Rapid Clearance of Sialylated Glycoproteins by the Asialoglycoprotein Receptor*

The asialoglycoprotein-receptor (ASGP-R) located on liver parenchymal cells was originally identified and characterized on the basis of its ability to bind glycoproteins bearing terminal galactose (Gal) or N-acetylgalactosamine (GalNAc); however, endogenous ligands for the ASGP-R have not to date been definitively identified. We have determined that the rat ASGP-R specifically binds oligosaccharides terminating with the sequence Siaα2,6GalNAcβ1,4GlcNAcβ1,2Man. Bovine serum albumin chemically modified with 10–15 tetrasaccharides with the sequence Siaα2,6GalNAcβ1,4GlcNAcβ1,2Man is cleared from the blood of the rat with a half-life of <1 min by a receptor located in the liver. We have isolated the receptor and identified it as the ASGP-R. Furthermore, we have determined that subunit 1 of the ASGP-R accounts for the binding of terminal Siaα2,6GalNAcβ. Based on the newly defined specificity of the rat ASGP-R we hypothesize that glycoproteins bearing structures that are selectively modified with terminal Siaα2,6GalNAcβ and are released into the blood may be endogenous ligands for the rat ASGP-R.

The rapid clearance of glycoproteins from the blood following removal of sialic acid (Sia) 1 residues and exposure of underlying galactose (Gal) residues was first reported by Ashwell and Morell in the early 1970s (1,2) and led to the discovery of the asialoglycoprotein-receptor (ASGP-R). The specificity and biochemical features of this endocytic receptor have been extensively investigated since that time (3,4). Although it is clear that the ASGP-R receptor binds oligosaccharides with terminal ␤-linked N-acetylgalactosamine (GalNAc) or Gal and can mediate the rapid clearance of glycoproteins bearing these terminal sugars from the circulation, endogenous ligands for this abundant receptor have not yet been identified.
We first described N-linked oligosaccharides containing ␤1,4-linked GalNAc on lutropin (LH) and other members of the glycoprotein hormone family of glycoproteins (5,6). The Gal-NAc is found almost exclusively in the form of GalNAc-4-SO 4 , reflecting the sequential action of a protein-specific ␤1,4-Nacetylgalactosaminyltransferse (␤1,4GalNAcT) and a GalNAc-4-sulfotransferase (GalNAc-4-ST1) (7,8). The terminal Gal-NAc-4-SO 4 is recognized by a receptor in hepatic endothelial cells that regulates the circulatory half-life of LH following its stimulated release into the blood (9 -12). The control of circulatory half-life is important for regulating estrogen production in vivo during implantation of the embryo (13). We subsequently described the presence of N-linked oligosaccharides terminating with Sia␣2,6GalNAc␤ on prolactin/growth (PLP) hormone family members that are synthesized by rat placenta spongiotrophoblasts between mid gestation and birth (14). The levels of protein-specific ␤1,4GalNAcT activity in rat placenta increase 150-fold between day 9 and 18 of gestation, whereas levels of ␣2,6sialyltransferase increase 5-fold during the same period (14). We have found significant levels of PLP hormone family members bearing terminal Sia␣2,6GalNAc␤ in the circulation of the pregnant rat late in gestation. Furthermore, the pregnancy-specific glycoprotein glycodelin (placental protein 14) that is found in the amniotic fluid of humans also bears this structure (15).
The selective addition of ␤1,4-linked GalNAc to the oligosaccharides on prolactin-like protein (PLP)-A, PLP-A, PLP-B, PLP-C, PRP, and placental lactogen I variant (14) as well as human glycodelin (15) raised the possibility that a receptor specific for Sia␣2,6GalNAc␤ is present and may regulate the circulatory half-life of glycoproteins bearing these structures and/or direct them to specific locations such as the amniotic fluid. We have examined this possibility in the pregnant rat and report that neoglycoconjugates bearing multiple tetrasaccharides terminating with the sequence Sia␣2,6GalNAc␤1, 4GlcNAc␤1,2Man are rapidly removed from the circulation by the ASGP-R. Thus, the PLP hormone family members that are released into the blood during pregnancy may represent examples of endogenous ligands for the ASGP-R.
