Advertisement

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Cross-talk between QseBC and PmrAB two-component systems is crucial for regulation of motility and colistin resistance in Enteropathogenic Escherichia coli

Abstract

The quorum sensing two-component system (TCS) QseBC has been linked to virulence, motility and metabolism regulation in multiple Gram-negative pathogens, including Enterohaemorrhagic Escherichia coli (EHEC), Uropathogenic E. coli (UPEC) and Salmonella enterica. In EHEC, the sensor histidine kinase (HK) QseC detects the quorum sensing signalling molecule AI-3 and also acts as an adrenergic sensor binding host epinephrine and norepinephrine. Downstream changes in gene expression are mediated by phosphorylation of its cognate response regulator (RR) QseB, and ʻcross-talksʼ with non-cognate regulators KdpE and QseF to activate motility and virulence. In UPEC, cross-talk between QseBC and TCS PmrAB is crucial in the regulation and phosphorylation of QseB RR that acts as a repressor of multiple pathways, including motility. Here, we investigated QseBC regulation of motility in the atypical Enteropathogenic E. coli (EPEC) strain O125ac:H6, causative agent of persistent diarrhoea in children, and its possible cross-talk with the KdpDE and PmrAB TCS. We showed that in EPEC QseB acts as a repressor of genes involved in motility, virulence and stress response, and in absence of QseC HK, QseB is likely activated by the non-cognate PmrB HK, similarly to UPEC. We show that in absence of QseC, phosphorylated QseB activates its own expression, and is responsible for the low motility phenotypes seen in a QseC deletion mutant. Furthermore, we showed that KdpD HK regulates motility in an independent manner to QseBC and through a third unidentified party different to its own response regulator KdpE. We showed that PmrAB has a role in iron adaptation independent to QseBC. Finally, we showed that QseB is the responsible for activation of colistin and polymyxin B resistance genes while PmrA RR acts by preventing QseB activation of these resistance genes.

Author summary

Enteropathogenic Escherichia coli (EPEC) is a human pathogen and the leading cause of diarrhoea in children under 5 years in low-income countries. EPEC, and specially the subgroup denominated atypical EPEC, has been associated with community-acquired persistent diarrhoea and also is an important agent of post-weaning diarrhoea in young piglets. Two-component systems (TCS) are signalling system used by bacteria to sense and adapt to different stimuli. Bacteria use TCS to detect host signals as cues to exert virulence in their preferred niche. QseBC is a TCS that has been linked to regulation of motility and virulence in other pathogenic E. coli. In this study we broaden our understanding of QseBC TCS and its role in regulating virulence traits such as motility in atypical EPEC. We also investigate its capabilities to interact with other TCS, named KdpDE and PmrAB, and how these interactions are responsible for regulation of motility and resistance to antimicrobials such as colistin.

Citation: Fernandez-Ciruelos B, Potmis T, Solomin V, Wells JM (2023) Cross-talk between QseBC and PmrAB two-component systems is crucial for regulation of motility and colistin resistance in Enteropathogenic Escherichia coli. PLoS Pathog 19(12): e1011345. https://doi.org/10.1371/journal.ppat.1011345

Editor: Thomas Guillard, Reims Champagne-Ardenne University Faculty of Medicine: Universite de Reims Champagne-Ardenne UFR de Medecine, FRANCE

Received: April 5, 2023; Accepted: November 13, 2023; Published: December 7, 2023

Copyright: © 2023 Fernandez-Ciruelos et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the manuscript and its Supporting Information files.

Funding: This project received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement number 765147 (awarded to JMW). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Enteropathogenic Escherichia coli (EPEC) is a common human pathogen that adheres to the small intestine and is considered the leading cause of diarrhoea in children under 5 years of age in low-income countries [1]. A recent European study reported a high prevalence of EPEC in samples collected from community-acquired gastroenteritis, suggesting this pathogen is re-emerging [2]. EPEC can also cause diarrhoeal disease in livestock [3] and companion animals [4,5], and can be zoonotic [6] posing an additional burden. EPEC strains are classified into typical (tEPEC) and atypical (aEPEC), depending on the presence of the pEAF virulence plasmid that encodes proteins involved in the formation of the bundle-forming pilus (Bfp) [7]. The Bfp mediates tEPEC adherence to intestinal epithelium and is responsible for the appearance of localised microcolonies on the cell surface otherwise known as a localized adherence (LA) pattern [8]. Due to the absence of bfp, aEPEC usually shows a diffused adherence (DA), aggregative adherence (AA) or localized adherence-like (LAL) pattern [9]. Even though tEPEC has been more extensively studied, aEPEC is a more common cause of diarrhoea [10], and is associated with a more persistent diarrheal disease than tEPEC [11,12]. On Spanish farms most cases of post-weaning diarrhoea are associated with enterotoxigenic E. coli (ETEC) but this is closely followed by aEPEC. Conversely, aEPEC was more prevalent in cases of diarrhoea in suckling pigs [3].

EPEC shares many similarities with Enterohaemorrhagic E. coli (EHEC) [13], associated with foodborne outbreaks that cause bloody diarrhoea in humans of which about 5% lead to haemolytic uremic syndrome (HUS) [14,15]. The main differences between EHEC and EPEC are the ability of EHEC to produce Shiga-Toxin and colonize the large intestine. EHEC and EPEC share a pathogenicity island encoding the locus of enterocyte effacement (LEE) which encodes a type 3 secretion system (T3SS) that injects different effectors into the mammalian intestinal host cell [16], causing microvilli effacement, intimate attachment to the host cell [17], actin pedestal formation, loss of epithelial tight junctions and apoptosis, all which leads to severe diarrhoea [18].

EPEC needs to sense different host signals to exert virulence in their preferred niche. One of the main signalling systems associated to virulence in both EPEC and EHEC is the QseBC two-component system (TCS) [19]. QseBC is a typical TCS comprising the histidine kinase (HK) QseC and the response regulator (RR) QseB [20]. In EHEC, QseC HK has been reported to detect adrenergic signals present in the intestinal tract (epinephrine and norepinephrine), and a novel bacterial quorum sensing molecule AI-3 and upon detection induce virulence and motility [21]. Regulation of virulence and motility by QseBC has been extensively studied using knock-out mutants. Deletion of QseC HK decreased motility, virulence [20,22] and expression of recA which is important for regulation of the SOS response [23]. This led to the hypothesis that QseC acts as an activator of virulence and motility, as well as other pathways. This classical theory of QseBC regulation (Fig 1A) theorized that, upon signal detection, QseC undergoes an autophosphorylation reaction, and in turn phosphotransfer to the cognate RR QseB, which upregulates flhCD the master regulator of flagella, activating motility [23]. Failing to replicate ΔqseC phenotypes in ΔqseB and ΔqseBΔqseC deletion mutants and other results led to the hypothesis that unphosphorylated QseB acts as a repressor of flhCD operon [23,24]. Other studies proposed that QseC also phosphorylates the non-cognate RRs KdpE to activate ler expression, and QseF to upregulate SOS response and Shiga-toxin expression [23] (Fig 1A).

thumbnail
Download:
Fig 1. Simplified theories of QseBC gene regulation in EHEC (A) and UPEC (B).

A. In EHEC, QseC phosphorylates three different RRs KdpE, QseB and QseF that in turn upregulate LEE, flhC and recA genes involved in A/E lesion formation, motility and SOS response, respectively [23]. B. In UPEC, it is theorized that QseB RR acts as a repressor of motility genes and other genes. QseB RR is phosphorylated by the non-cognate HK PmrB and de-phosphorylated by QseC that controls QseB repressor activity [25]. Created with Biorender.com.

https://doi.org/10.1371/journal.ppat.1011345.g001

In Uropathogenic E. coli (UPEC) a different mechanism for QseBC regulation of motility and virulence was proposed by Guckes et al [25] involving cross-regulation with the PmrAB TCS. PmrAB detects extracellular iron [26] and is involved in resistance to polymyxin B and colistin [27] in E. coli and Salmonella enterica. Guckes and co-workers hypothesized that the phenotypes observed in the ΔqseC mutant are due to upregulation and over-activation of QseB, that acts as a repressor of motility and other pathways. The authors showed that the PmrB HK can phosphorylate but not dephosphorylate QseB RR in vitro [25]. Conversely, QseC was shown to de-phosphorylate QseB in vitro and to also sequester it, so it is not available for regulation [28]. This model predicts that deletion of the QseC HK will lead to phosphorylation of QseB RR by the non-cognate PmrB HK, leading to auto-induced expression of QseB and repression of the motility regulator flhCD. A ΔqseCΔpmrB mutant was shown to recover the WT phenotype, fitting with the idea that in a ΔqseC background PmrB phosphorylates QseB RR [25] (Fig 1B).

