Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.enzyme.chem.msu.ru/~tishkov/Publications/B&B_99_64_187.pdf Äàòà èçìåíåíèÿ: Thu Mar 30 18:41:50 2000 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 23:04:38 2012 Êîäèðîâêà:

Pilot Scale Production and Isolation of Recombinant NAD+- and NADP+-Specific Formate Dehydrogenases
Vladimir I. Tishkov,1,2 Andrey G. Galkin,1,* Vladimir V. Fedorchuk,1 Pavel A. Savitsky,1 Alexandra M. Rojkova,1 Holger Gieren,2 Maria-Regina Kula2
1

Department of Chemical Enzymology, Faculty of Chemistry, M. V. Lomonosov Moscow State University, 119899 Moscow, Russian Federation; telephone: +7-(095)-939-32-08; fax: +7-(095)-939-27-42; e-mail: vit@enzyme.chem.msu.ru or tishkov@ibt022.ibt.kfa-juelich.de 2 Institute of Enzyme Technology, Heinrich-Heine University, Duesseldorf, D-52426 Juelich, Germany
Received 21 August 1998; accepted 15 December 1998

Abstract: The expression of the recombinant wild-type NAD+- and mutant NADP+-dependent formate dehydrogenases (EC 1.2.1.2., FDH) from the methanol-utilizing bacterium Pseudomonas sp. 101 in Escherichia coli cells has been improved to produce active and soluble enzyme up to the level of 50% of total soluble proteins. The cultivation process for E. coli /pFDH8a and E. coli / pFDH8aNP cells was optimized and scaled up to a volume of 100 L. A downstream purification process has been developed to produce technical grade NAD+- and NADP+-specific formate dehydrogenases in pilot scale, utilizing extraction in aqueous two-phase systems. © 1999
John Wiley & Sons, Inc. Biotechnol Bioeng 64: 187­193, 1999.

Keywords: formate dehydrogenase; Pseudomonas sp. 101; expression in E. coli; optimization of expression; large-scale cultivation; aqueous two-phase extraction

INTRODUCTION In methanol-utilizing microorganisms the NAD+-dependent formate dehydrogenase (EC 1.2.1.2, FDH) catalyses the last step in the dissimilatory pathway of methanol to carbon dioxide: HCOO- + NAD+ NADH + CO2 FDH consists of two identical subunits and does not contain metal ions or other prosthetic groups in the active center. FDH is of great interest from scientific and practical considerations. The enzyme belongs to the superfamily of Dstereospecific 2-hydroxy acid dehydrogenases (Vinals et al., 1993) and can be regarded as a model dehydrogenase cataPresent address: Institute for Chemical Research, Kyoto University, Uji, Kyoto-fu 611, Japan Correspondence to: V. Tishkov Contract grant sponsors: INTAS; Russian Foundation of Fundamental Investigations; Russian Federation; Russian Federal Scientific Program; Alexander von Humboldt Foundation Contract grant numbers: Grant 94-1309; RFFI 96-04-49927; RFFI 9615-97054; "Novel Methods of Bioengineering", Grant 5.1-92

lyzing the reaction with the simplest substrate: the formate ion. Structures of the apo- and holo-FDH from Pseudomonas sp. 101 have been solved with high resolution (Lamzin et al., 1994), and site-directed mutagenesis experiments have been carried out to clarify the role of different amino acid residues in the catalytic mechanism of hydride transfer (Tishkov et al., 1993b, 1996). From applied aspects FDH is the best catalyst for in-situ NADH regeneration in the enzymatic synthesis of amino acids, chiral hydroxy acids and esters, alcohols, and other fine chemicals using NAD+-dependent dehydrogenases (Hummel and Kula, 1989; Drauz and Waldmann, 1995; Peters, 1998). The reaction catalyzed by FDH is practically irreversible and 100% conversion of educts is possible in coupled reactions. Formate does not inhibit most other dehydrogenases and the co-product, carbon dioxide, can be easily removed from the reaction mixture. The enzyme has a wide pH optimum for activity and can be obtained from methanol-utilizing microorganisms in large amounts. A process for the production and purification of technical grade FDH from C. boidinii in large scale has been developed (Weuster-Botz et al., 1994) and the enzyme was used for the industrial production of L-tert-leucine (Bommarius et al., 1995). NAD+/NADH are mainly involved in biodegradation processes, while NADP+/NADPH participate in biosynthetic pathways. Many interesting compounds can be obtained only using NADP+-specific dehydrogenases requiring for economic reasons an efficient and inexpensive NADPH regeneration system, which was previously not available (Drauz and Waldmann, 1995). Wild-type FDH cannot be used for NADPH regeneration. Enzymes produced by yeasts do not accept NADP+ at all and FDHs from bacteria have very low specific activity (about 5% compared to the activity with NAD+) and very high Km values (0.10 versus 35 mM for NAD+ and NADP+, respectively). Using a rational design strategy we derived a NADP+-

© 1999 John Wiley & Sons, Inc.

