous growth factors, cytokines, and other stimuli can activate PKB in this manner.
Insulin utilizes the PKB signaling pathway to regulate many intracellular events including the stimulation of glucose uptake (via the translocation GLUT4 to the plasma membrane), glycolysis and glycogen synthesis, and alterations in gene expression. Constitutively active PKB mutants induce GLUT4 translocation in the absence of insulin, whereas dominant negative PKB mutants and ablation of PKB using small interfering RNA decrease insulin-stimulated glucose uptake (8,9), PKB␤ knock-out mice exhibit reduced insulin-stimulated glucose uptake into muscle and adipose tissue (10), and deregulation of PKB activation by insulin has been reported to be associated with insulin resistance in type II diabetes (11). PKB also plays a central role in the regulation of many other cellular processes, such as apoptosis and anoikis, neuronal development and degeneration, and the cell cycle (see Ref. 7 for a recent review).
Despite the central role of PKB in insulin action, many of the substrates that mediate the actions of PKB remain to be identified. Several groups, including our own, have used a commercially available PAS (phospho-Akt substrate) antibody raised against the reported minimal consensus phosphorylation site found in almost all known PKB substrates (RXRXX(pS/pT)) (12,13) to purify and identify new PKB substrates. For example, this antibody has been used to identify AS160 (14), ATP-citrate lyase (15), PRAS40 (16), PIKfyve (17), Yes-associated protein (18), and WNK1 (19) as PKB substrates. AS160 is a GTPase-activating protein for Rabs 2A, 8A, 10, and 14 and has been reported to play a role in insulin-stimulated GLUT4 translocation to the membrane (20,21). PIKfyve is a PtdIns-3-P 5-kinase, which also appears to play a role in regulating the intracellular trafficking of GLUT4 (17). The role of PRAS40 remains unclear, whereas Yes-associated protein attenuates p73-induced apoptosis (18) and WNK1 negative regulates insulin-stimulated mitogenesis (19).
Using this approach in the current study, we identified by mass spectrometry an insulin-stimulated phosphoprotein of an apparent molecular weight of 120 kDa in primary adipocytes as p122RhoGAP, a member of the extensive family of GTPase-activating proteins for the Rho, Rac, and Cdc42 family. We demonstrated that insulin stimulates p122RhoGAP phosphorylation on Ser-322 in both primary adipocytes and in an insulin-responsive cultured cell line (Chinese hamster ovary (CHO.T cells) and that PKB can directly phosphorylate p122RhoGAP on Ser-322 in vitro. We also showed that p122RhoGAP can be phospho-rylated in vitro by RSK1, another member of the AGC family of protein kinases, and that this also occurs in intact cells in an insulin-stimulated manner via the MEK/ERK signaling pathway. Thus, Ser-322 phosphorylation of p122RhoGAP acts as an integrator of two distinct signal transduction pathways with the relative contribution of the PKB and MEK pathways in insulin-stimulated p122RhoGAP phosphorylation clearly depending on the cellular background involved.
Transfection of Primary Adipocytes-Primary adipocytes were transfected as described previously (15) with a few adjustments. In short, freshly isolated cells were washed in intracellular Krebs-bicarbonate-HEPES buffer, pH 7.4 (4 mM NaCl, 125 mM KCl, 1 mM EGTA, 1 mM MgCl 2 , 2.5 mM NaH 2 PO 4 , 15.5 mM NaHCO 3 , 10 mM HEPES, and 11 mM glucose) with 1% bovine serum albumin. A 0.4-cm electrode gap Gene Pulser cuvette (Bio-Rad) was used to electroporate 500 l of cells (30% cytocrit) in the presence of 15 g of plasmid DNA. Electroporation was performed by administering five pulses at 500 V and a capacitance of 50 microfarads using a Gene Pulser transfection apparatus (Bio-Rad). After electroporation, the cells were transferred to a 30-ml tube (Bibby-Sterilin, Ltd., Staffs, UK) and incubated for 30 min at 37°C before replacing the medium with 4 ml of Dulbecco's modified Eagle's medium (containing 1% bovine serum albumin, 2 mM glutamine, 200 nM phenylisopropyladenosine, 100 g of gentamycin, and 25 mM HEPES, pH 7.4). The cells were incubated for an additional 17 h at 37°C and 5% CO 2 and subsequently washed into Krebsbicarbonate-HEPES buffer without bovine serum albumin prior to the experiment.
