Dr. Richard Allen



During his time in CoMPLEX, Richard worked on modelling the endothelial cell response to fluid flow. In his doctoral thesis, he describes the aim of his project and his contributions to the field of bio physics:

"In vitro endothelial cells respond to fluid flow by elongating in the direction of flow. How the mechanical signal is transformed into an organised and directed response is poorly understood. The most studied and crucial aspects to this response are: actin filament alignment, mechano-transduction, signal transduction, Rho GTPase localised activation and lamellipodium formation. The goal of this project is to understand how these separate facets interact and lead to a coordinated response. The flow is modelled over a 3D virtual cell, which naturally gives the force the flow exerts on the cell surface via a boundary integral representation. This force is coupled to a Kelvin-body model of mechano-transduction which links, via a focal adhesion associated protein, Src, to a partial differential equation model (PDE) of the Rho GTPases Rac and Rho. The PDEs are integrated over a 2D projection of the 3D cell giving a time course for protein concentration at any point in the cell.

It is demonstrated that a mechano-transducer that can respond to the normal component of the force is likely to be a necessary (though perhaps not sufficient) component of the signalling network. In some processes cross talk between the GTPases is thought to be important in forming spatially segregated zones of activation, for example in the front and back of migratory cells. This research shows that local signalling in endothelial cells could be initiated by the force normal to the surface of the cell and maintained by limited diffusion. Modelling indicates the EC signalling response to fluid flow may be attenuated by a change in morphology. Rac and Rho activation and deactivation are validated against experimentally reported time courses that have been taken for whole cell averages. However it will be demonstrated that these time courses do not characterise the process and therefore there is a need for more quantitative local measure of protein activation. "

Page last modified on 08 aug 10 21:45