The substrate specificity of bovine and feline lysosomal alpha-D-mannosidases in relation to alpha-mannosidosis

Lysosomal alpha-mannosidases were partially purified from bovine and feline liver and employed to digest a large number of oligosaccharides with structures corresponding to the oligomannosyl parts of complex, hybrid, and high-mannose glycans. The incubation products were identified by high pressure liquid chromatography with reference compounds of defined structure and by acetolysis. For all classes of substrates, the lysosomal alpha-mannosidases displayed a high degree of in vitro specificity with regard to the hydrolysis of mannose residues. Thus, in each case, 1 or at most 2 residues were always preferentially cleaved so that the degradative process proceeded down a well defined pathway. A comparison of the relative efficiency with which lysosomal alpha-mannosidases catalyzed the hydrolysis of particular oligosaccharides and of the structures of the resulting intermediates with those of the compounds accumulated in alpha-mannosidosis allows conclusions to be drawn regarding the nature of the enzymatic defect. In bovine alpha-mannosidosis, the oligosaccharides are those expected for a partial deficiency of normal lysosomal alpha-mannosidase, so that they correspond to intermediates in the normal catabolic pathway. In feline alpha-mannosidosis, in which the alpha-mannosidase deficiency is more severe than in cattle, the accumulated oligosaccharides primarily represent intact oligomannosyl moieties of N-linked glycans rather than the products of residual alpha-mannosidase activity.     

Evidence for an alpha-mannosidase in endoplasmic reticulum of rat liver

An alpha-mannosidase activity has been identified in a preparation of rat liver endoplasmic reticulum and shown to be distinct from the previously described Golgi alpha-mannosidases I and II and the lysosomal alpha-mannosidase. The enzyme was solubilized with deoxycholate and separated from other alpha-mannosidases by passage over concanavalin A-Sepharose to which it does not bind. The endoplasmic reticulum alpha-mannosidase cleaves alpha-1,2-linked mannoses from high mannose oligosaccharides and, unlike Golgi alpha-mannosidase I, is active against p-nitrophenyl-alpha-D-mannoside (Km = 0.17 mM). It has no activity toward GlcNAc-Man5GlcNAc2 peptide, the specific substrate of the Golgi alpha-mannosidase II. The endoplasmic reticulum alpha-mannosidase activity toward p-nitrophenyl-alpha-D-mannoside is relatively insensitive to swainsonine, an inhibitor of both the lysosomal alpha-mannosidase and Golgi alpha-mannosidase II. We propose that the endoplasmic reticulum alpha-mannosidase is responsible for the removal of mannose residues from asparagine-linked high mannose type oligosaccharides prior to their entry into the Golgi.     

Characterization of human liver alpha-D-mannosidase purified by affinity chromatography.

Human liver acidic alpha-D-mannosidase was purified 1400-fold by a relatively short procedure incorporating chromatography on concanavalin A-Sepharose and affinity chromatography on Sepharose 4B-epsilon-aminohexanoylmannosylamine. In contrast with the acidic enzymic activity the neutral alpha-mannosidase did not bind to the concanavalin A-Sepharose so the two types of alpha-mannosidase could be separated at an early stage in the purification. The only significant glycosidase contaminant after affinity chromatography on the mannosylamine ligand was alpha-L-fucosidase, which was selectively removed by affinity chromatography on the corresponding fucosylamine ligand. The final preparation was free of other glycosidase activities. The pI of the purified enzyme was increased from 6.0 to 6.45 on treatment with neuraminidase. Although the pI and the mol.wt. (220 000) suggested that alpha-mannosidase A had been purified selectively, ion-exchange chromatography on DEAE-cellulose indicated that the preparation consisted predominantly of alpha-mannosidase B. This discrepancy is discussed in relation to the basis of the multiple forms of human alpha-mannosidase. The purified enzyme completely removed the alpha-linked non-reducing terminal mannose from a trisaccharide isolated from the urine of a patient with mannosidosis. A comparison of the activity of the pure enzyme towards the natural substrate and synthetic substrates suggests that the same enzymic activity is responsible for hydrolysing all the substrates. These results validate the use of synthetic substrates for determining the mannosidosis genotype. They are also further evidence that mannosidosis is a lysosomal storage disease resulting from a deficiency of acidic alpha-mannosidase.     

