ntrC Contributes to Nitrogen Utilization, Stress Tolerance, and Virulence in Acidovorax citrulli

Bacterial fruit blotch (BFB), caused by Acidovorax citrulli, severely damages watermelon, melon, and other cucurbit crops worldwide. Nitrogen, one of the most important limiting elements in the environment, is necessary for the growth and reproduction of bacteria. As a nitrogen-regulating gene, ntrC plays an important role in maintaining bacterial nitrogen utilization and biological nitrogen fixation. However, the role of ntrC has not been determined for A. citrulli. In this study, we constructed a ntrC deletion mutant and a corresponding complementary strain in the background of the A. citrulli wild-type strain, Aac5. Through phenotype assays and qRT-PCR analysis, we investigated the role of ntrC in A. citrulli in nitrogen utilization, stress tolerance, and virulence against watermelon seedlings. Our results showed that the A. citrulli Aac5 ntrC deletion mutant lost the ability to utilize nitrate. The ntrC mutant strain also exhibited significantly decreased virulence, in vitro growth, in vivo colonization ability, swimming motility, and twitching motility. In contrast, it displayed significantly enhanced biofilm formation and tolerance to stress induced by oxygen, high salt, and copper ions. The qRT-PCR results showed that the nitrate utilization gene nasS; the Type III secretion system-related genes hrpE, hrpX, and hrcJ; and the pili-related gene pilA were significantly downregulated in the ntrC deletion mutant. The nitrate utilization gene nasT, and the flagellum-related genes flhD, flhC, fliA, and fliC were significantly upregulated in the ntrC deletion mutant. The expression levels of ntrC gene in the MMX-q and XVM2 media were significantly higher than in the KB medium. These results suggest that the ntrC gene plays a pivotal role in the nitrogen utilization, stress tolerance, and virulence of A. citrulli.


Introduction
Bacterial fruit blotch (BFB), caused by Acidovorax citrulli, is a seed-borne bacterial disease that can infect and cause significant economic losses to cucurbit crops, including watermelon (Citrullus lanatus) and melon (Cucumis melo). The disease was first reported in Georgia, USA, in 1965 and has been reported in many countries [1]. Acidovorax citrulli uses a variety of virulence factors, such as the Type III secretion system (T3SS), the Type VI secretion system (T6SS), Type IV pili (T4P), polar flagella, quorum sensing, and biofilms [2][3][4][5][6][7][8][9][10] to induce disease. Acidovorax citrulli strains can be divided into two major groups (I and II) [11,12]. Natural field experiments showed that A. citrulli groups (I and II) displayed a preferential association for different cucurbit hosts [13,14]. Despite efforts to develop strategies to mitigate the losses caused by BFB, the current management strategies are not effective, and outbreaks continue to occur sporadically, with economic consequences [15].
As one of the basic elements of life, nitrogen is necessary for the growth of bacteria [16]. Bacteria can utilize nitrogen-containing compounds such as ammonium salts and nitrates as nitrogen sources [16,17]. The utilization of nitrate and nitrite by bacteria mainly depends on the nitrate and nitrite assimilation reductase system, which can convert nitrate and nitrite in the environment into ammonium. This ammonium can then be used in the glutamate dehydrogenase reaction [16]. The utilization of ammonium salts from the environment by bacteria is mainly through the diffusion of NH 4 + across the cell membrane. The membrane protein AmtB has a high affinity for NH 4 + and plays an important role in the absorption of NH 4 + when it is limiting [16]. The two-component system, NtrBC, plays a key role in bacterial regulation of nitrogen. When nitrogen is limiting, the amount of glutamine inside the cells is reduced compared with 2-ketoglutarate, which causes uridylation of PII, the inhibition of NtrB phosphatase activity, and the activation of NtrC [17]. After being phosphorylated, NtrC, together with RNA polymerase and σ 54 , activates the expression of different genes [17]. The deletion of the response regulator ntrC gene in Rhizobium leguminosarum, Sinorhizobium fredii, Pseudomonas fluorescens, and Agrobacterium sp. had different effects on the utilization of nitrogen sources such as nitrate, ammonium salts, and urea [18][19][20][21].
In addition to regulating nitrogen, NtrC is also involved in many biological processes in bacteria, such as stress tolerance, extracellular polymer synthesis, and biofilm formation [22][23][24][25]. The deletion of ntrC reduced in vitro growth but enhanced oxygen-related stress tolerance in P. putida [22]. The deletion of ntrC resulted in significant changes in the tolerance of Aeromonas hydrophila to osmotic stress, heavy metal ions, oxidation, and different antibiotic stresses [23]. Interestingly, swimming motility and virulence decreased in the P. aeruginosa ntrB-ntrC deletion mutant [24]. The biofilm production of a Vibrio cholerae ntrC mutant was significantly higher than that of the wild-type strain [25]. Despite these observations, the role of ntrC in the virulence of plant pathogenic bacteria is not known. Therefore, in this study, we constructed an ntrC gene deletion mutant in A. citrulli and used it to investigate the effect of ntrC on virulence, nitrogen utilization, and stress tolerance. We also analyzed the regulatory network that ntrC participates in.

