Constitutive and Adaptive Detoxification of Nitric Oxide inEscherichia coli

Nitric oxide (NO⋅) is a naturally occurring toxin that some organisms adaptively resist. In aerobic or anaerobic Escherichia coli, low levels of NO⋅exposure inactivated the NO⋅-sensitive citric acid cycle enzyme aconitase, and inactivation was more effective when the adaptive synthesis of NO⋅-defensive proteins was blocked with chloramphenicol. Protection of aconitase in aerobically grown E. coli was dependent upon O2, was potently inhibited by cyanide, and was correlated with an induced rate of cellular NO⋅consumption. Constitutive and adaptive cellular NO⋅ consumption in aerobic cells was also dependent upon O2 and inhibited by cyanide. Exposure of aerobic cells to NO⋅ accordingly elevated the activity of the O2-dependent and cyanide-sensitive NO⋅ dioxygenase (NOD). Anaerobic E. coli exposed to NO⋅ or nitrate induced a modest O2-independent and cyanide-resistant NO⋅-metabolizing activity and a more robust O2-stimulated cyanide-sensitive activity. The latter activity was attributed to NOD. The results support a role for NOD in the aerobic detoxification of NO⋅ and suggest functions for NOD and a cyanide-resistant NO⋅ scavenging activity in anaerobic cells.

Nitric oxide (NO ⅐ ) is a naturally occurring toxin that some organisms adaptively resist. In aerobic or anaerobic Escherichia coli, low levels of NO ⅐ exposure inactivated the NO ⅐ -sensitive citric acid cycle enzyme aconitase, and inactivation was more effective when the adaptive synthesis of NO ⅐ -defensive proteins was blocked with chloramphenicol. Protection of aconitase in aerobically grown E. coli was dependent upon O 2 , was potently inhibited by cyanide, and was correlated with an induced rate of cellular NO ⅐ consumption. Constitutive and adaptive cellular NO ⅐ consumption in aerobic cells was also dependent upon O 2 and inhibited by cyanide. Exposure of aerobic cells to NO ⅐ accordingly elevated the activity of the O 2 -dependent and cyanide-sensitive NO ⅐ dioxygenase (NOD). Anaerobic E. coli exposed to NO ⅐ or nitrate induced a modest O 2 -independent and cyanide-resistant NO ⅐ -metabolizing activity and a more robust O 2 -stimulated cyanide-sensitive activity. The latter activity was attributed to NOD. The results support a role for NOD in the aerobic detoxification of NO ⅐ and suggest functions for NOD and a cyanide-resistant NO ⅐ scavenging activity in anaerobic cells.
Nitric oxide (NO ⅐ ) 1 is released by leukocytes and functions as an antibiotic (1)(2)(3). It may also be produced endogenously by bacteria during the reduction of NO 2 Ϫ by nitrate reductase (4). In addition, bacteria may encounter NO ⅐ released by competing microorganisms (5). Regardless of the source, NO ⅐ produced at sufficient levels directly or indirectly damages critical cell processes (3). Indeed, NO ⅐ is bacteristatic toward some bacteria (6), and NO ⅐ or NO ⅐ -derived species may display bactericidal activities in vitro and in infected animals (1)(2)(3)(7)(8)(9)(10).
Various organisms may benefit from adaptive mechanisms for NO ⅐ detoxification. Denitrifying bacteria (11,12) and fungi (13) are known to produce NO ⅐ -inducible (14) NO ⅐ -detoxifying NORs. NORs catalytically reduce NO ⅐ to produce nitrous oxide (N 2 O). NORs also increase the anaerobic energy production capacity of denitrifiers by catalyzing an essential step in the reduction of nitrate (NO 3 Ϫ ) and nitrite (NO 2 Ϫ ) to N 2 (15). Considerably less is known of the adaptations of nondenitrifying bacteria to NO ⅐ or of their normal exposures to NO ⅐ . An adap-tation of Escherichia coli to NO ⅐ under the transcriptional control of the antioxidant regulators SoxRS and OxyR has been suggested, since these global antioxidant regulators provide some survival and growth benefits against NO ⅐ (16,17) or nitrosothiols (18), respectively. Yet, it remains unclear how these regulators protect bacteria. E. coli does not appear to produce a typical NOR activity, but it does produce a multiheme nitrite reductase with NO ⅐ -reducing capacity (19) and a nitric oxide dioxygenase (NOD) that has been proposed to function in NO ⅐ detoxification (20).
