Proc. Nati. Acad. Sci. USA Vol. 85, pp. 8400-8404, November 1988 Biochemistry Negative and positive site-site interactions, and their modulation by pH, analogs, and monoclonal antibodies, are preserved in the purified insulin (receptor-affinity changes/negative cooperativity/insulin struct ction relationships) CHIH-CHEN WANG*, IRA D. GOLDFINEt, YOKo FUJITA-YAMAGUCHI4, HANS-GREGOR GATTNER§, DIETRICH BRANDENBURG§, AND PIERRE DE MEYTS*t *Department of , Endocrinology and , City of Hope National Medical Center, Duarte, CA 91010; tCell Biology Laboratory and Department of Medicine, Mount Zion Hospital and Medical Center, San Francisco, CA 94120; *Department of Molecular Genetics, Beckman Research Institute, City of Hope, Duarte, CA 91010; and §Deutsches Wollforschungsinstitut, D-5100 Aachen, Federal Republic of Germany Communicated by Rachmiel Levine, August 1, 1988 (receivedfor review April 25, 1988)

ABSTRACT The kinetic properties of the Although arguments have been raised against the negative were studied in solution after its purification to homogeneity. cooperativity hypothesis (23-25), no alternative model has Dissociation of "I-labeled insulin at a 1:50 dilution was not been proposed that is as compatible as the three-state model first order; unlabeled insulin at physiological concentrations outlined above with a variety of experimental findings using accelerated the dissociation rate with a maximal effect at "17 either insulin analogs, monoclonal antibodies, or alterations nM. At higher concentrations, the unlabeled insulin slowed the ofreceptors in clinical states, and with more recent studies of dissociation rate. Maximal acceleration was seen at pH 8.0. The alterations in the dimeric structure ofthe receptor (see ref. 22 ability to accelerate the dissociation rate was diminished with for review). While we believe that the experimental evidence [LeuB24]insulin and suppressed with desoctapeptide, [LeuB21], favors the site-site interactions model, it is clear that ultimate [LeuB24,B2S], desalanine-desasparagine, and desheptapeptide proof will have to come from the actual demonstration by , all of which slowed the dissociation at high concen- crystallographic or other physicochemical methods, of alter- trations. Monoclonal antibodies to the insulin receptor a native conformations of the purified receptor in various subunit (MA-5, MA-10, MA-20, and MA-51) all competed for liganded states, or by appropriate changes in receptor struc- insulin binding to the purified receptor. MA-10 and MA-51 ture and kinetics generated by site-directed mutagenesis. accelerated the dissociation of 12I-labeled insulin, while MA-5 As a first step in that direction, we have now characterized and MA-20 slowed the off rate. Thus, all the aspects of both in detail the kinetic behavior in solution of the insulin negatively and positively cooperative site-site interactions receptor purified to homogeneity from human placenta. One previously described in whole cells are present in solubilized of us previously demonstrated that Scatchard plots of 125I- purified receptors, demonstrating that these interactions rep- insulin binding to this purified preparation are curvilinear (1). resent intrinsic properties ofthe receptor molecule, most likely We now report that all the kinetic properties previously as a result of ligand-induced conformational changes. demonstrated with either whole cells or membrane prepara- tions have been preserved after extensive receptor purifica- The purification to homogeneity (1-3) and the cloning and tion, demonstrating that these properties are intrinsic to the expression (4, 5) of the insulin receptor, as well as the receptor. availability of monoclonal antibodies (6-12), have recently