Insight into the mechanism of a novel heterotrophic nitrifying – aerobic denitrifying bacterium, Bacillus subtilis strain H1, and its application potential in mariculture wastewater treatment

The nitrogen removal performance and mechanisms of Bacillus subtilis H1 isolated from a mariculture environment were investigated. Strain H1 e�ciently removed NH 4+ -N, NO 2− -N, and NO 3− -N in simulated wastewater with removal e�ciencies of 85.61%, 90.58%, and 57.82%, respectively. Strain H1 also e�ciently degraded mixed nitrogen and had removal e�ciencies ranging from 82.39–89.54%. Nitrogen balance analysis revealed that inorganic nitrogen was degraded by heterotrophic nitri�cation–aerobic denitri�cation (HN-AD) and assimilation. 15 N isotope tracing indicated N 2 O was the product of the HN-AD process, while N 2 as the �nal product was only detected during the reduction of 15 NO 2− -N. The nitrogen assimilation and dissimilation pathways by strain H1 were further clari�ed using complete genome sequencing, nitri�cation inhibitor addition, and enzymatic activity measurement, and the ammonium oxidation process was speculated as NH 4+ → NH 2 OH → NO → N 2 O. These results showed the application prospect of B. subtilis H1 in treating mariculture wastewater.


Introduction
The marine nitrogen cycle occupies an important position in the geo-biochemical nitrogen cycle (Shi et al., 2020).However, nitrogenous compounds from the aquaculture industry transported into marine ecosystems have been increasing over the years (Yun et al., 2019).Excess nitrogen in the forms of ammonium, oxidized nitrogen (e.g.nitrate, nitrite, etc.), and organic nitrogen produced by aquaculture activities not only disturb the natural equilibria of ecosystems but threaten the health of aquatic animals (Gao et al., 2014).
The biological nitrogen removal treatment with the advantages of high-effect and environmental friendliness is more popular for eliminating nitrogen pollution from the aquatic environment.This approach is mainly based on the different physiological characteristics and ecological functions of the microbes that participate in environmental nitrogen cycling to establish various microbial-nitrogen transformation processes (Zhu et al., 2015).The heterotrophic nitri cation-aerobic denitri cation (HN-AD) approach as the key to microbial-mediated nitrogen removal method for aquaculture systems plays an important role in treating nitrogen overload wastewater, it was gradually developed after the discovery of those heterotrophic bacteria which conducted simultaneously nitrify and denitrify within one reactor (Shu et al., 2022;Li et al., 2017).There two HN-AD patterns of converting ammonium (NH 4 + -N), nitrite (NO 2 − -N), and nitrate (NO 3 − -N) into nitrogenous gases have been described in Richardson's hypothesis (Richardson and Watmough, 1999).One exhibit NH 4 + -N was oxidized to NO 2 − -N or NO 3 − -N via hydroxylamine (NH 2 OH) and then convert to N 2 (Cui et al., 2021;Zhang et al., 2020a), while the other shows that NH 2 OH is directly oxidized to nitrogenous gases, without the intermediates of denitri cation (NO 2 − and NO 3 − ) (Zhao et al., 2012).The bacterial assimilation also makes a great contribution to nitrogen removal in aquatic ecosystems, apart from being removed from the system via nitri cation and denitri cation pathways (Li et al., 2021).During the process of nitrogen assimilation, NH 4 + is directly absorbed into cells for the synthesis of biomacromolecules, while NO 3 − and NO 2 − must be reduced to NH 4 + by assimilatory nitrate reductase and assimilatory nitrite reductase before being absorbed (Ren et al., 2014).
Currently, many functional bacteria from domestic or industrial wastewater that performed e cient nitrogen removal capacity have been isolated and fully explored, while due to the mariculture environment with the characterization of high dissolved oxygen, salinity, and containing a relatively low concentration of dissolved nitrogen (total nitrogen < 20 mg/L) those bacteria may not be suitably applied in aquaculture environments (Zhang et al., 2020b;Qi et al., 2009).Therefore, it is necessary to screen out indigenous strains as candidates for treating aquaculture wastewater.Recently, several functional bacteria that possess exible nitrogen metabolic pathways (such as nitri cation, bacterial assimilation and dissimilation) and exhibited excellent nitrogen removal performance, have been screened out and applied in treating mariculture wastewater, those of which play an important role in maintaining nitrogen balance in the aquatic ecosystem (Bai et al., 2019;Huang et al., 2017).A previous study reported that Bacillus subtilis H1 isolated from an aquatic ecosystem performed excellent nitrogen removal capacity across various environmental conditions (Zhao et al., 2019), but it exhibited no bacterial proliferation or nitrogen degradation without the addition of organic carbon (data are not listed).In this paper, strain H1 was cultured in synthetic wastewaters with individual and combined nitrogen sources to investigate the characteristics of bacterial nitrogen removal.Moreover, the nitrogen metabolic mechanisms were explored by methods of enzyme activity assays, nitrogen metabolites analysis, and nitrogen balance calculation.To validate the nitrogen removal mechanisms, the functional genes were identi ed by sequencing and characterizing the genome of B. subtilis H1.These results can be used to characterize the bacterial nitrogen metabolic behavior in mariculture system, and give theoretical support for B. subtilis H1 in treating mariculture wastewater.