We have observed, and now report, an increased susceptibility to NO ⅐ of the NO ⅐ -sensitive citric acid cycle enzyme aconitase (21,22) in aerobic or anaerobic E. coli inhibited for de novo protein synthesis. Protection of aconitase activity correlated with an increased rate of cellular NO ⅐ consumption and correlated with an increased NOD activity in cell-free extracts. The results support the proposed function of NOD in the constitutive and adaptive detoxification of NO ⅐ in aerobic E. coli. The possible role of the NO ⅐ -induced NOD activity in anaerobic E. coli is discussed.

MATERIALS AND METHODS
Cells and Reagents-E. coli strain DH5␣ was from Life Technologies, Inc. Mutants deficient in the terminal oxidases ECL936 (⌬cyo), ECL937 (⌬cyd), and the parent ECL933 were kindly provided by E. C. C. Lin (23). Compressed gas cylinders containing 1200 ppm (Ϯ 5%) NO ⅐ in ultrapure N 2 , 99.998% N 2 , and 99.993% O 2 were obtained from Praxair (Bethlehem, PA). NO ⅐ -saturated water was prepared by stirring N 2equilibrated water under 98.5% NO ⅐ gas (Aldrich), which was first bubbled through 1 N NaOH. Saturated NO ⅐ was stored at 4°C in a rubber septum-sealed glass tube. Sodium cyanide, NADP ϩ , FAD, chloramphenicol, glucose oxidase from Aspergillus niger, glucose-6-phosphate dehydrogenase from bakers' yeast, sodium nitrite, sodium nitrate, succinic acid, D-glucose, glucose 6-phosphate, L-arginine-HCl, thiamine HCl, Tris, and MnCl 2 were from Sigma. Nitrate reductase from Aspergillus niger and bovine liver catalase were purchased from Boehringer Mannheim. Tryptone and yeast extract were obtained from Difco.
Media, Growth of Bacteria, and Extract Preparation-The minimal salts medium was made up with tap water and contained 60 mM K 2 HPO 4 , 33 mM KH 2 PO 4 , 7.6 mM (NH 4 ) 2 SO 4 , 1.7 mM sodium citrate, 1 mM MgSO 4 , 10 M MnCl 2 , 10 g/ml thiamine Cl, 40 g/ml L-arginine, and 10 mM sodium succinate or 10 mM glucose as indicated. The phosphate-buffered LB medium was prepared with 10 g of tryptone, 5 g of yeast extract, and 10 g of NaCl per liter of 66 mM K 2 HPO 4 , 33 mM KH 2 PO 4 , and 10 M MnCl 2 . The pH of the phosphate-buffered LB and minimal salts media were adjusted to 7.0 with HCl or NaOH. MnCl 2 was routinely added to media to ensure full expression of the inducible O 2 . -scavenging manganese-containing superoxide dismutase, maximal activity of the O 2 . -sensitive aconitase, and thus optimal growth on citric acid cycle-dependent substrates (32). To achieve maximal gas exchange, cultures were routinely grown in a gyrorotary water bath shaking at Ͼ200 rpm at 37°C with a medium:flask volume ratio of at most 1:5, and growth was monitored by following the turbidity at 550 nm. Bacterial densities were determined by dilution, plating, and colony counting. An absorbance of 1.0 at 550 nm corresponded to 7 ϫ 10 8 bacteria/ml when bacteria were grown in the minimal medium. Anoxic growth of cultures was achieved by incubating cultures at 37°C in static stopper-sealed 50-ml Erlenmeyer flasks filled with 50 ml of medium. To minimize the disturbance of head space gases, culture aliquots were removed from gas-equilibrated culture flasks using a 1-ml tuberculin syringe connected via small tubing. Culture aliquots were immediately transferred to 1.5-ml Eppendorf tubes and were quickly centrifuged at 20,000 ϫ g for 25 s, the supernatant was aspirated, and the cell pellet was frozen on dry ice. Cell pellets were resuspended and lysed by sonicating in 0.1 ml of buffer containing 50 mM Tris-Cl, pH 7.4, 0.6 mM MnCl 2 , and 20 M barium dl-fluorocitrate, and the lysate was frozen on dry ice. Cell lysates were stored at Ϫ70°C for up to 2 weeks without noticeable loss of aconitase activity. Lysates were thawed in a 25°C water bath and clarified by centrifugation for 25 s at 20,000 ϫ g immediately prior to the assay of aconitase activity. Cell lysates were prepared for the assay of NOD activity essentially as described for the assay of aconitase except that the lysis buffer contained 50 mM potassium phosphate, pH 7.8, and 0.1 mM EDTA. Gas Exposures-A three-way gas proportioner (Cole-Parmer Instrument Co.) was used to produce various mixtures of O 2 , N 2 , and NO ⅐ at a constant flow rate of 30 ml/min, and gas mixtures were passed through a trap containing NaOH pellets to remove higher oxides of nitrogen.
