In Vitro Association of Phosphatidylinositol 3-Kinase Activity with the

1 downloads 0 Views 3MB Size Report
Jan 5, 2016 - U.S.C. Section 1734 solely to indicate this fact. .... CHO cells overexpressing wild-type (Ehina-type) (28) human IRs. (CHO. ..... 9, 1651-1658. 4.
THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1992 by The American Society for Biochemistry and Molecular Biology, h e .

Vol. 267, No. 1, Issue of January 5, pp. 440-446,1992 Printed in U.S.A.

In Vitro Association of Phosphatidylinositol 3-Kinase Activity with the Activated Insulin ReceptorTyrosine Kinase* (Received for publication, August 12, 1991)

Kazuyoshi YonezawaS, Koichi YokonoS, Kozui Shiis, WataruOgawaS, Akifumi Ando$, Kenta HaraS, Shigeaki BabaQ, Yasushi Kaburagill, RitsukoYamamoto-Hondall, Kaoru Momomurall, Takashi Kadowakill, and Masato KasugaSII From the $Second Department of Internal Medicine, Kobe University School of Medicine, 7-5-1 Kusurwki-cho, Chuo-ku, Kobe 650, Japan, the SHyogo Institute of Clinical Research, 13-70 Kitaoji-cho, Akashi 673, Japan, and the YThird Department of Internal Medicine, Faculty of Medicine, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113, Japan

We previously have shown that insulin treatment of novel phosphatidylinositol (PI)’ kinase that phosphorylates cells greatly increases the activity of phosphatidyli- the D-3position of the inositol ring to produce phosphatidylnositol (PI)3-kinase in immunoprecipitates made withinositol 3-phosphate (PI-3-P). This enzyme, phosphatidylian antibody to phosphotyrosine. However, the associ- nositol 3-kinase (PI 3-kinase), was shown to associate physically with p60””“ (3), polyoma middle T antigen (MTAg)ation of PI 3-kinase activity with the activated insulin p60”-”’“complexes (4, 5), the platelet-derived growth factor receptor is not significant under these conditions. In the present study, we have attempted to reconsti- (PDGF) receptor (6,7), thecolony-stimulating factor-1 (CSF1)receptor (8),~ 6 8 ~ ~ ~ ”p130gag-fp” ~ * ( 9 ) ,(9),p47gag-crk (9), MTAgtute the association of PI 3-kinase activity with the complexes (9), and the epidermal growth factor (EGF) activated insulin receptor in vitro.PI 3-kinase activity p62C-yeS does indeed associate with the autophosphorylated in- receptor (10). The enzyme was also shown to become tyrosinephosphorylated upon MTAg transformation ( l l ) , PDGF sulin receptor in our in vitro system. The autophos- stimulation (1, 6, l l ) , or CSF-1 stimulation of cells (8). In phorylation of the insulin receptor and/or its associated addition to phosphorylation of PI, this enzyme phosphorylconformational change appear to be necessary for the ates phosphatidylinositol 4-phosphate(PI-4-P) andphosphaassociation of PI 3-kinase activity with the receptor, tidylinositol 4,5-bisphosphate (PIP2) to produce the novel since kinase negative receptor failed to bind PI 3- phosphoinositides, phosphatidylinositol 3,4-bisphosphate kinase activity. After binding,PI 3-kinase or itsasso- (PI-3,4-P2)and phosphatidylinositol triphosphate (PIP3, ciated protein seems to be released from the activatedprobably phosphatidylinositol 3,4,5-triphosphate)(1).Recent receptor after the completion of its tyrosine phos- evidence suggests that the levels of these polyphosphoinosiphorylation by the receptor. Tyrseo in the juxtamem- tides with a phosphate at the D-3 position of the inositol ring brane region of the insulin receptor &subunit seems toare increased after stimulation of the cell with PDGF or CSFbe involved in the association of PI 3-kinase activity 1 (1, 8). The function of these polyphosphoinositides in cell with the receptor, butnot C terminus region of the 8- activation and proliferation remains unclear. Ithas been reported that these products appearto be resistant tohydrolsubunitincluding twotyrosineautophosphorylation sites (Tyr131sand T Y ~ ’ ~ ~The ’ ) . in vitro assay system ysis by phospholipase C-7, a phospholipase C capable of liberating inositol 1,4,5-triphosphate and diacylglycerol from for the association of PI 3-kinase activity with the PIPz in the “classic” PI turnover pathway (12, 13). insulin receptor can be utilized to study mechanism the Recently, PI 3-kinase has been purified to near homogeof interaction of these molecules and will bean useful method to detect other associated molecules with the neity, showing that the kinase contains two major subunits of 85 and 110 kDa (14, 15). cDNA cloning of the 85-kDa insulin receptor. protein has revealed that the protein contains one SH3 and two SH2 regions (src homology regions) (16-18). When expressed, the 85-kDa protein binds to and is a substrate for tyrosine-phosphorylated PDGF and EGF receptor kinases



Whitman and co-workers (1, 2) recently have identified a

* This work was supported by a grant-in-aid for Cancer Research from the Ministry of Education, Science and Culture of Japan, by a grant from the Cell Science Research Foundation, and by a grant for Diabetes Research from Ohtsuka Pharmaceutical Co., Ltd. (to M. K.) and by Research Grant 190831 from Juvenile Diabetes Foundation International (to T. K.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 11 To whom correspondence should be addressed: The Second Department of Internal Medicine, Kobe University School of Medicine, 7-5-1, Kusunoki-cho, Chuo-ku, Kobe 650, Japan. Tel.: 078-341-7451 (ext. 5520); Fax: 078-382-2080.

