Identification of regions of the follitropin (FSH) β-subunit that interact with the N-terminus region (residues 9–30) of the FSH receptor

Molecular and Cellular Endocrinology, 93 (1993) 39-46

39

0 1993 Elsevier Scientific Publishers Ireland, Ltd. 0303-7207/93/$06.00 MOLCEL 02979

Identification of regions of the follitropin (FSH) P-subunit that interact with the N-terminus region (residues 9-30) of the FSH receptor Bosukonda Dattatreyamurty

and Leo E. Reichert, Jr.

Department of Biochemistry and Molecular Biology (A-IO), Albany Medical College, Albany, NY 12208, USA

(Received 4 November 1992; accepted 25 January 1993)

Key words: Follicle-stimulating antibody

hormone receptor;

Follicle-stimulating

hormone p-subunit;

Synthetic peptide; Contact region; Anti-peptide

Summary

We have recently identified a region, N-terminus residues 9-30, in the extracellular domain of the follicle-stimulating hormone (FSH) receptor capable of binding FSH, but not luteinizing hormone (LH) or thyroid-stimulating hormone (FSH) (Dattatreyamurty and Reichert (1992) Mol. Cell. Endocrinol. 87, 9-17). The objectives of the present study were to examine the interaction between a synthetic peptide corresponding to this receptor sequence and the p-subunit of FSH, and to identify which FSH-/? regions are involved in the interaction. FSH-P subunit and synthetic FSH-P peptides 1-15, 71-85 and 101-111 effectively bound ‘251-labeled FSH ret-(9-30) peptide, and binding was inhibited by excess unlabeled FSH receptors. Scatchard analysis indicated that the synthetic FSH-P peptides had affinities for FSH ret-(9-30) peptide in the order of lo6 M-’ (K,), with the sum of individual peptide affinities (K, = 1.21 X 10’ M-‘) closely approximating that of the intact p-subunit (1.02 X 10’ M-‘1. Polyclonal antibodies raised against FSH ret-(9-30) peptide completely inhibited the binding of 1251-labeled receptor peptide to hFSH, hFSH-0, and hFSH+ peptides l-15, 71-85 and 101-111. Our results indicate that recognition of FSH-@ by N-terminus region (9-30) of the FSH receptor involves contact with residues in three discontinuous binding regions on FSH-P. The latter finding suggests that these three discontinuous sequence regions of FSH-P may be closely oriented on the hormone surface to form a contiguous region in the three-dimensional structure required for recognition by the N-terminus region (residues 9-30) of the FSH receptor.

Introduction

Follitropin (follicle-stimulating hormone, FSH) produced by the anterior pituitary gland belongs to a family of closely related glycoprotein hormones that includes lutropin (luteinizing hormone, LH), thyrotropin (thyroid-stimulating hormone, TSH) and chorionic gonadotropin (CG). These hormones are heterodimeric in nature, consisting of a common a-subunit in non-covalent association with a hormonespecific /?-subunit (Pierce and Parsons, 1981). Follitropin plays a critical role in regulating gonadal functions. The actions of follitropin are mediated by specific membrane-bound receptors that are coupled to

Correspondence to: L.E. Reichert, Jr., Department of Biochemistry and Molecular Biology (A-lo), Albany Medical College, Albany, NY 12208, USA. Tel. (518) 445-5365; Fax (518) 445-5365.

GTP-binding sites on G, protein (Dattatreyamurty et al., 1987; Zhang et al., 1988), resulting in activation of adenylate cyclase (Heindel et al., 1975; Zhang et al., 1991). It appears that each subunit of FSH is involved in receptor binding and stimulation, in vitro, of target cell steroidogenesis (Erickson et al., 1990; Reichert et al., 1991). Earlier studies from this laboratory (SantaColoma and Reichert, 1990; Santa-Coloma et al., 19901, using a synthetic peptide approach, have shown that the P-subunit of FSH contains multiple receptor-binding regions. Although peptides corresponding to these binding regions show lower individual binding affinities than FSH-/3, together they have been postulated to orient spatially to form contiguous domain(s) in the three-dimensional structure of FSH leading to productive interaction with the receptor (Reichert et al., 1991; Santa-Coloma et al., 1991; Keutmann, 1992). Follitropin receptor has been purified from bovine calf testis and its biological characteristics have been

40

Materials and methods

A:

rFSHR:

8 20 Cys SW Asn Arg Val Phe Leu Cys Gin Asp Ser LysVal Thr

rLHR :

Cys Pro Glu Pro Cys Asp Cyr Ala Pro Asp Gly Ala La

rTSHR:

Cys His Gin Glu Asp Asp Phe Arg VaJ Thr Cys Lyr Glu Leu

rFSHR:

Glu

rLHR :

Cys Pm Gly Pro Arg Ala Gly Leu Ala Arg Im Ser I&

rTSHR:

His Gin lie Pm Ser La

I

I

I

I

neproThr AspLEUPro

Arg

30 Arg Asn Ala Ile Glu Leu

I

/

Pm Pro Ser Thr Gin Thr Leu

B: rFSHR: hFSHR:

rFSHR: hFSHR:

10 Ser Asn Arg Val Phe Im

Cys Gin Asp Ser Lys

20 Val Thr Glu Ile Pro Thr Asp La . ser.

,

Glu

Pro Arg Am .