CCC 0006-3592/99/020187-07


specific FDH based on the FDH gene of the bacterium Pseudomonas sp. 101 (Tishkov et al., 1993a). The new enzyme was successfully employed in combination with an alcohol dehydrogenase (Seelbach et al., 1996) and a monooxygenase (Rissom et al., 1997) to synthesize R-alcohols and chiral lactones, respectively. For application in larger scale the enzyme has to be available in sufficient amounts. Presently the commercially available NAD+-dependent FDH is mainly produced from the yeast Candida boidinii (Weuster-Botz et al., 1994). Optimization of the cultivation medium resulted in a space/ time yield up to 3600 U/L per day (Weuster-Botz and Wandrey, 1995). Production of target proteins using recombinant E. coli strains has some advantages compared to systems based on wild strains. During expression of recombinant proteins in E. coli concentrations of 8.5 g/L have been described (Joly et al., 1998). This approach is very flexible and the same expression system could be used for the production of different mutant FDHs with changed coenzyme specificity and increased thermal stability. FDH genes from the bacteria Pseudomonas sp. 101 (Tishkov et al., 1993b), Mycobacterium vaccae N10 (Galkin et al., 1995), Moraxella C-1 (EMBL accession number O08375) and the yeast Hansenula polymorpha (Hollenberg and Janowicz, 1989), Candida methylica (Allen and Holbrook, 1995), and Candida boidinii (Sakai et al., 1997, Slusarczyk et al., in press) have been cloned and expressed in E. coli cells. In shake flask experiments the highest level of FDH expression, 15% of total soluble E. coli proteins, was achieved for the enzyme from C. methylica under the control of the tac promoter (Allen and Holbrook, 1995). Largescale production of FDH using recombinant E. coli cells has not been reported so far. The aim of the present work was to develop up to the pilot scale a low cost process for the production of a biocatalyst for NAD(P)H regeneration based on the recombinant wild-type NAD+- and mutant NADP+-specific formate dehydrogenases from Pseudomonas sp. 101. During this research we wanted to achieve (1) optimization of the FDH expression system in E. coli; (2) optimization of the largescale cultivation of the recombinant E. coli/pFDH strain to obtain high yield of biomass and high level of FDH expression; and (3) adaptation of a downstream process for FDH purification from E. coli homogenates. MATERIALS AND METHODS Enzymes and Biochemicals Restriction endonucleases, T4 DNA kinase, T4 DNA polymerase, and T4 DNA ligase (400 U/ L) were obtained from New England Biolabs (Schwalbach, Taunus, Germany). All chemicals used for genetic engineering experiments and site-directed mutagenesis were molecular biology grade. ATP and dNTPs were purchased from Pharmacia Biotech (Freiburg, Germany), and NAD+ and NADP+ ( 98% purity both) were from Sigma (Deisenhofen,Germany).

