Anaerobic Ammonia Oxidation in the Presence of Nitrogen Oxides (NO_x) by Two Different Lithotrophs

Abstract

The anaerobic ammonia-oxidizing activity of the planctomycete Candidatus “Brocadia anammoxidans” was not inhibited by NO concentrations up to 600 ppm and NO₂ concentrations up to 100 ppm. B. anammoxidans was able to convert (detoxify) NO, which might explain the high NO tolerance of this organism. In the presence of NO₂, the specific ammonia oxidation activity of B. anammoxidans increased, and Nitrosomonas-like microorganisms recovered an NO₂-dependent anaerobic ammonia oxidation activity. Addition of NO₂ to a mixed population of B. anammoxidans and Nitrosomonas induced simultaneous specific anaerobic ammonia oxidation activities of up to 5.5 mmol of NH₄⁺ g of protein⁻¹ h⁻¹ by B. anammoxidans and up to 1.5 mmol of NH₄⁺ g of protein⁻¹ h⁻¹ by Nitrosomonas. The stoichiometry of the converted N compounds (NO₂⁻/NH₃ ratio) and the microbial community structure were strongly influenced by NO₂. The combined activity of B. anammoxidans and Nitrosomonas-like ammonia oxidizers might be of relevance in natural environments and for technical applications.

Two different metabolic pathways for anaerobic ammonia oxidation by lithotrophic microorganisms have been described (8, 14). The anaerobic ammonia oxidation by the planctomycete Candidatus “Brocadia anammoxidans” is a nitrite-dependent reaction. The reaction mechanism of this anammox process was examined in a laboratory-scale fluidized bed reactor by using ¹⁵N-labeled compounds. Ammonia is oxidized with hydroxylamine as the most probable electron acceptor, and hydrazine is the first detectable intermediate (22). Subsequently, the hydrazine is oxidized to dinitrogen gas. The four reducing equivalents are transferred to nitrite, forming hydroxylamine. During the conversion of ammonia, some nitrate is formed from nitrite. This reaction could provide reducing equivalents necessary for the fixation of carbon dioxide (22). The overall nitrogen balance shows a ratio of about 1:1.32:0.26 for the conversion of ammonia, nitrite, and nitrate (18). The overall anammox reaction is presented in equation 1.

(1)

Under optimal growth conditions the specific anammox activity is about 3 mmol of NH₄⁺ g of protein⁻¹ h⁻¹. The affinity for the substrates ammonia and nitrite is high, and the K_s values are less than 10 μM (5, 19). Interestingly, nitrite concentrations of 5 mM or higher and oxygen concentrations as low as 2 μM completely, but reversibly, inhibit the anammox activity (17). During the conversion of ammonia and nitrite no other intermediates or products, such as NO or N₂O, could be detected.

The obligately anaerobic nature of B. anammoxidans is in sharp contrast with the more versatile metabolism of Nitrosomonas strains (1, 2, 10). The latter organisms are able to gain energy by aerobic or anaerobic ammonia oxidation or by denitrification by using hydrogen or organic compounds as the electron donor (2). The anaerobic ammonia oxidation by Nitrosomonas is a nitrogen dioxide-dependent reaction (12). It has been shown that N₂O₄, the dimeric form of NO₂, is used as an oxidant (13). About 50% of the nitrite produced is used as a terminal electron acceptor and is converted to N₂ (equation 2).

(2)

In a pure culture, the specific anaerobic ammonia oxidation activity of Nitrosomonas eutropha is about 0.15 mmol of NH₄⁺ g of protein⁻¹ h⁻¹. In contrast to B. anammoxidans, Nitrosomonas produces significant amounts of nitric oxide and nitrous oxide during ammonia oxidation. The nitrogen oxides, NO and NO₂, are obligatory intermediates in the oxidation of ammonia (14), and they have regulatory effects on the metabolism of the nitrifiers (16, 23).

NO has toxic effects on many microorganisms (4, 6, 15, 24), but the effect on the anammox activity of planctomycetes is still unknown. In natural and man-made environments (e.g., wastewater treatment plants) these organisms may be exposed to high NO concentrations and frequently also to high NO₂ concentrations The aim of this study was to investigate the effect of nitrogen oxides on the ammonia oxidation of B. anammoxidans. Furthermore, the influence of NO and NO₂ on the microbial community structure and the specific ammonia oxidation activities of B. anammoxidans and Nitrosomonas were evaluated.

