Cold shock protein A plays an important role in the stress adaptation and virulence of Brucella melitensis

Authors

  • Zhen Wang,

    1. Key Laboratory of Animal Epidemiology and Zoonosis of the Ministry of Agriculture, College of Veterinary Medicine, China Agricultural University, Beijing, China
    2. Animal Science and Technology College, Beijing University of Agriculture, Beijing, China
    Search for more papers by this author
  • Shuangshan Wang,

    1. College of Biological Sciences, Anyang Institute of Technology, Anyang, China
    Search for more papers by this author
  • Qingmin Wu

    Corresponding author
    1. Key Laboratory of Animal Epidemiology and Zoonosis of the Ministry of Agriculture, College of Veterinary Medicine, China Agricultural University, Beijing, China
    • Correspondence: Qingmin Wu, Key Laboratory of Animal Epidemiology and Zoonosis of the Ministry of Agriculture, College of Veterinary Medicine, China Agricultural University, Beijing 100193, China. Tel.: +86 10 6273 3901; fax: +86 10 6273 3901; e-mail: wuqm@cau.edu.cn

    Search for more papers by this author

Abstract

Brucella melitensis is a facultative intracellular pathogen that mainly resides within macrophages. The mechanisms employed by Brucella to adapt to harsh intracellular environments and survive within host macrophages are not clearly understood. Here, we constructed a cspA gene deletion mutant, NIΔcspA, that did not exhibit any discernible growth defect at a normal culture temperature (37 °C) or at a low temperature (15 °C). However, expression of the cspA gene in Brucella was induced by cold, acidic, and oxidative conditions, as determined via quantitative reverse transcription PCR. Unlike its parental strain, B. melitensis NI, the NIΔcspA mutant showed an increased sensitivity to acidic and H2O2 stresses, especially during the mid-log-phase, and these stress conditions would presumably be encountered by bacteria during intracellular infections. Moreover, macrophage and mouse infection assays indicated that the NIΔcspA mutant fails to replicate in cultured J774.A1 murine macrophages and is rapidly cleared from the spleens of experimentally infected BALB/c mice. These findings suggest that the Brucella cspA gene makes an essential contribution to virulence in vitro and in vivo, most likely by allowing brucellae to adapt appropriately to the harsh environmental conditions encountered within host macrophages.

Introduction

Brucella spp. are facultative intracellular pathogens that cause abortion and infertility in animals and severe debilitating febrile illness in humans. During the course of infection, Brucella reside and replicate within specialized compartments referred to as Brucella-containing vacuoles (BCVs) by their ability to resist bactericidal compounds released by the host (Martirosyan et al., 2011), such as reactive oxygen and oxygen intermediates. Within BCVs, Brucella have developed strategies to adapt to the harsh intracellular environment, including the acidic pH and the low nutrient availability (Baldwin & Winter, 1994). The ability of brucellae to survive and replicate within macrophages is essential for their virulence (Celli & Grovel, 2004). Although many stress-associated proteins (Latimer et al., 1992; Robertson & Roop, 1999; Robertson et al., 2000; Zhang et al., 2009) and virulence-related factors (Seleem et al., 2008) have been identified in Brucella, the mechanisms by which Brucella resist the harsh intracellular environment and survive and replicate within macrophages are not fully understood.

