INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

NADH oxidases are found in several microorganisms (9, 12, 20, 31, 41) and have been purified from at least nine bacterial species (7, 10, 13, 25, 30, 34, 40, 42, 44). There are two types of NADH oxidase, H₂O forming and hydrogen peroxide forming. We previously purified the hydrogen peroxide-forming NADH oxidases from aerobically grown Amphibacillus xylanus and Sporolactobacillus inulinus, both of which are facultatively anaerobic bacteria that lack a respiratory chain (25, 30). The physiological function of these enzymes was first thought to be the in vivo regeneration of NAD in aerobic metabolism of the bacteria (22-24, 30). The enzymes catalyze the reduction of oxygen by NADH to form hydrogen peroxide. However, in the presence of a 21-kDa disulfide-containing redox protein (AhpC), now commonly referred to as peroxiredoxin (Prx), the NADH oxidases also showed extremely high reductase activity for both hydrogen peroxide and alkyl hydroperoxides (26, 28-30). These NADH oxidases thus belong to a growing new family of peroxiredoxin oxidoreductases (PrxR) (38) and are involved not only in the regeneration of NAD but also in the removal of peroxides. Thus, in spite of lacking a respiratory chain and peroxide-scavenging enzymes such as catalase and heme-containing peroxidases, Amphibacillus xylanus and Sporolactobacillus inulinus can grow as well under aerobic conditions as do conventional aerobic bacteria (15, 24). In order to investigate the distribution of this type of NADH oxidase in bacteria, 15 bacterial strains were selected from aerobic, facultatively anaerobic, and anaerobic bacteria. The characteristics of the enzyme systems in these strains are reported here.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains and growth conditions. The bacterial strains used in this study are listed in Table 1. Alcaligenes faecalis NRIC 1001^T, Pseudomonas aeruginosa NRIC 1114^T, Bacillus licheniformis NRIC 1863, Corynebacterium glutamicum JCM 1318^T, Escherichia coli NRIC 1509, and Salmonella enterica serovar Typhimurium NRIC 1851 were grown under aerobic conditions with shaking. Corynebacterium glutamicum was cultured at 30°C, and Alcaligenes faecalis, P. aeruginosa, Bacillus licheniformis, E. coli, and Salmonella enterica serovar Typhimurium were cultured at 37°C. The growth medium of Alcaligenes faecalis, Corynebacterium glutamicum, and Salmonella enterica serovar Typhimurium consisted of 1.0% peptone, 1.0% beef extract, and 0.5% NaCl, pH 7.0. P. aeruginosa growth medium consisted of 0.63% peptone, 0.38% beef extract, and 0.63% NaCl, pH 7.0. Bacillus licheniformis growth medium consisted of 0.5% glucose, 0.2% yeast extract, 1.0% peptone, 0.2% ammonium nitrate, 0.1% sodium citrate, 0.6% KH₂PO₄, 1.4% K₂HPO₄, and 0.01% MgSO₄ · 7H₂O, pH 7.0. E. coli growth medium consisted of 1.0% tryptone, 0.5% yeast extract, and 0.5% NaCl, pH 7.2. Zymomonas mobilis subsp. mobilis NRIC 1158^T was grown with or without shaking at 30°C in a medium consisting of 2.0% glucose and 0.5% yeast extract, pH 6.6. Amphibacillus xylanus JCM 7361^T was grown at 39.5°C with shaking for aerobic growth or was grown anaerobically as described previously (24). (The type strain of Amphibacillus xylanus Ep01 was isolated as described previously [23] and has been deposited in the Japan Collection of Microorganisms, RIKEN, Wako, Saitama, Japan, as strain JCM 7361^T.) Amphibacillus xylanus growth medium consisted of 1.0% glucose, 0.3% yeast extract, 0.03% peptone, 0.2% ammonium nitrate, 0.1% K₂HPO₄, 1.0% CaCl₂ · 2H₂O, and 1.0% salt solution containing 20 mg of MgSO₄ · 7H₂O, 0.5 mg of MnSO₄ · 5H₂O, and 0.5 mg of FeSO₄ · 7H₂O per ml. The medium was adjusted to pH 10.0 by 10% of 1 M NaHCO₃-Na₂CO₃ buffer, pH 10.55 (24). Sporolactobacillus inulinus NRIC 1133^T and Lactobacillus delbrueckii subsp. delbrueckii NRIC 1053^T were grown at 37 and 30°C, respectively, with shaking for aerobic growth or by flushing the medium bottles with nitrogen gas for anaerobic growth. Lactococcus lactis subsp. lactis NRIC 1149^T was grown at 30°C with or without shaking. The growth medium of Sporolactobacillus inulinus, Lactobacillus delbrueckii, and Lactococcus lactis consisted of 1.0% glucose, 1.0% yeast extract, 1.0% peptone, 1.0% sodium acetate, 0.2% beef extract, and 0.5% salt B solution containing 40 mg of MgSO₄ · 7H₂O, 2 mg of MnSO₄ · 5H₂O, 2 mg of FeSO₄ · 7H₂O, and 2 mg of NaCl in 1 ml, pH 6.8. Bacteroides vulgatus JCM 5826^T, Bifidobacterium bifidum JCM 1255^T, Clostridium aminovalericum DSM 1283^T, and Clostridium butyricum JCM 1391^T were grown at 37°C by flushing the medium bottles with nitrogen gas for anaerobic growth; in addition, for aerobic growth, anaerobically growing cells were cultured with agitation for 15 min under aerobic conditions. The growth medium of Bacteroides vulgatus, Clostridium aminovalericum, and Clostridium butyricum consisted of 1.5% glucose, 1.0% yeast extract, 2.0% peptone, 0.22% beef extract, 0.5% soluble starch, 0.25% KH₂PO₄, 0.3% NaCl, 0.03% L-cysteine-HCl, 0.03% sodium thioglycolate, and 0.0001% resazurin, pH 7.3. The growth medium of Bifidobacterium bifidum consisted of 1.0% glucose, 0.5% yeast extract, 1.0% peptone, 0.5% beef extract, 0.1% Tween 80, 0.3% K₂HPO₄, 0.05% L-cysteine-HCl, and 1.0% sodium ascorbate, pH 6.8. All strains were harvested by centrifugation, washed with 50 mM sodium phosphate buffer, pH 7.0, containing 5 mM EDTA, and then stored at 80°C until use.

