Novel in silico Molecular Modeling for the Enhancement of Enzyme Activity

We proposed an effective molecular modeling concept for the enhancement of enzyme activity. The molecular dynamics of enzyme, substrate and cofactor was estimated with the CHARMM (Chemistry at HARvard Molecular Mechanics) and Generalized Born with a simple SWitching (GBSW). All variation of amino acid substitutions were introduced at the directly reacting residues with cofactor and the theoretical Delta Delta G values for each of the mutations was calculated. Also the correlation between changed surface area by the substitution and the Delta Delta G values was examined. Here we hypothesized that mutant enzymes revealed high Delta Delta G value at the same time does not show difference of surface area with wild type have higher activity than wild type enzyme. Therefore selected candidates were expressed in Escherichia coli, purified and its activity experimentally measured.

To verify the hypothesis, we used two enzymes, succinyl-CoA synthetase (SCS) from Escherichia coli and formate dehydrogenase (FDH) from Thiobacillus sp. SCS catalyzes a reversible reaction in the tricarboxylic acid (TCA) cycle. Succinate, CoA and NTP (N: adenosine or guanosine) were converted to succinyl-CoA, NDP and Pi by SCS and vice versa. Succinyl-CoA can be used as a precursor for the production of 5-aminolevulinate, porphyrins, heme, and chrorophyll. Therefore, increasing activity of SCS for succinyl-CoA production is needed for industrial usage. We discovered two SCS mutant enzymes that increased the production of succinyl-CoA two-fold than wild type. In addition, FDH is a NADH-dependent redox enzyme converting carbon dioxide and formate by ping-pong reaction. Oxidation activity of FDH to convert formate to carbon dioxide with NAD⁺ is stronger than reduction activity to convert formate to carbon dioxide with NADH. However, if its reductivity can be increased by our in silico molecular modeling, economical benefit is expected such as reducing greenhouse gas and production of valuable organic chemical. As a result, we found a mutant FDH with 60 percent higher reductivity than wild type. This novel in silico molecular modeling can be applicable to design novel enzymes with enhanced activity.