Recombinant rod domain of vimentin binds to spike protein to block SARS‐CoV‐2 replication in vitro

Abstract

The COVID‐19 pandemic, caused by SARS‐CoV‐2, has infected at least 85 million people worldwide and led to over 1.8 million deaths. To date, there are no approved efficacious treatments targeting SARS‐CoV‐2, with current tactics (e.g. modulating the host response & using repurposed medications) having mixed results. Thus, alongside vaccines, additional treatments to target viral infection are critical to fight the pandemic. SARS‐CoV‐2 uses its spike protein to bind to ACE2 on host cells. Therefore, developing a target to bind spike could prevent viral infection. We previously published on the development and testing of recombinant rod domain of vimentin (rhRod) to bind to P‐selectin to block leukocyte adhesion to platelets and endothelium. Given that the SARS‐CoV spike protein (which shares 76% sequence identity to SARS‐CoV‐2 spike) binds to vimentin, we hypothesized that rhRod would bind to SARS‐CoV‐2 spike protein to block viral adhesion. We used biolayer interferometry (OctetRed384) to measure the KD of rhRod (0‐1000 nM) to immobilized SARS‐CoV‐2 spike protein (S1S2ECD‐His). We found that rhRod had a strong affinity to the spike protein (180 ± 23 nM; R2 0.9858; Fig. 1 A‐B). Heat inactivation (100°C for 10 minutes) of rhRod (ΔrhRod) did not bind to spike protein (indeterminate; R2 0; Fig. 1 C‐D). We then tested whether rhRod blocked ACE2 from binding to spike protein using 2 methods (Fig. 1 E‐F). Immobilized spike protein sensors were first placed in rhRod (0‐1000 nM) containing wells then placed into wells with 200 nM ACE2. The IC50 of rhRod was 139 ± 17 nM (R2 0.9959). A separate series of experiments were done using immobilized ACE2 sensors placed in wells containing 125 nM spike protein mixed with increasing concentrations of rhRod (0‐500 nM). As the concentration of rhRod increased, binding to ACE2 sensors decreased, with an IC50 of 34 ± 8 nM (R2 0.9845). Based on our findings, we used in silico modeling (SwarmDock) to asses the predicted interactions between rhRod and spike. We found that rhRod was predicted to bind to spike protein in the same region as ACE2, consistent with our empirical data that rhRod blocks spike‐ACE2 interactions. Finally, to test whether rhRod decreased SARS‐CoV‐2 replication, we infected Vero E6 cells with SARS‐CoV‐2 (USA‐WA1/2020; MOI 0.1) in the presence of rhRod (0‐4,000 nM). Daily administration of rhRod decreased viral replication (PFU/mL) by 100‐1000‐fold (1000‐4000 nM) starting at 48 hours, through 96 hours (segmental linear regression with least squares fit; Fig. 1G). ΔrhRod (4000 nM) had no effect on viral replication. Our data show that rhRod binds to spike protein to block spike‐ACE2 interactions and decreases SARS‐CoV‐2 viral replication in vitro. Data from these studies will inform preclinical models of COVID‐19 to further develop a potential novel therapy for this devastating disease.