Due to the avirulent and non-motile phenotypes of a QseC mutant, QseC has been proposed as a good anti-virulence target [29]. Sperandio´s group developed LED209 as inhibitor of the QseC sensing domain which had a similar phenotype to a ΔqseC mutant [22]. However, for the further development of inhibitors of TCS as anti-virulence targets, it is important to have a good understanding of the cross-regulatory networks in different E. coli phenotypes [30]. Therefore, in this study, we explored the role of QseBC in virulence in atypical Enteropathogenic E. coli strain O125ac:H6 [31]. We also investigated potential cross-talk between QseBC and the TCSs PmrAB and KdpDE to regulate motility and expression of virulence genes. We found that, in aEPEC O125:H6, QseBC and PmrAB TCSs cross-regulate motility and virulence genes, and this crosstalk is the responsible of the aberrant phenotypes seen in an aEPEC QseC mutant as previously suggested by Guckes et al [25]. We found that KdpDE has a role in motility that does not involve cross-talk with QseBC TCS. Furthermore, we showed that in aEPEC colistin resistance is mediated via crosstalk between QseBC and PmrAB. Based on these results we suggest a theory for QseBC regulation in Enteropathogenic E. coli in which QseB acts as a master repressor of motility and virulence and activator of colistin resistance and is tightly regulated via interactions with the TCS PmrAB.

Results

Genetic manipulation of atypical EPEC str. O125ac:H6 using a two-plasmid CRISPR-Cas system

Atypical Enteropathogenic E. coli strain O125ac:H6 (DSM8711) was acquired from the Leibniz Institute DSMZ collection located in Germany. The presence of the restriction and modification system (HsdR) typically found in clinical strains of E. coli [32] was checked by PCR using primers pBM001+pBM002 and shown to be absent. The absence of this restriction-modification system in strain DSM8711 makes it a good candidate for genetic modification of atypical E. coli. To generate 6 individual mutants of the two-component system genes ΔqseB, ΔqseC, ΔpmrA, ΔpmrB, ΔkdpD and ΔkdpE we used the two-plasmid CRISPR-Cas system developed by Jiang et al [33]. Of particular interest was the cross-regulation between QseBC, reported to have a main role in virulence, and PmrAB and KdpDE, reported to interact with QseBC in UPEC and EHEC, respectively. To study these interactions, we constructed 10 double gene deletion mutants (ΔqseBΔqseC, ΔpmrAΔpmrB, ΔpmrAΔqseB, ΔpmrAΔqseC, ΔpmrBΔqseB, ΔpmrBΔqseC, ΔkdpDΔqseB, ΔkdpDΔqseC, ΔkdpEΔqseB and ΔkdpEΔqseC).

For each construct a guide was cloned into pTargetF and a 1kb repair template created by PCR as described in Material and Methods. All templates were designed to introduce in-frame deletions to avoid polar effects on expression of the remaining genes in the operon. For the response regulator gene mutants (qseB, pmrA and kdpE) at least 80% of both the receiver and DNA-binding domain were deleted. For the histidine kinase genes (qseC, pmrB and kdpD) around 1 kb was deleted from the cytoplasmic domain that contains the catalytic domain and other important residues for autophosphorylation (Fig 2), ensuring that neither phosphorylation nor de-phosphorylation of the response regulator by HKs would be possible. The sensory and transmembrane domains of HK were left intact.

thumbnail
Download:
Fig 2. The strategy used for constructing gene deletions showing the histidine kinase (QseC) and response regulator (QseB) deletion as an example.

In histidine kinases an in-frame deletion was inserted to eliminate more than 80% of the cytoplasmic domain that contains the catalytic domains. Sensory domain and transmembrane region were left functional. More than 80% of the response regulators were deleted, including parts of both the receiver domain and DNA-binding domain. In-frame deletions were used to avoid polar effects on expression of the rest of the operon. Created with Biorender.com.

https://doi.org/10.1371/journal.ppat.1011345.g002

Deletion of histidine kinase qseC slows growth rate in late-exponential phase

Most of the mutations do not affect growth in LB media compared to the WT control (Fig 3A and 3B). However, four mutants, ΔqseC (p<0.0001), ΔpmrAΔqseC (p<0.0001), ΔkdpDΔqseC (p<0.0001) and ΔkdpEΔqseC (p<0.0001) showed a delay in growth in late-exponential phase. Growth constants and standard deviation can be found in supporting S1 Table. As all affected mutants contain ΔqseC mutation, absence of QseC HK is most likely responsible for this phenotype. However, growth of ΔpmrBΔqseC was not affected.

thumbnail
Download:
Fig 3. Growth curves of EPEC O125ac:H6 deletion mutants compared to the WT.

Growth curves represent OD600 measurements recorded every 30 minutes for 15 hours for EPEC O125ac:H6 WT, ΔqseB, ΔqseC, ΔqseBΔqseC, ΔpmrA, ΔpmrB, ΔpmrAΔpmrB, ΔpmrAΔqseB, ΔpmrAΔqseC, ΔpmrBΔqseB, ΔpmrBΔqseC (A) and ΔkdpD, ΔkdpE ΔkdpDΔqseB, ΔkdpDΔqseC, ΔkdpEΔqseB and ΔkdpEΔqseC (B). Error bars represent the Standard Deviation, n = 4.

https://doi.org/10.1371/journal.ppat.1011345.g003

Response regulator QseB is a key repressor of motility and is activated by the histidine kinase PmrB in a ΔqseC mutant

The soft agar motility assay was used to determine the individual contributions of QseBC and PmrAB as well as their potential interactions on bacterial motility. Deletion of the QseB RR, alone and in combination with QseC HK, PmrA RR and PmrB HK leads to a significant increase in motility compared to the WT, indicating that QseB acts as a repressor of motility. Single and double mutants from PmrAB TCS show similar motility to the WT, which suggests that this TCS alone is not involved in regulating motility (Fig 4).

thumbnail
Download:
Fig 4. Relative motility of QseBC and PmrAB single and double mutants.

A. Relative motility compared to the WT of EPEC O125ac:H6 ΔqseB, ΔqseC, ΔqseBΔqseC, ΔpmrA, ΔpmrB, ΔpmrAΔpmrB, ΔpmrAΔqseB, ΔpmrAΔqseC, ΔpmrBΔqseB, ΔpmrBΔqseC. B. Relative motility compared to the WT of EPEC O125ac:H6 ΔqseB complemented with WT qseB, ΔpmrBΔqseC complemented with WT pmrb ΔqseC complemented with qseC, WT QseBD51A, ΔqseC QseBD51A and ΔqseC PmrBH152A. WT motility is calculated as the relative motility of the WT per plate compared to the average motility of the WT in the experiment. Motility plates that exemplify the seen motility are depicted. * p-value <0.05; ** <0.01; *** <0.001; **** <0.0001 via one-way ANOVA with Tukey’s multiple comparison (A) and via Krukal-Wallis test with Dunn’s multiple comparison (B), n = 8.

https://doi.org/10.1371/journal.ppat.1011345.g004

The ΔqseC mutant shows a reduction of motility. The deletion of both QseC HK and PmrA RR leads to even lower motility than the single QseC mutant. However, E. coli ΔpmrBΔqseC, with simultaneous deletion of the QseC and PmrB HKs, recovers the WT phenotype (Fig 4A). These data are consistent with the hypothesis that QseB RR acts as a repressor of motility and is activated by phosphorylation via PmrB in the QseC mutant. To prove that off-target effects are not responsible for the observed phenotypic changes we complemented the ΔqseB strain with qseB WT gene, leading to WT motility phenotypes (Fig 4B). As hypothesized ΔpmrBΔqseC mutant complemented with pmrB WT gene has reduced motility due to absence of qseC. In a second step, complementation with WT qseC gene restored the WT phenotypes (Fig 4B). These gene complementation studies show that off-target effects are not responsible for the changes in phenotypes seen in the mutants.

Finally, to study the contribution of phosphorylation in the cross-regulation between QseBC and PmrAB, we constructed non-phosphorylatable mutants of QseB and PmrB. QseBD51A shows a similar phenotype to WT indicating that the presence of unphosphorylated QseB is sufficient to repress motility to WT levels. PmrBH152A would be unable to phosphorylate QseB. Introduction of PmrBH152A in a ΔqseC mutant restores WT motility phenotypes, indicating that PmrB is likely responsible for phosphorylation of QseB in a ΔqseC mutant (Fig 4B). Introduction of the non-phosphorylatable QseBD51A into a ΔqseC background also restores WT levels of motility, corroborating the hypothesis that phosphorylation of QseB is responsible of the aberrant phenotypes in the ΔqseC mutant (Fig 4B).