Plant Materials, Bacterial Strains, and Plasmids
The watermelon cultivar 'Ruixin' (provided by China Vegetable Seed Technology Corporation, Beijing, China) was used in seed-to-seedling transmission assays as previously described [2,9,13]. The bacterial strains and plasmids used in this study are shown in Table S1.

Construction of the A. citrulli ntrC Mutant and Its Complementary Strain
The primer pair ntrC-F/R was designed on the basis of the ntrC gene (Aave_1445) in the AAC00-1 genome (GenBank accession number: CP000512.1), and the primer pairs, ntrC-1F/1R and ntrC-2F/2R, were designed on the basis of the upstream and downstream flanking sequences of the ntrC gene (Table S2). The nitrogen regulatory gene ntrC was deleted using the double homologous recombination approach, as previously described [2]. Briefly, the sequences 592 bp upstream and 532 bp downstream of the ntrC gene were amplified from the genomic DNA of Aac5-WT using KOD-Plus-Neo (TOYOBO, Osaka, Japan) and the primers ntrC-1F/ntrC-1R and ntrC-2F/ntrC-2R. These two flanking sequences were fused by an overlapping polymerase chain reaction (PCR) and then ligated into the suicide vector pK18mobsacB using the ClonExpress II One Step Cloning Kit (Vazyme, Shanghai, China) to generate pK18mobsacB-ntrC, which was then transformed into competent E. coli DH5α [26]. The vector pK18mobsacB-ntrC was confirmed by sequencing. The DNA sequencing in this study was performed by Liuhe BGI Co., Ltd. (Beijing, China). Then the vector pK18mobsacB-ntrC was introduced into Aac5-WT by triparental hybridization, and the individual crossover colonies were screened on the basis of their Amp and Kan resistance. The mutant was screened on an M9 agar medium with sucrose [27,28]. The mutant strain Aac5∆ntrC was verified by PCR and DNA sequencing. To construct the complementary strain, the ntrC gene sequence with its native promoter sequence was ligated into pBBR1MCS-2 to generate pBBR1MCS-ntrC. Then the pBBR1MCS-ntrC was introduced into the mutant Aac5∆ntrC by triparental hybridization and verified by PCR and DNA sequencing. To eliminate the effect of pBBR1MCS-2, the empty vector pBBR1MCS-2 was introduced into Aac5-WT and Aac5∆ntrC by triparental hybridization and verified by PCR and DNA sequencing.

Determination of the Nitrogen Utilization Capacity and In Vitro Growth Ability
To assess the ability of bacterial strains to use different nitrogen sources, (NH 4 ) 2 SO 4 (10 mmol·L −1 ), KNO 3 (20 mmol L −1 ), and CH 4 N 2 O (10 mmol·L −1 ) were added separately in the MMX basic medium [29]. Overnight cultures of Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp were adjusted to OD 600 = 0.6 after resuspension with sterilized distilled water, and the bacterial suspensions (10 µL) were added to 2 mL centrifuge tubes with 1 mL of the medium (MMX-(NH 4 ) 2 SO 4 , MMX-KNO 3 , and MMX-CH 4 N 2 O) (three replicates). The suspensions were cultured at 28 • C under shaking at 220 revolutions·min −1 for 96 h, and the OD 600 values were determined using a spectrophotometer (Biochrom, Cambridge, UK). The experiment was conducted three times.
The bacterial suspensions of Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp were resuspended with MMX-KNO 3 or KB medium, adjusted to OD 600 = 0.3, and diluted 100-fold using the MMX-KNO 3 or KB medium. Then the diluted bacterial suspensions (200 µL) were added to 100-well polystyrene plates. The plates were incubated at 28 • C with continuous shaking, and the OD 600 values was measured every 2 h for 96 h (Bioscreen C • PRO, Finland) [6]. Each treatment was replicated four times, and the experiment was conducted at least three times.

Hydrogen Peroxide Sensitivity Assay
Bacterial suspensions of Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp were generated in sterilized water and adjusted to a final concentration of 3 × 10 8 colony-forming units (CFU)·mL −1 (OD 600 = 0.3). Bacterial suspensions (2.5 mL) were mixed evenly with 50 mL of melted KB agar medium. After the medium had solidified, 5 µL of a 3% H 2 O 2 solution was placed on the plate (three replicates). The diameters of the zones of inhibition were measured 2 days after incubation at 28 • C. The experiment was conducted three times.

Sodium Chloride and Copper Sulphate Sensitivity Assays
Bacterial suspensions of Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp were adjusted to OD 600 = 0.3 and centrifuged, and the pellets were resuspended in KB media containing 4% NaCl or KB media containing 4 mmol CuSO 4 ·5H 2 O, and incubated at 28 • C with shaking at 220 r·min −1 for 20 h. Aliquots were serially diluted 10-fold and were plated on KB agar media (containing 1.5% agar) supplemented with appropriate concentrations of Amp and Kan. Colonies were counted 48 h after incubation at 28 • C. Each treatment was replicated 3 times and the experiment was conducted 3 times. Overnight cultures of Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp were adjusted to OD 600 = 0.3. For each treatment, four watermelon seedlings (4 weeks old) were sprayed with 10 mL of the bacterial suspension or water (negative control) and then bagged to maintain high relative humidity. The inoculated seedlings were placed in a growth chamber under the conditions of light at 28 • C for 16 h and darkness at 22 • C for 8 h, and a relative humidity of 80%. The disease index was evaluated and calculated at 5 days post-inoculation (dpi) [30]. The experiment was conducted three times.

