Enoate Reductases of Clostridia CLONING, SEQUENCING, AND EXPRESSION*

The enr genes specifying enoate reductases of Clostridium tyrobutyricum and Clostridium thermoaceticum were cloned and sequenced. Sequence comparison shows that enoate reductases are similar to a family of flavoproteins comprising 2,4-dienoyl-coenzyme A reductase from Escherichia coli and old yellow enzyme from yeast. The C. thermoaceticum enr gene product was expressed in recombinant Escherichia coli cells growing under anaerobic conditions. The recombinant enzyme was purified and characterized.

* This work was supported in part by grants from the Deutsche Forschungsgemeinschaft and the Fonds der Chemischen Industrie. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide and amino acid sequence(s) of the enr genes of C. tyrobutyricum and C. thermoaceticum and partial gene of C. kluyveri have been submitted to the GenBank TM  berg, Germany). The preparation of a rabbit antiserum against enoate reductase from C. tyrobutyricum has been reported earlier (12).
Microorganisms and Plasmids-Bacterial strains and plasmids used in this study are summarized in Table I.
Isolation of Clostridium DNA-Overnight cultures of Clostridium strains were lysed with sodium dodecyl sulfate (SDS). The lysates were treated with proteinase K and RNase A (19). Sodium chloride was added to a final concentration of 0.7 M, and cetyltrimethylammonium bromide was added to a final concentration of 1.25% (v/v). The mixture was extracted with one volume of chloroform/isoamyl alcohol (24:1) and subsequently with one volume of phenol/chloroform/isoamyl alcohol (25:24:1). High molecular weight DNA was obtained from the supernatant by isopropyl alcohol precipitation (19).
Plasmid DNA Isolation-Plasmid DNA was isolated with the plasmid DNA isolation kit from Qiagen (Hilden, Germany).
DNA Sequencing-DNA sequencing was performed by the automated dideoxy chain termination method (20) using a 377 Prism automated DNA sequencer from PerkinElmer Life Sciences (Norwalk, CT).
Purification of DNA Fragments-DNA fragments were isolated using the GeneClean II kit from Bio 101 Inc. (La Jolla, CA) or the PCR purification kit from Qiagen (Hilden, Germany).
Preparation of a DNA Linker-A reaction mixture containing 100 l of 10 mM Tris-HCl, pH 7.4, 10 mM MgCl 2 , 0.09 mM NaCl, 1.5 mM dithiothreitol, and 1 nmol of oligonucleotides LIG1 and LIG2 (Table IV) was incubated for 15 min at 94°C and for 15 min at 60°C. The mixture was stored at 4°C.
Preparation of a Vector for Direct Cloning of PCR Fragments-The plasmid vector pBluescript SKII Ϫ linearized with EcoRV and 3Ј-Ttailed as described by Mead et al. (22) was designated pBlue-EV-t.
Southern Blotting-Restriction fragments of chromosomal DNA were separated by agarose gel electrophoresis and transferred to Hybond N ϩ membranes by vacuum blotting. DNA/DNA hybridization was carried out as described in the ECL TM direct nucleic acid labeling and detection

3 101
a The primers were designed from peptide fragments and on the basis of sequence similarity of enoate reductase from C. tyrobutyricum with the enzymes listed in Table VII.
b The primers Eno5 and Eno6 were designed on the basis of the DNA sequence of the partial enr gene of C. tyrobutyricum from the plasmid pBlue-Eno3/4. The corresponding amino acid sequences of the oligonucleotides are written in bold letters (IUPAC one-letter code). Locations and orientation of these oligonucleotides and corresponding peptide fragments on the DNA, respectively, protein sequence of enoate reductase from C. tyrobutyricum are indicated with arrows. system from Amersham Pharmacia Biotech (Freiburg, Germany).