The two-component system KdpDE does not crosstalk with QseBC in the regulation of motility

KdpDE TCS has also been reported to interact with QseBC to regulate virulence in EHEC. Motility data of single and double mutants is shown in Fig 5. Absence of KdpD HK lowers motility significantly compared to the control, however, it is not comparable to the lack of motility showed by the QseC mutant. Double mutants ΔkdpDΔqseB and ΔkdpEΔqseB, and ΔkdpDΔqseC and ΔkdpEΔqseC show similar phenotypes to ΔqseB and ΔqseC, respectively, suggesting that there is no cross-regulation between these TCSs in the regulation of motility. Finally, KdpE RR single mutant shows similar motility to the WT, indicating that KdpD regulation of motility is through a third unidentified party.

thumbnail
Download:
Fig 5. Relative motility of QseBC and KdpDE single and double mutants.

Relative motility compared to the WT of eight replicates is depicted for EPEC O125ac:H6 ΔqseB, ΔqseC, ΔkdpD, ΔkdpE ΔkdpDΔqseB, ΔkdpDΔqseC, ΔkdpEΔqseB and ΔkdpEΔqseC. EPEC O125ac:H6 WT motility is calculated as the relative motility of the WT per plate compared to the average motility of the WT in the experiment. Motility plates that exemplify the seen motility are depicted. * p-value <0.05; ** <0.01; *** <0.001; **** <0.0001 via one-way ANOVA with Tukey’s multiple comparison, n = 8.

https://doi.org/10.1371/journal.ppat.1011345.g005

Motility of atypical E. coli is affected by iron but not by adrenergic signals or AI-3 in absence of the PmrAB TCS

QseBC signalling is activated by adrenergic hormones, such as epinephrine, norepinephrine and the autoinducer-3 [21,23] (Fig 6A), while PmrAB is activated by extracellular iron [26]. We explored the effects of these molecules and iron on EPEC O125ac:H6 WT motility by including different concentrations (500 nM, 5 μM, 50 μM and 500 μM) of each in the soft agar. In contrast to other findings reported in the literature [34,35], no significant change in motility was observed with epinephrine at any of the tested concentrations. AI-3 and Fe3+ also did not affect motility of the WT (Fig 6B).

thumbnail
Download:
Fig 6. Effect of adrenergic signals (epinephrine), AI-3 and iron in the motility of EPEC O125ac:H6.

A. Chemical structure of AI-3 [36] B. Relative motility compared to the non-treated control of EPEC O125ac:H6 in presence of 500 μM of epinephrine, AI-3 and Fe3+. C. Comparison between the relative motility of EPEC O125ac:H6 mutants compared to the WT in non-treated agar plates (green) with the relative motility of EPEC O125ac:H6 mutants compared to the WT in agar plates containing 500 μM Fe3+ (red). Motility plates of mutants that are affected by the presence of iron are depicted. * p-value <0.05; ** <0.01; *** <0.001; **** <0.0001 via one-way ANOVA with Tukey’s multiple comparisons, n = 5.

https://doi.org/10.1371/journal.ppat.1011345.g006

We then explored the effects of epinephrine, AI-3 and Fe3+ on the motility of single mutants from QseBC and PmrAB (ΔqseB, ΔqseC, ΔpmrA and ΔpmrB). No differences were found for epinephrine and AI-3. However, presence of Fe3+ significantly reduced the motility of ΔpmrA and ΔpmrB compared to the non-treated control. Motility of all mutants was then assayed in presence of 500 μM of Fe3+ (Fig 6C). Reduction of relative motility in the presence of iron was seen with mutants ΔpmrA, ΔpmrB, ΔpmrAΔpmrB, ΔpmrAΔqseB, ΔpmrBΔqseB and ΔpmrBΔqseC. This data suggests that PmrAB TCS has an independent role in adapting to metal ion stress that does not involve QseBC TCS, and in presence of excess iron, PmrAB is needed for maximum motility.

Differential gene expression using luciferase reporter assays, confirms cross-regulation of flagella and other virulence genes by QseBC and PmrAB TCS

To investigate the effects of TCS mutants on expression of genes associated with motility (flhC and fliA), LEE island expression (ler, main transcriptional regulator of the LEE island [37]) and regulation of the SOS response (recA) we constructed a luciferase reporter of the cognate promoters as explained in Material and Methods. ler expression was evaluated in DMEM medium containing low amounts of glucose, due to its poor expression in LB medium. The bla gene was used as a promoter control, as it was not expected to be regulated by TCSs.

Expression of all genes (except bla control), fliA, flhC, ler and recA was upregulated in ΔqseB, ΔqseBΔqseC, ΔpmrAΔqseB and ΔpmrBΔqseB in LB media (Fig 7A). Upregulation of the flagella regulatory genes matches with the observed increased motility (Fig 4). Expression of all tested genes in ΔpmrA, ΔpmrB, ΔpmrAΔpmrB ΔkdpD and ΔkdpE was similar to WT expression. Mutant ΔpmrBΔqseC also had a similar expression to WT in all tested genes, indicating that introduction of PmrB HK mutation into ΔqseC (HK) recovers both WT phenotype and gene expression. Finally, ΔqseC and ΔpmrAΔqseC show significant downregulation of flagella regulatory genes flhC and fliA and the stress regulator recA. In LB medium, the expression of ler in ΔqseB mutants is increased, indicating that in this condition QseB acts as a repressor. However, ler is not differently regulated compared to the WT in low glucose DMEM in any of the tested mutants, indicating that in activated conditions QseBC does not play a major role in regulation. To confirm this data we performed a qPCR in ΔqseB and ΔqseC mutants of ler and other related virulence gene espA (Fig 7B) in low glucose DMEM. Non-significant difference in expression with respect to the WT was found. This is different to what was reported in literature for EHEC [23]. Detailed graphs showing differential expression can be found in supporting data S1 Fig and raw luminescence data in S2 Table.

thumbnail
Download:
Fig 7. Differential expression of flhC, fliA, ler, recA and qseB in different EPEC O125ac:H6 QseBC, PmrAB and KdpDE mutants.

A. Differential expression according to the luciferase assay of the genes flhC, fliA, ler, recA and bla of all tested mutants. Significant differences are indicated by ++ > 2-fold; + 1-2-fold;. non-significant differences with WT; - 1–0.5-fold and—- < 0.5-fold B. Quantitative PCR of genes ler and espA in low glucose DMEM in EPEC O125ac:H6 H6 ΔqseB and ΔqseC C. qPCR of genes flhC and fliA in EPEC O125ac:H6 ΔqseB, ΔqseC, and ΔpmrAΔqseC to prove that luciferase assay was accurate. D. qPCR in EPEC O125ac:H6 ΔqseC and ΔpmrAΔqseC to check qseB expression in these mutants. * p-value <0.05; ** <0.01; *** <0.001; **** <0.0001 via one-way ANOVA with Tukey’s multiple comparisons, n = 4.

https://doi.org/10.1371/journal.ppat.1011345.g007

To check the validity of the promoter reporter assay we checked expression of flhC and fliA genes in the ΔqseB and ΔqseC mutants using qPCR. We also included ΔpmrAΔqseC because there was a significant decrease in luminescence with the bla promoter in this mutant compared to the WT. In the ΔqseB mutant flhC and fliA were upregulated but only fliA was significant. As expected, both genes were downregulated in the ΔqseC and ΔpmrAΔqseC mutants (Fig 7C), confirming the luminescence data. Differences in upregulation and downregulation seen between the luminescence reporter and qPCR data are probably due to the collection of culture for qPCR during mid-exponential phase, representing only one time point, while the luminescence reporter assay allows for kinetic measurements.

Finally, we investigated qseB expression in ΔqseC and ΔpmrAΔqseC using qPCR. QseB RR is highly upregulated in both mutants (around 100-times upregulation) (Fig 7D), supporting the hypothesis that qseB upregulation is one of the reasons for the aberrant phenotypes seen in ΔqseC and ΔpmrAΔqseC.