Assay of Aconitase, Protein, NO 2 Ϫ , and NO 3 Ϫ -Aconitase activity and protein were assayed as described previously (21). For the measurement of media NO 2 Ϫ and NO 3 Ϫ , cultures were clarified by centrifugation, and supernatants were incubated at 37°C for 2 h in a 0.1-ml reaction mixture containing 7 milliunits of nitrate reductase and 40 M NADPH in 100 mM Tris-Cl, pH 7.5. Samples were assayed for NO 2 Ϫ with the Griess reagent (24) using sodium nitrite as a standard.
NO ⅐ and O 2 Consumption Measurements-NO ⅐ consumption was measured at 37°C with an NO ⅐ microelectrode (Diamond General Inc.) fitted in a water-jacketed glass-stoppered 2-ml capacity cell (Gilson Inc.) equipped with a magnetic stirrer. One microliter of NO ⅐ was delivered to the cell with a Hamilton syringe from a saturated solution (2 mM) prepared in water. Rates of NO ⅐ consumption by bacteria were measured in minimal medium salts containing chloramphenicol and glucose or succinate as indicated. Rates of NO ⅐ consumption were determined from initial rates and were corrected for the background rate of NO ⅐ decomposition. O 2 consumption was measured at 37°C with a Clark-type O 2 electrode (Yellow Springs Instrument Co.) in a waterjacketed cell in a total volume of 2.0 ml of minimal succinate medium. The O 2 concentration for media saturated with air at normal atmospheric pressure and 37°C was taken to be 200 M (25).
Assay for NOD Activity-Cell-free extracts were assayed for NOD activity at 37°C in a 2-ml reaction mixture containing 50 mM potassium phosphate buffer, pH 7.8, 0.1 mM EDTA, 1 M FAD, 0.2 mM NADP ϩ , 0.5 units/ml glucose-6-phosphate dehydrogenase, 2.5 mM glucose 6-phosphate, and 1 M NO ⅐ . Initial rates of NO ⅐ disappearance from reaction mixtures were followed amperometrically with an NO ⅐ electrode and were corrected for the background rate of NO ⅐ decomposition. Where indicated, O 2 was removed by incubating the mixture with 10 mM glucose, 2 units/ml glucose oxidase, and 130 units/ml catalase for 5 min prior to the addition of NO ⅐ and extract. O 2 removal was followed amperometrically with an O 2 electrode.
Data Analysis-Results are representative of two or more independent experiments.