Genetic Engineering Methods All operations with DNA (digestion, analytical and preparative agarose gel electrophoresis, DNA extraction from gel, ligation, etc.) were carried according to Maniatis et al. (1982). The introduction of new restriction sites was accomplished by a site-directed mutagenesis using the method of Kunkel et al. (1991). All mutations were controlled by DNA sequencing of both strands of the entire FDH gene with an Applied Biosystems automated DNA sequencer 370A and PCR Taq dye-labeled terminators sequencing kit (Perkin Elmer­Applied Biosystems, Weiterstadt, Germany). Expression The level of the FDH expression in different E. coli/pFDH strains was studied in shake flasks experiments. A single colony was picked from a Petri dish and used to inoculate an overnight culture (37° C, 160 rpm) in 3 mL of 2YT medium (16 g of Bactotrypton, 10 g of yeast extract (Difco, Augsburg, Germany) and 5 g NaCl per liter, pH 7.0) containing 300 g/mL ampicillin. A 2-mL amount of the cell suspension was used to inoculate 200 mL of 2YT medium, 200 g/mL ampicillin in 1-L flasks. Cultivation was carried out at 37°C and 150 rpm. FDH biosynthesis was induced by supplementation of the medium with isopropyl-úthiogalactoside (IPTG final concentration 0.25 mM) after 4 h when cell absorbance A600 reached values about 2.0. At certain time intervals probes were taken and prepared for analysis as described by Klyushnichenko et al. (1997). Enzyme activity was measured spectrophotometrically at 340 nm in 0.1 M K-phosphate buffer (pH 7.0), at 30°C following the production of the reduced coenzyme. The concentrations of NAD+ or NADP+ (for the measurement of corresponding activity of NAD+- and NADP+-specific enzymes) were 1.5 mM, and the concentration of sodium formate was 0.3 M in the assay mixture. The FDH concentration in probes was also estimated taking into account a specific activity 10 and 2.5 U/mg of protein of homogeneous NAD+and NADP+-specific formate dehydrogenases, respectively (Tishkov et al., 1993c; Klyushynichenko et al., 1997). Total protein concentration in the probe was determined by the method of Bradford (1976). The level of FDH expression was expressed as a ratio of FDH and total protein concentrations in the probe. Some probes were subjected to SDS electrophoresis in polyacryl amide gel and the level of expression was estimated by scanning densitometry on the Shimadzu dual-wavelength Flying-Spot Scanner Model CS9000 (Shimadzu GmbH, Duesseldorf, Germany). Cultivation Laboratory and pilot-scale fed-batch cultivations of E. coli/ pFDH8a and E. coli/pFDH8aNP strains were performed in continuously stirred 20 and 300 L tank reactors (Chemap, Volketswil, Switzerland, working volumes 10 and 100 L,

188

BIOTECHNOLOGY AND BIOENGINEERING, VOL. 64, NO. 2, JULY 20, 1999


respectively). The pH was maintained at 7.0 ± 0.1 by the controlled addition of ammonia. The dissolved oxygen level was maintained at 20­25% of air saturation by progressive increase in the agitation rate (40­550 rpm), the sparge volume (air, 0.1­4 L/min per liter of cultivation medium) and the reactor vessel pressure (0­1.3 bar). Foaming was suppressed by the controlled addition of the antifoam agent DEHYSAN Z2111 (Henkel, Dusseldorf, Germany). The ¨ cultivation medium (CM) contained (g/L): KH2PO4, 3.0; Na2HPO4, 6.0; NaCl, 0.5; HCOONa, 2.0; NH4Cl, 0.25; MgSO4, 0.5, and CaCl2 2H20, 0.025. After heat sterilization and cooling to room temperature, a trace metal solution (g/L: ZnSO4 7H2O, 0.58; MnSO4 H2O, 0.30; H3BO3, 0.12; NaMoO4 2H2O, 0.10; CoCl2 6H2O, 0.10; KI, 0.17; CuSO4 5H2O, 0.25; FeCl3 6H2O, 6.0; EDTA, 0.5, and 4 mL of 0.25 M H2SO4 per liter) and thiamin solution (0.1 g/L) were added to the medium (1 mL/L of both solutions) by sterile filtration. Concentrated glucose solution (20% w/v) was added to the medium by sterile filtration just prior to inoculation to provide an initial concentration of 0.5 g/L. The feed medium added during the fed-batch process contained 650 g/L of glucose (heat sterilized), 12 mL/L of the trace element solution, and 12 g/L MgSO4. For induction, the feed medium was supplemented with 50 g/L lactose. Precultures were prepared by overnight cultivation at 37°C of the cells glycerol stock (50 L of stock per liter of medium) in CM containing 1 g/L of glucose and 400 mg/L of ampicillin to a density of A600 0.5­0.8. The inoculum volume was 5% and 1.3% of cultivation medium at laboratory and pilot scale, respectively. Cells were harvested by continuous flow centrifugation, and stored frozen at -20°C.

