Information-driven stepping in dimeric transport motors

Abstract

Dimeric transport motors are nanoscale protein complexes that move along cytoskeletal filaments. Here we introduce a theoretical model for their stepping dynamics, in which the two motor heads undergo Brownian motion with mobilities periodically switching in a position-dependent manner so that only one head moves at a time. Through consumption of chemical free energy, this mechanism produces directed motion. We characterize at steady state the motor's mean velocity and its energy and information flows. The motor operates as a pure information engine, where only information is transduced between its components, without energy exchange. For localized switching, the model yields a thermodynamically consistent expression for the mean motor velocity that reproduces experimentally observed behavior, and captures the stall force characteristic of tightly coupled motors. Finally, coarse-graining the model to a single mechanical degree of freedom produces a second-order non-Markovian dynamics, from which we compute the four distinct dwell-time distributions that can be directly observed in single-molecule experiments. Our findings highlight how information transduction, via implicit operation as a Maxwell demon, may underlie the remarkable performance of these molecular motors.