Minor Changes with Large Implications: Modeling Amino Acid Mutations in SARS‐CoV Monoclonal Antibodies (80R and 362) Towards the Design of More Universal Antibodies

Manikkuttiyil, Carol; Abbas, Feza; Abbas, Lyla; Alzamora, Carolina; Hunt, Matthew; Julakanti, Pujita; Likki, Sanjana; Manikkuttiyil, Christo; Schmitt Lavin, Emily; Sikora, Arthur

The SARS‐CoV‐2 virus is responsible for the COVID‐19 pandemic which continues to impact nearly every person on Earth, having caused over 1.8 million deaths. Two anti‐SARS‐CoV monoclonal antibodies (MAbs) 80R and 362 are known to bind to epitopes on the spike protein receptor‐binding domain (RBD) and neutralize the virus. To investigate this further and hypothesize structures for potentially more effective antibodies, undergraduate students cooperated in teams as part of the CREST (Connecting Researchers, Educators, and STudents) Program with the Center for Biological Modeling. Working collaboratively, students from eight universities nationwide applied their knowledge to build 3‐D printed models to explain a particular protein‐based molecular story using crystal structures of proteins described in the literature. The Nova Southeastern University (NSU) CREST team modeled and compared the 80R antibody that binds to SARS‐CoV‐1 and the MAb362 antibody that can bind to both SARS‐CoV‐1 and SARS‐CoV‐2. Students developed skills with protein visualization software including Pymol and Jmol to design models which showed the 80R and 362 antibodies binding to the RBD of the corresponding proteins. By studying the point mutation differences between the two antibodies (80R and 362), a potentially more universal antibody (named NSU1 in this study) was modeled. This hypothesized antibody was expected to bind more effectively to future mutations in the SARS spike protein. At the binding interface between these antibodies and the SARS spike protein, MAb362 mutations trend smaller and less polar including: Arg149Ser, Asn151Ser, Asp170Gly, and Trp213Ser. Due to the trend of smaller amino acids appearing in the MAb362 binding interface, it was hypothesized that more space in this area could allow antibodies to be more resistant to future SARS‐CoV spike protein structure variations. NSU1 was modeled based on MAb362 with the following four additional mutations: Asp103Gly, Trp104Leu, Gly170Ser, and Arg211Val. All of these except for Gly170 are mutations that decreased size and polarity of amino acid residues within the binding interface. Position 170 is Asp on the 80R structure and thus a mutation to Ser is still expected to maintain this trend of smaller residues in the antibody. Due to the additional space created due to these amino acid substitutions in the binding region between the antibody and RBD of the spike protein, NSU1 was predicted to be more resistant to spike protein mutations. These models allowed for deeper understanding of the impact that mutations in antibodies can have on binding interactions with viral proteins. Additionally, the modeling process also provided insight into the molecular structure of a potentially more universal antibody against variations in SARS‐CoV.