Loss of Ammonia Monooxygenase Activity in Nitrosomonas europaea upon Exposure to Nitrite

Abstract

Nitrosomonas europaea, an obligate ammonia-oxidizing bacterium, lost an increasing amount of ammonia oxidation activity upon exposure to increasing concentrations of nitrite, the primary product of ammonia-oxidizing metabolism. The loss of activity was specific to the ammonia monooxygenase (AMO) enzyme, as confirmed by a decreased rate of NH₄⁺-dependent O₂ consumption, some loss of active AMO molecules observed by polypeptide labeling with ¹⁴C₂H₂, the protection of activity by substrates of AMO, and the requirement for copper. The loss of AMO activity via nitrite occurred under both aerobic and anaerobic conditions, and more activity was lost under alkaline than under acidic conditions except in the presence of large concentrations (20 mM) of nitrite. These results indicate that nitrite toxicity in N. europaea is mediated by a unique mechanism that is specific for AMO.

Ammonia oxidation by Nitrosomonas europaea is mediated by two enzymes, ammonia monooxygenase (AMO), which catalyzes the oxidation of NH₃ to hydroxylamine (NH₂OH), and hydroxylamine oxidoreductase (HAO), which catalyzes the oxidation of NH₂OH to NO₂⁻ (26). For every molecule of NH₃ oxidized, one molecule of NO₂⁻ is produced. Additionally, ammonia oxidation is an acidogenic reaction. As the pH of the incubation medium decreases during ammonia oxidation, the NH₃-NH₄⁺ equilibrium is shifted away from NH₃, the true substrate of AMO (23). Ammonia oxidation ceases at about pH 6, partially because of the reduced concentration of NH₃. At the point of maximal cell growth in phosphate-buffered medium with an initial ammonium concentration of 50 mM, the concentration of nitrite is typically around 20 to 25 mM, and the rest of the ammonium remains unoxidized. In a previous study, we showed that this remaining NH₃-NH₄⁺ pool, or other substrates of AMO, had a specific protective effect on the AMO activity of N. europaea cells (22). In incubations where the ammonium was completely consumed, the cells lost up to 80% of their ammonia-oxidizing activity over a 24-h period (22). In the present paper, we show that this specific loss of AMO activity in cells of N. europaea is due to the toxicity of accumulated nitrite in the incubation medium.

Surprisingly, only a few reports describing the toxicity of nitrite in ammonia-oxidizing bacteria have been published, and the mechanism of toxicity remains unclear. In one study, nitrite toxicity in Nitrosomonas sp. occurred only at very high concentrations (greater than 30 mM) of nitrite, and this effect, as measured by the ability of the cells to consume O₂, was greater during the lag phase than during the log phase of growth at several different pH values tested (16). Nitrite was also toxic for cells in the log phase of growth, but only at acidic pH values, and the loss of activity was reversible upon washing of the cells (16). Other studies have shown that ammonia oxidizers are sensitive to nitrite accumulation in sewage treatment plants, but the mechanism of this sensitivity was not examined (2).

Because batch cultures of N. europaea accumulate large concentrations of nitrite, greater than 20 mM, and reach a pH of about 6, it is surprising that the cells are not more susceptible to nitrite toxicity. In cultures of Clostridium sporogenes, a food spoilage bacterium, 10 mM nitrite has profound bacteriocidal effects, especially at pH values below 7 (6). Many species of bacteria are susceptible to nitrite toxicity because of the formation of metal-nitrosyl complexes that occurs when NO⁺ or NO radicals interact with bacterial enzymes (28). The radicals, NO⁺ and NO, form spontaneously from nitrite most favorably in acidic media.

The present study shows that nitrite can cause a specific loss of ammonia-oxidizing activity in N. europaea cells at lower concentrations than those previously considered (5 to 20 mM). The loss of activity is specific to AMO and occurs under both acidic and alkaline conditions. In the presence of substrates of AMO, the loss of activity was not observed. Furthermore, the loss of activity was not reversible by washing the cells. Lastly, nitrite appears to specifically target the AMO enzyme in a manner different from that of characterized inactivators of AMO.

Analysis of batch incubations of N. europaea containing ammonium or nitrite.