Preparation of cell extracts. Bacterial cells were suspended in three volumes of 50 mM sodium phosphate buffer, pH 7.0, containing 5 mM EDTA. The suspensions were treated with lysozyme and then disrupted by three passages through a French pressure cell (SLM-AMINCO; Spectronic Instruments) at 1,400 kg/cm². Phenylmethylsulfonyl fluoride (final concentration, 2 mM) was added immediately after the first passage through the French pressure cell. The supernatant was centrifuged at 16,000 × g for 20 min, the pellet discarded and the supernatant ultracentrifuged at 187,000 × g for 90 min to obtain the cytoplasmic fraction. The cytoplasmic fractions were dialyzed against 50 mM sodium phosphate buffer, pH 7.0, containing 0.5 mM EDTA.

NADH oxidase assay. NADH oxidase activity was assayed at 30, 37, or 39.5°C (at the optimum growth temperature of each tested bacterium) in 3 ml of 50 mM sodium phosphate buffer, pH 7.0, containing 0.5 mM EDTA, 0.25 mM NADH, and cell extracts. The reaction was monitored with a Clark oxygen electrode (model 5331; Yellow Springs Instrument Co., Yellow Springs, Ohio). One unit of activity was defined as the amount of protein required to catalyze the consumption of 1 µmol of oxygen per min.

