Rosalind Kornfeld

Structure of Glycoproteins and Their Oligosaccharide Units

The presence of oligosaccharide chains covalently attached to the peptide backbone is the feature that distinguishes glycoproteins from other proteins and accounts for some of their characteristic physical and chemical properties. Glycoproteins occur in fungi, green plants, viruses, bacteria, and in higher animal cells where they serve a variety of functions. Connective tissue glycoproteins, such as the collagens and proteoglycans of various animal species, are structural elements as are the cell wall glycoproteins of yeasts and green plants. The submaxillary mucins and the glycoproteins in the mucous secretions of the gastrointestinal tract, which consist of numerous oligosaccharide chains attached at closely spaced intervals to a peptide backbone, serve as lubricants and protective agents. The body fluids of vertebrates are rich in glycoproteins secreted from various glands and organs. Constituents of blood plasma which are glycoproteins include the transport proteins transferrin, ceruloplasmin, and transcobalamin I as well as the immunoglobulins, all the known clotting factors, and many of the components of complement. Follicle-stimulating hormone, luteinizing hormone, and thyroid-stimulating hormone (secreted by the pituitary) and chorionic gonadotropin are all glycoproteins as are the enzymes ribonuclease and deoxyribonuclease (secreted by the pancreas) and α-amylase (secreted by the salivary glands). Fungi secrete a number of glycoprotein enzymes, for example, Taka-amylase and invertase. Another group of glycoproteins are those which occur as integral components of cell membranes in a variety of species. Enveloped viruses contain surface glycoproteins that are involved in the attachment of the virus to its host, and in eukaryotic cells the histocompatibility antigens are membrane glycoproteins. There is a growing body of evidence to suggest that cell surface glycoproteins are involved in a number of physiologically important functions such as cell-cell interaction, adhesion of cells to substratum, and migration of cells to particular organs, for example, the “homing” of lymphocytes to the spleen and the metastasis of tumor cells to preferred sites.

The effect of 1-deoxymannojirimycin on rat liver α-mannosidases

Summary The mannose analogue, 1-deoxymannojirimycin, has been tested for its effect on five α-mannosidase activities present in rat liver and shown to be a specific inhibitor of Golgi α-mannosidase I at low μmolar concentrations. Golgi α-mannosidases I and II were assayed in a highly purified Golgi membrane preparation. Endoplasmic* reticulum α-mannosidase activity was measured in a rough endoplasmic reticulum detergent extract. A purified soluble α-mannosidase activity which we believe is derived from the endoplasmic reticulum during tissue homogenization was also tested. And finally, the lysosomal or acidic α-mannosidase was measured in a postnuclear supernatant fraction obtained from rat liver. The results presented here show that 1-deoxymannojirimycin inhibits only Golgi α-mannosidase I, which is consistent with its effect on oligosaccharide processing in vivo (Fuhrmann et al. Nature 1984 307 :755–758).

Control of synthesis of guanosine 5′-diphosphate d-mannose and guanosine 5′-diphosphate l-fucose in bacteria☆

Abstract The control of the synthesis of GDP- d -mannose and GDP- l -fucose has been studied in various classes of bacteria. In those species which contain l -fucose but no d -mannose in their polysaccharides, GDP- l -fucose inhibits both GDP- d -mannose pyrophosphorylase (GTP: α- d -mannose-1-phosphate guanylyltransferase, EC 2.7.7.13) and GDP- d -mannose hydro-lyase. In the bacteria which have d -mannose but no l -fucose in their polysaccharides, GDP- d -mannose inhibits GDP- d -mannose pyrophosphorylase. Finaly, in bacteria which have both d -mannose and both d -mannose and l -fucose in their polysaccharides, GDP- l -fucose inhibits GDP- d -mannose hydro-lase and GDP- d -mannose inhibits GDP- d -mannose pyrophosphorylase. Thys variations occur in feedback control which allow each of these classes of bacteria to control eindependently the rate of synthesis of the nucleotide sugars which acts as glycosyl donors in the synthesis of polysaccharides.

Molecular cloning and functional expression of two splice forms of human N-acetylglucosamine-1-phosphodiester alpha-N-acetylglucosaminidase.

