Functional Roles of Four Conserved Charged Residues in the Membrane Domain Subunit NuoA of the Proton-translocating NADH-Quinone Oxidoreductase from Escherichia coli

The H (Na (cid:1) )-translocating NADH-quinone (Q) oxidoreductase (NDH-1) of Escherichia coli is composed of 13 different subunits (NuoA-N). Subunit NuoA (ND3, Nqo7) is one of the seven membrane domain subunits that are considered to be involved in H (cid:1) (Na (cid:1) ) translocation. We demonstrated that in the Paracoccus denitrificans NDH-1 subunit, Nqo7 (ND3) directly interacts with peripheral subunits Nqo6 (PSST) and Nqo4 (49 kDa) by using cross-linkers (Di Bernardo, S., and Yagi, T. (2001) FEBS Lett . 508, 385–388 and Kao, M.-C., Matsuno-Yagi, A., and Yagi, T. (2004) Biochemistry 43, 3750–3755). To investigate the structural and functional roles of conserved charged amino acid residues, a nuoA knock-out mutant and site-specific mutants K46A, E51A, D79N, D79A, E81Q, E81A, and D79N/E81Q were constructed by utilizing chromosomal DNA manipulation. In terms of immunochemical

dria, is a multiple subunit enzyme complex embedded in the cytoplasmic membrane (1). This enzyme represents the first step of the respiratory chain and links the electron transfer from NADH to quinone with the translocation of protons from the cytoplasmic phase to the periplasmic phase (1). The stoichiometry of H ϩ /2e Ϫ is considered to be 4 (2). The resulting membrane potential is utilized to drive energy required for processes like ATP synthesis or solute transport (3). Although mammalian mitochondrial complex I is composed of 46 unlike subunits (4), bacterial counterparts contain 14 different subunits (designated Nqo1-14 for Paracoccus denitrificans and Thermus thermophilus and NuoA-N for Escherichia coli) 2 (5,6). The bacterial NDH-1 contains cofactors (one FMN and 8 -9 iron-sulfur clusters) akin to complex I (7,8). Topological studies suggest that the NDH-1 can be divided into two sectors, the peripheral segment and the membrane segment (9). The peripheral segment is composed of 7 subunits (NuoB, -C, -D, -E, -F, -G, and -I). In the case of the E. coli NDH-1, subunits NuoC (30 kDa) and -D (49 kDa) are fused and form NuoCD. Among these peripheral subunits, the NuoB (PSST) and NuoI (TYKY) subunits are recognized to act as connector subunits between the peripheral and membrane segments (9,10). The membrane segment also consists of seven subunits (NuoA, -H, and -J-N) (11)(12)(13), which are homologues of mtDNA-encoded subunits (ND1-6 and 4L) (14,15). The peripheral segment protrudes into the cytoplasmic phase and is believed to house all the known cofactors (1). In contrast, the membrane domain is most likely involved in H ϩ (or Na ϩ ) translocation and inhibitor and quinone binding (16 -19). In recent years, complex I, in particular the mitochondrially encoded subunits, received much attention because of their involvement in many mitochondrial diseases (including sporadic Parkinson's disease) (20,21). It is known that point mutations of the ND1/nuoH, ND4/nuoM and ND6/ nuoJ genes are associated with Leber's hereditary optic neuropathy and suppress the respiratory chain activity of complex I (22,23). Furthermore, it has been reported that the defects of the ND5/NuoL subunit are involved in mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes syndrome and other encephalomyelopathies (24). A single point mutation (T10191C, this mutation substitutes Pro for Ser-45, human numbering) of the ND3/NuoA subunit has been reported to * This work was supported by United States Public Health Service Grants R01GM33712 (to A. M.-Y. and T. Y.) and R01GM55594 (to M. P.). Synthesis of oligonucleotides and DNA sequencing were supported in part by the Sam & Rose Stein Endowment Fund. This is publication 16127-MEM from The Scripps Research Institute, La Jolla, CA. 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.
§ Both authors contributed equally to this work. ʈ To whom correspondence may be addressed. E-mail: yagi@scripps. edu.
