Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.enzyme.chem.msu.ru/~tishkov/Publications/BiomolEng_2006.pdf Äàòà èçìåíåíèÿ: Tue Apr 4 14:43:36 2006 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 23:07:58 2012 Êîäèðîâêà: ISO8859-5

Biomolecular Engineering 23 (2006) 89-110 www.elsevier.com/locate/geneanabioeng

Review

Protein engineering of formate dehydrogenase
Vladimir I. Tishkov a,*, Vladimir O. Popov
a

b

Department of Chemical Enzymology, Faculty of Chemistry, M.V. Lomonosov Moscow State University, Moscow 119992, Russia b A.N. Bach Institute of Biochemistry, Russian Academy of Sciences, Leninsky pr. 33, Moscow 119071, Russia Received 23 November 2005; received in revised form 3 February 2006; accepted 6 February 2006

Abstract NAD+-dependent formate dehydrogenase (FDH, EC 1.2.1.2) is one of the best enzymes for the purpose of NADH regeneration in dehydrogenase-based synthesis of optically active compounds. Low operational stability and high production cost of native FDHs limit their application in commercial production of chiral compounds. The review summarizes the results on engineering of bacterial and yeast FDHs aimed at improving their chemical and thermal stability, catalytic activity, switch in coenzyme specificity from NAD+ to NADP+ and overexpression in Escherichia coli cells. # 2006 Elsevier B.V. All rights reserved.
Keywords: Formate dehydrogenase; Protein engineering; Pseudomonas sp.101; Candida boidinii; Stability; Mutagenesis

Contents 1. 2. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Approaches applied to FDH engineering . . . . . . . . . . . . . . . . . . . . . . 2.1. Structure analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Amino acid sequences alignment . . . . . . . . . . . . . . . . . . . . . . 2.3. Random mutagenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Catalytic mechanism studies and improvement of kinetic parameters . . 3.1. Switch in substrate specificity . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Enhancement of catalytic activity of FDH from C. boidinii . . . . Improvement of FDH operation stability . . . . . . . . . . . . . . . . . . . . . . 4.1. Improvement of chemical stability of FDHs from Pseudomonas 4.2. Improvement of chemical stability of FDH from C. boidinii . . . Improvement of FDH thermal stability . . . . . . . . . . . . . . . . . . . . . . . 5.1. Comparison of FDHs thermostability . . . . . . . . . . . . . . . . . . . 5.2. Improvement of PseFDH thermal stability . . . . . . . . . . . . . . . . 5.3. Improvement of CboFDH thermal stability . . . . . . . . . . . . . . . Change of coenzyme specificity. . . . . . . . . . . . . . . . . . . . . . . . . . . . Expression of FDH genes in E. coli cells . . . . . . . . . . . . . . . . . . . . . Alternative enzymes for NAD(P)H regeneration . . . . . . . . . . . . . . . . 8.1. Glucose dehydrogenase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2. Phosphite dehydrogenase . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..... ..... ..... ..... ..... ..... ..... ..... ..... sp.101 ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ... ... ... ... ... ... ... ... ... and ... ... ... ... ... ... ... ... ... ... ... ... ... .. .. .. .. .. .. .. .. .. M. .. .. .. .. .. .. .. .. .. .. .. .. .. ..... ..... ..... ..... ..... ..... ..... ..... ..... vaccae ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . N . . . . . . . . . . . . . .. .. .. .. .. .. .. .. .. 10 .. .. .. .. .. .. .. .. .. .. .. .. .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 91 91 91 95 95 95 97 98 99 100 100 100 102 102 103 104 105 105 105 106 107 107

3.

4.

5.

6. 7. 8.

9.

* Corresponding author. Tel.: +7 495 939 3208; fax: +7 495 939 2742. E-mail addresses: vit@enz.chem.msu.ru, vitishkov@gmail.com (V.I. Tishkov), vpopov@inbi.ras.ru (V.O. Popov). 1389-0344/$ - see front matter # 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.bioeng.2006.02.003


90

V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110

1. Introduction NAD+-dependent formate dehydrogenase (EC 1.2.1.2, FDH) catalyzes oxidation of formate ion into carbon dioxide coupled to the reduction of NAD+ to NADH: HCOOÐ ? NAD? ! NADH ? CO2 : The enzyme was first discovered in pea seeds more than 60 years ago (see Mathews and Vennesland, 1950; Davidson, 1951 and references therein). The intense studies began in 70s of the last century, mostly for formate dehydrogenase from methylotrophic bacteria and yeast. The interest originated from both practical application of FDH for the purposes of NADH regeneration in the enzymatic processes of chiral synthesis with NAD+-dependent dehydrogenases (Wichmann et al., 1981; Hummel and Kula, 1989), and fundamental studies on the dehydrogenase catalytic mechanism. FDH belongs to the superfamily of D-specific dehydrogenases of 2-hydroxy acids (Vinals et al., 1993). All enzymes of this family have similar structure and almost identical set of catalytically essential amino acid residues in the active center (Popov and Lamzin, 1994; Lamzin et al., 1995). The choice of FDH as a model enzyme was based on the fact that the enzyme catalyzes the simplest reaction among the other enzymes of the superfamily devoid of any proton release or abstraction steps. Last decade active sequencing of genomes resulted in the discovery of FDH genes in various organisms including pathogens such as Staphylococcus aureus (Baba et al., 2002), Mycobacterium avium subsp. paratuberculosis str.k10 (Li et al., 2005), different strains of Bordetella (Parkhill et al., 2003) and Legionella (Chien et al., 2004; Cazalet et al., 2004), Francisella tularensis subsp. tularensis SCHU S4 (Larsson et al., 2005), Histoplasma capsulatum (Hwang et al., 2003), Cryptococcus neoformans var. neoformans JEC21 (Loftus et al., 2005), etc. It was shown that, under specific conditions, FDH could play a key role in cell functioning. For instance, FDH appears to be a stress protein in plants. The enzyme localizes to mitochondria and its biosynthesis sharply increases (up to 9% of total mitochondrial proteins) under stressful conditions (Colas des Francs-Small et al., 1993). The analysis of FDH isoforms ratio was used to identify diseased trees (Weerasinghe et al., 1999). In the case of S. aureus, FDH is one of three overexpressed proteins, when the bacterium grows at biofilm conditions (Resch et al., 2005). Bacterial biofilm infections are particularly problematic because sessile bacteria can often withstand host immune responses and are generally much more tolerant to antibiotics, biocides and hydrodynamic shear forces than their planktonic counterparts. Expression of FDH gene is also phase specific in fungal pathogens (Hwang et al., 2003). The number of papers on FDH grows year by year, and the majority of the works describes FDH application for cofactor regeneration in the processes of chiral synthesis with NAD(P)+-dependent oxidoreductases. General scheme of NAD(P)H regeneration for cofactor coupled enzymatic synthesis of optically active compounds is presented in the next scheme: The main enzyme Ep (dehydrogenase, reductase, monooxygenase, etc.) catalyzes production of a chiral compound using reduced cofactor, while the second enzyme ER (for example, formate dehydrogenase) reduces oxidized coenzyme back to NAD(P)H. In some cases, the same enzyme can catalyze both reactions (Hummel and Kula, 1989). Numerous studies demonstrated that FDH is one of the best enzymes for the purposes of reduced cofactor regeneration (Shaked and Whitesides, 1980; Kula and Wandrey, 1987; Hummel and Kula, 1989; Liese and Villela, 1999; Burton, 2003; Liese, 2005; Wichmann and Vasic-Racki, 2005). The reaction catalyzed by FDH fits all the criteria for NAD(P)H regeneration. (1) The reaction is irreversible under normal conditions. This provides thermodynamic pressure to shift equilibrium of the main reaction and results in a 99-100% yield of the final product. (2) Formate-ion is a cheap substrate, and the reaction product, CO2, can be easily removed from the reaction mixture and does not interfere with the purification of the final product. (3) FDH exhibits a wide pH-optimum of catalytic activity (6.0- 9.0) (Mesentsev et al., 1997). (4) Methanol-utilizing yeast and bacteria provide a high scale enzyme production with a comparatively low production cost. (5) Bacterial and yeast FDHs are sufficiently stable to be used in flow-through reactors for a while. All the above factors determined the use of yeast FDH from Candida boidinii for the purpose of NADH regeneration in the first commercial process of chiral synthesis of tert-Lleucine with dehydrogenase realized by ``Degussa'' (Bommarius et al., 1995). The process is still the biggest one in production volume among the others used to produce optically active compounds with the help of dehydrogenases. Under the leadership of Profs. M.-R. Kula and C. Wandrey, the methods for cultivation of C. boidinii yeast and enzyme purification at the level of millions of activity units were developed (Weuster-Botz et al., 1994). Unfortunately, native FDHs have some disadvantages. First, their operational stability is rather low due to the presence of active Cys residues. Chemical modification or oxidation of these residues results in fast enzyme inactivation. Second, there are no native FDHs of the discussed family, which use NADP+ as a cofactor, and third, the production cost of FDH from native strains of methylotrophic bacteria or yeast was still high enough to use the enzyme in development of novel commercial processes of chiral synthesis. The review summarizes the experiments on FDH protein engineering based on directed and random mutagenesis which permitted to produce a new


V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110

91

generation of biocatalysts for NAD(P)H regeneration exhibiting improved and novel kinetic properties, increased chemical and thermal stability, and lower production costs. Since the key experiments have been performed with FDH from methylotropic yeast C. boidinii and bacterium Pseudomonas sp.101, these enzymes will be in the focus of the review. One can specify the following directions for FDH studies requiring mutagenesis: - catalytic mechanism studies and improvement of kinetic properties; - increase in chemical stability; - increase in thermal stability; - switch in coenzyme specificity; - crystallization and refinement of X-ray structure; - increase in the level and rate of FDH gene expression in Escherichia coli. The available information on FDH mutations is summarized in Table 1. Unfortunately, all experiments cannot be covered within a single review, therefore, we limit ourselves to consideration of the most important mutations which were critical for the production of new generation of recombinant FDH biocatalysts for NAD(P)H regeneration. 2. Approaches applied to FDH engineering 2.1. Structure analysis X-ray data analysis was used to select mutation positions for FDH from Pseudomonas sp.101 and highly homologous (different in only two aa residues) FDH from Mycobacterium vaccae N10. High resolution structures are available from PDB for apo-PseFDH (2NAC) and the ternary complex (PseFDHNAD+-azide) (2NAD) resolved in 1993 (Lamzin et al., 1994). Recently, some other complexes of the enzyme have been crystallized and respective structures solved (Table 2). The analysis of PseFDH structures in the complex with formate, ADP-ribose, NADH and (NADH + formate) shows their intermediate character between 2NAC and 2NAD structures, i.e. apo-enzyme transformation into a holo-enzyme. All the complexes have been obtained with native enzyme purified from Pseudomonas sp.101. The presence of seven additional amino acid residues at the C-terminus of recombinant wtPseFDH (Tishkov et al., 1991) interferes with crystallization. The deletion of these residues by means of mutagenesis resulted in production of crystals of recombinant FDH. The crystals of full size 400 aa polypeptide have been produced for two mutant forms with improved thermal stability, PseFDH GAV and PseFDH T7. High homology scores for FDH from different sources (Fig. 1) allowed high accuracy model structures to be obtained for the enzymes from C. boiidinii (Felber, 2001; Slusarczyk et al., 2000; Labrou et al., 2000; Labrou and Rigden, 2001), Candida methylica (Karaguler et al., 2004) and Saccharomyces cerevisiae (Serov et al., 2002). These structures were successfully used to plan mutations aimed at improving

chemical stability (Felber, 2001; Slusarczyk et al., 2000) and studying the catalytic mechanism (Labrou et al., 2000; Labrou and Rigden, 2001) of CboFDH, and for the switch in coenzyme specificity of FDH from S. cerevisiae (Serov et al., 2002). Numerous attempts to get wild-type CboFDH crystals suitable for the structure resolution failed. To get the required quality of CboFDH crystals, an approach based on the introduction of amino acid replacements in the regions of highly disordered structure, has been applied (Schirwitz et al., 2005). The prediction of these regions for new enzymes is performed using special programs and the structure of a homologous enzyme. In the case of CboFDH, this approach was used to introduce the following mutations: Lys47Val, Lys47Glu, Arg296Ala, Lys328Val and Lys338Ala. Replacement Lys47Glu in CboFDH resulted in preparation of highÀ quality enzyme crystals, which provided 1.9 A resolution of the apo-enzyme structure (Schirwitz et al., 2005). Noteworthy, Lys47 (Lys75 in PseFDH) is conserved through all 51 FDH complete and partial sequences known up to date (Fig. 1). 2.2. Amino acid sequences alignment The above approach is widely used for all enzymes. It is commonly used in combination with other methods. For instance, the alignment of FDH amino acid sequences from different sources was used to localize non-conserved Ser residues while improving PseFDH thermal stability with hydrophobization of a-helices (Rojkova et al., 1999) and optimization of polypeptide chain conformation (Serov et al., 2005). The approach has been also used to select the type of the introduced residue while increasing chemical stability of PseFDH (Tishkov et al., 1993; Odintseva et al., 2002), MycFDH (Yamamoto et al., 2005) and CboFDH (Slusarczyk et al., 2000) (see below). The comparative analysis of FDH amino acid sequences led us to decision to clone FDH from S. cerevisiae (Serov et al., 2002; Serov, 2002). It was the first enzyme with Lys and Val residues upstream catalytically important Gln313 and His332 (numbered as in PseFDH), respectively, whereas the majority of FDHs contain Pro residues in these positions (Fig. 1). Until recently, the effectiveness of this approach was limited by the small number of the cloned FDH sequences. In last 5 years, direct cloning of FDH genes and genome sequencing of different organisms resulted in a whole series of complete and partial sequences of the enzyme. Right now, 52 complete (17 from bacteria, 15 from plant, 10 from yeast and 10 from fungi) and more than 15 partial FDH sequences are known. Fig. 2 presents evolution tree for FDHs from different sources, generated with the Clustal X 1.83 program. The analysis did not include enzymes from M. vaccae N10 and C. methylica, since they differ in two replacements only (differences lower than 1%) from PseFDH and CboFDH sequences, respectively. Fig. 2 demonstrates that bacterial and plant FDHs form very compact groups, which are rather far from other FDHs. The biggest variety in sequences is observed for yeast and fungal FDHs. Nevertheless, FDH is an extremely conserved enzyme. Among all enzymes from all sources, 60 aa residues are absolutely


92

Table 1 Mutations performed to improve formate dehydrogenases properties Aim Probing of molecular mechanism Increase of specific activity

a

Mutation/source C23S/F285S CboFDH

Result/conclusion
fo 1.7-fold increase of specific activity, values of Km rmate NAD? and Km increased from 6 mM and 45 mM to 14 mM and 74 mM, respectively Mutations Glu141Gln and Glu141Asn induced 5.5fo and 4.3-fold increases in Km rmate values, 110- and 590-fold decreases in the kcat for reaction with formate and 9.5- and 85-fold increases in catalytic efficiency in reaction of glyoxylate reduction, respectively

