Asialofetuin Research Articles

Human lung adenocarcinoma alpha1,3/4-L-fucosyltransferase displays two molecular forms, high substrate affinity for clustered sialyl LacNAc type 1 units as well as mucin core 2 sialyl LacNAc type 2 unit and novel alpha1,2-L-fucosylating activity.

Human lung tumor alpha1,3/4-L-fucosyltransferase (FT) was purified (2000-fold, 29% recovery) from 290 g of tissue by including a chromatography step on Affinity Gel-GDP. Two molecular forms (FTA, larger size carrying 15% alpha1,4-FT activity; FTB, the major form with 85% activity) were separated by further fractionation on a Sephacryl S-100 HR column. A difference in the electrophoretic mobilities of these two activities was also found on native polyacrylamide gel electrophoresis (PAGE). Both forms were devoid of typical alpha1,2-fucosylating activity but were associated with the novel alpha1,2-fucosylating ability of converting the Lewis a determinant to Lewis b. Based on percentage activity toward 2-O-MeGalbeta1,3GlcNAcbeta-O-Bn, both forms exhibited the same extent of activity toward various acceptors, which included sulfated, sialylated, or methylated LacNAc type 1 or type 2 as well as mucin core 2 acceptors. However, FTA and FTB exhibited a difference in their ability to act on mucin core 2 3'-sialyl LacNAc (activities 24.2% and 40.8%, respectively, as compared to 2-O-MeGalbeta1,3GlcNAcbeta-O-Bn). The unsubstituted LacNAc type 1 acceptors were 15-20 times as active as the corresponding LacNAc type 2 acceptors. The 3-O-substitution on the beta1,4-linked Gal (methyl, sulfate, or sialyl) in mucin core 2 acceptors increased the efficiency of these acceptors five- to eightfold. The most efficient acceptor for FTA and FTB was 3-O-sulfoGalbeta1,3GlcNAcbeta-O-Al (K(m) 100 and 47 microM, respectively). The K(m) (mM) values for 2-O-methyl Galbeta1,3GlcNAcbeta-O-Bn and 3-O-sialyl Galbeta1,3GlcNAcbeta-O-Bn were 0.40 and 2.5 (FTA) and 0.16 and 0.67 (FTB), respectively. The 35-kDa glycoprotein ancrod (from Malayan pit viper venom) containing 36% complex N-glycans with the antennae NeuAcalpha2,3Galbeta1,3GlcNAcbeta- acted as the best macromolecular acceptor substrate (K(m): 45 microM), as examined with FTB. On desialylation the acceptor efficiency dropped to approximately 50% (K(m) for asialo ancrod: 167 microM). Sialylglycoproteins, such as carcinoembryonic antigen, fetuin, and bovine alpha(1)-acid glycoprotein, were better acceptors than asialo fetuin. On the contrary, fetuin triantennary glycopeptide containing predominantly NeuAcalpha2,3Galbeta1,4GlcNAcbeta- was only 55% active as compared to the asialo glycopeptide (K(m): 1.43 and 0.63 mM, respectively). Thus, the human lung tumor alpha1,3/4-L-FT has the potential to generate clustered sialyl Lewis a and Lewis b determinants in N-glycans and sialyl Lewis x determinant in mucin core 2 structures.

Differences in Zero-Force and Force-Driven Kinetics of Ligand Dissociation from β-Galactoside-Specific Proteins (Plant and Animal Lectins, Immunoglobulin G) Monitored by Plasmon Resonance and Dynamic Single Molecule Force Microscopy

