Carbohydrate-deficient glycoprotein syndrome type I: determination of the oligosaccharide structure of newly synthesized glycoproteins by analysis of calnexin binding.

Abstract

Carbohydrate-deficient glycoprotein syndromes (CDGS) are multisystemic disorders with severe mental and statomotor retardation (Jaeken 1991 ; Jaeken and Carchon 1993). They are based on incorrect glycosylation of many different glycoproteins (Harrison et al 1992). In type I, the most common form, incorrect glycosylation presumably occurs co-translationally in the endoplasmic reticulum (ER) (Yamashita et al 1993). In the ER, a so-called core oligosaccharide is transferred by oligosaccharyl transferase (OST) to special asparagine residues of a newly synthesized protein. In CDGS type I, glycosylation consensus sites randomly stay unglycosylated because the transfer of the core oligosaccharide does not occur (Yamashita et al 1994). As a consequence, a given glycoprotein population becomes heterogeneous. Whereas some newly synthesized glycoproteins are fully glycosylated, others have less or no oligosaccharides at all. OST, a multimeric enzyme complex located at the inner ER membrane, is an obvious candidate for the molecular defect in CDGS type I. However, OST activity in microsome preparations from CDGS fibroblasts was found to be normal (Knauer et al 1994). Since recent investigations have shown that the amount of fully assembled lipid precursors of the core oligosaccharide is lower in CDGS fibroblasts, current research is focusing on a possible assembly defect as the cause for CDGS (Powell et al 1994 ; Panneerselvam and Freeze 1995). The core oligosaccharide is a complicated structure of 14 monosaccharides (Figure 1) that is preassembled on dolichol before being transferred by OST. Since OST affinity is much lower for incompletely assembled oligosaccharides, an assembly defect could be the reason why glycosylation sites randomly stay unglycosylated. An assembly defect would cause transfer of truncated core oligosaccharides to newly synthesized glycoproteins. In this study we used calnexin, a resident ER chaperone with highly specific lectin-like properties, to determine the oligosaccharide structure of newly synthesized glycoproteins. The binding of calnexin to newly synthesized glycoproteins is highly selective for a specific terminal oligosaccharide structure. Already co-translationally extensive modifications of the transferred oligosaccharide take place that start place start with the step-by-step removal of the three terminal glucoses by glucosidases I and II (Figure 1). Binding of calnexin to a newly synthesized glycoprotein occurs if the carbohydrate has exactly one terminal glucose, i.e. if the outermost two glucoses have been trimmed (Hammond and Helenius 1995; Hebert et al 1995) (Figure 1). Isolation of carbohydrate assembly mutant has shown that the terminal glucose moities are not attached if assembly of the core oligosaccharide on the lipid precursor is arrested at some point during the synthesis (Chapman et al 1980; Freeze et al 1989). If truncated core oligosaccharides were attached to newly synthesized glycoproteins in CDGS type I, no binding to calnexin should occur.