Pyrococcus furiosus alcohol dehydrogenase D (AdhD) is a highly thermostable member of the well-characterized aldo-keto reductase (AKR) superfamily. Its preference for secondary alcohols and NAD(H) combined with promiscuous activity towards simple sugars, primary alcohols, and aldehydes makes it an attractive protein engineering candidate for applications such as biofuel cells and the development of artificial metabolic pathways. Here we present a kinetic analysis of the wild type enzyme and a double mutant with increased activity and broadened specificity.

Most known AKRs exhibit a strong preference for NADP(H) over NAD(H), while AdhD has a strong preference for NAD(H), and little detectible activity with NADP(H). To examine the cause of this reversal in specificity, site-directed mutagenesis was used to revert residues in the cofactor binding pocket to the consensus residues seen in the AKR superfamily. Changing a histidine residue involved in binding the adenine half of the cofactor back to the consensus arginine residue (in NADP(H) binding AKRs) resulted in an order of magnitude improvement in activity with NADP(H). Additionally, a lysine to glycine mutation in the cofactor binding pocket increased activity with both cofactors. Both the lysine in the wild type enzyme and the consensus arginine present in the mutant have been shown to form salt bridges with the 2'-phosphate of NADP(H), while the histidine seen in the wild type enzyme has been shown to stack with the adenine ring of either cofactor in a separate AKR mutant. The ionic interaction between the consensus arginine and the 2'-phosphate of NADP(H) is also associated with an observable fluorescence kinetic transient due to a multi-step rearrangement that takes place upon cofactor binding.

In order to understand the observed differences in activity, stopped-flow fluorescence spectroscopy and steady-state kinetics were used to determine the microscopic rate constants corresponding to each step in the reaction mechanism. Thus a direct comparison can be made between the wild type and mutant enzyme and the impacted steps identified. This information will be used to guide further engineering of cofactor specificity and catalytic activity.

Mutagenesis will also be performed within the substrate binding pocket in an attempt to improve activity with sugars or primary alcohols. Substrate specificity in the AKR superfamily is tailored through modification of surface loops near the active site. Multiple sequence alignments and homology modeling indicate these loops are significantly truncated in AdhD, and likely contribute to its thermostability and broad specificity. Typically these loops are necessary to differentiate between various steroids and other large substrates, and their reduced size should not significantly impact the affinity for smaller primary alcohols and monosaccharides. We report on the kinetic parameters for wild-type AdhD and mutants with improved activities with these substrates. The results of this work will be applied to the continued evolution of novel specificities for this enzyme.

Extended Abstract Status: Not Uploaded