RESULTS AND DISCUSSION Optimization of FDH Expression in E. coli Cells The native FDH promoter from Pseudomonas sp. 101 is not functional in E. coli, and it was necessary to select an expression system optimal for the host bacterial strain. Different expression systems can be used for the biosynthesis of recombinant proteins in E. coli, the most popular are based on lac, tac, and T7 RNA polymerase promoters (De Boer et al., 1983; Studier and Moffatt, 1986). To create an efficient expression system, high level of protein biosynthesis and high yield of biomass are required. In addition the level of the recombinant protein expression has to be considered, which depends on the initiation of transcription and translation. Furthermore it was desired to produce the recombinant enzyme as a soluble active protein. The soluble enzyme should be resistant to digestion by proteases in E. coli cells to avoid post-translational losses and the generation of fragments as contaminants. Figure 1 and Table I present data improving the FDH expression in E. coli. The plasmid pFDH2 was obtained cloning a 2.3 kbp HindIII­BglII fragment carrying the FDH gene (Tishkov et al., 1993b) in plasmid pUC19 under the control of the lac promoter. This insert contained additional 460 and 670 bp in the 5 -upstream and 3 -downstream regions, respectively, of the FDH gene (Fig. 1). Deletion of a 150 bp HindIII­NruI fragment in the 5 - region and insertion of the 81 bp BamHI-HindIII block with the tac promoter from the plasmid pDR540 (final plasmid pFDH4, see Fig. 1 and Table I) resulted in enzyme expression comprising 5­7% of the total soluble E. coli proteins. An EcoRI restriction site was introduced by mutagenesis behind the TGA stop codon (plasmid pFDH5) to remove 660 bp in the 3 - region. Plasmid pFDH6 was generated cloning from pFDH5 the 1635 bp BamHI­EcoRI fragment containing the tac promoter and the FDH gene, in the phagemid pUC119. Despite the use of the two strong promoters in tandem in pFDH6, the expression level of FDH was still rather low: 8­9% of total soluble protein. It could be explained by a low initiation of translation or by differences in the codon usage in Pseudomonas sp. 101 and E. coli. The (G+C) content in the FDH gene is 67% compared to 51% in the chromosome of E. coli. High level of FDH expression was achieved after substitution of the original ribosomal binding site (RBS) by a RBS with the enhancer of translation from the gene 10 of phage T7 (Olins and Rangwala, 1989) (plasmid pFDH6b). The high increase in the FDH expression level after optimization of the RBS in the plasmid pFDH6a shows that for plasmids pFDH2, pFDH4 and pFDH6 the initiation of translation was the limiting step in the FDH expression in E. coli. The best plasmid pFDH8a was obtained after optimization of the RBS and the exchange of some of the original codons in the starting region of the FDH gene for optimal codons from E. coli. This plasmid produced FDH corresponding to 45­50 % of the soluble proteins in three different E. coli strains: TG1, JM109, and BL21/pLysS. A functional FDH

Extraction in Aqueous Two-Phase Systems Aqueous two-phase systems were used for the large-scale purification of recombinant FDH. Frozen biomass was resuspended in 0.1 M potassium phosphate buffer, pH 8.5, to get 25% wet cell mass concentration. The cell suspension was passed two times through a bead mill desintegrator (Netzsch, Selb, Germany). The first aqueous two-phase system had the following composition: 50% (w/w) cell homogenate, 18% (w/w) PEG 1550, 6% (w/w) K2HPO4, 5% (w/w) sodium formate, and 21% additional water. After intensive mixing, the two-phase system was allowed to separate overnight. The upper phase was pumped to another vessel and the second two-phase system generated. It had a total composition of 50 % top phase, 3% PEG 20,000, 7.5% KH2PO4, 1.0% K2HPO4, 2.5% sodium formate, and 36% additional water. The bottom phase obtained after mixing and phase separation was desalted and concentrated using an ultrafiltration module Sartocon II (Sartorius AG, Gottingen, Ger¨ many) equipped with cellulose triacetate membranes (cutoff 20,000 Da), using 0.1 M sodium phosphate, 10 mM EDTA buffer, pH 7.0, for diafiltration.

TISHKOV ET AL.: PILOT-SCALE PRODUCTION OF FORMATE DEHYDROGENASES

189


Figure 1. Optimization of the expression system of FDH in E. coli cells. RBS, ribosome binding site. Structures of RBS1, RBS2, and RBS3 are presented in Table I. Plasmids pFDH2 and pFDH4 are based on plasmid pUC19, and plasmids pFDH6, pFDH6b, and pFDH8a are derived from phagemid pUC119. The lac promoter is a part of the original plasmid pUC19 and phagemid pUC119. The tac promoter was inserted as 81 bp BamHI­HindIII fragment from the plasmid pDR450. White boxes denote noncoding DNA sequences in 5 - upstream and 3 -downstream regions flanking the FDH gene in the Pseudomonas sp. 101 chromosome. At the 3 - end the inserted FDH gene is ligated through BglII or EcoRI restriction sites in the polylinker of plasmid pUC19 and phagemid pUC119, respectively.