Cloning of the enr Gene of C. tyrobutyricum-PCR experiments with C. tyrobutyricum DNA were performed using degenerate oligonucleotides (Table II) in various combinations. An experiment using the oligonucleotides Eno3 and Eno4 afforded an amplificate of 1.4 kb which was ligated into plasmid pBLUE-EV-t. The ligation mixture was transformed into E. coli XL1-Blue cells. The resulting plasmid pBlueEno3/4 served as template in a PCR reaction using the oligonucleotides Eno5 and Eno6 as primers (Table II). The amplificate was labeled with the ECL TM direct nucleic acid labeling and detection system (Amersham Pharmacia Biotech) and was used as probe for Southern blot analysis of EcoRI-digested C. tyrobutyricum DNA where it was shown to hybridize with fragments of 1.3, respectively, 3.7 kbp. These fragments were cloned by colony blotting from minilibraries of EcoRI fragments affording the plasmids pUC18E1.3kb and pUC18E3.7kb as shown in Fig. 2.
Cloning of the enr Gene of C. thermoaceticum-PCR experiments were performed with C. thermoaceticum DNA as template using consensus primers shown in Table III in various combinations. The primer pair ERNTERM2 and ER5 afforded an amplificate of 1.6 kbp (Fig. 3). This DNA segment was cloned and sequenced into the plasmid pBlue-EV-t yielding the plasmid pBlue-ERNTERM2/ER5 which was shown by sequencing to contain an insert of 1603 bp.
To extend that sequence in the 3Ј direction, a PCR reaction was performed using chromosomal DNA as template and the oligonucleotides Th-ER2a and the degenerate "anchor-primer" (Life Technologies, Inc.) as primers. The resulting 0.5-kb amplificate served as template in a second PCR amplification using the oligonucleotides Th-ER2 and "UAP-primer" (Life Technologies, Inc.) as primers (Fig. 3). The resulting DNA fragment was cloned and sequenced.
To extend the DNA sequence of the open reading frame in the 5Ј-direction, chromosomal DNA was digested with AatII, EcoRI, or MfeI. The DNA fragments resulting from each experiment were linked to the double strand DNA obtained by hybridization of the oligonucleotides LIG1 and LIG2 (see above). The ligation mixtures were used as template for PCR amplifications using the oligonucleotides LIG1 and Th-ERIa. A 0.3-kbp amplificate obtained from the experiment per-  Tables III and IV.
a The primers were designed from the DNA sequence of the enr gene of C. tyrobutyricum.

Enoate Reductases of Clostridia
formed with AatII-digested DNA was used as template in a second PCR amplification using the oligonucleotides LIG1 and Th-ERIIa (Fig. 3). The resulting DNA fragment was cloned and sequenced.
Cloning of the enr Gene of C. kluyveri-A major part of the C. kluyveri enr gene was obtained by the same approach as described above for the C. thermoaceticum gene. Amplification of C. kluyveri DNA with primers ER1 and ER4 (Table III) afforded a 1-kbp segment which was cloned and sequenced. Sequence extension in 5Ј direction was performed by ligation mediated two consecutive nested PCR experiments using the oligonucleotides LIG1, K-ER2 and K-ER2a (Table IV), as described above. The resulting 0.4-kb DNA fragment was cloned and sequenced.
Construction of an Expression Plasmid for the enr Gene of C. tyrobutyricum-The enr gene was assembled from two DNA segments prepared as described below. Amplification of C. tyrobutyricum DNA with the primers EnoA and Th-ERY1 (Table V) afforded a 0.4-bp fragment. This fragment served as template for a second amplification using the primers kEcoRI and Th-ERY1 (Table V). The resulting fragment was digested with EcoRI and BamHI. Similarly, PCR amplification of C. tyrobutyricum DNA with the primers ER2BamII and EnoB (Table V) afforded a 1.6-bp segment which was digested with BamHI.
The segments were mixed and ligated into plasmid pNCO113 which had been digested with EcoRI and BamHI. The ligation mixture was transformed into cells of E. coli XL1-Blue cells. Bacterial clones carrying inserts resulting from recombination of the two C. tyrobutyricum DNA segments were identified by restriction analysis followed by DNA sequence determination. The resulting plasmid pNCO-ERCTYR containing the entire enr gene under control of a T 5 promoter and lac operator was transformed into E. coli M15(pREP4) (23) cells affording the recombinant E. coli strain M15 pNCO-ERCTYR.
