The proton-translocating NADH-quinone oxidoreductase (NDH-1) of thermophilic bacterium Thermus thermophilus HB-8. Complete DNA sequence of the gene cluster and thermostable properties of the expressed NQO2 subunit.

The genes encoding the proton-translocating NADH-quinone oxidoreductase (NDH-1) of a thermophilic bacterium Thermus thermophilus HB-8 were cloned and sequenced. They constitute a cluster that is composed of 14 structural genes and contains no unidentified reading frames. All of the 14 structural genes, which are designated NQO1-14, encode subunits homologous to those of Paracoccus denitrificans NDH-1, respectively, and are arranged in the same order as other bacterial NDH-1 genes. T. thermophilus NDH-1 contains at most nine putative iron-sulfur cluster binding sites, eight of which are commonly found in other organisms. The T. thermophilus NQO2 subunit was expressed in Escherichia coli The expressed subunit bears a single [2Fe-2S] cluster whose optical and EPR properties are very similar to those of N1a cluster in the P. denitrificans NQO2 subunit (Yano, T., Sled', V.D., Ohnishi, T., and Yagi, T. (1994) Biochemistry 33, 494-499). These results strongly suggest that the T. thermophilus NDH-1 is similar to other NDH-1 enzyme complexes in terms of subunit and cofactor composition. The T. thermophilus NQO2 subunit displayed much higher stability than the mesophilic equivalent and its iron-sulfur cluster remained intact even after incubation for 3 h at 65°C under anaerobic conditions. With the advantage of thermostability, the T. thermophilus NDH-1 provides a great model system to investigate the structure-function relationship of the NDH-1 enzyme complexes.

coupling Site I (18). Later, it was shown that T. thermophilus HB-8 possesses two types of NADH-quinone oxidoreductase (19). One is a proton-translocating NADH-dehydrogenase (NDH-1), and the other is a small non-energy-coupled enzyme referred to as NDH-2. Both enzymes have been purified and characterized to some extent. The NDH-1 was found to be composed of at least 10 dissimilar polypeptides and contained one non-covalently bound FMN, 11-12 mol of non-heme iron and 7-9 mol of acid-labile sulfide/mol. The NDH-2 was shown to be a single polypeptide with non-covalently bound FAD and no iron-sulfur clusters. Both enzymes can utilize NADH and deamino-NADH as substrates (in contrast to mesophilic NDH-2) and catalyze NADH-oxidation with water-soluble electron acceptors such as K 3 [Fe(CN) 6 ] and Q 1 at high temperature (Ն80°C) (19). More recently, iron-sulfur clusters of the membrane-bound T. thermophilus HB-8 NDH-1 enzyme complex have been studied by EPR spectroscopy. At least three ironsulfur clusters (one binuclear and two tetranuclear clusters) and possibly two additional iron-sulfur clusters (one binuclear and one tetranuclear) with very low redox mid-point potentials were detected in NDH-1 (20). Notably, cluster N2 signals were not seen in the study. We expect that the thermostability will provide a great advantage for structural studies of the NDH-1 either by NMR or x-ray crystallography. Furthermore, since T. thermophilus contains only menaquinone, it will provide an interesting system for understanding the Site I energy coupling mechanism.
This paper reports molecular cloning, sequencing, and expression studies of the proton-translocating NADH-quinone oxidoreductase (NDH-1) of T. thermophilus HB-8. The genes encoding the T. thermophilus NDH-1 enzyme complex constitute a gene cluster that is composed of 14 structural genes (designated NQO1-14) 2 and no URFs. All of the 14 structural genes encode subunits homologous to the P. denitrificans NDH-1 enzyme complex. Comparison of the deduced amino acid sequences with those of other organisms revealed that T. thermophilus NDH-1 possesses principally the same molecular structure as other bacterial NDH-1 in terms not only of subunit composition but also of cofactor binding sites. The NQO2 subunit of T. thermophilus NDH-1 was expressed in E. coli, purified, and partially characterized.