The nature of the residual alpha-mannosidase in plasma in bovine mannosidosis.

Acidic alpha-mannosidase (EC 3.2.1.24), optimum pH 4.25, is absent from the plasma of Angus calves with mannosidosis, and the residual alpha-mannosidase activity has an optimum pH of 5.5, intermediate between that of the acidic and neutral alpha-mannosidases. This 'intermediate' alpha-mannosidase differs from the acidic form in its kinetic properties, its lack of marked inhibition by EDTA and its thermolability at 55 degrees C and physiological pH. Isoelectric focusing and ion-exchange chromatography show that it exists in at least two forms. The presence of a secondary peak at pH 5.5 in the pH/activity profile of normal plasma and the effect of heating at 55 degrees C indicate that such a form is present in normal plasma. The residual activity in the plasma of a calf with mannosidosis is therefore probably not the product of the defective gene. A differential assay, based on their different stabilities at 55 degrees C, has been developed for measuring the acidic and intermediate alpha-mannosidases in plasma. There was no correlation between the concentrations of the two enzymes in the plasma of Angus cows heterozygous for mannosidosis or in the plasma of normal animals. This precludes the use of the intermediate form as a reference enzyme for the acidic activity in a test for heterozygosity for mannosidosis based on the gene-dosage phenomenon. The concentrations of the intermediate activity were comparable in normal animals and animals homozygous or heterozygous for mannosidosis.     

The catabolism of mammalian glycoproteins. Comparison of the storage products in bovine, feline and human mannosidosis.

Analysis of the neutral urinary oligosaccharides in bovine, feline and human mannosidosis by thin-layer and gel-permeation chromatography has shown that the patterns of stored oligosaccharides in the three species are different. In bovine and feline mannosidosis the most abundant urinary oligosaccharide is also the most abundant in the tissues of each species. The predominant oligosaccharides were purified by a combination of gel-filtration, ion-exchange and thin-layer chromatography and shown to contain only mannose and N-acetylglucosamine by g.l.c. and g.l.c.--mass spectrometry. The probable composition and size of each oligosaccharide were predicted from its chromatographic properties, sugar composition and the known structure of asparagine-linked oligosaccharides. The bovine and feline oligosaccharides belonged to a homologous series of general composition Mann (GlcNAc)2, whereas the human oligosaccharides belong to a different series, MannGlcNAc. These structures suggest that lysosomal endohexosaminidase is not present in bovine and feline tissues. The predominant feline storage product, Man3(GlcNAc)2, was the expected storage product from the catabolism of complex asparagine-linked glycans. In contrast, the predominant bovine oligosaccharide, Man2(GlcNAc)2, probably lacks one of the alpha-linked mannose residues in the core region. A similar situation occurs in human mannosidosis. It is predicted that in these species either that the residual mutant alpha-D-mannosidase retains activity towards one of the core alpha-linked mannose residues or that another form of lysosomal alpha-D-mannosidase that is unaffected in these disorders occurs. It is concluded that the differences in storage products are due to differences in the catabolic pathways of glycoproteins among the species.     

A human lysosomal alpha-mannosidase specific for the core of complex glycans.

