A hydrophobic cluster controls long-range allostery in the TRMT2A RNA recognition motif

Abstract

TRMT2A has emerged as a disease-modifying target in polyglutamine (PolyQ) models, yet the conformational preferences and allostery of its RNA recognition motif (RRM) remain poorly resolved. Here we combine extensive atomistic molecular dynamics with Markov state modeling (BHMSM), transition path theory, and structure-based pocket analysis to map the conformational landscape of the human TRMT2A RRM. We resolve six metastable states and show that a hydrophobic cluster centered on F92-W134-L133 governs their interconversion. We further identify residues contributing to RNA strand recognition and reveal state-specific cryptic pockets consistent with reported TRMT2A RRM small molecule inhibitors. Together, these results support a hinge-gate model in which a soft, defectenabled 2 segment and a loop 5 hydrophobic cluster coordinate long-range communication between the RNP face and opposite side, yielding testable mutational predictions and state-specific opportunities for allosteric control of TRMT2A in polyQ disease contexts.