The RNA chaperone Hfq independently coordinates expression of the VirB type IV secretion system and the LuxR-type regulator BabR in Brucella abortus 2308

Abstract

The type IV secretion system encoded by the virB operon is required for full virulence of Brucella sp., and the present study links the RNA chaperone Hfq to wild-type expression of virB in Brucella abortus 2308. Studies employing virB-lacZ fusions, quantitative reverse transcription-PCR, and immunoblot analysis showed that both transcription and translation of virB are decreased in an isogenic hfq mutant compared to those in the parental strain. These results led to the hypothesis that Hfq regulation of virB is mediated through an intermediate transcriptional regulator. Subsequent experiments determined that expression of the gene encoding the putative Brucella quorum-sensing regulator BabR (also known as BlxR), a known virB regulator, is also controlled by Hfq at the posttranscriptional level, and a cis-acting element in the 5' untranslated region of the babR transcript responsible for this regulation was identified. Consistent with its role as a virB regulator, recombinant Brucella BabR binds to the virB promoter region in electrophoretic mobility shift assays. However, experiments employing a babR mutant strain determined that BabR is a repressor, not an activator, of virB transcription. These findings suggest that Hfq regulates virB expression through both BabR-dependent and BabR-independent pathways.

Fig 1

Hfq is required for maximal expression of the VirB system. Plasmid constructs pC³005 (A) and pC³003 (C) (Table 2) were transformed into B. abortus 2308, the hfq isogenic mutant (hfq), and the hfq mutant strain expressing hfq from a plasmid (hfq/Com), and following cultivation in low-pH (pH 4.5) defined medium, β-galactosidase activity was assessed. The asterisk denotes a statistically significant difference between the β-galactosidase activities of parental strain 2308 and the hfq mutant strain (P < 0.01). (B) Immunoblot analysis of VirB1 protein levels. Brucella strains were cultivated in low-pH minimal E medium. GroEL levels are shown as a protein loading control, and molecular weight markers are shown to the left. (D) Real-time RT-PCR analysis of virB1 transcript levels. Brucella strains were grown in low-pH (pH 4.5) defined medium, and total RNA was isolated. Oligonucleotide primers specific for virB1 or 16S rRNA were used to amplify the target genes by PCR, and quantification of the amplified DNA fragments was performed using SYBR green incorporation. The values represent the relative abundances of specific mRNAs, with the level of mRNA from parental strain 2308 designated 1.000.

Fig 2

Identification of an Hfq-regulated transcription regulator, BabR. Plasmid constructs pC³008 (A) and pC³006 (C) (Table 2) were transformed into B. abortus 2308, the hfq isogenic mutant (hfq), and the hfq mutant strain expressing hfq from a plasmid (hfq/Com), and following cultivation in brucella broth, β-galactosidase activity was assessed. The asterisk denotes statistically significant differences between the β-galactosidase activities of parental strain 2308 and the hfq mutant strain (P < 0.01). (B) Pictures of Brucella stains carrying either pC³008 (left) or pC³006 (right) grown on tryptic soy agar containing X-Gal. (D) Real-time RT-PCR analysis of babR transcript levels. Brucella strains were grown in brucella broth, and total RNA was isolated. Oligonucleotide primers specific for babR or 16S rRNA were used to amplify the target genes by PCR, and quantification of the amplified DNA fragments was performed using SYBR green incorporation. The values represent the relative abundances of specific mRNAs, with the level of mRNA in parental strain 2308 designated 1.000.

Fig 3

Mapping and characterization of the babR promoter region. (A) Schematic of the chromosomal region containing babR. The plus strand sequence of DNA corresponding to the region between the hyp gene and babR is shown below the schematic, and the identified transcriptional start site (+1) is labeled. This transcriptional start site is 119 nucleotides upstream of the babR start codon. The first 20 nucleotides of the 5′ UTR are boxed to indicate that this region was removed for the experiments shown in Fig. 5. The −10 and −35 elements of the babR promoter are also labeled. The genes in this region are metH (B12-dependent methionine synthase; BAB1_0188), hyp (hypothetical protein; BAB1_0189), babR (LuxR-type transcriptional regulator; BAB1_0190), and atf (aminotransferase; BAB1_0191). (B) Identification of the transcriptional start site of the babR mRNA. Primer extension (left panel) was performed using total RNA isolated from Brucella strains (wild-type 2308 and the hfq mutant) grown in brucella broth. The primer extension reaction products, along with DNA sequencing reaction products produced with the same radiolabeled primer, were separated on 6% denaturing polyacrylamide gels and visualized by autoradiography. 5′ RACE was also employed (right panel) using RNA isolated from B. abortus 2308 grown in brucella broth. Molecular weight standards and the 5′ RACE product were separated by 2.0% agarose gel electrophoresis. The indicated DNA band was extracted from the gel, cloned, and sequenced to determine the transcriptional start site. (C) Secondary structure predictions of the RNA containing the babR 5′ UTR and the first five codons of the babR coding region. The RBS of each predicted structure is underlined, while a box is placed around the start codon (AUG) of each predicted structure.

