Human Leptin Receptor

The leptin receptor (OB-R) is a member of the class I cytokine receptor family and mediates the weight regulatory effects of its ligand through interaction with cytoplasmic kinases. The extracellular domain of this receptor is comprised of two immunoglobulin-like and cytokine-receptor homology domains each and type III fibronectin domains. The extracellular domain of human leptin receptor was expressed in and purified from Chinese hamster ovary cells and was found to contain extensive N-glycosylation (approximately 36% of the total protein). The purified protein had a molecular weight of approximately 145,000 and exhibited ligand binding ability as evidenced by formation of ligand-receptor complex, followed by chemical cross-linking. The determined disulfide motif of the soluble leptin receptor contained several distinct cystine knots as well as 10 free cysteines. The N-glycosylation analysis revealed that Asn624 of the WSXWS motif (residues 622–626) within the C-terminal cytokine receptor homology domain was glycosylated, indicating that this region is solvent-exposed. On the other hand, the N-terminal WSXWS motif was not glycosylated.

The ob gene product, leptin, is an important circulating signal for regulation of body weight (1)(2)(3)(4). Weight reducing effects of recombinant leptin were observed in both normal mice and mice with diet-induced obesity (5). The biological function of leptin is mediated through its membrane-associated receptor, leptin receptor (OB-R). 1 Because of the biological and therapeutic importance of the leptin/OB-R system, numerous studies have been carried out to understand how this interaction is related to body weight regulation (6 -9).
Human OB-R is a membrane-spanning glycoprotein consisting of a signal sequence, two immunoglobulin domains, two cytokine receptor homology (CRH) domains each containing a WSXWS motif, fibronectin type III domains, a transmembrane region, and an intracellular domain. The predicted extracellular domain consists of 839 amino acid residues and shows significant similarity to members of the class I cytokine receptor family (6), in particular the gp-130 signal-transducing component of various cytokine receptors as well as the granulocytecolony-stimulating factor receptor (G-CSFR) (10). Although overall sequence identity between OB-R and gp-130 is low (approximately 24%), several key regions are conserved, particularly the second cytokine homology domain (CRH)-2 (6). In this paper, we report structural information including the disulfide motif and N-glycosylation map of the extracellular domain of OB-R.

EXPERIMENTAL PROCEDURES
Materials-Pepsin was obtained from Sigma. Trypsin and thermolysin were purchased from Boehringer Mannheim. Cross-linkers BS and DSS were obtained from Pierce and 4-HCCA from Sigma. The N-and O-glycanases were purchased from Genzyme (Cambridge, MA), and sialidase was from Boehringer Mannheim. Recombinant human leptin was prepared as described previously (11). All chemicals used were of reagent grade or analytical grade.
Antisera Preparation-New Zealand White rabbits were injected subcutaneously on day 1 with 0.2 mg of soluble OB-R and an equal volume of Freund's complete adjuvant. Further boosters (days 7, 21, 35, and 56) were given with the substitution of Freund's incomplete adjuvant. Antibody titers were monitored by enzyme-linked immunoassay. After the third booster, 20 ml of blood was obtained from each animal. The antisera for leptin was affinity purified on a leptin-Sepharose column and then conjugated to horseradish peroxidase using a Freezyme Kit (Pierce).
Western Blotting-Samples were electrophoresed under reducing conditions on 4 -20% SDS-PAGE gels (Novex, San Diego) and then transferred to BA83 nitrocellulose (Schleicher & Schuell) overnight at 10 V in 25 mM Tris, 192 mM glycine, 20% v/v methanol. After transfer, blots were treated for 1 h in 1ϫ PBS, Tween 20 (Bio-Rad) containing 5% w/v non-fat dry milk before adding antisera (1/2000 dilution) or antibody (0.1 g/ml) for an additional hour. The blots were then washed with PBS/Tween 20 four times, 10 min each. For detection of receptor, a secondary anti-rabbit Ig (Amersham Pharmacia Biotech.) was diluted 1/6000 in PBS/Tween 20 and reacted with the blot for 30 min, afterward repeating the washes. Blots were then treated with enhanced chemiluminescence detection reagents (Amersham Corp.) and exposed to Kodak AR film.