provided exciting means to investigate the exact mechanisms MATERIALS AND METHODS underlying the complex kinetics of insulin receptor binding. It is now well established that insulin binding departs mark- Porcine insulin (26.3 units/mg) was purchased from Sigma. edly from simple reversible mass action kinetics (13-22). [Tyr(125j)A14]insulin was prepared and purified by high- Scatchard plots are curvilinear in most systems studied, and performance liquid chromatography as described (26). Mono- the dissociation is not first order. Furthermore, dissociation clonal antibodies MA-5, MA-10, MA-20, and MA-51 have of 125I-labeled insulin (125I-insulin) at an "infinite" dilution is been described (6, 11-12, 22). Desalanine-desasparagine accelerated in the presence of unlabeled insulin. These basic insulin, desoctapeptide insulin, desheptapeptide insulin, findings, as well as the effects of numerous structural [Leun24]insulin, [LeuB25]insulin, and [LeuB24,B25]insulin modifications of the insulin molecule (16, 18-20) and more have been described (19, 27-29). Insulin receptors were recent studies with monoclonal antibodies (7, 8, 11, 12, 22), solubilized from human placental membranes and purified have led to the view (most explicitly developed in refs. 20 and 2400-fold by sequential affinity chromatography on wheat 22) that the insulin-receptor complex can shift reversibly germ agglutinin-Sepharose and insulin-Sepharose as de- between at least three interconvertible states by site-site scribed (1). This homogeneous preparation has full binding interactions, which depend on the nature of the ligand and kinase activity (1). occupying neighboring sites. We distinguished the initial, Association and dissociation of 125I-insulin were per- "empty" state with an apparent affinity denoted KAE; a lower formed in 50 mM Tris-HCl buffer containing 0.1% bovine affinity state, Kf, characterized by a faster dissociation rate serum albumin and 0.1% Triton X-100 (pH 7.4). Immuno- constant, induced by increased insulin occupancy-the phe- globulin and polyethylene glycol (PEG) were prepared in 50 nomenon known as "negative cooperativity"-and a higher mM Tris-HCl buffer (pH 7.4), except in the pH-dependence affinity state, which was denoted "Ksu.r" with a much determination in which 300 mM Tris HCl buffer was used to slower dissociation rate. ensure precipitation of the receptor in the pH range 7-8. Steady-state receptor binding was carried out as described The publication costs of this article were defrayed in part by page charge (1). Dissociation of 125I-insulin by dilution, with or without payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. Abbreviation: l25l-insulin, l25l-labeled insulin. 8400 Downloaded by guest on October 2, 2021 Biochemistry: Wang et al. Proc. Natl. Acad. Sci. USA 85 (1988) 8401

S~PECIFIC BINDING unlabeled insulin present, was performed as follows (for more 100 detailed description, see figure legends). Briefly, purified DILUTION ONLY insulin receptor (1-2 gg/ml) was incubated with 0.1 nM 50 '25I-insulin at pH 7.4 and 40C overnight. The dissociation was 0 carried out by diluting 1:50 an aliquot of the equilibrium X 20-DILUTION' mixture in the absence or presence of the indicated insulin +COILD INSULIN concentrations at 15TC. Then, at intervals, the 125I-insulin- (o-)0 10 60 120 180 receptor complex was precipitated by 0.05% immunoglobulin and 10% PEG at pH 7.4 for 10 min in an ice bath. The C')Iq receptor-bound radioactivity was separated from dissociated (n radioactivity by centrifugation and then counted. The dilu- c tion study was carried out in two different ways. In most 0 20 experiments, 20-pl aliquots from the equilibrium mixture z_1)0(Q were distributed in multiple duplicate