Protein-carbohydrate interactions are involved in diverse regulatory processes. To help understand the mechanics and kinetics of dissociation of receptor-ligand complexes, we have analyzed the separation of lactose and the N-glycan chains of asialofetuin (ASF) from three lectins and an immunoglobulin G fraction by surface plasmon resonance at zero force and by atomic force microscopy with variations of the external force. While the (AB)2 agglutinins from Ricinus communis (RCA) and Viscum album (VAA) show structural homology, the homodimeric galectin-1 from bovine heart (BHL) has no similarity to the two plant lectins except for sharing this monosaccharide specificity. The β-galactoside-binding immunoglobulin G (IgG) fraction from human serum provides a further model system with distinct binding-site architecture. The koff constants for the two plant agglutinins were independent of the nature of the ligand at 1.1–1.3 × 10−3 s−1, whereas the geometry of ligand and binding site presentation affected this parameter for BHL (0.5 × 10−3 s−1 for lactose and 1 × 10−3 s−1 for ASF) and IgG (1.3 × 10−3 s−1 for lactose and 0.55 × 10−3 s−1 for ASF). When assessing comparatively the rupture forces at a loading rate of 3 nN/s with lactose as ligand, 34 ± 6 pN (BHL), 36 ± 4 pN (IgG), 47 ± 7 pN (VAA), and 58 ± 9 pN (RCA) were measured. For the same loading rate the rupture forces for the receptor-ASF interactions were found to be 37 ± 3 pN (BHL), 43 ± 5 pN (VAA), 45 ± 6 pN (IgG), and 65 ± 9 pN (RCA). The variation of the pulling velocity revealed in all cases a linear dependence between the rupture force and the natural logarithm of the loading rate. Performing probability density and Monte Carlo calculations, the potential barrier widths, which determine the inverse dynamic dependence with the rate of force elevation, increased from 4 Å (RCA) and 7 Å (VAA and IgG) to 10 Å (BHL) for the receptor-lactose interactions. Presenting ASF as ligand potential widths of 4 Å for RCA and IgG and 6 Å for VAA and BHL were obtained. Since the dissociation kinetics at zero force apparently cannot predict the behavior in force-driven experiments, these results reveal new insights into biological functions. The dissociation kinetics under force helps to explain the difference in the toxic potency of VAA and RCA and points to a function of the galectin in cis-crosslinking and in transient trans-bridging.

Role of the macrophage galactose lectin in the uptake of desialylated LDL

Desialylated low density lipoprotein (LDL) is rapidly taken up and accumulated by both peripheral blood monocytes and cells isolated from human arterial intima consisting predominantly of smooth muscle cells. It is shown that thioglycollate (TG)-elicited mouse macrophages and mouse peritoneal macrophages stimulated with lipopolysaccharide (LPS) show increased expression of a membrane-bound, galactose-specific lectin that could be responsible for this uptake. In LPS-stimulated macrophages accumulation of desialylated LDL is increased ca. 2.6-fold. Accumulation of acetylated LDL in the same cells is reduced, suggesting that the galactose-specific lectin might be responsible for the uptake of desialylated LDL. Transfection of cells with the mouse macrophage Gal/GalNAc-specific lectin (MMGL) increased their capacity to take up asialofetuin (ASF) and, to a smaller extent, desialylated LDL. The uptake of desialylated LDL was small, most likely due to the high k d of MMGL for biantennary oligosaccharides as found on LDL, and low concentration of LDL achieved in tissue culture experiments. The data suggest that the expression of galactose-specific lectins can be elevated under inflammatory conditions, and that these receptors could contribute to foam cell formation under conditions of high desialylated LDL concentration, as might be found in arterial intima.

Forssman pentasaccharide and polyvalent Galβ1→4GlcNAc as major ligands with affinity for Caragana arborescens agglutinin

The binding properties of Caragana arborescens agglutinin (CAA, pea tree agglutinin) were studied by enzyme linked lectinosorbent assay (ELLSA) and by inhibition of CAA-glycan interaction. Among glycoproteins (gps) tested, CAA reacted strongly with asialo bird nest gp, asialo rat sublingual gp, human Tamm-Horsfall Sd(a +) urinary gp (THGP) and asialo THGP that are rich in GalNAcα1→, GalNAcβ1→ and/or Galβ1→4GlcNAc residues. CAA also bound tightly with multi-valent Galβ1→4GlcNAc (mII) containing glycoproteins (human blood group precursor gps, asialo fetuin) and asialo ovine salivary glycoprotein (Tn, GalNAcα1→Ser/Thr), but CAA reacted poorly or not at all with sialylated glycoproteins tested. Of the sugars tested for inhibition of binding, Forssman pentasaccharide (F p, GalNAcα1→3GalNAcβ1→3Galα1→4Galβ1→4Glc) was the best. It was about 2.3, 9.5 and 52.6 times more active than Galβ1→4GlcNAc, GalNAc and Gal, respectively, and about 1.9 times more active than tri-antennary Galβ1→4GlcNAc (Tri-II). These results suggest that this agglutinin is mainly specific for F p, mII and Tn clusters. This property can be used to detect human abnormal glycotopes related to F p and unmasked mII/Tn clusters and to study cell growth and differentiation given the lack of toxicity of this lectin toward mouse fibroblast cells.

