Allostery — the regulation of a protein's activity from a site distant to its functional region — underlies most cellular signalling and offers a selective route to therapeutics. Our group has helped reshape the modern view of allostery as a dynamic, ensemble-level phenomenon rather than a rigid conformational switch.
Working on the PDZ domain — a canonical signalling module — we uncovered the hidden electrostatic basis of dynamic allostery (PNAS, 2017) and showed that ligand binding or protonation triggers a perturbation-independent ‘universal response network’ (J. Phys. Chem. Lett., 2020), redistributing conformational entropy with little change in average structure. This provides a predictive, mechanistic framework for allosteric signalling.
We dissected how molecular switches operate in Rho GTPases — nucleotide-dependent conformational heterogeneity and competing interactions (Sci. Rep., 2017) — and how phosphorylation of RhoGDI alters protonation states to allosterically control the release of RhoA/Rac1 (Biophys. J., 2024; Protein Science, 2026), in collaboration with R. Nussinov.
Using mixed-solvent molecular dynamics, we discovered a druggable cryptic pocket in the C-terminal domain of PCSK9 (J. Phys. Chem. B, 2025) and showed how the bidirectional nature of allostery can disrupt the PCSK9–LDLR interaction (J. Chem. Inf. Model., 2024) — a major cardiovascular target. The same hotspot-mapping strategy revealed a cryptic pocket in the mitotic kinase PLK1 (ChemPhysChem, 2026). We distilled these advances into a roadmap from function to therapeutic innovation (J. Mol. Biol., 2025) and extended the toolkit to bacterial toxin–antitoxin systems (HipA; Biochemistry, 2023) and antibody–antigen mechanical stability (Protein Science, 2025).