BPS2026 – A multi-scale simulation approach to elucidate PRC1-mediated microtubule organization

Microtubule-associated proteins (MAPs) are critical for organizing the cytoskeleton to support specialized cellular functions. This work focuses on PRC1, a passive crosslinker that stabilizes the mitotic spindle by selectively bundling antiparallel microtubules. To understand how molecular interactions produce robust spindle-wide organization, we have developed a multi-resolution simulation pipeline that works in synergy with cryo-electron tomography (cryo-ET) data. Our approach is built on a bottom-up coarse-graining (CG) strategy where we develop minimal coarse-grained models. The interaction potentials for these CG models are systematically parametrized using relative-entropy minimization guided by Bayesian optimization to faithfully reproduce key structural and thermodynamic properties. To capture the dynamic nature of the cytoskeleton, the CG model also incorporates a stochastic kinetic Monte Carlo component (through aLENS) that allows for the spontaneous formation and breaking of crosslinker-mediated networks. This combined approach allows our simulations to accurately capture emergent bundling dynamics and, at larger scales, complex network mechanics. Ultimately, this work aims to provide a clearer, quantitative picture of how passive crosslinkers build robust biological structures, contributing to our fundamental understanding of spindle mechanics and offering a transferable approach for multi-scale modeling in other complex systems.