Cloning, Sequence Analysis, and Expression in Escherichia coli of Gene Encoding N-Benzyl-3-pyrrolidinol Dehydrogenase from Geotrichum capitatum

Cloning, Sequence Analysis, and Expression in Escherichia coli of Gene Encoding N-Benzyl-3-pyrrolidinol Dehydrogenase from Geotrichum capitatum

JOURNAL OF BIOSCIENCE AND BIOENGINEERING Vol. 104, No. 5, 379–384. 2007 DOI: 10.1263/jbb.104.379 © 2007, The Society for Biotechnology, Japan Clonin...

395KB Sizes 0 Downloads 3 Views

JOURNAL OF BIOSCIENCE AND BIOENGINEERING Vol. 104, No. 5, 379–384. 2007 DOI: 10.1263/jbb.104.379

© 2007, The Society for Biotechnology, Japan

Cloning, Sequence Analysis, and Expression in Escherichia coli of Gene Encoding N-Benzyl-3-Pyrrolidinol Dehydrogenase from Geotrichum capitatum Keiko Yamada-Onodera,1* Kazutaka Kojima,1 Yuhki Takase,1 and Yoshiki Tani2 Graduate School of Biological Sciences, Nara Institute of Science and Technology, 8916-5 Takayama, Ikoma 630-0192, Japan1 and Faculty of Bioenvironmental Science, Kyoto Gakuen, 1-1 Nanjyou-ohtani, Sogabe, Kameoka 621-8555, Japan 2 Received 19 June 2007/Accepted 2 August 2007

The gene encoding N-benzyl-3-pyrrolidinol dehydrogenase (DDBJ/EMBL/GenBank accession no. AB294179), a useful biocatalyst for producing (S)-N-benzyl-3-pyrrolidinol, was cloned from the genomic DNA of Geotrichum capitatum JCM 3908. The gene contained an open reading frame consisting of 1023 nucleotides corresponding to 340 amino acid residues. The subunit molecular weight was calculated to be 39,000. The predicted amino acid sequence did not have significant similarity to those of N-benzyl-3-pyrrolidinone reductases reported previously. From 30 mM N-benzyl-3-pyrrolidinone, (S)-N-benzyl-3-pyrrolidinol was obtained with a yield > 99.9% and an enantiomeric excess > 99.9% in 1-h and 2-h reactions without NADH addition by the resting cells of Escherichia coli HB 101 strains harboring the expression plasmids pSG-POBS and pSF-POBS that possess the glucose dehydrogenase gene and formate dehydrogenase gene as an NADH-reproducing system, respectively, besides the N-benzyl-3-pyrrolidinol dehydrogenase gene. N-Benzyl3-pyrrolidinol dehydrogenase activity (0.56 U/mg) was observed in E. coli (pSG-POBS), which was 17-fold the specific activity observed in G. capitatum JCM 3908. [Key words: N-benzyl-3-pyrrolidinol dehydrogenase, (S)-N-benzyl-3-pyrrolidinol production, N-benzyl3-pyrrolidinone reductase, Geotrichum capitatum]

those reported previously. We have purified and characterized N-benzyl-3-pyrrolidinone reductase/N-benzyl-3-pyrrolidinol dehydrogenase from Geotrichum capitatum JCM 3908 (5). Our enzyme has N-benzyl-3-pyrrolidinol oxidizing activity and requires NADH. The enzyme of Corynebacterium sp. was originally reported as an NADH-linked phenylacetoaldehyde reductase (3). The conversion of N-benzyl-3-pyrrolidinone to (S)-N-benzyl-3-pyrrolidinol using transformant cells that express this enzyme was described in a patent application (Asako, H. et al., Japanese patent P2004-350625A). In the medium that contained NAD+, (S)-N-benzyl-3-pyrrolidinol was produced with a conversion rate and an enantiomeric excess of 100% and more than 99%, respectively; however, it took 20 h and NAD+ addition was necessary. In this study, we report the cloning and sequence analysis of the gene encoding N-benzyl-3-pyrrolidinol dehydrogenase from G. capitatum JCM 3908 (DDBJ/EMBL/GenBank accession no. AB294179) in Escherichia coli and show that the expression system of this gene is convenient for the production of (S)-N-benzyl-3-pyrrolidinol.

