Measuring the effects of reaction coordinate and electronic treatments in the QM/MM reaction dynamics of Trypanosoma cruzi trans-sialidase
Doctoral Thesis
2016
Permanent link to this Item
Authors
Supervisors
Journal Title
Link to Journal
Journal ISSN
Volume Title
Publisher
Publisher
University of Cape Town
Department
Faculty
License
Series
Abstract
The free energy of activation, as defined in transition state theory, is central to calculating reaction rates, distinguishing between mechanistic paths and elucidating the catalytic process. Computational free energies are accessible through the reaction space that is comprised of the conformational and electronic degrees of freedom orthogonal to the reaction coordinate. The overarching aim of this thesis was to address theoretical and methodological challenges facing current methods for calculating reaction free energies in glycoenzyme systems. Tractable calculations balance chemical accuracy and sampling efficiency that necessitates simplification of these complex reaction spaces through quantum mechanics/molecular mechanics partitioning and use of a semi-empirical electronic method to sample an approximated reaction coordinate. Here I directly and indirectly interrogate both the appropriate levels of sampling as well as the accuracy of the semi-empirical method required for reliable analysis of glycoenzyme reaction pathways. Free Energies from Adaptive Reaction Coordinates Forces, a method that builds the potential of mean force from multiple iterations of reactive trajectories, was used to construct reaction surfaces and volumes for the glycosylation and deglycosylation reactions comprising the T. cruzi trans-sialidase catalytic itinerary. This enzyme was chosen for the wealth of experimental data available for it built from its significance as a potential drug target against Chagas disease. Of equal importance is the identification of an elimination reaction competing with the primary transferase activity. The identification of this side reaction, that is observable only in the absence of the trans-sialidase or sialic acid acceptor, presented the opportunity to study the means by which enzymes selectivity bias in favor of a single reaction path. I therefore set out to explore the molecular details of how T. cruzi transsialidase asserts a precision and selectivity synonymous with enzyme catalysis. The chemical nature of the transition sate, formally defined as a dividing hypersurface separating the reactant and product regions of phase space, was characterized for the deglycosylation reaction. More than 40 transition state configurations were isolated from reactive trajectories, and the sialic acid substrate conformations were analyzed as well as the substrate interactions with the nucleophile and catalytic acid/base. A successful barrier crossing requires that the substrate pass through a family of E₅, ⁴H₅ and ⁶H₅ puckered conformations, all of which interact slightly differently with the enzyme. This work brings new evidence to the prevailing premise that there are several pathways from reactant to product passing through the saddle and successful product formation is not restricted to the minimum energy path. Increasing the reaction space with use of a multi-dimensional (3-D) reaction coordinate allowed simultaneous monitoring of the hitherto unexplored competition between a minor elimination reaction and the dominant displacement reaction present in both steps of the catalytic cycle. The dominant displacement reactions display lower barriers in the free energy profiles, greater sampling of favorable reactant stereoelectronic alignments and a greater number of possible transition paths leading to successful crossing reaction trajectories. The effects on the electronic degrees of freedom in reaction space were then investigated by running density functional theory reactive trajectories on the semi-empirical free energy. In order to carry out these simulations Free Energies from Adaptive Reaction Coordinates Forces was ported as a Fortran 90 library that interfaces with the NWChem molecular dynamics package. The resulting B3LYP/6-31G/CHARMM crossing trajectory provides a molecular orbital description of the glycosylation reaction. Direct investigation of the underlying potential energy functions for B3LYP/6-31G(d), B3LYP/6-31G and SCC-DFTB/MIO point to the minimal basis set as the primary limitation in using self-consistent charge density functional tight binding as the quantum mechanical model for modeling of enzymatic reactions transforming sialic acid substrates.
Description
Keywords
Reference:
Rogers, I. 2016. Measuring the effects of reaction coordinate and electronic treatments in the QM/MM reaction dynamics of Trypanosoma cruzi trans-sialidase. University of Cape Town.