RESULTS
Growth-inhibitory Effects of NO ⅐ -We measured the effects of NO ⅐ on the growth of E. coli in order to gauge the capacity of cells to adapt to NO ⅐ under various growth conditions. Exposure of log phase cultures to an atmosphere containing 960 ppm NO ⅐ had little effect on the aerobic growth of E. coli in either the minimal succinate or the rich LB medium (Fig. 1, A and B, compare lines 1 and 2). Similarly, 480 ppm NO ⅐ exerted no discernible effect on the anaerobic growth of E. coli in a minimal glucose medium supplemented with 10 mM nitrate (data not shown). Interestingly, however, NO ⅐ exposure caused a small, but significant, decrease in the aerobic growth rate and the yield of E. coli in the LB medium at the end of the log phase (Fig. 1B, compare lines 1 and 2). The effect of growth phase on NO ⅐ inhibition was explored further. Exposure of late stationary phase E. coli to 960 ppm NO ⅐ strongly inhibited growth (Fig. 1, C and D, compare lines 1 and 2), and the growthinhibitory effects of NO ⅐ were more pronounced in the LB medium (Fig. 1D). Growth inhibition was readily reversible, as indicated by the ability of NO ⅐ -treated cells to resume normal growth following NO ⅐ removal (Fig. 1, C and D, compare lines 2 and 3). The susceptibility of stationary phase cultures to NO ⅐mediated growth inhibition suggests a requirement for nutritional resources, protein synthesis, or growth competency for NO ⅐ resistance. Moreover, the ability of cells to grow normally at NO ⅐ levels that were previously shown to potently inactivate the citric acid cycle enzyme aconitase (21) suggests that NO ⅐ resistance is due to the presence of adaptive mechanisms.
Adaptive Protection of Aconitase against NO ⅐ -mediated Inactivation in Aerobic and Anaerobic E. coli-The NO ⅐ -sensitive aconitase (21,22) was used to explore the mechanism of adaptation of E. coli to NO ⅐ . Thus, we supposed that adaptation to NO ⅐ should correlate with the protection of aconitase from inactivation. Indeed, aconitase activity was more sensitive to NO ⅐ in the presence of the protein synthesis inhibitor chloramphenicol than in its absence (Fig. 2, compare lines 1 and 2). Further, NO ⅐ was a more effective inactivator of aconitase in the absence of O 2 than in its presence (Fig. 2, compare lines 2  and 4), and, interestingly, chloramphenicol had no apparent effect on the susceptibility of aconitase to NO ⅐ -mediated inactivation in the absence of O 2 (Fig. 2, compare lines 3 and 4). The data clearly indicate inducible protective mechanisms for aconitase and demonstrate a role of O 2 in the adaptive mechanism.
We also measured the effects of chloramphenicol on NO ⅐mediated aconitase inactivation in anaerobic E. coli, since aerobic succinate-adapted E. coli are dependent upon O 2 for growth, respiration, and ATP production, which may have affected the ability of cells to adapt. Exposure of anaerobic glucose-adapted E. coli to 240 ppm NO ⅐ in N 2 caused a greater decline of aconitase activity in the presence of chloramphenicol than in its absence (Fig. 3, compare lines 1 and 2). It is noteworthy that the inducible protection of aconitase was relatively less effective under anaerobic than aerobic conditions. Thus, a ϳ20% loss of aconitase was observed following 60 min of 240 ppm NO ⅐ exposure in anaerobic cultures (Fig. 3, line 1), whereas 480 ppm NO ⅐ did not affect the aconitase activity in aerobic cultures (Fig. 2, line 1). Thus, anaerobic as well as aerobic E. coli protect aconitase from NO ⅐ -mediated inactivation. Moreover, while O 2 was not absolutely essential for adaptive protection, it did increase the apparent NO ⅐ detoxification capacity of cells.
Dioxygen Dependence and Cyanide Sensitivity of Aconitase Protection-We supposed that the decreased sensitivity of aconitase to NO ⅐ in chloramphenicol-treated E. coli in the presence of O 2 might be due, at least in part, to a lower exposure to NO ⅐ , since the O 2 -mediated oxidation of NO ⅐ to form NO 2 ⅐ and N 2 O 3 would be expected to decrease the steady-state NO ⅐ levels and increase NO 2 Ϫ and NO 3 Ϫ formation (26). Alternatively, the O 2 -dependent protection may be due to an NO ⅐ metabolic pathway, which was directly or indirectly dependent upon O 2 . For example, the protective effect of O 2 may be linked to the res-piration of E. coli. Thus, the homologous mitochondrial terminal oxidase, cytochrome c oxidase, is thought to metabolize NO ⅐ (27,28).