Previous research showed that a mutant containing a transposon insertion of an intergenic space between the genes encoding a cold shock protein (CSP) (BMEI0498) and soluble lytic transglycosylase (BMEI0499) in the genome of B. melitensis 16M displays significantly attenuated growth in macrophages (Wu et al., 2006); BMEI0499 was subsequently determined not to be associated with this phenotype. CSPs exist in all organisms, from prokaryotes to eukaryotes, and are required for cold adaptation through changing the composition of the cytoplasmic membrane and initiating the synthesis of a set of specific proteins (Phadtare, 2004). Many CSPs have been identified in Escherichia coli, among which the CspA family, which consists of nine proteins (CspA to CspI), has been studied in detail (Ermolenko & Makhatadze, 2002; Phadtare, 2004). The members of this protein family have been shown to act as RNA chaperones (Jiang et al., 1997) that regulate their own gene expression (Bae et al., 1997) and the expression of many other genes (Phadtare & Inouye, 2004). CspA, CspB, CspE, CspG, and CspI have been linked to the modulation of cold adaptation functions (Goldstein et al., 1990; Nakashima et al., 1996). Moreover, the CspC and CspE proteins have been implicated in chromosomal condensation, cell division, and the regulation of the RpoS and UspA stress response proteins (Bae et al., 1999; Phadtare & Inouye, 2001). In B. subtilis and Listeria monocytogenes, three Csps (CspA, CspB, and CspD) have been reported to be associated with the regulation of growth under both normal and cold conditions and stress adaptation responses (Graumann et al., 1997; Weber & Marahiel, 2002; Schmid et al., 2009). Therefore, we hypothesized that the CspA protein may be related to the virulence of Brucella through regulating the stress adaptation abilities of these bacteria.

In B. melitensis NI, there are four Csp-encoding genes: BMNI_I1459, BMNI_I1441, BMNI_I0436, and BMNI_I1229. The product of BMNI_I1459 (the gene ortholog of BMEI0498) shows 54% amino acid sequence identity to the CspA protein of E. coli, which is the highest percentage of identity among the nine members of the CspA family. Therefore, we designated BMNI_I1459 the cspA gene and its product the CspA protein. To further investigate the effects of CspA in Brucella, in the present study, we used B. melitensis NI as the parent strain and constructed a cspA gene deletion mutant. Then, the growth characteristics, cspA gene expression patterns, stress responses, and in vivo and in vitro survival of the cspA gene deletion mutant were assessed. Our results indicated that the CspA protein not only contributes to generalized stress resistance, but is also a major determinant of virulence in macrophages and mice. This study will shed light on the novel characteristics of CspA in Brucella spp.

Materials and methods

Ethics statement

All animal research was approved by the Beijing Association for Science and Technology. The approval ID is SYXK (Beijing) 2007-0001, and the animal research complied with the guidelines of Beijing laboratory animal welfare and ethics of Beijing Administration Committee of Laboratory Animals.

Bacterial strains and media

Brucella melitensis NI is an epidemic strain isolated from an aborted bovine fetus in China. This strain is also referred to as smooth virulent B. melitensis strain biovar 3 and induces abortion in pregnant cattle, sheep, and goats. Complete NI genomes have been sequenced and are available under GenBank accession numbers CP002931 and CP002932. All of the Brucella strains, including the parent strain and the derived mutant, were routinely grown or incubation in tryptic soy broth (TSB) or tryptic soy agar (TSA) or minimal medium (0.5% lactic acid, 3% glycerol, 0.75% NaCl, 1% K2HO4, 0.01% Na2S2O3·5H2O, 10 μg mL−1 Mg++, 0.1 μg mL−1 Fe++, 0.1 μg mL−1 Mn, 0.21 μg mL−1 thiamine·HCl, 0.2 μg mL−1 nicotinic acid, 0.04 μg mL−1 calcium pantothenate, 0.001 μg mL−1 biotin, 5 mg mL−1 glutamate; pH 6.8–7.0 with NaOH). The pH of minimal media was adjusted with HCl. E. coli strains were grown on Luria–Bertani (LB) plates overnight at 37 °C, with or without supplemental ampicillin (100 mg L−1) and chloromycetin (30 mg L−1; Supporting Information, Table S1). All work with the live virulent Brucella strains was performed in the biosafety level 3 facilities of China Agricultural University.

