The Brucella abortus general stress response system regulates chronic mammalian infection and is controlled by phosphorylation and proteolysis

Abstract

Background: Virulence of pathogenic bacteria is often determined by their ability to adapt to stress.

Results: The Brucella abortus general stress response (GSR) system is required for chronic mammalian infection and is regulated by phosphorylation and proteolysis.

Conclusion: The B. abortus GSR signaling pathway has multiple layers of post-translational control and is a determinant of chronic infection.

Significance: This study provides new, molecular level insight into chronic Brucella infection. Brucella spp. are adept at establishing a chronic infection in mammals. We demonstrate that core components of the α-proteobacterial general stress response (GSR) system, PhyR and σ(E1), are required for Brucella abortus stress survival in vitro and maintenance of chronic murine infection in vivo. ΔphyR and ΔrpoE1 null mutants exhibit decreased survival under acute oxidative and acid stress but are not defective in infection of primary murine macrophages or in initial colonization of BALB/c mouse spleens. However, ΔphyR and ΔrpoE1 mutants are attenuated in spleens beginning 1 month postinfection. Thus, the B. abortus GSR system is dispensable for colonization but is required to maintain chronic infection. A genome-scale analysis of the B. abortus GSR regulon identified stress response genes previously linked to virulence and genes that affect immunomodulatory components of the cell envelope. These data support a model in which the GSR system affects both stress survival and the interface between B. abortus and the host immune system. We further demonstrate that PhyR proteolysis is a unique feature of GSR control in B. abortus. Proteolysis of PhyR provides a mechanism to avoid spurious PhyR protein interactions that inappropriately activate GSR-dependent transcription. We conclude that the B. abortus GSR system regulates acute stress adaptation and long term survival within a mammalian host and that PhyR proteolysis is a novel regulatory feature in B. abortus that ensures proper control of GSR transcription.

Keywords: Bacterial Pathogenesis; Bacterial Signal Transduction; Bacterial Transcription; Brucella; General Stress Response; Infectious Diseases; PhyR; Stress Response.

FIGURE 1.

B. abortus phyR and rpoE1 are required for in vitro stress survival and chronic infection of a mammalian host. A, canonical model of PhyR-dependent regulation of transcription during stress (based on Ref. 10). The hybrid response regulator, PhyR, is composed of an N-terminal σ-like domain (orange) and C-terminal receiver (Rec) domain (blue). Stress-dependent phosphorylation of PhyR increases its affinity for the anti-σ factor, NepR (green). PhyR∼P·NepR binding releases an EcfG family σ factor (σ^E1; orange) to bind RNA polymerase and direct transcription. B, survival of wild type B. abortus 2308, ΔrpoE1, ΔnepR-rpoE1, ΔphyR, and phyR_VAA/ADD strains upon oxidative and acid stress treatment in stationary and logarithmic growth phases. ΔphyR and ΔnepR-rpoE1 null phenotypes are complemented by single copy integration of these genes plus their native promoter into the chromosome (phyR⁺ and nepR-rpoE1⁺). Cell viability was assessed by measuring cfu after 1 h of oxidative (5 mm H₂O₂) or 2 h of acid (pH 3.9) treatment and comparing with mock-treated controls. Error bars, S.D. of at least three independent replicates. Statistical significance of the survival difference between wild-type B. abortus and ΔphyR was assessed by one-way analysis of variance followed by Dunnett's postcomparison test (**, p < 0.001 between wild type and ΔphyR). C and D, kinetics of BALB/c mouse spleen colonization by wild-type B. abortus 2308 and its derivatives ΔrpoE1 and ΔphyR. Mice were infected intraperitoneally with ∼5 × 10⁴ cfu, and spleens were collected at the indicated time points. Error bars, S.D. Statistical significance of the spleen colonization difference between wild-type B. abortus and mutants was assessed by one-way analysis of variance followed by Bonferroni's multiple comparison test (p > 0.05 at week 2 and week 4; **, p < 0.001 at week 8 and week 12). Data are representative of two independent experiments with five mice per time point.

FIGURE 2.

Biophysical analysis of purified B. abortus His₆-NepR-σ^E1 complex and His₆-PhyR. A, organization of the B. abortus gene cluster encoding PhyR, NepR, σ^E1, and two predicted histidine kinases. B, SEC elution profile of NepR-σ^E1 complex; C, c(S) distribution calculated from sedimentation velocity measurements of the NepR-σ^E1 complex. D, SEC elution profile of PhyR; E, c(S) distribution calculated from sedimentation velocity measurements of PhyR. The estimated molecular mass based on SEC elution standards or c(M) calculations in Sedfit are indicated above the peaks; root mean square deviation of the fit sedimentation data is marked in C and E. Stars denote the SEC fractions analyzed by 14% SDS-PAGE.