Fig 4

Truncation of the babR 5′ UTR alleviates the requirement for Hfq. (A) Schematic of the babR-lacZ translational fusions used in this experiment. The full-length fusion construct pC³008 (also used in Fig. 2) contains the babR promoter region and the entire babR 5′ UTR cloned in frame with lacZ. The truncated fusion construct pC³014 (Table 2) contains the babR promoter region and a truncated babR 5′ UTR cloned in frame with lacZ. (B) β-Galactosidase activities of strains carrying the full-length and truncated babR 5′ UTR-lacZ translational fusions. B. abortus 2308 and the isogenic hfq mutant harboring pC³008 or pC³014 were growth in brucella broth, and β-galactosidase activity was measured. The asterisk indicates a statistically significant difference between the β-galactosidase activities of parental strain 2308 and the hfq mutant strain (P < 0.01). To the right of the graph is an image of Brucella stains carrying pC³008 or pC³014 grown on tryptic soy agar containing X-Gal. ND, no statistically significant difference.

Fig 5

Hfq binds the 5′ UTR of the babR transcript. The full-length 5′ UTR of the babR mRNA and a mutated version of the babR 5′ UTR (−20 nucleotides from the 5′ end) were transcribed in vitro and radiolabeled by [α-³²P]UTP incorporation. An EMSA was performed using the labeled in vitro-transcribed 5′ UTRs and purified recombinant B. abortus Hfq. Increasing concentrations of rHfq were incubated with the labeled RNA probe, and the binding reaction products were resolved in 5% native polyacrylamide gels and visualized by autoradiography.

Fig 6

BabR is a repressor of virB expression. (A) The virB-lacZ transcriptional fusion construct pC³003 (Table 2) was transformed into B. abortus 2308, the babR isogenic mutant (babR), and the babR mutant strain expressing babR from a plasmid (babR/Com), and following cultivation in low-pH (pH 4.5) defined medium, β-galactosidase activity was assessed. The asterisk denotes a statistically significant difference between the β-galactosidase activities of parental strain 2308 and the babR mutant strain (P < 0.01). (B) Real-time RT-PCR analysis of virB1 transcript levels. B. abortus 2308 strains were grown in low-pH (pH 4.5) defined medium, and total RNA was isolated. Oligonucleotide primers specific for virB1 or 16S rRNA were used to amplify the target genes by PCR, and quantification of the amplified DNA fragments was performed using SYBR green incorporation. The values represent the relative abundances of specific mRNAs, with the level of mRNA in parental strain 2308 designated 1.000.

Fig 7

BabR binds the virB1 promoter. An EMSA was employed to test the binding of purified rBabR to the virB promoter region. A DNA probe containing the first 150 bp of the virB upstream region was amplified by PCR, and the fragment was radiolabeled with [γ-³²P]ATP. Increasing concentrations of rBabR were incubated with the labeled DNA, and in some binding reaction mixtures, unlabeled specific (P_virB DNA) and nonspecific (BAB2_0612 coding sequence DNA) competitor DNA fragments were included as controls. The binding reaction products were resolved in 6% native polyacrylamide gels and visualized by autoradiography.

Fig 8

Model of Hfq-mediated regulation of babR and virB1. There is a clear relationship between Hfq and wild-type levels of VirB1 in B. abortus 2308. It was shown in this study that Hfq positively regulates the production of BabR, a known virB1 transcriptional regulator; however, BabR is a repressor of virB1 transcription. Therefore, BabR is not the sought-after genetic link between Hfq and virB1 expression. Two alternative means of virB1 regulation by Hfq may be taking place. Hfq may be directly altering the levels of VirB1 protein by binding to the virB1 mRNA and altering translation (Direct). Alternatively, Hfq may be affecting the levels of a transcriptional regulator protein (TscR), and altered levels of this regulator would lead to increased levels of virB1 mRNA, resulting in increased levels of VirB1 protein (Indirect).