Accurate Binding Configuration Prediction of a G‑Protein-Coupled
Receptor to Its Antagonist Using Multicanonical Molecular Dynamics-Based
Dynamic Docking
posted on 2021-09-22, 10:29authored byGert-Jan Bekker, Mitsugu Araki, Kanji Oshima, Yasushi Okuno, Narutoshi Kamiya
We
have performed dynamic docking between a prototypic G-protein-coupled
receptor (GPCR) system, the β2-adrenergic receptor,
and its antagonist, alprenolol, using one of the enhanced conformation
sampling methods, multicanonical molecular dynamics (McMD), which
does not rely on any prior knowledge for the definition of the reaction
coordinate. Although we have previously applied our McMD-based dynamic
docking protocol to various globular protein systems, its application
to GPCR systems would be difficult because of their complicated design,
which include a lipid bilayer, and because of the difficulty in sampling
the configurational space of a binding site that exists deep inside
the GPCR. Our simulations sampled a wide array of ligand-bound and
ligand-unbound structures, and we measured 427 binding events during
our 48 μs production run. Analysis of the ensemble revealed
several stable and meta-stable structures, where the most stable structure
at the global free energy minimum matches the experimental one. Additional
canonical MD simulations were used for refinement and validation of
the structures, revealing that most of the intermediates are sufficiently
stable to trap the ligand in these intermediary states and furthermore
validated our prediction results. Given the difficulty in reaching
the orthosteric binding site, chemical optimization of the compound
for the second ranking configuration, which binds near the pocket’s
entrance, might lead to a high-affinity allosteric inhibitor. Accordingly,
we show that the application of our methodology can be used to provide
crucial insights for the rational design of drugs that target GPCRs.