Assessment of nitrogen removal of B. subtilis H1
Nitrogen removal performance was measured in the different mediums containing individual (SNM-1, SNM-2, SNM-3, and SNM-4 ) and combined (CNM-1, CNM-2, and CNM-3) nitrogen sources.Strain H1 was pre-cultured in 2216E medium for 20 h, and the bacterial suspension was obtained and inoculated (v/v 5%) into a conical ask containing 200 mL nitrogen-containing synthesized wastewater (NSW) with a shakeing speed of 160 rpm (DO = 6.27 ± 0.09 mg/L) at 28°C, three parallel samples were conducted for each treatment.Bacterial cultivation was periodically taken out to monitor the concentrations of bacterial growth (OD 600 ), total inorganic nitrogen (TIN), total nitrogen in the system (TN), NH 4 + -N, NO 2 − -N, NO 3 − -N, NH 2 OH-N, and loss nitrogen (loss-N).

Addition of inhibitors allylthiourea (ATU)
Bacteria in the logarithmic growth phase were inoculated at 5% (v:v) into a conical ask containing 200 mL of SNM-1 for 24 h of incubation at 160 rpm and 28°C.The bacterial incubation in SNM-1 with/without the addition of 0.3 mM inhibitor allylthiourea (ATU) was investigated (Daims et al., 2015), and three parallel samples were conducted for each treatment.Bacterial cultivation was taken out per 6 h to monitor the OD 600 and concentrations of nitrogen.

Determination of gaseous nitrogen
SNM-1*, SNM-2*, and SNM-3* were formulated using 15 N-labelled 15 NH 4 Cl, Na 15 NO 2 , and Na 15 NO 3 (Sigma Corp., USA) as sole nitrogen instead of NH 4 Cl, NaNO 2 , and NaNO 3 in the NSW, respectively.The strain was pre-cultured in 2216E and further inoculated (v/v 5%) into 100 mL of 15 N-labelled NSW in 0.5 L serum bottles containing and sealed with a rubber stopper, the uninoculated medium served as a control.
Each treatment with the cultivated condition of 160 rpm and 28°C for 72 h and three parallel samples were set up.Gaseous nitrogen compounds ( 15 N 2 O and 15 N 2 ) in the headspace of the bottles were monitored using GC-MS (Thermo Fisher Finnigan DELTA V Advantage, USA).The calculation of relative 15 N isotopic abundance (δ 15 N) was followed by Denk et al. (2017).

