Antibodies to the Extracellular Receptor Domain Restore the Hormone ...

THEJOURNALOF BIOLOGICAL CHEMISTRY

Vol. 268,No. 15,Issue of May 25,pp. 11272-11277,1993 Printed in U.S. A.

Antibodies to the ExtracellularReceptor Domain Restore the Hormone-insensitive Kinaseand Conformation of the Mutant Insulin Receptor Valine 382” (Received for publication, January 7, 1993, and in revised form, February 16, 1993)

Christine LebrunSQ, Veronique Baron$, Perla KalimanSll, Nadine GautierS, Jacqueline DolaisKitabgiS, SimeonTaylor[[,Domenico Accili)), and EmmanuelVan ObberghenS ** From the Slnstitut National dela Sante et de la Recherche Medicale, Unite 145, Faculte de Medecine, 06107 Nice, Cidex 2, France and the 11 Diabetes Branch, National Instituteof Diabetes and Digestive and Kidney Diseases, National Institutesof Health, Bethesda, Maryland 20892

A mutation substituting a valine for phenylalanine at residue 382 in the insulin receptor a-subunit has been found in two sisters with a genetic form of extreme insulin resistance. This receptor mutation impairs the abilityof the hormone to activate autophosphorylation of solubilized receptors and phosphorylation of substrates (Accili, D., Mosthaf, L., Ullrich, A., and Taylor,S. I. (1991) J . Biol. Chem. 266,434-439). We have previously demonstratedthat in native receptors insulin induces a conformational change in the receptor @-subunit, which is thought to be necessary for receptor activation (Baron, V., Gautier, N., Komoriya, A., Hainaut, P., Scimeca, J. C., Mervic, M., Lavielle, S., Dolais-Kitabgi, J., and Van Obberghen, E. (1990)Biochemistry 29,4634-4641). Hence, it was thought that a defect in this conformational change might explainthefunctionaldefect of themutant receptor. This appears to be the case, as we demonstrate here that the mutant receptor is locked in its inactive configuration. However, we found twomonoclonal antibodies, directed to the extracellular domain, which are capable of restoring the mutant receptor kinase activity. The activationof the mutant receptor was accompaniedby restoration of conformational changes in the @-subunit C terminus. From these data, we draw the two following conclusions. (i) A causal link exists between receptor kinase activationandtheoccurrence of conformational changes. (ii) Ligands other than insulin, such as antibodies, which perturb the extracellular domain, can function as alternative ways to restore the mutant receptor kinase.

glycoprotein consisting of two 135-kDa extracellular a-subunits linked to two 95-kDa transmembrane @-subunits. The a-subunit contains the hormone binding site, whereas the @subunit cytoplasmic domain has an intrinsic tyrosine kinase activity (1, 2). Hormone binding to the a-subunit activates the receptor kinase, leading to the phosphorylation of tyrosine residuesin theP-subunitandphosphorylation of cellular substrates (3-5). How the receptor activation eventuates in the multiple hormonal responses is not yet clearly understood. However, the importance of the receptor kinase activity is now established, and a good correlation is observed between the level of this activity and altered insulin action. For example, a decreased tyrosine kinase activity hasbeen found to be associated with insulin resistance and type I1 diabetes (68). Conversely, an increase in receptorenzymic activity is linked to situationswith hypersensitivity to the hormone (9). Cloning of the receptor cDNA and determination of the amino acid sequence have led to major advances in our understanding of the receptor structure-function relationship. Moreover, this has allowed the identification of mutations responsible for a number of syndromes of extreme insulin resistance in humans. Several studies have shownthat natural mutations of the insulin receptor gene can render cells resistant to the biological actions of the hormone (10). In earlier work, Accili et al. (11-13) have identified a mutation in the insulin receptor gene obtained from two sisters with type A extreme insulin resistance,a genetic form of insulin-resistant diabetes associatedwithhyperandrogenism and acanthosis nigricans. This mutation results in the substitutionof valine for phenylalanine at residue 382 in the a-subunit and leads to a decreased number of cell-surface receptors. In cell-free systems, the mutation does not affect hormone binding but impairs theability of insulin to stimulate the receptor kinase, although the kinase retains the ability to be transactivated by Insulin is a key regulator of the metabolism of glucose, phosphorylation of its cytoplasmic domain by a wild type lipids, and proteins. The first step of its biological action is receptor (12). In the presentwork, we approached themolecular defect of binding to a specific cell-surface receptor. This receptor is a the Val”” receptor. We observed that the mutantreceptor is * This project was supported by Bayer-Pharma Grant 89038 and constrained in a conformation that is not changedby insulin by the Institut National de la Sante et de la Recherche Medicale. The binding. This is different from the wild type receptor, which costs of publication of thisarticle weredefrayed in part by the was previously described to undergo conformational changes in the @-subunit during its hormone-induced activation (14). payment of pagecharges. Thisarticlemustthereforebe hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 Interestingly, the Val382mutant receptor kinase can be actisolely to indicate this fact. vated by two monoclonal antibodies directed to the receptor § Supported by a fellowship from the Fondation pour la Recherche extracellular domain, and thisis concomitant with the occurMedicale, France. rence of conformational changes in the mutantreceptor. B Supported by a fellowship from the Ligue FranGaise contre le Cancer, Comite regional du Var. ** To whom correspondence should be addressed INSERM Unite 145, Faculte de Medecine, Ave. de Valombrose, 06107 Nice, Cedex 2, France. Tel.: 33-93-85-16-54; Fax: 33-93-92-07-13.