We have isolated and sequenced human cDNA and mouse genomic DNA clones encodingN-acetylglucosamine-1-phosphodiester α-N-acetylglucosaminidase (phosphodiester α-GlcNAcase) which catalyzes the second step in the synthesis of the mannose 6-phosphate recognition signal on lysosomal enzymes. The gene is organized into 10 exons. The protein sequence encoded by the clones shows 80% identity between human and mouse phosphodiester α-GlcNAcase and no homology to other known proteins. It predicts a type I membrane-spanning glycoprotein of 514 amino acids containing a 24-amino acid signal sequence, a luminal domain of 422 residues with six potential N-linked glycosylation sites, a single 27-residue transmembrane region, and a 41-residue cytoplasmic tail that contains both a tyrosine-based and an NPF internalization motif. Human brain expressed sequence tags lack a 102-base pair region present in human liver cDNA that corresponds to exon 8 in the genomic DNA and probably arises via alternative splicing. COS cells transfected with the human cDNA expressed 50–100-fold increases in phosphodiester α-GlcNAcase activity proving that the cDNA encodes the subunits of the tetrameric enzyme. Transfection with cDNA lacking the 102-base pair region also gave active enzyme. The complete genomic sequence of human phosphodiester α-GlcNAcase was recently deposited in the data base. It showed that our cDNA clone was missing only the 5′-untranslated region and initiator methionine and revealed that the human genomic DNA has the same exon organization as the mouse gene.

Purification and multimeric structure of bovine N-acetylglucosamine-1-phosphodiester alpha-N-acetylglucosaminidase.

N-Acetylglucosamine-1-phosphodiester α-N-Acetylglucosaminidase (EC 3.1.4.45; phosphodiester α-GlcNAcase) catalyzes the second step in the synthesis of the mannose 6-phosphate determinant required for efficient intracellular targeting of newly synthesized lysosomal hydrolases to the lysosome. A partially purified preparation of phosphodiester α-GlcNAcase from bovine pancreas was used to generate a panel of murine monoclonal antibodies. The anti-phosphodiester α-GlcNAcase monoclonal antibody UC1 was coupled to a solid support and used to immunopurify the bovine liver enzyme 670,000-fold in two steps to apparent homogeneity with an overall yield of 14%. The purified phosphodiester α-GlcNAcase has a specific activity of 498 μmol of [3H]GlcNAc-α-phosphomannose-α-methyl cleaved per h per mg of protein using 0.5 mm[3H]GlcNAc-α-phosphomannose-α-methyl as substrate. The subunit structure of the enzyme was determined using a combination of analytical gel filtration chromatography, SDS-polyacrylamide gel electrophoresis, and amino-terminal sequencing. The data indicate that bovine phosphodiester α-GlcNAcase is a 272,000-Da complex of four identical 68,000-Da glycoprotein subunits arranged as two disulfide-linked homodimers. A soluble form of the enzyme, isolated from fetal bovine serum, showed the same subunit structure. Both forms of the enzyme reacted with a rabbit antibody raised to the amino-terminal peptide of the liver enzyme, suggesting that phosphodiester α-GlcNAcase is a type I membrane-spanning glycoprotein with its amino terminus in the lumen of the Golgi apparatus.

Structure of the oligosaccharides of mouse immunoglobulin M secreted by the MOPC 104E plasmacytoma

Abstract The structures of the oligosaccharides present in the mouse immunoglobulin M secreted by the plasma cell tumor MOPC 104E have been determined making use of glycopeptides derived from purified unlabeled immunoglobulin and from [3H]mannose and [14C]glucosaminelabeled immunoglobulin M. The glycopeptides were fractionated on columns of Bio-Gel P-6 and concanavalin A-Sepharose and high mannose type oligosaccharides were released from glycopeptide with clostridial endo-β-N-acetylglucosaminidase CII. MOPC 104E immunoglobulin M was shown to contain four complex-type and one high mannose type oligosaccharide units per heavy chain. The glycopeptides with complex oligosaccharides were separated on concanavalin A-Sepharose into two classes with structures containing either two or three outer branches with the sequence ± NGNAα2 → 6Galβ1 → 4GlcNAcβ1 → attached to the 2 position or to the 2 and 4 positions of mannose in a core with the structure Manαl → 6(Manα1 → 3)Manβ1 → 4GlcNAcβ1 → 4(Fucα1 → 6)GlcNAc → Asn. The high mannose oligosaccharides liberated by endo-β-N-acetylglucosaminidase CII varied in size from Man8GlcNAc to Man5GlcNAc. A Man6GlcN Ac with the structure Manα → 6(Manα1 → 3)Manα1 → 6(Manα1 α 2Manα1 → 3)Manβ1 → 4GlcNAc was the predominant species. Some high mannose oligosaccharide was resistant to endo-β-N-acetylg]ucosaminidase CII release and apparently has a “hybrid” or atypical structure.