Enzyme activity assay
The pre-cultured bacterial suspension was respectively inoculated in different SNMs, and the cells were collected after inoculation of 12 hours.After being washed and resuspended with 10 mM PBS buffer (pH 7.5), the collected cells were fragmented by an ultrasonicator (SCIENTZ-IID, SCIENTZ, China), and the cellfree extracts were further obtained from lysed cells.The reaction system (1mL) for ammonia monooxygenase (AMO) assay contained 1 mM NH 4 + -N, 10 mM phosphate buffer, 0.02 mM NADH, and 200 µL enzyme extract was prepared.The reaction system (1 mL) used to determine hydroxylamine oxidoreductase (HAO) activity contained sterile water, 10 mM Tris-HCl buffer, 1 mM NH 2 OH, and 200 µL enzyme extract.The reaction system (1 mL) used for determined nitrate reductase (NR) and nitrite reductase (NIR) activities contained sterile water, 50 mM phosphate buffer, 100 µM NaNO 3 /NaNO 2 , and 200 µL enzyme extract.The speci c activities were determined based on the loss of nitrogen-containing substrates, and three parallel samples were conducted for each treatment (Crossman et al., 1997).Protein concentrations were measured by the previous method (Lowry et al., 1951).

Gene annotation and pathway analysis
The sequencing library was prepared using Illumina and PacBio by Huada Gene Technology Co. Ltd (Shenzhen, China).After sequencing, the raw data produced by HiSeq and PacBio RSII were ltered using the FastQC and NanoPack software to obtain high-quality clean data and high-quality subreads.The nal spliced sequences were obtained by applying the HGAP software and Bowtie2 software to lter the second-generation HiSeq data and the third-generation PacBio data for splicing and assembly, and then gap lling.The complete genome sequence data of B. subtilis H1 was obtained and uploaded, its accession number in NCBI GenBank was CP026662.The Prokka software, online comparison of BLASTn (http://blast.ncbi.nlm.nih.gov/Blast.cgi) and RAST (https://rast.nmpdr.org/)were used to annotate the predicted genes based on databases of Nr, KEGG, and COG.Circos software was used to draw the circular genome map of strain H1 to clarify the relationship between genome composition and its location.

Analytical measurements and data analysis
Measurement of bacterial growth concentration (OD 600 ) by using a spectrophotometer (DU800, BECKMAN COULTER, USA) (Li et al., 2013).The detection of TN, NH 4 + -N, NO 3 − -N, and NO 2 − -N in the cultures following their respective standard methods (Zhao et al., 2019).The previous method was used for detecting NH 2 OH (Frear and Burrell, 1955).The loss-N was calculated as the reduction of the initinal TN.The nitrogen removal rate and e ciency were computed using the calculation formulas (Rout et al., 2017).TBM SPSS 19.0 and Origin 2019 were used for analyzing and drawing the experimental data.ANOVA and t-test were used for comparing the signi cance of the difference in experimental results.

Ammonium removal characterization of B. subtilis H1
It is necessary to eliminate excess NH 4 + -N from aquatic ecosystems because of the ionized ammonia (NH 3 ) with toxicity for aquatic animals (Deng et al., 2021).The ammonium removal performance of B.
subtilis H1 was assessed in SNM-1.As shown in Fig.

