beta1-4Galactosyltransferase activity of mouse brain as revealed by analysis of brain-specific complex-type N-linked sugar chains.

We previously reported two brain-specific agalactobiantennary N-linked sugar chains with bisecting GlcNAc and alpha1-6Fuc residues, (GlcNAcbeta1-2)(0)(or)(1)Manalpha1-3(GlcNAcbeta1-2M analpha1-6)(GlcNA cbeta1-4)Manbeta1-4GlcNAcbeta1-4(Fucalpha1-6)Glc NAc [Shimizu, H., Ochiai, K., Ikenaka, K., Mikoshiba, K., and Hase, S. (1993) J. Biochem. 114, 334-338]. Here, the reason for the absence of Gal on the sugar chains was analyzed through the detection of other complex type sugar chains. Analysis of N-linked sugar chains revealed the absence of Sia-Gal and Gal on the GlcNAc residues of brain-specific agalactobiantennary N-linked sugar chains. We therefore investigated the substrate specificity of galactosyltransferase activities in brain using pyridylamino derivatives of agalactobiantennary sugar chains with structural variations in the bisecting GlcNAc and alpha1-6Fuc residues as acceptor substrates. While the beta1-4galactosyltransferases in liver and kidney could utilize all four oligosaccharides as substrates, the beta1-4galactosyltransferase(s) in brain could not utilize the agalactobiantennary sugar chain with both bisecting GlcNAc and Fuc residues, but could utilize the other three acceptors. Similar results were obtained using glycopeptides with agalactobiantennary sugar chains and bisecting GlcNAc and alpha1-6Fuc residues as substrates. The beta1-4galactosyltransferase activity of adult mouse brain thus appears to be responsible for producing the brain-specific sugar chains and to be different from beta1-4galactosyltransferase-I. The agalactobiantennary sugar chain with bisecting GlcNAc and alpha1-6Fuc residues acts as an inhibitor against "brain type" beta1-4galactosyltransferase with a K(i) value of 0.29 mM.

Systematic analysis of N-linked sugar chains from whole tissue employing partial automation.

A partially automated technique for the isolation and characterization of N-linked sugar chains from glycoproteins of crude tissue samples is established. The N-linked sugar chains from the acetone-extracted tissues are made free by a process of hydrazinolysis and subsequently N-acetylated by GlycoPrep 1000 (Oxford Glycosystems). These free sugar chains are further converted to pyridylamino derivatives by GlycoTag (Takara). Characterization of these sugar chains is achieved by a combination of HPLC columns using a highly sensitive fluorescence detector at femtomole levels. Tissue sample can be successfully pyridylaminated and analyzed to give highly reproducible results with consistent yield, requiring fewer purification steps, minimum skills, and less time. Moreover, fixed tissues can also be analyzed employing this technique, giving a similar sugar chain pattern compared to normal tissue samples. Using this method we show that the pattern of N-linked sugar chains present in human sera or in one small region of brain is strikingly similar among the different individuals. However, the absence of a highlighted peak in one of the samples suggests this method can be extrapolated to identify changes, if any, associated with disorders such as inflammation or cancer. Furthermore, this two-dimensional display of sugar chains would discover the function-specific molecules as we see in proteins.