Structure-Based Engineering of an Artificially Generated NADP+-Dependent d-Amino Acid Dehydrogenase

Abstract

A stable NADP⁺-dependent d-amino acid dehydrogenase (DAADH) was recently created from Ureibacillus thermosphaericusmeso-diaminopimelate dehydrogenase through site-directed mutagenesis. To produce a novel DAADH mutant with different substrate specificity, the crystal structure of apo-DAADH was determined at a resolution of 1.78 Å, and the amino acid residues responsible for the substrate specificity were evaluated using additional site-directed mutagenesis. By introducing a single D94A mutation, the enzyme's substrate specificity was dramatically altered; the mutant utilized d-phenylalanine as the most preferable substrate for oxidative deamination and had a specific activity of 5.33 μmol/min/mg at 50°C, which was 54-fold higher than that of the parent DAADH. In addition, the specific activities of the mutant toward d-leucine, d-norleucine, d-methionine, d-isoleucine, and d-tryptophan were much higher (6 to 25 times) than those of the parent enzyme. For reductive amination, the D94A mutant exhibited extremely high specific activity with phenylpyruvate (16.1 μmol/min/mg at 50°C). The structures of the D94A-Y224F double mutant in complex with NADP⁺ and in complex with both NADPH and 2-keto-6-aminocapronic acid (lysine oxo-analogue) were then determined at resolutions of 1.59 Å and 1.74 Å, respectively. The phenylpyruvate-binding model suggests that the D94A mutation prevents the substrate phenyl group from sterically clashing with the side chain of Asp94. A structural comparison suggests that both the enlarged substrate-binding pocket and enhanced hydrophobicity of the pocket are mainly responsible for the high reactivity of the D94A mutant toward the hydrophobic d-amino acids with bulky side chains.IMPORTANCE In recent years, the potential uses for d-amino acids as source materials for the industrial production of medicines, seasonings, and agrochemicals have been growing. To date, several methods have been used for the production of d-amino acids, but all include tedious steps. The use of NAD(P)⁺-dependent d-amino acid dehydrogenase (DAADH) makes single-step production of d-amino acids from oxo-acid analogs and ammonia possible. We recently succeeded in creating a stable DAADH and demonstrated that it is applicable for one-step synthesis of d-amino acids, such as d-leucine and d-isoleucine. As the next step, the creation of an enzyme exhibiting different substrate specificity and higher catalytic efficiency is a key to the further development of d-amino acid production. In this study, we succeeded in creating a novel mutant exhibiting extremely high catalytic activity for phenylpyruvate amination. Structural insight into the mutant will be useful for further improvement of DAADHs.

Keywords: NADP; Ureibacillus thermosphaericus; d-amino acid; d-phenylalanine; dehydrogenases; meso-diaminopimelate; phenylpyruvate.

FIG 1

Overall structure of U. thermosphaericus DAADH and close-up view of the C-terminal tail. (A) Overall structure of the apo-DAADH dimer. The dinucleotide-binding domain, dimerization domain, and C-terminal domain in one subunit are shown in cyan, green, and magenta, respectively. The C-terminal tail is in red. The adjacent subunit is in white. Residues 1, 155 to 157, 223, and 224 in subunit A and residue 1 in subunit B were disordered and not visible in the electron-density map. (B) C-terminal region of C-terminal His-tagged U. thermosphaericus DAPDH. (C) C-terminal region of nontagged U. thermosphaericus DAADH. (B and C) The C-terminal tail from the adjacent subunit (subunit A) is in yellow, and the ion-pair interactions around the C-terminal residues are shown as dotted lines.

FIG 2

Stereographic close-up of the meso-DAP-binding site in S. thermophilum DAPDH. (A) Hydrogen bonds (dotted lines) around the l-amino acid center of meso-DAP. (B) Hydrogen bonds around the d-amino acid center. The structure of NADPH–meso-DAP-bound S. thermophilum DAPDH (11) (subunit B; white residues and black labels with “/S”) is superimposed on that of U. thermosphaericus apo-DAADH (subunit B; green residues and red labels with “/U”).

FIG 3

Structure around the substrate-binding site in the D94A-Y224F mutant. The final model of the NADP⁺-bound D94A-Y224F mutant (PDB code 5GZ3) was composed of amino acid residues 2 to 326 in each subunit (residues 221 to 226 in subunit B were disordered and not visible in the electron-density map), two NADP⁺ molecules, four ethylene glycol molecules, and 370 water molecules. The final model of the NADPH-KACA-bound D94A-Y224F mutant (PDB code 5GZ6) was composed of amino acid residues 2 to 326 in each subunit (residues 156 to 157 and 259 to 263 in subunit A and 156 to 157, 245 to 246, and 259 to 263 in subunit B were disordered and not visible in the electron-density map), one NADPH molecule, one KACA molecule (see supplemental material), one acetate ion, two sulfate ions, and 397 water molecules. (A) The structure of the NADP⁺-bound D94A-Y224F mutant (subunit A; white) is superimposed on that of the NADPH-KACA-bound D94A-Y224F mutant (subunit A). Loops 1 to 4 in the NADPH-KACA-bound D94A-Y224F mutant are shown in cyan, green, purple, and orange, respectively. Phe224, Lys150, Met247, and Asn276 are shown as stick models. (B) Stereographic close-up of the KACA-binding site in the D94A-Y224F mutant. KACA (yellow) and NADPH (magenta) molecules are shown as stick models. The final σ_A-weighted (F_o – F_c) omitted electron-density map for KACA is shown at the 2.1σ level. The hydrogen bonds around KACA are shown as dotted lines.

FIG 4

Proposed model for phenylpyruvate binding. The NADPH-KACA-bound D94A-Y224F mutant and phenylpyruvate-bound model are shown in white and cyan, respectively. Phenylpyruvate and NADPH molecules in the phenylpyruvate-bound model are in green and magenta, respectively. The hydrogen bonds around phenylpyruvate are shown as dotted lines.