ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 252, No. 2, February 1, pp. 409-417,1987
Further Studies on the Covalent Crosslinking of Thyrotropin to Its Receptor: Evidence That Both the cxand 0 Subunits of Thyrotropin Are Crosslinked to the Receptor’ ROBERT
McQUADE,*p2
*Department
COLIN
of Biochemistry University
G. THOMAS,
and Nutrition Carolina,
AND
and fDepartment Chapel Hill, North
of North
Received
JR.,?
August
SHIHADEH of Surgery, Carolina
N. NAYFEH*r3
School 27514
of Medicine,
81986
Highly purified (Y- and P-subunits of thyrotropin were individually radioiodinated and, subsequently, recombined with their unlabeled complementary subunits. This procedure resulted in the formation of [‘251]thyrotropin(TSH) hybrid molecules which were labeled on only one hormone subunit. Characterization of the binding properties of these two hybrid molecules demonstrated that both yielded nonlinear Scatchard plots with Kd and B,,, values similar to those obtained with radioiodinated native TSH and that both were capable of interaction with the high- and low-affinity binding components of the TSH receptor. The recombined [‘251]TSH molecules were then crosslinked to the TSH receptor using disuccinimidyl suberate. Following electrophoresis and autoradiography, two labeled TSH-receptor complexes with M, of 68,000 and 80,000 were observed. These two complexes exhibited hormone specificity and electrophoretic mobility identical to those previously observed using native [‘?]TSH. Crosslinking with increasing concentrations of disuccinimidyl suberate suggested that the formation of the 68,000 and 80,000 complexes was sequential with the 68,000 appearing before the 80,000. Furthermore, the two bands were labeled regardless of which TSH subunit of the hybrid TSH was radioiodinated. These data strongly suggest that the 68,000 and 80,000 TSH-receptor complexes are the result of crosslinking to the TSH a-/3 dimer and not to one subunit in the case of the 68,000 complex and to the TSH a-/3 dimer in the case of the 80,000 complex, as had been hypothesized previously. o 1987 Academic press, I,,~. -
TSH molecule possesses an approximate molecular weight of 28,000 and is composed of two nonidentical subunits, (Y and /3, with molecular weights of 13,600 and 14,700, respectively (for review, see Ref. (1)). It has been demonstrated that within a single species, the TSH-a! subunit is identical or very similar to the (Y subunits of follitropin (FSH), lutropin (LH), and human chorionic gonadotropin (hCG). The TSH-fl subunit, although somewhat similar to the p subunits of the other three hormones, is unique
Thyrotropin (TSH)4 is a glycoprotein hormone, secreted by the anterior pituitary, which regulates thyroid function. The ‘This work was supported by U.S. Public Health Service Grants AM-23080 and CA-01915. ‘Present address: Schering-Plough Corporation, 60 Orange St., Bloomfield, NJ 07003. ’ To whom all correspondence should be addressed at 402A Faculty Laboratory Office Bldg. 231H, Department of Biochemistry, University of North Carolina, Chapel Hill, NC 27514. 4 Abbreviations used: TSH, thyrotropin; FSH, follitropin; hCG, human chorionic gonadotropin; LH, Intropin; DSS, disuccinimidyl suberate; DTT, dithiothreitol; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; DFP, diisopropylfluorophosphate; PMSF, phenylmethylsulfonyl fluo-
ride; STI, membranes; solubilized 409
soybean trypsin PPM, partially membranes. 0003-9861/8’7 Copyright All rights
inhibitor; PM, purified purified membranes; SM,
$3.00
Q 1987 by Academic Press, Inc. of reproduction in any form reserved.