Cells of N. europaea (ATCC 19178) were grown to late log phase, harvested, and washed as described previously (22). Cells (ca. 10⁹ cells ml⁻¹) were incubated in medium (25 ml) initially containing between 0 and 50 mM ammonium for 24 h. The changes in the ammonia and hydroxylamine oxidation activities, as measured by NH₄⁺- and NH₂OH-dependent O₂ uptake rates, respectively (12), and the nitrite concentrations in the incubation medium (9) were monitored. A decrease in the ammonia oxidation activity was observed over 24 h at intermediate ammonium concentrations, between 0 and 50 mM, with the greatest loss occurring at 15 to 25 mM (Fig. 1A). For example, in the incubation that contained no ammonium, the ammonia oxidation activity decreased to about 81% of the original level after 24 h. With 15 mM ammonium, the point of maximal activity loss, the ammonia oxidation activity decreased to approximately 18% of the initial level. In the incubation containing 50 mM ammonium, the ammonia oxidation activity decreased to only about 63% of the initial level. In all of the incubations, there was little change in the hydroxylamine oxidation activity after 24 h, confirming the previous observation that incubations containing different concentrations of ammonium specifically affect the ammonia oxidation activity of the cells (Fig. 1A) (22).

FIG. 1.

Ammonia and hydroxylamine oxidation activities and concentrations of nitrite and ammonium after a 24-h incubation of N. europaea cells in ammonium-containing media. Cells were incubated in growth medium containing from 0 to 50 mM ammonium. Washed cells were sampled for O₂ consumption rates, and the supernatant was assayed for pH and nitrite accumulation. Error bars represent the standard deviations from the averages of six replicate experiments. (A) Remaining NH₄⁺-dependent (■) and NH₂OH-dependent (□) O₂ uptake rates for cells incubated in media containing from 0 to 50 mM NH₄⁺ for 24 h relative to the amount of initial activity. The rates of NH₄⁺- and NH₂OH-dependent O₂ uptake at 100% activity were approximately 120 and 35 nmol of O₂ consumed min⁻¹ ml of cells⁻¹, respectively. (B) Measured concentration of nitrite (•) and calculated concentration of ammonium (○) (assuming minimal incorporation into cellular biomass) in the medium from the incubations in panel A after 24 h.

The trend of ammonia oxidation activity loss was correlated with the relative proportions of nitrite produced and ammonium remaining in the incubation medium after 24 h (Fig. 1B). The accumulation of nitrite in the medium was proportional to the initial concentration of ammonium up to 20 mM. In incubations containing 25 to 50 mM ammonium, nitrite accumulation ceased after reaching about 21 mM. At this point, the limiting pH for ammonia oxidation, 5.7 to 6.0, was reached. The greatest losses of ammonia oxidation activity were observed with the largest concentrations of nitrite, from 15 to 21 mM, and the smallest concentrations of ammonium, from 0 to 5 mM, remaining in the medium after the 24-h incubation.

Incubations were also conducted in the absence of ammonium and in the presence of increasing concentrations of nitrite, from 0 to 20 mM, in medium at a constant pH of 8. Increased loss of ammonia oxidation activity with increased nitrite concentration was observed over the 24-h incubation, although several hours were required for substantial losses of activity to occur (data not shown). The greatest loss of ammonia oxidation activity, approximately 62%, occurred with 20 mM nitrite in the incubation. Again, the hydroxylamine oxidation activity was not strongly affected by nitrite in any of the incubations. After the incubations with nitrite, cells were sedimented, washed, and resuspended in sodium phosphate buffer for 48 h to determine if the effect of nitrite was reversible. No activity was recovered during these incubations, suggesting an irreversible inactivation of the ammonia oxidation activity by nitrite (data not shown).

The effect of O₂ and CH₄ on the loss of ammonia oxidation activity.

To further characterize the mechanism of nitrite toxicity for ammonia oxidation activity, we incubated cells in concentrations of nitrite ranging from 0 to 20 mM both aerobically and anaerobically and with or without CH₄, an alternative substrate for AMO (Fig. 2). The incubations without O₂ were conducted in glass serum vials (160 ml) sealed with butyl rubber stoppers and aluminum crimp seals. The vials were evacuated and reequilibrated five times with O₂-free N₂. Complete anaerobiosis was monitored by gas chromatography (thermal conductivity detector) and the lack of nitrite production in the vials upon addition of ammonium. The amount of ammonia oxidation activity remaining after 24 h was measured in each treatment. The cells retained about 10% more activity on average in the absence than in the presence of O₂ in the incubations without CH₄; however, the differences were not substantial (Fig. 2A). Increasing amounts of nitrite resulted in increased losses of ammonia oxidation activity in both the aerobic and the anaerobic incubations. Thus, it did not appear that O₂ was required for the disabling effect of nitrite on ammonia oxidation activity.