MATERIALSANDMETHODS

Cell Lines-NIH 3T3 fibroblasts transfected with expression plasmids encoding the wild type human insulin receptor or the Val382

11272

Altered Conformation of the Mutant Insulin Receptor Vars2 receptor (Vals2) were produced by Accili et al. (11).The wild type and mutant insulin receptors correspond to isoforms lacking the 12 amino acids of the C terminus of the a-subunit (l),but we have used the numbering system including these residues (2). Both cell lines express approximately the same number of cell-surface receptors, i.e. 5 X 106/cell. Cells were maintained in Dulbecco'smodified Eagle's medium supplemented with 10% fetal calf serum and 1 mM L-glutamine. Antibodies to Synthetic Peptides-In this study, we used two antipeptide antibodies directed against the following proreceptor sequences: (i) 1306-1329, in the @-subunit C terminus; and (ii) 456465, located in the a-subunit. The antipeptide antibodies were produced as described earlier and were partially purified by chromatography on protein A-Sepharose. The anti-(1306-1329) was previously found to detect conformational changes in the receptor C terminus, induced by insulin binding or receptor autophosphorylation (14). The sequence 456-465 is located in the domain 450-601, originally described to represent an antigenic region for a series of autoantibodies found in type B insulin-resistant patients (15). The antipeptide does not immunoprecipitate the native insulin receptor but recognizes the denatured receptor in immunoblotting experiments. MonoclonalAntibodies t o the Insulin Receptor-Highly purified insulin receptors were prepared from fresh human placentas by chromatography on insulin bound to Sepharose following a procedure previously described (16). Purified receptor (1pg) emulsified in complete Freund's adjuvantwas used to inject female BALB/c mice. Ten days later, the procedure was repeated using incomplete Freund's adjuvant, and two other boosts were performed. The mice splenic lymphocytes were fused to NSI cells according to Milstein (17). Positive cells were expanded, cloned by limited dilution, and subsequently grown in tissue culture and ascites fluid in pristane-primed BALB/c mice. Screening was based on the capacity of antibodies to immunoprecipitate the insulin receptor. Several monoclonal antibodies to the extracellular domain of the insulin receptor were obtained. We used two of them, B6 andaM2, which are IgG2. and IgG, immunoglobulins, respectively. They are specific for the human insulin receptor and do not cross-react with insulin receptors from rat or mouse.These antibodies are able to immunoprecipitate the insulin receptor but do not recognize it in immunoblot experiments. Note that neither B6 nor aM2 recognizes the human IGF-I receptor. Receptor Purification-Cells expressing wild type human insulin receptor or Val3"were solubilized in 1% Triton X-100, 50 mM HEPES, 150 mM NaC1, 10 pg/ml leupeptin, 1.25 mM bacitracin, 100 units/ml Trasylol, and 1 mM phenylmethylsulfonyl fluoride, pH 7.6, for 90 min at 4 "C (18). The insoluble material was separated by ultracentrifugation at 100,000 X g for 1 h at 4 "C. The supernatant was applied to a wheat germ agglutinin-Sepharose column, and receptors were eluted with 0.1% Triton X-100, 50 mM HEPES, 150 mM NaC1, 0.3 M N-acetylglucosamine. To remove endogenous (murine) receptors, the wheat germ agglutinin-Sepharose column extract was then further purified by affinity chromatography using Sepharoselinked antibodies specific for human insulin receptor. Elution was performed with 1.5 M MgClZ, 120 mM sodium borate, 0.1% bovine serum albumin, 0.1% Triton X-100, pH 7.5, and immediately followed by a 7-fold dilution in 30 mM HEPES, 30 mM NaCl, 0.1% bovine serum albumin, 0.1% Triton X-100, pH 7.5 (final volume, 60 ml). A second chromatography on wheat germ agglutinin-Sepharose column was performed to eliminate the elution buffer and toconcentrate the purified receptors. The receptor concentration of each preparation was determined by Scatchard analysis of 1251-insulin binding. Preparation of 36S-Receptors-Cells expressing the VaP2receptor were washed twice with phosphate-buffered saline and placed in methionine-free Earle's minimal essential medium containing 10% fetal calf serum, L-glutamine, and ~ - [ ~ ~ S ] m e t h i o n(1 i n mCi/l5-cm e diameter dish). After 16 h, cells were rinsed with phosphate-buffered saline, solubilized, and the insulin receptors were partially purified by chromatography on wheat germ agglutinin agarose. Immunoprecipitation of the Val3" Mutant 35S-Receptors--35SReceptors were incubated either with insulin (0.1 p ~ or) with buffer alone for 3 h at 4 "C. Increasing amounts (20-220 fmol) of these receptors were exposed to antipeptide 1306-1329 (50 pg/ml) in buffer supplemented or not with insulin to maintain the hormone concentration. After 2 h at 4 "C, protein A was added for 1 h at 4 "C, and pellets were washed twice in 50 mM HEPES, 150 mM NaCl, 0.1% Triton X-100, pH 7.5, before addition of 1 ml of scintillation solution for counting (19). Some experiments were performed after a preincubation of 35S-receptorswith monoclonal antibodies aM2 or B6 at indicated ascites dilutions for 2 h at 4 "C.

