Thermostable Alcohol Dehydrogenase from Thermococcus kodakarensis KOD1 for Enantioselective Bioconversion of Aromatic Secondary Alcohols

Abstract

A novel thermostable alcohol dehydrogenase (ADH) showing activity toward aromatic secondary alcohols was identified from the hyperthermophilic archaeon Thermococcus kodakarensis KOD1 (TkADH). The gene, tk0845, which encodes an aldo-keto reductase, was heterologously expressed in Escherichia coli. The enzyme was found to be a monomer with a molecular mass of 31 kDa. It was highly thermostable with an optimal temperature of 90°C and a half-life of 4.5 h at 95°C. The apparent K_m values for the cofactors NAD(P)⁺ and NADPH were similar within a range of 66 to 127 μM. TkADH preferred secondary alcohols and accepted various ketones and aldehydes as substrates. Interestingly, the enzyme could oxidize 1-phenylethanol and its derivatives having substituents at the meta and para positions with high enantioselectivity, yielding the corresponding (R)-alcohols with optical purities of greater than 99.8% enantiomeric excess (ee). TkADH could also reduce 2,2,2-trifluoroacetophenone to (R)-2,2,2-trifluoro-1-phenylethanol with high enantioselectivity (>99.6% ee). Furthermore, the enzyme showed high resistance to organic solvents and was particularly highly active in the presence of H₂O–20% 2-propanol and H₂O–50% n-hexane or n-octane. This ADH is expected to be a useful tool for the production of aromatic chiral alcohols.

INTRODUCTION

Alcohol dehydrogenases (ADHs; EC 1.1.1.1), catalyzing reversible conversion of alcohols to their corresponding aldehydes or ketones, are widely distributed in the three domains of life and play an important role in many physiological processes (1). NAD(P)-dependent ADHs can be subdivided into three types: medium-chain zinc-containing ADHs, short-chain ADHs, and long-chain or iron-activated ADHs (1). Some of the NAD(P)-dependent dehydrogenases are related to another aldo-keto reductase (AKR) superfamily (2, 3), which has been proposed to exhibit divergent functions with the same active-site constellation (4).

In the last few decades, ADHs have been increasingly applied for the production of chiral alcohols that are important precursors or intermediates in the chemical and pharmaceutical industries, due to their remarkable chemo-, regio-, and enantioselectivities (5–7). The chiral alcohols can be obtained by either asymmetric reduction of the prochiral ketones or kinetic resolution of the racemic alcohols. Generally, the stability or robustness of the enzymes is of critical importance for reducing the cost of the practical operations, which often require organic solvents and elevated temperatures (6, 8). In this viewpoint, hyperthermophiles, the microorganisms that can optimally grow at temperatures higher than 80°C (9), are attractive sources for industrial enzymes (10–12). Besides thermostability, the enzymes from hyperthermophiles usually exhibit tolerance against denaturants such as detergents and organic solvents (12). All these features make them not only robust biocatalysts for practical applications but also interesting scientific subjects for understanding protein stabilization (13).

There have been many reports on thermostable ADHs from hyperthermophiles, but the substrates for most of them were limited to aliphatic alcohols/ketones and aromatic primary alcohols/aldehydes. These thermostable ADHs include iron-containing ADHs from Thermococcus litoralis (14), Thermococcus sp. strain ES-1 (15, 16), Thermococcus zilligii (17), Thermococcus hydrothermalis (18), and Thermotoga hypogea (19); zinc-containing ADHs from Pyrococcus furiosus (20), Sulfolobus solfataricus (21, 22), Sulfolobus tokodaii (23), and Thermococcus guaymasensis (24); short-chain ADHs from P. furiosus (25) and Thermococcus sibiricus (26); and an AKR from Thermotoga maritima (27). Despite the importance of aromatic chiral alcohols, such as (R)- and (S)-1-phenylethanols, as useful building blocks in pharmaceutical applications (5, 28), only four thermostable ADHs, namely, a zinc-containing ADH from Aeropyrum pernix (29, 30), two short-chain ADHs from Sulfolobus acidocaldarius (31, 32), and an AKR from P. furiosus (33), have been mentioned to be active toward aromatic ketones. Considering the value of aromatic chiral alcohols, it will be increasingly important to exploit more ADHs that can transform aromatic secondary alcohols or ketones with high enantioselectivity and thermostability from hyperthermophiles.

The hyperthermophilic archaeon Thermococcus kodakarensis KOD1 is an obligate anaerobic heterotroph that can grow optimally at 85°C (34). The genome information of this hyperthermophile (GenBank accession no. AP006878.1) (35) revealed the presence of several genes potentially encoding ADHs. However, the substrate spectrum of these probable ADHs cannot be determined only from the sequence information. In this study, we present the identification and characterization of an ADH showing activity toward aromatic secondary alcohols from T. kodakarensis (TkADH). TkADH could efficiently oxidize racemates of 1-phenylethanol and its derivatives with high enantioselectivity and exhibits extreme thermostability as well as high resistance to organic solvents. This study demonstrated the potential of TkADH as a versatile tool for the production of aromatic chiral alcohols.

MATERIALS AND METHODS

Chemicals.

General chemicals of analytical grade were obtained from Sigma-Aldrich (St. Louis, MO) or Acros Organics (Geel, Belgium). Restriction enzymes and other modifying enzymes were purchased from New England BioLabs (Beverly, MA) or Takara Bio (Otsu, Shiga, Japan).