In a previous paper (11), we determined the topology of the Paracoccus NuoA subunit. The Paracoccus NuoA subunit is composed of three transmembrane segments (designated TM1-3 from the N to the C terminus), and its N-and Cterminal regions are directed toward the cytoplasmic and periplasmic phases of the membrane, respectively (11). The predicted topology places two highly conserved carboxyl residues (Asp-79 and Glu-81, E. coli numbering) in the middle of the TM2 (11). More recently, our cross-linking study revealed direct interactions between subunits NuoA and NuoB and between subunits NuoA and NuoD (9,27). The NuoB subunit is considered to bear the center N2, which shows the highest midpoint redox potential values of all known cofactors in the NDH-1 (1, 6). Therefore, it was of interest to clarify structural and functional roles of these conserved and protonated residues in the NuoA subunit. For this purpose, we have constructed mutants of the residues of interest by using a gene manipulation technique of the E. coli chromosomal NDH-1 operon and characterized these mutants. Mutants E51A, D79A, D79N, E81A, and E81Q showed a partial decrease in the activities of deamino-NADH oxidase and deamino-NADH-dimethoxy-5methyl-6-decyl-1,4-benzoquinone (DB) reductase but retained the deamino-NADH-K 3 Fe(CN) 6 reductase activity comparable with the wild type. In addition, whereas the D79N/E81Q mutant is similar to the wild type in terms of both NADH dehydrogenase activity staining and immunochemical analyses of native gels, the energy-transducing NDH-1 activities of this double mutant were almost completely inactivated.

EXPERIMENTAL PROCEDURES
Materials-The pCRScript Cloning kit was from Stratagene. The gene replacement vector, pKO3 was a generous gift from Dr. George M. Church (Harvard Medical School, Boston, MA). Materials for PCR product purification, gel extraction, and plasmid preparation were obtained from Qiagen (Valencia, CA). Site-specific mutants were constructed using the GeneEditor mutagenesis kit from Promega (Madison, WI). The BCA protein assay kit and SuperSignal West Pico chemiluminescent substrate were from Pierce. NADH, deamino-NADH, DB, chloramphenicol, and spectinomycin (Spc) were from Sigma. p-Nitroblue tetrazolium was from CalBiochem. Capsaicin 40 (cap-40) and pET(EcoNuoE) bearing the E. coli nuoE gene were kind gifts from Dr. Hideto Miyoshi (Kyoto University, Kyoto, Japan) and Dr. Judy Hirst (MRC, Cambridge, United Kingdom), respectively.
Cloning and Mutagenesis of the E. coli nuoA Gene-The gene encoding the NuoA subunit together with a 1-kb DNA segment upstream and a 1-kb DNA segment downstream were cloned by PCR technology from E. coli DH5␣. To generate the restriction sites SmaI/NotI and NotI/SalI the sense/antisense primers 5Ј-GGTACGCCCGGGAAATCCTGCGTTT-TAATGATGAGG-3Ј with 5Ј-ACCTCGCGCGGCCGCGACCGCCTAAA-AACCGCC-3Ј and 5Ј-TATCTGGCGGCCGCGTTCTTCGTTATCTTCG-ACGTTG-3Ј with 5Ј-GTGTGCGTCGACGTTCGTCCATGCCGTGTAA-GTC-3Ј, respectively, were used, where the underlined bases were altered from E. coli DNA, and the italicized bases represented the restriction site sequence. The spectinomycin-encoding gene from transposon Tn554 of Staphylococcus aureus (28) was cloned by the PCR technology using the sense primer 5Ј-CGGGGGCGGCCGCTCAGTGG-AACGAAAACTCACG-3Ј and the antisense primer 5Ј-AAGGAGCGGC-CGCTTTCTATTTTCAATAGTTAC-3Ј both containing a NotI restriction site represented by italicized bases. The DNA fragments and the Spc cassette were cloned in pCRScript and finally assembled in pKO3. In the same way the sense primer 5Ј-GCATTCAAGATCTTGGTTACG-CCAGGAAAATCC-3Ј, which contains a BglII restriction site (italicized), was used together with the NotI-generating antisense primer to produce a DNA fragment that was cloned in pCRScript. Then the sense primer 5Ј-CCATGAATCGATGTGGCGTCC-3Ј, which contains a ClaI restriction site (italicized) was used together with the SalI-generating antisense primer to produce a DNA fragment used for the generation of nuoA point mutants. The DNA inserted in the pCRscript cloning plasmid was mutagenized with the mutagenesis primers shown in Table I. These fragments were also assembled in pCRScript and then cloned into pKO3.
