The relationships between the stringent response alarmone (p)ppGpp and persistence are dissimilar in different bacteria (Conlon et al., 2016; VandenBerg et al., 2016; Pacios et al., 2020). Previous research from our laboratory indicated that B. abortus was capable of forming multi-drug tolerant persister cells (Liu et al., 2023). B. abortus displayed a biphasic killing curve with a subpopulation of surviving persister cells during different antibiotic treatment, except with gentamicin. Moreover, deletion of the gene for Rsh (Δrsh) resulted in decreased persistence in either exponential phase or stationary phase with rifampicin treatment for 24 h (Liu et al., 2023). Therefore, we examined whether Rsh promoted persister cell production in B. abortus stationary phase during different antibiotic treatments for 36 h. Deletion of the gene for Rsh had no effect on persister cell formation in stationary phase in the presence of doxycycline (Figure 1A), polymyxin B (Figure 1B), ofloxacin (Figure 1C), or ampicillin (Figure 1D) compared with wild-type B. abortus. However, persister cell formation in stationary phase during ofloxacin treatment was reduced in the complemented strain (CΔrsh) compared with wild-type or Δrsh strains (p < 0.01) (Figure 1C). Moreover, the deletion strain showed significantly reduced persister cell levels in stationary phase during rifampicin and enrofloxacin treatment for 36 h (p < 0.01) (Figures 1E,F). These results indicate that Rsh participated in persister cell formation in B. abortus stationary phase during rifampicin and enrofloxacin treatment, but not when other antibiotic classes were tested.

We showed previously that persistence in Brucella was growth phase-dependent, by assessing exponential (12 h) and late stationary (72 h) phases during exposure to different antibiotic classes. These findings led us to assess whether Rsh affects persister cell production during different growth phases with different antibiotic treatments. Here, we used an analogous approach to investigate the extent to which Rsh impacted persistence in different growth phases following treatment with different antibiotics for 36 h. Wild-type and Δrsh strains were grown to exponential (12 h) and late stationary (72 h) phases (Liu et al., 2023). Aliquots were removed and cultured in TSB medium containing different antibiotics for 36 h. Similar with wild type, we found Rsh formed the time-dependent biphasic kill curves. There was a significant decrease in persister cells in Δrsh background compared with wild-type after doxycycline treatment in exponential phase (p < 0.01) (Figures 2A–D). However, no change in persister cell levels in the Δrsh mutant was detected in stationary phase (Figures 2A–D). After rifampicin or enrofloxacin treatment, persister cell formation in the Δrsh mutant was significantly lower than wild-type in both exponential and stationary phases (p < 0.01) (Figures 2E,F). In summary, we observed that persister cell formation by B. abortus was growth phase-dependent in both wild-type and Δrsh strains with different antibiotic treatments. These data suggest that persister cell production linked to Rsh depends both on growth phase and on antibiotic classes in B. abortus.

To assess whether reduced persister cell formation by the Δrsh strain was due to decreased sensitivity to antibiotics, we determined MICs of the mutant for rifampicin and enrofloxacin. The MIC values for wild-type, Δrsh, and CΔrsh strains were similar for both drugs (Table 1) which suggests that the persister cell defect in Δrsh is not due to increased antibiotic resistance. As rifampicin enters by diffusion through the cell wall (Namugenyi et al., 2017), we tested cell envelope permeability of the Δrsh mutant by ethidium bromide uptake assays. The Δrsh strain showed significantly elevated levels of ethidium bromide uptake relative to wild-type or CΔrsh strains (p < 0.01) (Figure 3A). The outer membrane protein Omp16 plays an important role in maintenance of cell wall integrity of Brucella (Zhi et al., 2020). Therefore, we tested expression of omp16 in wild-type, Δrsh and CΔrsh strains using qRT-PCR. Expression of omp16 was similar in the three strains in different growth phases (Figure 3B). Overall, these data suggest that reduced persister cell by Δrsh derivative is not due to enhanced susceptibility to rifampicin and enrofloxacin, but may be due to elevated cell envelope permeability which is unrelated to the Omp16 protein.