MATERIALS AND METHODS

Organisms.

An anammox enrichment culture containing about 80% B. anammoxidans (5 × 10⁸ cells ml⁻¹) was grown in a 2-liter sequencing batch reactor (18). Cells of N. eutropha strain N904 (isolated from cattle manure) were grown as described previously (12).

Medium and growth conditions.

Precultures of the anammox biomass were grown anaerobically at 28°C and pH 7.5. The mineral medium contained (per liter of demineralized water) 2.02 g of NaNO₂, 1.96 g of (NH₄)₂SO₄, 1.25 g of KHCO₃, 25 mg of KH₂PO₄, 300 mg of CaCl₂ · 2H₂O, 200 mg of MgSO₄ · 7H₂O, 6.25 mg of FeSO₄, 6.25 mg of EDTA, 25 ml of HCl (1 M), and 25 ml of a trace element solution. The trace element solution contained (per liter of demineralized water) 15 g of EDTA, 430 mg of ZnSO₄ · 7H₂O, 240 mg of CoCl₂ · 6H₂O, 990 mg of MnCl₂ · 4H₂O, 250 mg of CuSO₄ · 5H₂O, 220 mg of NaMoO₄ · 2H₂O, 190 mg of NiCl₂ · 6H₂O, 210 mg of NaSeO₄ · 10H₂O, and 14 mg of H₃BO₄. The influx of fresh medium into the 2-liter reactor was adjusted to 170 ml h⁻¹.

Precultures of N. eutropha were grown aerobically in a 10-liter sequencing batch reactor in mineral medium (12) aerated with 2 liters of air min⁻¹. The pH was kept at 7.5 by using a 20% Na₂CO₃ solution.

Experimental design.

Experiments were performed in a 2-liter laboratory-scale reactor with 1.7 liters of medium; 170 ml of fresh medium was added per hour, and the same volume was removed from the reactor. The effluent was filtered by using a hollow fiber membrane module (0.4 m²) with membranes consisting of polysulfone and a cutoff of 6 kDa (SPS 6005-6; Fresenius, St. Wendel, Germany) to achieve complete biomass retention (cross-flow filtration). While the cell-free medium (170 ml h⁻¹) was removed, the biomass was routed back into the reactor. Depending on the experiment the concentrations of substrates (ammonia and nitrite) in the fresh medium varied between 0 and 100 mM. The reactor was gassed with a 95% argon-5% CO₂ gas mixture supplemented with 0 to 1,000 ppm of NO or NO₂ at a rate of 0.5 to 1 liter min⁻¹ by using mass flow controllers. The level, temperature, dissolved oxygen content, and pH were permanently controlled (Applikon, Schliedam, The Netherlands). The temperature was maintained at 28°C. The pH was kept at 7.5. Due to cell growth the biomass concentration in the reactor increased. Every 3 days the biomass concentration was measured by protein determination, and the biomass concentration was adjusted to about 80 μg of protein ml⁻¹ (5 × 10⁸ cells ml⁻¹) by removing small amounts of biomass from the reactor.

Batch experiments were carried out in 100-ml glass bottles that were sealed with butyl rubber stoppers and contained 20 ml of mineral medium. The bottles were flushed with a 95% argon-5% CO₂ gas mixture supplemented with 0 to 1,000 ppm of NO. The anammox biomass (80% B. anammoxidans) was purified by density gradient centrifugation (99.6% ± 0.2% B. anammoxidans) as described previously (20). During the experiments (8 h) the cell suspension (5 × 10⁹ cells ml⁻¹) was stirred (800 rpm) to ensure efficient gas transfer. NO consumption was calculated on the basis of the difference in the NO concentrations at the beginning and at the end of an experiment. Control experiments were carried out with cell-free medium and heat-sterilized cell suspensions. The chemical conversion rates were low (less than 3% of the biological conversion rates).

Analytical procedures.

Measurements were carried out as described by Schmidt and Bock (12) (NH₄⁺, NO, and NO₂), Van de Graaf et al. (21) (NO₂⁻ and NO₃⁻), and Bradford (3) (protein concentration). Most-probable-number (MPN) analysis was performed as described by Mansch and Bock (7), and fluorescent in situ hybridization (FISH) was performed as described by Neef et al. (9). The following probes were used: NEU 653 (specific for Nitrosomonas-like microorganisms) and Amx 820 (specific for planctomycete-like bacteria). The FISH and MPN methods were used to determine the numbers of planctomycete and Nitrosomonas cells in the reactor. 4′,6′-Diamidino-2-phenylindole (DAPI) staining was used to determine the total cell number.

