Impact of Hfq on global gene expression and intracellular survival in Brucella melitensis

Abstract

Brucella melitensis is a facultative intracellular bacterium that replicates within macrophages. The ability of brucellae to survive and multiply in the hostile environment of host macrophages is essential to its virulence. The RNA-binding protein Hfq is a global regulator that is involved in stress resistance and pathogenicity. Here we demonstrate that Hfq is essential for stress adaptation and intracellular survival in B. melitensis. A B. melitensis hfq deletion mutant exhibits reduced survival under environmental stresses and is attenuated in cultured macrophages and mice. Microarray-based transcriptome analyses revealed that 359 genes involved in numerous cellular processes were dysregulated in the hfq mutant. From these same samples the proteins were also prepared for proteomic analysis to directly identify Hfq-regulated proteins. Fifty-five proteins with significantly affected expression were identified in the hfq mutant. Our results demonstrate that Hfq regulates many genes and/or proteins involved in metabolism, virulence, and stress responses, including those potentially involved in the adaptation of Brucella to the oxidative, acid, heat stress, and antibacterial peptides encountered within the host. The dysregulation of such genes and/or proteins could contribute to the attenuated hfq mutant phenotype. These findings highlight the involvement of Hfq as a key regulator of Brucella gene expression and facilitate our understanding of the role of Hfq in environmental stress adaptation and intracellular survival of B. melitensis.

Figure 1. Details of the B. melitensis hfq mutants and their growth characteristics.

A. Structure of the hfq locus on B. melitensis 16 M chromosome I. In the Δhfq, the coding region of hfq was deleted and replaced by a kanamycin resistance cassette. The coding region of hfq together with its native promoter cloned to pBBR1MCS5 yielding the complementation plasmid pBBR1-hfq. The percentage similarities were obtained using pairwise BLAST analyses that compared B. melitensis Hfq protein sequence with those of other Brucella spp., alpha-proteobacteria, Escherichia coli, and Salmonella typhimurium. B. Growth characteristics of B. melitensis Δhfq strain. B. melitensis wild-type, 16 MΔhfq, and 16 MΔhfq-C strain were cultured in TSB (pH7.0) at 37°C, and the OD600 was measured at 2 h intervals. Each graph represents the mean of three independent biological replicates grown on three different days. The error bars represent the standard deviation (SD) of the optical density at each time point.

Figure 2. Survival of the B. melitensis Δhfq mutant strain in in vitro environmental stress conditions, macrophages, and mice.

A. In vitro stress resistance analysis of B. melitensis Δhfq mutant strain. 16 M, 16 MΔhfq, and 16 MΔhfq-C were grown in TSB to the early logarithmic phase and then subjected to different stress conditions as described in the text. After the treatments, the surviving bacteria were enumerated by plating serial dilutions onto TSA plates. Bars represent mean percent survival compared to untreated controls, and error bars represent standard errors of percent survival from 3 replicates. Significant differences between the mutant and parent strain are indicated as follows: *, P<0.001. B. Survival capacity of B. melitensis Δhfq mutant strain in macrophage cells. RAW264.7 cells were infected with strains 16 M, 16 MΔhfq, or 16 MΔhfq-C at a MOI of 50∶1. Three wells were evaluated at each time point for every strain tested, and the colony forming units were determined by serial dilution and plating on TSA. The data was expressed as the mean log₁₀CFU ± SD (n = 3). Significant differences between the mutant and parent strain are indicated as follows: *, P<0.001. C. Survival capacity of B. melitensis Δhfq mutant strain in BALB/c mice. Groups of five BALB/c mice were infected intraperitoneally with 1×10⁷ CFU of 16 M, 16 MΔhfq, or 16 MΔhfq-C. At 7 and 28 days post-infection, the spleens were aseptically removed and the colony forming units were determined by plating serial dilutions on TSA plates. The data was expressed as the mean log₁₀CFU ± SD (n = 5). Significant differences between the mutant and parent strain are indicated as follows: *, P<0.001.

Figure 3. Determination of the in vitro induction conditions for hfq.

16 M was firstly cultured in TSB (pH7.0) to the stationary phase and then subjected to different stresses. RNA was isolated and transcription of hfq was quantified by qRT-PCR. Significant differences between the acidified minimum medium (GEM 4.0) and other in vitro stresses are indicated as follows: *, P<0.05.

Figure 4. Differentially expressed transcripts (upper graphs) and proteins (lower graphs) in the B. melitensis Δhfq mutant strain.

Histograms show the number of differentially expressed genes and their distributions in the B. melitensis chromosome. The functional categories according to the B. melitensis 16 M genome sequence annotation and the KEGG database is shown to the right in circle charts. The number of genes belonging to each category are shown in brackets.

Figure 5. Hfq upregulated the expression of Omp25/Omp31 and the Sigma factors.

A. Fold changes in the transcript abundances of omp25, omp25b, omp25c, and omp31 genes were detected by microarray and qRT-PCR in 16 MΔhfq, relative to 16 M. B, C. Transcript abundances of rpoH1 (B) and rpoE1 (C) were detected in the 16 M, 16 MΔhfq, and 16 MΔhfq-C during early logarithmic (EL), mid-logarithmic (ML), and stationary phases (SP). Significant differences between the transcription abundances of rpoH1 and rpoE1 in the mutant and parent strain are indicated as follows: *, P<0.001.