Expression and Purification of an Extracellular Domain of OB Receptor-DNA encoding the extracellular domain and signal sequence of the human leptin receptor (residues 1-839) (6) was cloned into the mammalian expression vector pDSR␣ (12) and then transfected into CHO D Ϫ cells (13). Individual cell colonies were selected based upon expression of a dihydrofolate reductase gene in the vector. The highest expressing cells were adapted to 30 mM methotrexate to stimulate amplification of soluble OB-R expression. After expansion in spinner flasks, roller bottles were inoculated with 2 ϫ 10 7 cells in 200 ml of Dulbecco's modified Eagle's medium:Ham's F12 supplemented with nonessential amino acids (Life Technologies, Inc.) and 5% fetal bovine serum (HyClone, Logan, UT). After the cells reached confluency (3-4 days), the medium was replaced with 200 ml of growth medium with no serum. Conditioned media were harvested after 7 days and immediately chilled on ice. CHO cell-conditioned media containing soluble OB-R were concentrated 10-fold with a Pellicon tangential flow ultra-* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Chemical Cross-linking Experiments-A stable complex was made by incubating soluble OB-R (100 g) with a 2-fold molar excess of recombinant human leptin at 4°C for 24 h. The ligand-receptor complex was purified by Superose 12 (10 ϫ 300 mm) equilibrated with 1ϫ PBS, pH 7.8. Cross-linking was performed by treating the complex with 1 mM DSS (dissolved in Me 2 SO) or BS (dissolved in water) for 1 h at room temperature. The reaction was terminated by addition of 0.1 M ammonium acetate. The reactions were analyzed by SDS-PAGE under reducing conditions. Cross-linked products were detected by both silver staining and Western blotting using an affinity purified, horseradish peroxidase-conjugated, polyclonal antibody raised against human leptin.
Proteolytic and Chemical Cleavages of Soluble OB-R-In order to obtain Cys-containing peptides, soluble OB-R (100 g) was digested with pepsin (2 g) in 0.02 N HCl, pH 2, for 20 h at 37°C. Under acidic conditions, disulfide rearrangements can be avoided. The digestion was terminated by direct injection into an HPLC column. Tryptic digestion of soluble OB-R was performed in the presence of 5 mM iodoacetate in 0.1 M Tris-HCl, pH 7.5. The digestion was carried out at 37°C for 24 h with an enzyme:substrate ratio of 1:50 (w/w). Thermolytic or chymotryptic digestions of peptide fragments were performed in 0.1 M Tris-HCl, pH 7.5, at 37°C for 20 h with an enzyme:substrate ratio of approximately 1:20 (w/w). Cyanogen bromide cleavage of OB-R (100 g) was performed with 2 mg of cyanogen bromide dissolved in 100 l of 70% formic acid. The reaction was allowed to proceed for 24 h at 25°C in the dark and was stopped by direct HPLC injection. The cyanogen bromide peptides were treated with N-glycanase (0.5 units) in 0.1 M Tris-HCl buffer, pH 7.5, and then digested with endoproteinase Asp-N, followed by endoproteinase Glu-C in the same buffer with an enzyme: substrate ratio of 1:50 (w/w).
Deglycosylation of Soluble OB-R-Soluble OB-R was reduced and denatured by adding SDS to 0.4% and ␤-mercaptoethanol to 50 mM and then heating at 95°C for 5 min. Nonidet P-40 was then added to 1.8% before diluting the reduced and denatured receptor with an equal volume of distilled water. Aliquots of 2.5 g were then digested with sialidase (5 milliunits), O-glycanase (0.5 units), N-glycanase (1 unit), or all three enzymes at 37°C for 18 h. A mock reaction containing no glycosidase was treated identically to the digested samples, all of which were then analyzed by SDS-PAGE on 8% polyacrylamide gels (Novex, San Diego) under reducing conditions. Protein bands were visualized by silver staining and Western blotting. Deglycosylation of glycopeptides was performed by treatment with N-glycanase (0.5 units) in 0.1 M Tris-Cl buffer, pH 7.5, at 37°C for 24 h. The digested material was purified by reversed phase HPLC for further analysis.