sets oftubes containing c 6:0 1 ml of buffer alone or buffer containing unlabeled insulin or I z 1 DILUTION + COLD INSULIN analogs; one set was precipitated immediately (time 0), while i WAJ 101 the other sets were precipitated at the indicated intervals. To Q obtain more precise initial time points, in subsequent exper- z iments studying time course of dissociation (Figs. 4 and 7) after determination of total binding, the remaining equilib- 51[ rium mixture was diluted 1:50 by the addition of buffer (time V) DILUTION ONLY 0), was thoroughly mixed, and 1-ml aliquots were precip- A £-". a A itated at various times (the first being 3 min after dilution). In en a _N all figures, the time indicated is the time when the insulin- DILUTION + COLD INSULIN receptor complex and PEG are put in contact (another 10 min elapses before centrifugation). In some dissociation experi- 60 120 180 ments, unlabeled insulin was replaced by various insulin MINUTES analogs or monoclonal antibodies. All experiments were performed at least three times with comparable results. FIG. 1. Dissociation of "25I-insulin from purified receptors in the absence and presence of unlabeled ("cold") insulin. Purified insulin RESULTS receptor (1-2 ,ug/ml) was incubated overnight at pH 7.4 and 4"C with 0.1 nM 125I-insulin. Three 20-sA aliquots were removed and precip- Kinetics of Dissociation of 125I-Insulin from Purified Recep- itated by adding 25 pA of 4% immunoglobulin and 2 ml of 10%o PEG. tors. Under our experimental conditions, 60-70% of the After 10 min at 4"C, the samples were centrifuged at 2000 x g for 20 125I-insulin tracer was bound to purified receptors after min. The counts in the receptor pellet represented 125I-insulin bound overnight equilibration. The dissociation of 125I-insulin was to receptors before dissociation [shown as T.B. (total binding) and not first order (Fig. 1). Unlabeled insulin markedly acceler- set as 100%1b]. Duplicate sets of 20 Al from the equilibration mixture ated the dissociation of 125I-insulin after a 50:1 dilution. After were then added to tubes containing 1 ml ofbuffer alone (e) or buffer 3 hr at 150C, 90% of bound tracer had dissociated in the containing unlabeled insulin (o) (100 ng/ml, 1.7 x 10-8 M) at 15'C. presence of 100 ng of insulin per ml (1.7 x 10-8 M) versus At the indicated intervals, contents of duplicate tubes were precip- itated by 25 Al of 4% immunoglobulin and 1 ml of ice-cold 20%o PEG only 60% in buffer only. To assess the contribution of the for 10 min in an ice bath. The samples were then centrifuged at 2000 nonspecific binding compartment (which includes some trap- x g for 20 min. The radioactivity in the pellet was counted in a ping of tracer in this system) in the observed dissociation V-counter after aspiration of the supernatants and is expressed as rates, we carried out in parallel the binding of 1251I-insulin in percent radioactivity bound before dissociation. Data shown are the the presence of 20 pug of unlabeled insulin per ml. As shown averages ofduplicate determinations. For dissociation ofnonspecific in Fig. 1, nonspecific binding was initially 20-25% ofthe total binding (N.S.B.), similar incubations were done except for the binding and after dilution had already dissociated to a presence of20 ,ug ofunlabeled insulin per ml (3.5 A±M) under the same residual binding of -3% at the first measured time point. conditions as for total binding, then diluted in the absence (v) and Thus, there was no difference in residual nonspecific binding presence (A) of unlabeled insulin. (Inset) Nonspecific binding was in the presence or absence of unlabeled insulin at any time subtracted from all time points. Duplicates in this and the following during dissociation, and