Biosynthetic labeling of insulin receptor - Europe PMC

1 downloads 0 Views 1MB Size Report
treme insulin resistance and acanthosis nigricans. In accord with previous data that insulin bound to receptor reduces the affinity of receptor for anti-receptor ...
Proc. Natl Acad. Sci. USA Vol. 78, No. 2, pp. 1052-1056, February 1981

Cell Biology

Biosynthetic labeling of insulin receptor: Studies of subunits in cultured human IM-9 lymphocytes (hormone receptor biosynthesis/autoantibodies against insulin receptors/down regulation/glycosidases)

EMMANUEL VAN OBBERGHEN*t, MASATO KASUGA*, ALPHONSE LE CAMt4, JosE A. HEDO*, AHUVA ITIN*,

AND LEN C. HARRISON*§

*Diabetes Branch, National Institute of Arthritis, Metabolism and Digestive Diseases, and tLaboratory of Molecular Biology, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20205

Communicated by J. Edward Rail, November 1, 1980

("2I-insulin) to receptor, is accompanied by loss of both bio-

ABSTRACT We have identified the subunits of the insulin receptor in cultured human lymphocytes (IM-9 line) by biosynthetic labelingwith [asS]methionine and specific precipitation with autoantibodies against the insulin receptor. IM-9 lymphocytes were cultured with [3S]methionine and extracted with Triton X100. Insulin receptors were concentrated and purified 20-fold by chromatography of the cell extract on wheat germ agglutinin-agarose, and then specifically precipitated by receptor antibodies after addition of a second antibody. Analysis of the immunoprecipitates by sodium dodecyl sulfate/polyacrylamide gel electrophoresis under reducing conditions followed by autoradiography revealed specific precipitation of two major bands with molecular weights of 130,000 and 90,000. Both species were precipitated by receptor antibodies from four different patients with the syndrome of extreme insulin resistance and acanthosis nigricans. In accord with previous data that insulin bound to receptor reduces the affinity of receptor for anti-receptor antibody, we found that preincubation of the wheat germ-purified cell extract with insulin (1.7 gsM) prior to immunoprecipitation caused a decrease in the appearance of both species. The decrease in insulin binding seen after incubation of the lymphocytes with insulin for 12 hr ("down regulation") was associated with a decrease in the labeling of both the 130,000 and 90,000 bands. The apparent molecular weight of both subunits was decreased after pretreatment with mixed glycosidases. In conclusion, we have biosynthetically labeled the insulin receptor with [NS]methionine and showed that the receptor consists of two major glycoprotein subunits with apparent molecular weights of 130,000 and 90,000.

synthetically labeled receptor subunits. MATERIALS AND METHODS Materials. Porcine insulin (lot lJM95AN) was purchased from Elanco (Indianapolis, IN); Na'25I, [35S]methionine (translation grade; specific activity 1020 Ci/mmol; 1 Ci = 3.7 x 1010 becquerels), Triton X-100, and En3Hance were from New England Nuclear. Aprotinin, phenylmethylsulfonyl fluoride (PhMeSO2F), and N-acetyl-D-glucosamine were obtained from Sigma; bacitracin was from Calbiochem; wheat germ agglutinin coupled to agarose was from Miles. The mixed glycosidases, which are a mixture of exo- and endoglycosidases isolated from Staphylococcus pneumoniae, were a gift from G. Ashwell. (It should be noted that no protease activity could be detected in this preparation of glycosidases.) All reagents for NaDodSOJ polyacrylamide gel electrophoresis were from Bio-Rad. The sera from patients with autoantibodies against the insulin receptor and from normal volunteers were obtained after an overnight fast, heated at 56°C for 30 min, and then stored at -20°C until used. IgGs were purified from pooled control serum and from sera obtained from patients with anti-receptor antibodies by acid elution from staphylococcal protein A-Sepharose (16). The patients with autoantibodies were designated B-2, B-5, B6, B-8, as described (15). Methods. Biosynthetic labeling of the cells: Human lymphocytes (IM-9 line) were grown in continuous suspension culture at 37°C in Eagle's minimal essential medium containing glutamine (0.29 mg/ml) and 10% fetal calf serum (17). When the cells had reached the stationary phase of growth, they were spun down (600 x g) and washed twice with methionine-free Eagle's minimal essential medium containing glutamine (0.29 mg/ml) and 10% dialyzed fetal calf serum. They were then resuspended in the same medium at 10 times their original concentration (5 x 107 cells per ml). For the labeling studies 5 mCi of [3S]methionine was added to flasks containing 100 ml of cell suspension for 14 hr at 37°C. After 14 hr the cells were spun down (600 X g) and washed twice with ice-cold Eagle's minimal essential medium. The viability of the cells was assessed by Trypan blue exclusion and was always greater than 90%.