Phosphorylase and uridinediphosphoglucose-glycogen transferase in pyridoxine deficiency

Abstract The total phosphorylase activity of the skeletal muscle of rats maintained on a pyridoxine deficient diet has been found to fall to 35% of the normal value. The phosphorylase a activity of the tissue of these rats has the normal value, and the glycogen content of the muscles of such deficient rats is not different from that of control animals. The apparent activity of uridine diphosphoglucose-glycogen transferase is not changed from the normal level in pyridoxine deficiency.

Intracellular site of synthesis of soluble blood group substance

Abstract The intracellular distribution of soluble blood group substance in hog gastric mucosa was investigated. The microsomal fraction contains 1–3% of the total intracellular blood group substance. Labeling experiments in vivo show this small pool to be newly synthesized material and the precursor of other intracellular pools. Attempts to detect a nucleotide-linked, fucose-containing oligosaccharide intermediate in the synthesis of the heterosaccharide chains of blood group substance were unsuccessful.

STRUCTURE OF MEMBRANE RECEPTORS FOR PLANT LECTINS

The plant lectins are a group of proteins that are capable of binding to specific carbohydrate determinants on the surfaces of mammalian cells. Depending on the cell type involved, lectin binding can induce a variety of biologic effects. Thus, lectins can cause resting lymphocytes to undergo blast transformation,’ evoke insulin-like effects in fat cells,2 cause insulin to be released from pancreatic islet cells,3 and simulate the effects of thrombin on platelets.? When one considers these observations, two features emerge. The first is that these biologic effects are not induced by all lectins. For example, the mushroom lectin is not mitogenic toward lymphocytes, yet it is the only lectin so far tested that induces insulin release from islet cells3 The lentil lectin is a lymphocyte mitogen and has insulin-like effects on fat cells, but it does not mimic the action of thrombin on platelets.” Wheat germ agglutinin has insulin-like effects on fat cells, but it is not a mitogen. The second noteworthy feature of these systems is that, in the cases examined up till now, it has been demonstrated that the biologic effect exerted by the lectin results from the binding of the lectin to its cell membrane receptor.5 Therefore, one explanation for the diverse biologic effects of the plant lectins is that they bind to different membrane receptors, and result in distinctive biologic effects. For this reason we have attempted to determine the nature of the cell membrane receptors for a number of plant lectins. The approach we have taken is first to determine the number of binding sites for a given lectin on the cell type of interest, and then to treat the cell with proteolytic enzymes in order to solubilize the receptor. The receptor was then purified and its structure determined.

The B4 lectin from Vicia villosa seeds interacts with N-acetylgalactosamine residues on erythrocytes with blood group Cad specificity

We have previously shown that the B4 lectin from Vicia villosa seeds interacts with N-acetylgalactosamine alpha-linked to serine or threonine in cell surface glycoproteins. In the present study, we show that the lectin also binds to Cad erythrocytes (0.44-2.78 X 10(6) sites/cell) with an association constant of 0.61-0.84 X 10(7)M-1. Variability in the number of B4 lectin binding sites in Cad erythrocytes from different individuals parallels reactivity of these erythrocytes with other N-acetylgalactosamine-binding lectins. Agglutination of Cad erythrocytes with B4 lectin is inhibited by urinary Tamm-Horsfall Sda-active glycoprotein. Since the Cad and Sda determinants share the terminal GalNAc beta 1.4----Gal sequence, our results indicate that Vicia villosa B4 lectin can also interact with terminal beta-linked N-acetylgalactosamine in closely-spaced oligosaccharide units of cell surface glycoproteins.