Isolation of the SO 4 -4-GalNAc (cid:98) 1,4GlcNAc (cid:98) 1,2Man (cid:97) -specific Receptor from Rat Liver*

Glycoproteins, such as the glycoprotein hormone lutropin (LH), bear oligosaccharides terminating with the sequence SO 4 -4GalNAc (cid:98) 1,4GlcNAc (cid:98) 1,2Man (cid:97) (S4GGnM) and are rapidly removed from the circulation by a re- ceptor present in hepatic endothelial cells and Kupffer cells. Rapid removal from the circulation is essential for attaining maximal hormone activity in vivo . We have isolated a protein from rat liver which has the pro- perties expected for the S4GGnM-specific receptor (S4GGnM-R). The S4GGnM-R is closely related to the macrophage mannose receptor (Man-R) both antigeni-cally and structurally. At least 12 peptides prepared from the S4GGnM-R have amino acid sequences that are identical to those of the Man-R. Nonetheless, the ligand binding properties of the S4GGnM-R and the Man-R differ in a number of respects. The S4GGnM-R binds to immobilized LH but not to immobilized mannose, whereas the Man-R binds to immobilized mannose but not to immobilized LH. When analyzed using a binding assay that precipitates receptor ligand complexes with polyethylene glycol, the S4GGnM-R is able to bind S4GGnM-bovine serum albumin (S4GGnM-BSA) conju- gates whereas the Man-R is not. In contrast both the S4GGnM-R and the Man-R are able to bind Man-BSA. Monosaccharides that inhibit binding of Man-BSA by the Man-R enhance amount of radiolabeled receptor, sufficient antisera to precipitate 50–70% of the 125 I-S4GGnM-R added, and in- creasing amounts of unlabeled receptor that had been quantitated by amino acid analysis. The amount of receptor in “unknown” samples was then determined by comparison to the standard inhibition curve.

Asn-linked oligosaccharides present on the glycoprotein hormones lutropin (LH) 1 and thyrotropin (TSH) terminate with the sequence SO 4 -4GalNAc␤1,4GlcNAc␤1,2Man␣ (S4GGnM), whereas those on follitropin and chorionic gonadotropin (CG) terminate with the sequence Sia␣2,3/6Gal␤1,4GlcNAc␤1, 2Man␣ (1)(2)(3)(4)(5). We have proposed that the sulfated oligosaccharides present on LH and TSH are critical for the expression of full biologic function by these hormones (6 -9). Terminal Gal-NAc-4-SO 4 does not influence binding to or activation of the LH/CG receptor itself (10) but does have a marked impact on the circulatory half-life of LH following secretion (7,11) due to recognition of the sulfated oligosaccharides by a receptor expressed at the surface of hepatic endothelial cells and Kupffer cells (11,12). The rapid removal of LH from the circulation in conjunction with its release from granules in response to gonadotropin releasing hormone accounts for the episodic rise and fall in hormone levels seen in the circulation. Since the LH/CG receptor is a G-protein-coupled receptor, which rapidly becomes refractory to further stimulation following ligand binding (13)(14)(15), episodic stimulation may provide for maximal activation during the preovulatory surge in circulating LH levels. TSH shows similar properties with respect to half-life and receptor activation (16 -19).
Glycoproteins bound by the S4GGnM-specific receptor are subsequently transported to lysosomes and degraded. There are roughly 600,000 S4GGnM-specific binding sites at the cell surface of hepatic endothelial cells, which bind LH through its sulfated oligosaccharides with an apparent K d of 2.7 ϫ 10 Ϫ7 M. Binding is pH-dependent, requiring a pH Ͼ 5.0, but does not require Ca 2ϩ (12). The location of the sulfate in the 4-position is critical since glycoconjugates bearing oligosaccharides terminating with the sequence SO 4 -3GalNAc␤1,4GlcNAc␤1,2Man␣ (S3GGnM) are not bound by hepatic endothelial cells. We have now identified and isolated a glycoprotein from rat liver that has the properties expected for the receptor, which mediates removal of LH from the circulation on the basis of its sulfated oligosaccharides.

Analytical Procedures
Protein concentrations were determined using the Bio-Rad protein assay kit (Bio-Rad) or ISS Protein-Gold (Integrated Separation Systems). Polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate (SDS-PAGE) was performed according to Laemmli (20). Following separation by SDS-PAGE on 5% or 7.5% acrylamide gels, proteins were transferred electrophoretically to polyvinylidene difluoride membranes using CAPS buffer as described by Matsudaira (21) for amino-terminal sequence determination, peptide mapping, and detection with specific antisera. Proteins detected by antisera were developed using 125 I-F(abЈ) 2 goat anti-rabbit IgG. M NaCl, 2.0 mM CaCl 2 , pH 7.8, containing 0.2% NaN 3 .
bLH-Sepharose-bLH-Sepharose was prepared in the same fashion with the following modifications. bLH was dissolved in 100 mM NaHCO 3 , 200 mM NaCl, pH 8.4, at a concentration of 3.8 mg/ml and added to cyanogen bromide-activated Sepharose 4B at a ratio of 6.5 mg/ml Sepharose. The coupling efficiency was Ͼ95% after 2 h at room temperature. Remaining active sites were quenched by incubation at 4°C overnight in 1.0 M ethanolamine, pH 8.3. bLH-Sepharose was stored in 20 mM Tris HCl, 0.15 M NaCl, 2.0 mM CaCl 2 , pH 7.8, containing 0.2% NaN 3 .