Peptide Isolation and Sequence Analysis-Peptides were separated by reversed phase HPLC using a Vydac C18 column (2.1 ϫ 150 mm) with two linear gradients from 2 to 40% solvent B over 30 min and from 40 to 60% solvent B over 10 min at a flow rate of 0.25 ml per min. Cyanogen bromide-generated peptides were separated on a Zorbax CN column (2.1 ϫ 150 mm) using a linear gradient from 2 to 60% solvent B over 30 min. The flow rate was the same as above. N-terminal sequence analyses of peptides and proteins were performed on a model 476A or 477A ABI gas sequencer system from Applied Biosystems (Foster City, CA). For analysis of diPTH-Cys, an ABI model 494 was used with a pulsed liquid program. Data analysis was performed with the Applied Biosystems model 610A data analysis system for protein sequencing, version 1.2.2. The peptides were loaded onto Biobrene-treated glass fiber filters for sequencing.
Partial Reduction and S-Alkylation of Cyanogen Bromide Peptide CB-1-The cyanogen bromide-generated peptide CB-1 that contained multiple disulfide bonds was partially reduced with 10 mM Tris-(2carboxyethyl)phosphine at 25°C for 5 min in 0.1% trifluoroacetic acid. The sequence was deduced from cDNA sequence (6). The double lined sequence shows a signal peptide.
The reduced peptide was alkylated with 20 mM N-ethylmaleimide for 30 min at room temperature after adjusting to pH 6.0 with 1 M Tris base as described previously (15). The modified peptides were sequenced after purification by reversed phase HPLC using a Zorbax CN column (2.1 ϫ 150 mm) as described above.
Mass Spectrometry-Matrix-assisted laser desorption ionization (MALDI) mass spectrometry of the peptides was performed using either a Kratos IV (Kratos Analytical) or Voyager mass spectrometer (PerSeptive Biosystems). Samples were dissolved in 0.1% trifluoroacetic acid, 50% acetonitrile and then spotted on the sample cartridge with a 4-HCCA matrix. Cys-containing peptides were also analyzed using a Sciex API triple quadrupole mass spectrometer with an ion-spray interface using a Michrom Biosource Ultrafast Michroprotein Analyzer. The carrier solvent was 50% acetonitrile/water with 0.1% trifluoroacetic acid flowing at 5 l/min. The scan range was 300 -2400 atomic mass unit with a step of 0.5 atomic mass unit. The mass units and standard deviation were calculated using Sciex hypermass software.

Molecular Properties of the Soluble OB-R-Human soluble
OB-R was expressed in CHO cells and purified from conditioned medium by Q-Sepharose, hydroxylapatite, and gel filtration. The purified soluble OB-R exhibited a single, broad band on SDS-PAGE, corresponding to approximately 145,000 daltons ( Fig. 1). Deglycosylation of the purified soluble OB-R with N-glycanase reduced the molecular mass to approximately 90,000 daltons, indicating extensive N-glycosylation al- The protein sample (100 g) was digested with pepsin (2 g) and directly injected to a Vydac C18 column (2.1 ϫ 150 mm) as described under "Experimental Procedures." though treatment with sialidase and O-glycanase slightly increased its mobility, suggesting that O-glycosylation is present but not significant. The N-terminal sequence of the purified protein was identified to be FXLSYPITP (residues 22-30), whereas the C-terminal sequence of DIEKHQSD (residues 832-839) was identified from a pepsin-generated fragment.
The observed N-and C-terminal sequences were identical to those predicted from the cDNA-deduced sequence reported by Tartaglia et al. (6), which is shown in Fig. 2. Thus, the soluble OB-R purified here contained 818 amino acid residues (from residues 22 to 839) with a calculated molecular weight of 93,501, indicating that the OB-R contained approximately 36% carbohydrates of its total mass.