We next examined whether environmental stresses induced Rsh-mediated persister cell formation in the stationary phase in B. abortus. Survival of Δrsh decreased significantly compared with wild-type or CΔrsh strains after exposure to acid, phosphate, NaCl, or heat stress for 1 h (Figures 4A–D). Moreover, the absence of rsh increased persister cell levels during acid and heat stress after rifampicin treatment for 24 h compared with wild-type or CΔrsh strains (Figures 4A,D). The Δrsh mutant also displayed significantly decreased persister cell production during phosphate stress after rifampicin treatment compared with wild-type or complemented strains (Figure 4B). No differences in persister cell levels were observed for the three strains during NaCl stress (Figure 4C). The results suggest that Rsh plays an important role in certain stress-induced persister cell formation in the stationary phase in B. abortus.

Brucella is an intracellular pathogen that produces persister cells following internalization by macrophages (Mode et al., 2022; Liu et al., 2023). To investigate whether Rsh is also involved in persistence in macrophages, we first utilized RAW264.7 cells to investigate the role of Rsh on adhesion and invasion by B. abortus. Wild-type, Δrsh and CΔrsh strains showed similar adhesion properties to RAW264.7 cells (Figure 5A). However, invasion by Δrsh mutant was less compared with wild-type and CΔrsh strains (p < 0.05) (Figure 5B). We next investigated persister cell formation by Δrsh mutant in RAW264.7 cells after ampicillin, rifampicin, or enrofloxacin treatment for 24 h. Similar to our preceding results in vitro, there was no difference in persister cell formation by the mutant strain after ampicillin treatment compared with wild-type strain. In contrast, persister cell levels of Δrsh mutant were reduced significantly in RAW264.7 cells compared with wild-type strain in the presence of rifampicin and enrofloxacin (Figure 5C). These results show that Rsh influences B. abortus persistence in macrophages in the presence of rifampicin and enrofloxacin.

Rsh is up-regulated in B. abortus A19 under various stress conditions (Liu et al., 2023). Type II TA systems also regulate stress reactions and persister cell production during antibiotic exposure (Lobato-Márquez et al., 2016). Certain TA modules are expressed during nutrient starvation which induces persister cell formation (Zhou et al., 2021). We screened known and predicted type II TA modules in B. abortus by RT-qPCR for induction of transcription by various treatment including acid, phosphate, NaCl, temperature, H₂O₂ exposure, persister cells formation and nutrition starvation. Low pH and NaCl stress increased transcription of the brnT (p < 0.01) and mbcA (p < 0.05) TA modules, whereas other TA genes showed no changes (Figures 6A,C). Moreover, expression of mbcA also was elevated >4-fold under phosphate and H₂O₂ stresses (p < 0.01) (Figures 6B,E). Expression of other TA modules genes did not vary significantly under these conditions. Interestingly, TA genes were up-regulated (p < 0.01) during temperature stress, except for relE locus (Figure 6D). The pemK was significantly increased (p < 0.01) during persister cells formation (Figure 6F). And all gene were up-regulated (p < 0.01) during nutrition starvation (Figure 6G). These results indicate that some, but not all, type II TA modules are induced under stress conditions, persister cells formation and nutrition starvation in stationary phase in B. abortus.

The levels of alarmone (p)ppGpp depend on nutrient availability as well as other parameters, including oxygen concentration, pH, osmotic shock, temperature shift, and environmental stresses (Pacios et al., 2020). As the hipT induces (p)ppGpp synthesis in E. coli (Vang Nielsen et al., 2019), we assessed whether there also was a link between (p)ppGpp and type II TA modules in stationary phase cells of B. abortus. Expression of five TA modules was detected by qRT-PCR in the Δrsh mutant in stationary phase. Moreover, the brnT type II TA module was up-regulated significantly in the mutant compared with wild-type or CΔrsh strains (Figure 7A). In contrast, expression of the mbcA, ratA and relE loci decreased in the Δrsh strain (Figures 7B,C,E), but expression of pemK was unchanged (Figure 7D). Our results demonstrate that there may be regulatory links between Rsh and certain type II TA modules in B. abortus.