Binding Assays
Binding assays (total volume of 150 l) contained 3-5 ng of S4GGnM receptor 125 I-S4GGnM-BSA (2-3 ϫ 10 5 dpm), and 90 g of hyaluronic acid and/or 90 g of fucoidin in 20 mM Tris-HCl, 0.15 M NaCl, 2 mM CaCl 2 , 1% (w/v) Triton X-100, pH 7.8. Hyaluronic acid is a weak inhibitor of S4GGnM-BSA binding, whereas fucoidin is a potent inhibitor (12). Incubations were performed in a 10 ϫ 75-mm glass tube at room temperature for 30 min. The reactions were terminated by adding 1.5 ml of ice-cold 10% (w/v) PEG 8000 (Sigma) in 20 mM Tris-HCl, 0.15 M NaCl, 2 mM CaCl 2 and mixing. After 30 min on ice, precipitated 125 I-S4GGnM-BSA⅐S4GGnM receptor complexes were collected by vacuum filtration on Whatman GF/C filter discs, which had been soaked in 20 mM Tris-HCl, 0.15 M NaCl, 2 mM CaC1 2 , 5 mg/ml bovine serum albumin. The filters were washed twice with 1.5 ml of ice-cold 20 mM Tris-HCl, 0.15 M NaCl, 2 mM CaCl 2 , 10% (w/v) PEG 8000, and the amount of 125 I determined by counting the filter in a ␥-counter. In the absence of added S4GGnM receptor, Ͻ5% of the added 125 I-S4GGnM-BSA was captured on the filter. One unit of activity is defined as the amount of S4GGnM receptor that is able to precipitate 1 ng of S4GGnM-BSA in the presence of hyaluronic acid above that precipitated in the presence of both hyaluronic acid and fucoidin.

S4GGnM Receptor Isolation
Step 1: Homogenate-Harlan Sprague Dawley rats, 150 -200 g each, were anesthetized and heparinized and their livers perfused with icecold 20 mM PO 4 , 0.15 M NaCl, pH 7.5, through the portal vein. Each liver was suspended in 20 ml of 0.25 mM EDTA, 0.02% NaN 3 (w/v) brought to pH 7.8 with solid NaHCO 3 and containing 50 units/ml aprotinin. The suspension was homogenized with three 20-s bursts of a Polytron homogenizer (Brinkman) at a setting of 5. Alternatively, 50 frozen rat livers weighing 300 g (Pel-Freez) were ground while frozen using a meat grinder and suspended in 400 ml of 0.25 mM EDTA, 0.02% NaN 3 (w/v) brought to pH 7.8 with solid NaHCO 3 and containing 50 units/ml aprotinin. The suspension was homogenized with four 15-s bursts of a Polytron homogenizer (Brinkman) at a setting of 8.
Step 2: Triton X-100 Solubilization-Sufficient 25% (w/v) Triton X-100 (Boehringer Mannheim) was added to bring the concentration of the homogenate to 10%, and the pH was adjusted to 7.7 with 1.0 M Tris-HCl, pH 7.8. After stirring for 1 h at 4°C, the Triton extract was passed through cheesecloth to remove connective tissue and sedimented at 7100 ϫ g for 20 min.
Step 3: PEG Pellet-Solid PEG 8000 (Sigma) was added to the supernatant to a final concentration of 10% (w/v). The extract was stirred for 15 min at 4°C and then allowed to stand for 30 min. Precipitated proteins were collected by sedimentation at 7100 ϫ g for 90 min. The supernatant was discarded.
Step 6: bLH-Sepharose-The WGA-Sepharose eluate was incubated with 0.5 ml of bLH-Sepharose/liver overnight at 4°C with rotation. Unbound material was removed by washing on a sintered glass funnel with 20 mM Tris-HCl, pH 7.8, 150 mM NaCl, 2 mM CaCl 2 , 1% Triton X-100. The bLH-Sepharose was then eluted with 50 mM sodium acetate, pH 4.0, 0.1% Triton X-100. The eluate was immediately adjusted to pH 7.0 by addition of 1.0 M Tris base and the volume reduced using a Centriprep-10 (Amicon).