410
MC
QUADE,
THOMAS,
to TSH. Neither the TSH-a nor the TSH@subunits possess any biological activity; the TSH molecule is only active when in the LY-/~dimer configuration. Therefore, although the TSH-/I subunit confers hormone specificity to the biological and immunological activities of TSH, the TSH-CY subunit must also be present in order to initiate action. The initial step in the regulation of thyroid function by TSH is the binding of TSH to a membrane-bound receptor. Several laboratories have recently attempted to examine the binding components of the TSH receptor using photoaffinity (2,3) and crosslinking (4,5) labeling techniques. Results from these studies demonstrated the formation of several complexes with molecular weights ranging between 30,000 and 250,000. Using the bifunctional crosslinker, disuccinimidyl suberate (DSS), we have recently shown that [‘251]TSH is crosslinked to two labeled complexes with molecular weights of 68,000 and 80,000 (5). The present study was initiated to determine whether these two complexes represent two distinct subunits of the receptor, each crosslinked to the TSH c+3 dimer, or simply represent the same receptor subunit (52,000-56,000) crosslinked to either the cy or the @subunit of TSH (68,000) or the TSH o+ dimer (80,000). Highly purified TSH-(U and -/3 subunits have been individually radioiodinated and recombined with their unlabeled complementary subunit to yield TSH hybrid molecules possessing only one labeled subunit. These TSH hybrid molecules were then utilized in binding and crosslinking experiments in an attempt to determine which TSH hybrid (a*-@ and or CP@*) is bound to thyroid plasma membranes and which yields the crosslinked 68,000 and 80,000 complexes. EXPERIMENTAL
PROCEDURES
Materials. Purified bovine TSH (40 IU/mg), as well as its (Y and B subunits, was a generous gift from Dr. John Pierce (UCLA). Purified TSH was iodinated with izsI (Amersham IM-30) to a specific radioactivity of 50-90 &i/pg by a modification of the lactoperoxidase method, as described previously (6). Partially purified TSH (TS-10) used in competitive binding assays was purchased from the Sigma Chemical Co. (St. Louis,
AND
NAYFEH
MO). Porcine thyroids were purchased from the PelFreeze Co. (Rogers, AR). Disuccinimidyl suberate was obtained from the Pierce Chemical Co. (Rockford, IL). All other reagents and chemicals were of the highest quality commercially available. Raditiodination of (Y- and (3-subunits of TSH and fomnation of TSH hybrids. Radioiodinated TSH and hybrid molecules labeled in only one subunit were prepared. Purified TSH-a and $3 subunits were incubated with ‘%I (1 mCi) in the presence of lactoperoxidase (0.1 mg/ml) and HzOz (0.03%) for 4 min and 40 s, respectively. The iodination was terminated by the addition of 0.1 M NaNa. TSH hybrids were made by a modification of the procedure of Sharp and Pierce (1’7). One labeled subunit was incubated with its unlabeled complementary subunit (10 fig) under Nz for 48 h at 37°C. The recombined TSH molecules were subsequently separated from unreacted subunits and free ‘%I by gel filtration chromatography on Sephadex G-100 at 4°C. The columns were eluted with 10 mM phosphate buffer/0.25% BSA, pH 7.4. Preparation of pwised and solutilized mwnbranes. Partially purified membranes (PPM) were prepared as previously described (5,6), except for the presence of protease inhibitors, 2.0 mM diisopropyl fluorophosphate (DFP), 1 mM phenylmethylsulfonyl fluoride (PMSF), and 0.1 mg/ml soybean trypsin inhibitor (STI), and were stored at -90°C. The addition of protease inhibitors was found necessary to prevent degradation of the labeled receptor bands. PPM were subsequently purified on discontinuous sucrose density gradients, as previously described (5). Purified membrane pellets (PM) were resuspended in 10 mM sodium phosphate buffer, pH 7.4, containing 0.5 mM DFP, 1 mM PMSF, and 0.1 mg/ml ST1 (Buffer A) at