E. coli ΔpmrAΔqseC is resistant to colistin and polymyxin B due to constitutive upregulation of colistin resistance regulon

PmrAB has been reported to regulate colistin and polymyxin B resistance in E. coli and other Gram-negative bacteria [38]. To test whether cross-regulation between PmrAB and QseBC TCSs would also affect expression of colistin and polymyxin B resistance genes we performed MIC assays with the single and double mutants. The double mutant ΔpmrAΔqseC has a MIC (4 μg/ml)16-times higher than the WT (0.25 μg/ml). The MIC of all the other mutants was 0.25 μg/ml for both colistin and polymyxin B, same as the WT (Fig 8A). As the PmrAB regulon includes, amongst others, the proteins EptA and ArnB, involved in LPS modification pathways [39], we hypothesized that increased resistance was due to upregulation of the PmrAB regulon when PmrA RR and QseC HK are deleted. Quantitative PCR in mutants ΔqseB, ΔqseC and ΔpmrAΔqseC showed that indeed genes arnB and eptA are constitutively upregulated in ΔpmrAΔqseC compared to the WT control, explaining increased colistin and polymyxin B resistance (Fig 8B). Upon inspection of the regulatory region of arn operon in the genome of E. coli O125ac:H6, we identified the consensus binding sequence of PmrA [(C/T)TTAA(G/T)-N5-(C/T)TTAA(G/T)] [40] with a slight modification, as well as the consensus binding sequence of QseB [G(C/T)(G/A)ACAAT(A/T)A-N4-TT] [24,41] also with a slight modification (Fig 8C). This suggests that both QseB and PmrA are implicated in regulating expression of the arn operon. To prove this hypothesis we performed an electrophoretic mobility shift assay with phosphorylated and non-phosphorylated QseB and a DNA fragment containing the predicted QseB binding site in the arn operon. We found that 3 μg of phosphorylated QseB completely shifts the DNA fragment containing the predicted binding site, indicating that phosphorylated QseB binds indeed to this region (Fig 8D). Non-phosphorylated QseB seems to bind the QseB site with lower affinity (Fig 8D).

thumbnail
Download:
Fig 8. Effect of mutations in QseBC, PmrAB and KdpDE genes in colistin and polymyxin B resistance.

A. MIC data of colistin and polymyxin B (PBM) for EPEC O125ac:H6 WT, ΔqseB, ΔqseC, ΔqseBΔqseC, ΔpmrA, ΔpmrB, ΔpmrAΔpmrB, ΔpmrAΔqseB, ΔpmrAΔqseC, ΔpmrBΔqseB, ΔpmrBΔqseC., ΔkdpD, ΔkdpE ΔkdpDΔqseB, ΔkdpDΔqseC, ΔkdpEΔqseB and ΔkdpEΔqseC. B. Differential expression using qPCR of colistin and PBM resistance genes arnB and eptA in the mutants EPEC O125ac:H6 WT, ΔqseB, ΔqseC and ΔpmrAΔqseC* p-value <0.05; ** <0.01; *** <0.001; **** <0.0001 via one-way ANOVA with Tukey multiple comparison test, n = 4. C. In the promoter region of the arn operon, binding motifs are found for both QseB (blue) and PmrA (orange). The -35 and -10 sites were predicted using BPROM [42], and a rpoD17 binding site was predicted within the PmrA binding site. The start (ATG, red) of arnB gene is also shown. D. Electrophoretic mobility shift assay (EMSA). To validate the technique a EBNA control provided by the kit is shown. Biotylinated DNA fragment containing the QseB binding site in the arn operon is found to shif with the presence of 3 μg of QseB-P, which is reversed by the presence of unlabeled QseB site. Finally unphosphorylated QseB is found to bind with lower affinity to the QseB binding site.

https://doi.org/10.1371/journal.ppat.1011345.g008

Discussion

In this study, we investigated the crosstalk between the three two-component systems QseBC, PmrAB and KdpDE in atypical Enteropathogenic E. coli strain O125ac:H6, and how this crosstalk affects motility and expression of relevant genes. KdpDE seems to have a role in motility independent of QseBC. Motility data of KdpD HK and KdpE RR mutants suggests that KdpD HK activates motility via an alternative response regulator to the cognate KdpE. Similar expression of flhC and fliA in the KdpD HK mutant to the WT suggests that KdpDE regulation of motility is likely through control of flagella late-regulatory genes. Unlike what was seen in EHEC [23], KdpDE does not seem to be involved in regulation of LEE in aEPEC.

We showed that inactivation of QseC HK results in lack of motility in aEPEC, as reported for EHEC, UPEC and other Gram-negative bacteria such as Salmonella enterica [20,29,35,43], which was consistent with previous studies suggesting that QseC was an activator of motility [44]. However, introduction of a PmrB HK deletion into ΔqseC background, resulted in similar motility to the WT strain, confirming the hypothesis suggested by Guckes et al [25] that QseB RR acts as repressor of motility and, in the absence of QseC HK, QseB is phosphorylated by the non-cognate PmrB HK.

In aEPEC, unlike what was reported in UPEC and EHEC, deletion of the response regulator qseB alone and in combination with ΔqseC, ΔpmrA and ΔpmrB leads to increased motility and expression of genes flhC, fliA, ler and recA. This confirms that the QseB response regulator acts as repressor of multiple genes. Introduction of a non-phosphorylatable QseB variant (QseBD51A) in aEPEC maintained WT motility phenotypes indicating that basal repression exerted by QseB RR is likely due to unphosphorylated QseB binding with low affinity to QseB targets, leading to weak repression of these genes.

Both ΔqseC and ΔpmrAΔqseC show lack of motility and downregulation of flhC, fliA and recA, together with high upregulation of qseB expression, likely responsible for these phenotypes. Over-expression of qseB is due to phosphorylated QseB autoregulating its own expression. Our hypothesis is that, in the WT, QseB can be phosphorylated by the non-cognate PmrB HK, but it is immediately de-phosphorylated or sequestered by QseC [25,28] (Fig 9A). In the absence of QseC HK, phosphorylated QseB induces its own expression and an excess of activated QseB in the cell in turn represses multiple genes involved in motility, stress response and probably more unidentified genes (Fig 9B). The introduction of QseBD51A in a ΔqseC background leads to WT phenotypes, suggesting that phosphorylation of QseB in the absence of QseC is responsible for the aberrant phenotypes. The introduction of PmrBH152A in the ΔqseC background also leads to WT phenotypes, suggesting that this phosphorylation occurs through PmrB. Further in vitro protein phosphorylation and de-phosphorylation studies with the aEPEC protein variants of PmrAB and QseCB would help to confirm this hypothesis. It is unknown if the positive feedback loop caused by QseB phosphorylation aims to increase repression via QseB, or to increase qseC expression that in turn would deactivate the cognate QseB. In the ΔpmrBΔqseC mutant, QseB cannot be phosphorylated leading to a recovery of WT phenotypes.

thumbnail
Download:
Fig 9. Proposed regulation of the TCS KdpDE, PmrAB and QseBC in EPEC.

A. In WT aEPEC, KdpD HK phosphorylates an unknown effector to activate flagella late-regulatory genes. PmrB activates its cognate RR PmrA, and the non-cognate QseB. QseC HK controls the level of activated QseB via de-phosphorylation and sequestration of activated QseB. Unphosphorylated QseB in EPEC WT is able to exert a basal repression of flhC, recA and ler genes. B. In ΔpmrAΔqseC mutant, QseB is phosphorylated via PmrB, and cannot be deactivated by QseC. P~QseB activates the expression of qseBC operon leading to QseB over-expression. In arn operon, PmrA usually blocks transcription, but in its absence phosphorylated QseB leads to activation of colistin resistance genes. Excess of phosphorylated QseB tightly represses flhC and recA genes. Created with Biorender.com.

https://doi.org/10.1371/journal.ppat.1011345.g009

Interestingly, ler regulator is upregulated in the ΔqseB mutant in LB media but ler expression is similar to the WT in the ΔqseC mutant. However, when the experiments were performed in low glucose DMEM all the mutants showed a similar expression to the WT. We confirmed this data by qPCR of two genes involved in adhesion to mammalian cells, ler and espA. We suggest that QseB is only able to repress ler expression in certain conditions. The tight regulation of LEE island by multiple factors may explain why in DMEM low glucose QseBC plays a minor role in regulation. Using the FIMO tool [45] from MEME Suite [46] with the binding motifs of PmrA and QseB we found that in the E. coli O125ac:H6 genome 16 genes harbour both binding sites in the promoter region including flhCD, arnB, ler and qseB. This indicates that these regulators directly regulate ler and flhCD expression as showed before in literature [24]. Other regulated genes such as fliA or recA do not contain the motifs suggesting indirect regulation. In fact, fliA is probably upregulated through the master regulator of flagella flhCD [47].

