Regulation of Hfq mRNA and Protein Levels in Escherichia coli and Pseudomonas aeruginosa by the Burkholderia cenocepacia MtvR sRNA

Abstract

Small non-coding RNAs (sRNAs) are important players of gene expression regulation in bacterial pathogens. MtvR is a 136-nucleotide long sRNA previously identified in the human pathogen Burkholderia cenocepacia J2315 and with homologues restricted to bacteria of the Burkholderia cepacia complex. In this work we have investigated the effects of expressing MtvR in Escherichia coli and Pseudomonas aeruginosa. Results are presented showing that MtvR negatively regulates the hfq mRNA levels in both bacterial species. In the case of E. coli, this negative regulation is shown to involve binding of MtvR to the 5′-UTR region of the hfq _Ec mRNA. Results presented also show that expression of MtvR in E. coli and P. aeruginosa originates multiple phenotypes, including reduced resistance to selected stresses, biofilm formation ability, and increased susceptibility to various antibiotics.

Introduction

Small non-coding RNAs (sRNAs) are increasingly recognized as important players in gene expression regulation in bacteria, and more recently, in the regulation of virulence determinants in bacterial pathogens [1]. Many of the sRNAs characterized so far exert their action by base-pairing with their target mRNAs in the region of the ribosome-binding site, thus affecting their stability and/or translation [2]. This mode of regulation has been shown to be advantageous for bacteria when a fast response to external stimuli is required, providing a fine tuning of gene expression [3]. This also enables the bacterial cell with a precise gene expression regulation and the means for a rapid adaptation in response to specific environmental changes [4]. These advantages are consistent with the involvement of many described bacterial sRNAs in response to stress [5]. In addition, sRNAs and the associated complex regulatory circuits are also used by bacterial pathogens to adapt to the host environment, and to coordinately express specific virulence determinants during different stages of infection [6].

Trans-encoded sRNAs are arguably the most extensively characterized class of bacterial sRNAs [6]. These sRNAs can target one or more mRNAs, sharing with them a limited complementarity. In order to interact with their targets, most of the trans-encoded sRNAs require the aid of the RNA chaperone Hfq, which plays a role as a facilitator of sRNA-mRNA interactions [7].

Most of the sRNAs recently described are only conserved among closely related bacterial species [8]. This is the case of MtvR, a 136 nucleotide-long sRNA identified in Burkholderia cenocepacia J2315, with homologues restricted to bacteria of the Burkholderia genus [9]. B. cenocepacia is a member of the so called Burkholderia cepacia complex (Bcc), a group of closely related species that emerged in the 1980s as important opportunistic pathogens among patients afflicted with the genetic disease cystic fibrosis [10], and more recently, among hospitalized non-cystic fibrosis patients, especially cancer patients (reviewed in [11]).

Previous work from our research group has shown that the sRNA MtvR targets multiple genes in the clinical isolate B. cenocepacia J2315, affecting cell size, growth and survival under nutrient deprivation, biofilm formation, antibiotics resistance, and virulence to the nematode Caenorhabditis elegans [9]. Due to the pleiotropic phenotypes observed, MtvR was proposed as a global regulatory sRNA in B. cenocepacia [9]. hfq was previously shown as one of the MtvR multiple targets in B. cenocepacia, with the sRNA negatively regulating hfq translation by specifically binding to its 5′ leader region [9]. In the present study, we investigated the possible roles played by MtvR in the non-Bcc organisms Escherichia coli and Pseudomonas aeruginosa, two bacterial species with no MtvR homologues, but harboring Hfq-encoding genes. Despite being an exogenous sRNA in both bacterial species, results here presented indicate that MtvR modulates the levels of the Hfq mRNA and protein in both species, affecting their resistance to selected stresses, including to antibiotics.

Materials and Methods

Bacterial Strains, Culture Conditions, Plasmids and Primers

Bacterial strains and plasmids are described in Table 1, and oligonucleotides are listed in Table 2. P. aeruginosa strains were maintained on PIA (Pseudomonas Isolation Agar, Becton Dickinson) plates, supplemented with 650 µg.ml⁻¹ trimethoprim, when appropriate. E. coli strains were maintained on Lennox Broth (LB) plates supplemented with 100 µg.ml⁻¹ trimethoprim or 50 µg.ml⁻¹ kanamycin or chloramphenicol, when appropriate. Unless otherwise stated, liquid cultures were carried out using LB liquid medium, supplemented with antibiotics when appropriate, with orbital agitation (250 rev min⁻¹) at 37°C. Expression of MtvR in E. coli or P. aeruginosa was achieved by induction of plasmid pCGR12 (Table 1) with 0.1% (v/v) L-arabinose. Plasmid pCGR34 was derived from pCR II by cloning into the XbaI-HindIII sites the E. coli MC4100 hfq (hfq _Ec) coding sequence, including the complete 5′ leader region and a 6×histidine tag encoded at the C-terminus. Plasmid pCGR35, containing the genetic fusion composed of the hfq _Ec leader -β-galactosidase, was engineered as follows: the LacZα fragment was amplified by PCR from pBBR1MCS and blunt-end ligated to the EcoRV site of pMLBAD. A 380 bp fragment containing the 5′-UTR region of the hfq _Ec gene was obtained by XbaI-BamHI restriction of a PCR fragment obtained using as template E. coli MC4100 chromosomal DNA. After restriction, the fragment was filled-in with the Klenow large fragment of DNA polymerase I (Invitrogen), and then blunt-ligated to the filled-in XbaI site of the cloned LacZα fragment. Plasmid pCGR36 derives from pCR II, and contains the hfq _Ec 5′ leader region together with the coding sequence in the XbaI-HindIII sites, allowing T7-driven transcription. Plasmid pCGR37 derives from pMLBAD, and contains the 255 nt fragment corresponding to the P. aeruginosa PA14 hfq (hfq _Pa, obtained by PCR using as template P. aeruginosa PA14 chromosomal DNA), cloned in the PstI-SalI sites (3′ to 5′ direction of the coding sequence). Plasmid pCGR38 derives from pMLBAD and contains the 140 bp cDNA fragment corresponding to the MtvR sRNA cloned in the EcoRI/XbaI sites and the 255 bp cDNA fragment corresponding to the P. aeruginosa PA14 hfq cloned in the SalI/PstI sites (tandem cloning, pBAD promoter control). All plasmid constructions were verified by DNA sequencing.

Table 1. Bacterial strains and plasmids used in this work.