FIG. 2.

Response of the ammonia oxidation activity in cells exposed to nitrite under aerobic or anaerobic conditions with or without CH₄. N. europaea cells were incubated in media containing the indicated concentrations of nitrite. (A) Incubations were conducted both aerobically (solid bars) and anaerobically (shaded bars). (B) Incubations were conducted aerobically (open bars) and anaerobically (hatched bars) in the presence of 25% (vol/vol of the gas headspace) CH₄. The ammonia oxidation activity after a 24-h incubation relative to the initial activity is shown. Error bars represent the standard deviations from the averages of three replicate experiments.

Incubations with and without 10 mM nitrite were conducted aerobically and anaerobically in the presence of 25% (vol/vol of the gas headspace) CH₄ to verify the protective effect of alternative substrates on ammonia oxidation activity as shown in a previous study (22). In all of the incubations, an average of 78% of the ammonia oxidation activity remained, indicating that CH₄ protected the activity both aerobically and anaerobically and in the presence of nitrite (Fig. 2B).

Effect of pH on nitrite-mediated loss of ammonia oxidation activity.

The loss of ammonia oxidation activity was 20% greater in incubations initially containing 15 mM ammonium than in incubations containing 20 mM nitrite alone. Because of this discrepancy and the dynamic changes in pH that occur in incubations with ammonium, the possibility of pH being a mediator of activity loss was investigated. Incubations containing 0, 5, 10, or 20 mM nitrite were conducted at a range of fixed pH values from 5.5 to 8. Without nitrite, the ammonia oxidation activity remaining after 24 h was the same regardless of the initial pH (Fig. 3). However, when nitrite was included in the incubations, the ammonia oxidation activity decreased more at alkaline (pH 7 to 8) than at acidic (pH 5.5 to 6.5) pH values, with up to 10 mM nitrite. Activity loss also occurred with a high concentration of nitrite, 20 mM, at an acidic pH (5.5 to 6). Only a slight loss of activity was observed with 10 mM nitrite at pH 5.5, indicating that, although nitrite toxicity does occur at both alkaline and acidic pH values, it is dependent upon the concentration of nitrite in the incubation medium.

FIG. 3.

Effect of nitrite on ammonia oxidation activity at different pH values. N. europaea cells were incubated in media with 0 (■), 5 (▴), 10 (•), or 20 (□) mM nitrite at a range of pH values from 5.5 to 8.0. The ammonia oxidation activity remaining after a 24-h incubation was determined relative to the initial activity. Error bars represent the standard deviations from four replicate experiments.

The extent of activity loss at alkaline pH was unexpected because most of the chemistry that transforms nitrite into active compounds, e.g., those that form metal-nitrosyl complexes, occurs at an acidic pH (28). Thus, traditional nitrite chemistry can explain the concentration-dependent loss of activity at pH 5.5, but another mechanism must also exist for the activity loss observed at pH 8. Furthermore, in incubations containing 15 mM ammonium (Fig. 1), the pH declined to about 6.6 and 80% of the ammonia oxidation activity was lost. In contrast, with 20 mM nitrite at the same pH, only about 40% of the activity was lost (Fig. 3). These results suggest that the interaction between pH and nitrite must be more complex during the dynamic process of ammonia oxidation.

Quantification of the active AMO polypeptide pool by ¹⁴C₂H₂.

The pool of active AMO enzyme molecules was examined by radiolabeling with ¹⁴C₂H₂ to determine whether the loss in ammonia oxidation activity, characterized in the preceding experiments, was due to the inactivation of the AMO enzyme. Exposure of N. europaea cells to ¹⁴C₂H₂ results in the specific incorporation of radiolabel into the 27-kDa, active-site containing subunit of AMO (12). The amount of radiolabel incorporated is proportional to the amount of active AMO enzyme at the time of exposure to ¹⁴C₂H₂. Cells of N. europaea were incubated with 0, 15, or 50 mM ammonium; 10 mM nitrite; or 10 mM nitrite without O₂. Cells (1 ml, sedimented) were taken from each of these treatments at an initial time point and after 24 h and were incubated for 2 h in sodium phosphate buffer (900 μl), ¹⁴C₂H₂ (∼300 μCi), and either hydrazine (2 mM), tetramethylhydroquinone (TMHQ; 0.5 mM), or endogenous reductant. Incubations containing 10 mM NO₂⁻ and 15 mM ammonium were also analyzed after incubation with ¹⁴C₂H₂ for 1 and 3 h to verify that the labeling reactions had reached completion after a 2-h incubation.

