RsaL is a self‐regulatory switch that controls alternative biosynthesis of two AHL‐type quorum sensing signals in Pseudomonas aeruginosa PA1201

Abstract Pseudomonas aeruginosa is a ubiquitous and metabolically versatile microorganism naturally found in soil and water. It is also an opportunistic pathogen in plants, insects, animals, and humans. In response to increasing cell density, P. aeruginosa uses two acyl‐homoserine lactone (AHL) quorum‐sensing (QS) signals (i.e., N‐3‐oxo‐dodecanoyl homoserine lactone [3‐oxo‐C12‐HSL] and N‐butanoyl‐homoserine lactone [C4‐HSL]), which regulate the expression of hundreds of genes. However, how the biosynthesis of these two QS signals is coordinated remains unknown. We studied the regulation of these two QS signals in the rhizosphere strain PA1201. PA1201 sequentially produced 3‐oxo‐C12‐HSL and C4‐HSL at the early and late growth stages, respectively. The highest 3‐oxo‐C12‐HSL‐dependent elastase activity was observed at the early stage, while the highest C4‐HSL‐dependent rhamnolipid production was observed at the late stage. The atypical regulator RsaL played a pivotal role in coordinating 3‐oxo‐C12‐HSL and C4‐HSL biosynthesis and QS‐associated virulence. RsaL repressed lasI transcription by binding the –10 and –35 boxes of the lasI promoter. In contrast, RsaL activated rhlI transcription by binding the region encoding the 5′‐untranslated region of the rhlI mRNA. Further, RsaL repressed its own expression by binding a nucleotide motif located in the –35 box of the rsaL promoter. Thus, RsaL acts as a molecular switch that coordinates the sequential biosynthesis of AHL QS signals and differential virulence in PA1201. Finally, C4‐HSL activation by RsaL was independent of the Las and Pseudomonas quinolone signal (PQS) QS signaling systems. Therefore, we propose a new model of the QS regulatory network in PA1201, in which RsaL represents a superior player acting at the top of the hierarchy.


INTRODUCTION
Many bacteria use chemical signals to communicate between cells in a process called quorum sensing (QS).QS allows bacteria to sense population density and coordinate their behavior through gene regulation 1 .Pseudomonas aeruginosa is a ubiquitous and metabolically versatile microorganism naturally found in soil and water.However, it is also an opportunistic pathogen in plants and animals, including insects and humans [2][3][4] .As a reflection of its versatile tropism and metabolism, P. aeruginosa possesses a complex QS network.The acyl-homoserine lactone (AHL)-and quinolonedependent QS (Pseudomonas quinolone signal [PQS]) systems have been extensively studied in the model strain PAO1, revealing a hierarchical network that regulates hundreds of genes in response to increasing cell density 5 .In AHL-dependent systems, two acyl-homoserine lactone (HSL) synthase enzymes, LasI and RhlI, are responsible for the biosynthesis of the QS signals N-3-oxo-dodecanoyl homoserine lactone (3-oxo-C12-HSL) and N-butanoyl-homoserine lactone (C4-HSL), respectively.3-oxo-C12-HSL and C4-HSL bind their cognate transcriptional factors LasR and RhlR to regulate the expression of downstream target genes 5,6 .Interestingly, many genes are exclusively regulated by one of the two acyl-HSLs, while others respond to both in the model strain PAO1 7 .In addition, QS activity in PAO1 is modulated by a number of regulatory factors, including Vfr, QteE, CdpR, the LuxR homologs QscR and VqsR, and RsaL, constituting a complex hierarchical regulatory network 5,[8][9][10] .
Intriguingly, when PAO1 is cultured in complex media, the 3-oxo-C12-HSL concentration plateaus during the late logarithmic growth phase, while the C4-HSL level increases continuously 8,11 .P. aeruginosa PA1201 was originally identified as a biocontrol strain in the rice rhizosphere 12 .Unlike the clinically isolated strain PAO1, PA1201 is less toxic to both human cell lines and Drosophila melanogaster 13 .The PA1201-derived strain UP46 has been used to industrially produce the biopesticide Shenqinmycin in China 13 .Our previous results on PA1201 confirmed a high level of C4-HSL in the late logarithmic growth phase 14 .However, how and why P. aeruginosa differentially regulates the production of two types of AHL QS signals, depending on its growth phase and specific environments, remain to be understood.
In PAO1, the small regulatory protein RsaL maintains 3-oxo-C12-HSL homeostasis.rsaL lies in the intergenic region between lasR and lasI and encodes a transcriptional regulator belonging to a subfamily of the tetrahelical superclass of helix-turn-helix proteins 8,15,16 .rsaL and lasI share an overlapping bidirectional promoter.RsaL represses lasI expression at high cell densities, which antagonizes LasR activity and limits 3-oxo-C12-HSL production to a physiological concentration in PAO1 8 .DNase I footprint analysis has identified a conserved RsaL-binding sequence (TATGnAAnTTnCATA) overlapping the -10 box of the lasI promoter (P lasI ) 8 .However, whether and how RsaL regulates the biosynthesis of another AHL-type QS signal, C4-HSL, remain unclear in P. aeruginosa.
Besides its role in QS, RsaL is also a global regulator controlling the expression of numerous genes involved in bacterial functions such as virulence, including the phz1 cluster, phzM, and cdpR, and in antibiotic resistance, including the narK1K2GHJI operon 8,14,[17][18][19] .Consequently, the rsaL mutation in the P. aeruginosa strain PAO1 enhances the production of virulence factors, such as elastase, hemolysins, hydrogen cyanide, and phenazine metabolites, increases ciprofloxacin and carbenicillin resistance and decreases the ability of the bacteria to cause chronic lung infections in mice 15,[19][20][21][22] .Therefore, understanding the RsaL-dependent regulation of gene expression may uncover new ways to control the harmful effects of P. aeruginosa in clinics and agriculture.
In this study, the rhizosphere P. aeruginosa strain PA1201 was used as a model to decipher the regulation of the two AHL QS signals, 3-oxo-C12-HSL and C4-HSL, and their effects on elastase and rhamnolipid biosynthesis during bacterial growth.We demonstrated that RsaL plays an essential role in the regulation of these alternative pathways and investigated the molecular mechanisms whereby RsaL represses its own and lasI expression and activates rhlI expression.These findings led to a new model in which RsaL acts as a molecular switch that coordinates alternative AHL QS signals in P. aeruginosa, which in turn triggers differential gene expression programs in response to cell density and environmental changes.
The levels of 3-oxo-C12-HSL and C4-HSL produced by PA1201 in PPM were further quantified using a previously described ultra-high performance liquid chromatographymass spectrometry (UPLC-MS) method 14 .This quantification confirmed that the 3-oxo-C12-HSL level reached a maximum at 24 hpi (0.77 μM/OD 600 ) and then declined progressively to 0.15 μM/OD 600 at 60 hpi (Figure 1C).In contrast, the C4-HSL level remained low at the early growth stage (i.e., between 12 and 24 hpi), increased rapidly from 36 hpi onwards, and peaked at 2.25 μM/OD 600 at 48 hpi (Figure 1C).These results suggest that PA1201 coordinates 3-oxo-C12-HSL and C4-HSL biosynthesis during growth and that this differential production operates a transcriptional switch during growth.

