Modelling the Dynamics of Actomyosin Networks

Abstract

In this dissertation, a new framework called bead-state model (BSM) for computer simulations of actin networks is presented, and its scientific validity and its usability and flexibility is demonstrated in three proof-of-concept scenarios. G-actin is a globular protein that can polymerize into semi-flexible actin fibers of several micrometer length. These actin fibers are one of the major constituents of the cell's cytoskeleton, and as such they play a main role in the cell's active motility as well as its response to external mechanical stimuli. Actin networks display a large variety of different network architectures and properties, allowing them to fulfill various different roles. One main driver for these differences is the interplay between actin and the vast family of actin binding proteins. Due to their importance, actin along with actin binding proteins are subject to many in vivo, in vitro and in silico studies. For in silico studies, open source frameworks like cytosim and MEDYAN exist. The BSM framework extends the available options by one that focusses on ease-of-use, large networks and interaction with external (i.e. non-actin) particles and potentials. In a first proof-of-concept scenario in this disseration, BSM is used to simulate isolated fibers to confirm that they behave like worm-like chains. The worm-like chain model is commonly used to describe the dynamics of isolated actin fibers. In a second set of simulations, large networks of actin fibers are simulated and studied via passive microrheology. Passive microrheology is a method where the trajectory of a large tracer particle is studied in order to infer viscoelastic properties of the network in which the tracer particle is embedded. The simulations yield characteristic properties that were predicted theoretically and confirmed experimentally, namely the power-laws that relate networks' plateau moduli to actin concentration in entangled and cross-linked actin networks. To further demonstrate the flexibility of the BSM framework in constructing systems with external particles and external potentials, the third scenario simulates a thin layer of actin network which gets indented by a sphere that is pressed into the network. These systems mimic indentation by an atomic force microscope. The resistance to the mechanical indentation and relaxation times are studied for networks without binding proteins (entangled networks) and with static cross-links, dynamic cross-links and active myosin motors.