11273

Receptor Autophosphorylation-Insulin receptors (40 fmol/sample) were incubated with varying dilutions of monoclonal antibodies aM2 or B6 for 2 h at 4 'C, and, when indicated, insulin (0.1 p M ) was added for 1 h at room temperature. Receptor autophosphorylation was initiated by addition of 15 p M [y3'P]ATP (10 mCi/mmol), 4 mM MnC12, 8 mM MgClZ.After 10 min, the reaction was stopped with Laemmli sample buffer containing 3% (v/v) SDS and 5% (v/v) 8mercaptoethanol(20). The samples were analyzed by SDS-polyacrylamide gel electrophoresis followed by autoradiography. Phosphorylation of Erogenous Substrates-Receptors (20 fmoll sample) were incubated with monoclonal antibodies for 2 h at 4 "C. Samples were exposed or not to insulin (0.1 p M ) for 1 h at room temperature. Phosphorylation was carried out by addition of poly(G1u-Tyr) at a final concentration of0.2 mg/ml and the phosphorylation mixture (15 p~ [y3'P]ATP, 4 mM MnClZ, 8 mM MgCld. The reaction was stopped after 45 min by spotting aliquots onto phosphocellulose papers (Whatman P-81). Papers were extensively washed in 1% orthophosphoric acid and dried. The radioactivity incorporated into the substrate was determined by Cerenkov counting. Insulin Binding to Solubilized Receptors-Purified receptors (20 fmol/sample) were preincubated with antibodies overnight at 4 "C. Tracer amounts of 1Z51-insulin(220 pCi/pg) were added for 3 h at 15 "C, and the receptor-hormone complexes were precipitated after 15 min on ice in the presence of 7-globulins (0.2%, w/v) and polyethylene glycol (25%, w/v). After centrifugation for 15 min at 13,000 X g, pellets were washed in polyethylene glycol (12.5%, w/v), and radioactivity was determined in a y counter. Nonspecific binding was measured in the presence of 0.1 p M unlabeled insulin. Purification of F(ab'Jz Fragments from B6 IgG-The preparative kit from Pierce Chemical Co. was used (ImmunopureTMF(ab')2 preparation kit). Briefly, B6IgGweredialyzed overnight at 4 "C against 20 mM sodium acetate, pH4.5 (digestion buffer), and concentrated toapproximately 20 mg/ml. Then, 0.5 ml of immobilized pepsin was washed in the digestion buffer and resuspended with 10 mgof IgG for 2 h at 37 "C in a shakingwater bath. Acolumn of immobilized protein A was equilibrated with the Pierce Chemical Co. binding buffer, and the crude digest was applied. This step allows separation of F(ab')2fragments from nondigested IgG and Fc fragments. F(ab')z were dialyzed against 30 mM HEPES, 60 mM NaCl, pH 7.4, overnight at 4 "C and kept for several weeks at -20 "C. RESULTS

