Poster AA13 - Alternative Cleavage of SARS-CoV-2 Spike Protein by Cathepsin Proteases May Affect ACE2 Binding and Cell Infectivity

Introduction: Many viral entry mechanisms require host proteases to enter the cell. For example, SARS-CoV-2 spike protein must be cleaved twice for the virus to infect human cells after initial binding to ACE2 receptors on the cell membrane. The first cleavage separates the binding (S1) and membrane fusion (S2) regions. The second cleavage exposes the membrane fusion peptide within the S2 region at the S2' site permitting viral entry. Beyond Furin and TMPRSS2, the most studied proteases implicated in spike cleavage and infectivity, other endogenous proteases such as cysteine cathepsins have been shown to cleave multiple sites on the spike protein, some of which may interfere with virus binding and fusion to human cells. Here we build on our previous work and build a multi-protease network modeling the kinetics of cathepsins K, L, S cleavage at multiple sites on SARS-CoV-2 spike protein while competing with the extracellular matrix (ECM) proteins that are also substrates for these cathepsins. We will test the hypothesis that multi-cathepsin cleavage of spike protein can prevent its cell binding and fusion. This will present additional therapeutic targets to prevent SARS-CoV-2 infection in diseases that upregulate cathepsins such as hypertension, emphysema, cardiovascular disease, and more.

Materials and

Methods: In this study, we first performed computational predictions based on modeling molecular dynamics and protease specificity to find potential binding and cleavage sites of spike protein. PACMANS was used to predict cleavage sites based on amino acid sequence. A molecular model was built using homology modeling (SwissProt) based on PDB 7TB4, on which surface accessibility calculations (Chimera) and molecular dynamics simulations (GROMACS) were performed to assess the viability of each cleavage site. We then tested these predictions by incubating monomeric or trimeric spike protein with recombinant cathepsins K, L, and S. The cleavage fragments were analyzed using a combination of western blotting and mass spectrometry. Western blots were incubated with antibodies corresponding to five different domains of the spike protein to see where cleavages took place on a macro scale. Fragments from SDS-PAGE gels were then trypsin digested and analyzed by LC/MS ESI-ToF mass spectrometry (Waters Synapt G2). These fragments were analyzed with MZMine3 and Waters Connect to match digested peptides to sequences and rebuild full protein sequences to verify exact cleavage locations. We then built a working model of cathepsin-mediated degradation of spike proteins using ordinary differential equations in Julia, to include a model that can simulate multiple cleavages of a spike protein.

Results, Conclusions, and Discussions: We computationally predicted sites in the NTD (N-terminal domain), RBD (receptor-biding domain), and S2 domain of SARS-CoV-2 spike protein that would be cleaved by cathepsins. Western blotting showed fragments that matched the predicted sizes of cleaved fragments. In parallel, mass spectrometry showed agreement of peptide molecular weights with the computational predictions. For the proteolytic network, increasing amounts of spike protein concentrations distracted cathepsins from hydrolyzing the ECM proteins, especially cathepsin K. Spike protein and the ECM proteins extended the activity life of the proteases as well, reducing their autodigestion binding and catalysis which removes them from the system.

Acknowledgements (Optional): Considering multiple cleavage sites on one tripeptide molecule involves 3D molecular structure, access upon unfolding, and competition with other protease substrates. This computational network model is helpful in predicting combinatorial strategies to maximize effective cleavage of spike protein in the presence of different types of tissues with different ECM compositions. Further validation with virus-like particles (VLPs) expressing SARS-CoV-2 spike protein to incorporate more of the 3-dimensional presentation and protease access to spike protein are planned.