Construction of an Expression Clone for the enr Gene of C. thermoaceticum-The enr gene of C. thermoaceticum was amplified by PCR using the primers ThPNV and ThPNH2 (Table V) and chromosomal DNA as template. The resulting fragment served as template in a second PCR round using the primers kEcoRI and ThPNH2 (Table V). The PCR product was digested with EcoRI and PstI and ligated into the EcoRI and PstI digested expression vector pNCO113 which had been treated with the same restriction enzyme. The resulting plasmid pNCO-ERCTHERM was transferred into E. coli M15(pREP4) cells which were maintained on medium containing kanamycin (50 g ml Ϫ1 ) and ampicillin (180 g ml Ϫ1 ).
Fermentation-The recombinant E. coli strain M15(pREP4) harboring the expression plasmid pNCO-ERCTHERM was grown under anaerobic conditions in terrific broth (18) using a 20-liter fermenter (type L1523 from Bioengineering AG, Wald, Switzerland). The autoclaved culture medium was cooled to 37°C under nitrogen atmosphere (2.5 bar). Ampicillin and kanamycin were added to final concentrations of 100, respectively, 10 g ml Ϫ1 .
The medium was inoculated with an anaerobic overnight culture of the recombinant strain at a ratio of 1:50. The culture was incubated at 37°C with weak agitation (60 rpm), and the pH was adjusted to 7.0 at intervals by the addition of 1 M sodium hydroxide.
At a cell density of 0.6 OD (600 nm), isopropyl-1-thio-␤-D-galactopyranoside was added to a final concentration of 2 mM, and incubation was continued overnight. Cells were harvested by centrifugation at 5,000 rpm in a GS3 rotor (Sorvall RC-5B Plus centrifuge, DuPont Instruments, Bad Homburg, Germany) for 30 min and were stored at Ϫ20°C.
Assay of Enoate Reductase-Enzyme activity was determined by initial rate measurements under anaerobic conditions at 37°C. Reaction mixtures contained 100 mM potassium phosphate, pH 6.8, 0.2 mM NADH, and 5-250 g of protein/ml. Reactions were started by the addition of the electron acceptor, (E)-2-methyl-2-butenoate, to a final concentration of 0.1 M. The reaction was monitored photometrically (334 nm). One unit of enzyme activity catalyzes the dehydrogenation of 1 mol min Ϫ1 of NADH at 37°C (24).
Protein Sequencing-Enoate reductase from C. tyrobutyricum was treated with 2-mercaptoethanol, subsequently alkylated with 4-vinylpyridine and treated with cyanogen bromide (25). The resulting peptides were separated by high performance liquid chromatography and N-terminal sequences were obtained by automated Edman degradation (25).
Estimation of Protein Concentration-Protein concentration was estimated by a dye-binding assay using bovine serum albumin as standard (26).
Western Blotting-Protein samples dissolved in 3% SDS containing 3% mercaptoethanol and 0.1% bromphenol blue were subjected to SDS-PAGE as described above. The proteins were electrophoretically transferred to Immobilon TM P transfer membranes using the Transblot SD from Bio-Rad (Munich, Germany) with a constant current of 2 mA cm Ϫ2 gel for 1 h (28). A rabbit polyclonal antiserum directed against enoate reductase from C. tyrobutyricum (12) (diluted 1:5000) was used as the

RESULTS
Cloning of the enr Gene of C. tyrobutyricum-Enoate reductase of C. tyrobutyricum (14) was treated with cyanogen bromide, and the resulting peptides were separated and analyzed by Edman degradation. Partial amino acid sequences of the N terminus and of 5 peptide fragments (Table VI) suggested sequence similarity between enoate reductase and several flavin oxidoreductases shown in Table VII. Degenerate oligonucleotides (Table II) were designed on the basis of the partial sequences. PCR amplification using oligonucleotides Eno3 and Eno4 as primers and chromosomal DNA of C. tyrobutyricum as template afforded an amplificate of ϳ1.4 kbp (Fig. 2) which was cloned into the pBluescript SKII Ϫ plasmid affording plasmid pBlueEno3/4. Sequence analysis indicated that the cloned fragment represented a gene segment specifying 491 amino acid residues of enoate reductase. Specifically, two of the known peptide fragments (peptides 3 and 4) ( Table VI) were reflected by the partial gene sequence (Fig. 4).
