Author’s Disclosure Block: Raheem Remtulla, Wisconsin Alumni Association Grant/research support, Wisconsin Alumni Association, Grant/research support; Leonard Levin, none

Abstract Body
Purpose: Anti-myelin oligodendrocyte glycoprotein (MOG) syndromes have been associated with clinically significant optic neuritis. Fortunately, novel phosphine-borane compounds (PBC) have demonstrated protective effects in retinal ganglion cells, through the suspected mechanism of disulfide bond cleavage. PBCs have demonstrated in vivo interactions with MOG. Using in silico modelling, we simulated the impact of disulfide bond cleavage on MOG by PBCs and its effect on anti-MOG antibody interactions. Study Design: Proteonomic Modelling Methods Methods: Molecular Docking: Computational structure of MOG was optimised to its lowest energy state. After this the MOG disulfide bond was modelled with a dataset of PBC.PBC structures were optimized to their lowest energy states at physiologic conditions. Using the Maestro's covalent docking feature, the interaction between PBC and MOG were simulated. The affinity for compound docking was assessed using docking scores and root-mean-square deviations. Methods: Antibody Modelling: Both the computational MOG and anti-MOG antibody structures were optimized to their lowest energy states. Using the protein docking features, the interaction between MOG and anti-MOG antibodies were simulated with both the MOG disulfide intact (MOG-DI) and MOG disulfide reduced (MOG-DR). Protein antibody stability was assessed using total complex energy, interface residuals, interface surface area, salt bridges, hydrogen bonds and non-contact bonds. RESULTS: Results Molecular Docking:All PBC compounds tested were able to directly interact with this disulfide bond. Multiple PBCs demonstrated a significantly greater affinity for docking than controls. PBCs that interacted with arg25 and lys80 had stronger affinities for docking than those that did not. These findings demonstrated that PBCs were able to interact with and cleage the disulfide bond of the MOG protein. Results Antibody Modeling: The MOG-DI antibody complex had a more favorable total complex energy compared to the MOG-DR complex by 1.06 kcal/mol. Furthermore, the MOG-DI antibody complex had 3 additional interface residuals, 53 Ųinterface area, 4 additional hydrogen bonds, and 32 additional non-contact bonds. These findings suggest that the MOG antibody is more likely to bind to the MOG protein when the disulfide bond is intact. CONCLUSIONS: PBC demonstrated the ability to cleave the disulfide bond of MOG and cleaving the disulfide bond on MOG reduced anti-MOG antibody complex favorability. This suggests that PBCs have the potential as a treatment in anti-MOG syndrome and that the potential neuroprotective mechanism of PBCs is through a MOG mediated pathway.