Logo Logo
Help
Contact
Switch language to German
Quantum-chemical investigations of enzymatic reaction mechanisms and benchmarking of range-separated density-functional theory within the random phase approximation
Quantum-chemical investigations of enzymatic reaction mechanisms and benchmarking of range-separated density-functional theory within the random phase approximation
Enzymes are efficient biocatalysts, which are not only essential for living organisms but are also employed for industrial syntheses of chemical compounds. Insights into the exact enzymatic reaction mechanism is valuable for many fields, e.g., controlling the reaction in industrial settings, or development of enzyme inhibitors. Furthermore, it also contributes to a comprehensive understanding of the entire biochemical process around the enzyme. The first part of this thesis deals with the elucidation of the catalytic mechanism of two enzymes. In a first project, the glycosylase activity of the bacterial DNA-glycosylase Fpg is investigated. An alternative, base-independent excision mechanism is found for the substrate 8OG. It is a variation of the previously found ribose-protonated excision mechanism that does not involve the excised base. With this ’base-independent’ ribose-protonated mechanism, Fpg can excise 8OG in both syn- and anti-bound conformations. This is in contrast to the previously found ’base-specific’ mechanism, which only proceeds with syn-bound 8OG. The energy profile of the rate-determining step of the base-independent excision mechanism, obtained with QM/MM calculations with a significant number of atoms in the QM sphere, results in a reaction barrier of similar height for syn- and anti-8OG, which is in good agreement with experimental measurements. From this it can be concluded that Fpg has no preference between syn- and anti-8OG in the base-independent excision mechanism. In a second project, the decarboxylation mechanism of 5caU by the enzyme IDCase is examined. For this purpose, QM/MM energy profiles of possible decarboxylation pathways are compared. The comparison reveals that the reaction catalyzed by IDCase most likely follows a direct decarboxylation mechanism. Detailed investigations on this mechanism confirm that the direct decarboxylation of 5caU by IDCase is a one-step mechanism with simultaneous proton transfer and the C-C bond opening. The description of an energy profile of a reaction mechanism requires an accurate method that performs well for thermodynamic and kinetic properties, as well as for non-covalent interactions. Semi-local and hybrid density functional theory rather gives varying results, especially for the description of reaction barrier heights. Range-separated hybrid DFT, based on a short-range PBE exchange-correlation functional, in combination with long-range random phase approximation correlation (RSHPBE+lrRPA) gives promising results in some small-scale studies and shows a relatively fast basis set size convergence. In the second part of this thesis RSHPBE+lrRPA is benchmarked on a large-scale to be able to assess its potential. Therefore RSHPBE+lrRPA is applied on the GMTKN55 data set. The results of the benchmark reveal that indeed a moderately sized triple-ζ basis set is mostly sufficient for RSHPBE+lrRPA. Furthermore, RSHPBE+lrRPA shows a stable performance over the complete test set with less fluctuations in the accuracy between the subsets than standard RPA. According to its results on the GMTKN55 data set, RSH-PBE+lrRPA is comparable to a double-hybrid functional without empirical dispersion correction.
8-Oxoguanine, Formamidopyrimidine DNA Glycosylase, base excision, Iso-Orotate Decarboxylase, decarboxylation, QM/MM, range-separated DFT, Random Phase Approximation
Kreppel, Andrea
2020
English
Universitätsbibliothek der Ludwig-Maximilians-Universität München
Kreppel, Andrea (2020): Quantum-chemical investigations of enzymatic reaction mechanisms and benchmarking of range-separated density-functional theory within the random phase approximation. Dissertation, LMU München: Faculty of Chemistry and Pharmacy
[img]
Preview
PDF
Kreppel_Andrea.pdf

13MB

Abstract

Enzymes are efficient biocatalysts, which are not only essential for living organisms but are also employed for industrial syntheses of chemical compounds. Insights into the exact enzymatic reaction mechanism is valuable for many fields, e.g., controlling the reaction in industrial settings, or development of enzyme inhibitors. Furthermore, it also contributes to a comprehensive understanding of the entire biochemical process around the enzyme. The first part of this thesis deals with the elucidation of the catalytic mechanism of two enzymes. In a first project, the glycosylase activity of the bacterial DNA-glycosylase Fpg is investigated. An alternative, base-independent excision mechanism is found for the substrate 8OG. It is a variation of the previously found ribose-protonated excision mechanism that does not involve the excised base. With this ’base-independent’ ribose-protonated mechanism, Fpg can excise 8OG in both syn- and anti-bound conformations. This is in contrast to the previously found ’base-specific’ mechanism, which only proceeds with syn-bound 8OG. The energy profile of the rate-determining step of the base-independent excision mechanism, obtained with QM/MM calculations with a significant number of atoms in the QM sphere, results in a reaction barrier of similar height for syn- and anti-8OG, which is in good agreement with experimental measurements. From this it can be concluded that Fpg has no preference between syn- and anti-8OG in the base-independent excision mechanism. In a second project, the decarboxylation mechanism of 5caU by the enzyme IDCase is examined. For this purpose, QM/MM energy profiles of possible decarboxylation pathways are compared. The comparison reveals that the reaction catalyzed by IDCase most likely follows a direct decarboxylation mechanism. Detailed investigations on this mechanism confirm that the direct decarboxylation of 5caU by IDCase is a one-step mechanism with simultaneous proton transfer and the C-C bond opening. The description of an energy profile of a reaction mechanism requires an accurate method that performs well for thermodynamic and kinetic properties, as well as for non-covalent interactions. Semi-local and hybrid density functional theory rather gives varying results, especially for the description of reaction barrier heights. Range-separated hybrid DFT, based on a short-range PBE exchange-correlation functional, in combination with long-range random phase approximation correlation (RSHPBE+lrRPA) gives promising results in some small-scale studies and shows a relatively fast basis set size convergence. In the second part of this thesis RSHPBE+lrRPA is benchmarked on a large-scale to be able to assess its potential. Therefore RSHPBE+lrRPA is applied on the GMTKN55 data set. The results of the benchmark reveal that indeed a moderately sized triple-ζ basis set is mostly sufficient for RSHPBE+lrRPA. Furthermore, RSHPBE+lrRPA shows a stable performance over the complete test set with less fluctuations in the accuracy between the subsets than standard RPA. According to its results on the GMTKN55 data set, RSH-PBE+lrRPA is comparable to a double-hybrid functional without empirical dispersion correction.