The interest in the use of redox enzymes for the construction of efficient biodevices has grown to achieve an environmentally friendly society. “Bioelectrocatalysis,” in which enzymatic reaction and electrode reactions are coupled, is a fundamental technology for various electrochemical biomimetics (e.g., biosensors, biofuel cells, and bioreactors). In particular, the reaction in which an enzyme directly shuttles electrons to an electrode without any external electron mediators is called a direct electron transfer (DET)-type reaction. Thanks to its mediator-less configuration, DET-type reaction is advantageous in biocompatibility and design freedom, enabling us to develop ideal bio-devices. However, the number of enzymes performing DET-type reactions (DET-type enzyme) is very limited, which is one of the major problems for developing the aforementioned biodevices. Thus, there has been considerable interest in studies for creating new DET-type enzymes using existing enzymes as templates. To tackle the issue, fundamental research on their catalytic reaction mechanisms is essential.D-Fructose dehydrogenase (FDH) from Gluconobacter japonicus NBRC3260, a membrane-bound heterotrimeric flavohemoprotein capable of intense DET-type bioelectrocatalysis, has been widely investigated. FDH forms heterotrimetric structures composed of the catalytic large subunit, the chaperonic small subunit, and the membrane-bound cytochrome c subunit. Although the catalytic center is flavin adenine dinucleotide (FAD), the details for substrate oxidation remain unclear. Thanks to its extremely high DET-type activity and its covalently-bound cofactors, FDH is regarded as the model DET-type enzyme. Several researchers have already revealed the enzyme properties, focusing on DET from the viewpoints of enzyme engineering and electrochemistry. In addition, the three-dimensional (3D) structure of FDH was first revealed in 2022 with cryo-electron microscopy (cryo-EM) analysis, enabling us to discuss the enzyme from the perspective of structural biology and bioinformatics. In this study, we intend to understand the catalytic mechanism of FDH, such as substrate recognition or catalysis.First, we performed enzyme-substrate docking simulation and homology search to estimate critical amino acid residues in the catalytic reaction. These results indicated that three amino acid residues around FAD (N1146, H1147, and N1190) were the critical residues. Site-directed FDH variants focused on the residues (Namely, N1146A, N1146Q, H1147A, and N1190A) were expressed, purified, and evaluated.Next, the electrochemical properties of variants were evaluated with enzyme-modified rotating disk electrodes at 2000 rpm, pH 4.5, and 25 °C. From the cyclic voltammogram, we found that the mutations to H1147 or N1190 brought remarkable declines in DET-type activities, implying the importance of the two residues for the catalytic reaction. Substrate concentration dependence of the variants on the catalytic currents (Michaelis-Menten plot) revealed that the mutation into N1146 or H1147 resulted in increases in Michaelis constants, which indicates that the two residues seemed to have roles in fructose recognition. Because the two N1146 variants (N1146A or N1146Q) maintained sufficient activities, we examined substrate characteristics for various sugars and confirmed that the relative activity of N1146Q with D-tagatose, the C4 epimer of D-fructose, was improved over that of recombinant FDH.Finally, we also discuss the properties of variants from structural biology. The structures of the variants were successfully analyzed with cryo-EM analysis, and the N1146A, N1146Q, H1147A, and N1190A resolutions were 2.4, 3.1, 2.8, and 3.0 Å, respectively. The overall quaternary structures remained almost unchanged, indicating that each point mutation did not disrupt the protein structure. The direction and position of the amino acid residues of each variant differed slightly around the mutation sites. When using their structure for enzyme-substrate docking simulation, a good relationship between docking scores and Michaelis constants of the variants was observed. This means that the decline in the affinity between the variants and the substrate can be explained in structural biology. In the simulation, on the other hand, we obtained different results when using in-silico variants (computationally constructed), implying room for improvement in the current protein structure prediction methods.In summary, we investigated the mechanisms underlying the catalytic activity of FDH using enzyme engineering, electrochemistry, structural biology, and bioinformatics. H1147 and N1190 are particularly important for catalytic activity, with H1147 mainly functioning as a basic catalyst. N1146, H1147, and N1190 contribute to the recognition of the fructose molecules. In the future, our understanding of the fundamental reaction mechanism is expected to lead to the computational design and creation of various DET-type enzymes with FDH as a template.
Read full abstract