As expression of the mbcTA and brnT TA genes was altered in the Δrsh strain, we assessed whether these changes were of functional relevance. The brnTA promoter is located in the upstream region (543 bp) of the brnT gene (Heaton et al., 2012). We constructed pBB-brnTAp-lacZ and pBB-mbcTAp-lacZ reporter fusion plasmids that contained the upstream region (343 bp) of brnTA and the putative promoter of mbcTA (312 bp), respectively, fused to lacZ transcriptional reporter gene (Supplementary Figure S1). The reporter plasmids were transformed into wild-type B. abortus A19 and Δrsh strains. Promoter activity was detected by blue-white selection on plates with X-gal and by β-galactosidase assays. Expression of mbcTA promoter in wild-type and Δrsh strains generated blue colonies in contrast to the negative control which produced white colonies (Figure 8A). Assays of β-galactosidase activity showed that mbcTA promoter was expressed more strongly in wild-type background compared with Δrsh strain (Figure 8B) (p < 0.05). RT-qPCR confirmed that expression of mbcTA promoter was weaker in mutant strain than wild-type (Figure 8C). In contrast, the activity of brnTA promoter was not detected in either β-galactosidase or blue-white assays (data not shown). The additional 200 bp upstream of brnTA gene that is missing in the reporter used here may be important for promoter activity (Heaton et al., 2012). In summary, our results indicate that Rsh positively regulates mbcTA promoter in B. abortus.

To examine whether Rsh accelerated persister cell formation by regulating the mbcTA promoter in stationary phase. The wild-type, ∆rsh and ∆rsh with pBB-mbcTAp-lacZ reporter plasmid (∆rsh-mbcTAp) strains were cultured in stationary phase for 72 h and then were treated with rifampicin for 24 h to induce persister cells. Persister cell numbers increased significantly in ∆rsh-mbcTAp strain compared with ∆rsh strain (Figure 8D) which indicate that Rsh promotes persister cells formation by positively regulating mbcTA in B. abortus stationary phase after rifampicin treatment.

We examined ATP levels in stationary-phase persister cells after rifampicin treatment and observed that ATP levels in these cells decreased significantly (Figure 9A). We next sought to examine whether Rsh was involved in this phenomenon. ATP levels of ∆rsh strain increased significantly compared with wild-type in stationary phase (Figure 9B). As Rsh promoted persister cell formation by regulating mbcTA promoter in stationary phase (Figure 8D), we further tested ATP concentrations in ∆rsh-mbcTAp strain in stationary phase. ATP levels in ∆rsh-mbcTAp strain decreased significantly compared with wild-type or ∆rsh backgrounds (Figure 9B). These results suggest that ATP levels in rifampicin-tolerant persister cells are depressed in stationary phase in B. abortus, but that these levels increase significantly in stationary phase ∆rsh compared with wild-type or C∆rsh strains. Moreover, the ATP concentration decreased when mbcTA promoter was overexpressed in ∆rsh background.

Plasmids and strains are listed in Table 2. The B. abortus vaccine strain A19 was supplied by the Shaanxi Veterinary Drug Supervision Institute (shaanxi province) and was cultured on tryptic soy broth (TSB) agar or in liquid medium at 37°C and 180 rpm to mid-exponential phase (12 h), early stationary phase (24 h), or late stationary phase (72 h). E. coli was grown in Luria-Bertani (LB) liquid or agar medium at 37°C and 180 rpm. Antibiotics were added when required according to the characteristics of strains and plasmids. Antibiotics used in persister assays were doxycycline, polymyxin B, ofloxacin, ampicillin, rifampin, and enrofloxacin. Antibiotic concentrations used are specified below for individual experiments.

The promoter fragments of brnTA (brnTAp) and mbcTA (mbcTAp) were amplified and cloned into pBB-Amp-lacZ expression vector by enzyme digestion and ligation. Plasmid constructs were confirmed by PCR amplification and/or sequencing (Tsingke Biotech Co., Ltd., Beijing, China). The wild-type and Δrsh strains were grown to exponential phase, the cultures were centrifuged, and washed in ddH₂O to make competent cells. The pBB-brnTAp-lacZ and pBB-mbcTAp-lacZ plasmids were electro-transformed into wild-type (WT-brnTAp, WT-mbcTAp) and Δrsh competent cells (ΔRsh-mbcTAp,ΔRsh-brnTAp). Strains obtained in this study are summarized in Table 2. All primers are shown in Table 3.