Role of sialylation in determining the pharmacokinetics of neutrophil inhibitory factor (NIF) in the Fischer 344 rat

1. Recombinant neutrophil inhibitory factor (NIF) is a glycoprotein. Its amino acid sequence remains constant and has a molecular weight of 28.9 kD. However, ~40% of the total molecular weight consists of glycans with variable structure. 2. The pharmacokinetics of 11 different NIF batches with varying extents and patterns of sialylation have been investigated in the Fischer 344 rat following intravenous administration. These data indicate that reducing the extent of NIF sialylation reduces the half-life of the molecule due to an increase in the systemic clearance. Also, an increase in the number of unsialylated or neutral glycans may increase the volume of distribution of NIF, although this effect is marginal. 3. Isolated perfused rat liver (IPRL) investigations have shown that sialylated NIF has a low hepatic extraction (< 1%), while asialo NIF has an extraction that is > 20-fold higher. Co-administration of asialo NIF with asialo fetuin (a protein cleared by hepatic asialoglycoprotein receptor (possibly galactose)-mediated uptake reduced the hepatic extraction of asialo NIF. 4. These data suggest that NIF molecules that have free sugar moieties (possibly galactose) interact with an asialoglycoprotein receptor (possibly galactose-mediated) in the liver (parenchymal cells/hepatocytes). Interaction with this receptor leads to cellular internalization and degradation.

Complement-mediated cytotoxicity and inhibition of ligand binding to hepatocytes by woodchuck hepatitis virus-induced autoantibodies to asialoglycoprotein receptor.

Hepadnavirus invasion in woodchucks has been identified as a potent inducer of autoantibodies against asialoglycoprotein receptor (anti-ASGPR), a molecule essentially unique to hepatocytes that mediate clearance of desialylated serum proteins. We evaluated the possible pathogenetic importance of anti-ASGPR triggered by woodchuck hepatitis virus (WHV), using anti-ASGPR-reactive serum immunoglobulins (Igs) from five animals with different stages of WHV hepatitis or self-limited WHV infection and isolated woodchuck hepatocytes or HepG2 cells as targets. The results revealed that WHV-induced anti-ASGPR can specifically inhibit asialoglycoprotein recognition by both homologous and heterologous liver cells, as tested in an asialofetuin (ASFN)-binding radioassay. However, the extent of the interference significantly varied (from 85% inhibition to none) for anti-ASGPR with similar titer from different animals, indicating a high degree of heterogeneity in the ASGPR epitope specificity and in the potential biological effects of these autoantibodies. The WHV-triggered anti-ASGPR also induced complement-mediated hepatocytolysis in a microculture tetrazolium (MTT) assay, which ranged from 8.9% +/- 0.3% to 33.6% +/- 3.6% (mean +/- SD) for different animals and target cell numbers. This cytopathic effect was strictly ASGPR-specific, complement-dependent, and was not related to the anti-ASGPR ability to inhibit ligand-hepatocyte binding. Our findings indicate that among pathways by which anti-ASGPR autoimmunity could cause liver damage, hepadnavirus-induced anti-ASGPR might impair hepatocytes by both disrupting clearance of desialylated proteins and activation of the complement-mediated cytolysis. These cytopathic effects might contribute to the pathogenesis, aggravate severity, and prolong recovery from liver injury in viral hepatitis.