The segment of the enr gene described above contains a EcoRI restriction site. Based on this information, a strategy was designed to clone the entire enr gene in two seperate parts. For this purpose, C. tyrobutyricum DNA was digested with EcoRI and analyzed by Southern blotting. A probe derived from the known gene segment recognized two fragments of 1.3 and 3.7 kbp. Minilibraries containing fragments of these respective lengths were then constructed and screened by colony hybridization affording two plasmids with 1.3-and 3.7-kbp inserts which were sequenced from both ends.
The resulting DNA sequence (GenBank accession number Y09960) comprised an open reading frame of 2001 bp preceded by a putative ribosomal binding site and followed by a putative palindromic terminator sequence (Fig. 4). The open reading frame predicted a protein of 667 amino acids with a molecular mass of 72.8 kDa in line with the relative mass of 73 kDa which had been reported earlier for enoate reductase of C. tyrobutyricum (14).
Cloning of the enr Genes from C. thermoaceticum and C. kluyveri-Consensus PCR primers (Table III) were designed on the basis of sequence similarity between enoate reductase of C. tyrobutyricum and the enzymes shown in Table VII. Specifically, the oligonucleotides were designed to match highly conserved segments of the enoate reductase sequence as shown in Fig. 5.
The oligonucleotides were used as primers in PCR experiments with chromosomal DNA from various Clostridia as templates (Table I). A 1-kb fragment of the putative enr gene was obtained from C. kluyveri DNA using the primers ER1 and ER4, and a 1.6-kb fragment was obtained from C. thermoaceticum DNA using the primers ERNTERM2 and ER5. These fragments were cloned into the plasmid vector pBlue-EV-t and sequenced.
The partial open reading frame from C. thermoaceticum was extended in both directions as described under "Experimental Procedures." The resulting DNA segment of 2142 bp (GenBank accession number Y16136) comprised an open reading frame of 2001 bp predicting a peptide of 667 amino acid residues with a mass of 73.0 kDa. The partial enr gene of C. kluyveri was similarly extended as described under "Experimental Procedures," but only a partial open reading frame of 1359 bp specifying 453 amino acid residues was obtained.
The predicted amino acid sequences of the enoate reductases from C. tyrobutyricum and C. thermoaceticum and the amino acid sequence predicted by the partial open reading frame C. kluyveri are highly conserved. The sequences of the proteins from C. thermoaceticum and C. tyrobutyricum show 59% identity. The identity between the partial enoate reductase from C. kluyveri and C. tyrobutyricum is 75%.  ( Fig. 6). Cell extracts had a specific enoate reductase activity of 0.1 mol min Ϫ1 mg Ϫ1 , pH 6.8, at 37°C with (E)-2-methylbutenoate and NADH as substrates. This value was similar to the enoate reductase in cell extracts of C. tyrobutyricum (0.2 mol min Ϫ1 mg Ϫ1 ).

Recombinant Expression of Enoate Reductase-The open
The recombinant enzyme was purified to apparent homogeneity by column chromatography as described under "Experimental Procedures" (Fig. 6). It had a specific activity of 2.4 mol min Ϫ1 mg Ϫ1 . Gel filtration experiments indicated a relative mass larger than 0.6 MDa. The enzyme showed activity in the pH range from 5.0 to 9.0 with a maximum around pH 7.3. NADPH could not serve as substrate. The purified protein migrated as a single band on SDS-PAGE gels at 73 kDa (Fig. 6). These characteristics are similar to those of enoate reductase from C. tyrobutyricum (24).
The N-terminal sequence (34 amino acids) of purified enoate reductase from C. thermoaceticum was determined by automated Edman degration. The amino acid sequence was identical to that predicted from the DNA sequence.