Fig. 1. A: Sequence comparison between the glycoprotein hormone receptors. Only N-terminus (residues 8-34) revealing the highly variable region among the receptors is shown here. For complete sequence comparison, see Salesse et al. (1991). rFSHR, rat FSH receptor; rLHR, rat LH receptor; rTSHR, rat TSH receptor. Identical residues Cys, Asp, Leu and Pro are indicated by connecting line. Highly variable (unique) region ~rres~nding to residues 9-30 in rFSHR (shown in bold letters) was chosen for study. 8: Sequence comparison between rFSHR and human FSH receptor (hFSHR). Only N-terminus region corresponding to 9-30 residues is shown here. Amino acid residues identical to those of rFSHR are indicated as dots.

described (Dattatreyamurty et al., 1990, 1992). Recent cloning studies using molecular probes derived from selected coding regions of the LH receptor cDNA have provided info~ation about the primary structure of rat FSH receptor (Sprengel et al., 1990). The deduced primary structure of the FSH receptor has several important features in common with other glycoprotein hormone receptors (Salesse et al., 1991). The receptor has a relatively large extracellular hydrophilic Nterminus domain suggested to be involved in hormone-receptor interaction (Braun et al., 1991). Recently, we identified a sequence, residues 9-30, in the extracellular domain of the FSH receptor having no homology with LH or TSH receptors (Fig. 1) and demonstrated that this region was capable of binding to FSH, but not to LH and TSH (Dattatreyamurty and Reichert, 1992). In the present study, we tested 11 overlapping synthetic peptides corresponding to the primary structure of hFSH+ (Fig. 2) to identify which FSH$ regions are involved in interaction with the N-terminus extracellular region represented by residues 9-30 of the FSH receptor. Our results suggest that the N-terminus region (residues 9-30) of the FSH receptor specifically interacts with three discontinuous receptorbinding regions of FSH-/3 which may form a contiguous domain in three-dimensional structure for recognition by the N-terminus (9-30) region of the FSH receptor.

Bovine calf testes were obtained from a local abattoir and kept at -20°C until used for plasma membrane preparation. Lactoperoxidase, ovalbumin, and ribonuclease-A were purchased from Sigma Chemical Company, St. Louis, MO, USA. Sephadex G-2.5 and Ultragel AcA 34 were from Pharmacia-Lo, Piscataway, NJ, USA. PVDF membranes (Immobilon-P) were purchased from Millipore, Bedford, MA, USA. Na”‘I (carrier-free) was obtained from Dupont-New England Nuclear, North Billerica, MA, USA. Solubilized FSH receptor was prepared from calf testes plasma membranes as described previously (Dattatreyamurty et al., 1986, 1990). Peptide synthesis A 23-mer peptide amide (Fig. 1A, in bold letters) corresponding to residues 9-30 in the N-terminus of rat FSH receptor sequence (Sprengel et al., 1990) was synthesized and kindly provided by Dr. Jean River, Clayton Foundation Laboratories for Peptide Biology, the SaIk Institute, San Diego, CA, USA, under contract No. l-HD-T-2907 from the Contraceptive Development Branch, Center for Population Research, NICHHD. The C-terminal tyrosine amide residue is not part of the selected sequence in rat FSH receptor but was included to allow radio-iodination of the peptide. Synthetic peptide amides corresponding to 11 overlapping regions (Fig. 2) in the primary amino acid sequence of hFSH ~-subunit (Watkins et al., 1987; Shome et al., 1988) were prepared by Multiple Peptide Systems (San Diego, CA, USA) utilizing the solid-phase method (Merrifield, 1963) and the tertbutoxycarbonyl protection scheme. The peptide amides were purified by preparative reverse-phase liquid chromatography on octa-decyl-$lica (Water Delta-Pak C1s, Milford, MA, USA; 100 A pore diameter) using a linear acetonitrile gradient (5-100%) in 0.1% trifluoroacetic acid at 30°C. Homogeneity of the peptide was checked by analytical high-performance liquid chromatography (HPLC). Amino acid composition (Bidlingmeyer et al., 1984)

41-55

21-35

I-15

_________“““““”

Cl-

___________“““” ____”

---------------

____________--- ___-----______” 11-25 81-95

-75

____“““““”

31-45

--------------51-65

101-111 ___________

“_““_----------

CAHHADSLYTYPVATQCHCGKCDSDSTDCTVRGLGPSYC5FGEMKE

_______________--------------71-85

91-105

Fig. 2. Overlapping synthetic peptides derived from the primary sequence of hFSH-/3. These peptides were used to identify which of the binding regions on FSH ~-subunit are recognized by N-te~inus region (residues 9-30) of the FSH receptor.

