Molecular Dynamics Simulations Support the use of Methyl Generalized Order Parameters in the “Entropy Meter”

Abstract

Conformational entropy is a potentially important thermodynamic parameter contributing to protein function. Unfortunately, conformational entropy has been difficult to measure experimentally. We have been working towards using measures of fast internal motion as a proxy for conformational entropy. Implementation of this approach has been impeded by the apparent need to employ a model-dependent interpretation of the obtained generalized order parameters. We have recently proposed that the quantitative relationship between generalized order parameters and conformational entropy could be determined empirically and effectively create an ‘entropy meter’ (Marlow et al. Nat. Chem. Biol. 6, 353). This approach has initially employed methyl symmetry axis order parameters that can be obtained relatively easily. The approach then rests on the coupling between motion of the methyl bearing side chains and the rest of the amino acids being sufficient to report on the whole protein. Here, we examine the generality of this assumption using molecular dynamics simulations of a number of proteins. The MD simulations performed with the NAMD suite showed excellent agreement between the calculated and experimental side chain order parameters for all the proteins, a marked improvement from previous studies. We find a general correlation between the internal dynamics of protein and the protein conformational entropy. The calculated protein conformational entropy showed excellent linear correlation to both the calculated NMR order parameter of methyl groups as well as the experimentally measured NMR methyl order parameters. Supported by NIH grant GM102447 and a grant from the Mathers Foundation.