P. putida Y-9 (Genbank No. KP410740), which performs NO₃^– assimilation, denitrification, and DNRA under aerobic conditions simultaneously (Huang et al., 2020), was used in this study.

A denitrification medium (DM) was used to assess the nitrate reduction abilities of strain Y-9. The DM (per liter, pH = 7.2) contained 7.0 g K₂HPO₄, 3.0 g KH₂PO₄, 5.13 g CH₃COONa, 0.10 g MgSO₄ ⋅ 7H₂O, 0.72 g KNO₃, and 0.05 g FeSO₄ ⋅ 7H₂O. Luria-Bertani (LB) medium used for bacterial enrichment contained 10 g NaCl, 10 g tryptone, and 5 g yeast extract per liter (per liter, pH 7.0–7.2). All of the mediums were autoclaved for 30 min at 121°C.

The preserved strain Y-9 bacteria were activated in the LB medium at 150 rpm and 15°C for 36 h. Cells in the logarithmic growth phase were inoculated into a DM medium to assess the effects of the carbon source, C/N ratio, pH, and DO on Y-9–driven NO₃^– reduction (Li et al., 2019; Yan et al., 2021).

In the carbon source experiments, one of the three carbon sources (sodium acetate, glucose, or sodium citrate) was added to 100 mL of DM medium. The C/N ratio, pH, and shaking speed were kept constant at 15, 7, and 150 rpm, respectively. In the C/N ratio experiments, 100 mL aliquots of the DM medium were amended with glucose to yield C/N ratios of 3, 6, 9, 12, or 15. The pH and shaking speed were kept constant at 7 and 150 rpm, respectively. In the pH experiments, the initial pH was adjusted using NaOH and HCl to 4, 6, 7, 8, or 9. The carbon source was glucose, and the C/N ratio and shaking speed were held constant at 9 and 150 rpm, respectively. To determine the effects of DO on NO₃^– reduction, the shaking speed was set to 0, 50, 100, 150, or 180 rpm according to previous studies (Ren et al., 2014; Lei et al., 2019; Chen et al., 2021; Yan et al., 2021). The carbon source was glucose, and the C/N ratio and pH were kept constant at 9 and 7, respectively. The cultures were incubated at 15°C for 4 d. All of the above experiments were performed in triplicate. Samples were taken every day from each system. The optical density at 600 nm (OD₆₀₀), NH₄⁺, NO₃^–, and total nitrogen (TN) were measured for each sample.

The modified Compertz model was used to describe the kinetics analysis of nitrate degradation by strain Y-9 (Chen et al., 2016). The kinetic equation was y=y0⁢(1-exp⁢(-exp⁢(e⁢Rmy0⁢(t0-t)⁢1))), where y is the NO₃^– concentration at different incubation times (mg/L); y₀ is the initial concentration of NO₃^– (mg/L), R_m is the maximum conversion rate (mg/L/h), t₀ is the lag time (h), t is the reaction time (h), and e is the mathematical constant.

Total RNA was extracted from strain Y-9 after 4 d of incubation under various conditions using a Trizol extraction kit (Invitrogen, United States), following the manufacturer’s instructions. The specific primers B1/B2 (F: CGCAACCATCTGCTCGTGT; R: CTGGCGGGTGTAGGAAAAGT) were designed based on the nirBD gene sequence (GenBank, MK561362). These primers were used to amplify the nirBD gene from the isolates. The 16S rRNA gene was used as an internal standard, as structural rRNA is present in cells at reasonably constant levels under normal growth conditions (Edwards and Saunders, 2010). The 16S rRNA gene was amplified using the forward primer GAACGCTAATACCGCATACGTCC and the reverse primer ATCATCCTCTCAGACCAGTTAC. The total RNA was reverse-transcribed using the RevertAid first-strand cDNA synthesis kit following the manufacturer’s instructions. Real-time quantitative PCRs were performed using the SYBR^® Premix Ex Taq™ II. Each real-time PCR was performed in triplicate. The PCR cycling conditions were as follows: initial denaturation at 95°C for 30 s; 38 cycles of 95°C for 15 s, 60°C for 30 s, and 72°C for 30 s; 1 cycle of 95°C for 15 s; and, finally, stepwise temperature increases from 55°C to 95°C to generate the melting curve. Standard curves were established using a dilution series of pMD19-T vectors containing the target gene.