Nitrogen transformation performance with multiple nitrogen sources
Excess inorganic nitrogen in aquatic ecosystems is normally present in multiple nitrogen forms (Zhang et al., 2020b), thus, the multiple nitrogen transformation performance of strain H1 was further evaluated.As previously reported, NH 4 + -N as one of the nitrogen sources was preferred utilizing by bacteria to conserve energy (Huang et al., 2019).Therefore, strain H1 preferred removing NH 4 + -N alone rather than using the multiple nitrogen sources synchronously when NH 4 + -N, NO 3 − -N, and NO 2 − -N were present in multiple nitrogen forms (Fig. 3).
When both NH 4 + -N and NO 2 − -N serve as nitrogen sources (CNM-1, Fig. 3-a), during 0-24 h of inoculation, with the bacterial growth concentration accumulated, NH 4 + -N and NO 2 − -N sharply decreased to nal concentrations of 2.15 mg/L and 2.28 mg/L at 24 h, respectively, and the average nitrogen removal rates were 0.791 mg/L/h and 0.838 mg/L/h.Although the removal performance of NH 4 + -N and NO 2 − -N by strain H1 cultured in CNM-1 was lower than it was in the treatment of sole nitrogen source, the nal TIN removal e ciency still reached 83.44%, and about 61.45% of TN was eventually lost into the air.This result indicated the HN-AD process in CNM-1 was promoted, despite the bacterial growth and maximum nitrogen removal rate might be inhibited by the dual toxic effects of NH 4 + -N and NO 2 − -N (He et al., 2016b).
The capability of simultaneously utilizing NH 4 + -N and NO 3 − -N of strain H1 in CNM-2 was illustrated (Fig. 3-b).The bacterial biomass in the medium increased to peak at 0.50, after undergoing the lag period (0-12h).With bacterial growth, the NH 4 + -N and NO 3 − -N were utilized with the nal concentrations of 3.29 mg/L and 4.23 mg/L at 72 h, and the corresponding maximum nitrogen removal e ciencies were 84.33% and 80.5%, respectively.During this process, about 82.39% of TIN was eliminated, moreover, about 52.21% of TN was eventually lost into the air.The average removal rate of NH 4 + -N obtained in CNM-2 medium (0.79 mg/L/h) at 24 h was lower than that in SNM-1 medium (1.48 mg/L/h), however, the NO 3 − -N and TIN removal performance in CNM-2 were better than treatments with single nitrogen sources.The results illustrated that strain H1 performed higher NO 3 − -N removal performance with simultaneous utilization of NH 4 + -N.
When strain H1 was cultured in CNM-3 (containing equal amounts of NH 4 + -N, NO 3 − -N, and NO 2 − -N), the dynamic changes of biomass and nitrogens in the medium was shown in Fig. 3-c.With the biomass accumulated, about 100% of NH 4 + -N and 92.48% of NO 2 − -N was degraded by strain H1 at 24 h, respectively.Subsequently, after the two inorganic nitrogen sources had been consumed, the strain began to utilize NO 3 − -N, and about 69.35% of the initial NO 3 − -N was utilized.After 72 h of processing, the nal TIN removal e ciency by strain H1 reached 89.31%, and about 55.07% of the initinal TN was lost.Notably, the corresponding TIN removal concentration (37.43 mg/L) by strain H1 was much higher than the general nitrogen concentration in aquatic ecosystems (TN < 20 mg/L) (Huang et al., 2017).These results demonstrated the application potential of treating multiple inorganic nitrogens by B. subtilis H1 in the aquatic ecosystem.