Strain or plasmid	Description	Reference or source
P. aeruginosa WT	P. aeruginosa PA14, clinical isolate	[37]
P. aeruginosa WT-hfqsil	P. aeruginosa PA14 with hfq Pa silenced after transformationwith pCGR37	This study
P. aeruginosa WT/phfq	P. aeruginosa PA14 expressing heterologous hfq Bc aftertransformation with pSAS3	This study
P. aeruginosa WT/phfq+MtvR	P. aeruginosa PA14 expressing heterologous hfq Bc andMtvR after transformation with pSAS3 and pCGR12	This study
P. aeruginosa WT/pMLBAD	P. aeruginosa PA14 after transformation with pMLBAD	This study
P. aeruginosa WT/MtvR	P. aeruginosa PA14 expressing MtvR after transformationwith pCGR12	This study
P. aeruginosa WT-hfqsil/MtvR	P. aeruginosa PA14 with hfq Pa silenced and expressing MtvR,after transformation with pCGR38	This study
E. coli WT	E. coli MC4100, laboratory strain	[38]
E. coli Δhfq	E. coli GS081, CmR	[39]
E. coli WT/phfq	E. coli MC4100 after transformation with pCGR34, expressingtagged hfq Ec	This study
E. coli WT/phfq+MtvR	E. coli MC4100 after transformation with pCGR34 and pCGR12,expressing tagged hfq Ec and MtvR	This study
E. coli WT/pMLBAD	E. coli MC4100 after transformation with pMLBAD	This study
E. coli WT/MtvR	E. coli MC4100 after transformation with pCGR12,expressing MtvR	This study
E. coli Δhfq/pMLBAD	E. coli GS081 transformed with pMLBAD	This study
E. coli Δhfq/phfq	E. coli GS081 transformed with pSAS3, expressingheterologous hfq Bc	This study
E. coli Δhfq/MtvR	E. coli GS081 transformed with pCGR12, expressing MtvR	This study
E. coli Δhfq/phfq+MtvR	E. coli GS081 transformed with pCGR34 and pCGR12,expressing tagged hfq Ec and MtvR	This study
	Plasmids
pCR II	AmpR; KmR; used for generating in vitro transcription templates	Invitrogen
pMLBAD	TmpR; used for inducible gene expression	[40]
pCGR4	pET23a+ with the hfq Bc encoding sequence cloned	[14]
pSAS3	pMLBAD with the hfq Bc encoding sequence cloned	[14]
pCGR12	pMLBAD with the 140 bp cDNA fragment corresponding tothe MtvR sRNA cloned in the EcoRI/XbaI sites(pBAD promoter control)	[9]
pCGR34	pCR II with the 609 bp cDNA fragment corresponding tothe E. coli MC4100 hfq full mRNA (5′-UTR and CDS) with6 histidines at the C-terminus, cloned in the XbaI/HindIIIsites (T7 promoter control)	This study
pCGR35	pMLBAD with the hfq Ec 5′-UTR-LacZ DNA fragment (pBABpromoter disrupted, only replicative)	This study
pCGR36	pCR II with the DNA fragment corresponding to E. coli MC4100hfq cloned in the XbaI/HindIII sites	This study
pCGR37	pMLBAD with the 255 bp cDNA fragment corresponding to theP. aeruginosa PA14 hfq cloned in the PstI/SalI sites	This study
pCGR38	pMLBAD with the 140 bp cDNA fragment corresponding tothe MtvR sRNA cloned in the EcoRI/XbaI sites and the255 bp cDNA fragment corresponding to the P. aeruginosaPA14 hfq cloned in the SalI/PstI sites(tandem cloning, pBAD promoter control)	This study

Table 2. Oligonucleotides and primers used in this work.

Name	Purpose	Sequence 5′ –3′	Source or reference
UF	Cloning mtvR	TTTCTAGATATTGACGGCGGCGGGT	[9]
LF	Cloning mtvR	TTAAGCTTAAATTATAGCGCCCCAATTA	[9]
NP	Northern analysis of MtvR	CTATCACCCGCCTGTGTCGCCA	[9]
HFQ	Northern blot probe for hfq	AAAGGGCAATTGTTACAAG	[12]
CGRO117	Amplification of P. aeruginosa hfq	TTCTGCAGACCGGACGGCTCGGTACCAC	This study
CGRO118	Amplification of P. aeruginosa hfq	TCTGTCGACTTCCGGAGCGAGACCGGAGT	This study
CGRO119	Amplification of E. coli hfq	CCTCTAGACCAGAACAGGCGCGTGACGA	This study
CGRO120	Amplification of E. coli hfq	TTAAGCTTGAAACCGGGCGAGACGGGAC	This study
CGRO123	Fwd primer hfq his-tag	TTTCTAGAGCACGTCCCGCAAGGGCTAG	This study
CGRO124	Rev primer hfq his-tag	TTGGATCCATTGTGGTGGTGGTGGTGGGACGAGGCTTCCGC	This study
M13FWD	LacZ amplification	GTAAAACGACGGCCAGT	Invitrogen
M13REV	LacZ amplification	AGCGGATAACAATTTCACACAGGA	Invitrogen
CGRO125	Fwd primer uhpA	CTGGGGCTGGAACCTGATTT	This study
CGRO126	Rev primer uhpA	CGCAGCAATGAGTTCATCCG	This study
CGRO127	Fwd primer uhpT	TTCCTGCCGTTCATGCTGAT	This study
CGRO128	Rev primer uhpT	GAGGCCATAAGATTCCGGGG	This study
5S	Northern blot probe for 5S rRNA	TTCGGGATGGGAAGGGGTGGGA	[12]

DNA Manipulation Techniques

Total DNA was obtained from the indicated cell cultures using the High pure PCR template preparation kit (Roche). PCR amplification was performed using TaqPlatinum (Invitrogen) DNA polimerase and adequate primers and DNA templates (Table 2). PCR products were purified using the NucleoExtract II kit (Nagel-Machery) as previously described [9]. After nuclease restriction, fragments were directionally cloned into the indicated plasmids. All plasmid constructions were confirmed by DNA sequencing.

RNA Manipulation Techniques

Total RNA isolation, quantification of RNA concentration and visual quality control, and RNA labeling with biotin were performed as previously described [12].

Northern Blot Analysis

The expression levels of MtvR, hfq _Ec or hfq _Pa were assessed by Northern blot analysis using 2 µg of total RNA and adequate oligonucleotide probes (Table 2), previously labeled with biotin, based on previously described methods [13]. Briefly, RNA samples were separated in 8% TBE-urea pre-cast gels (BioRad), using constant current of 200 mA. RNA was then transferred to a BrightStar plus membrane (Ambion) using the wet transfer system (BioRad), at 100 V for 1h. Samples were probed with adequate biotinylated oligonucleotides at 40°C, for 16 h. Hybridization signals were detected using the BrightStar Biodetect kit (Ambion) and Kodak MX X-ray films. The 5S RNA was used as loading control in all Northern blot experiments. Relative expression was estimated with the ImageJ software suite, using the band intensities of the 5S RNA for normalization.

