Modeling interactions between Heparan sulfate and proteins based on the Heparan sulfate microarray analysis.

Area of Science:

• Glycobiology
• Computational Chemistry
• Biochemistry

Background:

• Heparan sulfate (HS) is a crucial extracellular matrix polysaccharide involved in numerous biological processes through protein interactions.
• Understanding HS oligosaccharide binding selectivity is vital but challenging due to the vast number of possible structures.
• Current methods for analyzing HS-protein interactions are often limited in scope and practicality.

Purpose of the Study:

• To develop a computational pipeline for designing HS oligosaccharides with targeted protein affinity.
• To create predictive models for HS oligosaccharide-protein binding using Quantitative Structure-Activity Relationship (QSAR) analysis.
• To accelerate the discovery of HS oligosaccharides with specific binding properties for therapeutic and research applications.

Main Methods:

• A hybrid approach integrating microarray experiments and in silico modeling was employed.
• Two distinct structural representations of HS oligosaccharides were developed for modeling.
• Random Forest (RF) algorithm was used to build predictive QSAR models for fibroblast growth factor 2 (FGF2) affinity.

Main Results:

• The developed QSAR models demonstrated high predictivity for FGF2 affinity, with correct classification rates of 0.81-0.80 and positive predictive values up to 0.95.
• Virtual screening identified 15 potential high-affinity HS oligosaccharides, 11 of which were experimentally validated.
• The simplified disaccharide-based representation proved effective for predictive modeling.

Conclusions:

• The hybrid microarray and in silico pipeline enables the targeted design of HS oligosaccharides with specific protein interactions.
• This approach significantly advances the practical design of functional glycans.
• The findings provide a foundation for broader applications in glycobiology and drug discovery.