The first step of site-specific mutation of the nuoA gene of the E. coli NDH-1 operon was to construct a nuoA gene knock-out mutant. For this purpose, we employed the pKO3 system developed by Church and co-workers (29). In brief, the pKO3 vector contains a repA(Ts) (temperature-sensitive replication origin), a chloramphenicol-resistant gene (cat), and a Bacillus subtilis sacB gene encoding levansucrase. The pKO3 carrying nuoA-knock-out DNA was prepared as follows (see Fig.  1, a and b). DNA fragments, SmaI/NotI (1467 bp) and NotI/SalI (1269 bp), were amplified from E. coli chromosomal DNA by PCR and individually inserted in cloning vector pCRScript at the SrfI site as a blunt end fragment. A PCR-amplified spc cassette carrying NotI sites (1200 bp) was also inserted at the SrfI site of pCRScript. The two DNA fragments and the spc cassette were assembled in pKO3. The resulting plasmid, pKO3(nuoA::spc), lacks 90 bp of the nuoA gene, which have been replaced by the spc cassette.
The pKO3 vectors carrying mutated nuoA genes were prepared as shown in Fig. 1, c-e. The R and L fragments were cloned as blunt end fragments at the SrfI site in pCRScript. First, the R fragments of 1544 bp in which a SalI site was introduced at the 3Ј-end by the PCR amplification were cloned in pCRScript (designated pCRScript-R) (see Fig. 1, c). The pCRScript-R was used to generate the nuoA mutants. Then the L fragment of 1015 bp containing BglII and NotI sites at the 5Ј-and 3Ј-ends, respectively, was also cloned in pCRScript generating plasmid PCRScript-L (see Fig. 1, d). This plasmid was digested with HindIII/XhoI, blunted, and religated to remove a ClaI site in the multiple cloning site. The R fragments containing the mutations were isolated by ClaI and NotI (the latter is present in the multiple cloning site) and purified. The ClaI/NotI fragments were inserted into ClaI/ NotI-cleaved pCRScript-L. The resulting plasmids were designated pCRScript-nuoA(mutants). Each DNA fragment containing nuoA mutations were then isolated by BglII/SalI digestion from the pCRScript constructs and transferred to integration plasmid pKO3 at the BamHI/ SalI sites. The resulting plasmids are referred to as pKO3-nuoA (mutants) (see Fig. 1, e).
Preparation of Knock-out and Mutant Cells-E. coli strain MC4100 (F Ϫ , araD139, ⌬(arg F-lac)U169, ptsF25, relA1, flb5301, rpsL 150. Ϫ ) was transformed with pKO3 (nuoA::Spc) plasmid, and recombination was carried out as described in Link et al. (29). In brief, several well isolated colonies from LB agar plates containing 20 g/ml chloramphenicol and 100 g/ml Spc, grown overnight at 30°C, were transferred into 100 l of LB and serially diluted. The dilutions corresponding to 10 4 -10 6 were then plated on LB agar plates containing 20 g/ml chloramphenicol and 100 g/ml Spc, prewarmed at 43°C and grown overnight. The next day again several colonies (typically 5) were transferred from the 43°C plates into 100 l of LB, serially diluted, and plated on LB agar plates containing 5% sucrose at 30°C overnight. The surviving colonies were then replica-plated on LB agar plates containing 20 g/ml chloramphenicol and on LB plates containing 100 g/ml Spc and grown at 30°C overnight. Colonies sensitive to chloramphenicol but resistant to Spc were used for the PCR amplification of the nuoA region using the 5Ј-oligonucleotide CTGAACATGGCATTCAAC (chro5Ј) and the 3Ј-oligonucleotide AAGGAGCGGCGGCTTTCTATTTTCAATAGTTAC (spc3Ј). The chro5Ј oligonucleotide was designed to amplify DNA from within a Underline indicates mutation.
