Configurational entropy is thought to influence biomolecular processes, but there are still many open questions about this quantity, including its magnitude, its relationship to molecular structure, and the importance of correlation. The mutual information expansion (MIE) provides a novel and systematic approach to computing configurational entropy changes due to correlated motions from molecular simulations. Here, we present the first application of the MIE method to protein-ligand binding, using multiple molecular dynamics simulations (MMDSs) to study association of the UEV domain of the protein Tsg101 and an HIV-derived nonapeptide. The current investigation utilizes the second-order MIE approximation, which treats correlations between all pairs of degrees of freedom. The computed change in configurational entropy is large and is found to have a major contribution from changes in pairwise correlation. The results also reveal intricate structure-entropy relationships. Thus, the present analysis suggests that, in order for a model of binding to be accurate, it must include a careful accounting of configurational entropy changes.

The Protein Data Bank (PDB) is the global archive for structural information on macromolecules, and a popular resource for researchers, teachers, and students, amassing more than one million unique users each year. Crystallographic structure models in the PDB (more than 100,000 entries) are optimized against the crystal diffraction data and geometrical restraints. This process of crystallographic refinement typically ignored hydrogen bond (H‐bond) distances as a source of information. However, H‐bond restraints can improve structures at low resolution where diffraction data are limited. To improve low‐resolution structure refinement, we present methods for deriving H‐bond information either globally from well‐refined high‐resolution structures from the PDB‐REDO databank, or specifically from on‐the‐fly constructed sets of homologous high‐resolution structures. Refinement incorporating HOmology DErived Restraints (HODER), improves geometrical quality and the fit to the diffraction data for many low‐resolution structures. To make these improvements readily available to the general public, we applied our new algorithms to all crystallographic structures in the PDB: using massively parallel computing, we constructed a new instance of the PDB‐REDO databank (https://pdb-redo.eu). This resource is useful for researchers to gain insight on individual structures, on specific protein families (as we demonstrate with examples), and on general features of protein structure using data mining approaches on a uniformly treated dataset.

Molecular dynamics simulations reveal that the hydrophobic cavity in human cytokine Interleukin-1beta is hydrated and can dynamically accommodate between one and four water molecules. These waters have residence times >> 500 ps and can give rise to detectable NOEs, in agreement with NMR observations of Ernst et al. (Science 1995; 267:1813-1817). The waters also display high positional disorder within the cavity, which explains why they have not been resolved crystallographically. The average distribution of water molecules over time within the cavity matches well the low resolution electron density extracted by Yu et al. (Proc Natl Acad Sci 1999; 96:103-108). The water molecules hydrate the hydrophobic cavity preferentially as complex clusters. These clusters result from a combination of hydrogen bonds between the waters and stabilizing interactions between the waters and aromatic rings forming the cavity. Free energy estimates suggest that it takes 4-waters to hydrate the cavity in a thermodynamically stable manner leading to a gain in free energy of transfer from bulk of approximately approximately 3.6 kcal/mol. This arises from the existence of the water clusters in multiple hydrogen bonded states. In addition, the waters are found to migrate either individually or as clusters out of the cavity through several pathways. The upper limit for one-dimensional diffusion of the waters within the protein matrix is 4 A/ps (relative to 6 A/ps for bulk). Simulations reveal pathways in addition to those identified crystallographically, with motions controlled by the rotations of sidechains. We find that only when the hydrophobic cavity is hydrated, do correlated motions couple distant sites with the sites that make contact with the receptor and this data partly offers an explanation of experimental mutagenesis data. Simulations, together with recent observations based on mutagenesis by Heidary et al. (J Mol Biol 2005; 353:1187-1198) that hydrogen bond networks couple motions across long distances in interleukin-1beta, lead us to hypothesize that the hydration of the cavity (conserved across mammals) can thermodynamically enhance hydrogen bond networks to enable coupling across long distances by acting as a plug and this in turn enables a kinetic control of the rate of transmission of signals.

Publisher's copyright statement:Additional information: Use policyThe full-text may be used and/or reproduced, and given to third parties in any format or medium, without prior permission or charge, for personal research or study, educational, or not-for-pro t purposes provided that:• a full bibliographic reference is made to the original source • a link is made to the metadata record in DRO • the full-text is not changed in any way The full-text must not be sold in any format or medium without the formal permission of the copyright holders.Please consult the full DRO policy for further details. We present a conformational factorization approach. The theory is based on a superposition partition function, where the partition function is written as a sum over contributions from local minima. The factorisation greatly reduces the number of minima that need to be considered, by employing the same local configurations for groups that are sufficiently distant from the binding site. The theory formalises the conditions required to analyse how our definition of the binding site region affects the free energy difference between the apo and holo states. We employ basin-hopping parallel tempering to sample minima that contribute significantly to the partition function, and calculate the binding free energies within the harmonic normal mode approximation. A further significant gain in efficiency is achieved using a recently developed local rigid body framework in both the sampling and the normal mode analysis, which reduces the number of degrees of freedom. We benchmark this approach for human aldose reductase (PDB code 2INE). When varying the size of the rigid region, the free energy difference converges for factorisation of groups at a distance of 14Å from the binding site, which corresponds to 80% of the protein being locally rigidified. This approach is likely to be useful for estimating the binding free energy of protein-ligand complexes.