(S)-N-Benzyl-3-pyrrolidinol is an important building block as versatile chiral synthons for the asymmetric synthesis of drugs including antitumor, anesthetic, antispasmodic, hepatotoxic, anti-inflammatory, and anti-HIV products. This compound has been chemically synthesized from optically active compounds (Okada, Y. et al., Japanese patent S60-23328, 1985; 1, 2) and prochiral compounds (Iriuchijima, S. and Masuda, T., Japanese patent, H01-207266, 1988; Takahashi, S. et al., Japanese patent, H03-176462, 1991). Enzymes that can reduce N-benzyl-3-pyrrolidinone to (S)-N-benzyl-3-pyrrolidinol were isolated and their genes were cloned from Micrococcus luteus (Kizaki, N. et al., WO 02/10399, 2002) and Corynebacterium sp. (3, 4; DDBJ/EMBL/GenBank accession no. BAD51480). The isolated enzymes do not have N-benzyl-3-pyrrolidinol dehydrogenase activity. If there is an enzyme that can oxidize an isomer of N-benzyl-3-pyrrolidinol with strict stereospecificity, we can obtain another isomer with high optical purity after the removal of the undesired isomer by oxidation. The enzyme of M. luteus requires NADPH as a coenzyme. NADH has priority over NADPH because NADH is cheaper than NADPH. We aimed at isolating N-benzyl-3-pyrrolidinone reductase/N-benzyl3-pyrrolidinol dehydrogenase with features different from

MATERIALS AND METHODS Microorganisms, plasmids, and cultivation conditions G. capitatum JCM 3908 (5) was used as the source of chromosomal DNA. For the cloning of the gene of N-benzyl-3-pyrrolidinol dehydrogenase, pCR2.1-TOPO and E. coli TOP10 cells were used as

* Corresponding author. e-mail: [email protected] phone: +81-(0)743-72-5422 fax: +81-(0)743-72-5429 379

380

YAMADA-ONODERA ET AL.