Determination of ammonia oxidation activity.

The aerobic ammonia oxidation activity of N. eutropha was determined by short-term activity tests (2 h). The tests were performed in 100-ml Erlenmeyer flasks containing 10 ml of mineral medium (12). The protein concentration was measured, and the time-dependent consumption of ammonia was determined. The aerobic ammonia oxidation activity was used to estimate the numbers of Nitrosomonas cells in the anammox systems. The method was calibrated with anaerobically grown N. eutropha cells (13), and the correlation between cell number and the amount of ammonia converted was determined; 10⁷ cells oxidized 40 ± 3.3 nmol of NH₄⁺ h⁻¹. This method was used to confirm the results of the FISH and MPN analyses. Due to the presence of colonies the MPN method underestimates the number of cells (7); the activity method tends to overestimate the number of cells.

RESULTS

Effects of NO₂ and NO on anammox reactor systems.

The initial experiments were designed to investigate the influence of nitrogen oxides on the nitrogen conversion activities in anammox systems. For these experiments, anammox biomass precultured for 3 months in the absence of NO_x was supplemented with increasing amounts of NO₂ or NO. The NO_x concentration was increased every 2 weeks. Throughout the experiments the ammonia and nitrite concentrations were adjusted to about 4 and 0.5 mM, respectively. After the NO_x concentration was increased, the system adapted within 1 week to the new conditions. In the second week the nitrogen conversion activities were measured every day, and the mean values were calculated (the standard deviation was ±6.5%). The results of the experiments which were performed in the presence of NO₂ are shown in Fig. 1.

FIG. 1.

Specific ammonia (solid bars) and nitrite (open bars) consumption rates and nitrate production rates (shaded bars) in an anammox system supplemented with different concentrations of NO₂. According to the Mann-Whitney U test, the increases in activity as the NO₂ concentration was increased up to 50 ppm and the decreases in activity at NO₂ concentrations between 50 and 200 ppm are highly significant (error rate, 0.005). The experiments were repeated three times.

When the NO₂ concentration in the gas mixture used for gassing was increased, ammonia and nitrite consumption, as well as nitrate production, increased. The highest ammonia and nitrite conversion rates observed in the presence of 50 ppm of NO₂ were about two times higher than the conversion rates in the absence of NO₂. A further increase in the NO₂ concentration resulted in slowly decreasing anaerobic ammonia oxidation activities. In the absence of NO₂ the growth rate of B. anammoxidans was about 0.003 h⁻¹, in the presence of 50 ppm of NO₂ the growth rate was about 0.004 h⁻¹, and in the presence of 200 ppm of NO₂ the growth rate was about 0.0028 h⁻¹. According to a Mann-Whitney U test the differences in the growth rates are significant (error rate, 0.05).

To enlarge the statistical basis for the observed effects, short-term batch experiments were performed to examine the effect of NO₂ (0 to 200 ppm) on the anammox biomass (10 replicate experiments). Again, the highest activities were measured in the presence of 50 ppm of NO₂ (the activities were 1.5 times higher than those in an NO₂-free control). Interestingly, the ratio between ammonia and nitrite consumption changed depending on the NO₂ concentration (Table 1). When no gaseous NO₂ was added, the NO₂⁻/NH₃ ratio was about 1.3, as previously determined by Strous et al. (18). When the NO₂ concentration was increased to 200 ppm, the NO₂⁻/NH₃ ratio decreased to about 1.25.

TABLE 1.

NO₂⁻/NH₃ consumption ratios in anammox reactors in the presence of NO₂ or NO

NOx concn (ppm)	NO2−/NH3 ratio in the presence of NO2	NO2−/NH3 ratio in the presence of NO
0	1.33	1.30
10	1.32	1.44
25	1.30	1.39
50	1.28	1.37
100	1.29	1.49
150	1.28	1.34
200	1.25	1.50
600	NDa	1.53
1,000	ND	2.58

^aND, not determined.

The influence of NO₂ addition on the composition of the microbial community in the reactor was evaluated by FISH, MPN, and activity tests (Table 2). Throughout the experiment small amounts of the biomass were removed regularly (every 3 days) to keep the biomass content of the reactor constant. Therefore, the number of B. anammoxidans cells remained unchanged (about 5 × 10⁸ cells ml⁻¹). In contrast, the number of Nitrosomonas cells increased significantly in the presence of NO₂. The ratio of B. anammoxidans cells to Nitrosomonas cells changed from about 2 × 10⁶:1 to about 30:1.

