The DUF1013 protein TrcR tracks with RNA polymerase to control the bacterial cell cycle and protect against antibiotics

Abstract

How DNA-dependent RNA polymerase (RNAP) acts on bacterial cell cycle progression during transcription elongation is poorly investigated. A forward genetic selection for Caulobacter crescentus cell cycle mutants unearthed the uncharacterized DUF1013 protein (TrcR, transcriptional cell cycle regulator). TrcR promotes the accumulation of the essential cell cycle transcriptional activator CtrA in late S-phase but also affects transcription at a global level to protect cells from the quinolone antibiotic nalidixic acid that induces a multidrug efflux pump and from the RNAP inhibitor rifampicin that blocks transcription elongation. We show that TrcR associates with promoters and coding sequences in vivo in a rifampicin-dependent manner and that it interacts physically and genetically with RNAP. We show that TrcR function and its RNAP-dependent chromatin recruitment are conserved in symbiotic Sinorhizobium sp. and pathogenic Brucella spp Thus, TrcR represents a hitherto unknown antibiotic target and the founding member of the DUF1013 family, an uncharacterized class of transcriptional regulators that track with RNAP during the elongation phase to promote transcription during the cell cycle.

Keywords: Brucella; Caulobacter; RNA polymerase; TrcR; transcription.

Fig. 1.

Identification and phenotypic characterization of trcR mutant cells. (A) Schematic of C. crescentus division and morphogenesis along cell cycle and immunoblots showing steady-state levels of CtrA and TrcR during cell cycle progression of a synchronized WT population (time in minutes refers to the release of purified swarmer [G1] cells into M2G). During the Caulobacter cell cycle, capsulation is negatively regulated through the expression of hvyA that prevents capsulation in the Sw cell and is under the control of the transcriptional regulator CtrA. TrcR abundance peaks at the G1- to S-phase transition just prior the accumulation of CtrA. (B) DNA content (FL1-A channel) quantification, determined by FACS analysis, was performed during exponential growth phase in PYE. ΔtrcR and trcR::Tn populations show a strong decrease in G1 cell number (pink arrow) and cells that accumulate more than two chromosomes compared to WT cells (orange arrow). (C) The immunoblots on the left show the loss of the TrcR protein in ΔtrcR and trcR::Tn (NS245) cells. The immunoblots on the right show the reduction of CtrA and PilA in trcR mutant cells. MreB actin and CC_0164 serve a loading control for immunoblots. (D) Domain organization predicted for TrcR: DUF1013 is indicated in light blue, the putative helix-turn-helix domain within residues 107 through 140 is indicated in dark blue, and the predictive tertiary structure of this region based on structure homology modeling is shown on the right. Identified Tn-insertions in strains NS133 (NS245) and NS264 are indicated in red. (E) Motility (0.3% agar) plates inoculated with WT, ΔtrcR, and trcR::Tn (NS133, NS264, and NS245) (Left) and derivatives harboring empty vector (pMT335) or pMT335-trcR (Right). (F) Phase contrast light microscopy images of WT, ΔtrcR, trcR::Tn cells, and suppressor mutants during exponential growth in PYE. (Scale bar, 2 μm.)

Fig. 2.