Insulin receptors have been studied intensively by direct binding techniques and have been identified on cells of all vertebrates examined thus far (1, 2). Although the binding interaction between the receptor and insulin is well characterized, its molecular basis is poorly understood and only recently has some progress been made toward elucidating the subunit structure of the receptor (3-11). At present, with the exception of one study in which heavy isotopes were incorporated into the receptor (12), no direct information is available concerning the synthesis of the receptor, the assembly of its subunits, its intracellular transport, or its insertion into the cell membrane. In order to approach these latter questions we have, in the present study, labeled the insulin receptor biosynthetically with [3S]methionine and identified its subunit structure by taking advantage of specific autoantibodies to the receptor found in rare patients with extreme insulin resistance and acanthosis nigricans (13-15). Using this technique, we have been able not only to identify the major receptor subunits but also to show that insulin-induced loss of the insulin receptor, which produces a marked diminution in the binding of '"I-labeled insulin

Abbreviations: PhMeSO2F, phenylmethylsulfonyl fluoride; kDal, kilodalton(s). t Present address: Department of Experimental Medicine, Group Institut National de la Sante et de la Recherche Medicale U-145, Faculty of Medicine (Pasteur), Chemin de Vallombrose, 06034 Nice Cedex,

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.

France. § Present address: The Endocrine Laboratory, The Royal Melbourne Hospital, P. 0. 3050, Victoria, Australia.

1052

Cell Biology: Van Obberghen et aL Purification of solubilized insulin receptors: Washed cells were solubilized for 60 min at 240C in 50 mM Hepes buffer containing 1% Triton X-100, PhMeSO2F (1 mM), bacitracin (100 units/ml), and aprotinin (1000 trypsin inhibitor units/ml). This preparation was then centrifuged at 100,000 x gfor 90 min at 40C, and the insoluble pellet was discarded. The supernatant was applied to a wheat germ agglutinin-agarose column, the column was extensively washed, and bound glycoproteins were desorbed by using 0.3 M N-acetylglucosamine. This immobilized lectin allows a 20-fold purification with nearly 100% recovery of the insulin receptor as determined by 125I-insulin binding (18). Typically, 5 ml of N-acetylglucosamine was used and five 1-ml fractions were collected. The two fractions with the most radioactivity were pooled and bacitracin (100 units/ ml), PhMeSO2F (1 mM), and aprotinin (1000 trypsin inhibitor units/ml) were again added. over 95% of the radioactive material in the pooled fractions was precipitable by 5% trichloroacetic acid, indicating that the 35S was incorporated into proteins. The pooled fractions were used in the immunoprecipitation experiments. Immunoprecipitation of the insulin receptor: Immunoprecipitation of solubilized insulin receptors was performed according to the method previously described (19). In brief: (i) The solubilized receptors were incubated with normal serum or serum containing autoantibodies against the insulin receptor for 6 hr at 4°C. In one experiment normal IgGs from different patients were used instead of serum. (ii) Immunoprecipitation was then effected by addition of a slight excess of second antibody (sheep antiserum to human IgG; titer of 2 mg/ml). After 6 hr at 40C, the suspension was centrifuged for 5 min at 500 X g, and the pellet was washed twice with 50 mM Hepes buffer (pH 7.4) containing 0.1% Triton X-100. In some experiments incubation with anti-receptor serum was preceded by an incubation of the solubilized receptors with an excess of unlabeled insulin (1.7 ,uM) for 6 hr at 40C. Thereafter, the samples were treated as in i and ii. The immunoprecipitates were solubilized and boiled (3 min) in 10 mM sodium phosphate buffer containing 2% NaDodSO4, 100 mM dithiothreitol or 2% (vol/vol) 2-mercaptoethanol, and 0.01% bromophenol blue. Analysis of the immunoprecipitates: Aliquots of the solubilized immunoprecipitates were analyzed by one-dimensional NaDodSO4polyacrylamide slab gel electrophoresis using a discontinuous buffer system (20) with a 3% stacking gel and either a 7.5% resolving gel or a 5-15% linear gradient of acrylamide. The gels were stained with Coomassie blue (0.25%) dissolved in 50% (wt/vol) trichloroacetic acid, destained in 7% (wt/vol) acetic acid, fluorographed by using an autoradiography enhancer solution (En3Hance), and vacuum dried. Autoradiography was carried out by exposing the gel to Kodak X-Omat film. Quantitative analyses of labeled bands were performed by scanning the film in a Joyce-Loebl microdensitometer and measuring the peak area in arbitrary units. The molecular weights of the standards used in the 5-15% linear gradient of acrylamide gels were: myosin, 200,000; (3galactosidase, 116,000; phosphorylase B, 94,000; bovine serum albumin, 67,000; ovalbumin, 43,000; and carbonic anhydrase, 30,000. On the 7.5% polyacrylamide gels the standards were: filamin, 250,000; myosin; P( and 2 subunits of RNA polymerase, 165,000 and 155,000; phosphorylase B; bovine serum albumin; and ovalbumin. Down regulation of insulin receptors: Cells were incubated in methionine-free Eagle's minimal essential medium containing [35S]methionine (5 mCi/100 ml of cell suspension) and unlabeled insulin (0.1 p.M) for 12 hr at 37C. Thereafter, the cells were spun down (600 x g), washed once with sodium phosphate