A novel lysosomal alpha-mannosidase, with unique substrate specificity, has been partially purified from human spleen by chromatography through concanavalin A-Sepharose, DEAE-Sephadex, and Sephacryl S-300. This enzyme can catalyze the hydrolysis of only 1 mannose residue, that which is alpha(1----6)-linked to the beta-linked mannose in the core of N-linked glycans, as found in the oligosaccharides Man alpha(1----6)[Man alpha(1----3)] Man beta(1----4)GlcNAc and Man alpha(1----6)Man beta(1----4) GlcNAc. The newly described alpha-mannosidase does not catalyze the hydrolysis of mannose residues outside of the core, even if they are alpha(1----6)-linked, and is not active on the other alpha-linked mannose in the core, which is (1----3)-linked. The narrow specificity of the novel mannosidase contrasts sharply with that of the major lysosomal alpha-mannosidase, which is able to catalyze the degradation of oligosaccharides containing diverse linkage and branching patterns of the mannose residues. Importantly, although the major mannosidase readily catalyzes the hydrolysis of the core alpha(1----3)-linked mannose, it is poorly active towards the alpha(1----6)-linked mannose, i.e. the very same mannose residue for which the newly characterized mannosidase is specific. The novel enzyme is further differentiated from the major lysosomal alpha-mannosidase by its inability to catalyze the efficient hydrolysis of the synthetic substrate p-nitrophenyl alpha-mannoside, and by the strong stimulation of its activity by Co2+ and Zn2+. Similarly to the major mannosidase, it is strongly inhibited by swainsonine and 1,4-dideoxy-1,4-imino-D-mannitol, but not by deoxymannojirimycin. The presence of this novel alpha-mannosidase activity in human tissues provides the best explanation, to date, for the structures of the oligosaccharides stored in human alpha-mannosidosis. In this condition the major lysosomal alpha-mannosidase activity is severely deficient, but apparently the alpha(1----6)-mannosidase is unaffected, so that the oligosaccharide structures reflect the unique specificity of this enzyme.     

Biochemical studies on a case of feline mannosidosis.

Evidence is presented for the biochemical diagnosis of the first case of feline mannosidosis. A marked deficiency of acidic alpha-D-mannosidase in the brain, kidney and liver and excessive excretion of mannose-rich oligosaccharides in the urine were found in a kitten suffering from a nervous disorder. Residual acidic alpha-D-mannosidase, ranging from 2 to 5.5% of the normal activity, was observed in the tissues of the affected kitten. It has similar kinetic and physicochemical properties to the normal activity. The amount of mannose in the urine of the affected kitten was 19-fold greater than in a comparable control, and the molar ratio of mannose to N-acetylglucosamine was approx. 6 : 1. High concentrations of neutral oligosaccharides were detected in the urine. The predominant oligosaccharide appeared to be a hexasaccharide. The biochemical features of bovine, feline and human mannosidosis are compared, and it is concluded that feline mannosidosis may be a useful animal model for studying the human disease.     

Importance of glycosidases in mammalian glycoprotein biosynthesis

Processing glycosidases play an important role in N-glycan biosynthesis in mammalian cells by trimming Glc(3)Man(9)GlcNAc(2) and thus providing the substrates for the formation of complex and hybrid structures by Golgi glycosyltransferases. Processing glycosidases also play a role in the folding of newly formed glycoproteins and in endoplasmic reticulum quality control. The properties and molecular nature of mammalian processing glycosidases are described in this review. Membrane-bound alpha-glucosidase I and soluble alpha-glucosidase II of the endoplasmic reticulum remove the alpha1,2-glucose and alpha1,3-glucose residues, respectively, beginning immediately following transfer of Glc(3)Man(9)GlcNAc(2) to nascent polypeptides. The alpha-glucosidases participate in glycoprotein folding mediated by calnexin and calreticulin by forming the monoglucosylated high mannose oligosaccharides required for the interaction with the chaperones. In some mammalian cells, Golgi endo alpha-mannosidase provides an alternative pathway for removal of glucose residues. Removal of alpha1,2-linked mannose residues begins in the endoplasmic reticulum where trimming of mannose residues in the endoplasmic reticulum has been implicated in the targeting of malfolded glycoproteins for degradation. Removal of mannose residues continues in the Golgi with the action of alpha1, 2-mannosidases IA and IB that can form Man(5)GlcNAc(2) and of alpha-mannosidase II that removes the alpha1,3- and alpha1,6-linked mannose from GlcNAcMan(5)GlcNAc(2) to form GlcNAcMan(3)GlcNAc(2). These membrane-bound Golgi enzymes have been cloned and shown to have very distinct patterns of tissue-specific expression. There are also broad specificity alpha-mannosidases that can trim Man(4-9)GlcNAc(2) to Man(3)GlcNAc(2), and provide an alternative pathway toward complex oligosaccharide formation. Cloning of the remaining alpha-mannosidases will be required to evaluate their specific functions in glycoprotein maturation.     