Step 7: Mannan-Sepharose or Mannose-Sepharose-The bLH-Sepharose eluate was incubated with either mannan-Sepharose (0.05 ml/ liver) or mannose-Sepharose (0.05 ml/liver) and the unbound fraction taken. Bound proteins were eluted from the mannan-Sepharose and mannose-Sepharose by successive incubation with 50 mM galactose in 20 mM Tris, 150 mM NaCl, 0.05% Triton X-100, and 0.2 mM Pefabloc SC (Boehringer Mannheim) adjusted to pH 7.5 and 200 mM mannose and 5 mM EDTA in the same buffer.

Preparation of Rabbit Antisera to the S4GGnM Receptor
Polyclonal antisera to the S4GGnM receptor were raised in New Zealand White rabbits by immunization with the 180-kDa protein band isolated from an SDS-polyacrylamide gel. The protein band was excised, and, after extracting the gel pieces with 95% ethanol to reduce the SDS content, the gel was lyophilized. The dried gel was pulverized using a mortar/pestle and then emulsified with saline by passing through a series of successively smaller needles ranging from 18 to 23 gauge. The emulsified gel containing 5-10 g of protein was added to TDM-emulsion (RIBI) and injected intramuscularly into the hind leg and subcutaneously at four separate sites. The rabbit was boosted in the same manner 3 weeks later and sera obtained 8 -10 days later. The rabbit was subsequently boosted with antigen as required to maintain the titer of the antisera.

Radioimmunoassay for S4GGnM Receptor
The affinity-purified S4GGnM receptor was labeled with 125 I as described above. An antibody saturation curve was established using a constant amount of radiolabeled receptor (5-10 ng, 3 ϫ 10 5 cpm) and increasing amounts of antisera. Following an overnight incubation at 4°C, protein A-Sepharose antigen-antibody complexes were washed twice with 20 mM phosphate-buffered saline, pH 7.5, containing 0.1% BSA (w/v) and counted in a ␥ counter. Standard inhibition curves were constructed using a constant amount of radiolabeled receptor, sufficient antisera to precipitate 50 -70% of the 125 I-S4GGnM-R added, and increasing amounts of unlabeled receptor that had been quantitated by amino acid analysis. The amount of receptor in "unknown" samples was then determined by comparison to the standard inhibition curve.

RESULTS
Purification-We previously identified a receptor in rat liver that can account for the rapid removal of native LH bearing Asn-linked oligosaccharides terminating with the sequence S4GGnM from the circulation (12). The S4GGnM-R is located predominantly in hepatic endothelial cells and Kupffer cells and displays a high degree of specificity, recognizing S4GGnM-BSA but not S3GGnM-BSA. Fucoidin, a sulfated polysaccharide, inhibits binding of S4GGnM-BSA and LH by the receptor, whereas other sulfated and anionic polysaccharides do not inhibit binding or require much higher concentrations to do so. Glycoproteins bound to the S4GGnM-R are internalized, transported to lysosomes, and degraded. Binding is pH-dependent, requiring a pH above 5-6, and is not dependent on divalent cations even though divalent cations do enhance binding.
We established conditions that allowed us to detect and solubilize a binding activity with the properties we had described for the S4GGnM-R. Parenchymal cells (hepatocytes) and endothelial/Kupffer cells, prepared by collagenase perfusion as described previously (12,22), were disrupted by Dounce homogenization, and the nuclei and unbroken cells were collected by centrifugation. Soluble and total membrane fractions were obtained by sedimentation onto a 65% sucrose cushion. Fractions were brought to a final concentration of 1% (w/v) Triton X-100 and assayed for S4GGnM-BSA binding using the PEG precipitation assay. Binding activity was confined to the membrane fraction and the pellet containing nuclei and unbroken cells. None was found in the soluble fraction. As much as 80% of the binding activity was recovered in the membrane fraction after Dounce homogenization. The majority of the S4GGnM-BSA-specific binding activity, 64% of the total, was found in the endothelial and Kupffer cells, while 36% was in hepatocytes.
We examined the ability of Triton X-100 to solubilize the binding activity from membranes. Membranes were incubated with 125 I-S4GGnM-BSA in the presence of increasing amounts of Triton X-100 ( Fig. 1). Membranes were either collected by sedimentation in an Airfuge at 190,000 ϫ g (Beckman) in the absence of added PEG or by filtration on GF/C filters following addition of PEG. 125 I-S4GGnM-BSA was found in the membrane pellet (Fig. 1, ϪPEG) following addition of 0.05% Triton but not 0.01%, 0.5%, or 1.0% Triton. In contrast, 125 I-S4GGnM-BSA was bound in the presence of 0.05% Triton as well as 0.5% and 1.0% Triton when complexes were collected by precipitation with PEG (ϩPEG). At a Triton concentration of 0.05%, membrane vesicles become sufficiently permeable for the S4GGnM-R to be accessible to the 125 I-S4GGnM-BSA; however, the S4GGnM-R is not solubilized at this Triton concentration and can be sedimented in the absence of added PEG. At Triton concentrations above 0.5%, the S4GGnM-R is fully solubilized and requires PEG for precipitation of complexes. Using both the sedimentation assay in the presence of 0.05% Triton and the PEG precipitation assay in the presence of 1% Triton, we determined that: 1) 3-fold more 125 I-S4GGnM-BSA is bound at pH 7.5 than at pH 5.0 or below; 2) EDTA does not abolish binding of S4GGnM-BSA but does reduce it to 62% of that seen in the presence of 4 mM Ca 2ϩ ; and 3) fucoidin is a significantly more potent inhibitor of binding than other sulfated or anionic polysaccharides such as hyaluronic acid, heparin, chondroitin sulfate, and dextran sulfate. Thus, a binding activity with the properties expected for the S4GGnM-R could be detected in rat liver membranes and solubilized with Triton X-100. We therefore developed the isolation scheme summarized in Table I.