Proc. Nati Acad. Sci. USA 78 (1981)

1053

buffer, pH 6.0, containing 1% bovine serum albumin, and then washed three times with growth medium over a period of 2 hr at room temperature. This wash technique had been shown to remove residual bound insulin (21). The washed cells were resuspended in 50 mM Hepes buffer, samples were taken for measurement of '"I-insulin binding, and the remaining cells were solubilized and further processed as described above. The binding of l"I-insulin to control cells and down-regulated cells was measured as described (17). RESULTS Immunoprecipitation of 35S-Labeled Cell Proteins. [3S]Methionine was taken up and incorporated into proteins of IM-9 lymphocytes over a 14-hr period at 37C. More than 90% of the cell-associated radioactivity was incorporated into proteins as measured by precipitability in 5% trichloroacetic acid. Of the 50 gCi/ml added, 27% was present in cellular proteins at the end of the labeling period, corresponding to approximately 40 X 106 cpm per 106 cells. On the basis of theoretical calculations, of this 0.01-0.05% would be in the insulin receptor, if one assumes uniform distribution of the label. To identify the labeled receptor, the 'S-proteins were first solubilized and purified by affinity chromatography on wheat germ agglutinin-agarose, because in previous studies we have found that this step alone produces a 20-fold enrichment of insulin-

binding activity (18). The glycoprotein fraction that had been eluted from wheat germ agglutinin was then immunoprecipitated by using both A

0-

B

C

D

E

" "T,

kDal 200 -

116-

94-

q

-

3

67-

43- ! 30 -

kDal 130 90

~

WI

_ sw

FIG. 1. Immunoprecipitation of wS-labeled subunits of the insulin receptor. Cells were labeled with [MSlmethionine and solubilized with Triton X-100. The insulin receptors were purified on wheat germ agglutinin-agarose and then immunoprecipitated. After solubilization the immunoprecipitates were electrophoresed in a 5-15% linear acrylamide gradient. The gels were dried and subjected to autoradiography. The figure shows an autoradiograph of the gel. The different lanes correspond to the following immunoprecipitation conditions: A, normal serum (1:800 dilution or IgG at 25 lg/ml); B, anti-receptor serum B2 (1:800 dilution or IgG at 25 W/ml); C, preincubation with insulin (1.7 pM) followed by immunoprecipitation with anti-receptor serum B-2 (1:800 dilution or IgG at 25 jLg/ml); D, normal serum (1:400 dilution or IgG at 50 pg/ml); E, anti-receptor serum B-2 (1:400 dilution or IgG at 50 ;Lg/ml). 0, origin.