a vector and host strain, respectively, which were contained in a cloning kit (TOPO TA cloning; Invitrogen, Carlsbad, CA, USA). For the expression of the gene, the host strain (E. coli HB101 [6]) and vector plasmids (pSE420D [6], pSE-BSG1 [7], and pSE-MF26 [Kimoto, K. et al., Japanese patent 2004-49028, 2004]) were used. pSE-BSG1 and pSE-MF26 were used for expressing the glucose dehydrogenase gene from Bacillus subtilis (8) and the formate dehydrogenase gene from Mycobacterium vaccae (9), respectively. For the preparation of genomic DNA from G. capitatum JCM 3908, the yeast was cultivated at 30°C for 24 h in a liquid medium as described previously (5). For gene cloning and the expression of the gene, E. coli cells were cultivated at 37°C in Luria–Bertani (LB) medium (10) containing 50 µg of ampicillin per ml. For the induction of the gene under the control of the lac promoter in the expression plasmids, 0.1 mM isopropyl-β-D-thiogalactopyranoside (IPTG) was added to the LB medium. Expression of recombinant N-benzyl-3-pyrrolidinol dehydrogenase The recombinant cells were grown in 5 ml and 100 ml of LB medium to which ampicillin was added, with reciprocal and rotary shaking at 300 rpm and 150 rpm, respectively, at 37°C. When the optical density at 660 nm measured using a spectrophotometer (U1100; Hitachi, Tokyo) was found to be 0.5, IPTG was added to the medium. Afterwards, the cultures were incubated for another 3 h. The cells were harvested by centrifugation at 5200 ×g for 15 min. They were then used for the enzyme assay, protein assay, SDS–PAGE, and the synthesis of optical alcohols. Enzyme and protein assays The harvested transformant cells from the 5 ml culture medium were suspended in 1.2 ml of 50 mM potassium phosphate buffer (KPB; pH 7.0) and disrupted with a homogenizer (Multi Bead Shocker; Yasui Kikai, Osaka) for 7.5 min. The supernatant was obtained by centrifugation at 16,000 ×g for 30 min and used for enzyme and protein assays. Enzyme and protein assays were performed on the basis of a method from a previous study (5) unless otherwise specified. Enzyme activity was assayed at 25°C by measuring the change in NAD(P)H absorbance at 340 nm. One unit of the enzyme is defined as the amount of the enzyme that catalyzes a change of 1 µmol of NAD(P)H per min, using a molar absorption coefficient of 6220 M–1 ⋅ cm–1 at 340 nm. N-Benzyl-3-pyrrolidinone reducing activity was measured in a standard assay mixture that contained 0.2 mM NAD(P)H, 3 mM N-benzyl-3-pyrrolidinone, and an appropriate amount of the enzyme solution in a total volume of 1 ml of 50 mM KPB (pH 6.0). The reaction mixture without N-benzyl-3-pyrrolidinone was used as a reference. N-Benzyl-3-pyrrolidinol oxidizing activity was determined in the reaction mixture that contained 10 mM (S)- or (R)-N-benzyl3-pyrrolidinol, 2 mM NAD+, and an appropriate amount of the enzyme in a total volume of 1 ml of 50 mM KPB (pH 8.0). Glucose dehydrogenase and formate dehydrogenase activities were measured in the same reaction mixture as N-benzyl-3-pyrrolidinol oxidizing activity except that the reaction medium contained 10 mM glucose and 10 mM sodium formate, respectively, instead of N-benzyl-3-pyrrolidinol. The rate of N-benzyl-3-pyrrolidinol dehydrogenase production by the transformants relative to the total amount of soluble cellular proteins was expressed as a value relative to the specific activity (31 U/mg) of the purified N-benzyl-3-pyrrolidinone reductase (5). Enantioselective reduction of N-benzyl-3-pyrrolidinone by resting cells For the enantioselective reduction of N-benzyl3-pyrrolidinone, harvested cells (10 mg as dry cell weight) were added to 1 ml of a reaction mixture containing 30 mM N-benzyl3-pyrrolidinone, 50 mM glucose in 50 mM potassium phosphate buffer (KPB; pH 7.0 or 6.0). When E. coli (pSF-POBS) was used, 50 mM sodium formate was added to the medium instead of glucose. The reaction medium was purged with nitrogen gas before incubation to prevent the decomposition of N-benzyl-3-pyrrolidinone by oxygen. The reaction was performed at 30°C with recipro-