Multi-antennary Galβ1→4GlcNAc and Galβ1→3GalNAc clusters as important ligands for a lectin isolated from the sponge Geodia cydonium

The affinity of a lectin from the sponge Geodia cydonium (GCL-I) for multi-antennary Galβ1→4GlcNAc and Galβ1→3GalNAc ligands was studied by both the biotin/avidin-based microtiter plate lectin binding assay and the inhibition of lectin-glycoform interaction. Among the glycoforms tested for binding, GCL-I reacted strongly with three multi-antennary Galβ1→4GlcNAc clusters containing glycoproteins (asialo human and bovine α 1-acid gps and asialo fetuin), T (Galβ1→3GalNAc) rich glycoprotein from porcine salivary gland, asialo bird nest gp, and human blood group A active cyst gp, while human and bovine α 1-acid gps, fetuin, and Tn containing gps were inactive. Among the haptens tested for inhibition, tri-antennary Galβ1→4GlcNAc (Tri-II) was about 1500, 72, and 72 times more active than GalNAc, Galβ1→4GlcNAc (II), and Galβ1→3GalNAc (T), respectively. Based on the present and previous results, it is proposed that tri-antennary Galβ1→4GlcNAc and Galβ1→3GalNAc clusters, in addition to GalNAcα1→3GalNAc and GalNAcα1→3Gal, are also important ligands for binding; and sialic acid of glycoprotein does interfere with binding.

Increased liver uptake of liposomes and improved targeting efficacy by labeling with asialofetuin in rodents

To improve liposome-directed therapy of liver disease and gene delivery, it would be beneficial to selectively target hepatocytes. For this purpose, conventional liposomes (CL) were labeled with asialofetuin (AF), an asialoglycoprotein. The biodistribution of AF-labeled liposomes (AF-L) in mice and their incorporation into rat hepatocytes, and their potential use in acute liver injury, were investigated. AF-L displayed a quicker plasma clearance than CL, and 25.4%, 2.7%, and 1.2% of the injected dose remained in the plasma versus 47.0%, 26.1%, and 9.5% of CL, respectively at 2, 4, and 20 hours after the injection. Total liver uptake of AF-L (73%+/-3.9%) was markedly higher (P < .005) than CL (16.5%+/-1.8%) 4 hours after the injection. Liposomal radioactivity (cpm/mg) was greatly enhanced in the liver (11-fold) during the first 4 hours after the administration of 14C-AF-L, and was much higher than in 14C-CL-injected mice (1.5-fold). In vitro incubation of isolated rat hepatocytes with 14C-AF-L or intravenous injection of 14C-AF-L in rats resulted in higher hepatocyte-bound radioactivity compared with 14C-CL (P < .01-.005). AF-L-associated 1,1'-dilinoleyl-3,3,3',3'-tetramethylindocarbocyanine perchlorate (DiI) fluorescent signals were not only located in Kupffer cells, but also in hepatocytes, in which bile canaliculus networks were imaged. Intravenous administration of vitamin E (VE)-associated CL (VE-CL, 1 mg/mouse) significantly lowered alanine transaminase (ALT) levels in CCl4-treated mice (196+/-79 vs. 2,107+/-235 U/mL; P < .01). The ALT level in CCl4 + VE-AF-L group was decreased to 38+/-16 units/mL, which was significantly lower than the CC14 + VE-CL group (P < .05). In conclusion, labeling liposomes with AF led to a shortened liposome plasma half-life and greatly enhanced uptake of AF-L liposome by the liver. The enhanced uptake resulted from an increased incorporation of hepatocytes with AF-L liposomes. VE-associated AF liposomes further improved the protective effect of VE liposomes on CC14-induced acute liver injury in mice. Preferential hepatocyte incorporation of AF-L liposomes suggests a useful hepatocyte-targeting approach for drug delivery and gene transfection.

Dynamic liver imaging with iron oxide agents: effects of size and biodistribution on contrast.