The abbreviations used are: PI, phosphatidylinositol; MIgG, normalmouse immunoglobulins; Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid; PIPES, 1,4-piperazinediethanesulfonicacid; IR(A1018), in which LYS’~’’ was substituted with arginine; IR(F960) and IR(A960), in which TyrSmwas substituted with phenylalanine or alanine, respectively; IR(F1316, 1322), in which Tyrl3l6.132z were substituted with phenylalanine; IR(ACT36), in which the last 36 amino acids at theC terminus of the &subunit were deleted; PI-3-P, phosphatidylinositol 3-phosphate; MTAg, middle T antigen; PDGF, platelet-derived growth factor; CSF-1, colony-stimulating factor-1; EGF, epidermal growth factor; PI-4-P, phosphatidylinositol 4-phosphate; PIPz, phosphatidylinositol 4,5-bisphosphate; pI-3,4-P~,phosphatidylinositol3,4-bisphosphate;PIP3, phosphatidylinositol triphosphate; CHO, Chinese hamster ovary. In thisreport, the term “in viuo” refers to intact cells, whereas the term “in uitro” refers to a cell-free system. The numbering of amino acids in this paper corresponds to the sequence of Ullrich et al. (32).

440

In Vitro

Association of PI 3-Kinase Activity with Insulin

and thepolyoma virus MTAg-/pp6OC"'" complex but lacks PI 3-kinase activity (16, 18).The 110-kDa protein appears to be the catalytic subunit of the PI 3-kinase, and the 85-kDa protein appears to be the subunit that links PI 3-kinase to the ligand-activated receptor. The insulin receptor (IR) belongs to thefamily of structurally related transmembrane growth factor receptors with ligand-activated protein tyrosine kinase activity (19, 20). Several lines of evidence suggest that this kinase activity is required for normal insulin action (20, 21). Anumber of proteins (with molecular masses ranging from 15 to 250 kDa) have been reported to be tyrosine-phosphorylated in an insulin-dependent fashion (22,23). Recently, insulin treatment of cells has been found to increase the tyrosine phosphorylation of PI 3-kinase, suggesting that PI3-kinase is a substrate for the insulin receptor tyrosine kinase (24,25). Insulin treatment of cells was also found to stimulate the formation of polyphosphoinositides, such as PI-3,4-P2and PIP3 (25).However, unlike the othertyrosine kinases, only a small percentage (1-395)of the tyrosine-phosphorylated PI 3-kinase becomes associated with the insulin receptor (24-26). The treatment of insulin-stimulated intact cells with bifunctional cross-linkers has been found to cause a significant increase in the amount of PI 3-kinase activityassociated with the insulin receptor (26). In thepresent study,we have attempted toreconstitute the association of PI 3-kinase activity with the activated insulin in vivo study, P I 3-kinase receptor i n vitro. In contrast to the activity does associate with the autophosphorylated insulin receptor in our i n vitro assay system. Using this in vitro assay system, we analyzed the mechanism of interaction between P I 3-kinase activity and the insulin receptor in this paper. This assay system could be a method useful to detect other molecules associated with the activated insulin receptor in insulin signal transduction. EXPERIMENTALPROCEDURES

Materials-Nonidet P-40 was purchased from Sigma. Phosphatidylinositol (bovine liver) was purchased from Avanti Polar Lipid, Inc. (Birmingham AL), Protein G-Sepharose from Pharmacia, pork insulin from Novo (Bagsvaerd, Denmark), affinity-purified mouse IgG (MIgG) from Cappel (West Chester, PA), [y-"P] ATP from ICN, bifunctional cross-linkers disuccinimidyl suberate and dithiohis sulfosuccinimidylpropionate from Pierce. The monoclonal antibody py20, which specifically recognizes phosphotyrosine residues (27), was purchased from ICN. The monoclonal antibodies to the human insulin receptor, 5D9 and 29B4, were a giftfrom Dr. Richard A. Roth, at the Stanford University, Stanford, CA, and 2F3 and 3 B l l were developed by Kozui Shii andwill be described elsewhere. Transfected CHO cells overexpressing wild-type (Ehina-type) (28) human IRs (CHO. T) was as described previously (29). Preparation of Transfected CHO Cells Ouerexpressing Wild-type (Ullrich-type) and Mutant Insulin Receptors-Mutant human insulin receptors (IR(RlOl8),2 IR(F960),3 IR(A960): IR(F1316, 1322); and IR(ACT36)4) were constructed by a method using the polymerase chain reaction as described previously (30, 31).Wild-type human insulin receptor (Ullrich-type) (32) or mutantinsulin receptorcDNAs were ligated into SV40 expression vector, and CHO cells were transfected by the calcium phosphate precipitation method (33). IR(ACT36) was constructed using cDNA of Ebina-type human insulin receptor. Other mutants were constructed using cDNA ofU11rich-type human insulin receptor. Details will be described else~here.2.~ In Vitro Association of PI 3-kinase Activity with Insulin Receptor-

' A. Ando, K. Momomura, Y. Kaburagi, R. Yamamoto-Honda, T. Kadowaki, and M. Kasuga, manuscript in preparation. Y.Kaburagi, K. Momomura, R. Yamamoto-Honda, M. Kasuga, and T.Kadowaki, manuscript in peparation. R. Yamamoto-Honda, K. Momomura, T. Kadowaki, and M. Kasuga, manuscript in preparation.

