High-Throughput Simulations of Protein Dynamics in Molecular Machines: The ‘Link’ Domain of RNA Polymerase

Abstract

Complex molecular machines are involved in all information-processing steps in cells. The handling of information-bearing macromolecules (predominantly nucleic acids and proteins) requires a range of catalytic capabilities, such as the sequence-specific synthesis of macromolecular entities from smaller precursors and further processing by breaking/joining of pre-existing molecular components. These catalytic activities have to be spatially and temporally precisely controlled to ensure accuracy and efficiency levels that are commensurate with the associated biological functions (Lyubimov et al., 2011). Unlike metabolic enzymes, which are often medium-sized enzymes with freely accessible active sites, the enzymes associated with replication, transcription, translation, recombination are typically large multi-subunit protein complexes capable of a complex conformational spectrum. This conformational spectrum manifests itself through the dynamic features of individual domains within these molecular machines: many of the domains act explicitly as nanomechanical elements by serving as allosteric sensors, molecular hinges and motor units (Bustamante et al., 2011; Heindl et al., 2011). A deeper understanding of the functions of molecular machines is currently a high-priority topic and ultimate success will strongly depend on a combination of ‘wet-lab’ experimental techniques (X-ray crystallography, NMR, spectroscopic methods, site-directed mutagenesis and chemical cross-linking) with sophisticated computational simulations. Fully atomistic molecular dynamics (MD) simulations offer the most comprehensive and detailed insights into the behavior of molecular entities and are therefore the method of choice for attempts to unravel the structural basis of molecular mechanisms (reviewed in Karplus & McCammon, 2002; Freddolino et al., 2010; McGeagh et al. 2011; Schlick et al., 2011).