chorionic gonadotropin (CG) receptor and human CG for receptor-hormone interaction.

0013-7227/91/1283-1209103.00/0 Endocrinology Copyright © 1991 by The Endocrine Society

Vol. 128, No. 3 Printed in U.S.A.

Significance of the Glycan Moiety of the Rat Ovarian Luteinizing Hormone/Chorionic Gonadotropin (CG) Receptor and Human CG for Receptor-Hormone Interaction* ULLA E. PETAJA-REPO, WOLFGANG E. MERZ, AND HANNU J. RAJANIEMI Biocenter and Department of Anatomy (U.E.P.-R., H.J.R.), University of Oulu, Oulu, Finland, and Department of Biochemistry II (W.EM.)', University of Heidelberg, 6900-Heidelberg, FRG

ABSTRACT. The role of the glycan moiety of the rat ovarian LH/CG receptor and human CG (hCG) in high-affinity receptorhormone interaction was investigated by cross-linking and quantitative binding experiments. hCG and its derivatives, desialylated hCG and deglycosylated hCG were labeled either to the asubunit (12SI) or the /3-subunit (3H). The ligands were attached to ovarian membrane particles, which were treated with neuraminidase or peptide-iV-glycosidase F to remove terminal sialic acids or N-linked oligosaccharides of the receptor, respectively, and the complexes formed were solubilized, cross-linked with glutaraldehyde, and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. All of the ligands produced similar autoradiographic patterns with the native or glycosidasetreated receptor, and only the receptor-(a)hCG and receptor(a,/3)hCG complexes were detected. Moreover, quantitative binding studies indicated that all of the hormone derivatives had similar affinities for the native or glycosidase-treated receptor. In addition, the orientation of the carbohydrate side chains on the receptor-hormone complex was studied by digesting the

complex with the glycosidases. The molecular weight of the receptor, evidenced by ligand blotting, was reduced to the same extent, whether the membrane-bound free receptor or receptorhormone complex was treated with the glycosidases, suggesting that the oligosaccharide side chains of the receptor are apart from the hormone binding region. As pep tide -iV-glycosidase F treatment reduced the size of the Mr 90,000 receptor first to about Mr 67,000 and finally to about Mr 62,000, there may possibly be 2 N-linked carbohydrate chains per receptor polypeptide. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis of the glycosidase-treated receptor-[125I]hCG complex also revealed that neuraminidase was able to remove the sialic acids from both subunits of the receptor-bound hormone. In conclusion, the results suggest that hCG interacts with the polypeptide backbone of its ovarian receptor mainly through the peptide core of its a-subunit. Moreover, the carbohydrate side chains of both subunits of hCG are positioned on the outward face of the receptor-hormone complex. (Endocrinology 128: 1209-1217, 1991)

H

UMAN CG (hCG) and LH regulate gonadal functions by binding to a common high-affinity receptor on the surface of target cells (for review, see Ref. 1). These hormones, along with FSH and TSH, constitute a family of glycoprotein hormones that share a common a-subunit but differ in their hormone-specific /3-subunit (for review, see Ref. 2). Noncovalent association of the subunits is necessary for high-affinity receptor binding, and both of them have also been implicated in receptorhormone interaction (hCG: 3-7; LH: 8; FSH: 9; TSH: 10, 11). The carbohydrate moiety of the hormones is implicated in several actions, including the maintenance of intracellular stability, secretion, assembly, receptor Received September 6,1990. Address all correspondence and requests for reprints to: Dr. Ulla Petaja-Repo, University of Oulu, Department of Anatomy, Kajaanintie 52 A, SF-90220 Oulu, Finland. * This work was supported by the Academy of Finland, The Sigrid Juselius Foundation, the Cultural Foundation of Finland, and the Oulu University Scholarship Foundation.

binding, steroidogenesis, and modulation of plasma halflife (for review, see Refs. 12,13). Enzymatically or chemically deglycosylated hCG and LH are antagonists to the native hormones (for review, see Refs. 12,13). They bind to the receptor in the normal manner, but their relative potencies as activators of adenylyl cyclase and steroidogenesis are either lessened or abolished. The receptor for hCG and LH is also glycosylated, and about 30% of the molecular mass of the rat ovarian LH/ CG receptor is composed of sialylated, complex-type carbohydrate chains attached to asparagine residues (1417). Keinanen (14) and Ji et al. (18) have shown that enzymatic deglycosylation of the LH/CG receptor does not abolish its hormone binding ability, but otherwise the functional significance of the oligosaccharide side chains is unknown. The present work was undertaken to study the effect of the glycan moiety of the rat ovarian LH/CG receptor and hCG on receptor-hormone interaction and the ori-

1209

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 07:34 For personal use only. No other uses without permission. . All rights reserved.

