The human natural killer-1 (HNK-1) carbohydrate comprising a sulfated trisaccharide (HSO3-3GlcAβ1-3Galβ1-4GlcNAc-) is expressed on N-linked and O-mannose-linked glycans in the nervous system and involved in learning and memory functions. Although whole/core glycan structures and carrier glycoproteins for the N-linked HNK-1 epitope have been studied, carrier glycoproteins and the biosynthetic pathway of the O-mannose-linked HNK-1 epitope have not been fully characterized. Here, using mass spectrometric analyses, we identified the major carrier glycoprotein of the O-linked HNK-1 as phosphacan in developing mouse brains and determined the major O-glycan structures having the terminal HNK-1 epitope from partially purified phosphacan. The O-linked HNK-1 epitope on phosphacan almost disappeared due to the knockout of protein O-mannose β1,2-N-acetylglucosaminyltransferase 1, an N-acetylglucosaminyltransferase essential for O-mannose-linked glycan synthesis, indicating that the reducing terminal of the O-linked HNK-1 is mannose. We also showed that glucuronyltransferase-P (GlcAT-P) was involved in the biosynthesis of O-mannose-linked HNK-1 using the gene-deficient mice of GlcAT-P, one of the glucuronyltransferases for HNK-1 synthesis. Consistent with this result, we revealed that GlcAT-P specifically synthesized O-linked HNK-1 onto phosphacan using cultured cells. Furthermore, we characterized the as-yet-unknown epitope of the 6B4 monoclonal antibody (mAb), which was thought to recognize a unique phosphacan glycoform. The reactivity of the 6B4 mAb almost completely disappeared in GlcAT-P-deficient mice, and exogenously expressed phosphacan was selectively recognized by the 6B4 mAb when co-expressed with GlcAT-P, suggesting that the 6B4 mAb preferentially recognizes O-mannose-linked HNK-1 on phosphacan. This is the first study to show that 6B4 mAb-reactive O-mannose-linked HNK-1 in the brain is mainly carried by phosphacan.

Human natural killer-1 (HNK-1) carbohydrate (HSO3-3GlcAβ1-3Galβ1-4GlcNAc-R) is highly expressed in the brain and required for learning and neural plasticity. We previously demonstrated that expression of the HNK-1 epitope is mostly abolished in knockout mice for GlcAT-P (B3gat1), a major glucuronyltransferase required for HNK-1 biosynthesis, but remained in specific regions such as perineuronal nets (PNNs) in these mutant mice. Considering PNNs are mainly composed of chondroitin sulfate proteoglycans (CSPGs) and regulate neural plasticity, GlcAT-P-independent expression of HNK-1 in PNNs is suggested to play a role in neural plasticity. However, the function, structure, carrier glycoprotein and biosynthetic pathway for GlcAT-P-irrelevant HNK-1 epitope remain unclear. In this study, we identified a unique HNK-1 structure on aggrecan in PNNs. To determine the biosynthetic pathway for the novel HNK-1, we generated knockout mice for GlcAT-S (B3gat2), the other glucuronyltransferase required for HNK-1 biosynthesis. However, GlcAT-P and GlcAT-S double-knockout mice did not exhibit reduced HNK-1 expression compared with single GlcAT-P-knockout mice, indicating an unusual biosynthetic pathway for the HNK-1 epitope in PNNs. Aggrecan was purified from cultured cells in which GlcAT-P and -S are not expressed and we determined the structure of the novel HNK-1 epitope using liquid chromatography/mass spectrometry (LC/MS) as a sulfated linkage region of glycosaminoglycans (GAGs), HSO3-GlcA-Gal-Gal-Xyl-R. Taken together, we propose a hypothetical model where GlcAT-I, the sole glucuronyltransferase required for synthesis of the GAG linkage, is also responsible for biosynthesis of the novel HNK-1 on aggrecan. These results could lead to discovery of new roles of the HNK-1 epitope in neural plasticity.

No abstract