an approximate concentration of 1 mg/ml, and were stored at -90°C. Exact protein concentrations were determined by the method of Lowry et al, using bovine serum albumin as a standard (8). Solubilized membranes were prepared by incubation with 1% Triton X-100 in Buffer A for 1 h at room temperature (9), followed by centrifugation at 100,OOOg for 30 min in a Beckman SW 28 rotor at 4°C. Solubilized membranes (SM) were stored at -90°C at an approximate protein concentration of l-2 mg/ml. Protein concentrations were determined by the Markwell modification of the Lowry method (lo), which minimized the effects of Triton X-100 on color development. Binding and cross-linking of [‘p5I]TSH to pur$ed membranes. Binding of [iZIJTSH to thyroid PM was carried out as described previously (5). Radioiodinated TSH was covalently linked to PM by a modification of the disuccinimidyl suberate method of Pilch and Czech (ll), as described previously (5). PM (0.5 mg) were resuspended in Buffer A and labeled by incubation for 1 h at room temperature with [‘2bI]TSH (0.5-1.0 X lo6 cpm) in the presence or absence of excess
CROSSLINKING
OF
TSH
01 AND
unlabeled TSH. Unbound TSH was removed by centrifugation, and the membrane pellet was resuspended in 0.2 ml of cold 10 mM sodium phosphate, pH 7.4, containing 0.5 mM DFP, and 0.1 mM PMSF. Bound [?]TSH was covalently crosslinked by incubation of the labeled membranes for 15 min as described previously (5). Noncovalently bound TSH was dissociated by incubation for 1 h in 25 mM Tris-acetate buffer, pH 7.4, containing 1 M MgCl*. Following crosslinking, labeled membranes were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on 7.5 or 10% gels according to the method of Laemmli (12). Following electrophoresis, the gels were stained with Coomassie brilliant blue G, and labeled proteins were visualized by autoradiography on Kodak X-Omat XAE-5 X-ray film, using DuPont Cronex Lightning Plus intensifying screens.
12.0
r1
I
0 SUBUNITS
TO
TSH
RECEPTOR
411
RESULTS
Resolution Hybrids
of Radioiodinated TSH and Their Suhnits
Following radioiodination and recombination, the recombined TSH molecules were then separated from the remaining subunits by gel filtration chromatography on Sephadex G-100. Chromatography of the (Y,radioiodinated 0 (p*) recombination mixture resulted in the appearance of three distinct peaks of radioactivity (Fig. 1). The largest of these peaks had the same elution volume as that of native [lz51]TSH and was presumed to be the TSH hybrid, in which the p subunit was labeled (a-p*). The sec-
I
I
I
I
L
0 0 .
IO.0
FRACTION
TSH ag’ a*B
NUMBER
FIG. 1. Gel filtration chromatography of radioiodinated native TSH (0) and TSH hybrids composed of cz-[‘“I]/~(w@*) (0) and [‘%I]cu-P(a*-6) (A). Samples of native TSH and TSH hybrids, radioiodinated and prepared as described in the text, were applied to 1.5 X 72-cm Sephadex G-100 columns. The radioiodinated compounds were then eluted with 10 mM sodium phosphate/0.25% BSA, pH 7.4, at 4°C. Fractions (2 ml) were collected and aliquots (10 ~1) were counted.
412
MC QUADE,
THOMAS,
ond peak eluted after TSH and was presumed to represent free ,L?*.The third peak was free lz51. Chromatography of the radioiodinated CY((Y*),p recombination mixture resulted in the detection of only two peaks of radioactivity (Fig. l), although the larger peak possessed a distinct, reproclucible shoulder eluting before the main peak. This shoulder had the same elution volume as native [?]TSH and was presumed to contain the TSH hybrid in which the CY subunit was labeled a*-@). The primary peak elutecl after TSH and was presumed to be free (Y* (14). The second peak of radioactivity represented free ‘%I. Electrophoretic Characterization Radioiodinated Peptides
AND NAYFEH
TSH-
Pa-
of the