Receptor

44 1

Confluent 100-mm plates of either CHO. T cells, CHO cells overexpressing wild-type (Ullrich-type)or mutant receptors were starved in Ham's F-12 medium containing 20 mM Hepes (pH 7.6), for 12 h at 37 "C. Cells were then lysed, and the lysates from one-half of a plate were immunoprecipitatedwith10 pg of either control MIgG or monoclonal anti-insulin receptor antibody (either 5D9,29B4,2F3, or 3 B l l ) bound toProtein G-Sepharose (20-p1 beads) as described previously (24). Immunoprecipitates were washed twice with 50 mM Hepes (pH 7.6) buffered saline containing 0.1% Triton X-100 (buffer A). To phosphorylate the insulin receptor on the beads, the immunoprecipitate was mixed with 1p~ insulin, 1 mM ATP, 10 mM M e , 3 mM Mn'+, and 20 pg/ml aprotinin in buffer A. After incubating at room temperature for 30 min, the immunoprecipitates were washed three times with buffer A. Lysates for association of PI 3-kinase activity in uitro, were prepared from confluent, serum-starved cultures of parental CHO cells (100 mm dish). Cultures were lysed in 0.8 ml of 150 mM NaCI, 10 mM PIPES (pH 7.0), 0.1% NonidetP-40, 20 pg/ml aprotinin, 2 mM sodium vanadate, 1 mM phenylmethylsulfonyl fluoride (buffer B). The insoluble debris were removed by centrifugation at 15,000 X g for 10 min at 4 "C, and the cleared supernatant was used in the in uitro association assay. To associate PI 3-kinase activity with the insulin receptor, the autophosphorylatedinsulinreceptor (from one-half of a 100-mm plate) on the beads were exposed t o 400 pl of lysate from parental CHO cells. After 12 ha t 4 'C, during which time the tubes were mixed on a rotating platform, the immunoprecipitates were washed and PI 3-kinase activity was determined as described previously (24). To detect the tyrosine phosphorylation of PI 3-kinase or its associated protein by insulin receptor in uitro, the autophosphorylated insulin receptor on the beads were exposed to CHO cell lysates in the presence of 1 mM ATP, 10 mM M e , 3 mM MnZ+for 12 h at 4 "c. After the incubation, the lysates were removed and immunoprecipitated with 2 pg of monoclonal anti-phosphotyrosine antibody (py20) for 4 h at 4 'C. The anti-phosphotyrosine immunoprecipitates were washed and PI 3-kinase activity was measured as described previously (24). To examine the dissociation of PI 3-kinaseactivity from the activated insulin receptor, in vitro association of P I 3-kinase activity with the autophosphorylated insulin receptor was carried out on beads (20 p l ) as described above. After removing CHO cell lysates, the beads were washed twice with 1 ml of buffer B and then incubated with 80 pl of buffer B or the buffer B containing either M e and Mn" (final concentrations are10 and 3 mM, respectively), ATP alone(final concentrationis 1 mM), or both M g + , Mn'+, and ATP at room temperature for the indicated period of times. Then the beads were washed and PI 3-kinaseactivitiesremaining on the beads were determined as described previously (24). RESULTS

In our previous i n vivo study (24,26), a significant amount of PI 3-kinase activity associated with the insulin-activated receptor was not immunoprecipitated with monoclonal antiinsulin receptor antibodies (5D9, 29B4). However, the addition of bifunctional cross-linkers to the insulin-stimulated intact CHO .T cells caused the significant increase in the amount of PI 3-kinase activity associated with the activated insulin receptor i n vivo (26). To reconstitute the association of these proteins i n vitro, human insulin receptors were immunoprecipitated with monoclonal anti-insulin receptor antibodies from unstimulated C H 0 . T cells, incubated in the presence or absence of nonradioactive ATP, washed, and exposed to lysates of parental CHO cells not overexpressing the human insulin receptors in the presence or absence of cross-linkers a t 4 "C. After the incubation, nonassociated proteins were washed out, and receptor-associated PI 3-kinase activity was determined. PI 3-kinase activity bound to the insulin receptor even in the absence of cross-linkers, but only if the, receptor had been incubated with ATP before exposure to the cell lysate (Fig. l.4). Fig. 1B shows a typical autoradiogram of the thin layer chromatogram of PI 3-kinase activity associated with the non- or autophosphorylated insulin receptor i n vitro in the

In VitroAssociation of P I 3-Kinase Activity with Insulin Receptor

442

e

E-

.-x .->

-

2

2 2 .-

500 400

300 200 100

ii: 0

aulcphosphwylalion :

- +

-

crosslinker :

+

DSS

-

+

DTSSP

NP40

( R)

:

-

+

i-

i-

-

0.2

0.5

FIG. 2. Effect of detergent on PI 3-kinase activity associated with autophosphorylated insulin receptor.I n vitro association of PI 3-kinase activitywas carried out asdescribed in thelegend to Fig. 1. Then PI 3-kinase assaywas done in theabsence or presence of Nonidet P-40( N P 4 0 ) (final concentration, 0.2 or 0.5%). Shown are amounts of PI 3-kinase activities in 3 B l l immunoprecipitates from which PI 3-kinase activities in control MIgG immunoprecipitates were subtracted.

Auto-

phosphorylation

autophosphorylation :

'

- +

FIG. 1. In vitro association of PI 3-kinase activity with the insulin receptor. CH0.T cellswere lysed, and the lysates were immunoprecipitated with 10 pg of either controlMIgG or monoclonal anti-insulin receptor antibody 3Bll. Then the immunoprecipitates were mixed with 10 mM M F , 3 mM Mn2+, and20 pg/ml aprotinin in theabsence (open column) or thepresence (closed column) of 1 p M insulin and1 mM ATP for 30min a t room temperature. After washing, the immunoprecipitates were exposed to parental CHO cell lysates for 12 h a t 4 "C in the absence or presence of either 75 pg/ml of disuccinimidyl suberate ( D S S ) or 75 pg/ml of dithiobios sulfosuccinimidyl propionate (DTSSP),washed, and the PI 3-kinase activities were determined. A, data shown are PI 3-kinase activities in antiinsulin receptor immunoprecipitates from which PI 3-kinase activities in MIgG immunoprecipitates were subtracted. B, shownis a typical autoradiogram of thin layer chromatogram of PI 3-kinase activity associated with non- or autophosphorylated insulin receptor in vitro in the absenceof cross-linkers.