1054

Proc. Nad Acad. Sci. USA 78 (1981)

Cell Biology: Van Obberghen et aL

normal serum and anti-receptor serum B-2 and subjected to NaDodSO4polyacrylamide gel electrophoresis under reducing conditions in a 5-15% linear gradient of acrylamide. The labeled protein bands were identified by autoradiography. When the solubilized labeled glycoproteins were immunoprecipitated with serum from a normal individual, a few minor bands were observed ranging in molecular mass from 25 to 150 kilodaltons (kDal) (Fig. 1, lane A). Immunoprecipitation with serum (B-2) containing autoantibodies against the insulin receptor revealed four additional bands on the autoradiographs: two major bands of apparent molecular mass 130 and 90 kDal, and two minor bands of molecular mass 200 and 70 kDal (Fig. 1, lane B). Further evidence suggests that both the 130- and the 90-kDal bands, and possibly the 200-kDal band, are the components of the insulin receptor. We have previously shown that insulin bound to receptor reduces anti-receptor binding to and immunoprecipitation of the insulin receptor (14, 19). Preincubation of solubilized [3S]methionine-labeled glycoproteins with an excess of unlabeled insulin (1.7 ,uM) for 6 hr prior to incubation with anti-receptor serum markedly decreased the appearance of both the 130- and the 90-kDal bands (Fig. 1, lanes B and C). Quantitative scanning of the autoradiographs reveals that preincubation with insulin caused decreases of 58% and 35% in the area of the 130- and 90-kDal bands, respectively (Fig. 2), and a 25% decrease in the 200-kDal band (Table 1). Insulin was without significant effect on the other bands (Table 1). Conversely, a higher concentration of anti-receptor antibodies (50 ,ug/ml) produced a moderate increase in the precipitation of the 130- and 90-kDal bands (Fig. 1, lanes B and E, and Fig. 2). This concentration of anti-receptor antibodies and normal IgG was used in the subsequent experiments. Note that the 3S specifically immunoprecipitated by the anti-receptor serum represent 0.01-0.02% of the cell-associated 3S radioactivity, which is of the same order of magnitude as predicted from theoretical considerations. Immunoprecipitation of 3'S-Labeled Cell Proteins with Autoantibodies Against the Insulin Receptor Derived from Different Patients. Autoantibodies against the insulin receptor are found in very rare patients with extreme insulin resistance and acanthosis nigricans (14, 15). It was therefore important to

:1 ~ 1 90 kDal

.10

0 =1

-6

2 50 L .!9 Od Q

.-4

4) ;..4

V- 130

kDal-n

Table 1. Effect of preincubation with insulin on the precipitation of 'S-labeled protein by receptor antibodies 35S precipitated Without With Molecular preincubation preincubation mass of Ratio with insulin with insulin protein, kDal 1.53 90 79.8 52.2 2.4 130 27.3 11.4 1.3 200 2.0 1.5 70 4.1 1.02 4.0 60 16.8 1.03 16.3 43 61.4 1.17 52.0 25 7.1 0.88 8.1

The data were derived from scanning densitometry of the autoradiographs shown in Fig. 1, lanes B and C. The surface area of the peak corresponding to the different bands was expressed in arbitrary units.

ascertain whether the antibodies from different patients with the same syndrome would precipitate the same 'S-labeled cell proteins. Fig. 3 is an autoradiograph of a 7.5% polyacrylamide gel after electrophoresis of 'S-labeled cell proteins immunoprecipitated with normal IgG (lane A), and with anti-receptor IgG obtained from four different patients with the syndrome (lanes B-E). Compared to the normal IgG, the four anti-receptor IgGs precipitated the same two major bands (130 and 90 kDal). Furthermore, in all cases the 90-kDal band was more intense than the 130-kDal. Thus all four antisera recognized the same labeled cellular proteins. In this experiment, the intensity of the minor 200- and 70-kDal bands appeared to be similar with both anti-receptor IgG and normal IgG, and thus it is unclear to what extent, if any, they contain proteins that are specifically immunoprecipitated by the various antisera. Biosynthetic Labeling During Down Regulation of the Insulin Receptor. Incubation in vitro of IM-9 lymphocytes, as well as other cells, with insulin has been shown to result in a A

0

B I

-

kDal 250 200 165 83 * * I155 ¢r1

C

D

E

* %

kDal

-

..

94 -

i f w

1#,. --4 ,^

_

*-

t b0__0

go

67 -

43-

co ko CO

II.

OL

Anti-R, /g/'ml: 50 Insulin, pg/'ml: 0

25 0

25 10

50

25

25

0

0

10

FIG. 2. Quantitative analysis of the 3S-labeled insulin receptor subunits immunoprecipitated by autoantibodies against the insulin receptor (Anti-R). The data are derived from the autoradiograph shown in Fig. 1. The autoradiographs were scanned, and the peak areas corresponding to the 130- and 90-kDal bands were calculated and expressed in arbitrary units.