3-oxo-C12-HSL-and C4-HSL-dependent virulence factors are alternatively induced during growth
To test whether PA1201 can alternatively regulate gene expression through the control of 3-oxo-C12-HSL and C4-HSL biosynthesis, QS-regulated gene expression and virulence factor production were monitored during growth.In P. aeruginosa, lasB, which encodes the predominant protease elastase, is one of the main target genes activated by the LasI/LasR system 24 .We cloned the lasB promoter (P lasB ) and generated a lacZ fusion reporter strain, PA1201::P lasB -lacZ, to monitor lasB expression.This experiment showed that P lasB activity increased from 790 Miller units (M.U.) at 12 hpi to 1277.5 M.U. at 24 hpi and then decreased to 650 M.U. at 36 hpi (Figure 2A).Consistently, in the PA1201 culture, lasB-dependent elastase activity was significantly higher at 24 hpi (0.49 OD 450 ) than that at 48 hpi (0.24 OD 450 ) (Figure 2B).Thus, during PA1201 growth, the peaks of lasB expression and elastase activity coincided with the highest level of 3-oxo-C12-HSL.
The rhlAB operon encoding rhamnosyltransferase 1, an enzyme involved in the synthesis of the surfactant monorhamnolipid, is activated by the RhlI/RhlR system 25 .We generated the reporter strain PA1201::P rhlA -lacZ by cloning the promoter of rhlA (P rhlA ) upstream of the lacZ reporter gene to monitor rhlA expression.The peak of P rhlA activity was observed at 36 and 48 hpi (Figure 2C).Similarly, the relative rhamnolipid level at 24 hpi (0.14 OD 450 ) was significantly lower than that at 48 hpi (0.34 OD 450 ) (Figure 2D).Thus, the kinetics of rhlA expression and rhamnolipid production mirrored that of C4-HSL.Taken together, these results suggest that PA1201 coordinates the production of elastase at the early growth stage and of rhamnolipid at the late growth stage during growth.