Incubation of Transfected Rat Epididymal Adipocytes-Cells were washed into Krebs-bicarbonate-HEPES buffer, pH 7.4, without bovine serum albumin and left untreated or incubated with 100 nM wortmannin, 10 M UO126, or 200 nM rapamycin for 30 min at 37°C prior to stimulation with 87 nM insulin for the times indicated. The reaction was terminated by extracting the cells 1:1 (packed cell volume/volume) in ice cold Nonidet P-40 extraction buffer (50 mM Tris, pH 7.5, containing 1% Nonidet P-40, 1 mM EDTA, 120 mM NaCl, 50 mM NaF, 40 mM ␤-glycerophosphate, 1 mM benzamidine, 1 M microcystin, 10 mM sodium orthovanadate, and 1 g/ml each of pepstatin, leupeptin, and antipain). Cell extracts were centrifuged at 10,000 ϫ g for 10 min at 4°C, and the infranatant was taken for subsequent analysis.
Culture, Transfection, and Incubation of CHO.T Cells-CHO.T cells (CHO cells stably expressing the human insulin receptor) were cultured in Ham's F-12 medium containing 5% (v/v) fetal calf serum, 200 units/ml benzylpenicillin, 100 g/ml streptomycin, and 0.25 mg/ml G-418. CHO.T cells at 70% confluence were transfected in 60-mm dishes with 5 g of each plasmid (GFP-tagged wild-type RhoGAP, p110CAAX, myrPKB, or MEK-MANE) using a charge ratio of 3 l of FuGENE 6 reagent (Roche Applied Science) per g of plasmid DNA, according to the manufacturer's instructions. After 5 h, the cells were serum-starved for 16 h, treated with insulin (87 nM) for the time period indicated in the figure legends, and washed twice in ice-cold phosphatebuffered saline before extraction by scraping into 500 l of ice-cold Nonidet P-40 extraction buffer (50 mM Tris, pH 7.5, containing 1% Nonidet P-40, 1 mM EDTA, 120 mM NaCl, 50 mM NaF, 40 mM ␤-glycerophosphate, 1 mM benzamidine, 1 M microcystin, 10 mM sodium orthovanadate, and 1 g/ml each of pepstatin, leupeptin, and antipain). Cell lysates were centrifuged at 10,000 ϫ g for 10 min at 4°C, and the supernatant was taken for subsequent analysis.
Immunoprecipitation-The GFP-labeled p122RhoGAP was immunoprecipitated by rotating 250 l of total cell extract with 5 l of anti-GFP antibody and 10 l of protein G-Sepharose (50% w/v) at 4°C. The protein G-Sepharose beads were isolated by centrifugation and washed three times in Nonidet P-40 extraction buffer. Subsequently, Laemmli sample buffer was added, and proteins were separated by SDS-PAGE for immunoblotting.
Immunoblotting-Proteins were separated by SDS-PAGE using 3-8% gradient gels (p122RhoGAP immunoprecipitates) or 7.5% gels (total lysates) and transferred to polyvinylidene difluoride membranes (Immobilon). The membranes were blocked in 10% (w/v) bovine serum albumin dissolved in Tris-buffered saline with 0.1% Tween (TBS-T; 20 mM Tris, pH 7.6, 137 mM NaCl, 0.1% Tween) and subsequently incubated with primary and secondary antibodies, which were diluted in TBS-T containing 5% (w/v) bovine serum albumin. The blots were washed at least five times for 5 min after each antibody incubation and developed using an Enhanced ChemiLuminescence detection system (AP Biotech, Amersham Biosciences). Primary antibodies were used at a concentration of 1 g/ml. Horseradish peroxidase-conjugated secondary antibody (Amersham Biosciences) was diluted 1:10,000 for all antibodies.
