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

J. Biol. Chem., Vol. 278, Issue 26, 23472-23479, June 27, 2003

This Article Abstract Full Text (PDF) All Versions of this Article: 278/26/23472    most recent M212232200v1 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 Ballou, L. M. Articles by Lin, R. Z. Search for Related Content PubMed PubMed Citation Articles by Ballou, L. M. Articles by Lin, R. Z. Social Bookmarking                      What's this?

Activated G_q Inhibits p110 Phosphatidylinositol 3-Kinase and Akt^*

Lisa M. Ballou , Hong-Ying Lin  , Gaofeng Fan , Ya-Ping Jiang  and Richard Z. Lin  ¶ ||

From the ^¶Research Service, Department of Veterans Affairs Medical Center, Northport, New York 11768 and the Department of Medicine, Stony Brook University, Stony Brook, New York 11794

Received for publication, December 2, 2002 , and in revised form, April 14, 2003.

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Some G_q-coupled receptors have been shown to antagonize growth factor activation of phosphatidylinositol 3-kinase (PI3K) and its downstream effector, Akt. We used a constitutively active G_q(Q209L) mutant to explore the effects of G_q activation on signaling through the PI3K/Akt pathway. Transient expression of G_q(Q209L) in Rat-1 fibroblasts inhibited Akt activation induced by platelet-derived growth factor or insulin treatment. Expression of G_q(Q209L) also attenuated Akt activation promoted by coexpression of constitutively active PI3K in human embryonic kidney 293 cells. G_q(Q209L) had no effect on the activity of an Akt mutant in which the two regulatory phosphorylation sites were changed to acidic amino acids. Inducible expression of G_q(Q209L) in a stably transfected 293 cell line caused a decrease in PI3K activity in p110 (but not p110) immunoprecipitates. Receptor activation of G_q also selectively inhibited PI3K activity in p110 immunoprecipitates. Active G_q still inhibited PI3K/Akt in cells pretreated with the phospholipase C inhibitor U73122 [GenBank] . Finally, G_q(Q209L) co-immunoprecipitated with the p110-p85 PI3K heterodimer from lysates of COS-7 cells expressing these proteins, and incubation of immunoprecipitated G_q(Q209L) with purified recombinant p110-p85 in vitro led to a decrease in PI3K activity. These results suggest that agonist binding to G_q-coupled receptors blocks Akt activation via the release of active G_q subunits that inhibit PI3K. The inhibitory mechanism seems to be independent of phospholipase C activation and might involve an inhibitory interaction between G_q and p110 PI3K.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Phosphatidylinositol 3-kinase (PI3K)¹ mediates many of the cellular actions of receptor tyrosine kinases, including effects on glucose metabolism, cell survival, and cytoskeletal rearrangements (1). Of the three classes of PI3K, only the class I enzymes preferentially phosphorylate phosphatidylinositol (PI) 4,5-bisphosphate in vivo. Class I PI3Ks are divided into two groups: I_A enzymes are heterodimers between a p110 catalytic subunit and a p85 or p55 (or their splice variants) regulatory subunit. Mammals have three class I_A p110 catalytic subunits (, , and ) that can associate with at least seven regulatory subunits that are generated by alternative splicing of three different genes (p85, p85, and p55). p110 and p110 are ubiquitously expressed, whereas p110 is present almost exclusively in leukocytes. Class I_A PI3Ks can be activated by receptor tyrosine kinases, and the p110 isoform is also activated by some G protein-coupled receptors (2). The p85 regulatory subunit contains two SH2 (Src homology 2) domains that bind to specific phosphotyrosine motifs on receptor tyrosine kinases or their substrates. This binding leads to translocation of p110 to the membrane and enhances its catalytic activity. The class I_B PI3K consists of a p110 catalytic subunit and a p101 regulatory protein and is activated by some receptors coupled to heterotrimeric G proteins.

An important downstream effector of PI3K is the serine/threonine protein kinase Akt. Akt is activated by phosphorylation of Thr³⁰⁸ in the activation loop of the kinase domain and Ser⁴⁷³ in the C-terminal tail (reviewed in Ref. 3). It is believed that phosphorylation of both sites requires an interaction between the N-terminal pleckstrin homology domain of Akt and membrane phosphoinositides generated by PI3K. Although a few cell treatments have been reported to activate Akt in a PI3K-independent manner (4, 5), receptor tyrosine kinases that activate Akt also in general activate PI3K.

Cell-surface receptors that transmit signals through heterotrimeric G proteins regulate the PI3K/Akt pathway in a variety of ways. Receptors that couple to proteins in the G_i/o family can increase PI3K activity. This effect is mediated mainly by G heterodimers, which activate both the p110 and p110 PI3Ks (6–8). Recent reports indicate that G_i/o-coupled receptors also active Akt (9–11). Receptors that couple to G_s have also been shown to activate Akt. In some cases, this response is unusual in that it appears to be independent of PI3K (4, 12), whereas in other cases, it is thought to be due to activation of PI3K by G subunits (11).

The data regarding regulation of PI3K and/or Akt by G_q-coupled receptors are more controversial. One body of evidence suggests that some receptors that can couple to G_q might activate PI3K/Akt signaling in certain cellular settings (13–19). Indeed, we reported that norepinephrine stimulation of ₁-adrenergic receptors in human aortic smooth muscle cells increases PI3K activity (20). However, this effect is completely blocked by pretreatment with pertussis toxin, indicating that it is not mediated by G_q (20). A second body of evidence suggests that G_q-coupled receptors do not activate PI3K or Akt and in fact antagonize activation of these enzymes by growth factors that act through tyrosine kinase receptors (11, 21–25). In agreement with these reports, we found that stimulation of the _1A-adrenergic receptor in Rat-1 cells with the agonist phenylephrine (PE) does not increase PI(3,4,5)P₃ levels or PI3K activity and does not activate Akt (26). Furthermore, the _1A-adrenergic receptor inhibits activation of PI3K/Akt in response to platelet-derived growth factor (PDGF), insulin, and insulin-like growth factor I (26, 27). To further explore the mechanism by which the _1A-adrenergic receptor and other G_q-coupled receptors inhibit PI3K signaling, we used G_q(Q209L) to test whether activated G_q inhibits Akt activation by inhibiting PI3K.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Materials—Human recombinant PDGF-A/B, PE, doxycycline, insulin, PI, U73122 [GenBank] , GTPS, and monoclonal antibody to FLAG were purchased from Sigma. [-³²P]ATP (3000 Ci/mmol) and myo-[³H]inositol (10–25 Ci/mmol) were from PerkinElmer Life Sciences. Rabbit polyclonal antibodies against G_q and Akt (H-136) were from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-p110, anti-p110, and anti-p85 antibodies were from Upstate Biotechnology, Inc. (Lake Placid, NY). Anti-phospho-Ser⁴⁷³ Akt antibody was from Cell Signaling Technology (Beverly, MA). Anti-HA and anti-Myc antibodies were from Covance (Richmond, CA).