Roles of Glu-81 and Asp-79 of NuoA Subunit in E. coli NDH-1 the E. coli chromosome and the spc3Ј oligonucleotide from within the Spc cassette. In this way the presence of the Spc cassette and its location in the genomic DNA was confirmed. The knocked-out MC4100 cells where then stored as glycerol stocks at Ϫ80°C. Knocked-out MC4100 competent cells were then employed to introduce nuoA mutated DNA in the E. coli genome using a similar procedure except that the identification of recombinants was carried out by screening for spectinomycin sensitivity in addition to chloramphenicol sensitivity.
To confirm the presence of the mutations, the sense oligonucleotide chro5Ј and the antisense oligonucleotide CATACGCTCGCGGCGTG (nuoA3Ј), which is located inside the nuoA gene, were used as primers for PCR amplification of the nuoA DNA fragments. The nuoA DNA fragments produced were subjected to direct sequencing.
Antibody Production-Antibodies directed against a 12-amino-acid oligopeptide corresponding to the C-terminal region of the E. coli NuoA subunit was produced as follows. An oligopeptide H-CNPETNSIAN-RQR-OH was synthesized (designated NuoAc) and conjugated to maleimide-activated bovine serum albumin (Pierce) according to the manufacturer's protocol. It should be noted that, for the purpose of conjugation with bovine serum albumin, a cysteine residue was added to the N terminus. For raising antibodies specific to subunits NuoB, NuoE, NuoF, NuoG, and NuoI inclusion bodies of the overexpressed subunits were used as described previously (7). The antibodies were affinity-purified according to Han et al. (30).
Cell Growth and Membrane Preparation-For the preparation of membranes suitable for enzymatic assays wild type, knock-out, and point mutants were grown in 250 ml of Terrific Broth medium until A 600 was ϳ2. The cells were then harvested in a GSA rotor at 6000 rpm for 10 min. The cell pellet was resuspended at 10% (w/v) in a buffer containing 10 mM Tris-HCl (pH 7.0), 1 mM EDTA, 1 mM dithiothreitol, 1 mM phenylmethanesulfonyl fluoride, and 15% (w/v) glycerol. The cell suspension was then passed once in a French press at 25,000 p.s.i. and centrifuged again in the GSA rotor at 12,000 rpm for 10 min. Cell debris was discarded, and the supernatant was then ultracentrifuged in a 70Ti rotor at 50,000 rpm for 30 min. The pellet was resuspended in the same buffer and was used immediately for enzymatic activity measurements.
Gel Electrophoresis and Western Blot Analysis-To confirm the expression of the NDH-1 subunits Western blot experiments were carried out. Antibodies against the NuoAc and the peripheral subunits NuoE, NuoF, NuoG, and NuoI reacted with a 16-, a 20-, a 50-, a 91-, and a 21-kDa band of the E. coli membranes, respectively.
Membranes were subjected to blue-native PAGE according to Schagger (31). Briefly, the cholate-treated E. coli membranes (800 g of protein) were prepared as described previously (7) and resuspended in 40 l of 750 mM aminocaproic acid, 50 mM Bistris-HCl (pH 7.0). Then, 8 l of 10% dodecylmaltoside and 50 g/ml DNase were added, and the preparation was left on ice for 1 h. After the incubation on ice the samples were centrifuged at 149,000 ϫ g in a Beckman Airfuge for 5 min. The supernatant was recovered (ϳ40 l), and 16 l of 5% Coomassie Blue in 500 mM aminocaproic acid was added to the samples. The samples were then loaded on a 7% gel and run in the cold room at 75 V until the dye entered the separating gel. Subsequently the voltage was raised to 200 V, and the gel was run for another 3 h. After completion of the electrophoresis the gel was incubated in 2 mM Tris-HCl (pH 7.5) containing 150 M NADH and 2.5 mg/ml p-nitroblue tetrazolium at 37°C for 2 h in a shaking incubator. The reaction was stopped with 7% acetic acid.
Other Analytical Procedures-Protein concentrations were estimated by the BCA protein assay kit with bovine serum albumin as the standard according to the manufacture's instruction. Any variations from the procedures and details are described in the figure legends.

