(712j) Graph Network Analysis of Protein-Osmolyte Preferential Interactions | AIChE

(712j) Graph Network Analysis of Protein-Osmolyte Preferential Interactions

Osmolytes are known to influence protein conformation and solubility. Some like sorbitol and trehalose stabilize the native protein conformation, while others like urea shift the conformation equilibrium towards unfolded state. The exact molecular mechanism of action of osmolytes is, however, not yet completely understood. The alteration of water structure by osmolytes and/or the direct interaction of the osmolytes with protein are postulated as the plausible mechanisms. The direct mechanism suggests that the difference in the competing interaction of water and osmolytes between the folded and unfolded protein conformations determines the conformational changes in presence of osmolytes. These preferential interactions are fairly difficult to measure directly; it can be computed from molecular simulations through the local spatial distribution of the water and osmolytes molecules. Such computations, however, miss out the difference in the molecular size and the associated correlation length scale between water and osmolytes. Consequently the large osmolytes like trehalose are identified to preferentially exclude from the protein surface undermining the direct protein-osmolyte interactions. Our key objective is to identify a uniform metric to compare protein-water and protein-osmolyte interactions independent of the difference in the molecular sizes. In this context, we employed graph network analysis, routinely used for analyzing social interaction network and transport, to understand how interaction in protein solutions gets modified in presence of osmolytes. This provides the complete solution level picture instead of local vicinal distributions. From the molecular simulation trajectory, we construct the interaction graphs based on hydrogen bonding and spatial distribution; the extent of interactions is analyzed through various graph measures. In this talk, we will compare our network results against a random network providing a new perception on preferential interactions and its implication on protein conformation.