FIG. 3. Immunoprecipitation of 35S-labeled components of the insulin receptor by different anti-receptor IgGs. Cells were labeled with [85S]methionine and solubilized with Triton X-100, and the insulin receptors were purified and immunoprecipitated. The figure shows an autoradiograph of a 7.5% polyacrylamide slab gel. The different lanes correspond to the following immunoprecipitation conditions: A, normal IgG; B, anti-receptor IgG from patient B-2; C, anti-receptor IgG from patient B-5; D, anti-receptor IgG from patient B-6; E, anti-receptor IgG from patient B-8. All the IgGs were present at the same final concentration, 120 pyg/ml.

Proc. NatL Acad. Sci. USA 78 (1981)

CeH Biology: Van Obberghen et aL A

0-

B

C

D

A

kDal 250 - 200

kDal 200 -

-

W

_

116-

B

.

W

w

__

1055

.

.,-.

_O.

_ _-_

kDal 130

' "PR

67--a

kDal 130

-

165 155

-

94

-

67

-

43

a-

43-

30-

FIG. 4. Effect of down regulation of insulin receptors on the 35Slabeling of the insulin receptor subunits. Cells were down regulated with insulin and processed as described under Materials and Methods. An autoradiograph of a 5-15% acrylamide linear gradient slab gel is shown. The different lanes of the gel are the following: A, down-regulated cells, immunoprecipitation with normal serum; B, down-regulated cells, immunoprecipitation with anti-receptor serum B-2; C, control cells, immunoprecipitation with normal serum; D, control cells, immunoprecipitation with anti-receptor serum B-2.

loss of insulin receptors as measured by insulin binding, a process that has been termed "down regulation." To investigate the effects of down regulation on receptor subunits, IM-9 lymphocytes were incubated for 12 hr with 0.1 uM insulin together with [3S]methionine for receptor labeling. These cells showed an 85% decrease in the concentration of cell surface insulin receptors by insulin-binding studies when compared to control cells grown in the absence of insulin (data not shown). After down regulation, solubilization, and wheat germ purification, precipitation by anti-receptor serum (B-2) failed to reveal either the 130-kDal or the 90-kDal band (Fig. 4). In fact, after down regulation the autoradiographs of the immunoprecipitates with normal serum and anti-receptor serum were indistinguishable. Note that the radioactivity incorporated into total cellular proteins was the same in control cells and down-regulated cells. Effect of Glycosidase Treatment on Mobility of Subunits. Although both major receptor subunits were observed after lectin chromatography, it was not clear if both are in fact glycoproteins or if one is simply copurified with the other due to strong interactions that persist in the Triton-solubilized material. To study this question, the solubilized labeled glycoproteins desorbed from wheat germ agglutinin agarose were treated with mixed glycosidases and then immunoprecipitated with serum containing anti-receptor antibodies. Both major subunits, after glycosidase exposure, migrated faster than their counterparts not treated with glycosidases; the 130- and 90-kDal bands of the untreated sample corresponded to bands with molecular masses of approximately 120 and 80 kDal, respectively (Fig. 5). These findings indicate that both the 130- and the 90-kDal bands are glycoproteins.

DISCUSSION It is now widely accepted that the first step in insulin action is binding to specific receptors in target cell surface membranes

FIG. 5. Effect of glycosidase treatment on mobility of subunits. Solubilized labeled glycoproteins desorbed from wheat germ agglutinin-agarose were incubated in the absence (lane A) and presence (lane B) of mixed glycosidases overnight at 37°C. The untreated and treated samples were immunoprecipitated with anti-receptor serum B-2 and subjected to electrophoresis in a NaDodS04/7.5% polyacrylamide gel. The figure shows an autoradiograph of the gel.