Hydroxylamine provides reductant to AMO via HAO (26), TMHQ likely provides reductant through a ubiquinone pathway (20), and endogenous reductant is likely mediated through NADH (25). Because each of the reductant sources may involve different electron carriers at some point in the electron transport pathway, a difference in label incorporation would suggest that an electron transport molecule between AMO and HAO, such as cytochrome c₅₅₄ or a quinone, may be damaged by nitrite rather than by AMO itself. After exposure to ¹⁴C₂H₂, the cells were sedimented, resuspended in sodium dodecyl sulfate-polyacrylamide gel electrophoresis loading buffer (100 μl), and frozen at −80°C. Cell extracts were thawed and vortexed (1 min), and the polypeptides were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (12% polyacrylamide). Incorporation of ¹⁴C into polypeptides was analyzed on a PhosphorImager (Molecular Dynamics, Sunnyvale, Calif.). The apparent masses of labeled polypeptides were determined by comparison with R_f values for molecular mass markers as described previously (12). Densitometric values were determined with ImageQuant software.

The amount of radiolabel incorporation into polypeptides after 24 h relative to the amount at the initial time point was determined (Fig. 4). Treatments containing 15 mM ammonium or 10 mM nitrite with or without O₂ (treatments 2, 4, and 5, respectively) resulted in about a 15 to 40% decrease in the amount of label incorporation into the 27-kDa polypeptide after 24 h. However, these same treatments exhibited a 50 to 80% decrease in ammonia oxidation activity as measured by the O₂ electrode. These results indicate that nitrite does not initially cause a complete inactivation of AMO molecules but rather that nitrite lowers the rate of AMO activity. However, some molecules of AMO are completely inactivated, as indicated by the loss of 15 to 40% of radiolabeled polypeptide in these same treatments.

FIG. 4.

¹⁴C₂H₂ labeling of cells exposed to ammonia or nitrite for 24 h. Cells of N. europaea were incubated in media with the following treatments: 1, no additions; 2, 15 mM ammonium; 3, 50 mM ammonium; 4, 10 mM nitrite; and 5, 10 mM nitrite and without O₂. Aliquots of cells were exposed to ¹⁴C₂H₂ at the beginning of the treatments and after 24 h, as described in the text, to radiolabel the active AMO population. The difference in radiolabel accumulation between the initial and 24-h time points is shown for each treatment as an average of two replicate experiments. Labeling reaction mixtures contained 2 mM hydrazine (solid bars), 0.5 mM TMHQ (shaded bars), or endogenous reductant (open bars).

There were no major differences in label incorporation with the different reductant sources for each treatment. The same amount of label was incorporated with endogenous reductant as with added hydrazine or TMHQ, indicating that very little reductant was necessary for the complete labeling of the AMO polypeptide population by ¹⁴C₂H₂. Therefore, even if an electron carrier was debilitated by nitrite, the low level of endogenous reductant flow would circumvent the crippled pathway. This result alone was unable to verify whether AMO or an accessory molecule was the target for nitrite. As a better test for the involvement of AMO, cells were incubated in the presence of the copper chelator thiourea.

The role of copper in mediating the toxicity of nitrite.

Based on several lines of evidence, AMO is proposed to contain copper (7, 19, 27). Thiourea chelates cuprous copper, which leads to the inactivation of copper-containing enzymes, including AMO (3, 10). Cells were incubated with 0, 10, or 20 mM nitrite either in the presence or in the absence of thiourea (100 μg ml⁻¹). The loss of AMO activity by thiourea is partially reversible by stringent washing and addition of copper. Once the thiourea-treated cells were washed and copper was reintroduced, about 50 to 60% of the ammonia oxidation activity was recovered (Fig. 5). The initial activity of the cells which were not exposed to nitrite or thiourea was designated as 100%, and the activity remaining after 24 h in all of the incubations was relative to this 100% activity level. Cells exposed to thiourea maintained approximately the same amount of activity after 24 h with or without nitrite in the incubations, indicating that copper had to be present for nitrite to inflict damage on the rate of O₂ consumption by the cells. Because copper plays a major role in maintaining AMO activity, it is likely that the removal of copper from AMO resulted in the inability of nitrite to inactivate the ammonia-oxidizing activity of the cell.

FIG. 5.