Ref. Felber (2001)

Role of loop between b8-sheet and aA-helix in substrate specificity

Glu141Gln, Glu141Asn, ParFDH

Shinoda et al. (2005)

Increase of operational stability Change of ``essential'' Cys, controlling PseFDH operational stability. Cys255 is located above the plane of adenine moiety of NAD and occupies conservative position Change of surface Cys354 Change of Cys145 near catalytically important Asn146

V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110

Cys255Met, Cys255Ser, Cys255Ala, PseFDH

Stable at least a month (200-fold increase in chemical stability) NAD? increased seven-fold Decreased thermostability. Km for Met, three-fold for Ser and is the same as for WT for Ala; formate binding is unchanged for Ala and Ser and is three-fold decreased for Met Provided best thermal stability 1000-fold increased operational stability No changes No changes decrease of No changes of chemical in kinetic parameters and thermal stability in kinetic parameters and 10% thermal stability in kinetic parameters and increase stability >1000-fold

Tishkov et al. (1993), Odintseva et al. (2002)

Cys354Arg, Cys354Ser, Cys354Ala, PseFDH Cys255Ala/Cys354Ala, PseFDH Cys145Ser Cys145Ala, PseFDH Cys255Ala/Cys145Ser, Cys255Ala/Cys145Ala, PseFDH

Odintseva et al. (2002) Odintseva et al. (2002) Own data

Change of essential Cys in MycFDH (PseFDH and MycFDH differ by only two residues in positions 35 and 61)

Cys6Ser, Cys145/Ser, Cys255Ala/Ser/Val, C146S/C256V, C6A/C146S/C256V, MycFDH

No data about kinetic properties and thermal stability. Increase of chemical stability was estimated by tolerance to inactivation by substrate ethyl 4-chloroacetoacetate and the yield of synthesis of ethyl (S)-4-chloro-3-hydroxybutanoate No change of kinetic parameters, increased chemical stabiliy No change of kinetic parameters, diminished chemical stabiliy No change of kinetic parameters, substantially decreased thermostability but operational stability under biotransformation conditions increased an order of magnitude PseFDH and MycFDH differ by only two residues in positions 35 and 61 Four- to six-fold lower thermostability of MycFDH is caused by electrostatic repulsion between Asp43 and Glu61 residues Mutation changed temperature dependence of thermal inactivation rate constant 1.20-fold increase of thermal stability compared to wt-PseFDH 1.24-fold increase of thermal stability compared to wt-PseFDH 1.40-fold decrease of thermal stability compared to wt-PseFDH

Yamamoto et al. (2005)

Change of all available cysteines in CboFDH

Cys23(52)Ser, CboFDH Cys262(288)Val, CboFDH Cys23Ser/Cys262)Ala, CboFDH

Slusarczyk et al. (2000), Felber (2001)

Increase of thermal stability Optimization of electrostatic interactions (effect of amino residues in positions 43 and 61 on thermal stability of bacterial FDH)

Glu61Gln, Glu61Pro Glu61Lys, MycFDH

Galkin et al. (1995), Fedorchuk et al. (2002)

Lys61Arg, PseFDH Hydrophobization of a-helices Ser131Ala Ser160Ala Ser168Ala

Rojkova et al. (1999)


Ser184Ala Ser228Ala Ser(131,160)Ala Ser(184,228)Ala Ser(131,160,184,228)Ala (mutant T4) PseFDH Tyr62Phe Tyr165Phe, PseFDH Minimization of conformational tensions in polypeptide chain His263Gly Ala191Gly Asn234Gly Asn136Gly Tyr144Gly Tyr144Gly + T4, PseFDH Cys23Ser CboFDH (SM CboFDH) Arg178Ser SM CboFDH Arg178Gly, SM CboFDH Asp149Glu, Arg178Ser, SM CboFDH Glu151Asp, Arg178Ser, SM CboFDH Glu151Asp, Lys356Glu, Glu151Asp, Lys306Arg, Glu151Asp, Lys306Arg, Arg178Ser, SM CboFDH Arg178Ser, Lys356Glu, SM CboFDH Arg178Ser, Thr315Asn, SM CboFDH

1.13-fold increase of thermal stability compared to wt-PseFDH 1.2-fold increase of thermal stability compared to wt-PseFDH 1.40-fold increase of thermal stability compared to wt-PseFDH 1.28-fold increase of thermal stability compared to wt-PseFDH 1.60-fold increase of thermal stability compared to wt-PseFDH The same kinetic properties as for wt-PseFDH No change of thermal stability compared to wt-PseFDH 17.6-fold decrease of thermal stability compared to wt-PseFDH 1.30-fold decrease of No significant effect No significant effect 1.20-fold increase of 1.40-fold increase of 2.30-fold increase of thermal stability on the stability on the stability thermal stability thermal stability thermal stability compared to wt-PseFDH

Serov and Tishkov (2002) Serov et al. (2005)

compared to wt-PseFDH compared to wt-PseFDH compared to wt-PseFDH V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110 Slusarczyk et al. (2000) Slusarczyk et al. (2003)

Improvement of thermal stability of CboFDH by directed evolution

Decrease of thermal stability 6.7-fold and Tm 58 compared to wtCboFDH Increase of thermal stability 3.1-fold and Tm 38 compared to SM Increase Increase mutation Increase mutation Increase mutation Increase of thermal stability 2.2-fold and Tm 28 compared to SM of thermal stability 6.7-fold and Tm 58 compared to SM, Asp149Glu provides increase of thermal stability 2.15-fold of thermal stability 27.6-fold and Tm 98 compared to SM, Glu151Asp provides increase of thermal stability 9.0-fold of thermal stability 18-fold and Tm 88 compared to SM, Lys356Glu provides decrease of thermal stability 1.5-fold of thermal stability 18-fold and Tm 88 compared to SM

Cys23Ser, Cys262A, CboFDH (DM CboFDH) Lys306Arg, Thr315Asn, Lys356Glu, DM CboFDH Glu18Asp, Lys35Arg, Arg187Ser, DM CboFDH Glu18Asp, Lys35Arg, Glu151Asp, Arg187Ser, Phe285Tyr, DM CboFDH Investigation of the role of conservative prolines in thermal stability Role of ``charge-relay'' system in thermal stability Testing of role of Thr169 and Thr226 in stability of C. methylica FDH Pro288(312)Thr, CboFDH Gln287(313)Glu/His311(332)Gln, CboFDH Thr169Val Thr226Val Thr169Val/Thr226Val, CmeFDH

Increase of thermal stability 36-fold and Tm 108 compared to SM, mutations Lys306Arg and Thr315Asn provide increase of thermal stability 1.5-fold. This mutant is 5.7-fold more stable than wt-CboFDH Decrease of thermal stability 35-fold and Tm 108 compared to wtCboFDH Increase of thermal stability 3.8-fold and Tm 48 compared to DM Increase of thermal stability 3.8-fold and Tm 48 compared to DM, mutations Glu18Asp and Lys35Arg provide increase of thermal stability 1.1-fold Increase of thermal stability 47-fold and Tm 118 compared to DM. This mutant is 1.4-fold more stable than wt-CboFDH Thermal inactivation rate increased 18-fold Neutral mutation Stability increased 1.6-fold at 55 8C Decrease of k 4-fold compared to wt-CmeFDH

Slusarczyk et al. (2000) Slusarczyk et al. (2003) Slusarczyk et al. (2003) Slusarczyk et al. (2003) N. Labrou N. Labrou
b

b

cat

Karaguler et al. (2004)

No change of kinetic parameters compared to wt_CmeFDH Decrease of enzyme stability by Ð4 kcal/mol due to remove of hydrogen bond between this residues 93


94

Table 1 (Continued ) Aim Change of coenzyme specificity Change of coenzyme specificity of FDH from C. methylica from NAD+ to NADP Change of coenzyme specificity of CboFDH from NAD+ to NADP+ Mutation/source Asp195(221)Ser CmeFDH Asp195Ser Asp195Ser/Tyr196His Asp195Ser/Tyr196His/Lys356(379)Thr CboFDH Asp196(221)Ala/Tyr197Arg, SceFDH PseFDH T5M8 Result/conclusion Decrease in coenzyme preferencec for NAD+ from 2.5 Ò 105 to 410 The mutant enzyme still retained specificity for NAD+ Activity with Activity with Activity with Final mutant wt-CboFDH. NAD+ and NADP+ 1.5 and 0.083 U/mg, respectively NAD+ and NADP+ 1.3 and 0.19 U/mg, respectively NAD+ and NADP+ 1.3 and 0.36 U/mg, respectively is 276-fold more active with NADP+ compared to No data about Km for formate and coenzymes Ref. Gul-Karaguler et al. (2001) Rozzell et al. (2004)

V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110

+

Change of SceFDH Change of PseFDH

coenzyme specificity of from NAD+ to NADP+ coenzyme specificity of from NAD+ to NADP+

Extending of pH-optimum of NADP binding for mutant PseFDH

+

PseFDH T5M9-10

Shift in coenzyme preference for NAD+ from >3 Ò 109 to 0.43-0.67 resulted in NADP+-specific enzyme Shift in coenzyme preference for NAD+ from 2.4 Ò 103 NAD ? to 0.29 resulted in NADP+-specific enzyme, Km P is constant in pH range 6.0-7.0. The mutant enzyme has specific activity with NADP+ 2.5 U/mg NADP? Km is constant in pH range 6.0-9.0

Serov et al. (2002) Serov et al. (2002)

InnoTech MSU (2005) Shwirwitz et al. (2004)

Preparation of enzyme crystals for x-ray analysis Preparation of mutant CboFDH producing crystals suitable for X-ray analysis
a b c

Lys47(75)Glu CboFDH

Ä À Determination of apo-CboFDH structure with resolution 1.9 A

Numbering of the residues refers to particular enzyme, in parenthesis--numbering for PseFDH. Prof. N. Labrou, personal communication. ? ? The value of coenzyme preference for NAD+ is expressed as ?kcat =Km îNAD =?kcat =Km îNADP .


V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110 Table 2 Study of formate dehydrogenase structures by X-ray analysis Enzyme form apo-PseFDH PseFDH + NAD+ + azide PseFDH + NADH PseFDH + formate PseFDH + ADP-ribose PseFDH + NADH + formate PseFDH GAV PseFDH T7 apo-CboFDH apo-Moraxella FDH apo-ArabidopsisFDH À Resolution (A) 1.8 2.0 2.1 2.2 1.5 2.3 2.0 2.0 1.9 2.4 2.0-2.2 Remarks

95

Crystals prepared from native enzyme from Pseudomonas sp.101. Structure 2NAC (Lamzin et al., 1994) Crystals prepared from native enzyme from Pseudomonas sp.101. Structure 2NAD (Lamzin et al., 1994) Native enzyme. Conformation similar to apoPseFDH (2NAC) (Filippova et al., 2005) Native enzyme. Formate is bound with Arg201, which is a residue responsible for binding of pyrophosphate moiety of coenzyme. Cys354 is oxidized (Filippova et al., 2006) Native enzyme. Only part of ADP-ribose can be seen. Conformation similar to apoPseFDH except movement of residues 121-123 Native enzyme. NADH is not visible in active site. Cys354 is oxidized. Conformation is transient between apo-PseFDH and ternary complex with NAD+ and azide (Filippova et al., 2006) Mutant recombinant full size enzyme. Conformation similar to apoPseFDH Mutant recombinant full size enzyme. Conformation similar to apoPseFDH Mutant CboFDH Lys47Glu (Schirwitz et al., 2005) Recombinant enzyme expressed in E. coli. Active site has more open conformation than in apo-PseFDH Recombinant enzyme expressed in E. coli. Coenzyme binding domain is similar to one in apo-PseFDH

conserved, and within the individual groups, the homology score is higher than 75%. 2.3. Random mutagenesis The above approach was successfully used to improve thermal stability (Slusarczyk et al., 2003), increase catalytic activity (Felber, 2001; Slusarczyk et al., 2003) and switch the coenzyme specificity (Rozzell et al., 2004) of CboFDH. Mutations were introduced with error prone PCR. The analysis of E. coli cell libraries with mutant CboFDH (up to 200,000 clones) (Felber, 2001; Slusarczyk et al., 2003) were performed in two steps. Primary qualitative screening of clones obtained after transformation was carried out directly on solid agar. This step yields only those clones that produced active enzyme. Clones were selected in accordance with the protocol used for the enzyme activity staining in PAAG (Felber, 2001). The produced NADH was oxidized by phenazine etho-sulfate, and the reduced form of the latter reacted with nitrotetrazoleum blue to generate insoluble colored product. The yield of active clones was only 0.1%, i.e. only one mutation out of 1000 did not result in enzyme inactivation. At the second step, the selected clones were cultivated in 96-well microtiter plates. Then, the cells were lysed and the homogeneous enzyme was produced using affinity microchromatography in 96-well microtiter plates. To screen for mutant CboFDH with enhanced thermal stability, the enzyme preparations were incubated at 50-58 8C for 15 min, and the residual activity was determined. To screen for the mutants with high specific activity, the activity of free enzyme was measured in the presence of a fixed concentration of an inhibitor, Procion MX-R, which bound the enzyme equimolarly (Felber, 2001). To get NADP+-dependent CboFDH (Rozzell et al., 2004), amino acid residues were replaced using error prone PCR, however, in this case, the yield of active clones was much higher than 70%, compared to the previous experiments (Felber, 2001). All clones from the library (up to 20,000) were cultivated in 96-well microtiter plates, and then, the enzyme was isolated and its activity was analyzed with NADP+ in 96- or 384-well microtitre plates.

