Quorum sensing in Pseudomonas aeruginosa mediated by RhlR is regulated by a small RNA PhrD

Pseudomonas aeruginosa is a highly invasive human pathogen in spite of the absence of classical host specific virulence factors. Virulence factors regulated by quorum sensing (QS) in P. aeruginosa cause acute infections to shift to chronic diseases. Several small regulatory RNAs (sRNAs) mediate fine-tuning of bacterial responses to environmental signals and regulate quorum sensing. In this study, we show that the quorum sensing regulator RhlR is positively influenced upon over expression of the Hfq dependent small RNA PhrD in Pseudomonas. RhlR transcripts starting from two of the four different promoters have same sequence predicted to base pair with PhrD. Over expression of PhrD increased RhlR transcript levels and production of the biosurfactant rhamnolipid and the redox active pyocyanin pigment. A rhlR::lacZ translational fusion from one of the four promoters showed 2.5-fold higher expression and, a 9-fold increase in overall rhlR transcription was seen in the wild type when compared to the isogenic phrD disruption mutant. Expression, in an E. coli host background, of a rhlR::lacZ fusion in comparison to a construct that harboured a scrambled interaction region resulted in a 10-fold increase under phrD over expression. The interaction of RhlR-5′UTR with PhrD in E. coli indicated that this regulation could function without the involvement of any Pseudomonas specific proteins. Overall, this study demonstrates that PhrD has a positive effect on RhlR and its associated physiology in P. aeruginosa.

SCienTifiC RePoRts | (2019) 9:432 | DOI: 10.1038/s41598-018-36488-9 regulation of RhlR is shown by means of transcriptional assays, reporter gene fusions and physiological assays in strains expressing altered levels of PhrD as well as studies in E. coli as a heterologous system.