Clearance and Tissue Distribution Studies-Clearance studies were performed as previously described (11). Rats were anesthetized by intraperitoneal injection of ketamine (87 mg/kg) and xylazine (13 mg/ kg). The left jugular vein was cannulated, and the rats were heparinized prior to injection of 4 ϫ 10 6 cpm of SiaGGnM-[ 125 I]BSA in 100 l of saline. Blood was withdrawn at designated times, and the amount of SiaGGnM-[ 125 I]BSA or S3GGnM-[ 125 I]BSA present per milligram of blood was determined by ␥-counting. Rats were euthanized at the termination of the clearance study by metofane inhalation. Organs were removed and weighed, and the amount of SiaGGnM-[ 125 I]BSA present was determined by ␥-counting.
Preparation of Rat Liver Membranes-Rat livers were homogenized with a Polytron homogenizer (Brinkmann Instruments, Westbury, NY) in 5 volumes of buffer containing 25 mM HEPES, 50 mM KCl, 2 mM magnesium acetate, 1 mM dithiothreitol, and 10% (w/v) sucrose. Homogenates were sedimented at 1,500 ϫ g for 5 min, and the resulting supernatants were layered over a 65% (w/v) sucrose cushion prior to sedimentation at 100,000 ϫ g at 4°C for 75 min. The membrane fraction at the interphase and the soluble fraction in the upper phase were collected separately and stored at Ϫ80°C. Protein concentrations were determined using the Bradford method (Bio-Rad, Richmond, CA).
Binding Assays-Binding studies were performed in 100 l of Buffer A containing 0.5% Triton X-100, 2 mM CaCl 2 , 2-5 ϫ 10 4 cpm of SiaG-GnM-[ 125 I]BSA, and 200 g of membrane or soluble liver protein. Following incubation for 30 min at 25°C, 1.5 ml of ice-cold 10% (w/v) PEG 8000 in Buffer B (25 mM Tris-HCl, 150 mM NaCl, 2 mM CaCl 2 , pH 7.4) was added, and the reaction was vortexed for 2 s. Precipitated ligandreceptor complexes were collected after incubation for 30 min at 4°C by vacuum filtration on Whatman GF/C filter discs that had been incubated in Buffer B containing 5 mg/ml BSA. The filtrates were washed three times with 1 ml of ice-cold 10% (W/V) PEG 8000 in Buffer B and counted in a ␥-counter. Assays were also performed using a 96-well MultiScreen GF/C filter plate (Millipore, Bedford, MA); however, the binding reaction was reduced to75 l, and the ligand-receptor complexes were precipitated by adding 75 l of 20% (w/v) PEG 8000 in Buffer B prior to filtration. Individual filters were removed using a MultiScreen punch tip (Millipore) and counted in the ␥-counter.
Receptor Purification-Rat liver membrane proteins (400 mg) prepared as described above were resuspended in 50 ml of 50 mM Tris-HCl at pH 7.4 with 150 mM NaCl, 2 mM CaCl 2 , and 3% Triton X-100 using a Dounce homogenizer. Insoluble material was removed by sedimentation for 2 min at 700 ϫ g. The supernatant was incubated with 1 ml of GalNAc-␤-BSA-Sepharose (5 mg of GalNAc-␤-BSA/ml of Sepharose) for 16 h at 4°C. After washing five times with 5 ml of Buffer B containing 1% Triton X-100, the bound protein was eluted with Buffer A containing 100 mM GalNAc. Fractions containing SiaGGnM-BSA binding activity were pooled, dialyzed against Buffer B, concentrated, and stored at Ϫ20°C.