In conclusion of this chapter, we note that the rational design based on the analysis of the enzyme structure and amino acid sequences alignment saves one a lot of time and reagents to generate new mutants. However, the modern level of computer calculations for the effect of the introduced mutations does not yield all possible candidatures for mutations. This can be demonstrated by the improvement of CboFDH specific activity (see below). Therefore, the maximum effect can be achieved by combination of both approaches. 3. Catalytic mechanism studies and improvement of kinetic parameters 3.1. Switch in substrate specificity The model for FDH catalytic mechanism was proposed using the structures of the apo-enzyme and of enzyme-NAD+azide ternary complex (Lamzin et al., 1994). The detailed analysis of the effects of amino acid replacements on the enzyme catalytic mechanism can be found in the review (Popov and Tishkov, 2003). In addition to the mutations reviewed there the effects of Glu141Gln and Glu141Asn mutations in FDH from Paracoccus sp.12-A (Shinoda et al., 2005) are discussed. As mentioned above, FDH belongs to the superfamily of D-specific dehydrogenases of 2-hydroxy acids and structure of D-lactate dehydrogenase is very similar to one for PseFDH. The Asn97Asp replacement in D-lactate dehydrogenase from Lactobacillus pentosus has minimum effect on protein overall folding and catalytic activity, however, the Km value for lactate increases 70-fold. The Asn97 residue is located in the loop, which covers the enzyme active center from the solvent. This residue is highly conserved for majority of D-specific dehydrogenases of 2hydroxy acids. In the case of FDH Glu141 occupies the equivalent position in 52 out of 53 known sequences. The exception is the FDH from barley, which like E. coli D-lactate dehydrogenase, contains Arg residue in the same position. The Glu141Gln and Glu141Asn replacements in ParFDH resulted in an increase of Km value for formate 5.5- and 4.3fold, a decrease in kcat for the reaction with formate 110- and


96

V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110

Fig. 1. The alignment of amino acid sequences of formate dehydrogenases from bacteria Pseudomonas sp.101 (PseFDH) (Tishkov et al., 1991), Thiobacillus sp.KNK65MA (TbaFDH) (Nanba et al., 2003a), Sinorhizobium meliloti (SmeFDH) (Barnett et al., 2001), Bordetella bronchiseptica RB50 (BbrFDH) (Parkhill et al., 2003), Legionella pneumophila (LegFDH) (Chien et al., 2004), uncultivated g-proteobacterium EBAC31A08 (UmgFDH) (Beja et al., 2000), Mycobacterium avium


V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110

97

Fig. 2. Evolution tree for formate dehydrogenase. Full names of organisms from top to bottom: Emericella nidulans (Aspergillus nidulans), Aspergillus fumigatus Af293, Ajellomyces capsulatus, Mycosphaerella graminicola (Septoria tritici), Gibberella zeae PH-1, Magnaporthe grisea, Botrytis cinerea, Neurospora crassa, Saccharomyces cerevisiae, Candida albicans SC5314, Yarrowia lipolytica CLIB99, Pichia angusta (Hansenula polymorpha), Pichia pastoris, Candida boidinii, Cryptococcus neoformans var. neoformans JEC21 (Filobasidiella neoformans), Ustilago maydis 521, soya G. max izoenzyme 2 and 1, Zea mays, rice Oryza sativa, barley Hordeum vulgare, apple tree Malus Ò domestica, English oak Quercus robur, tomato Lycopersicon esculentum, potato Solanum tuberosum, Arabidopsis thaliana, Streptomyces avermitilis, Mycobacterium avium subsp. paratuberculosis str.k10, Burkholderia sp.383, Bordetella bronchiseptica RB50 (Alcaligenes bronchisepticus), Bordetella parapertussis strain 12822, Bordetella pertussis strain Tohama I, uncultivated marine g-proteobacterium EBAC31A08, uncultured marine a-proteobacterium HOT2C01, Legionella pneumophila subsp. pneumophila str. Philadelphia 1, Sinorhizobium meliloti, Hyphomicrobium strain JT-17 (FERM P-16973), Paracoccus sp.12-A, Moraxella sp., Ancylobacter aquaticus, Thiobacillus sp.KNK65MA and Pseudomonas sp.101.

590-fold, and an increase in catalytic efficiency in the glyoxylate reduction 9.5- and 85-fold, respectively (Shinoda et al., 2005). These results demonstrate the possibility to change substrate specificity of the enzymes of superfamily of D-specific dehydrogenases of 2-hydroxy acids. 3.2. Enhancement of catalytic activity of FDH from C. boidinii One of FDH disadvantages is its low catalytic activity. The highest specific activity was reported for bacterial

enzymes. At 30 8C, the specific activity of PseFDH is ca. 10 U per mg of protein. Enzymes from other sources are less active than bacterial FDHs (Tishkov and Popov, 2004). The activity of CboFDH is ca. 6.1-6.3 U/mg (30 8C) (Slusarczyk et al., 2000; Labrou et al., 2000; Felber, 2001). However, due to the difference in the molecular mass of these enzymes (44,000 and 40,370 Da for bacterial and yeast enzymes, respectively), the values of kcat differ by two-fold, 7.3 and 3.7 sÐ1 for PseFDH and CboFDH, respectively. The CboFDH activity was increased up to 9.1 U/mg with random mutagenesis (Slusarczyk et al., 2003). Out of 200,000 clones generated by random

subsp. paratuberculosis str.k10 (MavFDH) (Li et al., 2005), Streptomyces avermitilis (SavFDH) (Omura et al., 2001), plants Arabidopsis thaliana (AraFDH) (Olson et al., 2000), potato Solanum tuberosum (PotFDH) (Colas des Francs-Small et al., 1993), English oak Quercus robur (OakFDH, GeneBank Accession AJ577266), barley Hordeum vulgare (BarFDH, EMBL Accession D88272) and soya G. max (SoyFDH1); yeasts Candia boidinii (CboFDH) (Sakai et al., 1997; Slusarczyk et al., 2000; Labrou et al., 2000) and Pichia angusta (HanFDH, former Hansenula polymorpha, EMBL P33677), Yarrowia lipolytica strain CLIB99 (YarFDH) (Dujon et al., 2004), Candida albicans (CabFDH) (Jones et al., 2004), S. cerevisiae (SceFDH, EMBL Z75296), fungi Ajellomyces capsulatus (AjeFDH) (Hwang et al., 2003), Aspergillus nidulans (AspFDH) (Saleeba et al., 1992), Magnaporthe grisea (MagFDH, GeneBank Accession AY850352), Gibberella zeae PH-1 (CzeFDH, GenBank XM_386303), Cryptococcus neoformans (CryFDH) (Loftus et al., 2005) and Ustilago maydis (UstFDH, GeneBank Accession XM_402785). Catalytically important residues are shown in white letters on black background, conservative residues in bold. Residues subjected to mutagenesis marked by grey background.


98

V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110

Fig. 3. Position of Phe311PseFDH (Phe285CboFDH) (marked by pink color) in ternary complex (PseFDH-NAD+-azide) (structure 2NAD). NAD+ and azide are marked by dark blue and grey blue colors, respectively. Picture was created using WebLab ViewerPro 3.7 software (Molecular Simulations Inc.).

16 bacterial FDHs and only in FDH from Streptomyces avermitilis this residue is substituted by homologous Tyr (Fig. 1). Among 16 plant FDHs, there are 6 Phe, 4 Tyr, 5 Asn and 1 Asp (Arabidopsis thaliana) residues in this position. In 20 sequences of yeast and fungal FDHs, Phe residue is found 12 times, Asp 5 times, Pro twice and Tyr once. The Phe285 (311 PseFDH) residue in FDH is in Ð2 position with respect to the catalytically important Gln287 (313 PseFDH) residue, which is located at the entrance to the active center of the enzyme at the site of substrate-binding channel (Fig. 3). Increase in CboFDH activity up to 9.1 U/mg resulting from the Phe285Ser replacement did not effect the thermal stability of the enzyme, but slightly worsened the Km values both for fo coenzyme and formate. The Km rmate value grew from 6 to NAD? 14 mM, while Km increased from 45 to 73 mM (Felber, 2001). We note that kcat value for the mutant CboFDH Phe285Tyr, 6.1 sÐ1, is still lower than that for PseFDH, 7.3 sÐ1. 4. Improvement of FDH operation stability The main reason for FDH inactivation at elevated temperatures (up to 40-45 8C) is the oxidation of SH-groups of cysteine residues. Chemical modification of Cys residues may occur due to the impurities present in substrates or directly by substrates containing active groups: for instance, ethyl 4-chloro-acetoactate (ECAA) is used as a substrate for the synthesis of (S)-ethyl 4chloro-3-hydroxybutanoate ((S)ECHB), the key intermediate in LipitorTM synthesis (Rozzell et al., 2004; Yamamoto et al., 2005). Table 3 presents the data on the occurrence of Cys residues in FDHs from various sources. As can be seen, bacterial FDHs exhibit the highest content of Cys residues compared to the

mutagenesis, 1500 clones expressing the active enzyme have been selected. Among the latter, four clones have been identified with the enzyme specific activity higher than that of the wild-type CboFDH. Sequencing showed that all four clones had the same replacement, Phe285Ser. The authors produced also a mutant CboFDH Phe285Tyr. The mutation did not affect the enzyme activity, but increased its thermal stability (see below). The Phe residue in this position (Phe311 in PseFDH) is highly conserved for bacterial FDHs: it is present in 15 from
Table 3 Location of cysteine residues in formate dehydrogenases Position 3 5 52 74 81 116 117 140 145 171 182 196 199 215 235 248 255 273 288 345 354 368 Bacteria 1 15 8 0 3 0 0 1 10 3 13 1 0 0 0 13 11 0 12 1 15 0 (17) (17) (17) (17) (17) (17) (17) (17) (17) (17) (17) (17) (17) (17) (17) (16) (16) (16) (16) (16) (16) (16)
a

Plants 0 0 14 0 6 0 1 1 0 0 0 0 0 13 1 16 0 3 0 0 0 0 (13) (13) (15) (15) (15) (15) (15) (17) (17) (16) (16) (16) (16) (15) (14) (16) (16) (16) (16) (15) (15) (13)

Yeasts 0 0 7 0 0 0 7 0 0 0 0 0 0 0 1 1 6 0 7 0 0 1 (9) (9) (9) (9) (9) (10) (10) (10) (10) (10) (10) (10) (10) (8) (10) (9) (9) (9) (9) (9) (9) (8)

Fungi 0 0 0 1 0 1 2 0 2 0 0 0 1 11 8 9 11 0 1 0 0 0 (9) (9) (10) (10) (10) (10) (10) (10) (10) (10) (10) (10) (10) (11) (11) (11) (11) (11) (11) (11) (11) (11)

Total 1 14 28 1 9 1 9 2 12 3 12 1 1 23 9 39 27 3 20 1 14 1 (48) (48) (51) (51) (51) (52) (52) (54) (54) (53) (53) (53) (53) (51) (52) (52) (52) (52) (52) (51) (51) (48)

Alternative residues

b

Val(6,0,6,6), Ile(9,13,3,1) Ala(1,0,0,4), Met(1,0,2,0), Gly(0,13,0,0), Leu(0,0,7,3) Ser(9,1,0,0), Thr(0,0,2,8), Ala(0,0,0,1) Ala(2,1,0,1), Ile(0,0,1,0), Asp(0,10,3,0), Asn(1,0,0,0), Pro(0,3,0,0), Ser(14,1,4,8), Thr(0,0,1,0) Ser(14,9,9,10) Leu(14,16,8,9), Ile(0,0,1,0), Ser(0,0,1,0), Met(3,0,0,0) Ala(16,0,0,8), Leu(0,14, 2,0), Ile(1,0,0,0), Leu(0,0,1,0) Ala(13,15,1,9), Val(2,0,0,0), Met(1,0,0,0), Ser(0,1,0,0), Leu(0,0,6,0), Thr(0,0,3,0), Tyr(0,0,0,1) Ser(7,17,10,8) Trp(9,0,0,0), Gln(1,16,10,9), Ile(4,0,0,0), Met(0,0,0,1) Ala(3,0,2,4), Asp(1,0,0), Ile(0,13,3,1), Val(0,2,5,5), Met(0,1,0,0) Thr(10,16,10,10), Val(4,0,0,0), Ser(1,0,0,0), Gly(1,0,0,0) Gly(1,1,1,0), Ala(16,15,9,0), Val(0, ,0,8), Ser(0,0,0,1) Val(13,0,0,0), Pro(0,0,8,0,0), Met(2,0,0,0), Thr(1,0,0,0), Leu(1,1,0,0), Trp(0,1,0,0) Leu(12,0,1,0), Ala(3,13,3,3), Val(2,0,4,0), Ile(0,0,1,0) Ala(0,0,6,0), Val(3,0,0,0), Ser(0,0,2,0), Leu(0,0,0,1), Trp(0,0,0,1) Ala(1,0,3,0), Ser(1,0,3,0), Thr(1,13,3,0), Met(0,2,3,0), Val (1,0,0,0), Ile (1,0,0,0) Met(10,9,6,11), Phe(6,0,3,0), Leu(0,5,0,0) Val(2,13,0,10), Ala(1,3,0,0), Thr(1,0,2,0) Ala(15,14,9,11), Thr(0,1,0,0), Arg(0,15,0,1), Ser(0,0,6,10), Glu(0,0,2,0), Asp(1,0,0,0), Asn (0,0,01), (Val(0,0,1,0) Val(16,13,3,11), Leu(0,0,3,0), Ile(0,0,1,0)

Bold numbers show positions with high occurence of Cys and unique positions of Ser. a 14 (16) means that 14 enzymes of 16 have Cys residue in this position. b Values in parentheses show number of alternative residue in bacteria, plant, yeasts and fungi, respectively.


V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110

99

enzymes from other sources. There are 14 positions in amino acid sequences of bacterial FDHs where Cys residues can be found; the probability of Cys occurrence in 7 positions is as high as 70%. In plant enzyme Cys is found in eight positions, and in three particular positions 52, 215 and 248 (numbered in accordance with PseFDH sequence), the probability of their occurrence is 93, 87 and 100%, respectively. In sequences of yeast and fungal FDHs, Cys residues are found in 7 and 11 positions, respectively, among which only four are highly conserved (Table 3). Bacterial FDHs show the highest variability of Cys residues content. The enzyme from Legionella pneumophila has the highest number of Cys residue among all FDHs, nine per subunit, and FDH from S. avermitilis has the lowest content, one residue per subunit. Among 17 known bacterial FDH sequences, 10 contain from six to eight Cys residues per subunit. In plant and yeast FDHs, the average content of Cys residues is from two to five per subunit. As one can judge from Table 3, no correlation can be found between Cys occurrence and FDH source. Only in one position (248 in PseFDH) Cys residue is present in bacteria, plants, yeast and fungi. There is also no specific preference for the residue type for the positions that can be occupied by Cys, except positions 81 and 145, which alternatively show Ser only (Fig. 1, Table 3). Cys residues show different activity and accessibility for the solvent. Using apo-PseFDH structure (PDB2NAC) as basic, one can mark out three groups of Cys residues. Most solvent accessible are Cys81, 171, 255 and 354; much lesser accessible are Cys residues in positions 52, 140, 145, 196, 248, 288 and 345. All others are located deep inside the protein globule. The most critical for the enzyme activity are Cys145 and Cys255. Cys145 is adjacent to Asn146, which participates in formate binding in the enzyme active center, and Cys255 is located in the coenzyme-binding domain and contacts with the adenine moiety of NAD+ (Lamzin et al., 1994). 4.1. Improvement of chemical stability of FDHs from Pseudomonas sp.101 and M. vaccae N10 Each PseFDH subunit has seven Cys residues in positions 5, 145, 182, 248, 255, 284 and 354 (Fig. 1, Table 3). Chemical modification experiments performed with PseFDH in the end of 70s and beginning of 80s of the last century demonstrated that modification of a single Cys residue per subunit was sufficient to inactivate the enzyme. Amino acid sequencing proved this residue to be Cys255 (Popov et al., 1990). The PseFDH Cys255Ser and Cys255Met mutants produced in 1993 were absolutely stable toward Hg2+ ion inactivation and showed a 100-fold decrease in the rate of inactivation with DTNB (Tishkov et al., 1993). However, the mutants produced showed inferior Km for substrates compared to wt-PseFDH, and thermal inactivation rate increased four- to eight-fold (Tishkov et al., 1993). The PseFDH Cys255Ala mutant produced later had the same kinetic parameters as the wildtype enzyme, but its thermal stability dropped four-fold