TABLE 2.

Numbers of B. anammoxidans and Nitrosomonas cells in reactors supplemented with NO₂ or NO^a

NOx (ppm)	No. of cells (ml−1) in the presence of:
	NO2				NO
	B. anammoxidans (FISH analysis)b	Nitrosomonas (FISH analysis)c,d	Nitrosomonas (MPN analysis)d	Nitrosomonas (activity test)e,f	B. anammoxidans (FISH analysis)b	Nitrosomonas (FISH analysis)	Nitrosomonas (MPN analysis)g	Nitrosomonas (activity test)e,g
0	5.1 × 108	NDh	1.7 × 103	2.3 × 102	6.9 × 108	ND	2.6 × 103	5.9 × 102
10	5.2 × 108	ND	6.3 × 103	6.1 × 104	7.2 × 108	ND	2.7 × 103	5.3 × 102
25	5.0 × 108	ND	1.3 × 104	4.7 × 105	6.5 × 108	ND	2.0 × 103	5.7 × 102
50	5.0 × 108	4.5 × 105	3.3 × 104	1.6 × 106	6.9 × 108	ND	2.2 × 103	5.0 × 102
100	5.1 × 108	1.2 × 106	9.1 × 104	4.7 × 106	6.5 × 108	ND	1.6 × 103	3.9 × 102
150	5.1 × 108	2.7 × 106	2.9 × 105	1.2 × 107	7.6 × 108	ND	8.4 × 102	1.9 × 102
200	5.2 × 108	6.4 × 106	7.6 × 105	1.8 × 107	6.9 × 108	ND	4.9 × 102	1.1 × 102

^aThe number of B. anammoxidans cells was controlled by biomass removal.

^bThe standard deviation was 7%. According to the Mann-Whitney U test the constant cell number is highly significant (error rate, 0.005).

^cThe standard deviation was 18%.

^dAccording to the Mann-Whitney U test the increasing cell number with increasing NO₂ concentration is significant (error rate, 0.05).

^eThe standard deviation was 9%.

^fAccording to the Mann-Whitney U test the increasing cell number with increasing NO₂ concentration is highly significant (error rate, 0.005).

^gAccording to the Mann-Whitney U test the decreasing cell number with increasing NO concentration is significant (error rate, 0.05).

^hND, not detectable.

The effect of nitric oxide was investigated by using a similar experimental design, but the reactors were gassed with a gas mixture that consisted of argon (95%), carbon dioxide (5%), and NO. The NO concentration was increased every 2 weeks until the final value of 1,000 ppm was reached after 16 weeks (Fig. 2). The specific ammonia consumption activity remained unchanged in the presence of concentrations of NO between 0 and 600 ppm. In contrast, the specific nitrite consumption activity increased significantly as the NO concentration was increased. As a consequence, the NO₂⁻/NH₃ ratio increased (Table 1). Furthermore, increasing amounts of NO were consumed (Fig. 2). Strong inhibition of the specific ammonia, nitrite, and NO consumption activities was observed with 1,000 ppm of NO (Fig. 2). The growth rate remained unchanged at 0.003 h⁻¹ in the presence of 0 to 600 ppm of NO but decreased to about 0.001 h⁻¹ in the presence of 1,000 ppm of NO. Throughout the experiments the number of B. anammoxidans cells was controlled and remained 5 × 10⁸ cells ml⁻¹ (Table 2). The number of active Nitrosomonas cells decreased about fivefold (from 5.9 × 10² to 1.1 × 10² cells ml⁻¹). NO consumption (detoxification) by B. anammoxidans was also examined in short-term batch experiments with purified biomass (purity, 99.6% ± 0.2%; concentration, 5 × 10⁸ cells ml⁻¹). In three replicate experiments NO consumption rates of 0, 0.1, and 0.4 mmol of NO g of protein⁻¹ h⁻¹ and ammonia oxidation activities of 2.3, 2.4, and 1.4 mmol g of protein⁻¹ h⁻¹ were measured in the presence of 0, 100, and 1,000 ppm of NO, respectively. Interestingly, the NO consumption rate increased as the NO concentration increased.

FIG. 2.