Western and Ligand Blots-Proteins were separated in the presence of SDS without reduction on NuPAGE Bis-Tris gels and electrophoretically transferred to Immobilon-P (Millipore). Duplicate gels were stained with Coomassie Blue or silver. For two-dimensional electrophoresis, isoelectric focusing was performed with PROTEAN IEF system (Bio-Rad) using pH 3-10 ReadyStrip IPG strips. Separation in the second dimension was performed without reduction on 10% Criterion TM Tris-HCl gels as recommended by the manufacturer. Rabbit anti-rat ASGP-R (1:10,000) and mouse anti-rabbit IgG:horseradish peroxidase (1:100,000) were used for Western analysis. For ligand blots, the Immobilon-P membranes were blocked with 5% milk protein for 30 min, washed once with Buffer A containing 25 mM EDTA, and washed three times with Buffer A containing 0.5% Triton X-100 and 2 mM CaCl 2 . The washed Immobilon-P membrane was incubated overnight at 4°C in the presence of 2 ϫ 10 5 cpm of SiaGGnM-[ 125 I]BSA in Buffer A containing 0.5% Triton X-100 and 2 mM CaCl 2 , washed three times with Buffer A containing 0.5% Triton X-100 and 2 mM CaCl 2 , and exposed to film for 7-15 days. Proteins corresponding to the region reactive with SiaGGnM-[ 125 I]BSA were excised and examined by MALDI-TOF at the Johns Hopkins University Mass Spectrometry Facility.
Cloning and Expression of Rat ASGP-R Subunit 1 and Subunit 2-Total RNA was isolated from rat liver with TRIzol reagent per the manufacturer's instructions. First-strand cDNA was synthesized using oligo(dT 12 ) and Superscript TM II RNase H Ϫ reverse transcriptase (Life Technologies). Rat ASGP-R subunits 1 and 2 were amplified using KlenTaq LA DNA polymerase mix (Sigma Chemical Co., St. Louis, MO) and the gene specific primers RHL1-F (CGG GAT CCC ATC ATG ACA AAG GAT TAT CAA GAT TTC C) and RHL1-⌬R (5Ј-CCC ATT GGC CTT GCC CAA CTC TG), RHL2-F (CGG GAT CCC ATC ATG GAG AAG GAC TTT CAA GAT ATC C), and RHL2-⌬R (5Ј-CCC GTA GGT GAT GTC CCG TTT G), respectively. Amplified products were subcloned into pcDNA3.1/V5His-TOPO and sequenced. The cDNAs were designated as RHL1V5His and RHL2V5His, respectively.
CHO/Tag and 293/Tag cells grown on a 100-mm plate were transfected with 13 and 6 g, respectively, of RHL1V5His or RHL2V5His using 35 g of LipofectAMINE (Life Technologies) in serum-free medium for 5 h according to the manufacturer's protocol. The pcDNA3.1/V5His-TOPO cloning vector was used as a control for negative expression. Forty-eight hours after transfection, the cells were washed with PBS (8 g of NaCl, 0.2 g of KCl, 1.44 g of Na 2 HPO 4 , and 0.24 g of KH 2 PO 4 , pH 7.4) and incubated for 10 min at Ϫ20°C. Cells were thawed and scraped off in 1 ml of Buffer A containing 50 mM EDTA. Cells were pelleted by centrifugation for 2 min at 12,000 ϫ g and washed three times with 1 ml of Buffer A. Cells were incubated with 250 l of Buffer A containing 1% (w/v) Triton X-100 per 100-mm diameter culture plate for 30 min at 4°C with mixing. The mixture was centrifuged at 12,000 ϫ g for 10 min at 4°C. The supernatant was collected and store at Ϫ20°C until needed.