In a further confirmation of these TCSs interaction, we saw that the ΔpmrAΔqseC mutant is 16-times more resistant to polymyxin B and colistin than the WT due to constitutive upregulation of colistin resistance regulon. This increase in resistance was not seen in any of the other mutants. The presence of binding sites for both QseB and PmrA in arn operon suggests that both play a role in regulation. Direct regulation was confirmed for QseB using an EMSA assay. We suggest that PmrA binds to the arn promoter, blocking binding of the RNA polymerase and the initiation of the transcription (Fig 9A). Only when PmrA is not bound, and QseB is upregulated and activated via PmrB, the arn operon is expressed. (Fig 9B). Taken together, all these findings suggest that PmrA and QseB act together to tightly regulate certain genes and QseB can act as activator or repressor depending on the target gene.

We did not find aEPEC to respond with higher motility to adrenergic signals, such as epinephrine, nor the bacterial signalling molecule AI-3. However, we did find that, in presence of excess Fe3+, strains containing mutations in either ΔpmrA or ΔpmrB show significantly lower motility than the WT. This data indicates that in absence of PmrAB, bacteria cannot respond correctly to excess iron. In fact, PmrAB has been linked to adaptation to excess environmental iron [27] and mutations in PmrAB have also been shown to increase sensitivity to low pH in excess iron [48]. In Uropathogenic Escherichia coli it has been shown that the iron homeostasis protein Fur leads to motility repression [49]. Thus it is possible that absence of PmrAB prevents iron homeostasis, leading to an excess intracellular iron and an effect of Fur on motility. When the mutations ΔpmrA or ΔpmrB were combined with either ΔqseB or ΔqseC, relative motility was lower than the non-treated control, but we still observed higher motility in ΔpmrAΔqseB and ΔpmrBΔqseB than in the single mutants ΔpmrA and ΔpmrB. This suggests that the adaptation of motility to excess iron via PmrAB is not related to the QseBC regulation of motility and that PmrAB has an independent role in iron homeostasis.

PmrAB and QseBC are phylogenetically very close to each other, with a 45.83% identity between RRs, and 32.61% identity between the cytoplasmic domains of HKs. In the evolution of these two TCS the sensing domains have diverged (no homology found), such as they sense different stimuli, but the protein sequences involved in the cross-talk between the two HK cytoplasmic domains and RRs have been maintained. PmrAB and QseBC are also present in other enteric Gram-negative bacteria with high degree of identity to the proteins in aEPEC (80–98%), such as in UPEC, EHEC, Salmonella enterica, Shigella and Klebsiella pneumoniae, indicating their importance. Percentages of identity between QseBC and PmrAB from these bacteria is similar to the one found in aEPEC, and it is possible that cross-regulation occurs in other bacteria than E. coli. Point mutations in PmrAB TCS in Klebsiella are associated to resistance to colistin [50,51]. It is should be investigated if these point mutations lower affinity of PmrA to promoters of genes involved in colistin resistance, thereby allowing QseB to upregulate colistin resistance. For other Gram-negative bacteria, such as Acinetobacter and Pseudomonas only the presence of PmrAB TCS has been reported. Here, we demonstrated that regulation in aEPEC and UPEC is similar, however discrepancies have been reported with EHEC regulation found in literature [21,23]. PmrA and QseB regulators are identical in aEPEC and UPEC, and they belong to the B2 phylo-group, EHEC possesses some different amino acids (S2 Fig) and belongs to E phylo-group possibly accounting for the differences in regulation.

In conclusion we suggest that PmrAB and QseBC TCSs may have co-evolved to tightly regulate gene expression. It is unknown in which conditions QseB repression is activated. It is possible that the bacteria uses PmrA and QseB interactions as internal mechanism to regulate expression in a time dependent manner. It is also possible that QseBC signalling has evolved to activate QseB repression only in certain situations where decreased flagella production and stress response are beneficial for the bacteria. Further research is needed to fully understand QseBC and PmrAB regulation. This research also highlights the importance of selecting specific TCS as targets for development of anti-virulence compounds against Gram-negative bacteria.

Material and methods

Bacterial strains and growth conditions

Atypical Enteropathogenic Escherichia coli O125ac:H6 (DSM8711) was purchased from the Leibniz Institute DSMZ and cultured in Luria-Bertani (LB) medium (Merck Millipore) and when necessary supplemented with kanamycin (50 μg/ml) or spectinomycin (100 μg/ml), and incubated at 37°C. Chemically competent Escherichia coli TOP10 (Invitrogen) was used for cloning. The whole genome sequence of E. coli O125ac:H6 was obtained using Illumina-sequencing (MicrobesNG https://www.microbesng.com). Accession number PRJNA930679.

Construction of deletion mutants, site-directed mutagenesis and mutant complementation using CRISPR-Cas

For the construction of CRISPR-Cas deletion mutants in EPEC O125ac:H6 we used the two-plasmid system developed by Jiang et al [33]. First, guides targeting the gene of interests were designed using Benchling software. To introduce the guides in pTargetF, we amplified pTargetF using primer pBM004 that contains a XhoI restriction site, and guide primers (pBM005, pBM013, pBM021, pBM029, pBM037, pBM045) that anneal with the plasmid just after the 20-bp CRISPR guide complementary to the genome. These primers include a SpeI recognition site and the new guide as an overhang. All plasmids and primers used can be found in S3 and S4 Tables. Both pTargetF and amplification product were cut using SpeI FD and XhoI FD (Molecular biology, ThermoScientific), cleaned and ligated using T4 DNA ligase (Promega). Ligation was transformed into chemically competent TOP10 E. coli and plated on LB agar containing spectinomycin. Plasmid from colonies growing in spectinomycin was extracted using QIAprep spin miniprep kit (QIAGEN) and introduction of the new guide was confirmed via Sanger sequencing, using primer pBM003 and GATC LightRun tubes services (Eurofins Genomics, Ebersberg, Germany).

The PCR product containing the repair template was constructed using PCR SOEing [52] to generate an amplicon with 500-bp regions of homology upstream and downstream of the region to be deleted. The repair template was designed to retain several amino-acids of the deleted gene resulting in an in-frame deletion. The homologous fragments were amplified by PCR and then purified using Invisorb Fragment CleanUp (Invitek, Berlin, Germany). The homologous repair template was generated by SOEing PCR using 50 ng of each homologous region as template and external primers for amplification and was purified using Invisorb Fragment CleanUp kit.

To complement the mutations ΔqseB and ΔpmrBΔqseC, we used CRISPR-Cas9 to introduce the WT gene (amplified using primers pBM007+pBM010 for qseB, pBM007+pBM010 for qseC and pBM007+pBM010 for pmrB). Guides used for complementation can be found in S4 Table. Finally, to construct the non-phosphorylatable mutants we used CRISPR-Cas9 to change the conserved aspartic acid 51 to alanine in QseB and the conserved histidine 152 to alanine in PmrB. To do this a CRISPR guide containing the amino-acid of interest was selected. Construction of the templates for the non-phosphorylatable variant was performed using primers pBM0072+pBM0073 and pBM075+pBM0076 that contain an overhang correspondent to the selected guide where silent mutations were introduced in the codons to make it different enough from the guide for the Cas9 not to cut after homologous recombination.

To construct the mutants, electrocompetent cells of EPEC O125ac:H6 were prepared as reported by Jiang et al [33], and transformed by electroporation with 100 ng of the temperature sensitive plasmid pCas (0.2 cm-gap cuvettes (Bio-Rad), in a Gene pulser Xcell electroporation system (Bio-Rad)). After electroporation 1 ml SOC media (Invitrogen) was added to the cuvettes and transferred to 1.5 ml Eppendorfs. Cells were left to recover at 30°C for 3 hours at shaking 180 r.p.m. and then plated in LB kanamycin (50 μg/ml) plates and incubated at 30°C. Colonies containing pCas were selected by colony PCR using the primers pBM053 and pBM054.

Electrocompetent cells of EPEC O125ac:H6 pCas were prepared as previously reported with slight modifications [33]. Cells were grown in 50 ml tubes with LB and 50 μg/ml kanamycin at 30°C with shaking at 200 r.p.m. to OD600 0.1, at which point arabinose was added to a final concentration of 10 mM to induce the λ-red recombination system. After induction, bacteria were grown to OD600 0.5, pelleted by centrifugation at 4°C, washed twice with cold sterile demi-water and stored in 200 μl aliquots cold 12% glycerol. 50 μl aliquots of electrocompetent cells were incubated for 30 minutes with 100 ng pTargetF with the guide of interest and 500 ng of template and electroporated. After 3-hour incubation at 30°C cells were plated in LB plates with kanamycin and spectinomycin and incubated at 30°C. Deletion mutants were selected by colony PCR using primer annealing outside the designed template.