RNA Decay Experiments

The RNA decay rate was assessed by Northern blot analysis using 2 µg of total RNA, purified from cells of E. coli (Ec) or P. aeruginosa (Pa), harvested from 24 h-cultures, immediately before (t₀) or after the addition of rifampicin to induce transcriptional arrest. This antibiotic was used at final concentrations of 250 µg ml⁻¹. Aliquots taken after 5, 10, 15 and 20 min (for hfq), or 2, 5, 10 and 15 min (for MtvR) of transcription arrest, were processed and analyzed by Northern blot, as described above. RNA decay rates were calculated based on the exponential fit expression: t_1/2 (min): ae^−bt, and using the slope of semi-log plots.

RNA in vitro Transcription and Labeling

The DNA templates for in vitro transcription of the hfq _Ec mRNA full transcript, the 5′-UTR of hfq _Ec, the hfq _Ec coding region (CDS), and the MtvR sRNA were obtained by endonuclease restriction of the appropriate plasmids (Table 1). All RNA transcripts were generated from the T7 promoter, using the MEGAshortscript kit (Ambion). The transcripts MtvR (136 nt), 5′-UTR of hfq _Ec (155 nt), and the CDS of hfq _Ec (307 nt) were purified from 8%–7 M urea polyacrylamide gels, while the hfq _Ec full transcript (462 nt) was purified from a 1% TBE/agarose gel. RNA transcripts were processed and labeled based on previously described methods [13]. Signal intensity was tested using the dot-blot procedure, and detected using the Bright Star Biodetect Kit (Ambion).

Electrophoretic Mobility Shift Assays

EMSA experiments to assess the binding affinity of MtvR to the hfq _Ec 5′-UTR, the hfq _Ec coding region (CDS), and the hfq _Ec full mRNA, were performed as previously described [12]. Briefly, 25 nM of the MtvR, together with 0, 0.5, 1, 5, 10, 50 or 100 nM of the hfq _Ec full transcript, or with 0, 5, 10, 25, 50, 75, 100, 150, 200, 250 or 500 nM of the hfq _Ec CDS, or with 0, 0.5, 1, 5, 10, 50 or 100 nM the 5′-UTR of hfq _Ec, were incubated in 25 µl of RNA binding buffer (20 mM Tris.Cl, pH 8.0, 1 mM DTT, 1 mM MgCl₂, 10 mM KCl and 10 mM KH₂PO₄) [14] for 30 min at 25°C. The ability of 25 nM of the 5′-end labeled (with 11-UTP-biotin) hfq _Ec 5′-UTR to bind to 50, 100 or 250 nM of B. cenocepacia J2315 Hfq₆ (Hfq_Bc), of 100 nM of hfq _Ec full RNA to bind to Hfq_Bc (concentrations ranging from 10 to 1000 nM of Hfq_Bc), or to Hfq_Bc (concentrations ranging from 0.5 to 250 nM) in the presence of MtvR (concentrations ranging from 0.1 to 100 nM), were also evaluated by EMSA experiments using native 8% polyacrylamide gel containing 10% glycerol.

Non-labeled yeast tRNA (Ambion) was added in excess to each sample to minimize non-specific binding. Incubation, resolution of RNA-RNA and RNA-protein complexes, and detection of band-shifts, were performed as previously described [14]. The curves generated from the data plots were fitted by non-linear least squares regression assuming a bimolecular model in which the K _d values represent the protein concentration at half the maximal RNA binding.

Reverse Transcription-PCR Experiments

Total RNA was extracted from cells of exponentially growing cultures of E. coli WT strain, E. coli WT expressing MtvR, or E. coli Δhfq mutant, using the already described methods. Reverse transcription reactions were performed using the First Strand cDNA synthesis kit (Fermentas), with an incubation of 75 min at 45°C, in a total volume of 50 µL, and 500 ng of total RNA. The cDNA samples were used in PCR experiments with TaqMed (Citomed) DNA polymerase, and the oligonucleotide pairs CGRO125 and CGRO126 (uhpA), or CGRO127 and CGRO128 (uhpT) (Table 2). Cycling conditions were as follows: 45 cycles of denaturation at 95°C for 1 min, annealing at 55°C for 1 min, extension at 72°C for 0.5 min, and a final extension at 72°C for 10 min. cDNA pools were resolved in 2%-TBE agarose gels. Control reactions using DNA were included. 500 ng of total RNA from each strain were also loaded on the same gel, as loading controls.

Western Blot Analysis

The effect of MtvR expression on the levels of the Hfq_Ec protein was assessed by Western blotting. For this purpose, plasmid pCGR34 was introduced into the E. coli Δhfq mutant strain. Plasmid pCGR34 is able to drive the transcription of hfq _Ec as a derivative containing a 6×His-tag at the C-terminus (Hfq_Ec-His), from its own promoter. This strain was further transformed with plasmid pCGR12, which allows MtvR expression. The cultures were grown for 16 h in LB liquid medium, supplemented with 150 µg ml⁻¹ trimethoprim, 50 µg ml⁻¹ kanamycin and 0.1% L-arabinose. Total proteins from culture aliquots of 1 ml were purified with Illustra TriplePrep kit (GE Healthcare), based on previously described methods [13]. Aliquots containing 10 µg of total protein were used for Western blot detection of the Hfq_Ec-His protein, using the polyclonal pentaHis-IgG-HRP antibody (Invitrogen). The α-GroEL, used as loading control in Western-blot experiments, was detected with a goat anti-GroEL antibody (SicGen, Portugal). Signals were detected using a ECL system (GE Healthcare). Fold-change values were estimated using as unitary value the number of pixels counted for the reference, normalized with the internal standard.

β-galactosidase Assays

β-galactosidase assays were performed based on previously described methods [15] using a SpectraMax 250 microtiter plate reader (Molecular Devices). Briefly, E. coli strains WT, WT transformed with the plasmid that allows MtvR expression, or transformed with the control vector pMLBAD, were grown at 37°C in LB medium for 24 h with antibiotics (when appropriate) and diluted 1000-fold into 50 mL of fresh medium at 37°C. Cultures were grown with agitation to an OD₆₄₀ of 0.1 before inducing MtvR expression by the addition of L-arabinose (0.1% final concentration). Specific β-galactosidase activities (OD₆₄₀ ∼1.0) were calculated using the formula V_max/OD₆₀₀. The reported results represent data from at least four independent experiments.

