Dual Repression of the Multidrug Efflux Pump CmeABC by CosR and CmeR in Campylobacter jejuni

Abstract

During transmission and intestinal colonization, Campylobacter jejuni, a major foodborne human pathogen, experiences oxidative stress. CosR, a response regulator in C. jejuni, modulates the oxidative stress response and represses expression of the CmeABC multidrug efflux pump. CmeABC, a key component in resistance to toxic compounds including antimicrobials and bile salts, is also under negative regulation by CmeR, a TetR family transcriptional regulator. How CosR and CmeR interact in binding to the cmeABC promoter and how CosR senses oxidative stress are still unknown. To answer these questions, we conducted various experiments utilizing electrophoretic mobility shift assays and transcriptional fusion assays. CosR and CmeR bound independently to two separate sites of the cmeABC promoter, simultaneously repressing cmeABC expression. This dual binding of CosR and CmeR is optimal with a 17 base pair space between the two binding sites as mutations that shortened the distance between the binding sites decreased binding by CmeR and enhanced cmeABC expression. Additionally, the single cysteine residue (C218) of CosR was sensitive to oxidation, which altered the DNA-binding activity of CosR and dissociated CosR from the cmeABC promoter as determined by electrophoretic mobility shift assay. Replacement of C218 with serine rendered CosR insensitive to oxidation, suggesting a potential role of C218 in sensing oxidative stress and providing a possible mechanism for CosR-mediated response to oxidative stress. These findings reveal a dual regulatory role of CosR and CmeR in modulating cmeABC expression and suggest a potential mechanism that may explain overexpression of cmeABC in response to oxidative stress. Differential expression of cmeABC mediated by CmeR and CosR in response to different signals may facilitate adaptation of Campylobacter to various environmental conditions.

Keywords: Campylobacter jejuni; CmeABC; CmeR; CosR; efflux pump; oxidative stress; transcriptional regulation.

FIGURE 1

Sensitivity of CosR to oxidation by hydrogen peroxide. Binding of 250 ng of rCosRWT (lanes 2–5) or rCosRC218S (lanes 6–9) to DIG-labeled 11168 cmeABC promoter DNA (0.05 pmol) by EMSA. No protein was added to lane 1 as a probe-only control. The promoter probe and protein were incubated for 30 min prior to addition of hydrogen peroxide. Hydrogen peroxide was added for final concentrations of 0 nM (lanes 2 and 6), 5 nM (lanes 3 and 7), 10 nM (lanes 4 and 8), or 20 nM (lanes 5 and 9). After addition of hydrogen peroxide, all reactions were incubated for an additional 30 min before electrophoresis.

FIGURE 2

Dual binding of CosR and CmeR to variants of the cmeABC promoter sequence. (A) Alignment of cmeABC promoter sequences from strains NCTC 11168 (11168), 81–176, and X7199, along with the in silico designed P14D probe. The CosR binding site is indicated in bold and the CmeR binding site indicated in lowercase, underlined italics. Mutations in the CmeR binding site are indicated in bold, lowercase, underlined italics. (–) indicates a deleted base. (B) EMSA of the cmeABC promoter probes from 81–176 (lanes 1–4) and X7199 (lanes 5–8) incubated with 200 ng of rCmeRSS (lanes 2 and 6), 400 ng of rCosRC218S (lanes 3 and 7), or both proteins (lanes 4 and 8). No protein was added to lanes 1 and 5 as probe-only controls. (C) EMSA of the cmeABC promoter probes from 81 to 176 (lanes 1–4) and P14D (lanes 5–8) incubated with rCmeRSS, rCosRC218S (lanes 3 and 7), or both proteins as in (B). No protein was added to lanes 1 and 5 as probe-only controls. All promoter probes were 0.05 pmol per reaction.

FIGURE 3

Effect of CosR inhibition by anti-cosR-PNA on transcription from the cmeABC promoter in the presence or absence of CmeR. Expression from 11168, 81–176, or X7199 cmeABC promoters or the Cj0561c promoter (561) in β-galactosidase assays. Presence or absence of CmeR was determined by using 81–176 wild-type (A) and 81–176ΔcmeR (B) for the transcriptional fusions. Cultures were incubated with (+ anti-cosR-PNA) or without (- anti-CosR-PNA) 1.5 μM of the anti-cosR-PNA. The Cj0561c promoter was a control for regulation solely by CmeR, not by CosR. Data represents +means with standard error from three independent experiments. Unpaired Student’s t-test with Welch’s correction was used for comparison. ^∗∗p < 0.01 and ^∗∗∗p < 0.001, NS, not significant; p > 0.05 of the means with significance set at 0.05.