Specific ammonia (solid bars), nitrite (open bars), and NO (•) consumption rates and nitrate production rates (shaded bars) in a reactor supplemented with different concentrations of NO. According to the Mann-Whitney U test, the specific ammonia consumption activity was constant as the NO concentration was increased up to 600 ppm (error rate, 0.05), and the specific nitrite consumption activity increased (error rate, 0.1). Two replicated experiments were performed.

Influence of Nitrosomonas on anammox reactor systems.

The experiments in which NO₂ was added showed that high numbers of active Nitrosomonas cells can have a strong influence on the N conversion rates in anammox reactors. This influence was studied in laboratory-scale reactors inoculated with a mixed population of B. anammoxidans (about 10⁸ cells ml⁻¹) and N. eutropha (0, 10⁷, 5 × 10⁷, or 10⁸ cells ml⁻¹). The experiments were performed for 2 days in the presence of 200 ppm of NO₂ and an excess of ammonia; nitrite was limiting. The ammonia, nitrite, and NO₂ consumption rates, as well as the nitrate and NO production rates, were determined, and the NO₂⁻/NH₃ ratio was calculated. The results are shown in Table 3. When the anammox system was supplemented with N. eutropha, the conversion of nitrogen compounds changed significantly. Ammonia and NO₂ consumption and NO production increased. Nitrite consumption remained unchanged. The NO₂⁻/NH₃ ratio decreased from 1.3 to 0.82, indicating that Nitrosomonas produced nitrite, which was directly consumed by B. anammoxidans.

TABLE 3.

N conversion rates, NO₂⁻/NH₃ ratios, and N loss in 2-liter reactors^a

No. of N. eutropha cells (ml−1)	Conversion rates (μmol reactor−1 h−1)					NO2−/NH3 ratio	Sp act (μmol of NH4+ g of protein−1 h−1)
	Ammonia	Nitrite	NO2b	NO	Nitrate		B. anammoxidansc	N. eutrophad
0e	141.4	187.9	0	0.01	36.8	1.33	4,423
107	149.9	191.4	14.4	13.5	38.1	1.28	4,579	1,080
5 × 107	183.8	183.2	70.6	71.2	41.3	1.00	4,963	1,462
108	226.2	185.8	129.4	119.1	49.3	0.82	5,525	1,545

^aThe reactor with the anammox biomass was supplemented with 0, 10⁷, 5 × 10⁷, or 10⁸ N. eutropha cells ml⁻¹. The consumption and production rates were calculated on the basis of the substrates added to the reactor and the products leaving the reactor. The standard deviation for five replicate experiments was less than ± 5%.

^bIn sterile control experiments, the NO₂ consumption resulting from the reaction of NO₂ with water was determined. The values obtained were subtracted from the data obtained with cells.

^cThe specific ammonia oxidation activity of B. anammoxidans was calculated on the basis of nitrate production.

^dThe specific ammonia oxidation activity of N. eutropha was calculated on the basis of NO₂ consumption and NO production.

^eThe anammox biomass contained about 2 × 10³ Nitrosomonas cells ml⁻¹.

The anoxic activity of N. eutropha resulted in levels of ammonia and NO₂ consumption and NO production similar to the expected 1:2:1 ratio shown in equation 2 (13). About 50% of the ammonia consumed should have been converted to nitrite. However, in the presence of excess ammonia B. anammoxidans immediately converted this nitrite to N₂.

Only when ammonia was the limiting substrate did B. anammoxidans and N. eutropha have to compete for ammonia. Due to the ammonia limitation, the additional nitrite produced by N. eutropha could not be converted by B. anammoxidans and therefore should have accumulated in the reactor. The N conversion in an ammonia-limited anammox reactor (10⁸ cells ml⁻¹) supplemented with nitrite and 250 ppm of NO₂ is shown in Fig. 3. After 120 h N. eutropha (10⁸ cells ml⁻¹) was added. Before N. eutropha was present, B. anammoxidans consumed all of the ammonia and nitrite added with the fresh medium; the concentrations of these compounds in the reactor were low, and increasing concentrations of nitrate were detected. After N. eutropha was added, the concentration of nitrite increased rapidly (Fig. 3), NO₂ was consumed, and NO was produced (data not shown), indicating an NO₂-dependent ammonia oxidation activity of N. eutropha. The decreasing nitrate concentration indicated that the activity of B. anammoxidans was reduced, which was most likely due to competition with N. eutropha for ammonia. When the ammonia limitation in the reactor was eliminated, the nitrite concentration decreased to less than 10 μM and the nitrate concentration reached about 850 μM (data not shown).