Mice

Female BALB/c mice (aged 4–6 weeks) were purchased from the Wei Tong Li Hua Laboratory Animal Services Centre (Beijing, China), bred in individually ventilated cage rack systems, and subsequently transferred to the biosafety level 3 facilities of China Agricultural University at the beginning of the experiments. All experiments involving animals followed the regulations enacted by the Beijing Administration Office for Laboratory Animals.

Construction of the cspA deletion mutant

The procedures for constructing the cspA deletion mutant were performed according to a previous study (Wang et al., 2013). Briefly, 5′ and 3′ fragments of approximately 500 bp flanking the cspA gene were amplified from B. melitensis NI genomic DNA with the primers shown in Table S2, then gel-purified, and used as templates for a second round of overlapping PCR. The forward primer for the 5′ fragment and the reverse primer for the 3′ fragment were utilized in a second round of PCR to join the 5′ and 3′ fragments. Then, the ΔcspA fragment was cloned into pEX18Ap to generate the recombinant plasmid pEX18Ap-ΔcspA. Finally, pEX18Ap-ΔcspA was introduced into B. melitensis NI via electroporation. Brucella colonies that were sensitive to ampicillin (Amps) were selected on a sucrose-containing medium (Sucr). The cspA deletion mutant was then verified by PCR and sequencing analysis and is hereafter referred to as NIΔcspA.

Construction of the complemented strain of the cspA deletion mutant

To construct the complemented strain of NIΔcspA, the complete cspA gene and the upstream and downstream noncoding sequences were amplified from B. melitensis NI genomic DNA via PCR using the primers shown in Table S2. The resulting PCR products were digested with SmaI and HindIII and then ligated into the pBBR1MCS plasmid (Kovach et al., 1994) digested with the same enzymes. The resultant recombinant vector, pBBRcspA, was subsequently electroporated into the NIΔcspA deletion mutant. The complementation strain, loaded with pBBRcspA, was selected on TSA plates containing chloromycetin. Finally, the selected complementation strain was verified through PCR and designated NIΔcspA PBBRcspA.

Growth characteristics of the cspA deletion mutant at different temperatures

To investigate the effect of the cspA gene deletion on the growth of B. melitensis, one colony from each strain (the NIΔcspA mutant, the NIΔcspA PBBRcspA complementation strain, and the B. melitensis NI parent strain) was inoculated into 5 mL of TSB medium and grown to mid-log-phase in a shaking incubator at 37 °C. These cultures were then adjusted to the same concentration (CFU mL−1) and subsequently used for the growth curve analysis. Briefly, a 20-μL sample of each strain was inoculated into 5 mL of TSB medium, followed by incubation in a shaking incubator at 37 °C or 15 °C. The CFUs mL−1 were determined at 0, 24, 48, 72, and 96 h postincubation. Growth characteristics were evaluated through analysis of the growth curves of the cspA deletion mutant at different temperatures.

Stress response assays

To investigate the effect of the cspA gene deletion on the stress adaptation ability of B. melitensis, acid challenge and oxidation resistance assays were performed in NIΔcspA, NIΔcspA PBBRcspA, and B. melitensis NI, as previously described with some modifications (Kulakov et al., 1997; Robertson & Roop, 1999; Zhang et al., 2009; Roset et al., 2013). Each strain was grown to both mid-log- and stationary phases in 5 mL of TSB medium (pH 7.3, 37 °C, shaking) following inoculation of a single colony, and the cultures were then adjusted to the same concentration (CFU mL−1), which was used for the later experiments.

Acid stress experiment

A 200-μL sample of each strain that had been cultured for the various time periods stated above was concentrated, washed, resuspended in 200 μL of pH 4.4 minimal medium (for submitted cultures), and incubated for 2 h at 37 °C or 15 °C. Next, the submitted cultures were washed and resuspended in pH 3.4 minimal medium for challenge and incubated for 2 h at 37 °C or 15 °C. The cultures were then immediately serially diluted and plated to determine their viability postchallenge (Robertson & Roop, 1999; Zhang et al., 2009). The percent survival was calculated by dividing the CFUs obtained 2 h postacid challenge by those obtained prior to acid challenge for each time point and temperature, multiplied by 100.