Strategy for Constructions of NuoA Mutants-The E. coli
NDH-1 operon is predicted to be ϳ15-kb long (35,36). Because this length does not allow incorporation of the whole operon into expression vectors, a site-specific mutation is traditionally carried out by complementation of a cassette-inserted gene with a mutated gene in the expression plasmid (designated in trans complementation) (37)(38)(39)(40)(41)(42)(43). However, the in trans complementation procedure presents some problems when applied to a gene cluster. For example, the cassette inserted in chromosomal DNA might interrupt the expression of the downstream genes (44). The only case that does not suffer this polar effect is when the mutated gene is at the last position in the operon (38). Another problem is that the mutated gene is under the control of a promoter in the expression plasmid, which often leads to overexpression of the mutated subunit. In mutation studies of the NDH-1 using the in trans complementation, it has been reported that the enzyme activities were significantly low (ϳ20% of the wild type cells) even when the unmutated gene was used (42,43,45). An alternative method of site-directed mutation is to introduce mutations directly in chromosomal DNA as detailed in this work (designated chromosomal DNA mutation, see Fig. 1) (29,42). In this procedure, expression of all genes of the operon are regulated by the authentic promoter. Although chromosomal DNA mutation is laborious and time-consuming, we adopted this technique to produce NuoA mutants to minimize any complications derived from disruption of the operon. As anticipated, the NDH1 was apparently expressed at the same level in all mutants as in the wild type (see below).
Sequence Analysis of the NuoA Subunit- Fig. 2A is an amino acid sequence comparison between the E. coli NuoA subunit and its counterparts of various organisms. In terms of hydropathy plots, the E. coli NuoA subunit is akin to its counterpart of P. denitrificans. Fig. 2B is a hypothetical topology of the E. coli NuoA subunit deduced from topological studies of the P. denitrificans NuoA subunit (11). The E. coli NuoA subunit is predicted to contain the three transmembrane segments (designated TM1-3 from N to C terminus). The N-and C-terminal regions are also predicted to be directed toward the cytoplasmic and periplasmic phases of the membrane, respectively. In addition, a long loop (L1) between TM1 and TM2 is exposed to the periplasmic side. As far as our data base search is concerned (more than 250 organisms), Asp-79 (E. coli numbering) is conserved except for its homologues of Cyanidium caldarium mitochondria (Asp-79 3 C, CAA88774) and Pseudomonas aeruginosa (Asp-79 3 G, D83410). On the other hand, Glu-81 is perfectly conserved. Asp-79 and Glu-81 (E. coli numbering) seem to be located in the middle of the TM2. Carboxyl residues are rarely located in the middle of TM of the hydrophobic polypeptides. Therefore, it has been generally recognized that FIG. 1. a-e, schematic representation of  carboxyl residues present in the TM may play important roles in cation translocation of the membrane-associated enzyme complexes (46,47). One well known example is a perfectly conserved carboxyl residue in the center of a transmembrane helix of the N,NЈ-dicyclohexylcarbodiimide (DCCD)-binding protein (also called subunit c or proteolipid subunit) of the ATP synthase (48). This carboxyl residue is clearly involved in the mechanism of proton translocation catalyzed by the membrane sector of the ATP synthase (49). It has been demonstrated that DCCD also inhibits energy-transducing electron transfer of the NDH-1/complex I (50,51). It is therefore possible that the Asp-79 and/or Glu-81 may be involved in the proton translocation of the energy coupling site 1. To clarify the structural and functional roles of these carboxyl residues, we have constructed site-directed mutants of these two residues (D79A, D79N, E81A, E81Q, D79N/E81Q) by using chromosomal DNA mutation procedures instead of the in trans complementation method. In addition, it has been reported that a heteroplasmic mutation of S34P and S45P (human numbering) in the human ND3 subunit (a homologue of the NuoA) drastically reduced the complex I activity and caused encephalopathy (25,26). These Ser residues appear to be located in the Loop 1 directed against the cytoplasmic phase and periplasmic phase in eukaryotes and bacteria, respectively (see Fig. 2A). Unfortunately, neither Ser-34 nor Ser-45 is conserved in the bacterial NDH-1. Therefore, we constructed mutations in charged residues that are well conserved (K46A, E51A) to assess whether this loop is involved in function of the NDH-1/complex I. Seven site-specific mutants (nuoA-K46A, -E51A, -D79A, -D79N, -E81A, -E81Q, -D79N/E81Q) were generated. The mutations were confirmed by direct DNA sequencing analyses.