To evaluate the contribution of O 2 -mediated NO ⅐ oxidation to the protection, we measured the concentration of O 2 required for aconitase protection in E. coli exposed to NO ⅐ and compared it with that required for NO ⅐ oxidation as detected by the formation of NO 2 Ϫ and NO 3 Ϫ in the culture medium. Surprisingly, aconitase was near maximally protected from the inactivating effect of 240 ppm NO ⅐ by the lowest O 2 concentration tested (ϳ17 M O 2 ) (Fig. 4, closed circles). However, at this O 2 level, NO 2 Ϫ and NO 3 Ϫ formation was only a fraction of that achievable via NO ⅐ autoxidation (Fig. 4, open circles). Thus, the O 2 -mediated decomposition of NO ⅐ does not appear to account for the protective effects of O 2 .
The role of respiration and the terminal respiratory oxidases in the O 2 -dependent protection of aconitase was assessed by measuring the effects of the inhibitor cyanide. The addition of cyanide (25 M) to aerobic cultures completely blocked the protection of aconitase by O 2 (Fig. 5A, compare lines 1 and 2), while cyanide was without effect in the absence of O 2 (compare lines 3 and 4). Importantly, cyanide was effective at decreasing aconitase protection at much lower concentrations than those required for the inhibition of respiration. Thus, while ϳ5 M NaCN was saturating in its effect on the O 2 -dependent protection of aconitase (Fig. 5B), half-maximal inhibition of respiration required Ͼ50 M NaCN (Fig. 5B, inset). The results clearly demonstrate a cyanide-sensitive mechanism of aconitase protection; however, the difference in cyanide sensitivities indicates a mechanism of inhibition independent of cell respiration and the terminal respiratory oxidases.
We also investigated the effects of O 2 and cyanide on the NO ⅐ sensitivity of aconitase in naive aerobic cultures and compared these effects with those in NO ⅐ -treated cultures to determine whether the adaptive protection displayed a similar O 2 dependence and cyanide sensitivity. Indeed, the induced protection of aconitase was O 2 -dependent; however, this protection appeared less sensitive to cyanide than the constitutive activity (Fig. 6). We were unable to test the effects of higher cyanide concentrations, because aconitase activity was sensitive to cyanide at Ͼ25 M.  2. Effects of chloramphenicol and O 2 on the NO ⅐ sensitivity of aconitase in aerobically grown E. coli. Aerobically grown cultures of DH5␣ were exposed for 60 min to an atmosphere containing various concentrations of NO ⅐ in N 2 with 21% O 2 (lines 1 and 2) or in N 2 only (lines 3 and 4). Exposures were in the presence (lines 2 and 4) or absence (lines 1 and 3) of chloramphenicol (200 g/ml). Chloramphenicol treatment and pre-equilibration under the corresponding atmosphere was for 10 min prior to NO ⅐ exposure. Cells were harvested, and extracts were prepared and assayed for aconitase activity and protein as described under "Materials and Methods." Cultures were initiated with a 30% inoculum from an overnight culture grown in minimal succinate medium and were grown to an A 550 of ϳ0.5 in fresh minimal succinate medium for exposures. Percentage of aconitase activity is expressed relative to that at time 0. 100% aconitase activity was equal to 310 Ϯ 21 milliunits/mg protein.

FIG. 3. Effect of chloramphenicol on the NO ⅐ sensitivity of aconitase in anaerobically grown E. coli.