The S4GGnM-R was solubilized using 10% Triton X-100, concentrated by precipitation with PEG 8000, and solubilized in 10% Triton X-100 prior to incubation with WGA-Sepharose. The S4GGnM-R bound to WGA-Sepharose and was eluted with 300 mM GlcNAc. The WGA-Sepharose eluate containing the S4GGnM-R was incubated with bLH-Sepharose. S4GGnM-R that bound to bLH-Sepharose was eluted by reducing the pH to 4.0 with acetate buffer. When this material was examined by SDS-PAGE, a major band was found to be present, which had an M r of 180,000 ( Fig. 2A). Additional proteins with apparent molecular weights of 75-80,000, 60,000, 43,000, and 35,000 were also present. Following electrophoretic transfer to Immobilon-P, a ligand blot was performed using 125 I-S4GGnM-BSA, which demonstrated that only the protein with an M r of 180,000 was reactive (Fig. 2B). An NH 2 -terminal sequence of LK(Y)S(Q)YQFLIYNE was obtained for the protein with an M r of 180,000, suggesting it was closely related to the murine macrophage mannose receptor (Man-R), which has an NH 2terminal sequence of LLDARQFLIYNE (23).
S4GGnM-R that had been eluted from bLH-Sepharose was incubated with mannan-Sepharose or mannose-Sepharose. Neither the S4GGnM-BSA-specific binding activity nor the protein with an M r of 180,000 was bound by immobilized mannan or mannose, whereas the other proteins in the eluate from bLH-Sepharose were bound to either mannan-Sepharose or mannose-Sepharose and removed. The mannose-Sepharose unbound fraction was homogeneous and consisted of a single band, migrating with an M r of 180,000 when analyzed by SDS-PAGE (see Fig. 4), which was designated the S4GGnM-R.
The NH 2 -terminal sequence of the S4GGnM-R suggested a close relationship to the macrophage Man-R. However, the inability of the S4GGnM-R to bind to either immobilized mannan or mannose, which are used for affinity-based purification of the Man-R from lung (24), placenta (25), and macrophages (26,27), indicated the S4GGnM-R is distinct from the Man-R. We examined the potential relationship between the S4GGnM-R and the Man-R by isolating the macrophage Man-R from rat lung by affinity chromatography on Mannose-Sepharose as described by Lennartz et al. (24) for direct comparison with the S4GGnM-R isolated from rat liver. The Man-R isolated from lung by this procedure is homogeneous and has an Membranes prepared from rat liver were incubated with 125 I-S4GGnM-BSA in the presence of increasing amounts of Triton X-100 as indicated. 125 I-S4GGnM-BSA⅐S4GGnM-R complexes were collected either by sedimentation at 190,000 ϫ g for 5 min in an Airfuge in the absence of added PEG (ϪPEG) or by filtration of GF/C glass fiber filters following addition of PEG (ϩPEG). Precipitation with PEG was not performed for the 0.01% Triton concentration. Nonspecific binding was determined by performing incubations in the presence of 100 g/ml fucoidin and has been subtracted.

TABLE I
Isolation of the S4GGnM-specific receptor from rat liver Protein values were determined using the Bio-Rad protein assay kit or the ISS Protein-Gold kit. Binding units were determined using 125 I-S4GGnM-BSA and the PEG precipitation assay. The amount of receptor protein present was determined by a RIA in which unlabeled S4GGnM-R was used to inhibit binding of 125 I-S4GGnM-R by a rabbit antibody raised to the purified S4GGnM-R.
Step a ND, the amount of protein determined to be in the homogenate and the PEG pellet resuspended in Triton X-100 was not considered accurate and is not included.