Constructs—An epitope-tagged Akt construct (Akt-HA) was obtained from Dr. Richard Roth (Stanford University, Stanford, CA). PCR with primers 5'-CGCCTCGAGGCCACCATGGGCAGCGACGTGGCTATTGTGAAG (forward) and 5'-CGCGATATCTCAGGCCGTGCTGCTGGCCGA (reverse) was used to amplify Akt from Akt-HA, and the cDNA fragment was subcloned into pBluescript II SK. Akt was excised using XbaI and XhoI and subcloned into pcDNA3.1/Myc-His (Invitrogen) to obtain Akt-Myc. The Akt(T308D/S473D) double mutant (referred to as AktDD-Myc) was constructed using the QuikChange site-directed mutagenesis kit (Stratagene) and forward primers 5'-GGTGCCACCATGAAGGACTTTTGCGGCACACCT (T308D) and 5'-CACTTCCCCCAGTTCGACTACTCGGCCAGC (S473D).

Mouse wild-type p110 and constitutively active myristoylated (myr) p110 in pUSEamp were purchased from Upstate Biotechnology, Inc. FLAG-p110 was obtained by subcloning p110 from pUSEamp into the XbaI and BamHI sites of p3XFLAG-CMV-10 (Sigma). Mouse p85 was obtained by PCR from mouse liver cDNA (Clontech) using primers 5'-GCGGAATTCATGAGTGCAGAGGGCTACCA (forward) and 5'-GCGGGATCCTCATCGCCTCTGTTGTGC (reverse). The p85 cDNA fragment was subcloned into the BamHI and EcoRI sites of pBluescript II SK and then subcloned into pcDNA3.1 using XbaI and EcoRV. PCR with primers 5'-CGCGGTACCGCCACCATGGCATACCCCTACGACGTGCCCGACTACGCCACTCTGGAGTCCATCATGGC (forward) and 5'-CGCGGATCCTTAGACCAGATTGTACTCCTTCAG (reverse) was used to obtain HA-G_q from Swiss mouse 3T3 cell cDNA. The cDNA fragment was subcloned into pcDNA3.1 using KpnI and EcoRV. HA-G_q(Q209L) was constructed from HA-G_q using the QuikChange site-directed mutagenesis kit and the forward primer 5'-GTCGATGTAGGGGGCCTAAGGTCAGAGAGAAG. PCR with primers 5'-CGCGGATCCGCCACCATGACTCTGGAGTCCATCATGGC (forward) and 5'-CGCGGATCCTTAGACCAGATTGTACTCCTTCAG (reverse) was used to obtain G_q(Q209L) from HA-G_q(Q209L). G_q(Q209L) was subcloned into pcDNA5/FRT/TO (Invitrogen) using NotI and ApaI.

Cell Culture—Rat-1 fibroblasts stably expressing the human _1A-adrenergic receptor (28) or the human insulin receptor (29), human embryonic kidney (HEK) 293 cells, and COS-7 cells were maintained in Dulbecco's modified Eagle's medium (Mediatech, Herndon, VA) with 10% fetal bovine serum (Sigma) in 5% CO₂ at 37 °C. Before agonist treatments, the cells were incubated overnight in serum-free medium. COS-7 cells were transfected using LipofectAMINE (Invitrogen). After 5 h, the transfection solutions were replaced with growth medium, and cell lysates were prepared 2 days later. Rat-1 cells were transfected using TransIT-LT1 (Mirus, Madison, WI) following a protocol supplied by the manufacturer. Cells at 50% confluence were transfected in a mixture containing 3 µl of TransIT-LT1 for each microgram of DNA in growth medium. The cells were left in the transfection solution for 24 h and then incubated in serum-free medium for another 16–18 h prior to the experiment.

Stable HEK 293 cell lines expressing G_q(Q209L) under the control of a doxycycline-inducible promoter and control cells were generated using the Flp-In T-REx system (Invitrogen) following the protocol supplied by the manufacturer. Briefly, Flp-In T-REx/293 cells growing in medium containing Zeocin were cotransfected with either G_q(Q209L)/pcDNA5/FRT/TO or the empty vector pcDNA5/FRT/TO together with pOG44. Transfected cells were then selected for Zeocin and hygromycin resistance. Clones with inducible expression of G_q(Q209L) were confirmed by Western blotting.

Cell Lysate Preparation—After treatment, cells were rinsed with ice-cold phosphate-buffered saline and scraped into lysis buffer (50 mM HEPES (pH 7.5), 50 mM NaCl, 5 mM EDTA, 50 mM NaF, 10 mM pyrophosphate, 1 mM sodium orthovanadate, 0.5 mM phenylmethylsulfonyl fluoride, and 10 µg/ml each aprotinin and leupeptin) with either 1% Triton X-100 or 1% Nonidet P-40 plus 0.25% sodium deoxycholate. Homogenates were centrifuged at 15,000 x g for 15 min at 4 °C, and protein concentrations were determined using the Bradford assay (Bio-Rad).

Immunoprecipitation—Cell lysates containing equal amounts of protein were incubated with the appropriate antibody for 2 h and then with 25 µl of protein A- or protein G-agarose (Sigma) for 1 h. The beads were either washed three times with lysis buffer and used for immunoblotting or washed three times with lysis buffer and once with the appropriate kinase assay buffer prior to performing kinase assays.

