AraC-type regulators HilC and RtsA are directly controlled by an intestinal fatty acid to regulate Salmonella invasion

Abstract

Invasion of the intestinal epithelium is an essential but energetically expensive survival strategy and is, therefore, tightly regulated by using specific cues from the environment. The enteric pathogen Salmonella controls its invasion machinery through the elegant coordination of three AraC-type transcription activators, HilD, HilC, and RtsA. Most environmental signals target HilD to control invasion, whereas HilC and RtsA are known only to augment these effects on HilD. Here we show that a fatty acid found in the murine colon, cis-2-hexadecenoic acid (c2-HDA), represses Salmonella invasion by directly targeting HilC and RtsA, in addition to HilD. c2-HDA directly binds each of these regulators and inhibits their attachment to DNA targets, repressing invasion even in the absence of HilD. Fatty acid binding, however, does not affect HilC and RtsA protein stability, unlike HilD. Importantly, we show that HilC and RtsA are highly effective in restoring HilD production and invasion gene expression after elimination of the repressive fatty acid c2-HDA. Together, these results illuminate a precise mechanism by which HilC and RtsA may modulate invasion as Salmonella navigates through different regions of the intestine, contributing to our understanding of how this enteric pathogen senses and adapts to a diverse intestinal environment while maintaining its virulence.

Keywords: Salmonella; enteric diseases; fatty acids; host-pathogen interactions.

Figure 1.. c2-HDA directly interacts with HilC and RtsA

A. Structure of c2-HDA. B. Increasing concentrations of HilC or RtsA were added to 96‐well plates pre-adsorbed with c2-HDA. Binding was detected by anti-His tag antibody. All measured values represent the average of at least three independent experiments (mean ± S.D.). Binding constants (K_d values) were determined by nonlinear curve fitting using Kaleidagraph (Synergy Software) with error values calculated automatically.

Figure 2.. c2-HDA prevents the binding of HilC and RtsA to their target DNA

A-C. EMSA of hilA promoter DNA (25 nM) in the presence of 200 μM HilC (A), RtsA (B) or HilD (C) with DMSO and increasing concentrations of c2-HDA. In all images, lane 1: DNA, lane 2: DNA + protein, lane 3–8: DNA + protein + DMSO or c2-HDA.

Figure 3.. c2-HDA represses HilC- and RtsA-mediated SPI-1 activation and invasion

A-B. Salmonella strains with tetRA-hilC-3XFLAG chromosomal fusion (ΔhilD ΔrtsA) (A) and tetRA-rtsA-3XFLAG chromosomal fusion (ΔhilD ΔhilC) (B) and sopB-luxCDABE reporter plasmid were grown in presence of 20 μg ml⁻¹ tetracycline and 2 μM concentration of different fatty acids. DMSO and Ethanol (EtOH) were used as solvent controls. Area under curve (AUC) of normalized luminescence (raw luminescence divided by bacterial culture density) of cultures that received DMSO were set to 100% and other treatments were normalized accordingly. Data show mean AUC ±SD of five replicates. Difference between DMSO vs. other treatments were calculated by Kruskal-Wallis test. *, ** and *** indicate p<0.05, p<0.01 and p<0.001 respectively. C. Salmonella strains lacking two out of three regulators with the remaining regulator under inducible tetracycline promoter or lacking all three regulators ΔhilD ΔhilC ΔrtsA (ΔΔΔ), were grown in the presence of DMSO or 20 μM concentration of different fatty acids and invasion of HEp-2 cells was measured by gentamicin protection assay. Invasion by wild type (WT) Salmonella strain treated with DMSO was set to 100% and other treatments were normalized accordingly. Data show mean ±SD (n=4). Difference between DMSO vs. other treatments for each strain were calculated by two-way ANOVA. * indicates p<0.05.

Figure 4.. Binding of c2-HDA does not promote degradation of HilC and RtsA

A-C. Western blots of HilD (A), HilC (B) and RtsA (C) degradation assays. Salmonella strains lacking two of the three regulators with the remaining regulator under a tetracycline inducible promoter and C-terminal 3XFLAG were grown in the presence of DMSO or 20 μM c2-HDA. Protein expression was paused using an antibiotic cocktail and cultures were collected at indicated time points to analyze the stability of the regulator by western blotting. Respective calculated half-lives are indicated. ND is Not Determined.

Figure 5.. HilC and RtsA are necessary for the restoration of SPI-1 expression after exposure to c2-HDA

A. Summary of the protocol used to study restoration of invasion gene expression. Fully grown overnight cultures were sub-cultured and grown till they reached OD₆₀₀ of 1. For each strain, equal number of bacteria were added to three tubes, one received DMSO and the other two received 5 μM c2-HDA. These were grown for an hour at 37°C after which one of the cultures receiving c2-HDA was washed and resuspended in LB containing DMSO. B. Wild type or ΔhilC ΔrtsA Salmonella strains with hilD-3XFLAG plasmid were grown as summarized in A. Aliquots were collected at indicated time-points and HilD protein levels were detected by western blotting using anti-FLAG antibody. Intensities were analyzed in ImageJ. HilD band intensity at each time-point in cultures that stayed in DMSO (top-panel) were set to 100%, and intensities obtained in cultures that recovered from brief exposure to c2-HDA, at corresponding time-points were normalized accordingly. C. Wild type or ΔhilC ΔrtsA Salmonella strains with sipB::lacZ transcriptional reporter fusion were grown as summarized in A. Aliquots were collected at indicated time points and expression of the invasion gene sipB was measured by β-galactosidase assays. For each strain, at individual time-points, sipB expression in cultures that stayed in DMSO was set to 100% and c2-HDA treatments and recovery were normalized accordingly. Bars represent mean ±SD (n = 3). Differences between respective c2-HDA treatments vs. recovery were calculated by t test, ** indicates p<0.01. D. Wild type or ΔhilC ΔrtsA Salmonella strains with hilA-luxCDABE reporter plasmid were grown as summarized in A and hilA expression was measured for 10 hours in a lux assay. AUC of each strain that stayed in DMSO was set to 100% and c2-HDA treatments and recovery were normalized accordingly. Bars represent mean AUC ±SD (n = 5). Difference between recoveries of wild type vs. ΔhilC ΔrtsA were calculated by Mann-Whitney test, ** indicates p<0.01.