J. BIOSCI. BIOENG.,

cal shaking at 300 rpm. The reaction medium was centrifuged at 8000 ×g and the supernatant was used for the determination of reaction products by HPLC. Detection of N-benzyl-3-pyrrolidinol and N-benzyl-3-pyrrolidinone N-Benzyl-3-pyrrolidinol and N-benzyl-3-pyrrolidinone concentrations were measured by HPLC as described previously (5). Sequencing of amino acid residues and nucleotides The internal amino acid sequences of peptide fragments that were separated on an SDS–PAGE gel (5) were analyzed by Aproscience (Naruto). Nucleotide sequencing and analysis of the sequence data were performed as described previously (10). Cloning of gene of N-benzyl-3-pyrrolidinol dehydrogenase Chromosomal DNA was prepared from G. capitatum JCM3908 by the method of Cryer et al. (11). For the core region of the gene of N-benzyl-3-pyrrolidinol dehydrogenase, PCR was performed with primer 1 and primer 2 as described in the manual for Ex Taq DNA polymerase (Takara, Kyoto). According to two internal amino acid sequences (DDBJ/EMBL/GenBank accession no. AB294179), AFGSYVIAVDP and ITFDLNHLAF, two primers, namely, primer 1,5′-TTYGGIAGYTAYGTNATHGCNGT-3′ corresponding to F-2 to V-6 and primer 2,5′-AANGCIARRTGRTTIARRTCRAA-3′ corresponding to sequences complementary to F-3 to F-10 were synthesized, respectively. The PCR product was ligated into vector pCR2.1TOPO. DNA sequences that flank a core region were cloned using a Takara LA PCR in vitro cloning kit, according to the manufacturer’s manual. For 3′-flanking regions, the primer (5′-AACGAAG TTTACGCCAAACTC-3′) consisted of sequences identical to 652 through 672 in DNA sequences (DDBJ/EMBL/GenBank accession no. AB294179); for 5′-flanking regions, the primer (5′-AAGT CAAAGGTACCTTGGGAAC-3′) consisted of sequences complementary to 716 through 737 in DNA sequences. The sequence data of the fragments obtained after amplification were analyzed. Construction of expression plasmids of gene of N-benzyl3-pyrrolidinol dehydrogenase Primers POBS-F and POB-R were prepared for PCR to obtain the open reading frame (ORF) of the gene of N-benzyl-3-pyrrolidinol dehydrogenase. Primer POBS-F, 5′-CATGCCATGGCCGAAATCCCC-3′, consisted of CATG, an NcoI site and C + 5 to C + 15 at the 5′ end of the ORF; primer POB-R, 5′-GCTCTAGATTAGTTATCTTCATTGTGGACCAAAA-3′, consisted of GC, an XbaI site, and 26 nucleotides complementary to the 3′ end of the ORF, from T +998 to A +1023. Amplification by PCR was carried out according to the manual for Takara Ex Taq DNA polymerase. Amplified DNA fragments were digested with NcoI and XbaI. The resultant product was purified and ligated to pSE-BSG1, already digested with NcoI and XbaI, using a ligation kit (Takara). Plasmids that were inserted by their PCR products were designated as pSG-POBS. DNA fragments amplified with the primers POBS-F and POB-R and digested with NcoI and XbaI were also ligated to pSE420D already digested with NcoI and XbaI, as described above. The resultant product was designated as pSE-POBS. For the construction of pSF-POBS, the primers FPOBS-F and FPOBS-R were used as forward and reverse primers, instead of POBS-F and POBS-R, respectively. Primer FPOBS-F, ACCGGA ATTCTAAAATGGCCGAAATCCCC, consisted of ACCG, an EcoRI site, TAAA, and A +1 to C +15 at the 5′ end of the ORF, and primer FPOBS-R, GGCCCAAGCTTATTTAGTTATCTTCATTGT GGACC, consisted of GGCCC, a HindIII site, AT, and 22 nucleotides complementary to the 3′ end of the ORF, from G + 1002 to A+1023. EcoRI and HindIII were used for the digestion of the amplified DNA and pSE-MF26. These digestion products were ligated, and the resultant plasmid was designated as pSF-POBS.

N-BENZYL-3-PYRROLIDINOL DEHYDROGENASE

VOL. 104, 2007

RESULTS AND DISCUSSION The internal amino acid sequences of the peptides with molecular weights of 41,000 and 39,000 separated on an SDS–PAGE gel (5) were determined; AFGSYVIAVDP and ITFDLNHLAF from the former peptide; FGSYVIAVDP and PVGLQ from the latter peptide. Because FGSYVIAVDP was found in both peptides, the latter peptide might be the degradation product of the former peptide. The core region (C + 606 to T + 804) of the gene of N-benzyl-3-pyrrolidinol dehydrogenase was cloned by PCR using primers synthesized on the basis of the two inside amino acid sequences, AFGSYVIAVDP and ITFDLNHLAF. Another inside amino acid sequence PVGLQ was found in the deduced amino acid sequences of the core region. A possible ORF of 1023 nucleotides encoded a polypeptide of 340 amino acid residues with a calculated molecular weight of 39,000. In addition to ORF, the promoter region of the gene of N-benzyl3-pyrrolidinol dehydrogenase was sequenced. Two possible promoter sequences, TATA (positions −215 to −212, and −66 to −63), and a putative CAAT motif (positions −281 to −278, −276 to −273, −44 to −41, −27 to −24, and −12 to −9) were found upstream of the initiation codon. Proteins with sequences similar to the deduced amino acid sequence were searched using BLAST software (National Center for Biotechnology Information, NIH, USA) (Fig. 1). The de-

