Tissue-Specific Regulation of HNK-1 Biosynthesis by Bisecting GlcNAc.

Human natural killer-1 (HNK-1) is a sulfated glyco-epitope regulating cell adhesion and synaptic functions. HNK-1 and its non-sulfated forms, which are specifically expressed in the brain and the kidney, respectively, are distinctly biosynthesized by two homologous glycosyltransferases: GlcAT-P in the brain and GlcAT-S in the kidney. However, it is largely unclear how the activity of these isozymes is regulated in vivo. We recently found that bisecting GlcNAc, a branching sugar in N-glycan, suppresses both GlcAT-P activity and HNK-1 expression in the brain. Here, we observed that the expression of non-sulfated HNK-1 in the kidney is unexpectedly unaltered in mutant mice lacking bisecting GlcNAc. This suggests that the biosynthesis of HNK-1 in the brain and the kidney are differentially regulated by bisecting GlcNAc. Mechanistically, in vitro activity assays demonstrated that bisecting GlcNAc inhibits the activity of GlcAT-P but not that of GlcAT-S. Furthermore, molecular dynamics simulation showed that GlcAT-P binds poorly to bisected N-glycan substrates, whereas GlcAT-S binds similarly to bisected and non-bisected N-glycans. These findings revealed the difference of the highly homologous isozymes for HNK-1 synthesis, highlighting the novel mechanism of the tissue-specific regulation of HNK-1 synthesis by bisecting GlcNAc.

The K346T mutant of GnT-III bearing weak in vitro and potent intracellular activity.

BACKGROUND: N-Acetylglucosaminyltransferase-III (GnT-III, also designated MGAT3) catalyzes the formation of a specific N-glycan branch, bisecting GlcNAc, in the Golgi apparatus. Bisecting GlcNAc is a key residue that suppresses N-glycan maturation and is associated with the pathogenesis of cancer and Alzheimer's disease. However, it remains unclear how GnT-III recognizes its substrates and how GnT-III activity is regulated in cells. METHODS: Using AlphaFold2 and structural comparisons, we predicted the key amino acid residues in GnT-III that interact with substrates in the catalytic pocket. We also performed in vitro activity assay, lectin blotting analysis and N-glycomic analysis using point mutants to assess their activity. RESULTS: Our data suggested that E320 of human GnT-III is the catalytic center. More interestingly, we found a unique mutant, K346T, that exhibited lower in vitro activity and higher intracellular activity than wild-type GnT-III. The enzyme assays using various substrates showed that the substrate specificity of K346T was unchanged, whereas cycloheximide chase experiments revealed that the K346T mutant has a slightly shorter half-life, suggesting that the mutant is unstable possibly due to a partial misfolding. Furthermore, TurboID-based proximity labeling showed that the localization of the K346T mutant is shifted slightly to the cis side of the Golgi, probably allowing for prior action to competing galactosyltransferases. CONCLUSIONS: The slight difference in K346T localization may be responsible for the higher biosynthetic activity despite the reduced activity. GENERAL SIGNIFICANCE: Our findings underscore the importance of fine intra-Golgi localization and reaction orders of glycosyltransferases for the biosynthesis of complex glycan structures in cells.