RsaL is required for switching between the two HSL-dependent QS signals and between elastase versus rhamnolipid production in PA1201
To test the role of RsaL in the regulation of HSL-dependent QS signals, elastase activity, and rhamnolipid production, we created the rsaL in-frame deletion mutant ΔrsaL.In the PPM medium, this mutant produced more 3-oxo-C12-HSL than wild-type PA1201 during all growth phases (Figures 1C,D and 3A), but rsaL deletion had no significant effect on growth (Figure 1B).This result contrasted with the significant decrease in the 3-oxo-C12-HSL level observed after 24 hpi in PA1201 (Figures 1C,D and 3B).3-oxo-C12-HSL overproduction in ΔrsaL could be complemented by integrating a single copy of rsaL in ΔrsaL (Figure 3A,B), confirming that RsaL acts as a repressor of 3-oxo-C12-HSL production.
When we evaluated C4-HSL production in ΔrsaL cultured in PPM, we found that C4-HSL production was almost completely abolished at all growth stages (Figures 1C,D and 4A,B).C4-HSL production was restored by integrating a single copy of rsaL in ΔrsaL, demonstrating that RsaL is an essential activator of C4-HSL production (Figure 4A,B).These results suggest that RsaL is a key regulator of the alternative biosynthesis of 3-oxo-C12-HSL and C4-HSL during PA1201 growth.
Next, we studied the effect of rsaL deletion on elastase activity and rhamnolipid production.rsaL deletion resulted in a dramatic increase in lasB expression level at the early growth stage (12-36 hpi) compared to the wild-type expression level but had no significant effect at 48 hpi (Figure 2A).The increase in elastase activity caused by rsaL deletion at 24 hpi could be reversed by complementing ΔrsaL with a single copy of rsaL (Figure 2B).
At 12 and 24 hpi, rhlA expression in ΔrsaL and PA1201 was comparable.In contrast, at 36 and 48 hpi, ΔrsaL failed to upregulate rhlA expression, as in PA1201 (Figure 2C).Consistent with these results, at 48 hpi, rhlA-dependent rhamnolipid production was significantly lower in ΔrsaL than in PA1201 but not at 24 hpi (Figure 2D).Complementation of ΔrsaL with a single copy of rsaL fully restored rhamnolipid production to the wild-type level (Figure 2D).These results suggest that RsaL is also essential for controlling the alternative production of elastase and rhamnolipid during PA1201 growth.
RsaL binds the -10 and -35 boxes in the lasI promoter (P lasI ) to repress lasI transcription RsaL has been shown to repress 3-oxo-C12-HSL production by binding specifically an AT-rich region (TATGAAATTTG-CATA) within P lasI in P. aeruginosa PAO1 8 .In this study, we generated the reporter strains PA1201::P lasI -lacZ and ΔrsaL:: P lasI -lacZ and tested the effect of rsaL deletion on lasI expression in PA1201.Indeed, at 24 hpi, P lasI -dependent galactosidase activity significantly increased in ΔrsaL (Figure 3C), confirming that RsaL could repress lasI expression.
Using a DNA probe covering the lasI-rsaL intergenic region, a DNase I footprint assay identified an RsaL-binding site in P lasI (BS-1:ATACGTTTAAAGTATTAAAATATTGAT), which included a previously described RsaL-binding site (ATACGTTTAAAGTAT) in PAO1, and covered the -10 and -35 boxes of P lasI (Figure 3D).These results indicate that RsaL may repress lasI expression by binding BS-1.

RsaL activates rhlI transcription by binding a DNA region encoding the 5′-untranslated region (UTR) of rhlI mRNA
Since rsaL was essential to C4-HSL production (Figures 1D  and 4A,B), we hypothesized that rsaL could control C4-HSL production by activating rhlI.In keeping with this hypothesis, at 48 hpi, the rhlI promoter (P rhlI )-dependent galactosidase activity in ΔrsaL (67,595 M.U.) was significantly lower than that in wild-type PA1201 (138,547 M.U.) (Figure 4C).This result was confirmed at the protein level by western blot, using in-house generated polyclonal antibodies against RhlI to compare RhlI expression in wild-type PA1201, ΔrsaL, and ΔrsaL::rsaL (Figure 4D).
Electrophoretic mobility shift assay (EMSA) showed that RsaL could bind a 227-bp DNA probe, PRO rhlI , covering the rhlI promoter P rhlI (Figure 5A).Further DNase I footprint analysis identified a 26-bp RsaL-binding site, TGTGTGCTGG TATGTCCTCCGACTGA, within the probe PRO rhlI (Figure 5B,C).When this 26-bp DNA sequence was deleted from PRO rhlI to generate the variant probe PRO rhlI-Δ , RsaL lost the ability to bind PRO rhlI-Δ (Figure 5D).Previously, RsaL has been shown to bind the promoter region of several genes, including lasI, phzM, and phz1 8,14,18 .Multiple sequence alignment analyses of known RsaL binding sites identified an AT-rich core sequence (Figure 6A).Point mutations of the conserved TAT into CCC within the PRO rhlI resulted in a loss of RsaL-binding activity (Figure 6B).Consistently, point mutations of the conserved TAT into CCC in the PA1201 genome led to a loss in C4-HSL production in the mutated strain PA1201-CCC (Figure 6C).
As shown in Figure 5C, P rhlI has been well characterized 26 .The RsaL binding site identified in the present study is located within the DNA region encoding the 5′-UTR of rhlI mRNA (Figure 5C), raising the question of whether RsaL could be an mRNA-binding protein.To address this question, we generated Cy5-labeled single-stranded DNA or mRNA fragments corresponding to the probe PRO rhlI .EMSA analysis revealed that RsaL had no binding activity to single-stranded DNAs or mRNAs derived from PRO rhlI (Figure S1).Taken together, these results demonstrate that RsaL binds the double-stranded DNA sequence corresponding to the 5′-UTR of rhlI mRNA to positively regulate rhlI transcription.