FIG. 3.

Ammonia (▪), nitrite (◊), and nitrate (▴) concentrations in a reactor before (0 to 120 h) and after (120 to 142 h) the addition of N. eutropha. The NO₂ concentration was adjusted to 200 ppm. The cell concentrations were adjusted to 10⁸ B. anammoxidans cells ml⁻¹ and 10⁸ N. eutropha cells ml⁻¹.

B. anammoxidans-N. eutropha community with ammonia and gaseous NO₂ as substrates.

On the basis of equations 1 and 2, we speculated that ammonia and gaseous NO₂ are suitable substrates for a combined Brocadia-Nitrosomonas community. Nitrosomonas seems to be able to produce nitrite, N₂, and NO by combined ammonia oxidation and denitrification (equation 2). Ammonia and the nitrite produced may be further converted to nitrate and N₂ by B. anammoxidans (equation 1). The specific ammonia oxidation activity of B. anammoxidans is dependent on the rate of nitrite production by N. eutropha. Such a coculture was examined in a reactor containing mineral medium (10 mM ammonia). The reactor was inoculated with 10⁸ cells of B. anammoxidans ml⁻¹ and 10⁸ cells of N. eutropha ml⁻¹. The reactor was gassed at a rate of 0.5 liter min⁻¹ with a 95% argon-5% carbon dioxide gas mixture supplemented with 250 ppm of NO₂. Immediately after inoculation ammonia and NO₂ consumption started (Fig. 4). Throughout the experiment nitrate and NO were produced, but nitrite did not accumulate (the concentration was always less than 20 μM). On the basis of NO₂ consumption, the specific anaerobic ammonia oxidation activity of N. eutropha was calculated to be about 1.6 mmol g of protein⁻¹ h⁻¹. The production of nitrate was used to calculate the specific activity of B. anammoxidans (about 0.9 mmol of NH₄⁺ g of protein⁻¹ h⁻¹). In sterile control experiments the chemical formation of nitrite and nitrate by reaction of NO₂ with water was determined. The values obtained in the control experiments were taken into account when the specific activity of B. anammoxidans was calculated.

FIG. 4.

Ammonia (▪), nitrite (◊), and nitrate (▴) concentrations, NO₂ consumption (•) (NO₂ fresh gas minus NO₂ off gas), and NO production (○) (NO concentration in the off gas) in a B. anammoxidans-N. eutropha system with ammonia and NO₂ as the only substrates. The cell concentrations were adjusted to 10⁸ B. anammoxidans cells ml⁻¹ and 10⁸ N. eutropha cells ml⁻¹.

DISCUSSION

The results presented here show that Brocadia and Nitrosomonas are able to coexist under anoxic conditions. While B. anammoxidans uses nitrite as a substrate and Nitrosomonas uses gaseous NO₂ as a substrate, these organisms compete for ammonia. Key elements used to analyze the results of this study are shown in Fig. 5 and in equations 1 and 2 describing anaerobic ammonia oxidation by Brocadia and Nitrosomonas. It is important to note that Nitrosomonas produces nitrite during anaerobic ammonia oxidation, which can serve as a substrate for Brocadia (Fig. 5). The assumption that this additional nitrite is consumed by Brocadia when ammonia is not the limiting substrate was confirmed in this study. As a consequence, the NO₂⁻/NH₃ ratio in such a mixed population changes depending on the metabolic activities of the two ammonia oxidizers. Since this is important for interpretation of the results, it is illustrated below with a theoretical example (equations 3 to 5 [equation 5 is the sum of equations 3 and 4]).

(3)

(4)

(5)

FIG. 5.

Schematic model describing the competition and cooperation between B. anammoxidans and Nitrosomonas under anoxic conditions. (a), ammonia addition (fresh medium); (b), ammonia oxidation by Nitrosomonas; (c), additional nitrite in the reactor due to ammonia oxidation; (d), nitrite addition (fresh medium); (e), anammox reaction; med, medium; rec, reactor. End products are enclosed in boxes.