DISCUSSION
The amino acid sequences predicted by enr genes of C. tyrobutyricum, C. kluyveri, and C. thermoaceticum indicate that enoate reductases belong to a family of pyridine nucleotide-dependent flavoproteins (Table VII). An alignment of the sequences of enoate reductase from C. tyrobutyricum and the other protein sequences with the exception of old yellow enzyme from yeast is shown in Fig. 5. All sequences showed homology over their entire lengths (29 -34% identity to enoate reductase of C. tyrobutyricum). A region with 4 conserved cysteine residues is present in all sequences. The consensus pattern C-(2X)-C-(2-3X)-C-(11-12X)-C is similar to Fe 4 S 4 clusters in ferredoxins and other iron-sulfur proteins (30, 31). X-ray crystallographic studies of trimethylamine dehydrogenase indicated that these 4 cysteins are ligands for the Fe 4 S 4 cluster (32). With the exception of trimethylamine dehydrogenase and dimethylamine dehydrogenase, all sequences show two other well conserved motifs (amino acid residues 404 -431 and 529 -546 in enoate reductase) with similarity to the binding sites for the ADP moiety of FAD and NAD(P)H in other flavoproteins (32)(33)(34)(35)(36)(37).
The N terminus of enoate reductases (residue 1-368) shares 26% identity (44% similarity) with old yellow enzyme from Saccharomyces cerevisiae (EC 1.6.99.1) which was recently shown to reduce 2-enals and methyl-3-en-ketones (41). The stereochemical course of the reaction is trans, opposite to that of enoate reductase. Old yellow enzyme does not contain ironsulfur clusters. It catalyzes the transfer of the pro-R hydrogen from NADPH to the ␤-position of the substrate (42) in contrast to enoate reductase, which transfers the pro-S hydrogen (24).
Enoate reductases have no detectable similarity to enoyl-CoA reductases (43). This is in line with the different stereochemistry of the reaction products formed by the addition of hydrogen by enoate reductases and enoyl-CoA reductases (Fig.  1).
With the exception of trimethylamine dehydrogenase and dimethylamine dehydrogenase, the oxidoreductases with similarity to enoate reductases have subunit molecular masses the in the range of 71-73 kDa, but the subunit structure varies widely. All enzymes have been shown either directly or indirectly to have iron and acid-labile sulfide, and all contain at least one molecule of flavin. Reduced pyridine nucleotides served as electron donors with the exception of trimethylamine dehydrogenase and dimethylamine deydrogenase, which have trimethylamine or dimethylamine as electron donors, respectively. These facts correspond well to the features across the amino acid sequences concerning the putative FeS-cluster-, FAD-, and NAD-binding sites (Table VIII). A comparison of the amino acid sequences of enoate reductases showed strong conservation between C. tyrobutyricum and C. thermoaceticum (59% identity), although the GC content of the DNA sequences from these microorganisms differs widely (30 and 54%, respectively (44,45)).
Recombinant expression of enoate reductase from C. tyrobutyricum in aerobically grown E. coli host strains directed the formation of insoluble, inactive protein. The expression level was very low, so that the protein could only be detected on SDS-PAGE gels as a weak band and by Western blot analysis. This is likely because of the low GC content of the C. tyrobutyricum DNA (30%), which is also observed in the open reading frame of the enr gene. This fact is reflected in a much different codon usage of E. coli and C. tyrobutyricum, which results in up to 90 so-called regulatory codons in the open reading frame (46).
Recombinant enoate reductase from C. thermoaceticum from aerobically grown E. coli host cells was insoluble and enzymatically inactive too, but was expressed to a level of about 20% of total cell protein (data not shown). This is likely due to the different GC content resulting in only 20 regulatory codons identified across the DNA sequence of the enr gene from C. thermoaceticum. However, in anaerobically grown E. coli host cells recombinant enoate reductase was expressed in soluble and enzymatically active form. The recombinant enzyme was purified nearly to homogenity and preliminary studies on characterizations were performed indicating both differences and similarities in properties of the highly conserved enoate reductases from C. tyrobutyricum and C. thermoaceticum. The catalytic activity of recombinant enoate reductase from C. thermoaceticum (2.4 mol min Ϫ1 mg Ϫ1 ) is similar to that of the enzyme isolated from wild type C. tyrobutyricum (10.6 mol min Ϫ1 mg Ϫ1 ).