Developmental regulation of asparagine-linked oligosaccharide synthesis in Dictyostelium discoideum

In the preceding report we demonstrated that the expression of two developmentally regulated alpha-mannosidase activities is induced in Dictyostelium discoideum during its differentiation from single-cell amoebae to multicellular organism (Sharkey, D. J., and Kornfeld, R. (1991) J. Biol. Chem. 266, 18477-18484). These activities, designated membrane alpha-mannosidase I (MI) and membrane alpha-mannosidase II (MII), were shown to have several properties in common with rat liver Golgi alpha-mannosidases I and II, respectively, suggesting that MI and MII may play a role in the processing of asparagine-linked oligosaccharides in developing D. discoideum. In this study we analyzed the structures of the asparagine-linked oligosaccharides synthesized by D. discoideum at various stages of development to determine the timing and extent of asparagine-linked oligosaccharide processing. Cells were labeled with [2-3H] mannose, and then total cellular glycoproteins were digested with Pronase to generate glycopeptides that were fractionated on concanavalin A-Sepharose. Glycopeptides from each fraction were digested with endoglycosidase H, both before and after desulfation by solvolysis, and the released, neutral oligosaccharides were sized by high pressure liquid chromatography. At early stages of development, D. discoideum contain predominantly large high mannose-type oligosaccharides (Man9GlcNAc and Man8GlcNAc). Some of these are modified by GlcNAc residues attached beta 1-4 to the mannose-linked alpha 1-6 to the beta-linked core mannose (the "intersecting" position), as well as by fucose, sulfate, and phosphate. In contrast, the oligosaccharides found at late stages of development (18-24 h) have an array of sizes from Man9GlcNAc to Man3GlcNAc. These are still modified by GlcNAc, fucose, sulfate, and phosphate, but the percent of larger high mannose oligosaccharides that are modified with GlcNAc in the intersecting position decreases after 6 h of development, in parallel with the decrease in the intersecting GlcNAc transferase activity. Similarly, the changes in the size of asparagine-linked oligosaccharides synthesized during development correlate well with the appearance of MI and MII activities and suggest that these developmentally regulated alpha-mannosidase activities function in the processing of these oligosaccharides. This is supported further by the observation that oligosaccharide processing was inhibited in late stage cells labeled in the presence of either deoxymannojirimycin, an inhibitor of MI, or swainsonine, an inhibitor of MII.     

Rapid characterization of asparagine-linked oligosaccharides isolated from glycoproteins using a carbohydrate analyzer