Enzymes. NADH oxidases from Amphibacillus xylanus and from Sporolactobacillus inulinus were purified as described previously (30, 33). NADH oxidase from Bacillus licheniformis was purchased from Nacalai Tesque, Inc. (Kyoto, Japan). AhpC from A. xylanus was isolated as described previously (29).

Southern blot analysis. The genomic DNA of all bacterial strains was isolated by the method of Saito and Miura (43). The DNA was digested with EcoRI, electrophoresed on an agarose gel, and then transferred to a positively charged nylon membrane. The membrane was hybridized with digoxigenin-11-dUTP-labeled random-primed Amphibacillus NADH oxidase DNA probes which were prepared using a DIG DNA labeling kit (Boehringer Mannheim GmbH Biochemica, Mannheim, Germany) according to the manufacturer's instructions.

Steady-state kinetics. The apparent K_m value for oxygen was determined from Lineweaver-Burk plots of the kinetic data obtained at 25°C and various oxygen concentrations in 50 mM sodium phosphate buffer, pH 7.0, containing 0.5 mM EDTA and 120 µM NADH. The decrease in the A₃₄₀ was monitored with a spectrophotometer.

Hydroperoxide reductase activity assays. The activities reported in Table 2 were obtained as follows. The reaction mixture (2 ml) containing 50 mM sodium phosphate buffer (pH 7.0), 2 mM EDTA, 4% ammonium sulfate, 2.5 mM NADH, 1 mM t-butyl hydroperoxide, cell extracts, and 30 µM Amphibacillus xylanus AhpC was loaded into a 2-ml microtube, which was closed with a cap. Alkyl hydroperoxide reduction was performed by incubation for 60 min at 30, 37, or 39.5°C (optimum growth temperatures of the individual bacteria). The reaction mixture was then subjected to high-performance liquid chromatographic analysis on an Inertsil ODS-2 column, with 5 mM potassium phosphate buffer (pH 7.0)-acetonitrile (9:1) as the mobile phase, and the absorbance of hydroperoxide at 230 nm was monitored. Activity was expressed in milliunits per milligram, where 1 mU is defined as reduction of 1 nmol of hydroperoxide per min.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Distribution of alkyl hydroperoxide reductase activity in the presence of AhpC in bacteria. The AhpC-linked alkyl hydroperoxide reductase activities in cell extracts of bacteria were examined using t-butyl hydroperoxide as the substrate. By high-performance liquid chromatographic analysis, the reduction of t-butyl hydroperoxide was assayed in the presence of various combinations of the cell extract, AhpC from A. xylanus, and NADH. Although in the absence of additional AhpC, every bacterial cell extract showed some reductase activity for alkyl hydroperoxide, the activity was increased by the addition of A. xylanus AhpC in most of the tested strains. Since AhpC proteins are widely distributed in bacteria, including aerobic, facultatively anaerobic, and anaerobic bacteria, the cell extracts of each tested strain presumably contain some AhpC protein (2, 3, 5, 19, 21). In the preparation and assay of the cell extracts, the concentrations of endogenous AhpC were diluted about 15-fold and may not have been high enough for the full observation of AhpC-linked alkyl hydroperoxide reductase activity compared with that in vivo, suggesting that the observed increments of activity by addition of AhpC can be attributed to intrinsic AhpC-linked hydroperoxide reductase activity. AhpC-linked t-butyl hydroperoxide reductase activities were detected in most of the tested strains (Table 2). Especially high increments of activity (over 4.0 mU/mg more than that found for the archetypal Salmonella enterica serovar Typhimurium) were observed in six species, including the aerobic bacteria: Alcaligenes faecalis and Bacillus licheniformis, facultatively anaerobic bacteria having a respiratory chain (E. coli and Salmonella enterica serovar Typhimurium), and facultatively anaerobic bacteria lacking a respiratory chain (Amphibacillus xylanus and Sporolactobacillus inulinus). The two aerobic bacteria and two of the facultatively anaerobic bacteria have a respiratory chain and also catalase. The other two species belonging to the facultatively anaerobic class lack a respiratory chain and also lack catalase. These data suggest that the AhpC-linked reductase system functions as an effective system for removing hydroperoxides in bacteria growing under aerobic conditions, whether the bacteria have or lack catalase (see Discussion).

