R. aquatilis JZ-GX1 is a plant growth-promoting bacterium isolated from the rhizosphere soil of 28-year-old P. massoniana in Nanning, Guangxi, and is stored in the Culture Preservation Center of China (CCTCC, No: M2012439). After activation, strain JZ-GX1 was cultured overnight in LB liquid medium at 28°C (Kong et al., 2020). LB solid medium was prepared with 10 g of peptone, 5 g of yeast powder, 10 g of NaCl, and 15–20 g of agar (pH 7.2), and the NaCl concentration of the medium was set to 0, 3, 6, and 9% (0, 0.51, 1.02, and 1.53 mol/L, respectively). The activated strains were inoculated into the prepared media and cultured to the logarithmic phase at 28°C with shaking 200 r/min.

Strain JZ-GX1 was inoculated on plates with different salt concentrations, and its growth on was observed after 24–48 h. Then, 50 μL of an activated bacterial solution was inoculated into test tubes containing 4,950 μL of LB medium with different salt concentrations and shaken. Subsequently, the culture broth was pipetted into each well of a 96-well plate and cultured in a microplate reader (Bioscreen C, FP-110-C, Finland) at 28°C to determine the growth curves, with measurements automatically collected every 2 h for 48 h. The activated strain was inoculated (1%) in LB medium with different salt concentrations, and after cultivating for 24, 48, and 72 h, the cells were collected by centrifugation and dried to a constant weight to determine the dry weight.

Scanning electron microscopy observations of the tested strain JZ-GX1 were obtained as described by Castelijn et al. (2012). For sample preparation, the cells were first fixed with glutaraldehyde, washed with 0.2 M phosphate buffer (pH 7.4) and dehydrated with gradient ethanol. The dehydrated sample was dried with a CO₂ critical point desiccator (liquid CO₂, EMITECH K850), after which the dried sample was divided into fragments of suitable length, glued to the table and sprayed with gold (HITACHE-1010). Subsequently, the morphology of strain JZ-GX1 cells in the presence of different salt concentrations was observed with a scanning electron microscope (QUANTA200, FEI, United States).

The EPS content in the culture medium was determined using the phenol-sulfuric acid method (Wang, 2015).

LB media with different salinities (0, 3, 6, and 9%) was prepared. The bacteria were inoculated (1%) and cultured on a shaking table (28°C, 180 rpm) for 1, 2, 3, and 4 days. Subsequently, 500 μL of bacterial suspension was collected and centrifuged at 10,000 rpm for 3 min, and the supernatant was discarded.

Samples were prepared to assess the proline content of cells using a ratio of bacteria (10⁴cfu/mL) to extract volume (mL) of 500:1 (ultrasonic power 200 w, 3 s, interval 10 s, repeated 30 times). Thee samples were extracted in a 95°C water bath for 10 min. Then, after cooling, the samples were centrifuged at 25°C for 10 min at 10,000 rpm, and the supernatant was reserved. Samples were prepared to assess the trehalose content of cells by lysing bacteria with an ultrasonicator at a ratio of bacteria (10⁴ cfu/ml) to extract volume (mL) of 500:1 (ultrasonic power 200 w, 3 s, interval 10 s, repeated 30 times). Then, after incubating at room temperature for 45 min, the samples were shaken 3–5 times, cooled, centrifuged at 8,000 rpm at 25°C for 10 min, and the supernatant was reserved. Samples were prepared to assess the betaine contents of cells by continuously shaking the samples with 1.2 mL of water for 60 min prior to extraction for 30 s. Subsequently, 400 μL of extract was added, mixed thoroughly, and centrifuged at 10,000 rpm at 25°C for 10 min to obtain the supernatant.

Three replicates of each treatment were evaluated, and the proline, betaine and trehalose contents accumulated by strain JZ-GX1 in the presence of different salt concentrations were determined using a commercial kit (Suzhou Keming Biotechnology Co., Ltd.).

Na⁺ and K⁺ were extracted according to Nagata’s method, and the concentrations of Na⁺ and K⁺ were determined by flame photometry (Nagata et al., 2005).

For quantitative RT-PCR, strain JZ-GX1 was cultured in LB medium for 72 h with shaking, after which RNA was extracted from cells using the TRIzol method. cDNA samples were prepared using HiScript II Q Select RT SuperMix for qPCR (China). The expression of salt tolerance-related genes was determined by qRT-PCR with an ABI 7500 instrument (Applied Biosystems, United States), and atpD was used as an internal control (Li et al., 2019). Three genes related to salt tolerance were selected from the genome of R. aquatilis JZ-GX1 (unpublished) for analysis (Table 1). The relative gene expression changes were calculated with the 2^−ΔΔCT method. The RT-PCR assay consisted of three independent experiments, with three replicates performed for each experiment.

The siderophile production capacity of strain JZ-GX1 was determined as described in the literature (Schywn and Nielands, 1987); the indole acetic acid production capacity was assessed in accordance with the literature (Glickmann and Dessaux, 1995); and the phosphorus dissolving ability was determined according to the method reported by Li et al. (2012).

