Making waves: Harnessing anammox bacteria coupled with dissimilatory nitrate reduction to ammonium for sustainable wastewater management

Highlights

• •Nitrification-partial DNRA coupled anammox, PNA, and PdNA were compared firstly.
• •Anammox bacteria-mediated partial DNRA is the key in this nitrogen removal system.
• •Partial DNRA-anammox has thermodynamic advantages in oxidizing NH₄⁺ than that of anammox alone.
• •Partial DNRA-anammox appears to be feasible for energy-positive wastewater treatment.

Abstract

Anaerobic ammonia oxidation (anammox) which converts nitrite and ammonium to dinitrogen gas is an energy-efficient nitrogen removal process. One of the bottlenecks for anammox application in wastewater treatment is the stable supply of nitrite for anammox bacteria. Dissimilatory nitrate reduction to ammonium (DNRA) is a process that converts nitrate to nitrite and then to ammonium. Significantly, it has been reported that some anammox bacteria can perform DNRA by reducing nitrate to nitrite and ammonium nitrogen with little low-molecular-weight organic acids such as volatile fatty acids. Here, we propose an innovative nitrogen removal process, i.e., nitrification and anammox coupled with partial DNRA (i.e., NPDA), and make a theoretical comparison with previously accepted partial nitrification and anammox (PNA) and partial denitrification and anammox (PdNA) for nitrogen removal. Under similar conditions of oxygen consumption, removal efficiency, external carbon source addition, and greenhouse gas emission, the novel NPDA process can better facilitate resource-effective and environment-friendly wastewater treatment. Thermodynamic analysis indicates that partial DNRA-anammox appears to be preferred, oxidizing per mole of NH₄⁺produces higher energy gain than that of conventional anammox alone. The carbon source limitation rather than nitrate limitation is the key to the realization of NPDA process. This perspective highlights the positive role of DNRA for sustainable wastewater management.

Graphical abstract

1. Nitrite supply for anammox bacteria

Previous studies considered that heterotrophic denitrification was the main nitrogen removal pathway in natural habitats (Xia et al., 2018). However, the discovery of anaerobic ammonia oxidation (anammox) broke this long-held view and suggested that anammox process can be the main pathway for the removal of reactive nitrogen (Nr) in wetlands, soils and marine environments (Li, et al., 2021; Wang et al., 2020). Anammox refers to the microbial anaerobic oxidation of ammonium (NH₄⁺) by nitrite (NO₂⁻), which forms dinitrogen gas (N₂) and nitrate (NO₃⁻) (Eq.1).

$$\begin{array}{c}\hfill {\text{NH}}_4^{+}+{1.32\text{NO}}_2^{-}+0.13{\mathrm{H}}^{+}+{0.066\text{HCO}}_3^{-}\to 2.03{\mathrm{H}}_2\mathrm{O}+0.066{\mathrm{CH}}_2{\mathrm{O}}_{0.5}{\mathrm{N}}_{0.15}+1.02{\mathrm{N}}_2+0{.26\text{NO}}_3^{-}\end{array}$$ (1)

Anammox-based nitrogen removal saves up to 100 % of organic carbon source requirements compared to conventional heterotrophic denitrification. However, one of the bottlenecks for anammox application is the lack of stable supply of nitrite for anammox bacteria (Adams et al., 2025). To feed nitrite to anammox bacteria, except for the direct addition of nitrite, partial nitrification (PN, NH₄⁺ → NO₂⁻) and partial denitrification (PdN, NO₃⁻ → NO₂⁻) are two previously recognized options, where nitrite is an intermediate product of both processes (Cao et al., 2019; Liu et al., 2024). As a consequence, coupling either PN with anammox (i.e., PNA) or PdN with anammox (i.e., PdNA) for nitrogen removal has been widely studied in wastewater treatment at bench- and pilot scales (Zhang et al., 2021; Zheng et al., 2023). Nevertheless, for PNA, the long-term stability and total nitrogen removal efficiency (TNRE) of PNA has been questioned due to the difficulty in inhibiting diverse nitrite-oxidizing bacteria (NOB) found in wastewater systems as well as the final products of PNA still include nitrate, which cannot be subsequently utilized, leading to a significant nitrate accumulation in the PNA process (Su et al., 2023; Wang et al., 2021a). For PdNA, 100 % complete TNRE is theoretically achievable, but its nitrite dependent on nitrate reduction in the presence of carbon source (Izadi et al., 2023). However, anammox activity can be suppressed at a high carbon/nitrogen (C/N) ratio yet denitrifiers become limited at a low C/N ratio. This means that a feasible operational boundary for the PdNA process may not be easily achieved. Providing stable nitrite supply for anammox bacteria in advanced anammox-based wastewater treatment processes is therefore a crucial task.

