Cloning, functional expression and biochemical ... - Springer Link

2 downloads 0 Views 68KB Size Report
showing 29% identity to a C α-dehydrogenase of Pseudo- monas paucimobilis ... of a stereoselective alcohol dehydrogenase from Pseudomonas fluorescens ...
Appl Microbiol Biotechnol (2002) 59:483–487 DOI 10.1007/s00253-002-1036-2

S H O RT C O N T R I B U T I O N

P. Hildebrandt · A. Musidlowska · U. T. Bornscheuer J. Altenbuchner

Cloning, functional expression and biochemical characterization of a stereoselective alcohol dehydrogenase from Pseudomonas fluorescens DSM50106 Received: 20 February 2002 / Revised: 22 April 2002 / Accepted: 26 April 2002 / Published online: 26 June 2002 © Springer-Verlag 2002

Abstract Sequencing of a genomic library prepared from Pseudomonas fluorescens DSM 50106 identified an orf showing 29% identity to a C α-dehydrogenase of Pseudomonas paucimobilis and high homology to several sequences with unknown functions derived from genome projects. The corresponding gene adhF1 encodes a dehydrogenase of 296 amino acids with a calculated molecular mass of 31.997 kDa. The gene was functionally expressed in E. coli using a rhamnose inducible expression system. The resulting recombinant enzyme was active in the pH range 6–10 (best pH 8) and at 5–25 °C. This dehydrogenase converts cyclic ketones to the corresponding alcohols utilizing the cofactor NADH. The highest activity was found for cyclohexanone. The enzyme also exhibits high stereoselectivity in the desymmetrization of the prochiral ketone acetophenone, producing optically pure (R)-α-phenyl ethanol (>99%ee) at high conversion (95%).

Introduction Dehydrogenases are enzymes belonging to the class of oxidoreductases (E.C.1.x). Within this class, alcohol dehydrogenases (E.C.1.1.1.1, also named keto-reductases) represent an important group of biocatalysts, because they can be used efficiently in the synthesis of optically pure compounds by reduction of prostereogenic ketones to the corresponding optically active alcohols. From a practical point of view, only those dehydrogenases which use NADH as cofactor are of importance, because for biocatalysts depending on NADPH much less efficient cofactor recycling systems are available.

A range of alcohol dehydrogenases useful for organic synthesis has been already described in the literature (Drauz and Waldmann 1995; Faber 2000; Hummel and Kula 1989; Peters 2000; Wong and Whitesides 1994). Amongst these, the most frequently used are from yeast, horse liver and Thermoanaerobium brockii, which also differ considerably in their substrate specificity and stereopreference. Only a few ADHs from Pseudomonas spp. have been described so far. An enzyme from strain ATCC49688 (Shen et al. 1990) exhibits only limited substrate tolerance, whereas an enzyme from Pseudomonas sp. PED (Bradshaw et al. 1992) accepts a wide range of substrates similar to ADHF1 and also shows activity in the presence of isopropanol. However, none of these ADHs is available in recombinant form. Recently, we identified the gene encoding a lactonespecific esterase from Pseudomonas fluorescens DSM 50106 within its genomic library (Khalameyzer et al. 1999). The deduced amino acid sequence of two other Cterminal truncated orfs showed homology to a cyclohexanone monooxygenase and an alkane hydroxylase, respectively. Cycloalkanone monooxygenases are known to catalyze the enzymatic Baeyer-Villiger oxidation (Stewart et al. 1998) (thus they are also named BaeyerVilliger monooxygenases, BVMO), yielding lactones, which in turn are esterase substrates. Further sequencing identified a third orf showing 29% identity to a C α-dehydrogenase of Pseudomonas paucimobilis (Masai et al. 1993) and high homology to sequences derived from genome projects, which presumably encode for putative dehydrogenases. In this paper, we describe the sequencing, cloning, and biochemical characterization of this alcohol dehydrogenase.

