Condensed Matter Physics Seminar

Microtubules are key components of force transmission in cells, andplay a role in cell locomotion, transport, and mitosis. Experiments inTanmay Lele's group at UF have shown that microtubules severed bylaser ablation do not straighten, as would be expected from the largebending moments along their lengths. Instead, segments near newlycreated minus ends typically increased in curvature followingsevering, while segments near new microtubule plus ends depolymerizebefore any observable change in shape. However, in dynein-inhibitedcells, segments near the cut straightened rapidly following severing.These observations suggest that microtubules are subject tosignificant tangential forces, and that lateral motion of themicrotubule is opposed by a large effective friction rather thanelastic forces. To help interpret the experimental results we havedeveloped a new numerical model for intracellular microtubulemechanics, accounting for dynein-generated forces on the microtubules.At the length scales of mammalian cells, microtubules behave assemi-flexible filaments and can be coarse-grained using the Kirchofftheory for elastic rods. We have supplemented the Kirchoff model[1]with the stochastic growth and collapse of microtubules[2] (thedynamic instability), and by a model for dynein generated forces[3].Numerical simulations of the buckling of a single microtubule canexplain both the enhanced buckling at the minus end of a severedmicrotubule and the apparently frozen shape of the plus end. Ourresults suggest that microtubule shapes in vivo reflect a dynamicforce balance, where bending moments are opposed by dynein-motorforces, including an effective friction from the stochastic bindingand unbinding of the motors. I will present simulations of thedynamics of the centrosome, driven by the motion of ~ 100microtubules. The results are consistent with a mechanism forcentrosome centering driven by pulling forces exerted by dyneinmotors. I will explain how tension on the centrosome can be reconciledwith buckled filaments near the cell periphery. The simulations spantime scales of about 14 orders of magnitude using a projection method[4] combined with parallelization to speed up the simulations.[1] A. J. C. Ladd and G. Misra, J. Chem. Phys. 130:124909, 2009.[2] T. Mitchison and M. Kirschner Nature 312: 237-42, 1984.[3] R. B. Dickinson, Private Communication, 2010.[4] I. G. Kevrekidis, C. W. Gear and G. Hummer, AIChE J. 50:1346, 2004.