RsaL regulation of C4-HSL biosynthesis is independent of the Las and PQS QS systems in PA1201
PA1201 possesses at least three QS signaling pathways that regulate its virulence and adaptation to different environmental conditions 27 .To determine whether the RsaL-dependent regulation of C4-HSL biosynthesis is regulated by the Las and PQS QS systems, we generated a range of QS genedeleted mutants.As expected, 3-oxo-C12-HSL depended on LasI activity, as the ΔlasI and the ΔlasIΔrsaL strains failed to   produce 3-oxo-C12-HSL (Figure 7A).In contrast, deleting pqsR had no effect on 3-oxo-C12-HSL biosynthesis.Deletion of rsaL in the ΔpqsR or ΔpqsRΔrhlI strains significantly increased 3-oxo-C12-HSL biosynthesis (Figure 7A).These results suggest that RsaL-dependent repression of 3-oxo-C12-HSL biosynthesis is independent of the Rhl and PQS pathways.
ΔlasI produced a similar level of C4-HSL to that produced by wild-type PA1201, while ΔrsaLΔlasI produced a similar level of C4-HSL to that produced by ΔrsaL (Figure 7B).In the same way, deleting pqsR, a key regulator of the PQS pathway, had no effect on C4-HSL biosynthesis.Deleting rsaL in the strains ΔpqsR or ΔpqsRΔlasI resulted in disrupted C4-HSL biosynthesis (Figure 7B).These results suggest that the RsaL-dependent regulation of C4-HSL biosynthesis is independent of the Las and PQS signaling systems.

RsaL acts as an auto-repressor in PA1201
To monitor rsaL expression during growth, the rsaL promoter (P rsaL ) was cloned upstream of the lacZ coding sequence to generate the reporter plasmid P rsaL -lacZ subsequently integrated into wild-type PA1201 and ΔrsaL to generate the reporter strains PA1201::P rsaL -lacZ and ΔrsaL::P rsaL -lacZ (Figure 8A).PA1201::P rsaL -lacZ displayed P rsaL -dependent galactosidase activity that was significantly higher at 24 hpi (741.5 M.U.) and 36 hpi (602 M.U.) than at 12 hpi (275.5 M.U.) and 48 hpi (324 M.U.) (Figure 8B).To confirm the correlation between the gene reporter activity and the presence of the RsaL protein, we generated polyclonal antibodies (anti-RsaL) against a recombinant RsaL protein produced in Escherichia coli and purified by chromatography (Figure S2).Using this new antibody, we confirmed that the RsaL protein level at 24 and 36 hpi was significantly higher than that at 12 and 48 hpi when PA1201 was cultured in PPM (Figure 8C).

RsaL binds to the −35 box of the rsaL gene promoter P rsaL
To study the molecular mechanism of RsaL self-repression, first, we characterized P rsaL by rapid amplification of cDNA ends (RACE) analysis.This experiment identified the transcription start site (+1, C) in P rsaL (Figure 9A).Further promoter prediction using the SAPPHIRE program identified the -10 box (TTGCTA) and the -35 box (GATAGA) in P rsaL (Figure 9A).The functionality of this predicted promoter was tested by mutagenesis.The -10 box TTGCTA was mutated into CCGCTA on the PA1201 chromosome.The resulting mutant, PA-10M, displayed a phenotype similar to ΔrsaL (i.e., increased 3-oxo-C12-HSL biosynthesis and disrupted C4-HSL production; Figure 9B).These results confirmed that the predicted -10 box was constitutive of P rsaL (Figure 9A).
Next, we investigated the potential binding sites of RsaL on P rsaL using a DNase I protection footprint assay on both strands of a DNA fragment encompassing the entire rsaL-lasI intergenic region.In addition to BS1 (Figure 3D), this analysis identified a second binding site, BS-2 (AAATGAGATAGATTTC), which could correspond to the RsaL binding site responsible for P rsaL regulation.BS-2 partially overlapped the -35 box of P rsaL and contained a palindromic sequence (Figure 9C).This symmetrical motif suggests that RsaL could bind P rsaL as a dimer.The functionality of this binding site was assessed in EMSA experiments using the Cy5-labeled probe I (PRO-I), excluding BS-1 and BS-2, and the Cy5-labeled probe II (PRO-II), encompassing BS-2 and flanking sequences but excluding BS-1 (Figure 9A).RsaL bound only PRO-II and not PRO-I (Figure 9C).Mutation of the BS-2 palindromic sequence ATTTC into GCCCC (PRO-II-p) disrupted RsaL binding to PRO-II (Figure 9C), demonstrating that this DNA motif was responsible for BS-2 recognition by RsaL.