Antibiotic Susceptibility and Biofilm Formation Experiments

The susceptibility of E. coli WT, P. aeruginosa WT, and of the respective derivatives to the antibiotics chloramphenicol, ciprofloxacin, tetracycline, tobramycin, gentamycin and ampicillin, was assessed by the broth micro-dilution method, in Mueller-Hinton medium (Gibco) supplemented with 0.1% arabinose, based on previously described methods [16], and following the CLSI guidelines [17]. The ability of E. coli WT, P. aeruginosa WT and the respective derivatives to form biofilms was assessed after 24, 48 and 72-h of growth on LB liquid medium at 28°C, based on previously published methods [18]. Briefly, appropriate volumes of overnight liquid cultures of bacterial strains were used to inoculate LB liquid medium and grown at 28°C with orbital agitation, until the mid-exponential phase was reached. The cultures were subsequently diluted to a standardized culture OD₆₄₀ of 0.5, and 20 µl of these cell suspensions were used to inoculate the wells of a 96-well polystyrene microtiter plate (Greiner Bio-One) containing 180 µl of LB medium. Plates were incubated at 28°C without agitation for 24, 48 and 72 h. The biofilm formed was quantified by measuring the absorbance at 590 nm using a VERSAmax microplate reader (Molecular Devices). All measurements were performed in triplicate, using biological duplicates.

Growth and Nutrient Deprivation Kinetics

Cultures of E. coli WT, P. aeruginosa WT and the respective derivatives were carried out at 37°C in liquid LB supplemented with 0.1% L-arabinose. Growth was followed spectrophotometrically at 640 nm. The ability of E. coli WT, P. aeruginosa WT and respective derivatives to survive to nutrient limiting conditions was assayed in M9 minimal medium supplemented with 0.1% L-arabinose instead of glucose, based on previously described methods [12]. Determination of total CFU’s was performed by plating aliquots of each culture in solid LB supplemented with 0.1% L-arabinose. All measurements were performed in quadruplicate, using biological triplicates.

Stress Susceptibility Experiments

The susceptibility of E. coli WT or P. aeruginosa WT and derivative strains to the stresses imposed by growth in LB solid medium containing 0.1% (w/v) L-arabinose, and supplemented with 0.05% SDS, 2.5% (w/v) NaCl, 2.5% (v/v) ethanol, or 25 µM methyl viologen, were performed as previously described [12]. Results are the means of at least 5 independent experiments.

Statistics Analysis

An unpaired two-tailed chi-square test was used to calculate the P values for β-galactosidase assays. Analysis of data from Northern and Western blotting was performed using a paired one-tailed t test to calculate the P values (P<0.01 [*]; p<0.005 [**]). Error bars represent the means of the standard deviation. Images shown are representative of the experiments performed. All experiments were repeated independently at least 4 times, using biological triplicates, with a minimum n value of 12.

Bioinformatics

BLAST searches were performed using the Integrated Microbial Genomes (IMG) webserver [19], using an E-value ≤1e⁻⁵⁰ as cut-off. MtvR putative targets were predicted using the sRNATarget [20], the TargetRNA [21], and the RNAPredator [22] programs, with a cut-off of 1, a minimum seed of 7 nucleotides, and an hybridization target region window size of −100 to +30 around the translation start site. The MtvR sRNA sequence was aligned with the RNA sequence of the putative mRNA targets, using the sequences retrieved from NCBI for the Escherichia coli str. K12 substr. DH10B and the Pseudomonas aeruginosa UCBPP-PA14 genome sequences, using the RNAHybrid web tool [23], using a minimum seed of 7 nucleotides [24] in the region around the start codon.

Results

MtvR Regulates the mRNA Levels of E. coli and P. aeruginosa hfq Genes by Promoting their Accelerated Decay

We have investigated if MtvR affects hfq expression in E. coli and P. aeruginosa. For this purpose, we transformed E. coli and P. aeruginosa WT strains with plasmid pCGR12, which allows MtvR expression. Results in Fig. 1A and 1C confirm that the sRNA is expressed, respectively, in E. coli and P. aeruginosa. The ectopic expression of MtvR for 2, 8 or 24 h led, respectively, to a reduction of 1.4±0.09, 1.29±0.08 or 1.21±0.03 fold of the hfq _Ec mRNA levels (Fig. 1A), and to a reduction of the hfq _Pa mRNA levels of 2.60±0.07, 2.30±0.13 or 1.8±0.3 fold (Fig. 1C). Expression of MtvR also affected the stability of the E. coli and P. aeruginosa hfq mRNAs, leading to the reduction of the hfq _Ec mRNA half-life time (t_1/2) from 13.4±1.0 to 8.6±0.5 min (Fig. 1B), and of the hfq _Pa mRNA t_1/2 from 18.6±1 to 9.9±0.4 min (Fig. 1D). Unlike the single band corresponding to hfq _Ec or hfq _Pa mRNAs observed for the respective controls, the expression of MtvR resulted in the detection of two bands corresponding to hfq _Ec or hfq _Pa mRNA decay products, suggesting that the processing of both mRNAs in the presence of MtvR involves pathways distinct from those used by the E. coli or P. aeruginosa WT strains. Interestingly, our results also indicate that MtvR requires the Hfq proteins from E. coli or P. aeruginosa for stability, as suggested by the decrease of the sRNA t_1/2 from 17.3±0.7 min in the E. coli WT strain to 8.5±0.4 min in the E. coli hfq mutant (Fig. 1B), and from 5.2±0.7 min in the P. aeruginosa WT strain to 2.6±0.4 min in the P. aeruginosa strain with the hfq _Pa gene silenced (Fig. 1D). Together, these results indicate that MtvR is stabilized by Hfq_Ec and Hfq_Pa, while the mRNAs hfq _Ec and hfq _Pa are targets of this sRNA.

Figure 1. MtvR regulates the mRNA levels of the E. coli and P. aeruginosa hfq genes.

(A) Levels the hfq_Ec mRNA and the MtvR transcript in cells of the E. coli WT or the E. coli Δhfq mutant, expressing (+pMtvR) or not the sRNA. (B) Stability of the hfq _Ec mRNA (left panel) in cells expressing (+pMtvR) or not MtvR, and of the MtvR transcript (right panel) in cells of the WT or Δhfq mutant, expressing MtvR (+pMtvR). (C) Levels of the hfq_Pa mRNA and the MtvR transcript in cells of the P. aeruginosa WT or WT with the hfq gene silenced (hfq ^sil), expressing (+pMtvR) or not MtvR. (D) Analysis of the stability of the P. aeruginosa hfq_Pa mRNA and of the MtvR transcript, in cells of the P. aeruginosa strains WT or the WT with the hfq gene silenced (hfq^sil), expressing (+pMtvR) or not MtvR. The levels of the 5S rRNA were used as loading control.

The MtvR sRNA Binds to the 5′ Leader Region of the hfq mRNA

Based on both experimental and bioinformatics analyses indicating that hfq _Ec mRNA is a target of MtvR, we have conducted further experiments to gain additional insights into the molecular details of the observed negative regulatory effects exerted by MtvR on the hfq _Ec mRNA.

The interaction between MtvR and the hfq _Ec 5′ leader region was investigated using the RNAHybrid software, which predicted a strong interaction between the two RNA molecules, suggesting that these two RNAs are able to form extended RNA duplexes, occluding the Ribosome Binding Site (green letters in Fig. 2A) and the start codon (Red letters in Fig. 2A).