41

and peptide sequence were determined synthesis. Reduction and S-acetamidomethylation peptide amide

to verify the

of FSH receptor

The receptor-(9-30) peptide amide contains a cysteine residue at position 15. To prevent formation of a peptide dimer, the receptor peptide amide was reduced and alkylated as previously described (Dattatreyamurty and Reichert, 1992). The alkylated peptide was purified by gel filtration through Sephadex G-25. Preparation of radio-iodinated receptor peptide amide

The alkylated receptor peptide amide (100 pg) was radio-iodinated with carrier-free 2.5 mCi of Na”‘I as described previously (Dattatreyamurty and Reichert, 1992) using the lactoperoxidase method (Miyachi et al., 1972) with some modifications (Dattatreyamurty et al., 1986). We separated radio-iodinated peptide from free iodide by gel filtration through a column of Sephadex G-25. The column was eluted with 50 mM phosphate buffer, pH 7.5, and 1 ml fractions were collected into tubes containing 1 ml of 0.1% ovalbumin. The specific activity of the alkylated and radio-iodinated receptor peptide amide was 24-27 &i/pg. Binding studies

(1) Experiments to examine the binding of lz51labeled ret-(9-30) peptide amide to hFSH, hFSH-P and synthetic peptides of hFSH-P were carried out at least 4 times under standard assay conditions as previously described (Dattatreyamurty and Reichert, 1992). 10 pg of hFSH-/3, synthetic peptide amides of hFSH+? (Fig. 2), hFSH (which serves as positive control), unrelated proteins such as ovalbumin and ribonuclease-A, or unrelated synthetic 15-mer peptide amide (which serve as negative controls) were each dissolved in 10 ~1 of 25 mM Tris-HCl buffer, pH 7.0, and immobilized on PVDF membranes (Immobilon-P) by using a slot-blot apparatus (VacuSlot-VS). The sample-containing membranes were incubated for 14 h at 4°C in 20 mM Hepes buffer, pH 7.4, containing 3% ovalbumin to block excess protein binding sites on the membrane. The blocking buffer was then replaced with 20 mM Hepes buffer, pH 7.4, containing 0.5% ovalbumin and 10 mM MgCl,, and the membrane samples were again incubated with 1251-labeled FSH ret-(9-30) (730,000 cpm/ml) in the presence or absence of excess solubilized FSH receptor for 18 h, with slow shaking, at room temperature. Finally, the membranes were washed 3 times with 25 mM Hepes buffer, pH 7.4, containing 5 mM MgCI,, air dried and allowed to develop (autoradiography) overnight. The autoradiograms were then subjected to densitometric scanning as described below.

(2) The quantitative binding of 1251-labeled FSH ret-(9-30) to hFSH P-subunit or various synthetic peptides was determined through use of a solid-phase assay technique (Sternberg and Nygren, 1988) as previously adapted to our studies (Santa-Coloma and Reichert, 1991). Filtration plates of 96 wells, containing surfactant-free cellulose ester membranes (Millititer HA, Millipore, Bedford, MA, USA) were used as solid support for hormone subunit or peptides. The wells were prewetted with 25 mM Tris-HCI buffer, pH 7.4 (Tris buffer). 10 pg of unlabeled hormone subunit or synthetic peptide amides were dissolved in 10 ~1 of Tris-HCI buffer, pH 7.4, and each sample (in triplicate) was added to the wells. After incubation for 30 min at room temperature, the wells were again incubated for 14 h at 4°C with Tris buffer containing 2% ovalbumin to block excess protein binding sites on the membrane. The blocking buffer was then removed by vacuum filtration, and the wells containing the samples were further incubated with assay buffer (20 mM Hepes buffer, pH 7.4, 0.5% albumin and 10 mM MgCl,) containing increasing concentrations of 1251-labeled FSH ret-(9-30) in the presence or absence of excess solubilized FSH receptor for 21 h at 4°C and then 1 h at room temperature with slow shaking to reach equilibrium. For separation of bound ‘251-labeled FSH rec(9-30) from free peptide, we utilized a method of rapid filtration with vacuum. The wells were washed 3 times with 25 mM Hepes buffer, pH 7.4, containing 5 mM MgCl,, air dried and punched to collect the membranes which were counted in an autogamma counter. The affinity constant, K,, was determined from specific binding data using the LIGAND program (Munson and Rodbard, 1980). The criteria used for distinguishing between single and two-site models were as indicated in the LIGAND program. (3) The specificity of the binding of 1251-labeled FSH ret-(9-30) peptide to hFSH+? and related synthetic peptides was confirmed by inhibition studies using a rabbit polyclonal antibody raised against the synthetic FSH ret-(9-30) peptide amide. lXI-Labeled FSH ret-(9-30) peptide (- 14 x lo6 cpm in 170 ~1 of 0.05 M phosphate-buffered saline (PBS)/O.l% ovalbumin) was pre-incubated with 375 ~1 of rabbit antiserum to receptor peptide for 2 h at room temperature, and then overnight at 4°C. The incubation mixture was passed through a column of Ultrogel ACA 34 previously equilibrated and washed with PBS/ovalbumin to separate the labeled receptor peptide-antibody complex from free 1251-labeled receptor peptide. 10 pg of hFSH, hFSH+ or related peptide amides were immobilized on PVDF membranes and the sample-containing membranes were then incubated with 3% ovalbumin to block excess protein binding sites on the membranes, as described above. The membrane-containing samples were then incubated with 1251-labeled FSH