Distribution of hydrogen peroxide-forming NADH oxidase activity in bacteria. NADH oxidase activity is defined as the ability to reduce oxygen with NADH, regardless of whether the reduction product is H₂O₂ or H₂O. Cell extracts of bacteria with a respiratory chain also show oxygen consumption caused by the addition of NADH, but this consumption should be inhibited by KCN, while the oxygen consumption by NADH oxidases is not (25, 30). Oxygen consumption on the addition of NADH was observed in cell extracts of all strains used in this study (Table 3). These activities were in general not inhibited or, in some cases, only marginally inhibited by KCN (1 mM), suggesting that NADH oxidases were present in all of the tested strains (Table 3). NADH oxidases so far studied are divided into two types, hydrogen peroxide- and H₂O-forming enzymes, both of which have been found in several bacterial species (1, 7, 10, 13, 25, 30, 34, 40-42, 44). Hydrogen peroxide-forming NADH oxidases are known to be markedly accelerated in activity by addition of free flavin adenine dinucleotide (FAD) (10, 30, 33). In contrast, NADH oxidase activity by addition of free FAD is only increased slightly with a H₂O-forming NADH oxidase (10). NADH oxidase activities which were accelerated over fivefold by the addition of 50 µM FAD were observed in strains of the following species: Alcaligenes faecalis and Bacillus licheniformis (aerobic bacteria); E. coli and Salmonella enterica serovar Typhimurium (facultatively anaerobic bacteria having a respiratory chain); Amphibacillus xylanus, Sporolactobacillus inulinus, and Lactobacillus delbrueckii (facultatively anaerobic bacteria lacking a respiratory chain); and the anaerobic bacterium Bacteroides vulgatus. Hydrogen peroxide-forming NADH oxidase activity is thus presumably present in these eight strains, including the six strains showing high AhpC-linked t-butyl hydroperoxide reductase activities.

Reaction products in the absence or presence of AhpC. To confirm the presence of hydrogen peroxide-forming NADH oxidase in these eight strains, oxygen consumption experiments were performed with and without the addition of catalase. The purified alkyl hydroperoxide reductase flavoprotein of Amphibacillus xylanus, which catalyzed the reduction of oxygen to hydrogen peroxide, showed 50% reformation of oxygen on addition of catalase (25); however, in the presence of AhpC no reformation of oxygen was detected on addition of catalase (29). The enzyme thus catalyzes the four-electron reduction of oxygen to water in the presence of additional AhpC. Even in cell extract experiments with A. xylanus, in which the presence of AhpC has been shown unambiguously (29), hydrogen peroxide is found to accumulate on oxidation of NADH, as shown by O₂ liberation on addition of catalase, but does not accumulate in the presence of added AhpC (data not shown). In the cell extracts from Sporolactobacillus inulinus, Lactobacillus delbrueckii, and Bacteroides vulgatus, reformation of oxygen on addition of catalase was observed in the absence of AhpC, and in the presence of AhpC, no hydrogen peroxide production was observed (results not shown). These results show that a hydrogen peroxide-forming NADH oxidase was present and could function as a peroxide reductase in these bacteria in the presence of AhpC. All of these bacteria lack catalase. As expected, in the absence of AhpC, reformation of oxygen was not observed in the cell extracts from bacterial species having catalase: Alcaligenes faecalis, Bacillus licheniformis, E. coli, and Salmonella enterica serovar Typhimurium. Since acceleration of NADH oxidase activity on addition of FAD indicates the presence of hydrogen peroxide-forming NADH oxidases in these bacteria, hydrogen peroxide produced by the NADH oxidase reaction is presumably scavenged by the endogenous catalase.

