Tuesday, July 11
Computer Simulation as a Tool for Exploring Cytoskeletal Dynamics
9:15 AM - 10:00 AM
Room: Imperial Ballroom - ML
Chair: Leah Edelstein-Keshet, University of British Columbia, Canada
I will discuss a computer simulation model that demonstrates how simple mechanochemical interactions among the myriad protein parts of phylogenetically universal cytoskeletal genetic networks can yield emergent dynamics that animate cells. These realistic models involve numerical solution of tens to hundreds of thousands differential equations to resolve newtonian force balance laws characterizing tens of thousands of individual cytoskeletal parts, undergoing brownian motion, wandering freely in 3-D space and interacting with anything they bump into by exchanging appropriate forces. These are large computations, programmed with many threads in Java, using object-oriented programming to represent interactions of tens of thousands of independent 'agents', one per cytoskeletal part, run on a cluster of multiple processor linux computers. The example I present models the phenomenon of pronuclear fusion and centration in C. elegans (nematode worm) embryos in which microtubules grow, from centrosomes attached initially to the male pronucleus, throughout the cell's cytoplasm. Many dynein molecular motor proteins can attach to the cell cortex and to nuclear membranes, and exert force to walk along these microtubules toward the centrosomes. The emergent consequence is to pull the male and female pronuclei, which start at opposite ends of the egg cell, together and then center them in the cell. When we experimentally knock out dyneins in real embryos, pronuclear fusion and centration fail. I am therefore attempting an in silico reconstitution of these phenomena involving a minimal parts list: nuclei, centrosomes, microtubules, the cell's cortex, and dynein/dynactin motors plus the effects of CLIP170 on microtubule catastrophe rates. Computer-animated movies show the lifelike behavior that emerges from the molecular interactions, and these turn out, mysteriously, to include realistic pronuclear fusion, rotation, and centration. Of more general biological interest is using this model to explore a CLIP170-related mechanism by which collision with a cell's cortex might induce catastrophes that cause polymerizing microtubule tips to depolymerize instead. A mechanism for regulating the lengths of dynamically unstable microtubules so that, on average, they stop elongating when they encounter the cell cortex, is of much more general cell biological interest than the quirky C. elegans phenomena of pronuclear fusion and centration. I address those phenomena only as a proving ground for the general cytoskeletal dynamics computational tool I am developing. Despite frequent references to 'nuclear' and 'fusion', this has nothing to do with weaponry.
University of Washington