Computational Study of Polyligand Complexes of Aspirin with Human Serum Albumin Using Docking and Molecular Dynamics Methods

An important feature of biochemical reactions of molecules is the possibility of binding of several ligands to a biomolecular target. This effect should be taken into account in the analysis of protein–ligand interactions and in the estimates of drug distribution in the living systems. This work describes molecular simulations of successive parallel steps of binding of two aspirin (As) molecules at the known binding sites 1–3 of human serum albumin with differing affinity. The experimental data on multiple binding of aspirin to albumin is inconclusive. Docking of aspirin anion As– to albumin predicts that stability of the complexes at the binding sites changes as 1 > 3 > 2. Molecular dynamics simulations have further shown that the complexes at site 3 are unstable. The free energies of ligand binding ΔGb have been calculated using extended linear interaction energies method with additional contributions of the entropy of ligand binding. The results have shown that the most probable reaction path corresponds to binding of As– at site 1 with ΔGb1= ‒8.2 kcal·mol–1 and after that to site 2 with ΔGb2= ‒4.5 kcal·mol–1. The calculated values of ΔGb agree with the known experimental data. The stoichiometry of the albumin–As–complexes is 2. Negative cooperative effect is found for binding of two As– molecules with albumin. The used molecular model and computational approaches can be further employed in the studies of binding of different medicinal molecules that are transported by serum albumin.

protein–ligand complexes, binding constant, serum albumin, aspirin, molecular dynamics