Models of Vimentin Organization Under Actin-Driven Transport

TL;DR

This study models vimentin intermediate filament dynamics in cells, revealing that their organization is best explained by spatially dependent transport speed or trapping, driven mainly by actin mechanisms when microtubules are disrupted.

Contribution

It introduces a novel mathematical model of vimentin dynamics incorporating state switching and spatially variable transport, validated against experimental data.

Findings

Vimentin moves towards the cell center after microtubule disruption.

Actin-driven transport primarily governs vimentin movement in these conditions.

Spatially dependent transport parameters best explain the observed filament behavior.

Abstract

Intermediate filaments form an essential structural network, spread throughout the cytoplasm and play a key role in cell mechanics, intracellular organization and molecular signaling. The maintenance of the network and its adaptation to the cell's dynamic behavior relies on several mechanisms implicating cytoskeletal crosstalk which are not fully understood. Mathematical modeling allows us to compare several biologically realistic scenarios to help us interpret experimental data. In this study, we observe and model the dynamics of the vimentin intermediate filaments in single glial cells seeded on circular micropatterns following microtubule disruption by nocodazole treatment. In these conditions, the vimentin filaments move towards the cell center and accumulate before eventually reaching a steady-state. In absence of microtubule-driven transport, the motion of the vimentin network is primarily driven by actin-related mechanisms. To model these experimental findings, we hypothesize that vimentin may exist in two states, mobile and immobile, and switches between the states at unknown (either constant or non-constant) rates. Mobile vimentin are assumed to advect with either constant or non-constant velocity. We introduce several biologically realistic scenarios using this set of assumptions. For each scenario, we use differential evolution to find the best parameter sets resulting in a solution that most closely matches the experimental data, then the assumptions are evaluated using the Akaike Information Criterion. This modeling approach allows us to conclude that our experimental data are best explained by a spatially dependent trapping of intermediate filaments or a spatially dependent speed of actin-dependent transport.