Seed-to-Seedling Transmission Assay
Overnight cultures of Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp were adjusted to OD 600 = 0.3. Then watermelon seeds (three replicates per treatment, 10 seeds per replicate) were soaked in the bacterial suspension for 1 h. Seeds were soaked in sterilized water as a negative control. Inoculated seeds were then air-dried and planted in nutritive substrates (Guangdahengyi, Beijing, China). The watermelon seedlings were cultured in a growth chamber under the conditions described above. The disease index was evaluated and calculated at 14 days after sowing [31]. The experiment was conducted three times.

In Vivo Growth Ability
The concentrations of the bacterial suspensions of the strains Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp were adjusted to 10 6 CFU·mL −1 with sterilized water. Each bacterial suspension was injected into watermelon cotyledons (three replicates per treatment, 15 cotyledons per replicate) with a 1 mL syringe and incubated in a growth chamber under the same incubation conditions as described above. Sterilized water was used as a negative control. At 1, 24, 48, 72, and 96 h post-inoculation (hpi), watermelon cotyledons were selected for observation and photography. One leaf disk (0.8 cm in diameter) for each cotyledon was sampled, and each replicate constituted three cotyledons. Three leaf disks were macerated in 500 µL of sterilized water in a microcentrifuge tube. The macerates were serially diluted 10-fold, and plated on KB agar media with Amp and Kan. Colonies were counted 48 h after incubation at 28 • C [32]. The assays were repeated three times.

Assays of Swimming and Twitching Motility Ability
To assess swimming motility, bacterial suspensions of Aac5, Aac5∆ntrC, and Aac5∆ntrC comp (OD 600 = 0.3, 3 µL) were gently inoculated onto the surfaces of a 0.3% semisolid medium (tryptone, 0.3 g; yeast extract, 0.3 g; agar, 3 g) [33] and incubated at 28 • C for 48 h. Each colony halo was photographed and its diameter was measured. Every treatment had four replicates and the experiment was conducted three times.
To assess twitching motility, fresh colonies of Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp were streaked onto KB agar media. Corrugated tracks around each colony were observed with an IX83 inverted microscope (OLYMPUS, Tokyo, Japan) after 4 days of incubation at 28 • C [5]. The ratio of the halo's diameter compared with the colony's diameter was determined. Each treatment was replicated six times and the experiment was conducted three times.

Biofilm Formation Assay
Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp liquid cultures suspended in KB broth were adjusted to OD 600 = 0.3, and 2 mL of each suspension was added to 12-well polystyrene plates with three replicates for each treatment. The plates were incubated at 28 • C for 72 h, and then the liquid was slowly removed with a pipette, washed slowly with sterilized water, and fixed in an oven at 80 • C for 30 min. The wells were stained with 0.1% crystal violet for 50 min and rinsed three times with sterilized water. The plates were air-dried at 37 • C and photographed. Two milliliters of 95% ethanol needed to be added to dissolve the biofilm formed by Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp, and the OD 575 values were measured using a spectrophotometer. The experiment was conducted three times [7].

Determination of the Expression of the ntrC Gene in Different Media
The expression of ntrC by Aac5 in the KB medium, the T3SS induction XVM2 medium, and the MMX-q medium was determined as described above, and rpoB was used as an internal reference gene. Each treatment had three replicates. This experiment was performed three times independently.

Data Analysis
The experimental data were recorded and calculated using Excel (Microsoft, Redmond, WA, USA), and graphs were plotted using GraphPad Prism 7 (GraphPad, San Diego, CA, USA). For statistical analysis, one-way analysis of variance (ANOVA) was conducted using GraphPad Prism 7 (with 95% confidence intervals).

Confirmation of the Mutant and Complementary Strains
The full length of the ntrC gene is 1650 bp long and is located at genomic nucleotide positions 1592112 to 1593761 of A. citrulli Group II strain AAC00-1. Aac5∆ntrC was verified by PCR using the ntrC-specific verification primer ∆ntrC-F/∆ntrC-R, the A. citrullispecific primer WFB1/WFB2, and the pBBR1MCS-2 plasmid detection primer Kan-F/Kan-R (Table S2), and by sequencing. The complementary strain Aac5∆ntrCcomp showed resistance to Kan, and was also verified by PCR and sequencing.