RESULTS
Sia␣2,6GalNAc␤1,4GlcNAc␤1,2Man␣-BSA Is Rapidly Removed from the Circulation by the Liver-We had previously observed that PLP hormones bearing oligosaccharides terminating with the unique sequence Sia␣2,6GalNAc␤ are present in the blood of pregnant rats between mid gestation and birth. We introduced SiaGGnM-[ 125 I]BSA into the blood of pregnant rats between days 16 and 18 of gestation to determine if a receptor that recognizes this carbohydrate structure is expressed. We found that the injected SiaGGnM-[ 125 I]BSA was cleared from the circulation in less than 1 min (Fig. 1), suggesting the presence of a receptor specific for terminal Sia␣2,6GalNAc␤. In contrast, S3GGnM-[ 125 I] BSA, which bears the identical trisaccharide with a SO 4 linked to the C-3 hydroxyl of the GalNAc, was cleared from the blood at a much slower rate.
The rats were autopsied to determine the tissue distribution of the SiaGGnM-[ 125 I]BSA. Essentially all of the radiolabel (Ͼ90%) was present in the liver when rats were sacrificed 10 min after injection of the SiaGGnM-[ 125 I]BSA (Table I). The cpm/mg of tissue in the liver, kidneys, and lungs was 47-, 6-, and 5-fold greater (not shown), respectively, than that in the large intestine, indicating that the SiaGGnM-[ 125 I]BSA was concentrated in the liver as compared with other highly vascular tissues. Although the amount of label in the liver had declined by 45 min as compared with 10 min following injection, the amount of label in the kidneys had increased. This suggested that SiaGGnM-[ 125 I]BSA was taken up exclusively by receptors in the liver and subsequently degraded, resulting in the release of radiolabel into the blood and its appearance in the kidneys and urine. There was no evidence of binding to or uptake by the uterus, placenta, or pups.
Cation-dependent Binding of SiaGGnM-BSA by a Receptor in Rat Liver-The rapid and efficient clearance of SiaGGnM-BSA by the liver indicated that a receptor specific for terminal Sia␣2,6GalNAc␤ was present in either hepatocytes or hepatic endothelial cells. The membrane fraction from liver, but not the soluble fraction, displayed cation-dependent binding of SiaG-GnM-[ 125 I]BSA (Fig. 2). Binding was pH-dependent, reaching a maximum between pH 5.5 and 7.0, and declining markedly below pH 5.0 (Fig. 3A). Binding did not occur in the absence of cations (Fig. 3B). Whereas Cd 2ϩ or Co 2ϩ enhanced SiaGGnM-[ 125 I]BSA binding as compared with Ca 2ϩ , other cations with the exception of Ni 2ϩ and Mn 2ϩ were significantly less effective.
A range of monosaccharides were tested as inhibitors of SiaGGnM-[ 125 I]BSA binding (Fig. 3C). At a concentration of 20 mM, GalNAc was the most potent inhibitor of binding (3% of control). Lactose (Gal␤1,4Glc) and Gal both were good inhibitors, although not as potent as GalNAc, indicating the binding site can accommodate either Gal or GalNAc. Free sialic acid was also an inhibitor (36% of control) suggesting that the binding site accommodated both the terminal sialic acid and the ␤-linked GalNAc. The lack of inhibition by Man, GlcNAc, and ␣-methyl-Man indicated the receptor was not likely the Man/GalNAc-4-SO 4 -receptor found in hepatic endothelial cells.