Strains were grown to late stationary phase (72 h) in 8 mL of fresh TSB liquid medium at 37°C with shaking at 180 rpm. The cultures were plated onto TSB agar and the resulting CFUs were used as initial values. Doxycycline (1.2 μg/mL), polymyxin B (40 μg/mL), ofloxacin (40 μg/mL), ampicillin (8 μg/mL), rifampicin (40 μg/mL), or enrofloxacin (12 μg/mL) were added and samples were removed at 12 h, 24 h, and 36 h, and centrifuged at 6, 000 × g for 10 min. The pellets were collected and washed once in PBS and then diluted in PBS for viable counts.

In other experiments, bacteria were grown to exponential phase (12 h), (24 h) or late stationary phase (72 h) in 8 mL of fresh TSB liquid medium at 37°C with shaking at 180 rpm. Antibiotics were added as described above and incubation continued for 36 h. Samples before and after antibiotic treatment were collected by centrifugation at 6, 000 × g for 10 min. Cell pellets were washed once in PBS and then diluted in PBS for viable counts on TSB agar plates. The surviving colonies were determined in treated and untreated samples after 72 h of incubation at 37°C.

For adhesion assays, bacteria were incubated with macrophage cells for 1 h at 37°C in 5% CO₂ atmosphere. Infected cells were washed once in PBS (pH 7.2), collected, and then lysed with 0.1% Triton X-100. The lysates were removed and 10-fold serial dilutions were made with PBS to assess viable counts as the initial time point. The lysates were centrifuged, collected, and resuspended in TSB containing ampicillin (4 μg/mL), rifampicin (20 μg/mL), or enrofloxacin (6 μg/mL) for detection of intracellular persister cell formation at 2 h, 6 h, and 24 h. Samples were collected, washed once in PBS, and then 10-fold serially diluted in PBS for viable counts. The dilutions were plated onto TSB agar plates and incubated at 37°C for 72 h.

The MICs of wild-type B. abortus A19 and the Δrsh and CΔrsh derivatives on rifampicin and enrofloxacin were determined as previously reported using two-fold serial dilutions of antibiotics with TSB broth (Liu et al., 2023). The strains were grown to OD₆₀₀ = 0.6 in TSB broth and diluted to OD₆₀₀ = 0.01 in 5 mL fresh medium. Antibiotics were dispensed in 96-well polyethylene plates and cultures were added. Cultures without antibiotics served as controls. The cultures were incubated at 37°C for 36–48 h. The MIC was taken as the minimum concentration of antibiotic that significantly inhibited growth of bacteria relative to that of the positive control.

Strains were cultured to stationary phase in TSB medium at 37°C with shaking for 72 h. Stationary phase cultures were adjusted to 1–2 × 10⁸ CFU/mL. The cultures were exposed to different stress conditions for 1 h. For acid stress, cultures were incubated in TSB medium with the addition of HCl (pH 4.5). Phosphate starvation involved incubation in 0.01 M PBS (pH 7.0). For NaCl stress, the cultures were incubated in TSB medium with the addition of 2% NaCl. Temperature stress involved growth in TSB medium at 25°C whereas oxidative stress involved growing in TSB containing H₂O₂ (1 mM). For nutrition starvation, the cultures were harvested and incubated in PBS with with 0.05% Tween 80 for 4 h (Li et al., 2022). Untreated cultures in TSB medium were used as controls. The cultures were centrifuged, washed once in PBS, and the cell pellets were resuspended in TSB liquid medium containing rifampicin (40 μg/mL) with shaking at 37°C for 24 h for detection of persister formation. Persister cell assays were conducted as outlined above.