381

duced amino acid sequences of our enzyme showed the highest identity (44%) with the (S)-specific secondary alcohol dehydrogenase of Candida parapsilosis (DDBJ/EMBL/ GenBank accession no. BAA24528) among proteins in the NCBI databank. The identity between phenylacetaldehyde reductase from Corynebacterium sp. and our enzyme was 31%. Our enzyme did not have a significant similarity with the enzyme of M. luteus (Kizaki, N. et al., WO 02/10399, 2002), showing an identity under 20%. In the deduced amino acid sequences, GXGXXG and DXXXX, which are conserved throughout the medium chain dehydrogenase/reductase (MDR) family (12) and a putative coenzyme binding site (13), respectively, were also observed. Although both sorbitol dehydrogenase and alcohol dehydrogenase are zinccontaining enzymes, alcohol dehydrogenase contains two zinc atoms per subunit, a structural atom and a catalytic atom, whereas sorbitol dehydrogenase has only a single catalytic zinc atom (14). In our enzyme, both zinc-coordinating residues were observed in the deduced amino acid sequence, as shown in the boxes in Fig. 1. From these results and the molecular weight of the subunits, (S)-N-benzyl-3-pyrrolidinol dehydrogenase seems to belong to the MDR family, as well as glycerol dehydrogenase of Hansenula polymorpha DL-1 (10). A possible ORF of 1023 bp was inserted into the expression vector pSE-BSG1, which contained the gene of NAD(P)-

FIG. 1. Multiple sequence alignment between the N-benzyl-3-pyrrolidinol dehydrogenase of G. capitatum JCM 3908 and other dehydrogenases. Residues that have generally been conserved throughout the MDR family (GXGXXG, where X is any amino acid), putative zinc-coordinating residues (for catalytic zinc [C............HE] and for structural zinc [C...C...C........C]), and the putative coenzyme-recognizing site (DXXXX, where X is any amino acid) are framed. G. capitatum JCM 3908 (DDBJ/EMBL/GenBank accession no. AB294179); Corynebacterium sp. (BAD51480); C. parapsilosis (BAA24528); Geobacillus stearothermophilus (P42327); Bacillus clausii (YP_173550); Sphingomonas sp. (ZP_01303305). POBDH, N-Benzyl-3-pyrrolidinol dehydrogenase; PAR, phenylaldehyde reductase; ADH, alcohol dehydrogenase.

382

J. BIOSCI. BIOENG.,

YAMADA-ONODERA ET AL.

TABLE 1. Expression of genes of (S)-N-benzyl-3-pyrrolidinol dehydrogenase, glucose dehydrogenase, and formatedehydrogenase in E. coli HB101 Reaction

Cofactor

pSE-BSG1 pSG-POBS N-Benzyl-3-pyrrolidinone reducing activity NADH ND 0.56 NADPH ND ND ND 0.20 (S)-N-Benzyl-3-pyrrolidinol oxidizing activity NAD+ ND ND (R)-N-Benzyl-3-pyrrolidinol oxidizing activity NAD+ 15 37 Glucose oxidizing activity NAD+ NT NT Formate oxidizing activity NAD+ The activity is expressed as the specific activity (U/mg protein). NT, Not tested; ND, not detected.

linked glucose dehydrogenase of B. subtilis (8) for NADH regeneration (Fig. 2). In the cell extracts from E. coli (pSGPOBS), significant glucose oxidizing activity and NADHlinked N-benzyl-3-pyrrolidinone reducing activity were found; only glucose oxidizing activity was detected in E. coli HB101 (pSE-BSG1) (Table 1). (S)-N-Benzyl-3-pyrrolidinol oxidizing activity was detected in E. coli HB101 (pSG-POBS), not towards the (R)-isomer. The amount of N-benzyl-3-pyrrolidinol dehydrogenase produced by E. coli HB101 (pSGPOBS) corresponded to about 1.8% of the total amount of soluble cellular proteins. The expression vectors pSE-POBS and pSF-POBS were also constructed by inserting the N-benzyl-3-pyrrolidinol dehydrogenase gene of 1023 bp into pSE420D and pSF-MF26, respectively (Fig. 2). pSE-POBS did not co ntain an NADH regeneration system. pSF-POBS contained the gene of the NAD-linked formate dehydrogenase of M. vaccae N10 (9) as an NADH regenerator. In the cell extracts of the transformants harboring pSE-POBS and pSF-POBS, significant N-benzyl-3-pyrrolidinone reducing activity and (S)-N-benzyl-3-pyrrolidinol oxidizing activity were observed (Table 1). Formate oxidizing activity was found in E. coli (pSF-POBS) at the same level as that in E. coli (pSE-MF26). The reduction of N-benzyl-3-pyrrolidinone by G. capitatum JCM 3908, E. coli (pSE-POBS), E. coli (pSG-POBS), and E. coli (pSF-POBS) was performed at an initial pH of

FIG. 2. Restriction maps of pSG-POBS (a), pSE-POBS (b), and pSF-POBS (c).