Subunit Assembly of NDH-1 in NuoA Mutants- Fig. 3 illustrates Western blot analyses of the membranes isolated from wild type and the NuoA null mutant with affinity-purified antibodies to the NuoAc and peripheral subunits NuoE, NuoF, NuoG, and NuoI of the E. coli NDH-1. The antibody to the NuoAc reacted with the wild type but did not react with nuoA::spc mutant (knock-out mutant). In addition, mutants   FIG. 2. A, comparison of the deduced amino acid sequence of the E. coli NuoA subunit with its counterparts from various organisms. The comparison was conducted with the PILEUP programs of GCG software (55). Pd, P. denitrificans Nqo7 subunit (M93015); Rc, Rhodobacter capsulatus NuoA subunit (AF029365); Tt, T. thermophilus HB-8 Nqo7 subunit (U52917); Hs, human mitochondrial ND3 subunit (NP_776058); Ec, E. coli K-12 NuoA subunit (U00096/b2288-b2276). Amino acid residues mutated in this study are marked by asterisks. Three predicted transmembrane segments (TM1-3) are surrounded by blocks according to Di Bernardo et al. (11). Underlined segment marked NuoAc indicates the oligopeptide region used to raise the antibody specific to the C-terminal region of the E. coli NuoA. It should be noted that the E. coli NuoA subunit has a significantly longer C-terminal stretch than other organisms. The positions of the Ser whose mutations to Pro are involved in human mitochondrial diseases are illustrated by arrowheads. B, proposed topology of the E. coli NuoA subunit. The prediction has been performed on the basis of the reported topology of its Paracoccus counterpart (11). As described under "Results," the three transmembrane segments of the E. coli NuoA subunit from the N terminus to the C terminus are tentatively designated TM1, TM2, and TM3, respectively. The loops between the TM1 and TM2 and between TM2 and TM3 are designated L1 and L2, respectively. The N terminus and C terminus of the subunit are exposed to the cytoplasmic side and the periplasmic side of the membrane, respectively. The mutated amino acid residues are displayed by squares.
K46A, E51A, D79A, D79N, E81A, E81Q, and D79N/E81Q were recognized by the NuoAc antibody. In contrast, membranes isolated from the wild type and all available site-specific NuoA mutants seem to bear similar amounts of peripheral subunits NuoE, NuoF, NuoG, and NuoI. Subunits NuoE and NuoF (the NADH-binding subunit) are known to be essential for deamino-NADH-K 3 Fe(CN) 6 reductase activity (52). These results suggest that site-specific NuoA mutants apparently remain intact in subunit assembly in all NuoA mutants examined. To further confirm this point, isolated membranes were treated with dodecylmaltoside and subjected to blue native polyacrylamide gel electrophoresis. Then, the gels were stained for NADH dehydrogenase activity using NADH and p-nitroblue tetrazolium (Fig. 4). Two bands appeared in the wild type membranes. The upper band, but not the lower band, was recognized by the antibody to the peripheral subunit NuoE. In addition, the antibody specific to NuoAc reacted with the upper band (data not shown). Furthermore, the membranes isolated from the knockout mutant lacked the NADH dehydrogenase activity and reactivity with the NuoE antibody in the upper band position. The data indicate that the NADH dehydrogenase activity of upper band is because of NDH-1. As shown in Fig. 4, all of the site-specific nuoA mutants showed a comparable NADH dehydrogenase activity band because of the NDH-1. In addition, the site-specific NuoA mutants are similar to the wild type in terms of relative molecular size of the NDH-1 bands. It seems likely that constructed site-specific mutants are similar to wild type in terms of subunit assembly.
Effects of NuoA Mutation on the NDH-1 Activity-We measured activities of NDH-1 using membranes prepared from wild type and NuoA mutants (Table II). E. coli membranes contain a second type of NADH dehydrogenases (NDH-2). To eliminate contribution from the NDH-2, deamino-NADH was used as the substrate in all assays, because NDH-2 cannot utilize this compound (32). First, it should be noted that deamino-NADH-K 3 Fe(CN) 6 reductase activity of all site-specific mutants was comparable with that of the wild type, whereas the activity of the knock-out mutant was almost null. These results are consistent with the data from the NADH dehydrogenase activity staining of the native gels, suggesting that none of the mutations affected the assembly of the NDH-1 subunits. Second, deamino-NADH oxidase activity and deamino-NADH-DB reductase activity behaved in a similar fashion among the mutants tested, indicating that the inhibitory effect observed was solely because of the NDH-1 mutation.