FIGURE 3.

Atypical interaction between B. abortus NepR and unphosphorylated PhyR. PhyR elutes as a single peak by SEC (pink line; 10.9 ml); NepR elutes as multiple species (dashed green line; 10.9 and 11.8 ml); an equimolar mixture of individually purified NepR and PhyR (10 μm each) elutes as a single peak with a volume shift that is consistent with complex formation (10.7 ml). Stars indicate the fractions analyzed by 14% SDS-PAGE (inset).

FIGURE 4.

B. abortus NepR forms a high affinity complex with unphosphorylated PhyR; phosphorylation of PhyR yields an extremely high affinity, long-lived PhyR∼P·NepR complex. Sensorgrams and fit residuals from SPR experiments measuring association and dissociation of surface-immobilized NepR to analyte proteins PhyR, PhyR∼P, and σ^E1. Immobilized ligand (R_ligand) was ∼400 response units for PhyR and PhyR∼P binding experiments; R_ligand was ∼100 for the σ^E1 binding experiment. Analytes were flowed at marked concentrations. Black lines indicate global fit of association and dissociation curves using a 1:1 binding model (R_max local) with drifting base line. The table presents equilibrium affinities (K_D) calculated from association (K_a) and dissociation (K_d) rate constants.

FIGURE 5.

Regulated proteolysis of PhyR protein and mutant variants. A, immunoblot analysis of B. abortus wild-type PhyR and the C-terminal mutant variant PhyR_VAA/ADD in stationary growth phase and logarithmic phase. PhyR concentration [PhyR] relative to WT log phase cells is normalized to the loading control band. B, blot measuring B. abortus PhyR in oxidative stress (5 mm H₂O₂) and acid stress (pH 3.9) versus a mock-treated control; a ΔphyR control is included. [PhyR] relative to mock-treated cells is normalized to the loading control band. C, immunoblot of C. crescentus wild-type PhyR in the presence of an activating peroxide stress (0.2 mm H₂O₂) versus a mock-treated control. D (top), to probe PhyR stability in vivo, B. abortus cells from early stationary phase were treated 10 min poststress with chloramphenicol (Cm) to halt protein synthesis. PhyR decay was assessed by immunoblot using protein samples collected across a 120-min window (after chloramphenicol treatment). PhyR and PhyR_VAA/ADD protein decay as a function of 5 mm H₂O₂ treatment was compared with a mock-treated control. D (bottom), as above, C. crescentus PhyR stability was assessed across a 120-min window (after chloramphenicol treatment). Protein decay was measured as a function of peroxide stress treatment. E, B. abortus PhyR stability was assessed by immunoblot as a function of acid treatment (pH 3.9) across a 60-min window (after chloramphenicol treatment). F, B. abortus ClpX protein as a function of 5 mm H₂O₂ treatment and acid pH treatment (pH 3.9) was assessed by immunoblot. In all immunoblots, a nonspecific loading control band is shown below the PhyR band.

FIGURE 6.

Assessing the functional role of PhyR degradation and phosphorylation. A, relative rpoH1 and dps transcript levels in phyR_VAA/ADD versus WT strains during unstressed logarithmic growth. Transcript levels were measured by quantitative RT-PCR and normalized to rplK mRNA. B, cell viability of B. abortus 2308 and mutant derivatives upon treatment with 5 mm H₂O₂. Assessed strains include the ΔphyR in-frame deletion mutant and strains in which the wild-type phyR allele was replaced with aspartyl phosphorylation site point mutants, D191A and D191E. Survival was measured relative to a mock-treated control. Error bars, S.D.

FIGURE 7.

Proposed molecular model of the PhyR-NepR-σ^E1 GSR regulatory system of B. abortus. PhyR protein concentration is maintained at lower levels in the absence of stress by an ATP-dependent protease, ClpXP; PhyR proteolysis helps to ensure that NepR (orange) remains bound to σ^E1 (blue). Under stress and stationary growth conditions, the PhyR is phosphorylated and stabilized and binds NepR. This releases σ^E1 to activate transcription of genes required for stress survival and chronic animal infection.