Geometric Sampling Framework for Exploring Molecular Walker Energetics and Dynamics

Abstract

The motor protein kinesin is a remarkable natural nanobot that moves cellular cargo by taking 8 nm steps along a microtubule molecular highway. Understanding kinesin's mechanism of operation continues to present considerable modeling challenges, primarily due to the millisecond timescale of its motion, which prohibits fully atomistic simulations. Here we describe the first phase of a physics-based approach that combines energetic information from all-atom modeling with a robotic framework to enable kinetic access to longer simulation timescales. Starting from experimental PDB structures, we have designed a computational model of the combined kinesin-microtubule system represented by the isosurface of an all-atom model. We use motion planning techniques originally developed for robotics to generate candidate conformations of the kinesin head with respect to the microtubule, considering all six degrees of freedom of the molecular walker's catalytic domain. This efficient sampling technique, combined with all-atom energy calculations of the kinesin-microtubule system, allows us to explore the configuration space in the vicinity of the kinesin binding site on the microtubule. We report initial results characterizing the energy landscape of the kinesin-microtubule system, setting the stage for an efficient, graph-based exploration of kinesin preferential binding and dynamics on the microtubule, including interactions with obstacles.