Quorum Sensing N-acyl Homoserine Lactones-SdiA Suppresses Escherichia coli-Pseudomonas aeruginosa Conjugation through Inhibiting traI Expression

Conjugation is a key mechanism for horizontal gene transfer and plays an important role in bacterial evolution, especially with respect to antibiotic resistance. However, little is known about the role of donor and recipient cells in regulation of conjugation. Here, using an Escherichia coli (SM10λπ)-Pseudomonas aeruginosa (PAO1) conjugation model, we demonstrated that deficiency of lasI/rhlI, genes associated with generation of the quorum sensing signals N-acyl homoserine lactones (AHLs) in PAO1, or deletion of the AHLs receptor SdiA in the donor SM10λπ both facilitated conjugation. When using another AHLs-non-producing E. coli strain EC600 as recipient cells, deficiency of sdiA in donor SM10λπ hardly affect the conjugation. More importantly, in the presence of exogenous AHLs, the conjugation efficiency between SM10λπ and EC600 was dramatically decreased, while deficiency of sdiA in SM10λπ attenuated AHLs-inhibited conjugation. These data suggest the conjugation suppression function of AHLs-SdiA chemical signaling. Further bioinformatics analysis, β-galactosidase reporter system and electrophoretic mobility shift assays characterized the binding site of SdiA on the promoter region of traI gene. Furthermore, deletion of lasI/rhlI or sdiA promoted traI mRNA expression in SM10λπ and PAO1 co-culture system, which was abrogated by AHLs. Collectively, our results provide new insight into an important contribution of quorum sensing system AHLs-SdiA to the networks that regulate conjugation.