EXPERIMENTAL PROCEDURES
Genomic DNA Preparation-T. thermophilus HB-8 (ATCC27634) was grown in Thermus nutrient medium at 70°C to late exponential phase according to Hon-nami and Oshima (21). The cells were harvested by centrifugation in a SS34 rotor at 6,000 rpm for 10 min. The cells were washed once with 10 mM Tris-HCl (pH 8.0) containing 0.15 M NaCl and 0.1 mM EDTA and resuspended in 2.0 ml of 10 mM Tris-HCl buffer (pH 8.0) containing 0.1 mM EDTA (TE buffer). The cell suspension was then treated with 3 mg/ml lysozyme at 42°C for 20 min. Twelve ml of 0.1 M Tris-HCl (pH 8.0) containing 0.1 M NaCl and 1% (w/v) SDS were added, and the cell suspension was freeze-thawed three times with liquid nitrogen and a 60°C water bath. DNA was extracted with 14 ml of phenol/chloroform/TE for 10 min, followed by centrifugation three times. The aqueous phase was carefully transferred and dialyzed against 1 liter of 50 mM Tris-HCl (pH 8.0) containing 10 mM NaCl overnight with one buffer exchange. The DNA solution was incubated with 100 g/ml RNase A at 37°C for 3 h and then with 100 g/ml proteinase K and 0.1% (w/v) SDS at 50°C for 2 h. The solution was successively treated with an equal volume of phenol/chloroform/TE and then with chloroform/isoamyl alcohol. DNA was collected with a glass rod after the addition of 2.5 volumes of cold ethanol and suspended in TE buffer.
Cloning and Sequencing of the T. thermophilus NDH-1 Genes-To date, two complete (P. denitrificans and E. coli) and three partial DNA sequences (S. typhimurium, R. capsulatus, and Synechocystis) of the NDH-1 enzyme complexes have been deduced and are available in GenBank™ (6 -10, 13-16). In the two bacteria, the NDH-1 genes are organized as a gene cluster. Therefore, it could be anticipated that T. thermophilus NDH-1 genes also constitute a gene cluster. Since no amino acid sequence information was available in T. thermophilus NDH-1, we decided to amplify DNA fragments by polymerase chain reaction (PCR) with oligonucleotides designed on the basis of sequence similarities between organisms whose sequences had already been determined. The NADH-binding subunit of T. thermophilus NDH-1 was previously identified as the 47-kDa subunit by photoaffinity labeling with [ 32 P]NAD(H), and its amino acid composition was found to be very similar to those of P. denitrificans NQO1 and bovine heart Complex I FP 51-kDa subunit (22). We selected the NADH-binding subunit as the target, since it is expected that the primary sequence of the 47-kDa subunit is well conserved. The following oligonucleotides were derived from amino acid sequences that were highly conserved among several organisms (P. denitrificans NQO1, bovine heart complex I FP 51 kDa subunit, N. crassa 51-kDa subunit, E. coli NUOF, and S. typhimurium ). The amplification was performed in a thermocycler with a reaction program that consisted of one cycle of hot start (94°C for 10 min, 60°C for 5 min), 35 amplification cycles that were composed of denaturing (94°C for 1 min), annealing (50°C for 1.5 min), and elongation (72°C for 2 min), and finally one cycle of termination (72°C for 10 min). Amplified DNA was purified with QIAquick PCR purification kit (QIAGEN) and run on a 1.0% (w/v) agarose gel. The DNA fragments of desired molecular sizes were excised and purified from gel with QIAquick gel extraction kit (QIAGEN) and then subcloned into TA cloning vector based on pBluescript KS(ϩ) (Stratagene) that had been prepared according to Papp et al. (23). When two sets of oligonucleotides, TT1F (sense)/TT3R (antisense) and TT2F (sense)/TT3R (antisense), were used as primers, DNA fragments with molecular sizes of 540 and 480 bp were specifically amplified by PCR, respectively (data not shown). The 540-bp fragment contained an identical sequence to that of the 480-bp DNA fragment as expected. When the putative amino acid sequence derived from 540-bp DNA fragment was compared with those of the NADH-binding subunits of other organisms, a high degree of sequence similarity was found. Therefore, we concluded that the amplified DNA fragments encoded a part of the T. thermophilus NDH-1 NADH-binding subunit and could be used as a probe to clone the gene cluster. The standard cloning techniques used were according to Sambrook et al. (24). Southern blotting and colony hybridization were performed with DNA probes non-radioactively labeled with Genius 1 DNA labeling and detection kit (Boehringer Mannheim) according to the manufacture's protocol. With the 540-bp DNA fragment as the first probe, a 5.5-kilobase pair HindIII fragment was cloned (designated as pTTH-1) from a mini library constructed from 5.0 -6.0-kilobase pair HindIII fragments digested from genomic DNA of T. thermophilus HB-8. The clone contained genes that were homologous to the downstream half of the PdNQO4, the entire region of the Pd-NQO2, NQO1, and NQO3 and the upstream part of the PdNQO8 (Fig.  1). DNA encompassing the entire T. thermophilus NDH-1 gene cluster was subsequently obtained as five clones that were designated as pTTH1-5, respectively, as shown in Fig. 1. DNA sequencing was performed using the dideoxy-chain termination method with the Applied Biosystems dye terminator cycle sequencing kit and T3, T7, and synthetic unique oligonucleotide primer. The DNA sequence of total 14,886 bp (from SacII to BamHI in Fig. 1) and the predicted amino acid sequences were deposited in GenBank™ under an accession number of TAU52917. Sequence Analysis-GCG software programs were used to analyze the DNA and amino acid sequences (25). Sequence comparison of the polypeptides were conducted with BESTFIT and PILEUP programs. The FASTA and PROFILESEARCH programs were used to search the GenBank™/EMBL sequence data bases for proteins having some homology to the polypeptides. Homology search was also carried out by using the BLAST program running at the National Center for Biotechnology Information. Coding regions were searched with CODONPREF-ERENCE, and hydrophobic profiles were examined with HYDROPLOT.