Ligand Binding Activity and Preparation of Ligand-Receptor Complex-The leptin-OB-R complex was prepared by mixing the receptor with a 2-fold molar excess of leptin. After incubation at 4°C for 24 h, the mixture was subjected to size exclusion chromatography using a Superose 12 column (10 ϫ 300 mm) equilibrated with PBS, pH 7.8. Two peaks were obtained, corresponding to complex and free leptin. From the elution position in Superose 12 column, it is likely that the complex contained only one molecule of OB-R. The complex peak was subjected to reversed phase HPLC using a Zorbax CN (2.1 ϫ 150 mm), which indicated that it contained a 1:1 ratio of leptin and OB-R (Fig. 3).
Chemical Cross-linking of the Ligand-Receptor Complex-In order to examine ligand binding specificity of OB-R, the purified complex was cross-linked with 1 mM BS and DSS, followed by SDS-PAGE analysis. One gel was silver-stained and other electroblotted to nitrocellulose and detected with an affinity   purified, polyclonal antibody raised against recombinant human leptin. In the silver-stained gel (Fig. 4A), a protein band just above that corresponding to free receptor is observed. Fig.  4B shows the Western blot of this gel using an 8% gel, verifying the presence of leptin (cross-linked to the OB-R) in this band. Pepsin-generated Peptide Map of the Soluble OB-R for Determining Disulfide Linkages-The soluble OB-R was fragmented by peptic digestion, resulting in numerous peptide peaks separated by reversed phase HPLC (Fig. 5). Therefore, sequence analysis was performed on all peptic peptides and indicated that more than 15 Cys-containing peptides were isolated. Table  I shows the summary of the sequence analyses of the Cyscontaining peptides including both disulfides and free cysteines. Peak 8 contained two peptides, VRX (residues 611-613) (P8a) and LTKMTXRW (residues 442-449) (P8b), in which X was expected to be a cysteine residue according to the cDNA sequence. MALDI mass spectrometry showed that these peptides (P8a and P8b) were not linked together through a disulfide bond but existed as free cysteines, Cys 447 and Cys 613 , respectively. Although the two peptides are different in size, they co-eluted from the HPLC column under the conditions used.
Peptic peptide P16 contained the single sequence MKND-SLSX (residue 668 -674), in which X is an unidentified residue but is a Cys residue based on the cDNA sequence. Since this peptide was solely isolated, Cys 674 was assigned to be in the    free state. Mass spectrometry of this peptide revealed it not to be glycosylated although it contained an N-glycosylation sequence, NXS. Two peptides, P18 and P19, had the same sequence of RYVINHHTSXNGTW (residues 678 -691) where N denotes an N-glycosylation site. Mass spectrometry of the intact peptides showed a larger mass than expected. After deglycosylation with N-glycanase, these peptides showed the same mass of 1689.0, which is nearly identical to the expected mass 1687.9. The observed difference in HPLC retention time is evidently due to carbohydrate heterogeneity, and the data indicated that Cys 687 was in a free state. Peptide P21 contained two peptides, HXIYKKENKIVPSKE (residues 351-365) and AVYXXNEHEXHHRY (residues 409 -422), where X indicates a cysteine residue. Since these two peptides contain four cysteine residues, their co-elution suggests one inter-chain disulfide with the remaining two forming an intra-chain disulfide. Mass spectrometry of this peptide subsequently suggested that these two peptides were linked together through a disulfide bond. Fig. 6 shows mass spectral data of peptide P21, indicating the mass 3573.3 for intact peptide and a signal at 1761.7 derived from one of the components. Exact disulfide linkages were determined later from tryptic fragments.