a correction for this nonspecific experiments did not differ by more than 10%6. compartment in further experiments was unnecessary. Unlabeled insulin accelerated dissociation at all pH values, Curves corrected for nonspecific binding are shown in Fig. 1 but the difference was maximal at pH ==8.0 and decreased on (Inset). both sides of this optimum. The dissociation in the presence Effect of Insulin Concentration on Tracer Dissociation. Acceleration of 1251I-insulin dissociation was dependent on of unlabeled insulin was altogether much less pH sensitive the concentration of unlabeled insulin in the medium (Fig. 2). than dissociation by dilution alone, especially in the alkaline These concentrations were well in the physiological range. range. Insulin (1.7 x 10-11 M) decreased the fraction of 125I-insulin Effect of Insulin Analogs on 1"I-Insulin Dissociation. The still associated after 2 hr ofdissociation by 15%, and maximal degree of acceleration induced by [LeuB24]insulin at 10 nM acceleration of dissociation was seen at 1.7 x 10-8 M. As was smaller than that induced by native insulin (Fig. 4). The described with whole cells and membrane preparations (14- dissociation rate of "25I-insulin was slowed down in the 16, 30), this accelerating effect was reversed when the presence of [LeuB25]insulin, desoctapeptide insulin, deshep- concentration of insulin was further increased from 1.7 x tapeptide insulin, [LeuB24,B25]insulin, and desalanine- 10-8 M to 10-5 M. At the highest insulin concentrations used, desasparagine insulin. At the latest time points, a slight dissociation of 125I-insulin was slower than that induced by acceleration was seen with [LeuB25]insulin essentially be- dilution alone. cause no further dissociation of 125I-insulin by dilution alone Effect of pH on '5I-Insulin Dissociation. Dissociation of appears to occur between 2 and 3 hr (probably due to some '25I-insulin by dilution alone was markedly dependent on pH rebinding occurring in the absence of competing ligand). (Fig. 3). Dissociation was slowest at pH -8.0 and became The dose-response curves for the various analogs are more rapid at either more acidic or more basic pH values. shown in Fig. 5. Again, while [LeuB24]insulin showed some Downloaded by guest on October 2, 2021 8402 Biochemistry: Wang et al. Proc. Natl. Acad. Sci. USA 85 (1988)

- 120 100 0 -4 z 0 P:

-4 (3 CS) 50h LL 2 +DAA g 0 2 80 Lsj 0co + [LeUB24,2]INS t+DHpP La 9 0 LLj DILUTION ONLY InJ +DOP -4 :> co 2n (I) 60 Q 30F S +[Leue25]INS a]l 2 7 4[Leu824] INS

0 C,) , DILUTION *COLD INSULIN 00 In

10-1 lo, 103 105 (ng/ml) 0 60 120 180 MINUTES lo-10 1o-8 10-6 (M) INSULIN CONCENTRATION FIG. 4. Effect of insulin and various analogs on the dissociation of 125I-insulin. The conditions for association were identical to those FIG. 2. The dose-response curve for insulin-induced accelera- in Fig. 1. After determination of total binding, the remaining tion of 125I-insulin dissociation. Association of 125I-insulin to purified equilibrium mixture was diluted 1:50 by the addition of buffer alone receptors was performed overnight as in Fig. 1. Multiple aliquots in (e) or buffer containing the indicated analog (10 nM) or buffer duplicate were diluted 1:50 in buffer alone or in buffer containing the containing unlabeled (cold) insulin. DAA, desalanine-desasparagine; indicated concentrations of unlabeled insulin. All samples were left DHPP, desheptapeptide; DOP, desoctapeptide; INS, insulin. At the to dissociate for 120 min and were processed together for determi- indicated intervals, 1-ml aliquots were removed, precipitated, and nation for residual binding as in Fig. 1. Residual bound radioactivity processed as described above. Radioactivity in the pellet was in the samples containing insulin was expressed as a fraction of that expressed as percent of that bound before dissociation. in dilution alone. 