To curate the mutants of pTargetF, the bacteria were grown in LB medium containing kanamycin and 0.5 mM IPTG at 30°C to induce expression of a pMB1 guide present in pCas that targets pTargetF plasmid. Loss of pTargetF was confirmed by streaking the culture in kanamycin and spectinomycin plates. After elimination of pTargetF, pCas was cured by growing the mutants at 37°C, the non-permissive temperature for replication. Loss of pCas was confirmed by plating of the cultures in kanamycin plates. The PCR to confirm gene deletion was then repeated.

Growth curve

Growth curves of the different mutants were performed by adding 200 μl of a 1:100 dilution into 96-well plates. Condensation was avoided by adding 1 ml anti-condense agent (0.05% Triton X-100 in 20% ethanol) to the lid of the plate and leaving it to dry. Incubation at 37°C was performed in SpectraMax M5 (Molecular Devices LLC, San Jose, CA, USA) and OD600 measured every 30 minutes for 15 hours. Experiments were repeated four independent times.

Motility assays

Motility assays were performed using a soft agar plate motility assay. Briefly, motility plates were made by pouring 20 ml of LB soft agar (LB+0.3% agar (Alfa Aesar, Haverhill, MA, USA)) into 90x16 mm Petri dishes (VWR, Radnor, PA, USA). When necessary, agar was pre-mixed with 500 nM, 5 μM, 50 μM or 500 μM epinephrine (Sigma-Aldrich), AI-3 or FeCl3. Experiments with epinephrine were protected from the light at all times to prevent oxidation. After solidification, plates were left to dry for 15 minutes. Single colonies from the mutant of interest were picked with a 10 μl pipette tip and then poked into the motility plate. Plates were incubated for 18 hours at 37°C, and motility measured in cm with a ruler. We assayed the motility of two mutants and the WT per motility plate. Motility of each mutant was calculated as the relative motility compared to the WT on the same plate. WT variability was accounted by measuring the relative motility of the WT in each plate compared to the average motility of the WT on different plates from the same experiment. Pictures of the motility plates were made using GelDoc XR+ gel documentation system (Bio-Rad). Experiments were performed 8 times.

AI-3 synthesis

Fig 10 shows the synthetic scheme of autoinducer AI-3. Starting material (3-chloro-2,5-dimethylpyrazine) was purchased from commercial vendor (Enamine, Frankfurt, Germany) and used without additional purification.

thumbnail
Download:
Fig 10. Synthetic scheme of autoinucer AI-3.

https://doi.org/10.1371/journal.ppat.1011345.g010

Construction of luciferase reporter plasmids

For the construction of the luciferase reporter plasmid, first we amplified the luxAB luciferase genes from Photorhabdus luminescens from pXen-1 plasmid [53] using the primers pBM055+pBM056 that contain a SpeI and EcoRI site overhang respectively. This was cloned between the SpeI and EcoRI sites of pTargetF so that luxAB was under the control of the constitutive promoter J23119 that in pTargetF is flanked by BamHI and SpeI sites, creating in this way pTarget_lux plasmid.

The promoters located upstream the genes of interest flhC, fliA, bla, ler and recA were amplified by PCR (≈1 kb fragments) using the genomic DNA of EPEC O125ac:H6 as template, except for bla that was amplified form plasmid pXen-1. The ampicillin resistance gene bla gene was selected as a promoter control. To assure correct positioning of the promoter fragments the primers were designed to include a few amino-acids from the gene of interest that would be in-frame with the luciferase gene after annealing. The primers (pBM059 to pBM068) contain BamHI and SpeI overhangs to facilitate replacement of the constitutive promoter in pTarget_lux. After cloning the promoter fragments into pTarget_lux, the insertion of the promoter was checked by colony PCR using the primers pBM057 and pBM058.

Differential expression using luciferase assay

Mutants were transformed by electroporation with the luciferase reporter plasmids: pTargeF_flhC, pTargetF_fliA, pTargetF_bla. pTargetF_ler, pTargetF_recA. For the differential expression assay, 200 μl of 1:100 dilution in LB+spectinomycin or low glucose DMEM+spectinomycin for pTargetF_ler (GIBCO) of the mutants with the reporter plasmid were transferred to a clear bottom 96-well plate (Corning Incorporated). To avoid cross-luminescence the wells around each culture were left empty. As substrate for the luciferase, 50 μl of 1% nonanal (Sigma-Aldrich) in mineral oil (Sigma-Aldrich) [54] was added to all the gaps between the wells. Evaporation was avoided by spreading 1 ml of anti-condense agent to the lid and letting it dry. Plates were incubated at 37°C in a SpectraMax M5 reader, and luminescence was measured every 30 minutes for 18 hours. Experiments were repeated in quadruplicate.

The peak-luminescence (highest luminescence found during the 18 hours) was recorded. Per promoter, relative luminescence in each experiment was calculated relative to the average of the WT. WT variability was calculated as the luminescence of individual WT measurements relative to the average luminescence for the WT in the experiment.

Differential expression using qPCR

EPEC O125ac:H6 WT, EPEC O125ac:H6 ΔqseB, EPEC O125ac:H6 ΔqseC and EPEC O125ac:H6 ΔqseCΔpmrA were grown in 6 ml of LB media to OD600 0.4 or 10 ml low glucose DMEM to OD600 0.2. Cultures were centrifuged at 4°C, 4.000 r.p.m. for 10 minutes, supernatant discarded, and pellets frozen using liquid nitrogen. RNA extraction from the pellets was performed using RNeasy Mini Kit (QIAGEN) following manufacturer´s manual. RNA concentration was measured using Qubit RNA broad range kit (Invitrogen) in a Qubit 4 fluorometer (Invitrogen). QuantiTect reverse transcription kit (QIAGEN) was used for cDNA synthesis from 1 μg RNA. Quantitative PCR was performed using GoTaq qPCR Master Mix (Promega) and primers shown in S5 Table. gyrA was chosen as control gene. Differential expression was calculated using the 2-ΔΔCt method [55]. Experiments were repeated 4 times.

Minimal inhibitory concentration (MIC) assay

Minimal inhibitory concentrations against EPEC O125ac:H6 were performed as described by European Committee on Antimicrobial Susceptibility Testing [56]. Briefly, double dilutions of colistin (Duchefa Biochemie, Haarlem, The Netherlands) and polymyxin B (Sigma-Aldrich) were performed in Mueller Hinton broth (MHB) horizontally in a 96-well plate over a range of concentrations 16–0.015 μg/ml in a final volume 100 μl. A 1:100 dilution of the desired mutant was performed and 100 μl added to the plate for a final volume of 200 μl. Plates were incubated overnight at 37°C. The next day, OD600 was measured using SpextraMax M5 and MIC recorded as the minimal concentration preventing any visible bacterial growth.

Electrophoretic mobility shift assay (EMSA)

The QseB response regulator cloned in pNIC28-Bsa4 [57] plasmid containing His6-tag was expressed and purified in E. coli BL21(DE3) using His-affinity (GE Healthcare) and size exclusion chromatography. EMSA assays were performed using the LightShift Chemiluminescent EMSA kit (Thermo Scientific). First, 20 μg of QseB were phosphorylated by addition of 0.1 M of lithium potassium acetyl phosphate (Sigma-Aldrich) in 50 mM TrisCl, 100 mM KCl and 10 mM MgCl2 and incubated at room temperature for 45 minutes. Biotinylated primers in the 5’ region from the QseB-binding site in the arn promoter region were purchased from Integrated DNA Technologies (Leuven, Belgium) (Fw-/5BiosG/ATTTACGTCTAATATTAGTTTCTTAAGGTTAAGTTAATATTCTATCCTTAAAATTTTGCT; Rv-/5BiosG/AGCAAAATTTTAAGGATAGAATATTAACTTAACCTTAAGAAACTAATA TTAGACGTAAAT). Oligos were annealed by mixing equimolar concentrations in TE buffer (10 mM Tris-Cl, 1 mM EDTA) and heating to 95°C followed by cooling at 1°C per second in a thermocycler. Binding reactions were performed by adding 0. 1 or 3 μg of phosphorylated or non-phosphorylated QseB, and 20 fmol of the biotinylated QseB-site region. Controls provided by the kit and a control with non-biotylinated QseB site (200 pmol) were included in the assays. Binding reactions were loaded with the provided loading dye in 8% polyacrylamide in 0.5X TBE (Thermofisher) gels. A mini trans-blot cell (Bio-Rad) was used to transfer proteins and DNA to Biodyne B positively charged nylon membranes (ThermoFisher) for 45 minutes at 100 V. Cross-link was performed by exposing the membranes to UV light for 15 minutes in a GelDoc XR+ (Bio-Rad). Blocking, incubation with streptavidin-horseradish peroxidase conjugate, washing and revealing were performed according to the protocol. Membranes were imaged using ChemiDoc XRS+ (Bio-Rad).