In vivo effective relaxation rates in normal rat liver were evaluated for four dextran coated iron oxide agents: monocrystalline iron oxide nanocolloid (MION) with a mean particle diameter of 3.9 nm, a polycrystalline agent (PION) with a larger mean diameter of 12 nm, and these two agents labeled with the asialofetuin (ASF) protein for high hepatocytic receptor binding affinity (MION-ASF and PION-ASF). Using echo planar imaging at 2 Tesla, dose response was measured with measured with high temporal resolution for 3 h after injection of agent, and by comparing with relaxivities in vitro and in brain, dominant in vivo contrast phenomena were elucidated. While transverse relaxivity for PION-ASF exceeded that for MION-ASF by almost a factor of 2 in solution, relaxation rates in vivo became equivalent. Liver relaxation using non-ASF agents was consistent with rapid water exchange between vascular and extravascular compartments, which dominated relaxation as a result of agent accumulation in Kupffer cells.

Covalent grafting of fibronectin and asialofetuin at surface of poly(ethylene terephthalate) track-etched membranes improves adhesion but not differentiation of rat hepatocytes.

The adhesion and differentiation of rat hepatocytes onto track-etched poly(ethylene terephthalate) (PET) membranes were studied. Native or [3H]methylated fibronectin (FN) or asialofetuin (ASF) were immobilized onto oxidized PET membranes, preactivated or not with carbodiimide (WSC). Radiochemical assay and ELISA indicated that upon washing in SDS or in serum-containing medium, significantly less FN was released from WSC-activated membranes that from unactivated ones. These differences were abolished when the NH2 functions of FN were fully acetylated, suggesting that part of FN was covalently grafted on WSC-activated PET. Although weaker, a comparable effect was observed with ASF. Hepatocytes adhered faster on membranes on which FN was grafted than adsorbed, reaching values comparable to collagen-coated PET; after 24 h these differences decreased. Experiments with cycloheximide or at 4 degrees C suggested that this results from secretion of extracellular matrix adsorbing on PET. Phase I or II biotransformation activities of cells cultured for 1-4 days on FN-, ASF-, or collagen-treated substrates were not significantly different. These results indicate that stable immobilization of FN (covalent grafting) onto PET membranes significantly accelerates adhesion of hepatocytes but does not affect their differentiation. This may result from a progressive surface reconditioning by neosynthesized extracellular matrix.

Covalent grafting of fibronectin and asialofetuin at surface of poly(ethylene terephthalate) track‐etched membranes improves adhesion but not differentiation of rat hepatocytes

The adhesion and differentiation of rat hepatocytes onto track-etched poly(ethylene terephthalate) (PET) membranes were studied. Native or [3H]methylated fibronectin (FN) or asialofetuin (ASF) were immobilized onto oxidized PET membranes, preactivated or not with carbodiimide (WSC). Radiochemical assay and ELISA indicated that upon washing in SDS or in serum-containing medium, significantly less FN was released from WSC-activated membranes than from unactivated ones. These differences were abolished when the NH2 functions of FN were fully acetylated, suggesting that part of FN was covalently grafted on WSC-activated PET. Although weaker, a comparable effect was observed with ASF. Hepatocytes adhered faster on membranes on which FN was grafted than adsorbed, reaching values comparable to collagen-coated PET; after 24 h these differences decreased. Experiments with cycloheximide or at 4°C suggested that this results from secretion of extracellular matrix adsorbing on PET. Phase I or II biotransformation activities of cells cultured for 1–4 days on FN-, ASF-, or collagen-treated substrates were not significantly different. These results indicate that stable immobilization of FN (covalent grafting) onto PET membranes significantly accelerates adhesion of hepatocytes but does not affect their differentiation. This may result from a progressive surface reconditioning by neosynthesized extracellular matrix. © 1996 John Wiley & Sons, Inc.

Structural elucidation of o-linked glycopeptides by high energy collision-induced dissociation

