ST8Sia IV mRNA corresponds with the biosynthesis of alpha2,8sialyl polymers but not oligomers in rat oligodendrocytes.

As oligodendrocytes mature they progress through a series of distinct differentiation steps characterized by the expression of specific markers. One such marker, polysialic acid found on the neural cell adhesion molecule (NCAM), is detected by antibodies and is present on progenitor oligodendrocytes, but is not detected to the same extent on mature oligodendrocytes. Two closely related polysialyltransferases, ST8Sia II (STX) and ST8Sia IV (PST) have been cloned previously and shown to synthesize polysialic acid on NCAM and other glycoproteins. To determine whether or not polyalpha2,8sialyltransferases are downregulated during the differentiation of oligodendrocytes, the enzyme activity and expression of ST8Sia II and ST8Sia IV mRNA at two stages of maturation in JS12/1 and JS3/16 oligodendrocytes were examined. Differentiation in both oligodendroglial cell lines was accompanied by more than a 50% reduction in the biosynthesis of polymers of alpha2,8sialic acid when fetuin was used as substrate. Most interestingly, extracts of JS12/1 mature cells synthesized 60% more short oligomers of alpha2,8sialic acid than the progenitor cells, whereas JS3/16 mature cells synthesized barely detectable amounts of the short oligomers. Transcripts for ST8Sia IV mRNA were present in both JS12/1 and JS3/16 and were reduced when the biosynthesis was markedly reduced. In contrast ST8Sia II mRNA was barely detectable in JS3/16 cells and although detectable in JS12/1 cells, there was no clear modulation with maturation. These results were supported by the examination of the brains of rats from embryonic to Day 21 ages. The enzyme activity and mRNA experiments show that polyalpha2,8sialyltransferase itself is down regulated to cause the reduction in sialyl polymers on mature oligodendrocytes. Moreover, ST8Sia IV is responsible for the polysialylation of NCAM in oligodendrocytes.

Activity of fucosyltransferases and altered glycosylation in cystic fibrosis airway epithelial cells.

Cystic fibrosis (CF) glycoconjugates have a glycosylation phenotype of increased fucosylation and/or decreased sialylation when compared with non-CF. A major increase in fucosyl residues linked alpha 1,3 to antennary GlcNAc was observed when surface membrane glycoproteins of CF airway epithelial cells were compared to those of non-CF airway cells. Importantly, the increase in the fucosyl residues was reversed with transfection of CF cells with wild type CFTR cDNA under conditions which brought about a functional correction of the Cl(-) channel defect in the CF cells. In contrast, examination of fucosyl residues in alpha 1,2 linkage by a specific alpha 1,2 fucosidase showed that cell surface glycoproteins of the non-CF cells had a higher percentage of fucose in alpha 1,2 linkage than the CF cells. Airway epithelial cells in primary culture had a similar reciprocal relationship of alpha 1,2- and alpha 1,3-fucosylation when CF and non-CF surface membrane glycoconjugates were compared. In striking contrast, the enzyme activity and the mRNA of alpha 1,2 fucosyltransferase did not reflect the difference in glycoconjugates observed between the CF and non-CF cells. We hypothesize that mutated CFTR may cause faulty compartmentalization in the Golgi so that the nascent glycoproteins encounter alpha 1,3FucT before either the sialyl- or alpha 1,2 fucosyltransferases. In subsequent compartments, little or no terminal glycosylation can take place since the sialyl- or alpha 1,2 fucosyltransferases are unable to utilize a substrate, which is fucosylated in alpha 1,3 position on antennary GlcNAc. This hypothesis, if proven correct, could account for the CF glycophenotype.

Purification of an alpha-2,8-sialyltransferase, a potential initiating enzyme for the biosynthesis of polysialic acid in human neuroblastoma cells.

alpha-2,8-Sialyltransferase has been purified from human neuroblastoma CHP-134 cells greater than 2900-fold. The key step in the purification was a substrate affinity column utilizing immobilized colominic acid. Several kinetic parameters of the enzyme were defined. Fetuin but not asialofetuin served as substrate. The product of the enzyme reaction was characterized as containing sialyl residues in alpha-2,8-linkage with the use of recombinant sialidases. It is suggested that the purified enzyme is an initiating enzyme for the biosynthesis of polysialic acid since these cells also have the activity of poly alpha-2,8-sialyltransferase and contain polysialic acid. This alpha-2,8-sialyltransferase may be a new member of a family of alpha-2,8-sialyltransferases recently described, since it differs in substrate specificity reported for the cloned and expressed enzymes.

Expression of GDP-L-Fuc: Gal(beta 1-4)GlcNAc-R (Fuc to GlcNAc) alpha-1,3-fucosyltransferase and its relationship to glycoprotein structure in a human erythroleukemia cell line, HEL.

Terminal glycosylation may be a mechanism to control the function of specific biologically active glycoproteins. The biosynthesis of terminal sialyl and fucosyl residues on certain glycoproteins has been linked to the expression of the respective glycosyltransferase. In contrast, a human erythroleukemia cell line, HEL, contained a highly active GDP-L-Fuc: Gal(beta 1-4)GlcNAc-R (Fuc to GlcNAc) alpha-1,3-fucosyltransferase (alpha-1,3-fucosyltransferase) but no detectable alpha-1,3-linked fucosyl residues on the glycoproteins. The alpha-1,3-fucosyltransferase gave apparent Km values for Fuc(alpha 1-2)Gal(beta 1-4)GlcNAc beta-O-benzyl, Gal(beta 1-4)GlcNAc and GDP-fucose of 0.04, 0.68 and 0.12 mM, respectively. The lack of detectable fucosyl residues in alpha-1,3-linkage to GlcNAc on the [3H]fucose-labeled glycoproteins was shown with the use of almond alpha-1,3/4-fucosidase and internal controls to verify that the enzyme was active. Using Western-blot analysis, HEL cell glycoproteins reacted with blood group H type-2 antibody, confirming the presence of Fuc(alpha 1-2)Gal(beta 1-4)GlcNAc as reported by others and the presence of the preferred substrate for the enzyme. It is proposed that controls for terminal glycosylation in addition to glycosyltransferase expression are operative in HEL cells and that they are part of a multi-regulated process controlling terminal modifications of glycoproteins.

Biosynthesis of lipid-linked oligosaccharides by microsomes from virus induced Hepatoma MC-29. Dependence of some glycosyl transferases on dolichyl phosphates of different chain length.

Dolichyl phosphates of various chain length ranging from 7 to 22 isoprene units were tested as lipid acceptors in transglycosylation reactions in chicken liver and Hepatoma MC-29. In the presence of exogenous dolichyl phosphate mixture (18 and 19 isoprene units) the synthesis of dolichyl pyrophosphate N-acetylglucosamine and dolichyl phosphate mannose increased 3 times both in the liver and Hepatoma MC-29, while the formation of dolichyl phosphate glucose was 4 fold higher in the liver and 6-fold higher in Hepatoma MC-29. In liver microsomes the maximum rate of the stimulation of glycosylation was achieved by exogenous dolichyl phosphates, containing 18 and 19 isoprene units, while glycosyl transferases in microsomes from Hepatoma MC-29 did not show any structural requirements to the chain length of dolichyl phosphates.