10-6 and 10-' M that minor experimental differences are accelerating ability with an optimum (like native insulin) of markedly amplified. 10-8 M, the other analogs, in contrast, all markedly slowed Effect of Monoclonal Antibodies Against the Insulin Recep- down the dissociation as their concentration increased. Note tor on 1251-Insulin Binding and Dissociation. The ability offour that the curve for insulin appears more shallow than that monoclonal antibodies to compete for 125I-insulin binding at shown in Fig. 2 because of the different scale of the graph. steady state was assessed in steady-state overnight binding Also the upward part ofthe insulin curve is so steep between experiments (Fig. 6). Antibody MA-10 was the most potent in competing, with a potency equal to =50% ofthat ofinsulin. o 30 In contrast, MA-51, MA-5, and MA-20 were 50-1000 times 0 DILUTION ONLY less potent than insulin. MA-5, MA-20, and MA-51 tended to z enhance 125I-insulin binding at the lower concentrations, but 200r 3DAA 2 P 3q F:i K [Leu 824,25] INS

[LeuB25] INS *DOP Z. v [ Leu B24] INS

Va DILUTION + COLD INSULIN eN D INSULIN 6.0 7.0 8.0 9.0 In pH FIG. 3. pH dependence of "25I-insulin dissociation. Association of '25I-insulin to purified receptors was performed overnight as in Fig. 1. Total binding was performed in duplicate and set as 100%o. Multiple aliquots (in duplicate) were diluted 1:50 either in buffer (M) alone (e) or in buffer containing unlabeled (cold) insulin (o) (100 ng/ml, 1.7 x 10-8 M); the final pH in the various diluted samples FIG. 5. Dose-response curve for acceleration of l25l-insulin ranged from 6.3 to 9.1. All samples were left to dissociate for 120 min dissociation by various analogs of insulin. The experimental condi- and were processed together for determination of residual binding as tions and expression of results are the same as in Fig. 2, except that in Fig. 1. Residual bound radioactivity is expressed as percent ofthat various analogs were present in the dilution medium at the concen- bound before dissociation. trations indicated. Abbreviations are the same as in Fig. 4. Downloaded by guest on October 2, 2021 Biochemistry: Wang et al. Proc. Natl. Acad. Sci. USA 85 (1988) 8403 solution with the purified receptor. We find that all the properties previously observed in both whole cells and 100 membrane preparations have been remarkably preserved A after purification. LL Unlabeled insulin accelerated the dissociation of 1251 0 80 I \ MA2O insulin after a 1:50 dilution. The dose-response curve was bell shaped as previously shown (13, 15, 16). Overall, the 0N2 60 \ MA5 phenomenon demonstrated a higher sensitivity to insulin than :3\ MA51 in vivo: 10-11 M induced a very significant effect and the 2 40 optimum was seen at ==10-8 M, corresponding to a 10-fold MAIO increase in sensitivity after purification. This observation is 2 20 of interest since there had been speculation that the "two- INSULIN dimensionality" of membranes may result in much higher , local effective insulin concentrations than deduced from the 10-10 lo-9 o108 10-7 concentration in solution and that it may bring the local (M) insulin concentration in the dimerizing range (36). Our results are the opposite of what would be predicted if that specula- FIG. 6. Competition of mornoclonal antibodies for 1251-insulin tion was correct: the acceleration of dissociation is observed binding. Purified receptor (1-2 pqg/ml) was incubated with 0.2 ng of at lower insulin concentration in solution. Also, although we "'I-insulin per ml (3 x 10`1 M)) with increasing concentrations (0- had first