2. Dissimilatory nitrate reduction to ammonium mediated by anammox bacteria

Dissimilatory nitrate reduction to ammonium (DNRA) is known as an important metabolic pathway of nitrate reduction in addition to denitrification, in which DNRA bacteria use organic or inorganic substrates as electron donors to reduce nitrate to nitrite and then to ammonium (NO₃⁻→NO₂⁻→NH₄⁺). Recent studies revealed that some anammox bacteria (such as Candidatus ‘Kuenenia stuttgartiensis’ and Candidatus ‘Brocadia fulgida’) can perform DNRA by utilizing low-molecular-weight organic acids such as volatile fatty acids (VFA) (Feng et al., 2018; Kartal and Leltjens, 2016; Wang et al., 2022). In the absence of exogenous NH₄⁺, anammox bacteria convert NO₃⁻ to NO₂⁻ and NH₄⁺ through whole DNRA, and then utilize the formed NO₂⁻ and NH₄⁺ for anammox (Wang et al., 2021b); in the presence of exogenous NH₄⁺, the second step of the DNRA reaction, the conversion of NO₂⁻ to NH₄⁺, is more difficult compared to the conversion of NO₃⁻ to NO₂⁻ (Ji et al., 2021), which provides the possibility for NO₂⁻ accumulation. Therefore, it appears feasible to apply such anammox bacteria to perform partial DNRA and simultaneously meet nitrite supply requirements in anammox-based nitrogen removal processes, for which we here propose nitrification and anammox coupled with partial DNRA (i.e., NPDA).

3. Comparison of NPDA with conventional PNA and PdNA for nitrogen removal

Fig. 1 illustrates a comprehensive assessment of a mainstream wastewater treatment process with the proposed NPDA in comparison to PNA and PdNA under the same influent and operating conditions. All three systems use a high rate activated sludge (HRAS) system for COD recovery, and the main reactors and process layouts are essentially the same, but COD and nitrogen fluxes are transferred in different ways. The assessment is conducted based on a treatment capacity of 100,000 m³ domestic wastewater per day, containing typical 400 mg/L chemical oxygen demand (COD) and ammonium nitrogen of 40 mg/L.

Fig. 1.

(A) Schematic diagram of mainstream nitrogen removal system for a sustainable municipal wastewater treatment with NPDA using anammox bacteria. DNRA represents dissimilatory nitrate reduction to ammonium. (B&C) Mainstream nitrogen removal system with PNA and PdNA under the same conditions. Taking a full-scale wastewater treatment plant with a flow rate of 100,000 m³/d as an example, the initial concentrations of NH₄⁺-N and COD are 40 mg/L and 400 mg/L, respectively. HRAS represents high rate activated sludge system, which recovers approximately 60 % of the organic matter, according to Gulhan et al. (2023). PdN represents partial denitrification.

Compared to PNA, in the presence of organic carbon, both NPDA and PdNA processes reduce NO₃⁻ to NO₂⁻, and then realize nitrogen removal by anammox process with NH₄⁺. The NO₃⁻ generated by anammox process can be further used as sbustrate for partial DNRA or PdN, which suggests that the NPDA and PdNA can theoretically achieve 100 % TNRE. The PNA system, on the other hand, has a maximum TNRE of only 89 % due to the accumulation of nitrate as a product of anammox reaction.

Oxygen supply is the major energy consumption in wastewater treatment. The NPDA and PdNA process consumes 22,700 kg of oxygen which is slightly higher than 22,460 kg of oxygen consumption in the PNA theme. The reason is that the NPDA and PdNA processes require complete nitrification to obtain nitrate, whereas the PNA requires only PN to obtain nitrite. In addition, NPDA (1150 kg/d) has the lowest biomass production (Fig. 1A), followed by PNA (1797 kg/d) (Fig. 1B), while PdNA has a high biomass production of 3850 kg /d (Fig. 1C). There are two main reasons for this, one is that some of the COD in the influent bypasses the nitrification unit without being oxidized in the NPDA and PdNA processes; and the other is that in a NPDA system, anammox bacteria tend to oxidize organic matter to CO₂ via the acetyl CoA pathway rather than the heterotrophic bacteria needing to assimilate a portion of the COD into biomass, as in PdNA system (Kartal and Leltjens, 2016; Lawson et al., 2021). Therefore, compared to the PNA and PdNA process, the surplus sludge production of NPDA system can be kept at a lower level.