Construction of an Expression Vector for the T. thermophilus NQO2 (TthNQO2) Subunit-The NQO2 gene was obtained from the pTTH-1 plasmid by PCR using two oligonucleotides to introduce NdeI site and BglII restriction sites at the translation initiation site and near the stop codon, respectively: NQO2F (sense), 5Ј-GGC ATA TGG GGT TCT TTG CGA C-3Ј; NQO2R (antisense), 5Ј-AAA GAT CTG GGC CAG TCA TAC CTC C-3Ј (The italic bases were altered from T. thermophilus DNA, and the underlined bases indicate the newly introduced restriction sites, NdeI site for NQO2F and BglII site for NQO2R). The amplification was performed as described previously. The amplified DNA fragment (560 bp) was purified by agarose gel electrophoresis and subcloned into the pBluescript KS(ϩ) vector. The integrity of the entire NQO2 gene was verified by sequencing. The construct was digested with NdeI and BglII and ligated into NdeI and BamHI sites of the pET16b expression vector. The final construct was designated pET16b(TthNQO2) and was used for the expression study.
Expression and Purification of the TthNQO2 Subunit from E. coli-Expression of the TthNQO2 subunit was conducted according to Yano et al. (26). The cells were suspended in 50 mM Tris-HCl (pH 7.4) containing 300 mM NaCl, 1.0 mM DTT, 1.0 mM PMSF, and 20 g/ml leupeptin. The cell suspensions were treated by freeze-thawing twice using liquid nitrogen and a 30°C water bath and were then passed twice through a Parr cell disruption bomb (Ͼ1,000 kg/cm 2 ). The resulting suspension was centrifuged in a SS34 rotor at 12,000 rpm for 10 min. The cell-free extracts thus obtained were degassed and purged with oxygen-free argon and then brought into an anaerobic chamber. Afterward, all procedures were performed under anaerobic conditions to prevent destruction of iron-sulfur clusters by oxygen. The cell-free extracts were centrifuged in a 60Ti rotor at 48,000 rpm for 60 min. The TthNQO2 subunit was purified from the cytoplasmic fraction following two different procedures. For Protocol 1, the cytoplasmic fractions were dialyzed against 1 liter of degassed 50 mM potassium phosphate buffer (pH 7.0) containing 1.0 mM PMSF, and 20 g/ml leupeptin for 6 h and then applied onto a nickel chelation column (1.0 ϫ 1.5 cm) equilibrated with the same buffer. The column was washed with 50 ml of degassed 50 mM potassium phosphate buffer (pH 7.0) containing 0.5 M NaCl, 1.0 mM PMSF, 20 g/ml leupeptin, and 20 g/ml pepstatin A, 80 mM imidazole, 10% (v/v) glycerol, and 0.1% (w/v) Tween 20, 50 ml of the same buffer without glycerol and Tween 20, and then the adsorbed TthNQO2 subunit was eluted with degassed 50 mM potassium phosphate buffer (pH 7.0) containing 0.5 M NaCl, 1.0 mM PMSF, and 400 mM imidazole. The purified subunit was dialyzed against 1 liter of 50 mM potassium phosphate buffer (pH 7.0) containing 1.0 mM DTT, and 0.1 mM PMSF overnight. For Protocol 2, the cytoplasmic fractions were incubated at 65°C for 60 min under an anaerobic condition and chilled on ice for 10 min. The resultant solutions were then centrifuged in a SS34 rotor at 12,000 rpm for 15 min to remove denatured proteins. The transparent supernatants thus obtained were dialyzed against 1 liter of degassed 50 mM potassium phosphate buffer (pH 7.0) containing 1.0 mM PMSF, and 20 g/ml leupeptin for 6 h and then applied onto a nickel chelation column (1.0 ϫ 1.5 cm). The TthNQO2 subunit was purified in the same way as Protocol 1.