42

ret-(9-30) peptide-antibody complex (complex containing N 600,000 cpm/ml; final dilution of antibody was 1:50) for 18 h, with slow shaking, after which the membranes were washed, dried and kept for autoradiography overnight. Autoradiograms were subjected to densitometric scanning as described below. Control experiments were done using equivalent amounts and final dilutions of pre-immune serum. Densitomet~c scanning

Quantitation of bands reflecting the binding of ‘2”I-iabeled FSH ret-(9-30) peptide amide to hFSH/hFSH-P-related peptides in autoradiograms was done by densitometric scanning of the bands by using Bio-Image model 60s (Millipore) equipped with Visage 4.6 software system. Results were expressed in arbitrary densitometric units (ADU). Results birding of 12’I-labeled FSH ret- f9-301 peptide amide to hFSH, hFSH$ and hFSH-j3 peptide amides

When ‘251-labeled ret-(9-30) peptide amide was incubated with hFSH and hFSH+ immobilized on PVDF membrane (Immobilon-P), it effectively bound to the hormone and subunit, as determined by analysis of bands following autoradiography and quantitative scanning densitometry (Fig. 3). When equal amounts (6.67 nmol) of 11 overlapping synthetic peptides representing the entire primary structure of hFSH$ were immobilized on PVDF membranes and incubated with r%labeled FSH ret-(9-30) peptide amide, the Nterminus FSH receptor peptide bound most effectively to FSH$ peptides l-15, 71-85 and 101-111 (Fig. 3). Negligible or no binding was observed to the other eight peptides tested. This, however, was not due to the absence of peptides on the PVDF membranes, since parallel blots after staining with Coomassie blue showed the presence of these peptides on them. i2’Ilabeled FSH ret-(9-30) peptide did not bind to unrelated proteins (URP) (Fig. 3) including bovine serum albumin, ribonuclease-A or a 15-mer unrelated synthetic peptide similarly immobilized on PVDF membrane. Moreover, 1251-labeled FSH ret-(9-30) peptide binding to hFSH, hFSH-/3 and hFSH-@ peptides was completely inhibited by excess solubilized follitropin receptor, providing further evidence of a high degree of specificity for the interactions. Affinity constants (KJ of the binding of ‘251-labeled FSH ret-(9-30) peptide to hFSH-p and hFSH-P peptides

The quantitative binding of ‘251-labeled FSH ret-(930) peptide to hFSH-8 or hFSH+ peptides 1-15, 71-85 and 101-111 was determined utilizing a solidphase assay. A constant amount of FSH-/3 or related

:

-s

0

t-15

21-3s 1 l-25

f%

61-75

41-5s 31-45

51-65

81-95 77-85

101-11, Vl--105

2 ‘;

$ LL

$+ 3

A B-MSH

paptides

or VW

(6.67

nmol)/

B-hFSH

(0.56

nmoI)/

hFSH (0.3

nmol)

Fig. 3. Direct binding of lZSI-labeled FSH receptor peptide amide to immobilized hFSH, hFSHj3, 11 overlapping synthetic peptide amides of hFSH+, unrelated synthetic peptides and unrelated proteins (URP) ribonuclease-A and bovine serum albumin. Each of these preparations was applied in duplicates (10 ,ug/slot) and immobilized on PVDF membranes. The procedure for the binding experiments, subsequent autoradiography and densitometric scanning are as described in Materials and methods. Experiments were carried out at least 4 times. The figure represents a summary of the data (mean-f SD) from densitometric scanning of autoradiograms. Values in arbitrary densitometric units (ADU) (mean+SD) for the FSH-fi peptides l-15, 71-8.5, 101-111, hFSH and FSH-/3 were 1.271~0.06, 0.868rt 0.08, 0.939kO.04, 0.65 kO.01, and 0.884+ 0.04, respectively. Results are expressed in ADU with 100% assigned for FSH-P l-15 peptide.

synthetic peptides was incubated with increasing concentrations of ‘%labeled FSH ret-(9-30) peptide in the presence or absence of excess solubilized intact FSH receptor, as described in Materials and methods. The binding of ‘“I-labeled FSH ret-(9-30) peptide to FSH-@ or related synthetic peptides was dose-dependent and saturable. Scatchard analysis of the quantitative binding data indicated that FSH-fi peptides 1-15, 71-85 and 101-111 bound ‘251-labeled FSH ret-(9-30) peptide with similar affinities (K,) (Table 1). The individual K, values of the three hFSH-p peptides were lower than that of intact hFSH+ (1.02 x 10’ M-l), but sum of the individual peptide-binding affinities (1.2 X lo7 M-’ appro~mated that of the intact subunit. Inhibition of the binding of ‘~‘I-labeled FSH ret-(9-30) peptide to hFSH, hFSH-p and hFSH-/3 peptides by antibodies raised against receptor peptide

