Ribosomal tRNA release decoded via multiscale simulations at biological timescales

Understanding how molecular machines like the ribosome operate over long biological timescales remains a major challenge for computational biology. Large systems and rare, slow events, such as ligand dissociation or domain rearrangements, often occur on the order of milliseconds to seconds, far beyond the reach of conventional atomistic molecular dynamics simulations. Here, we introduce MUON (MUlti-scale multi-bOdy accelerated sampliNg), a novel coarse-grained, constrained dynamics framework that overcomes these limitations by modeling biomolecular assemblies with atomistic resolution while enforcing rigid-body domains connected by flexible joints. This enables temperature-accelerated sampling without unphysical distortions, allowing access to biologically relevant timescales in dramatically reduced wall-clock time. We applied MUON to study tRNA dissociation from the E-site of the Escherichia coli ribosome, a process that occurs on the timescale of hundreds of milliseconds to seconds. Our simulations captured over 100 dissociation events across multiple stages of the ribosomal translocation cycle, cumulatively corresponding to approximately 15 minutes of biological time. From this data, we extracted free energy landscapes, transition pathways, and kinetic estimates, revealing the multichannel nature of tRNA release and offering consistent explanations for several seemingly conflicting intermediates observed by independent cryo-EM experimental studies. The simulations suggest that the dynamic interplay among the L1 stalk, the small subunit head, and the deacylated tRNA collectively governs the tRNA release pathway. Our study demonstrates that rigid-body accelerated sampling provides a practical path to simulate large biomolecular machines over physiological timescales and offers a general strategy for probing rare transitions in structural biology.