Peptide P27a contained the single sequence ED-SPLVPQKGSFQMVHXNXSVHEXXEXL (residues 170 -197). Since this peptide contains five cysteines according to the DNA sequence, at least one must be in the free state. Disulfide linkages in this peptide were determined after cyanogen bromide cleavage, as discussed later. Additionally, Asn 187 was glycosylated since no PTH-Asn was detected at this cycle. Meanwhile, peptide 27b contained two sequences, KLSXMP-PNSTY (residues 34 -44) and SKTTFHXXFRSEQDRNXS-LXAD (residues 83-104), containing one of three cystine knots in the receptor. However, disulfide linkages in these peptides could not be determined from this digestion. Further digestion of this peptide with thermolysin gave the three sequences, LSXMPPNSTY (residues 35-44), FHXX (residues 87-90), and RSEQDRNXSLXAD (residues 92-104). Asn 41 and Asn 98 were assigned to be N-glycosylation sites, as determined by sequence analysis. Table II shows sequence analysis of peptide P27b TH1. Upon PTH analysis of this peptide, the third cycle did not show significant diPTH-Cys, indirectly suggesting that Cys 37 is not disulfide-linked to Cys 89 but probably to Cys 90 . The complete disulfide structure of this Cys cluster was eventually determined by tryptic digestion as discussed later.
Peptide P28a corresponded to the same sequence as peptide P21 but contained another seven N-terminal amino acids. Peptide P28b showed a single peptide MCLK (residues 211-214) with one free cysteine, Cys 212 . Sequence analysis of peptide P29 showed two sequences, PKDXY (residues 485-489) and YEXIFQ (residues 496 -501), clearly demonstrating the presence of a disulfide linkage between Cys 488 and Cys 498 . This result was confirmed by mass spectrometry, experimental mass 1425.0 versus the theoretic mass 1423.0. Similarly, peptide P33 showed two sequences, NIQXW (residues 128 -132) and IXY (residues 141-143), again using mass spectrometry to confirm that Cys 131 and Cys 142 are linked together. Peptide P34b contained the two sequences, YXSDIPSIHPISE (residues 472-484) and DSPPTXVLPDS (residues 523-533), and similarly the molecular mass of this peptide was 2583.4, close to the Cyanogen bromide digest of OB-R was subjected to reversed phase HPLC using a Zorbax CN column (2.1 ϫ 150 mm). The peptides were eluted with a linear gradient from 2 to 60% B over 30 min. theoretic mass of 2587.2, clearly demonstrating a disulfide linkage of Cys 473 -Cys 528 . A glycopeptide P36, SAYPLNSSC (residues 745-753) contained a free cysteine, Cys 753 , although mass spectrometry of this peptide was not successful due to its limited amount. Since peptide P37 had the single sequence VSLPVPDLXAVYA (residues 596 -608), confirmed by mass spectrometry, it contained a free cysteine Cys 604 . Meanwhile, peptide P38 contained two cysteines in a single peptide ISX-ETDGYLTKMTXRWSTSTIRSLA (residues 434 -458), and both Cys residues indicated by X might exist in the free state since the second cysteine was previously detected as such in the peptide P8b.
Determination of the Disulfide Linkages of Three Cystine Knots-In order to determine the disulfide linkages of three cystine knots present in OB-R, intact receptor was digested with trypsin in the presence of 5 mM iodoacetate. The peptide map shown in Fig. 7 gave several key Cys peptides for disulfide assignments, and sequences of selected tryptic peptides are summarized in Table III. Tryptic fragment T1 contained five cysteines in two peptides, one of which (Cys 102 ) might be in the free state since Cys of the human receptor is substituted by Leu in both murine and rat OB receptors. Disulfide linkages were determined by PTH analysis after further digestion of this peptide with endoproteinase Glu-C, resulting in a new fragment (T1-E1). Peptide T1-E1 consisted of three sequences, LSCMPPNSTYDYFLLPAGLSK (residues 35-55), TTFHC-CFRSE (residues 85-94), and QDRNCSLCADNIE (residues 95-107), respectively. When sequenced, the fifth cycle corresponding to Cys 89 and Cys 99 showed significant yields (ϳ50%) of diPTH-Cys (Table IV) (16,17). To confirm this result, a thermolytic digest of peptide T1 provided two sequences LSC-MPPNSTYDY (residues 35-46) and HCCFRSEQDRNCS(residues 88 -100) which also showed significant recovery of diPTH-Cys at cycle 3 corresponding to Cys 37 and Cys 90 . Subsequently, these results indicated that Cys 89 was linked to Cys 99 , supporting another linkage Cys 37 -Cys 90 . Cys 102 was assigned to be a free state.
