Amino acid variants of SARS-CoV-2 papain-like protease have impact on drug binding

Abstract

The novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has caused both a health and economic crisis around the world. Its papain-like protease (PLpro) is one of the protein targets utilized in designing new drugs that would aid vaccines in the fight against the virus. Although there are already several potential candidates for a good inhibitor of this protein, the degree of variability of the protein itself is not taken into account. As an RNA virus, SARS-CoV-2 can mutate to a high degree, but PLpro variability has not been studied to date. Based on sequence data available in databases, we analyzed the mutational potential of this protein. We focused on the effect of observed mutations on inhibitors' binding mode and their efficacy as well as protein's activity. Our analysis identifies five mutations that should be monitored and included in the drug design process: P247S, E263D-Y264H and T265A-Y268C.

Fig 1. Structure of PLpro and location of the mutations.

Upper panel: cartoon representation of the protein with inhibitor S43 shown in sticks and zinc ion as a sphere (based on PDB ID: 7e35). Teal region of the protein indicates N-terminal ubiquitin-like domain (Ubl). Lower panel: cartoon representation of the protein with residues colored according to their mutation frequency (percentage of sequences with a given mutation out of all mutated sequences).

Fig 2. Structure of PLpro and variants within noncovalent inhibitor binding site.

The rates of mutations of these residues are presented in Table 1, along with the residues they are replaced to. It is worth noting that the residues forming the BL2 also can be mutated (aa 267–270). Based on PLpro-S43 complex from PDB ID: 7e35.

Fig 3. Selected variants found in PLpro.

(A) Location of amino acids subject to mutation relative to the ligand (S43) binding site (based on PDB ID: 7d7t). (B) Representative S43 conformations bound to PLpro. Calculations are based on frames sampled every 100 ps from joint Molecular Dynamics trajectories. Conformations found in less than 5% of the frames are not shown for clarity. (C) Frequency of protein-ligand hydrogen bonds in different trajectories. Left: GRL-0617, right: S43. Calculations are based on frames sampled every 100 ps from Molecular Dynamics trajectories. Dashed rectangle points to the amino acid interacting via hydrogen bond with GRL-0617 in the crystal structure (PDB ID: 7jrn), the one with solid line with S43 (PDB ID: 7d7t). In the case of Q269 the ligand can bind to either its main chain (M) or side chain (S). The red asterisk indicates the trajectories in which the ligand dissociated from the binding site. (D) Ligands MMGBSA binding energy (kcal/mol) to PLpro. Left: GRL-0617, right: S43. Calculations are based on frames sampled every 100 ps from Molecular Dynamics trajectories—the averages and standard errors are shown. The trajectories in which the ligand dissociated from the binding site are marked with red asterisk.

Fig 4. Characterization of activity and inhibitor binding properties of recombinant SARS-CoV-2 PLpro variants.

Assay conditions: [E]=100 nM, [S]=10 μM. (A) Determination of activity of the PLpro mutants toward tetrapeptide fluorogenic substrate. Experiment was repeated at least two times. (B) IC50 determination for variants that possessed enzymatic activity of GRL-0617 and S43 inhibitors. Enzyme was incubated in assay buffer for 10’ followed by enzyme incubation with inhibitor for 30’ prior to the measurement. Measurements were performed three times.