RlOl8

WI1.D

FIG. 3. In vitro association of PI 3-kinase activity with the kinase negative insulin receptor. The insulin receptors fromCHO cells overexpressing wild-type (Ullrich-type) human insulin receptor or IR(R1018),which has a putative ATP binding siteLys'OLRreplaced with arginine, were immunoprecipitated with control MIgG or monoabsence of cross-linkers. The binding of P I 3-kinase activity clonal anti-insulin receptor antibody 5D9, then incubated with 10 was detected to the autophosphorylated insulin receptor pre-mM M F , 3 mM Mn2+, and20 pg/ml aprotinin in the absence (open cipitated by each of two monoclonal antibodies (5D9, 3 B l l ) column) or the presence (closed column) of 1 pM insulin and 1 mM ATP for 30 min a t room temperature. The assay for the in vitro to a-subunit of theinsulinreceptorand twomonoclonal antibodies (29B4, 2F3) to the intracellular epitope of @-sub- association of PI 3-kinaseactivity with the insulinreceptor was carried out as described in the legend to Fig. 1. Results are PI 3unit (data not shown). kinase activities in the 5D9 immunoprecipitates from which PI 3T o confirm that the enzymatic activity associated with thekinase activities in control MIgG immunoprecipitates were subtracted insulin receptor is due to the PI 3-kinase activity, theeffect and havebeennormalized to 100% for the PI 3-kinase activity of nonionic detergent on the associated activity was examined, associated with the autophosphorylatedwild-type receptor.

since the PI 3-kinase activity is inhibited completely by the presence of more than 0.2% Nonidet P-40 (24). The addition theinsulinreceptorand/orits associated conformational of Nonidet P-40 to the reaction mixture caused a complete change arenecessary for theassociation of PI 3-kinaseactivity with the insulinreceptor. inhibition of the activity to the basal activity associated with the nonautophosphorylated receptor(Fig. 2), indicating that As shown in previous study (26), PI 3-kinase activity was the associated activity is due to PI 3-kinase. detectedinanti-phosphotyrosineimmunoprecipitates reTo test whether the autophosphorylation of the insulin covered from insulin-stimulated cell cytosol. To reproduce receptor and/or itsassociated conformational change arenec- this in uitro, the effect of ATP, M e , and Mn2' during the essary for the association of P I 3-kinase activity,we employed incubation of the activated insulin receptor with CHO cell a mutant receptor, IR(R1018), which has a ATP binding site lysates was examined on the appearance of PI 3-kinase activLys1OI8 replaced with arginine. This mutant previously has ityinanti-phosphotyrosineimmunoprecipitates recovered been shown to be devoid of kinase activity and be to incapable from cell lysates. The insulin receptor was immunoprecipiof mediating biological responses (34). This kinase negative tated with anti-insulin receptor antibodies on the beads, aureceptor failed to bind PI 3-kinase activity even if it was tophosphorylated a t room temperature for 30 min in the preincubated with ATP prior to in uitro association (Fig. 3). presence of 1 mM ATP, 10 mM M$+, and 3 mM Mn2', and beads was washed three This is consistent with only the autophosphorylated wild-type this activated insulin receptor on the receptor being capable of associating P I 3-kinase activity as times with 1 ml buffer B. Then CHO cell lysates were added shown in Fig. 1, suggesting that the autophosphorylation of and incubated in the absence or presence of 1 mM ATP, 10

443

I n VitroAssociation of P I 3-Kinase Activity with Insulin Receptor mM M$+, and 3 mM Mn2+a t 4 “C for 12 h. Finally, the lysates were separated from the beads and immunoprecipitated with anti-phosphotyrosine antibody (py20) and then PI 3-kinase activity in the anti-phosphotyrosine immunoprecipitates recovered from the lysates and PI 3-kinase activity associated with the activated receptor on the beads were determined separately. The presence of ATP, Mg+,and Mn2+during the incubation of the activated insulin receptor with CHO cell lysates showed no significant effect on the amount of PI 3kinase activity associated with the activated receptor on the beads compared to the results in the absence of ATP, M$+, and Mn2+ (Fig. 4A), however, did show the increase in the amount of PI 3-kinase activity in the anti-phosphotyrosine immunoprecipitates recovered from the CHOlysates (Fig. 4B). Next, to test whether receptor-associated PI 3-kinase is released from the activated insulin receptor, the following experiment was performed. The insulin receptor was immunoprecipitated on beads, autophosphorylated, andthen washed. CHO cell lysates were added to associate PI 3-kinase activity with the insulin receptor. After incubation for 12 h at 4 “C, lysates were removed, the beads were mildly washed, and thebuffer containing either ATP and/orMf and Mn2+ was added to the beads. The beads were incubated at room temperature for the indicated time, washed, and receptorassociated PI 3-kinase activity was determined on the beads. Incubation with either buffer or 10 mM M P and 3 mM Mn2+ for 10 min showed a 25% decrease of PI 3-kinase activity associated with the activated insulinreceptor compared to the initial associated PI 3-kinase activity a t 0 min, whereas incubation with 1mM ATP for 10 min showed a smallerdecrease (17%) of the associated PI 3-kinaseactivity (Fig. 5).In

2

0

6

4

8

1

0

1

2

Time (min)

FIG. 5. In vitro dissociation of PI 3-kinase activity from the activated insulin receptor. In vitro association of PI 3-kinase activity with the autophosphorylated insulin receptor was carried out on beads (20 p l ) as described in the legend to Fig. 1. After washing out CHO cell lysates, the beads were incubated with 80 pl of either buffer (open circle), M e , and Mn2+(closed circle) (final concentrations are 10 and 3 mM, respectively), ATP alone (open square) (final concentration is 1 mM), or M e , Mn2+,and ATP (closed square) at room temperature for the indicated period of time and then washed and PI 3-kinase activities remained on the beads were determined. Results are PI 3-kinase activities associated with the autophosphorylated insulin receptor from which PI 3-kinase activity associated with nonautophosphorylatedinsulinreceptor was subtractedand have been normalized to 100% for the initial PI 3-kinase activity at 0 min.