The OD₆₀₀ was determined based on the absorbance at 600 nm, which was measured using a spectrophotometer. The contents of the different forms of nitrogen were determined as described by Huang et al. (2019). TN was measured in the suspension. The concentration of NH₄⁺, NO₃^–, and NO₂^– was measured in the supernatant, which was obtained by centrifuging each sample at 8,000 rpm for 5 min. Three replicates were analyzed per sample, and the results are presented as means ± the standard deviation of the mean (SD). The TN and NO₃^– removal efficiencies were calculated as follows:R_V = (T₁−T₂)/T₁×100%, where R_v is the removal efficiency of TN or NO₃^– (%), and T₁ and T₂ are the initial and final concentrations of TN or NO₃^– in the system, respectively.

One-way analyses of variance (ANOVAs), followed by Duncan’s Multiple Range Tests were performed using SPSS 22, and the differences among means were considered statistically significant at P < 0.05. Graphs were drawn using Origin 8.6 and GraphPad Prism 6.

A carbon source is typically essential for the growth of heterotrophic microorganisms, and it acts as an electron donor for nitrogen cycling (Sun et al., 2016). In this study, strain Y-9 grew vigorously and reached the stationary cell growth phase after 2 d when sodium acetate, glucose, or sodium citrate was used as the sole carbon source (Figure 1A). Moreover, sodium acetate, glucose, and sodium citrate were suitable carbon sources for NO₃^– removal, with removal efficiencies of 74.75, 89.79, and 100%, respectively, at 4 d (Figure 1B). These results were consistent with those of Guo et al. (2016), who reported that sodium acetate, glucose, and sodium citrate enhanced the NO₃^– removal capacity of Enterobacter cloacae strain HNR. Furthermore, the nitrate degradation rate followed the modified Compertz model (R²> 0.90), and the maximum NO₃^– conversion rates were 1.60, 4.92, and 44.35 mg/L/h in media containing sodium acetate, glucose, and sodium citrate, respectively (Table 1).

Previous results have shown that strain Y-9 performs DNRA and nitrate assimilation under aerobic conditions (Huang et al., 2020). Based on the duration of cultivation (4 d), the detectable NH₄⁺ in the supernatant might have resulted from DNRA and nitrate assimilation followed by mineralization (Figure 1C). It is worth noting that after 3 d of cultivation, when all the cells were in the stationary phase, the detectable NH₄⁺ began to decrease in the glucose-containing system. However, it continued to increase in the media containing sodium acetate and sodium citrate. These results demonstrated that NO₃^– reduction by strain Y-9 differed when glucose was the sole carbon source compared to the other two carbon sources.

The TN decreases in our system were due to the denitrification activities of strain Y-9, and nirBD in strain Y-9 controls DNRA and nitrate assimilation (Huang et al., 2020). The TN in the media supplemented with different carbon sources tended to decrease (Figure 1D). The maximum TN decrease (60.27 mg/L) was found in the sodium citrate-containing medium, and the minimum TN decrease occurred (22.77 mg/L) in the glucose medium. This was in accordance with data from Yang et al. (2012) who reported that Pseudomonas stutzeri D6 most effectively removed TN when sodium citrate was the carbon source. Moreover, the nirBD expression level in strain Y-9 peaked when glucose was the carbon source (Figure 2). These findings demonstrated that glucose addition promoted DNRA and nitrate assimilation, effectively removing most of the NO₃^– from the system (up to 89.79%) while inhibiting denitrification (i.e., the total nitrogen lost from the system was 22.77 mg/L).