Specificity of the binding of radio-iodinated receptor peptide to hFSH, hFSH-j3 and hFSH+ peptides was confirmed by inhibition studies utilizing polyclonal rabbit antibodies raised against FSH ret-(9-30) peptide. When rabbit antibody was pre-incubated with ‘251-labeled FSH ret-(9-30) peptide, it effectively abolished binding of radio-iodinated receptor peptide to hFSH, hFSH-/3, hFSH$ peptides I-15, 71-85 and 101-l 11 immobilized on PVDF membrane (Fig. 4).

43 TABLE 1

Discussion

ASSOCIATION CONSTANTS W,) OF THE BINDING OF ‘=ILABELED FSH rec_(9-30) PEPTIDE TO THE hFSH+ AND hFSH-@ SYNTHETIC PEPTIDE AMIDES Subunit/peptide

Association constant (K,) (+SDl

amides

1.02~10’ M-’ (+0.22) *

hFSH+ hFSH-j3 hFSH-#I hFSH-P hFSH+? hFSH+ hFSH-/3 hFSH$ bFSH-P hFSH-P hFSH$ hFSH+

3.52 x lo6 M-’ (&-0.88) * NB NB NB NB NB NB 4.6~10~ M-‘(+1.26)* NB NB 4.02~10” M-r (+1.02) *

1-15 1l-25 21-35 31-45 41-55 51-65 61-7.5 71-85 81-95 91-105 101-111

NB, negligible or no specific binding. * Between bFSH-@ and hFSH-P peptides, P < 0.05.

Pre-immune rabbit serum which serves as control in this experiment did not inhibit the binding of ‘251labeled ret-(9-30) peptide amide to the FSH/subunit/peptide preparations (Fig. 4). b

z

200 *

8 e; ng

cna

h

hFSH hFSH (0.3

1-15

beta-FSH nmol)/

B-hFSH

(0.56

nmol)/

0

+ Pre-immune

I

+ Antibody against receptor (g-30)

71-E B-hFSH

peptides

serum

peptide

101-111 (6.67

nmol)

Fig. 4. Inhibition of the binding of 1251-labeled ret-(9-30) peptide amide to hFSH, bFSH-P and hFSH-P peptides by antibodies against receptor peptide sequence 9-30. Unlabeled hFSH, hFSH-8 and hFSH+ 1-15, 71-85 and 101-111 peptides each in duplicate (10 pg/slot) were immobilized on PVDF membranes. After blocking with buffer containing 3% ovalbumin, the membranes were incubated with preformed ‘251-labeled ret-(9-30) peptide amide + antibody to receptor peptide complex (- 600,000 cpm/ml, final dilution of antibody was 1: 50; see Materials and methods for details). The procedure for binding experiment, and subsequent autoradiography and densitometric scanning are as described in Materials and methods. Experiments were carried out at least 3 times. Data (mean f SD) are from densitometric scanning of replicate autoradiograms. Values in ADU (mean+ SD) for hFSH, FSH-P, FSH-@ peptides l-15, 71-85 and 101-111 were 1.59+0.30, 2.16f0.22, 2.83kO.15, 1.94+0.29, and 2.1 f0.29, respectively. Results are expressed in ADU witb 100% assigned for hFSH. * Between hFSH-@ 1-15 and 71-85 peptide, P < 0.05. * Between hFSH$? 1-15 and 101-111 peptide, P < 0.05.

Understanding structure-function relationships of the follitropin receptor requires identification of interacting regions between hormone and receptor and elucidation of their role in activation of post-binding events. The present study utilized a synthetic peptide approach to examine the interaction of the extracellular N-terminus region of the FSH receptor, corresponding to residues 9-30 (FSH ret-(9-30)) with discrete domains of FSH p-subunit. Our results are consistent with and extend our previous findings that FSH p- ubunit has multiple binding regions important for specific recognition by its receptor (Santa-Coloma and Reichert, 1990; Santa-Coloma et al., 1990; Reichert et al., 1991). The primary structure of the FSH receptor (Sprengel et al., 1990), as with receptors for other pituitary glycoprotein hormones (Tsai-Morris et al., 1990; Xie et al., 1990; Nagayama and Rapoport, 1992), includes a large extracellular N-terminus domain thought to be involved in the hormone binding (Braun et al., 1991). The N-terminus region, residues 9-30 of the FSH receptor examined in the present study, has several features characteristic of a hormone-binding region. (1) Secondary structure predictions for N-terminus region (residues 9-30) of FSH receptor by the Chou and Fasman analysis (1978) suggest that this region is hydrophilic and predominantly random-coil oriented. (2) Polyclonal antibodies raised against a synthetic peptide corresponding to this N-terminus region (9-30) of the FSH receptor specifically recognize the native follitropin receptor (Dattatreyamurty and Reichert, unpublished) indicating that this domain is surface-oriented and antigenic. (3) Comparison of the primary amino acid sequences of the glycoprotein hormone receptors reveals that the N-terminus region (9-30) of the FSH receptor represents a unique region which has no sequence homology with LH or TSH receptors (Fig. 1A), although it is well conserved between rat and human follitropin receptors (Fig. 1B) (Sprengel et al., 1990; Minegish et al., 1991). Recently, we have presented multiple lines of evidence to indicate that this N-terminus region of the FSH receptor specifically binds FSH, but not LH or TSH, suggesting a high degree of hormone specificity (Dattatreyamurty and Reichert, 1992). Another noteworthy feature of this N-terminus region (residues 9-30) of the FSH receptor is its ability to recognize the hormone-specific P-subunit of FSH. We have shown this by direct binding of ‘251-labeled FSH ret-(9-30) peptide to FSH-P. Scatchard analysis of the binding data indicated that the iodinated receptor peptide amide bound to hFSH$ with an association constant of 1 x 10’ M-l. The specificity of receptor peptide amide binding to FSH-P was demonstrated