Chromatographic methods were developed for the separation and characterization of acidic (sialylated) and neutral (asialo-complex and high-mannose) oligosaccharides released from glycoproteins with peptide N-glycosidase F. endo-beta-N-acetylglucosaminidase F and endo-beta-N-acetylglucosaminidase H using a carbohydrate analyzer (Dionex BioLC). All the carbohydrate separations were carried out on a polymeric pellicular anion-exchange column HPIC-AS6/CarboPac PA-1 (Dionex) using only two eluants namely, 0.5 M NaOH and 3% acetic acid/NaOH pH 5.5, which were mixed with water to generate various gradients. Developed conditions for quantitative detection of carbohydrates with pulsed amperometry were necessary to obtain steady baselines at 0.1-0.3 microA output with suitable sensitivity (less than 5 pmol) in separations employing a variety of acidic and alkaline sodium acetate gradients. Oligosaccharides released from heat-denatured and trypsin-treated glycoproteins were purified initially from large-scale digestion (greater than 0.1 g) by extraction of peptide material into phenol/chloroform and finally by ion-exchange chromatography of the acqueous phase. Oligosaccharides isolated from the peptide N-glycosidase digests of bovine fetuin, human transferrin and alpha 1-acid glycoprotein gave multiple peaks in each charge group in separations based on the charge content at pH 5.5. Alkaline sodium acetate gradients were developed to obtain oligosaccharide maps of the glycoproteins within 60 min, in which separated oligosaccharides eluted in the order of neutral, mono-, di-, tri- and tetra-sialylated species based on both charge, size and structure. Baseline separations were obtained with neutral oligosaccharide types but mixtures of high-mannose and complex types were poorly resolved. The high-mannose peaks were eliminated specifically from complex oligosaccharides by digesting with alpha-mannosidase. Treatment with beta-galactosidase, beta-N-acetylglucosaminidase and alpha-mannosidase resulted in a decrease of the oligosaccharide elution times corresponding to the number of sugar residues lost, the profile of changes was highly reproducible. In contrast, treatment with alpha-L-fucosidase, endo-beta-N-acetylglucosaminidase F and endo-beta-N-acetylglucosaminidase H resulted in an increase in their corresponding oligosaccharide retention times similar to the presence of an additional sugar residue. Conditions developed for separation of the reduced oligosaccharides and also a mixture of monosaccharide to oligosaccharide containing about 15 sugar residues within 30 min were useful in determining the effect of endo- and exo-glycosidases on porcine thyroglobulin oligosaccharides. Changes in elution time of the oligosaccharides following specific glycosidase digestions combined with methylation analysis provided a rapid and sensitive tool for confirmation of the carbohydrate primary structures present in thyroglobulin.     

Characterization and subcellular localization of human neutral class II alpha-mannosidase [corrected

A glycosyl hydrolase family 38 enzyme, neutral alpha-mannosidase, has been proposed to be involved in hydrolysis of cytosolic free oligosaccharides originating either from ER-misfolded glycoproteins or the N-glycosylation process. Although this enzyme has been isolated from the cytosol, it has also been linked to the ER by subcellular fractionations. We have studied the subcellular localization of neutral alpha-mannosidase by immunofluorescence microscopy and characterized the human recombinant enzyme with natural substrates to elucidate the biological function of this enzyme. Immunofluorescence microscopy showed neutral alpha-mannosidase to be absent from the ER, lysosomes, and autophagosomes, and being granularly distributed in the cytosol. In experiments with fluorescent recovery after photo bleaching, neutral alpha-mannosidase had slower than expected two-phased diffusion in the cytosol. This result together with the granular appearance in immunostaining suggests that portion of the neutral alpha-mannosidase pool is somehow complexed. The purified recombinant enzyme is a tetramer and has a neutral pH optimum for activity. It hydrolyzed Man(9)GlcNAc to Man(5)GlcNAc in the presence of Fe(2+), Co(2+), and Mn(2+), and uniquely to neutral alpha-mannosidases from other organisms, the human enzyme was more activated by Fe(2+) than Co(2+). Without activating cations the main reaction product was Man(8)GlcNAc, and Cu(2+) completely inhibited neutral alpha-mannosidase. Our findings from enzyme-substrate characterizations and subcellular localization studies support the suggested role for neutral alpha-mannosidase in hydrolysis of soluble cytosolic oligomannosides.     

Mannosidosis in Angus cattle. The enzymic defect

Normal calf alpha-mannosidase activity exists in at least three forms separable by chromatography on DEAE-cellulose and by starch-gel electrophoresis. Two components, A and B, have optimum activity between pH3.75 and 4.75, but component C has an optimum of pH6.6. Components A and B are virtually absent from the tissues of a calf with mannosidosis and the residual activity is due to component C. The acidic and neutral forms of alpha-mannosidase differ in their molecular weights and sensitivity to EDTA, Zn(2+), Co(2+) and Mn(2+). An acidic alpha-mannosidase component (pH optimum 4.0) accounts for most of the activity in normal plasma but it is absent from the plasma of a calf with mannosidosis. Although the acidic alpha-mannosidase component is probably related to tissue components A and B, it can be distinguished from them by ion-exchange chromatography and gel filtration. The optimum pH of the low residual activity in the plasma from a calf with mannosidosis is pH5.5-5.75. The results support the hypothesis that Angus-cattle mannosidosis is a storage disease caused by a deficiency of lysosomal acidic alpha-mannosidase activity.     