Preparation of T. thermophilus HB-8 Membranes-T. thermophilus HB-8 membranes were prepared according to Yagi et al. (19). Cells were grown at 75°C in Thermus nutrient medium to late exponential phase according to Hon-nami and Oshima (21). Cells were harvested by centrifugation and suspended in 10 mM Tris-SO 4 buffer (pH 7.5) containing 10 mM MgSO 4 and 1.0 mM EDTA. The cell suspension was passed through a Parr cell disruption bomb (Ͼ1,000 kg/cm 2 ) twice. The resultant suspension was centrifuged at 30,000 ϫ g for 90 min. The pellet was homogenized with the same buffer and centrifuged. The pellet was resuspended in the same buffer. The membrane fractions were frozen in liquid nitrogen and stored at Ϫ80°C until used.
EPR Measurements-EPR spectra were recorded by a Bruker ESP300 E spectrometer operating in the X band (9.2 GHz). The sample temperature was varied using an Oxford instrument ESR-9 helium flow cryostat. The magnetic field was calibrated using a strong or weak pitch standard. Spin quantitations were performed under non-power-saturated conditions using 500 M Cu-EDTA as a standard.
Other Analytical Procedures-UV-visible absorption spectra were recorded on a SLM-Aminco DW-2000 spectrophotometer at room temperature. Protein concentration was estimated by the method of Lowry et al. (27) or by Biuret in the presence of 1 mg/ml sodium deoxycholate (28). SDS-polyacrylamide gel electrophoresis was carried out by a modified method of Laemmli (29). Immunoblotting was conducted as described previously (30 -32). Non-heme iron and acid-labile sulfide were determined according to Doeg and Ziegler (33) and Fogo and Popowski (34), respectively. DNA sequences were determined according to Sanger et al. (35). Any variations from the procedures and other details are described in the figure legends.
Materials-Acrylamide, N,NЈ-methylenebis(acrylamide), SDS, SDS-PAGE marker proteins, and Coomassie Brilliant Blue R-250 were from Bio-Rad; DTT, and horseradish peroxidase-conjugating affinity-purified antibodies to rabbit IgG were from Calbiochem; expression vectors, pET16b, and E. coli strain BL21(DE3)pLysS were from Novagen. All other chemicals were of the highest grade available from Sigma.

RESULTS
Structural Properties of the Gene Cluster-One of the unusual features of the T. thermophilus NDH-1 genes is their very high GϩC content (an average of 68.3%). High preference of GϩC in the third codon position (93.9%) results in a unique codon usage (17,36). In order to identify the reading frames of the cloned DNA, we utilized (i) a codon preference table that had been built on the basis of 45 T. thermophilus DNA coding sequences reported in GenBank™, (ii) amino acid sequence similarities to homologous subunits of other organisms (human, bovine heart, N. crassa, P. denitrificans, E. coli, S. typhimurium, R. capsulatus, etc.), and (iii) relative distance of putative translation initiation sites to ribosome binding site (Shine-Dalgarno sequence) (37). It was found that the cluster is composed of 14 open reading frames and contains no URFs ( Fig. 1 and Table I). Five structural genes utilize unique translation start codons. GTG start codons were predicted in NQO6, NQO8, NQO11, and NQO13, and TTG start codon in the NQO7 gene (Table I). To our knowledge, nine genes have been found to utilize GTG start codon out of 45 T. thermophilus DNA sequences in GenBank™. However, no TTG start codon has been reported to date in T. thermophilus. These rare start codons should be considered as tentative, and the actual translation start positions of the open reading frames must be experimentally identified in the future. The NDH-1 gene cluster contains a promoter-like sequence in its upstream region. The sequence 5Ј-CCCCTTGCG-3Ј and 5Ј-TAAGAT-3Ј are homologous to those of promoter regions of other T. thermophilus genes that have been empirically studied (38). These sequences are tentatively assigned to be Ϫ35 and Ϫ10 boxes, respectively. The gene cluster ends with a stem-loop-like structure that starts 49 bp downstream of the NQO14 stop codon. In addition, all 14 structural genes are so compactly organized in the cluster that there is almost no intergenic space where promoter or terminator-like sequences can be present. These properties strongly suggest that the T. thermophilus NDH-1 genes constitute an operon.