In an anammox reactor system without active Nitrosomonas cells (equation 3), the NO₂⁻/NH₃ ratio is about 1.3. If actively ammonia-oxidizing Nitrosomonas cells are present (equation 4), the NO₂⁻/NH₃ ratio of the system decreases to about 0.4 (equation 5) when the two organisms contribute equally to ammonia consumption. According to this calculation the presence of ammonia-oxidizing Nitrosomonas cells should be indicated by a decrease in the NO₂⁻/NH₃ ratio. Hence, on the basis of this ratio the contributions of B. anammoxidans and Nitrosomonas to the total ammonia conversion in a mixed population can be calculated. Our results (Table 3) indicate that the NO₂⁻/NH₃ ratio was 0.82 in the presence of 10⁸ cells of B. anammoxidans ml⁻¹ and 10⁸ cells of N. eutropha ml⁻¹. According to equations 3 to 5 B. anammoxidans accounts for about 75% and N. eutropha accounts for about 25% of the total anaerobic ammonia oxidation. The NO₂⁻/NH₃ ratios presented in Tables 1 and 3 provide a strong indication that there is simultaneous anaerobic ammonia oxidation by B. anammoxidans and Nitrosomonas.

When the number of N. eutropha cells in the mixed population supplied with NO₂, ammonium, and nitrite was increased, the total ammonia consumption of the system was stimulated and the NO₂⁻/NH₃ ratio decreased significantly (from 1.33 to 0.82). Furthermore, both NO₂ consumption and NO production were detectable, indicating that anaerobically ammonia-oxidizing N. eutropha cells were present. When N. eutropha oxidized ammonia, nitrite was produced, but nitrite did not accumulate in the reactor. It is very likely that B. anammoxidans consumed the additional nitrite, resulting in a higher specific ammonia oxidation activity, which was also indicated by an increased rate of production of nitrate (Table 3 and Fig. 5). When ammonia was the limiting substrate (Fig. 3), B. anammoxidans was not able to adapt its activity to the increased nitrite supply (nitrite in the fresh medium plus nitrite produced by Nitrosomonas), and therefore, nitrite accumulated in the system.

The results shown in Table 3 were used to calculate the anaerobic ammonia oxidation activities of B. anammoxidans and N. eutropha. Nitrate production is directly correlated to the activity of B. anammoxidans (equation 1), because N. eutropha is unable to produce or consume nitrate. According to the nitrate production data (Table 3), the specific anaerobic ammonia oxidation activity of B. anammoxidans ranged from 4.4 to 5.4 mmol of NH₄⁺ g of protein⁻¹ h⁻¹. On the basis of NO₂ consumption and NO production, a specific activity for N. eutropha of between 1.1 and 1.5 mmol of NH₄⁺ g of protein⁻¹ h⁻¹ was calculated. Interestingly, the specific anaerobic ammonia oxidation activity of N. eutropha in this mixed culture was about 10 times higher than that measured in a pure culture (12). In pure cultures, NO₂ concentrations of more than 50 ppm were toxic for N. eutropha. In anammox systems NO₂ concentrations as high as 250 ppm were not inhibitory. The possibility that N. eutropha cells in a mixed population are supplied with higher NO₂ concentrations might explain the high anaerobic ammonia oxidation activity of such cells.

Simultaneous anaerobic ammonia oxidation activities of B. anammoxidans and N. eutropha were also detectable when ammonia was the limiting substrate (Fig. 3). An anammox system was grown with low ammonium and nitrite concentrations in the reactor. Nitrate was produced and accumulated in the medium. When N. eutropha (10⁸ cells ml⁻¹) was added, the consumption of NO₂ and the production of nitrite and NO indicated that there was anaerobic ammonia oxidation activity by N. eutropha. Since the ammonia supply was not increased, B. anammoxidans and N. eutropha had to compete for this limited resource. As a result of the competition, the specific activity of B. anammoxidans decreased, as indicated by a decreasing nitrate concentration.