Plasmid pSE-POBS 0.68 ND 0.23 ND ND ND

pSE-MF26 ND ND ND ND NT 0.54

pSF-POBS 0.27 ND 0.089 ND NT 0.85

7.0 in the reaction medium (Fig. 3a, b, c, and d, respectively). Only (S)-N-benzyl-3-pyrrolidinol was produced from N-benzyl-3-pyrrolidinone by all strains. G. capitatum JCM 3908 completed the reduction in 8 h (Fig. 3a), whereas the reduction by E. coli (pSE-POBS) stopped in 4 h with a conversion rate of 65% (Fig. 3b). E. coli (pSG-POBS) completely reduced N-benzyl-3-pyrrolidinone in a 1-h incubation (Fig. 3c), whereas E. coli (pSF-POBS) completed the reduction in a 3-h incubation (Fig. 3d). Generally speaking, a low pH is favorable for reduction. The final pH of the reaction medium with E. coli (pSF-POBS) was 7.7 when the initial pH was 7.0 (Fig. 3d). When the reaction started at pH 6.0, the final pH was 7.0 and the reaction was completed in 2 h (Fig. 3e). The efficiency was improved slightly. The N-benzyl-3-pyrrolidinone reductase activities in G. capitatum JCM 3908, E. coli (pSE-POBS), E. coli (pSG-POBS), and E. coli (pSF-POBS) were 0.04, 0.62, 0.75, and 0.30 U for every 1 ml of the reaction medium, respectively. The efficiency in the resting-cell reactions seems to be due to the total enzyme activity in the medium and the existence of the NADH regeneration system. In this study, the gene of N-benzyl-3-pyrrolidinol dehydrogenase was cloned from G. capitatum JCM 3908. The deduced amino acid sequence of the enzyme was not similar to those of N-benzyl-3-pyrrolidinone reductases reported previously. N-Benzyl-3-pyrrolidinol dehydrogenase activity was observed in E. coli (pSG-POBS), which was 17-fold the specific activity observed in G. capitatum JCM 3908. By using the transformants, which have NADH regeneration system, (S)-N-benzyl-3-pyrrolidinol productivity was improved in comparison with those of G. capitatum JCM 3908 and the transformant having no NADH regeneration system. Formate dehydrogenase and glucose dehydrogenase produce carbon dioxide and gluconic acid from formate and glucose, respectively, in the reaction medium. Gluconic acid production decreases the pH of the medium. Carbon dioxide is released into the air and the medium becomes alkali. Generally speaking, formate dehydrogenase seems to be suitable from the viewpoint that the removal of by-products is unnecessary. The (S)-N-benzyl-3-pyrrolidinol productivity of the transformants that express formate dehydrogenase gene was the second highest. The transformant cells that express the N-benzyl-3-pyrrolidinone reductase genes of M. luteus (Kizaki, N. et al., WO 02/10399, 2002) and Corynebacterium sp. (Asako, H. et al., Japanese patent P2004350625A) produce (S)-N-benzyl-3-pyrrolidinol from N-benzyl-3-pyrrolidinone. The results in the resting-cell reactions without coenzyme addition were not described. Our trans-

N-BENZYL-3-PYRROLIDINOL DEHYDROGENASE

VOL. 104, 2007

383

FIG. 3. (S)-N-Benzyl-3-pyrrolidinol production by G. capitatum JCM 3908, E. coli HB101 (pSE-POBS), E. coli HB101 (pSG-POBS), and E. coli HB101 (pSF-POBS). The resting-cell reactions by (a) G. capitatum JCM 3908, (b) E. coli HB101 (pSE-POBS), (c) E. coli HB101 (pSG-POBS), and (d) E. coli HB101 (pSF-POBS) at an initial pH of 7.0, and that by (e) E. coli HB101 (pSF-POBS) at an initial pH of 6.0 were performed as described in Materials and Methods.

formants did not require NAD+ addition in the reaction medium. We produced 30 mM (5.3 g/l) (S)-N-benzyl-3-pyrrolidinol from N-benzyl-3-pyrrolidinone with a yield >99.9% and an enantiomeric excess > 99.9% in a 1-h reaction. In the next stage, we will produce (S)-N-benzyl-3-pyrrolidinol at a higher concentration.