Predicted Amino Acid Sequence of Individual Subunits of the T. thermophilus NDH-1-The deduced primary sequences of the T. thermophilus NDH-1 subunits were compared with those of other organisms (e.g. bovine heart, N. crassa, P. denitrificans, and E. coli) and with those of some phylogenetically related enzymes (e.g. Alcaligenes eutrophus NAD ϩ -reducing hydrogenase, E. coli formate hydrogen lyase). All of the 14 subunits of the T. thermophilus NDH-1 are homologous to their mitochondrial and bacterial counterparts. Here, several important and interesting features of individual subunits are described in conjunction with the latest information of mitochondrial Complex I and bacterial NDH-1.
TthNQO1 Subunit-The TthNQO1 subunit is homologous to bovine FP 51-kDa, PdNQO1, and EcNuoF subunits ( Fig. 2 and Table II). This subunit plays an essential role in initiating the redox reaction by oxidizing NADH. As described above, the 47-kDa subunit was photoaffinity-labeled by [ 32 P]NAD(H), indicating that the NADH-binding site is located in the subunit (22). The TthNQO1 subunit conserves a typical nucleotide binding sequence motif (G 64 XG 66 XXG 69 ) with remote acidic amino acid residues D 94 E 95 . A tetranuclear iron-sulfur cluster (N3) is thought to be located in the NADH-binding subunits (39,40). The TthNQO1 subunit contains five conserved cysteine residues (Cys 182 , Cys 353 , Cys 356 , Cys 359 , and Cys 400 ). Four of them form a cluster with a typical motif for a [4Fe-4S] cluster binding site (C 353 XXC 356 XXC 359 ), and Cys 400 is followed by Pro (Fig. 2). Therefore, Cys 400 seems more likely to be the fourth ligand residue of the [4Fe-4S] cluster than Cys 182 . The Tth-NQO1 subunit and its homologues are also thought to accommodate an FMN that is an indispensable component for NADH oxidation (40). As discussed in previous papers (2,6,41), however, it is difficult to predict the FMN-binding site on the basis of sequence data at the present time mainly due to lack of consensus sequence motifs for this cofactor.
TthNQO3 Subunit-The TthNQO3 subunit is homologous to bovine IP 75-kDa, PdNQO3, and EcNuoG subunits ( Fig. 3 and Table II). This subunit generally contains 12 invariant cysteine residues in the N-terminal region that are possible iron-sulfur cluster ligands (8,14,43). Recently, expression studies of the PdNQO3 subunit showed that the subunit contains multiple iron-sulfur clusters: one [2Fe-2S] cluster (N1b), one [4Fe-4S] cluster (N4), and possibly another [4Fe-4S] cluster (44). Interestingly, it was found that the TthNQO3 subunit conserves   (Fig. 3). It is worth noting that this cysteine residue is also absent from the HoxU subunit of A. eutrophus NAD ϩ -reducing hydrogenase that contains the other 11 cysteine residues (Fig. 3) (45). Assuming three iron-sulfur clusters (one binuclear and two tetranuclear clusters) in the subunit, this implies that at least one of the clusters has to utilize a non-cysteinyl ligand residue. Currently, it is difficult to predict which iron-sulfur cluster utilizes a non-cysteinyl ligand residue. However, it can be anticipated that one of the [4Fe-4S] clusters is coordinated by four cysteine residues that form a FIG. 2. Comparison of the amino acid sequences of the T. thermophilus NQO1 subunit with its homologues from various organisms. The comparison was conducted using the PILEUP program. Bovine, bovine heart mitochondrial Complex I; P.d., P. denitrificans; T.th., T. thermophilus; HoxU, A. eutrophus NAD ϩ -reducing hydrogenase ␥ subunit. The putative nucleotide binding sites are indicated by #. * indicates the conserved cysteine residues. An arrow (a) indicates a proline residue that is unique to the T. thermophilus NQO1 subunit. Boxed sequences correspond to those used to synthesize the PCR primers TT1F, TT2F, and TT3R (indicated by arrows: 3, 4).  3 If another cluster, [4Fe-4S], is present, it must utilize a non-cysteinyl ligand within a novel binding motif. Furthermore, the TthNQO3 subunit contains an additional cysteine cluster (C 256 XXC 259 XXXC 263 . . . C 291 ) that is also found in E. coli and S. typhimurium NuoG subunits but not in other NDH-1 enzyme complexes (Fig. 3). It has been reported that the E. coli NDH-1 enzyme contains an additional EPR detectable [2Fe-2S] cluster (three [2Fe-2S] clusters are designated as N1a, N1b, and N1c), although the exact location of the third [2Fe-2S] cluster is not known (5,46). Thus, it is of interest to investigate if another [2Fe-2S] cluster is located in the TthNQO3 subunit and, if so, what role it plays in electron transfer. In order to address these issues, expression of the TthNQO3 subunit is in progress in our laboratory.