Organisms, plasmids, and medium.

T. kodakarensis KOD1 was grown anaerobically at 85°C in a rich growth medium composed of a 1.25-fold dilution of artificial seawater (ASW; Senju Seiyaku, Osaka, Japan), 10 g · liter⁻¹ of yeast extract, and 5 g · liter⁻¹ of tryptone (ASW-YT). Into this medium (ASW-YT), 2 g · liter⁻¹ of elemental sulfur (ASW-YT-S⁰) or 5 g · liter⁻¹ of pyruvate (ASW-YT-Pyr) was added. pET-21b(+) (Novagen, Madison, WI) was used for overexpression of the gene encoding TkADH, and Escherichia coli Rosetta(DE3) (Novagen, Madison, WI) was used as the host strain. The recombinant E. coli cells were cultivated at 37°C or 30°C in Luria-Bertani (LB) medium containing 100 μg · ml⁻¹ of ampicillin and 34 μg · ml⁻¹ of chloramphenicol.

Purification of TkADH from the crude extract of T. kodakarensis KOD1.

T. kodakarensis KOD1 was cultivated for 16 h at 85°C in a total volume of 5 liters. The cells were harvested by centrifugation (5,000 × g for 10 min at 4°C), washed once with 0.8× ASW, resuspended in buffer A (50 mM Tris-HCl buffer, pH 8.5), and then disrupted by sonication for 2 min. After removal of debris by centrifugation (8,000 × g for 10 min at 4°C), a soluble protein extract was obtained by further centrifugation (15,000 × g for 20 min at 4°C) and was kept at 4°C prior to purification.

All purification steps were performed with an AKTA prime chromatography apparatus (GE Healthcare, Uppsala, Sweden) at room temperature. The soluble protein extract was loaded on a HiLoad 16/10 Q Sepharose HP column (20 ml; GE Healthcare) equilibrated with buffer A. Proteins were eluted with a linear gradient of 0.0 to 1.0 M NaCl, and the fractions at 0.4 to 0.6 M NaCl were pooled and then concentrated using a Vivaspin 20 centrifugal concentrator (Sartorius Stedim Biotech, Göttingen, Germany). The concentrated protein solution was further applied to a HiLoad 16/10 phenyl Sepharose HP column (20 ml; GE Healthcare) equilibrated with buffer A containing 1.33 M (NH₄)₂SO₄. A decreasing linear gradient of 1.33 to 0.0 M (NH₄)₂SO₄ was applied, and TkADH was eluted at 0 M (NH₄)₂SO₄. The fractions with ADH activity were pooled, concentrated, and then loaded on a Superdex 200 10/30 HR column (24 ml; GE Healthcare) equilibrated with buffer A containing 0.15 M NaCl. The peak fractions with ADH activity were collected and stored at 4°C for further analysis.

Determination of N-terminal amino acid sequence.

The purified TkADH was subjected to 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and electroblotted onto a polyvinylidene difluoride (PVDF) membrane. A membrane piece for the purified enzyme was subjected to a protein sequencer (Shimadzu PPSQ-21; Shimadzu Co., Kyoto, Japan) according to the manufacturer's instructions.

Expression of TkADH gene in E. coli and purification of recombinant enzyme.

DNA manipulations were carried out according to general procedures. The gene encoding TkADH (tk0845), identified from the N-terminal sequence of the purified protein, was amplified from T. kodakarensis genomic DNA with a primer pair of TK0845-F (5′-GCCGTCGCATATGAAGAAGGTTAGGATTTTTAACG-3′) and TK0845-R (5′-CTGGAATTCTCAGACACACCCCCTTGCGTTCT-3′) (the underlined sequences indicate NdeI in TK0845-F and EcoRI in TK0845-R). The amplified fragment of 837 bp was digested by NdeI and EcoRI and then ligated with pET-21b(+) at the corresponding sites. The resulting expression plasmid was designated pETTK0845.

Recombinant E. coli Rosetta(DE3) harboring pETTK0845 was cultivated at 37°C in 1 liter LB medium containing antibiotics. After the optical density at 600 nm reached 0.6, isopropyl-β-d-thiogalactopyranoside (IPTG) was added to a final concentration of 0.1 mM. Following further cultivation for 17 h at 30°C, the cells were harvested by centrifugation (8,000 × g for 10 min at 4°C), washed and resuspended in 50 mM Tris-HCl buffer (pH 8.0), and then disrupted by sonication for 5 min. The supernatant obtained after centrifugation (15,000 × g for 20 min at 4°C) was heated at 85°C for 20 min and subsequently centrifuged for 10 min at 15,000 × g. The supernatant was applied on a HiLoad 16/10 Q Sepharose HP column (20 ml) equilibrated with 50 mM Tris-HCl buffer (pH 8.0), and the proteins were eluted with a linear gradient of 0.0 to 1.0 M NaCl. The active peak fractions of TkADH that eluted at 0.4 to 0.6 M NaCl were combined and stored at −80°C until further use.

ADH assay.