Western Blotting—Immunoprecipitates or equal amounts of cell lysate protein were subjected to SDS-PAGE, followed by electrophoretic transfer onto nitrocellulose or polyvinylidene difluoride membranes. Signals were visualized using horseradish peroxidase-linked secondary antibodies (Amersham Biosciences) and a chemiluminescence kit (PerkinElmer Life Sciences). Blots were stripped as described (26).

PI3K and Akt Assays—PI3K and Akt activities were assayed following methods described previously (26).

Phospholipid Analysis—Rat-1 cells were labeled with myo-[³H]inositol as described (26) and treated with agonists. Phospholipids were extracted and analyzed as described previously (26), except the HPLC gradient was modified to better separate the PI(3,4,5)P₃ derivative from other compounds (30). The solutions used to develop the gradient were ultrapure water (solution A) and 1.25 M (NH₄)₂HPO₄ (pH 3.8) (solution B). The gradient was developed as follows: 0% solution B from 0 to 10 min, increased to 15% from 10 to 48 min; 15% solution B from 48 to 60 min, increased to 25% from 60 to 75 min, and increased to 30% from 75 to 100 min. The flow rate was 1 ml/min, and 1-ml fractions were collected. The PI(3,4,5)P₃ derivative emerged after 82 min. Elution positions were confirmed using ³²P-labeled standards as described (26).

Purification of Recombinant p110-p85—Baculovirus expressing human p85 was purchased from Orbigen (San Diego, CA). To produce baculovirus expressing p110, mouse p110 from pUSEamp was subcloned into pBlueBacHis2A (Invitrogen). Sf9 cells were cotransfected with the plasmid and linear Autographa californica nuclear polyhedrosis virus viral DNA using the Invitrogen BAC-N-BLUE transfection kit, and a baculovirus clone expressing recombinant p110 protein was isolated. To make the p110-p85 complex, a 400-ml culture of Sf9 cells was co-infected with the two baculoviruses. Two days later, the cells were pelleted and lysed, and the lysate was passed over a 5-ml HiTrap QFF column (Amersham Biosciences). The proteins were eluted with a gradient of NaCl; PI3K activity emerging at 180–270 mM NaCl was pooled. Material from three QFF columns was passed over a column of Ni²⁺-nitrilotriacetic acid-agarose (QIAGEN Inc.). After extensive washing with 50 mM imidazole, the PI3K complex was eluted with 200 mM imidazole. Fractions with the highest activity contained equivalent amounts of two major Coomassie Blue-stained proteins corresponding to recombinant p110 and p85.

Quantitation of Inositol Phosphates—Rat-1 cells expressing the human _1A-adrenergic receptor were seeded in 24-well plates at 2.5 x 10⁴ cells/well. The next day, cells were labeled for 16–24 h in 1 ml of serum-free medium containing 3 µCi/ml myo-[³H]inositol. After labeling, the cells were washed with phosphate-buffered saline and incubated in 0.5 ml of buffer A (120 mM NaCl, 0.5 mM CaCl₂, 5 mM KCl, 5.6 mM glucose, 0.4 mM MgCl₂, 20 mM LiCl, and 25 mM PIPES (pH 7.2)) with or without 5 µM U73122 [GenBank] for 30 min. Cells were then treated with agonists for another 30 min. COS-7 cells were seeded at 2.5 x 10⁴ cells/well in 24-well plates 1 day before transfection with FLAG-p110 and p85 in the presence of either empty vector or HA-G_q(Q209L). Twenty-four hours after transfection, cells were incubated in 1 ml of growth medium containing 3 µCi/ml myo-[³H]inositol for 16–24 h. Cells were then washed and incubated in buffer A with or without 5 µM U73122 [GenBank] for 1 h. After cell treatments, 1 ml of 16 mM HCl in methanol was added to each well, and the solutions were applied to 500-µl columns of Dowex 1-X8 (formate form). The columns were washed with 5 mM sodium tetraborate and 60 mM sodium formate, and then total inositol phosphates were eluted with 2 ml of 1 M ammonium formate and 100 mM formic acid. Radioactivity in the eluted material was determined by scintillation counting. Assays were performed in triplicate.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
G_q(Q209L) Inhibits Growth Factor Activation of Akt—Stimulation of G protein-coupled receptors leads to GTP loading of the G subunit and release of the G subunits. The G and G subunits can then independently exert their effects on downstream effectors. Although the _1A-adrenergic receptor couples mainly to proteins in the G_q/11 family, the receptor can also activate proteins in other families (31, 32). Inhibition of Akt upon PE stimulation of the _1A-adrenergic receptor in Rat-1 cells is probably not due to coupling of the receptor to G_i or G_s because the effect was not blocked by pretreatment with pertussis toxin or 2',5'-dideoxyadenosine, an adenylyl cyclase inhibitor, respectively (data not shown). Because G subunits have been shown to activate PI3K, we decided to explore a possible role for G_q in PI3K/Akt inhibition. To do this, we used an HA-tagged version of the Q209L mutant of G_q, which is GTPase-deficient and therefore constitutively activates downstream effectors such as phospholipase C without agonist stimulation of a receptor (33).

Rat-1 cells were cotransfected with Akt-Myc and with either HA-G_q(Q209L) or empty vector as a control. The cells were treated with or without PDGF, and Akt activity was assayed in Myc immunoprecipitates. PDGF treatment stimulated Akt activity 3.6-fold in control cells, and this response was suppressed 58% by expression of HA-G_q(Q209L) (Fig. 1A, graph). Western blotting showed that Akt-Myc and HA-G_q(Q209L) were expressed appropriately (Fig. 1A, blots). An additional experiment using a different growth factor was done to test the generality of Akt inhibition by HA-G_q(Q209L). Rat-1 cells stably expressing the human insulin receptor (Rat-1/HIR cells) were transfected with either HA-G_q(Q209L) or empty vector as a control. The cells were treated with or without insulin, and the activity of endogenous Akt was assayed in Akt immunoprecipitates. Insulin activated Akt 3.4-fold in control cells, and this response was suppressed 38% by expression of HA-G_q(Q209L) (Fig. 1B, graph). Western blot analysis showed that the reduction in Akt activity was paralleled by a reduction in Ser⁴⁷³ phosphorylation (Fig. 1B, upper blot). HA-G_q(Q209L) did not affect the expression of the Akt protein (Fig. 1B, middle blot). The blot was also probed with anti-HA antibody to demonstrate that HA-G_q(Q209L) was appropriately expressed (Fig. 1B, lower blot). Thus, expression of HA-G_q(Q209L) opposes the activation of Akt induced by both the PDGF and insulin receptors. The apparently stronger effect of HA-G_q(Q209L) in Fig. 1A compared with Fig. 1B is due to the fact that Akt-Myc was cotransfected with HA-G_q(Q209L) in the former experiment, whereas endogenous Akt was analyzed in the latter. Transfection of Rat-1/HIR cells with a vector expressing green fluorescent protein showed that the transfection efficiency under the conditions used in Fig. 1B was 60% (data not shown).