(1, 2). Although the interaction of insulin with its receptor has been well characterized in both intact cells and membrane preparations, very little is known concerning the exact molecular mechanism of the interaction, because purification and direct chemical characterization of the receptor have been difficult. Recently we, and others, have obtained data on the structure of the insulin receptor by using NaDodSO4polyacrylamide gel electrophoresis after purification on immobilized insulin (7) or immobilized antibody (10), after covalent labeling of the binding site with 125I-insulin (5, 6, 11), and after precipitation of "2I-labeled membrane proteins by receptor antibody (9). The advance in the present study is the demonstration of biosynthetic labeling of receptor subunits. Using [3S]methionine labeling and specific immunoprecipitation of the insulin receptor, we find two major subunits with molecular weights of 130,000 and 90,000, and possibly two minor components of 200,000 and 70,000. A number of antireceptor sera, obtained from patients with the syndrome of severe insulin resistance and acanthosis nigricans, have previously been shown by various criteria to be specific for the insulin receptor (13-15); all these sera, but not control sera, precipitate two major 3S-labeled lymphocyte proteins with molecular weights of 130,000 and 90,000. The quantitative precipitation of both subunits correlates well with the relative titers of autoantibodies found in the four sera used (15); the higher the titer the more intense the 130- and 90-kDal bands on autoradiography. The present finding of a 130-kDal subunit agrees with that identified by others as an insulin-binding site(s), but in addition we clearly identify a major 90-kDal subunit. There could be at least three possible explanations for the observation that autoantibodies against the receptor precipitate both the 130- and the 90-kDal proteins. First, assuming that

1056

Proc. Natl. Acad. Sci. USA 78 (1981) Cell Biology: Van Obberghen et aLPrcNaLAd.SiUA78(91

there exist strong interactions between the 90- and the 130-kDal species, the binding of an antibody against either species could precipitate both in the native state. Second, it is possible that the 130- and 90-kDal species have common antigenic determinants, which are recognized by a single antibody population. Finally, the sera are polyclonal in nature and may contain different antibody populations, which separately recognize the two species. The two first possibilities seem the most likely, on the basis of the observation that sera from four different patients with anti-receptor antibodies all precipitate the same two components. A number of lines of evidence support the idea that the two major bands are specific subunits of the insulin receptor and that one, if not both, of them contains the binding site(s) for insulin. First, when the solubilized material was exposed to an excess of unlabeled insulin and thereafter to anti-receptor antibodies, there was, by both inspection and quantitative scanning, a significant decrease in the precipitation of both the 130and the 90-kDal bands, as would be expected from previous studies which showed that insulin bound to receptor reduced the affinity of receptor for anti-receptor antibody. Second, both 130- and 90-kDal bands were observed when sera from four different patients with autoantibodies against the insulin receptor were used. Third, the decrease in 125I-insulin binding seen after incubation of the cells with insulin (down regulation) was associated with a concomitant decrease in labeling of both the 130- and the 90-kDal bands. It is well established that the phenomenon of down regulation is due to regulatory events in the cell, rather than occupancy of the receptors by insulin, and is highly specific-i.e., in these lymphocytes insulin will down regulate only its own receptors and will not affect, for example, receptors for growth hormone (21). Consistent with this exclusive effect is the observation in the down-regulated cells that only the 130- and 90-kDal bands disappear, whereas other bands representing other labeled cell proteins remain unchanged. In the last couple of years different investigators using a variety of techniques have identified in NaDodSO,4polyacrylamide gel electrophoresis an insulin-binding subunit with a molecular mass of 125 to 135 kDal, but there is much less consensus concerning the existence of smaller subunits of the receptor (3-11). For example, Pilch and Czech (5, 6) and Kasuga et aL (11), using a chemical crosslinking agent to affinity label the receptor, found very heavy labeling of a 125- to 135-kDal protein, but none or minimal labeling of a 90-kDal band. In a partially purified receptor, Jacobs et aL (7) have also found that the major subunit is about 130 kDal, although minor components with lower molecular mass exist. By contrast, we have identified specific 130- and 90-kDal receptor subunits after labeling of cell surface proteins with "~I and immunoprecipitation with antibodies to the insulin receptor (unpublished data), and Yip et aL have observed a 90-kDal subunit of the receptor after photoaffinity labeling (4). These data suggest that, de-

pending on the method of labeling, the 90-kDal subunit is more

less "accessible" to detection. It also raises the question as to whether both the 130- and the 90-kDal components are binding sites for insulin (but one has more available reactive groups), or alternatively, whether only one of these (the 130-kDal) is the binding site and the other a closely associated protein. Due to the presence of three inhibitors of proteolysis (PhMeSO2F,

or

bacitracin, and aprotinin) throughout the experimental procedures we think that it is unlikely that the 90-kDal component is a degradation product derived from the 130-kDal component, although definitive proof of this is lacking at the-present time. Preliminary results of peptide mapping of both bands show major differences in these two proteins.