Southern blot analysis. To demonstrate the distribution of the prxR gene in bacteria, we performed Southern blot analysis of genomic DNA of the tested bacterial strains which had been digested with EcoRI and probed with Amphibacillus xylanus prxR. Southern blot analysis revealed a single band in six bacterial strains (Fig. 1). Weak bands were revealed in Alcaligenes faecalis, E. coli, Clostridium aminovalericum, and Bacteroides vulgatus. Particularly strong bands were revealed in Bacillus licheniformis and Sporolactobacillus inulinus, indicating that the DNA fragment, which has a high degree of similarity to the Amphibacillus xylanus NADH oxidase gene, is in the nuclear DNA of Bacillus licheniformis and Sporolactobacillus inulinus. In these bacteria, including Amphibacillus xylanus, as mentioned above, marked acceleration of NADH oxidase activity on addition of FAD is observed, and high AhpC-linked t-butyl hydroperoxide reductase activity is also observed in the presence of AhpC, suggesting that the NADH oxidases from these bacteria have enzymatic properties similar to those of enzymes from Amphibacillus xylanus, i.e., they are all members of the peroxiredoxin oxidoreductase (PrxR) family. A comparison of the purified NADH oxidases from these bacteria is presented below.

View larger version (47K): [in this window] [in a new window]   FIG. 1.   Southern blot analysis of the DNA of tested bacterial strains. The DNAs were digested with EcoRI. Hybridization was done as described in Materials and Methods. Lane 1, Bacillus licheniformis NRIC 1863; lane 2, Corynebacterium glutamicum JCM 1318T; lane 3, Alcaligenes faecalis NRIC 1001T; lane 4, P. aeruginosa NRIC 1114T; lane 5, Amphibacillus xylanus JCM 7361T; lane 6, Sporolactobacillus inulinus NRIC 1133T; lane 7, Lactobacillus delbrueckii subsp. delbrueckii NRIC 1053T; lane 8, Lactococcus lactis subsp. lactis NRIC 1149T; lane 9, E. coli NRIC 1509; lane 10, Salmonella enterica serovar Typhimurium NRIC 1851; lane 11, Z. mobilis subsp. mobilis NRIC 1158T; lane 12, Bifidobacterium bifidum JCM 1255T; lane 13, Clostridium aminovalericum DSM 1283T; lane 14, Clostridium butyricum JCM 1391T; lane 15, Bacteroides vulgatus JCM 5826T. DNA size markers are on the right.

Characterization of purified hydrogen peroxide forming-NADH oxidases from Amphibacillus xylanus, Bacillus licheniformis, and Sporolactobacillus inulinus. In our previous reports of Amphibacillus xylanus NADH oxidase, the K_m value for oxygen was found to be 1.7 mM, too high to catalyze effectively the reoxidation of NADH by oxygen in the cell (33). In the presence of additional FAD, the NADH oxidase activity was accelerated markedly, and the K_m value for oxygen was greatly diminished; indeed, it was too low to allow accurate determination of its value by the usual assay method (33). The NADH oxidase activity in cell extracts from Bacillus licheniformis and Sporolactobacillus inulinus was also found to be accelerated in the presence of additional FAD. Further analyses were carried out with the purified enzymes. Lineweaver-Burk plots of the NADH oxidase activities in the presence of 120 µM NADH are shown in Fig. 2. The apparent K_m values for oxygen with the NADH oxidases from Bacillus licheniformis and from Sporolactobacillus inulinus were 3.3 and 1.3 mM, respectively. The Lineweaver-Burk plots of the NADH oxidase activities from these strains were similar to that of Amphibacillus xylanus PrxR. In the presence of FAD, the NADH oxidase activities were accelerated, and the apparent K_m values for oxygen were greatly diminished and were too low to allow accurate determination of values (Fig. 2, inset).