The effects of the C/N ratio on the nitrate reduction conducted by strain Y-9 were further studied. Strain Y-9 growth improved as the C/N ratio increased (Figure 3A). This result was consistent with previous studies (Kim et al., 2008; Liu et al., 2016), that the growth of P. putida AD-21 and Marinobacter strain NNA5 increased as the relative proportion of carbon increased in the medium. This might have been because electron transfer slowed when carbon concentrations were low, providing insufficient energy for microbial growth (Kim et al., 2008; Zhao et al., 2018). Greater than 80% of the NO₃^– was removed at C/N ratios of 9–15. However, the removal efficiency of NO₃^– did not exceed 30.46 and 54.00% when the C/N ratio was 3 and 6, respectively (Figure 3B). Furthermore, nitrate degradation rates at C/N ratios of 9–15 were consistent with the predictions of the modified Compertz model (R² > 0.90), and the NO₃^– conversion rate was maximum at a C/N ratio of 12 (Table 1).

The decrease of TN in this system generally mirrored the change in NO₃^– (Figure 3D). The reduction in TN at the extremely high C/N ratio of 15 was lower than the reduction in TN at a C/N ratio of 12, suggesting that a C/N ratio of 12 was optimal for denitrification. Our results indicated that the influences of C/N ratios on Y-9-driven denitrification agreed with many previous studies. They showed that extremely low or high carbon concentrations suppressed microorganismal denitrification (Kim et al., 2008; Guo et al., 2016; Zhao et al., 2018). NH₄⁺ concentration in the supernatant initially increased and then decreased during NO₃^– reduction (Figure 3C), consistent with our carbon source analysis (Figure 1C). When the C/N ratio was 9, strain Y-9 removed most of the NO₃^– (removal efficiency 81.78%) via DNRA and NO₃^– assimilation (nirBD in strain Y-9 was most strongly expressed (Figure 4)). Notably, the denitrification performance of strain Y-9 at a C/N ratio of 9 was significantly weaker than that at a C/N ratio of 12 (P < 0.05).

The impacts of the initial pH on the nitrate reduction performance of strain Y-9 are shown in Figure 5. At an initial pH of 4, the bacterial density did not noticeably increase, and NO₃^– reduction was minimal throughout the experiment (Figures 5A,B), suggesting that an overly acidic environment was detrimental to these bacteria. However, in the pH range 7–9, strain growth and NO₃^– removal were significantly improved (P < 0.05) (Figures 5A,B). These results are in agreement with the general finding that neutral or alkaline environments are beneficial for bacteria growth and bacterium-driven NO₃^– removal (Li et al., 2017; Rout et al., 2017). The NO₃^– removal efficiency was significantly positively correlated with the growth of strain Y-9 (P < 0.01) (Figure 5), indicating that pH might control the NO₃^– removal efficiency by influencing the growth of strain Y-9. However, this possibility requires further study. The TN concentration in the suspension decreased as the initial pH increased, and the TN decreased by 35.01 mg/L at pH 9 (Figure 5D). This indicated that alkaline environments favored the denitrification in strain Y-9 under aerobic conditions.

After 3 d of culture, a negligible amount of NH₄⁺ was detected at pH 4. However, the accumulation of NH₄⁺ at pH 7–9 was higher than 5.0 mg/L (Figure 5C). At the end of the experiment, the nirBD expression level in strain Y-9 at pH 7–9 was better than at pH 4 or 6 (Figure 6). These results showed that the initial pH affected the expression of nirBD in strain Y-9, and this might influence NH₄⁺ production from DNRA and NO₃^– assimilation as well as its subsequent mineralization (Huang et al., 2020). The results of previous studies on the effects of pH on NO₃^– reduction by soil microorganisms are widely contradictory (Nägele and Conrad, 1990; Stevens et al., 1998). Here, strain Y-9 effectively performed NO₃^– assimilation, DNRA, and denitrification at pH 7–9. This finding was inconsistent with a previous study (Yoon et al., 2015) that suggested that a low pH was more favorable for denitrification, while a high pH promoted the production of NH₄⁺ via DNRA. This discrepancy indicated that the effects of pH on the microbial nitrogen cycle were complex and required further study. Our results suggested that a neutral pH was most favorable for NO₃^– removal and nitrogen retention.