DISCUSSION
The P. aeruginosa strain PA1201 has become a model rhizosphere bacterium for studying the regulation of PCA or PCN biosynthesis [12][13][14] .Our previous research showed that PA1201 possesses all three QS pathways and differentially regulates PCA biosynthesis 27 .Here, we further show that when PA1201 is cultured in PPM, the 3-oxo-C12-HSL level plateaus at 24 hpi and then drops.In contrast, the C4-HSL level is low at 12 hpi and starts to increase from 24 hpi onward, reaching peak production at 48 hpi (Figure 1).These alternative productions are correlated with the highest elastase activity at 24 hpi and the highest rhamnolipid level at 48 hpi (Figure 2).These results are consistent with previous findings in PAO1, in which the 3-oxo-C12-HSL level plateaued in the late logarithmic growth phase, whereas the C4-HSL level continued to increase 8 .Thus, P. aeruginosa has the ability to coordinate the alternative biosynthesis of 3-oxo-C12-HSL and C4-HSL during growth.To explore the underlying mechanism of this regulation, Sun et al. 14 proposed a hypothesis based on substrate competition.Since the lactone rings of 3-oxo-C12-HSL and C4-HSL both derive from the same precursor, S-adenosylmethionine, excessive production of 3-oxo-C12-HSL at the early growth stage likely drains the pool of S-adenosylmethionine necessary for C4-HSL biosynthesis, thus leading to reduced C4-HSL production 14 .However, the present study demonstrated for the first time that the transition from 3-oxo-C12-HSL to C4-HSL QS signals is regulated by an intrinsic mechanism involving the global regulator RsaL.The RsaL-dependent regulation of 3-oxo-C12-HSL and C4-HSL biosynthesis is independent of the Las and PQS QS systems (Figure 7).Nevertheless, we could not rule out the minor role of substrate competition in the regulation of 3-oxo-C12-HSL and C4-HSL biosynthesis in PA1201.
Previous results have shown that RsaL is required for 3-oxo-C12-HSL homeostasis 8 .The present study further suggests that RsaL is also essential for C4-HSL homeostasis.Deletion of rsaL abolished C4-HSL biosynthesis in PA1201 (Figures 1 and 4).RsaL activated rhlI expression by binding a DNA region encoding the 5′-UTR of rhlI mRNA (Figure 5).However, we could not find evidence that RsaL could directly interact with single-stranded DNAs or mRNAs corresponding to rhlI 5′-UTR.How RsaL could have antagonistic effects on DNA binding and its biological significance deserves further investigation.Thomason et al. 26 conducted a term-seq analysis of PAO1 and identified an sRNA named RhlS, which is derived from the 5′-UTR of rhlI.RhlS is required for the production of normal C4-HSL levels by promoting rhlI translation.An antisense RNA (asRhlS) to the 5′-terminal of the rhlI open reading frame has also been identified in PAO1.RhlS may act antagonistically to the asRhlS to regulate rhlI translation 26 .The experiments in the present study could not detect binding activity between RsaL and single-stranded rhlI fragments or rhlI mRNA (Figure S1).Whether RsaL regulates the transcription of RhlS or asRhlS and whether two small RNAs mediate the RsaL regulation of 3-oxo-C12-HSL and C4-HSL biosynthesis deserve further study.
Taking these results together, we propose a new hypothetical model of QS regulation, integrating RsaL autorepression and RsaL-dependent regulation of 3-oxo-C12-HSL and C4-HSL biosynthesis in PA1201.During the early growth phase (Figure 10A), rsaL is transcribed at a basal low level, RsaL binds the -35 box of its own promoter, which limits rsaL transcription, no RsaL binds the P lasI promoter, and lasI transcription is induced via a 3-oxo-C12-HSL-dependent LasR system.lasB transcription is induced by the Las QS system, which promotes high elastase activity.rhlI is transcribed at a basal level that is not sufficient to induce C4-HSL-dependent rhamnolipid biosynthesis.During the late growth phase (Figure 10B), a signal molecule X, as yet to be identified, which informs the bacteria that a population threshold has been reached, could serve as a ligand for RsaL.The RsaL/X complex could release RsaL from P rsaL , which would increase rsaL transcription and result in a high level of RsaL.In parallel, RsaL/ X complexes could outcompete the LasR/3-oxo-C12-HSL complex for binding to P lasI and repress lasI transcription.The RsaL/X complex could also bind the DNA region encoding the 5′-UTR of the rhlI promoter, inducing rhlI transcription, which in turn could induce the transcription of the rhlAB cluster for rhamnolipid production.