View larger version (28K): [in this window] [in a new window]   FIG. 1. Effect of Gq(Q209L) on Akt activation. A, Rat-1 cells were cotransfected with Akt-Myc and with either HA-Gq(Q209L) or empty vector as a control. Serum-starved cells were treated for 5 min with or without 25 ng/ml PDGF. Akt activity was measured in Myc immunoprecipitates (graph). Data shown are means ± S.E. from three independent experiments. Expression of Akt-Myc and HA-Gq(Q209L) was confirmed by Western blot analysis of cell lysate proteins using anti-epitope antibodies (blots). B, Rat-1/HIR cells were transfected with HA-Gq(Q209L) or empty vector as a control. Serum-starved cells were treated for 5 min with or without 100 nM insulin, and endogenous Akt activity was measured in Akt immunoprecipitates (graph). Data shown are means ± S.E. from three independent experiments. A Western blot of cell lysate proteins was probed sequentially with antibodies to phospho-Ser473 Akt (P-S473 Akt), total Akt, and HA (blots). Western blotting was repeated with similar results.

G_q(Q209L) Inhibits Myristoylated p110 Activation of Akt—The inhibitory effect of G_q(Q209L) on Akt activation induced by two distinct receptors raised the possibility that active G_q might have a direct inhibitory effect on either Akt or PI3K. To test whether G_q(Q209L) targets Akt, HEK 293 cells were cotransfected with HA-G_q(Q209L) or empty vector as a control and with an activated form of Akt in which the two phosphorylation sites were mutated to Asp (AktDD-Myc). Cells were serum-starved overnight prior to assaying Akt activity in Myc immunoprecipitates. Extracts from cells transfected with AktDD-Myc had four to six times more Akt kinase activity than extracts from cells transfected with Akt-Myc (Fig. 2A). The presence of HA-G_q(Q209L) did not affect the activity of AktDD-Myc (Fig. 2A). These results suggest that activated G_q does not have a direct inhibitory effect on Akt.

View larger version (29K): [in this window] [in a new window]   FIG. 2. Effect of Gq(Q209L) on myr-p110 activation of Akt. A, HEK 293 cells were cotransfected with Akt-Myc or AktDD-Myc and with combinations of myr-p110, HA-Gq(Q209L), or empty vector as a control. Cells were serum-starved overnight, and Akt activity was assayed in Myc immunoprecipitates. Data shown are means ± S.E. from three independent experiments. B, cell extracts from A (representing the first, fourth, and fifth bars) were analyzed on a Western blot probed sequentially with antibodies to phospho-Ser473 Akt (P-S473 Akt-myc), Myc, HA, and p110. Western blotting was repeated with similar results.

Expression of membrane-localized myr-p110 is sufficient to trigger downstream signaling events in the absence of growth factors. It is believed that when p110 is directed to the membrane, the increased availability of lipid substrates leads to increased production of PI(3,4,5)P₃ and subsequent activation of Akt, even though the enzymatic activity of myr-p110 is no higher than that of wild-type p110 (34). We tested whether activated G_q can inhibit myr-p110 signaling to Akt. HEK 293 cells were cotransfected with Akt-Myc in the presence or absence of myr-p110 or HA-G_q(Q209L). Cells were serumstarved overnight, and extracts were analyzed on a Western blot to detect phospho-Ser⁴⁷³ in Akt-Myc. (Akt-Myc migrated more slowly than wild-type endogenous Akt on SDS-polyacrylamide gels, so the two proteins were well separated on Western blots.) The level of Ser⁴⁷³ phosphorylation was strongly increased in cells expressing myr-p110, and coexpression of HA-G_q(Q209L) reduced the phosphorylation at this site (Fig. 2B, upper panel). Consistent with these results, Akt activity in Myc immunoprecipitates was 40% lower in cells expressing both myr-p110 and HA-G_q(Q209L) than in cells expressing only myr-p110 (Fig. 2A). The blot was reprobed with antibodies to Myc, HA, and p110 to demonstrate that Akt-Myc, HA-G_q(Q209L), and myr-p110 were appropriately expressed (Fig. 2B, lower three panels). Together, these results suggest that G_q(Q209L) might have a direct inhibitory effect on PI3K, but not on Akt.

Activated G_q Inhibits p110 PI3K Activity—To further examine the effect of constitutively active G_q on PI3K, a stable cell line that expresses untagged G_q(Q209L) in a doxycycline-responsive manner was constructed using a commercially available system (see "Experimental Procedures"). Flp-in T-REx/HEK 293 cells containing the empty vector (control) or G_q(Q209L) were treated overnight with 1 µM doxycycline. Western blotting of cell lysates using antibody to G_q detected the wild-type endogenous protein in control cells and a stronger signal comprising the wild-type and mutant proteins in G_q(Q209L) cells (Fig. 3A). Cell extracts were mixed with polyclonal antibody to either p110 or p110, and the immunoprecipitates were assayed for PI3K activity. The activity in p110 immunoprecipitates from cells expressing G_q(Q209L) was only 39% of that in immunoprecipitates from control cells (Fig. 3B). In contrast, PI3K activities in p110 immunoprecipitates from control cells and cells expressing G_q(Q209L) were not significantly different (Fig. 3B). Western blotting showed that expression of the p110 and p110 proteins was the same in control and G_q(Q209L) cells (Fig. 3A).