The activity of the native TkADH at each purification step was assayed by determining the consumption of (RS)-1-phenylethanol in the cofactor-regenerating reaction system with thermostable NAD(P)H oxidase from T. kodakarensis (TkNOX) (36). The reaction mixture was composed of 50 mM Tris-HCl buffer (pH 8.5), 20 mM (RS)-1-phenylethanol, 2 mM NAD(P)⁺, 0.17 mM flavin adenine dinucleotide (FAD), 0.36 U · ml⁻¹ of TkNOX, and an appropriate amount of the native TkADH in a total volume of 0.6 ml. The reactions were carried out at 50°C for 3 h in a capped 15-ml tube with shaking at a rate of 150 strokes · min⁻¹. The (R)- and (S)-1-phenylethanol remaining after the reaction was quantified by high-performance liquid chromatography (HPLC) as described previously (36). One unit of ADH activity corresponds to the oxidation of 1 μmol (RS)-1-phenylethanol per minute.

The activity of the recombinant TkADH was spectrophotometrically determined at 70°C by monitoring the substrate-dependent change in the absorbance of NAD(P)⁺ or NAD(P)H at 340 nm (ε₃₄₀ = 6.22 mM⁻¹ cm⁻¹). Oxidation of alcohols was performed in a reaction mixture composed of 50 mM glycine-NaOH buffer (pH 9.0), 100 mM substrate, and 1.0 mM NAD(P)⁺. Reduction of ketone/aldehyde was carried out in 50 mM sodium phosphate buffer (pH 6.0) containing 100 mM substrate and 0.2 mM NAD(P)H. The reaction was started by addition of an appropriate amount of the enzyme into the reaction mixture in a total volume of 1 ml. One unit of ADH activity was defined as the amount of enzyme that reduces 1 μmol NAD(P)⁺ or oxidizes 1 μmol NAD(P)H per minute. The spontaneous degradation of NAD(P)H was corrected.

Protein concentrations were measured by the Bradford method with bovine serum albumin as the standard.

Characterization of recombinant TkADH. (i) Molecular mass determination.

The subunit molecular mass of TkADH was determined by SDS-PAGE, and the native molecular mass was determined by a Superdex 200 10/30 HR gel filtration column equilibrated with 50 mM Tris-HCl buffer (pH 8.0) containing 0.15 M NaCl. The standard proteins used for calibration were blue dextran 2000 (>2,000 kDa), conalbumin (75 kDa), ovalbumin (44 kDa), carbonic anhydrase (29 kDa), RNase A (13.7 kDa), and aprotinin (6.5 kDa).

(ii) Optimal pH.

The effect of pH on the oxidation of (RS)-1-phenylethanol was determined at 70°C in 50 mM Tris-HCl buffer (pH 7.5 to 9.0), 50 mM glycine-NaOH buffer (pH 9.0 to 10.5), and 200 mM disodium hydrogen phosphate-NaOH buffer (pH 10.0 to 12.0). For the reduction of acetoin, the buffers used were 50 mM glycine-HCl buffer (pH 2.0 to 3.0), 50 mM acetate buffer (pH 3.0 to 6.0), and 100 mM sodium phosphate buffer (pH 6.0 to 8.0). The pH of each buffer was adjusted at 25°C, and the temperature corrections were made.

(iii) Optimal temperature and thermostability.

The temperature dependency of TkADH was determined at temperatures ranging from 30 to 95°C in 50 mM HEPPS {3-[4-(2-hydroxyethyl)-1-piperazinyl]propanesulfonic acid} buffer (pH 8.5) containing 100 mM (RS)-1-phenylethanol as the substrate. For evaluation of the thermostability of TkADH (0.48 mg · ml⁻¹ in 50 mM Tris-HCl buffer [pH 8.0]), the residual activity was measured by the standard assay for oxidation direction after incubation for appropriate time intervals at 85°C and 95°C.

(iv) Kinetics.

For determining the kinetic parameters of TkADH, the enzyme assays were carried out at 70°C with various substrate concentrations of NAD⁺ and NADP⁺ (0 to 1.0 mM), meso-2,3-butanediol (0 to 250 mM), (RS)-1-phenylethanol (0 to 200 mM), NADH and NADPH (0 to 0.4 mM), and acetoin (0 to 200 mM). The concentrations of the cosubstrates were kept constant. The kinetic parameters were calculated by nonlinear regression to the Michaelis-Menten equation using Microcal Origin software (Microcal Software Inc., Northampton, MA). The k_cat parameter was calculated on the basis of monomeric TkADH with a subunit molecular mass of 31 kDa.

(v) Effects of metals, chemical additives, and organic solvents.

The oxidation activity of TkADH toward (RS)-1-phenylethanol and the reduction activity toward acetoin were assayed at 70°C in the presence of various metal ions (K⁺, Ca²⁺, Mg²⁺, Zn²⁺, Mn²⁺, Fe³⁺, Cu²⁺, Ni²⁺, Co²⁺, and Ag⁺), dithiothreitol (DTT), or EDTA at a final concentration of 1 mM, to evaluate their effects on the initial activity of TkADH.

The effects of organic solvents on the activity of TkADH were assessed for the oxidation direction at 70°C in the presence of 20% (vol/vol) water-miscible solvents (dimethyl sulfoxide, dimethylformamide, methanol, ethanol, acetone, and 2-propanol) and 50% (vol/vol) water-immiscible solvents (ethyl acetate, octanol, n-hexane, and n-octane). The solvent tolerance of TkADH was assessed by measuring the residual activity after incubation of the enzyme with the solvents at 60°C for 4 h.

(vi) Enantioselectivity.