Isolation of a Sia␣2,6GalNAc␤-specific Receptor-The marked inhibition of binding seen with GalNAc (Fig. 3C) suggested that an affinity matrix containing ␤-linked GalNAc could be utilized for purification of the Sia␣2,6GalNAc␤-spe-cific receptor. Rat liver membrane proteins were solubilized using Triton X-100 and incubated with GalNAc␤-BSA-Sepharose. The SiaGGnM-[ 125 I]BSA specific binding activity was retained by the affinity matrix and was eluted with 100 mM GalNAc. When examined by SDS-PAGE on a 10% gel under non-reducing conditions, the GalNAc-eluted material contained two major protein species with mobilities equivalent to 33 and 25 kDa when stained with Coomassie Blue (Fig. 4, lane  3). Because we expected the ASGP-R to be present in the material that was bound by GalNAc␤-BSA-Sepharose and eluted with 100 mM GalNAc, Western blot analysis with rabbit anti-rat ASGP-R antibody was used to identify the ASGP-Rderived proteins. Proteins with mobilities equivalent to 33, 70, 130, and Ͼ170 kDa reacted with anti-ASGP-R antibody in both the unfractionated Triton X-100 solubilized membrane fraction (Fig. 4, lane 4) and the affinity purified fraction (Fig. 4, lane 5). The same protein species bound SiaGGnM-[ 125 I]BSA when a ligand blot was performed on the affinity-purified fraction (Fig.  4, lane 6). We also examined the proteins in the affinity-purified fraction by two-dimensional electrophoresis (Fig. 5). A ligand blot using SiaGGnM-[ 125 I]BSA (Fig. 5B) indicated that proteins migrating with pI values between 4.6 and 5.6 and molecular masses of ϳ33, 75, and 150 kDa (Fig. 5A) accounted for the Sia␣2,6GalNAc␤ specific binding activity. The material migrating at 33 kDa was visualized by silver staining (Fig. 5A) and identified as the ASGP-R by MALDI-TOF analysis of tryptic fragments. Six out of a total of 31 potential tryptic peptides for subunit 1 of the ASGP-R were identified, accounting for 17% the subunit 1 sequence. The ASGP-R represents the highest non-trivial protein detected in the sample with a molecular weight search (MOWSE) score of 197 for subunit 1. The only other non-trivial protein detected was keratin. Because the SiaGGnM-[ 125 I]BSA used for the binding assays and the ligand blots does not contain any terminal GalNAc, the results indicated that the rat ASGP-R is capable of binding saccharides with terminal Sia␣2,6GalNAc␤.
Subunit 1 Accounts for Sia␣2,6GalNAc␤-specific Binding by the Rat ASGP-R-cDNAs encoding subunit 1 (RHL1) and subunit 2 (RHL2) of the rat ASGP-R were amplified from rat liver mRNA. The stop codons were mutated to permit expression as chimeric proteins containing the V5 epitope followed by six His residues at the carboxyl terminus. CHO-Tag and 293-Tag cells  Table I.  were transfected with pcDNA3.1-RHL1V5His or pcDNA3.1-RHL2V5His. Equal fractions of the transfected cells were examined by Western blot analysis with anti-V5 antibody following SDS-PAGE on a 10% gel under reducing conditions. RHL1V5His is expressed at higher levels than RHL2V5His in both cell lines. In addition RHL1V5His is more heterogeneous with major bands migrating with molecular masses of 52 and 57 kDa, whereas RHL2V5His has a single band migrating with a molecular mass of 57 kDa (Fig. 6) . This is consistent with the observations of other groups that lectin activity is primarily associated with RHL1 (17)(18)(19). Nonetheless, it is clear that recombinant RHL1V5His is able to bind both SiaGGnM-BSA and Gal-BSA as was seen with the affinity-purified ASGP-R prepared from rat liver above.

DISCUSSION
The ASGP-R was originally identified on the basis of its ability to mediate the rapid clearance of glycoproteins such as ceruloplasmin from the blood following removal of sialic acid and exposure of underlying Gal moieties. Numerous studies have supported this conclusion, and sialylated glycoproteins are typically used to demonstrate the specificity of the ASGP-R for terminal Gal and GalNAc moieties. Our observation that the rat ASGP-R is able to bind saccharides with the terminal sequence Sia␣2,6GalNAc␤ is both unexpected and remarkable. The binding activity we have observed does not reflect the presence of terminal GalNAc on the SiaGGnM-BSA glycoconjugate, because less than 1% of the SiaGGnM-[ 125 I]BSA preparation was bound by Wistaria fluoribunda agglutinin-agarose (not shown), a lectin that would detect the presence of saccharides with a single terminal ␤-linked GalNAc (20). Furthermore, the same amount of SiaGGnM-[ 125 I]BSA was precipitated by the ASGP-R before and after neuraminidase digestion to remove the sialic acid (not shown), indicating that removal of the sialic acid did not result in a significant increase in binding by the ASGP-R.