P. Hildebrandt · A. Musidlowska · U.T. Bornscheuer (✉) Institute of Chemistry and Biochemistry, Department of Technical Chemistry and Biotechnology, Greifswald University, Soldmannstrasse 16, 17487 Greifswald, Germany e-mail: [email protected] Tel.: +49-3834-864367, Fax: +49-3834-864346

Materials and methods

J. Altenbuchner Institute of Industrial Genetics, Stuttgart University, Allmandring 31, 70569 Stuttgart, Germany

E. coli JM109 (Yanish-Perron et al. 1985) was used as host for transformation of plasmid DNA, E. coli HB101 F'lac(Tn1739tnpR) (Altenbuchner 1993) for infection with λRES-phages. The strains

Bacterial strains, plasmids and growth conditions

484 were grown in LB liquid media or on LB agar plates at 37°C (Sambrook et al. 1989). Media were supplemented with 100 µg/ml ampicillin or 50 µg/ml kanamycin for selection of plasmids. The vector pIC20H was used for cloning and DNA sequencing experiments (Marsh et al. 1984) and the vector pJOE2775 with the rhamnose-inducible rhaBAD promoter (Bornscheuer et al. 1998) for expression of ADHF1 in E. coli JM109. Dehydrogenase production was induced upon addition of rhamnose (final concentration 0.2%) and cultivation continued for 5 h. Cells were collected by centrifugation (Heraeus Labofuge 400R, 4,000 g, 10 min, 4°C) and washed twice with sodium phosphate buffer (50 mM, pH 7.5, 4°C). Cells were disrupted by sonification on ice for 12 min at 50% pulse and 50% power (Bandelin HD 2070, MS73, Berlin, Germany). Cell debris was removed by centrifugation (4,000 g, 30 min, 4°C) and the supernatant was directly used for reductions or lyophilized and stored at 4°C. Protein content was determined using a bicinchoninic acid kit (Pierce, Rockford, Ill., USA) with bovine serum albumin as protein standard. DNA manipulation techniques Restriction enzymes and DNA modifying enzymes were obtained from Roche Applied Science Mannheim. For restriction enzyme analysis and cloning experiments, standard procedures were used (Sambrook et al. 1989). Plasmid DNA was isolated according to a published protocol (Kieser 1984). E. coli was transformed with plasmid DNA as described by Chung et al. (1989). PCR reactions were done in a volume of 100 µl with 1 ng plasmid DNA, 30 pMol primer, 0.2 mM dNTP-mix, 10% DMSO, 2.5 units Pwo polymerase and 1× reaction buffer provided by the supplier. The following primers were used S2995: 5′-AAA ACA TAT GAA GTC ATT CAA CGG CC-3′ and S2996: 5′-AAA AGG ATC CGA GAC GGG GCT CGT CGT T-3′. DNA was first heated to 100°C for 2 min and then amplified in a Minicycler (Biozym Diagnostic GmbH) in 30 cycles, 1 min denaturation at 94°C, 1.5 min annealing (5°C below the melting temperature of the primer) and 1.5 min extension at 72°C.

ny) to a nitrocellulose membrane. The detection of the His-tagged protein was done on the nitrocellulose using the Ni–NTA conjugate according the manufacturer's instructions (QIApress detection system, Ni-NTA alkaline phosphate-conjugate; Qiagen, Hilden, Germany). The carboanhydrase (29 kDa) present in the low-molecular-weight standard mixture reacts with the Ni–NTA conjugate and thus served as a control. Substrate specificity of ADHF1 and kinetic data Dehydrogenase activity of the lyophilized enzyme was determined using acetophenone as model compound. The standard reaction mixture (250 µl) consisted of 6.4 µmol substrate dissolved in isopropanol, 1.25 mg protein (crude extract) in Tris-HCl (0.1 M, pH 8.0) and 20% (v/v) isopropanol as substrate for the NADH-recycling dehydrogenase. All reactions were performed at room temperature unless stated otherwise. All values were determined in triplicate. At the end of the reaction, the mixture was extracted with two volumes chloroform and the organic phase was dried over anhydrous Na2SO4. The reaction components were then analyzed by gas chromatography (Shimadzu GC 14A, Tokyo, Japan, equipped with a flame ionization detector, Integrator C5RA) using a chiral column (Heptakis (2,6-O-methyl-3-O-pentyl)-β-cyclodextrin, 25 m×0.25 mm ID; Macherey and Nagel, Düren, Germany) and the following temperature programs: (1) 120°C isothermal, retention times: acetophenone, 1.55 min, (R)-α-phenylethanol, 2.77 min, (S)-α-phenylethanol, 2.99 min; Absolute configurations were assigned using commercially available (R)-α-phenylethanol; (2) temperature program 90°C 1 min heating at 5°C/min to 120°C, retention times: cyclopentanone, 0.95 min, cyclopentanol, 1.37 min; cycloheptanone, 2.85 min, cycloheptanol, 4.29 min; (3) temperature program 90°C 5 min heating at 5°C/min to 120°C, retention times: cyclohexanone, 1.42 min, cyclohexanol, 2.12 min. Various attempts to obtain active purified ADHF1 based on the His6-tag or using conventional purification methods failed, although enrichment of the protein from the soluble fraction after cell breakage was possible (data not shown). Temperature and pH profiles, conversion of ketones, and determination of kinetic data were thus performed using the crude cell extract.