Absence of Insulin-induced Conformational Change in Mutant InsulinReceptor V~l~~*-First, we searched for a possible link between the defect in the insulin-dependent activation of the receptor tyrosine kinase and structural modifications of the mutantreceptor. To do this, we studied the interaction of both receptors with antibodies to theextracellular domain: (i) anantipeptide antibody to receptor sequence 456-465, and (ii) themonoclonal antibodies a M 2 and B6. The effect of these antibodies on insulin binding to wild type and mutant receptors was measured by incubation of antibodies with the receptors before addition of tracer amounts of '251-insulin.In these conditions, an alteration in hormone binding could reflect a modification in receptor affinity for the hormone or in thenumber of available binding sites. Results obtainedwith the antipeptide to sequence 456465 are presented in Fig. 1. This antibody inhibits insulin binding to the mutant receptor, whereas it has no effect on the wild type receptor. Monoclonal antibodies a M 2 and B6 also had different effects upon ligand binding to mutant and wild type receptors (Fig. 2). Indeed, a M 2 inhibited hormone wild binding t o the mutantreceptor more strongly than to the type one. In contrast, B6 reduced hormone binding to the wild type receptor to a considerable extent butexerted only a slight effect on the mutant receptor. Taken together, these data show that our antibodies distinguish between the two receptors and suggest that the extracellular conformation of the mutant receptor and that of the wild type receptor are different. We have previously shown that insulin binding induces a

Altered Conformation of the Mutant Insulin Receptor Val3s2

11274

conformational change in thewild type receptor C terminus, and we proposed that this change could be important in the mechanism leading to receptor activation. Then, we hypothesized that the Val3" receptormightnot undergo such a hormone-induced change, explaining the failureof insulin t o stimulate its kinase activity. To test this view, increasing amounts of 35S-receptors were incubatedwithorwithout insulin before addition of an antipeptide to the C-terminal domain 1306-1329. This antipeptide is able to detect a conformational change after hormone binding to the wild type receptor. Indeed, insulin binding results ina 50% decrease in immunoprecipitation compared with non-occupied receptors (at saturatingreceptorconcentrations) (19). However, as shown in Fig. 3, immunoprecipitation of the Val3" receptor is not significantly affected by insulin binding. Thus, theVal"' andthe wild typereceptorshavedistinctbehaviorswith respect toantipeptide recognition. Ourdataindicatethat interaction of the hormone with the mutant receptor does not change thereceptor C-terminal configuration in the same way as has been shown for the wild type receptor. We conclude that the Val3" receptor is locked in a conformationthat cannot be changed by insulin and appears be to different from that of the wild type receptor. Monoclonal AntibodiestoReceptorRestoretheTyrosine Kinase Activity of the Mutant Insulin Receptor VaPa2"The Val3" receptor kinase has been shown to be unstimulated by hormone binding to the receptor a-subunit but to be able to undergo activation upon transphosphorylation of itscytoplasmic domain. This indicates that the intrinsic kinase is intact (12). We were therefore interested to assess theeffect