(Odintseva et al., 2002). Some native bacterial FDHs contain Ala (Sinorhizobium meliloti) (Barnett et al., 2001), Val (Thiobacillus sp.KNK65MA) (Nanba et al., 2003a), Ser (M. avium subsp. paratuberculosis str.k10) (Li et al., 2005)orThr (S. avermitilis)(Omura et al., 2001) in 255 position, instead of Cys (Fig. 1). It was found that FDH from Thiobacillus sp.KNK65MA exhibited higher chemical stability against inactivation with a-haloketones compared to the enzymes from Ancylobacter aquaticus and C. boidinii (Nanba et al., 2003a). The chemical modification of PseFDH Cys255Ser and Cys255Met mutants with DTNB showed the importance of an additional Cys residue for the catalytic activity of PseFDH, however, this second Cys was less reactive than Cys255 (Tishkov et al., 1993). A decrease in the inactivation rate for the PseFDH Cys255Ser mutant under the action of DTNB in the presence of formate-ion pointed to the residue localization in the substrate-binding domain of the active center. This second residue appeared to be Cys145, adjacent to Asn146 necessary for formate binding (Tishkov et al., 1991). Table 3 shows that all plant, yeast, fungal and six of bacterial FDHs have Ser residue in this position. The Cys145Ser replacement in PseFDH had no effect on kinetic parameters and thermal stability, while the double mutant PseFDH Cys145Ser/ Cys255Ala exhibited at least a 1000-fold increase in chemical stability compared to the wild-type enzyme. Single replacement Cys145Ala slightly (10%) increased the rate of thermal inactivation. The analysis of apo- and holo-PseFDH structures demonstrates the Cys354 accessibility for the solvent. In plant FDHs, this position is occupied by Arg, and in yeast and fungi by Ser (Table 3). The study of PseFDH mutant forms, Cys354Ala, Cys354Ser and Cys354Arg shows that these replacements increase Km for formate two- for four-fold, and decrease thermal stability 2.5-, 3- and 10-fold, respectively (Odintseva et al., 2002). X-ray analysis of (PseFDH + NADH + formate) complex (two molecules per elementary crystallographic cell) shows oxidized forms of sulfur in Cys residues: SO in one subunit and SO3Ð in three others (Filippova et al., 2006). This observation proves that Cys354 is not essential for chemical stability and explains the appearance of different PseFDH isoforms upon storage, due to multiple oxidation forms of sulfur in this residue. Improvement of chemical stability with directed mutagenesis was achieved for FDH from M. vaccae N10 as well (Yamamoto et al., 2005). As we mentioned before, this enzyme differs from PseFDH in two amino acid residues (Galkin et al., 1995). In addition to Cys255Ala and Cys255Ser mutations, by analogy with TbaFDH, the Cys255Val replacement was made (Yamamoto et al., 2005). As for Cys145, all three mutations were made, i.e. Cys145Ser, Cys145Ala and Cys145Val. In addition, Cys5 was replaced for Ala, Val and Ser to generate single, double and triple mutants. Unfortunately, the authors did not analyze the properties of each mutant in detail. It was shown that the introduced mutations resulted in a drop of enzyme activity in cell-free extracts from 2- to 16-fold, and that the activity of a triple mutant, Cys(5, 255, 354)Ser was only 0.011


100

V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110

compared to 3.24 U/mg for wt-MycFDH (Yamamoto et al., 2005). The effectiveness of either mutation was evaluated by the yield of the final product in the synthesis of (S)ECHB from ECAA, and by the stability against 20 mM ECAA induced inactivation at 25 8C. Analyzing the results of the work (Yamamoto et al., 2005), one can conclude that Cys5 is not essential for chemical stability of FDH. For instance, (S)ECHB yield with MycFDH C5A/C145S/C255V triple mutant and C145S/C255V double mutant was 32.2 and 31.0 g lÐ1, respectively. The value of residual activity for the triple and double mutants after 20 min incubation in the presence of 20 mM ECAA was 108 and 104%, respectively. These values allow us to conclude that there is no recorded change within the experimental error. In addition, for Cys145Ser/Cys255Val double mutant, the activation effect in 5% ethyl acetate was 187% compared to 137% for C5A/C145S/C255V triple mutant. The biggest activation was observed for C5A/ C145A/C255V triple mutant (219%), however, in this case, the yield of the final product (S)ECHB was almost 20% lower than for that of C5A/C145S/C255V mutant MycFDH. Activation affect and increase of affinity for formate in water-organic solvents were also shown for PseFDH (Demchenko et al., 1990). 4.2. Improvement of chemical stability of FDH from C. boidinii CboFDH contains two Cys residues per subunit, Cys23 and Cys262 (Ser52 and Cys288 in PseFDH, respectively) (Fig. 1). Single Cys23Ser and Cys262Val, and double mutants Cys23Ser/Cys262Val and Cys23Ser/Cys262Ala have been produced (Slusarczyk et al., 2000; Felber, 2001). Cys23 plays a more important role in chemical stability of CboFDH (Felber, 2001). For instance, in the presence of 150 mM hydrogen peroxide, half-life periods (t1/2) for wt-CboFDH and its Cys23Ser and Cys262Val mutants were 3.3, 7.3 and 2.4 min, respectively, and in the presence of 50 mM CuSO4, t1/2 values were equal to 38, 657 and 20 min, respectively (Felber, 2001). These data are in good agreement with the results of computer modeling of CboFDH structure (Slusarczyk et al., 2000; Felber, 2001). In accordance with the model, Cys23 is more solvent accessible than Cys262. The most visual effect of CboFDH chemical stabilization is observed at conditions for tert-Lleucince production (40 8C and pH 8.2). Under these conditions, the half-life time for Cys23Ser and Cys23Ser/ Cys262Ala CboFDH mutants increased more than five-fold compared to the recombinant wt-enzyme (Slusarczyk et al., 2000; Felber, 2001). Values of Km for NAD+ and formate for single and double mutants were unchanged compared to wt-CboFDH, however, the specific activity decreased from 6.3 to 4.9-5.5 (Slusarczyk et al., 2000). In addition, the introduced mutations resulted in significant decrease in thermal stability of CboFDH. If single Cys255Ala, Cys354Ala and Cys354Ser, and double Cys255Ala/Cys354Ser mutations in PseFDH resulted in a 4, 2.5-, 3.0- and 10-fold increase in the rate of thermal inactivation compared to the wild-type enzyme, respectively,

for single Cys23Ser, Cys262Val and double Cys23Ser/ Cys262Val 4 Cys23Ser/Cys262Ala of CboFDH, the rate of thermal inactivation increased 6.7, 21.6, 93.7 and 35.1 times, respectively, compared to the wild-type enzyme (Felber, 2001). Triple C145A/C255A/C354S PseFDH mutant exhibited comparable thermal stability at 58 8C, and surpassed wtCboFDH at lower temperatures. Thus, mutagenesis of Cys residues in FDH molecule results in significant improvement of chemical stability coupled to the decrease in thermal stability. To compensate the latter effect, additional studies were needed to improve the enzyme thermal stability. 5. Improvement of FDH thermal stability 5.1. Comparison of FDHs thermostability There are many approaches in the literature to quantitatively characterize enzyme thermal stability. In the case of FDHs, many authors used the residual enzyme activity upon incubation at a fixed temperature for a fixed time interval (15-30 <4>) (Galkin et al., 1995; Shinoda et al., 2002; Nanba et al., 2003a,b), or introduced the value of Tm, the temperature which provides with 50% inactivation in 20 min (Slusarczyk et al., 2000, 2003; Felber, 2001). The disadvantage of the first approach is the difference in thermal inactivation mechanisms for the enzymes from different sources, and inactivation kinetics may be rather complicated. Therefore, the choice of different time intervals could give opposite results. Moreover, the mechanism of enzyme inactivation may change at elevated temperatures. For instance, FDH from S. cerevisiae inactivates reversibly at temperatures below 42 8C, while at elevated temperatures, its inactivation mechanism includes both reversible and irreversible steps (Serov, 2002). Complex inactivation mechanism may cause serious difference in Tm-profiles for the same mutant series when different time intervals are used. Moreover, Tm values give no clue to quantitatively estimate enzyme thermal stability at temperatures other than Tm. The most rational approach to characterize enzyme thermal stability is to monitor the enzyme inactivation kinetics at different temperatures, or to use differential scanning calorimetry (DSC). The former approach gives quantitative characteristics of enzyme stability at different temperatures. The second method, DSC, allows the determination of the heat of transfer between native and denatured states of the protein globule. Thermal stability of wt-CboFDH and its mutants was studied in Slusarczyk et al. (2000, 2003) and Felber (2001). Quantitative effects were presented as Tm and half-life period at 50 8C. In this laboratory, the inactivation kinetics of wild-type and mutant FDHs from bacteria Pseudomonas sp.101 (Rojkova et al., 1999; Fedorchuk et al., 2002), M. vaccae N10 (Fedorchuk et al., 2002) and Moraxella sp. as well as from yeast C. boidinii (Sadykhov et al., 2006)and S. cerevisiae (Serov, 2002) and plants A. thaliana and siya Glycine max have been studied (Sadykhov et al., 2006). It was found that thermal inactivation of all FDHs except that of S. cerevisiae is irreversible and follows kinetics


V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110 Table 4 Tm values and first order inactivation rate constants of formate dehydrogenases at 55 8C Type of enzyme wt-FDH Thiobacillus sp.KNK65MAa (Nanba et al., 2003a) wt-FDH Ancylobacter aquaticusa (Nanba et al., 2003b) wt-FDH Paracoccus sp.12-Aa (Shinoda et al., 2002) wt-FDH Candida boidiniib Mutant Candida boidinii FDH C23Sb (Slusarczyk et al., 2000; Felber, 2001) Mutant Candida boidinii FDH C262Vb (Slusarczyk et al., 2000; Felber, 2001) Mutant Candida boidinii FDH C23S/C262Ab (Slusarczyk et al., 2000; Felber, 2001) Mutant Candida boidinii FDH C23S/C262A/E18N/K35R/E151D/R187S/F285Tb (Slusarczyk et al., 2003) Mutant Candida boidinii FDH C23S/E151D/R178S/K306R/T315Nb (Slusarczyk et al., 2003) wt-FDH Moraxella sp. (own data) wt-FDH M. vaccae N10 (Fedorchuk et al., 2002) Mutant M. vaccae N10 FDH E61K (Fedorchuk et al., 2002) wt-FDH Pseudomonas sp.101 (Fedorchuk et al., 2002) Mutant NAD+-specific FDH Pseudomonas sp.101 GAV (own data) Mutant NAD+-specific FDH Pseudomonas sp.101 T7 (own data) Mutant NADP+-specific FDH Pseudomonas sp.101 T5M9-10 (own data) kin (Ò106 sÐ1) 1330 996 385 183 1224 3960 6430 131 25.8 122 10.4 5.81 4.86 2.01 0.097 2.03 kin =k
PseFDH in

101

Tm (8C) 52.5 53 56 56.8 51.7 49.1 47.6 58 62 58 62 62.8 63 64.5 68 64.5

PseFDH Tm Ð Tm (grad)

274 205 79 38 252 814 1320 27.9 5.3 25 2.1 1.2 1.0 0.41 0.020 0.41

Ð10.5 Ð10 Ð7 Ð6.2 Ð11.3 Ð13.9 Ð15.4 Ð5 Ð1 Ð5 Ð1 Ð0.2 0 +1.5 +5 +1.5

Line for wild-type PseFDH is marked in bold because this enzyme used as a reference to show differences in stability. a Inactivation rate constants were calculated from data presented in this references supposing monomolecular mechanism of thermal denaturation. b kin for mutant CboFDHs were obtained by division of kin for wt-CboFDH at 55 8C by value of stabilization (destabilization) effect calculated from dependence in Fig. 6 and Tm values from Felber (2001) (see text).

of first-order reactions. The dependence of the inactivation rate constant kin on temperature T is described by the equation of the transition state theory: kB T kin Ì e h
Ð
DH 6Ì RT

Ð

DS 6Ì R

(1)

where T is the absolute temperature in K, kB and h the constants of Boltzmann and Plank, respectively and R is the universal thermodynamic constant. DH6Ì and DS6Ì are the activating parameters of changes in enthalpy and entropy for the process of enzyme thermal inactivation. This dependence can be linearized using [ln(kin/T)] Ð 1/T plot. Values of DH6Ì for PseFDH and CboFDH determined from slope of corresponding plots were 930 Ö 30 and 662 Ö 40 kJ/mol, respectively. Higher DH6Ì value for PseFDH compared to that for CboFDH shows that the change in the rate constant upon temperature rising for the bacterial enzyme is much bigger than that for yeast FDH. At the same time upon temperature decrease the inactivation rate constant for PseFDH will drop faster than one for CboFDH, i.e. thermal stability of bacterial enzyme will also grow faster than for CboFDH. MycFDH is a very good example of the fact. DH6Ì value 900 Ö 40 kJ/mol for this enzyme is similar to one for PseFDH. At 62 8C MycFDH and the most stable mutant CboFDH have the same values of Tm and kin (Table 4), but at 55 8C MycFDH is 2.5-fold more stable as mutant CboFDH due to higher DH6Ì value (Table 4). Table 4 presents the values for rate constants of thermal inactivation of PseFDH and CboFDH at 55 8C. Based on the data obtained by the other authors, we calculated the rate constants for thermal inactivation of FDH from bacteria Paracoccus sp.12-A (Shinoda et al., 2002), Thiobacillus sp.KNK65MA (Nanba et al., 2003a) and A. aquaticus (Nanba

et al., 2003b). In addition, we give the values of Tm for these enzymes. As seen from Table 4, FDH from Thiobacillus sp.KNK65MA it the least stable enzyme. Taking into account the results of mutagenesis of Cys residues in the other FDHs (see above), one can suggest that the reason for such low stability is the presence of Val and Ala residues in 255 and 288 positions in TbaFDH amino acid sequence instead of Cys residues (Fig. 1). Thermal denaturation studies of PseFDH, MorFDH, CboFDH and SceFDH using DSC also prove PseFDH to be the most thermostable enzyme among the known FDHs. The details of these experiments will be published elsewhere. Fig. 4

Fig. 4. Differential scanning calorimetry of wt-CboFDH, wt-PseFDH, mutant PseFDH T7 with increased thermal stability and mutant PseFDH GAV with increased chemical and thermal stability. Normalized melting curves. Protein concentration 1 mg/ml, 0.1 M phosphate buffer, pH 7.0, heating rate 0.1 grad per min.


