The Canonical Long-Chain Fatty Acid Sensing Machinery Processes Arachidonic Acid To Inhibit Virulence in Enterohemorrhagic Escherichia coli

Abstract

The mammalian gastrointestinal tract is a complex biochemical organ that generates a diverse milieu of host- and microbe-derived metabolites. In this environment, bacterial pathogens sense and respond to specific stimuli, which are integrated into the regulation of their virulence programs. Previously, we identified the transcription factor FadR, a long-chain fatty acid (LCFA) acyl coenzyme A (acyl-CoA) sensor, as a novel virulence regulator in the human foodborne pathogen enterohemorrhagic Escherichia coli (EHEC). Here, we demonstrate that exogenous LCFAs directly inhibit the locus of enterocyte effacement (LEE) pathogenicity island in EHEC through sensing by FadR. Moreover, in addition to LCFAs that are 18 carbons in length or shorter, we introduce host-derived arachidonic acid (C_20:4) as an additional LCFA that is recognized by the FadR system in EHEC. We show that arachidonic acid is processed by the acyl-CoA synthetase FadD, which permits binding to FadR and decreases FadR affinity for its target DNA sequences. This interaction enables the transcriptional regulation of FadR-responsive operons by arachidonic acid in EHEC, including the LEE. Finally, we show that arachidonic acid inhibits hallmarks of EHEC disease in a FadR-dependent manner, including EHEC attachment to epithelial cells and the formation of attaching and effacing lesions. Together, our findings delineate a molecular mechanism demonstrating how LCFAs can directly inhibit the virulence of an enteric bacterial pathogen. More broadly, our findings expand the repertoire of ligands sensed by the canonical LFCA sensing machinery in EHEC to include arachidonic acid, an important bioactive lipid that is ubiquitous within host environments.IMPORTANCE Polyunsaturated fatty acids (PUFAs) play important roles in host immunity. Manipulation of lipid content in host tissues through diet or pharmacological interventions is associated with altered severity of various inflammatory diseases. Our work introduces a defined host-pathogen interaction by which arachidonic acid, a host-derived and dietary PUFA, can impact the outcome of enteric infection with the human pathogen enterohemorrhagic Escherichia coli (EHEC). We show that long-chain fatty acids including arachidonic acid act as signaling molecules that directly suppress a key pathogenicity island in EHEC following recognition by the fatty acyl-CoA-responsive transcription factor FadR. Thus, in addition to its established effects on host immunity and its bactericidal activities against other pathogens, we demonstrate that arachidonic acid also acts as a signaling molecule that inhibits virulence in an enteric pathogen.

Keywords: FadR; PUFA; arachidonic acid; enterohemorrhagic E. coli (EHEC); fatty acid; host-pathogen interactions; infection; locus of enterocyte effacement (LEE); omega 6; virulence regulation.

FIG 1

LCFAs inhibit the LEE pathogenicity island in EHEC. (A) Schematic of Ler regulation of the LEE pathogenicity island in EHEC. (B to E) EHEC was grown microaerobically under LEE-inducing conditions in the presence of 8 µM palmitic acid (PA), 8 µM arachidonic acid (AA), or the vehicle control (V). (B) Relative expression of the LEE-carried gene espA in EHEC as assessed by targeted qRT-PCR. (C) EHEC secretion of the LEE effector EspA at late log phase as assessed by Western blotting (right) and densitometry (left). (D) Relative expression of representative genes from each of the 5 LEE operons in EHEC as assessed by targeted qRT-PCR. (E) EHEC secretion of EspA and EspB at late log phase as assessed by Western blotting (right) and densitometry (left). (F) Relative expression of espA in EHEC in response to a range of arachidonic acid (AA) doses as assessed by targeted qRT-PCR. (G) EHEC secretion of EspA at late log phase in response to a range of arachidonic acid (AA) doses as assessed by Western blotting (right) and densitometry (left). LC, loading control. All data are represented as the mean ± SEM from at least 3 independent experiments. P values were determined by Student’s unpaired t test (B and D) Mann-Whitney test (C and E), one-way ANOVA (F), or Kruskal-Wallis test (G). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

FIG 2

Arachidonic acid does not impact EHEC growth. (A) Aerobic growth kinetics of EHEC in LB medium with the indicated concentrations of arachidonic acid or vehicle control (Veh). (B and C) EHEC was grown microaerobically under LEE-inducing conditions in the presence of 8 µM arachidonic acid (A) or the vehicle control (V). (B) Microaerobic growth kinetics of EHEC. (C) Quantitative culture of EHEC at the indicated time points. (D) Aerobic growth kinetics of S. aureus in brain heart infusion (BHI) medium with the indicated concentrations of arachidonic acid or vehicle control. (E) Quantitative culture of EHEC grown microaerobically in minimal medium in the absence of a carbon source (V) or in the presence of glucose (glu), 8 80 µM palmitic acid (P), or 1.2 mM arachidonic acid (A) as sole carbon source. All data are represented as the mean ± SEM from at least 3 independent experiments. The dashed horizontal line represents the CFU/ml of EHEC recovered without a carbon source.