TrcR regulates transcriptome and antibiotic resistance. (A) Volcano plot showing the global changes in transcript levels of ΔtrcR cells versus WT cells grown in PYE in exponential phase. The abundance of 553 transcripts is affected more than twofold in the ΔtrcR cells compared to WT. All transcripts retained during the analysis are plotted. Each circle represents one transcript. The log2 fold change in ΔtrcR versus WT is represented on the x-axis. The y-axis shows the −log10 of the false discovery rate (FDR) value. Red dots show the genes that have an FDR value of 0.05, and gray lines indicate twofold changes in gene expression. (B) Table showing specific CtrA-dependent transcripts down-regulated in the ΔtrcR cells as determined by RNA-Seq (Supplementary Dataset 1). (C) EOP assay of WT and trcR mutant cells on plates with and without Nal (20 μg/mL). The plates below show antibiotic sensitivity tests by antibiotic disk diffusion assays with discs containing the quinolones Nal (NAL), flumequine (UBN), norfloxacin (NOR), and ciprofloxacin (CIP). The scheme on the Left shows the arrangement of the discs. (D) EOP assays with WT and ΔtrcR cells harboring pMT335–empty or pMT335–trcR on PYE plates supplemented with gentamycin and vanillic acid with or without Nal (20 μg/mL). (E) EOP assays of two ΔtrcR Nal-resistant suppressor mutants harboring mutations in rpoB (rpoB^P642L and rpoB^P575S), encoding the β-subunit of the RNAP, compared to WT and ΔtrcR cells on PYE plates and PYE with Nal (20 μg/mL). (F) Motility assays on 0.3% agar with WT, ΔtrcR, ΔtrcR rpoB^P642L, and ΔtrcR rpoB^P575S cells. (G) Antibiotic susceptibility test by disks diffusion assays with discs containing rifampicin (30 μg) and novobiocin (15 μg) in WT and mutant cells embedded in 6 mL soft (0.3%) PYE agar overlaid on a PYE plate. The scheme on the left shows the arrangement of the discs on the plate.

Fig. 3.

TrcR tracks and associates with RNAP in vivo. (A) RpoC-TAP pull-down samples were assessed for the presence of TrcR, CtrA, and GcrA by immunoblotting. FljK and MreB are used as negative control, as they are not known to associate with the transcriptional machinery. Specificity was assessed using extracts from untagged WT cells subjected to TAP-based pull-down. Asterisks indicate the expected migration position for the marker proteins detected by the antibodies. (B) GFP-TrcR pull-down sample probed for the presence of the RNAP by immunoblotting. (C) Genome-wide occupancies of TrcR (Upper) and RNAP (Bottom) in different conditions (PYE and PYE Rif₃₀) in WT and ΔtrcR cells as determined by ChIP-Seq. The x-axis represents the nucleotide position on the genome (Mbp), whereas the y-axis shows the normalized ChIP-Seq read abundance in reads per million (rpm). Control ChIP-seq in the ΔtrcR cells (orange graph on the right) shows only one unspecific peak. The inset shows the TrcR steady-state levels in the presence or absence of Rif as determined by immunoblotting. (D) Venn diagram comparing the 524 TrcR significant (narrow) promoter peaks to the significant 922 RNAP promoter peaks as determined by ChIP-Seq analysis. (E) Heatmap of TrcR and RNAP ChIP-Seq reads aligned on the translation initiation codon (ATG) of Coding DNA Sequence in the center. The alignment covers an 800-bp range around the translation initiation codon in WT and ΔtrcR cells treated or not with Rif and sorted with reference to the relative abundance of reads compared to WT (highest read abundance at the top). ChIP-Seq signals are normalized in read per million (rpm) and the color bar (intensity) indicates the enrichment in rpm. Gray arrows indicate the presence or absence of binding downstream of the ATG. (F) Comparison of TrcR and RNAP ChIP-Seq traces, including the narrow promoter peak at the ctrA locus in WT cells treated or not with Rif. Coding sequences are represented as boxes on the upper part of the graph. ChIP-Seq traces of ΔtrcR cells are shown as a control.

Fig. 4.