Inactivation of ntrC Affects A. citrulli's Ability to Assimilate Nitrogen and Grow In Vitro
Aac5∆ntrC was unable to utilize KNO 3 as a sole nitrogen source, but could utilize (NH 4 ) 2 SO 4 and CH 4 N 2 O as a sole nitrogen source. In contrast, Aac5 and Aac5∆ntrCcomp cells grew on KNO 3 , (NH 4 ) 2 SO 4 , and CH 4 N 2 O (Figure 1a). We further examined the in vitro growth dynamics of Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp on KNO 3 as the sole nitrogen source. During the 96 h incubation period, both Aac5 and Aac5∆ntrCcomp reached a plateau (OD 600 = 0.3) after a logarithmic growth period, while the OD 600 value of Aac5∆ntrC remained unchanged from 0 h to 96 h ( Figure 1b). The results showed that ntrC was critical for A. citrulli to utilize nitrate. In the KB medium, Aac5∆ntrC entered the logarithmic growth phase about 8 h later than Aac5. However, after entering the stationary phase, Aac5∆ntrC reached a higher OD 600 than Aac5 ( Figure 1c). Additionally, the time when the complementary strain Aac5∆ntrCcomp entered the logarithmic phase was between those of Aac5 and Aac5∆ntrC, but the growth in the stationary phase was the same as that of Aac5∆ntrC and higher than that of Aac5 ( Figure 1c). growth dynamics of Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp on KNO3 as the sole nitrogen source. During the 96 h incubation period, both Aac5 and Aac5ΔntrCcomp reached a plateau (OD600 = 0.3) after a logarithmic growth period, while the OD600 value of Aac5ΔntrC remained unchanged from 0 h to 96 h ( Figure 1b). The results showed that ntrC was critical for A. citrulli to utilize nitrate. In the KB medium, Aac5ΔntrC entered the logarithmic growth phase about 8 h later than Aac5. However, after entering the stationary phase, Aac5ΔntrC reached a higher OD600 than Aac5 ( Figure 1c). Additionally, the time when the complementary strain Aac5ΔntrCcomp entered the logarithmic phase was between those of Aac5 and Aac5ΔntrC, but the growth in the stationary phase was the same as that of Aac5ΔntrC and higher than that of Aac5 (Figure 1c). , and CH4N2O (10 mmol·L −1 ) were added separately to the basic MMX medium. Each treatment had three replicates, and the experiment was conducted three times. Asterisks indicate significant differences; error bars represent standard errors of the means (p < 0.05, one-way ANOVA test). (b) Growth curve of the strains Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp in the MMX-KNO3 medium at 28 °C for 96 h. The OD600 values were measured every 2 hours. Each treatment had four replicates, and the experiment was conducted three times. (c) Growth curves of Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp in the KB medium at 28 °C for 96 h. Each treatment had four replicates, and the experiment was conducted three times. (a) The OD 600 value of the strains Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp after 96 h of incubation in a basic MMX medium. Nitrogen sources (NH 4 ) 2 SO 4 (10 mmol·L −1 ), KNO 3 (20 mmol L −1 ), and CH 4 N 2 O (10 mmol·L −1 ) were added separately to the basic MMX medium. Each treatment had three replicates, and the experiment was conducted three times. Asterisks indicate significant differences; error bars represent standard errors of the means (p < 0.05, one-way ANOVA test). (b) Growth curve of the strains Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp in the MMX-KNO 3 medium at 28 • C for 96 h. The OD 600 values were measured every 2 hours. Each treatment had four replicates, and the experiment was conducted three times. (c) Growth curves of Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp in the KB medium at 28 • C for 96 h. Each treatment had four replicates, and the experiment was conducted three times.

ntrC Deletion Affects the Stress Tolerance of Aac5
The average diameters of the inhibition zones of Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp were 25.08 mm, 16.42 mm, and 18.17 mm, respectively, on KB agar. The diameter of Aac5∆ntrC was significantly smaller than that of Aac5 (p < 0.05). However, the diameter of the inhibition zone of the complementary strain Aac5∆ntrCcomp was larger than that of Aac5∆ntrC but less than that of Aac5 (Figure 2b). The results showed that deletion of the ntrC gene enhanced the ability of A. citrulli to tolerate oxygen stress.
population of Aac5ΔntrC was significantly higher than that of Aac5 (p < 0.05), while that of Aac5ΔntrCcomp could be restored to the level of Aac5 (Figure 2d). The results showed that the deletion of ntrC enhanced the ability of A. citrulli to tolerate high salt stress.
After incubation in a KB medium with 4 mmol CuSO4·5H2O for 20 h, the average surviving population of Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp was 3.63 × 10 6 CFU·mL −1 , 1.53 × 10 7 CFU·mL −1 , and 6.1 × 10 6 CFU·mL −1 , respectively. The surviving population of Aac5ΔntrC was significantly higher than that of Aac5 (p < 0.05). The surviving population of Aac5ΔntrCcomp was between that of Aac5ΔntrC and Aac5 (Figure 2f). The results showed that deletion of the ntrC gene significantly increased tolerance to copper in A. citrulli. Figure 2. Effect of ntrC on tolerance to oxidative stress, high salt stress, and Cu 2+ stress in A. citrulli. (a) The ability of Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp to tolerate oxidative stress was determined using the inhibitory halo method. After the medium had solidified, 5 μL of a 3% H2O2 solution was placed on the plate. The plates were photographed 2 days after plating. Each treatment included three replicates and the experiment was conducted three times. (b) The average diameter of the inhibition zone of Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp measured after 2 days of incubation. (c,d) The viable bacterial counts of Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp after 20 h of culture in a KB medium containing 4% sodium chloride. (e,f) The viable bacterial counts of Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp after 20 h of culture in a KB medium containing 4 mmol CuSO4·5H2O. Each treatment had three replicates, and the assays were repeated three times. Asterisks indicate significant differences; error bars represent the standard errors of the means (p < 0.05, one-way ANOVA test). The tolerance of Aac5 and its derived strains to high salt stress was determined using a KB medium containing 4% sodium chloride. After incubation under high salt stress for 20 h, the average surviving population of Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp was 7.67 × 10 5 CFU·mL −1 , 2.0 × 10 6 CFU·mL −1 , and 5.0 × 10 5 CFU·mL −1 , respectively. The surviving population of Aac5∆ntrC was significantly higher than that of Aac5 (p < 0.05), while that of Aac5∆ntrCcomp could be restored to the level of Aac5 (Figure 2d). The results showed that the deletion of ntrC enhanced the ability of A. citrulli to tolerate high salt stress.