1210

ROLE OF CARBOHYDRATES IN LH/CG RECEPTOR-hCG INTERACTION

entation of the carbohydrate side chains on the receptorhormone complex. In a previous paper from our laboratory (7) it was shown that the a-subunit of hCG seems to be in closer contact with the receptor. Materials and Methods Chemicals and hormone derivatives Peptide-iV-glycosidase F (PNGase F; EC 3.2.2.18) was obtained from Boehringer-Mannheim (Mannheim, F.R.G.) and neuraminidase from either Calbiochem (San Diego, CA; EC 3.2.1.18, from Vibrio cholerae) or Sigma (Deisenhofen, F.R.G.; Type X-A, from Clostridium perfringens, immobilized on agarose beads). AT-ethylmaleimide (NEM) and phenylmethylsulfonyl fluoride (PMSF) were products of Sigma (St. Louis, MO). PMSG (2,820 IU/mg) and hCG (10,900 IU/mg), which were used to induce pseudopregnancy, were obtained from Diosynth (Oss, Netherlands). Na125I and NaB3H4 were from Amersham (Bucks, U.K.) and the equipment and reagents for electrophoresis and blotting from Bio-Rad (Richmond, CA). All the other reagents were of analytical grade and obtained from commercial suppliers. Purified hCG (12,000 IU/mg) was prepared from crude hCG (Schering, Berlin, F.R.G.) according to Canfield et al. (19) as modified by Merz et al. (20). Desialylated hCG (DS-hCG) was obtained by digestion (16-18 h, 37 C) of purified hCG with neuraminidase from Clostridium perfringens in a buffer containing 50 mM sodium phosphate and 50 mM citric acid (pH 5.6) according to Cassidy et al. (21) and desalting on a Biogel P-30 column (Bio-Rad, Miinchen, F.R.G.; 50 mM NH4HCO3). Deglycosylated hCG (DG-hCG) (70% of sugar residues removed) was prepared as described by Merz (22). Briefly, purified hCG was treated in a closed Teflon apparatus with anhydrous HF for 60 min at 0 C. After the HF had been removed under vacuum, DG-hCG was recovered in 0.5 M NH4HCO3, incubated overnight at 37 C, and subjected to gel filtration (Biogel P-60 column, Bio-Rad; 50 mM NH4HCO3). The hormone derivatives were radioiodinated with Na125I by the chloramine-T method (23) as described by Markkanen et al. (24). The labeled DS-hCG and DG-hCG were separated from free 125 I by gel filtration using a Sephadex G-50 column (Pharmacia, Uppsala, Sweden) and 10 mM PBS (pH 7.4) containing 0.1% BSA as an eluent. The specific activities of the preparations were 60-80 Ci/g, as determined by self-displacement assay (25). The fraction of labeled hormone derivatives capable of binding to an excess of ovarian membranes was 40-50%. 3H-Labeled hCG was prepared by a modification of the periodate-NaB3H4 method (26), essentially as described previously (7). The specific activity of the preparations averaged 1 Ci/g, and the binding activity was about 30%. In 125I-labeled hormone derivatives about 90% of the radioactivity was in the a-subunit, and in 3H-hCG about 70% was in the /?-subunit, as described previously (7).

Endo • 1991 Voll28«No3

ized ovaries were collected 7-9 days after hCG injection, freed from extraneous tissue, frozen in liquid nitrogen, and stored at —70 C before use. Preparation of ovarian membrane particles and solubilization of the receptor and receptor-hormone complex Crude ovarian membrane particles were prepared as described previously (7). The luteinized ovaries were homogenized in buffer A (PBS containing 5 mM EDTA, 5 mM iV-ethylmaleimide, and 0.2 mM PMSF; 2 ml/ovary) in a tight-fitted allglass homogenizer wir,h 100 strokes, after which the homogenate was centrifuged at 100 x g for 10 min and the supernatant further centrifuged at 27,000 x g for 30 min. The pellet containing the crude ovarian membranes was suspended in buffer A (0.25 ml/ovary) with mild homogenization and incubated with the labeled hormone derivatives ([125I]hCG, 3H-hCG, [125I] DS-hCG, or [125I]DG-hCG; 3 x 106 cpm/ovary) at 37 C for 1 h or at 20 C for 16 h. For control purposes, the labeling was also performed in the presence of an excess of unlabeled hCG (20 Mg/ovary). The unbound hormone was removed by washing twice with buffer A (L.5 ml/ovary). To obtain the solubilized receptor-hormone complex for sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) or ligand blotting analyses, the final pellet was solubilized in buffer A (0.5 ml/ ovary) containing 1% Triton X-100 or 1% Triton X-100 and 20% glycerol, respectively, by stirring on ice for 30 min. The suspension was further diluted 1:1 with buffer A or buffer A containing 20% glycerol and centrifuged at 100,000 X g for 60 min. To obtain the solubilized receptor for ligand blotting analysis, the pellet containing the crude membranes was solubilized as described above without attaching the labeled hormone to the membranes. All the above steps were performed at 0-4 C unless stated otherwise. Desialylation and deglycosylation of the membrane-bound receptor, receptor- f1251JhCG complex and free

The membrane-bound receptor was treated with glycosidases as described previously (14). Crude membrane particles (100 ng protein) were suspended in 0.1 M sodium phosphate (pH 7.0) containing 0.2 mM PMSF, after which 16 n\ neuraminidase (16 mU), PNGase F (3.2 U), or buffer (control) was added, and the mixtures were incubated at 30 C for various times as indicated in the figure legends. The reaction was quenched by washing twice with buffer A, and the membranes were solubilized as described above. Crude membrane particles (100 ng protein) containing the receptor-[125I]hCG complex and free [125I]hCG in the presence of an equivalent amount of down-regulated membrane particles (28) were treated in the similar way. After glycosidase treatment, the free hormone was separated from the down-regulated membranes by centrifugation (27,000 X g, 30 min), and the membranes containing the receptor-hormone complex were solubilized as described above.

Animals

Covalent cross-linking of the receptor-hormone complex and radioiodinated hormone derivatives

Immature 27-day-old female Sprague-Dawley rats were rendered pseudopregnant by sequential injections of PMSG (40 IU) and hCG (25 IU) as described by Parlow (27). The lutein-

The receptor-hormone complex was cross-linked as described before (7) with 2 mM glutaraldehyde (final concentration) for 60 min at 20 C. The cross-linking reaction was quenched by



1211

adding 0.125 vol 1 M Tris-HCl (pH 7.4 at 4 C). The solubilized receptor-hormone complex was cross-linked after diluting the Triton X-100 extract with buffer A containing 0.5% Triton X100 (100 ng protein/ml). In control experiments, labeled hormone derivatives ([125I]hCG, [125I]DS-hCG, or [125I]DG-hCG; 10,000 cpm in PBS) were cross-linked without attaching them to the receptor.

cross-linked receptor-hormone complex (2-5 ng protein) were mixed with 2 vol SDS-sample buffer and subjected to SDSPAGE under nonreducing conditions as described above but without prior heating. The proteins resolved were transferred electrophoretically from the gels onto nitrocellulose sheets (100 V, 60 min, Bio-Rad Mini-Trans Blot apparatus) and processed further as described previously (ligand blotting; 34).