102

V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110

shows melting curves for wt-CboFDH, wt-PseFDH and its mutant with improved thermal stability (see below). In the case of wt-PseFDH, heat of protein globule melting in the course of transfer from native to denatured state is by 310 kJ/mol higher that that for wt-CboFDH (2020 and 1710 kJ/mol for bacterial and yeast enzymes, respectively). 5.2. Improvement of PseFDH thermal stability To improve PseFDH thermal stability, the following approaches were used: hydrophobization of a-helices (Rojkova et al., 1999), increase in hydrophobicity of the protein globule, optimization of electrostatic interactions (Fedorchuk et al., 2002) and optimization of polypeptide chain conformation (Serov and Tishkov, 2002; Serov et al., 2005). Selection of mutation points was based on the X-ray analysis data and FDH sequences alignment for the enzymes from different sources. Note the stabilization effect for a single replacement was not high, usually from 10 to 50%. However, in all cases, the stabilization effect was additive (Rojkova et al., 1999; Serov et al., 2005), i.e. the final value of the stabilization effect in the multi-point mutant (nfin) was equal to the product of multiplication of individual stabilization effects nn for each singlepoint mutation: n
fin

Ì n1 Ò n2 Ò n3 Ò ÑÑÑ Ò n

n

It was found that for many single and multi-point PseFDH mutants (>90%) values DH6Ì are the same as for wild-type enzyme, i.e. stabilizing effect of mutations was due to change of DS6Ì (Rojkova et al., 1999; Fedorchuk et al., 2002; Serov et al., 2005). In some cases, the introduced replacements (such as Lys61Pro) had no effect on the thermal stability, but improved the stability in high ionic strength solutions (Fedorchuk et al., 2002). Combination of seven best mutations resulted in production of PseFDH T7 mutant with a 50-fold lower thermal inactivation rate constant compared to wt-PseFDH (Table 4), and the melting temperature in DCS experiments increased by 6.68 (Fig. 4). The mutations improving operational and thermal stability were combined in PseFDH GAV mutant. The combination compensated the decrease in thermal stability resulting from Cys replacement, and in addition, improved the overall thermal stability 2.5-fold compared to wt-PseFDH (Fig. 4). Moreover, the mutant showed two-fold increase in affinity for NAD+ (InnoTech MSU, 2006). The construction of mutant PseFDHs with increased thermal stability allowed the step of heat treatment of cellfree extract to be introduced into the purification protocol for the recombinant enzyme. Incubation of cell-free extract at 60 8C for 20-30 min increases the purity of the PseFDH GAV preparation from 50 to 80-85% without any loss of enzyme activity (Fig. 5). 5.3. Improvement of CboFDH thermal stability An improvement of thermal stability of CboFDH was obviously a more complicated task because the simultaneous

Fig. 5. SDS-analytical electrophoresis of E. coli cell-free extract with mutant PseFDH GAV before and after heat treatment at 60 8C (lines 1 and 2, respectively).

replacement of two Cys residues in each subunit resulted in a significant decrease in the enzyme thermal stability (35-94fold) (Slusarczyk et al., 2003) compared to that for PseFDH (10-fold). To increase the thermal stability of Cys23Ser and Cys23Ser/Cys262Ala CboFDH mutants, the method of ``directed evolution'' has been applied (Slusarczyk et al., 2003). The screening of two libraries, 200,000 clones each, yielded three clones derived from Cys23Ser/Cys262Ala double mutant and seven clones derived from Cys23Ser CboFDH single mutant. The stabilizing replacements increased Tm values by 10-118 compared to the original mutants (from 52 to 62, and from 47 to 58 8C for Cys23Ser and Cys23Ser/ Cys262Ala CboFDH, respectively) (Slusarczyk et al., 2003). As it was mentioned above, Tm values do not give quantitative assessment of the stabilization effects. To determine the quantitative parameters for the stabilization effects we employed the transition state theory (see above). By analogy with PseFDH, we assumed that the introduction of mutations into CboFDH did not change DH6Ì value. If this is the case, the dependence of the inactivation rate constant kin, determined for the set of mutants at the same temperature, will be linear in a [ln(kin/Tm)] Ð 1/Tm plot, where Tm is expressed in Kelvin. The work (Felber, 2001) presents the values for half-life periods t1/2 for wt-CboFDH and its Cys mutants at 50 8C. The


V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110

103

Fig. 6. Dependence of half-life periods t1/2 for wt-CboFDH and its Cys mutants at 50 8C on Tm in coordinates [ln(1/(t1/2 Ò Tm)] Ð 1/Tm. Values of t1/2 and Tm were taken from Felber (2001).

Fig. 7. pH-profiles for Km for NADP+ for the first generation of mutant NADP+specific PseFDH M9 and the second generation mutants PseFDH M9-10 and PseFDH M9-11. Reproduced with permission from InnoTech MSU (2005).

t1/2 value is reciprocally proportional to kin, therefore, a [ln(1/ (t1/2 Ò Tm)] Ð 1/Tm plot should be linear as well (Tm has to be expressed in grad K). Fig. 6 demonstrates that a true linear dependence (error < 3%, correlation coefficient R > 0.999) between half-life times of inactivation at 50 8C and Tm is observed. It was the linear dependence that allowed us to get correlation between kin and Tm and quantitatively evaluate the role of individual mutations in enzyme stabilization (Table 1). Among all mutations introduced, the biggest effects were observed for replacements Glu151Asp, Arg178Ser, Arg178Gly 4 Asp149Glu (the stabilization effects were 9.7, 3.4, 2.3 and 2.15, respectively). We want to highlight the Glu151Asp mutation, which contributed most to the stabilization. This replacement can be predicted from alignment of FDHs amino acid sequences. Among 20 FDHs from yeast and fungi only CboFDH has Glu residue in position 151. In other cases, there are 15 Asp and 4 Asn residues in this position. In equivalent position, bacterial FDHs have only Asp residue (Fig. 1). At the same time Arg in position 178 (numeration according CboFDH aa sequence) is absolutely conservative residue for FDHs from yeast and fungi (Fig. 1). The analysis of Asp149, Glu151 and Arg178 positions in the model structure of CboFDH shows their location in region of intersubunit contacts. The stabilization effect of the other replacements, e.g. Glu18Asp, Lys35Arg, Phe285Tyr, Lys306Arg, Thr315Asn and Lys356Glu, did not increase more than 1.5-fold (Table 1). Thus, the problem of CboFDH stabilization has been successfully solved. The value of thermal stability of Cys23Ser and Cys23Ser/Cys262Ala CboFDH mutants was increased 48and 56-fold, respectively, and if compared to the wild-type enzyme, 7.1- and 1.6-fold, respectively. The data on the role of the other amino acid residues in FDHs from C. boidinii and C. methylica are illustrated in Table 1. 6. Change of coenzyme specificity Formate dehydrogenase is a highly specific enzyme with respect to NAD+ (Tishkov and Popov, 2004). The data on

for coenzyme preference ?kcat =Km îNAD =?kcat =Km îNADP FDHs from Pseudomonas sp.101, C. methylica and S. cerevisiae were presented in Gul-Karaguler et al. (2001) and Serov et al. (2002). The analysis of kinetic properties of plant FDHs from A. thaliana and soya G. max expressed in E. coli cells in our laboratory shows the similarity in their coenzyme preference with that of PseFDH. Mutant PseFDH with coenzyme specificity changed from NAD+ to NADP+ was prepared in 1993. The enzyme was successfully used in synthesis of chiral alcohols and e-lactones using alcohol dehydrogenases and cyclohexanone monooxygenases (Seelbach et al., 1996; Rissom et al., 1997; SchwarzLinek et al., 2001). Unfortunately, contrary to NAD+-specific PseFDH, which has Km for NAD+ unchanged in pH range 6.0-9.0 (Mesentsev et al., 1997), the first generation of NADP+dependent mutant enzymes (version PseFDH T5M8) demonstrated the constant value of Km for NADP+ only in the pH range NADP? of 6.0-7.4 (InnoTech MSU, 2005). At pH ! 8.0, the Km value increases 10-fold and higher. Recently, new NADP+specific formate dehydrogenase PseFDH T5M9-10 have been prepared and this mutant enzyme has extended pH optimum for NAD ? Km P (pH range 6.0-9.0) (Fig. 7) (InnoTech MSU, 2005). The analysis of experiments resulting in the change of coenzyme specificity of FDHs from Pseudomonas sp.101, C. methylica and S. cerevisiae was reviewed earlier (Serov et al., 2002; Tishkov and Popov, 2004). Additional information about new NADP+-specific PseFDHs can be found in InnoTech MSU (2005). Herein, we will discuss the recently published results on the change in coenzyme specificity of CboFDH (Rozzell et al., 2004). CboFDH mutant active with NADP+ was prepared by directed evolution based on the Asp195Ser mutants described in Gul-Karaguler et al. (2001). The resultant forms were CboFDH double D195S/Y196H and triple D195S/Y196H/ K356T mutants. The activity of single, double and triple CboFDH mutants with NAD+ decreased (1.5, 1.3 and 1.3 U/mg, respectively) as compared to the activity wt-CboFDH (2.2 U/ mg). Introduction of replacements resulted in the increase of enzyme activity with NADP+ from 0.0013 for wild-type enzyme to 0.083, 0.19 and 0.36 U/mg for single, double and

?

?


104

V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110

In accordance with our model of holo-CboFDH structure, Tyr196 residue is not oriented towards adenosine ribose (Fig. 8), while in the model structure for SceFDH, this residue forms a hydrogen bond with 30 -OH group of adenosine ribose (Serov et al., 2002). Most likely, Tyr196His replacement provides an additional positive charge in the coenzyme-binding domain, necessary to compensate the negative charge of 30 phosphate group of NADP+. Thus, in the result of three replacements only, the authors were able to get CboFDH mutant with sufficiently high activity towards NADP+. This enzyme was used for NADPH regeneration in (S)-ethyl 4-chloro-3-hydroxybutanoate production with NADP+-specific ketoreductase (Rozzell et al., 2004).
Fig. 8. Orientation of Asp195, Tyr196 and Lys356 towards NAD+ in model structure of binary complex (CboFDH-NAD+). Picture was created using WebLab ViewerPro 3.7 software (Molecular Simulations Inc.).

7. Expression of FDH genes in E. coli cells Production of individual enzymes even partially purified is a costly process. Therefore, to lower the production cost, one constructs recombinant strains superproducing the target enzyme. Currently, FDH from bacteria, Pseudomonas sp.101 (Tishkov et al., 1991, 1999), M. vaccae N10 (Fedorchuk et al., 2002; Yamamoto et al., 2005), Moraxella sp., Hyphomicrobium strain JT-17 (FERM P-16973) (Mitsunaga et al., 2000), Paracoccus sp.12-A (Shinoda et al., 2002), Thiobacillus sp.KNK65MA (Nanba et al., 2003a), A. aquaticus (Nanba et al., 2003b), yeast C. methylica (Allen and Holbrook, 1995), C. boidinii (Sakai et al., 1997; Slusarczyk et al., 2000; Labrou et al., 2000; Felber, 2001) and baker's yeast S. cerevisiae (Serov et al., 2002; Serov, 2002) are successfully cloned and expressed in E. coli. In this laboratory, plant FDHs, from soybean and A. thaliana (the genes were kindly provided by Profs. N. Labrou and J. Markwell, respectively) have been expressed in E. coli as active enzymes. Noteworthy, the first plant FDH genes were cloned 8-12 years ago, but there were no data reported on their expression in E. coli in active and soluble form. Some details

triple mutants, respectively. Unfortunately, there is no data about the values of Km for NADP+ and formate. The activity of enzymes was measured at room temperature. If recalculated for 30 8C, the activity of best mutant CboFDH with NADP+ should be ca. 1.0 U/mg. This value is 2.5-fold lower if compared to the activity of NADP+-specific PseFDH (Serov et al., 2002; InnoTech MSU, 2005). The analysis of holo-CboFDH model structure shows that Asp195 and Lys356 form hydrogen bonds with 20 - and 30 -OH groups of adenosine ribose (Fig. 8). The Asp195Ser replacement results in the removal of the negative charge, and the Lys356Thr replacement, probably, provides additional room for the phosphate group. Note, PseFDH contains His379 in the position equivalent to Lys356 in CboFDH (Fig. 1). One may suggest that the positively charged His residue with lesser volume of the side chain compared to that of Lys, could participate in NADP+ binding.
Table 5 Expression of formate dehydrogenases in E. coli cells Source of gene Pseudomonas sp.101 NAD+-specifica Pseudomonas sp.101 NADP+-specific (Tishkov et al., 1999) M. vaccae N10 (Yamamoto et al., 2005) M. vaccae N10 a Moraxella sp. a Paracoccus sp.12-A (Shinoda et al., 2002) Ancylobacter aquaticus (Nanba et al., 2003b) Thiobacillus sp.KNK65MA (Nanba et al., 2003a) Candida methylica (Allen and Holbrook, 1995) Candida boidinii (Slusarczyk et al., 2000; Labrou et al., 2000) Candida boidinii (Felber, 2001) Candida boidinii (Rozzell et al., 2004) Candida boidiniia Saccharomyces cerevisiae (Serov, 2002) Soya G. maxa Arabidopsis thalianaa
a b c

Level of expression (% of soluble E. coli proteins) 50-55 50-55 30-35 50-55 50-55 12 44 n.d. 15 18 15 20-40 35-40 30-35 25-30 30-35

Inducer Lactose Lactose IPTG Lactose Lactose IPTG IPTG IPTG IPTG IPTG IPTG n.d Lactose Lactose Lactose Lactose

Production scale per run Megaunits Hundred kilounits n.d.b Dozen kilounits Hundred kilounits n.d. n.d. n.d. n.d. n.d. Megaunitsc n.d. Dozen kilounits Dozen kilounits Dozen kilounits Dozen kilounits

Own data. n.d., no data. Personal communication of Dr. T. Daussmann.