Although both the NPDA and PdNA require carbon source, most previous studies on PdNA process demonstrated that the best removal performance was achieved when the influent COD/NO₃⁻ ratio was maintained at 3. Since anammox bacteria do not derive energy through its mediated DNRA pathway, partial DNRA performed by anammox bacteria tends to occur at a lower C/N ratio (a C/N ratio of 1.2 was sufficient to meet the requirements, or even lower) (Castro-Barros et al., 2017). This gives anammox bacteria an advantage in competing with denitrifiers at low C/N ratio. With the same HRAS capture efficiency, before the stream flows into the partial DNRA-anammox process, the content of NO₃⁻ is 2000 kg N/d and the content of COD is 2440 kg COD/d (Fig. 1A). This C/N ratio adequately fulfills the requirement of partial DNRA-anammox process. Whereas to meet the requirement of C/N ratio of 3 for PdNA (Fig. 1C), it would rely on an additional carbon source of approximately 3560 kg COD/d to maintain an efficient partial denitrification and microbial activity. If sodium acetate is used as an additional carbon source to meet the COD requirement, 4561 kg/d of sodium acetate would need to be added, at a cost of approximately $3910 per day.

Nitrous oxide (N₂O) is a potent greenhouse gas and a by-product and intermediate of nitrification and denitrification in biological nitrogen removal processes. Factors affecting the release of N₂O are complex, and therefore, the emission of N₂O was not calculated in Fig. 1. However, based on previous reports, it is known that N₂O emission was about 2.6 %−6.6 % of total influent nitrogen in a two-stage PNA system, with a median value of 5.11 % (Liu et al., 2023). While the anammox stage accounted for only 4 % of all N₂O production in the PNA system, with the most of N₂O resulting from the accumulation of hydroxylamines and nitrite in PN (Hausherr et al., 2022). The N₂O emission factor range of the PdNA process was 0.22 % to 2.80 %, with a median of 1.3 %, which was four times lower than that of the two-stage PNA process (Liu et al., 2023). In the context of the NPDA system, since complete nitrification produces less N₂O than PN and different from denitrification, N₂O is not reported to be an intermediate product of DNRA (Aalto et al., 2021), which significantly reduces the greenhouse gas N₂O emission.

4. Thermodynamic perspective

In general, microorganisms are able to harvest more energy per unit of substrate consumed, which in principle represents a competitive advantage (Kästner et al., 2024). In cultures of non-H₂-forming bacteria, the concentration ratio of products to exogenous substrates is generally ≥ 10⁻² at the beginning of growth phase and ≤ 10² towards the end. The effectively utilizable energy of a metabolic process △G^’ thus deviates by at most 11.7 kJ/mol from its standard free energy △G⁰ (Thauer et al., 1977). Therefore, the △G⁰ was calculated from the standard free energy of information at available for compounds of biological interest to assess the thermodynamics of the following biological processes (Table 1). For the typical nitrate reduction process (denitrification, DNRA, and partial DNRA/denitrification), the denitrification has the highest energy production capacity (△G⁰ = −99.56 kJ) and the DNRA process is the most unfavorable process (△G⁰ = −62.51 kJ) based on the energy required to transfer each electron. However, the energy released in NO₃⁻ reduction to NO₂⁻ (△G⁰ = −69.05 kJ) is higher compared with the reduction to NH₄⁺ (△G⁰ = −62.51 kJ) while the same electron is transferred in the process. As such, partial DNRA appears as favored in the systems. In addition, comparing the substrate consumed per unit of various biological processes shows that the theoretical Gibbs energy consumed per mole of NH₄⁺ by partial DNRA-anammox can produce higher energy gain than that of conventional anammox alone. This result suggests that partial DNRA-anammox in two steps can be more thermodynamically driven in oxidizing NH₄⁺. Although the obtained energy per mole of NO₃⁻ consumed by denitrification, DNRA, partial DNRA-anammox, and complete DNRA-anammox is similar, the energy of 1983.56 kJ generated by partial DNRA-anammox per mole of acetate consumed exceeds that of denitrification and complete DNRA-anammox. This means that the limitation of acetate rather than nitrate is the key to achieving partial DNRA-anammox in a mainstream anammox systems.

Table 1.

Catabolic reaction stoichiometry, standard Gibbs free energy change, electron transferred, and stoichiometric ratio COD/NO₃⁻ for different pathways (taking CH₃COO⁻ as an example).