I00 80 60 40 20

0 A

autophosphaylslion : additirm of ImM ATP :

- + + - - +

B

C

D

WILD

FIG. 6. In vitro association of PI 3-kinase activity with mutant insulin receptors. The insulin receptors from CHO cells overexpressing wild-type and mutant receptors were immunoprecipitated with control MIgG or monoclonal anti-insulin receptor antibody 5D9, then nonautophosphorylated (open column) or autophosphorylated (closed column). The assay for the in vitro association of PI 3-kinase activity with the insulinreceptor was carried out as described in the legend to Fig. 1. IR(F960) ( A ) , IR(A960) ( E ) , and IR(F1316,1322) (C) compared to thewild-type (Ullrich-type) insulin receptor, whereas IR(ACT36) (D)was compared to the wild-type (Ebina-type) receptor. Results are PI 3-kinase activities in the 5D9 immunoprecipitates from which PI 3-kinase activities in control MIgG immunoprecipitates were subtracted and have been normalized to 100% for the PI 3-kinase activity associated with the autophosphorylated wild-type receptor (either Ullrich or Ebina-type).

FIG. 4. In vitro tyrosine phosphorylation of PI 3-kinase or its associated protein by the insulin receptor. The nonautophosphorylated or autophosphorylated insulin receptor from CHO. T cells were prepared as shown in the legend to Fig. 1. The nonautophosphorylated receptor (shaded column) was exposed to CHO cell lysates without any addition, and the autophosphorylated receptor was exposed to the lysates in the absence (open column) or the presence (closed column) of 1mM ATP, 10mM M$+, and 3 mM Mn2+ for 12 h at 4 “C. PI 3-kinase activities associated with the insulin receptor on the beads were determined as described under “Experi- contrast, incubation with 1mM ATP, 10 mM M e , and 3 mM mental Procedures.” The CHO cell lysates were removed after the Mn2+for 10 min reduced the amount of receptor-associated exposure to thereceptor and immunoprecipitated with 2pg of monoclonal anti-phosphotyrosine antibody (py20) bound to beads for 4 h PI 3-kinase activityto less than 50% of the initial (Fig. 5). Finally, we studied the region of the insulin receptor essenat 4 “C.The anti-phosphotyrosine immunoprecipitates were washed and PI3-kinase activitieswere measured as described under “Exper- tial for the association of PI 3-kinase activity. To this end, imental Procedures.” A , PI 3-kinase activities associated with the we employed four mutant insulin receptors: IR(F960) and insulin receptor. Results are PI 3-kinaseactivities in anti-insulin IR(A960), in which Tyrg6’was substituted with phenylalanine receptor precipitates from which PI 3-kinase activities in control or alanine, respectively, IR(F1316, 1322), in which both tyMIgG were subtracted. B , PI 3-kinase activities in anti-phosphotyrosine immunoprecipitates recovered from CHO lysates. Data shown rosine residues a t 1316 and 1322 weresubstituted with phenylalanine, and IR(ACT36), in which the last 36 amino acids are from which PI 3-kinase activities in anti-phosphotyrosine immunoprecipitates recovered from CHO cell lysates exposed to control at theC terminus of the @-subunitwere deleted. All mutant MIgG immunoprecipitates were subtracted. receptors have normal insulin binding, almost normal insulin-

444

Vitro In Association

of 3-Kinase PI Activity with Insulin Receptor

stimulated autophosphorylation and kinase activity in vitro compared to the wild-type insulin r e ~ e p t o r . ~ Tsame h e receptor number of each mutant receptor or the wild-type receptor was immunoprecipitated on beads with monoclonal anti-insulin receptor antibodies, then preincubatedwith ATP, washed, and exposed to lysates of CHO cells, and receptorassociated PI 3-kinaseactivity was compared. IR(F1316, 1322), in which two tyrosyl autophosphorylation sites of the insulin receptor a t C terminus (35-37) were substituted with phenylalanine, showed the same level of associated PI 3kinaseactivity as the wild-type receptor (Fig. 6, lane D). IR(ACT36), whose truncation removes these two autophosphorylation sites and a threonine phosphorylation site (Thr'"') (38, 39) at C terminus, also showed the same level of associated PI 3-kinase activity as the wild-type receptor (Fig. 6, lane C). Mutations in the juxtamembrane region of the insulin receptor @-subunitby substitution of Tyr"" with phenylalanine or the deletion of 12 amino acidsaround TyrS6O impair normalinsulin-stimulatedtyrosine phosphorylationof pp185 and several biological responses (40,41). IR(F960) and IR(A960) exhibited impaired insulin-stimulated tyrosine phosphorylation of pp 18L6These mutants also showed reduced PI 3-kinase activity associated with the receptor in vitro compared to thewild-type insulin receptor (Fig. 6, lanes A and B ) . DISCUSSION

