Molecular mechanism of decision-making in glycosaminoglycan biosynthesis

Two major glycosaminoglycan types, heparan sulfate (HS) and chondroitin sulfate (CS), control many aspects of development and physiology in a type-specific manner. HS and CS are attached to core proteins via a common linker tetrasaccharide, but differ in their polymer backbones. How core proteins are specifically modified with HS or CS has been an enduring mystery. By reconstituting glycosaminoglycan biosynthesis in vitro, we establish that the CS-initiating N-acetylgalactosaminyltransferase CSGALNACT2 modifies all glycopeptide substrates equally, whereas the HS-initiating N-acetylglucosaminyltransferase EXTL3 is selective. Structure-function analysis reveals that acidic residues in the glycopeptide substrate and a basic exosite in EXTL3 are critical for specifying HS biosynthesis. Linker phosphorylation by the xylose kinase FAM20B accelerates linker synthesis and initiation of both HS and CS, but has no effect on the subsequent polymerisation of the backbone. Our results demonstrate that modification with CS occurs by default and must be overridden by EXTL3 to produce HS.

Automated summary

Heparan sulfate (HS) and chondroitin sulfate (CS) control development and physiology in a specific manner. The modification of core proteins with HS or CS is carried out selectively by N-acetylglucosaminyltransferase EXTL3 and equally by N-acetylgalactosaminyltransferase CSGALNACT2, respectively. Phosphorylation accelerates linker synthesis and the initiation of HS and CS.