gene based on Tishkov et al. (1993b) but containing some site-directed mutations that altered the coenzyme specificity and thermostability of the FDH was cloned in plasmids with the expression systems described in Table I. Details of the construction of this gene and characterization of the different mutant proteins will be described elsewhere (Tishkov, Egorov, and Galkin, manuscript in preparation; Galkin, Rojkova, Fedorchuk, Serov, Kulakova, Matorin, Savitsky, Kula, and Tishkov, manuscript in preparation). There were no detectable differences in the expression of wild-type NAD+-dependent and mutant NADP+-specific FDHs. The enzyme was expressed solely as a soluble protein and efficiently extracted in solution. No evidence for inclusion bodies or aggregated FDH was found in the debris fraction after cell disruption and centrifugation as judged by SDSelectrophoresis in polyacryl amide gel (data not shown). Pilot Scale Cultivation of FDH Plasmids pFDH8a and pFDH8aNP containing the genes of NAD+- and NADP+-specific enzymes, respectively, were
Table I.

used for pilot scale cultivation employing E. coli TG1, TG1/ pLysS, or BL21/pLysS cell lines as the host strain. The strains TG1 as well as BL21 can be grown to high cell density (Korz et al., 1995). The use of the second plasmid pLysS expressing lysozyme from bacteriophage T7 facilitates cell disruption after cultivation (Studier, 1991). Ampicillin (400 mg/L) was used only to supplement the medium of the preculture but was omitted from the main culture. In the pilot cultivation in 100-L scale a sensitive glucose analyzer was not available. The addition of glucose was therefore based on the experimental data relating growth rate and biomass yield found in preliminary cultivations in 10-L scale (data not shown). The concentration of oxygen was maintained at 20­25% of saturation. Before induction cells were grown at 37°C. With the start of the induction the temperature was decreased to 30°C to prevent premature cell lysis. Lower growth temperature also facilitates the production of soluble recombinant proteins and prevents production of inclusion bodies (Schein and Noteborn, 1988). Instead of IPTG FDH biosynthesis was

Optimization of expression of the recombinant FDH from Pseudomonas sp. 101 in E. coli cells. Distance between promoter and RBS, bp 460 360 360 90 45 Expression level, % of soluble protein ATG ATG ATG ATG ATG 2­3 5­7 8­9 40­43 45­50

Plasmid pFDH2 pFDH4 pFDH6 pFDH6b pFDH8a

Promoter lac + + + +

Ribosome binding site AGG AGG AGG TTAACTTTA ag AAGG A AG a AG a AG a AG a GGA tgcccgc tgcccgc tgcccgc t at a c at tcc a t a t

lac lac lac lac

tac tac tac tac

190

BIOTECHNOLOGY AND BIOENGINEERING, VOL. 64, NO. 2, JULY 20, 1999


Table II. Dependence of total and specific yield of NADP+-dependent FDH in 10-L fermentations on the induction at different cell absorption. Cell absorbance at induction, A600 4.2 5.1 6.2 7.0 8.1 10.0 Biomass yield (dry weight), g 48 60 71 71 86 92 Total yield, units 17300 21300 24850 25350 18100 4970 Specific yield, U/g of dry biomass 360 353 350 357 210 54

technical reasons the cultivation was stopped after 21 h when cells growth and FDH expression were still in the log phase (Fig. 3B). Therefore it should be possible to increase the biomass yield and total activity at least two times. As it was shown in our previous work (Klyushnichenko et al., 1997), both, NAD+- and NADP+-specific FDH are resistant to E. coli proteases. Using MALDI-TOF techniques (matrix-assisted laser desorption/ionisation time of flight mass spectrometry) it was demonstrated that FDH from Pseudomonas sp. 101 expressed in E. coli is a full size 400-aa protein and is not digested during the whole cultivation period (Klyushnichenko et al., 1997). Aqueous Two-Phase Extraction System Aqueous two phase extraction was successfully used for the large scale production (up to mega units range) of technical grade NAD+-dependent formate dehydrogenase from the yeast C. boidinii (Weuster-Botz et al., 1994) as well as numerous other enzymes. This method of purification is very flexible. It avoids difficult solid/liquid separation steps for the removal of cell debris and can be easily scaled up 10­10,000 times (Hustedt et al., 1985; Kula, 1990). Recombinant NAD+- and mutant NADP+-dependent FDHs from Pseudomonas sp. 101 expressed in E. coli were isolated by extraction in aqueous two phase systems based on polyethylene glycol (PEG)/salt (Table III). Identical conditions were used for the purification of NAD+- and NADP+dependent enzymes and about equal results were obtained. In the first step FDH was extracted to the upper, PEG-rich phase (partition coefficient 100) and separated from cell debris and DNA. The main difference of the first system compared to the isolation of FDH from C. boidinii is the use of PEG 1550 only instead of a mixture of PEG 400 and PEG