Exploration of HN-AD pathways in B. subtilis H1
The process of converting NH 4 + -N to NH 2 OH by AMO is believed to be the rst step in the aerobic ammonia oxidation process (Wang et al., 2019).Therefore, ATU as the AMO inhibitor was added to the NH 4 + -N transformation process by B. subtilis H1 (Fig. 4-a).ATU contains a C = S structure that chelates with the AMO active center, this inhibits the nitri cation process but does not affect the nitrogen assimilation process (McCarty, 1999).The result indicated that about 37.3% and 87.0% of NH 4 + -N concentration in the ATU-added and non-ATU-added NSW were removed after 24 h, respectively.
Furthermore, NH 2 OH signi cantly accumulated in the non-ATU-added medium during the entire NH 4 + -N removal process.By comparison, after 12 h of incubation, NH 2 OH appeared in the ATU-added medium when the inhibitory effect of ATU was decreased.The result indicated that the ammonia oxidation process by strain H1 was signi cantly hindered due to the inhibition by AMO activity.
NH 2 OH is the typical intermediate product during the ammonium oxidation process, so the utilization of NH 2 OH by strain H1 was further detected (Fig. 4-b).The OD 600 reached a value of 0.363 and approximately 38.66% of NH 2 OH was eliminated by strain H1, after 24 h of incubation, with 20.12% of nal TN lost into the air.During the entire NH 2 OH oxidation process, neither NO 2 − -N nor NO 3 − -N accumulation was observed (< 0.01 mg/L).Previous research found that many HN-AD bacteria were unable to convert NH 2 OH to intracellular nitrogen (Joo et al., 2005;Zhao et al., 2010a).In this study, partial NH 2 OH was eliminated by strain H1 with bacterial growth, which demonstrated the application potential of strain H1 for treating NH 2 OH-containing wastewater.
Generally, whether N 2 O or/and N 2 are produced depended largely on oxygen tension and the tolerances of various bacterial nitrous oxide reductases (Lloyd et al., 1987).The nal gaseous products of the HN-AD process of strain H1 were clari ed using the 15 N-labeling method.When 15 NH 4 Cl (SNM-1*), Na 15 NO 2 (SNM-2*), and Na 15 NO 3 (SNM-3*) were used as the sole nitrogen source, the δ 15 N of N 2 O in these 15 Nlabeling medium detected as 1556.1%,6663.81%, and 2059.67% were respectively much higher than the blank control (8.68%, 323.36%, and 143.73%) (Online Resource, Fig. S1).Interestingly, the δ 15 N of N 2 was only detected in the headspace of SNM-2* with a higher value (12.85%) than the blank control (3.07%), although the δ 15 N 2 O was detected in all of the 15 N-labeling group (Online Resource, Fig. S1).A similar observation was also depicted in B. methylotrophicus L7, which transformed NO substrates.Of course, the denitri cation process of strain H1 needs to be investigated in greater depth to con rm these results.

Nitrogen metabolism by B. subtilis H1
The analysis of bacterial whole-genome sequencing was obtained to investigate the molecular mechanisms of nitrogen removal by B. subtilis H1.The whole genome of strain H1 only contained a circular chromosome sequence with a total length of 4,356,270 bp and a GC content of 44.16%.When annotated by the Prokka software, RAST online alignment, and BLAST algorithm, about 4686 proteincoding genes were eventually annotated (Fig. 5-a).According to the COG fetch results, the vast majority of the annotated genes (43.51%) were classi ed as metabolic category, and approximately 21.96% belonged to the cellular pathway and signaling classi cation (Online Resource, Fig. S2).The genomic information also was illustrated by creating a circular genome map of B. sutilis strain H1 (Online Resource, Fig. S3).
The total number of 33 putative genes involved in nitrogen metabolism, including nar, nir, nas, gln, and glt, and other functional genes related to the cellular transport and transformation processes in strain H1 (Online Resource, Table .S2).The substrate-binding protein-coding genes amt and narK, expressed in the NH 4 + -N and NO 2 − -N/NO 3 − -N transport system, respectively, facilitated the uptake of nitrogen by strain H1 into the cell for further transformation (Zhu et al., 2019).The gene clusters of assimilated nitrate reductase (nasAB) and membrane-bound nitrate reductase (narGHJI) were annotated, both of which were responsible for converting NO ).The nitrite reductase gene clusters (nirBD) were also found, which sequentially arranged and formed a nas gene cluster with nasAB in the bacterial genome where they can cooperate and complete the assimilation of nitrate (Rajakumar et al., 2008).Meanwhile, intracellular NH 4 + -N assimilated to various nitrogen-containing substances can be accomplished by the expression of gudB (which encodes glutamate dehydrogenase in GDH pathway) and glnA/gltBD (which encodes glutamine synthetase/ glutamate synthase in GS/GOGAT pathway, respectively).Furthermore, the gene tnrA encoding a transcriptional regulator, which activates nitrogen assimilation genes during nitrogen limitation in strain H1, was uncovered (Yoshida et al., 2003).Under the KEGG database, the functional genes of nitrogen assimilation and dissimilation pathways were depicted (Fig. 5-c).
Interestingly, those enzyme-coding genes related to the canonical ammonia oxidation (amoCBA and hao) and denitri cation (nirK/nirS, norBC, and nosZ) were not uncovered through comparisons with the RAST and KEGG databases.Nevertheless, a putative gene coding for AMO with 72.22% similarity of sequences to NG74_01082 and BAPNAU_2723 was successfully annotated using the BLAST tool between 3630881-3632035 bp in the genome.Therefore, combined with the inhibiting result of ammonia oxidation, the ammonia oxidase in strain H1 with the similar active center of AMO (encoded by amoCBA) might be encoded by this novel functional gene.Moreover, as described previously, the reduction of NO 2 − -N to NO as the crucial step in the typical denitri cation pathway was regulated predominantly by nirK/nirS, encoding NO 2 − -N reductase.However, recent studies found that the operon nirBDCcysG was also likely responsible for this reduction (NO 2 − -N→NO) (Li et al., 2021).Thus, the nirBDC gene cluster in the strain H1 genome (Fig. 5-b) not only plays an important role in assimilatory NO  Tables  The growth and nitrogen transformation performance of B. subtilis H1 when ammonium was added as a sole nitrogen source (SNM-1) under aerobic conditions