The terminal sequence Sia␣2,6GalNAc␤1,4GlcNAc␤1,2Man that we described on PLP hormones produced by rat spongiotrophoblasts from mid gestation to birth is a unique structure that is present on only a limited number of glycoproteins (14,15,21). It is not yet clear if the terminal sequence Sia␣2, 6Gal␤1,4GlcNAc␤1,2Man that is present on large numbers of glycoproteins will also be recognized by the ASGP-R. Our observation that SiaGGnM-[ 125 I]BSA binding by the ASGP-R is partially inhibited by 20 mM sialic acid suggests that the sialic acid may actively participate in binding to the RHL1 subunit. The crystal structure of the Gal/GalNAc binding site of subunit 1 of the human ASGP-R has been resolved (22). It will be of interest to determine how Sia␣2,6GalNAc␤1,4GlcNAc␤ is bound as compared with either GalNAc␤1,4GlcNAc␤ or Gal␤1, 4GlcNAc␤. Because the pH and calcium dependence for binding are similar for binding Sia␣2,6GalNAc␤1,4GlcNAc␤ and Gal/ GalNAc terminal ligands, it is likely that GalNAc coordinates with the same calcium (site 2) and other contact residues in the presence or absence of the ␣2,6-linked sialic acid.
The rapid clearance of the neoglycoconjugate SiaGGnM-BSA by the ASGP-R was unanticipated, because the ASGP-R has long been assumed to exist for the clearance of ligands whose terminal sialic acid has been removed to expose underlying Gal or GalNAc moieties. Because SiaGGnM-BSA is multivalent, we have not yet been able to establish the affinity of RHL1 for Sia␣2,6GalNAc␤1,4GlcNAc␤ as compared with GalNAc␤1, 4GlcNAc␤ or Gal␤1,4GlcNAc␤. The K d values for glycoproteins bearing 1, 2, 3, or more terminal Sia␣2,6GalNAc␤1,4GlcNAc␤ sequences will have to be determined to establish how rapidly such glycoproteins would be cleared from the circulation. The results we have obtained with the Man/GalNAc-4-SO 4 -receptor are, however, instructive.
The Man/GalNAc-4-SO 4 -receptor is found predominantly in the form of a dimer at the surface of hepatic endothelial cells (9). Each cysteine-rich domain at the amino terminus of the receptor is able to engage a single terminal GalNAc-4-SO 4 (23), and two terminal GalNAc-4-SO 4 moieties on separate N-linked oligosaccharides must be engaged simultaneously to form a stable complex with the affinity seen for binding bovine LH to isolated endothelial cells through its sulfated carbohydrate chains. The neoglycoconjugate SO 4 -4-GalNAc␤1,4GlcNAc␤1, 2Man␣-BSA, which like SiaGGnM-BSA has a high density of conjugated saccharides, is cleared from the circulation of the rat in less than 2-3 min, whereas LH with three N-linked oligosaccharides bearing terminal GalNAc-4-SO 4 moieties is cleared with a half-life of 7.5 min (11,12). The clearance rate is consistent with the K d of 1.6 ϫ 10 Ϫ7 M for binding to the Man/GalNAc-4-SO 4 -receptor; however, LH in the circulation never exceeds a concentration of 1 ϫ 10 Ϫ9 M (12).
The clearance rate seen for LH in vivo reflects the enormous amount of Man/GalNAc-4-SO 4 -receptor that is expressed in the liver. With 4 ϫ 10 8 endothelial cells in the liver, Ͼ600,000 receptors at the surface of each cell, and a rapid rate of endocytic uptake (t1 ⁄2 of 20 s) there is sufficient capacity to clear all of the LH from the blood even if only 0.5% of the receptor is occupied. Furthermore, the rate of clearance will remain constant at all concentrations of LH below its K d of 1.6 ϫ 10 Ϫ7 M for Man/GalNAc-4-SO 4 -receptor.