carefully envisaged and dismissed the hypothesis 1.7 x 10-' M) of unlabeled insulin or monoclonal antibodies at 40C that membrane-associated unstirred layers may play a role in overnight. The reaction was interrpted by precipitation with 0.04% immunoglobulin and 10% PEG for 10 mmb in an ice bath, and the the observed phenomena(14), this hypothesis has been raised mixture was centrifuged at 2000 x g for 20 min. The radioactivity in again later (25). The present findings definitively exclude that the pellet was counted after asxpiration of the supernatant and is peculiar membrane-associated phenomena explain the accel- expressed as percent radioactivitty bound in the absence ofcompeting erated dissociation. Moreover, as also observed in whole ligand. Nonspecific binding has not been subtracted. cells, the accelerated dissociation disappears and dissocia- tion in fact becomes slower than by dilution alone as the this effect was, for unknowri reasons, variable from experi- insulin concentration is brought into the dimerizing range (13, ment to experiment. 30). The effect of the various antibodies on 125I-insulin disso- The pH dependence of dissociation is also remarkably ciation kinetics is shown in IFig. 7. While MA-10 and MA-51 similar to that previously described (15). One difference is a both accelerated the dissociation (although to a smaller clear increase in dissociation by dilution as pH increases from extent than insulin at the studied concentrations), MA-5 and 8.0 to 9.0 with the purified receptor (compare Fig. 3 with MA-20 slowed the dissociatiion markedly. figure 7 of ref. 15). This observation may imply that the dissociation rate is affected by side-chain groups that titrate in that pH range and become exposed after purification, DISC'USSION suggesting that groups close to the membrane play a role in Shortly after our demonstration ofthe negative cooperativity the pH dependence of dissociation. of insulin receptors in intactt cells (13) and membrane prep- We previously reported (refs. 16-19; reviewed in ref. 20) arations (15), several groupss demonstrated similar phenom- that negative cooperativity is intimately linked to a cluster of ena (curved Scatchard plotss, accelerated I251-insulin disso- invariant residues on the surface of insulin that comprise ciation by unlabeled insulin) in solubilized or partially puri- AsnA21 and GlyB23-PheB24-PheB25-TyrB26 ("cooperative fled receptors (31-34). Fuijita-Yamaguchi et al. (1, 35) site"). reported that the insulin Ireceptor retained a curvilinear Negative cooperativity is markedly altered or abolished Scatchard plot even after piurification to homogeneity. We with insulin analogs where this domain has been tampered report here the first detailedI study of the kinetic and struc- with, such as desalanine-desasparagine insulin lacking ture-function properties olf the interaction of insulin in AsnA21 and AlaB30 (13, 16), desheptapeptide insulin lacking B24-B30 (¶; J. L. Gu and P.D.M., unpublished data), des- 0 octapeptide insulin lacking B23-B30 (13, 16), and mutant r-100 [LeuB24]-, [LeuB25]_, or [LeuB24,B25linsulins (19, 29). The structure-activity relationships previously described remain very similar with purified receptors. Also, as shown 50 1 To + MA5 before, the analogs with impaired or absent negative coop- 0 +MA2 erativity in fact slow down the dissociation rate. This phe- c nomenon depends on GlyB23 since, as shown here as well as DILUTION ALONE in whole cells (ref. 20; J. L. Gu and P.D.M., unpublished i30 ¶; c +MA51 data), the marked positive effect seen with desheptapeptide -Z 20 H insulin is largely lost in desoctapeptide insulin or in analogs i xMAIO of desheptapeptide insulin in which GlyB23 is substituted by I other residues. wI