3. 2 5 (
Nitrate and nitrite removal characterization of B. subtilis H1The B. subtilis H1 exhibited slow growth in SNM-2 because of the inhibition of high NO 2 − -N concentrations, and the bacterial growth reached a steady growth period until after 48 h of inoculation (Fig. 2-a).Strain H1 could convert approximately 90.59% of initial NO 2 − -N during 0-24 h with the average NO 2 − -N removal rate that occurred was 1.58 mg/L/h.The e cient NO 2 − -N removal capacity was better than 0.93 mg/L/h exhibited by Rhodococcus sp.CPZ24 and 1.16 mg/L/h showed by Bacillus cereus GS-Chen et al., 2012; Rout et al., 2017).During NO 2 − -N removal process, high concentrations of nitrite induced the expression of nitrite oxidase resulting in the NO 3 − -N accumulated, which peaked at 10.45 mg/L in the initial 24 h and then was degraded in a nal concentration of 4.95 mg/L at 72 h.The production of NO 3 − -N might re ect the adaptive mechanism utilized by strain H1 to combat nitrite toxicity (Zhang et al., 2010).Although NH 4 + -N as an intermediate product was not detected during the reduction of nitrite, the increased biomass (OD 600 ) was achieved through assimilatory nitrite reduction.Moreover, about 58.54% of TN was lost, which indicated that denitri cation mainly contributes to e cient NO 2 − -N degradation.Due to its stability, NO 3 − -N is usually di cult to be transformed completely(He et al., 2016a).The NO 3 − -N removal characterization was investigated when bacteria were cultured in SNM-3 (Fig.2-b).With the bacterial growth concentration gradually increased, a maximum NO 3 − -N removal rate (0.918 mg/L/h) of strain H1 occurred in the early bacterial growth phase at 12 h.When the biomass accumulation reached a steady growth period, a maximum NO 3 − -N removal e ciency of 56.68% was observed at 48 h.Traces of NO 2 − -N, the main product of NO 3 − -N reduction, persisted throughout the process and accumulated gradually with the consumption of carbon sources in the later stages of strain growth(Ji et al., 2015).Although the strain H1 exhibited a lower NO 3 − -N removal capability than its utilization of NH 4 + -N or NO 2 − -N, it still transformed about 24.16 mg/L NO 3 − -N after 72 h, moreover, about 22.41% of the TN was lost during the process of NO 3 − -N utilization.The nitrogen loss and increased biomass (OD 600 ) suggested the NO 3 − -N degradation by strain H1 was probably accomplished through the denitrifying and assimilatory pathways.

