Discovery and characterization of broad-spectrum entry inhibitors of coronaviruses

Abstract

A decade after SARS-CoV outbreak, MERS-CoV has emerged in 2012 and caused higher mortality for reported cases. With the continuous threat of MERS-CoV and the risk of re-emergence of SARS-like CoVs, the WHO thus puts coronaviruses (CoVs) on the list of “Top 8 diseases likely to cause severe epidemics” in order to raise public concerns. No vaccine or clinically approved antiviral therapy is currently available. It is important and urged for us to develop broad-spectrum inhibitors for CoVs to fight against any emerging CoVs in the future. 108 anti-SARS-CoV compounds has been previously identified from a high-throughput screening (HTS) (1). In this study, those hits were further tested against live MERS-CoV and pseudo-MERS-CoV (pMERS-CoV) to identify broad-spectrum entry inhibitors for CoVs. Finally, I identified three potential hits that were CA-616, CA-607 and CA-603. Their antiviral activities on various human coronaviruses (HCoVs) and other commonly circulating respiratory viruses were examined together with their cytotoxicity on Vero cells. These three compounds were found to specifically inhibit CoVs at low micromolar level with TC50 higher than 400 μM on Vero cells, so their selectivity index (SI) value was generally higher than 80. To have better SI value, structure-activity relationship study was conducted with their analogs. Structural moieties contributing to the efficacy of the corresponding lead compounds have been probed, facilitating future structural optimization of the compounds. CA-616, CA-607 and CA-603 were then applied for further mechanistic study. Time of addition assay indicated that they inhibited CoVs in the early stage of infection. This result was also supported by the pseudovirion assay in which they specifically inhibited pseudo-CoVs. Immunofluorescence assay further revealed that virions were trapped in late endosomes at 2 h pi in the presence of the compounds. This phenomenon suggested that virions were able to enter cells but their fusion process was probably blocked. I, therefore, hypothesized that the selected compounds may inhibit the fusion process by either targeting the Spike protein (S) of CoVs or prohibiting cathepsin L (CTSL) activity essential to S activation in late endosomes. I found that they neither physically interacted with the purified S of SARS-CoV in differential scanning fluorimetry, nor blocked cell-cell membrane fusion in polykaryon formation assay. Instead, they possessed cell-free and cell-based inhibitory activity on CTSL, meaning that they probably acted as CTSL inhibitors. Further enzymatic studies using pro-CTSL and mature CTSL indicated that our compounds preferentially targeted pro-CTSL as their potency towards pro-CTSL was higher than the mature form. This finding was also supported by the molecular docking analysis. It was found that our compounds mainly interacted with the residues on the propeptide of CTSL that is absent in the mature CTSL after its autoactivation, probably explaining the difference in their potency towards pro-CTSL and the mature form. This study not only highlights the biological significance of CTSL in CoV infection, but also provides lead compounds for the development of broad-spectrum anti-CoV agents. Most importantly, this is the first study to demonstrate CTSL inhibitors as potential broad-spectrum antivirals for CoVs.