Truncating the C terminus of formate dehydrogenase leads to improved preference to nicotinamide cytosine dinucleotide

Abstract

Formate dehydrogenase (FDH) is widely applied in regeneration of redox cofactors. There are continuing interests to engineer FDH for improved catalytic activity and cofactor preference. In the crystal structure of FDH from Pseudomonas sp. 101 (pseFDH), the C terminus with 9 amino acid residues cannot be resolved. However, our earlier work showed mutations at C terminus led pseFDH variants to favor a non-natural cofactor nicotinamide cytosine dinucleotide (NCD). Here, we investigated the role of C-terminal residues on cofactor preference by truncating their corresponding C terminus of pseFDH variants. Sequence comparison analysis showed that C-terminal residues were barely conservative among different FDHs. pseFDH and mutants with their C termini truncated were constructed, and the resulted variants showed improved preference to NCD mainly because NAD-dependent activity dropped more substantially. Further structure analysis showed that these pseFDH variants had their cofactor binding domains reconstructed to favor molecular interactions with NCD. Our work indicated that C-terminal residues of pseFDH affected enzyme activity and cofactor preference, which provides a new approach for ameliorating the performance of redox enzymes.

Keywords: C terminus; Cofactor preference; Directed evolution; Formate dehydrogenase; Non-natural cofactor.

Fig. 1

Structures of NAD and NCD (a) and the application of FDH in NADH regeneration (b).

Fig. 2

Analysis of C-terminal residues of pseFDH and mutants. Mutant A2, C256I/E261P/S381I; Mutant 3A3, V198I/C256I/P260S/E261P/S381N/S383F. The ligand in WT is NAD, and the ligand in mutant A2 or 3A3 is NCD.

Fig. 3

Sequence alignment of FDH from different sources and structure analysis. a, pseFDH, FDH from Pseudomonas sp. 101 (UniProtKB/Swiss-Prot: P33160.3). MorFDH, FDH from Moraxella sp.C2 (PDB: 2GSD); MvaFDH, FDH from Mycolicibacterium vaccae (GenBank: AAB36206.1); LtuFDH, FDH from Legionella tucsonensis (WP_058520322.1); FduFDH, FDH from Fluoribacter dumoffii (WP_010653742.1); SnoFDH, FDH from Starkeya novella (WP_013168047.1); OsaFDH, FDH from Oryza sativa (GenBank: BAA77337.1); VumFDH, FDH from Vigna umbellata (GenBank: ALI89099.1); GsoFDH, FDH from Glycine soja (GenBank: KHN47171.1); ZmaFDH, FDH from Zea mays (GenBank: ACG39798.1); CboFDH, FDH from Candida boidinii (GenBank: ABE69165.2); RsoFDH, FDH from Rhizoctonia solani 123E (GenBank: KEP50943.1); SceFDH, FDH from Saccharomyces cerevisiae S288C (NP_015033.1); TruFDH, FDH from Trichophyton rubrum (GenBank: OAL65030.1). b, Crystal structure of pseFDH (white) and pseFDH model (cyan) with residue 2 to 400. c, Highlighted C terminus of chain A in pseFDH crystal structure.

Fig. 4

Cofactor binding domains of pseFDH variants (a-c) and molecular interaction analysis (d-f). Domain and interactions analysis are shown based on homologous modeling and molecular docking.

Fig. 5

Interactions analysis of pseFDH variants with cytosine ribose of NCD. a, Crystal structure profile of N-3A3 (PDB: 6JUJ); b and c, Structure analysis based on homologous models of N-3C4 and N-3C4-392. Hydrogen bonds are shown with dotted green lines and are measured in angstrom.