Bacterial strains and growth conditions
All the bacterial strains used in this study are listed in Table S1.Unless stated otherwise, PA1201 and all isogenic mutants were grown at 28°C in 50 ml of PPM (22 g/l tryptone, 20 g/l glucose, 5 g/l KNO 3 , pH 7.5) in 250-ml flasks.E. coli strains were grown in shake culture at 37°C in Luria-Bertani broth.When required, the following antibiotics were added to the media: spectinomycin (Spe, 50 µg/ml), kanamycin (Kan, 50 µg/ml), and gentamycin (Gen, 100 µg/ml for P. aeruginosa strains and 20 µg/ml for E. coli strains).

DNA engineering and generation of in-frame deletion mutants
DNA preparation and agarose gel electrophoresis were performed following the protocols described by Sambrook and Russell 28 .The plasmids and oligonucleotides used in this study are listed in Tables S2 and S3.The gene deletion mutants of PA1201 were generated as previously described 29 .Briefly, the ∼500 bp upstream and downstream regions of the target gene were fused by PCR.The PCR product was then subcloned into the plasmid pK18mobsacB.The resultant recombinant plasmid was further integrated within the target gene of XC1.The resultant strain was finally plated on an NYG agar plate with 50 μg/ml Spe and 5% (w/v) sucrose.The generated mutants were verified by colony PCR and subsequent DNA sequencing.Single-copy gene complementation was performed following the protocol described by Jittawuttipoka et al. 30 Briefly, the fragment corresponding to the promoter and coding region of a target gene was amplified and cloned into a mini-Tn7T-Gm transposon, which was subsequently integrated into PA1201 at the neutral site attTn7.

Construction of the promoter-lacZ fusion reporter strain and β-galactosidase assays
Construction of lacZ-dependent reporter strains in PA1201 was conducted following a previously described protocol 25 .Briefly, approximately 500 bp promoter regions with 30 bp coding sequences from the target genes were PCR-amplified and cloned into mini-CTX-lacZ vectors 31 .The resulting plasmids were integrated into PA1201-derived strains.β-Galactosidase activity of the constructed reporter strains was determined according to the protocol described by Miller 28 .

Protein expression and purification
RsaL protein expression and purification has been previously described 14 .Briefly, the rsaL gene was amplified by PCR and then cloned into the plasmid pET28a.The resulting plasmid was introduced into E. coli BL21 (DE3, pLysS).RsaL expression was induced with 0.1 mM isopropyl β-D-1thiogalactopyranoside at OD 600 = 0.6, and the culture was grown in a 16°C shaker for 12 h.The recombinant protein was purified by Ni 2+ -affinity chromatography according to the manufacturer's instructions.Briefly, bacterial cells were lysed in the lysis buffer with 10 mM Tris-buffer saline, 100 mM NaCl, 100 mM (NH 4 ) 2 SO 4 , and 10 mM imidazole, pH 7.2.Histagged RsaL proteins were eluted in an elution buffer with 10 mM Tris-buffer saline, 100 mM NaCl, 100 mM (NH 4 ) 2 SO 4 , and 250 mM imidazole, pH 7.2.The eluted RsaL protein was finally loaded into a Superdex 200 gel filtration column (GE Healthcare) and equilibrated with a nonimidazole buffer containing 10 mM Tris-HCl, 150 mM NaCl, 10% (v/v) glycerol, pH 7.5.

EMSA
EMSA was performed using the Thermo Scientific Light Shift Chemiluminescent EMSA Kit.Briefly, the probe DNA fragments were biotin-labeled at the 5′-end.The labeled probes were then incubated with RsaL protein in binding buffer (10 mM Tris-HCl, 1 mM ethylenediaminetetraacetic acid, 5 mM MgCl 2 , 50 mM KCl, 2.5% [v/v] glycerol, 30 μg/ ml poly(dI-dC), 0.05% [v/v] NP-40, pH 8.0).After incubation at room temperature for 20 min, the reaction mixtures were loaded on a 5% (w/v) polyacrylamide gel under nondenaturing conditions.The PAGE gel was then submitted for scanning for fluorescent DNA using a Starion FLA-9000 Scanner (FujiFilm).

DNase I protection footprint sequencing
DNase I protection footprint sequencing was performed at Shanghai Biotechnology Corporation.Briefly, the fluorescent probes were PCR amplified.The products were then purified using a PCR Clean-Up System (Promega).100 ng of probe was incubated with different amounts of RsaL in a total volume of 40 µl reaction mixtures.After incubation for 30 min at room temperature, 10 µl of the solution containing 0.015 units of DNase I (Promega) and 100 nM CaCl 2 was added, and the reaction mixtures were further incubated for 1 min at 25°C.The reaction was terminated by adding 140 µl of stop solution.Samples were extracted with phenol/chloroform, precipitated with ethanol, and the pellets were dissolved in 30 µl of water.Gel electrophoreses were performed and the data were analyzed using a 3130XL DNA analyzer and the Peak Scanner Software v1.0 (Applied Biosystems).