Results
In silico analysis of PhrD. PhrD, a 74 nt highly conserved small RNA 16 was predicted to have two stem loops in its secondary structure (Fig. 2a). Putative targets of PhrD were predicted by RNA Predator program and confirmed by IntaRNA. A list of its first 50 putative targets can be found in Supplementary Table S1. Most of the predicted targets showed base pairing to the second stem (II) of PhrD. Out of the indicated targets, the major transcriptional activator of quorum sensing, RhlR, was chosen to study for its regulation by PhrD owing to its role in quorum sensing and regulation of pathogenicity genes 17 . rhlR gene reportedly has four promoters 17 (Fig. 3), and the sequences coinciding with the transcription start site of the P3 promoter (downstream of P4) were predicted to bind with PhrD (−159 to −134, 'G' at position −158 being the transcription start site, Fig. 2b).
Over expression and disruption of phrD and its levels under different nutrient conditions. phrD without its promoter was over expressed in Pseudomonas under the pBAD promoter using the shuttle vector pHERD30T. Two prominent transcripts of 160 nt and 74 nt were expressed from the cloned plasmid and the chromosomal copy respectively in comparison with the control strain (Fig. 4a). The longer transcript derived from  phrD, cloned downstream to the pBAD promoter was shown to base pair with RhlR transcript at the same nucleotides as would the chromosomal PhrD, as analyzed by IntaRNA program (Supplementary Fig. S1). Expression analysis of phrD in the wild type (WT) strain indicated a steady increase along the growth curve after normalization with 16S rRNA gene (Fig. 4b). Chromosomal phrD gene was disrupted by the introduction of a gentamicin resistance gene (Gm R ) by homologous recombination. Disruption of phrD was confirmed by PCR with multiple primers (see Supplementary Fig. S2) and absence of transcript in northern blot (Fig. 4a, lane 3). PhrD was predicted to base pair with RhlR mRNA, at the transcription start site of its σ 54 (involved in N regulation) dependent P3 promoter which has been reported to express under phosphate limited conditions 17 . Expression analysis of phrD in Luria broth, nitrogen limited MMP and phosphate limited PPGAS medium showed that phrD expressed under all the studied conditions ( Fig. 4c) with maximum levels in MMP (5-6 fold higher than the other two) (Fig. 4d).
PhrD positively influences rhlR expression. The predicted interaction region of PhrD with RhlR coincides with the transcription start of P3 promoter of rhlR gene, at −159 to −134 bp upstream of the start codon of RhlR mRNA. phrD over expression in strain PAO1 influenced a 6-fold increase on RhlR in qRT-PCR assays (Fig. 5a). Disruption of phrD (phrDΩGm) in PAO1, reduced rhlR expression to 0.2-fold while its complementation with pUCP-phrD restored the levels of RhlR (Fig. 5b).
With an objective of establishing a correlation between PhrD and RhlR levels, lacZ reporter translational fusions were constructed from promoter P3-rhlR ( Fig. 3 construct A-pME6013) and a time course measurement of β-galactosidase activity in PPGAS, LB and MMP media was correlated with RhlR transcript levels. β-galactosidase activity of the fusion exhibited a specific increase along the log phase up to 10 h in the WT and the disruption strains. The increase in WT, as compared to the disruption, was 1.5 and 2.5-fold in PPGAS and LB respectively ( Fig. 6a(i), b(i)). This paralleled with a 6-9 fold increase in the overall transcript levels of RhlR in the WT PAO1 with respect to the disruption, in these media ( Fig. 6a(ii), b(ii)). Comparing the increase observed in qRT-PCR with that of P3-rhlR::lacZ fusion explains that the difference in the fold increase might arise as a consequence of the modulation of transcripts arising from P4 promoter also.
Complementation of the disruption strain by the cloned PhrD restored its β-galactosidase levels to ~75% of that of the WT (Fig. 6b(i)). Expression of B-pME6013 bearing a non-specific scrambled interaction sequence did not show any increase even in the WT (phrD + ) background ( Fig. 6b(i)). These results corroborated the positive influence of PhrD with a sequence specific interaction to RhlR transcripts.
The expression of P3-rhlR::lacZ was very high and maximum in nitrogen limited MMP medium as compared to other media, and had no difference between WT and disruption strains ( Fig. 6c(i)). This result perhaps indicated that under amino acid starvation there is an involvement of stringent response mediated by ppGpp in the high expression of rhlR 18,19 , alleviating the requirement of PhrD under these conditions. RhlR expression is regulated by PhrD in a heterologous system. Subsequent to the positive influence of PhrD on rhlR expression, the specificity of their interaction was assayed in the heterologous host E. coli with the constructs A-pME6013 and B-pME6013. A-pME6013 harboured the intact interaction region (Fig. 2b) whereas B-pME6013 contained a scrambled stretch of the same 25 nucleotides (Fig. 2c) instead of the predicted interaction region and served as a negative control. When introduced into an E. coli containing the PhrD over expressing plasmid, β-galactosidase activity of the fusion with intact interaction region showed a specific increase of 10-fold with growth ( Fig. 7). No significant increase was observed in the expression of P3-rhlR::lacZ fusion with the scrambled PhrD-RhlR mRNA interaction region, and in A-pME6013 with vector plasmid (no PhrD). These results proved the specific involvement and significance of base pairing interaction of PhrD with RhlR mRNA at the stretch of 25 nucleotides at the 5′UTR. Also, regulation of RhlR by PhrD in a heterologous system like E. coli proved that the interaction between PhrD and RhlR could be carried out without the assistance of any Pseudomonas specific proteins.
PhrD positively influences rhamnolipid and pyocyanin production. RhlR, in complex with its cognate autoinducer C 4 -HSL, stimulates the synthesis of important virulence factors like rhamnolipids and pyocyanin pigment 6 . Therefore, an effect on their production, in relation to PhrD could be partially attributed to the regulation of rhlR by this small RNA. Rhamnolipids are known to be produced maximally under elevated C/N ratio under nitrogen exhaustion 20 . Measurement of rhamnolipid levels in M9 defined medium with 0.4% glycerol and low nitrogen source (0.1 g/L KNO 3 ) showed an increase of 2.5-fold under phrD over expression (Fig. 8a). The disruption strain that showed 0.5-fold level of rhamnolipid was restored to 1.8-fold when complemented with multi-copy PhrD. The rhamnolipid levels in these strains are in correlation with that of RhlR and PhrD levels in M9 (Fig. 8b,c). Likewise, the pyocyanin production in LB that was increased by 4-fold in pHERDphrD strain is also in correlation with the corresponding RhlR levels (Figs 9 and 5a).