Production of Recombinant GST-p122RhoGAP--p122RhoGAP was expressed as a recombinant fusion with GST in Escherichia coli. Briefly, 2 liters of cells were grown to an optical density (A 600 ) of 0.8, and protein expression was induced over 2 h by the addition of 0.5 mM isopropyl ␤-D-thiogalactopyranoside. The cells were harvested by centrifugation and resuspended in 40 ml of ice-cold lysis buffer (50 mM Tris, pH 7.5, 1% Triton X-100, 150 mM NaCl, 5 mM MgCl, 1 mM dithiothreitol, 0.2 mM phenylmethylsulfonyl fluoride), and the cells were then broken by sonication. Insoluble material was removed by centrifugation, and GST-p122RhoGAP protein was purified through binding to glutathione-Sepharose beads. Purified GST-p122RhoGAP protein was stored on beads in aliquots at Ϫ80°C.

RESULTS
Identification of p122RhoGAP as an Insulin-stimulated Phosphoprotein-To identify novel substrates for PKB, we stimulated primary adipocytes with insulin in the absence and presence of the PI 3-kinase inhibitor wortmannin. The cell extracts were fractionated by Mono Q anion exchange chromatography, and fractions were immunoblotted with the PAS antibody. As shown in Fig. 1A, insulin stimulated an increase in phosphorylation of several proteins that eluted from the Mono Q column between 300 and 370 mM sodium chloride (fractions 20 -22). Phosphorylation of the majority of these proteins was reduced in the presence of wortmannin. Coomassie staining of fraction 22 revealed the presence of numerous proteins (Fig. 1B), and therefore we subjected this fraction to further purification by size exclusion chromatography (Superose 12). A phosphoprotein reactive with the PAS antiserum and with an apparent molecular weight of ϳ120 kDa was abundant and well separated from other proteins upon Coomassie staining. 3 This protein was excised from the gel, eluted, and digested with trypsin. Analysis of the tryptic peptide masses using the ProFound data base led to the identification of this protein as p122RhoGAP (25).
Insulin Stimulates the Phosphorylation of p122RhoGAP at Ser-322 in Primary Adipocytes and CHO.T Cells-Analysis of the amino acid sequence of p122RhoGAP using the prediction program Scansite (scansite.mit.edu) suggested the presence of high stringency putative PKB phosphorylation sites at Ser-320 (VTRTRS 320 LST) and Ser-559 (RLRWHS 559 FQS). Manual inspection of the amino acid sequence surrounding Ser-320 revealed an additional potential phosphorylation site at Ser-322 (VTRTRSLS 322 T) conforming to part of the consensus sequence RXRXX(S/T) for PKB phosphorylation.
To investigate whether one or more of these sites are phosphorylated upon insulin stimulation, we transfected primary adipocytes and CHO.T cells with p122RhoGAP constructs containing substitutions of Ser-320, Ser-322, and Ser-5593 Ala. Fig. 2 demonstrates that insulin stimulation leads to the phosphorylation of p122RhoGAP in both primary adipocytes ( Fig. 2A) and CHO.T cells (Fig. 2B), as determined by immunoblotting with the PAS antiserum. Although mutation of Ser-320 and Ser-559 to Ala had no significant effect on the phosphorylation of p122RhoGAP, mutation of Ser-322 abrogated the reactivity of   p122RhoGAP with the PAS antibody in both cell types (Fig. 2). These results demonstrate that Ser-322 is the major insulin-stimulated phosphorylation site recognized by the PAS antibody on p122RhoGAP in both primary adipocytes and CHO.T cells.