Oxidative resistance assay

A 200-μL sample of each strain cultured for the various time periods stated above was supplemented with 200 μL of H2O2 (freshly diluted in PBS) at final concentrations of 1 mM, 2.5 mM, and 5 mM. After incubation for 1 h at 37 °C or 15 °C, the cells were rapidly diluted with PBS and plated onto TSA plates. The results are expressed as the percentage of viable bacteria relative to the bacterial numbers prior to the H2O2 challenge. In addition, the bacterial resistance to oxidative stress was measured through a disk diffusion assay. Cultures of Brucella in TSB medium were spread onto TSA plates. A 5-mm-diameter Whatman 3M paper disk containing H2O2 (5 μL of a 10% solution per disk) was placed in the center of each plate. After 3 days of incubation at 37 °C under 5% CO2 and 10 days of incubation at 15 °C, the diameter of the bacteria-free zone was determined as a measure of resistance.

CspA gene expression levels in B. melitensis NI under different conditions

The level of the cspA gene expression in B. melitensis NI was detected via RT-PCR. Briefly, B. melitensis NI was grown to mid-log- and stationary phases in 40 mL of TSB at 37 °C in a shaking incubator. For each culture phase, 10 mL of the bacterial sample was challenged and incubated at 15 °C for 2 h; 10 mL of the bacterial sample was challenged under acidic conditions at 37 °C, as described above; 10 mL of the bacterial sample was challenged under oxidative conditions (1 mM H2O2) at 37 °C; and 10 mL of the sample was maintained at 37 °C. Then, total RNA samples were prepared for the challenged and unchallenged B. melitensis strains from each set of conditions. DNA was removed using the DNase (Ambion, Foster City, CA), and the RNA samples were reverse-transcribed into cDNA using random oligonucleotide hexamers and the Fermentas First-Strand cDNA synthesis kit (Thermo Fisher Scientific, Bremen, Germany), according to the manufacturer's protocol. The obtained RNA and cDNA were quantified using a NanoDrop spectrophotometer (NanoDrop, Wilmington, DE). The resulting cDNA samples were subjected to quantitative real-time PCR using the Power SYBR Green PCR System. The sense and antisense primers for cspA are shown in Table S2. The level of the translation initiation factor IF-1 gene expression was used to normalize all of the obtained values. Control reactions were performed without RT to detect genomic DNA contamination. PCR amplification was performed according to a previous report (Liu et al., 2012). Each cDNA was analyzed in triplicate, after which the average threshold cycle (Ct) was calculated per sample. Ratios were calculated using the ΔΔCt method (Livak & Schmittgen, 2001).

Cell infection assay

To investigate the intracellular survival of the NIΔcspA mutant, the NIΔcspA PBBRcspA complementation strain, and B. melitensis NI, J774.A1 murine macrophage infection assays were performed, as previously reported with some modifications (Robertson & Roop, 1999; Zhang et al., 2009). Briefly, monolayers of cells were cultured in 24-well plates and infected with the Brucella strains at a multiplicity of infection (MOI) of 200 CFU. To synchronize the infection, the infected plates were centrifuged at 200 g. for 5 min at room temperature, followed by a 20-min incubation at 37 °C in an atmosphere containing 5% (v/v) CO2. Then, the cells were washed three times with PBS and incubated in medium containing gentamycin (50 μg mL−1) at 37 °C under 5% CO2 until the end of the infection period. At 1, 4, 12, 24, 36, and 72 h p.i., the cells were washed and lysed in sterile 0.5% (v/v) Tween 20 water, and the number of surviving intracellular bacteria was determined through serial dilutions, followed by plating on TSA or TSA supplemented with chloromycetin. Three replicate wells for each strain were evaluated at each time point, and the presented results represent the averages from at least three separate experiments.