Figure 2. MtvR regulates hfq _Ec expression through interaction with its 5′-leader region.

(A) Schematic representation of the nucleotide interaction between the 5′-UTR of hfq _Ec and the MtvR sRNA, highlighting in green and red lettering, respectively, the RBS and the AUG translation start site in the hfq _Ec 5′-UTR. (B) EMSA experiments using 25 nM of biotin-labeled MtvR transcript and the indicated concentrations of: left panel, the hfq _Ec full mRNA; center panel, the hfq _Ec coding region (CDS); right panel, the 5′-UTR of hfq _Ec. C) EMSA experiments performed to assess the ability of Hfq_Bc to bind to: left panel, 25 nM of the biotin-labeled hfq _Ec full mRNA; center panel, 25 nM of the biotin-labeled hfq _Ec 5′-UTR. The ability of mixtures containing Hfq_Bc together with the tested concentrations of the MtvR transcript to bind to 25 nM of the biotin-labeled hfq _Ec full mRNA is shown in the right panel. H_(n), Hfq_Bc(n); R₁, MtvR; R₂, hfq _Ec RNA; k_DP, affinity constant for the protein; k_DR, affinity constant for the RNA species.

This RNA-RNA interaction was experimentally confirmed by EMSA experiments. For this purpose, MtvR was transcribed in vitro and biotin-labeled, and the hfq _Ec RNA was transcribed from a T7 promoter using as template plasmid pCGR36, yielding the complete transcriptional unit of hfq _Ec, composed of 475 nt corresponding to the 5′ leader region of hfq _Ec, and 306 nt corresponding to the hfq _Ec coding sequence (CDS). Results presented in Fig. 2B (left panel) show that MtvR binds to the hfq_Ec RNA, with an apparent K_D of 51.4 nM. Similar band-shift assays performed with the hfq_Ec CDS instead of the full RNA, revealed that MtvR requires the hfq_Ec 5′-UTR for binding (Fig. 2B, center panel), since no MtvR displacement could be detected, at least for the concentrations used of the hfq _Ec CDS. An additional experiment was performed using the sRNA together with the 5′-UTR of hfq_Ec. Results shown in Fig. 2B (right panel) indicate that MtvR binds to the 5′-UTR of hfq_Ec, with an apparent K_D of 6.0 nM.

The E. coli Hfq protein was previously shown to be auto-regulated, through the interaction of the protein with two distinct binding sites within the 5′ leader of the mRNA, resulting in the inhibition of the formation of the translation initiation complex [25]. Since the action of sRNAs is often mediated by the Hfq RNA chaperone and the Hfq protein is involved in auto-regulation in E. coli, apparently without the requirement for sRNAs, we have conducted EMSA experiments with the hfq_Ec RNA (composed of the 5′-UTR and the coding sequence), or the hfq_Ec 5′-UTR together with the Hfq of B. cenocepacia J2315 (Hfq_Bc, which lacks amino acid residues beyond position 79). Results obtained indicate that Hfq_Bc can form complexes with the hfq_Ec RNA only in relatively high concentrations (500 nM or higher) (Fig. 2C, left panel), with an apparent K_D of 11.6 µM. The Hfq_Bc protein also needs to be present in concentrations as high as 250 nM in order to bind to the hfq_Ec 5′-UTR RNA (Fig. 2C, center panel). Interestingly, in the presence of the MtvR sRNA, the Hfq_Bc protein is able to bind to the two RNAs at a concentration 500-fold lower compared to the hfq_Ec RNA alone (Fig. 2C, left panel).

These results suggest that MtvR together with Hfq_Bc can synergistically bind to the hfq _Ec mRNA, more efficiently than Hfq_Bc alone.

MtvR Expression Affects E. coli Hfq Translation

Our data strongly suggests that MtvR might also impact the Hfq_Ec levels. To further investigate the impact of MtvR on the levels of Hfq_Ec, we have transformed the E. coli WT strain and the strain expressing MtvR with plasmid pCGR34, which allows expression of Hfq_Ec-His from its own promoter. This was performed since the polyclonal anti-Hfq IgG that was previously raised (unpublished data) is specific for the Hfq_Bc, and no signals on Western blots could be detected for the Hfq_Ec using this antibody (data not shown).

Results from Western blot analysis showed that MtvR expression reduced hfq _Ec translation by 1.87-, 2.09-, 2.30- or 2.77-fold, at 2, 4, 8 or 16 h, respectively (Fig. 3A).

Figure 3. MtvR affects E. coli Hfq translation.

(A) Western blot analysis of the levels of the 6×His-tagged Hfq_Ec protein in cells of the E. coli Δhfq mutant expressing only the tagged protein (left panel, Δhfq/phfq), or also expressing the MtvR transcript (Right panel, Δhfq/phfq+pMtvR). The GroEL levels were used as loading control. (B) β-galactosidase relative activity in cells of the E. coli strains WT (black bars), WT expressing MtvR (+pMtvR, white bars), or WT with the empty vector (pMLBAD, grey bars), harboring the hfq _Ec 5′-UTR-LacZ fusion (p_hfq-LacZ). C) β-galactosidase activity of the indicated E. coli strains, grown in solid LB media, supplemented with X-Gal and 0.1% L-arabinose. The WT strain transformed with pMLBAD was used as control.

The in vivo assessment of MtvR interaction with the 5′ leader region of hfq _Ec was achieved by measuring the β-galactosidase activity of the 5′-UTR-LacZ fusion. With this purpose, E. coli derivatives expressing MtvR were transformed with plasmid pCGR35. The WT strain was transformed with pMLBAD and used as control. Results obtained (Fig. 3B and 3C) show that the lowest values of β-galactosidase activity were registered when MtvR was expressed. The inability of MtvR to induce a complete translational blocking might derive from its dependence on a RNA chaperone for increased stability. These results are consistent with a negative regulatory action exerted by MtvR on hfq _Ec mRNA, occurring by binding to its 5′ leader region, which most probably prevents an efficient translation of the messenger RNA, by coupling enhanced mRNA decay with impaired ribosome loading.

MtvR Expression Affects E. coli and P. aeruginosa Growth Abilities

Data presented so far indicates that MtvR is able to regulate hfq _EC and hfq _Pa, and, at least, Hfq_Ec. This led us to investigate the roles of MtvR on the growth abilities of E. coli and P. aeruginosa. The specific growth rate and biomass yield reached after 24 h of batch growth of the E. coli expressing MtvR were reduced to approximately the same level as that of the Δhfq mutant strain (Fig. 4A). In the case of P. aeruginosa, the expression of MtvR reduced the specific growth rate and biomass yield reached after 24 h of batch growth to levels even lower than those observed for the strain with the hfq _Pa silenced (Fig. 4B). When challenged with long-term carbon starvation, E. coli and P. aeruginosa strains expressing MtvR exhibited a reduction of their survival ability more pronounced than the reduction observed for the E. coli Δhfq mutant and the P. aeruginosa with the hfq gene silenced (hfq ^sil), respectively.