of our extracellular targeted antibodies on the mutant receptor kinase activity. In these experiments, antibodies were incubated with Val382 receptorsandthenwithorwithoutinsulinatsaturating concentration. Finally, the substrate poly(G1u-Tyr) and the phosphorylationsolution were added. As expected, in the absence of antibody, insulin had little effect on the phosphorylation of poly(G1u-Tyr) (Fig. 4). In the lower panel, results obtained with aM2 in the presence of insulin are shown. The antibody plus the hormone stimulate the Val382 receptor about 2-fold (220% of basal a t optimal concentration), indicating that combination of the two effectorsrestores the Val3" kinase. Next, we tested the effect of the other monoclonal antibody B6 (upper panel). This antibody also stimulates 2-fold the mutant receptor kinase activity, but, in contrast to aM2, B6 alone appears to be sufficient. The 2fold stimulation obtained with aM2 plus insulin or with B6 corresponds approximately to the insulin-induced stimulation of the wild type receptor, i.e. 2-%fold (data not shown). We then analyzed whether this could be correlated witha n increase in receptor autophosphorylation. These experiments were performed as described above, except that noexogenous substrate was added before the phosphorylationmixture. The samples were analyzed by SDS-polyacrylamide gel electrophoresis under reducing conditions followed by autoradiography. As seen in Fig. 5, there is no 32Pincorporation in the receptor /3-subunit in theabsence of antibody, whether insulin is present or not. After incubation with aM2 (panel A ) , only a slight increasein phosphorylation isobserved in theabsence of insulin, whereas in the presence of the hormone, aM2 enhanced considerably the mutant receptor autophosphorylation. By contrast, B6 alonewas able to stimulate theVal3S2 receptor autophosphorylation, and phosphate incorporation was not enhanced by addition of insulin (panel B ) . T o summarize, we have obtained two monoclonal antibodies capable of restoring the Val3'* receptor kinase for autophosphorylation and phosphorylation of the synthetic substrate poly(G1u-Tyr). Interestingly, whereas B6 alone is sufficient, aM2 needs insulin toproduce the effect. Finally, our results indicated that it ispossible to stimulate the mutantreceptor by perturbing its extracellular domain. 25 0 L Monoclonal Antibodies to Receptor Alter the Conformation of the Mutant Insulin Receptor VaP'-Based on ourpreceding 0 0.2 0.4 0.6 0.0 results, we speculated that the restoration of the kinase activity by monoclonal antibodies could be due to the appearance anti(456-465) (mglml) FIG. 1. Effect of antipeptide anti-(456-465) on '251-insulin of C-terminal conformational changes normally induced in binding to its receptor. Solubilized wild type (open symbol) or the wild type receptor by insulin binding. Therefore, we mutant (closed symbol) receptors were incubated for 2 h at 4 "C with measured the effect of aM2 and B6 on 35S-receptor recogniincreasing concentrations of partially purified antipeptide anti-(456tion by the antipeptide to theC terminus. The antipeptide to 465). Thereafter, radiolabeled ligand was added at tracer concentra- sequence 456-465 was used as a control, since it does not tion for 3 h at 15"C. The nonspecific binding was determined in the presence of 0.1 PM unlabeled insulin. '251-Insulinbound to preimmune stimulate thereceptor kinase activity (data not shown). First, with Ig-treated receptor was defined as 100%.A representative experiment we made sure that the three antibodies did not interfere immunoprecipitation by the antipeptide to the C terminus. is shown.

-

B

t

FIG. 2. Effect of monoclonal antibodies on '261-insulinbinding to its receptor. Experiments were performed with aM2 and B6 as described in the legend to Fig. 1. O L

I

rnm

1 m

l / 4 0 0

1 m

aMZ dllutlon

tnw

OL

I

1 / 1 m In#, l/aa

lrn

86 dllutlon

1/1w

Altered Conformation of the Mutant Insulin Receptor VUEs2 A

A

B

C

11275 E

F

G

+ -

+

- +

D

H

0-

1

20

60

100

1a0

140

Val, 35 S-receptor (frnol) FIG.3. Anti-(1306-1329)-induced immunoprecipitation of Val3"mutant 35S-receptor occupied or not by insulin. "'SReceptors were incubated with 0.1 p~ insulin (open symbols) or with buffer (closed symbols) for 3 h a t 4 "C. Increasing amounts (20-180 fmol/sample) of the "'!+receptors were then added to 50 pg/ml antipeptide to sequence 1306-1329 for 2 h a t 4 "C. After precipitation with protein A, the pellets were washed and resuspended in scintillation solution. The data shown correspond to the mean of three independent experiments f S.E. They are expressed as a percentage of the maximal immunoprecipitation obtained in each experiment.

C

0

200,

B

- + -

111200 11400 11100 A

-

ti addition:

insulin ascites dilution

113200 111600

none insulin

11800

11100

B6 (ascites dilution)

113200 111600 11800 11400

addition: none insulin

11200

1/200

uM2 (ascites dilution)+ insulin

FIG. 4. Effect of monoclonal antibodies to insulin receptor on the receptor kinase activity. Val'' receptors (20fmol/sample) were incubated for 2 h at 4 "C either with control Ig or with monoclonal antibodies prior to addition of insulin (0.1p M ) for 1 h a t room temperature. Poly(Glu-Tyr) was then added to a final concentration of 0.2 mg/ml, and phosphorylation was performed as indicated under "Materials and Methods." After 45 min a t room temperature, the reaction was stopped, and the samples were analyzed using a filter paper assay. Resultsare expressed as a percentage of thebasal substrate phosphorylation obtained in the absence of effectors. We (upper and lowerpanels, show the meanof three and two experiments respectively) where each point was run in triplicate.