FIG 3

Inhibition of the LEE by arachidonic acid is dependent on FadR. (A) Schematic of canonical long-chain fatty acid (LCFA) sensing in Escherichia coli. (B to D) EHEC was grown microaerobically under LEE-inducing conditions in the presence of 8 µM arachidonic acid (AA) or the vehicle control (V). (B) Secretion of the LEE effectors EspA and EspB at late log phase by EHEC WT or ΔfadL strain as assessed by Western blotting (left) and densitometry (right). (C) Secretion of EspA and EspB by EHEC WT or ΔfadR strain as assessed by Western blotting (left) and densitometry (right). (D) Secretion of EspA and EspB by EHEC WT or ΔfadE strain as assessed by Western blotting (right) and densitometry (left). LC, loading control. All data are represented as the mean ± SEM from at least 3 biological replicates. P values were determined by Kruskal-Wallis test. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

FIG 4

FadD catalyzes the production of arachidonoyl-CoA in EHEC. In vitro acyl-CoA synthesis activity assay with EHEC WT or ΔfadD bacterial lysates using 20 µM oleic acid (A) or 20 µM arachidonic acid (B) as the substrates. All data are represented as the mean ± SEM from 6 biological replicates. P values were determined by Mann-Whitney test.

FIG 5

Arachidonoyl-CoA directly interacts with FadR to inhibit binding at its DNA targets. (A) Thermal unfolding of recombinant FadR is monitored using SYPRO Orange. Data were collected in the presence or absence of the indicated long-chain fatty acid (LCFA) or acyl-CoA at 20 µM, leading to a rightward shift in the unfolding transition. All data depict representative curves from 3 independent experiments with 3 technical replicates. (B and C) Isothermal titration calorimetry (ITC) isotherms of the EHEC FadR protein with approximately 0.5 mM arachidonoyl-CoA (AA-CoA) (B) or palmitoyl-CoA (PA-CoA) (C) at 20°C. The raw thermogram of each experiment is shown. The lower panel in panels B and C indicates the titration curve fitted to the one-site model. Residuals between the data and the fit lines are shown in the lowest plot. All data were integrated using NITPIC and analyzed in SEDPHAT. (D and E) Electrophoretic mobility shift assays (EMSAs) for EHEC FadR in the presence of the LCFAs palmitic acid (PA) and arachidonic acid (AA) or their respective acyl-CoAs at 20 µM using the FadR binding sites within the fadL (D) and ler (E) promoters. All images are representative of 2 independent experiments.

FIG 6

FadR acts as a transcriptional activator of the LEE. (A and B) EHEC was grown microaerobically under LEE-inducing conditions in the presence of 8 µM arachidonic acid (AA), 8 µM palmitic acid (PA), or the vehicle control (V). (A) ChIP-qPCR of N-terminally tagged FadR. Probes are designed to amplify the promoter regions of fadL (positive control), rpoZ (negative control), or ler (LEE1). Data are displayed as percentages of the protein input. (B) Relative expression of fadL as assessed by targeted qRT-PCR. (C to F) EHEC WT or ΔfadR strain was grown aerobically in LB medium. Relative expression of fadL (C), fabB (D), ler (E), or espA (F) in EHEC WT or ΔfadR strain at late log growth phase as assessed by qRT-PCR. All data are represented as the mean ± SEM from 3 biological replicates. P values were determined by two-way ANOVA (A) or Student’s unpaired t test (B to F). *, P < 0.05; **, P < 0.01; ***, P < 0.001.

FIG 7

Arachidonic acid inhibits epithelial EHEC infection in a FadR-dependent manner. (A) Schematic of EHEC in vitro infection model. (B to E) EHEC was grown microaerobically under LEE-inducing conditions (DMEM-low glucose) with 8 µM arachidonic acid (AA) or the vehicle control (V). At late log phase, arachidonic acid- or vehicle-treated EHEC was transferred to a coculture system with HeLa cells to initiate EHEC infection. (B) Representative confocal microscopy images of LEE-dependent pedestal formation (white arrowheads) on epithelial cells by mCherry-expressing EHEC. DNA (blue) is stained with DAPI, and actin (green) is stained with FITC-phalloidin. Images at 40×. (C) Percentage of epithelial cells infected with EHEC pedestals at 4 h postinfection. (D) Quantity of EHEC pedestals per infected epithelial cell at 4 h postinfection. At least 275 cells in 17 fields at 40× were enumerated for each group. (E) Quantitative culture of EHEC WT and its isogenic mutants recovered from an adhesion assay with HeLa cells at 4 h postinfection. (F) EHEC was grown microaerobically under LEE-inducing conditions (DMEM-low glucose) with arachidonic acid (A) or the vehicle control (V). At late log phase, arachidonic acid- or vehicle-treated EHEC was transferred to a coculture system with Caco-2 cells to initiate EHEC infection. Quantitative culture of EHEC and its isogenic mutants recovered from an adhesion assay with Caco-2 monolayers at 5 h postinfection. (G) Schematic of model depicting long-chain fatty acid (LCFA) regulation of EHEC virulence and the canonical FadR regulon. All data are represented as the mean ± SEM from 3 independent experiments. P values were determined by Mann-Whitney test (C and D) or one-way ANOVA (E and F). **, P < 0.01; ***, P < 0.001.