Apparently normal extracellular acidic alpha-mannosidase in fibroblast cultures from patients with mannosidosis.

Fibroblasts from patients with mannosidosis, cultured in medium supplemented with fetal calf serum from which acidic alpha-mannosidase (alpha-D-mannoside mannohydrolase, E.C.3.2.1.24) has been removed, secreted a normal amount of apparently unaffected acidic alpha-mannosidase into fetal calf serum-free medium. Both the intracellular and extracellular acidic alpha-mannosidase activities were completely precipitated by antiserum to placenta alpha-mannosidase B. In contrast to the heat-lability of the intracellular acidic alpha-mannosidase and its low affinity for artificial mannoside substrate, the extracellular enzyme exhibited both normal thermostability and normal kinetics. Mixing experiments with the intercellular enzymes suggested that the decreased activity in the patients' fibroblasts is not the effect of an inhibitor or absence of an activator. However, incubation of the mannosidosis extracellular enzyme with either normal or patient cell lysate resulted in a partial loss of activity, whereas an additive value was observed with the normal extracellular enzyme. In contrast to normal culture medium, the medium from mannosidosis cell culture was unable to enhance the rate of reduction of intracellular radioactivity in mucolipidosis type II fibroblasts precultured in the presence of radiolabeled mannose. These findings suggest that the defect in mannosidosis is expressed only after the enzyme has been delivered to lysosomes and presumably undergone some form of processing there.     

1,4-Dideoxy-1,4-imino-D-mannitol inhibits glycoprotein processing and mannosidase

1,4-Dideoxy-1,4-imino-D-mannitol (DIM) was synthesized chemically from benzyl-alpha-D-mannopyranoside [Fleet et al (1984) J. Chem. Soc. Chem. Commun., 1240-1241], and was tested in vitro as an inhibitor of various alpha-mannosidases and in cell culture as an inhibitor of glycoprotein processing. DIM proved to be an effective inhibitor of jack bean alpha-mannosidase, with 50% inhibition requiring 25 to 50 ng/ml inhibitor. It also inhibited lysosomal alpha-mannosidase, but in this case 50% inhibition required about 1 to 2 micrograms/ml. In both cases, the inhibition was of the competitive type when p-nitrophenyl-alpha-D-mannopyranoside was used as the substrate. The inhibition was better at higher pH values, suggesting that DIM was more effective when the nitrogen in the ring was in the unprotonated form. In addition, rat liver processing mannosidase I was also inhibited by DIM as measured by the release of [3H]mannose from [3H]mannose-labeled Man9GlcNAc. Glycoprotein processing was examined in influenza virus-infected MDCK cells. Infected cells were incubated in various concentrations of DIM and labeled with [2-3H]mannose. Viral and cell pellets were digested with Pronase and glycopeptides were isolated by gel filtration on columns of Bio-Gel P-4. The glycopeptides were then treated with endoglucosaminidase H (Endo H) and rechromatographed on the Bio-Gel column in order to distinguish complex from high-mannose structures. As the DIM concentration in the medium was raised, more and more of the [3H]mannose was incorporated into high-mannose oligosaccharides, and less and less radioactivity was in the complex chains. Most of the Endo H-released oligosaccharides induced by DIM were of the Man9GlcNAc structure, as determined by gel filtration, HPLC, and digestion by alpha-mannosidase. Thus, DIM also appears to inhibit mannosidase I in cell culture. However, about 15% of the Endo H-released oligosaccharides appear to be hybrid types of oligosaccharides, suggesting that DIM may also inhibit mannosidase II.     

Mannosyltransferase activity in calf pancreas microsomes. Formation of 14C-labeled lipid-linked oligosaccharides from GDP-D-[14C]mannose and pancreatic dolichyl beta-D-[14C]mannopyranosyl phosphate.