Hormone binding assays

Protein determination

The hormone binding characteristics of the solubilized, native or glycosidase-treated receptor were determined by measuring the binding of an increasing amount (0.1-20 ng) of 125Ilabeled hCG derivatives ([125I]hCG, [125I]DS-hCG, or [125I]DGhCG) to the solubilized receptor in the absence or presence of an excess of unlabeled hormone (2 ng). Crude membrane particles were solubilized in PBS containing 1% Triton X-100 and 20% glycerol, and samples of the extract (60-70 ng protein) were incubated with the hormones in a final vol of 250 n\ PBS containing 0.1% Triton X-100 and 0.1% BSA for 16 h at 20 C. The receptor-bound and free hormone were separated by polyethylene glycol precipitation as described previously (29), with some modifications. Five hundred fifty microliters of 0.1% bovine 7-globulin (in PBS) was added to the assay tubes, followed by 800 n\ 24% polyethylene glycol (also in PBS). The tubes were mixed, incubated on ice for 30 min, and centrifuged at 1500 x g for 15 min at 4 C. The supernatants were carefully decanted and the precipitates dissolved in 250 /xl PBS containing 0.1% Triton X-100. After 10 min on ice the receptorhormone complexes were reprecipitated and recentrifuged as described above and the pellets counted for 125I-radioactivity. The binding data were transformed to Scatchard plots (30) to obtain equilibrium dissociation constants (Ka) and maximal binding capacities (BmQX). All assays were performed in triplicate.

Protein was assayed by the method of Lowry et al. (35) as modified by Dulley and Grieve (36).

SDS-PAGE and ligand blotting SDS-PAGE was performed according to Laemmli (31), using 4% stacking gels and 7.5% or 13.5% separating gels. Samples containing the labeled hormone derivatives or solubilized receptor-hormone complex were dialyzed against diluted (1:41:10) SDS-sample buffer (pH 6.8) containing 2% SDS, 125 mM Tris-HCl, 10% glycerol, and 0.001% Bromophenol blue for 20 h at 4 C and concentrated using a Savant Speed-Vac concentrator (Savant Instruments, Farmingdale, NY). Before electrophoresis the samples were diluted with SDS-sample buffer to obtain the same amount of radioactivity in each well and heated at 95 C for 2 min in the presence of 2% 2-mercaptoethanol. Molecular weight markers (SDS-6H, SDS-6, Sigma) detected by staining with Coomassie brilliant blue (32) or prestained molecular weight markers (SDS-7B, Sigma) were used to calibrate the slab gels. The calibration was important, because individual strips of different autoradiograms were combined in some of the figures. For the detection of radioactivity, the gels were dried and exposed at -70 C for 2-30 days using Kodak XAR-5 film (Eastman Kodak, Rochester, NY) with or without an intensifying screen (Kodak X-Omatic). The gels containing 3H-labeled samples were treated for fluorography (33) before drying. The samples containing the solubilized receptor or non-

Results Interaction of a-subunit and fi-subunit-labeled hCG with the native, desialylated, and deglycosylated LH/CG receptor The ability of glycosidases to digest the membranebound LH/CG receptor was first tested by subjecting the rat ovarian membranes to glycosidase treatment and identifying the receptor by ligand blotting. The solubilized membrane proteins were resolved with SDS-PAGE and transferred onto nitrocellulose sheets. Incubation of the sheets containing the untreated membrane proteins with [125I]hCG revealed specific binding to a Mr 90,000 polypeptide, representing the native receptor (Fig. 1). Treatment of the membranes with neuraminidase decreased the Mr of the receptor to about 79,000, the desialylation being complete in 30 min (Fig. 1A). Treatment with PNGase F, on the other hand, decreased the Mr first to about 67,000 and finally to about 62,000 (Fig. IB). The deglycosylation was virtually complete in 120 min, and this incubation time was used in the subsequent experiments. Incubation of the membranes without any enzymes had no effect on the Mr of the receptor. In order to determine the effect of the glycan moiety of the receptor on receptor-hormone interaction, the asubunit-labeled hCG ([125I]hCG) and /3-subunit-labeled hCG (3H-hCG) (see Ref. 7) were attached to glycosidasetreated ovarian membrane particles. The complexes formed were solubilized in Triton X-100 and cross-linked with glutaraldehyde. SDS-PAGE analysis of the native receptor-hormone complex under reducing conditions gave a specific band of Mr 137,000 on the autoradiograms irrespective of the labeled ligand used (Fig. 2, A and B). A smaller specific band of Mr 106,000 was also detected, but only when [125I]hCG was used (Fig. 2A). The Mr 137,000 complex contains the a,/3-dimer of hCG bound to a Mr 90,000 receptor component, and the Mr 106,000 complex contains only the a-subunit bound to the receptor (see Ref. 7). The hormone that was dissociated from the receptor appeared on the autoradiograms as a Mr 49,000 band, and the free subunits migrated at the dye front. Desialylation of the receptor with neuraminidase reduced the 137,000 and 106,000 complexes in size by



1212

Endo • 1991 Voll28»No3

A

A

neuraminidase

control

M r x10" 3

control

neuraminidase PNGaseF I Mrx10"3

-205

-180 -116 45

0 -

-84

-58 -48.5

0 240 10 30 60120240 240 incubation (min) + - - - - - - + unlabeled hCG

control

unlabeled hCG

PNGase F I

M r x10" 3 control -205

neuraminidase PNGase F

-116 -97.4 -66

I

M r x10" 3

-45

0 -

0 24010 30 60120240240 + - - - - - +

incubation (min) unlabeled hCG

FIG. 1. Identification of the rat ovarian LH/CG receptor by ligand blotting after neuraminidase (panel A) and PNGase F (panel B) treatment of the ovarian membrane particles. Crude membrane particles (100 fig protein) were incubated alone, with neuraminidase (16 raU), or with PNGase F (3.2 U) at 30 C for the periods of time indicated. After digestion, the membranes were washed and solubilized in 1% Triton X-100/20% glycerol. The solubilized receptor was subjected to SDS-PAGE under nonreducing conditions (7.5% slab gel) and electroblotted onto nitrocellulose sheets, which were incubated with [125I] hCG in the absence or presence of an excess of unlabeled hCG for 16 hours at 20 C. Molecular weight markers are indicated on the right (from the top: myosin, #-galactosidase, phosphorylase b, BSA, and egg albumin).