The enantioselectivity of TkADH for the oxidation of aromatic secondary alcohols was investigated by a reaction with cofactor regeneration. The standard reaction mixture was 50 mM glycine buffer (pH 9.0) containing 15 mM (RS)-1-phenylethanol or its derivatives, 2 mM NAD⁺, 100 mM NaCl, 0.17 mM FAD, 0.3 U · ml⁻¹ of TkADH, and 0.3 U · ml⁻¹ of TkNOX in a total volume of 0.13 ml. The enantioselectivity for the reduction of an aromatic ketone was determined in the mixture composed of 50 mM sodium phosphate buffer (pH 6.0), 5 mM 2,2,2-trifluoroacetophenone, 5 mM NADH, and 0.3 U · ml⁻¹ of TkADH in a total volume of 0.13 ml. The reaction mixture was incubated in a 1.5-ml microtube at 40°C and 300 rpm in a Thermomixer comfort device (Eppendorf AG, Germany). A 100-μl portion was periodically taken and extracted with 1:2 (vol/vol) ethyl acetate. The organic phase was directly subjected to gas chromatography (GC; GC-2010 plus; Shimadzu Co.) equipped with a chiral column (length, 10 m; diameter, 25 μm; CP-Chirasil-DEX CB; Varian). The absolute configurations of the product alcohols were identified by comparing the chiral GC chromatograms of the authentic samples using the temperature programs in the literature (37, 38).

RESULTS

Purification of TkADH and identification of the gene.

We first examined the ADH activity toward the simplest aromatic secondary alcohol [(RS)-1-phenylethanol] in the extract of T. kodakarensis KOD1. As the activity was too low to be determined spectrophotometrically, batch reaction with NAD(P)⁺ regeneration by TkNOX (36) was applied at 50°C and the average reaction rate was determined from the conversion ratio after 3 h. When NADP⁺ and NAD⁺ were used as the cofactor, the TkADH activities toward (S)-1-phenylethanol were determined to be 0.30 and 0.20 mU · mg⁻¹, respectively, in the extracts of the cells grown in ASW-YT-Pyr medium, while they were 0.17 and 0.11 mU · mg⁻¹, respectively, in those of the cells grown in ASW-YT-S⁰ medium. Significant activity toward (R)-1-phenylethanol could not be detected in either of the extracts.

The native TkADH was then purified from the cells grown on ASW-YT-Pyr by a combination of anion-exchange, hydrophobic, and gel-filtration chromatographies, leading to the successful identification of a protein of 33 kDa, likely corresponding to TkADH, by SDS-PAGE analysis (Fig. 1A). The final preparation showed a specific activity of 253 mU · mg⁻¹, which indicated 827-fold purification with 29.7% yield from the initial cell extract. The N-terminal sequence of the 33-kDa protein was determined to be MKKVXIFNDLKWI (where X indicates an unidentified residue), and the following homology search against the T. kodakarensis KOD1 genome identified a protein encoded by tk0845. The gene encodes a protein of 278 amino acids with a calculated molecular mass of 31,464 Da, which was consistent with the result of SDS-PAGE.

Fig 1.

SDS-PAGE analyses of TkADH in each purification step. The enzymes were subjected to a 12% Tris-glycine polyacrylamide gel. (A) Native TkADH. Lane A1, crude extract of T. kodakarensis KOD1; lane A2, Q-Sepharose fractions; lane A3, phenyl-Sepharose fractions; lane A4, Superdex fractions; lane M, molecular mass standards. (B) Recombinant TkADH. Lane B1, crude extract of the recombinant E. coli; lane B2, heat-treated supernatant; lane B3, Q-Sepharose fractions; lane M, molecular mass standards.

The amino acid sequence deduced from the identified gene revealed that TK0845 was a member of the AKR superfamily. It shared high identities (79 to 100%) to the homologs from the closely related hyperthermophilic archaea in the order Thermococcales, including PF1960 from P. furiosus, which was previously reported to be a thermostable alcohol dehydrogenase, AdhD (33). Moderately homologous proteins with 40 to 70% identities were also found in various archaea and bacteria, such as aldo-keto reductases from the thermoacidophilic archaeon S. solfataricus (SSO3209; 45% identity) and hyperthermophilic bacterium T. maritima (TM1743; 42% identity), and numerous putative reductases with low identities (35 to 40%) were present in various eucarya. These AKR superfamily members have been described to have a common (α/β)₈-barrel fold and a highly conserved catalytic tetrad (4, 39–41) that was seen as Asp58, Tyr63, Lys89, and His121 in TkADH. In the T. kodakarensis genome (35), TK0845 is the only one encoding a member of the AKR superfamily.

Production and general properties of recombinant TkADH.

The identified TkADH gene, tk0845, was inserted into pET-21b(+) and overexpressed in E. coli Rosetta(DE3). The recombinant form of TkADH was purified to homogeneity by heat treatment and successive anion-exchange chromatography. The apparent subunit molecular mass of TkADH was confirmed to be 31 kDa by SDS-PAGE analysis (Fig. 1B). The native molecular mass was determined to be 27 kDa using gel-filtration chromatography, indicating that TkADH appears to be functional as a monomer form.

The optimal pH was determined to be 9.0 for the oxidation reaction. For the reduction reaction, TkADH showed a maximal activity at pH 3.0 and 50% activity at pH 6.0. Considering the stability of the cofactor, the reduction reactions were carried out at pH 6.0 in further experiments. TkADH was highly thermostable, as its oxidation activity increased with elevated temperature up to 90°C and its half-life was 16.2 and 4.8 h at 85°C and 95°C, respectively. The activation energy of the oxidation reaction was determined to be 37.8 kJ · mol⁻¹.