DNA sequence analysis DNA sequencing of the 4.3 kb MunI/BamHI fragment was done according to the chain termination method with double stranded plasmid DNA as template. Two strategies were employed. Fragments generated with NaeI and MscI were subcloned in pIC20H. Plasmid pFIS5 and the deletion derivatives were sequenced using the Cy5-labelled M13 universal and reverse primer with the ALFexpress AutoRead sequencing kit (Amersham Pharmacia Biotech). Primer walking was performed using oligonucleotides from MWG Biotech, Ebersberg, and the Cy5-dATP labeling mix in combination with the ALFexpress AutoRead sequencing kit. The reaction products were separated on a 5.5% Hydrolink Long Ranger gel matrix in an ALFexpress DNA sequencer for 12 h at 55 °C, 800 volt and 0.5×TBE buffer. The nucleotide sequence was analysed with the GCG program package (Devereux et al. 1984), version 8.01. Database searches were done with the programs BLASTX, BLASTP and BLASTN using the electronic mail server from the National Center for Biotechnology Information, Bethesda, Md., USA (Altschul et al. 1990). The sequence of ADHF1 was deposited in the GenBank database under the accession code AF090329. Electrophoresis and blotting Proteins were analyzed by a 12% sodium dodecyl sulfate polyacrylamide gel (Sambrook et al. 1989). The low-molecular-weight standard mixture from Sigma was used as a source of reference proteins. After electrophoresis the gel was stained with Coomassie brilliant blue or the proteins were electrophoretically transferred (1 h at 1 mA/cm2 gel) using a semi-dry blotting system (Panther semi-dry electroblotter, Model HEP-1; Peqlab, Erlangen, Germa-

Results The construction of a genomic library from P. fluorescens in the λ RESIII phage and the isolation of the plasmid pJOE2967 exhibiting esterase activity in E. coli have been described elsewhere (Khalameyzer et al. 1999). A 3.2 kB MunI/BamHI fragment of pJOE2967 was sequenced and revealed the presence of the esterase gene estF1 somewhere in the center of the fragment and two C-terminal truncated ORFs (ORF1 and ORF2). The deduced protein sequences of ORF1 and ORF2 showed similarities to cyclohexanone monooxygenases and alkane hydroxylases (Khalameyzer et al. 1999). To complete the ORF1 DNA sequence, a 1.3 kb ApaI fragment was isolated from pJOE2967 which overlapped the 3.2 kb MunI/BamHI fragment in 150 bp. The fragment was completely sequenced. Together with the data from the MunI/BamHI fragment, a continuous DNA sequence of 4,360 bp was obtained. The results of the sequence analysis are shown in Fig. 1. The completed ORF1 encodes a protein of 512 amino acids with a calculated molecular mass of 57.317 kDa and is transcribed in opposite orientation to estF1 and ORF2. Another open reading frame (ORF3) was identi-