about 9,000 and deglycosylation with PNGase F by about 20,000. (Fig. 2). Densitometric scans of the autoradiograms revealed that the proportions of the bands were unchanged. No other complexes that could have represented the receptor-(/?)hCG complex were detected after either enzyme treatment. To evaluate further the role of the carbohydrates of the receptor in high-affinity hormone binding, quantitative binding assays were performed. The binding affinity of [125I]hCG for the solubilized, native, or glycosidasetreated receptor was essentially unaltered (Table 1).

-

+

-

+

-

+

unlabeled hCG

FIG. 2. Covalent cross-linking of [125I]hCG (panel A) and 3H-hCG (panel B) to the native, desialylated, and deglycosylated rat ovarian LH/CG receptor. Crude membrane particles (100 fig protein) were incubated alone for 120 min, with neuraminidase (16 mU) for 30 min or with PNGase F (3.2 U) for 120 min at 30 C. After digestion, [12SI] hCG or 3H-hCG was attached to the membranes in the presence or absence of an excess of unlabeled hCG, and the complexes formed were solubilized in 1% Triton X-100. The diluted extracts (5-10 fig protein) were treated with 2 mM glutaraldehyde and analyzed by SDS-PAGE (7.5% slab gel) under reducing conditions. Note that both labeled subunits of hCG migrate at the dye front on the 7.5% gel. Numbers on the right denote molecular weight markers (from the top: a2-macroglobulin, j8-galactosidase. fructose-6-phosphate kinase, pyruvate kinase, and fumarase).

Interaction of 125I-labeled DS-hCG and DG-hCG with the native, desialylated, and deglycosylated LH/CG receptor Two hCG derivatives were used to elucidate the effect of the glycan moiety of hCG on receptor-hormone interaction: enzymatically desialylated hCG and chemically deglycosylated hCG. SDS-PAGE analysis of the crosslinked 125I-labeled DS-hCG and DG-hCG revealed that


ROLE OF CARBOHYDRATES IN LH/CG RECEPTOR-hCG INTERACTION TABLE 1. Binding of [126I]hCG, [125I]DS-hCG, and [125I]DG-hCG to the native, desialylated, and deglycosylated rat ovarian LH/CG receptor Bman

Hormone preparation

Receptor preparation

Kd (M X 10" )

(mol/mg protein x 10~13)

hCG hCG hCG DS-hCG DS-hCG DS-hCG DG-hCG DG-hCG DG-hCG

Native receptor DS-receptor DG-receptor Native receptor DS-receptor DG-receptor Native receptor DS-receptor DG-receptor

4.20 ± 0.67 4.23 ± 1.15 4.08 ± 0.88 4.89 ± 0.90 5.86 ± 0.43 4.86 ± 0.30 6.15 ± 1.00 6.44 ± 1.71 5.76 ± 1.66

2.21 ± 0.16 2.57 ± 0.43 2.21 ± 0.42 1.99 ± 0.75 2.11 ± 0.24 1.57 ± 0.22 2.35 ± 0.69 2.71 ± 1.21 1.77 ± 0.40

10

Crude ovarian membrane particles (0.5 mg protein/ml) were incubated alone for 120 min, with neuraminidase (80 mU/ml) for 30 min or with PNGase F (16 U/ml) for 120 min at 30 C, washed and solubilized in 1% Triton X-100/20% glycerol. The solubilized extracts (20-30 ng protein/assay) were incubated with an increasing amount of radiolabeled hCG derivatives for 16 h at 20 C, and the free and receptor-bound hormone were separated by polyethylene glycol precipitation. Equilibrium dissociation constants (Kd) and maximal binding capacities (Bmax) were derived from Scatchard plots (30). Data are represented as means ± SD of 3 separate determinations. 125

!-hCG 125i_ I-DS- iCG 1-DGJhCG M r x10' 3

1213

molecular weights of these bands were about 6,000 and 14,000 smaller than the ones obtained with the native hormone, the differences in the band positions are most likely due to variations in the electrophoretic mobility of the hormone derivatives. Thus they clearly represent receptor-(a,/?)hCG and receptor-(a)hCG complexes. Similarly, the complexes formed after attaching 125Ilabeled DS-hCG and DG-hCG to the neuraminidase and PNGase F-treated receptor were about 9,000 and 20,000 smaller, respectively, than the ones obtained with these hormone derivatives and the native receptor (Fig. 4). Again the proportions of the bands were unchanged, as revealed by densitometric scanning. Moreover, the binding affinities of the hCG derivatives for the solubilized, native, or glycosidase-treated receptor were unaltered (Table 1). Orientation of the carbohydrate side chains on the LH/ CG receptor-hCG complex The orientation of the oligosaccharides on the receptor-hormone complex was studied by testing the ability of glycosidases to remove carbohydrates of both the hormone and the receptor from the complex. Ovarian membrane particles bearing the receptor-[125I]hCG complex were treated with glycosidases, solubilized, and transferred onto nitrocellulose sheets after SDS-PAGE. The bound hormone dissociated from the receptor during SDS-PAGE, as the complex was not cross-linked, and control

neuraminidase PNdasdF

r —

—

—

+

+ 125

+

M r x10" 3

cross-linking 125

FlG. 3. Covalent cross-linking of [ I]hCG, [ I]DS-hCG, and [125I] DG-hCG. The radioiodinated hormone derivatives were treated with 0 mM or 2 mM glutaraldehyde and analyzed by SDS-PAGE (13.5% slab gel) under reducing conditions. Note that in the absence of the crosslinker almost all of the a,j8-dimer of hCG derivatives dissociated, and the radioactivity appeared in the labeled a-subunits.