PKB Phosphorylates p122RhoGAP in Vitro on Ser-322-To determine whether p122RhoGAP is a direct substrate for PKB in vitro, we incubated recombinant GST-p122RhoGAP in the presence and absence of recombinant active PKB and ATP. Phosphorylation of wildtype p122RhoGAP was substantially increased in the presence of PKB (as shown in Fig. 3, left panel, lane 2). PKB also phosphorylated several lower molecular mass breakdown products of p122RhoGAP in the range of 50 -122 kDa. The ability of PKB to phosphorylate p122RhoGAP on the site recognized by the PAS antibody was completely lost in the Ser-3223 Ala mutant, further demonstrating that Ser-322 is the phosphorylation site recognized by this antibody (Fig. 3).
p122RhoGAP Is Phosphorylated in a PI 3-Kinase-dependent Manner in Primary Adipocytes-We next investigated whether insulin-stimulated p122RhoGAP Ser-322 phosphorylation was regulated downstream of PI 3-kinase and PKB in intact cells. To do this, we transfected primary adipocytes with wild-type p122RhoGAP and subsequently incubated them with or without the PI 3-kinase inhibitor wortmannin. We also examined the effect of UO126, which inhibits the activation of ERK, and rapamycin, which inhibits the activation of mTOR and S6K, the latter a ribosomal protein S6 kinase. Fig. 4A shows that insulin-stimulated p122RhoGAP phosphorylation was completely inhibited in the presence of wortmannin, whereas UO126 and rapamycin had only very weak inhibitory effects. In Fig. 4, B-D, we confirmed the efficacy and selectivity of the inhibitors for insulin-stimulated phosphorylation of PKB (Fig. 4B), ERK (Fig. 4C), and S6K (Fig. 4D). These results demonstrate that the major insulin-stimulated pathway leading to p122RhoGAP Ser-322 phosphorylation in primary adipocytes is PI 3-kinase-dependent.
Expression of Constitutively Active PI 3-Kinase and PKB Mutants Stimulate p122RhoGAP Phosphorylation in Primary Adipocytes Independently of Insulin-To investigate whether activation of PI 3-kinase and/or PKB is sufficient for phosphorylation of Ser-322 on p122RhoGAP, we co-transfected primary adipocytes with membranetargeted constitutively active mutants of either PI 3-kinase (p110CAAX) or PKB (myrPKB) in conjunction with the p122RhoGAP expression vector. Fig. 5 demonstrates that both p110.CAAX and myr-PKB induced an insulin-independent increase in the phosphorylation of p122RhoGAP on Ser-322. These results demonstrate that activation of PI 3-kinase or PKB is sufficient to stimulate Ser-322 phosphorylation of p122RhoGAP in primary adipocytes.
Insulin-stimulated Phosphorylation of p122-RhoGAP in CHO.T Cells Is Dependent on the MEK Signaling Pathway-In contrast to the results in primary adipocytes, we found that insulin-stimulated p122RhoGAP Ser-322 phosphorylation in CHO.T cells was only marginally inhibited by pretreatment with wortmannin (Fig. 6A) under conditions where insulin-stimulated PKB phosphorylation was completely blocked (Fig.  6B). In these cells, the MEK inhibitor UO126 reduced p122RhoGAP Ser-322 phosphorylation to 28 Ϯ 8% (n ϭ 5, average Ϯ S.E.) of control, suggesting that a kinase downstream of ERK is primarily responsible for the phosphorylation of this site in this cellular background. The residual insulin-stimulated p122RhoGAP phosphorylation observed in the presence of a maximally effective UO126 concentration was blocked by the further addition of wortmannin. This suggests that the MEK/ERK pathway plays a dominant role over the PI 3-kinase/PKB pathway, with the contribution of the latter pathway only becoming apparent upon complete inhibition of MEK (Fig. 6A). The mTOR inhibitor rapamycin had no significant effect on p122RhoGAP Ser-322 phosphorylation under any condition (Fig. 6A). Fig. 6, B-D, again confirms the efficacy of these inhibitors under the conditions used. These results demonstrate that insulin-stimulated p122RhoGAP Ser-322 phosphorylation in CHO.T cells is dependent on both PI 3-kinase and MEK but is independent of mTOR.