In order to confirm the presumed iclentity of the raclioiodinatecl peptides, each peak was analyzed by SDS-PAGE. Aliquots of each peak were incubated in 2% SDS for 30 min at 25°C and were then analyzed on 10% polyacrylamide gels. The free radioiodinated subunits ((Y* and B*) (Fig. 2, lanes E and F) and native [lZI]TSH (Fig. 2, lane G) were run as standards. Each of the standards migrated as a single band with [lZIJIYSH having the slowest mobility, followed by p* and then (Y*. If [lZI]TSH was boiled in SDS prior to electrophoresis, the subunits dissociated and two bands were observed which comigratecl with the free raclioiodinatecl subunits (data not shown). The radioactivity from the first Sephadex G-100 peak fraction of the (cu-p*) TSH hybrid (column fraction 29) exhibited two bands on SDS-PAGE: the more intensely labeled band comigrated with [‘?]TSH, while the lighter band migrated with the p* subunit (Fig. 2, lane B). The second peak (fraction 36) yielded a major band comigrating with the ,f3* subunit and a minor band which corresponded to [1251]TSH (Fig. 2, lane D). When the peaks from the (a*-@) TSH hybrid were similarly analyzed, the shoulder which preceded the main peak of radioactivity (fraction 29) exhibited a major band with the mobility of [1Z51JTSH and a minor one which corresponded to the a*-subunit (Fig. 2, lane A). The material from this shoulder was used
ABCDEFG
FIG. 2. Autoradiograms of SDS-PAGE gels of radioiodinated TSH and TSH hybrids, (Y*+ and (Y-P*, obtained from Sephadex G-100 chromatography. Aliquots (lOO,OOO-200,000 cpm) of the different radioiodinated peptides, obtained after gel filtration chromatography, were incubated in 2% SDS for 30 min at 25°C and subjected to SDS-PAGE on 10% polyacrylamide gels which were dried and analyzed by autoradiography. The samples were taken from fraction 29 of the radioiodinated native TSH (lane G), fractions 29 (lane A), and 32 (lane C) of the a*+3 hybrid and fractions 23 (lane B) and 36 (lane D) of the (Y-B*hybrid. Radioiodinated LY(lane E) and /3 (lane F) subunits, which had not undergone recombination, were included as standards.
for the binding and crosslinking studies described below. The main peak itself (fraction 32) consisted mainly of the (Y*subunit and a small amount of (a*$) TSH hybrid (Fig. 2, lane C). On the other hand, if the raclioioclinatecl (a*-@) and ((Y-B*) TSH hybrids were boiled in SDS prior to electrophoresis, only one band corresponding to a* or ,L?*,respectively, was seen on the autoradiograph (data not shown). These results demonstrate that the recombination process resulted in radioioclinatecl hybrid molecules which had the same mobility as native [‘?]TSH and that only one subunit was labeled in each hybrid species.
CROSSLINKING
OF
Characterization of the Binding of the TSH Hybrid Molecules
TSH
OL AND
,f3 SUBUNITS
Activity
Studies were next initiated to determine the binding activity of the [lZI]TSH hybrid molecules. Under the present experimental conditions, neither of the labeled free subunits (a* or fl*) exhibited any significant binding activity to thyroid membranes. However, the radioiodinated hybrids were highly active, and in equilibrium saturation analyses were comparable to the radioiodinated native TSH. Both TSH hybrid molecules yielded curvilinear Scatchard plots: the approximate KD values for the a-p* recombinant were 0.7 nM for the highaffinity component and 70 nM for the lowaffinity component (Fig. 3A), while the (Y*0 recombinant gave approximate values of 0.4 and 80 nM (Fig. 3B). These values are similar to those previously reported using native [lz51]TSH (6). To ensure further that the recombinant [“251]TSH hybrid molecules possessed binding properties similar to those of the native hormone, we next examined the ability of these molecules to interact with the highand low-affinity binding components of the TSH receptor. Solubilized thyroid membranes were incubated with a*-@, CY-p*, or native [1251]TSH in the presence or absence of excess unlabeled hormone, and the highand low-affinity binding components were
0.1
0.3
0.5
O.?