In the present studies, we established the in vitro assay system in which we could demonstrate the physical association of PI 3-kinase activity with the activated insulin receptor. This is in contrast to thein vivo study in which a significant amount of PI 3-kinase activity associated with the insulinactivated receptor was not immunoprecipitatedwith antiinsulin receptor antibodies ( 2 4 26). In this in vitro assay system, we employed a lower temperature (4 "C) for the association of PI 3-kinase activity with the insulin receptor compared to the intact cell (37 "C). As for the concentration of ATP, during the association of P I 3-kinase activity with the receptor, the concentration in the cytosol in vivo is approximately 100 p M to 1 mM, whereas that in in vitro assay is estimated to be less than 10 nM. Because, after the autophosphorylation reaction in the presence of 1 mM ATP, the activated insulin receptor on beads (20 pl) was washed out 3 times with 1 ml of buffer, it islikely that these lower temperatures and lower concentrations of ATP in in vitro assays reduce the kinase reaction by the insulin receptor and keep the PI 3-kinaseactivity being associatedwith the insulin receptor. We also demonstrated that the addition of 1 mM ATP, 10 mMMg", and 3 mM MnZ+during the incubation of the activated insulin receptor with CHO cell lysates resulted in the appearance of tyrosine-phosphorylated P I 3-kinase activity in the cell lysates, suggesting that the addition of ATP, Mg2+, and Mn2+ to the cell lysates initiated thekinase reaction by the activated insulin receptor toward PI 3-kinase or its associated protein and resulted in the increase in the amount of tyrosine-phosphorylated PI 3-kinase activity in CHO cell lysates. No significant decrease of the amount of PI 3-kinase activity complexed with the activated insulin receptor on the beads was observed in this experiment. This could be due to, after PI 3-kinase activitybeing released from the receptor, PI 3-kinase activitybeing continuously supplied to theactivated

' Y. Kaburagi, A. Ando, R. Yamamoto-Honda, K. Momomura, M. Kasuga, and T. Kadowaki, unpublished observations. 'Y. Kaburagi, K.Momomura, M. Kasuga, and T. Kadowaki, unpublished observations.

insulin receptor from the cell lysates. To arrest this supply from the cell lysates, the dissociation of receptor-associated PI 3-kinase activityfrom the activated receptorwas examined after washing out CHO cell lysates. The incubation of PI 3kinase activity associated with the activated insulin receptor in the presence of 1 mM ATP, 10 mM M e , and 3 mM Mn *+ decreased the association of PI 3-kinase activity with the receptor compared to the incubation in the absence or presence of either M e and Mn2' or ATP alone. It is not known why higher amount of receptor-associated PI 3-kinase activity was observed in the presence of ATP alone compared to the incubation in the absence or presence of MgZ+ and Mn2+. These results are consistent with our hypothesis, in which PI 3-kinase or its associated protein being released from the activatedreceptor afterthe completion of tyrosinephosphorylation by the receptor. In the case of the association of EGF receptor and phospholipase C-y, Margolis et al. (42) have suggested that the association represents an enzymesubstrate intermediate in which the non-tyrosine-phosphorylated phospholipase C-y is complexed at thesubstrate-binding site of EGF receptor and that, once the kinase reaction is completed, the phosphorylated phospholipase C-y product is released. It is possible that theone associatedwith the insulin receptor is either non-tyrosine-phosphorylated or less tyrosine-phosphorylated form of P I 3-kinase or its associated protein, whereas the released one from the receptor is a fully tyrosine-phosphorylated form. The present studies showed that the preincubation of the insulin receptor with ATP is necessary for the association of P I 3-kinase activity withthe receptor, and thekinase negative receptor failed to bind PI 3-kinase activity. These results suggest that the autophosphorylation of the insulin receptor and/or itsassociated conformationalchange are necessary for the association of P I 3-kinase activity withthe insulin receptor. Therefore, there could be at least two models for interaction between PI 3-kinase activity and the insulin receptor. One is that tyrosyl autophosphorylation sites are involved in the association of PI 3-kinase activity withthe insulin receptor. The otheris that the conformational change induced by the receptor autophosphorylation allows PI 3-kinase activity to associate with the receptor. It is not known whether one or both of these models are involved in this interaction. In the case of PDGF receptor, PI 3-kinase directly associates with a specific sequence of the receptor containing phosphotyrosine (43). Recently, Cantley et al. (44) have proposed TyrMet-X-Met motif as a consensus sequence for tyrosine phosphorylation sites thatbind SH2 domain on the85 kDasubunit of PI 3-kinase and have suggested that the sequence around Tyr13'* on the insulin receptor is a candidate of the binding site for PI 3-kinase. We studiedthis possibility using amutant insulin receptor IR(F1316, 1322), in which two tyrosyl autophosphorylation sites a t C terminus (34-36) were substituted with phenylalanine by site-directed mutagenesis. However, this mutant showed the same level of associated PI 3-kinase activity as the wild-type receptor as well as another deletion mutant IR(ACT36). These results indicate that the C-terminal region, including two major tyrosyl autophosphorylation sites, isnot involved in theassociation of PI 3-kinase activity. On the other hand, IR(F960) and IR(A960), in which Tyrg6' was substituted with phenylalanine or alanine, respectively, exhibited reduced PI 3-kinase activity associated with the receptor compared to the wild-type receptor. These results are consistent with those of studies of PI 3-kinase activity in anti-phosphotyrosine immunoprecipitates from insulin-stimulated cells expressing mutant receptors at theC terminus or

