Regulatory Cross Talk between Motility and Interbacterial Communication in Salmonella enterica Serovar Typhimurium

Abstract

FliA is a broadly conserved σ factor that directs transcription of genes involved in flagellar motility. We previously identified FliA-transcribed genes in Escherichia coli and Salmonella enterica serovar Typhimurium, and we showed that E. coli FliA transcribes many unstable, noncoding RNAs from intragenic promoters. Here, we show that FliA in S Typhimurium also directs the transcription of large numbers of unstable, noncoding RNAs from intragenic promoters, and we identify two previously unreported FliA-transcribed protein-coding genes. One of these genes, sdiA, encodes a transcription factor that responds to quorum-sensing signals produced by other bacteria. We show that FliA-dependent transcription of sdiA is required for SdiA activity, highlighting a regulatory link between flagellar motility and intercellular communication.IMPORTANCE Initiation of bacterial transcription requires association of a σ factor with the core RNA polymerase to facilitate sequence-specific recognition of promoter elements. FliA is a widely conserved σ factor that directs transcription of genes involved in flagellar motility. We previously showed that Escherichia coli FliA transcribes many unstable, noncoding RNAs from promoters within genes. Here, we demonstrate the same phenomenon in Salmonella Typhimurium. We also show that S Typhimurium FliA directs transcription of the sdiA gene, which encodes a transcription factor that responds to quorum-sensing signals produced by other bacteria. FliA-dependent transcription of sdiA is required for transcriptional control of SdiA target genes, highlighting a regulatory link between flagellar motility and intercellular communication.

Keywords: FliA; Salmonella; SdiA; flagellar gene regulation; quorum sensing; sigma 28.

FIG 1

ChIP-seq of FliA in a motility promoter-deficient strain identifies novel binding sites. (A) Examples of FliA binding sites identified by ChIP-seq in a motility promoter-deficient strain. The schematic shows the position of genes in three genomic regions. FLAG₃-FliA (JP095) ChIP-seq sequence read coverage (for one of two replicates) is shown, with positive values indicating coverage on the plus strand and negative values indicating coverage on the minus strand. RNA-seq data are shown for fliA⁺ (LT2 and TH15903) and ΔfliA (TH6827 and TH22673) motility promoter-deficient and wild-type strains. The scale on the y axis differs for the ChIP-seq and RNA-seq data and differs between each of the three panels. The scale on the y axis for RNA-seq data is the same for the four data sets shown within each panel. The values indicated to the left of each panel are the highest read depth values for ChIP-seq or RNA-seq data on that strand in the genomic range shown. (B) Extended enriched sequence motif identified in 255 of the 265 FliA-bound regions identified by ChIP-seq using MEME (31). The y axis represents the information content at each position of the motif, and the relative height of each letter indicates the relative proportion of the corresponding base at each position. Positions 1 to 19 (to the left of the dashed line) are not part of the motif identified by MEME. (C) Heatmap showing sequence read coverage in previously described ChIP-seq data for FLAG₃-FliA in S. Typhimurium 14028s (1). Sequence read coverage, summed for both strands and across two replicates, is shown for 1-kb regions surrounding each of the 255 FLAG₃-FliA ChIP-seq peak centers identified from the motility promoter-deficient strain, with regions ranked by the level of FLAG₃-FliA ChIP-seq signal (“occupancy”) in the motility promoter-deficient strain. Yellow indicates high sequence read coverage, and blue indicates zero coverage.

FIG 2

Many novel FliA binding sites represent transcriptionally active FliA promoters inside genes. (A) Pie chart showing the distribution of the 255 FliA binding sites with an associated binding site relative to annotated genes. Sense and antisense refer to the orientation of the predicted FliA promoter relative to the orientation of the overlapping gene. Upstream refers to predicted FliA promoters with a predicted promoter that is ≤300 bp upstream of the start of an annotated gene in the same orientation. (B) β-Galactosidase activity of wild-type (LT2) or ΔfliA (TH6827) cells containing transcriptional fusions of 200-bp regions upstream of 10 putative FliA promoters (plasmids pJRP54 to pJRP72). Gene names not in parentheses indicate intergenic promoters, with the gene being the first downstream gene in the same orientation as the putative promoter. Gene names in parentheses indicate intragenic promoters, with the gene listed being that which encompasses the putative promoter. Numbers in parentheses indicate the relative ChIP-seq sequence read coverage (occupancy scores; see Materials and Methods for details) for each putative promoter. Asterisks indicate statistical significance: **, P < 0.01; *, P < 0.05 (one-tailed t test).

FIG 3

FliA-dependent transcription of the STM1300 and sdiA protein-coding genes. (A) FliA binding sites upstream of STM1300, sdiA, ycgO, and ybeM. The schematic shows the position of genes in the four genomic regions. FLAG₃-FliA ChIP-seq sequence read coverage (for one of two replicates) is shown, with positive values indicating coverage on the plus strand and negative values indicating coverage on the minus strand. RNA-seq data are shown for fliA⁺ and ΔfliA motility promoter-deficient and wild-type strains. The scale on the y axis differs for the ChIP-seq and RNA-seq data and differs between each of the four panels. The scale on the y axis for RNA-seq data is the same for the four data sets shown within each panel. The values indicated to the left of each panel are the highest read depth values for ChIP-seq or RNA-seq data on that strand in the genomic range shown. (B) Luciferase activity of wild-type (LT2) or ΔfliA (TH6827) S. Typhimurium containing translational reporter fusions for STM1300, sdiA, ycgO, or ybeM (plasmids pJRP33 and pJRP34, pJRP37 to pJRP40, and pJRP43 to pJRP46). The putative FliA-dependent promoters are wild type or were mutated to prevent RNAP:FliA binding. In the case of ybeM, there are two predicted FliA-dependent promoters, and mutations were made in each individually (p1 or p2) or both simultaneously (p1 plus p2). Statistically significant differences (P < 0.01 by t test) between wild-type fliA and ΔfliA strains or between wild-type and mutant reporter fusions are indicated by a double asterisk. (C) Western blot showing SdiA-FLAG₃ in wild-type (LT2; untagged control), SdiA-FLAG₃ (JP085), or SdiA-FLAG₃ ΔfliA (JP089) S. Typhimurium. The expected position of SdiA-FLAG₃ is indicated. Asterisks indicate nonspecific bands.

FIG 4

fliA is required for cells to accumulate normal levels of active SdiA. (A) Luciferase activity of wild-type (LT2), ΔfliA (JP096), or ΔsdiA (JP087) S. Typhimurium containing a reporter fusion for rck (pBA428). Cells were grown with either DMSO or 3OC₆HSL (C6 AHL), as indicated. (B) Luciferase activity of wild-type (LT2) or ΔsdiA (JP087) S. Typhimurium containing a reporter fusion for rck (pBA428) and either empty pBAD24 or pBAD24 expressing fliA (pJRP81). Cells were grown with either DMSO or 3OC₆HSL (C6 AHL), as indicated. Statistically significant differences are indicated by a single asterisk (P < 0.05 by t test) or a double asterisk (P < 0.01 by t test).