485

Fig. 1 Physical map of the 4,360 bp fragment encoding the dehydrogenase ADHF1

fied downstream from ORF1. The start codon of ORF3 overlapped the stop codon of ORF1 in two base pairs (ATGA), indicating a translational coupling of the two ORFs. ORF3 encodes a putative protein of 296 amino acids with a calculated molecular mass of 31.997 kDa. The ORF1- and ORF3-encoded proteins share a high amino acid sequence identity with the putative gene products of two adjacent genes on the chromosome of Pseudomonas aeruginosa (Stover et al. 2000); GenBank accession number H83452; a flavin-containing monooxygenase (55% identity) and an oxidoreductase (60% identity). However, in contrast to the DNA sequence obtained from P. fluorescens, no esterase gene could be identified upstream of the flavin-containing monooxygenase-encoding gene of P. aeruginosa. By database searching with the BLAST program, many other proteins with significant identities in amino acid sequences to ORF3 were found, especially from genome sequencing projects. The highest similarity to proteins where the enzymatic activity was experimentally determined was to the cyclohexanone monooxygenase of Acinetobacter spp. (Chen et al. 1988, 26% identity to ORF1) and to the C α-dehydrogenase of P. paucimobilis (Masai et al. 1993, 29% identity to ORF3).

Fig. 2 SDS-PAGE analysis (left gel) and Ni–NTA–AP conjugate assay (right gel) to confirm expression of ADHF1. Lane 1: low range molecular weight standard (Sigma); lane 2: 10 µg crude protein extract before induction; lanes 3, 4: 10 µg ADHF1 crude protein extract, 3.5 h after induction; lane 5: 5 µg ADHF1 crude protein extract; lane 6: Sigma marker (carboanhydrase, 29 kDa, reacts with the Ni–NTA conjugate) Table 1 Results of the reductions of cyclic ketones and acetophenone catalyzed by ADHF1. All reactions were performed in the presence of NADH and 20% (v/v) isopropanol at 20°C for 20 h (see Materials and methods for details) Substrate

Conversiona (%)

Cyclopentanone Cyclohexanone Cycloheptanone Acetophenone

53 100 51 95b

aAs determined by GC-analysis b>99%ee (R)-α-phenylethanol.

Expression of the adhF1 gene in E. coli The dehydrogenase gene was PCR-amplified from plasmid pJOE2967 using the primers S2995 and S2996. By these means a NdeI site was introduced at the ATG start codon and a BamHI site just before the translation stop codon. The fragment was digested with NdeI and BamHI and inserted into the L-rhamnose-inducible expression vector pJOE3075 (Stumpp et al. 2000), which was cut with the same enzymes. Thereby the C-terminal end of the oxidoreductase gene was fused to six histidine codons of the vector. E. coli JM109 was transformed with this new plasmid pJOE4016 and was grown at 37°C in 500 ml LB media supplemented with ampicillin (100 µg/ml) until the early exponential phase (OD600=0.5–0.6). After induction with rhamnose, cultivation was continued for 5 h and harvested cells were disrupted by sonification, yielding active enzyme in the supernatant (450 µg protein/mg lyophilized extract). Expression was confirmed by SDS-PAGE analysis (Fig. 2, left) indicating the presence of a 33 kDa protein (calcu-

lated: 32.819 kDa) and the Ni–NTA–AP conjugate assay (Fig. 2, right). Substrate specificity of ADHF1 and determination of kinetic data As shown in Fig. 1, ORF3 encoding the gene of ADHF1 is preceded by ORF1, having high sequence homology to a cyclohexanone monooxygenase (BVMO). If both proteins were involved in the metabolic pathway, it can be assumed that cyclic alcohols would be good substrates for ADHF1, yielding the corresponding ketones, which are known to be converted by BVMOs in a Baeyer-Villiger oxidation (Stewart et al. 1998). Consequently, the reduction of several cyclic ketones was investigated and it was found that they are efficiently converted to the corresponding alcohols (Table 1). Moreover, cyclohexanone was reduced with the highest activity and in a quantitative manner.