Anaerobically grown DH5␣ were exposed to 240 ppm NO ⅐ in N 2 in the absence (line 1) or presence (line 2) of chloramphenicol (200 g/ml). Cultures were pretreated with chloramphenicol for 10 min prior to the NO ⅐ exposure. Cells were harvested at intervals, extracts were prepared, and aconitase activity and protein were assayed as described under "Materials and Methods." The percentage of aconitase activity was calculated relative to the corresponding control. Control aconitase activity was maintained at 22 Ϯ 2 milliunits/mg protein in the presence or absence of chloramphenicol. Anaerobic DH5␣ cultures were grown up in minimal glucose medium overnight and were washed, resuspended at a density of A 550 ϭ 0.8 in fresh minimal salts-glucose medium, and then incubated for 30 min with vigorous shaking at 37°C under an atmosphere of N 2 prior to the exposures. Inducible NO ⅐ Consumption in Aerobic and Anaerobic Cultures of E. coli-To determine whether aconitase protection in E. coli was associated with an increased rate of NO ⅐ metabolism, we measured the rate of NO ⅐ consumption by aerobic and anaerobic cells. We also measured the effects of O 2 and cyanide on these rates. As shown by the data in Fig. 7A, aerobic E. coli consumed NO ⅐ . Moreover, NO ⅐ consumption was sensitive to cyanide and dependent upon O 2 . Further, NO ⅐ consumption was induced approximately 13-fold in aerobic cultures exposed to 480 ppm NO ⅐ for 60 min (Fig. 7B). The induced rate of NO ⅐ consumption was also sensitive to cyanide and dependent upon O 2 . Control anaerobic cultures did not express an NO ⅐ consumption activity, whereas anaerobic cells exposed to 960 ppm NO ⅐ for 60 min produced an NO ⅐ -consuming activity that did not require O 2 and that was insensitive to cyanide (Table I,   Experiment A). Interestingly, however, the presence of O 2 in the assay revealed a cryptic NO ⅐ consumption activity that was cyanide-sensitive and was also induced ϳ12-fold by the NO ⅐ exposure. NO 3 Ϫ (10 mM) also induced an NO ⅐ -consuming activity in anoxic cultures with similar properties as that induced by gaseous NO ⅐ (Table I, Experiment B).
Induction of NOD by Nitric Oxide-Prompted by the aforementioned results, we identified an O 2 -dependent cyanide-sensitive NO ⅐ -converting activity in extracts of E. coli that was NOD/flavohemoglobin and has been described elsewhere (20). Further, we supposed that this NOD activity might account for the NO ⅐ inducible, cyanide-sensitive, and O 2 -dependent protection of aconitase and NO ⅐ consumption by cells. As shown by the data in Table II, NOD activity in DH5␣ cells was induced ϳ36-fold following an exposure to 960 ppm NO ⅐ . Extract NOD activity was dependent upon O 2 , was potently inhibited by cyanide, and displayed cofactor requirements consistent with the flavohemoglobin/NOD activity (20). NOD activity was also measured in anaerobic cultures exposed to NO ⅐ or NO 3 Ϫ as described in the legend to Table I. NOD activity was increased 33-fold (588 milliunits/mg versus 18 milliunits/mg) following exposure to 960 ppm NO ⅐ and 40-fold (644 milliunits/mg versus 16 milliunits/mg) during anoxic growth with NO 3 Ϫ . Thus, NO ⅐induced NOD activity levels correlate with the NO ⅐ -induced aconitase protection and NO ⅐ consumption in both aerobic and anaerobic cells.

DISCUSSION
Our results demonstrate aerobic and anaerobic pathways for NO ⅐ detoxification that appear to differ from the NORs described in denitrifiers (11-13, 14, 15). Foremost among these differences was the requirement for O 2 . Moreover, the similar NO ⅐ inducibility, O 2 dependence, and cyanide sensitivity between the NOD activity in extracts (Table II) and the inducible and constitutive aconitase protection and NO ⅐ consumption by intact E. coli strongly supports a role for the recently described NOD (20) in the observed aerobic NO ⅐ removal and detoxification pathway. We can estimate that the NOD activity measured in cell extracts is within range of the NO ⅐ consumption activity of intact aerobic E. coli. Thus, assuming roughly 10 Ϫ13 g of soluble protein per E. coli, we can calculate the NOD activity in extracts from control and NO ⅐ -exposed E. coli to equal 0.1 and 2.1 nmol of NO ⅐ /min/10 8 cells, respectively. These values are comparable with the respective values of 0.2 and 3.2 measured for intact aerobic DH5␣ grown under similar conditions (Fig. 7). Furthermore, NO ⅐ consumption by cells does not appear to involve the terminal respiratory oxidases as proposed for mitochondria (27,28), since cyanide was without effect on the respiration of E. coli at levels that inhibited NO ⅐ detoxification (Figs. 5 and 7). Moreover, E. coli strains deficient in either of the terminal respiratory oxidases (23) expressed normal levels of the constitutive and inducible aerobic NO ⅐ consumption activity (data not shown), whereas NOD/flavohemoglobin-deficient E. coli express essentially no constitutive or inducible aerobic NO ⅐ consumption activity (20).