INTRODUCTION
The acquisition of antibiotic resistance by pathogenic microorganisms is a threat to public health worldwide. Horizontal gene transfer, especially conjugative transfer of plasmids that carry resistance genes, is the primary cause of bacterial antibiotic resistance and-on the larger scalebacterial evolution (Zatyka and Thomas, 1998;Arthur et al., 2011). The self-transmissible plasmids, such as the well-studied fertility F-plasmids and IncP plasmid RP4 (also known as RK2), generally present a mobilization (MOB) region which includes the origin of transfer (oriT) and the relaxase gene. The relaxase, identified as being TraI in RP4, initiates conjugation by cleaving the oriT in a site-and strandspecific manner (Carballeira et al., 2014). Other plasmids, termed mobilizable, are incapable of initiating conjugation, but can transfer by using the conjugative apparatus of another plasmid (Zatyka and Thomas, 1998). Mobilizable plasmids are more frequently found in natural environment; therefore, replication and mobilization can be considered as important mechanisms that influence plasmid promiscuity (Fernández-López et al., 2014).
Many Gram-negative bacteria utilize N-acyl homoserine lactones (AHLs) as signal molecules to enable individual bacteria to coordinate their behavior in populations; such quorum sensing (QS) enables bacteria to not only sense members of their own species but other species as well (Smith et al., 2011). The essential constituents of QS include a signal producer, or synthase, and a cognate transcriptional regulator that responds to the accumulated signal molecules (Bassler and Losick, 2006). The opportunistic animal and plant pathogen Pseudomonas aeruginosa possesses one of the best-studied models of QS, and two different AHL systems, las and rhl, have been identified (Wagner et al., 2003). In the las QS system, the lasI gene product directs formation of the diffusible extracellular signal N-(3-oxododecanoyl)-L-HSL (3-oxo-C12-HSL), which interacts with LasR to activate a number of virulence genes including the LasA and LasB elastases, exotoxinA, and alkaline protease (Toder et al., 1991;Gambello et al., 1993;Jones et al., 1993;Passador et al., 1993). In the rhl system, the rhlI gene product catalyzes the synthesis of N-butanoyl-L-HSL (C4-HSL). This diffusible signaling molecule, together with RhlR, activates directly some virulence genes like those encoding rhamnolipids and pyocyanin, and represses those genes responsible for assembly and function of the type III secretion system (Bleves et al., 2005;Jimenez et al., 2012). Besides the fact that the las and rhl systems are hierarchically connected, both rhlR and rhlI are positively regulated by the las system (Wagner et al., 2003). The roles of QS in diverse biological processes, such as virulence, biofilm formation and metabolism in P. aeruginosa have attracted research attention (Pearson et al., 1994;Hassett et al., 1999;Whiteley et al., 1999;García-Contreras, 2016). However, as the cell-to-cell communication system, the influence of QS on interspecies conjugation remains largely unknown.
Some organisms, such as Escherichia, Klebsiella, Salmonella and Shigella lack AHL synthase and therefore do not produce AHLs; however, they possess a LuxR homolog known as SdiA that can bind AHLs produced by other microorganisms and affect gene expression (Smith and Ahmer, 2003;Yao et al., 2006;Sabag-Daigle et al., 2015). Case et al. described the phenomenon of non-AHLs-producing microorganisms binding and utilizing AHLs produced by other organisms as eavesdropping (Case Abbreviations: QS, quorum sensing; AHLs, N-acyl homoserine-lactones; P. aeruginosa, Pseudomonas aeruginosa; E. coli, Escherichia coli; qPCR, quantitative real-time PCR. et al., 2008). Although SdiA can bind to DNA and regulate transcription in the absence of AHLs, the structural studies of SdiA suggest a double mode of action for AHLs on SdiA activity, by increasing both protein stability and DNA-binding affinity (Nguyen et al., 2015;Ishihama et al., 2016). Besides, a neighborjoining tree analysis revealed that SdiA of E. coli did not cluster with the LuxR homologs found in other enterobacterial species, but was closely related to the RhlR of P. aeruginosa (Gray and Garey, 2001).
Herein, we clarified the effect of QS on conjugation and investigated the underlying mechanisms by employing a mobilizable plasmid and E. coli-P. aeruginosa conjugation model. We found that QS signal molecules produced by P. aeruginosa inhibited interspecies conjugation by activating E. coli SdiA, resulting in decreased mRNA expression of traI in E. coli. Blockade of AHL-SdiA signaling using strains deficient in lasI, rhlI or sdiA significantly enhanced conjugative transfer. These findings provide new insight into the regulatory networks of conjugation, and offer novel potential targets for antibiotic resistance.

Growth Curves
The indicated bacterial strains were cultured in LB overnight (8∼10 h) at 37 • C, then diluted to 0.5 MCF (McFarland standard) and 3 mL cultures were grown at 37 • C with shaking at 200 rpm. The samples were collected at the indicated time points and the OD 600 values were determined.

Plasmid Construction
The plasmid pUCP24T was constructed by inserting the oriT fragment into pUCP24 (West et al., 1994), which contains a gene cassette (aacC1) conferring gentamycin resistance in recipient cells. As a result, pUCP24T is not able to transfer on its own, but can transfer by using the conjugative apparatus of E. coli SM10λπ. Details of construction of the plasmids used to delete sdiA gene or express SdiA are described in the Supporting Materials and Methods.

Construction of PAO1 lasI or rhlI and E. coli SM10λπ sdiA Deficient Mutants
The phage λ Red recombination system was employed for sdiA deletion in E. coli SM10λπ, while the sacB-based suicide vector system was adapted for knockout of lasI or rhlI in PAO1 (Zeng et al., 2016); further details are provided in the Supporting Materials and Methods.

Conjugation Experiments
For the conjugation assays, the same amount (0.5 × 10 7 CFU/mL, counted using the Sysmex UF-1000i TM Automated Urine Particle Analyzer; Tokyo, Japan) of mid-logarithmic phase donor (E. coli SM10λπ harboring plasmid pUCP24T) and recipient cells (PAO1 or EC600) were mixed in 200 µL LB with or without the indicated HSLs in 96-well plates. After 6 h mating at 37 • C, the cultures were vigorously mixed and 30 µL aliquots of each conjugation mixture were spread on LB agar containing 30 µg/mL Gm plus 100 µg/mL Amp for SM10λπ-PAO1 or 30 µg/mL Gm plus 50 µg/mL Rif for SM10λπ-EC600 transconjugants. The numbers of transconjugant colonies were counted after overnight incubation at 37 • C.