44

by several different approaches. First, binding of 1251labeled receptor peptide amide to FSH-/3 was completely inhibited by excess soluble FSH receptor. Second, the radio-iodinated receptor peptide did not bind to unrelated proteins such as bovine serum albumin and ribonuclease-A, or to an unrelated synthetic peptide. Finally, when rabbit antibody against FSH ret-(930) peptide or pre-immune serum was pre-incubated with radio-iodinated receptor peptide amide, the rabbit antibody, but not pre-immune serum which served as control, effectively inhibited the binding of ‘*?-labeled FSH ret-(9-30) peptide to hFSH-P as well as to hFSH. Recent studies from this laboratory (Santa-Coloma and Reichert, 1990; Santa-Coloma et al., 1990, 19911, and others (Vakharia et al., 1990; Campbell et al., 1991) indicated the presence of multiple receptor-binding regions on the FSH P-subunit. Similar studies on receptors for LH and TSH also indicated that each receptor contains more than one structural domain involved in hormone recognition (Atassi et al., 1991; Ji and Ji, 1991; Nagayama and Rapoport, 1992; Roche et al., 1992). Localization of the region-to-region contacts between FSH and its receptor has not yet been reported, but studies to this end are critical to understanding the molecular basis for hormone recognition by the FSH receptor. Since we observed that the Nterminus region (residues 9-30) of FSH receptor is capable of recognizing FSH-P, we undertook to identify FSH-/l regions involved in the interaction. iZIlabeled FSH ret-(9-30) peptide bound to three FSH-P peptides l-15, 71-85 and 101-111 out of 11 overlapping peptides derived from the primary sequence of hFSH-P and tested simultaneously. As with intact hFSH P-subunit, binding of ‘251-labeled FSH ret-(930) peptide to these FSH-P peptides was specific, being completely inhibited by the presence of excess solubilized intact FSH receptor. Rabbit polyclonal antibodies raised against FSH ret-(9-30) peptide, but not pre-immune serum (control), completely inhibited binding of radio-iodinated receptor peptide to FSH-P peptides 1-15, 71-85 and 101-111. These results support the notion that recognition of FSH+l by the N-terminal region (9-30) of the FSH receptor involves multiple contacts. The fact that these contact regions are present in the variable regions of both FSH-P and FSH receptor (Ward et al., 1990; Salesse et al., 1991) suggests that they may play a discriminatory role in specific recognition of FSH by the receptor. Earlier studies from our laboratory have shown that synthetic peptides corresponding to hFSH-P 33-53 and 81-95 strongly inhibited FSH binding to membranebound receptors (Santa-Coloma and Reichert, 1990; Santa-Coloma et al., 1990), while the FSH-P 1-15 peptide was able to induce FSH receptor-mediated uptake of Ca *+ by proteoliposomes (Grass0 et al., 1991). hFSH+? peptides 31-45 and 71-85 possessed