Quantitative determination of 3-oxo-C12-HSL and C4-HSL levels using biosensor strains
The synthesis of C4-HSL and 3-oxo-C12-HSL in PA1201 and derived mutants was assessed by semiquantitative diffusion plate assay using C. violaceum strain CV026 and A. tumefaciens CF11 biosensors, respectively, as previously described 23 .Purple or blue spots indicated that the diffusible QS signal molecules were detected by the biosensor strains.The relative level of C4-HSL or 3-oxo-C12-HSL was proportional to the diffusion distance of the QS molecules in PPM cultures of their respective sensor strains.
Extraction and quantification of 3-oxo-C12-HSL and C4-HSL by UPLC-MS 3-oxo-C12-HSL and C4-HSL extraction and quantification were performed following a previously described Figure 10.Model for the RsaL-dependent control of differential AHL QS-dependent virulence in Pseudomonas aeruginosa PA1201.During the early growth phase (A), rsaL is transcribed at a low level; RsaL binds the -35 box and represses its own transcription; no RsaL binds the P lasI promoter, and lasI is activated by the 3-oxo-C12-HSL-dependent LasR system, while lasB transcription and elastase activity are activated by the Las QS system.rhlI is transcribed at a low level, which is not enough to promote C4-HSL-dependent rhamnolipid biosynthesis.During the late growth phase (B), a signal molecule X, as yet to be identified, informs the bacteria that a population threshold has been reached, could bind RsaL and release it from P rsaL , thereby increasing rsaL transcription.In parallel, RsaL/X complexes could outcompete LasR/3-oxo-C12-HSL complex on P lasI and repress lasI transcription.RsaL/X complex could also bind to the DNA region encoding the 5′-UTR of the rhlI promoter, inducing rhlI transcription, which in turn induces the transcription of the rhlAB cluster for rhamnolipid production.QS, quorum sensing.protocol 14 .Briefly, 270 µl of culture with an OD 600 of 1.8-6.9 was acidified to pH 4.0 with 6 M HCl, and extracted with an equal volume of ethyl acetate.A total of 100 µl of ethyl acetate extract was evaporated at 40°C, and the resulting residue was dissolved in 500 µl of methanol.Then 10 μl of the extract was then injected into an UPLC-MS (Agilent UPLC1290-TOF-MS6230), under the following conditions: Agilent Zorbax XDB C18 reversephase (5 µm, 4.6 × 150 mm) system separated by gradient ACN with 0.5% acetic acid and H 2 O with 0.5% (v/v) acetic acid at 0.4 ml/min.The MS analysis was performed under positive mode.The concentration of 3-oxo-C12-HSL and C4-HSL was quantified using the peak area (A) of the specific extracted ion chromatogram in the total ion chromatogram according to the established formula 14 .

Western blot analysis
RsaL proteins were electrotransferred onto a polyvinylidene difluoride membrane (Roche).After blocked with 5% (w/v) nonfat milk powder, the membranes were incubated with the rabbit polyclonal antibodies against RsaL at a 1:3000 dilution.The membrane was then washed three times with TBST buffer (20 mM Tris, 0.15 M NaCl, and 0.1% [v/v] Tween 20).A horseradish peroxidase-conjugated goat antirabbit IgG (#M21001; Abmart) diluted at 1:6500 was used as the secondary antibody.After membrane washing, the luminescent signal was detected using an ECL kit and a ChampChemi 610 Plus instrument (Sage Creation Science).

Mapping of rsaL transcription start site
The transcription start site of rsaL was identified using the 5′-RACE system (Invitrogen).Briefly, total RNA was prepared from PA1201 PPM culture at 24 hpi.The cDNA synthesis was performed following the manufacturer's protocol.After cDNA synthesis, RNase was added to remove the residual RNA.An oligo-dC tail was added to the generated cDNA.The dCtailed cDNA was amplified by PCR and was then cloned into the pGEM-T Easy vector for sequencing.

Statistical analysis
All of the experiments were performed in triplicate.The statistical significance of the differences observed in mean invasion frequency was determined by calculating the p values using the two-tailed Student t-test for unpaired data sets.

Figure 2 .
Figure 2. Monitoring of lasB and rhlA expression and corresponding elastase activity and rhamnolipid production in PA1201-derived strains.(A) Relative transcriptional activity of the LacZ reporter gene under the control of the lasB promoter in the strains PA1201::P lasB -lacZ and ΔrsaL:: P lasB -lacZ in Miller units (M.U.) at different growth time points.(B) Elastase activity in the strains PA1201, ΔrsaL, and ΔrsaL::rsaL at 24 and 48 hours postinoculation (hpi).(C) Relative transcriptional activity of the LacZ reporter gene under the control of the rhlA promoter in the strains PA1201::P rhlA -lacZ and ΔrsaL::P rhlA -lacZ at different growth time points.(D) Rhamnolipid levels in the strains PA1201, ΔrsaL, and ΔrsaL::rsaL at 24 and 48 hpi.Three independent experiments were conducted; averages and standard deviations are shown.Statistically significant differences are indicated by one asterisk (p < 0.05).M.U., Miller units.

