The Answer Lies in the Energy: How Simple Atomistic Molecular Dynamics Simulations May Hold the Key to Epitope Prediction on the Fully Glycosylated SARS-CoV-2 Spike Protein

Betacoronavirus SARS-CoV-2 is posing a major threat to human health and its diffusion around the world is having dire socioeconomical consequences. Thanks to the scientific community’s unprecedented efforts, the atomic structure of several viral proteins has been promptly resolved. As the crucial mediator of host cell infection, the heavily glycosylated trimeric viral Spike protein (S) has been attracting the most attention and is at the center of efforts to develop antivirals, vaccines, and diagnostic solutions.Herein, we use an energy-decomposition approach to identify antigenic domains and antibody binding sites on the fully glycosylated S protein. Crucially, all that is required by our method are unbiased atomistic molecular dynamics simulations; no prior knowledge of binding properties or ad hoc combinations of parameters/measures extracted from simulations is needed. Our method simply exploits the analysis of energy interactions between all intra-protomer aminoacid and monosaccharide residue pairs, and cross-compares them with structural information (i.e., residueresidue proximity), identifying potential immunogenic regions as those groups of spatially contiguous residues with poor energetic coupling to the rest of the protein.Our results are validated by several experimentally confirmed structures of the S protein in complex with anti- or nanobodies. We identify poorly coupled sub-domains: on the one hand this indicates their role in hosting (several) epitopes, and on the other hand indicates their involvement in large functional conformational transitions. Finally, we detect two distinct behaviors of the glycan shield: glycans with stronger energetic coupling are structurally relevant and protect underlying peptidic epitopes; those with weaker coupling could themselves be poised for antibody recognition. Predicted Immunoreactive regions can be used to develop optimized antigens (recombinant subdomains, synthetic (glyco)peptidomimetics) for therapeutic applications.