Conformational Sampling during the Reaction of the SARS-CoV‑2 Methyltransferase nsp16

The
coronavirus 2′-O-methyltransferase nsp16 catalyzes the
methylation of the viral RNA cap structure, playing an essential role
in viral RNA immune evasion. Unusually, nsp16 forms a heterodimer
with a second viral protein, nsp10, which appears to be essential
for activity. Here, we use a combination of density functional theory
(DFT) modeling of the nsp16 active site to investigate the methyl
transfer reaction and molecular dynamics (MD) simulations to investigate
substrate binding and dynamics. The active site cluster models give
barrier heights of 76–120 kJ mol–1, with
much of the variation appearing to come from large differences in
the relative (de)stabilization of the reactant state. The lower barriers
are in good agreement with experiment, suggesting that conformational
sampling of nonreactive conformations of the RNA substrate occurs
during MD simulations. An analysis of interaction energies shows that
nsp10 stabilizes nsp16-RNA interactions, but we see only modest changes
in nsp16 structure and dynamics upon removal of nsp10, with these
changes centered around the active site loops and the dimer interface.
We also observe a considerable conformational sampling of RNA substrates
within the active site. The population of potentially reactive substrate
configurations is relatively low, and we see no significant effect
of nsp10, but differences between RNA substrates 7-methyl-GpppA and
7-methyl-GpppAUU; the larger substrate appears to more frequently
sample potentially reactive configurations. This conformational sampling
of the RNA substrate is consistent with X-ray crystal structures of
substrate (Michaelis) complexes of nsp16 nsp10, where soaking the
RNA fragment into a crystal of SAM-bound nsp16 nsp10 can prevent productive
sampling of the RNA conformational space within the active site.