Based on the example of LH and the Man/GalNAc-4-SO 4receptor, our studies suggest that a different group of circulating glycoproteins may be the endogenous ligands for the ASGP-R in vivo. Like the Man/GalNAc-4-SO 4 -receptor, ASGP-R is highly expressed at the cell surface (500,000 receptors/cell) (24,25) of an abundant cell and the hepatocyte (6 -7 ϫ 10 8 cells per rat liver) (26) and is rapidly endocytosed (27)(28)(29). Thus, if glycoproteins with multiple terminal Sia␣2,6GalNAc␤ moieties can be engaged with affinities in the range of 1 ϫ 10 Ϫ7 M or below, they would likely be removed from the blood with half- FIG. 5. Analysis of affinity-purified SiaGGnM-BSA specific binding activity by two-dimensional gel electrophoresis. The proteins eluted from the GalNAc-␤-BSA-Sepharose column with 100 mM GalNAc were separated by two-dimensional gel electrophoresis. A, silver stain of 3.5 g of eluate that was separated in the first dimension using ReadyStrip IPG, pH 3-10 (Bio-Rad Laboratories, Hercules, CA) and in the second dimension using a non-reducing Criterion 10% Tris-HCl gel (Bio-Rad Laboratories). B, a duplicate two-dimensional gel that was electrophoretically transferred to a polyvinylidene difluoride membrane and probed with SiaGGnM-[ 125 I]BSA (2 ϫ 10 3 cpm/ml). The material in A indicated by the white box corresponds to the ϳ33-kDa region that bound SiaGGnM-BSA in B and was excised for analysis by MALDI-TOF following trypsinization. lives of 5-20 min that would not change with the changing concentration of each glycoprotein below its K d . Because ␤1,4linked Gal and GalNAc are both recognized by the ASGP-R, it is possible that Sia␣2,6Gal␤ may also be recognized. This raises the possibility that the ASGP-R may function to actually clear glycoproteins with terminal Sia␣2,6GalNAc␤ or Sia␣2,6Gal␤ moieties in vivo rather than ones with terminal Gal or GalNAc.
Our current studies have revealed that glycoproteins bearing terminal Sia␣2,6GalNAc␤ are recognized by the rat ASGP-R. Thus, the PLP hormones that bear these structures may represent examples of endogenous ligands for the ASGP-R. Notably, in mice the levels of ASGP-R have been reported to increase between mid gestation and birth and return to the levels seen in non-pregnant animals within 24 h after delivery (30). This pattern of expression further supports a potential relationship between the addition of these carbohydrate structures to PLP hormones and the regulation of their circulatory halflife by the ASGP-R.
The specificity of the ASGP-R may also provide insight into other observations. For example, terminal Sia␣2,6GalNAc␤ is found on glycodelin, a human pregnancy-specific glycoprotein with potent immunosuppressive and contraceptive activities that is found in the amniotic fluid (15,31). The immunosuppressive effects seen with glycodelin have been attributed to the Sia␣2,6GalNAc␤-bearing oligosaccharides that may specifically block adhesive and activation-related events mediated by CD22, the human B cell-associated receptor. Because the ASGP-R is present on late stage spermatids (32), the contraceptive activity associated with glycodelin may be mediated through the carbohydrate recognition via the ASGP-R subunit 1.
Our observations clearly demonstrate that the rat ASGP-R is capable of binding oligosaccharides terminating with Sia␣2,6GalNAc␤ as well as ones terminating with GalNAc or Gal. Furthermore, glycoproteins bearing oligosaccharides terminating with Sia␣2,6GalNAc␤ are rapidly removed from the blood by the ASGP-R. Although PLPs bearing 2-4 termini with the sequence Sia␣2,6GalNAc␤ would not likely be removed from the circulation as rapidly as SiaGGnM-BSA, their halflives, like that of LH, may reflect clearance by the ASGP-R. Our observations raise the possibility that oligosaccharides bearing sialic acid in ␣2,6-linkage to GalNAc and possibly also to Gal may represent endogenous ligands for the ASGP-R.