486

Discussion

Fig. 3 Influence of temperature on the activity of ADHF1 towards cyclohexanone (■) and acetophenone (▲)

Reactions at different temperatures revealed that ADHF1 surprisingly exhibited highest activities in the reduction of cyclohexanone and acetophenone below 30°C (Fig. 3) although P. fluorescens is known as a mesophilic organism. For cyclohexanone, even at 5°C no decrease in activity was observed. ADHF1 shows a sharp pH profile in the reduction of acetophenone, ranging from pH 6 to 10 with highest activities at pH 8 (data not shown). In principle, purification of ADHF1 should be easy, as the His6-tag cloned to the mature enzyme should allow an efficient separation of ADHF1 from other enzymes present in the crude extract. However, various attempts using different commercial matrices and protocols failed to obtain active enzyme, despite an enrichment being possible. In addition, conventional purification methods (i.e., affinity chromatography using NADH-specific resins) failed. On the other hand, we were able to show by control experiments with cell extracts of E. coli DH5α lacking the gene encoding ADHF1, that none of the ketones used in this study were reduced to the corresponding alcohols. Therefore, it is acceptable to use the crude cell extract for biotransformation reactions, as unwanted side-reactions can be excluded. Based on this, the specific activity for acetophenone was determined to 1.1 nmol min–1 mg–1 lyophilisate. From Eadie-Hofstee plots, Vmax was calculated to ~5.5 nmol min–1, at a Km-value of ~157 µM. Although these values are quite low, the efficient reduction of cyclic ketones, (Table 1) as well as the stereoselective conversion of acetophenone, is possible. Here, the biotransformation gave (R)-α-phenylethanol with excellent optical purity (>99%ee) at high conversion (~95%). Beside this substrate, a broad range of other aryl-substituted acetophenones and 3-oxobutyric acid methylester were also converted with high to excellent stereoselectivity and activity (Hildebrandt et al. 2001).

In this work we describe the functional expression of a novel alcohol dehydrogenase from P. fluorescens. This enzyme efficiently reduces a broad range of ketones yielding the corresponding alcohols. The physiological role of ADHF1 might be in the degradation of cyclic alcohols as the encoding orf is surrounded by sequences encoding for an esterase and a Baeyer-Villiger monooxygenase. An interaction between the monooxygenase and the dehydrogenase seems to be very likely as ADHF1 can generate a ketone from a cyclic alcohol, which in turn can be converted into a cyclic lactone. Next, this can be hydrolyzed by the esterase ESTF1, for which it has already been shown that it exhibits exceptionally high activity towards lactones (Khalameyzer et al. 1999). Thus, a stepwise conversion of cyclohexanol to ω-hydroxyhexanoic acid is possible. Furthermore, ADHF1 also shows high enantioselectivity in the reduction of acetophenone as well as a broad range of other prostereogenic compounds and therefore can be considered as a useful catalyst for the production of optically pure alcohols. Unfortunately, it was impossible to obtain highly purified ADHF1 despite various efforts based on the Histag or by conventional purification methods. We assume, that the enzyme is rather sensitive and therefore cannot be recovered in homogeneous and active form. On the other hand, ADHF1 can be directly used as a convenient biocatalyst in form of its crude (lyophilized) cell extract. Although this crude extract also contains an endogenous alcohol dehydrogenase produced by E. coli, this enzyme did not interfere with the substrate specificity and stereoselectivity of ADHF1. From a practical point of view, this is even advantageous, as an efficient cofactor recycling can be achieved by the simple addition of isopropanol. This is oxidized by the E. coli dehydrogenase yielding NADH, which in turn is consumed by ADHF1 (Hildebrandt et al. 2001). A similar principle of efficient cofactor recycling has recently been published for an alcohol dehydrogenase from Rhodococcus ruber (Stampfer et al. 2002), which was also stable up to 50% (v/v) isopropanol. Acknowledgement Financial support from Degussa AG (Hanau, Germany) is gratefully acknowledged.