p122RhoGAP Ser-322 Phosphorylation Can Be Stimulated by Either Constitutively Active PKB or Constitutively Active MEK in CHO.T Cells-As the MEK pathway appears to play a dominant and necessary role in insulin-stimulated Ser-322 phosphorylation, we next asked whether MEK activation was sufficient to induce Ser-322 phosphorylation in CHO.T cells using a constitutively active MEK mutant (MEK-MANE). Fig. 7 demonstrates that this was indeed the case. Co-expression of MEK-MANE with p122RhoGAP resulted in an increase in the phosphorylation of p122RhoGAP to an extent that was almost indistinguishable from that induced by either insulin or, interestingly, the constitutively active myrPKB. Thus, although the PI 3-kinase/PKB pathway plays a minor role in insulin-stimulated Ser-322 phosphorylation, this pathway is competent at inducing the phosphorylation of this site. Taken together, the results demonstrate that activation of either PKB  FEBRUARY 24, 2006 • VOLUME 281 • NUMBER 8 or MEK is sufficient to stimulate Ser-322 phosphorylation of p122RhoGAP but that the MEK pathway predominates in insulin-stimulated Ser-322 phosphorylation in intact CHO.T cells.

Phosphorylation of p122RhoGAP (DLC-1) by PKB and RSK
RSK1 Phosphorylates p122RhoGAP on Ser-322-Upon activation in cells, MEK phosphorylates and activates the downstream extracellular signal-regulated kinases 1 and 2 (ERK1/2). However, it is unlikely that Ser-322 is directly phosphorylated by ERK1/2 kinases, as this site does not lie within the consensus sequence phosphorylation by these kinases (Pro-Xaa-(Ser/Thr)-Pro). A good candidate for phosphorylation of Ser-322 on p122RhoGAP downstream of ERK1/2 is ribosomal S6 kinase-1 (RSK1), which, similar to PKB, is a member of the AGC kinase family. RSK1 is activated downstream of ERK1/2 and recognizes the same consensus motif as PKB (12,26). To determine whether RSK1 phosphorylates p122RhoGAP on Ser-322, we incubated recombinant wild-type p122RhoGAP and p122RhoGAP containing a Ser-3223 Ala mutation with active RSK1 and ATP and analyzed p122RhoGAP phosphorylation using the PAS antibody. Incubation of wild-type p122RhoGAP with active RSK1 led to phosphorylation of p122RhoGAP (Fig. 8). This phos-phorylation was reduced, although not blocked, in the Ser-3223 Ala mutant. These results demonstrate that RSK1 is able to phosphorylate p122RhoGAP on Ser-322 in vitro.

DISCUSSION
In this study, we have demonstrated that the small GTPase-activating protein p122RhoGAP is a novel insulin-stimulated PKB substrate in two different cell types, primary adipocytes and CHO.T cells. We show that PKB directly phosphorylates p122RhoGAP on Ser-322 in vitro and that expression of active mutants of PI 3-kinase or PKB in primary adipocytes leads to Ser-322 phosphorylation. As the PI 3-kinase inhibitor wortmannin blocks insulin-stimulated p122RhoGAP phosphorylation in these cells, we propose that PKB is both necessary and sufficient for the insulin-stimulated phosphorylation of Ser-322 on p122RhoGAP in primary adipocytes. In contrast, we found that insulin-stimulated p122RhoGAP phosphorylation in CHO.T cells, which overexpresses the insulin receptor, depends largely on MEK activity, as the MEK inhibitor UO126 strongly reduced p122RhoGAP phosphorylation. In these cells, wortmannin had little significant effect unless added in combination with UO126. Overexpression of a constitutively active MEK in CHO.T cells stimulates p122RhoGAP phosphorylation to the same extent as observed with insulin or a constitutively active PKB. These results demonstrate that the MEK/ERK pathway is both necessary and sufficient for the insulin-stimulated phosphorylation of Ser-322 on p122RhoGAP in CHO.T cells. In conclusion, insulin can stimulate p122RhoGAP Ser-322 phosphorylation via both the PI 3-kinase/PKB and MEK/ERK pathways, with the relative contribution of each pathway dependent on the cellular context.