Virulence in BALB/c mice

Twenty-five mice were intraperitoneally inoculated with a dose of 106 CFU in 0.1 mL of phosphate-buffered saline (PBS) of either the NIΔcspA mutant or the B. melitensis NI parent strain. Five infected mice from each infected group were randomly selected and euthanized via carbon dioxide asphyxiation at 1, 3, 6, 9, and 11 weeks postinoculation. At each time point, the spleens were collected aseptically, homogenized in 1 mL of PBS, and then serially diluted (1/10, 1/100, and 1/1000). A 200-μL aliquot of each dilution and undiluted spleen homogenates was plated on TSA plates, incubated for 3–5 days at 37 °C with 5% (v/v) CO2, and checked daily for growth. The bacteria recovered from the spleens were enumerated to evaluate the survival of each strain in mice (Kahl-McDonagh & Ficht, 2006). The results are presented as the mean number of CFU per spleen ± the standard deviation (SD) in each group. If no bacteria grew in the undiluted homogenized sample, the spleen was assumed to contain less than five bacteria, that is, falling below the limit of detection of 5 CFU/spleen.

Statistical analysis

Student's t-test was performed to analyze the data from the mouse virulence experiments, and anova was performed for growth curve analysis, cellular infections, and environmental stress response assays. A P < 0.05 was considered to represent a significant difference.

Results

Construction and growth characteristics of the Brucella cspA deletion mutant

The B. melitensis NI cspA gene deletion mutant was constructed via a double recombination event and confirmed through PCR using the primers listed in Table S2 and sequencing analysis (data not shown). The genetic complementation strain was constructed by electroporating pBBRcspA into the mutant NIΔcspA and confirmed via PCR (data not shown). Because most CSPs contribute to the growth phenotype of bacteria at cold temperatures (Schmid et al., 2009), we initially examined the growth characteristics of NIΔcspA, NIΔcspA pBBRcspA, and B. melitensis NI in TSB medium at 37 °C and 15 °C. No differences were observed in the growth of NIΔcspA, NIΔcspA pBBRcspA, and B. melitensis NI at 37 °C or 15 °C (data not shown). However, the growth rates of all the Brucella strains at 15 °C were slower than at 37 °C. These results suggested that the CspA protein did not affect the growth of Brucella at normal and low temperatures.

Cold-, acid-, and oxidative stress-associated induction of cspA gene expression in Brucella

The effect of different conditions on the expression pattern of the cspA gene in B. melitensis was examined. To determine the expression patterns of cspA, mid-log- and stationary-phase B. melitensis NI cells were adapted to cold, acid, and oxidative conditions. As a control, an equal aliquot of the same B. melitensis culture was submitted to optimal temperature conditions. The total RNA isolated from these samples was used to determine the level of cspA gene expression through quantitative RT-PCR assays. A summary of the relative cspA gene expression levels observed in different culture phases and conditions is presented in Fig. 1. As shown in Fig. 1, there was low constitutive expression of cspA transcripts in B. melitensis NI cells maintained at 37 °C. However, cspA gene transcription was significantly induced in cold-, acid-, and oxidative stress-submitted B. melitensis NI cells (< 0.05). The level of the cspA gene expression was highest under acidic conditions and was also dependent on the physiological phase, as cspA levels were increased about 1.5-fold at mid-log-phase compared to stationary phase.

Figure 1.

Real-time PCR analysis of cspA gene expression levels in Brucella melitensis NI under different conditions. The translation initiation factor IF-1 gene was used as a reference gene to normalize the expression levels of the target gene under different conditions. Fold change is expressed as a ratio of normalized gene expression levels under different conditions (to those at normal culture condition (37 °C)). The presented values represent the means of three experiments performed in duplicate, and the error bars indicate the SD. Significant differences between the strains are indicated by an asterisk (P < 0.05).