Single-residue mutations introduced into the Loop 1 (Lys-46 and Glu-51) or the middle of the TM2 (Asp-79 and Glu-81) resulted in either no inhibition or partial (up to 70%) inactivation. However, when the two carboxyl residues in the TM2 were mutated simultaneously (D79N/E81Q), NDH-1 activities were almost completely abolished (see also Fig. 5). It was, therefore, of particular interest to examine the sensitivity of the NDH-1 activity of these mutants to DCCD. As anticipated, mutations in the Loop 1 region (Lys-46 and Glu-51) showed the same degree of DCCD inhibition as the wild type. Of the two mutants in the TM2, Asp-79 had the same DCCD sensitivity as the wild type. In contrast, the Glu-81 mutants were less sensitive to DCCD treatment than other mutants and the wild type. It remains to be seen whether Glu-81 is one of the target sites of DCCD binding.
Miyoshi and co-workers (34) reported that cap-40 acts as a competitive inhibitor for quinone in the NDH-1/complex I and suppresses only the energy-coupled activity. We found that all NuoA mutants described above were almost completely inhibited by cap-40 (data not shown). Furthermore, the I 50 values of cap-40 for the mutants were about the same as that of the wild type (Table II), suggesting that the quinone-binding site is not modified by these mutations. DISCUSSION The membrane domain of the NDH-1 is composed of seven subunits, which are homologues of mitochondrially encoded ND subunits. This domain is involved in H ϩ (or Na ϩ ) translocation in the coupling site 1 (18) but lacks any cofactors. In addition, DCCD, known to specifically modify carboxyl residues located in the hydrophobic environment, inhibits the energycoupled activities of the NDH-1/complex I. It is therefore speculated that conserved carboxyl residues located in the middle of membranes might participate in cation translocation in the coupling site 1. On the basis of the deduced primary sequence analysis, we predicted that there are eight highly conserved FIG. 3. Immunoblotting of membrane preparations from the wild type (W), nuoA knock-out (KO), and site-specific nuoA mutants by using antibodies specific to the NuoAc, NuoE, NuoF, NuoG, and NuoI. The membranes (10 g of protein/lane) were loaded on a 13% Laemmli SDS-polyacrylamide gel. For immunoblotting for subunit NuoA, the membranes were suspended in the Laemmli's sample buffer with an additional 4 M urea, and the samples were incubated on the boiling water for 10 min before loading on the SDS gels. If these harsh conditions were not used, the NuoA(D79A) and NuoA(E81A) subunit bands became remarkably thinner than the band of the wild type and the mobility of the NuoA(D79A) subunit band on the SDS gels decreased. After electrophoresis the proteins were transferred to nitrocellulose membranes, and Western blotting was carried out with SuperSignal West Pico system according to Han et al. (30). The secondary antibody used for detection was goat anti-rabbit IgG horseradish peroxidase conjugate (Pierce).
FIG. 4. NADH dehydrogenase activity staining and immunostaining of the blue native polyacrylamide gels of the E. coli wild type (W), knock-out (KO), and NuoA mutant membranes. A, NADH dehydrogenase activity staining of the blue native polyacrylamide gel. The arrow indicates the NADH dehydrogenase activity band because of the NDH-1. The electrophoresis and NADH diaphorase staining were performed as described under "Experimental Procedures." B, immunoblotting of the E. coli membrane proteins by using the affinity-purified antibody specific toward the E. coli NuoE subunit. After blue native polyacrylamide gel electrophoresis, the E. coli membrane proteins were transferred to nitrocellulose membranes. Subsequently, the nitrocellulose membranes were immunostained with the affinity-purified NuoE antibody as described in the legend for Fig. 3. The arrow shows the location of the immunoblotting bands recognized by the anti-NuoE antibody.