The identity of the inducible anaerobic pathway for NO ⅐ metabolism and detoxification is less clear, since there is no evidence for a typical NOR in E. coli. The demonstration of NOD induction and a cryptic O 2 -stimulated and cyanide-sensitive NO ⅐ consumption activity in anaerobically grown E. coli exposed to NO ⅐ or NO 3 Ϫ suggests an additional role of NOD in anaerobic NO ⅐ detoxification. However, both the failure of cyanide to inhibit the anaerobic activity in vivo (Table I) and the failure of the induced NOD to metabolize NO ⅐ in the absence of O 2 (Table II) suggests a limited role of NOD/flavohemoglobin in anaerobic NO ⅐ metabolism and detoxification. Nevertheless, NOD may have other anaerobic protective functions or may prepare the cell for more efficient NO ⅐ detoxification upon exposure to O 2 . It will now be important to assess the overall contributions of NOD and other potential NO ⅐ -metabolizing systems, such as the NO ⅐ -reducing multiheme nitrite reductase (19), to NO ⅐ detoxification in E. coli using deletion strains and overexpressors under various growth conditions. Finally, it is hoped that greater knowledge of the function and regulation of NO ⅐ detoxification systems, including the E. coli NOD, may reveal the true antibiotic potential of NO ⅐ toward a variety of organisms. For example, the greater ability of NO ⅐ to inhibit the growth of stationary phase E. coli (Fig. 1) suggests that latent or dormant pathogens will be more susceptible to the antibiotic action of NO ⅐ than healthy growing microorganisms simply because they are unable to induce their NO ⅐ detoxification systems. Indeed, this may account for the activation of latent Mycobacterium tuberculosis and Leishmania spp. infections in nitric-oxide synthase-deficient mice (1,2). The ability of O 2 to increase the NO ⅐ detoxification capacity of cells through the action of an NOD activity, like that of the (flavo)hemoglobin (20), may also explain the sensitivity of intraerythrocytic malaria parasites to NO ⅐ at low O 2 tensions (31). Furthermore, greater knowledge of the NODs, NORs, and other NO ⅐ detoxification systems may facilitate the design of new drugs that target these systems and allow the expression of the full potential of NO ⅐ as a natural broad spectrum antibiotic (3).  ) were grown overnight in stoppered Erlenmeyer flasks filled to capacity with minimal glucose medium supplemented with amino acids. Cells were washed and resuspended at their original density (A 550 ϭ ϳ0.5) in fresh medium and were incubated under an atmosphere of N 2 for 30 min prior to a 60-min exposure to 960 ppm NO ⅐ in N 2 or to N 2 alone. For Experiment B, static overnight anoxic cultures of DH5␣ were grown under similar conditions except that amino acids were omitted and 10 mM NaNO 3 was added as indicated. Cells were harvested and washed, and NO ⅐ consumption was measured in minimal glucose medium as described under "Materials and Methods." Growth was initiated with 2% inocula from aerobic log phase cultures grown in phosphate-buffered LB medium. Results represent the average Ϯ S.D. of three measurements.  TABLE II NO ⅐ consumption by cell-free extracts of E. coli NO ⅐ consumption activity was measured in extracts prepared from strain DH5␣ grown aerobically in minimal salts succinate medium to an A 550 ϭ ϳ0.5. Cells were exposed either to air or to 960 ppm NO ⅐ in 21% O 2 balanced with N 2 for 60 min prior to harvest. The effects of omission of individual components of the reaction mixture, removal of O 2 , or the addition of 0.25 mM NaCN on the induced activity were measured. Cells were grown, extracts were prepared, and NO ⅐ consumption and protein were assayed as described under "Materials and Methods."