0.9
TO
TSH
RECEPTOR
413
separated by centrifugation on discontinuous sucrose density gradients, as described previously (9). In agreement with our previous finding (9), prelabeling solubilized membranes with native [‘251]TSH resulted in the appearance of two specifically bound peaks of radioactivity (Fig. 4). The peak near the top of the gradient represented the high-affinity while the one near the bottom represented the low-affinity binding component. When solubilized membranes were prelabeled with the [‘251]TSH hybrid molecules (a*-B or ry-/3*), the same two peaks were observed regardless of which TSH subunit was radioiodinated (Fig. 4). Taken together, these data demonstrate that the binding properties of the recombined [1251]TSH molecules are virtually indistinguishable from those of the native hormone. Covalent Crosslinking of Radioiodinated TSH Hybrids to the TSH Receptor We have shown previously that the formation of the crosslinked 68,000 and 80,000 TSH-receptor complexes was dependent on the concentration of the crosslinker DSS (15). While both complexes were observed at high concentrations of DSS (0.3-1.0 mM), only the 68,000 complex was detectable at lower concentrations of crosslinker (0.050.010 rnM).
I.1
0.1
0.3
0.5
0.7
0.9
I.1
BOUND(nMI FIG. 3. Equilibrium binding analysis of radioiodinated TSH hybrid molecules. Aliquots (150,000200,000 cpm) of the recombined TSH molecules, cy-@* (A) and n*-fl (B), were incubated with porcine thyroid PPM (40 pg protein) in the presence of increasing concentrations of unlabeled TSH for 1 h at 25°C. Bound hormone was separated from free by ultrafiltration on cellulose-acetate filters and the filters were counted. Data were expressed according to the method of Scatchard (25), as described previously (6).
414
MC
QUADE,
. 0 .
2600
OL
TOP
’
THOMAS,
AND
NAYFEH
TSH Q8’ a’8
I
I
I
I
3
5
7
9
FRACTION NUMBER
I II
BOTTOM
‘FIG. 4. Separation of the high- and low-affinity TSH binding components following labeling with TSH hybrids. Porcine thyroid SM (0.5 mg) were incubated with [12SI]TSH (6 nM) (0) or with the ‘%Ilabeled (approximately 500,000 cpm) TSH hybrids, a-/3* (0) or (Y*-@ (A), in the presence or absence of excess native TSH (1.25 IU) for 1 h at 25°C. The labeled membranes were applied to discontinuous sucrose gradients of 5,10,20, and 50% sucrose and were centrifuged at 221,000g for 2 h at 4°C. The gradients were fractionated from the bottom and bound radioactivity was determined by polyethylene glycol precipitation and filtration on cellulose-acetate filters. Specific bound [1251]TSH was determined by subtracting the number of counts bound in the presence of excess TSH from those bound in the absence of unlabeled hormone.
To determine which of the hormone subunits (a and/or /3) are bound to the 68,000 and 80,000 labeled complexes, attempts were made to crosslink the radioiodinated TSH hybrids (a*-/3 or a-@*) to thyroid plasma membranes using three different concentrations of DSS. Following SDSPAGE and autoradiography, the crosslinked complexes were compared to those observed with native [‘?]TSH. At 0.1 mM DSS, both [1251]TSH hybrid species labeled a single band with an M, of 68,000 (Fig. 5, lanes A-B). This band comigrated with the band labeled by native [?]TSH (Fig. 5, lane C) and was eliminated by incubation
with excess unlabeled hormone (Fig. 5, lane D). Intermediate concentrations of DSS (0.3 mM) again resulted in the appearance of the 68,000 band (Fig. 5, lanes E-G); however, a second band, with M, of 80,000, was also specifically labeled, albeit faintly. These two specifically labeled bands were observed with both ‘%I-labeled hybrid species (a*-@ or cu-p*), as well as with native [‘251Jl”SH. At higher concentrations of DSS (0.5 mM), the two bands were more clearly visible (Fig. 5, lanes I-K), and were again observed with both the recombinant and native TSH molecules. Since both the 68,000 and 80,000 complexes were labeled,
CROSSLINKING
OF
TSH
(Y AND
fl SUBUNITS
TO
TSH
RECEPTOR
415
80,000 bands (data not shown). Additional evidence supporting this conclusion has recently been obtained using the alkaline-cleavable crosslinker bis[2-(succinimidooxycarbonyloxy)ethyl] sulfone (BSOCOES) (see Discussion). DISCUSSION
ABCDEFGHIJKL
FIG. 5. Crosslinking of radioiodinated native and recombined TSH to thyroid PM. Porcine thyroid PM (0.5 mg) were labeled with [‘251]TSH (lanes C-D, GH, and K-L), cu*-fi (lanes A, E, and I) or a-P* (lanes B, F, and J). Nonspecific binding was monitored by the inclusion of excess (1.25 IU) native TSH (lanes D, H, and L). The labeled membranes were then crosslinked with 0.1 mM (lanes A-D), 0.3 mM (lanes E-H) and 0.5 mM (lanes I-L) DSS for 15 min at 25°C. The crosslinked membranes were then washed, boiled in 2% SDS in the presence of 50 mM DTT, and subjected to SDS-PAGE on 7.5% gels. The gels were subsequently dried and autoradiographed.