View larger version (27K): [in this window] [in a new window]   FIG. 3. Effect of Gq(Q209L) on PI3K activity. Flp-in T-REx/HEK 293 cells containing Gq(Q209L) or empty vector as a control were treated overnight with 1 µM doxycycline prior to making extracts. A, expression of Gq(Q209L) was confirmed by Western blotting with antibody to Gq (upper panel). The blot was reprobed with antibodies to p110 (middle panel) and p110 (lower panel). B, PI3K activity was assayed in p110 or p110 immunoprecipitates (IP). The autoradiogram shows a typical result from four independent experiments. Spots containing radioactive PI(3)P were scraped off the silica gel plates and quantitated by scintillation counting. Mean p110 activity in cells expressing Gq(Q209L) was 39% of the activity in control cells containing the empty vector (S.E. = 2.3%; p < 0.01 by t test). Mean p110 activity in cells expressing Gq(Q209L) was 99% of the activity in control cells (S.E. = 2.5%; not statistically significant).

We next investigated whether receptor activation of G_q also affects p110 activity. Rat-1 cells expressing the _1A-adrenergic receptor were treated with or without PE in the presence of PDGF, and PI3K activity was measured in p110 or p110 immunoprecipitates. PE treatment caused a 53% reduction in PI3K activity in p110 immunoprecipitates compared with cells not treated with PE (Fig. 4A). Similar to what was observed in Flp-in T-REx/HEK 293 cells expressing G_q(Q209L), PE activation of G_q did not have a significant effect on PI3K activity in p110 immunoprecipitates (Fig. 4A). We also assayed overall PI3K signaling by measuring levels of PI(3,4,5)P₃ in Rat-1 cells following PE and PDGF treatment. The amount of PI(3,4,5)P₃ in cells treated with PE alone was not significantly increased above control levels, whereas PDGF induced a 2.6-fold increase in PI(3,4,5)P₃ levels (Fig. 4B). In cells cotreated with PE, the PDGF-induced increase in PI(3,4,5)P₃ was reduced by 64% (Fig. 4B). Taken together, these results suggest that activated G_q inhibits p110 PI3K activity, resulting in decreased growth factor-mediated PI(3,4,5)P₃ production and Akt activation.

View larger version (19K): [in this window] [in a new window]   FIG. 4. Effect of the 1A-adrenergic receptor on PI3K activity. A, Rat-1 cells expressing the 1A-adrenergic receptor were treated with or without 10 µM PE for 5 min in the presence of 50 ng/ml PDGF. PI3K activity was measured in p110 or p110 immunoprecipitates (IP). The autoradiogram shows a representative result from three independent experiments. Mean p110 activity in PE-treated cells was 47% of the activity in cells not treated with PE (S.E. = 3.2%; p < 0.01 by t test). Mean p110 activity in PE-treated cells was 98% of the activity in control cells (S.E. = 3.1%; not statistically significant). B, Rat-1 cells expressing the 1A-adrenergic receptor were labeled with myo-[3H]inositol and then treated for 5 min with 50 ng/ml PDGF in the presence or absence of 10 µM PE. Phospholipids were extracted, deacylated, and analyzed by HPLC (see "Experimental Procedures"). The amount of PI(3,4,5)P3 was normalized to the total amount of myo-[3H]inositol-labeled material recovered from the column. Data shown are means ± S.E. from three independent experiments. *, significant difference between PDGF and PE + PDGF (p < 0.01). Data were analyzed by one-way analysis of variance, and pairwise comparisons were made using Fisher's post-hoc tests.

Involvement of Phospholipase C in PI3K/Akt Inhibition— Activated G_q directly stimulates phospholipase C, leading to production of diacylglycerol and inositol 1,4,5-trisphosphate, release of Ca²⁺ from intracellular stores, and activation of protein kinase C (35). We noted earlier that PE still inhibits Akt activation by insulin-like growth factor I in cells depleted of intracellular Ca²⁺ (26). We therefore did additional experiments to assess the involvement of phospholipase C signaling in PI3K/Akt inhibition by activated G_q. First, we tested whether the phospholipase C inhibitor U73122 [GenBank] blocks the inhibitory effect of the _1A-adrenergic receptor on insulin activation of Akt. Insulin-induced phosphorylation of Akt was unaffected by pretreatment of Rat-1 cells with U73122 [GenBank] (Fig. 5A). More importantly, the inhibitory effect of PE on Akt phosphorylation was still largely intact in cells pretreated with U73122 [GenBank] (Fig. 5A). Similarly, treatment of Rat-1 cells for 24 h with 100 nM phorbol 12-myristate 13-acetate to down-regulate diacylglycerol-dependent protein kinase C isoforms did not abrogate the inhibitory effect of PE on insulin-induced Akt phosphorylation (data not shown). We also tested the effect of U73122 [GenBank] on the inhibition of PI3K by HA-G_q(Q209L). In control COS-7 cells cotransfected with FLAG-p110 and p85, expression of HA-G_q(Q209L) caused a 53% reduction in PI3K activity in FLAG immunoprecipitates (Fig. 5B). U73122 [GenBank] had a small inhibitory effect by itself on PI3K activity. However, expression of HA-G_q(Q209L) still caused a 59% decrease in PI3K activity in cells pretreated with U73122 [GenBank] (Fig. 5B). These data suggest that inhibition of PI3K by activated G_q occurs independently of the phospholipase C pathway.

View larger version (43K): [in this window] [in a new window]   FIG. 5. U73122 resistance of PI3K/Akt inhibition by PE and Gq(Q209L). A, serum-starved Rat-1 cells expressing the 1A-adrenergic receptor were incubated with 5 µM U73122 [GenBank] or vehicle for 30 min prior to stimulation with 1 µM insulin in the presence or absence of 10 µM PE for 5 min. Phospho-Ser473 Akt (P-S473 Akt; upper panel) and total Akt (lower panel) were detected on a Western blot of cell lysate proteins. Control experiments (see "Experimental Procedures") showed that PE caused 6.8- and 2.9-fold increases in inositol phosphate levels in control and U73122 [GenBank] -treated cells, respectively. B, COS-7 cells were cotransfected with FLAG-p110 and p85 in the presence of HA-Gq(Q209L) or empty vector. Cells were treated 48 h later with 5 µM U73122 [GenBank] or vehicle for 1 h. PI3K activity was then assayed in FLAG immunoprecipitates. Control experiments (see "Experimental Procedures") showed that expression of HA-Gq(Q209L) caused 7.7- and 4.7-fold increases in inositol phosphate levels in control and U73122 [GenBank] -treated cells, respectively. The experiments were repeated with similar results.