I We have attributed the negative cooperativity as well as A 10 D n" INSULIN the positive effect of various analogs to the induction of alternative affinity (conformational) states by site-site inter- '-6 actions. We and others have recently shown that the so- HQ 0 6060 120I20 180I80 called Kf and Ksu r states of the receptor induced by various insulin analogs are also stabilized by a variety of monoclonal FIG. 7. Effect of monoclonal antibodies on '25I-insulin dissocia- tion. The experimental conditions and expression of results are the IGu, J. L., De Meyts, P., Keefer, L. M., Piron, M. A., Chu, S. C., same as in Fig. 4, except the different monoclonal antibodies or Wang, C. C., Gattner, H. G. & Brandenburg, D., Annual Meeting unlabeled insulin are present at 10 nM in the dilution medium. of the Endocrine Society, 1981, Cincinnati, OH, abstr. 253, p. 146. Downloaded by guest on October 2, 2021 8404 Biochemistry: Wanig et A Proc. Natl. Acad. Sci. USA 85 (1988) antibodies (7, 8, 11, 12, 22), some of which behave as R. & Goldfine, I. D. (1987) J. Biol. Chem. 262, 4134-4140. antagonists ofthe negative cooperativity (22). The interaction 12. Forsayeth, J. R., Caro, J. F., Sinha, M. K., Maddux, B. A. & of monoclonal antibodies with the purified receptor showed Goldfine, I. D. (1987) Proc. Natl. Acad. Sci. USA 84, 3448- the same properties as with cells and membranes. MA-10 and 3451. 13. De Meyts, P., Roth, J., Neville, D. M., Jr., Gavin, J. R., III, MA-51 mimicked the accelerating effect of insulin, while & Lesniak, M. A. (1973) Biochem. Biophys. Res. Commun. 55, MA-5 and MA-20 slowed down the dissociation. While the 154-161. effects of the antibodies on receptor kinetics were similar to 14. De Meyts, P. & Roth, J. (1975) Biochem. Biophys. Res. those seen with whole cells (11, 12, 22), there were significant Commun. 66, 1118-1126. differences in their potency in competing for I251-insulin 15. De Meyts, P., Bianco, A. R. & Roth, J. (1976) J. Biol. Chem. binding. In IM-9 lymphocytes, MA-10 was equipotent with 251, 1877-1888. insulin, while MA-51, MA-5, and MA-20 ranged from =50% 16. De Meyts, P., Van Obberghen, E., Roth, J., Brandenburg, D. to 1% in their relative potencies. These results suggest that & Wollmer, A. (1978) Nature (London) 273, 504-509. separating the receptor from its membrane environment may 17. De Meyts, P. (1980) in and Cell Regulation, eds. result in some degree ofconformational alteration. However, Dumont, J. E. & Nunez, J. (North-Holland, Amsterdam), pp. the relative potencies reported in the present work are closer 107-121. 18. Piron, M. A., Michiels-Place, M., Waelbroeck, M. & De to those seen in Hep-G2 cells (see figure 1 of ref. 11), except Meyts, P. (1980) in Insulin: Chemistry, Structure and Function for the much weaker competition observed with MA-20 in our of Insulin and Related Peptides, eds. Brandenburg, D. & purified receptor system. Thus, the observed differences may Wollmer, A. (De Gruyter, Berlin), pp. 371-391. reflect some tissue specificity (e.g., glycosylation), as re- 19. Keefer, L. M., Piron, M. A., De Meyts, P., Gattner, H. G., cently suggested (37), rather than conformational alterna- Diaconescu, C. & Brandenburg, D. (1981) Biochem. Biophys. tions due to receptor purification. These differences, how- Res. Commun. 100, 1229-1236. ever, do not appear to change the major characteristics ofthe 20. Gammeltoft, S. (1984) Physiol. Rev. 64, 1321-1378. receptor kinetics. 21. Ilondo, M. M., Courtoy, P. J., Geiger, D., Carpentier, J. L., Several groups have observed that the curvilinear Scatch- Rousseau, G. G. & De Meyts, P. (1986) Proc. Natl. Acad. Sci. ard plot (33-35, 38, 39) and the accelerated dissociation (33, USA 83, 6460-6464. 