their molecular weights were about 7,000 and 16,000 smaller, respectively, than that of the native hCG and that these preparations did not contain detectable amounts of native hormone (Fig. 3). 125 I-Labeled DS-hCG and DG-hCG were attached to ovarian membrane particles, and the complexes formed were solubilized and treated with glutaraldehyde. SDSPAGE analysis of the complexes gave 2 high molecular weight bands on the autoradiograms (Fig. 4). As the

125

l-hCG

125

l-DS-hCG 125 l-DG-hCG

FIG. 4. Covalent cross-linking of [125I]hCG, [125I]DS-hCG, and [125I] DG-hCG to the native, desialylated, and deglycosylated rat ovarian LH/CG receptor. Crude membrane particles (100 jig protein) were incubated alone for 120 min, with neuraminidase (16 mU) for 30 min or with PNGase F (3.2 U) for 120 min at 30 C before hormone binding. The 1% Triton X-100-solubilized receptor-hormone complexes were cross-linked and analyzed as described in Fig. 2.


1214


the receptor could be visualized by subjecting the nitrocellulose sheets to incubation with [125I]hCG (Fig. 5). The hormone that dissociated during SDS-PAGE was seen as a Mr 49,000 band on the autoradiograms, as it did not dissociate extensively into subunits under nonreducing conditions. As the Mr of the receptor decreased from 90,000 to about 79,000 after neuraminidase treatment of the complex and to about 62,000 after PNGase F treatment, the bound hormone did not prevent the glycosidases from digesting the receptor (Fig. 5). Interestingly, neuraminidase was able to desialylate the receptor-bound hCG, as the Mr of the hormone molecule was reduced from 49,000 to about 42,000 (Fig. 5; see also Fig. 3). The receptor-bound hCG, on the other hand, was unaffected by PNGase F treatment (Fig. 5). The ability of neuraminidase to digest the oligosaccharides of the individual subunits of hCG was studied further. Ovarian membrane particles containing the receptor-[125I]hCG complex or free [125I]hCG in the presence of an equivalent amount of down-regulated ovarian membranes were treated with neuraminidase. SDSPAGE analysis of the solubilized noncross-linked receptor-[125I]hCG complex and the free [125I]hCG revealed that both the a- and 0-subunits of the free and receptorbound hormone were desialylated (Fig. 6). The Mr values of the a- and /:?-subunits decreased from 24,000 and 37,000 to 23,000 and 32,000, respectively, in 30 min. The 0-subunit could be detected on autoradiograms after long exposure, as radioiodination of hCG to a specific activity of 60-80 Ci/g introduced about 5-7% of the label into it (see Ref. 7). PNGase F was not able to deglycosylate the free or receptor-bound hormone, not even when incubated for 22 h (data not shown). This was to be expected,

M

2 I 3

4 I5

6 I7

8\9

1OM1 I 2 I

M r x10" 3

. -205

jg| n':IH' "45 unlabeled hCG

FIG. 5. Ability of neuraminidase and PNGase F to digest the carbohydrate chains from the rat ovarian LH/CG receptor- [125I]hCG complex. Crude membrane particles (100 fig protein) containing the free receptor (lanes 1, 3, 5, 7, 9,11) or receptor-[125I]hCG complex (lanes 2, 4, 6, 8, 10, 12) were incubated alone for 120 min, with neuraminidase for 30 min, or with PNGase F for 120 min at 30 C and solubilized in 1% Triton X-100/20% glycerol. Samples were subjected to ligand blotting as described in Fig. 1.

125


l-DS-hCG free

125

!-hCG IOC

receptor-bound

1-hCG

(a,B)-hCGd (B)-hCG(Z (a)-hCGCZ neuraminidase FIG. 6. Treatment of thu free and receptor-bound [125I]hCG with neuraminidase. Crude ovarian membrane particles (100 ng protein) containing the receptor-bound [125I]hCG and free [125I]hCG in the presence of an equivalent amount of down-regulated membranes were incubated at 30 C alone or with neuraminidase (16 mU) for 30 min. The free hormone was separated from the down-regulated membrane particles by centrifugation, and the membranes containing the receptor-bound hormone were solubilized in 1% Triton X-100. Samples were analyzed by SDS-PAGE (13.5% slab gel) under reducing conditions. For contol purposes, [125I]DS-hCG was run on the same gel.

as deglycosylation of glycoprotein hormones with PNGase F has been shown to require denaturation (37).

Discussion The evidence presented here indicates that the carbohydrate moiety of neither the LH/CG receptor nor hCG is essential for high-affinity receptor-hormone interaction. The binding affinity of hCG, DS-hCG, and DGhCG to the native, neuraminidase or PNGase F-treated receptor was essentially unaltered. This is in accordance with previous reports showing that progressive removal of carbohydrates from the glycoprotein hormones by either enzymatic or chemical means has little effect on receptor binding even though it lowers biological activity (for review, see Refs. 12, 13). In addition, Matzuk et al. (38) report that the nonglycosylated hCG obtained by site-directed mutagenesis can fully bind to its receptor. On the other hand, little is known of the effect of the glycan moiety of the receptor on receptor-hormone interaction. Keinanen (14) has shown that the enzymatic removal of carbohydrates does not abolish the hormone binding ability of the rat ovarian LH/CG receptor, and similar results have been obtained recently with mouse Leydig tumor cells and rat granulosa cells (18). A few authors have also reported on a modulation of gonadotropin binding after incubation of target membranes with neuraminidase, but the more pronounced hormone binding was shown to be due to an increase in the number of binding sites without any alteration in affinity (39, 40, 41). LH/CG receptor-hCG interaction seems to be of a