M r of 180,000 when examined by SDS-PAGE (see Fig. 4). As will be presented in greater detail below, the S4GGnM-R and Man-R have a number of features that indicate they are distinct and other features that indicate they are closely related. For example the following features indicate that the receptors differ. 1) The S4GGnM-R will bind to immobilized ligands containing terminal GalNAc-4-SO 4 but not ligands with terminal Man, while the Man-R will bind to immobilized ligands containing terminal Man but not those containing terminal GalNAc-4-SO 4 . 2) The S4GGnM-R is able to bind soluble ligands terminating with S4GGnM as well as ligands with terminal Man or Fuc in the PEG precipitation assay, whereas the Man-R will bind ligands with terminal Man or Fuc but not those with terminal S4GGnM in the same assay. Features indicating the receptors are structurally related include the following observations. 1) The S4GGnM-R and the Man-R both react with 125 I-Man-BSA in ligand blots. 2) Antibodies raised to either the purified S4GGnM-R from liver or the Man-R from lung react with either receptor in Western blots. 3) The receptors have similar peptide maps, and multiple peptides prepared from the S4GGnM-R have sequences that are identical to those of the murine macrophage Man-R (23).
The S4GGnM-R and the Man-R Provide Comparable Peptide Maps-The S4GGnM-R (200 g) and the Man-R (150 g) were subjected to electrophoretic separation on 5% polyacrylamide gels and electrophoretically transferred to Immobilon (Millipore) in CAPS buffer. After staining with Ponceau Red, the regions containing the transferred protein were excised for analysis. Peptides were released by digestion with LysC or trypsin in the presence of reduced Triton X-100 and separated by reverse phase chromatography. The separations are shown in Fig. 3 for peptides released by LysC digestion of the S4GGnM-R and the Man-R. The profiles were nearly identical. Only peaks 57 and 75 of the S4GGnM-R were not also present in the Man-R. Peaks 45, 62, and 91 of the S4GGnM-R appeared to be identical to peaks 37, 54, and 84 of the Man-R, respectively. Peaks that appeared to be identical and peaks that appeared to differ between the S4GGnM-R and the Man-R were analyzed. The results of these analyses are summarized in Table II. Sequence was obtained for 12 peptides from three different S4GGnM-R preparations. Nine of the peptide sequences obtained from the S4GGnM-R were identical to peptide sequences predicted to be present in the macrophage Man-R (23). The sequences obtained originate from a number of different regions and encompass the entire 1365 amino acids of the Man-R extracellular domain (Table II). As expected from the similarity of the peptide maps, the predominant peptides present in peaks 45, 62, and 91 of the S4GGnM-R were identical to those of peaks 37, 54, and 84 of the Man-R, respectively. Only in the case of peaks 57 and 78 of the S4GGnM-R, which had no equivalent in the Man-R peptide profile, were sequences obtained that could not be identified in the Man-R sequence. These were minor components, however, and were identified within the asialoglycoprotein receptor (ASGP-R) sequence. Since small amounts of terminal ␤1,4-linked GalNAc are present on bLH and would be recognized by the ASGP-R, the presence of low levels of ASGP-R in the final preparation of the S4GGnM-R would not be unexpected. We did not, however, detect ASGP-R in the S4GGnM-R preparation by Western blot analysis using ASGP-R-specific antisera.
The peptide sequence analyses summarized in Table II, in conjunction with the nearly identical peptide maps obtained for the S4GGnM-R and the Man-R in Fig. 3, support the conclusion that the S4GGnM-R from liver is closely related to the macrophage Man-R isolated from rat lung. Furthermore, the S4GGnM-R and the Man-R both react with antisera raised to either receptor in Western blots (data not shown). Despite the strong evidence of a close structural relationship between the S4GGnM-R and the Man-R across their entire extracellular regions, we had evidence that indicated they would differ in specificity and ligand binding properties. We therefore examined their ligand binding properties in greater detail.

The S4GGnM-R and Man-R Differ in Their Ability to Bind to Immobilized GalNAc-4-SO 4 and Mannose-When purified
S4GGnM-R was incubated with Man-BSA conjugated to Sepharose (Fig. 4), Ͼ80% of the S4GGnM-R was present in the unbound fraction when analyzed by SDS-PAGE and staining with Coomassie Blue (panel A). Ligand blotting of the S4GGnM-R with 125 I-S4GGnM-BSA (panel B) and 125 I-Man-BSA (panel C) demonstrated that the S4GGnM-R in the unbound fraction could react with soluble S4GGnM-BSA and with soluble Man-BSA despite its inability to bind Man-BSA conjugated to Sepharose. The Man-R, which had been isolated from lung by affinity chromatography on mannose-Sepharose, did not bind to bLH-Sepharose, which contains multiple oligosaccharides terminating with S4GGnM (Fig. 5). Thus, the S4GGnM-R isolated from liver and the Man-R isolated from lung differ in their ability to bind to immobilized ligands containing terminal S4GGnM and Man. The S4GGnM-R and the Man-R represent the major forms of these receptors in liver and lung, respectively, since significant amounts of Man-R cannot be isolated from liver by affinity chromatography on Man-Sepharose following solubilization with Triton X-100 as described for isolation of the Man-R from lung (24), nor can significant amounts of S4GGnM-R be isolated from lung using the procedures we developed for isolation of the S4GGnM-R from liver. This conclusion is supported by analysis of the various steps during purification of the S4GGnM-R and the Man-R by Western blotting with receptor-specific antisera (data not shown).
