Brucella MucR acts as an H-NS-like protein to silence virulence genes and structure the nucleoid

Abstract

Histone-like nucleoid structuring (H-NS) and H-NS-like proteins coordinate host-associated behaviors in many pathogenic bacteria, often through forming silencer/counter-silencer pairs with signal-responsive transcriptional activators to tightly control gene expression. Brucella and related bacteria do not encode H-NS or homologs of known H-NS-like proteins, and it is unclear if they have other proteins that perform analogous functions during pathogenesis. In this work, we provide compelling evidence for the role of MucR as a novel H-NS-like protein in Brucella. We show that MucR possesses many of the known functions attributed to H-NS and H-NS-like proteins, including the formation of silencer/counter-silencer pairs to control virulence gene expression and global structuring of the nucleoid. These results uncover a new role for MucR as a nucleoid structuring protein and support the importance of temporal control of gene expression in Brucella and related bacteria.

Keywords: Brucella; ChIP-seq; H-NS; H-NS-like; Hi-C; MucR; counter-silencer; nucleoid structuring; virulence.

Fig 1

MucR binding sites in the B. abortus 2308 btaE (A), bpdB (B), and babR (C) promoters identified by deletion mapping (Fig. S1). The cross-hatched bar on the schematic shows the DNA region evaluated in EMSAs. Nucleotides shown in bold are the MucR binding regions and those shown in red are AT-rich regions with multiple TA “steps” (shown in boxes). The –35, –10, and +1 start sites for btaE, bpdB, and babR are shown in green, and the start codons are shown in blue. The transcriptional start sites for btaE (49) and babR (50) were determined by primer extension, and the transcriptional start site for bpdB was determined by differential RNA-seq (Caswell and Roop, unpublished data). Binding sites identified for MdrA in the btaE promoter region are underlined (49).

Fig 2

(A) Distribution of MucR binding sites detected by ChIP-seq analysis on Chromosomes 1 and 2 of B. abortus 2308. The x-axis shows genome position in kilobases (kb). The y-axis represents the fold enrichment (ChIP/Input). For each sample, the sequencing reads at each position were normalized to the total number of reads before plotting. The data were plotted in 1 kb bins. Origin of replication for both replicons has been centered on the graph at position 0; therefore, the plots start at 950 kb on Ch1 and 550 kb on Ch2. (B) AT-rich nature of MucR peaks determined by MEME analysis of the top 150 MucR peaks identified in the B. abortus 2308 genome by ChIP-seq. AT-rich sequence shown is present in 71/150 peaks.

Fig 3

MucR ChIP-seq peaks are associated with genes encoding known B. abortus 2308 virulence determinants including the autotransporter BmaC (A), the T4SS (B), T4SS effectors (C and D), and VjbR (D). ChIP-seq read depth is shown above annotated open reading frames (ORFs) within each region. Scale bars of the x-axis are shown below each graph.

Fig 4

(A) Competitive displacement of MucR from the B. abortus 2308 btaE promoter by the MarR-type transcriptional activator MdrA in an EMSA. (B) Deoxycholate inhibits the ability of MdrA to displace MucR from the btaE promoter. SC denotes the addition of unlabeled specific competitor DNA. A 192 bp btaE promoter fragment (F3/R3, Table S4) was used in both panels A and B.

Fig 5

MucR ChIP-seq peaks extend across Region 16 (top) and Core 8 (bottom) genomic islands described by Wattam et al. (65). ORFs falling within/without the described genomic islands are shown in black/gray, respectively. Select gene names are shown. ChIP-seq read depth is shown above annotated ORFs within each region. Scale bars of the x-axis are shown below each graph.

Fig 6

(A) Normalized Hi-C contact frequency maps of the genomes of B. abortus 2308 (left) and an isogenic mucR mutant (right). The x-and-y axes correspond to the genome position in kb. The data are plotted in 5 kb bins. Chromosome 1 (Ch1, green bar), Chromosome 2 (Ch2, blue bar), and their respective origins (ori1, ori2) are labeled. Origins have been situated to the middle of each replicon for better visualization of the interactions within these regions. The starting position for Ch1 is 950 kb and for Ch2 is 550 kb, but the Hi-C axes are shown as a contiguous number to represent the distance in kb. The scale bar for Hi-C interaction scores (contact frequency) is shown on the right. The Hi-C map can be divided into four parts: a Ch1 interaction map in the bottom left quadrant (green box), a Ch2 interaction map in the top right quadrant (blue box), and a Ch1-Ch2 interaction map with identical, mirrored copies in the top left and bottom right quadrants (red boxes). The strong primary diagonal lines on the map from bottom left to top right demonstrate the short-range interactions along the chromosomes. The weak secondary diagonal lines extending from top left to bottom right demonstrate interactions between loci on opposite arms of each chromosome. (B) Hi-C contact probability decay curves show the averaged contact frequency between every pair of loci on the chromosome separated by set distances. The x-axis indicates the genomic distance of separation in kb. The y-axis represents the averaged contact frequency in a logarithmic scale. The data are generated using 1 kb resolution. The red ovals (labeled 1–4) indicate differences between 2308 and the mucR mutant. Two biological replicates of each strain are shown. (C) Log₂ ratio plots comparing different Hi-C matrices. Log₂(matrix A/matrix B) was calculated and plotted in the heatmaps. The ratio of two biological replicates (Wt rep1/Wt rep2) is shown on the left and the ratio of ∆mucR to Wt (∆mucR/Wt) is on the right. The color scale is shown. Black arrows point to a few examples of strong signatures that overlap with MucR ChIP-seq peaks that had fold enrichment greater than 10.

Fig 7

(A) Plasmid-borne E. coli hns can complement the growth defect of B. abortus 2308 mucR mutant. B. abortus 2308 (above) and an isogenic ΔmucR mutant (below) carrying pSRKKm with mucR, hns, or no insert (empty) were grown and spotted onto SBA with (+) or without (−) IPTG (see Materials and Methods). (B) Plasmid-borne expression of E. coli hns can complement elevated btaE expression of B. abortus 2308 ΔmucR. RNA was isolated from B. abortus ΔmucR carrying either pSRKKm alone (empty, left), pSRKKm mucR (middle), or pSRKKm hns (right) grown in the presence (+) or absence (−) of IPTG and used for RT-PCR (see Materials and Methods). Relative expression is shown as a Log2 fold-change for each treatment relative to Wt B. abortus carrying empty pSRKKm. * indicates a significant difference from Wt (one sample t-test; P-value < 0.05).