Substrate spectra.

The substrate specificity of TkADH for the oxidation reaction was determined using a variety of alcohols with NAD⁺ as the cofactor (Table 1). (RS)-1-Phenylethanol was chosen as the standard substrate, and the specific activity of TkADH toward this alcohol (19.0 U · mg⁻¹) was set at 100%. Considering the importance of aromatic chiral alcohols as useful blocks in pharmaceutical applications (5, 28), we tested a series of 1-phenylethanol derivatives as substrates. TkADH well accepted 1-phenylethanols with substitutions at the meta and para positions on the aromatic ring with an activity range of 119 to 315%; and the highest activity of 315% was found toward 1-(m-bromophenyl)ethanol. In contrast, very low activities of 16.4 and 11.3% were observed toward o-chloro- and o-bromo-substituted derivatives, respectively. Substitutions at the β position (1-phenyl-1,2-ethanediol, 2-chloro-1-phenylethanol, and 1-phenyl-1-propanol) also drastically decreased the enzyme activity to 0.3 to 13.1%, probably due to the steric effects on the interaction of the substrate with the protein. TkADH showed almost no activity toward 2-phenyl-1-ethanol (0.3%) but rather high activity toward 1-phenyl-2-propanol (39.1%), indicating that this enzyme is a secondary ADH. Various aliphatic alcohols were also examined as substrates for TkADH. The activities toward 1-alkanols (methanol to octanol, C₁ to C₈) (0.2 to 3.3%) were much lower than those toward 2-alkanols (2-propanol to 2-octanol, C₃ to C₈) and cyclohexanol (5.3 to 61.6%), which again indicated the preference of TkADH for secondary alcohols. d-Arabinose, l-arabinose, and acetoin also acted as substrates for TkADH, despite moderate activity (4.5 to 21.4%). Among the aliphatic substrates examined, the highest activity was obtained toward meso-2,3-butanediol (272%).

Table 1.

Substrate specificity of TkADH in the oxidation reaction

Substrate	Concn (mM)	Relative activitya (%)
Aromatic alcohols
(RS)-1-Phenylethanol	15	100 ± 1b
2-Phenylethanol	15	0.3 ± 0.0
1-Phenyl-1,2-ethanediol	15	3.4 ± 0.3
2-Chloro-1-phenylethanol	15	0.3 ± 0.1
1-Phenyl-1-propanol	15	13.1 ± 0.4
1-Phenyl-2-propanol	15	39.1 ± 0.2
1-(p-Fluorophenyl)ethanol	15	119 ± 12
1-(p-Methylphenyl)ethanol	15	189 ± 3
1-(p-Chlorophenyl)ethanol	15	151 ± 0
1-(m-Chlorophenyl)ethanol	15	205 ± 8
1-(o-Chlorophenyl)ethanol	15	16.4 ± 2.1
1-(p-Bromophenyl)ethanol	15	167 ± 3
1-(m-Bromophenyl)ethanol	15	315 ± 23
1-(o-Bromophenyl)ethanol	15	11.3 ± 2.8
Aliphatic alcohols
(RS)-1-Phenylethanol	100	100 ± 5c
d-Arabinose	100	21.4 ± 0.0
l-Arabinose	100	5.2 ± 0.0
meso-2,3-Butanediol	100	272 ± 22
Cyclohexanol	100	52.3 ± 3.2
Acetoin	100	4.5 ± 0.0
Methanol	100	0.2 ± 0.1
Ethanol	100	0.5 ± 0.3
1-Propanol	100	1.2 ± 0.3
1-Butanol	100	2.3 ± 0.3
1-Pentanol	100	2.8 ± 0.2
1-Hexanol	10	2.8 ± 0.1
1-Heptanol	10	3.3 ± 0.0
1-Octanol	10	2.3 ± 0.0
2-Propanol	100	5.3 ± 0.2
2-Butanol	100	22.3 ± 1.2
2-Pentanol	100	30.2 ± 0.7
2-Hexanol	100	38.7 ± 0.3
2-Heptanol	10	61.6 ± 0.4
2-Octanol	10	45.8 ± 0.1

^aData are means from three reactions.

^bRelative activity of 100% corresponds to 3.7 U · mg⁻¹ measured with 15 mM (RS)-1-phenylethanol under the standard condition.

^cRelative activity of 100% corresponds to 19.0 U · mg⁻¹ measured with 100 mM (RS)-1-phenylethanol under the standard condition.

The substrate spectrum of TkADH in the reduction direction toward several types of ketones and aldehydes was examined with NADH as the cofactor, as shown in Table 2. Acetoin was chosen as the standard substrate with a specific activity of 19.5 U · mg⁻¹. TkADH tended to show high reduction activity toward aliphatic aldehydes and ketones having a hydroxyl or carbonyl group at the α position, while aldoses did not act as substrates. The highest activity of TkADH was observed toward pyruvic aldehyde (328%). With regard to aromatic compounds, TkADH was active toward aldehyde, such as phenyl acetaldehyde, and α-ketoester, such as methyl benzoylformate. Acetophenone and p-chloroacetophenone did not act as substrates, whereas high activity of 180% toward 2,2,2-trifluoroacetophenone was observed.

Table 2.

