Predictive methods for simulating redox reactions and charge transfer in proteins - Project Summary/Abstract At Boston University (BU) the PI has developed a broad interdisciplinary computational and theo- retical chemistry program focused on photo-induced and electron-mediated processes. The program integrates method development and phenomenon-driven research closely tied to experiment through collaboration with experimental groups at BU and elsewhere. Below the main developments pertain- ing to the proposed program are summarized followed by an outline of the future research directions. The PI has developed a first principle protocol that enables predictive simulations of redox po- tentials of co-factors in proteins from the first principles. While electrostatic embedding QM/MM is still a standard in the field, the PI has shown that polarizable embedding is often critical for reliable estimates of redox potentials. The proposed scheme also correctly accounts for long-range electrostatic interactions, rectifying an oversight of earlier approaches. It has been shown that this model yields accurate estimates of redox potentials and electron transfer rates by using cryptochrome protein as an example. To enable fast high-throughput the PI has developed a web-based software, eMap, that enables automated mapping of ET channels in proteins only based on their crystal structure. eMap parses a PDB structure file, identifies all relevant redox-active moieties, constructs distance maps, and uses graph theory to identify the most efficient electron transfer pathways. The software is now being actively used by research groups outside BU for guiding the experimental studies as evidenced by citations. In the next five years, the PI plans to push the frontiers of the predictive simulations with the final goal of building an umbrella platform that enables simulations of redox chemistry and electron transfer from the first unbiased qualitative predictions to accurate quantification of observables. The PI plans to take eMap analysis to a new level by enabling large-scale screening of protein families for shared electron transfer pathways to explore the role of conserved electron transfer channels for protein functions’ evolution. The PI also will enhance the predictive power of eMap through the integration of the chemically-relevant energetic parameters without sacrificing the efficacy of the method. The PI will be addressing the remaining challenges in the quantitative description of the redox potentials of co-factors in proteins as well as the energetics of biological charge transport. These advances together with the prior developments will allow the PI to bring the computational program in the research group to the next level by significantly expanding the exploratory mechanistic computational studies enabled by new software functions and level of accuracy.