Calf pancreas microsomes incorporated radioactive D-mannose from GDP-D-[14C]mannose into lipid-bound oligosaccharides extracted with chloroform/methanol/water (10/10/2.5, v/v). Several products, which probably differed in the size of the oligosaccharide moiety, were labeled. These could be partially resolved by thin layer chromatography and DEAE-cellulose chromatography. The labeled lipid-bound oligosaccharides were retained on DEAE-cellulose more strongly than synthetic dolichyl alpha-D-[14C]mannopyranosyl phosphate. They were stable to mild alkali, but labile to acid and hot alkali. Acid treatment yielded a neutral 14C-labeled oligosaccharide fraction which was estimated by gel filtration to have a minimum of 8 monosaccharide residues. Hot alkali treatment yielded a mixture of neutral and acidic 14C-labeled oligosaccharides which could be transformed into neutral products by alkaline phosphatase. The D-[14C]mannose residues were alpha-linked at the nonreducing terminus of the oligosaccharides since they could be removed completely with alpha-mannosidase. Most of the D-[14C]mannose-labeled oligosaccharides were retained on concanavalin A Sepharose and eluted with methyl alpha-D-mannopyranoside. Pancreatic dolichyl beta-D-[14C]mannopyranosyl phosphate incubated with calf pancreas microsomes in the presence of sodium taurocholate was efficiently utilized as donor of alpha-D-mannosyl residues in lipid-bound oligosaccharides. The products formed from dolichyl beta-D-[14C]mannopyranosyl phosphate were identical with those formed from GDP-D-[14C]mannose, and evidence was obtained to show that the dolichyl beta-D-[14C]mannopyranosyl phosphate was serving as donor without prior conversion to GDP-D-[14C]mannose. Transfer of mannose from dolichyl beta-D-[14C]mannopyranosyl phosphate to lipid-bound oligosaccharides took place at a pH optimum of 7.3, whereas transfer to the precipitate containing glycoproteins was greatest at pH 6.0 in Tris/maleate buffer. The addition of divalent cation was not required, but low concentrations of EDTA were extremely inhibitory. The carbohydrate composition of the lipid-bound oligosaccharides of microsomal membranes was investigated by gas-liquid chromatography and by reduction with sodium borotritide. A heterogeneous mixture of oligosaccharides containing N-acetyl-D-glucosamine, D-mannose, and D-glucose varying in proportions from approximately 1/2.5/0.5 to 1/5/1.5 was obtained with glucosamine at the reducing end. Acid treatment of the lipid-bound oligosaccharide fraction yielded dolichyl pyrophosphate, suggesting that at least some of the oligosaccharides were linked to dolichol through a pyrophosphate group.     

Characterization of lipid-linked octa-, nona-, and decasaccharides formed during in vitro synthesis of mammary glycoproteins

The lipid-linked octa-, nona-, and decasaccharides, isolated from incubations of a membrane preparation from the lactating bovine mammary tissue with GDP-[14C]mannose and UDP-N-acetylglucosamine were subjected to mild acid hydrolysis and purified extensively by repeated gel filtration and paper chromatography. Structural characterization of the oligosaccharides containing six to eight mannose residues linked to an N,N'-diacetylchitobiose unit utilizing digestions with alpha-mannosidase, beta-mannosidase, endo-beta-N-acetylglucosaminidase, D, H, and L, acetolysis, and methylation analysis revealed the presence of several isomers within each size species. Supplementation of the incubations with 0.1 mM dolichol phosphate reduces the number of isomers within these oligosaccharides; the predominant isomers of saccharides from these incubations appear to be similar to the saccharides isolated from in vivo preparations of Chinese hamster ovary cells.     

Structures of the asparagine-linked sugar chains of human chorionic gonadotropin.