Substrate specificity of TkADH in the reduction reaction

Substrate	Concn (mM)	Relative activitya (%)
Acetoin	100	100 ± 1b
Pyruvic aldehyde	100	328 ± 15
Ethyl pyruvate	100	294 ± 5
Dihydroxyacetone	100	34.8 ± 0.3
Cyclohexanone	100	19.1 ± 2.5
2-Hexanone	100	3.6 ± 1.5
Acetone	100	4.1 ± 0.0
d-Arabinose	100	0
l-Arabinose	100	0
Hexanal	10	12.9 ± 0.0
Methyl benzoylformate	10	67.0 ± 9.9
Phenylacetaldehyde	10	18.6 ± 0.2
Acetophenone	10	0
p-Chloroacetophenone	10	0
2-Bromoacetophenone	10	8.7 ± 3.1
2,2,2-Trifluoroacetophenone	10	180 ± 7

^aData are means from three reactions.

^bRelative activity of 100% corresponds to 19.5 U · mg⁻¹ measured for 100 mM acetoin under the standard condition.

Enantioselectivity.

Kinetic resolution of racemic 1-phenyl-ethanol derivatives with NAD⁺ regeneration mediated by TkNOX was performed at 40°C in order to minimize the experimental error caused by volatile loss of the alcohols. The enantiomeric ratio, E, which is the ratio of the specificity constants for two competing enantiomers, was calculated to evaluate the enantioselectivity of TkADH (42). The enzyme showed a high preference for the (S)-1-phenylethanol, with an E value of >604, yielding the (R)-alcohol in optical purities of greater than 99.8% enantiomeric excess (ee) (Table 3). The high enantioselectivity was also observed toward the meta- and para-substituted derivatives of 1-phenylethanol with excellent E values (>290 to >1,000). Relatively lower E values were observed with o-substituted substrates and phenylpropanol derivatives (>134 to >273), although the ee values of the remaining alcohols were still satisfactory (>98%). The asymmetric reduction of 2,2,2-trifluoroacetophenone by TkADH produced the corresponding (R)-alcohol with an optical purity of higher than 99.6% ee.

Table 3.

Enantioselectivity of TkADH^a

Direction and substrate	% ee (configuration)b	E
Oxidation
1-Phenylethanol	>99.8 (R)	>604
1-(p-Chlorophenyl)ethanol	>99.8 (R)	>1,000
1-(m-Chlorophenyl)ethanol	>99.8 (R)	>633
1-(o-Chlorophenyl)ethanol	>98.0 (R)	>261
1-(p-Bromophenyl)ethanol	>99.8 (R)	>360
1-(m-Bromophenyl)ethanol	>99.8 (R)	>1,000
1-(o-Bromophenyl)ethanol	>98.4 (R)	>273
1-(p-Fluorophenyl)ethanol	>99.8 (R)	>1,000
1-(p-Methylphenyl)ethanol	>99.8 (R)	>290
1-Phenyl-1-propanol	>98.4 (R)	>134
1-Phenyl-2-propanol	>98.4 (R)	>181
Reduction, 2,2,2-trifluoroacetophenone	>99.6 (R)

^aAll the reactions were carried out at 40°C. Data are means from three reactions.

^bEnantiomeric excess and configuration of the remaining alcohols for oxidation reaction or those of the produced alcohol for reduction direction.

Kinetic parameters.

Table 4 presents the kinetic parameters of TkADH determined with several substrates. The changes of K_m and k_cat values depending on the cofactor [NAD(H) or NADP(H)] were very small, again indicating that TkADH could utilize both cofactors. Although the enzyme showed a higher apparent K_m value toward (RS)-1-phenylethanol (244 mM) than that toward meso-2,3-butanediol (61 mM), the V_max values toward these compounds were not drastically different from each other [69 and 85 U · mg⁻¹ for (RS)-1-phenylethanol and meso-2,3-butanediol, respectively].

Table 4.

Kinetic parameters for TkADH^a

Substrate (concn range [mM])	Cosubstrate (concn [mM])	Apparent Km (mM)	Apparent Vmax (U · mg−1)	kcat (s−1)	kcat/Km (s−1 mM−1)
NAD+ (0–1.0)	meso-2,3-Butanediol (100)	0.127 ± 0.005	59.7 ± 0.8	30.8 ± 0.4	243 ± 10
NADP+ (0–1.0)	meso-2,3-Butanediol (100)	0.113 ± 0.006	58.0 ± 0.9	30.0 ± 0.5	265 ± 15
NAD+ (0–1.0)	(RS)-1-Phenylethanol (100)	0.089 ± 0.004	20.2 ± 0.3	10.4 ± 0.1	117 ± 6
NADP+ (0–1.0)	(RS)-1-Phenylethanol (100)	0.066 ± 0.004	20.4 ± 0.3	10.5 ± 0.2	159 ± 10
meso-2,3-Butanediol (0–250)	NAD+ (1.0)	61.3 ± 5.2	84.8 ± 2.5	43.8 ± 1.3	0.71 ± 0.06
(RS)-1-Phenylethanol (0–200b)	NAD+ (1.0)	244 ± 23	69.0 ± 4.2	35.7 ± 2.2	0.15 ± 0.02
NADH (0–0.4)	Acetoin (100)	0.105 ± 0.008	30.2 ± 0.8	15.6 ± 0.4	149 ± 12
NADPH (0–0.4)	Acetoin (100)	0.113 ± 0.011	33.4 ± 1.2	17.3 ± 0.6	153 ± 15
Acetoin (0–200)	NADH (0.2)	30.2 ± 1.7	26.5 ± 0.5	13.7 ± 0.3	0.45 ± 0.03

^aAll the reactions were carried out at 70°C. Data are means from three reactions.