polypeptide nature, the hormone being bound to the polypeptide backbone of the receptor mainly through the peptide core of its a-subunit. This was confirmed by cross-linking experiments, as all of the hCG derivatives produced similar autoradiographic patterns whether they were cross-linked to the native, desialylated or deglycosylated receptor. The receptor-(a)hCG and receptor(a,/3)hCG complexes were the only components detected. In a recent report, Sairam (42) describes cross-linking of [125I]DG-hCG to mouse Leydig tumor cells (MA-10) with similar results. In contrast, Nishimura et al. (43) and Ji and Ji (44) show that [125I]DG-hCG can be crosslinked to a different receptor component from the native hormone. The last two authors had earlier suggested an oligomeric structure for the LH/CG receptor, however. Since the absence of the bulky N-linked oligosaccharides of hCG does not alter receptor-hCG interaction, the oligosaccharides of the a- and /3-subunits are probably positioned on the outward face of the hormone when it is bound to the receptor. This was confirmed by the ability of neuraminidase to remove terminal sialic acids from the oligosaccharide side chains of both subunits of the receptor-bound hCG. In addition, this showed that although the a-subunit seems to be in closer contact with the receptor, it is not buried under the /3-subunit. We have shown previously (7) that the peptide core of the asubunit is also at least partially exposed on the receptorhormone complex, as (a)hCG antibodies were able to recognize the complex in immunoblotting. The orientation of the carbohydrate side chains of hCG on the receptor-hormone complex is an important question and needs to be resolved in order to understand the manner in which carbohydrates take part in signal transduction mechanisms. The oligosaccharides may either contribute to the active conformation of the hormone with respect to receptor activation (45-47) or they may directly participate in the activation of target cell responses by interacting with the receptor or other plasma membrane components (48). It is interesting that the extracellular part of the rat LH/CG receptor, which includes the hormone binding region (49), expresses sequence similarity with soybean lectin (50). It remains to be solved, whether this putative carbohydrate binding region of the receptor is of any functional significance. As the removal of carbohydrates does not change the binding affinity of glycoprotein hormones, direct interaction of hCG with this lectin-like region seems to be unlikely. On the other hand, Thotakura et al. (51) have shown in a recent report that N-linked oligosaccharide chains extracted from various glycoproteins inhibit the binding of hCG to its receptor, indicating that they may perturb receptor-hCG interaction through the carbohydrate binding site of the receptor. There are 6 putative glycosylation sites on the extra-

1215

cellular portion of the rat LH/CG receptor (50). The finding that PNGase F treatment of the receptor resulted initially in the formation of a Mr 67,000 peptide, which was further converted to a Mr 62,000 one suggests that there are at least 2 N-linked carbohydrate side chains per receptor polypeptide with differing susceptibilities to PNGase F treatment. Moreover, the glycosylation sites of the receptor are not very close to the hormone binding region, as PNGase F was able to remove the carbohydrates of the receptor from the receptor-hormone complex. The same conclusion can be drawn from the experiments performed by Kress and Spiro (52) concerning the TSH receptor, in which it was found that both the free receptor and the hormone-bound receptor, but not the free hormone, were able to bind to a lectin, Bandeiraea simplicifolia I. As hCG is a large molecule, the finding that the oligosaccharide side chains of the receptor seem to be apart from the hormone binding region was unexpected. On the other hand, it must be remembered that the extracellular part of the LH/CG receptor is also exceptionally large among G-protein-coupled receptors, 341 amino acid residues out of the total of 674 in the rat (50). Although the carbohydrate moiety of the LH/CG receptor does not seem to have any effect on hormone binding it may play a role in some other functions such as intracellular transport of the receptor or signal transduction, actions implicated for another G-protein-coupled receptor, the /?-adrenergic receptor (53-55). Azhar and Menon (56) have shown that pretreatment of ovarian cells with lectins concanavalin A and wheat germ agglutinin blocks the hCG and LH-induced cAMP and progesterone responses. Thus, further research is required to assess whether a nonglycosylated or deglycosylated LH/CG receptor is capable of activating adenylyl cyclase and possibly other plasma membrane effectors. In conclusion, the present results show that hCG interacts with the polypeptide backbone of the LH/CG receptor mainly through the peptide portion of its asubunit. The a-subunit is not buried under the /3-subunit, however, and the carbohydrate side chains of both subunits as well as those of the receptor are on the outward face of the receptor-hormone complex.

Acknowledgments The skillful technical assistance of Ms. Paula Soininen, Ms. Aino Kuha, Mr. Eero Oja, and Ms. Sirkka Martti is gratefully acknowledged.

References 1. Rajaniemi HJ, Keinanen KP, Kellokumpu S, Petaja-Repo UE, Metsikko MK 1989 Molecular structure of the luteinizing hormone receptor. Biol Reprod 40:1-12 2. Pierce JG, Parsons TF 1981 Glycoprotein hormones: structure and function. Annu Rev Biochem 50:465-495