Figure 4. MtvR expression affects E. coli and P. aeruginosa growth kinetics and survival to prolonged nutrient deprivation.

Growth curves in liquid LB medium supplemented with 0.1% L-arabinose (A, B) and survival in M9 minimal medium supplemented with 0.1% L-arabinose, at 37°C for 30 days (C, D), of (A, C) E. coli strains WT (circles), WT expressing MtvR (squares) or the Δhfq _Ec mutant (triangles), or (B,D) P. aeruginosa strains WT (circles), WT expressing MtvR (squares) or the WT strain with the hfq _Pa gene silenced (triangles).

Since the growth rate and biomass yield or resistance to nutrient deprivation registered for the WT strains of E. coli or P. aeruginosa and the respective transformants harboring pMLBAD were similar, these results were not included in Fig. 4 to keep it simpler.

MtvR Affects the Biofilm Formation Ability of E. coli and P. aeruginosa

In a previous study MtvR was shown to play a role on biofilm formation ability in Bcc bacteria [9]. Therefore, we decided to investigate the effect of expressing MtvR on the biofilm formation ability of E. coli and P. aeruginosa. Since for the different strains under study differences in the biomass yield were registered, results were expressed as relative biofilm amount, i.e., the ratio of biofilm amount estimated using the crystal violet and the biomass concentration assessed by the OD₆₄₀ of cultures. Results obtained are presented in Figs. 5A and 5B and indicate that MtvR expression affected negatively the ability of both E. coli and P. aeruginosa to form biofilms in vitro, particularly evident at 72 h. The observed effects might result from the down-regulation of hfq _Ec and hfq _Pa, since the strains E. coli Δhfq and P. aeruginosa hfq ^sil formed relative biofilm amounts comparable to those formed by the respective WT strains expressing MtvR.

Figure 5. MtvR expression in E. coli and P. aeruginosa reduces biofilm formation ability and increases susceptibility to stresses.

Relative biofilm formation ability (panels A, B) and susceptibility to the stress imposed by growth on the surface of LB solid medium supplemented or not (LB) with the indicated concentrations of methyl viologen (MV), ethanol (etOH), NaCl or SDS (SDS) (panels C,D), of strains of (panels A, C) E. coli WT (WTEc), WT expressing MtvR (WTEc+MtvR), or the Δhfq _Ec mutant (EcΔhfq), and (panels B, D) P. aeruginosa WT (WTPa), WT expressing MtvR (WTPa+MtvR), or the WT strain with the hfq _Pa gene silenced (Pahfqsil). Relative biofilm formation was estimated by dividing the total amount of biofilm formed by the total amount of biomass (see Materials and Methods section). Panels E and F show photographs illustrative of results from a single representative susceptibility experiment with the E. coli (panel E) and P. aeruginosa (panel F) strains WT (WT), WT expressing MtvR (+MtvR), and the E. coli Δhfq _Ec mutant (Δhfq), or the P. aeruginosa WT with the hfq _Pa gene silenced (hfq ^sil). Susceptibilities were assessed by spot inoculation of serially diluted bacterial suspensions with an initial OD640 of 1.0. Error bars represent standard deviation of the means. Numbers above bars in panels (A) and (B) are the estimated P-values.

E. coli and P. aeruginosa Exhibit Enhanced Susceptibility to Stresses When Expressing MtvR

The effects of MtvR expression on the susceptibility of E. coli and P. aeruginosa strains to oxidative, osmotic and membrane stresses were investigated by spot-inoculation of bacterial culture aliquots on LB solid media supplemented with methyl viologen, NaCl, ethanol or SDS. Illustrative photographs of a set of results obtained for E. coli and P. aeruginosa are shown, respectively, in Figs. 5E and 5F. Methyl viologen is a charged quaternary ammonium compound that generates reactive oxygen species under aerobic conditions [26], used to generate oxidative stress conditions. Ethanol and NaCl were used to increase the medium osmolarity, while SDS was used as a membrane integrity-perturbing agent. These stressors were chosen to mimic the environmental conditions oxidative stress, high osmolarity, and extracytoplasmic stress, conditions that are faced, for instance, when P. aeruginosa colonizes/infects the cystic fibrosis lung [10].

Results presented in Fig. 5C show that when compared to the WT strain, the numbers of total CFUs of the E. coli strains expressing MtvR or the Δhfq _Ec mutant, exposed to methyl viologen or SDS, were reduced by more than 1 log. This reduction was higher, 2 to 3 logs, for cells of the strains expressing MtvR or Δhfq _Ec when exposed to stressors NaCl or ethanol, respectively. Under non-stress conditions, about one log reduction in the total CFU were registered for the E. coli strains expressing MtvR and the mutant Δhfq _Ec when compared to the WT strain (Fig. 5C). It is worth to note that the reductions in total CFUs registered for the E. coli strains expressing MtvR or the Δhfq mutant were quite similar, suggesting that the observed increased susceptibility to the tested stressors might result from the down-regulation of hfq _Ec.

In the case of P. aeruginosa, no differences on the total CFUs were registered for the strains WT, WT expressing MtvR, and the WT with the hfq _Pa gene silenced under non-stress conditions (Fig. 5D). However, when compared to the P. aeruginosa WT strain, approximately 2, 3, or 4-log reduction in the total CFU were registered for the strains P. aeruginosa expressing MtvR and P aeruginosa hfq ^sil when exposed, respectively, to methyl viologen or SDS, NaCl, or ethanol (Fig. 5D). Since no significant differences in the total CFUs were observed for the E. coli strains WT and WT transformed with pMLBAD, and P. aeruginosa WT and P. aeruginosa transformed with pMLBAD, only the results obtained for the E. coli and P. aeruginosa WT strains are shown, respectively, in Figs. 5C to 5F.

MtvR Expression Enhances Antibiotic Susceptibility in E. coli and P. aeruginosa

The sRNA MtvR was recently shown to play a role on resistance to antibiotics in Bcc [9]. Therefore, we investigated the effects of MtvR expression on the E. coli and P. aeruginosa susceptibility to the antibiotics chloramphenicol, ciprofloxacin, tetracycline, tobramycin, gentamycin and ampicillin. Results presented in Table 3 show that MtvR expression reduced E. coli MIC values by 8-fold for tetracycline, 4-fold for chloramphenicol and ciprofloxacin, and 2-fold for tobramycin, gentamycin and ampicillin. The observed increased susceptibility to these antibiotics in the Δhfq _Ec mutant strain was identical to the observed due to MtvR expression, except for chloramphenicol and tetracycline. For these two antibiotics, the susceptibility only increased by 2- and 4-fold, respectively. The effect of MtvR expression in the Δhfq _Ec led to an even more drastic effect in susceptibility, especially to tobramycin and gentamycin, with MIC values lowering 16- and 128-fold, respectively (Table 3).