Statistical analysis

Growth curves were fitted to a logistic growth and the rate of growth was calculated per each curve (n = 3 per mutant). A one-way ANOVA with Tukey’s multiple comparison test was performed to test significance. Motility (n = 8) and gene expression data (n = 4) were first confirm to be normal using a Shapiro-Wilk test. Significance was tested using a one-way ANOVA with Tukey’s multiple comparison test. When data was not normal a Kurksal-Wallis test with Dunn’s multiple comparison was used (adjusted p-value<0.05 considered significant).Statistical analysis was performed using Prism 9 (Graphpad Software, San Diego, USA).

Supporting information

S1 Table. Growth rate of WT and Escherichia coli O125ac:H6 mutants together with standard deviation (SD).

Growth rate was calculated by fitting growth curves into a logistic growth equation (n = 3).

https://doi.org/10.1371/journal.ppat.1011345.s001

(PDF)

S2 Table. Raw luminescence data expressed in RIL x100.

Individual values of four replicates are represented in the peak-luminescence of each strain and each gene.

https://doi.org/10.1371/journal.ppat.1011345.s002

(PDF)

S3 Table. Plasmids used in this study.

https://doi.org/10.1371/journal.ppat.1011345.s003

(PDF)

S4 Table. Primers used in this study for cloning.

https://doi.org/10.1371/journal.ppat.1011345.s004

(PDF)

S5 Table. Primers used in this study for qPCR.

https://doi.org/10.1371/journal.ppat.1011345.s005

(PDF)

S1 Fig. Differential expression of flhC, fliA, ler, recA and bla in different EPEC O125ac:H6 QseBC, PmrAB and KdpDE mutants.

Differential expression according to the luciferase assay of the genes flhC, fliA, ler in LB or low glucose DMEM and recA of all tested mutants is depicted (A, B, C, D, E and F). * p-value <0.05; ** <0.01; *** <0.001; **** <0.0001 via one-way ANOVA with multiple comparisons, n = 4.

https://doi.org/10.1371/journal.ppat.1011345.s006

(TIF)

S2 Fig. QseB and PmrA alignment between atypical Enteropathogenic Escherichia coli O125ac:H6, Uropathogenic Escherichia coli UTI89 and Enterohemorrhagic Escherihica coli O157:H7.

Light blue shows identical amino-acids, dark blue shows changes in the amino-acid sequence.

https://doi.org/10.1371/journal.ppat.1011345.s007

(TIF)