The NIΔcspA mutant is highly sensitive to environmental stresses

To determine the effect of the CspA protein on environmental stress tolerance, brucellae survival was evaluated under reduced pH and exogenous hydrogen peroxide (H2O2) conditions. For the acid sensitivity assays, after a 2 h exposure of mid-log-phase cultures to a pH of 3.4 at 37 °C, NIΔcspA and B. melitensis NI showed decreases in bacterial viability of 66% and 23.5%, respectively, whereas in stationary cultures, decreases of 43.5% and 39% were observed, respectively, which indicated that NIΔcspA is more sensitive to acidic conditions during mid-log-phase (Fig. 2, P < 0.05) than during stationary phase compared to the parent strain (Fig. 2, P > 0.05). However, at 15 °C, NIΔcspA was more susceptible to acidic conditions compared to the B. melitensis NI parent strain, independent of the growth phase (Fig. 2). Introduction of pBBRcspA into the mutant NIΔcspA restored the resistance of NIΔcspA to acidic conditions to parental strain levels during both periods, suggesting that the CspA protein contributes to resistance to acidic conditions, especially during mid-log-phase and in cold conditions (15 °C).

Figure 2.

Survival rates of NIΔcspA, NIΔcspA PBBRcspA, and Brucella melitensis NI under acidic conditions. After a 2-h exposure to pH 3.4 minimal media, the survival rates of each strain were calculated. The presented values represent the means of three experiments performed in duplicate, and the error bars indicate the SD. Significant differences between the strains are indicated by an asterisk (P < 0.05).

The survival rates of NIΔcspA were also determined in the oxidation resistance assays. After a 1-h exposure to H2O2 at concentrations of 1, 2.5, and 5 mM at 37 °C, the survival rates of NIΔcspA were 3.34%, 0%, and 0%, respectively, during mid-log-phase and 44.4%, 5.8%, and 0%, respectively, during stationary phase, which were significantly lower than the rates of the parent strain under the same conditions (Fig. 3, P < 0.05). However, both NIΔcspA and the parent strain were more sensitive to oxidative killing during mid-log-phase than during stationary phase (Fig. 3, P < 0.05). There was no significant difference in the survival rates of NIΔcspA under oxidative conditions at 15 °C compared to the survival rates at 37 °C. The parent strain showed increased survival rates, especially at concentrations of 2.5 and 5 mM of H2O2, compared to the survival rates at 37 °C. In addition, we also determined the inhibition zone diameters for NIΔcspA in disk diffusion assays, and we observed significantly different inhibition zone diameters between the mutant and the parent strain (Table 1). Introduction of pBBRcspA restored the resistance of NIΔcspA to H2O2 conditions to parental strain levels during each period, thus establishing a positive correlation between the presence of the cspA gene and H2O2 susceptibility. These results indicated that the CspA protein also contributes to oxidative stress resistance in Brucella.

Table 1. Sensitivity of Brucella strains to oxidative conditions
StrainSensitivity to oxidative conditions (mm)a
Mid-log-phaseStationary phase
37 °C15 °C37 °C15 °C
  1. a

    Sensitivity to oxidative stress was studied in disk diffusion assays using 5 μL of 10% peroxide hydrogen (H2O2) for Brucella strains, as described in the Materials and methods. The presented values represent the inhibition zone diameters (mm).

  2. b

    P < 0.05 (significant) compared to the parent strain B. melitensis NI under the same conditions.

  3. c

    P < 0.05 (significant) compared to stationary-phase cultures at the same temperature.

B. melitensis NI10.2 ± 0.359.1 ± 0.286.3 ± 0.075.1 ± 0.21
NIΔcspA23.4 ± 0.28b,c25.7 ± 0.42b,c13.8 ± 0.49b15.2 ± 0.28b
NIΔcspA pBBRcspA10.5 ± 0.229.3 ± 0.186.0 ± 0.155.2 ± 0.34
Figure 3.