TthNQO4 and TthNQO5 Subunits-The TthNQO4 and Tth-NQO5 subunits are homologous to the bovine IP 49-kDa, Pd-NQO4, and EcNuoC subunits and to bovine IP 30-kDa, Pd-NQO5, and EcNuoB subunits, respectively (Table II) lus subunits. The TthNQO4 subunit contains only one nonconserved cysteine residue (Cys 385 ), whereas the TthNQO5 subunit does not contain cysteine residues. The functional roles of these subunits are not known.
TthNQO6 Subunit-Among the 14 subunits, the TthNQO6 subunit showed the highest sequence identities to its bovine, P. denitrificans, and E. coli equivalents (Fig. 4A and Table II). These subunits are homologous to the HycG subunit of E. coli formate hydrogen lyase. This subunit contains conserved cysteine residues that are also present in TthNQO6 subunit (Cys 45 , Cys 46 , Cys 111 , and Cys 140 ). The cysteine arrangement in the NQO6 subunit and its homologues has been suggested to be analogous to that in the small subunit of nickel-iron hydrogenases (48,49). In the latter enzymes, involvement of the corresponding cysteines in ligating a [4Fe-4S] cluster has recently been revealed by the three-dimensional crystal structure of Desulfovibrio gigas hydrogenase (50). It remains to be determined whether the NQO6 subunit contains an iron-sulfur cluster.
TthNQO9 Subunit-The TthNQO9 subunit is homologous to the bovine TYKY, PdNQO9, and EcNuoI subunits ( Fig. 4B and  (51,52). Our preliminary results from an expression and reconstitution study of the PdNQO9 subunit have indicated that NQO9 indeed contains two [4Fe-4S] clusters. 4 Iron-sulfur cluster N2, which probably plays a key role in the electron transfer pathway as a direct electron donor to quinone, has not yet been assigned to a subunit. The hypothetical cluster in NQO6 and either of those in NQO9 are possible candidates. ND2 gene products and to the other seven subunits of bacterial NDH-1 (Table II). These subunits are extremely hydrophobic and are thought to constitute the integral membrane part of the enzyme complex. The T. thermophilus subunits contain a number of hydrophobic stretches that are predicted to span the cytoplasmic membrane (Fig. 5). As has been suggested previously (1,2,5,53,54), the membrane assembly probably plays an essential role in quinone reduction and proton translocation. The T. thermophilus membrane subunits conserve several charged amino acid residues, some of which may be involved in the energy-tranducing mechanism. It is very difficult, however, to discuss specific functional roles of the individual subunits based only on the sequence data. Little is known about the structures of these subunits, and quinone-binding site(s) have not been identified. Thus, much future work is needed before any conclusions can be made.
Expression of the T. thermophilus NQO2 (TthNQO2) Subunit in E. coli-As mentioned previously, the TthNQO2 subunit contains the four conserved cysteine residues that were shown to ligate a [2Fe-2S] cluster in PdNQO2. The PdNQO2 subunit is one of the well characterized subunits in NDH-1 (26,42,55) and has similar physicochemical properties to Clostridium pasteurianum binuclear ferredoxin (56). Therefore, to explore the thermostable properties, we attempted to express the Tth-NQO2 subunit in E. coli and purify it. Its physicochemical properties could be directly compared with those of the Pd-NQO2 subunit. The TthNQO2 subunit was expressed as a soluble protein and could be purified with nickel chelation column chromatography (Fig. 6A). We raised polyclonal antibodies against the expressed TthNQO2 subunit and utilized them to examine the identity of the subunit with the T. thermophilus membrane preparation. The antibodies recognized a single polypeptide with an apparent molecular weight of M r ϭ 21,000 (Fig. 6B). These results indicate that the sequences were precisely analyzed and that the subunit was correctly expressed in E. coli. The purified TthNQO2 subunit was reddish-brown in color as previously observed with the PdNQO2 subunit. Its absorption spectrum showed great resemblance to that of the PdNQO2 subunit. Characteristic absorption peaks of the iron-sulfur cluster (330, 420, 460, and 550 nm) could be seen in its oxidized form (Fig. 7A). EPR measurements provided further evidence that the TthNQO2 subunit contains a single [2Fe-2S] cluster with similar properties to the PdNQO2 subunit (Fig. 7B).