I n Vitro Association of PI 3-Kinase Activity with Insulin Receptor

445

White, M., Cantley, L. & Roberts, T. (1987) Cell 60, 1021Tyrm (26, 45),7 suggesting that these phenomena are also 1029 observed in intact cells as well as in uitro. Impairment of the 12. Lips, D. L., Majerus, P. W., Gorda, F. R., Young, A. T. & association of PI 3-kinase activity by substitution of Tyrg60 Benjamin, T. L. (1989) J. Biol. Chem. 2 6 4 , 8759-8763 with phenylalanine or alanine might implicate that the se- 13. Serunian, L. A., Haber, M. T., Fukui, T., Kim, J. W., Rhee, S. quence around Tyrg60is the specific one for the direct associG.. Lowenstein. J. M. & Cantlev. - . L. C. (1989) J. Biol. Chem. 264,17809-17815 ation of PI 3-kinase with the insulin receptor via SH2 domain of its 85-kDa subunit. However, this remains unknown be- 14. CaDenter. C. L.. Duckworth. B. C.. Auger. K. R.. Cohen. B.. &haffl;ausen,B. S. & Canyley, L. c . ( l & O ) J . Bwl.’Chem. 265; cause there is no direct evidence for the tyrosine phosphoryl19704-19711 ation of Tyrg60 during insulin stimulation (35-37). 15. Shibasaki, F., Homma, Y. & Takenawa, T. (1991) J . Bwl. Chem. Recently, one of major substrates for the insulin receptor 266,8108-8114 tyrosine kinase, IRS-1 (insulin receptor substrate-1) hasbeen 16. Escobedo, J. A., Navankasattusas, S., Kavanaugh, W. M., Milfay, D., Fried, V. A. & Williams, L. T. (1991) Cell 6 5 , 75-82 purified and its cDNA has been cloned out by White et al. (46, 47). The IRS-1 has characteristics similar to those of 17. Skolnic, E. Y., Margolis, B., Mohammadi, M., Lowenstein, E., Fischer, R., Drepps, A., Ullrich, A. & Schlessinger, J. (1991) pp185 (48-50) and contains consensus sequences, Tyr-MetCell 65,83-90 X-Met, for tyrosine phosphorylation sites that bind SH2 18. Otsu, M., Hiles, I., Gout, I., Fry, M. J., Ruiz-Larrea, F., Panayodomain on the 85-kDa subunit of PI 3-kinase. In addition, tou, G., Thompson, A., Dhand, R., Hsuan, J., Totty, N., Smith, the polyclonal anti-IRS-1 antibody is found to be capable of A. D., Morgan, S. J.,Courtneidge S. A., Parker, P. J. & coimmunoprecipitating PI 3-kinase activity (47). Therefore, Waterfield, M. D. (1991) Cell 65,91-104 it has been hypothesized that the IRS-1 is a molecule that 19. Kasuga, M., Fujita-Yamaguchi, Y., Blithe, D. & Kahn, C.R. (1983) Proc. Natl. Acad. Sci. U. S. A . 80, 2137-2141 links PI3-kinase to theinsulin receptor. Our present data are 20. Kahn, C. R. &White, M. F. (1988) J. Clin. Invest. 82,1151-1156 consistent with this hypothesis, since mutant receptors 21. Becker, A. B. & Roth, R. A. (1990) Annu. Rev. Med. 41,99-115 IR(F960) and IR(A960) showed the impairment of insulin- 22. Kasuga, M., Izumi, T., Tobe, K., Shiba, T., Momomura, K., induced tyrosine phosphorylation of ~p185.~However, further Tashiro-Hashimoto, Y. & Kadowaki, T. (1990) Diabetes Care 13,317-326 studies are required to clarify the molecular basis of the 23. Roth, R.A., Steele-Perkins, G., Hari, J., Stover, C., Pierce, S., association of PI 3-kinase activity withthe insulin receptor. Turner, J., Edman, J. C . & Rutter, W. J. (1988) Cold Spring In conclusion, the present study demonstrated that 1) PI Harbor Symp Quant Biol. 53,537-542 3-kinase activity associates with the insulin receptor in vitro, 24. Endemann, G., Yonezawa, K. & Roth, R. A. (1990) J. Biol. Chem. 2) the autophosphorylation of the receptor and/or its associ265,396-400 ated conformational change appear to be necessary for the 25. Ruderman, N., Kapeller, R., White, M. F. & Cantley, L. C. (1990) association of PI 3-kinase activity, 3) PI 3-kinase or its Proc. Natl. Acad. Sci. U. S. A . 8 7 , 1411-1415 associated protein seems to be released from the activated 26. Yonezawa, K., Pierce, S., Stover, C., Aggerbeck, M., Rutter, W. J. & Roth, R. A. (1991) in Insulin, Insulin-like Growth Factors receptor after thecompletion of its tyrosine phosphorylation and Their Receptors (Raizada, M., and LeRoith, D., eds) by the receptor, and 4) Tyrg6’in the juxtamembrane region of Plenum Press, New York, in press the receptor @-subunitmay be involved in the association of 27. Glenney, J. R., Jr., Zokas, L. & Kamps, M. P. (1988) J.Immunol. PI 3-kinase activity with the insulin receptor, but not C Methods 109,277-285 terminus region of the @-subunitincluding two tyrosil auto- 28. Ellis, L., Clauser, E., Morgan, D. O., Edery, M., Roth, R. A. & Rutter, W. J. (1986) Cell 4 5 , 721-732 phosphorylation sites. Finally, in vitro assay system for the association of PI 3-kinase activity with the insulin receptor 29. Ebina, Y., Ellis, L., Jarmagin, K., Edery, M., Graf, L., Clauser, E., Ou, J . H., Masiarz, F., Kan, Y. W., Goldfine, I. D., Roth, R. can be utilized to study the mechanism of interaction of these & Rutter, W. (1985) Cell 46, 747-758 molecules and will be an useful method to detect other asso- 30. Higuchi, R., Krummel, B. & Sakai, R. K. (1989) Nucleic Acids ciated molecules with the insulin receptor. Res. 16,7351-7367 Acknowledgments-We thank Dr. Richard A. Roth for a gift of monoclonal anti-insulin receptor antibody (5D9 and 29B4). We also would like to thank Drs. K. Doi, S. Higashi, M. Kishimoto, H. Ueda, K. Tobe, 0. Koshio, Y. Akanuma, and Y. Yazaki for their support and useful discussions. REFERENCES 1. Auger, K. R., Serunian, L. A., Soltoff, S. P., Libby, P. & Cantley, L. C. (1989) Cell 57, 167-175 2. Whitman, M., Downes, C. P., Keeler, M., Keller, T. & Cantley, L. (1988) Nature 332, 644-646 3. Fukui, Y. & Hanafusa, H. (1989) Mol. Cell. Biol. 9, 1651-1658 4. Courtneidge, S. A. & Heber, A. (1987) Cell 50, 1031-1037 5. Whitman, M., Kaplan, D.R., Shaffhausen, B., Cantley, L. & Roberts, T. M. (1985) Nature 3 1 5 , 239-242 6. Coughlin, S. R., Escobedo, J. A. & Williams, L. T. (1989) Science 2 4 3 , 1191-1194 7. Williams, L. T. (1989) Science 2 4 3 , 1564-1570 8. Varticovski, L., Druker, B., Morrison, D., Cantley, L. & Roberts, T. (1989) Nature 342, 699-702 9. Fukui, Y., Kornbluth, S., Wang, S-H. & Hanafusa, H. (1989) Oncogene Res. 4,283-292 10. Bjorge, J. E., Chan, T-O., Antczak, M., Kung, H-T. & Fujita, D. J. (1990) Proc. Natl. Acad. Sci. U. S. A. 87,3816-3820 11. Kaplan, D. R., Whitman, M., Schaffhausen, B., Pallas, D. C.,