Discussion
This paper reports the functional characterization of a sRNA PhrD and its influence on quorum sensing in P. aeruginosa under different nutrient conditions that mimic host physiology in pathogenesis. PhrD was shown to base pair at the transcription start site of the σ 54 dependent P3 promoter of rhlR (downstream of P4) which was reported to express under phosphate limited PPGAS medium 17 . A steady increase in the transcript levels of RhlR in the presence of PhrD indicated its role in modulation of rhlR expression from P3 and P4 promoters ( Fig. 6a(ii), b(ii)). Enhancement of expression of P3-rhlR::lacZ fusion by PhrD, observed in PPGAS and Luria broth ( Fig. 6a(i),b(i)), signifies its role in regulation of rhlR expression under phosphate deficient and nutrient rich conditions. Although rhlR expression is reported to be dependent on LasR when grown in Luria broth 21 , our results indicate that PhrD is an additional regulator of RhlR under these conditions. PhrD had no influence on rhlR expression in MMP medium ( Fig. 6c(i), c(ii)), its role perhaps being overruled by stringent response under this condition. Amino acid starvation in P. aeruginosa, promptly elicits the RelA protein mediated stringent response (SR) that synthesizes the signalling molecule (p)ppGpp. The QS-based response is mediated by RelA, since a (p)ppGpp-null SR mutant (spoT relA) shows reduced rhlI, rhlR, lasI, and lasR expression 19 . Further, overexpression of RelA leads to the early transcription of the lasR and rhlR genes, production of QS signals, and overproduction of QS-dependent virulence factors rhamnolipids, pyocyanin, elastase etc. 18 The positive influence of PhrD on the expression of rhlR::lacZ fusion observed in E. coli host background proves the specific interaction between RhlR and PhrD without the requirement of P. aeruginosa specific proteins (Fig. 7). This interaction was shown to be sequence specific, and a direct base pairing between PhrD sRNA and RhlR mRNA manifested this regulation. The Pseudomonas background however facilitated better expression of rhlR than by E. coli, probably owing to better recognition of P3 promoter by Pseudomonas sigma factors or involvement of additional Pseudomonas proteins. A similar heterologous E. coli host system was used earlier for in vivo identification of nucleotides important for regulation of HapR mRNA by Qrr sRNA of V. cholerae 22 .
The increase induced by PhrD in the production of rhamnolipid and pyocyanin mediated by RhlR, reflected an indirect regulation of these pathogenicity factors by this sRNA (Figs 8a and 9). The haemolytic activity as well as high surface activities of rhamnolipids cause lyses of polymorphonuclear leukocytes and monocyte-derived macrophages leading to necrosis 3 . Strains lacking in rhamnolipid production are rapidly cleared from the lungs of Pseudomonas infected mice 23 . Pyocyanin influences pathogenicity by causing goblet cell metaplasia and hyperplasia, airway fibrosis, and alveolar airway destruction in cystic fibrosis lungs 24 . It damages human cells by causing inhibition of cellular respiration, ciliary function, epidermal cell growth, prostacyclin release and disruption of calcium homeostasis 2 . PhrD-directed increase in the synthesis of these RhlR associated virulence factors may enhance invasiveness and colonization of the pathogen.
PhrD regulates quorum sensing under different conditions in a generalized manner in P. aeruginosa unlike other previously characterized sRNAs: RsmY regulates rhl system under nutrient rich conditions 11 . PhrS activates PqsR quorum sensing regulator under oxygen limited conditions leading to increased pyocyanin production 13 . Base pairing of NrsZ sRNA with rhlA activates rhamnolipid production under nitrogen limitation 14 . Our findings show that sRNA PhrD enhances both, rhamnolipid and pyocyanin production, by positively influencing RhlR. This work establishes a substantial route of positive regulation of RhlR transcripts generated from multiple promoters upstream to the PhrD interaction region. PhrD thus proves out to be a sRNA that assists this bacterium to tune in to the environmental and host induced stimuli.