ntrC Contributes to the Virulence of A. citrulli Aac5
To elucidate the role of the ntrC gene in the virulence of A. citrulli, we carried out watermelon spray-inoculation assays and seed-to-seedling transmission assays.
The disease symptoms on seedlings inoculated with Aac5∆ntrC were less severe than in seedlings inoculated with Aac5 at 5 dpi (Figure 3a). The disease index for watermelon seedlings inoculated with Aac5∆ntrC was 28.80, which was significantly lower than that of watermelon seedlings inoculated with Aac5 (61.77, p < 0.05). The disease index of Aac5∆ntrCcomp recovered to 49.99 (Figure 3b).

ntrC Contributes to the Virulence of A. citrulli Aac5
To elucidate the role of the ntrC gene in the virulence of A. citrulli, we carried out watermelon spray-inoculation assays and seed-to-seedling transmission assays.
The disease symptoms on seedlings inoculated with Aac5ΔntrC were less severe than in seedlings inoculated with Aac5 at 5 dpi (Figure 3a). The disease index for watermelon seedlings inoculated with Aac5ΔntrC was 28.80, which was significantly lower than that of watermelon seedlings inoculated with Aac5 (61.77, p < 0.05). The disease index of Aac5ΔntrCcomp recovered to 49.99 (Figure 3b).
The virulence of each strain was also determined using seed-to-seedling transmission assays. The disease symptoms on seedlings inoculated with Aac5ΔntrC were less severe than those for seedlings inoculated with Aac5 14 days after sowing (Figure 3c). The disease index of watermelon seedlings inoculated with Aac5ΔntrC was 66.20, significantly lower than that of watermelon seedlings inoculated with Aac5 (99.54, p < 0.05). The disease index of Aac5ΔntrCcomp recovered to 86.11 (Figure 3d). (c) Watermelon seeds were soaked in sterilized water (CK), and Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp suspensions for 1 h, and photos were taken at 14 dpi. Each treatment had three replicates and the experiment was conducted three times. (d) The disease index of watermelon seedlings 14 days after sowing. Asterisks indicate significant differences; the error bars represent the standard errors of the means (p < 0.05, one-way ANOVA test).

Deletion of ntrC Reduces the In Vivo Growth of A. citrulli Aac5
Cotyledons infiltrated with strains Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp gradually developed BFB symptoms, but cotyledons treated with Aac5ΔntrC showed only mild symptoms. As expected, cotyledons treated with sterilized water showed no symptoms (Figure 4a). The bacterial population in cotyledons of Aac5ΔntrC was significantly lower than that of Aac5 between 24 hpi and 96 hpi (p < 0.05), while the in vivo growth of Aac5ΔntrCcomp basically returned to the wild-type levels (Figure 4b). (c) Watermelon seeds were soaked in sterilized water (CK), and Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp suspensions for 1 h, and photos were taken at 14 dpi. Each treatment had three replicates and the experiment was conducted three times. (d) The disease index of watermelon seedlings 14 days after sowing. Asterisks indicate significant differences; the error bars represent the standard errors of the means (p < 0.05, one-way ANOVA test).
The virulence of each strain was also determined using seed-to-seedling transmission assays. The disease symptoms on seedlings inoculated with Aac5∆ntrC were less severe than those for seedlings inoculated with Aac5 14 days after sowing (Figure 3c). The disease index of watermelon seedlings inoculated with Aac5∆ntrC was 66.20, significantly lower than that of watermelon seedlings inoculated with Aac5 (99.54, p < 0.05). The disease index of Aac5∆ntrCcomp recovered to 86.11 (Figure 3d).

Deletion of ntrC Reduces the In Vivo Growth of A. citrulli Aac5
Cotyledons infiltrated with strains Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp gradually developed BFB symptoms, but cotyledons treated with Aac5∆ntrC showed only mild symptoms. As expected, cotyledons treated with sterilized water showed no symptoms (Figure 4a). The bacterial population in cotyledons of Aac5∆ntrC was significantly lower than that of Aac5 between 24 hpi and 96 hpi (p < 0.05), while the in vivo growth of Aac5∆ntrCcomp basically returned to the wild-type levels (Figure 4b).

Deletion of ntrC Impairs Swimming and Twitching Motility in A. citrulli
The average diameters of the Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp colonies were 14.33 mm, 10.17 mm, and 9.42 mm, respectively. Compared with Aac5, the swimming motility of Aac5ΔntrC was significantly reduced, and complementation did not restore this ability (Figure 5a,b). The twitching motility of Aac5ΔntrC was also significantly reduced compared with Aac5, while that of Aac5ΔntrCcomp was restored to the wild-type levels (Figure 5c,d).