V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110

105

about expression of FDHs in E. coli cells are presented in Table 5. In all above experiments, FDH was synthesized in E. coli as active enzyme. The level of expression varied from 12-15 to 50-55% of the total soluble E. coli protein. The lower expression level for plant and yeast FDHs is likely to be caused by the presence of Arg codons rare for E. coli, i.e. AGA and AGG. To improve yield of recombinant CboFDH, Rozzell et al. (2004) optimized CboFDH gene sequence for E. coli codon usage and synthesized the modified gene. In addition, a Gly residue has been added just after the Nterminal Met. As for PseFDH, the gene sequence had no effect on the expression level, nevertheless, the optimization of a number of codons resulted in a two-fold increase in the enzyme biosynthesis rate. Commercial production of recombinant FDH was developed for the enzymes from C. boidinii and Pseudomonas sp.101. For PseFDH, the maximum yield of the enzyme was ca. 35 kU lÐ1 of cultural medium. Fermentation was performed in a fed-batch mode during 19 h at 25 8C in the absence of antibiotics. Lactose served as an inducer. Time/space yield was ca. 1850 U lÐ1 hÐ1, and the enzyme specific activity in the cell-free extract was equal to 4.5-5.5 U/mg of protein. Introduction of mutations improving thermal and chemical stability of PseFDH, had no effect on cultivation results. Optimization of large scale preparation of recombinant CboFDH is described in Felber (2001). Fermentation was performed for 36-41 h at 30 8C with subsequent IPTG induction. The maximum yield was ca. 60 kU lÐ1, time/space yield 1600 U lÐ1 hÐ1 and the specific activity 0.8-1.0 U/mg. These results were the same for the wild-type and C23S, C262V, C23S/C262V and C23S/C262A mutant enzyme forms. Accounting for the improved catalytic activity of CboFDH, from 6.0 to 9.1 U/mg (Felber, 2001), the enzyme yield can be expected up to 90 kU lÐ1. As could be seen from the data provided, process of recombinant mutant CboFDH production gives higher time/ space yield of active enzyme than PseFDH. However, the prolonged duration and lower expression level of CboFDH compared to those for PseFDH results in higher production costs: preparation of one activity unit of CboFDH requires 8-10-fold higher glucose expense than preparation of one activity unit of PseFDH; the lower temperature of PseFDH cultivation and use of a cheaper inducer, lactose instead of IPTG, also helps to reduce the production costs. In addition, the higher content of PseFDH in the biomass (50-55% of the total soluble protein compared to 15% for CboFDH) significantly simplifies, and therefore, lowers the reagent consumption and the cost of purification. 8. Alternative enzymes for NAD(P)H regeneration Many enzymes were tested and used for NAD(P)H regeneration in processes of enzyme chiral synthesis and formate dehydrogenase is still the gold standard in this area (van der Donk and Zhao, 2003). Detailed information about most successful examples can be found in review (Wichmann and Vasic-Racki, 2005). Here, we will shortly describe two

alternatives to FDH enzymes, glucose dehydrogenase and phosphite dehydrogenase. The first one is already widely used in practice. Phosphite dehydrogenase is a newcomer in this area and looks a promising candidate for coenzyme regeneration. 8.1. Glucose dehydrogenase Gluconolactone, product of reaction catalyzed by glucose dehydrogenase, GDH, is spontaneously hydrolysed to gluconic acid. This makes the overall reaction irreversible and enables to use GDH for cofactor regeneration. Majority of GDHs show dual coenzyme specificity, however with a preference towards one of the coenzyme forms, either NAD+ or NADP+. Several isoenzymes of GDH can be found in one strain. For example, strain Bacillus megaterium IAM1030 harbours four GDH isoenzymes. Two isoenzymes prefer NAD+ as a coenzyme, while the other two as NADP+ (Nagao et al., 1992). Wide abundance of GDH in nature enables to search for the enzymes showing maximal activity at extreme conditions, e.g. under high temperatures (Bright et al., 1993), acidic pH (Angelov et al., 2005) or very high salt concentration (Bonete et al., 1996). A wide number of commercial GDH preparations are available from various sources (for example, see product catalogs of Biocatalytics Inc. and Julich Chiral Solutions). Higher specific activity of GDH (20-100 U/mg) compared to FDH (2.5-10 U/ mg) is the clear advantage of GDH-based NAD(P)H regeneration system over FDH-based. Glucose and ammonium formate have similar prices, but reducing equivalent capacity (REC) of glucose is about four-fold less. Reduction of one mole of NAD(P)H requires 172 g of glucose and only 45 g of formate (in calculation of REC we took into account only molecular mass of formate ion and did not consider molecular mass of ammonium ion because in the reaction it acts only as a buffer component). The other disadvantage of GDH-based regeneration systems results from the necessity to purify the end-product from gluconic acid. All microorganisms have a system of active transport of glucose (as well as formate) inside the cell. Therefore, during the last years GDH is actively used for coenzyme regeneration in the processes where whole cells are utilized as biocatalysts (Endo and Koizumi, 2001; Kataoka et al., 2003, 2004). To produce such a biocatalyst the main enzyme and GDH can be expressed in two separate strains (Liu et al., 2005; Xu et al., 2005), as well as coexpressed in one strain (Kataoka et al., 1997; Wada et al., 2003; Yun et al., 2005). In this whole-cell approach even intracellular pool of NAD(P)+ is enough to achieve the necessary level of cofactor regeneration (Ishige et al., 2005) and gluconic acid, product of glucose oxidation, can be further utilized by the cell as a carbon source. Similar recombinant E. coli strains, coexpressing formate dehydrogenase and two or more enzymes, were also constructed (Galkin et al., 1997a,b; Ernst et al., 2005). 8.2. Phosphite dehydrogenase Phosphite dehydrogenase from Pseudomonas stutzeri WM88 (PTDH) has been recently proposed as an alternative enzyme for NAD(P)H regeneration (Vrtis et al., 2005; Relyea and van der Donk, 2005). The enzyme catalyzes reaction of


106

V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110

phosphite oxidation to phosphate with corresponding reduction of NAD+ to NADH. The reaction is irreversible and can be used for coenzyme regeneration. Wild-type PTDH has the same kcat value as PseFDH, 7.3 sÐ1 (Costas et al., 2001). Km value for phos NAD+ does not depend on pH, while kcat =Km phite value shows a maximum in the pH range of 7.0-7.6. Above and below this range, dehydrogenase activity drops off steeply (Relyea et al., 2005). Primordial PTDH exhibits poor thermal stability. Temperature optimum of activity for the wild-type enzyme is as low as 35 8C (Costas et al., 2001). The random mutagenesis generated PTDH preparations showing an improved thermal stability, approximately 2.5-fold better than that of wild-type CboFDH (Johannes et al., 2005). Due to this fact the mutant PTDH was more effective than wt-CboFDH in the synthesis of tert-L-leucine (Johannes et al., 2005). In future it would be also interesting to compare mutant PTDH with PseFDH GAV, which is commercially available since 2001. At 50 8C PseFDH GAV has 100-fold higher thermal stability compared to wt-CboFDH and consequently 40-fold compared to current version of mutant PTDH. Wild-type PTDH is NAD+-specific, but exhibits rather high activity with NADP+. Value of kcat with NADP+ is only 50% of kcat value with NAD+ (Woodyer et al., 2003). Site-directed mutagenesis resulted in production of NADP+-specific PTDH (Woodyer et al., 2005). Its application to NADPH regeneration in the reaction of xylitol production from xylose catalyzed by xylose reductase showed a four-fold increase in the rate of synthesis of the final product as compared to the NADP+dependent PseFDH (version T5M8) and wt-PTDH. One has to admit, that the use of reaction of xylitol synthesis as a reference to compare the efficiency of NADPH regeneration is not the best choice, as the main enzyme, xylose reductase, is apparently the only one that is reported to be inhibited by formate (Neuhauser et al., 1998). Therefore, the lower reaction rate of xylitol production in the system using NADP+-specific PseFDH for coenzyme regeneration, in this particular case, can be attributed to the xylose reductase inhibition itself. Phosphite-ion has about two-fold better reducing equivalent capacity in comparison with glucose and is two-fold inferior compared to formate-ion. An advantage of PTDH over other dehydrogenases is the ease with which it can be used for the preparation of deuterated compounds. Preparation of deuterTable 6 Costs of formate dehydrogenases from different companiesa Enzyme Biocatalytics Inc. US$ per 1000 U NAD+-specific from C. boidiniib NAD+-specific from Pseudomonas sp.101c NADP+-specific from Pseudomonas sp.101
a b c d

ated phosphite from normal incubation in D2O at pH 2.0 et al., 2005). Therefore, the lower compared to prices for particularly glucose. 9. Conclusion

substrate can be performed by followed by liophylization (Vrtis price of deuterated phosphite is deuterated formate, alcohols and

Chiral compounds synthesis is a rapidly growing area in biotechnology. NAD(P)+-dependent dehydrogenases and reductases are the most effective biocatalysts for such type of processes. First, the use of these enzymes allows optically active molecules to be produced from non-chiral substrates. Second, the reactions catalyzed by dehydrogenases are extremely stereospecific (LaReau and Anderson, 1989; Weinhold et al., 1991), and the yield of the final product may reach 100%, while the processes based on kinetic resolution of racemic mixtures can provide the theoretic yield of 50%. The main disadvantage of the dehydrogenase-catalyzed process is the high cost of NADH and especially NADPH. The development of regeneration systems for reduced coenzymes and methods for their retention in bioreactors made great impact to use dehydrogenases in synthesis of chiral alcohols from ketones with alcohol dehydrogenases and natural and artificial amino acids with keto-acid dehydrogenases as well as to use monooxygenases for hydroxylation and epoxidation. The processes were reviewed in detail in Liese and Villela (1999), Liese (2005) and Wichmann and Vasic-Racki (2005). In the last decade, the improvement of formate dehydrogenase properties and development of large scale production with recombinant E. coli strains significantly reduced the cost of FDH in the overall production cost of target compounds. Production of NADP+-specific FDH opened the possibility for its application for the purposes of NADPH regeneration (Seelbach et al., 1996; Rissom et al., 1997; Schwarz-Linek et al., 2001; Maurer et al., 2003). Recombinant CboFDH is available in large volumes from Julich Chiral Solutions (Julich Fine Chemicals before January 2006) in Europe and from Biocatalytics Inc. in the USA (Table 6). Unfortunately, all mutations providing improvement of CboFDH properties are covered by patents and at the moment only wild-type enzyme is commercially available. Recombinant PseFDH GAV with improved chemical and thermal stability is available from Julich Chiral Solutions and Innovations and High Technologies MSU

Julich Chiral Solutions (Julich Fine Chemicals) US$ per 10,000 U 1950 - - s per 1000 U 290 - 680 s per 10,000 U 750 950 3400

Innovations and High Technologies MSU s per 1000 U - 150 420 s per 10,000 U - 570 2000

d

390 - -

Prices for December 2005. Recombinant wild-type enzyme. Mutant PseFDH GAV with increased chemical and thermal stability and improved affinity for NAD+. Mutant enzyme with extended pH-optimum for NADP+ and increased chemical and thermal stability.


V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110

107

(InnoTech MSU). Mutant NADP+-specific PseFDH is also offered by these companies (Table 6). In our opinion, if trying to predict the further development in this research area, the NADH regeneration with high purity enzymes will get broader commercial impact than that of NADPH regeneration. This prediction is based on high cost and low stability of NADP+ compared to NAD+. The NAD+dependent enzyme can be derived from the corresponding NADP+-specific analog. As an example of the concept, we cite here the successful change of coenzyme specificity of reductase in cytochrome P-450 monooxygenase from B. megaterium (P450 BM-3) (Urlacher and Schmid, 2004). Wild-type and mutant monooxygenase were successfully used for hydroxylation of poorly soluble compounds in combination with NADP+and NAD+-specific PseFDHs in two-phase water-cyclohexane system (Hofstetter et al., 2004; Maurer et al., 2003, 2005; Urlacher et al., 2005). Another prospective trend for FDH application research is metabolic engineering of recombinant strains. Expression of FDH gene in a recombinant strain gives an additional supply of intracellular NADH and NADPH when growing in the presence of formate. This helps to release other metabolic pathways producing NADH or NADPH and redirect them to the synthesis of the target product (Berrios-Rivera et al., 2002a,b, 2004; San et al., 2002; Kaup et al., 2004; Sanchez et al., 2005). In conclusion, we think that the existence of numerous FDH genes from various sources opens new horizons in improvement of enzyme properties with gene shuffling. In the first turn, the attention will be paid to further increase in catalytic activity because FDH is still a ``slow'' enzyme compared to other dehydrogenases. Acknowledgments Authors thanks Dr. I. Gazaryan for help in manuscript preparation, Dr. S. Felber for presentation of his Ph.D. dissertation. This work was supported by grants from Russian Foundation for Basic Research (project a05-04-49073), NATO (grant no. LST.CLG 977839) and Russian Federal Agency for Science and Innovations (FASI contract 02.435.11.3005). References
Allen, S.J., Holbrook, J.J., 1995. Isolation, sequence and overexpression of the gene encoding NAD-dependent formate dehydrogenase from the methylotrophic yeast Candida methylica. Gene 162, 99-104. Angelov, A., Futterer, O., Valerius, O., Braus, G.H., Liebl, W., 2005. Properties of the recombinant glucose/galactose dehydrogenase from the extreme thermoacidophile, Picrophilus torridus. FEBS J. 272, 1054-1062. Baba, T., Takeuchi, F., Kuroda, M., Yuzawa, H., Aoki, K.-I., Oguchi, A., Nagai, Y., Iwama, N., Asano, K., Naimi, T., 2002. Genome and virulence determinants of high virulence community-acquired MRSA. Lancet 359, 1819- 1827. Barnett, M.J., Fisher, R.F., Jones, T., Komp, C., Abola, A.P., Barloy-Hubler, F., Bowser, L., Capela, D., Galibert, F., Gouzy, J., Gurjal, M., Hong, A., Huizar, L., Hyman, R.W., Kahn, D., Kahn, M.L., Kalman, S., Keating, D.H., Palm, C., Peck, M.C., Surzycki, R., Wells, D.H., Yeh, K.C., Davis, R.W., Federspiel, N.A., Long, S.R., 2001. Nucleotide sequence and predicted