O-linked glycopeptides that bear a GalNAc core with and without the presence of sialic acid have been analyzed by high energy collision-induced dissociation (CID). We show that the CID spectra from the glycosylated precursor ions contain sufficient information to identify the peptide sequence and to determine the glycosylated site(s). Asialo O-linked glycopeptides, previously prepared from a tryptic digest of bovine fetuin were studied. One of the glycopeptides contained only a single Hex (hexose)-HexNAc (N-acetylhexosamine) substitution at Thr(262), whereas the other exhibited Hex-HexNAc moieties at both Thr(262) and Ser(264). In addition, sialo and asialo fetuin glycopeptides from a pronase digest were derivatized with t-butoxycarbonyl-tyrosine, and characterized by high energy CID analysis. The presence of a Galβ(1,3)GalNAc core structure at Ser(264) was confirmed by using the substrate specificity of endo-α-N-acetylgalactosaminidase. These studies revealed the presence of a β-galactosidase specific for β(1,4) linkages in the endo-α-N-acetylgalactosaminidase preparation employed. Finally, the relative stability of N-and O-glycosyl bonds to high energy CID is addressed based upon comparison of the behavior of a synthetic N-linked glycopeptide with analogous O-linked structures.

High-Affinity Ligand Binding to Subunit HI of the Asialoglycoprotein Receptor in the Absence of Subunit H2

The hepatic asialoglycoprotein receptor is a hetero-oligomer composed of two homologous subunits. The specificity and affinity of ligand binding depends on the number and spatial arrangement of several galactose-binding sites within the receptor complex. Previous studies indicated that both subunits are required for high-affinity ligand binding, i.e. for the simultaneous interaction with three galactose residues within an N-linked glycan. However, we found that asialoorosomucoid (ASOR) and asialofetuin (ASF) bind to transfected COS-7 cells expressing subunit H1 in the absence of the second subunit H2. ASOR binding occurred with a dissociation constant of approximately 40 nM, approximately four-times higher than the Kd of ASOR binding to the hetero-oligomeric receptor. Normalized to the amount of H1 expressed, approximately 10–times fewer binding sites were produced by H1 alone. A glycopeptide with a single tri-antennary N-linked glycan purified from ASF bound to the hetero-oligomeric receptor, but did not bind detectably to HI-expressing COS-7 cells. H1 is thus unable to simultaneously recognize all three galactose residues in a glycan. From this, we conclude that, at a sufficiently high density of HI on the cell surface, high-affinity binding of ASOR and ASF is the result of two or more glycans interacting with H1 oligomers with low affinity in a bivalent manner.

High-affinity ligand binding to subunit H1 of the asialoglycoprotein receptor in the absence of subunit H2.

The hepatic asialoglycoprotein receptor is a hetero-oligomer composed of two homologous subunits. The specificity and affinity of ligand binding depends on the number and spatial arrangement of several galactose-binding sites within the receptor complex. Previous studies indicated that both subunits are required for high-affinity ligand binding, i.e. for the simultaneous interaction with three galactose residues within an N-linked glycan. However, we found that asialoorosomucoid (ASOR) and asialofetuin (ASF) bind to transfected COS-7 cells expressing subunit H1 in the absence of the second subunit H2. ASOR binding occurred with a dissociation constant of approximately 40 nM, approximately four-times higher than the Kd of ASOR binding to the hetero-oligomeric receptor. Normalized to the amount of H1 expressed, approximately 10-times fewer binding sites were produced by H1 alone. A glycopeptide with a single tri-antennary N-linked glycan purified from ASF bound to the hetero-oligomeric receptor, but did not bind detectably to H1-expressing COS-7 cells. H1 is thus unable to simultaneously recognize all three galactose residues in a glycan. From this, we conclude that, at a sufficiently high density of H1 on the cell surface, high-affinity binding of ASOR and ASF is the result of two or more glycans interacting with H1 oligomers with low affinity in a bivalent manner.

Isolation and Characterization of a &amp;beta; -Galactoside-Binding Protein (14 kD) from Rat Liver Nuclei

A soluble carbohydrate-binding protein with galactoside-binding properties (CBP14) was isolated from liver nuclei of adult rats. Proteins were extracted from membrane-depleted nuclei; they contained 3 major fractions (Mr 43,000, 14,000 and 10,000) which reacted with antibodies raised against human placental β-galactoside-binding lectin (HP 14) as well as with asialofetuin (ASF). Enrichment (72- to 143-fold) of CBP14 was carried out by affinity chromatography, using anti-HP 14 antibodies and ASF as consequent ligands. CBP14 retained by affinity column chromatography was characterized by the molecular mass, haemagglutinating activity and immuno-chemical cross-reactivity. By fluorescence technique using fluorescein-labelled neoglycoprotein (carrying lactose) as a probe, lactose-binding protein(s), e.g. the CBP14 described here, can be detected in a homogeneous staining pattern within the entire nucleus of CV-1 cells in the stationary phase of the cells. In case of logarithmically proliferating cells the lactose-binding protein(s) are arranged in a spot-like pattern along the rims of the cell nuclei. The putative ligand for the CBP14 is associated with the nuclear surface, irrespective of the proliferation state of the cells.