Although PKB is the kinase most likely to phosphorylate Ser-322 downstream from PI 3-kinase, we demonstrated that RSK1, a member of the AGC kinase family, is the most likely candidate to phosphorylate p122RhoGAP downstream of MEK/ERK, as it directly phosphorylates p122RhoGAP in vitro. The majority of phosphorylation by RSK1 is Cells were subsequently stimulated with (ϩ) or without (Ϫ) insulin (87 nM) for 10 min prior to extraction. Anti-GFP immunoprecipitates (A) or total lysates (B-D) were subjected to SDS-PAGE followed by immunoblotting with the indicated antibodies (anti-PAS, anti-PKB pThr-308, anti-ERK pThr-202/pTyr204, anti-p70 S6K pThr-389). The membranes were stripped and reprobed with the appropriate antibodies to confirm loading. The bar graph (A) represents quantification of the phosphorylation of p122RhoGAP (ratio of phospho/total) expressed as percentage of the signal obtained with insulin alone (mean Ϯ S.E., n ϭ 4). A significant difference with respect to the insulin response, as determined by the two-tailed paired Student's t test, is indicated by asterisks (*, p Ͻ 0.05). FIGURE 5. Constitutively active mutants of PI 3-kinase and PKB stimulate p122RhoGAP Ser-322 phosphorylation in primary adipocytes. Primary adipocytes were co-transfected with p122RhoGAP and p110CAAX or myrPKB, as indicated. Cells were subsequently stimulated with or without insulin (87 nM) for 10 min prior to extraction. Anti-GFP immunoprecipitates were subjected to SDS-PAGE followed by immunoblotting with the PAS antibody. The membrane was stripped and reprobed with the p122RhoGAP antibody to confirm equal loading. Shown is a representative experiment of two performed. absent in the Ser-3223 Ala mutant, demonstrating that RSK1 phosphorylates p122RhoGAP at Ser-322. As the Ser-3223 Ala mutant showed some residual phosphorylation by RSK1, this kinase must phosphorylate at least one other site. However, this site(s) was not phosphorylated by insulin in intact CHO.T cells as the reactivity of p122RhoGAP toward the PAS antibody was completely abrogated in the Ser-3223 Ala mutant (Fig. 2B). Further support for the hypothesis that RSK1 (and not mitogen-and stress-activated protein and MAPK-inte- The membranes were stripped and reprobed with the appropriate antibodies to confirm loading. The bar graph (A) represents quantification of the phosphorylation of Ser-322 on p122RhoGAP (ratio of phospho/total) expressed as percentage of the signal obtained with insulin alone (mean Ϯ S.E., n ϭ 4 -7). A significant difference with respect to the insulin response, as determined by the twotailed paired Student's t test, is indicated by asterisks (*, p Ͻ 0.05, and **, p Ͻ 0.005). FEBRUARY 24, 2006 • VOLUME 281 • NUMBER 8 grating kinase) phosphorylates Ser-322 in p122RhoGAP comes from studies using the stress stimulus anisomycin. This agent strongly activates p38␣ mitogen-activated protein kinase in intact cells, resulting in the activation of mitogen-and stress-activated protein and MAPK-integrating kinase but not RSK (27). p122RhoGAP does not become phosphorylated in the presence of anisomycin, even though p38 mitogenactivated protein kinase is strongly activated, 4 suggesting that mitogenand stress-activated protein and MAPK-integrating kinase are not involved in p122RhoGAP phosphorylation.