Tomato seeds with similar breakage, size and fullness were selected, surface-sterilized with 10% hydrogen peroxide for 10 min, and washed thoroughly with aseptic distilled water to remove the residual hydrogen peroxide, after which the seeds were air-dried. Two groups were established: the control group (H₂O and LB) and the experimental group (JZ-GX1). The surface-sterilized seeds were soaked in sterile water and then cultured in LB liquid medium with 1 × 10⁷ cfu/mL JZ-GX1 for 24 h. Subsequently, seeds of the same size were evenly distributed in a Petri dish (20 seeds in each dish) containing 10 mL of 100 mmol/L NaCl solution. Four replicates of each treatment were included. Starting on the 2nd day of observation of the artificial climate box, the number of germinated seeds based on a radicle length of 0.2 cm as the germination standard was counted every day. On the 7th day, 10 plants from each treatment were randomly selected to measure their germination length, root length, and fresh weight and to calculate their germination rate and germination energy (Fu et al., 2017).

The data were analyzed by analysis of variance and Duncan’s multiple comparison using SPSS 17.0, and the standard errors of the mean values were calculated (P < 0.05).

The growth of bacteria in media with high salt concentrations is an important indicator of their salt tolerance. R. aquatilis JZ-GX1 was able to grow on plates containing 0–9% NaCl, indicating that strain JZ-GX1 is a moderately halophilic bacterium. In the presence of 9% NaCl concentration of, the growth rate was slow, and the colonies were small. Thus, the decrease in the growth rate of strain JZ-GX1 was related to the increase in salt concentration (Figure 1A).

A microplate reader was used to generate growth curves of strain JZ-GX1 cultured in LB medium with different salt concentrations. Strain JZ-GX1 had the shortest lag phase when cultured in the presence of 1%, while the lag phase reached 18 h in the presence of 9% NaCl. After entering the logarithmic phase of growth, the fastest growth rate of strain JZ-GX1 was obtained with a salt concentration of 3%, and the lowest growth rate of strain JZ-GX1 was observed with a salt concentration of 9%. After entering the stationary phase, the maximum OD600 value exceeded 1.0 for strain JZ-GX1 when cultured in the presence of 3% NaCl, with this phase lasting longer when cultured in medium containing 0 and 3% NaCl than 6% NaCl. Increasing the NaCl concentration to 9% significantly prolonged the lag phase, decreased the growth rate at each growth period, and significantly inhibited the growth of R. aquatilis JZ-GX1. In addition, the lowest OD value observed after entering the stationary phase was detected at this salt concentration, with a maximal OD value of 0.241. Furthermore, almost no growth in the logarithmic phase was detected at this salt concentration, and the growth was seriously inhibited (Figure 1B). A further increase in the salt concentration reduced the tolerance and survival rate of the bacteria.

In terms of biomass, salt stress (3, 6, and 9%) increased the biomass of strain JZ-GX1 compared to that obtained in the absence of salt stress. At a salt concentration of 0%, no significant difference in the cell dry weight was observed over time. At a salt concentration of 3% NaCl, the dry weight of JZ-GX1 decreased slowly over time. At salt concentrations of 6 and 9%, the dry weight of strain JZ-GX1 first increased and then decreased over time (Figure 1C).

As shown in Figure 2, the cell morphology of R. aquatilis JZ-GX1 was altered under different levels of salt stress. In the presence of 0% NaCl, the cells were relatively short and regular in shape and presented many visible folds on the surface. In the presence of 3% NaCl, the bacteria became slightly longer and showed a sparse arrangement, and the visible folds on the cell surface were decreased and relatively smooth. In medium containing 6% NaCl, the bacteria adsorbed and adhered together, presumably due to the secretion of EPSs by the cells. In the presence of 9% NaCl, the bacteria were notably longer and arranged in a disorderly manner, cell fragments were detected, the cell morphology was damaged, and breakage was observed during the cell division. In general, increases in the degree of salt stress were associated with increases in cell length. The length and width of the cells treated with 9% salt were 5.12- and 1.41-fold higher than those observed with 0% salt, respectively. The results showed that strain JZ-GX1 can resist a high salt environment by altering its cell morphology.

The production of EPSs by many halotolerant bacteria is a growth strategy that can promote the formation of a protective layer around cells under adverse conditions to protect cells from the external environment. Therefore, the potential role of EPSs in the salt tolerance of JZ-GX1 was evaluated. As shown in Figure 3, strain JZ-GX1 could secrete EPSs in the presence of different salinities, and the EPS yield increased with increasing salt concentration (0, 3, 6, and 9% NaCl), reaching maximum levels at 9% NaCl (60.6983 mg/L).