Demonstration of the interaction between the CD23 molecule and the galactose residue of glycoproteins

The CD23 molecule is a low-affinity receptor for IgE and has a marked homology in amino acid sequence with C-type animal lectins, including asialoglycoprotein receptor. We tested whether the CD23 antigen can indeed interact with the sugar chain of glycoproteins. Detergent extract of the membrane component from Epstein-Barr Virus (EBV)-transformed human B-cell line, L-KT9 cells, was incubated with asialofetuin (ASF)-coupled Sepharose, and bound proteins were effectively eluted by 0.3 M lactose or galactose which were among the competitive sugars tested. In this eluate, the CD23 molecule was detected by an immunoblotting technique. Because fetuin has both an N- and O-type sugar chain on the molecule, we tested which type of sugar chain can interact with CD23. The CD23 molecule interacted with asialocasein having a sugar chain with the Gal-GalNAc structure with asialobovine submaxillary mucin having the GalNAc structure, and also with ASF; however, it faintly interacted with ASF after removal of the O-type sugar chain by β-elimination. These results showed that the CD23 molecule can, indeed, interact with the galactose residue, especially with the Gal-GalNAc rather than the Gal-GlcNAc structure of the terminal sugar chain of glycoproteins.

Differences in the cross-linking activities of native and recombinant Erythrina corallodendron lectin with asialofetuin. Evidence for carbohydrate-carbohydrate interactions in lectin-glycoprotein complexes.

A previous study showed that several multivalent galactose-specific lectins including the 14-kDa lectin from calf spleen and the lectins from Erythrina indica, Erythrina cristagalli, and soybean agglutinin formed specific cross-linked complexes with the glycoprotein asialofetuin (ASF) [Mandal, D. K., & Brewer, C. F. (1992) Biochemistry 31, 8465-8472]. In the present study, we have used quantitative precipitation analysis to compare the cross-linking activities of the Gal/GalNAc-specific lectin from Erythrina corallodendron (ECorL) and the recombinant protein (rECorL) which lacks the covalently linked heptasaccharide chains of the native lectin, with ASF. At low concentrations of ASF relative to the lectin, native dimeric ECorL binds to each of the three terminal Gal residues of the three N-linked triantennary chains of ASF and precipitates as a cross-linked complex at a ratio of 1:9 ASF/lectin (monomer). With increasing concentrations of ASF, the 1:9 complex changes to a 1:3 ASF/lectin complex, and at higher ASF concentrations, a 1:1 cross-linked complex forms. However, rECorL, which possesses the same specificity and binding affinity as the native lectin, forms only the 1:9 and 1:3 ASF/lectin complexes. Other Erythrina lectins examined, all of which have covalently attached carbohydrate and are structurally similar to ECorL, show the same cross-linking behavior as native ECorL. On the other hand, the dimeric 14-kDa calf spleen lectin which lacks covalently attached carbohydrate forms only 1:9 and 1:3 cross-linked complexes with ASF [Mandal, D. K., & Brewer, C. F. (1992) Biochemistry 31, 8465-8472].(ABSTRACT TRUNCATED AT 250 WORDS)

Effect of asialofetuin on prolongation of skin allograft survival by donor bone marrow cells.