References Altenbuchner J (1993) A new λRES vector with a built-in Tn1721-encoded excision system. Gene 123:63–68 Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic local alignment search tool. J Mol Biol 215:403–410 Bornscheuer UT, Altenbuchner J, Meyer HH (1998) Directed evolution of an esterase for the stereoselective resolution of a key intermediate in the synthesis of epothilones. Biotechnol Bioeng 58:554–559 Bradshaw CW, Fu H, Shen GJ, Wong C-H (1992) A Pseudomonas sp. alcohol dehydrogenase with broad substrate specificity and unusual stereospecificity for organic synthesis. J Org Chem 57:1526–1532

487 Chen YC, Peoples OP, Walsh CT (1988) Acinetobacter cyclohexanone monooxygenase: gene cloning and sequence determination. J Bacteriol 170:781–789 Chung CT, Niemela SL, Miller RH (1989) One-step preparation of competent Escherichia coli: transformation and storage of bacterial cells in the same solution. Proc Natl Acad Sci USA 86:2172–2175 Devereux J, Haeberli P, Smithies O (1984) A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res 12:387–395 Drauz K, Waldmann H (1995) Enzyme catalysis in organic synthesis, vols 1 & 2. VCH, Weinheim Faber K (2000) Biotransformations in organic chemistry, 4th edn. Springer, Berlin Heidelberg New York Hildebrandt P, Riermeier T, Altenbuchner J, Bornscheuer UT (2001) Efficient resolution of prostereogenic arylaliphatic ketones using a recombinant alcohol dehydrogenase from Pseudomonas fluorescens DSM50106 expressed in E. coli. Tetrahedron: Asymmetry 12:1207–1210 Hummel W, Kula M-R (1989) Dehydrogenases for the synthesis of chiral compound. Eur J Biochem 184:1–13 Khalameyzer V, Fischer I, Bornscheuer UT, Altenbuchner J (1999) Screening, nucleotide sequence and biochemical characterization of an esterase from Pseudomonas fluorescens with high activity toward lactones. Appl Environ Microbiol 65:477–482 Kieser T (1984) Factors affecting the isolation of ccc DNA from Streptomyces lividans and Escherichia coli. Plasmid 12:19–36 Marsh JL, Erfle M, Weykes EJ (1984) The pIC plasmid and phage vectors with versatile cloning sites for recombinant selection by insertional inactivation. Gene 32:481–485 Masai E, Kubota S, Katayama Y, Kawai S, Yamasaki M, Morohoshi N (1993) Characterization of the C alpha-dehydrogenase gene involved in the cleavage of beta-aryl ether by Pseudomonas paucimobilis. Biosci Biotechnol Biochem 57:1655–1659

Peters J (2000) Dehydrogenases – characteristics, design of reaction conditions, and applications. In: Rehm HJ, Reed G, Pühler A, Stadler PJW, Kelly DR (eds) Biotechnology series, vol 8b. Wiley-VCH, Weinheim, pp 391–474 Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: a laboratory manual, 2nd edn. Cold Spring Harbor, N.Y. Shen GJ, Wang YF, Wong CH (1990) A new NAD-dependent alcohol dehydrogenase with opposite facial selectivity useful for asymmetric reduction and cofactor regeneration. J Chem Soc Chem Commun 9:677–679 Stampfer W, Kosjek B, Moitzi C, Kroutil W, Faber K (2002) Biocatalytic asymmetric hydrogen transfer. Angew Chem Int Ed Engl 41:1014–1017 Stewart JD, Reed KW, Martinez CA, Zhu J, Chen G, Kayser MM (1998) Recombinant baker's yeast as a whole-cell catalyst for asymmetric Baeyer-Villiger oxidations. J Am Chem Soc 120: 3541–3548 Stover CK, Pham XQ, Erwin AL, Mizoguchi SD, Warrener P, Hickey MJ, Brinkman FS, Hufnagle WO, Kowalik DJ, Lagrou M, Garber RL, Goltry L, Tolentino E, WestbrockWadman S, Yuan Y, Brody LL, Coulter SN, Folger KR, Kas A, Larbig K, Lim R, Smith K, Spencer D, Wong GK, Wu Z, Paulsen IT (2000) Complete genome sequence of Pseudomonas aeruginosa PA01, an opportunistic pathogen. Nature 406:959–964 Stumpp T, Wilms B, Altenbuchner J (2000) Ein neues L-Rhamnose-induzierbares Expressionssystem für Escherichia coli. BIOspektrum 6:33–36 Wong C-H, Whitesides GM (1994) Enzymes in synthetic organic chemistry. Pergamon, Oxford Yanish-Perron C, Vieira J, Messing J (1985) Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC vectors. Gene 33:103–119