induced by the controlled addition of lactose. Usually, in shaking flasks and batch cultivation, induction is started at a cell absorbance at 600 nm (A600) of less than 1. At higher A600 values cells start to loose plasmids rather quickly. In fed-batch cultivation the limited growth rate provides better plasmid maintenance. In our case it was found that the optimal cell absorbance for induction is A600 6­7 (Table II). Induction at higher cell absorption resulted in a decrease of the total and cell-specific FDH yield. Induction at lower cell absorption did not influence the level of expression but decreased significantly the growth rate. Besides early cell lysis was observed. After optimization of the cultivation conditions in 10-L scale, the level of FDH expression was about 50% of total soluble proteins (Fig. 2A). Data for the cultivation of NADP+-dependent FDH in 100 L are presented in Fig. 3. Taking into account that the total volume increased to 120 L by the feeding regime, the biomass yield was 8 g dry cells/L, the activity yield was 1650 U/L and the total yield about 200,000 U. The last value corresponds to 800,000 U for the NAD+-specific FDH. The enzyme content was more than 30% of total soluble protein (Fig. 2B, line 5). Due to

Figure 2. Expression of the mutant NADP+-specific FDH in E. coli/pFDH8aNP strain during fed-batch fermentations. (A) Cultivation in 10 L. M, molecular weight markers (Boehringer Mannhem); lines 1­6, 0, 3, 6, 9, 12, and 15 h after induction, respectively; line 7, purified FDH. (B) Cultivation in 100 L. M, molecular weight markers (Sigma); lines 1­5, 0, 3, 7, 11, and 15 h after induction, respectively; line 6, purified FDH; line 7, FDH preparation after extraction in aqueous two-phase systems.

TISHKOV ET AL.: PILOT-SCALE PRODUCTION OF FORMATE DEHYDROGENASES

191


was subjected to diafiltration to concentrate the protein and remove salts as well as residual PEG. The purity of the FDH preparations after these steps is about 45­55% and depends on the level of FDH expression in E. coli. Line 7 in Fig. 2B shows the NADP+-specific FDH purified from biomass obtained by cultivation in 100-L scale. The final enzyme preparations had a specific activity of 0.9 and 3.6 U/mg for NADP+- and NAD+-specific FDHs, respectively (Table III). High yield at each step ( 90%) resulted in 75% overall recovery of activity. The enzyme can be used for in situ coenzyme regeneration processes. After addition of 20% glycerol to prevent bacterial contamination, technical grade FDH may be stored at +4°C at least 10­12 months without loss of activity. If desired, the specific activity of the final FDH preparation can be increased by heat treatment of the cell suspension before the first extraction step. In preliminary small scale experiments it was found that 5 min heat treatment at 60°C in the presence of 10% w/w sodium formate removed 17­20% of E. coli proteins by heat coagulation without any loss of FDH activity. CONCLUSION
Figure 3. Fermentation of E. coli/pFDH8aNP by glucose fed-batch operation followed by induction with a glucose/lactose feed mixture. (A and B) normal and logarithmic ordinate scales, respectively.

1550 for the yeast enzyme (Weuster-Botz et al., 1994). In the second extraction step FDH was partitioned to the bottom, salt-rich phase. To achieve a low partition coefficient (k < 0.06) we had to decrease the pH value to 6.0­6.2 instead of pH 8.0 used for C. boidinii FDH (Weuster-Botz et al., 1994). After phase separation the enzyme solution

The mutant FDH from Pseudomonas sp. 101 is one of the best catalyst for NAD(P)H regeneration. This enzyme has higher specific activity, better affinity to formate and the highest stability compared to FDH from other sources. The system developed for the pilot scale production of recombinant NADP+- as well as NAD+-dependent FDH from Pseudomonas sp. 101 gives higher space time yield (U/L per day), activity yield (U per g of dry biomass), and higher enzyme recovery after purification compared to the continuous production of wild type FDH with the yeast C. boidinii

Table III. Purification of NAD+- and NADP+-specific formate dehydrogenases using aqueous two-phase systems.
a