Survival of skin allografts made to antilymphocyte serum (ALS)-treated recipients is prolonged by posttransplant intravenous injection of donor strain bone marrow cells (BMC). If asialofetuin (ASF) is coinjected with the BMC, the prolonged graft survival is augmented (e.g., median survival time increased from 43 days to 72 days by injection of ASF). We have confirmed that, like peanut agglutinin-binding stem cells, the active BMC are at risk for hepatic sequestration after injection, possibly via hepatic asialoglycoprotein receptors. A fraction enriched for these active cells binds to liver sections in vitro and localizes to liver in vivo. This binding and localization can be partly inhibited by ASF. Although injected cells could also be found in the spleen, the beneficial effect of ASF could be demonstrated in previously splenectomized mice. Also, splenectomy 2 hr after BMC injection (without ASF) had little effect on the BMC-induced prolonged graft survival, while transfer of cells from the removed spleens to secondary ALS-treated recipients did not transfer such prolongation. In contrast, transfer of BMC from primary, donor marrow-injected recipients did transfer graft-prolonging activity, especially if both primary and secondary recipients were ASF injected. Our results suggest that recipient marrow, but not spleen, is the site of short-term localization of graft survival-prolonging BMC. Augmentation of allograft survival with ASF in ALS-treated recipients appears to result from the diversion of BMC away from a microenvironment not conducive to the expression of their graft-prolonging activity.

17 alpha-ethinylestradiol increases transcytosis of asialoglycoproteins in rat liver.

We have measured the hepatobiliary transcytosis of asialofetuin (ASF) and low density lipoprotein (LDL) in rats following treatment with 5 mg of 17 alpha-ethinylestradiol/kg of body weight for 3 days. The rate of appearance of both ligands in bile was increased 5-fold in estradiol-treated rats compared with controls. In contrast, no increase was observed in the transcytosis of dimeric IgA. The majority of the ASF recovered from the bile of estradiol-treated rats was undegraded, indicating transcytic delivery rather than lysosomal discharge. No increase in the rate of endocytosis of ASF was observed. When transcytosis of 125I-LDL was measured in the presence of excess ASF the rate decreased by approximately 70%. Moreover, sialidase treatment of LDL in vitro increased the biliary appearance of LDL 2-fold in estradiol-treated rats, suggesting that the effect of estradiol on both ligands was mediated via the asialoglycoprotein receptor. We conclude that 17 alpha-ethinylestradiol increases the transcytosis of ASF and LDL through mechanisms that alter the intracellular trafficking of the asialoglycoprotein receptor.

Hepatocyte horseradish peroxidase uptake is saturable and inhibited by mannose-terminal glycoproteins.

In the liver, horseradish peroxidase (HRP) is thought to be taken up via mannose receptor-mediated endocytosis by non-parenchymal cells (NPC) and via fluid-phase endocytosis by hepatocytes. When we attempted to inhibit NPC uptake of HRP with mannan in the whole perfused rat liver, > 80% of HRP uptake was eliminated. Liver cell fractionation revealed that mannan not only inhibited HRP uptake by NPC (91%) but also by hepatocytes (81%). In isolated hepatocytes, HRP uptake was linear over 60 min and saturable in the range of 0 to 200 mg/l (Vmax = 4.3 ng.mg protein-1.min-1; Km = 8.3 mg/l). Mannan inhibited uptake competitively (Ki = 2.0-2.5 mg/l). At high concentrations of HRP, a nonsaturable component of HRP uptake became evident (k = 2.8 pg.mg protein-1.min-1.mg HRP-1.l-1). Hepatocyte uptake of HRP was inhibited by other glycoproteins and glycopeptides with mannose-terminal groups, as well as by mannan, but not by asialofetuin (ASF) or bovine serum albumin. Hepatocyte uptake of 125I-labeled ASF, which is taken up via the asialoglycoprotein receptor, was saturable and not inhibited by mannan. HRP binding to hepatocytes, determined at 4 degrees C, was also inhibited by mannan. Quantification of contamination of the parenchymal cell fraction by NPC by cell counting and by pronase digestibility suggested our results could not be explained by contamination of hepatocytes by NPC. At concentrations used for most morphological studies (1,000-10,000 mg/l), fluid-phase endocytosis accounts for much of HRP uptake. However, at low concentrations, a saturable low-capacity mechanism is responsible for most HRP uptake by the hepatocyte.(ABSTRACT TRUNCATED AT 250 WORDS)