View larger version (18K): [in this window] [in a new window]   FIG. 2.   Lineweaver-Burk plots of steady-state kinetic analyses of NADH oxidases from the indicated species. Assay conditions were 50 mM sodium phosphate buffer, pH 7.0, containing 0.5 mM EDTA and 120 µM NADH with or without 150 µM FAD at 25°C.

1

Induction of NADH oxidase activity under aerobic conditions. Nine bacterial species, including facultatively anaerobic bacteria lacking a respiratory chain (Z. mobilis, Amphibacillus xylanus, Sporolactobacillus inulinus, Lactobacillus delbrueckii, and Lactococcus lactis) and anaerobic bacteria (Bacteroides vulgatus, Bifidobacterium bifidum, Clostridium aminovalericum, and Clostridium butyricum), were cultivated under anaerobic or aerobic conditions in this study. The NADH oxidase activities which were induced over twofold under aerobic conditions compared to under anaerobic conditions were observed only in the facultatively anaerobic bacteria lacking a respiratory chain: Z. mobilis, Amphibacillus xylanus, Sporolactobacillus inulinus, and Lactobacillus lactis (Table 5). Induction of NADH oxidase activities over fourfold were especially evident in Amphibacillus xylanus and Sporolactobacillus inulinus, both of which can grow well under aerobic conditions as well as under anaerobic conditions. However, induction of the NADH oxidase activities was not observed in anaerobic bacteria, which cannot grow well under aerobic conditions, including Bacteroides vulgatus, Bifidobacterium bifidum, Clostridium aminovalericum, and Clostridium butyricum (Table 5). The induced NADH oxidase is thus possibly playing an important role in aerobic metabolism of the bacteria (see below).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

AhpC-linked t-butyl hydroperoxide reductase activities were observed in cell extracts from most of the tested strains (Table 2). The bacteria growing well under aerobic conditions, except Corynebacterium glutamicum and P. aeruginosa, showed high AhpC-linked reductase activity. In contrast, AhpC-linked t-butyl hydroperoxide reductase activities were low in bacteria which cannot grow well under aerobic conditions. These results suggest that the peroxiredoxin oxidoreductase system is present mainly in bacteria capable of good aerobic growth. Although Corynebacterium glutamicum and P. aeruginosa grow well under aerobic conditions, high AhpC-linked t-butyl hydroperoxide reductase activities were not observed in their cell extracts. The NADH oxidase activities of enzymes belonging to the peroxiredoxin reductase family are markedly accelerated on addition of free FAD (30, 33), but the NADH oxidase activities of the above two organisms were only slightly accelerated by addition of FAD, suggesting that the content of peroxiredoxin oxidoreductase is low or absent in these strains. Conventional heme peroxidases or catalase probably function to remove peroxides in such bacteria.

All of the tested strains showing PrxR activity in the cell extracts, except for Salmonella enterica serovar Typhimurium, revealed a single band in Southern blot analyses probed with the Amphibacillus xylanus prxR gene. Salmonella enterica serovar Typhimurium was the first organism shown to possess the peroxiredoxin reductase-peroxiredoxin system (11) and the nucleotide sequence of the gene encoding the flavoprotein component, prxR (formerly known as AhpF), has already been determined (48). Regardless of their similar catalytic function, the similarity of their nucleotide sequences is presumably not sufficient (54.8%) for rigorous hybridization. Particularly strong bands were revealed in the tested gram-positive bacteria, including Amphibacillus xylanus, Sporolactobacillus inulinus, and Bacillus licheniformis, suggesting high similarity in their nucleotide sequences. In a previous article, we reported that Sporolactobacillus NADH oxidase catalyzes the reduction of oxygen to hydrogen peroxide and, in the presence of AhpC, of hydrogen peroxide to water (30). The purified enzyme from Bacillus licheniformis also showed high reductase activity for both hydrogen peroxide and alkyl hydroperoxide. Their V_max values were similar to those of the NADH oxidases from Amphibacillus xylanus and Sporolactobacillus inulinus (Table 4).