^bThe maximum concentration was set to 200 mM due to the solubility.

Effects of metals, chemical additives, and organic solvents.

The effects of metal ions, DTT, and EDTA on TkADH activity were tested at a concentration of 1 mM. In the oxidation reaction, Ag⁺, Co²⁺, and Cu²⁺ decreased the enzyme activity by 10 to 25%, while Mg²⁺ and Ca²⁺ had a small inhibitory effect on the reduction reaction. DTT and EDTA had no effect on either oxidative or reductive activity, suggesting that neither a metal ion nor a disulfide bridge was involved in the activity of TkADH.

When 20% (vol/vol) water-miscible solvents was added to the reaction mixture, the oxidation activity of TkADH was roughly correlated with the octanol–water partition coefficients (log P_o/w values) of the solvents (8), except for acetone (Fig. 2A). High activity (90% of the activity without solvents) was observed in the case of 2-propanol, demonstrating a potential for a cofactor regeneration-coupled reaction by using 2-propanol as the sacrificial substrate. In the presence of 50% (vol/vol) water-immiscible solvents, the enzyme activity was high with hydrophobic solvents such as n-hexane and n-octane. These solvents were expected to be useful in the biphasic system. TkADH showed high stability in all the solvents examined (Fig. 2B), regardless of the log P_o/w values of the solvents.

Fig 2.

Effects of organic solvents on the activity (A) and stability (B) of TkADH. The amounts of water-miscible and water-immiscible solvents added were 20% (vol/vol) and 50% (vol/vol), respectively. The log P_o/w values are given in parentheses. (A) Relative activity of 100% corresponds to 19.0 U · mg⁻¹ for the oxidation of (RS)-1-phenylethanol determined under the standard condition. (B) Stability was determined by measuring the residual activity after incubation with organic solvents at 60°C for 4 h. DMSO, dimethyl sulfoxide; DMF, dimethylformamide.

DISCUSSION

This study identified and characterized a novel thermostable ADH from the hyperthermophilic archaeon T. kodakarensis KOD1 (TkADH) which was active and enantioselective toward aromatic secondary alcohols. TkADH belongs to the AKR superfamily, which differs from conventional ADH groups. The recombinant form of TkADH was highly thermostable and could oxidize 1-phenylethanol and its derivatives having halo substituents on the phenyl group, 1-phenyl-1-propanol and 1-phenyl-2-propanol, with high enantioselectivity. It could also catalyze oxidation of various aliphatic alcohols with preference for secondary alcohols rather than primary ones. With regard to the reduction direction, TkADH could accept various ketones and aldehydes as substrates and particularly showed high activity toward carbonyl compounds having an α substituent and a α-dicarbonyl structure, such as acetoin and pyruvic aldehyde, respectively.

Aromatic chiral alcohols are important intermediates for the synthesis of pharmaceuticals, such as anticancer drugs and antibiotics (5, 28). Although several thermostable ADHs have been identified from various hyperthermophiles, their potential for enantioselective bioconversion of aromatic alcohols/ketones has not been well considered. Table 5 summarizes some properties of TkADH and four thermostable ADHs, focusing on their use in enantioselective conversion of aromatic compounds. The ADH from A. pernix was shown to be active toward 4-methoxyphenyl acetone, but its enantioselectivity toward this aromatic compound was not studied (29). The short-chain ADH from S. acidocaldarius showed high enantioselectivity for the reduction of benzyl to (R)-benzoin (98% ee); however, enantioselectivity to acetophenone was less satisfactory (92% ee) (32). There have been two reports regarding the characterization of thermostable AKRs from the hyperthermophilic archaeon P. furiosus (33, 43, 44) and the bacterium T. maritima (27). It has been pointed out that the functions and catalytic properties of the AKR superfamily members are quite diverse, despite the common three-dimensional structures and active sites (4, 39, 45). Indeed, the ADH enzyme from T. maritima (TmADH) sharing 42% identity with TkADH showed a substrate specificity markedly different from that of TkADH. TmADH could accept a very limited range of substrates, which are aldehydes having a phenyl or partially hydrogenated phenyl group (benzaldehyde, 1,2,3,6-tetrahydrobenzaldehyde, and p-anisaldehyde) for reduction direction. The ADH of the AKR superfamily from P. furiosus (PfADH) is highly homologous (85% identity) and showed properties rather similar to those of TkADH, such as high activity toward 2,3-butanediol and acetoin for oxidation and reduction direction, respectively. PfADH was reported to be capable of reducing a series of aromatic ketones and keto esters with high enantioselectivity (43, 44). However, the reaction required a large amount of the enzyme and a long reaction time due to the quite low activity of PfADH toward the substrates examined in the previous study (0.013 to 0.19 U · mg⁻¹ at 37°C) (43). In the present study, we found that (RS)-1-phenylethanol and a series of derivatives could be oxidized by TkADH with both high activity (1.0 to 3.1 U · mg⁻¹ at 40°C) and high (S)-enantioselectivity, allowing more efficient bioconversion. Interestingly, TkADH had an affinity toward NAD(P)⁺ (K_m value range, 66 to 127 μM) similar to that toward NAD(P)H (K_m value range, 105 to 113 μM), which was not observed in other known thermostable ADHs, including highly homologous PfADH (33). This property could be an advantage for kinetic resolution of racemates of aromatic secondary alcohols using TkADH (46). Moreover, TkADH showed a high reduction activity toward 2,2,2-trifluoroacetophenone with high enantioselectivity, leading to efficient preparation of (R)-2,2,2-trifluoro-1-phenylethanol, a useful chiral agent, with high optical purity (>99.6% ee).