functions of the entire Sinorhizobium meliloti pSymA megaplasmid. Proc. Natl. Acad. Sci. U.S.A. 98, 9883-9888. Beja, O., Aravind, L., Koonin, E.V., Suzuki, M.T., Hadd, A., Nguyen, L.P., Jovanovich, S.B., Gates, C.M., Feldman, R.A., Spudich, J.L., Spudich, E.N., DeLong, E.F., 2000. Bacterial rhodopsin: evidence for a new type of phototrophy in the sea. Science 289, 1902-1906. Berrios-Rivera, S.J., Bennett, G.N., San, K.Y., 2002a. The effect of increasing NADH availability on the redistribution of metabolic fluxes in Escherichia coli chemostat cultures. Metab. Eng. 4, 230-237. Berrios-Rivera, S.J., Bennett, G.N., San, K.Y., 2002b. Metabolic engineering of Escherichia coli: increase of NADH availability by overexpressing an NAD+-dependent formate dehydrogenase. Metab. Eng. 4, 217-229. Berrios-Rivera, S.J., Sanchez, A.M., Bennett, G.N., San, K.Y., 2004. Effect of different levels of NADH availability on metabolite distribution in Escherichia coli fermentation in minimal and complex media. Appl. Microbiol. Biotechnol. 65, 426-432. Bommarius, A.S., Schwarm, M., Stingl, K., Kottenhahn, M., Huthmacher, K., Drauz, K., 1995. Synthesis and use of enantiometrically pure tert-leucine. Tetrahedron-Asymmetry 6, 2851-2888. Bonete, M.J., Pire, C., Lorca, F.I., Camacho, M.L., 1996. Glucose dehydrogenase from the halophilic Archaeon Haloferax mediterranei: enzyme purification, characterisation and N-terminal sequence. FEBS Lett. 383, 227-229. Bright, J.R., Byrom, D., Danson, M.J., Hough, D.W., Towner, P., 1993. Cloning, sequencing and expression of the gene encoding glucose dehydrogenase from the thermophilic archaeon Thermoplasma acidophilum. Eur. J. Biochem. 211, 549-554. Burton, S.G., 2003. Oxidizing enzymes as biocatalysts. Trends Biotechnol. 21, 543-549. Cazalet, C., Rusniok, C., Bruggemann, H., Zidane, N., Magnier, A., Ma, L., Tichit, M., Jarraud, S., Bouchier, C., Vandenesch, F., Kunst, F., Etienne, J., Glaser, P., Buchrieser, C., 2004. Evidence in the Legionella pneumophila genome for exploitation of host cell functions and high genome plasticity. Nat. Genet. 36, 1165-1173. Chien, M., Morozova, I., Shi, S., Sheng, H., Chen, J., Gomez, S.M., Asamani, G., Hill, K., Nuara, J., Feder, M., Rineer, J., Greenberg, J.J., Steshenko, V., Park, S.H., Zhao, B., Teplitskaya, E., Edwards, J.R., Pampou, S., Georghiou, A., Chou, I.-C., Iannuccilli, W., Ulz, M.E., Kim, D.H., Geringer-Sameth, A., Goldsberry, C., Morozov, P., Fischer, S.G., Segal, G., Qu, X., Rzhetsky, A., Zhang, P., Cayanis, E., De Jong, P.J., Ju, J., Kalachikov, S., Shuman, H.A., Russo, J.J., 2004. The genomic sequence of the accidental pathogen Legionella pneumophila. Science 305, 1966-1968. Ä Colas des Francs-Small, C., Ambard-Bretteville, F., Small, I.D., Remy, R., 1993. Identification of a major soluble protein in mitochondria from nonphotosynthetic tissues as NAD-dependent formate dehydrogenase. Plant Physiol. 102, 1171-1177. Costas, A.M., White, A.K., Metcalf, W.W., 2001. Purification and characterization of a novel phosphorus-oxidizing enzyme from Pseudomonas stutzeri WM88. J. Biol. Chem. 276, 17429-17436. Davidson, D.C., 1951. Studies on plant formic dehydrogenase. Biochem. J. 49, 520-526. Demchenko, A.P., Rusyn, O.I., Egorov, A.M., Tishkov, V.I., 1990. The solvent effects on the kinetics of bacterial formate dehydrogenase. Biochim. Biophys. Acta 1039, 290-296. Dujon, B., Sherman, D., Fischer, G., Durrens, P., Casaregola, S., Lafontaine, I., De Montigny, J., Marck, C., Neuveglise, C., Talla, E., Goffard, N., Frangeul, L., Aigle, M., Anthouard, V., Babour, A., Barbe, V., Barnay, S., Blanchin, S., Beckerich, J.M., Beyne, E., Bleykasten, C., Boisrame, A., Boyer, J., Cattolico, L., Confanioleri, F., De Daruvar, A., Despons, L., Fabre, E., Fairhead, C., Ferry-Dumazet, H., Groppi, A., Hantraye, F., Hennequin, C., Jauniaux, N., Joyet, P., Kachouri, R., Kerrest, A., Koszul, R., Lemaire, M., Lesur, I., Ma, L., Muller, H., Nicaud, J.M., Nikolski, M., Oztas, S., OzierKalogeropoulos, O., Pellenz, S., Potier, S., Richard, G.F., Straub, M.L., Suleau, A., Swennen, D., Tekaia, F., Wesolowski-Louvel, M., Westhof, E., Wirth, B., Zeniou-Meyer, M., Zivanovic, I., Bolotin-Fukuhara, M., Thierry, A., Bouchier, C., Caudron, B., Scarpelli, C., Gaillardin, C., Weissenbach, J., Wincker, P., Souciet, J.L., 2004. Genome evolution in yeasts. Nature 430, 35-44.


108

V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110 3-hydroxybutanoate: asymmetric reduction of ethyl 4-chloro-3-oxobutanoate by an Escherichia coli transformant expressing the aldehyde reductase gene from yeast. Appl. Microbiol. Biotechnol. 48, 699-703. Kataoka, M., Kita, K., Wada, M., Yasohara, Y., Hasegawa, J., Shimizu, S., 2003. Novel bioreduction system for the production of chiral alcohols. Appl. Microbiol. Biotechnol. 62, 437-445. Kataoka, M., Kotaka, A., Thiwthong, R., Wada, M., Nakamori, S., Shimizu, S., 2004. Cloning and overexpression of the old yellow enzyme gene of Candida macedoniensis, and its application to the production of a chiral compound. J. Biotechnol. 114, 1-9. Kaup, B., Bringer-Meyer, S., Sahm, H., 2004. Metabolic engineering of Escherichia coli: construction of an efficient biocatalyst for D-mannitol formation in a whole-cell biotransformation. Appl. Microbiol. Biotechnol. 64, 333-339. Kula, M.-R., Wandrey, C., 1987. Continuous enzymatic transformation in an enzyme-membrane reactor with simultaneous NADH regeneration. Methods Enzymol. 136, 9-21. Labrou, N.E., Rigden, D.J., Clonis, Y.D., 2000. Characterization of the NAD+ binding site of Candida boidinii formate dehydrogenase by affinity labelling and site-directed mutagenesis. Eur. J. Biochem. 267, 6657- 6664. Labrou, N.E., Rigden, D.J., 2001. Active-site characterization of Candida boidinii formate dehydrogenase. Biochem. J. 354, 455-463. Lamzin, V.S., Dauter, Z., Popov, V.O., Harutyunyan, E.H., Wilson, K.S., 1994. High resolution structures of holo and apo formate dehydrogenase. J. Mol. Biol. 236, 759-785. Lamzin, V.S., Dauter, Z., Wilson, K.S., 1995. How nature deals with stereoisomers. Curr. Opin. Struct. Biol. 5, 830-836. LaReau, R.D., Anderson, V.E., 1989. Lactate dehydrogenase displays absolute stereospecificity in the transfer of the prochiral hydrogen of NADH. J. Biol. Chem. 264, 15338-15343. Larsson, P., Oyston, P.C., Chain, P., Chu, M.C., Duffield, M., Fuxelius, H.H., Garcia, E., Halltorp, G., Johansson, D., Isherwood, K.E., Karp, P.D., Larsson, E., Liu, Y., Michell, S., Prior, J., Prior, R., Malfatti, S., Sjostedt, A., Svensson, K., Thompson, N., Vergez, L., Wagg, J.K., Wren, B.W., Lindler, L.E., Andersson, S.G., Forsman, M., Titball, R.W., 2005. The complete genome sequence of Francisella tularensis, the causative agent of tularemia. Nat. Genet. 37, 153-159. Li, L., Bannantine, J.P., Zhang, Q., Amonsin, A., May, B.J., Alt, D., Banerji, N., Kanjilal, S., Kapur, V., 2005. The complete genome sequence of Mycobacterium avium subspecies paratuberculosis. Proc. Natl. Acad. Sci. U.S.A. 102, 12344-12349. Liese, A., Villela, M., 1999. Production of fine chemicals using biocatalysis. Curr. Opin. Biotechnol. 10, 595-603. Liese, A., 2005. Technical application of biological principles in asymmetric catalysis. Adv. Biochem. Eng./Biotechnol. 92, 197-224. Liu, Y., Xu, Z., Jing, K., Jiang, X., Lin, J., Wang, F., Cen, P., 2005. Asymmetric reduction of ethyl 4-chloro-3-oxobutanoate to ethyl (R)-4-chloro-3-hydroxybutanoate with two co-existing, recombinant Escherichia coli strains. Biotechnol. Lett. 27, 119-125. Loftus, B.J., Fung, E., Roncaglia, P., Rowley, D., Amedeo, P., Bruno, D., Vamathevan, J., Miranda, M., Anderson, I.J., Fraser, J.A., Allen, J.E., Bosdet, I.E., Brent, M.R., Chiu, R., Doering, T.L., Donlin, M.J., D'Souza, C.A., Fox, D.S., Grinberg, V., Fu, J., Fukushima, M., Haas, B.J., Huang, J.C., Janbon, G., Jones, S.J., Koo, H.L., Krzywinski, M.I., Kwon-Chung, J.K., Lengeler, K.B., Maiti, R., Marra, M.A., Marra, R.E., Mathewson, C.A., Mitchell, T.G., Pertea, M., Riggs, F.R., Salzberg, S.L., Schein, J.E., Shvartsbeyn, A., Shin, H., Shumway, M., Specht, C.A., Suh, B.B., Tenney, A., Utterback, T.R., Wickes, B.L., Wortman, J.R., Wye, N.H., Kronstad, J.W., Lodge, J.K., Heitman, J., Davis, R.W., Fraser, C.M., Hyman, R.W., 2005. The genome of the basidiomycetous yeast and human pathogen Cryptococcus neoformans. Science 307, 1321-1324. Mathews, M.B., Vennesland, B., 1950. Enzymic oxidation of formic acid. J. Biol. Chem. 186, 667-682. Maurer, S.C., Schulze, H., Schmid, R.D., Urlacher, V., 2003. Immobilisation of P450BM-3 and an NADP+ cofactor recycling system: towards a technical application of heme-containing monooxygenases in fine chemical synthesis. Adv. Synth. Catal. 345, 802-810.

Endo, T., Koizumi, S., 2001. Microbial conversion with cofactor regeneration using genetically engineered bacteria. Adv. Synth. Catal. 343, 521-526. Ernst, M., Kaup, B., Muller, M., Bringer-Meyer, S., Sahm, H., 2005. Enantioselective reduction of carbonyl compounds by whole-cell biotransformation, combining a formate dehydrogenase and a (R)-specific alcohol dehydrogenase. Appl. Microbiol. Biotechnol. 66, 629-634. Fedorchuk, V.V., Galkin, A.G., Yasny, I.E., Kulakova, L.B., Rojkova, A.M., Filippova, A.A., Tishkov, V.I., 2002. Influence of interactions between amino acid residues 43 and 61 on thermal stability of bacterial formate dehydrogenases. Biochemistry (Mosc.) 67, 1145-1151. ? Felber, S., 2001. Optimierung der NAD-abhangigen Formiatdehydrogenase aus ? Candida boidinii fur den Einsatz in der Biokatalyse. Ph.D. Thesis. HeinrichHeine University of Duesseldorf. URL: http://diss.ub.uni-duesseldorf.de/ ebib/diss/file?dissid=78. Filippova, E.V., Polyakov, K.M., Tikhonova, T.V., Stekhanova, T.N., Boiko, K.M., Popov, V.O., 2005. Structure of a new crystal modification of the bacterial NAD+-dependent formate dehydrogenase with a resolution of À 2.1 A Crystallogr. Rep. 50, 796-800. Filippova, E.V., Polyakov, K.M., Tikhonova, T.V., Stekhanova, T.N., Boiko, K.M., Sadykhov, E.G., Tishkov, V.I., Popov, V.O., 2006. Crystal structure of NAD+-dependent formate dehydrogenase from Pseudomonas sp.101 in complex with formate. Crystallogr. Rep. 51, N2. 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, 479-483. Galkin, A., Kulakova, L., Yoshimura, T., Soda, K., Esaki, N., 1997a. Synthesis of optically active amino acids from alpha-keto acids with Escherichia coli cells expressing heterologous genes. Appl. Environ. Microbiol. 63, 4651- 4656. Galkin, A., Kulakova, L., Yamamoto, H., Tanizawa, K., Tanaka, H., Esaki, N., Soda, K., 1997b. Conversion of alpha-keto acids to D-amino acids by coupling of four enzyme reactions. J. Ferment. Bioeng. 83, 299-300. Gul-Karaguler, N., Sessions, R.B., Clarke, A.R., Holbrook, J., 2001. A single mutation in the NAD-specific formate dehydrogenase from Candida methylica allows the enzyme to use NADP. Biotechnol. Lett. 23, 283-287. Hofstetter, K., Lutz, J., Lang, I., Witholt, B., Schmid, A., 2004. Coupling of biocatalytic asymmetric epoxidation with NADH regeneration in organic- aqueous emulsions. Angew. Chem. Int. Ed. 43, 2163-2166. Hummel, W., Kula, M.R., 1989. Dehydrogenases for the synthesis of chiral compounds. Eur. J. Biochem. 184, 1-13. Hwang, L., Hocking-Murray, D., Bahrami, A.K., Andersson, M., Rine, J., Sil, A., 2003. Identifying phase-specific genes in the fungal pathogen Histoplasma capsulatum using a genomic shotgun microarray. Mol. Biol. Cell 14, 2314-2326. Innovations and High Technologies MSU (InnoTech MSU Ltd.), 2005. NADP+specific formate dehydrogenase from Pseudomonas sp.101. Product description. URL: http://www.innotech-msu.com/products/NADPFDH.pdf. Innovations and High Technologies MSU (InnoTech MSU Ltd.), 2006. Mutant formate dehydrogenase from Pseudomonas sp.101 with improved affinity for NAD+ and increased chemical and thermal stability (version PseFDH GAV) overexpressed in E. coli. Product description. URL: http://www.innotech-msu.com/products/NADFDHGAV.pdf. Ishige, T., Honda, K., Shimizu, S., 2005. Whole organism biocatalysis. Curr. Opin. Chem. Biol. 9, 174-180. Johannes, T.W., Woodyer, R.D., Zhao, H., 2005. Directed evolution of a thermostable phosphite dehydrogenase for NAD(P)H regeneration. Appl. Environ. Microbiol. 71, 5728-5734. Jones, T., Federspiel, N.A., Chibana, H., Dungan, J., Kalman, S., Magee, B.B., Newport, G., Thorstenson, Y.R., Agabian, N., Magee, P.T., Davis, R.W., Scherer, S., 2004. The diploid genome sequence of Candida albicans. Proc. Natl. Acad. Sci. U.S.A. 101, 7329-7334. Julich Fine Chemicals, 2005. Product Catalog. Karaguler, N.G., Sessions, R.B., Moreton, K.M., Clarke, A.R., Holbrook, J.J., 2004. Estimating the energetic contribution of hydrogen bonding to the stability of Candida methylica formate dehydrogenase by using double mutant cycle. Biotechnol. Lett. 26, 1137-1140. Kataoka, M., Rohani, L.P., Yamamoto, K., Wada, M., Kawabata, H., Kita, K., Yanase, H., Shimizu, S., 1997. Enzymatic production of ethyl (R)-4-chloro-