Biological process	Reaction	e- transf.	△G0 (kJ/mol/K) a	△G0 (kJ/e−/K)	△G0 (kJ/mol NH4+/K)	△G0 (kJ/mol NO3−/K)	△G0 (kJ/mol CH3COO−/K)	Stoichiometric COD/NO3−-N (g COD (gN)−1)
Denitrification	8NO3− + 5CH3COO− + 8H+ → 5CO2 + 4N2 + 5HCO3− + 9H2O	40	−3982.18	−99.56	/	−497.77	−796.44	2.86
DNRA	NO3− + CH3COO− + 2H+ → CO2 + NH4+ + HCO3−	8	−500.09	−62.51	/	−500.09	−500.09	4.57
Anammox	NO2− + NH4+ → N2 + 2H2O	3	−357.82	−119.27	−357.82	/	/	/
Partial DNRA/Denitrification	4NO3− + CH3COO−→ CO2 + 4NO2− + HCO3− +H2O	8	−552.42	−69.05	/	−138.11	−552.42	1.14
Partial DNRA+ anammox b	4NO3− + CH3COO− +4NH4+→ CO2 + 4N2 + HCO3− + 9H2O	20	−1983.56	−99.17	−495.89	−495.89	−1983.56	1.14
Complete DNRA + anammox c	8NO3− + 5CH3COO−+ 8H+→ 5CO2 + 5HCO3− + 4N2 + 9H2O	40	−3983.94	−99.60	/	−497.99	−796.79	2.86

^aConsidering the condition of pH 7 and 298.15 K. Liquid concentration 1 M and 1 bar for gas components (N₂ and CO₂).

^bThe NH₄⁺ comes from the environment.

^cThe NH₄⁺ comes from the full DNRA process.

5. How to stimulate anammox bacteria to perform DNRA?

The anammox bacteria driven-DNRA is difficult to proceed under normal autotrophic circumstances (Kartal and Leltjens, 2016). Small-molecule organic acids can stimulate anammox bacteria to perform DNRA but also promote the proliferation of heterotrophic microorganisms. Most DNRA process could be done by specialist or non-specialist heterotrophs or autotrophs, which could compete with anammox bacteria that most likely only using DNRA as a surviving strategy. Therefore, it is critical to inhibit heterotrophic microorganisms in the NPAD process.

As described in Section 4, Gibbs free energy analysis shows that limiting the C/N ratio enhances the competitive advantage of partial DNRA-anammox. The results of batch tests using NH₄⁺, NO₃⁻ and CH₃COO- as substrates (C/N ratio ≈ 1.2) showed that the contribution of anammox bacteria to nitrate reduction was up to 68.1 % (Information from Zhao et al., unpublished data). Transcriptome analysis showed that the DNRA pathway was activated in mixotrophic anammox consortium (C/N ratio = 0.3 or 0.5) (Feng et al., 2018). In addition, previous studies have shown that anammox bacteria have different metabolic potential for nitrate and VFA. For example, Candidatus Jettenia showed the highest nitrate reduction activity, followed by Candidatus Brocadia and Candidatus Scalindua, while Candidatus Kuenenia had the poor nitrate reduction ability (Naufal and Wu, 2024). Acetate could significantly up-regulate the acetyl-CoA pathway in Candidatus Brocadia, but had no stimulation on Candidatus Jettenia (Feng et al., 2019). Therefore, keeping specific anammox bacteria with high DNRA metabolic potential is also one of the strategies to achieve the stability of NPDA system.

Electron shuttle is a chemical substance that can mediate electron transfer through autoxidation and reduction, such as humus, carbonaceous materials, and anthraquinone-2,6-disulfonate (Sun et al., 2023). They can accelerate the transfer of electrons from microorganisms to the outside of the cell. It has been hypothesized the electron shuttles may enhance the nitrogen removal efficiency by strengthening the DNRA pathway in anammox bacteria (Wang et al., 2022). With the addition of biochar, anammox bacteria could achieved up to 93.7 % of TNRE at very low C/N ratios (0.1-0.5). Anammox bacteria have extracellular electron transfer capacities (Zhen et al., 2024). In this case, biochar acted as an electron shuttle to improve the efficiency of extracellular electron transfer of anammox bacteria. Easily available electrons can enhance the expression of DNRA-related genes, thereby improving the TNRE. Therefore, selective stimulating DNRA function mediated by anammox bacteria may be a suitable strategy. Whether other electron shuttles similarly use this pathway to enhance the efficiency in anammox systems remains to be further explored.

In sum, the new NPDA concept appears to be a promising path forward for promoting anammox application aiming at sustainable wastewater management. In contrast to the generally accepted viewpoint, the positive role of DNRA mediated by anammox bacteria is key and should be more focused on in future research. The NPDA process may hold engineering application of mainstream anammox in municipal wastewater systems and reduce obstacles associated with its regulation.

CRediT authorship contribution statement

Yiyi Zhao: Writing – original draft, Validation, Conceptualization. Min Zheng: Writing – review & editing. Bing-Jie Ni: Writing – review & editing. Shou-Qing Ni: Writing – review & editing, Funding acquisition.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request