In an anammox reactor system without added N. eutropha cells (Fig. 1 and 2) the total activity of the system increased significantly when up to 50 ppm of NO₂ was added (Fig. 1). A further increase in the NO₂ concentration led to decreases in activity. The growth rate of B. anammoxidans increased only 1.4-fold, while the specific ammonia oxidation activity increased about 2-fold (Fig. 1) when the NO₂ concentration was increased from 0 to 50 ppm. Perhaps parts of the reducing equivalents generated during ammonia oxidation are necessary for the conversion (reduction) of NO₂. In contrast to the stable number of B. anammoxidans cells (the cell number was adjusted to about 5 × 10⁸ cells ml⁻¹ every 3 days), the number of Nitrosomonas cells increased from about 10³ to 10⁷ cells ml⁻¹ (Table 2) (the growth rate was about 0.005 h⁻¹). The ammonia oxidation activity of Nitrosomonas provided the system with additional nitrite, which was consumed by B. anammoxidans. Obviously, the increased specific activities of both B. anammoxidans (more nitrite available) and Nitrosomonas (NO₂ available) led to higher N conversion activities in the reactor (Fig. 1) (0 to 50 ppm of NO₂). Apparently, concentrations of NO₂ higher than 50 ppm were inhibitory for B. anammoxidans in this system (Fig. 1). The cell number (Table 2) and the specific activity (Table 3) of Nitrosomonas were too low to compensate for the NO₂-inhibited activity of B. anammoxidans (NO₂ concentrations greater than 50 ppm). When the reactor system was incubated without supplemental NO₂, Nitrosomonas was not able to oxidize ammonia (12, 14), and the conversion rates of the N compounds were dependent only on the activity of B. anammoxidans.

The situation was less complex when NO was added. The activity of the anammox system was stable at NO concentrations between 0 and 600 ppm. At higher NO concentrations the activity decreased. Under anoxic conditions in the presence of NO, Nitrosomonas was not able to oxidize ammonia (12). As a consequence, Nitrosomonas did not contribute to the ammonia oxidation activity of the system. Therefore, the N conversion rates were dependent only on the activity of B. anammoxidans. It is interesting that higher NO concentrations added to the anammox system led to increased specific NO consumption activity of B. anammoxidans (Fig. 2). The specific nitrate and NO production activities increased almost parallel to the specific nitrite consumption activity (Fig. 2). Since part of the nitrite oxidation activity provides B. anammoxidans with reducing equivalents (22), it seems likely that the electrons derived from nitrite were used to reduce NO. This might have been possible via the hydroxylamine oxidoreductase, which has some NO-reducing capacity, or via the cytochrome c nitrite reductase, producing ammonia (11). Another possibility is disproportion of NO to N₂O and nitrite.

The final experiment in this study showed that B. anammoxidans and Nitrosomonas formed stable cocultures under anoxic conditions when they were supplemented with a surplus of ammonium and gaseous NO₂ (Fig. 4). Under these conditions, ammonia was oxidized by N. eutropha to dinitrogen and nitrite. The nitrite produced was further converted by B. anammoxidans with ammonia as an electron donor. The model for this process is shown in Fig. 5. The ammonia oxidation activity of N. eutropha was indicated by the NO₂ consumption and NO production profiles. The activity of B. anammoxidans, which was dependent on the nitrite production of N. eutropha, was indicated by the nitrate production rate. Under anoxic conditions N. eutropha converts only 50 to 60% of the ammonia consumed to nitrite (12). Because of this nitrite limitation the specific ammonia oxidation activity of B. anammoxidans (0.9 mmol g of protein⁻¹ h⁻¹) was lower than the specific ammonia oxidation activity of N. eutropha (1.6 mmol g of protein⁻¹ h⁻¹).

The results of this study provide strong indications that the anaerobic ammonia-oxidizing planctomycetes (B. anammoxidans) are not sensitive to NO concentrations up to 600 ppm and that the N conversion rates of an anammox reactor system increase about twofold in the presence of 50 ppm of NO₂. Further experiments demonstrated that addition of NO₂ leads to increases in the number of cells and the specific ammonia oxidation activity of Nitrosomonas under anoxic conditions. Although the two groups of ammonia oxidizers compete for ammonia, cooperation also seems to be possible when the NO₂-dependent ammonia oxidation of Nitrosomonas supplies B. anammoxidans with nitrite as an oxidant. This might have ecological relevance. On the one hand, the ammonia oxidation of B. anammoxidans is restricted to anoxic environments. On the other hand, Nitrosomonas-like microorganisms need an oxidizing agent (O₂ or NO₂) that is available only under oxic conditions. The oxic-anoxic interface might be a suitable environment for both groups of ammonia oxidizers. The availability of ammonia and nitrite is decisive for coexistence. At the oxic-anoxic interface the ammonia oxidation of Nitrosomonas-like microorganisms might be the major source of the nitrite necessary for the nitrite-dependent anammox metabolism. The products of this cooperation are mainly N₂ and small amounts of nitrate. When ammonia is the limiting substrate, the two groups of ammonia oxidizers compete for ammonia (Fig. 3), and the specific affinities for ammonia might be decisive for the outcome of the competition.