Comparison of Man-BSA Binding by the S4GGnM-R and the Man-R-The ability of the S4GGnM-R to bind to immobilized bLH but not immobilized Man-BSA, and the ability of the Man-R to bind immobilized Man-BSA but not immobilized bLH, suggested that the S4GGnM-R and the Man-R would display different specificities for soluble ligands when examined in the precipitation assay. Notably, the S4GGnM-R binds both S4GGnM-BSA and Man-BSA in a concentration-dependent manner as determined by precipitation with PEG (Fig. 6A). In contrast, the Man-R binds Man-BSA but does not bind S4GGnM-BSA using the PEG precipitation assay (Fig. 6B). The inability of the Man-R to precipitate S4GGnM-BSA is consistent with its inability to bind to immobilized bLH, which bears multiple Asn-linked oligosaccharides terminating with the sequence S4GGnM (3). The ability of the S4GGnM-R to precipitate soluble Man-BSA was not expected since the S4GGnM-R is not able to bind to immobilized ligands containing terminal mannose such as mannan-Sepharose, Man-Sepharose, and Man-BSA-Sepharose.
The S4GGnM-R and the Man-R both react with Man-BSA

TABLE II
Comparison of peptides obtained from the S4GGnM-R and the Man-R Peptides were released from the S4GGnM-R or the Man-R by LysC digestion and fractionated by reverse phase chromatography as shown in Fig.  3. The peptides in the peaks indicated had their amino acid sequences determined and were found to be present in either the macrophage mannose receptor or the heptocyte asialoglycoprotein receptor (ASGPP-R). The sequences and their locations within these receptors are indicated. PK-41, PK-46, and PK-61 were obtained from an earlier analysis using a different preparation of the S4GGnM-R and release by digestion with trypsin. when examined by ligand blotting with 125 I-Man-BSA following SDS-PAGE (Fig. 7). When equal units of Man-BSA-specific binding activity, as measured by the PEG precipitation assay, were examined by ligand blotting with 125 I-Man-BSA, the S4GGnM-R in lane 3 was more intensely labeled than the Man-R in lane 4. This suggested that even though 5-6-fold more S4GGnM-R than Man-R was required to precipitate the same amount of soluble Man-BSA (compare panels A and B of Fig. 6), this difference in binding capacity was not retained in the ligand blot following SDS-PAGE. In support of this conclusion, we found that equal amounts of the S4GGnM-R and Man-R reacted with equal intensity when examined by ligand blotting with 125 I-Man-BSA following SDS-PAGE (Fig. 7, lanes  1 and 2). The difference in binding efficiency measured in the PEG precipitation assay for the S4GGnM-R and the Man-R is not retained in ligand blots following separation by SDS-PAGE. This suggests that binding in the soluble assay may reflect binding to a different site or in a different manner than when the same receptor is probed with ligand following SDS-PAGE. In light of the structural similarities between the S4GGnM-R and the Man-R, it is notable that they bind Man-BSA with equal intensity following SDS-PAGE even though they show marked differences in Man-BSA binding in their native state.
A remarkable feature of the macrophage Man-R is its ability to bind ligands with terminal Man, GlcNAc, Glc, and Fuc (28,29). These same monosaccharides can be utilized as inhibitors of binding by the Man-R. We therefore compared inhibition of Man-BSA binding by monosaccharides for both the S4GGnM-R and the Man-R. As reported by others (28,30,31), we found binding of Man-BSA by the Man-R in the PEG precipitation assay is inhibited by Man, Fuc, Glc, and GlcNAc, whereas Gal inhibits weakly and GalNAc not at all (Fig. 8). In contrast, low concentrations of Man enhance Man-BSA binding by the S4GGnM-R. Concentrations of Man as high as 200 mM are not inhibitory, although they do reduce binding as compared with Man concentrations ranging from 25-100 mM (Fig. 8). Fuc, Glc, and GlcNAc have similar effects whereas GalNAc is without effect (Fig. 8). Gal, which is a poor inhibitor of binding by the Man-R, enhances binding by the S4GGnM-R at a concentration of 200 mM but not at 50 mM. Thus, the same monosaccharides affect binding by the S4GGnM-R and the Man-R; however, they enhance binding by the S4GGnM-R and inhibit binding by the Man-R. Even though the S4GGnM-R and the Man-R both are able to bind Man-BSA, the properties of this binding reaction for the native receptors differ dramatically with respect to the effect of monosaccharides and the amount of receptor required to precipitate a given amount of Man-BSA.