Activated G_q Interacts with p110 in Vivo—It has been reported that endogenous G_q/11 forms a complex with p110 PI3K (17, 36). We wondered whether G_q(Q209L) might directly bind to and inhibit PI3K. To test for binding, COS-7 cells were cotransfected with FLAG-p110 and p85 in the presence or absence of HA-G_q(Q209L). Equal amounts cell lysate protein were immunoprecipitated with anti-HA antibody, and the immunoprecipitates were examined on a Western blot probed with antibody to the FLAG epitope. A significant amount of FLAG-p110 was detected in the HA-G_q(Q209L) immunoprecipitate (Fig. 6A, upper panel). Conversely, cell lysate proteins were immunoprecipitated with anti-FLAG antibody, and the immunoprecipitates were examined on a Western blot probed with antibody to the HA epitope. HA-G_q(Q209L) was found to coprecipitate with FLAG-p110 (Fig. 6B, upper panel). As expected, p85 was associated with FLAG-p110 in both HA and FLAG immunoprecipitates (Fig. 6, A and B, lower panels). In addition, the PI3K activity measured in FLAG immunoprecipitates from cells expressing HA-G_q(Q209L) was only 44% of that recovered from control cells (Fig. 6C).

View larger version (29K): [in this window] [in a new window]   FIG. 6. Gq(Q209L) co-immunoprecipitates with p110 A–C, COS-7 cells were transfected with FLAG-p110 and p85 in the presence or absence of HA-Gq(Q209L) or empty vector as a control. A, HA immunoprecipitates (IP) and total cell lysates were examined on a Western blot probed with antibodies to FLAG (upper panel) and p85 (lower panel). B, FLAG immunoprecipitates and total cell lysates were analyzed on a Western blot probed with antibodies to HA (upper panel) and p85 (lower panel). The experiment was repeated three times with similar results. C, shown is an autoradiogram illustrating PI3K activity measured in FLAG immunoprecipitates (left panel). The activity in cells expressing HA-Gq(Q209L) was 44% of the activity in control cells (mean activities from duplicate assays were 48,953 and 111,194 cpm, respectively). FLAG immunoprecipitates were probed with anti-FLAG antibody to show that the two assays contained similar amounts of p110 protein (right panel). D, highly purified recombinant p110-p85 complex (see "Experimental Procedures") was mixed with HA immunoprecipitates from COS-7 cells transfected with HA-Gq(Q209L) or empty vector. Mixtures were then assayed for PI3K activity. To make the HA immunoprecipitates, COS-7 cells were seeded at 7.5 x 105 cells/10-cm plate and transfected with 8 µg of HA-Gq(Q209L) or empty vector the next day. Two days later, cell lysate proteins were immunoprecipitated with anti-HA antibody as described under "Experimental Procedures." The immunoprecipitates were washed twice with lysis buffer and twice with PI3K assay buffer (26). Washed immunoprecipitates were suspended in 40 µl of PI3K assay buffer containing 1 mM GTPS, 10 mM MgCl2 and 6 ng of purified PI3K. The mixtures were rotated at 4 °C overnight. The next day, PI3K assays were started by adding 5 µl of 10 mg/ml PI and 5 µl of 400 µM ATP, 2.5 µCi of [-32P]ATP, and 100 mM MgCl2 in PI3K assay buffer. Decreased PI3K activity in the presence of immunoprecipitated HA-Gq(Q209L) was seen in three experiments. An autoradiogram of a representative result is shown (left panel). In a parallel experiment, mixtures that had been rotated at 4 °C overnight were probed with anti-p110 antibody to show that the two assays contained similar amounts of p110 protein (right panel).

We next tested whether HA-G_q(Q209L) inhibits PI3K in vitro. HA immunoprecipitates were prepared from COS-7 cells transfected with either HA-G_q(Q209L) or empty vector as a control. The immunoprecipitates were incubated overnight with highly purified recombinant p110-p85 complex prior to performing PI3K assays. The results show that PI3K activity in the presence of HA-G_q(Q209L) was lower than in the control reaction (Fig. 6D). Together, these results suggest that activated G_q might interact with p110, either directly or indirectly through p85, to inhibit its activity.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
In this study, we investigated whether the inhibitory effect of G_q-coupled receptors on the PI3K/Akt signaling pathway might be mediated by the activated G_q subunit inhibiting PI3K. We found that expression of G_q(Q209L) inhibited Akt activation induced by growth factors and constitutively active PI3K, but did not inhibit the activity of a constitutively active Akt mutant. We also demonstrated that expression of the activated G_q mutant resulted in decreased activity of PI3K in p110 (but not p110) immunoprecipitates. Furthermore, G_q(Q209L) bound to the p110-p85 PI3K heterodimer and appeared to inhibit its activity in vitro, suggesting that there might be an inhibitory interaction between p110 and activated G_q.

As mentioned in the Introduction, several reports indicate that some G_q-coupled receptors can antagonize PI3K/Akt signaling. Because G protein-coupled receptors such as the _1A-adrenergic receptor can couple to more than one type of G subunit, assessing the involvement of a particular G protein can be difficult. Therefore, transfection studies using active G subunit mutants have been done to examine what effect the individual proteins have on Akt. In one report using COS-7 cells, transfection of activated mutants of G_q or G_i caused an increase in Akt activity, whereas G_s and G₁₂ had no effect (37). However, in a separate study, expression of activated mutants of G_s, G_i, G₁₂, or G_q in COS-7 or HEK 293 cells did not activate Akt (11). Similarly, G_s, G_i, G₁₂, G₁₃, or G_q mutants did not activate Akt in NIH 3T3 cells (38). In agreement with the latter two reports, we found that G_q(Q209L) did not promote an increase in Akt activity in two different Rat-1 cell lines (Fig. 1).