The asparagine-linked sugar chains of human chorionic gonadotropin were released from the polypeptide moiety by hydrazinolysis followed by N-acetylation and NaB3H4 reduction. More than 90% of the released radioactive oligosaccharides contained N-acetylneuraminic acid residues. After removal of N-acetylneuraminic acid residues by sialidase treatment, two neutral oligosaccharide fractions were obtained by paper chromatography. Sequential exoglycosidase digestion revealed that one of them was a mixture of two neutral oligosaccharides. The complete structures of the three oligosaccharides were elucidated by methylation analysis. It was confirmed that all the N-acetylneuraminic acid residues of the asparagine-linked sugar chains of human chorionic gonadotropin occur as NeuAc alpha 2 leads to 3Gal groupings by comparing the methylation analysis data for the acidic oligosaccharide mixture before and after sialidase treatment. Based on these results, the structures of the asparagine-linked sugar chains of human chorionic gonadotropin were confirmed to be +/- NeuAc alpha 2 leads to 3Gal beta 1 leads to 4GlcNAc beta 1 leads to 2Man alpha 1 leads to 6(NeuAc alpha 2 leads to 3Gal beta 1 leads to 4GlcNAc beta 1 leads to 2Man alpha 1 leads to 3)Man beta 1 leads to 4GlcNAc beta 1 leads to 4(+/- Fuc alpha 1 leads to 6)GlcNAc and Man alpha 1 leads to 6(NeuAc alpha 2 leads to 3 Gal beta 1 leads to 4 GlcNAc beta 1 leads to Man alpha 1 leads to 3)Man beta 1 leads to 4 GlcNAc beta 1 leads to 4GlcNAc.     

Carbohydrate structures of HVJ (Sendai virus) glycoproteins

The carbohydrate structures of two membrane glycoproteins (HANA protein and F protein) of HVJ have been determined on materials purified from virions grown in the allantoic sac of embryonated chicken eggs. Both glycoproteins contain fucose, mannose, galactose, and glucosamine but not galactosamine, indicating that their sugar chains are exclusively of the asparagine-linked type. The radioactive oligosaccharide fractions obtained from the two glycoproteins by hydrazinolysis followed by NaB[3H]4 reduction gave quite distinct fractionation patterns after paper electrophoresis. More than 75% of the oligosaccharides from F protein were acidic and separated into at least four components by paper electrophoresis. Only 18% of the oligosaccharide from HANA protein was an acidic single component. These acidic oligosaccharides could not be converted to neutral oligosaccharides by sialidase digestion. Structural studies of the neutral oligosaccharide fractions from HANA and F proteins revealed that both of them are mixtures of a series of high mannose type oligosaccharides and of complex type oligosaccharides with Gal beta 1 leads to (Fuc alpha 1 leads to 3) GlcNAc group in their outer chain moieties.     

Comparative study of the oligosaccharides released from baby hamster kidney cells and their polyoma transformant by hydrazinolysis

The asparagine-linked sugar chains of the membrane of baby hamster kidney cells and their polyoma transformant were quantitatively released as oligosaccharides by hydrazinolysis and labeled by NaB3H4 reduction. The radioactive oligosaccharides thus obtained were fractionated by paper electrophoresis. The neutral oligosaccharides of both cells were exclusively of high mannose type. The acidic oligosaccharides were bi-, tri-, and tetraantennary complex-type sugar chains with Man alpha 1----6 (Man alpha 1----3) Man beta 1----4 GlcNAc beta 1----4 (+/- Fuc alpha 1----6) GlcNAc as their cores and Gal beta 1----4 GlcNAc and various lengths of Gal beta 1----4 GlcNAc repeating chains in their outer-chain moieties. Prominent features of these acidic oligosaccharides are that all sialic acid residues were N-acetylneuraminic acid and were linked exclusively at C-3 of the nonreducing terminal galactose residues of the outer chains. Comparative study of oligosaccharides of the two cells by Bio-Gel P-4 column chromatography revealed that transformation of baby hamster kidney cells leads to a reduction in high mannose-type oligosaccharides and an increase in tetraantennary oligosaccharides. Increase of the outer chains linked at C-6 of the Man alpha 1----6 residue of the core is the cause of increase in the relative amount of highly branched oligosaccharides in the polyoma transformant.