B

C

-

D

E

-

F

G

H

-

insulin + + + + ascites 1 11200 11400 11100 dilution FIG.5. Effect of monoclonal antibodies to insulin receptor on receptor autophosphorylation. Vala2 receptors (40 fmol/sample) were incubated for 2 h a t 4 "C with control Ig or with the monoclonal antibodies, a M 2 (panel A ) or B6 (panel B ) . After exposure of receptors to insulin (0.1 p M ) for 1 h a t room temperature, phosphorylation was performed for 15 min. Thereafter, the reaction was stopped with Laemmli samplebuffer, and sampleswere analyzed by SDS-polyacrylamide gel electrophoresis. The autoradiograph of one representative experiment isshown.

The antipeptide to sequence 456-465 does not immunoprecipitate the insulin receptor, and aM2 does not bind to protein A, since it is a murine IgG,. However, we had to produce F(ab'), fragments of B6, because this antibody is recognized by protein A. Having done this,we confirmed that B6 F(ab'), retained the capacity to inhibitligand binding and to stimulate kinase activity of mutant receptor at the concentration used in the immunoassay (10 pg/ml) (data not shown). The assay was performed as follows. Samples of '"S-receptors at increasing concentrations were incubated with antibodies to the extracellular domainin the condition where they exhibit the optimal stimulation of the kinase (ie. in the presence of insulin and aM2). The incubation was continued with the antipeptide to the C-terminal receptor sequence 1306-1329, and immunoprecipitationwas achievedby addition of protein A. The presence of anti-(456-465) did not interfere with the precipitation of Val'*' receptor by anti-(1306-1329), indicatingthatthisextracellularantibody does not provoke Cterminal conformational changes (data not shown). In contrast, aM2 is able to induce an improved recognition of the mutant receptor by the anti-C-terminal antibody (Fig. 6). In the absenceof insulin, aM2 has no or little effect on mutant receptor conformation (data not shown). Finally, F(ab'), B6 also induces a clear-cut increase inrecognition by the anti-Cterminal antipeptide (Fig. 7). In summary, the monoclonal antibodies aM2 and B6pro-

Altered Conformation Mutant theof

11276

b

18080

I

I

I

I

I

Insulin Receptor Val3"

I

120

I

220

Val382 receptor (fmol) FIG. 6. Anti-(1306-1329)-inducedimmunoprecipitation of ValsE2 mutant3sS-receptors occupied or not by a M 2 antibodwere incubated withbuffer (closedsymbols) or with ies. 35S-Receptors aM2 (1/400)plus insulin (open symbols) for 2 h at 4 'C. Increasing amounts of thesereceptors were then exposed to antipeptideto sequence 1306-1329 (50 pg/ml). After 2 h at 4 "C, samples were immunoprecipitated by protein A, andpellets were washed twice before addition of the scintillation solutionfor counting. The results shown are the mean of two experiments where each point was run in duplicate.Resultsareexpressed as a percentage of the maximal immunoprecipitation obtained in the presence of aM2. As a control, we verified that aM2 did not immunoprecipitatethe 35S-receptorin the absence of anti-C-terminalantipeptide. 100

I

20 b

60 140

100

.

I

I

180

220

Val382 35Sreceptor (tmol) FIG. 7. Anti-(1306-1329)-inducedimmunoprecipitation of ValsE2 mutant36S-receptoroccupied or not by F(ab')2 B6. 35SReceptors were incubated either with F(ab'), B6 (10 pg/ml) (open symbols) or with buffer alone (closed symbols) for 2 h at 4 "C. The rest of the procedure was performed as described in the legend t o Fig. 6. Results are expressed as a percentage of the maximal immunoprecipitation obtained in the presence of B6. The figure represents the mean of three independent experimentswhere each point was run in duplicate.

voke conformational changes in the mutant receptor C terminus under conditions where they stimulate the receptor kinase activity, whereasanti-(456-465) does not affect the Cterminal conformation and also is without effect on the kinase activity. DISCUSSION