Adhesion and invasion assays were performed using macrophage cell line RAW264.7 (National Collection of Authenticated Cell Cultures). Briefly, bacterial strains were grown in TSB medium at 37°C with shaking for 72 h. The cultures were centrifuged at 6,000 × g for 10 min at 4°C, washed three times with PBS, and resuspended in Dulbecco’s modified Eagle medium (DMEM; Hyclone, United States). RAW264.7 cells were cultured in DMEM with 10% unheated fetal bovine serum (FBS) using 12-well culture plates at a density of 8 × 10⁵ cells per well. Stationary phase bacterial cultures were added at a multiplicity of infection of 200. Infected cells were centrifuged at 100 × g for 5 min. The adherence of bacteria to macrophages was continued for 1 h at 37°C in 5% CO₂ atmosphere. Infected cells were washed once in PBS (pH 7.2), collected, and lysed with 0.1% Triton X-100. The lysates were removed for 10-fold serial dilutions in PBS for viable counts. Invasion of B. abortus in macrophage cells was performed by adding gentamycin (50 μg/mL) to medium and incubating for an additional 1 h at 37°C in 5% CO₂ atmosphere. Infected cells also were lysed with 0.1% Triton X-100 for viable counts. Dilutions were plated onto TSB agar plates and incubated at 37°C for 72 h.

Ethidium bromide uptake was tested as described previously with some minor modifications (Namugenyi et al., 2017). B. abortus strains were grown to stationary phase (72 h). The cultures were centrifuged, washed once with PBS, and resuspended in PBS to OD₆₀₀ = 0.6–0.8. Ethidium bromide was added to a final concentration of 2 μg/mL. Uptake was measured using 96-well microplates with excitation at 544 nm and emission at 590 nm. The uptake rates were calculated using data in the linear range between 0 and 30 min. Each experiment was repeated at least three times.

Total RNA was extracted from stationary-phase cultures of B. abortus after 72 h of growth using Trizol (Takara, Bio, Dalian, China) by conventional methods. The cDNA samples were generated with the PrimeScript™ RT reagent Kit (Takara) according to the manufacturer’s instructions. Real-time quantitative primers are listed in Table 3. Quantitative real-time PCRs (qRT-PCR) was performed using SYBR^® green enzyme (Vazyme, Nanjing, China). Fold change was calculated by the 2^−ΔΔCt method. Target genes were normalized internally to 16S rRNA.

In brief, the pBB-brnTAp-lacZ and pBB-mbcTAp-lacZ report plasmids were transformed into wild-type and ∆rsh competent cells. Single colonies were selected and cultured to stationary phase. The cultures were collected and washed twice with cold PBS. 100 μL of bacterial suspensions were added to 2 mL EP tubes. 600 μL of pre-cooled Z buffer (100 mM Na₂HPO₄, 40 mM NaH₂PO₄, 10 mM KCl, 1 mM MgSO₄, 5.4 μL of β-mercaptoethanol, pH 7.5) were added to permeabilize bacteria. 200 μL of resuspended cultures were used to determine OD₆₀₀ values. 50 μL of 0.1% SDS and 100 μL of chloroform were added. The samples were mixed and incubated at room temperature for 5 min. O-nitrophenyl-β-D-galactoside (ONPG) (200 μL of 4 mg/mL) was added and incubated at room temperature for 10 min. 1 M Na₂CO₃ (500 μL) was added to terminate the reactions. Samples were centrifuged at 12,000 × g for 5 min and 200 μL of supernatants were collected to determine OD₄₂₀, OD₅₅₀, and OD₆₀₀ values. Miller units were calculated using the formula: Miller units = 1,000 × (OD₄₂₀–1.75 × OD₅₅₀)/T (min) × V (mL) × OD₆₀₀. In addition, stationary phase cultures were diluted 10-fold and spotted onto TSB agar plates containing X-gal (2 mg/mL) and incubated at 37°C. Colony color was observed.

ATP levels of persister cells in stationary phase and stationary cultures were measured using an ATP assay kit (Beyotime Biotechnology, Shanghai, China) according to the manufacturer’s instructions.

All experiments were performed as three biological replicates. GraphPad Prism 8.0 (San Diego, CA, USA) was used for data graphing. Data were analysed with one-way or two-way analysis of variance (ANOVA) using SPSS 22.0 (IBM, Chicago, Ill, USA). p > 0.05 (ns) was considered not significant. p < 0.05 (*) was considered significant. p < 0.01 (**) was considered highly significant. Data points indicate the mean values of the results of at least three biological replicates and error bars indicate standard errors of the mean.