The AhpC-linked two-electron reduction of hydrogen peroxide and alkyl hydroperoxides has been observed with the following enzymes: NADH oxidases from Amphibacillus xylanus (26, 28, 29), Sporolactobacillus inulinus (30), Streptococcus mutans (39), and Bacillus licheniformis (this work); NADH dehydrogenase from an alkaliphilic Bacillus strain (14); and alkyl hydroperoxide reductase flavin component (AhpF) from Salmonella enterica serovar Typhimurium (26, 36, 37, 38). All of these enzyme systems belong to the peroxiredoxin oxidoreductase family and show extremely high reductase activities for both hydrogen peroxide and alkyl hydroperoxides (14, 26, 28-30). Their V_max values are similar to the rate constant for reduction of the flavin component of the Amphibacillus PrxR by NADH, suggesting that this was the rate-determining step of the overall reaction (27). The K_m values for hydroperoxides were too low to allow accurate determination of their values in our experiments. No other peroxide-scavenging enzyme, catalase or peroxidase, so far studied has been reported to show such high turnover numbers and low K_m values for both hydrogen peroxide and alkyl hydroperoxide as described here (4, 6, 8, 11, 17, 32, 35, 47). A recent report also describes the isolation and function of peroxiredoxin and peroxiredoxin reductase from a thermophile, Thermus aquaticus (18). The flavoprotein component, PrxR, was originally isolated as an NADH oxidase (7). This system, however, shows much lower catalytic activity than the ones described above (18).

Amphibacillus xylanus and Sporolactobacillus inulinus, lacking both catalase and conventional heme-containing peroxidase, can grow as well under aerobic conditions as aerobic bacteria that have catalase and peroxidase, suggesting that the PrxR-Prx system can function in the removal of peroxides in these organisms. The NADH oxidases from these organisms thus catalyze the four-electron reduction of oxygen to water in the presence of AhpC (28, 30). This activity is also thought to function in vivo to regenerate NAD from NADH produced in aerobic metabolism of Amphibacillus xylanus and Sporolactobacillus inulinus, both of which lack a respiratory chain (Fig. 3) (28-30). On the other hand, the NADH oxidase activities of Salmonella AhpF and Bacillus NADH dehydrogenase are low compared with those of enzyme from Amphibacillus xylanus and Sporolactobacillus inulinus. The PrxR-Prx systems presumably function physiologically to remove peroxides in Salmonella and Bacillus, because these organisms have a respiratory chain and do not need the oxidase activity for the regeneration of NAD. Indeed, the only functional difference in the NADH-linked flavoprotein components appears to be in the low reactivities of the Salmonella and Bacillus enzymes with molecular oxygen. Acceleration of NADH oxidase activities by addition of FAD and AhpC-linked peroxide reductase activities was widely detected in most of the aerobically growing bacteria, indicating that the AhpC-linked NADH oxidase system is distributed widely not only in facultative anaerobes that lack a respiratory chain but also in aerobes having a respiratory chain. This conclusion has also been reached on the basis of BLAST searches of GenBank and other databases, indicating a wide distribution of PrxR and Prx proteins (38).

View larger version (26K): [in this window] [in a new window]   FIG. 3.   Metabolic pathways of Amphibacillus xylanus and Sporolactobacillus inulinus.

In conclusion, the NADH oxidases belonging to the peroxiredoxin oxidoreductase family are widely distributed and probably play an important role both in peroxide-scavenging systems and in effective regeneration of NAD to maintain oxidative and reductive balance in bacterial cells that can grow well under aerobic conditions, including not only bacteria lacking a respiratory chain, catalase, and conventional peroxidases but also bacteria having all of these.