Markov State Models for Electrostatic Channeling in Bioelectrocatalysis

Abstract

Multistep biocatalytic cascades benefit from channeling mechanisms that guide intermediate transport between active sites. Channeling approaches such as electrostatic interactions and steric confinement are best exemplified in natural biocatalytic complexes such as tryptophan synthase and malate dehydrogenase–citrate synthase, but are also amenable to application in de novo cascades designed to accomplish chemical conversion and energy production.We have recently conducted an extensive computational study of the Malate Dehydrogenase–Citrate Synthase (MDH-CS) complex, which displays a positively charged patch on its surface, capably of electrostatically channeling a negatively-charged oxaloacetate intermediate. Our collaborators have demonstrated experimentally that mutation of residues in the positively-charged patch can strongly effect intermediate transport, as measured by lag time, the time required for the second of a two-reaction cascade to approach steady state [1].Using classical molecular dynamics approaches to study the transport of oxaloacetate in the presence of the charged surface, we created a Markov State model that effectively maps the prevalence of the intermediate across the enzyme surface as well as in bulk solution. Building the model involved seeding the possible states of the intermediate using metadynamics to force the intermediate out of low-energy states and to explore the overall energy landscape. Once assembled, this model enables us to study the dominant pathways of intermediate transport, for example contrasting transport along the charged enzyme surface with diffusion into the bulk solution. A “hub score” approach enabled identification of key residues that control intermediate traversal over the charged surface, and we are able to calculate pathway efficiency that is directly relatable to lag time [2].We will discuss a new, fast finite-difference approach to predict lag times for these multidimensional complexes. The steady-state transition matrix of the Markov state model is easily converted to a material balance for each node in the network. Using the model, the lag time of MDH-CS was determined computationally to be comparable to experiment for both the original and mutant complex. Using the model, the the dynamics of each of the four possible reaction pathways between the the two source (MDH) active sites and the two sink (CS) sites could be studied independently. This analysis provides a dynamic model for intermediate transport in an electrostatically channeled system, and can be used as a predictive tool to provide mechanistic insight into path dominance. B. Bulutoglu, K. E. Garcia, F. Wu, S. D. Minteer and S. Banta, ACS Chem. Bio., 11, 2847–2853 (2016). doi:10/f9c5cpY. Xie, S. D. Minteer, S. Banta and S. C. Barton, ACS Nanosci. Au, 2, 414 (2022). doi:10/gqkvtq