Purification step Cell homogenate after bead-mill desintegration Phase system 1

System components

Step yield, % 100 92

Total yield, % 100 92

Phase system 2

Diafiltration

25% wet weight cell suspension in 0.1 M potassium phosphate buffer, pH 8.5 Cell homogenate, 50% PEG 1550, 18% K2HPO4, 6% HCOONa, 5% Additional water, 21% pH 8.0­8.5 Partition coefficient,b 100 Upper phase from system 1, 50% PEG 20000, 3% KH2HPO4, 7.5% K2HPO4, 1.0% HCOONa, 2.5% Additional water, 36% pH 6.0­6.2 Partition coefficientb <0.06 0.1 M potassium phosphate buffer, 10 mM EDTA, pH 7.0

90

83

90

75

a b

All percentage values are shown as a w/w %. Ratio of FDH concentration in top and bottom phases.

192

BIOTECHNOLOGY AND BIOENGINEERING, VOL. 64, NO. 2, JULY 20, 1999


as described by Weuster-Botz et al. (1994). Large-scale cultivations and two-phase extraction isolation of our recombinant Pseudomonas FDH and the formate dehydrogenase from C. boidini (Weuster-Botz et al., 1994) were done in the same equipment and our calculations show that the production cost of the recombinant bacterial NAD+- and NADP+dependent FDHs should be similar or less than the same wild type enzyme from C. boidinii (Weuster-Botz et al., 1994).
Different parts of this work was supported by the INTAS Grant 94-1309, by the grant of the Russian Foundation of Fundamental Investigations RFFI 96-04-49927, by the grant of the President of Russian Federation for young professors RFFI 96-15-97054, by the grant 5.1-92 of the Russian Federal Scientific Program "Novel Methods of Bioengineering" and by a fellowship for VIT from the Alexander von Humboldt Foundation.

References
Allen SJ, Holbrook JJ. 1995. Isolation, sequence and overexpression of the gene encoding NAD-dependent formate dehydrogenase from the methylotrophic yeast Candida methylica. Gene 162:99­104. Bommarius AS, Schwarm M, Stingl K, Kottenhanh M, Huthmacher K, Drauz K. 1995. Synthesis and use of enantiomerically pure tertleucine. Tetrahedron: Asymmetry 6:2851­2888. Bradford MM. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein­dye binding. Anal. Biochem. 72:248­252. De Boer HA, Coomstock LJ, Vasser M. 1983. The tac promoter: A functional hybrid derived from trp and lac promoters. Proc Natl Acad Sci USA 80:21­25. Drauz K, Waldmann H, editors. 1995. Enzyme Catalysis in Organic Synthesis. Weinheim, Germany: VCH. p 597. Galkin A, Kulakova L, Tishkov V, Esaki N, Soda K. 1995 Cloning of formate dehydrogenase gene from a methanol-utilizing bacterium Mycobacterium vaccae N10. Appl Microbiol Biotechnol 44:479­483. Hollenberg CP, Janowicz Z. 1989. DNA-molecules coding for FMDH control regions and structured gene for a protein having FMDHactivity and their uses. European Patent Application EP 0 299 108 A1, Bulletin 89/03. Hummel H, Kula M-R. 1989. Dehydrogenases for the synthesis of chiral compounds. Eur J Biochem 184:1­13. Hustedt H, Kroner KH, Kula M-R. 1985. Applications of Phase Partitioning in Biotechnology. In: Walter H, Brooks DE, Fisher D, editors. "Partitioning in Aqueous Two-Phase Systems: Theory, Methods, Uses and Applications to Biotechnology." New York: Academic Press. p 529­587. Joly JC, Leung WS, Swartz JR. 1998. Overexpression of Escherichia coli oxidoreductases increases recombinant insulin-like growth factor-I accumulation. Proc Natl Acad Sci USA 95:2773­2777. Klyushnichenko V, Tishkov V, Kula M-R. 1997. Rapid SDS-gel capillary electrophoresis for the analysis of recombinant NADP+-dependent formate dehydrogenase during expression in E. coli cells and purification. J Biotechnol 58:187­195. Korz DJ, Rinas U, Hellmuth K, Sanders EA, Decker W-D. 1995. Simple fed-batch technique for high cell density cultivation of Escherichia coli. J Biotechnol 39:59­65. Kula M-R.1990. Trends and future prospects of aqueous two-phase extraction. Bioseparation 1:181­189. Kunkel TA, Bebenek K, McClary J. 1991. Efficient site-directed mutagenesis using uracil-containing DNA. Methods Enzymol 204:125­139.