A limited number of similar experiments have been performed to investigate the inhibitory effects of G proteins on Akt activation. Bommakanti et al. (11) found that expression of G_q(Q209L) inhibited Akt activation induced by coexpression of G subunits or activated Ras in COS-7 cells. In addition, the activated mutant of G_q (but not of G_s, G_i, or G₁₂) inhibited Akt activation induced by insulin-like growth factor I in HEK 293 cells. Here, we found that G_q(Q209L) inhibited the activation of Akt promoted by PDGF or insulin treatment of Rat-1 cells as well as by expression of myr-p110 in HEK 293 cells (Figs. 1 and 2). Taken together, our results and those of Bommakanti et al. (11) indicate that G_q(Q209L) antagonizes Akt activation induced by a surprisingly diverse set of stimuli. One explanation for this general effect could be that G_q(Q209L) directly inhibits Akt. However, we found that the activity of constitutively active AktDD was not affected by G_q(Q209L) (Fig. 2A). In addition, because the inhibitory effect of G_q(Q209L) on Akt activity was accompanied by decreased phosphorylation of a critical regulatory site (Figs. 1B and 2B), we thought it was more likely that G_q(Q209L) might inhibit a kinase or activate a phosphatase that controls Akt phosphorylation.

We showed previously that growth factor activation of PI3K in Rat-1 cells is antagonized by stimulation of the _1A-adrenergic receptor (26). Those results led us to suspect that G_q(Q209L) might inhibit Akt activation by blocking the activity of its upstream regulator, PI3K. However, because PI3K activity in our previous study was measured in phosphotyrosine immunoprecipitates, an alternative explanation for our results could be that G_q(Q209L) interferes with protein tyrosine phosphorylation. To distinguish between these possibilities, in this study, we examined the effect of activated G_q on PI3K activity in p110 and p110 immunoprecipitates. We found that expression of G_q(Q209L) in an HEK 293 cell line caused a 61% reduction in p110 activity, but no change in p110 activity (Fig. 3). Similarly, release of GTP-bound G_q upon stimulation of the _1A-adrenergic receptor in Rat-1 cells caused a 53% decrease in p110 activity, but had no effect on p110 (Fig. 4A). The reduction in p110 activity in Rat-1 cells was accompanied by a 64% decrease in PI(3,4,5)P₃ production (Fig. 4B) and blunted Akt activation (26). Thus, activated G_q exerts a stable inhibitory effect on p110 that can be measured even after immunoprecipitation of the enzyme. We were surprised that activated G_q did not appear to have such an effect on p110 since Bommakanti et al. (11) found that G_q(Q209L) inhibited Akt activation induced by expression of G subunits in COS-7 cells. Activation of Akt by G in these cells is presumably mediated by the p110 isoform of PI3K, as COS-7 cells express p110, but not p110 (38). Thus, if activated G_q does inhibit p110, it may either bind weakly to this PI3K or use a distinct mechanism that does not allow us to measure a stable change in activity using our in vitro assay.

Our data suggest that inhibition of PI3K by activated G_q is not mediated by the phospholipase C signaling pathway. Activated G_q still exerted an inhibitory effect on Akt and p110-p85 in the presence of a phospholipase C inhibitor (Fig. 5) and in cells depleted of protein kinase C (data not shown) or intracellular Ca²⁺ (26). G_q(Q209L) stimulation of GLUT4 translocation in 3T3-L1 adipocytes is also thought to be independent of phospholipase C signaling (39). We therefore postulated that activated G_q might bind to PI3K and directly inhibit its catalytic activity. In support of this hypothesis, we found that the p110-p85 heterodimer coprecipitated with G_q(Q209L) (Fig. 6, A and B). In addition, the PI3K activity measured in these immunoprecipitates was lower than in control reactions that did not contain G_q(Q209L) (Fig. 6C). Finally, incubation of immunoprecipitated G_q(Q209L) with p110-p85 in vitro also led to a decrease in PI3K activity (Fig. 6D). More studies are needed to determine the nature and relevance of the physical interaction between G_q and p110.

In direct contrast to our hypothesis, Olefsky and co-workers (17) have proposed that stimulation of the G_q-coupled ET_A receptor in 3T3-L1 adipocytes causes G_q to activate p110 PI3K. Acute treatment of these cells with endothelin-1 increased p110 activity almost as much as insulin did, but there was no corresponding increase in Akt phosphorylation. Stimulation with endothelin-1 also increased the amount of endogenous G_q/11 that coprecipitated with p110 (17). Similarly, adenoviral expression of G_q(Q209L) increased the PI3K activity in p110 and p110 immunoprecipitates, and yet Akt phosphorylation did not increase. Unlike our findings, G_q(Q209L) modestly inhibited insulin activation of p110 and Akt, whereas PDGF-induced activation of the two enzymes was not affected (36). The differences between our results and those cited above could be due to cell type variation. For example, 3T3-L1 adipocytes might contain unique proteins that join the G_q-p110 complex to cause a stimulatory effect, or a posttranslational modification of G_q might cause the interaction with p110 to be stimulatory rather than inhibitory. Indeed, it was reported that tyrosine phosphorylation of G_q/11 increases following treatment of 3T3-L1 adipocytes with endothelin-1 or insulin (17, 36). On the other hand, some of the data in Refs. 17 and 36 are controversial, as Pessin and co-workers (39) believe that activation of G_q in 3T3-L1 adipocytes does not lead to an increase in PI3K activity.

Activation of G_q can have profound effects on cell survival. Simon and co-workers (40) reported that expression of the constitutively active G_q(R183C) mutant in COS-7 or Chinese hamster ovary cells promotes apoptosis. We used the inducible Flp-In T-REx system to study G_q(Q209L) (Fig. 3) because we were unable to isolate cell clones stably expressing the active mutant in a constitutive manner.² In addition, we found that _1A-adrenergic receptor activation of G_q promotes apoptosis in Rat-1 cells (26). Finally, transgenic mice overexpressing G_q were found to die of heart failure, and cardiomyocytes from these animals showed an increased rate of apoptosis (41). We speculate that direct inhibition of p110 by active G_q could be a common mechanism used by the _1A-adrenergic receptor and other G_q-coupled receptors to inhibit Akt activity and to promote apoptosis. This prediction could be tested by attempting to reverse the apoptotic effect of G_q(Q209L) by expression of constitutively active Akt or a p110 mutant that no longer binds to G_q.

In conclusion, our results demonstrate that activated G_q modulates Akt activity by inhibiting PI3K. Further investigation into how G_q-coupled receptors inhibit this signaling pathway might reveal important physiological insights into diseases such as diabetes mellitus, heart failure, and cancer.

