Session: 17-15-01: Society-Wide Micro/Nano Poster Forum

Paper Number: 99132

99132 - A Steered Molecular Dynamics Study of Unbinding of Human Cell Receptors From Sars Cov-2 Viral Protein Receptor Binding Site 

The unbinding of several potential ligands in terms of human body receptors with the SARS CoV-2 spike protein receptor binding domain (RBD) have been investigated using external stimulus as a harmonic force. These receptor molecules are chosen based on their half-maximal inhibitory concentration (IC50) values. First, we performed molecular docking to predict the ligand binding configurations and their corresponding binding energy with the spike protein RBD, which is a critical first step in the early stage structure-based drug discovery. As an additional screening, we use Machine Learning algorithms to predict potential inhibitors to the spike protein RBD. Based on the results from these two complementary tools, the most potential inhibitors are further considered to perform steered molecular dynamics (SMD) simulations for a period of 600 picoseconds to calculate the mean unbinding force and center of mass displacement of corresponding ligand molecules. Umbrella sampling technique is adopted to generate several sample trajectory configurations after the SMD simulations to predict the potential of mean force (PMF) along the reaction co-ordinates with a specified constant co-ordinate spacing. We have applied weighted histogram analysis method (WHAM) to calculate the PMF and this is being used to develop the calculations for the binding free energies (ΔG_binding) of each of the ligand molecules. Our results predicted that N-acetyl neuraminic acid (Sialic acid) shows the best binding conformational strength in terms of staying bounded with the spike protein RBD for more than 50-100 ps than rest of the other four ligand receptors during SMD simulations and the pulling force required to unbind it completely by breaking all short- and long-range interactions with spike protein RBD is also the highest. The temporal change of corresponding protein-ligand short range coulombic and Lennard-Jones (LJ) interaction energies, root mean square deviations of protein and ligand with respect to their initial configurations, calculations for hydrogen bond formation and results from the binding free energy calculation suitably agree with the earlier findings. Parallel studies utilizing RBD-ligand binding inhibition assays similarly observed a low IC₅₀ (ligand concentration at 50% binding inhibition) value for sialic acid (19 nM). Based on insights from the SMD studies, a set of test ligands with similar structure to sialic acid was tested: identifying drug candidate molecules with a further order of magnitude reduction in IC₅₀. The results of this work can be very useful in developing a thorough understanding of how effectively these ligands can be used as a potential drug candidate by restricting the active site of the spike protein of SARS CoV-2 to minimize its human cell inhibition capabilities and also this can lead the way the determining the binding-unbinding characteristics for other potential spike protein inhibitors in the future.

Presenting Author: Tanumoy Banerjee Lehigh University

Presenting Author Biography: I am a Graduate Research Assistant from Spring 2021 at Lehigh University.

Authors:

Tanumoy Banerjee Lehigh University
Agnivo Gosai Corning Inc.
Niloofar Yousefi University of Central Florida
Craig Neal University of Central Florida
Elyaraja Kolnathai University of Central Florida
Ozlem Garibay University of Central Florida
Sudipta Seal University of Central Florida
Ganesh Balasubramanian Lehigh University