1216


3. Kusuda S, Dufau ML 1986 Purification and characterization of the rat ovarian receptor for luteinizing hormone. Structural studies of subunit interaction. J Biol Chem 261:16161-16168 4. Charlesworth MC, McCormick DJ, Madden B, Ryan RJ 1987 Inhibition of human choriogonadotropin binding to receptor by human choriogonadotropin a peptides. J Biol Chem 262:1340913416 5. Schwarz S, Berger P, Nelboeck E, Khashabi D, Panmoung W, Klieber R, Wick G 1988 Probing the receptor-interaction of glycoprotein hormones with monoclonal antibodies. J Recept Res 8:437-453 6. Salesse R, Bidart J-M, Troalen F, Bellet D, Gamier J 1990 Peptide mapping of intersubunit and receptor interactions of human choriogonadotropin. Mol Cell Endocrinol 68:113-119 7. Petaja-Repo UE, Rajaniemi H 1990 Interaction of human chorionic gonadotropin (hCG) with the rat ovarian luteinizing hormone (LH)/CG receptor. Mol Cell Endocrinol 72:43-53 8. Alonso-Whipple C, Couet ML, Doss R, Koziarz J, Ogurno EA, Crowley Jr WF 1988 Epitope mapping of human luteinizing hormone using monoclonal antibodies. Endocrinology 123:1854-1860 9. Coloma TAS, Dattatreyamurty B, Reichert Jr LE 1990 A synthetic peptide corresponding to human FSH /3-subunit 33—53 binds to FSH receptor, stimulates basal estradiol biosynthesis, and is a partial antagonist of FSH. Biochemistry 29:1194-1200 10. Morris III JC, Jiang N-S, Charlesworth MC, McCormick DJ, Ryan RJ 1988 The effects of synthetic a-subunit peptides on thyrotropin interaction with its receptor. Endocrinology 123:456-462 11. Endo Y, Tetsumoto T, Nagasaki H, Kashiwai T, Tamaki H, Amino N, Miyai K 1990 The distinct roles of a- and /3-subunits of human thyrotropin in the receptor-binding and postreceptor events. Endocrinology 127:149-154 12. Sairam MR 1985 Protein glycosylation and receptor-ligand interactions. In: Conn PM (ed) The Receptors. Academic Press, New York, vol 2:307-340 13. Ryan RJ, Keutmann HT, Charlesworth MC, McCormick DJ, Milius RP, Calvo FO, Vutyavanich T 1987 Structure-function relationships of gonadotropins. Recent Prog Horm Res 43:383-429 14. Keinanen KP 1988 Effect of deglycosylation on the structure and hormone binding activity of lutropin receptor. Biochem J 256:719724 15. Kusuda S, Dufau ML 1988 Characterization of ovarian gonadotropin receptor. Monomer and associated form of the receptor. J Biol Chem 263:3046-3049 16. Minegishi T, Delgado C, Dufau ML 1989 Phosphorylation and glycosylation of the luteinizing hormone receptor. Proc Natl Acad Sci USA 86:1470-1474 17. Sojar HT, Bahl OP 1989 Characterization of rat ovarian lutropin receptor. Role of thiol groups in receptor activation. J Biol Chem 264:2552-2559 18. Ji I, Slaughter RG, Ji TH 1990 N-linked oligosaccharides are not required for hormone binding of the lutropin receptor in a Leydig tumor cell line and rat granulosa cells. Endocrinology 127:494-496 19. Canfield RE, Morgan FJ, Kammerman S, Bell JJ, Agosto GM 1971 Studies of human chorionic gonadotropin. Recent Prog Horm Res 27:121-164 20. Merz WE, Hilgenfeldt U, Dorner M, Brossmer R 1974 Biological, immunological and physical investigations on human chorionic gonadotropin. Hoppe-Seyler's Z Physiol Chem 355:1035-1045 21. Cassidy JT, Jourdian GW, Roseman S 1966 Sialidase from Clostridium perfringens. Methods Enzymol 8:680-685 22. Merz WE 1988 Evidence for impaired subunit interaction in chemically deglycosylated human choriogonadotropin. Biochem Biophys Res Commun 156:1271-1278 23. Hunter WM, Greenwood FC 1962 Preparation of iodine-131 labelled human growth hormone of high specific activity. Nature 194:495-496 24. Markkanen S, Tollikko K, Jaaskelainen K, Rajaniemi H 1980 Effect of radioiodination on structural integrity, receptor and antibody-binding activity and circulatory behavior of human choriogonadotropin. Horm Res 12:32-45 25. Ketelslegers J-M, Knott GD, Catt KJ 1975 Kinetics of gonadotropin binding by receptors of the rat testis: analysis by a non-linear


curve-fitting method. Biochemistry 14:3075-3083 26. Gahmberg CG, Andexsson LC 1977 Selective radioactive labeling of cell surface sialoglycoproteins by periodate-tritiated borohydride. J Biol Chem 252:5888-5894 27. Parlow AF 1961 Bio-assay of pituitary luteinizing hormone by depletion of ovarian ascorbic acid. In: Albert A (ed) Human Pituitary Gonadotropins. CC Thomas, Springfield, pp 300-310 28. Lakkakorpi JT, Keiniinen KP, Rajaniemi HJ 1990 Production and characterization of polyclonal antiserum to rat luteinizing hormone/chorionic gonadotropin receptor and its immunohistochemical application for studying receptor location and down-regulation. Endocrinology 127:513-522 29. Dufau ML, Charreau EH, Catt KJ 1973 Characteristics of a soluble gonadotropin receptor from the rat testis. J Biol Chem 248:69736982 30. Scatchard G 1949 The attractions of proteins for small molecules and ions. Ann NY Acad Sci 51:660-672 31. Laemmli UK 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-685 32. Fairbanks G, Steck TL, Wallach DFH 1971 Electrophoretic analysis of the major polypeptides of the human erythrocyte membranes. Biochemistry 10:2606-2617 33. Bonner WM, Laskey RA 1974 Afilmdetection method for tritiumlabelled proteins and nucleic acids in polyacrylamide gels. Eur J Biochem 46:83-88 34. Keinanen KP, Kellokumpu S, Metsikko MK, Rajaniemi HJ 1987 Purification and partial characterization of rat ovarian lutropin receptor. J Biol Chem 262:7920-7926 35. Lowry OH, Rosebrough NJ, Farr AL, Randall, RJ 1951 Protein measurement with the' Folin phenol reagent. J Biol Chem 193:265275 36. Dulley JR, Grieve PA 1975 A simple technique for eliminating interference by detergents in the Lowry method of protein determination. Anal Biochem 64:136-141 37. Ronin C, Papandreou M-J, Canonne C, Weintraub BD 1987 Carbohydrate chains of human thyrotropin are differentially susceptible to endoglycosidajie removal on combined and free polypeptide subunits. Biochemistry 26:5848-5853 38. Matzuk MM, Keene JL, Boime I 1989 Site-directed specificity of the chorionic gonadotropin N-linked oligosaccharides in signal transduction. J Biol Chem 264:2409-2414 39. Danforth DR, Stouffer RL 1985 Evidence for two populations of masked gonadotropin binding sites in the corpus luteum of the Rhesus monkey (Mactxa mulatto). Endocrinology 117:747-754 40. Azhar S, Menon KMJ 1981 Receptor-mediated gonadotropin action in ovary. Possible regulatory role of cell-surface sialic acids in gonadotropin interaction to purified bovine corpus luteum plasma membranes. Eur J Biochem 118:25-31 41. Nishimori K, Yamoto M, Nakano R 1989 In vitro membrane desialylation in porcine granulosa cells unmasks functional receptors for follicle-stimulating hormone. Endocrinology 124:26592665 42. Sairam MR 1990 Complete dissociation of gonadotropin receptor binding and signal trunsduction in mouse Leydig tumour cells. Biochem J 265:667-674 43. Nishimura R, Raymond MJ, Ji I, Rebois RV, Ji TH 1986 Photoaffinity labeling of the gonadotropin receptor with native, asialo, and deglycosylated choriogonadotropin. Proc Natl Acad Sci USA 83:6327-6331 44. Ji I, Ji T 1990 Differential interactions of human choriogonadotropin and its antagonistic aglycosylated analog with their receptor. Proc Natl Acad Sci USA 87:4396-4400 45. Keutmann HT, Johnson L, Ryan RJ 1985 Evidence for a conformational change in deglycosylated glycoprotein hormones. FEBS Lett 185:333-338 46. Rebois RV, Liss MT :.987 Antibody binding to the 0-subunit of deglycosylated chorionic gonadotropin converts the antagonist to an agonist. J Biol Chem 262:3891-3896 47. Hattori M, Hachisu T, Shimohigashi Y, Wakabayashi K 1988 Conformation of the /J subunit of deglycosylated human chorionic gonadotropin in the interaction at receptor sites. Mol Cell Endocrinol 57:17-23


