Motor Proteins Have Highly Correlated Brownian Engines

TL;DR

This paper models two-headed motor proteins like kinesin as correlated Brownian ratchets, explaining their discrete stepping behavior and reduced variance, and compares the model's predictions with experimental data.

Contribution

It introduces a correlated Brownian ratchet model for motor proteins, capturing their discrete steps and reduced variance, and extends to more realistic mechanical models aligning with experimental observations.

Findings

Model reproduces discrete 8nm steps of kinesin

Reduced variance rules out flashing ratchet models

Mechanical models show qualitative agreement with experiments

Abstract

Two headed motor proteins, such as kinesin and dynein, hidrolyze environmental ATP in order to propel unidirectionally along cytoskeletal filaments such as microtubules. In the case of kinesin, protein heads bind primarily on the alpha tubulin site of asymmetric alpha-beta 8nm-long tubulin dimers that constitute the microtubular protofilaments. Kinesin dimers overcome local binding forces up to 5pN and are known to move on protofilaments with ATP concentration-dependent speeds while hydrolizing on average one ATP molecule per 8nm step. The salient features of protein trajectories are the distinct abrupt usually 8nm-long steps from one tubulin dimer to the next interlaced with long quiescent binding periods at a tubulin site. Discrete walks of this type are characterized by substantially reduced variances compared to pure biased random walks, and as a result rule out flashing-type ratchet models as possible mechanisms for motor movement. On the other hand, simple additive correlated brownian ratchets that present exactly these discrete trajectory patterns with reduced variances are compatible with the general features of protein motion. In the simplest such model the protein is simplified to a single particle moving in a periodic non-symmetric tubulin-derived potential and the environmental and ATP interaction is included in a correlated additive noise term. For this model we show that the resulting protein walk has features resembling experimental data. Furthermore, in more realistic mechanical models of two masses connected by a spring we find qualitative agreement with recent experimental facts related to motion of protein chimaeras formed through kinesin motor domains with non-claret disjunctional (ncd) neck regions.