Contractility in an extensile system.

Essentially all biology is active and dynamic. Biological entities autonomously sense, compute, and respond using energy-coupled ratchets that can produce force and do work. The cytoskeleton, along with its associated proteins and motors, is a canonical example of biological active matter, which is responsible for cargo transport, cell motility, division, and morphology. Prior work on cytoskeletal active matter systems showed either extensile or contractile dynamics. Here, we demonstrate a cytoskeletal system that can control the direction of the network dynamics to be either extensile, contractile, or static depending on the concentration of filaments or weak, transient crosslinkers through systematic variation of the crosslinker or microtubule concentrations. Based on these new observations and our previously published results, we created a simple one-dimensional model of the interaction of filaments within a bundle. Despite its simplicity, our model recapitulates the observed activities of our experimental system, implying that the dynamics of our finite networks of bundles are driven by the local filament-filament interactions within the bundle. Finally, we show that contractile phases can result in autonomously motile networks that resemble cells. Our results reveal a fundamentally important aspect of cellular self-organization: weak, transient interacting species can tune their interaction strength directly by tuning the local concentration to act like a rheostat. In this case, when the weak, transient proteins crosslink microtubules, they can tune the dynamics of the network to change from extensile to contractile to static. Our experiments and model allow us to gain a deeper understanding of cytoskeletal dynamics and provide an new understanding of the importance of weak, transient interactions to soft and biological systems.