’ K. Yonezawa, Y. Kaburagi, A. Ando, T. Kadowaki, and M. Kasuga, unpublished observations.

31. Kadowaki, H., Kadowaki, T., Wondisford, F. E. & Taylor, S. I. (1989) Gene (Amst.) 7 6 , 161-166 32. Ullrich, A., Bell, J. B., Chen, E. Y., Herrera, R., Petruzzelli, L. L. M., Dull. T. J., Gray, A., Coussens, L., Liao, Y-C., Tsubokawa, M., Mason, A., Seeburg, P. H., Grunfeld,. C., Rosen, 0. M. & Ramachandran, J. (1985) Nature 313,756-761 33. Yamamoto-Honda, R., Koshio, O., Tobe, K., Shibasaki, Y., Momomura, K., Odawara, M., Kadowaki, T., Takaku, F., Akanuma, Y. & Kasuga, M. (1990) J. Biol. Chem. 2 6 5 , 14777-14783 34. Ebina, Y., Araki, E., Taira, M., Shimada, F., Mori, M., Craik, C. S., Siddle, K., Pierce, S. B., Roth, R. A. & Rutter, W. J . (1987) Proc. Natl. Acad. Sci. U. S. A. 8 4 , 704-708 35. Tornqvist, H. E., Pierce, M. W., Frackelton, A. R., Nemenoff, R. A. & Avruch, J. (1987) J. Biol. Chem. 262,10212-10219 36. Tornqvist, H. E., Gunsalus, J. R., Nemenoff, R. A., Frackelton, A.R., Pierce, N. W. & Avruch, J. (1988) J. Bwl. Chem. 263, 350-359 37. White, M.F., Shoelson, S. E., Keutmann, H. & Kahn, C. R. (1988) J. Biol. Chem. 2 6 3 , 2969-2980 38. Koshio, O., Akanuma, Y. & Kasuga, M. (1989) FEBS Lett. 2 5 4 , 22-24 39. Lewis, R., Cao, L., Perregaux, D. & Czech, M. P. (1990) Biochem&tv2 9 , 1807-1813 40. White, M. F., Livingston, J. N., Backer, J. M., Lauris, V., Dull, T. J., Ullrich, A. & Kahn, C. R. (1988) Cell 5 4 , 641-649 41. Backer, J . M., Kahn, C. R., Cahill, D.A., Ullrich, A. & White, M. F. (1990) J . Biol. Chem. 2 6 5 , 16450-16454 42. Margolis, B., Bellot, F., Honegger, A. M., Ullrich, A., Schhlessinger, J. & Zilberstein, A. (1990) Mol. Cell. Biol. 1 0 , 435-441 43. Escobedo, J. A., Kaplan, D. R., Kavanaugh, W. M., Turck, C. W.

446

I n Vitro Association of P I 3-Kinase Activity with Insulin

& Williams, L. T.(1991) Mol. Cell. Biol. 11, 1125-1132 44. Cantley, L. C., Auger, K. R., Carpenter, C., Duckworth, B., Graziani, A., Kappeller, R. & Soltoff, S. (1991) Cell 6 4 , 281302 45. Myers, M. G., Jr., Backer, J. M., Siddle, K. & White, M. F. (1991) J. Biol. Chem. 2 6 6 , 10616-10623 Lane, W. S., Karasik, A., Backer, J., White, 46. Rothenberg, P. L., M. & Kahn, C. R. (1991) J. Biol. Chem. 266,8302-8311 47. Sun, X. J., Rothenberg, P., Kahn, C. R., Backer, J. M., Araki, E.,

Receptor

Wilden, P. A., Cahill, D. A., Goldstein, B. J. & White, M. F. (1991) Nature 3 6 2 , 73-77 48. White, M. F., Maron, R. & Kahn, C. R. (1985) Nature 318,183186 49. White, M. F., Stegmann, E. W., Dull, Y. J., Ullrich, A. & Kahn, C. R. (1987) J. Biol. Chem. 262,9769-9777 50. Kadowaki, T., Koyasu, S., Nishida, E., Tobe, K., Izumi, T., Takaku, F., Sakai, H., Yahara, I. & Kasuga, M. (1987) J. Biol. Chem. 262,7342-7350