Simone Furini (University of Siena)

May 8, 2019 – 11:00 AM

DIISM, Artificial Intelligence laboratory (room 201), Siena SI

Description

Ion channels are membrane proteins that control the fluxes of ions across cell membranes. In order to perform their biological functions, these proteins need to select the different ion species with high accuracy. For instance, K+ channels are highly permeable by potassium ions, with conductances in the range 1-100 pS, and at the same time they are almost completely impermeable by sodium ions, with selectivity ratios as high a 1:10000, while the opposite behavior is exhibited by sodium channels. In the last two decades, the release of the experimental atomic structures of several potassium and sodium channels revealed important details about conduction and selectivity in these two families of membrane proteins. However, experimental structures provide mainly static information, while conduction is an intrinsically dynamic process. Molecular Dynamics simulations allow to reproduce the dynamics of complex biological molecules, as ion channels, at the atomic scale. The comparison between atomic trajectories and experimental data is prevented by the high-computational cost of this technique, and the consequent severe limitations on the accessible time scale. I will discuss the methods used to accelerate sampling of the configurational space in Molecular Dynamics simulations, and to extract useful information that could be directly related to experimental behaviors from atomic trajectories. The applications of these methods to potassium and sodium channels will be presented, together with the current hypothesis about the atomic mechanisms of conduction and selectivity.