Over the last 20 years, the insulin receptor has been recognized as a target for pathological processes in human dis-

ease. More recently, mutations in the insulin receptor gene have been identified in patients withgenetic forms of extreme insulin resistance. Identification of the mutations has been essential for the elucidation of molecular mechanisms responsible for altered insulin action. Moreover, these studies have provided considerable insights in the role of defined receptor domains. Mutations have been localized in virtually every domain of the receptor (10). Several mutations in the Nterminal half of the a-subunit impaired intracellular transport throughthe endoplasmic reticulumand Golgi apparatus, Among thosemutations,the Phe382 to Val3" mutationis particularly intriguing, as it affects the functioning of the intracellular @-subunit, which has become an insulin-insensitive kinase. The Phe3" to Val3" mutation was described by Accili et al. (11) who cloned and expressed the mutant receptor in NIH 3T3 cells. They have shown that the Val382mutation impairs post-translational processing and retards receptor transport totheplasmamembrane,thus decreasing thenumber of receptors on thecell surface. This is due to the binding of the mutant receptor to the immunoglobulin heavy chain-binding protein in theendoplasmic reticulum, a process described for other mutant receptors (13). Moreover, the Val3'* mutation impairs the ability of insulin to activate receptor autophosphorylation and phosphorylationof other peptide substrates (12). Because the mutant receptor can be transactivated by phosphorylation of its cytoplasmic domain, it was proposed that the defect of the mutant kinase activity was due to a defectin the molecular mechanism of signal transmission from a- to @-subunits rather than to an intrinsic alteration of the kinase. As a whole, the data presented in this study are in favor of a transductional defect as we show that the interactions of the mutant and wild type receptors with antipeptide to sequence 456-465 or with monoclonal antibodies are distinct. Indeed, the antibodies inhibit insulin binding to a different extent depending on the receptor studied (Figs. 1 and 2). Further, the mutant receptor appearsto be locked inits inactive configuration. This is reflected at themolecular level by the inabilityof insulin toinduce achange in the C terminus of Valq8',as illustrated inFig. 3. The lack of hormone-induced conformational changes probably explains the unresponsiveness of the mutant receptor to insulin and the inability of the hormone to activate the receptor kinase. It has been shown that the receptor domain surrounding the residue glycine 390 is partof the insulin binding site (21). Thefactthatthe Val3" receptorexhibitsnormalinsulin binding suggests that the binding sitemay be unaltered and that other extracellular domainsinvolved in the signal transduction from the a- to the @-subunitmay be affected by the mutation. Our data emphasize the importanceof the molecular process involving structural changes in signal transmission. To be specific, we find that an alteration in the receptor a-subunit induces inactivation of the receptor, leading at the cellular level to insulin resistance withsevere physiological consequences. The correlation that we observed betweenthe capacity of certain antibodies to activate the mutant receptor kinase activity and their ability to induce conformational changes reinforces the notion that this representsa major mechanism for receptor activation. An intriguing findingof this study concerns the discrepancy between the effect of certain antibodies on theconfiguration of the Val3" receptor and the effect of insulin on wild type receptor (Figs. 6and 7). Indeed, the antibodiesincrease Val3" recognition by the anti-C-terminal antipeptide, whereas for