Peptide fragment T2 contained two peptides when sequenced and was further digested with chymotrypsin to obtain specific peptides, in which cysteine residues are located at the same position (16,17). From the chymotryptic digest, two Cys-containing peptides were isolated as follows: T2-CT1 containing HXIY (residues 351-354) and XXNEHEXHHR (residues 412-421), and another peptide T2-CT2, which contained XIY (residues 352-354) and XXNEHEXHHR (residues 412-421), where X denotes a Cys residue. As shown in Table V, mass spectrom-   etry of peptide T1-CT2 showed a signal of 1662.4, which is identical to the theoretical mass (Fig. 8). When peptide T2-CT1 was sequenced, no significant diPTH-Cys was detected at any cycle upon PTH analysis, whereas peptide T2-CT2 showed a significant diPTH-Cys at the first cycle corresponding to Cys 352 and Cys 412 . Recovery of diPTH-Cys was approximately 95%, demonstrating the presence of Cys 352 -Cys 412 . Since mass spectrometry indicated that peptide T2-CT2 did not contain any carboxymethylcysteine, the remaining cysteine linkage was assigned to be Cys 413 -Cys 418 . Peptide T3 has a sequence GSFQMVHCNCSVHECCE-CLVPVPTAK (residues 179 -204) in which an N-glycosylated asparagine is underlined (Table III). This single peptide contained five cysteine residues and, as before, one must be in the free state. Since the peptide was recovered in low yield, we isolated a corresponding peptide after CNBr cleavage of intact soluble OB-R. Fig. 9 shows the HPLC peptide map of the CNBr fragments. The peptide CB1 contained a sequence VHCNCS-VHECCECLVPVPTAKLNDTLLM (residues 184 -211). Mass spectral analysis revealed that deglycosylated CB1 has a mass of 3346, corresponding to a single glutathione adduct (Table  VI). To determine the two disulfide bonds, the CB1 was partially reduced with 10 mM Tris-(2-carboxyethyl)phosphine at 25°C for 5 min, followed by alkylation with NEM in order to identify a preferentially reduced disulfide bond. Fig. 10A shows an HPLC profile after partial reduction and alkylation of peptide CB-1. Sequence analysis of the peptide CB1-NEM1 revealed that only Cys 188 and Cys 193 were detected as PTH-Nethylsuccinimidocysteines appearing as double peaks between PTH-Pro and PTH-Met. Peptide CB1-NEM2 contained four NEM-modified cysteines, Cys 186 , Cys 188 , Cys 193 , and Cys 196 , respectively. Finally, the peptide CB1-NEM3 contained five NEM-modified cysteines including additional Cys 194 (Table  VI). These results suggested the presence of one disulfide linkage, Cys 186 -Cys 196 , and thus suggested another linkage, Cys 188 -Cys 193 . Cys 194 was therefore assigned to be glutathione adduct. This assignment was confirmed by direct isolation of Cys peptides from peptide CB1. The CB1 peptide was deglycosylated with N-glycanase treatment and was digested with endoproteinase Asp-N, followed by endoproteinase Glu-C. The peptides were purified by reversed phase HPLC as shown in Fig. 10B. Peptide CB1-D.E-1 contained the sequence DCS-VHECC (residues 187-194), showing a small amount of diPTH-Cys at cycle 7 corresponding to Cys 193 (Table VII). This suggests a presence of disulfide linkage Cys 188 -Cys 193 . Mass spectrometry indicated one of the cysteines (presumably Cys 194 ) was modified probably with ␥-EC (mass 250), a breakdown product of glutathione. Meanwhile, peptide peak CB1-D.E-2 contained two peptides VHC (184 -186) and CLVPVPT (196 -202), indicating a disulfide linkage Cys 186 -Cys 196 , as confirmed by mass spectrometry (Table VII).