significant but lesser receptor-binding activity (SantaColoma and Reichert, 1990). In our earlier studies which utilized intact, in situ holo-receptor rather than the N-terminus 9-30 domain studied in this report, no binding inhibitory activity was seen with synthetic peptides FSH-P 1-15 or 101-111. Noort et al. (1992) reported that synthetic peptides corresponding to FSH-@ 27-45, but not 87-95, inhibited FSH stimulation of CAMP synthesis in Sertoli cells from immature rats. Campbell et al. (19911, using chimeric recombinant hormones, reported that hFSH-P residues 88-108, but not 33-52, were critical for FSH binding to receptor and steroidogenic activity. Utilizing immunologic techniques, Vakharia et al. (1990) reported that FSH-/? sequences 33-53, 49-67 and 66-85 contained determinants required for receptor binding. The reasons for such variable results are not clear, but are probably related to different experimental approaches utilized. Nevertheless, it seems clear that multiple sites of interaction are required for specific recognition of hFSH p-subunit by its receptor. Taken together, these results suggest a model whereby initial specific binding of FSH involves interaction of its P-subunit with the extracellular N-terminus 9-30 region of the receptor, presumably involving FSH-P regions l-15,71-85 and 101-111, followed by folding of the hormone-receptor complex to allow interaction of other subunit regions, such as hFSH+? 33-53 and 81-95, with the FSH receptor. Each of the latter peptides has been shown to stimulate estradiol synthesis in cultured rat Sertoli cells (Santa-Coloma and Reichert, 1990; Santa-Coloma et al., 1990). The failure of 1251-labeled ret-(9-30) peptide to interact with hFSH-P peptides 11-25, 41-55 and 51-65 is noteworthy because these FSH-P peptides were previously shown to represent a-subunit contact regions in formation of the FSH heterodimer (SantaColoma and Reichert, 1991). Although synthetic hormone peptides typically show binding affinities lower than those of intact hormones, formation of a contiguous domain by several binding regions would be expected to substantially increase affinity for receptor and allow for productive interactions with receptor (Reichert et al., 1991; Keutmann, 1992). Thus, Noort et al. (19921 observed a synthetic peptide corresponding to three discontinuous regions of FSH to have an agonist activity at a concentration, lo-’ M, which was significantly greater than that of the individual peptides, the agonist activity being defined as the ability to stimulate basal CAMP levels in Sertoli cell membranes from immature rats. Formation of such a contiguous domain in three-dimensional structure was suggested in our earlier studies (SantaColoma et al., 1991) wherein two discontinuous regions of hFSH-@, namely residues 33-53 and 81-95, were synthesized as a single peptide (omitting the intervening sequence 54-80) and found to bind receptor with

45

an affinity higher than that of the individual peptides. In this study, FSH ret-(9-30) peptide bound to FSH-/3 peptides l-15,71-85 and 101-111, with relatively lower affinities (K,, in the order lo6 M-l) than to hFSH$. However, the sum of affinities calculated for FSH-/3 peptides 1-15, 71-85 and 101-111 approximated that observed for FSH-P (K,, 1.02 X 10’ M-i). By virtue of the extensive folding and disulfide linking within hormone subunit (between Cys3 and Cyss4, Cys” and cy?, cys*o and CJJS~‘~)(Ward et al., 1990), important residues from the three receptor-binding regions 1-15, 71-85 and 101-111 along the linear hFSH+? chain may closely orient on the hormone surface to form a contiguous region in three-dimensional structure of FSH, for productive recognition by the N-terminus (9-30) region of the FSH receptor. Our results suggest that the unique N-terminus region (residues 9-30) of the FSH receptor (i.e., not found in LH or TSH receptors) specifically binds FSH P-subunit. Also, we present evidence that recognition of FSH-P by this region of its receptor occurs through contacts between the latter and multiple discontinuous binding regions of FSH-p, including regions l-15, 7185 and 101-111. It is not yet known which other potential hormone-binding region(s) of the FSH receptor are in contact with the FSH p-subunit. Computerbased predictions using a sense-antisense similarity approach suggest that other FSH P-subunit contact regions may reside in the conserved regions of the FSH receptor (Slootstra and Roubos, 1991). Available evidence is consistent with a model wherein the interactions of FSH with its receptor involve an initial specific recognition through contacts between N-terminus (residues 9-301 region of the FSH receptor and multiple binding regions on hFSH-P (regions 1-15, 71-85 and 101-111) followed by a conformational change permitting receptor interaction with additional domains of FSH-/3 such as regions 33-53 and 81-95. Previously, FSH-P peptides 33-53 and 81-95 were shown to bind to membrane-bound holo-receptors as well as stimulate estradiol synthesis by cultured rat Sertoli cells (SantaColoma and Reichert, 1990; Santa-Coloma et al., 1990). Acknowledgments

We wish to thank Carol Kowal for excellent technical assistance, and the Division of Molecular Medicine, Department of Medicine, Albany Medical College, for use of the Bio-Image Model 60s Densitometric Scanner. Supported by NIH Grant HD-13938. References Atassi, M.Z., Manshouri, T. and Sakata, S. (1991) Proc. Natl. Acad. Sci. USA 88, 3613-3617.