Altered Conformationof the Mutant InsulinReceptor Val3s2

11277

the wild type receptor, insulin decreases this interaction (19). interest to study whether our antibodies are able to restore However, conformational changes in the wild type receptor such responses. 0-subunit have been found tobe insulin- or phosphorylationare grateful to Drs. Y. Le Marchand-Brusinduced, in a two-step mechanism (14). The hormone-evoked telAcknowledgments-We and R. Ballotti for critical reading of the manuscript. We thank receptor form is an intermediary one, and the phosphorylaG. Visciano for photographs and J. Duch for secretarial assistance. tion-inducedonecorrespondstotheactive receptor. This REFERENCES conformation is accompaniedby a n increased immunoprecip1. Ullrich, A,, Bell, J. R., Chen, E. Y., Herrera, R., Petruzzelli, L. M., Dull, T. itation by anti-(1306-1329). Thus,itis possible that our J., Gray, A., Coussens, L., Liao, Y.-C., Tsubokawa, M., Mason, A., Seeburg, P. H., Grunfeld, C., Rosen, 0. M., and Ramachandran, J. (1985) monoclonal antibodies induce a conformational change simiNature 3 1 3 , 756-761 lar to the changeinduced by phosphorylation, mimicking the 2. Ebina Y., Edery M., Ellis, L., Standring, D., Beaudoin, J., Roth, R. A,, andRutter, W.'J. (1985) Proc. Natl. Acad. Sci. U. S. A. 8 2 , 8014-8018 activating process. This couldalso explain the stimulating 3. Kasuga, M., Karlsson, F. A,, and Kahn, C. R. (1982) Science 2 1 5 , 185-187 properties of these antibodies. 4. Kasuga, M., Zick, Y., Blithe, D. L., Crettaz, M., and Kahn, C. R. (1982) Nature 298,667-669 Until now, the only known way of stimulating this mutant 5. Van Obberghen E., and Kowalski A. (1982) FEES Lett. 143,179-182 6. Grigorescu, F., h e r , J. S., and Kahn, C. R. (1984) J. Eiol. Chem. 2 5 9 , receptor was by transphosphorylation of its cytoplasmic do15003-15006 main. Here, we demonstrate that it is possible to achieve 7. Grunber er, G , Zick, Y., and Gorden, P. (1984) Science 223,932-934 8. Le Marcfand-Brustel, Y., G r h e a u x , T., Ballotti, R., and Van Obberghen, activation by interactingatthe level of theextracellular E. (1985) Nature 315,676-679 domain.The monoclonal antibody B6 stimulatesreceptor 9. Debant A. Guerre-Millo M. Le Marchand-Brustel, Y., FreychetP., Lava;, hi., and Van Obberghen, E. (1987) A n . J. Physiol. 2 5 2 , E h 3 autophosphorylation and kinase activity, without additivity R977 .. with insulin. In contrast, aM2 needs the hormone to exerta 10. TaJlor, S . I., Cama, A,, Accili, D., Barbetti, F., Quon, M. J., Luz Sierra, M., uzukl, Y., Koller, E., LeyToledano, R., Wertheimer, E., Moncada, V. stimulatory effect. Moreover,this is in apparent contradiction Y., Kadowaki H. and Ka owakl, T. (1992) Endocr. Reu. 13,566-595 with the strong inhibitionof this antibody on insulin binding 11. Accili D.Fra iLr 6. Mosthaf L. McKeon, C. Elhein S. C. Permutt M. A., Ramos, E., ianher, E., UilriAh, A., and Tiylor, S.'I. (1489) EMEO J . to Val"' receptor.This could be duetoourexperimental A. 2snq-m 7 -, 12. Accili, D., Mosthaf, L., Ullrich, A,, and Taylor, S. I. (1991) J. Biol. Chem. conditions, since hormone binding was measured at tracer 266. 434-4.19 "_, ." ligand concentration, whereas saturating hormone concentra-13. Accili, D., Kadowaki, T., Kadowaki,H., Mosthaf, L., Ullrich, A., and Taylor, S. I. (1992) J. Bid. Chem. 267,586-590 tion was used in the kinaseassay. It seems that aM2 aloneis 14. Baron, V., Kaliman, P., Gautier, N., and Van Ohherghen, E. (1992) J. Bid. Chem. 2 6 7 , 23290-23294 not potent enough to activate,by itself, the mutant receptor. 15. Zhang, B., and Roth, R. A. (1991) Proc. Natl. Acad. Sci. U. S. A. 88,9858a correlation existsbetween We conclude fromour data that 9862 Fu Ita Yamaguchi Y. Choi, S., Sakamoto, Y., and Itakura, K. (1983) J . 16. receptor kinase activation and the occurrence of conformabioi Chem., 258,6045-5049 tional changes. In sucha causal relationship, it would appear 17. Kohler, G., and Milstein, C. (1975) Nature 2 5 6 , 495-498 18. Van Ohherghen, M., Kasuga, M., Le Cam, A,, Hedo, J. A,, Itin, A., and that some antibodiestothe receptor extracellulardomain Harrison, L.C. (1981) Proc. Natl. A c d . Scr. U. S. A. 78, 1052-1056 Baron, V., Gautier N., Komoriya, A. Hainaut P. Scimeca, J. C., Mervic, 19. representalternativemeanstoactivatethemutant Val3*' M., Lavielle, S.,'Dolais-Kitabgi, J.: and Van' Obberghen, E. (1990) Biochemrstry 29,4634-4641 receptor kinase. Cells expressing Va138' receptor do not re20. Laemmli, U. K. (1970) Nature 227,680-685 spond to insulin for a number of biological actions such as 21. Fabry, M., Schaefer E. Ellis, L. Kojro E., Fahrenholz, F., and Brandenburg, D. (1992) J.'Bidl. Chem. 267,8550-8956 thymidine incorporation, c-Jun expression, and receptor au22. Quon, M. J., Cama, A., and Taylor, S. I. (1992) Biochemistry 3 1 , 9947tophosphorylation in intact cells (22). Hence, it would be of 9954 "

-"I-

__"