2 (
Zhang et al., 2012).Thus, the speculated HN-AD pathway of B. subtilis H1 might be as follows: NH 4 + -N → NH 2 OH → NO → N 2 O, the NH 4 + -N oxidation process with N 2 O as the nal gaseous product and without NO 2 − -N or NO 3 − -N accumulated (He et al., 2020; He et al., 2022); aerobic denitrifying process with N 2 O or N 2 as nal gaseous products, respectively, when NO 3 − -N or NO 2 − -N as 3 − -N to NO 2 − -N in the periplasm and intracellularly, respectively (Rothery et al., 2010) (Fig. 5-b).The gene clusters of moa and mob also played important roles in the formation of narJ.Furthermore, the oxidation reaction of nitrite (NO 2 − -N→NO 3 − -N) might have been regulated by katE, which encodes catalase in the H1 genome (Krych-Madej and Gebicka, 2017

4 .B
. subtilis H1 isolated from the mariculture environment exhibited excellent performance in removing NH 4 + -N, NO 2 − -N, NO 3 − -N, and mixed nitrogen.The nitrogen assimilation and dissimilation mechanisms of strain H1 were determined using nitrogen balance calculation, intermediate product detection, nitri cation inhibitor addition, complete genome sequencing, and enzymatic activity measurement.Of which, HN-AD pathways included a heterotrophic nitrifying process (NH 4 + -N → NH 2 OH → NO → N 2 O), and an aerobic denitrifying process with N 2 O or N 2 as the nal gaseous products, respectively, when NO 3 − -N or NO 2 − -N were utilized.The results revealed the nitrogen metabolic mechanism of strain H1 and demonstrated the bacterial application potential in mariculture wastewater treatment.Declarations Author Yumeng Xie: Conceptualization, Investigation, Formal analysis, Visualization, Writing -original draft.Yang Liu: Data curation, Investigation, Resources.Kun Zhao: Conceptualization, Methodology, Investigation, Data curation.Yongmei Li: Investigation, Resources.Kai Luo: Resources.Bo Wang: Visualization.Shuanglin Dong: Validation, Supervision.Xiangli Tian: Conceptualization, Validation, Writing -Review & Editing, Supervision.

Figure 3 The
Figure 3
Overall, the nitrogen loss and the increase in OD 600 indicated B. subtilis H1 could e ciently remove NH 4 + -N probably through nitrifying and assimilatory pathways.
and reached 1.64 mg/L in 36 h before it was further utilized.However, there were no signi cant accumulations of NO 2 − -N or NO 3 − -N during the whole procedure of NH 4 + -N removal, which contradicted the previous ndings that nitrite will inevitably accumulate during nitri cation (Kundu et al., 2014; Zhang et al., 2014).It could be speculated that NO 2 − -N and NO 3 − -N as the ammonium oxidation intermediates were never present or immediately utilized by denitrifying enzymes with higher activity (Huang et al., 2019).
3 − -N reduction, but may also regulate the reduction of NO 2 − -N to NO.Furthermore, the gene coding for NO reductase activator (norQ), NO reductase transcriptional regulators (norG), and NO response transcriptional regulators (nsrR) were all successfully annotated based on the KEGG database.In addition, this study showed that strain , these processes respectively catalyzed by AMO, HAO, NR, and NIR, and their speci c activities were quanti ed as 0.0701, 0.0474, 0.0626, and 0.4379 µmol/min/mg protein, respectively (Table.1).The determination of these enzyme activities demonstrated the successful expression of their key functional genes.According to all the aforementioned information, the putative nitrogen assimilation and dissimilation pathways of B. subtilis H1 can be illustrated in Fig.6.

Table 1
Activity detection of four key enzymes during the heterotrophic nitrifying-aerobic denitrifying process.