Strain Y-9 growth and NO₃^– reduction increased gradually as the shaking speed increased (Figures 7A,B). Changes in the NO₃^– degradation rates at various rotating speeds were consistent with the predictions of the modified Compertz model (R² > 0.80), and the NO₃^– conversion rate achieved its maximum at 150 rpm (Table 1). These results suggested that increasingly aerobic conditions improved strain growth and NO₃^– reduction. TN decreased gradually throughout the incubation process, irrespective of the DO concentration. However, TN decreased at low shaking speeds (≤50 rpm) was significantly greater than those at high shaking speeds (≥100 rpm) (P < 0.05). The decrease in the TN at the end of the experiment reached the maximum (34.58 mg/L) at a shaking speed of 50 rpm (Figure 7D). These results suggested that the denitrification performance of strain Y-9 first increased and then decreased as the DO concentration increased. This finding was consistent with the results of Zhao et al. (2018); Rout et al. (2017), and Huang and Tseng (2001). Previous studies demonstrated that the denitrification performance remained stable as long as the DO concentration remained within a fixed range. Nevertheless, the denitrification enzyme activity levels improved noticeably when the DO concentration decreased below a threshold value (Song et al., 2011). For strain Y-9, 50 rpm might be the threshold DO value that affects denitrification enzymes, although this possibility requires further testing.

The quantitative PCR amplification results indicated that the expression of nirBD increased with an increase in the shaking speed (Figure 8). These results, in conjunction with the NO₃^– reduction performance of strain Y-9 (Huang et al., 2020), indicated that high DO concentrations stimulated the expression of nirBD in strain Y-9, promoting NO₃^– assimilation as well as DNRA, and thus releasing more NH₄⁺ into the supernatant (Figure 7C). Consistent with this, Yang et al. (2012) and Zhao et al. (2018) found that NH₄⁺ production increased with the DO content during NO₃ ^– reduction by P. stutzeri D6 and P. stutzeri strain XL-2. Variance analyses indicated that the amounts of NO₃^– removal from the culture media and NH₄⁺ accumulated in the culture media at high shaking speeds (≥100 rpm) differed obviously from those at low rotation speeds (≤50 rpm) (P < 0.05). These results indicated that good aeration effectively promoted NO₃^– removal and NH₄⁺ production by strain Y-9.

The heavy application of chemical nitrogen fertilizers leads to an accumulation of highly mobile nitrate in upland soils and significantly increases the risk of nitrogen loss (Lin et al., 2020; Vidal et al., 2020). Therefore, it is essential to control NO₃^– concentrations in soil. The nitrogen cycle conducted by microorganisms plays a critical role in regulating nitrate concentrations in soil, compared to artificially limiting the application of ammonium and nitrate fertilizers (Shao et al., 2011; Song et al., 2014; Zhang et al., 2015; Pandey et al., 2019; Wang et al., 2020). Denitrification effectively removes excess NO₃^– from soil systems but leads to nitrogen losses in the form of nitrogen gas or the greenhouse gas N₂O (Stein and Klotz, 2016). For example, Putz et al. (2018) showed that approximately 70–78% of all N₂O originated from denitrification in annual cereal soils. Both the DNRA and NO₃^– assimilation processes can decrease soil NO₃^– concentration and facilitate soil nitrogen conservation by reducing NO₃^– to NH₄⁺ via NO₂^– (Kuypers et al., 2018; Wang et al., 2020). Thus, to pursue the minimum loss of nitrogen and maximize nitrogen fertilizer efficiency, strategies that strengthen DNRA as well as NO₃^– assimilation while weakening denitrification in surface soils should be pursued. Previously, we found that strain Y-9 performs simultaneous nitrate assimilation, DNRA, and denitrification under aerobic conditions. It has also been clarified that the gene, nirBD, controls NO₃^– assimilation and DNRA process in strain Y-9 (Huang et al., 2020). In this study, we further explored the environmental factors that affect the nitrate removal pathways of strain Y-9. Our results provide a theoretical reference for technical studies of nitrate removal and nitrogen conservation in farmland soils.