Deletion of ntrC Impairs Swimming and Twitching Motility in A. citrulli
The average diameters of the Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp colonies were 14.33 mm, 10.17 mm, and 9.42 mm, respectively. Compared with Aac5, the swimming motility of Aac5∆ntrC was significantly reduced, and complementation did not restore this ability (Figure 5a,b). The twitching motility of Aac5∆ntrC was also significantly reduced compared with Aac5, while that of Aac5∆ntrCcomp was restored to the wild-type levels (Figure 5c,d).

Deletion of ntrC Impairs Swimming and Twitching Motility in A. citrulli
The average diameters of the Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp colonies were 14.33 mm, 10.17 mm, and 9.42 mm, respectively. Compared with Aac5, the swimming motility of Aac5ΔntrC was significantly reduced, and complementation did not restore this ability (Figure 5a,b). The twitching motility of Aac5ΔntrC was also significantly reduced compared with Aac5, while that of Aac5ΔntrCcomp was restored to the wild-type levels (Figure 5c,d).

ntrC Contributes to Biofilm Formation in A. citrulli Aac5
We determined the biofilm formation ability of Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp by culturing them in a KB medium. Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp formed visible biofilms on the inner walls of the 12-well cell culture plates (Figure 6a). The OD 575 values of Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp were 0.74, 1.27, and 0.41, respectively. The OD 575 value of Aac5∆ntrC was significantly higher than that of Aac5, while the absorbance of the complementary strain was significantly lower than that of Aac5 (p < 0.05) (Figure 6b).

ntrC Contributes to Biofilm Formation in A. citrulli Aac5
We determined the biofilm formation ability of Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp by culturing them in a KB medium. Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp formed visible biofilms on the inner walls of the 12-well cell culture plates (Figure 6a). The OD575 values of Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp were 0.74, 1.27, and 0.41, respectively. The OD575 value of Aac5ΔntrC was significantly higher than that of Aac5, while the absorbance of the complementary strain was significantly lower than that of Aac5 (p < 0.05) (Figure 6b).

Effect of ntrC on the Expression of Select A. citrulli Genes
The expression of the nitrate utilization-related gene nasS was significantly decreased while nasT was significantly increased in the MMX-q medium. The expression of the key T3SS genes hrpE, hrpX, and hrcJ was significantly decreased in the XVM2 medium in Aac5ΔntrC compared with Aac5. However, the expression of hrpG was not significantly different. When cultured in the KB medium, the expression of the Aac5ΔntrC flagellumrelated genes flhD, flhC, fliA, and fliC increased significantly in Aac5ΔntrC compared with Aac5, while the pili-related gene pilA decreased significantly. The expression levels of fliM and pilN were not significantly different (Figure 7). Overall, there appeared to be a regulatory relationship of A. citrulli ntrC with T3SS-related genes, nitrate utilization-related genes, flagellum-related genes, and pili-related genes. rpoB was used as an internal reference gene. Each treatment had three replicates, and the experiment was conducted three times. Asterisks indicate significant differences; error bars represent the standard errors of the means (p < 0.05, one-way ANOVA test).

Effect of ntrC on the Expression of Select A. citrulli Genes
The expression of the nitrate utilization-related gene nasS was significantly decreased while nasT was significantly increased in the MMX-q medium. The expression of the key T3SS genes hrpE, hrpX, and hrcJ was significantly decreased in the XVM2 medium in Aac5∆ntrC compared with Aac5. However, the expression of hrpG was not significantly different. When cultured in the KB medium, the expression of the Aac5∆ntrC flagellumrelated genes flhD, flhC, fliA, and fliC increased significantly in Aac5∆ntrC compared with Aac5, while the pili-related gene pilA decreased significantly. The expression levels of fliM and pilN were not significantly different (Figure 7). Overall, there appeared to be a regulatory relationship of A. citrulli ntrC with T3SS-related genes, nitrate utilization-related genes, flagellum-related genes, and pili-related genes.

ntrC Contributes to Biofilm Formation in A. citrulli Aac5
We determined the biofilm formation ability of Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp by culturing them in a KB medium. Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp formed visible biofilms on the inner walls of the 12-well cell culture plates (Figure 6a). The OD575 values of Aac5, Aac5ΔntrC, and Aac5ΔntrCcomp were 0.74, 1.27, and 0.41, respectively. The OD575 value of Aac5ΔntrC was significantly higher than that of Aac5, while the absorbance of the complementary strain was significantly lower than that of Aac5 (p < 0.05) (Figure 6b).