Table 3. Antibiotic susceptibility of E. coli WT and derivative strains.

Strain	CHL	CIP	TET	NM	GNT	AMP
E. coli WT	4	1	2	8	16	4
E. coli Δhfq	2*	0.25	0.25	2	8	2
E. coli WT/phfq	16	8	32	32	128	64
E. coli WT/phfq+MtvR	2	0.25	0.5	2	1	0.5
E. coli WT/pMLBAD	4	1	2	8	16	4
E. coli WT/MtvR	1	0.125	0.125	4	8	2
E. coli Δhfq/pMLBAD	64*	≤0.125	0.125	1	8	2
E. coli Δhfq/phfq	>512*	1	2	8	16	4
E. coli Δhfq/MtvR	8*	≤0.125	≤0.125	0.5	0.125	1
E. coli Δhfq/phfq+MtvR	64*	0.125	0.125	4	8	2

CHL, chloramphenicol; CIP, ciprofloxacin; TET, tetracycline; NM, tobramycin; GNT, gentamycin; AMP, ampicillin. Numbers with asterisks represent strains that have a chromosomal chloramphenicol resistance cassette.

We have also investigated the role of MtvR on the P. aeruginosa susceptibility to chloramphenicol, ciprofloxacin, tetracycline, tobramycin, gentamycin and ampicillin. Results in Table 4 indicate that the antibiotic susceptibility to the tested antibiotics increased in cells expressing MtvR. MtvR expression induced a 4-fold reduction in most of the MIC values, with the exception of gentamycin, for which a MIC value 2-fold lower than the WT strain was registered (Table 4). In the strain with the hfq _Pa gene silenced, the antibiotic susceptibility was also affected, although to a lesser extent. The MIC values for chloramphenicol and ampicillin remained unchanged in the strain with the hfq silenced (Table 4). A more drastic effect on antibiotic susceptibility was observed in the strain with the hfq gene silenced and expressing MtvR. In fact, an impressive 64-fold reduction in the MIC value for chloramphenicol, 32-fold for tobramycin and gentamycin, and 16-fold for ampicillin were registered for this strain (Table 4).

Table 4. Antibiotic susceptibility of P. aeruginosa WT and derivative strains.

Strain	CHL	CIP	TET	NM	GNT	AMP
P. aeruginosa WT	64	1	16	16	128	>512
P. aeruginosa hfqsil	64	0.5	8	4	64	512
P. aeruginosa WT/phfq	128	1	64	128	256	>512
P. aeruginosa WT/phfq+MtvR	16	0.125	2	4	16	64
P. aeruginosa WT/pMLBAD	64	1	16	16	128	>512
P. aeruginosa WT/MtvR	16	0.125	2	4	16	64
P. aeruginosa hfqsil/MtvR	1	0.125	1	0.5	8	32

CHL, chloramphenicol; CIP, ciprofloxacin; TET, tetracycline; NM, tobramycin; GNT, gentamycin; AMP, ampicillin.

MtvR has Additional Putative Targets in E. coli

Some of the phenotypes here reported for the E. coli strain expressing MtvR differ from those observed for the E. coli Δhfq _Ec mutant, suggesting that MtvR might regulate additional mRNAs in this bacterium. Therefore, we have used the programs TargetRNA [21], RNAPredator [22] and sRNATarget [20] to predict possible MtvR additional mRNA targets within the genome of E. coli K12 strain MG1655. A total of 11 (Table S1 in File S1) and 3 (Table S2 in File S1) putative mRNA targets were predicted by TargetRNA and RNAPredator, respectively, assuming only putative hybridizations in the −20 to +20 nt region around the start codon. Using the sRNATarget with the same restrictions, and excluding genes of unknown functions, 52 distinct mRNA targets were predicted (Table S3 in File S1). Interestingly, several targets were predicted to interact with MtvR in more than one region (Table S4 in File S1).

The gene uhpA of E. coli was the only common target of MtvR predicted by TargetRNA and sRNATarget. The uhpA encodes the response regulator of a two-component regulatory system where UhpB is a histidine kinase that controls the synthesis of the sugar-phosphate transporter UhpT [27]. These findings prompted us to investigate the effects of MtvR expression on the transcript levels corresponding to the uhpA mRNA in E. coli using RT-PCR. We have also investigated the levels of the uhpT mRNA, which is regulated by the UhpA-UhpB two-component regulatory system [27]. Results presented in Fig. 5 show that MtvR expression led to undetectable uhpA-derived cDNA, opposed to the observed for the WT strain (Fig. 6A). However, cDNA corresponding to uhpA mRNA could be detected in the Δhfq _Ec mutant strain, although with a reduced intensity relative to the WT strain. These observations suggest that Hfq might be involved in uhpA regulation.

Figure 6. The MtvR sRNA also targets uhpA gene in E. coli.

(A) Reverse-transcription analysis of the effect of MtvR on the mRNA levels of uhpA. Total RNA was obtained from late-exponentially growing cells of the E. coli strains WT and WT expressing MtvR (+pMtvR), or the Δhfq mutant (Δhfq). Reverse-transcription experiments were also performed for uhpT, induced by UhpA. The 16S and 23S rRNA bands were used as loading controls. PCR experiments were also performed using DNA, for reference. Images shown are representative of 3 independent experiments. B) Schematic representation of the nucleotide interaction between the 5′-UTR of uhpA and MtvR, highlighting in red lettering the AUG translation start site in the 5′-UTR of uhpA.

The levels of cDNA corresponding to uhpT mRNA were highly reduced in E. coli cells expressing MtvR when compared to those observed for the WT strain (Fig. 6A). This observation is consistent with the requirement of UhpA (in its phosphorylated form) for the transcriptional activation of uhpT [27].

Since the levels of cDNA corresponding to uhpT mRNA were higher in the Δhfq _Ec mutant than in the strain expressing MtvR, we concluded that Hfq_Ec is unlikely to be involved in the direct regulation of uhpT mRNA. The regulation exerted by MtvR on uhpA mRNA might involve partial base-pairing of the two molecules, predicted to occur within the 5′ region of uhpA mRNA, leading to the formation of a RNA duplex (with a predicted energy of −117.0 kcal/mol) that occludes part of the RBS and start codon (Fig. 6B).

We also have used the TargetRNA [21], the RNAPredator [22] and the sRNATarget [20] programs to predict possible MtvR mRNA targets within the genome of P. aeruginosa UCBPP-PA14 (Tables S5 and S6 in File S1). No common targets were predicted.

