Unraveling the complexity of biological processes from protein native dynamics to cell motility in molecular simulations

This dissertation consists of two major parts. Both are dedicated to studying biological molecular processes with computer simulations, but differ in the scale of the studied processes. In one project we investigated the dynamics of cellular organelles involved in cell motility -- the filopodia. The other project is zooming in to the scale of single molecule, elucidating the organization of protein molecule native state.;Some motile cells use special fingerlike probes of their environment for guiding their motion called filopodia. They are bundles of parallel actin filaments protruding from the cell body and enveloped by cell's membrane. They are highly dynamic, constantly growing an retracting, randomly, or in response to the change in the environment. These dynamics are governed by the cell's regulatory proteins and by external chemical cues or mechanical obstacles. The previous models predicted that a filopodium grows to a stationary length of about 1 mum with miniscule fluctuations around. (i) We found that capping proteins (they attach to the barbed ends of actin filaments and stop polymerization) can induce macroscopic oscillation of filopodial length -- the growth-retraction cycles. The retraction can be complete. This is the first model that predicts finite lifetimes for filopodia. The lifetimes are consistent with experimental observations. (ii) In the model, however, the maximal filopodial lengths of several microns are still limited by the diffusional transport of actin monomers to the filopodial tip and are far below experimentally observed lengths of up to 100 mum. Assuming the obvious solution for the problem of slow transport in cell, the molecular motors, that are known to be present inside filopodia, we found that a naive addition of motors does not increase the lengths much. In order to have an efficient active transport, two rules must be observed: the motors should not sequester the cargo and the rails for motors should be kept from being clogged by motors. Protein Ena/VASP that is known to be actively transported to the filopodial tip by molecular motors may be a way to fight sequestration.;On the scale of a single macromolecule we studied the organization of protein native state. It is not a single structure, but an ensemble of constantly interconverting conformations. It is essential for a deep insight into protein functioning to know thermodynamics of these substates and dynamical regime of their exploration. (i) In all-atom MD simulations we constructed a 2D free energy surface for a protein Trp-cage and using the FES for Brownian dynamics investigated the nature of dynamical behavior of Trp-cage in its native state. We found that the dynamical regime is borderline between liquid and supercooled liquid. (ii) We developed a general technique for calculating free energy difference between two polymer conformations in explicit solvent simulations and used the Trp-cage 2D FES for testing of this technique, revealing remarkable accuracy and computation efficiency.