regardless of which TSH subunit was radioiodinated, these results indicate that both complexes were crosslinked to the TSH cr-@ dimer. Further evidence for this conclusion has come from studies with native [1251]TSH and the thiol-cleavable crosslinking agent dithiobis(succinimidylpropionate) (DSP). This reagent has reactivity towards nucleophilic residues analogous to DSS (16). Crosslinked hormonereceptor complexes were analyzed by twodimensional SDS-PAGE as described previously (17). The DSP was cleaved prior to the second dimension with DTT and although there was a great deal of streaking, the resulting autoradiograph demonstrated that both the (Y and p subunits of TSH were crosslinked to the 68,000 and
In the present study, we have demonstrated that recombination of one radioiodinated subunit of TSH (CY* or ,L3*) with its unlabeled complement (a! or 6) yields radioiodinated TSH hybrid molecules that are labeled in one subunit and are very similar in electrophoretic mobility and binding properties to the native hormone. This is consistent with previous reports showing that appropriate recombination of the isolated CYand 0 subunits of glycoprotein hormones generated hormone molecules that are virtually indistinguishable from the native hormone, viz., receptor binding, adenylate cyclase activation, and cellular response (see Ref. (1) for details). DSS crosslinking of radioiodinated TSH hybrid molecules ((Y*$ or a-p*) to thyroid plasma membranes resulted in the specific labeling of two complexes with molecular weights (68,000 and 80,000) and hormone specificity similar to those obtained with native [1251]TSH (Fig. 4), regardless of which TSH hybrid (a*$ or CY-/3*) was used. These results confirm and extend our previous conclusion (5) that the 68,000 and 80,000 complexes represent two different forms of the TSH receptor that are crosslinked to the TSH a-P dimer, and not the same receptor subunit crosslinked to the TSH LY-fl dimer in the case of the 80,000 and to one of the hormone subunits in the case of the 68,000 species. This conclusion is similar to that reported for the interaction of [‘251]hCG and [1251]FSH with their receptors on granulosa cell membranes (18-20), but differs from results previously reported for the interaction of TSH with its receptor on porcine thyroid plasma membranes (3). Using N-hydroxysuccinimidyl-4-azidobenzoate (HSAB) as a crosslinker, Buckland et al. (3) observed two labeled bands, with M, of 74,000 and 59,000. If the crosslinked complexes were immu-
416
MC
QUADE,
THOMAS,
noprecipitated with anti-TSH-@ antibody, the resulting pellet contained both complexes, indicating the /3subunit was present in both bands. If, however, the crosslinked complexes were immunoprecipitated with anti-TSH-a, the pellet contained only the 74,000 band. These authors (3) interpreted these results as meaning that the (Ysubunit of TSH was present only in the 74,000 band. It was suggested, therefore, that the 74,000 band represented a receptor subunit crosslinked to both cx and 6, while the 59,000 band was the same receptor subunit labeled with only the /3 peptide. Similar studies have recently been attempted by Dr. Gennick in our laboratory, using intact cultured rat thyroid cells (FRTL-5) instead of solubilized porcine thyroid membranes (S. E. Gennick, C. G. Thomas, Jr., and S. N. Nayfeh, manuscript in preparation). Attempts to immunoprecipitate the 68,000 and 80,000 complexes following crosslinking [lz51]TSH to solubilized porcine thyroid membranes have been unsuccessful because of heavy streaking on the gels due to the presence of noncovalently bound [‘251]TSH. In contrast to the results of Buckland et al. (3), both the 68,000 and 80,000 complexes were immunoprecipitated by either anti-TSH-cy or anti-TSH,f3, thus indicating that under our experimental conditions both the (Y and /3 subunits are present in the 68,000 and 80,000 complexes. The cause for this discrepancy is currently unknown. Both laboratories utilized antibodies from the same source (kindly provided by Dr. John G. Pierce, UCLA). The only apparent difference is that Buckland et al. used HSAB while Gennick et al used DSS as a crosslinker. It is possible that HSAB crosslinked TSH to its receptor in such a way as to bury the (Y subunit within the smaller receptor peptide, thus preventing its interaction with anti-TSH-a. A similar proposal was made by Hwang and Menon for the interaction between human chorionic gonadotropin and its receptor (21). The demonstration that the accumulation of both the 68,000 and 80,000 bands is dependent on the concentration of DSS and that the appearance of the 68,000 complex precedes that of the 80,000 complex sug-
