Identification of Polysialyltransferase Sequences Required for Substrate Recognition and Polysialylation

ST8Sia II (STX) and ST8Sia IV (PST) represent the two human polysialyltransferases (polySTs) that synthesize polysialic acid (polySia) on a limited number of glycoproteins. Substrate polysialylation has been shown to be both beneficial and detrimental, as this process is necessary for proper development of the mammalian nervous system, has roles in memory formation, olfaction, and neuron repair, but also aberrantly occurs in several human tumors where polySia is believed to promote their metastatic potential. Despite this knowledge, the precise mechanism by which the polySTs recognize and polysialylate their substrates remains poorly understood. The aim of this work is to identify sequences within the polySTs that are responsible for effective substrate recognition and subsequent polysialylation, thus contributing to our knowledge of the mechanism of polysialylation. Initially focusing on the polysialylation of neural cell adhesion molecule (NCAM), the major polysialylated substrate, we determine, through competition and binding analyses utilizing a catalytically inactive full-length PST (PST H331K) and a series of truncated PST mutants, that residues 71-127 of PST contain sequences critical for NCAM recognition. To identify residues necessary for substrate recognition in PST and STX, as well as compare criteria across polysialylated substrates, we examined the role of a polybasic region (PBR, PST residues 71-105 and STX residues 86-120) in the recognition of NCAM, synaptic cell adhesion molecule 1, and neuropilin-2. Focusing specifically on the PBR basic residues, each was mutated individually to alanine in the wild-type PST/STX or catalytically inactive PST H331K/STX H346K background for substrate polysialylation and competition/recognition studies, respectively. We find that in each polyST-substrate pairing examined, basic residues within the polyST PBR are required for effective substrate recognition. Interestingly, we also find that recognition of each substrate by a polyST requires a set of overlapping PBR basic residues unique to the polyST-substrate pairing. Cumulatively, these data provide evidence that the polyST PBR is directly involved in substrate recognition and thus is required for effective substrate polysialylation. Furthermore, these results provide compelling evidence that substrate polysialylation is a highly protein-specific process and suggests that the polyST-substrate interaction may serve as a target for therapeutic interventions.