Thermostability of the TthNQO2 Subunit-It has been reported that the isolated T. thermophilus NDH-1 enzyme complex exhibits extremely thermostable properties and is capable of catalyzing NADH oxidation at temperatures over 65°C. In a preliminary experiment, the thermostability of the expressed TthNQO2 subunit was tested by incubating crude preparations (cytoplasmic fraction) at 65°C for various time under anaerobic conditions. The solutions were centrifuged to remove denatured proteins, and the resultant supernatants were subjected to SDS-PAGE. As shown in Fig. 8A, while E. coli proteins were readily denatured by the heat treatment, the TthNQO2 subunit remained soluble in the supernatant. These results show that the TthNQO2 subunit is more heat-resistant than E. coli proteins. After subsequent purification by nickel chelation column chromatography, the subunit exhibited the characteristic red-brown color, indicating that the [2Fe-2S] cluster also survived the treatment (data not shown). The TthNQO2 subunit was further used in order to assess the degree of thermostability. The purified TthNQO2 subunit (by Protocol 1) was incubated at various temperatures (50 -90°C) for 30 min under anaerobic conditions followed by centrifugation at 14,000 rpm for 10 min. Absorption spectra of the supernatants were measured in the visible region. Almost no precipitation was observed up to 65°C, and the absorption spectra of the subunits were unaltered (Fig. 8B, line 1). Notably, no spectral changes occurred at 65°C during incubation up to 3 h (data not shown). When the temperature was elevated to 75-80°C, the subunit started to show some instability such that some portion of the protein was denatured and precipitated (Fig. 8B, lines 2 and 3). At 85°C and above, the denaturing process was further accelerated. The absorption spectrum of the soluble subunits was significantly modified (Fig. 8B, line 4). On the other hand, when the PdNQO2 subunit was heat-treated in the same way, the protein was immediately denatured even at 30°C (not shown). These results clearly demonstrate that the T. thermophilus NDH-1 subunits are extremely stable and thus more suitable to express and study as single proteins than are their mesophilic equivalents. It should be noted that the iron-sulfur cluster in the TthNQO2 subunit was stable only under oxygenfree conditions. In the presence of oxygen, the subunit readily lost its reddish-brown color, indicating that the iron-sulfur cluster is vulnerable to attack by oxygen. This is often seen also in other iron-sulfur proteins (57). DISCUSSION The present study has revealed that the T. thermophilus NDH-1 (eukaryotic Complex I equivalent) genes constitute an operon-like structure that is composed of 14 structural genes (NQO1- 14) with no URFs. The T. thermophilus NDH-1 is composed of seven predicted transmembranous subunits (NQO7, NQO8, and NQO10 -14) and seven peripherally located subunits (NQO1-6 and NQO9), all of which are very homologous to the corresponding subunits of mitochondrial Complex I and other bacterial NDH-1. The order of the 14 structural genes is identical to those of other NDH-1 gene clusters (P. denitrificans and E. coli) and similar to those encoding other related enzymes, suggesting a phylogenetic relationship between bacterial NDH-1, NAD-reducing hydrogenase, and some related enzymes. These results further substantiate that the 14-subunit structure is the minimal and ubiquitous functional unit of Complex I and NDH-1.