AND
NAYFEH
gests that the two crosslinked complexes may be spatially related to one another. This phenomenon is characteristic of incremental crosslinking of oligomeric proteins (22) and is similar to that recently reported for both the hCG/LH and the FSH receptors on granulosa cells (18-20) as well as the prolactin receptor in ovaries (1’7). Using increasing concentrations of the alkaline-cleavable crosslinker BSOCOES in the presence or absence of reducing agents, Ji and co-workers (19) demonstrated the formation of three crosslinked complexes with molecular weights of 65,000, 83,000, and 117,000. Results from studies using two-dimensional SDS-PAGE following cleavage of the crosslinked complexes by alkaline hydrolysis or by reduction with dithiothreitol suggest that these complexes were formed by incremental crosslinking first of FSH cu-fi dimer (43,000) to 22,000 peptide (65,000) which was then crosslinked to 18,000 (83,000) and in turn to a 34,000 peptide (117,000). The receptor is therefore believed to be composed of three disulfidelinked peptides with molecular weights of 22,000, 18,000, and 34,000, and only the 22,000 species was involved in binding to the FSH a-P dimer. Similar results have recently been obtained in our laboratory using BSOCOES to crosslink [?]TSH to its receptor on FRTL-5 rat thyroid cells (S. E. Gennick, C. G. Thomas, Jr., and S. N. Nayfeh, manuscript in preparation). Three complexes with molecular weights of 65,000, 82,000, and 145,000 were formed by incremental crosslinking. Results from cleavage of these complexes with dithiothreitol prior to two-dimensional SDSPAGE suggested that like the FSH receptor, the TSH receptor is composed of three disulfide-linked peptides with molecular weights of 31,000, 17,000, and 63,000, of which only the 31,000 binds to TSH a-/3 dimer. It is unlikely that one or more of the crosslinked complexes so far identified in thyroid, as well as in granulosa cell membranes, resulted from crosslinking of the hormone to membrane components close to, but not part of, the receptor itself, since the same three complexes have been shown to occur in intact cells, crude plasma mem-
CROSSLINKING
OF
TSH
a AND
branes, highly purified membranes, and solubilized membranes (5, 15, M-20). It is also unlikely that they were proteolytic fragments of larger molecules, since proteolysis results in smaller molecular weight bands when membranes are labeled in the absence of protease inhibitors. Another possible explanation for the accumulation of the 80,000 complex in the present study or of the 82,000 and 145,000 complexes in FRTL-5 cells as a function of the concentration of crosslinkers is that they are formed as artifacts resulting from crosslinking more than one TSH subunit or molecule to the receptor. Although previously this has been considered unlikely to occur with the hCG/LH or FSH receptors (18, 19), final resolution of this possibility must await elucidation of the structure of the receptor following purification to homogeneity. The possibility that only one out of two or more subunits of the TSH receptor is involved in hormone binding may be analogous to the insulin receptor (29). Both the TSH and insulin receptors appear to be large oligomeric proteins with molecular weights of 450,000 (5) and 350,000 (24), respectively. The insulin receptor is composed of two (Y and two /I subunits of which only the (Y is involved in binding, while the /3 is a tyrosine kinase. Whether the 17,000 and/or 63,000 subunits of the TSH receptor play a special function, such as interaction with the guanine nucleotide regulatory proteins or some other component of the cell machinery, remains to be determined. REFERENCES 1. PIERCE, J. G., AND PARSONS, T. F. (1981) Annu Rev. Biochem 50,465-495. 2. BUCKLAND, P. R., HOWELLS, R. D., RICHARDS, C. R., AND REES SMITH, B. (1985) Biochem J. 225,753760. 3. BUCKLAND, P. R., STRICKLAND, T. W., PIERCE, J. G., AND REES SMITH, B. (1985) Endocrinology (Baltimore) 116,2122-2124.