Man-BSA and S4GGnM-BSA Bind to the S4GGnM-R Independently-We next determined if there was any relationship between the mannose-and S4GGnM-specific binding sites of the S4GGnM-R. Like the Man-R, we found that the S4GGnM-R is able to bind Fuc-BSA as well as Man-BSA (Fig. 9). Addition of excess Man-BSA inhibited binding of 125 I-Man-BSA and 125 I-Fuc-BSA by the S4GGnM-R. Man-BSA had no effect on the binding of 125 I-S4GGnM-BSA by the S4GGnM-R (Fig. 9). Fuc-BSA was also able to inhibit binding of both 125 I-Man-BSA and 125 I-Fuc-BSA by the S4GGnM-R, suggesting Man-BSA and Fuc-BSA compete for the same sites on the receptor. In contrast excess S4GGnM-BSA inhibits binding of 125 I-S4GGnM-BSA by the S4GGnM-R but not binding of either 125 I-Man-BSA or 125 I-Fuc-BSA (Fig. 9). The addition of either 40 mM mannose or fucose enhanced binding of Fuc-BSA and to an even greater extent than Man-BSA. At a concentration of 200 mM, binding of Man-BSA and Fuc-BSA were reduced as compared with that seen in the presence of 40 mM monosaccharide but not to the levels seen in the complete absence of added monosaccharides. Neither mannose nor fucose had any effect on S4GGnM-BSA  binding at either concentration (Fig. 9). Thus S4GGnM-BSA appears to bind to a site on the S4GGnM-R that is distinct from and independent of the Man/Fuc-specific binding site.
Kinetics of Man-BSA and S4GGnM-BSA Binding-The kinetics of binding of S4GGnM-BSA and Man-BSA in the presence and absence of GlcNAc were assessed for both the S4GGnM-R and the Man-R. The saturation curves obtained were analyzed using the Ligand program (32) as summarized in Table III. At saturation, 0.06 mol of S4GGnM-BSA/mol of S4GGnM-R was bound with an apparent K d of 3.0 ϫ 10 Ϫ9 M. The presence of 50 mM GlcNAc had no impact on the kinetics of S4GGnM-BSA binding. Man-BSA was bound by the S4GGnM-R with an apparent K d of 3.9 ϫ 10 Ϫ9 M and a mole ratio of 0.11 at saturation. In the presence of 50 mM GlcNAc, the apparent K d for binding of Man-BSA by the S4GGnM-R was reduced to 1.7 ϫ 10 Ϫ8 M while the mole ratio at saturation increased to 1.06. The apparent K d for binding of Man-BSA by the Man-R from lung was 3.1 ϫ 10 Ϫ9 M with a mole ratio of 0.77 at saturation. As with the S4GGnM-R, 50 mM GlcNAc reduced the apparent K d to 1.5 ϫ 10 Ϫ8 M and increased the mole ratio to 1.05 at saturation.
Thus, even though the S4GGnM-R is able to bind both S4GGnM-BSA and Man-BSA, the kinetics seen for binding of Man-BSA differ from those seen for binding of Man-BSA by the Man-R. Since both the apparent K d and the B max for binding of Man-BSA by the S4GGnM-R are influenced by addition of monosaccharides such as GlcNAc, the effects of monosaccharides on binding by the S4GGnM-R must be considered complex and will require more detailed analysis to be understood. It would appear, however, that the S4GGnM-R is capable of binding and internalizing ligands with terminal GlcNAc, Man, or Fuc as well as those with terminal S4GGnM. Thus, it is not clear at present if the receptor responsible for clearance of glycoproteins bearing oligomannose type oligosaccharides or neoglycoproteins such as Man-BSA from the blood (33) is the S4GGnM-R or the Man-R. DISCUSSION We have isolated a protein that has the properties predicted for the S4GGnM-R in hepatic endothelial cells and Kupffer cells. The S4GGnM-R mediates the rapid clearance of glycoproteins bearing oligosaccharides with the terminal sequence S4GGnM, for example LH and TSH, from the circulation (12). We have proposed that this function is critical for the expression of hormone biologic activity in vivo (6 -9, 34). Peptide maps, amino acid sequence of multiple peptides, and immune cross-reactivity indicate the S4GGnM-R is closely related to the previously characterized macrophage Man-R (23,35). The Man-R is one member of a family of structurally related membrane proteins, which includes DEC-205 (36), the phospholipase A 2 receptor (37,38), and a "novel" type C lectin expressed in fetal but not adult liver (39). Each consists of a signal sequence that is cleaved, a cysteine-rich (Cys-R) domain, a domain with fibronectin type II repeats (FN-II), 8 -10 type C carbohydrate recognition domains (CRD) separated by "stalks," a transmembrane domain, and a cytosolic domain. The S4GGnM-R peptides for which sequence was obtained (Table  II)

TABLE III
Binding studies of the S4GGnM-R and the Man-R Saturation curves were obtained for binding of 125 I-S4GGnM-BSA and 125 I-Man-BSA using the PEG precipitation assay. Saturation curves for Man-BSA were performed in the absence and presence of 50 mM GlcNAc. The data was analyzed using the program LIGAND (32). The B max is expressed as a mole ratio based on a molecular weight of 180,000 for the S4GGnM-R and Man-R, and 70,000 for the BSA conjugates. S4GGnM-BSA has an average of 13  How closely related are the S4GGnM-R and the Man-R and what is the structural basis for the differences in their ligand specificities? The data we have obtained indicates that the structural relationship between the S4GGnM-R and the Man-R is extensive. Possible mechanisms that could result in two such closely related receptors include: 1) the existence of two genes encoding closely related proteins, 2) a post-transcriptional alteration in the mRNA sequence producing different forms of the receptor, and 3) a post-translational modification which alters the specificity of the receptor. Any differences in the structure of the S4GGnM-R and the Man-R must account for differences in the ability to bind ligands terminating with S4GGnM and for differences in the characteristics for binding of ligands containing terminal mannose or fucose.