Table 5.

Properties of selected ADHs from hyperthermophiles

Species	Classification	Cofactor(s)	Optimal temp (°C)	Stability (half-life [h])	Standard substrates and acceptable racemic alcohols/prochiral ketones (relative activity)	Enantioselectivity	GenBank accession no.	Reference(s) or source
Aeropyrum pernix	Zn-containing ADH	NAD(H)	≥95	0.5 (90°C)	Oxidation, benzylalcohol (Vmax, 1.53 U · mg−1; 60°C) (100%), 2-alkanols (C3–C5) (24–59%); reduction, benzaldehyde (Vmax, 1.83 U · mg−1; 60°C) (100%), 4-methoxyphenylacetone (6%), 2-alkanones (C5–C10) (33–89%)	92–95% ee (S) for the reduced products of 2-alkanones (C8–C10); 37–79% ee (S) for the reduced products of 2-alkanones (C5–C7)	BAA81251	29, 30
Sulfolobus acidocaldarius	Short-chain ADH	NAD(H)	78	0.5 (88°C)	Oxidation, isoborneol (Vmax, 36.9 U · mg−1; 65°C) (100%), 1-phenylethanol (2%), 1-(p-chlorophenylethanol) (2%), 1-phenyl-1-propanol (2%), α-(trifluoromethyl)benzyl alcohol (2%), 3-methylcyclohexanol (10%), 2-alkanols (C5–C8) (1–2%); reduction, 1-phenyl-1,2-propanedione (Vmax, 11.8 U · mg−1; 65°C) (100%), benzil (62%), 2,2,2-trifluoroacetophenone (35%), 2,2-dichloroacetophenone (30%), 3-methylcyclohexanone (60%)	98 and 91% ee (R) for the reduced products of benzil and 2,2′-dichlorobenzil, respectively; 92% ee (S) for the reduced product of acetophenone	AAY80578	32
Pyrococcus furiosus	Aldo-keto reductase	NAD(H) (preferred), NADP(H)	≥100	2.17 (100°C)	Oxidation, 2,3-butandiol (Vmax, 108.3 U · mg−1; 70°C) (100%), acetoin (18%); reduction, acetoin (Vmax, 22.5 U · mg−1; 70°C) (100%), diacetyl (∼150%), dihydroxyacetone phosphate (82%), 2-hexanone (0.2%); the enzyme could catalyze reduction of substituted acetophenones, α-choloroacetophenones, and α- and β-ketoesters, despite the low activity (0.01–0.19 U · mg−1; 37°C)	>99% ee for the reduced products of the substituted acetophenones, α-choloroacetophenones, and aromatic α- and β-ketoesters	AAL82084	33, 43, 44
Thermotoga maritima	Aldo-keto reductase	NADP(H)	≥70	40% loss for 5 h (80°C)	Reduction, 1,2,3,6-tetrahydrobenzaldehyde (Vmax, 5.5 U · mg−1; 50°C), benzaldehyde, para-anisaldehyde	Not demonstrated	AAD36088	27
Thermococcus kodakarensis	Aldo-keto reductase	NAD(H), NADP(H)	90	16.4 (85°C)	Oxidation, 1-phenylethanol (Vmax, 69.0 U · mg−1; 70°C) (100%), 1-(p-chlorophenyl)ethanol (151%), 1-(m-chlorophenyl)ethanol (205%), 1-(o-chlorophenyl)ethanol (16%), meso-2,3-butandiol (272%), 2-heptanol (62%); reduction, acetoin (Vmax, 26.5 U · mg−1; 70°C) (100%), 2,2,2-trifluoroacetophenone (180%), 2-bromoacetophenone (9%), 2-hexanone (4%)	>99.8% ee (R) for the remaining substituted 1-phenylethanols; >99.6% ee (R) for the reduced product of 2,2,2-trifluoroacetophenone	BAD85034	This work

TkADH showed not only thermostability but also high resistance to both hydrophobic and hydrophilic solvents. The similar solvent resistance was also reported for PfADH, but the activity in the solvent system was not mentioned. Here, we clarified that TkADH showed high stability to all the solvents examined and retained its activity, despite the decreasing trend in the presence of solvents. Among the solvents examined, TkADH particularly exhibited high activity in miscible H₂O–20% 2-propanol and biphasic H₂O–n-hexane as well as H₂O–n-octane (Fig. 2). As TkADH showed weak but actual activity toward 2-propanol as a substrate, the 2-propanol solvent system would allow efficient conversion of water-immiscible compounds with in situ regeneration of the reduced cofactor (7). The use of n-hexane and n-octane would be also beneficial for the conversion of hydrophobic substrates in the biphasic reaction systems (47).

The results obtained in the present study demonstrated that TkADH is a useful tool for the production of valuable aromatic chiral alcohols. The reaction mechanisms and structural properties of this enzyme responsible for the unique substrate and cofactor specificities, as well as the physiological roles of this enzyme in the hyperthermophile, will be interesting scientific subjects for further exploration.