V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110 Maurer, S.C., Kuehnel, K., Kaysser, L.A., Eiben, S., Schmid, R.D., Urlacher, V.B., 2005. Catalytic hydroxylation in biphasic systems using CYP102A1 mutants. Adv. Synth. Catal. 347, 1090-1098. Mesentsev, A.V., Lamzin, V.S., Tishkov, V.I., Ustinnikova, T.B., Popov, V.O., 1997. Effect of pH on kinetic parameters of NAD+-dependent formate dehydrogenase. Biochem. J. 321, 475-480. Mitsunaga, T., Tanaka, Y., Yoshida, T., Watanabe, K., 2000. New Hyphomicrobium sp. formate dehydrogenase gene for producing formate dehydrogenase of high specific activity, high temperature stability and high pH stability. Patent of Japan JP245471A2, 12.09.2000. Nagao, T., Mitamura, T., Wang, X.H., Negoro, S., Yomo, T., Urabe, I., Okada, H., 1992. Cloning, nucleotide sequences, and enzymatic properties of glucose dehydrogenase isozymes from Bacillus megaterium IAM1030. J. Bacteriol. 174, 5013-5020. Nanba, H., Takaoka, Y., Hasegawa, J., 2003a. Purification and characterization of an alpha-haloketone-resistant formate dehydrogenase from Thiobacillus sp. strain KNK65MA, and cloning of the gene. Biosci. Biotechnol. Biochem. 67, 2145-2153. Nanba, H., Takaoka, Y., Hasegawa, J., 2003b. Purification and characterization of formate dehydrogenase from Ancylobacter aquaticus strain KNK607M, and cloning of the gene. Biosci. Biotechnol. Biochem. 67, 720-728. Neuhauser, W., Steininger, M., Haltrich, D., Kulbe, K.D., Nidetzky, B., 1998. A pH-controlled fed-batch process can overcome inhibition by formate in NADH-dependent enzymatic reductions using formate dehydrogenase-catalyzed coenzyme regeneration. Biotechnol. Bioeng. 60, 277-282. Odintseva, E.R., Popova, A.S., Rojkova, A.M., Tishkov, V.I., 2002. Role of cysteine residues in stability of bacterial formate dehydrogenase. Bull. Moscow Univ., Ser. 2 Chem. 43, 356-359. Olson, B.J., Skavdahl, M., Ramberg, H., Osterman, J.C., Markwell, J., 2000. Formate dehydrogenase in Arabidopsis thaliana: characterization and possible targeting to the chloroplast. Plant Sci. 159, 205-212. Omura, S., Ikeda, H., Ishikawa, J., Hanamoto, A., Takahashi, C., Shinose, M., Takahashi, Y., Horikawa, H., Nakazawa, H., Osonoe, T., Kikuchi, H., Shiba, T., Sakaki, Y., Hattori, M., 2001. Genome sequence of an industrial microorganism Streptomyces avermitilis: deducing the ability of producing secondary metabolites. Proc. Natl. Acad. Sci. U.S.A. 98, 12215-12220. Parkhill, J., Sebaihia, M., Preston, A., Murphy, L.D., Thomson, N., Harris, D.E., Holden, M.T.G., Churcher, C.M., Bentley, S.D., Mungall, K.L., CerdenoTarraga, A.M., Temple, L., James, K., Harris, B., Quail, M.A., Achtman, M., Atkin, R., Baker, S., Basham, D., Bason, N., Cherevach, I., Chillingworth, T., Collins, M., Cronin, A., Davis, P., Doggett, J., Feltwell, T., Goble, A., Hamlin, N., Hauser, H., Holroyd, S., Jagels, K., Leather, S., Moule, S., Norberczak, H., O'Neil, S., Ormond, D., Price, C., Rabbinowitsch, E., Rutter, S., Sanders, M., Saunders, D., Seeger, K., Sharp, S., Simmonds, M., Skelton, J., Squares, R., Squares, S., Stevens, K., Unwin, L., Whitehead, S., Barrell, B.G., Maskell, D.J., 2003. Comparative analysis of the genome sequences of Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica. Nat. Genet. 35, 32-40. Popov, V.O., Shumilin, I.A., Ustinnikova, T.B., Lamzin, V.S., Egorov, Ts.A., 1990. NAD-dependent formate dehydrogenase from methylotrophic bacterium Pseudomonas sp.101. I. Amino acid sequence. Bioorgan. Chem. 16, 324-335 (Russ.). Popov, V.O., Lamzin, V.S., 1994. NAD+-dependent formate dehydrogenase. Biochem. J. 301, 625-643. Popov, V.O., Tishkov, V.I., 2003. NAD+-dependent formate dehydrogenase. From a model enzyme to a versatile biocatalyst. In: Uversky, V.N. (Ed.), Protein Structures: Kaleidoscope of Structural Properties and Functions. Research Signpost, Kerala, pp. 441-473. Relyea, H.A., van der Donk, W.A., 2005. Mechanism and applications of phosphite dehydrogenase. Bioorg. Chem. 33, 171-189. Relyea, H.A., Vrtis, J.M., Woodyer, R., Rimkus, S.A., van der Donk, W.A., 2005. Inhibition and pH dependence of phosphite dehydrogenase. Biochemistry 44, 6640-6649. Resch, A., Rosenstein, R., Nerz, C., Gotz, F., 2005. Differential gene expression profiling of Staphylococcus aureus cultivated under biofilm and planktonic conditions. Appl. Environ. Microbiol. 71, 2663-2676. Rissom, S., Schwarz-Linek, U., Vogel, M., Tishkov, V.I., Kragl, U., 1997. Synthesis of chiral epsilon-lactones in a two-enzyme system of cyclohex-

109

anone mono-oxygenase and formate dehydrogenase with integrated bubblefree aeration. Tetrahedron-Asymmetry 8, 2523-2526. Rojkova, A.M., Galkin, A.G., Kulakova, L.B., Serov, A.E., Savitsky, P.A., Fedorchuk, V.V., Tishkov, V.I., 1999. Bacterial formate dehydrogenase. Increasing the enzyme thermal stability by hydrophobization of alpha helices. FEBS Lett. 445, 183-188. Rozzell, J.D., Hua, L., Mayhew, M., Novick, S., 2004. Mutants of enzymes and methods for their use. US Patent Application Publication US2004/0115691, 17.06.2004. Sadykhov, E.G., Serov, A.E., Voinova, N.S., Uglanova, S.V., Petrov, A.S., Alexeeva, A.A., Kleimenov, S.Yu., Popov, V.O., Tishkov, V.I., 2006. Comparable study of thermal stability of formate dehydrogenases from microorganisms and plants. Appl. Biochem. Microbiol. 42, N3. Sakai, Y., Murdanoto, A.P., 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, 4480-4485. Saleeba, J.A., Cobbett, C.S., Hynes, M.J., 1992. Characterization of the amdAregulated aciA gene of Aspergillus nidulans. Mol. Gen. Genet. 235, 349- 358. San, K.Y., Bennett, G.N., Berrios-Rivera, S.J., Vadali, R.V., Yang, Y.T., Horton, E., Rudolph, F.B., Sariyar, B., Blackwood, K., 2002. Metabolic engineering through cofactor manipulation and its effects on metabolic flux redistribution in Escherichia coli. Metab. Eng. 4, 182-192. Sanchez, A.M., Bennett, G.N., San, K.Y., 2005. Effect of different levels of NADH availability on metabolic fluxes of Escherichia coli chemostat cultures in defined medium. J. Biotechnol. 117, 395-405. Schwarz-Linek, U., Krodel, A., Ludwig, F.A., Schulze, A., Rissom, S., Kragl, U., Tishkov, V.I., Vogel, M., 2001. Synthesis of natural product precursors by Baeyer-Villiger oxidation with cyclohexanone monooxygenase from Acinetobacter. Synthesis-Stuttgart 33, 947-951. Seelbach, K., Riebel, B., Hummel, W., Kula, M.-R., Tishkov, V.I., Egorov, A.M., Wandrey, C., Kragl, U., 1996. A novel, efficient regenerating method of NADPH using a new formate dehydrogenase. Tetrahedron Lett. 37, 1377-1380. Serov, A.E., Popova, A.S., Fedorchuk, V.V., Tishkov, V.I., 2002. Engineering of coenzyme specificity of formate dehydrogenase from Saccharomyces cerevisiae. Biochem. J. 367, 841-847. Serov, A.E., 2002. Structure-function relationship in recombinant formate dehydrogenases from baker's yeast and methylotrophic bacterium. Ph.D. Thesis. M.V. Lomonosov Moscow State University, Moscow, Russian Federation. Serov, A.E., Tishkov, V.I., 2002. Role of Pro residues in stability of prokaryotic and euacaryotic formate dehydrogenases. Bull. Moscow Univ., Ser. 2 Chem. 43, 345-349. Serov, A.E., Odintzeva, E.R., Uporov, I.V., Tishkov, V.I., 2005. Use of Ramachandran plot for increasing thermal stability of bacterial formate dehydrogenase. Biochemistry (Mosc.) 70, 804-808. Shaked, Z., Whitesides, G.M., 1980. Enzyme-catalyzed organic synthesis: NADH regeneration by using FDH. J. Am. Chem. Soc. 102, 7104- 7105. Shinoda, T., Satoh, T., Mineki, S., Iida, M., Taguchi, H., 2002. Cloning, nucleotide sequencing, and expression in Escherichia coli of the gene for formate dehydrogenase of Paracoccus sp.12-A, a formate-assimilating bacterium. Biosci. Biotechnol. Biochem. 66, 271-276. Shinoda, T., Arai, K., Shigematsu-Iida, M., Ishikura, Y., Tanaka, S., Yamada, T., Kimber, M.S., Pai, E.F., Fushinobu, S., Taguchi, H., 2005. Distinct conformation-mediated functions of an active site loop in the catalytic reactions of NAD-dependent D-lactate dehydrogenase and formate dehydrogenase. J. Biol. Chem. 280, 17068-17075. Schirwitz, K., Schmidt, A., Lamzin, V.S., 2005. Site-directed mutagenesis and high-resolution structures of Candida boidinii formate dehydrogenase. HASYLAB Annual Report, DESY, Hamburg. http://www-hasylab.desy.de/ science/annual_reports/2005_report/part2/contrib/72/14698.pdf. Slusarczyk, H., Felber, S., Kula, M.R., Pohl, M., 2000. Stabilization of NADdependent formate dehydrogenase from Candida boidinii by site-directed mutagenesis of cysteine residues. Eur. J. Biochem. 267, 1280-1289.


110

V.I. Tishkov, V.O. Popov / Biomolecular Engineering 23 (2006) 89-110 hydroxy-2,2,6-trimethylcyclohexanone, by two-step enzymatic asymmetric reduction. Appl. Environ. Microbiol. 69, 933-937. Weerasinghe, P.A., Weerasekera, M.L.M.C., van Holm, L.H.J., 1999. Use of isoenzymes to differentiate growth categories of Pericopsis mooniana trees. Biol. Plant. 42, 41-547. Weinhold, E.G., Glasfeld, A., Ellington, A.D., Benner, S.A., 1991. Structural determinants of stereospecificity in yeast alcohol dehydrogenase. Proc. Natl. Acad. Sci. U.S.A. 88, 8420-8424. Weuster-Botz, D., Paschold, H., Striegel, B., 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, 131- 137. Wichmann, R., Wandrey, C., Buckmann, A.F., Kula, M.R., 1981. Continuous enzymatic transformation in an enzyme membrane reactor with simultaneous NAD(H) regeneration. Biotechnol. Bioeng. 23, 2789- 2802. Wichmann, R., Vasic-Racki, D., 2005. Cofactor regeneration at the lab scale. Adv. Biochem. Eng./Biotechnol. 92, 225-260. Woodyer, R., van der Donk, W.A., Zhao, H., 2003. Relaxing the nicotinamide cofactor specificity of phosphite dehydrogenase by rational design. Biochemistry 42, 11604-11614. Woodyer, R., Zhao, H., van der Donk, W.A., 2005. Mechanistic investigation of a highly active phosphite dehydrogenase mutant and its application for NADPH regeneration. FEBS J. 72, 3816-3827. Xu, Z., Liu, Y., Fang, L., Jiang, X., Jing, K., Cen, P., 2005. Construction of a two-strain system for asymmetric reduction of ethyl 4-chloro-3-oxobutanoate to (S)-4-chloro-3-hydroxybutanoate ethyl ester. Appl. Microbiol. Biotechnol. Published on-line: September 21, 2005. Yamamoto, H., Mitsuhashi, K., Kimoto, N., Kobayashi, Y., Esaki, N., 2005. Robust NADH-regenerator: improved alpha-haloketone-resistant formate dehydrogenase. Appl. Microbiol. Biotechnol. 67, 33-39. Yun, H., Choi, H.L., Fadnavis, N.W., Kim, B.G., 2005. Stereospecific synthesis of (R)-2-hydroxy carboxylic acids using recombinant E. coli BL21 overexpressing YiaE from Escherichia coli K12 and glucose dehydrogenase from Bacillus subtilis. Biotechnol. Prog. 21, 366-371.

Slusarczyk, H., Felber, S., Kula, M.-R., Pohl, M., 2003. Novel mutants of formate dehydrogenase from Candida boidinii. US Patent Application Publication US2003/0157664, 21.09.2003. Tishkov, V.I., Galkin, A.G., Egorov, A.M., 1991. NAD-dependent formate dehydrogenase of methylotrophic bacteria Pseudomonas sp.101: cloning, expression, and study of the gene structure. Dokl. Akad. Nauk SSSR 317, 745-748. Tishkov, V.I., Galkin, A.G., Marchenko, G.N., Egorova, O.A., Sheluho, D.V., Kulakova, L.B., Dementieva, L.A., Egorov, A.M., 1993. 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, 976-981. Tishkov, V.I., Galkin, A.G., Fedorchuk, V.V., Savitsky, P.A., Rojkova, A.M., Gieren, H., Kula, M.-R., 1999. Pilot scale production and isolation of recombinant NAD+- and NADP+-specific formate dehydrogenases. Biotechnol. Bioeng. 64, 187-193. Tishkov, V.I., Popov, V.O., 2004. Catalytic mechanism and application of formate dehydrogenase. Biochemistry (Mosc.) 69, 1252-1267. Urlacher, V.B., Schmid, R.D., 2004. Protein engineering of the cytochrome P450 monooxygenase from Bacillus megaterium. Methods Enzymol. 388, 208-224. Urlacher, V.B., Makhsumkhanov, A., Schmid, R.D., 2005. Biotransformation of beta-ionone by engineered cytochrome P450 BM-3. Appl. Microbiol. Biotechnol Published online: July 7, 2005. van der Donk, W.A., Zhao, H.M., 2003. Recent developments in pyridine nucleotide regeneration. Curr. Opin. Biotechnol. 14, 421-426. 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, 182-188. Vrtis, J.M., White, A.K., Metcalf, W.W., van der Donk, W.A., 2005. Phosphite dehydrogenase: versatile cofactor-regeneration enzyme. Angew. Chem. Int. Ed. Engl. 41, 3257-3259. Wada, M., Yoshizumi, A., Noda, Y., Kataoka, M., Shimizu, S., Takagi, H., Nakamori, S., 2003. Production of a doubly chiral compound (4R,6R)-4-