The macrophage Man-R displays multiple specificities, being able to bind carbohydrate moieties terminating with Man, GlcNAc, or Fuc (28 -30). CRDs 4 -8 together account for the high affinity binding of ligands such as Man-BSA and mannan. CRD 4 plays a predominant role in binding and is the only CRD that is able to bind mannose containing ligands in the absence of other CRDs (28). Binding of Man-BSA and Fuc-BSA by the S4GGnM-R, like binding by the Man-R, is Ca 2ϩ -and pH-dependent. Furthermore, the same monosaccharides have an impact on binding by both receptors; however, in the case of the Man-R these monosaccharides are strictly inhibitory, whereas for the S4GGnM-R they have a complex effect resulting in enhanced rather than reduced binding. This suggests that one or more of the CRDs that mediate Man-BSA binding by the Man-R are altered in the S4GGnM-R. Based on peptide sequences from the S4GGnM-R the Cys-R region, FN-II region, CRD 2, and CRD 3 are present and likely have the same sequence as in the Man-R. The functional significance of the Cys-R region and FN-II region is not known for the Man-R or other members of this family, nor is there evidence of ligand binding by CRDs 1-3 of the Man-R (28). Should the S4GGnM-R prove to have the same overall structure as the Man-R throughout its extracellular domain, there would be an ample number of regions, which could potentially account for the independent binding of ligands terminating with S4GGnM and those terminating with either mannose or fucose.
The S4GGnM-R may be the first example of a carbohydratespecific receptor that can bind unrelated oligosaccharide structures at independent sites. Since binding of S4GGnM-BSA does not require Ca 2ϩ , it is likely that the structural motif which accounts for binding of GalNAc-4-SO 4 -containing ligands differs from the Ca 2ϩ -dependent CRDs, which are characteristic of the Man binding sites. If distinct regions account for S4GGnM-and Man-specific binding by the S4GGnM-R, it would seem likely that structural differences between the S4GGnM-R and the Man-R would have to involve more that one region. An intriguing feature of the S4GGnM-R is that the same peptide bearing different carbohydrate moieties could be bound by the same receptor at different sites and with differing kinetics. This could result in differing kinetics of clearance from the circulation. For example, LH bearing high mannose type structures may be cleared more rapidly than LH bearing oligosaccharides terminating with S4GGnM. Should this be the case it would suggest that the precise rate of clearance as determined by the structure of the oligosaccharide is indeed critical for maintaining biologic activity in vivo.
A number of studies have suggested that a receptor with the same properties as the Man-R isolated from lung and placenta is present in liver endothelial cells and Kupffer cells (30, 40 -42). The relationship of the hepatic receptor to the macrophage Man-R present in alveolar macrophages and other tissues may have to be re-evaluated in light of the current findings. It is not surprising that it has been difficult to purify the Man-R from liver using affinity chromatography on immobilized ligands containing terminal mannose, procedures that are effective for isolation of the Man-R from lung (24), macrophage lines (26), and placenta (25), in light of the properties of the S4GGnM-R. It now seems likely that the S4GGnM-R we have isolated from liver accounts for a major fraction of binding and internalization of ligands containing terminal Man, Fuc, or GlcNAc by hepatic endothelial cells and Kupffer cells (41)(42)(43). The characteristics of binding and internalization of Man-BSA by the S4GGnM-R are likely to differ from those encountered with the Man-R of alveolar macrophages.
Many issues remain to be addressed. How does the S4GGnM-R differ from the Man-R structurally? What region accounts for binding of S4GGnM? Is the region accounting for S4GGnM binding structurally related to the Ca 2ϩ -dependent C type lectin motif, or does it represent a new binding motif? Is expression of the S4GGnM-R, like that of the Man-R, highly regulated. Is expression regulated by estrogen? Is the S4GGnM-R expressed in other cells and in other tissues? These are some of the issues we will address in our future studies. The answers promise to reveal new insights about the biologic significance of oligosaccharides terminating with S4GGnM for the glycoprotein hormones as well as other glycoproteins, which are continually being added to the list family of glycoproteins bearing S4GGnM structures.