Effect of thiourea on the loss of ammonia oxidation activity in cells exposed to nitrite. N. europaea cells were incubated in media containing 0, 10, or 20 mM nitrite either with or without the copper chelator thiourea (100 μg ml⁻¹). The initial activity of cells incubated without thiourea (solid bars) and with thiourea (open bars) and the amount of activity remaining after 24 h without thiourea (shaded bars) and with thiourea (hatched bars) are shown. Error bars represent the standard deviations from the averages of three replicate experiments.

Specificity of nitrite in mediating the loss of ammonia-oxidizing activity.

We attempted to determine whether the loss of ammonia oxidation activity was specific to nitrite or if other chemicals related to nitrite chemistry could also induce activity loss. Nitric oxide (NO) and nitrous oxide (N₂O) are produced by N. europaea from either the reduction of nitrite or the oxidation of hydroxylamine during ammonia metabolism (1, 11, 24). In many systems, NO is known to be toxic, especially through the formation of metal-nitrosyl complexes when NO interacts with metal-containing enzymes (5, 28). Furthermore, it has been reported recently that a putative iron moiety within the AMO enzyme can interact with NO in cell extracts (27). There is no evidence that N₂O is toxic to N. europaea, but because it is a product of nitrite reduction, we tested whether it could mediate activity loss. Addition of NO or N₂O to incubations had little effect on the ammonia oxidation activity of N. europaea, and the observed effects were not specific for AMO (data not shown). Similarly, no loss of activity was observed in the presence of NO₃⁻, a compound which is similar to nitrite in size, composition, and charge. Therefore, these results suggest that nitrite itself, rather than a known product of nitrite chemistry, causes the activity loss.

Taking all of the results together, we suggest that nitrite lowers the ammonia oxidation activity of N. europaea by specifically debilitating AMO. Mechanism-based inactivators of AMO require the presence of O₂, because the product of oxidation, rather than the substrate itself, is the toxic agent (21). Therefore, nitrite cannot be considered a mechanism-based inactivator of AMO, and the low rate of inactivation suggests that nitrite also does not belong to two other known classes of AMO inactivators, copper chelators (e.g., thiourea) or metabolic inactivators (e.g., nitrapyrin) (3, 17). Furthermore, nitrite does not cause a complete loss of activity as do many of these other inactivators. However, inactivation of AMO by CS₂ does not require O₂ and potentially involves a reversible interaction with a nucleophilic amino acid that is in close proximity to the putative copper molecule in the AMO active site (13). Although nitrite might target AMO in a manner similar to that of CS₂, the loss of activity mediated by nitrite is not reversible and occurs much slower than inactivation by CS₂, suggesting again that nitrite is a unique inactivator of AMO.

Although we cannot discount the possibility that nitrite may interact with AMO as NO₂⁻, it seems more logical that nitrite would be converted to a reactive species before interacting with the enzyme, especially if the target is a metal-containing moiety (4, 28). One possible mechanism is that nitrite could be reduced by AMO to a reactive species (e.g., NO or a nitrosyl radical) which targets a copper or iron molecule in close proximity to the enzyme active site, forming a covalently bound metal-nitrosyl complex. Once the complex is formed, the catalytic rate of the enzyme may be lowered initially by a debilitation of O₂ binding and activation, substrate binding and oxidation, or reduction of the active site. Furthermore, the debilitation of AMO by nitrite may render the enzyme more susceptible to complete inactivation under the appropriate conditions (Fig. 4). Nitrite most likely interacts with the reduced form of AMO because, under anaerobic conditions, most of the enzyme will be in a reduced state and nitrite is still toxic in this situation (Fig. 2). Although there is only indirect evidence for the involvement of AMO as the target of nitrite, the ability of substrates to protect the activity and the role of copper in mediating toxicity imply that AMO, and not an accessory enzyme (e.g., an electron carrier), is the target.

The loss of ammonia-oxidizing activity experienced by N. europaea upon exposure to nitrite may influence the ecology of ammonia oxidizers. For example, although the interactions between ammonia- and nitrite-oxidizing bacteria are dependent upon several factors including substrate availability, O₂, and pH (14), the toxic effect of nitrite on ammonia oxidizers could be ameliorated by the close physical association between the two bacterial groups as observed in soils and bioreactors (15, 18). The close physical positioning of ammonia and nitrite oxidizers in the natural environment is useful for both energetic reasons and the prevention of substrate accumulation that could be toxic or could lead to the formation of toxic by-products, especially in environments where nitrite cannot simply diffuse away (8). Therefore, this study suggests that the challenge for ammonia-oxidizing bacteria in natural systems is not only survival with an inconstant energy source but also avoidance of toxic product formation once the ammonia has been converted to nitrite.