References

  1. 1. Ochoa TJ, Contreras CA. Enteropathogenic Escherichia coli infection in children. Curr Opin Infect Dis. 2011 Oct;24(5):478–83. pmid:21857511
  2. 2. Spina A, Kerr KG, Cormican M, Barbut F, Eigentler A, Zerva L, et al. Spectrum of enteropathogens detected by the FilmArray GI Panel in a multicentre study of community-acquired gastroenteritis. Clinical Microbiology and Infection. 2015 Aug;21(8):719–28. pmid:25908431
  3. 3. García-Meniño I, García V, Mora A, Díaz-Jiménez D, Flament-Simon SC, Alonso MP, et al. Swine enteric colibacillosis in Spain: pathogenic potential of mcr-1 ST10 and ST131 E. coli isolates. Front Microbiol. 2018 Nov 5;9.
  4. 4. Watson VE, Jacob ME, Flowers JR, Strong SJ, DebRoy C, Gookin JL. Association of atypical Enteropathogenic Escherichia coli with diarrhea and related mortality in kittens. J Clin Microbiol. 2017 Sep;55(9):2719–35.
  5. 5. Kjaergaard AB, Carr AP, Gaunt MC. Enteropathogenic Escherichia coli (EPEC) infection in association with acute gastroenteritis in 7 dogs from Saskatchewan. Can Vet J. 2016 Sep;57(9):964–8. pmid:27587889
  6. 6. Arais LR, Barbosa A v, Andrade JRC, Gomes TAT, Asensi MD, Aires CAM, et al. Zoonotic potential of atypical Enteropathogenic Escherichia coli (aEPEC) isolated from puppies with diarrhoea in Brazil. Vet Microbiol. 2018 Dec;227:45–51. pmid:30473351
  7. 7. Trabulsi LR, Keller R, Gomes TAT. Typical and atypical Enteropathogenic Escherichia coli. Emerg Infect Dis. 2002 May;8(5):508–13.
  8. 8. Girón JA, Ho ASY, Schoolnik GK. An inducible bundle-forming pilus of Enteropathogenic Escherichia coli. Science (1979). 1991 Nov;254(5032):710–3.
  9. 9. Hernandes RT, Elias WP, Vieira MAM, Gomes TAT. An overview of atypical Enteropathogenic Escherichia coli. FEMS Microbiol Lett. 2009 Aug;297(2):137–49.
  10. 10. Ochoa TJ, Barletta F, Contreras C, Mercado E. New insights into the epidemiology of Enteropathogenic Escherichia coli infection. Trans R Soc Trop Med Hyg. 2008 Sep;102(9):852–6.
  11. 11. Nguyen RN, Taylor LS, Tauschek M, Robins-Browne RM. Atypical Enteropathogenic Escherichia coli infection and prolonged diarrhea in children. Emerg Infect Dis. 2006 Apr;12(4):597–603.
  12. 12. Afset JE, Bevanger L, Romundstad P, Bergh K. Association of atypical Enteropathogenic Escherichia coli (EPEC) with prolonged diarrhoea. J Med Microbiol. 2004 Nov 1;53(11):1137–44.
  13. 13. Hartland EL, Leong JM. Enteropathogenic and Enterohemorrhagic E. coli: ecology, pathogenesis, and evolution. Front Cell Infect Microbiol. 2013;3.
  14. 14. Rangel JM, Sparling PH, Crowe C, Griffin PM, Swerdlow DL. Epidemiology of Escherichia coli O157:H7 outbreaks, United States, 1982–2002. Emerg Infect Dis. 2005 Apr;11(4):603–9.
  15. 15. Hedican EB, Medus C, Besser JM, Juni BA, Koziol B, Taylor C, et al. Characteristics of O157 versus non-O157 Shiga toxin–producing Escherichia coli infections in Minnesota, 2000–2006. Clinical Infectious Diseases. 2009 Aug;49(3):358–64.
  16. 16. Elliott SJ, Wainwright LA, McDaniel TK, Jarvis KG, Deng Y, Lai LC, et al. The complete sequence of the locus of enterocyte effacement (LEE) from Enteropathogenic Escherichia coli E2348/69. Mol Microbiol. 2002 May 1;28(1):1–4.
  17. 17. Moon HW, Whipp SC, Argenzio RA, Levine MM, Giannella RA. Attaching and effacing activities of rabbit and human Enteropathogenic Escherichia coli in pig and rabbit intestines. Infect Immun. 1983 Sep;41(3):1340–51.
  18. 18. Wong ARC, Pearson JS, Bright MD, Munera D, Robinson KS, Lee SF, et al. Enteropathogenic and Enterohaemorrhagic Escherichia coli: even more subversive elements. Mol Microbiol. 2011 Jun;80(6):1420–38.
  19. 19. Weigel WA, Demuth DR. QseBC, a two-component bacterial adrenergic receptor and global regulator of virulence in Enterobacteriaceae and Pasteurellaceae. Mol Oral Microbiol. 2016 Oct 20;31(5):379–97.
  20. 20. Sperandio V, Torres AG, Kaper JB. Quorum sensing Escherichia coli regulators B and C (QseBC): a novel two-component regulatory system involved in the regulation of flagella and motility by quorum sensing in E. coli. Mol Microbiol. 2002 Feb;43(3):809–21.
  21. 21. Clarke MB, Hughes DT, Zhu C, Boedeker EC, Sperandio V. The QseC sensor kinase: a bacterial adrenergic receptor. Proceedings of the National Academy of Sciences. 2006 Jul 5;103(27):10420–5. pmid:16803956
  22. 22. Rasko DA, Moreira CG, Li DR, Reading NC, Ritchie JM, Waldor MK, et al. Targeting QseC signaling and virulence for antibiotic development. Science (1979). 2008 Aug 22;321(5892):1078–80. pmid:18719281
  23. 23. Hughes DT, Clarke MB, Yamamoto K, Rasko DA, Sperandio V. The QseC adrenergic signaling cascade in enterohemorrhagic E. coli (EHEC). PLoS Pathog. 2009 Aug 21;5(8):e1000553.
  24. 24. Clarke MB, Sperandio V. Transcriptional regulation of flhDC by QseBC and σ28 (FliA) in Enterohaemorrhagic Escherichia coli. Mol Microbiol. 2005 Sep;57(6):1734–49.
  25. 25. Guckes KR, Kostakioti M, Breland EJ, Gu AP, Shaffer CL, Martinez CR, et al. Strong cross-system interactions drive the activation of the QseB response regulator in the absence of its cognate sensor. Proceedings of the National Academy of Sciences. 2013 Oct 8;110(41):16592–7. pmid:24062463
  26. 26. Wösten MMSM Kox LFF, Chamnongpol S Soncini FC, Groisman EA. A signal transduction system that responds to extracellular iron. Cell. 2000 Sep;103(1):113–25.
  27. 27. Kato A, Chen HD, Latifi T, Groisman EA. Reciprocal control between a bacterium’s regulatory system and the modification status of its lipopolysaccharide. Mol Cell. 2012 Sep;47(6):897–908. pmid:22921935
  28. 28. Breland EJ, Zhang EW, Bermudez T, Martinez CR, Hadjifrangiskou M. The histidine residue of QseC is required for canonical signaling between QseB and PmrB in Uropathogenic Escherichia coli. J Bacteriol. 2017 Sep 15;199(18).
  29. 29. Curtis MM, Russell R, Moreira CG, Adebesin AM, Wang C, Williams NS, et al. QseC Inhibitors as an antivirulence approach for Gram-negative pathogens. mBio. 2014 Dec 31;5(6). pmid:25389178
  30. 30. Theuretzbacher U, Piddock LJV. Non-traditional antibacterial therapeutic options and challenges. Cell Host Microbe. 2019 Jul 10;26(1):61–72. pmid:31295426
  31. 31. Barros SF, Abe CM, Rocha SPD, Ruiz RM, Beutin L, Trabulsi LR, et al. Escherichia coli O125ac:H6 encompasses atypical Enteropathogenic E. coli strains that display the aggregative adherence pattern. J Clin Microbiol. 2008 Dec;46(12):4052–5.
  32. 32. Hobson N, Price NL, Ward JD, Raivio TL. Generation of a restriction minus enteropathogenic Escherichia coli E2348/69 strain that is efficiently transformed with large, low copy plasmids. BMC Microbiol. 2008 Dec 5;8(1):134.
  33. 33. Jiang Y, Chen B, Duan C, Sun B, Yang J, Yang S. Multigene editing in the Escherichia coli genome via the CRISPR-Cas9 system. Appl Environ Microbiol. 2015 Apr;81(7):2506–14.
  34. 34. Bansal T, Englert D, Lee J, Hegde M, Wood TK, Jayaraman A. Differential effects of epinephrine, norepinephrine, and indole on Escherichia coli O157:H7 chemotaxis, colonization, and gene expression. Infect Immun. 2007 Sep;75(9):4597–607.
  35. 35. Bearson BL, Bearson SMD. The role of the QseC quorum-sensing sensor kinase in colonization and norepinephrine-enhanced motility of Salmonella enterica serovar Typhimurium. Microb Pathog. 2008 Apr;44(4):271–8.
  36. 36. Kim CS, Gatsios A, Cuesta S, Lam YC, Wei Z, Chen H, et al. Characterization of autoinducer-3 structure and biosynthesis in E. coli. ACS Cent Sci. 2020 Feb 26;6(2):197–206.
  37. 37. Elliott SJ, Sperandio V, Girón JA, Shin S, Mellies JL, Wainwright L, et al. The locus of enterocyte effacement (LEE)-encoded regulator controls expression of both LEE- and non-LEE-encoded virulence factors in enteropathogenic and enterohemorrhagic Escherichia coli. Infect Immun. 2000 Nov;68(11):6115–26.
  38. 38. Olaitan AO, Morand S, Rolain JM. Mechanisms of polymyxin resistance: acquired and intrinsic resistance in bacteria. Front Microbiol. 2014 Nov 26;5. pmid:25505462
  39. 39. Chen HD, Groisman EA. The biology of the PmrA/PmrB two-component system: the major regulator of lipopolysaccharide modifications. Annu Rev Microbiol. 2013 Sep 8;67(1):83–112. pmid:23799815
  40. 40. Tamayo R, Prouty AM, Gunn JS. Identification and functional analysis of Salmonella enterica serovar Typhimurium PmrA-regulated genes. FEMS Immunol Med Microbiol. 2005 Feb;43(2):249–58.
  41. 41. Tierrafría VH, Rioualen C, Salgado H, Lara P, Gama-Castro S, Lally P, et al. RegulonDB 11.0: comprehensive high-throughput datasets on transcriptional regulation in Escherichia coli K-12. Microb Genom. 2022 May 18;8(5).
  42. 42. Solovyev V, Slamov A. Automatic annotation of microbial genomes and metagenomic sequences. In: Li RW, editor. Metagenomics and its applications in agriculture, biomedicine and environmental studies. New York: Nova Science Publishers; 2011. p. 61–78.
  43. 43. Kostakioti M, Hadjifrangiskou M, Pinkner JS, Hultgren SJ. QseC-mediated dephosphorylation of QseB is required for expression of genes associated with virulence in Uropathogenic Escherichia coli. Mol Microbiol. 2009 Sep;73(6):1020–31.
  44. 44. Clarke MB, Sperandio V. Transcriptional regulation of flhDC by QseBC and σ 28 (FliA) in enterohaemorrhagic Escherichia coli. Mol Microbiol. 2005 Sep;57(6):1734–49.
  45. 45. Grant CE, Bailey TL, Noble WS. FIMO: scanning for occurrences of a given motif. Bioinformatics. 2011 Apr 1;27(7):1017–8. pmid:21330290
  46. 46. Bailey TL, Johnson J, Grant CE, Noble WS. The MEME Suite. Nucleic Acids Res. 2015 Jul 1;43(W1):W39–49. pmid:25953851
  47. 47. Fitzgerald DM, Bonocora RP, Wade JT. Comprehensive mapping of the Escherichia coli flagellar regulatory network. PLoS Genet. 2014 Oct 2;10(10):e1004649.
  48. 48. Hagiwara D, Yamashino T, Mizuno T. A genome-wide view of the Escherichia coli BasS–BasR two-component system implicated in iron-responses. Biosci Biotechnol Biochem. 2004 Jan 22;68(8):1758–67.
  49. 49. Kurabayashi K, Agata T, Asano H, Tomita H, Hirakawa H. Fur represses adhesion to, invasion of, and intracellular bacterial community formation within bladder epithelial cells and motility in uropathogenic Escherichia coli>. Infect Immun. 2016 Nov;84(11):3220–31.
  50. 50. Zafer MM, El-Mahallawy HA, Abdulhak A, Amin MA, Al-Agamy MH, Radwan HH. Emergence of colistin resistance in multidrug-resistant Klebsiella pneumoniae and Escherichia coli strains isolated from cancer patients. Ann Clin Microbiol Antimicrob. 2019 Dec 12;18(1):40.
  51. 51. Azam M, Gaind R, Yadav G, Sharma A, Upmanyu K, Jain M, et al. Colistin resistance among multiple sequence types of Klebsiella pneumoniae is associated with diverse resistance mechanisms: a report from India. Front Microbiol. 2021 Feb 22;12.
  52. 52. Horton RM. PCR-mediated recombination and mutagenesis. Mol Biotechnol. 1995 Apr;3(2):93–9.
  53. 53. Francis KP, Joh D, Bellinger-Kawahara C, Hawkinson MJ, Purchio TF, Contag PR. Monitoring bioluminescent Staphylococcus aureus infections in living mice using a novel luxABCDE construct. Infect Immun. 2000 Jun;68(6):3594–600.
  54. 54. Bachmann H, Santos F, Kleerebezem M, van Hylckama Vlieg JET. Luciferase detection during stationary phase in Lactococcus lactis. Appl Environ Microbiol. 2007 Jul 15;73(14):4704–6.
  55. 55. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2−ΔΔCT method. Methods. 2001 Dec;25(4):402–8.
  56. 56. The European Committee on Antimicrobial Susceptibility Testing (EUCAST). Broth microdilution—EUCAST reading guide version 4.0 [Internet]. 2022 [cited 2023 Jan 26]. Available from: https://www.eucast.org/ast_of_bacteria/mic_determination.
  57. 57. Gileadi O, Burgess-Brown NA, Colebrook SM, Berridge G, Savitsky P, Smee CEA, et al. High throughput production of recombinant human proteins for crystallography. Methods Mol Biol. 2008;221–46. pmid:18542867