Rational Design And Development Of Polysialic Acid-Binding Peptides

Polysialic acid (PSA) is a unique post-translational modification found on the neural cell adhesion molecule (NCAM) and plays an important physiological role in nervous system development through modulation of neuronal interactions. While the size and hydrophilicity of PSA result in a large excluded volume that serves its non-adhesive properties well, a number of proteins act as specific ligands to PSA, including Siglec-11 – a sialic-acid specific lectin that recognizes alpha-2,8-linked N-acetylneuraminic acid. Understanding of biomolecular interactions involving PSA may be applied in the targeting of PSA in biotechnological applications, as for the development of ligands, such as peptides, to serve as biosensors, modulators in infections caused by PSA-expressing microbes, or novel affinity agents for neural stem cell targeting. In addition to being small, relatively stable molecules, peptides offer the possibility to study and exploit natural interactions as the basis for rational design. In this work, we demonstrate the development of PSA-binding peptides through rational design, beginning with epitope mapping of Siglec-11 for understanding of PSA-lectin interaction regions and identification of lead candidates. High-throughput synthesis and screening of peptide libraries, with screening in a spatially addressable manner on peptide microarrays, enabled study of peptide-PSA binding interactions. Lead peptides isolated show overlap with the binding pocket of Siglec-11, as evident from molecular dynamics (MD) simulations of PSA with a homology model of the Siglec-11 N-terminal domain. Peptides were further studied through mutational analyses for examination of critical binding residues and for understanding the affinity and selectivity of PSA-peptide interactions. Binding data from microarray studies were corroborated with surface plasmon resonance (SPR) spectroscopy and fluorescence anisotropy assays. Library level data, combined with MD simulations of peptide-PSA interactions, suggest the importance of spatial positioning along with identity of peptide residues. That is, while the electrostatic component of binding is high, as expected, the geometry of residues plays an important role and hence offers potential for development of selectivity. For purposes of comparison to our natural-partners based rational approach, phage display screening of a random peptide library was also conducted and lead peptides were studied through mutational analyses as above. The approach used in development and study of PSA-binding peptides thus contributes to the understanding of carbohydrate-peptide interactions and has implications for small-molecule targeting of specific glycosylation patterns in disease-related applications.