ROLE OF CARBOHYDRATES IN LH/CG RECEPTOR-hCG INTERACTION 48. Calvo FO, Ryan RJ 1985 Inhibition of adenylate cyclase activity in rat corpora luteal tissue by glycopeptides of human chorionic gonadotropin and the a-subunit of human chorionic gonadotropin. Biochemistry 24:1953-1959 49. Kellokumpu S, Rajaniemi H 1985 Involvement of plasma membrane enzymes in the proteolytic cleavage of luteinizing hormone receptor. Endocrinology 116:707-714 50. McFarland KC, Sprengel R, Phillips HS, Kohler M, Rosemblit N, Nikolics K, Segaloff DL, Seeburg PH 1989 Lutropin-choriogonadotropin receptor: an unusual member of the G protein-coupled receptor family. Science 245:494-499 51. Thotakura NR, Weintraub BD, Bahl OP 1990 The role of carbohydrate in human choriogonadotropin (hCG) action. Effects of Nlinked carbohydrate chains from hCG and other glycoproteins on hormonal activity. Mol Cell Endocrinol 70:263-272 52. Kress BC, Spiro RG 1986 Studies on the glycoprotein nature of the thyrotropin receptor: interaction with lectins and purification

53.

54. 55.

56.

1217

of the bovine protein with the use of Bandeiraea (Griffonia) simplicifolia I affinity chromatography. Endocrinology 118:974-979 Benovic JL, Staniszewski C, Cerione RA, Codina J, Lefkowitz RJ, Caron MG 1987 The mammalian /3-adrenergic receptor: structural and functional characteristics of the carbohydrate moiety. J Recept Rec 7:257-281 Boege F, Ward M, Jiirss R, Hekman M, Helmreich EJM 1988 Role of glycosylation for /?2-adrenoceptor function in A431 cells. J Biol Chem 263:9040-9049 Cervantes-Olivier P, Delavier-Klutchko C, Durieu-Trautmann O, Kaveri S, Desmandril M, Strosberg AD 1988 The /32-adrenergic receptors of human epidermoid carcinoma cells bear two different types of oliogosaccharides which influence expression on the cell surface. Biochem J 250:133-143 Azhar S, Menon KMJ 1981 Receptor-mediated gonadotropin action in the ovary. Inhibitory actions of concanavalin A and wheatgerm agglutinin on gonadotropin stimulated cyclic AMP and progesterone responses in ovarian cells. Biochem J 200:153-159


CG receptor.

CG Receptors.

CG receptor complementary DNA.

Kathon CG and cosmetic products.

[Radial connections of the central gray (CG) and escape response induced by stimulation of the CG].

CG receptor cDNAs from human thyroid: regulation by tissue specific splicing.

Ovarian stimulation with human chorionic gonadotropin and equine chorionic gonadotropin affects prostacyclin and its receptor expression in the porcine oviduct.

Purification of human chorionic gonadotropin-receptor complex by immunoaffinity chromatography.

Interaction of echinomycin with An.Tn. and (AT)n regions flanking its CG binding site.

Free β-human chorionic gonadotropin, total human chorionic gonadotropin and maternal risk of breast cancer.

Determination of human chorionic gonadotropin.

Human lutropin (hLH) and choriogonadotropin (CG) are assembled by different pathways: a model of hLH assembly.

Characterization of aromatic dehalogenases of Mycobacterium fortuitum CG-2.

Human chorionic gonadotropin enhances trophoblast-epithelial interaction in an in vitro model of human implantation.

NICE CG 174: intravenous fluid therapy in adults in hospital.

Feedback mechanism of human chorionic gonadotropin secretion.

Carboxypeptidase digestion of human chorionic gonadotropin.

Ectopic human chorionic gonadotropin in breast carcinoma.

Beta-human Chorionic Gonadotropin-secreting Lung Adenocarcinoma.

A chorionic gonadotropin-secreting human pituitary cell.

CG-containing oligonucleotides and transcription factor-binding motifs are enriched in human pericentric regions.

Recognition of N-glycoforms in human chorionic gonadotropin by monoclonal antibodies and their interaction motifs.

Universal CG cloning of polymerase chain reaction products.

uPy: a ubiquitous CG Python API with biological-modeling applications.