34) require the intact a2J2 dimeric receptor structure and are 22. Gu, J. L., Goldfine, I. D., Forsayeth, J. R. & De Meyts, P. detailed observations of (1988) Biochem. Biophys. Res. Commun. 150, 694-701. lost in the isolated monomer. Our 23. Pollet, R. J., Standaert, M. L. & Haase, B. A. (1977) J. Biol. intact negative and positive site-site interactions with a Chem. 252, 5828-5834. homogeneous preparation of purified insulin receptor bring 24. Corin, R. E. & Donner, D. B. (1982) J. Biol. Chem. 257, 104- us closer to a mechanistic interpretation ofthese phenomena. 110. 25. Wheeler, F. B., Santora, A. C. & Elsas, L. J. (1980) Endocri- This work was supported by National Institutes of Health Grants nology 107, 195-207. DK 29770 and DK 34427 to Y.F.-Y. and DK 26667 to I.D.G. and by 26. Markussen, J. & Larsen, V. D. (1980) in Chemistry, Structure Cancer Center Support Grant CA 33572 to City of Hope. and Function ofInsulin and Related Hormones, eds. Branden- burg, D. & Wollmer, A. (De Gruyter, Berlin), pp. 161-168. 1. Fujita-Yamaguchi, Y., Choi, S., Sakamoto, Y. & Itakura, K. 27. Carpenter, F. H. (1966) Am. J. Med. 40, 750-758. (1983) J. Biol. Chem. 258, 5045-5049. 28. Gattner, H. G., Danho, W., Behn, C. & Zahn, H. (1980) 2. Kasuga, M., Fujita-Yamaguchi, Y., Blithe, D. L. & Kahn, Hoppe-Seyler's Z. Physiol. Chem. 361, 1135-1138. C. R. (1983) Proc. Natl. Acad. Sci. USA 80, 2137-2141. 29. Jonckzyk, A., Keefer, L. M., Naithani, V. K., Gattner, H. G., 3. Petruzelli, L. M., Herrera, R. & Rosen, 0. (1984) Proc. Nat!. De Meyts, P. & Zahn, H. (1981) Hoppe-Seyler's Z. Physiol. Acad. Sci. USA 81, 3327-3331. Chem. 362, 557-561. 4. Ullrich, A., Bell, J. R., Chen, E. Y., Herrera, R., Petruzelli, 30. Taylor, S. I. & Leventhal, J. (1983) J. Clin. Invest. 71, 1676- L. M., Dull, T. J., Gray, A., Coussens, L., Liao, Y.-C., 1685. Toubokawa, M., Mason, A., Seeburg, P H., Grunfeld, C., 31. Harrison, L. E., Billington, T., East, I. J., Nichols, R. J. & Rosen, 0. M. & Ramachandran, J. (1985) Nature (London) 313, Clark, S. (1978) Endocrinology 102, 1485-1495. 756-761. 32. Ginsberg, B. H., Kahn, C. R., Roth, J. & De Meyts, P. (1976) 5. Ebina, Y., Ellis, L., Jannagin, K., Edery, M., Graf, L., Biochem. Biophys. Res. Commun. 73, 1068-1074. Clauser, E., Ou, J. H., Mascarz, F., Kan, Y. W., Goldfine, 33. Baron, M. D. & Sonksen, P. H. (1982) Biosci. Rep. 2,785-793. I. D., Roth, R. A. & Rutter, W. J. (1985) Cell 40, 747-758. 34. Deger, A., Kramer, H., Rapp, R., Koch, R. & Weber, U. (1986) 6. Roth, R. A., Cassel, D. J., Wong, K. Y., Maddux, B. A. & Biochem. Biophys. Res. Commun. 135, 458-464. Goldfine, I. D. (1982) Proc. Natl. Acad. Sci. USA 79, 7312- 35. Fujita-Yamaguchi, Y. & Harmon, J. T. (1988) Biochemistry 27, 7316. 3252-3260. 7. Siddle, K., Soos, M. A., O'Brien, R. M., Granderton, R. H. & 36. De Lisi, C. (1979) in Physical Chemical Aspects ofCell Surface Taylor, R. H. (1986) Biochem. Trans. 15, 47-51. Events in Cellular Regulation, eds. De Lisi, C. & Blumenthal, 8. Soos, M. A., Siddle, K., Baron, M. D., Heward, J. M., Luzio, R. (Elsevier/North-Holland, Amsterdam), pp. 251-292. J. P., Bellatin, J. & Lennox, E. S. (1986) Biochem. J. 235, 199- 37. Caro, J. F., Raju, S. M., Sinha, M. K., Goldfine, I. D. & 208. Dohm, G. L. (1988) Biochem. Biophys. Res. Commun. 151, 9. Taylor, R., Soos, M. A., Wells, A. Argyraki, M. & Siddle, K. 123-129. (1987) Biochem. J. 242, 123-129. 38. Boni-Schnetzler, M., Scott, W., Waugh, S. M., Di Bella, T. & 10. O'Brien, R. M., Soos, M. A. & Siddle, K. (1987) EMBO J. 6, Pilch, P. F. (1987) J. Biol. Chem. 262, 8395-8401. 4003-4010. 39. Sweet, L. J., Morrison, B. D. & Pessin, J. E. (1987) J. Biol. 11. Forsayeth, J. R., Montemurro, A., Maddux, B. A., De Pirro, Chem. 262, 6939-6942. Downloaded by guest on October 2, 2021