It is estimated that about 50% of proteins are glycosylated. However, not all sequons Asn-aa-Ser/Thr are glycosylated and it is not clear how to differentiate between occupied and unoccupied sites. Some studies have examined the amino acid sequence in the positions close to the linked Asn. However, 3D structures of glycoproteins complexes can also contribute to a better comprehension of these events.
The Protein Data Bank (PDB) is the largest freely available source of biomolecular 3D structures determined mainly by x-ray crystallography or NMR. Most of entries in the PDB are proteins, but around 7% of total entries are glycoproteins and protein-carbohydrate complexes. In this study, we collect and statistically analyse such entries in order to examine the interactions established between amino acids and N-linked carbohydrate residues. The analysis is carried out by the program GlyVicinity. GlyVicinity is a publicly available tool that examines the frequency with which amino acids are present within a user-defined distance of glycan residues.
Our results show that the presence of amino acids interacting with carbohydrate residues vary depending on monosaccharide type and position in the N-glycan structural core. For instance, aromatic and polar uncharged amino acids are overrepresented for β-D-GlcpNAc residues. However, the results vary depending on whether β-D-GlcpNAc is positioned in the first or second place in the N-glycan structural core. In the first case, not only the amino acids Ser and Trp are overrepresented (as expected), but also Trp and Asn, despite that the first Asn was not considered. This pattern changes for the β-D-GlcpNAc in the second position, where Tyr prevails and the presence of Thr and Ser becomes underrepresented.
GlyVicinity is freely available at http://www.glycosciences.de/tools/glyvicinity/