Methods
Bacterial strains, plasmids and growth conditions. The strains and plasmids used in the study are mentioned in Table 1. For routine culturing, bacterial cultures were grown in Luria broth (HiMedia, Mumbai, India) at 37 °C and supplemented with antibiotics as and when required. All the antibiotics were bought from HiMedia, Mumbai, India and were added to a final concentration of 30 µg/ml of gentamicin(Gm), 300 µg/ml of carbenicillin(Cb), 100 µg/ml of tetracycline for P. aeruginosa and 10 µg/ml of gentamicin, 100 µg/ml of ampicillin(Ap) and 25 µg/ml of tetracycline(Tc) for E. coli. For rhamnolipid measurements, cultures were grown in M9 minimal medium (1X M9 salts, 2 mM MgSO 4 , 0.4% glucose, 100 µM CaCl 2 ). 5X M9 salts contained Na 2 HPO 4 .  In case of MMP, the cells were first grown to the desired OD in N-rich medium and resuspended in N-limited MMP. There was no net growth owing to nitrogen starvation. β-galactosidase activity of the disruption strain complemented with phrD over expression plasmid is measured in Luria broth. The data is represented as the mean ± SD of three individual experiments and each sample was analyzed in triplicates. Luria broth. Cells grown to logarithmic phase in N-rich medium were shifted down to N-limited MMP for the β-galactosidase assay.
Bioinformatics. The sequence of PhrD sRNA was retrieved from Pseudomonas Genome database 27 and its secondary structure was determined using the mfold Web Server 28 . RNA Predator 29 and IntaRNA program 30 were used to predict the putative targets of PhrD.

Construction of plasmids.
All the cloning procedures were carried out as per the standard molecular biology protocols 25 . A 134 bp region containing the phrD gene without its endogenous promoter was amplified with primer pair PhrDF and PhrDR (Table 2)    RhlR mRNA (primers AM_103 and 104), by overlap extension PCR (Fig. 3). The PCR fragment was fused to the eighth codon of lacZ at EcoRI-PstI digested pME6013 to get translational fusion of rhlR with lacZ, A-pME6013. A similar construct B-pME6013 that served as a negative control was made where in the interaction region was Figure 9. Pyocyanin production under altered levels of PhrD. Pyocyanin levels in LB were measured as µg/ml/ OD 600 of the culture. **and *indicate statistically significant data at P < 0.01 and P < 0.05, as analyzed by one way ANOVA performed on data obtained from three independent experiments. substituted with a commercially synthesized scrambled sequence (primers AM_111 and AM_112). Neither of these constructs included the P4, P2 and P1 promoters mentioned by Medina and colleagues 17 .

Strains/Plasmids
Construction of phrD disruption mutant. phrD gene was disrupted by gentamycin resistance marker, cloned into plasmid pTZ57R/T and transformed into PAO1 strain to achieve disruption of chromosomal phrD by homologous recombination. An 855 bp long gentamicin marker was amplified with GmF and GmR primer pair using pBBRMCS5 plasmid as the template. The 734 bp and 792 bp long upstream and downstream regions consisting of 5′ and 3′ ends of phrD respectively were amplified using primers FupPhrD and RupPhrD-Gm and FdnPhrD-Gm and RdnPhrD respectively from genomic DNA of strain PAO1. The reverse and forward primers of the above primer pairs shared a 20 bp homology with the gentamicin marker. An overlap extension PCR was done to fuse the above amplified fragments and the resulting gene disruption cassette of phrD was cloned in pTZ57R/T plasmid using TA overhangs. The disrupted gene in the recombinant plasmid was recombined into strain PAO1 containing the recombinant proficient vector pUCP18-RedS expressing the λ red recombinase 32 . Gentamicin resistant, carbenicillin sensitive colonies were screened with PCR to confirm the chromosomal disruption of phrD (Supporting gels in Supplementary Fig. S2). For complementation of phrD disruption mutant, a 146 bp long XbaI-PstI released fragment from pHERD-phrD(Gm r ) was sub-cloned into plasmid pUCP18(Cb r ) under lac promoter and transformed into gentamicin resistant disruption strain. Complementation of phrDΩGm was achieved by expressing phrD from pUCP18.