Effect of ntrC on the Expression of Select A. citrulli Genes
The expression of the nitrate utilization-related gene nasS was significantly decreased while nasT was significantly increased in the MMX-q medium. The expression of the key T3SS genes hrpE, hrpX, and hrcJ was significantly decreased in the XVM2 medium in Aac5ΔntrC compared with Aac5. However, the expression of hrpG was not significantly different. When cultured in the KB medium, the expression of the Aac5ΔntrC flagellumrelated genes flhD, flhC, fliA, and fliC increased significantly in Aac5ΔntrC compared with Aac5, while the pili-related gene pilA decreased significantly. The expression levels of fliM and pilN were not significantly different (Figure 7). Overall, there appeared to be a regulatory relationship of A. citrulli ntrC with T3SS-related genes, nitrate utilization-related genes, flagellum-related genes, and pili-related genes. Figure 7. Analysis of the expression of key genes in Acidovorax citrulli strains Aac5 and Aac5ΔntrC. rpoB was used as an internal reference gene. Each treatment had three replicates, and the experiment was conducted three times. Asterisks indicate significant differences; error bars represent the standard errors of the means (p < 0.05, one-way ANOVA test). rpoB was used as an internal reference gene. Each treatment had three replicates, and the experiment was conducted three times. Asterisks indicate significant differences; error bars represent the standard errors of the means (p < 0.05, one-way ANOVA test).

Expression of the A. citrulli ntrC Gene in Different Media
The expression levels of Aac5 ntrC in the MMX-q and XVM2 media were significantly higher than in the KB medium (p < 0.05). The expression of ntrC by Aac5 in the XVM2 medium was significantly higher than in the MMX-q medium (p < 0.05) (Figure 8).

Expression of the A. citrulli ntrC Gene in Different Media
The expression levels of Aac5 ntrC in the MMX-q and XVM2 media were significantly higher than in the KB medium (p < 0.05). The expression of ntrC by Aac5 in the XVM2 medium was significantly higher than in the MMX-q medium (p < 0.05) (Figure 8).

Figure 8.
Relative expression of ntrC in Acidovorax citrulli strain Aac5 in KB, MMX-q, and XVM2 media. qRT-PCR assays were conducted, and rpoB was used as an internal reference gene. Each treatment had three replicates, and the experiment was conducted three times. Asterisks indicate significant differences; error bars represent the standard errors of the means (p < 0.05, one-way ANOVA test).

Discussion
In this study, we constructed a ntrC deletion mutant and a corresponding complemented strain in the background of the A. citrulli wild-type strain Aac5. Through phenotype assays and qRT-PCR analysis, we investigated the role of the ntrC gene in nitrogen utilization, stress tolerance, and virulence against watermelon seedlings in A. citrulli.
We observed that the ntrC gene deletion mutant was unable to use potassium nitrate as the sole nitrogen source, but could grow with ammonium sulfate and urea as the sole nitrogen sources. In addition, the expression of the nitrate utilization-related gene nasS was significantly downregulated, and the expression of nasT was significantly upregulated in Aac5ΔntrC. NasS and NasT comprise the two-component system that regulates bacterial nitrate utilization [38]. We speculate that ntrC affects the activation of the nasS promoter and further affects the utilization of nitrate, which is consistent with previous studies [38]. The expression levels of ntrC by Aac5 in the MMX-q and XVM2 media were significantly higher than in the KB medium, indicating that when nutrients are limiting, the ntrC gene is highly expressed to regulate the critical intracellular metabolic pathways. In addition, in the MMX-q medium, deletion of the ntrC gene upregulated the expression of ntrB in A. citrulli, which promoted the phosphorylation of NtrC to cope with nitrogen deficiency stress (results not shown). In summary, ntrC plays a key role in the nitrogen metabolism of A. citrulli.
The A. citrulli ntrC gene deletion mutant showed significantly heightened tolerance to oxidative, high salt, and copper ion-induced stress relative to Aac5, which is in agreement with previous reports [22,23]. In Azospirillum brasilense, phosphorylated NtrC is a transcriptional activator of the genes involved in nitrogen metabolism, which can promote the expression of the organic hydrogen peroxide resistance protein, Ohr [39]. Ohr belongs to the OsmC superfamily and has a detoxification effect on organic hydrogen peroxide [39]. In Aeromonas hydrophila, when ntrC was deleted, the expression levels of the NhaP type Na + /H + and K + /H + antiporter A0KFD8 were downregulated. This indicates that NtrC may use Na + /H + and K + /H + anti-transporters to cope with the stress of high osmotic Figure 8. Relative expression of ntrC in Acidovorax citrulli strain Aac5 in KB, MMX-q, and XVM2 media. qRT-PCR assays were conducted, and rpoB was used as an internal reference gene. Each treatment had three replicates, and the experiment was conducted three times. Asterisks indicate significant differences; error bars represent the standard errors of the means (p < 0.05, one-way ANOVA test).