	FOOTNOTES

 
^* This work was supported by the American Federation for Aging Research Paul Beeson Physician Scholar Award and National Institutes of Health Grant R01 DK62722 (to R. Z. L.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Present address: Inst. of Genetics, Roemerstr. 164, University of Bonn, 53117 Bonn, Germany.

^|| To whom correspondence should be addressed: Dept. of Medicine, Div. of Hematology, zip = 8151, Stony Brook University, Stony Brook, NY 11794. Tel.: 631-444-2059; Fax: 631-444-7530; E-mail: richard.lin{at}sunysb.edu.

¹ The abbreviations used are: PI3K, phosphatidylinositol 3-kinase; PI, phosphatidylinositol; PE, phenylephrine; PDGF, platelet-derived growth factor; GTPS, guanosine 5'-O-(3-thiotriphosphate); HA, hemagglutinin; myr-, myristoylated; HEK, human embryonic kidney; HPLC, high pressure liquid chromatography; PIPES, 1,4-piperazinediethanesulfonic acid; HIR, human insulin receptor.

² L. M. Ballou, H.-Y. Lin, G. Fan, Y.-P. Jiang, and R. Z. Lin, unpublished data.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 

1. Katso, R., Okkenhaug, K., Ahmadi, K., White, S., Timms, J., and Waterfield, M. D. (2001) Annu. Rev. Cell Dev. Biol. 17, 615–675[CrossRef][Medline] [Order article via Infotrieve]
2. Roche, S., Downward, J., Raynal, P., and Courtneidge, S. A. (1998) Mol. Cell. Biol. 18, 7119–7129[Abstract/Free Full Text]
3. Alessi, D. R., and Cohen, P. (1998) Curr. Opin. Genet. Dev. 8, 55–62[CrossRef][Medline] [Order article via Infotrieve]
4. Sable, C. L., Filippa, N., Hemmings, B., and Van Obberghen, E. (1997) FEBS Lett. 409, 253–257[CrossRef][Medline] [Order article via Infotrieve]
5. Yano, S., Tokumitsu, H., and Soderling, T. R. (1998) Nature 396, 584–587[CrossRef][Medline] [Order article via Infotrieve]
6. Stoyanov, B., Volinia, S., Hanck, T., Rubio, I., Loubtchenkov, M., Malek, D., Stoyanova, S., Vanhaesebroeck, B., Dhand, R., Nurnberg, B., Gierschik, P., Seedorf, K., Hsuan, J. J., Waterfield, M. D., and Wetzker, R. (1995) Science 269, 690–693[Abstract/Free Full Text]
7. Stephens, L. R., Eguinoa, A., Erdjument-Bromage, H., Lui, M., Cooke, F., Coadwell, J., Smrcka, A. S., Thelen, M., Cadwallader, K., Tempst, P., and Hawkins, P. T. (1997) Cell 89, 105–114[CrossRef][Medline] [Order article via Infotrieve]
8. Kurosu, H., Maehama, T., Okada, T., Yamamoto, T., Hoshino, S., Fukui, Y., Ui, M., Hazeki, O., and Katada, T. (1997) J. Biol. Chem. 272, 24252–24256[Abstract/Free Full Text]
9. Tilton, B., Andjelkovic, M., Didichenko, S. A., Hemmings, B. A., and Thelen, M. (1997) J. Biol. Chem. 272, 28096–28101[Abstract/Free Full Text]
10. Franke, T. F., Kaplan, D. R., Cantley, L. C., and Toker, A. (1997) Science 275, 665–668[Abstract/Free Full Text]
11. Bommakanti, R. K., Vinayak, S., and Simonds, W. F. (2000) J. Biol. Chem. 275, 38870–38876[Abstract/Free Full Text]
12. Moule, S. K., Welsh, G. I., Edgell, N. J., Foulstone, E. J., Proud, C. G., and Denton, R. M. (1997) J. Biol. Chem. 272, 7713–7719[Abstract/Free Full Text]
13. Kowalski-Chauvel, A., Pradayrol, L., Vaysse, N., and Seva, C. (1996) J. Biol. Chem. 271, 26356–26361[Abstract/Free Full Text]
14. Saward, L., and Zahradka, P. (1997) Circ. Res. 81, 249–257[Abstract/Free Full Text]
15. Graness, A., Adomeit, A., Heinze, R., Wetzker, R., and Liebmann, C. (1998) J. Biol. Chem. 273, 32016–32022[Abstract/Free Full Text]
16. Takahashi, T., Taniguchi, T., Konishi, H., Kikkawa, U., Ishikawa, Y., and Yokoyama, M. (1999) Am. J. Physiol. 276, H1927–H1934[Medline] [Order article via Infotrieve]
17. Imamura, T., Ishibashi, K., Dalle, S., Ugi, S., and Olefsky, J. M. (1999) J. Biol. Chem. 274, 33691–33695[Abstract/Free Full Text]
18. Eguchi, S., Iwasaki, H., Ueno, H., Frank, G. D., Motley, E. D., Eguchi, K., Marumo, F., Hirata, Y., and Inagami, T. (1999) J. Biol. Chem. 274, 36843–36851[Abstract/Free Full Text]
19. Tang, X., Batty, I. H., and Downes, C. P. (2002) J. Biol. Chem. 277, 338–344[Abstract/Free Full Text]
20. Hu, Z. W., Shi, X. Y., Lin, R. Z., and Hoffman, B. B. (1996) J. Biol. Chem. 271, 8977–8982[Abstract/Free Full Text]
21. Batty, I. H., and Downes, C. P. (1996) Biochem. J. 317, 347–351[Medline] [Order article via Infotrieve]
22. Velloso, L. A., Folli, F., Sun, X. J., White, M. F., Saad, M. J., and Kahn, C. R. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 12490–12495[Abstract/Free Full Text]
23. Folli, F., Kahn, C. R., Hansen, H., Bouchie, J. L., and Feener, E. P. (1997) J. Clin. Invest. 100, 2158–2169[Medline] [Order article via Infotrieve]
24. Hajduch, E., Rencurel, F., Balendran, A., Batty, I. H., Downes, C. P., and Hundal, H. S. (1999) J. Biol. Chem. 274, 13563–13568