Measuring Disorder and Dynamical Properties of FG-Nucleoporins

Abstract

The nuclear pore complex (NPC) is a massive protein structure located on the nuclear envelope serving as a selective gateway between the cellular cytoplasm and nucleoplasm. The NPC is composed of numerous protein subunits referred to as nucleoporins, half of which are intrinsically disordered proteins (IDPs) containing numerous phenylalanine-glycine repeats. These so-called FG-nucleoporins (FG-Nups) naturally exist in an unfolded state without tertiary structure or persistent secondary structure, yet they are believed to mediate the selective gating function of the NPC. IDPs such as FG-Nups are challenging to characterize by molecular dynamics (MD) simulations because most available analysis tools were designed for protein simulations sampling relatively small conformational spaces. Our previous molecular dynamics simulations focused on exploring average static structural properties of FG-Nups such as extension, shape, and contact maps. These studies yielded significant differences in the average shapes of different FG-Nups supporting the new “forest” model for FG-Nup organization in the NPC. We are now extending this work to study the dynamical properties of FG-Nups, with the goal of developing statistical measures that can characterize different types and degrees of disorder from MD trajectories. We are evaluating established metrics for MD convergence and sampling efficiency (such as structural decorrelation time) as well as novel methods (such as metric scaling) to characterize the rate and extent of conformational sampling in IDPs. A significant challenge to the use of MD to study IDP dynamics is the potentially dramatic undersampling of the vast conformational space available to unstructured proteins. Therefore we are studying how to apply these analysis tools on structural ensembles derived from multiple independent simulations and accelerated MD methods. Additionally, to allow much longer MD trajectories, we are evaluating the accuracy of coarse grained force fields, such as the MARTINI model, for simulating IDPs.