Discussion

The sRNA MtvR was recently identified as a trans-encoded sRNA that occurs exclusively among members of the Burkholderia genus [9]. In these bacteria, MtvR acts as a global regulatory RNA, and strains with the sRNA silenced or overexpressed exhibited pleiotropic phenotypes related to growth and survival when challenged with stress, motility, biofilm formation, resistance to antibiotics and virulence [9]. In addition, MtvR was shown to regulate the levels of at least 17 mRNA targets in the cystic fibrosis isolate B. cenocepacia J2315 [9]. Results presented in this work also show that, at least in E. coli, MtvR targets other genes besides hfq _Ec, as is the case of uhpT.

Several trans-encoded sRNAs have been described as regulating multiple targets, most probably due to the limited complementarities shared with their mRNA targets [6]. This limited base-pairing is thought to justify the need of most of the trans-encoded sRNAs to bind to the RNA chaperone Hfq to effectively interact with their targets [6].

Despite the absence of homologues to MtvR in E. coli or P. aeruginosa, here we present evidence that this sRNA is able to regulate the levels of the hfq mRNA in both species.

The Hfq protein of Burkholderia is 83% identical to the E. coli [14]. The E. coli Hfq is composed of 102 amino acid residues, while the P. aeruginosa and B. cenocepacia Hfq proteins are composed of 82 and 79 amino acid residues, respectively. All 3 proteins contain the conserved Sm1 and Sm2 motifs, and have a secondary structure compose by a N-terminal α-helix, followed by 5 β-strands. The amino acid residues 8–68 correspond to the conserved core of the 3 proteins.

Earlier studies on the E. coli Hfq have shown that the protein relative abundance is growth-phase dependent [28], being hfq _Ec transcription regulated by multiple mechanisms [29]. More recently, Vecerek et al. (2005) presented evidence indicating that in E. coli the synthesis of Hfq_Ec is auto-regulated at the translational level [25]. These authors have shown that Hfq_Ec binds to the two Hfq_Ec-binding sites in the 5′-UTR region of hfq _Ec mRNA, inhibiting the formation of the translation initiation complex [30], in an interaction that involves the C-terminal region of the protein. This interaction occurs in vitro at protein concentrations ranging 50–200 nM [30]. In the case of Bcc bacteria, the Hfq_Bc protein lacks the C-terminal region [12] suggesting a lack of auto-regulation. In fact, the B. cenocepacia hfq _Bc mRNA was recently shown to be regulated through a mechanism involving sequestration of the RBS by MtvR, leading to accelerated mRNA decay and reduced protein translation [9].

Interestingly, the 5′-UTR region of the Bcc hfq _Bc is 55% identical to the E. coli hfq _Ec 5′-UTR (data not shown). Our results show that the B. cenocepacia shorter protein is also able to bind to the E. coli hfq _Ec 5′ leader region, but with a ∼100-fold lower affinity. We also show that the sRNA can act synergistically with Hfq_Bc, increasing its affinity to hfq _Ec mRNA by more than 500-fold.

A few examples of functional analysis of sRNAs on heterologous systems have been reported. For instance, AbdelRahman et al. (2010) have used a heterologous co-expression system to demonstrate that the Chlamydia trachomatis non-coding RNA CTIG270 regulates the expression of FtsI by inducing ftsI mRNA degradation [31].

The data presented here show that despite being absent in E. coli, MtvR regulates at least, the levels of the hfq, uhpA and uhpT mRNAs. In B. cenocepacia J2315, MtvR affects the mRNA levels of at least 17 genes, among 309 predicted targets [9]. It is therefore possible that MtvR also interacts with other mRNAs, as suggested by the observed phenotypes of cells expressing the sRNA.

Perhaps the most interesting phenotypes observed due to MtvR expression in E. coli and P. aeruginosa are those related to increased antibiotic susceptibility. Bacterial resistance to antibiotics is a problem of increasingly concern, since data from the Centers for Disease Control and Prevention evidence a rapidly increasing rate of infections due to fluoroquinolone-resistant P. aeruginosa [32].

A recent study revealed that the involvement of MtvR in the regulation of an hfq-like gene also impacted the bacterium resistance to several antibiotics, leading to a phenotype conversion from resistant to susceptible [9]. A study by Yamada et al. (2010) revealed that mutations in the hfq gene from E. coli resulted in susceptibility to acriflavine, benzalkonium, cefamandole, chloramphenicol, crystal violet, nalidixic acid, novobiocin, oxacillin and rhodamine 6G [33]. In addition, Moon & Gottesman (2009) reported on the requirement of Hfq for resistance to polymyxin B through a mechanism involving the sRNA MgrR [34]. In Stenotrophomonas maltophilia, Hfq was also shown to play a role in resistance to tobramycin and amikacin, most likely due to the regulation of efflux pumps [35].

Anti-sense acting oligonucleotides are being used as components of peptide-morpholino oligonucleotide conjugates (PMO) that can act as bactericidal agents. For instance, a PMO targeting the highly conserved region of the E. coli gyrA was recently shown to effectively inactivate several species of Gram-positive and Gram-negative bacteria, when used in the micro-molar range [36].

Results presented in this work show that when expressing MtvR, the MIC values for the studied antibiotics were reduced by 2 to 128 fold in E. coli and P. aeruginosa, pointing out this sRNA as an interesting molecule, with potential to be exploited as an adjuvant in antimicrobial therapies. In fact, multi-target sRNAs, like MtvR, are potential candidates for the development of PMOs that can be used as antimicrobials, or in combination with already available antibiotics, to fight infections by multi-resistant bacteria, as is the case of P. aeruginosa.

Supporting Information

File S1

Supporting Information File combining Supplementary Methods, Supplementary Results, Table S1 (Putative MtvR targets in the genome of E. coli predicted by TargetRNA), Table S2 (Putative MtvR targets in the genome of E. coli predicted by RNAPredator), Table S3 (Putative MtvR targets in the genome of E. coli predicted by sRNATarget), Table S4 (Regions with homology to within the genome of E. coli), Table S5 (Putative MtvR targets in the genome of P. aeruginosa UCBPP-PA14 predicted by TargetRNA), and Table S6 (Putative MtvR targets in the genome of P. aeruginosa UCBPP-PA14 predicted by RNAPredator), and Supplementary References.

(PDF)

Data Availability

The authors confirm that all data underlying the findings are fully available without restriction. All data are included within the manuscript.

Funding Statement

This work was partially supported by Fundação Ciência e Tecnologia, Portugal (contracts PTDC/BIA-MIC/119091/2010 and PTDC/BBB-BIO/1958/2012), a post-Doc grant to CGR, and doctoral grants to AMG and JRF, respectively. SAS and PJPDC acknowledge research grants from projects PTDC/BBB-BIO/1958/2012 and PTDC/BIA-MIC/119091/2010, respectively. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.