p SUBUNITS
TO
TSH
RECEPTOR
417
4. KOHN, L. D., VALENTE, W. A., LACETTI, P., COHEN, J. L., ALOJ, S. M., AND GROLLMAN, E. F. (1983) Life Sci 32,15-30. 5. MCQUADE, R., THOMAS, C. C., JR., AND NAYFEH, S. N. (1986) Arch. Biochem. Biophys. 246, 5262. 6. POWELL-JONES, C. H. J., THOMAS, C. G., ANDNAYFEH, S. N. (1980) J. Biol. Chem. 255,4001-4010. 7. SHARP, S. B., AND PIERCE, J. (1981) Endocrinology (Baltimore) 109,1052-1060. N. J., FARR, A. L., 8. LOWRY, 0. H., ROSEBROUGH, AND RANDALL, J. (1951) J. Biol. Chem. 193,265275. 9. DRUMMOND, R. W., MCQUADE, R., GRUNWALD, R., THOMAS, C. G., AND NAYFEH, S. N. (1982) Proc. Natl. Acad. Sci USA 79,2202-2206. N. A., HAAS, S. M., BIEBER, L. L., AND 10. MARKWELL, TOLBERT, N. E. (1978) Anal. B&hem. 87,206210. 11. PELCH, P. F., AND CZECH, N. P. (1979) J. Bid Chem 254,3375-3381. 12. LAEMMLI, U. K. (1970). Nature (Lundwn) 227,680685. 13. POWELL-JONES, C. H. J., MCQUADE, R. D., THOMAS, C. G., JR., AND NAYFEH, S. N. (1982) Mol. Cell. Endocrinol. 28,275-287. 14. LIAO, T., AND PIERCE, J. (1970) J. Biol. Chem. 245, 3275-3283. 15. GENNICK, S. E., THOMAS, C. G., AND NAYFEH, S. N. (1986) Biochem Biophys. Res. Commun 135, 208-214. 16. REBOIS, R. V., OMEDEO-SALE, F., BRADY, R. O., AND FISHMAN, P. H. (1981) Proc. Nat1 Acd Sti USA 78,2086-2089. 17. BONIFACINO, J. S., AND DUFAU, M. L. (1984) J. Biol Chem. 259,4542-4549. 18. JI, I., BOCK, J. H., AND Jr, T. H. (1985) J. Biol Chem 260,12815-12821. 19. SHIN, J., AND JI, T. H. (1985) J. Biol. Chem. 260, 12822-12827. 20. SHIN, J., AND JI, T. H. (1985) J. Biol. Chem. 260, 12828-12831. 21. HWANG, J., AND MENON, K. M. J. (1984) Natl Ad Sci. USA 81,4667-4671. 22. MIDDAUGH, C. R., AND JI, T. H. (1980) Eur. J. Biochem 110,587-592. 23. JACOBS, S., AND CUATRECASAS, P. (1981) Endocr. Rev. 2,251-263. 24. MASSAGUE, J., PILCH, P. F., AND CZECH, M. P. (1980) Proc. Natl Acad Sci. USA 77,7137-7141. 25. SCATCHARD, G. (1949) Ann N. Y Acad Sci 51,660674.