TraDIS-Xpress: a high-resolution whole-genome assay identifies novel mechanisms of triclosan action and resistance

Abstract

Understanding the genetic basis for a phenotype is a central goal in biological research. Much has been learnt about bacterial genomes by creating large mutant libraries and looking for conditionally important genes. However, current genome-wide methods are largely unable to assay essential genes which are not amenable to disruption. To overcome this limitation, we developed a new version of "TraDIS" (transposon directed insertion-site sequencing) that we term "TraDIS-Xpress" that combines an inducible promoter into the transposon cassette. This allows controlled overexpression and repression of all genes owing to saturation of inserts adjacent to all open reading frames as well as conventional inactivation. We applied TraDIS-Xpress to identify responses to the biocide triclosan across a range of concentrations. Triclosan is endemic in modern life, but there is uncertainty about its mode of action with a concentration-dependent switch from bacteriostatic to bactericidal action unexplained. Our results show a concentration-dependent response to triclosan with different genes important in survival between static and cidal exposures. These genes include those previously reported to have a role in triclosan resistance as well as a new set of genes, including essential genes. Novel genes identified as being sensitive to triclosan exposure include those involved in barrier function, small molecule uptake, and integrity of transcription and translation. We anticipate the approach we show here, by allowing comparisons across multiple experimental conditions of TraDIS data, and including essential genes, will be a starting point for future work examining how different drug conditions impact bacterial survival mechanisms.

Figure 1.

Validation of TraDIS-Xpress. Identification of known targets using the TraDIS-Xpress approach incorporating an inducible outward-facing promoter that identifies the impact of both essential and nonessential genes on survival and growth. (A) Genetic map of the relative gene positions is shown at the bottom of the panel. Above this, each row of vertical red or blue lines (plotted with red behind) indicates the position of mapped reads, and the height of the bar represents the relative number of reads mapped. Red indicates transposon insertions wherein the transposon-encoded kanamycin-resistance gene is oriented 5′ to 3′ left to right, and blue indicates the opposite right-to-left orientation. (A,B) Data split with induced and uninduced libraries shown at a single concentration of triclosan; this makes the impact of induction obvious. (C,D) Results from different drug concentrations (all with induction). (A) Inserts predicted to result in up-regulation of fabI in the presence of triclosan. (B) Mutants up-regulating skp, lpxD, fabZ, lpxA, and lpxB. (C) Mutants inactivating acrR and up-regulating acrAB being enriched by triclosan. (D) Mutants positioned antisense to rpoN being selected in the presence of triclosan. The top row in each plot shows untreated control cultures; the three rows below are for cultures grown in the presence of 2×, 1×, and 0.5× the MIC of triclosan (with IPTG induction for the outward facing promoter).

Figure 2.

Dose-dependent activity of triclosan. (A) Viability of BW25113 from LB broth cultures supplemented with different concentrations of triclosan, immediately after inoculation (top) and after incubation for 24 h (bottom). Samples from the cultures were serially diluted 10-fold, and 5 µL of all dilutions was spotted onto LB agar and incubated. The level of dilution (from 10⁻¹ to 10⁻⁷) is shown on the right side of the photographs, and the concentrations of triclosan in the cultures are shown along the top. The results show a bactericidal effect >0.5 mg/L; triclosan at 0.25 mg/L caused a 10-fold reduction in growth over the 24-h period, but at ≤0.125 mg/L, there was no growth inhibition. (B) Heat map highlighting differences in reads mapped to genes at the different triclosan concentrations shown along the bottom of the chart. Each row represents a gene, and genes are ordered from top to bottom. The intensity scale reflecting the size of difference is shown on the right of the map. (C) Network produced by the AlbaTraDIS software illustrating the genes identified as important at different concentrations of triclosan. Nodes represent genes (blue) or triclosan concentrations (indicated by colored nodes); edges show links between conditions and genes. The similarities of responses between sub-MIC and supra-MIC exposures are clearly visible.

Figure 3.

Novel importers contributing to triclosan sensitivity. (A,B) Insert patterns indicating inactivation of trkAH is beneficial for survival in the presence of triclosan. (C) Increase in insertions within the first part of rbsB in triclosan-exposed libraries. (D–F) Growth curves for BW25113 and isogenic mutants grown in LB broth supplemented with 0.125 mg/L of triclosan. Mutants were replicated on multiple plates. Each line indicates the average from an individual plate; independent plates are represented by symbols (triangles, circles, or squares; allowing comparison between the mutant and parent from each individual plate). Red lines indicate BW25113; blue lines, mutants. The format of the genetic maps is as described in Figure 1.

Figure 4.

Impact of lon and pcnB on triclosan sensitivity. (A,B) Insert patterns indicating inactivation of lon and pcnB is beneficial for survival in the presence of triclosan. The format of these genetic maps is as described in Figure 1. (C,D) Growth curves for BW25113 and isogenic mutants grown in LB broth supplemented with 0.125 mg/L of triclosan. Mutants were replicated on multiple plates. Each line indicates the average from an individual plate; independent plates are represented by symbols (triangles, circles, or squares; allowing comparison between the mutant and parent from each individual plate). Red lines indicate BW25113; blue lines, mutants.

Figure 5.

Altering expression of targets impacts triclosan sensitivity. The coding sequences of selected genes were cloned into the pBAD30 expression vector under transcriptional control of the arabinose-inducible promoter. This was to modify their expression and test the predicted impact on triclosan sensitivity. Genes fabI, infB, marAB, and fabZ were cloned in the forward orientation, allowing their up-regulation after induction (under the green bar). In contrast, fabA, rpoN, pstA, and lacA were cloned in the reverse direction, allowing inducible production of an antisense transcript and consequent repression of expression (under the red bar). The lacA and pBAD30 vector alone constructs were included as negative controls. Duplicate recombinants (Rec1, Rec2) for each gene were tested by spotting cultures onto LB-agar (top), LB agar supplemented with 0.04 µg/mL triclosan (middle), or LB agar supplemented with 0.04 µg/mL triclosan and 0.5% arabinose. All recombinants grew without selection, and none grew in the presence of triclosan without arabinose. With triclosan and arabinose induction, the fabI, infB, and marAB recombinants consistently grew better than the controls.

Figure 6.

Overview of the major cellular pathways revealed as being involved in triclosan sensitivity by TraDIS-Xpress. Proteins with blue name labels appear beneficial to survival in the presence of triclosan; those with red name labels contribute to triclosan sensitivity. Dotted lines indicate potential routes of triclosan movement across the inner membrane.