Exploration of Protein Conformational Change with PELE and Meta- Dynamics Benjamin P. Cossins, † Ali Hosseini, † and Victor Guallar* ,†,‡ † Joint BSC-IRB Research Program in Computational Biology, Barcelona Supercomputing Center, c/Jordi Girona 29,08034 Barcelona, Spain ‡ Institució Catalana de Recerca i Estudis Avanç ats (ICREA), Passeig Lluís Companys 23, 08010 Barcelona, Spain * S Supporting Information ABSTRACT: Atomistic molecular simulation methods are now able to explore complex protein or protein-ligand dynamical space in a tractable way with methods such as meta-dynamics or adaptive biasing force. However, many of these methods either require a careful selection of reaction coordinates or the knowledge of an initial pathway of some kind. Thus, it is important that effective methods are developed to produce this pathway data in an efficient fashion. PELE, a proven protein-ligand sampling code, has been developed to provide rapid protein sampling in highly flexible cases, using a reduced network model eigen problem approach. The resulting method is able to rapidly sample configuration space with very general driving information. When applied to ubiquitin, PELE was able to reproduce RMSD and average force data found in molecular dynamics simulations. PELE was also applied to explore the opening/closing transition of T4 lysozyme. A meta-dynamics exploration using a low energy pathway validated that the configurations explored by PELE represent the most populated regions of phase space. PELE and meta-dynamics explorations also discovered a low free energy region where a large cross-domain helix of T4 lysozyme is broken in two. There is previous NMR evidence for the validity of this unfolded helix region. 1. INTRODUCTION Recent years have brought the realization of important milestones in atomistic protein simulation. Simulations of relatively large time scale events such as protein folding, protein-ligand association, and large scale conformational change are becoming tractable and predictive. 1-5 These advances rely on accelerated methods which use simplified pathway coordinates to explore complex many-dimensional processes and/or molecular simulation techniques able to efficiently use large numbers of computer processors. 6,7 Whatever the combination of methods used, there is a need to perform hundreds of nanoseconds worth of conformational sampling. The scale of computer power needed for systems of interest is not available to all, and so for the majority, the problem of rapidly obtaining realistic dynamic information on proteins remains. Methods able to quickly probe large scale protein conforma- tional changes based on molecular dynamics such as steered MD 8,9 (SMD) and essential dynamics sampling (EDS)) 10-12 have been used to direct MD in a direction of interest through clever constraints or restraints. There are many examples of steered or biased MD simulations being used to find a pathway for further free energy analysis with umbrella sampling or other such methods. 9,13,14 A recently developed method, temper- ature-enhanced essential dynamics replica exchange, seems able to steer large biomolecular MD simulations through temper- ature control of specific essential space modes while maintaining Boltzmann weighting. 15 Other advanced methods such as the finite temperature string method 16 and transition path sampling 17,18 attempt to sample defined pathways using molecular dynamics. Alternative pathway building methods have been developed based on minimization rather than MD. A family of methods based on the nudged elastic band method (NEB) 19-22 have been used to find pathways between two experimental structures of the same protein. NEB methods in general work on the basis of the minimization of a series of intermediate configurations between two end point protein structures. Every intermediate configuration is connected to the previous and next by springs which keep the structure of the path while allowing minima to be found. NEB methods have been used to find probable low energy pathways of protein conformational change, which can then be used in conjunction with free energy methods to give predictive information. 23 We present here a novel methodology capable of producing accurate and quick conformational sampling, and of providing reliable initial pathways for free energy methods. The methodology is a new development of the Protein Energy Landscape Exploration (PELE) program. PELE, a Monte Carlo (MC) based method, has thus far been used to characterize the exit pathways of bound molecules from proteins and for protein-ligand docking. 24-26 We introduced a new protein perturbation step based on anisotropic network model methodologies, capable of providing significant backbone motion. These PELE developments have been tested on two systems: ubiquitin (Ubi) and T4 lysozyme (T4lyz). Both systems were chosen due to their small size and the amount of experimental and computational studies on their dynamics. For Ubi, a 76 residue regulatory protein, we have compared the PELE Received: September 25, 2011 Published: January 27, 2012 Article pubs.acs.org/JCTC © 2012 American Chemical Society 959 dx.doi.org/10.1021/ct200675g | J. Chem. Theory Comput. 2012, 8, 959-965