To determine the type and content of compatible solutes in the JZ-GX1 strain, the accumulation of proline, betaine and trehalose in the presence of different concentrations of NaCl (0, 3, 6, and 9%) was assessed. As shown in Figure 4, the accumulation of proline and trehalose gradually increased with increases in the NaCl concentration on the 1st day, but no obvious change in the betaine content was detected. The highest accumulation of proline, betaine and trehalose was detected with 9% NaCl on the 1st, 2nd, and 3rd days, respectively, and on the 4th day of fermentation with 6% NaCl. Compared to the results obtained under salt-free conditions, the contents of proline, betaine and trehalose increased by 28.15, 46.55, and 66.63%, respectively. In addition, on the 1st day, strain JZ-GX1 primarily relied on proline to resist salt stress, while trehalose was relied upon at later stage, indicating that were selective in the accumulation of soluble solutes. These results suggest that strain JZ-GX1 may alleviate salt stress by regulating the content of compatible solutes in vivo, and the accumulation of trehalose is the primary protective mechanism used by this bacterium against high salt stress.

R. aquatilis JZ-GX1 cells were analyzed by flame photometry, and a large amount of Na⁺ was detected in the intracellular/extracellular spaces (Figure 5). The concentration of Na⁺ in strain JZ-GX1 was lower than that in LB liquid medium (CK), a finding that was observed both intracellularly and extracellularly. Analysis of the accumulation of K⁺ showed that the intracellular K⁺ concentration was lower than that of CK, whereas the extracellular K⁺ concentration was higher than that of CK. Further analysis of the extracellular Na⁺/K⁺ concentrations showed that the intracellular and extracellular Na⁺/K⁺ concentrations of strain JZ-GX1 were lower than those of CK under the four salt concentration gradients. These findings show that R. aquatilis JZ-GX1 can accumulate K⁺ and increase the Na⁺/K⁺ ratio to cope with high salt environments, demonstrating another osmotic adaptation mechanism of R. aquatilis JZ-GX1.

To further elucidate the molecular mechanism underlying the salt tolerance of JZ-GX1, the target genes NhaB (encoding Na⁺/H⁺ antiporter NhaB), lgbT (encoding L-proline/glycine) and betB (encoding betaine) were selected to analyze the effect of NaCl on the expression of salt tolerance-related genes in strain JZ-GX1 at the transcriptional level. As shown in Figure 6, the level of NhaB transcription decreased with increasing salinity (0, 3, and 6% NaCl). In contrast, whereas the transcriptional levels of lgbT and betB first increased and then decreased with increasing salinity, the maximal values of these genes were obtained with 3% NaCl and reached 2.23- and 2.94-fold higher values than those observed in the absence of NaCl, respectively.

Some PGPRs can promote plant growth under stress by secreting IAA. Therefore, the ability of strain JZ-GX1 to secrete IAA was also assessed in the present study, and the results are shown in Figure 7A. Quantitative analysis showed that strain JZ-GX1 could secrete a high concentration of IAA, even in TSB medium with a salt concentration of 3%. Siderophores are bioactive substances secreted by PGPRs that can increase the absorption of iron in plants and inhibit the growth of pathogens. In the present study, the ability of strain JZ-GX1 to synthesize siderophores was determined. As shown in Figure 7B, the siderophore production of strain JZ-GX1 first increased and then decreased as the salt concentration increased, and the highest content of siderophores (44.93%) was obtained with a salt concentration of 1%. Appropriate salt concentrations could promote the secretion of siderophores by strain JZ-GX1, but the level of secreted siderophores began to decrease when the salt concentration reached 4%. By producing iron carriers, strain JZ-GX1 can help plants obtain iron from soil while controlling plant pathogens. An analysis of the ability of strain JZ-GX1 to dissolve inorganic phosphorus under different salt concentrations showed that increases in the salt concentration significantly improved the phosphorus-solubilizing ability of strain JZ-GX1 (Figure 7C).

To determine whether strain JZ-GX1 can promote plant growth under salt stress, we used tomato as the target plant (Figure 8A). After being subjected to salt stress for 7 days, the fresh weight of the seedlings treated with strain JZ-GX1 was 23.97 and 50.005% higher than those obtained with the H₂O and LB treatments, respectively (Figure 8B). In addition, the primary root length of the JZ-GX1 strain-treated seedlings was approximately 1.9- and 3.2-fold longer than those of the seedlings subjected to the H₂O and LB treatments, respectively (Figure 8C). Furthermore, the stem length of the seedlings treated with strain JZ-GX1 was 41.88 and 144.64% higher than those obtained with the H₂O and LB treatments, respectively (Figure 8D). As shown in Figures 8E,F, treatment with strain JZ-GX1 significantly increased the germination rate and germination energy of tomato seedlings under salt stress. The results showed that the germination percentage and germination potential of tomato seeds treated with strain JZ-GX1 were significantly higher than those of the seeds treated with H₂O and LB, and the fresh weight, primary root length and stem length of the tomato seedlings were significantly increased by treatment with strain JZ-GX1.