4.

5.

ACKNOWLEDGMENTS We are grateful to Dr. H. Yamamoto for valuable advice on the determination of internal amino acid sequences of the purified enzyme.

REFERENCES 1. Shibata, T., Iino, K., and Sugiura, Y.: Synthesis of optically active 3-mercaptophyrrolidine derivatives. Synthetic intermediates of carbapenem RS-533 and its isomer. Heterocycles, 24, 1331–1346 (1986). 2. Harris, B. D., Bhat, K. L., and Joullié, M. M.: Synthesis of 3S-pyrrolidinol from L-glutamic acid. Synth. Commun., 16, 1815–1822 (1986). 3. Itoh, N., Morihama, R., Wang, J., Okada, K., and Mizuguchi, N.: Purification and characterization of phenylacetaldehyde

6.

7.

8.

9.

reductase from a styrene-assimilating Corynebacterium strain, ST-10. Appl. Environ. Microbiol., 63, 3783–3788 (1997). Wang, J.-C., Sakakibara, M., Liu, J.-Q., Dairi, T., and Ito, N.: Cloning, sequence analysis, and expression in Escherichia coli of the gene encoding phenylacetaldehyde reductase from styrene-assimilating Corynebacterium sp. strain ST-10. Appl. Microbiol. Biotechnol., 52, 386–392 (1999). Yamada-Onodera, K., Fukui M., and Tani, Y.: Purification and characterization of an alcohol dehydrogenase reducing N-benzyl-3-pyrrolidinone from Geotrichum capitatum. J. Biosci. Bioeng., 103, 174–178 (2007). Yamada-Onodera, K., Yamamoto, H., Kawahara, N., and Tani, Y.: Expression of the gene of glycerol dehydrogenase from Hansenula polymorpha DL-1 in Escherichia coli for the production of chiral compounds. Acta Biotechnol., 22, 355– 362 (2002). Yamada-Onodera, K., Kawahara, N., Tani, Y., and Yamamoto, H.: Synthesis of optically active diols by Escherichia coli transformant cells that express the glycerol dehydrogenase gene of Hansenula polymorpha DL-1. Eng. Life Sci., 4, 413–417 (2004). Lampel, K. A., Uratani, B., Chaudhry, G. R., Ramaley, R. F., and Rudikoff, S.: Charactarization of the developmentally regulated Bacillus subtilis glucose dehydrogenase gene. J. Bacteriol., 166, 238–243 (1986). Galkin, A., Kulakova, L., Tishkov, V., Ezaki, N., and Soda,

384

YAMADA-ONODERA ET AL.

K.: Cloning of formate dehydrogenase gene from a methanol-utilizing bacterium Mycobacterium vaccae N10. Appl. Microbiol. Biotechnol., 44, 479–483 (1995). 10. Yamada-Onodera, K., Yamamoto, H., Emoto, E., Kawahara, N., and Tani, Y.: Characterization of glycerol dehydrogenase from a methylotrophic yeast, Hansenula polymorpha DL-1, and its gene cloning. Acta Biotechnol., 22, 337–353 (2002). 11. Cryer, D. R., Eccleshall, R., and Marmur, J.: Isolation of yeast DNA. Methods Cell Biol., 12, 39–44 (1975). 12. Scrutton, N. S., Berry, A., and Perham, R. N.: Redesign of

J. BIOSCI. BIOENG.,

the coenzyme specificity of a dehydrogenase by protein engineering. Nature, 343, 38–43 (1990). 13. Watanabe, S., Kodaki, T., and Makino, K.: Complete reversal of coenzyme specificity of xylitol dehydrogenase and increase of thermostability by the introduction of structural zinc. J. Biol. Chem., 280, 10340–10349 (2005). 14. Jeffery, J., Chesters, J., Mills, C., Sadler, P. J., and Jornvall, H.: Sorbitol dehydrogenase is a zinc enzyme. EMBO J., 3, 357–360 (1984).