Carbohydrate-based drug design: Recognition fingerprints and their use in lead identification

Abstract

Carbohydrate-based therapeutics is a rapidly developing theme, resulting directly from the recent increase in interest in the biological roles of carbohydrates. A fundamental requirement for successful drug design is a detailed understanding of protein-carbohydrate interactions. This article reports a structural bioinformatics study of several carbohydrate binding proteins to identify common minimum principles required for the recognition of mannose, glucose and galactose, which indeed form much or the basis for recognition of higher sugars. The study identifies all aspartic acid -04 sugar hydroxyl interaction to be highly conserved, which appears to be crucial for recognition of all three sugars. Other interactions are specific to particular sugars, leading to individual fingerprints. These fingerprints have then been used in the identification of lead compounds, using fragment-based design approaches. The results obtained by such guided design protocols are found to be more focused than those obtained from comparable <i>ab-initio </i>design protocols. These studies, apart from providing clues about the usable pharmacophore space for these structures, also prove that the use of fingerprints in a fragment-based ligand design, leads to the design of a sugar-like ligand, mimicking the natural carbohydrate ligand in each of the eight examples studied.