Phosphorylation of p122RhoGAP (DLC-1) by PKB and RSK
The finding that Ser-322 on p122RhoGAP is a dual PKB and RSK1 substrate in vivo is of significant interest, as it can be added to a small but growing list of similar phosphorylation sites on proteins. For example, Ser-9 of glycogen synthase-3 has been reported to be a substrate for both PKB and RSK in cells (28,29). Similarly, both PKB and RSK are reported to be able to phosphorylate Ser-939, Thr-1462, and Ser-1798 of TSC2 (30), Ser-52 of a calcium-regulated heat-stable protein called CRHSP24 (31), and Ser-27 of the tRNA methylase METTL1 (32). All of these proteins are phosphorylated on the same site by PKB and RSK in response to different agonists that activate either the PI 3-kinase or the classical mitogen-activated protein kinase pathway, respectively. The contribution of each of these pathways to p122RhoGAP phosphorylation on Ser-322, therefore, may depend on their relative extent of activation by insulin. The ERK1/2 pathway, for example, is strongly activated in CHO.T cells, whereas its activation is weaker in primary adipocytes (compare Figs. 4C and 6C). This may explain the greater contribution of the ERK1/2 pathway in p122RhoGAP phosphorylation in CHO.T cells compared with primary adipocytes. Clearly, however, both the PI 3-kinase/PKB and MEK/ERK/RSK pathways are activated in primary adipocytes and CHO.T cells in response to insulin; therefore, another mechanism may be involved. For example, differences in the subcellular localization of PKB and RSK1, relative to p122RhoGAP, may be responsible for determining the relative contribution of each pathway in these cell types. These possibilities warrant further exploration.
GTPase-activating proteins (GAPs) are involved in regulating the activity of small GTP-binding proteins, which switch between an active GTP-bound form and an inactive GDP-bound form. GAP proteins promote GTP hydrolysis leading to inactivation of the small G-protein.
p122RhoGAP has been previously identified as a PLC␦ 1 -binding protein that contains a sterile ␣-motif domain (a protein-protein interaction domain) at the N terminus and a GTPase-activating (GAP) domain and a StAR-related lipid transfer domain toward the C terminus (25). The GAP domain has been reported to be selective in stimulating the intrinsic GTPase activity of RhoA but not of Rac (25,33). Interestingly, three other GAP proteins have previously been identified as insulin-stimulated phosphoproteins and in vitro substrates for PKB using a similar proteomic approach to our own. These are the tuberous sclerosis complex-2 (TSC2), which is a Rheb GAP (34 -36), AS160, which is a GAP for Rab2A, -8A, -10, and -14 (14,20), and pp250, which possesses a predicted C-terminal GAP domain for Rheb and Rap (37). PKB phosphorylation of TSC2 has been proposed to inhibit its GAP activity toward Rheb, which would lead to increased Rheb GTP binding and by implication, Rheb activity (38 -41). This is consistent with the fact that insulin increases the amount of GTP-bound Rheb in cells (38). Rheb subsequently activates mTOR by an unknown mechanism leading to an increase in translation and cell growth (38 -40). Phosphorylation of AS160 by PKB occurs on multiple serine residues, and an AS160 mutant lacking two or more of these phosphorylation sites has been reported to markedly inhibit insulin-stimulated GLUT4 translocation (21). This dominant inhibitory effect on GLUT4 translocation was largely reversed by introducing an additional mutation that inactivates the GAP domain, suggesting that insulin inhibits the GAP activity of AS160 (21). As insulin increases GTP-binding to RhoA in both rat adipocytes (42) and CHO.T cells, 4 phosphorylation of p122RhoGAP by PKB may also inhibit its GAP activity as has been proposed to be the case for TSC2 and AS160. Thus, direct phosphorylation of GAP proteins by PKB may be a more generic mechanism by which the GAP activity of these proteins is regulated by insulin and other factors that activate PKB. This hypothesis requires confirmation, as there is currently no direct evidence that PKB phosphorylation alters the GAP activity of any of these GTPase-activating proteins.