Bidlingmeyer, B.A., Cohen, S.A. and Tarvin, T.L. (1984) J. Chromatogr. 336, 93-104. Braun, T., Schofield, P.R. and Sprengel, R. (1991) EMBO J. 10, 1885-1890. Campbell, R.K., Dean-Emig, D.M. and Moyle, W.R. (1991) Proc. Natl. Acad. Sci. USA 88, 760-764. Chou, P.Y. and Fasman, G.D. (1978) Annu. Rev. B&hem. 47, 251-276. Dattatreyamurty, B. and Reichert, Jr., L.E. (1992) Mol. Cell. Endocrinol. 87, 9-17. Dattatreyamurty, B., Schneyer, A. and Reichert, Jr., L.E. (1986) J. Biol. Chem. 261, 13104-13113. Dattatreyamurty, B., Figgs, L.W. and Reichert, Jr., L.E. (1987) J. Biol. Chem. 262, 11737-11745. Dattatreyamurty, B., Zhang, S.-B. and Reichert, Jr., L.E. (1990) J. Biol. Chem. 265, 5494-5503. Dattatreyamurty, B., Smith, R.A., Zhang, S.-B., Santa-Coloma, T.A. and Reichert, Jr., L.E. (1992) J. Mol. Endocrinol. 9, 83-92. Erickson, L.D., Rizza, S.A., Bergert, E.R., Charlesworth, M.C., McCormick, D.J. and Ryan, R.J. (1990) Endocrinology 126, 25552560. Grasso, P., Santa-Coloma, T.A. and Reichert, Jr., L.E. (1991) Endocrinology 128, 2745-2751. Heindel, J.J., Rothenberg, R., Robinson, G.A. and Steinberger, A. (1975) J. Cyclic Nucleotide Res. 1, 69-79. Ji, I. and Ji, T.H. (1991) Endocrinology 128, 2648-2650. Keutmann, H.T. (1992) Mol. Cell. Endocrinol. 86, Cl-C6. Merrifield, R.B. (1963) J. Am. Chem. Sot. 85, 2149-2154. Minegish, T., Nakamura, K., Takakura, Y., Ibuki, Y. and Igarashi, M. (1991) Biochem. Biophys. Res. Commun. 175, 1125-1130. Miyachi, Y., Vaitukaitis, J.L., Nieschlag, E. and Lipsett, M.B. (1972) J. Clin. Endocrinol. Metab. 43, 23-28. Munson, P.J. and Rodbard, D. (1980) Anal. Biochem. 107, 220-239. Nagayama, Y. and Rapoport, B. (1992) Mol. Endocrinol. 6, 145-156. Noort, M.H., Puuk, WC., Plasman, H.H., Kuperus, D., Schaaper, W.M.M., Beekman, N.J.C.M., Grootegoed, J.A. and Maloen, R.H. (1992) Proc. Natl. Acad. Sci. USA 89, 3922-3926. Pierce, J.G. and Parsons, T.F. (1981) Annu. Rev. Biochem. 50, 465-490. Reichert, Jr., L.E., Dattatreyamurty, B., Grasso, P. and Santa-Coloma, T.A. (1991) Trends Pharmacol. Sci. 12, 199-203. Roche, P., Ryan, R.J. and McCormick, D.J. (1992) Endocrinology 131, 268-274. Salesse, R., Remy, J.J., Levin, J.M., Jallal, B. and Garnier, J. (1991) Biochemie 73, 109-120. Santa-Coloma, T.A. and Reichert, Jr., L.E. (1990) J. Biol. Chem. 265,5037-5042. SantaColoma, T.A. and Reichert, Jr., L.E. (1991) J. Biol. Chem. 266, 2759-2762. Santa-Coloma, T.A., Dattatreyamurty, B. and Reichert, Jr., L.E. (1990) Biochemistry 29, 1194-1200. Santa-Coloma, T.A., Crabb, J.W. and Reichert, Jr., L.E. (1991) Mol. Cell. Endocrinol. 78, 197-204. Shome, B., Parlow, A.F., Liu, W.-K, Nahm, H.S., Wen, T. and Ward, D.N. (1988) J. Protein Chem. 7, 325-339. Slootstra, J.W. and Roubos, E.W. (1991) B&hem. Biophys. Res. Commun. 179,266-271. Sprengel, R., Braun, T., Nikolics, K., Segaloff, D.L. and Seeburg, P.H. (1990) Mol. Endocrinol. 4, 525-530. Stenberg, M. and Nygren, H. (1988) J. Immunol. Methods 113,3-15. Tsai-Morris, C.H., Buczko, E., Wang, W. and Dufau, M.L. (1990) J. Biol. Chem. 265, 19385-19388. Vakharia, D.D., Dias, J.A., Thakur, A.N., Andersen, T.T. and O’Shea, A. (1990) Endocrinology 127, 658-666. Ward, D.N., Bousfield, G.R. and Mar, A.O. (1990) in StructureFunction Relationship of Gonadotropins (Bellet, D. and Bidart, J.M., eds.), Vol. 65, pp. 1-19, Raven Press, New York.

46 Watkins, P.C., Eddy, R., .Beck, A.K., Velucci, V., Leverone, B., Tanzi, R.E., Gusella, J.F. and Shows, T.B. (1987) DNA 6, 201% 212. Xie, Y.-B., Wang, H. and Segaloff, D.L. (1990) J. Biol. Chem. 265, 21411-21414.

Zhang, S-B., Dattatreyamur~, B. and Reichert, Jr., L.E. (1988) Mol. Endocrinol. 2, 148-158. Zhang, S.-B., Dattatreyamurty, B. and Reichert, Jr., L.E. (1991) Endocrinology 128, 295-302.