Recruitment of the TrcR·RNAP complex to the promoter of the acrAB–nodT operon induced by Nal. (A) Genome-wide ChIP-Seq profile of TrcR chromosomal occupancy in WT cells before (Left) and after a 10-min treatment with Nal (20 μg/mL, Right). The acrAB–nodT locus is indicated as acrA2 (corresponding to its annotation in the genome) to avoid confusion with another annotated acrA gene. (B) Venn diagram showing comparison of genes containing significant promoter (narrow) peaks of TrcR binding as determined by ChIP-Seq analysis from WT before and after exposure to Nal. (C) Ratio of normalized TrcR ChIP-Seq profiles between untreated and Nal-treated conditions. (D) Comparison of TrcR, RNAP, and RpoD ChIP-Seq traces at the acrAB–nodT locus between WT cells treated or not with antibiotics (Nal, Rif) or in ΔtrcR cells. Total input (Ti) chromatin DNA (prior pull-down) was sequenced as a control for sequence bias, and the corresponding traces are shown as control. Genes encoded are represented as boxes on the upper part of the graph (annotations or CCNA gene annotation numbers of the sequenced NA1000 reference genome).

Fig. 5.

Conservation of TrcR activity in alphaproteobacteria and its essentiality for Brucella viability. (A) Motility assays on soft (0.3%) agar inoculated with WT and ΔtrcR containing the empty vector (pMT335-) or plasmids with trcR orthologs from two different rickettsiales (Upper) and S. meliloti (Bottom). Complementation of the ΔtrcR mutant with trcR orthologs under control of the P_van promoter on pMT335 improves the motility defect of the ΔtrcR mutant cells. Numbers shown correspond to the legend described in B. (B) EOP assay on PYE plates and PYE Nal (20 μg/mL) supplemented with gentamycin (1 μg/mL) and vanillate 50 μM in WT and ΔtrcR cells containing the empty vector (pMT335-) or plasmids with trcR orthologs from two different rickettsiales (Upper) and S. meliloti (Bottom). (C) Quantification of Tn miniHimar reads from the B. ovis trcR locus (locus ID BOV_RS08235) and from the B. abortus trcR locus (locus ID BAB_RS24265). DNA for sequencing was prepared from pooled B. ovis ATCC 25840 or B. abortus ATCC 2308 miniHimar mutant libraries. All possible TA dinucleotide miniHimar insertion sites at the locus are marked in vertical orange lines. Gene orientation is shown. (D) Phase contrast micrographs (63×) of WT B. ovis carrying an IPTG-inducible empty plasmid vector (WT/pSRK-EV; Left) and B. ovis ∆trcR carrying plasmid pSRK-trcR (∆trcR/pSRK-trcR; Right) were captured from cells harvested from agar medium lacking IPTG. Single cell contours were extracted from micrographs of WT/pSRK-EV (n = 721) and ∆trcR/pSRK-trcR (n = 931). (Scale bar, 5 μm.) (E, Left) Smoothed histogram showing marginal distribution of cell area (in μm²) for all WT/pSRK-EV (gray) and ∆trcR/pSRK-trcR (blue) single cell contours. (Right) Two principal shape modes account for over 90% of variance in each strain population. Percentage of variance accounted by each mode is shown. For each mode, mean shape ± 2 SD is shown. (Scale bar, 0.5 μm.) Cell contour extraction, cell area, and shape mode calculations were conducted using Celltool (61). (F) Antibiotic sensitivity assay for WT/pSRK-EV (1), WT/pSRK-trcR (2), and ∆trcR/pSRK-trcR (3). Nal (20 μg/mL) or Rif (0.025 μg/mL) was added to plates, and cells were plated in a log₁₀ dilution series (starting at a top cell titer of OD [optical density]₆₀₀ of 0.01). Plates images are inverted to enhance contrast and colony identification. IPTG was added at 1 mM final concentration where indicated.

Fig. 6.

Model for TrcR acting together with RNAP. TrcR forms a stable complex with RNAP and is brought to promoters and coding sequence via this interaction. TrcR association and tracking with the RNAP allow gene transcription and cell cycle progression. In absence of TrcR, transcription at selected cell cycle–regulated promoters is impaired and σ⁷⁰ occupancy is reduced. Compensatory mutations in rpoB* restore σ⁷⁰ occupancy and permit adequate transcription of cell cycle genes, thus supressing the cell cycle and other defects, including the antibiotic sensitivity.