Finally, the results of this study add further insights into the mechanisms of down regulation. First, they show that down regulation cannot be accounted for simply by inactivation of the binding function in situ but must involve a loss of total cellular receptors. Second, the present observations are not consistent with the notion that a redistribution of receptors intracellularly or a block in their recycling plays a major role in down regulation. Thus, the decreased labeling of both components of the insulin receptor during down regulation is most consistent with decreased synthesis or increased degradation of the receptors or both. In conclusion, we have biosynthetically labeled the insulin receptor with ['Slmethionine and identified its subunit structure. The insulin receptor consists of two major subunits with apparent molecular weights of 130,000 and 90,000. Evidence suggests that the 130,000 species contains the binding site(s) for insulin; the 90,000 species is an authentic subunit but its precise function remains to be elucidated. The ability to biosynthetically label the insulin receptor adds a dimension to the study of receptor structure, function, and turnover not only under physiologic conditions but, more importantly, in disease states with altered insulin action. We are indebted to Drs. Jesse Roth, C. Ronald Kahn, and Kenneth Yamada for helpful comments and criticisms. We thank Ms. Carol Guiwell for her excellent secretarial assistance. E.VO. was supported by the Kroc Foundation, A. L.C. was the recipient of a Fogarty International Fellowship (FO5 TW02597-02), and M.K. is the recipient of a Fogarty International Fellowship (FO5 TW02818) while on leave from the Third Department of Internal Medicine, University of Tokyo,

Japan.

1. Kahn, C. R. (1976)1. Cell Biol. 70, 261-286. 2. Roth, J. (1979) in Endocrinology, ed. De Groot, L. (Grune & Stratton, New York), Vol. 3, pp. 2037-2054. 3. Yip, C. C., Yeung, C. W. T. & Moule, M. L. (1978) J. Biol. Chem. 253, 1743-1745. 4. Yip, C. C., Yeung, C. W. T. & Moule, M. L. (1980) Biochemistry 19, 70-76. 5. Pilch, P. F. & Czech, M. P. (1979) J. Biol. Chem. 254, 3375-338 1. 6. Pilch, P. F. & Czech, M. P. (1980) J Biol. Chem. 255, 1722-1731. 7. Jacobs, S., Hazumn, E., Shechter, Y. & Cuatrecasas, P. (1979) Proc. Natl. Acad. Sci. USA 76, 4918-4921. 8. Wisher, M. H., Baron, M. D., Jones, R. H., Sonksen, P. H., Saunders, D. J., Thamm, P. & Brandenburg, D. (1980) Biochem. Biophys. lRes. Commun. 92, 492-498. 9. Lang, U., Kahn, C. R. & Harrison, L. C. (1980) Biochemistry 19, 64-70. 10. Harrison, L. C. & Itin, A. (1980)J. Bol. Chem. 255, 12066-12072. 11. Kasuga, M., Van Obberghen, E., Harrison, L. C. & Yamada, K. (1981) Diabetes, in press. 12. Reed, B. C. & Lane, D. M. (1980) Proc. Natl. Acad. Sci. USA 77, 285-289. 13. Flier, J. S., Kahn, C. R., Roth, J. & Bar, R. 5. (1975) Science 190, 63-65. 14. Kahn, C. R., Flier, J. S., Bar, R. S., Archer, J. A., Gorden, P., Martin, M. M. & Roth, J. (1976) N. Engl. J. Med. 294, 739-745. 15. Kahn, C. R. & Harrison, L. C. (1981) in Carbohydrate Metabolism and Its Disorders, eds. Randle, P. J., Steiner, D. F. & Whelan, W. J.(Academic, London), in press. 16. Goding, J. W. (1978)1J. Immunol. Methods 20, 241-253. 17. De Meyts, P. (1976) in Methods in Receptor Research, ed. Blecher, M. (Dekker, New York), pp. 301-383. 18. Hedo, J. A., Harrison, L. C. & Roth, J. (1980) Diabetes Suppl. 2, 29, 38A (abstr.). 19. Harrison, L. C., Flier, J. S., Roth, J., Karlsson, F. A. & Kahn, C. R. (1979) J. Clin. EndocrinoL Metab. 48, 59-65. 20. Laemmli, U. K. (1970) Nature (London) 227, 680-685. 21. Kosmakos, F. C. &Roth,J. (1980)J BioL Chem. 255, 9860-9869.