Previous biochemical studies of the isolated T. thermophilus NDH-1 have shown that the enzyme contains one FMN and several iron-sulfur clusters as redox components (19). The T. thermophilus HB-8 membrane protein were loaded on a 15% SDS-polyacrylamide gel. After electrophoresis the proteins were transferred onto a nitrocellulose membrane. Immunoblotting was carried out using antibodies against the expressed TthNQO2 subunit and horseradish peroxidase-conjugated anti-rabbit IgG antibody as described in Refs. 30 -32, except that the detection was carried out using the ECL kit (Amersham). The molecular sizes are indicated on the left. thermophilus NDH-1 contains only 40 cysteine residues (while the P. denitrificans NDH-1 contains 64 cysteine residues), all of which are located in the extrinsic subunits (Table III). Thirty-two out of 40 cysteine residues are fully conserved among mitochondrial and bacterial Complex I whose amino acid sequences have been reported. As described under "Results," the T. thermophilus NDH-1 has the potential of bearing eight ironsulfur clusters in common with mitochondrial and bacterial enzyme complexes and probably an additional [2Fe-2S] cluster in the TthNQO3 subunit, which is unique to T. thermophilus, E. coli, and S. typhimurium. It has been shown by EPR studies that at least three NADH-reducible iron-sulfur clusters (one binuclear and two tetranuclear) and possibly two iron-sulfur clusters (one binuclear and one tetranuclear) with very low redox midpoint potentials are present in the membrane-bound T. thermophilus NDH-1 (20). Although exact assignment of these EPR-detectable iron-sulfur clusters remains to be made, these results together with the sequence information in the present study indicate that the T. thermophilus NDH-1 possesses at least as many iron-sulfur clusters as its mitochondrial and bacterial equivalents.
However, T. thermophilus NDH-1 seems to be unique among the known Site I enzyme complexes from a bioenergetic point of view, because T. thermophilus contains only menaquinone-8 (58). Menaquinone-8 has a lower redox midpoint potential than ubiquinone (E m,7 ϭ Ϫ75 mV for menaquinone/menaquinol couple; E m,7 ϭ ϩ100 mV for ubiquinone/ubiquinol couple). Thus, the free energy available from electron transfer from NADH to menaquinone is much less than when using ubiquinone. Therefore, comparative studies of T. thermophilus NDH-1 with ubiquinone-utilizing complexes may provide new insights into Site I energy coupling mechanism.
Thermostability of proteins isolated from not only T. thermophilus but also from several other thermophilic microorganisms has recently received considerable attention. Several attempts have been made to clarify the origin of the thermostability, and some principles have been proposed (59 -61). Based on the comparative structural studies of thermophilic and mesophilic proteins, it seems likely that structural stability of thermophilic proteins are reinforced by an increment of intrapolypeptide interactions throughout the protein molecule (62)(63)(64). The present study has shown that the expressed TthNQO2 subunit is thermostable, since the subunit and its iron-sulfur cluster survived at 65°C for 3 h. The same observation has been made for other T. thermophilus NDH-1 subunits and subcomplexes expressed in E. coli. 5 These results indicate that heat resistance resides in individual subunits. The purified NQO2 subunit and other expressed subunits remained stable at room temperature (22-25°C) for weeks in an anaerobic chamber without loosing the iron-sulfur cluster. The availability of these stable subunits give us the possibility to 5 T. Yano and T. Yagi, unpublished data.
FIG. 8. Thermostability measurement of the expressed Tth-NQO2 subunit. Panel A, the cytoplasmic fractions (ϳ20 -25 mg/ml) prepared from E. coli harboring pET16b(TthNQO2) in 50 mM Tris-HCl buffer (pH 7.4) containing 300 mM NaCl, 1.0 mM DTT, 1.0 mM PMSF, and 20 g/ml leupeptin in 1.5-ml microtubes were incubated at 65°C for various times (0, 5, 10, 15, 30, and 60 min, lanes 1-6, respectively) under an anaerobic condition and then chilled on ice for 10 min. The preparations were centrifuged at 14,000 rpm for 15 min to remove denatured proteins, and the transparent supernatants were transferred into new microtubes. The supernatants were mixed with an equal volume of SDS-PAGE sample buffer that contained 160 mM Tris-HCl (pH 6.8) and 12% (w/v) SDS, boiled for 10 min, and then subjected to 14% (w/v) SDS-polyacrylamide gel electrophoresis. Lane 7 represents a purified TthNQO2 subunit (1.0 g). Panel B, 700 l of the purified TthNQO2 subunit (0.25 mg/ml) in microtubes were incubated for 30 min at various temperatures (50 -90°C) and then chilled on ice for 10 min. After the solutions were centrifuged at 14,000 rpm for 15 min, the absorption spectra of the supernatants were recorded at room temperature. Lines 1, 2, 3, and 4 represent the absorption spectra of the subunit treated at 65, 75, 80, and 85°C, respectively. investigate subunit-subunit interaction by reconstituting the entire enzyme complex from individual components as demonstrated for ATPase (65). Furthermore, significant stability is also a great advantage for structural studies of the complex using x-ray crystallography or NMR spectroscopy.