The biological consequences for the proposed PI 3-kinase/PKB/ p122RhoGAP/RhoA signaling pathway is not yet known. RhoA is a key regulator of the actin cytoskeleton, in particular actin stress fiber formation, and the regulation of several downstream effector kinases, including ROK (Rho-kinase) and PKN/PRK (protein kinase novel) (43). Indeed, overexpression of p122RhoGAP in 3T3-L1 and CHO.T cells resulted in the dissolution of actin stress fibers, profound changes in cell morphology (cell rounding), and eventual loss of cells from the substratum, 4    . Active PKB and active MEK stimulate p122RhoGAP phosphorylation in CHO.T cells. CHO.T cells were co-transfected with p122RhoGAP and myrPKB or MEK-MANE, as indicated. The cells were subsequently stimulated with (ϩ) or without (Ϫ) insulin (87 nM) for 10 min prior to extraction. Anti-GFP immunoprecipitates were split in two and subjected to SDS-PAGE followed by immunoblotting with the PAS antibody and p122RhoGAP antibody, respectively. Shown is a representative experiment of three performed.
precluded our attempts to examine the role of p122RhoGAP phosphorylation in cytoskeletal rearrangements and proliferation in CHO.T cells and in GLUT4 translocation in 3T3-L1 adipocytes. In our hands, GLUT4 translocation cannot be measured in transfected primary rat adipocytes as they do not retain adequate insulin responsiveness during long term culture.
RhoA has been implicated in the regulation of glucose uptake and GLUT4 translocation to the plasma membrane by insulin, although this is controversial. For example, the C3 toxin from Clostridium botulinum, which inhibits RhoA function, has been reported to mimic (44), inhibit (45,46), as well as have no effect (47) on insulin-stimulated glucose uptake and GLUT4 translocation.
p122RhoGAP is the rat homologue of human DLC1, which has been reported to undergo loss of heterozygosity or genomic silencing in a significant proportion of human primary hepatocellular carcinomas (48,49), non-small cell lung carcinomas (50), and breast carcinomas (51). This suggests that the DLC1 and p122RhoGAP proteins function as tumor suppressors. This would be consistent with our hypothesis that the phosphorylation of Ser-322 in response to activation of the MEK/ERK/RSK pathway by growth factors inhibits GAP activity toward RhoA, leading to RhoA activation and then enhanced cell growth and/or motility. Importantly, the sequence surrounding Ser-322 in rodent p122RhoGAP (SPVTRTRSLS 322 TCNKR) is almost completely conserved in the region surrounding the equivalent serine Ser-329 of human DLC1 (SPVTRTRSLS 329 ACNKR), making it likely that this phosphorylation event is also important in regulating human DLC1 function.
In NRK (normal rat kidney) cells, endogenous p122RhoGAP is localized to focal adhesions via a region covering residues 117-533 (22). As Ser-322 falls within this domain, its phosphorylation may be involved in regulating the targeting of p122RhoGAP to focal adhesions. However, despite the profound morphological changes alluded to above, we have been able to demonstrate that Ser-3223 Ala and Ser-3223 Asp mutants both showed similar localizations to focal adhesions as the wild-type p122RhoGAP. 5 Furthermore, DLC-2, which does not have an equivalent to Ser-322, was also found localized in focal adhesions. Phosphorylation of Ser-322 is therefore unlikely to control targeting of p122RhoGAP to focal adhesions.
In conclusion, we have demonstrated that p122RhoGAP Ser-322 is a novel PKB and RSK1 substrate and insulin-stimulated phosphoprotein in physiologically relevant cells as well as immortalized cells in culture. This site has the potential to integrate the activities of two different signal transduction pathways in a manner dependent on the cellular context. Future studies will focus on identifying the function of p122RhoGAP phosphorylation in signaling events downstream of PKB, such as GLUT4 translocation, the activation of RhoA effectors, and cellular transformation.