carboxyl residues in the transmembrane regions of the NDH-1/complex I, namely, two each in subunits NuoA/ND3, NuoH/ ND1, NuoK/ND4L, and NuoL/ND5. It has been reported that the two carboxyl residues in subunit NuoH are not essential for the energy-coupled activities of the NDH-1 (37). Recently, Kervinen et al. (43) demonstrated that one of the two conserved carboxyl residues of subunit NuoK (Glu-36, E. coli numbering) is indispensable for the energy-transducing NDH-1 activities. In this paper, we have investigated the two conserved carboxyl residues, Asp-79 and Glu-81, of the E. coli NuoA subunit. As it turned out, mutating only one of them caused either little or partial inhibition of energy-coupled activities (deamino-NADH-O 2 and deamino-NADH-DB). However, these activities were almost completely lost when both carboxyl groups were removed. There are at least two explanations for the results obtained with the NuoA mutants. One is that the residues Asp-79 and Glu-81 synergistically contribute to the maintenance of an intact architecture of the NDH-1, and thus disruption of both but not either one of them results in drastic change in the structure. However, this seems unlikely because in all mutants examined the assembly of the whole enzyme seems to be normal, and the deamino-NADH-Fe(CN) 6 activity remained unchanged. Another possibility is that Asp-79 and Glu-81 are both involved in the mechanism of ion translocation, but they may work in a compensatory manner. In other words, subunit NuoA needs to have at least one carboxyl group in the area where the two residues are located for the NDH-1 to function as a pump. It can be further postulated that Glu-81 may be in a more favorable position than Asp-79, because omission of the latter has much less impact on the coupled activities. In fact, Glu-81 is perfectly conserved in all sequences available in data bases, whereas Asp-79 is replaced with C in C. caldarium mitochondria (CAA88774) and with G in P. aeruginosa (D83410). The significance of the presence of two carboxyl groups in NuoA is somewhat similar to that in NuoK. As described above, the NuoK subunit has two highly conserved glutamic acids presumably in the middle of transmembrane segments. Both residues seem to be required for the optimal activity and removal of one of them leads to a partial or substantial loss of coupled activity. Although we do not know the relative positioning of the two subunits, NuoA and NuoK, it might be possible that together they provide negatively charged groups that constitute the H ϩ or Na ϩ binding site as reported for certain cation transporters (47,53,54).
Recently, it has been reported that two mutations of the ND3 gene (S34P, T10158C; S45P, T10191C) induced infantile mitochondrial encephalopathy (26) and a progressive mitochondrial disease (25). The two Ser residues are not conserved and located in Loop 1 segment of the NuoA/ND3 subunit (see Fig. 2). The mutants had moderately reduced amounts (44 -65%) of complex I but drastically reduced amounts of complex I activity (1-11% remaining) (26). According to the predicted topology of the NuoA subunit, the Loop 1 segment is localized in the periplasmic phase and, therefore, may not interact with the peripheral segment of NDH-1/complex I. We have constructed and characterized mutants of highly conserved charged resi-  6 reductase, deamino-NADH oxidase, and deamino-NADH-DB reductase activities between the wild type and D79N/E81Q mutant membranes. dNADH indicates deamino-NADH. Left, wild type membranes (80 g/ml). Right, D79N/E81Q mutant membranes (80 g/ml). Where indicated, 0.15 mM dNADH and 10 M cap-40 were added. The reaction medium was composed of 10 mM potassium phosphate (pH 7.0) containing 1 mM EDTA. When required, 10 mM KCN, 100 M DB, and 1 mM K 3 Fe(CN) 6 were added. The dNADH oxidase and dNADH-DB reductase activities were observed at 340 nm. The dNADH-K 3 Fe(CN) 6 reductase activity was measured at 420 nm.
dues Lys-46 and Glu-51, which are present in the Loop 1 segment. Although the K46A mutation did not affect any NDH-1 activities, E51A mutation showed a significant inhibition (ϳ70% inhibition) of the energy-transducing NDH-1 activities. On the other hand, neither mutation induced any drastic modification of subunit assembly or capsaicin sensitivity. The data suggest that the Loop 1 segment may be involved in the NDH-1/complex I activity, although this loop faces the periplasmic phase (the cytoplasmic phase in mitochondria). In eukaryotes, it is not easy to manipulate mitochondrial DNA. In contrast, gene manipulation procedures of bacterial DNA are well advanced. Although there are some limitations, the bacterial NDH-1 is a useful model to clarify functional roles of amino acid residues of interest, especially residues involved in human mitochondrial diseases, in the membrane domain ND subunits of complex I.