Figure 3 .
Figure 3. RsaL represses lasI transcription and 3-oxo-C12-HSL biosynthesis.(A) Relative 3-oxo-C12-HSL levels in the strains PA1201, ΔrsaL, ΔrsaL::rsaL, and ΔlasL evaluated with the biosensor strain CF11.(B) 3-oxo-C12-HSL quantification in PA1201, ΔrsaL, and ΔrsaL::rsaL cultures by UPLC-MS.(C) Relative transcriptional activity of the LacZ reporter gene under the control of the lasI promoter in the strains PA1201 and ΔrsaL at 24 hours postinoculation, measured as β-galactosidase activity expressed in M.U.(D) Mapping of two RsaL-binding sites (BS), with BS-1 in the promoter P lasI and BS-2 in the promoter P rsaL in the intergenic region between lasI and rsaL, by a DNase I footprint protection assay.The BS-1 sequence corresponding to the RsaL binding sites in P lasI is shown in blue.Three independent experiments were conducted; averages and standard deviations are shown.Statistically significant differences are indicated by one asterisk (p < 0.05).

Figure 4 .
Figure 4. RsaL positively regulates C4-HSL biosynthesis and rhlI expression.(A) Relative C4-HSL levels in the strains PA1201, ΔrsaL, ΔrsaL::rsaL, and ΔrsaL(rsaL) evaluated with the biosensor strain CV026.(B) C4-HSL quantification of PA1201, ΔrsaL, and ΔrsaL::rsaL cultures at different growth time points, measured by UPLC-MS.(C) Relative transcriptional activity of the LacZ reporter gene under the control of the rhlI promoter in the strains PA1201 and ΔrsaL at 48 hours postinoculation, measured as β-gal activity expressed in M.U.(D) RhlI protein levels in the strains PA1201, ΔrsaL, and ΔrsaL::rsaL.Three independent experiments were conducted; averages and standard deviations are shown.Statistically significant difference is indicated by one asterisk (p < 0.05).

Figure 5 .
Figure 5. RsaL directly binds a DNA region encoding the 5′ untranslated region of rhlI mRNA.(A) Electrophoretic mobility shift assay (EMSA) results showing that RsaL binds the probe PRO rhlI , encompassing the rhlI promoter.2000 ng of unlabeled probe was added as a specific competitor.(B) The RsaL binding site in PRO rhlI identified by a DNase I protection assay.(C) DNA sequence of the rhlI promoter region; the RsaL-binding site is shown in red and is located in the region encoding the 5′ untranslated region of rhlI mRNA.(D) EMSA results comparing RsaL-binding capacity to wild-type PRO rhlI and PRO rhlI-Δ , carrying a deletion of the RsaL putative binding site in the rhlI promoter.EMSA, electrophoretic mobility shift assay.

Figure 6 .
Figure 6.Functional characterization of the RsaL binding site in the DNA region encoding the 5′-UTR of rhlI mRNA.(A) Alignment analysis of the rhlI promoter (P rhlI ) with other gene promoters containing an RsaL-binding motif and identification of the conserved residues defining a consensus RsaL binding site.(B) Impairment of RsaL binding to the probe PRO rhlI mutated in the RsaL putative binding site (PRO rhlI-ccc ), demonstrated by EMSA.(C) Relative C4-HSL levels in wild-type PA1201 and PA1201 carrying the CCC mutation in P rhlI (PA1201-CCC) evaluated by the biosensor strain CV026.

Figure 8 .
Figure 8. Kinetics of rsaL expression during PA1201 growth.(A) Verification of the reporter strains PA1201::P rsaL -lacZ and ΔrsaL::P rsaL -lacZ on the PPM agar plate supplemented or not with the β-galactosidase substrate X-gal.(B) Kinetics of rsaL expression during PA1201 growth, estimated by measuring β-galactosidase activity expressed in M.U.Data are expressed as means ± standard deviations obtained in three independent assays.(C) Kinetics of RsaL protein production in PA1201 grown in PPM medium assessed by western blot using in-house polyclonal antibodies against RsaL (upper blot) and a commercially available antibody against the α-subunit of RNA polymerase as a control for sample loading (lower blot).

Figure 9 .
Figure 9. RsaL binding in the rsaL promoter (P rsaL ).(A) Analysis of the promoters located in the intergenic region between lasI and rsaL in PA1201.The lasR, rsaL, and lasI genes are depicted as arrows.The corresponding sequence is annotated for the RsaL binding sites and the -10 and -35 boxes of the rsaL and lasI promoters.(B) Assessment of 3-oxo-C12-HSL (left) and C4-HSL (right) production using the biosensors CF11 (left) and CV026 (right) in wild-type PA1201 and PA-10M mutated in the -10 box of the rsaL promoter (P rsaL ).(C) EMSA results showing the functionality of the RsaL binding site in BS-2.The probes PRO-I and PRO-II are shown in (A).PRO-II-p corresponds to PRO-II, with a point mutation in the RsaL binding site 2 located in the -35 box of P rsaL .