Determination of N-Glycosylation Sites-Sequence analyses of both peptic and tryptic fragments provided information on the extensive N-glycosylation of the OB-R. Determination of the glycosylation sites relied upon no detection of Asn, obtained in sequence analysis. As indicated in Table I, the Cys-containing peptides have eight N-glycosylation sites. The remaining 10 N-glycosylation sites were identified by sequence analysis and mass spectrometry of other peptic or tryptic peptides. Some peptides were re-treated with N-glycanase in order to obtain FIG. 11. Disulfide structure and N-glycosylation map of the extracellular domain of OB-R. Three cystine clusters were schematically drawn as well as free cysteine residues. N-Glycosylation sites were indicated as rhombic marks. CRH, cytokine receptor homology domain; WSXWS, a common motif in cytokine receptor.

TABLE VIII
Potential N-glycosylation sites of human OB-R their exact mass. Table VIII shows all the possible N-glycosylation sites of the human receptor and a comparison with the corresponding murine sequence. Almost every N-glycosylation sequence was found to be modified by carbohydrates and only two sites were not, Asn 433 and Asn 670 . Interestingly, residue Asn 624 located at the second WSXWS motif was found to be N-glycosylated. Since the WSXWS motif has been shown to be important in the folding of interleukin-2 receptor or at least its ligand binding region (19), N-glycosylation at Asn 624 may affect the folded structure of OB-R. From peptide mapping, 18 of 20 N-glycosylation sequences might be occupied by N-glycans. An overall structural diagram of the extracellular domain of human OB-R detailing 9 disulfide linkages, 10 free cysteines, and 18 N-glycosylation sites is shown in Fig. 11. These disulfide and N-glycosylation motifs are not homologous with any known cytokine receptors including G-CSFR (18). DISCUSSION In comparing the sequences of the human and murine OB-R, the extracellular domain of the murine receptor lacks several cysteines (Cys 102 , Cys 194 , Cys 212 , and Cys 687 ) present in the human receptor (Fig. 11). This leads us to believe that these four cysteines may be in the free state. This study revealed that all of these cysteine residues are indeed present in the free state and, additionally, that Cys 436 , Cys 447 , Cys 604 , Cys 613 , Cys 674 , and Cys 753 are also in the free state. Although 10 free cysteine residues are present in the human OB-R, most of them are largely unreactive to alkylating reagents, suggesting that they are buried within the protein molecule or otherwise protected from alkylation by O-or N-glycosylation. Possible functional roles of these free cysteines were not addressed in this study.
OB-R contains an unusual N-glycosylation motif, indicated as an NCS sequence, where the cysteine residue is involved in the disulfide formation. This sequence was found in several proteins including protein C and G-CSFR (18,20), and there are two sites in OB-R. Several other unusual N-glycosylation sites containing cysteine have been reported for proteins including CD69, immunoglobulin chain, and von Willebrand factor in which the atypical sequences NXC or NGGT are also found to be N-glycosylated (20 -23). However, these sequences contained a Ser or Thr residue at the second or fourth residue, for instance NSCX or NXCS, where X denotes variable residues.
Both the G-CSF receptor and gp-130 have significant sequence homology to the OB-R (6). The CRH domains of these and other receptors include a WSXWS motif, which is thought to play an important role in folding and is located between two ␤ sheet strands FЈ and GЈ (24,25). Furthermore, another sequence motif, SSFY, has been postulated to be important in the ligand-binding region of interleukin-6 receptor (26). A similar sequence was also found in OB-R, SSLY (residues 469 -472). Although OB-R contains two repeating Ig-CRH domains, the disulfide structures of these two domains were found to be slightly different from each other (Fig. 11) in that the FЈ-GЈ loop of the C-terminal CRH contained two free cysteines, suggesting a different conformation from the N-terminal CRH which did not have any such free cysteines. These findings suggest that domains CRH I and II may have different mechanisms for ligand interaction or perhaps other implications for biologic activity. To study further this structure-function relationship site-directed or deletion mutagenesis of the receptor would be useful.