Distinct Protein Hydration Water Species Defined by Spatially Resolved Spectra of Intermolecular Vibrations

Abstract

In this molecular dynamics simulation study, we analyze intermolecular vibrations in the hydration shell of a solvated enyzme, the membrane type 1–matrix metalloproteinase, with high spatial resolution. Our approach allows us to characterize vibrational signatures of the local hydrogen bond network, the translational mobility of water molecules, as well as the molecular entropy, in specific local environments. Our study demonstrates the heterogeneity of water properties within the hydration shell of a complex biomolecule. We define a classification scheme based on the vibrational density of states that allows us to distinguish separate classes of hydration water species and facilitates the description of hydration water properties at distinct hydration sites. The results demonstrate that no single characteristic of the protein surface is sufficient to determine the properties of nearby water. The protein surface geometry, quantified here by the number of protein atoms in the vicinity of a hydration water molecule, as well as the chemical nature of a solvated protein functional group, influences dynamic and thermodynamic properties of solvating water molecules.

Disassembling solvation free energies into local contributions—Toward a microscopic understanding of solvation processes

“…However, in addition to the thermodynamic information, the analysis implicitly provides detailed spatially resolved information on local dynamic properties of the solvent, such as short‐time diffusion (zero‐frequency response of the VDOS) and the spectrum of local vibrations. The latter has been used to classify distinct species of hydration water molecules in the vicinity of small molecular solutes as well as in the hydration shell of proteins …”

“…The minimum sampling rate for resolving these vibrations corresponds to 10-20 fs intervals, while previous implementations have used sampling rates as low as 8 fs. 21,52,53 In order to calculate the required time-correlation functions, a history of at least several hundred time frames is required, which eventually determines the frequency resolution of the VDOS that is central to the entropy evaluation. If the analysis is carried out as an a posteriori analysis of simulation trajectories, the required time resolution results in large disk space requirements for the storage of the trajectories that include coordinates and velocities.…”

“…The latter has been used to classify distinct species of hydration water molecules in the vicinity of small molecular solutes 21 as well as in the hydration shell of proteins. 52,53 6 | CONCLUSIONS In general, the information provided by the various methods described in this article provides vital insights into the interactions of molecular solutes with surrounding water. This information is crucial to formulate improved implicit solvent models, which are able to accurately describe solvent-mediated interactions beyond current solvent-accessible-surface area-based approximations that do not capture the impact of chemical context.…”

Solvent Entropy Contributions to Catalytic Activity in Designed and Optimized Kemp Eliminases

“…There has also been some analysis of the role of the solvent in mediating activity 11 . However, in the present study a novel spatially resolved variant of the two-phase thermodynamics approach [12][13][14] allows us to analyze the entropy of solvent molecules in great detail in the vicinity of the catalytic center and on the protein surface. We compare results obtained for the designed Kemp Eliminase enzymes and their mutated counterparts obtained from laboratory directed evolution (LDE).…”

“…In this study, we utilized a spatially resolved version of the 2PT method, which allows studying contributions of individual hydration sites to the solvation entropy of a solute (3D-2PT). 14 With this definition, we obtain contributions to the solvation entropy from a given region in the hydration environment of the enzyme/enzyme-substrate complex as a sum of averaged voxel contributions times the product of the local average water number density ̅ l ( ) and the volume V:…”