Discussion
In this study, we constructed a ntrC deletion mutant and a corresponding complemented strain in the background of the A. citrulli wild-type strain Aac5. Through phenotype assays and qRT-PCR analysis, we investigated the role of the ntrC gene in nitrogen utilization, stress tolerance, and virulence against watermelon seedlings in A. citrulli.
We observed that the ntrC gene deletion mutant was unable to use potassium nitrate as the sole nitrogen source, but could grow with ammonium sulfate and urea as the sole nitrogen sources. In addition, the expression of the nitrate utilization-related gene nasS was significantly downregulated, and the expression of nasT was significantly upregulated in Aac5∆ntrC. NasS and NasT comprise the two-component system that regulates bacterial nitrate utilization [38]. We speculate that ntrC affects the activation of the nasS promoter and further affects the utilization of nitrate, which is consistent with previous studies [38]. The expression levels of ntrC by Aac5 in the MMX-q and XVM2 media were significantly higher than in the KB medium, indicating that when nutrients are limiting, the ntrC gene is highly expressed to regulate the critical intracellular metabolic pathways. In addition, in the MMX-q medium, deletion of the ntrC gene upregulated the expression of ntrB in A. citrulli, which promoted the phosphorylation of NtrC to cope with nitrogen deficiency stress (results not shown). In summary, ntrC plays a key role in the nitrogen metabolism of A. citrulli.
The A. citrulli ntrC gene deletion mutant showed significantly heightened tolerance to oxidative, high salt, and copper ion-induced stress relative to Aac5, which is in agreement with previous reports [22,23]. In Azospirillum brasilense, phosphorylated NtrC is a transcriptional activator of the genes involved in nitrogen metabolism, which can promote the expression of the organic hydrogen peroxide resistance protein, Ohr [39]. Ohr belongs to the OsmC superfamily and has a detoxification effect on organic hydrogen peroxide [39]. In Aeromonas hydrophila, when ntrC was deleted, the expression levels of the NhaP type Na + /H + and K + /H + antiporter A0KFD8 were downregulated. This indicates that NtrC may use Na + /H + and K + /H + anti-transporters to cope with the stress of high osmotic pressure [40]. In Escherichia coli, the expression of RpoS, one of the RNA polymerase σ factors, was upregulated, and the expression of RpoS-dependent genes contributed to the universal resistance of cells [41].
ntrC has been deeply studied in nitrogen-fixing bacteria. In recent years, it has been reported that ntrC also plays an important role in the process of virulence in pathogenic bacteria. An ntrC mutant of Pseudomonas aeruginosa showed a reduced ability to invade and cause damage in a high-density abscess model in vivo [24]. In this study, ntrC deletion significantly impaired the virulence of A. citrulli. Virulence-related phenotypic assays showed that in vivo colonization, in vitro growth, swimming motility, and twitching motility of the A. citrulli ntrC mutant were significantly reduced, while the biofilm-forming ability was significantly enhanced. The weakening of swimming and twitching motility not only directly affected the ability of A. citrulli to infect plant tissue; it also affected the formation of biofilms [42,43]. The decrease in vivo colonization and in vitro growth may be related to the metabolic slowdown after the deletion of ntrC. Although the ntrC A. citrulli mutant entered the logarithmic growth phase 8 h later than Aac5 in vitro, the population of the mutant strain was higher than that of Aac5 after entering the stationary phase. This observation was consistent with the phenotype associated with the deletion of the flagellum-related gene flgM in A. citrulli [44]. In Aac5∆ntrC, the expression of flagellum-related genes (flhD, flhC, fliA, and fliC) was significantly upregulated except for filM. Similarly, in the A. citrulli flgM deletion mutant, flhD and fliC genes were also significantly upregulated [44]. We speculate that flagellum-related genes may be closely related to the growth ability of A. citrulli. However, the swimming ability of the complemented strain was not restored to wild-type levels and was weaker than that of the ntrC mutant. These results were confirmed by the qRT-PCR assay. The expression levels of flhD, flhC, fliA, and fliC were not restored in the complemented strain, Aac5∆ntrCcomp, but were more upregulated than in Aac5∆ntrC (results not shown). In order to further explain this observation, the flagella of Aac5, Aac5∆ntrC, and Aac5∆ntrCcomp were observed, and their swarming motility was tested. We found that the deletion of ntrC did not affect the production of flagella nor the swarming motility of A. citrulli (results not shown). A recent study also found that swimming ability was completely lost after the deletion of a transcriptional regulatory factor, OxyR, in A. citrulli, and the complementary strain did not restore swimming ability [3]. One possibility is that the levels of NtrC may require precise regulation, possibly involving anti-sigma factors and anti-anti-sigma factors, for the proper function of flagellum-related genes [44][45][46]. In addition, the expression of ntrC was different between complementary and wild-type strains because the complementation is provided by a plasmid. The regulation of flagellum-related genes may require more precise regulation. Therefore, the regulatory mechanism of ntrC on swimming motility needs further investigation. In addition, the expression levels of the key A. citrulli T3SS genes hrpE, hrpX, and hrcJ were significantly downregulated in Aac5∆ntrC, while the change in the expression of hrpG was not significant. However, it was reported that in Xanthomonas oryzae pv. Oryzicola, the expression of hrpE was not regulated by hrpG and hrpX [47], and the iron transport family regulator Zur could regulate the expression of hrcC and hrpX, but not hrpG [48]. Therefore, we speculate that ntrC is a potential regulator upstream of hrpE and hrpX.
In conclusion, ntrC not only plays an important role in nitrogen utilization and stress tolerance, but also contributes to the virulence of A. citrulli. Because of its important role in the regulation of the nitrogen metabolism, ntrC has been widely studied in nitrogen-fixing bacteria, but there are few studies on virulence in pathogenic bacteria. Future work should continue to search for the downstream targets of ntrC in A. citrulli and explore its regulatory network to identify new targets to improve the management of pathogenic bacteria such as A. citrulli.