RNA isolation and Northern Blot.
Total RNA was extracted from stationary phase cultures grown in different media by acid guanidinium thiocyanate-phenol-chloroform method 33 for Northern Blot, and Roche High Pure RNA purification kit for real time PCR, as per manufacturer's instructions. For northern blots, total RNA was separated on 6% polyacrylamide/6 M urea gels and electroblotted on to the nylon membrane. The RNA was UV cross-linked to the membrane followed by hybridization with PhrD probe, labelled by Digoxygenin-dUTP using DIG-labeling kit as per manufacturer's instructions (Roche, USA). DNA fragment released with XbaI-PstI digestion from pHERDphrD was used as a template for PhrD probe preparation. The hybridized probes were immunodetected using anti-digoxygenin-AP, Fab fragments and visualized with the chemiluminescence substrate CSPD (Roche, USA) on X-ray films.
Real time PCR assays. Real time PCR assays were performed using total RNA isolated from cultures grown in Luria broth to an OD 600 of 2 as per the manufacturer's instructions (High Pure RNA isolation kit, Roche, Basel, Switzerland) to study the effect of altered levels of PhrD on rhlR expression. For expression analysis of PhrD under nitrogen limited or phosphate deficient conditions, RNA was extracted from cells grown to an OD 600 of 1.5 in MMP and PPGAS media respectively. Influence of PhrD on rhlR expression was studied by performing qRT-PCR. RNA was extracted from cells harvested every 2 h from WT and phrDΩGm cultures grown in Luria broth, PPGAS and MMP.  cDNA first strand synthesis was done using Genei RT-PCR kit (Bangalore Genei, Bangalore, India). Real time PCR was performed using Maxima SYBR Green/ Rox qPCR Master Mix (2×) (Thermo Scientific, Massachusetts, USA) and fold expression was determined after normalization with 16S rRNA gene by 2 −ΔΔCt method 34 . The WT strain consisting of empty vector pHERD30T was used as the calibrator for rhlR expression analysis whereas expression of PhrD in MMP and PPGAS medium was compared to that in Luria broth.
Rhamnolipid measurement. Culture supernatant from bacterial cultures grown in M9 minimal media for 24 h was adjusted to pH 2.5 ± 0.2 using 1 N HCl. The acidified sample was then extracted with 5 volumes of chloroform. 4 ml of chloroform extract was allowed to react with freshly prepared methylene blue solution containing 200 µl of 1 g/L of methylene blue (prepared in 10 mM borax buffer pH 10.5 and stabilized by adjusting pH to 5.5) and 4.9 ml of distilled water. The samples were mixed vigorously and allowed to stand. The absorbance of the chloroform phase was read at 638 nm with chloroform as blank with Beckman Coulter DU ® 720 spectrophotometer.
The absorbance values were normalized with A 600 of the cultures 35 . Each sample was analyzed in triplicates and results are average of three independent experiments. Pyocyanin assay. 5 ml of stationary phase grown cultures in Luria broth were extracted with 3 ml chloroform. The pyocyanin containing chloroform layer was acidified with 0.1 N HCl and absorbance of the pink colored acid fraction was measured at 520 nm. Pyocyanin levels were expressed as µg/ml/A 600 of culture supernatant 13 . The experiment was done thrice in triplicates.
β-galactosidase assay. 200 µl cultures were periodically withdrawn from cells growing in Luria broth, MMP or PPGAS media. β-galactosidase activity was determined as described previously after normalization with protein in mg 36 . The Miller units are represented as mean of three independent experiments.

Statistical analyses.
All the experiments were performed in triplicates and performed thrice (n = 3). Data analyses was carried out by either Paired t test or one way analysis of variance (ANOVA) followed by Tukey's multiple comparison test using Graph Pad Prism 6.0 (CA, USA).