Survival rates of NIΔcspA, NIΔcspA PBBRcspA, and Brucella melitensis NI under oxidative conditions. After a 1-h exposure to H2O2, the survival rates of each strain were calculated. The presented values represent the means of three experiments performed in duplicate, and the error bars indicate the SD. Significant differences between the strains are indicated by an asterisk (P < 0.05).

CspA is required for the replication of B. melitensis NI in J774.A1 macrophages

The in vitro sensitivity of NIΔcspA to environmental stresses that are predicted to be encountered in host macrophages prompted us to investigate the effect of the CspA protein on the virulence of B. melitensis NI. First, we assessed the intracellular survival of the NIΔcspA mutant. CFUs were determined at six time points: 1, 4, 12, 24, 36, and 60 h postinfection. As shown in Fig. 4, at 1 h postinfection, the macrophages infected with NIΔcspA showed similar bacterial loads compared to the macrophages infected with B. melitensis NI and NIΔcspA PBBRcspA, which indicated that there was no significant variation in the ability of the bacteria to invade macrophages. At 4 h postinfection, all of the intracellular bacterial loads of the three Brucella strains were decreased. However, after 4 h postinfection, the CFUs of B. melitensis NI and NIΔcspA PBBRcspA increased rapidly, while numbers of intracellular bacterial CFUs in NIΔcspA-infected cells decreased continuously. At the end of the test, the number of recovered NIΔcspA strain bacteria was four orders of magnitude lower than those of B. melitensis NI and NIΔcspA PBBRcspA (the bacterial loads of NIΔcspA, B. melitensis NI, and NIΔcspA PBBRcspA were 5 × 102 CFU, 5 × 106 CFU, and 5 × 106 CFU, respectively). These data showed that mutant NIΔcspA has difficulty replicating inside J774.A1 macrophages, and the mutant exhibits reduced virulence in vitro. Thus, it was clear that the CspA protein of B. melitensis NI is required for survival inside macrophages.

Figure 4.

Multiplication of NIΔcspA (●), NIΔcspA PBBRcspA (■), and Brucella melitensis NI (▲) in J774.A1 macrophages. At the indicated hour p.i., the number of intracellular bacteria was measured and expressed as log10 CFU/well. The presented values represent the means of three experiments performed in duplicate, and the error bars indicate the SD.

CspA contributes to the virulence of B. melitensis NI in mice

The in vivo survival and bacterial loads of NIΔcspA in mice were compared with B. melitensis NI. As shown in Fig. 5, we observed a large reduction (almost a 2-log difference) in the splenic bacterial load at 1, 3, and 6 weeks postinoculation in mice infected by NIΔcspA compared to mice inoculated with B. melitensis NI, and NIΔcspA showed accelerated clearance from spleens. At 9 weeks postinfection, complete clearance was observed in mice infected with NIΔcspA, whereas more than 103 CFUs of Brucella remained in the spleens of mice infected with B. melitensis NI. These results indicated that the NIΔcspA mutant was attenuated in mice, which is consistent with the results of the macrophage infection assay.

Figure 5.

Kinetics of NIΔcspA (●) and Brucella melitensis NI (▼) replication in mice. Twenty-five mice were inoculated with each strain at a dose of 106 CFU/mouse. Five mice/group were euthanized at 1, 3, 6, 9, and 11 weeks postinoculation, and the virulence of each strain was determined based on the number of CFUs recovered from the spleen, which was expressed as the mean ± SD (n = 5) of individual log10 CFU/spleen values.