Lamzin VS, Dauter Z, Popov VO, Harutyunyan EH, Wilson KS. 1994. High resolution structures of holo and apo formate dehydrogenase. J Mol Biol 236:759­785. Maniatis T, Fritsch EE, Sambrook J. 1982. Molecular cloning: A laboratory approach. Plainview, NY: Cold Spring Harbor Laboratory Press. Olins P, Rangwala SH. 1989. A novel sequence element derived from bacteriophage T7 mRNA acts as an enhancer of translation of the lacZ gene in Escherichia coli. J Biol Chem 264:16973­16976. Peters J. 1998. Dehydrogenases: Characteristics, design of reaction conditions, and applications. In: Rehm H-J, Reed G, editors. Biotechnology, 2nd edition, Vol. 8a. Weinheim, Germany: Wiley-VCH. p 393­474. Rissom S, Schwarz-Linek U, Vogel M, Tishkov VI, Kragl U. 1997. Synthesis of chiral -lactones in a two-enzyme system of cyclohexanone monooxygenase and formate dehydrogenase with integrated bubblefree aeration. Tetrahedron: Assymetry 8:2523­2526. Sakai Y, Murdanoto AP, Konishi T, Iwamatsu A, Kato N. 1997. Regulation of the formate dehydrogenase gene, FDH1, in the methylotrophic yeast Candida boidinii and growth characteristics of an FDH1disrupted strain on methanol, methylamine, and choline. J Bacteriol 179:4480­4485. Seelbach K, Riebel B, Hummel W, Kula M-R, Tishkov VI, Egorov AM, Wandrey C, Kragl U. 1996. A novel, efficient regeneration method of NADPH by using new formate dehydrogenase. Tetrahedron Lett 37: 1377­1380. Schein CH, Noteborn MHM. 1988. Formation of soluble recombinant proteins in Escherichia coli is favored by lower growth temperature. Bio/Technology 6:291­294. Sluscarzyk H, Pohl M, Kula M-R. Cloning and stabilization of NADdependent formate dehydrogenase from Candida boidinii by sitedirected mutagenesis. In: Ballesteros A, Plou FJ, Iborra JL, Halling P, editors. Stability and Stabilization of Biocatalysts. Amsterdam: Elsevier. [in press] Studier FW 1991. Use of bacteriophage T7 lysozyme to improve an inducible T7 expression system. J Mol Biol 219:37­44. Studier FW, Moffatt BA. 1986. Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J Mol Biol 189:113­130. Tishkov VI, Galkin AG, Egorov AM. 1993a. Recombinant NAD+dependent formate dehydrogenase: New properties for biotechnology application. International Conference on Enzyme Engineering XII, Deu Ville, France. Tishkov VI, Galkin AG, Marchenko GN, Tsygankov YD, Egorov AM. 1993b. Formate dehydrogenase from methylotrophic bacterium Pseudomonas sp. 101: Gene cloning and expression in Escherichia coli. Biotechnol Appl Biochem 18:201­207. Tishkov VI, Galkin AG, Marchenko GN, Egorova OA, Sheluho DV, Kulakova LB, Dementieva LA, Egorov AM. 1993c. Catalytic properties and stability of a Pseudomonas sp. 101 formate dehydrogenase mutants containing Cys-255-Ser and Cys-255-Met replacements. Biochem Biophys Res Commun 192:976­981. Tishkov VI, Matorin AD, Rojkova AM, Fedorchuk VV, Savitsky AP, Dementieva LA, Lamzin VS, Mezentzev AV, Popov VO. 1996. Sitedirected mutagenesis of formate dehydrogenase active centre: Role of the is332­Gln313 pair in enzyme catalysis. FEBS Lett 390:104­108. Vinals C, Depiereux E, Feytmans E. 1993. Prediction of structurally conserved regions of D-specific hydroxy acid dehydrogenases by multiple alignment with formate dehydrogenase. Biochem Biophys Res Commun 192:182­188. Weuster-Botz D, Paschold H, Gieren H, Kula M-R, Wandrey C. 1994. Continuous computer controlled production of formate dehydrogenase (FDH) and isolation on a pilot scale. Chem Eng Technol 17:131­137. Weuster-Botz D, Wandrey C. 1995. Medium optimization by genetic algorithm for continuous production of formate dehydrogenase. Process Biochem 30:563­571.

TISHKOV ET AL.: PILOT-SCALE PRODUCTION OF FORMATE DEHYDROGENASES

193