Discussion

CspA, which is the major CSP in E. coli, is dramatically induced upon temperature downshift. CspA has previously been shown to act as an RNA chaperone (Jiang et al., 1997) and is linked to the regulation of both normal growth and stress adaptation processes as well as to promotion of the survival of stationary-phase cells in some organisms (Graumann et al., 1996, 1997; Yamanaka et al., 1998; Weber & Marahiel, 2002; Schmid et al., 2009). To explore the effect of CspA on the virulence and stress adaptation of Brucella, in this study, we constructed a cspA gene deletion mutant, NIΔcspA, through a novel positive/negative selection method (Wang et al., 2013). As described above, NIΔcspA failed to replicate in cultured J774.A1 macrophages and was rapidly cleared from infected BALB/c mice, which demonstrated that CspA was required for the virulence of B. melitensis NI both in vivo and in vitro. However, in the tests of growth characteristics, we observed that NIΔcspA did not show any discernible growth defect in vitro at normal culture temperatures or at cold temperatures, although cspA gene expression was cold inducible. These results indicated that the function of CspA in Brucella may be associated with multiple phenotypes, rather than just growth characteristics which is similar to L. monocytogenes that some of the members of the Csp family are not required for the viability or efficient growth under optimal temperature conditions (Schmid et al., 2009).

Within host macrophages, Brucella replicate in BCVs associated with the endoplasmic reticulum. While early acidification of vacuolar compartments has been shown to be essential for the intracellular survival of Brucella (Detilleux et al., 1991; Porte et al., 1999), the pH of phagocytic vacuoles has been observed to decrease rapidly to 4.0–4.5 (Porte et al., 1999), which indicates that Brucella must adapt to the low-pH environment. Therefore, we first determined the acid resistance of NIΔcspA during different culture periods at different temperatures. The obtained results indicated that NIΔcspA was more sensitive to acidic conditions during mid-log-phase than during stationary phase and as compared to the parent strain, especially at cold (15 °C) temperatures. This phenomenon can be explained by the significantly induced expression of the cspA gene observed under acidic and cold conditions. In addition to exposure to acidic conditions, intracellular brucellae are exposed to lethal reactive oxygen intermediates (ROIs) (Baldwin & Winter, 1994). Therefore, we also determined the oxidative resistance of NIΔcspA and observed a strong increase in the sensitivity of the NIΔcspA mutant to exogenous H2O2 stress compared to the parent strain, especially at 15 °C, in both mid-log- and stationary phases. Moreover, the cspA gene was also found to be induced by oxidative stress. Although the functional contributions of CspA to stress tolerance in Brucella are not yet clear, research has shown that Brucella that are grown in macrophages undergo a significant change in protein expression (Lin & Ficht, 1995; Rafie-Kolpin et al., 1996). Because the RNA chaperone Hfq acts as a global regulator in Brucella (Caswell et al., 2012) and CspA is also an RNA chaperone (Jiang et al., 1997) that is linked to the regulation of the expression of multiple genes in E. coli (Bae et al., 1997; Phadtare & Inouye, 2004), an intriguing possibility is that CspA may participate in the regulation of some or all of these proteins involved in stress tolerance. In fact, CspA is related to the generalized stress response, and it should contribute to resistance to hostile environments both in vivo and in vitro. Therefore, we inferred that the inability of the NIΔcspA mutant to replicate within host macrophages and the attenuation of its growth in mice might be due to the sensitivity of this strain to acid and oxidative stress, which would presumably be encountered by bacteria during infection. In addition, similar to its function in E. coli, where CspA significantly regulates genes in the flagellar operon (Phadtare & Inouye, 2004), we also found that some flagellar assemble genes were differentially expressed in NIΔcspA mutant (data not shown), which contribute to the virulence of Brucella (Fretin et al., 2005).

In summary, this study is the first to describe the impact of the CspA protein on resistance to harsh environmental stress, including acidic and oxidative stresses, as well as the virulence of Brucella. However, elucidating the mechanism underlying the effect of CspA on the stress tolerance and the virulence of Brucella will require further investigation.

Acknowledgements

We thank Prof. Baoli Zhu for his critical reading of the manuscript. This work was supported by the National Basic Research Program of China (973 Program; 2010CB530202) and the National Natural Science Foundation of China (Project No. 31372446). The authors have declared that no conflict of interest exists.

Ancillary