Quantification of HSLs by HPLc-MS/MS
Supernatants of PAO1, PAO1 lasI, and PAO1 rhlI cultures were collected for HPLC-MS/MS detection of HSLs; full details are provided in the Supporting Materials and Methods.
β-Galactosidase Assays β-Galactosidase activities were performed on cells in the mid-log phase of growth according to the modified Miller's method (Giacomini et al., 1992). All tests were performed in triplicate.

Electrophoretic Mobility Shift Assays (EMSA)
His-SdiA fusion protein was expressed in E. coli BL21 (DE3) and purified via Ni-chelating affinity chromatography. Gel shift assays were carried out using the Lightshift Chemiluminescent EMSA kit according to the manufacturer's instructions (Thermo Scientific, Waltham, MA, USA), details are provided in the Supporting Materials and Methods.

Real-Time PCR
Total RNA was extracted using total RNA isolation reagent (Promega, Madison, WI, USA). Reverse transcription (1 µg of total RNA) was performed with the PrimeScript RT reagent Kit (Takara, Dalian, Liaoning, China). The cDNA was subjected to qPCR on a ViiA TM 7 Dx system (Applied Biosystems, Foster, CA, USA) using SYBR Green qPCR Master Mixes (ThermoFisher Scientific). The expression levels of the target genes were normalized to the expression of the internal control gene (rpoD), using the 2 − Ct method. The sequences of the primers are listed in Table S1.

Statistical Analysis
Data are expressed as the mean ± standard error of the mean (SEM) of at least three independent experiments. The differences between groups were analyzed using the Student's t-test when two groups were compared or one-way ANOVA when more than two groups were compared. All analyses were performed using GraphPad Prism, version 5 (GraphPad Software, Inc., San Diego, CA, USA). All statistical tests were two-sided; P < 0.05 was considered statistically significant.

Deficiency of lasI or rhlI in P. aeruginosa Promotes SM10λπ-PAO1 Conjugation
To elucidate the biological significance of the QS system in P. aeruginosa conjugation, we first constructed lasI or rhlI single gene-deficient mutants, named PAO1 lasI and PAO1 rhlI, respectively. In P. aeruginosa, lasI catalyzes the formation of 3-oxo-C12-HSL, which positively regulates the expression of RhlI. RhlI directs the synthesis of C4-HSL, which subsequently regulates pyocyanin production (O'Loughlin et al., 2013). In this study, despite the existence of rhlI in the genome of PAO1 lasI, both 3-oxo-C12-HSL and C4-HSL were barely detectable in the conditioned medium of this mutant strain using HPLC-MS/MS analysis. For PAO1 rhlI, the deficiency of rhlI in the genome led to an absence of C4-HSL in the conditioned medium of this mutant strain, whereas lasI and its product 3-oxo-C12-HSL were present at similar levels as the WT strain ( Figure 1A and Figure  S1). Furthermore, as a result of mutation of the QS system, both PAO1 lasI and PAO1 rhlI lost the ability to express pyocyanin, which could be rescued by exogenous addition of 3-oxo-C12-HSL or C4-HSL ( Figure 1B). Taken together, these results confirmed the successful creation of PAO1 strains deficient in lasI or rhlI.
We subsequently examined the growth and conjugation ability of PAO1 lasI and PAO1 rhlI. Compared to the WT strain, deficiency of lasI or rhlI hardly affected the growth of PAO1 ( Figure S2), but significantly promoted SM10λπ-PAO1 conjugation ( Figure 1C). Furthermore, exogenous 3-oxo-C12-HSL or C4-HSL attenuated the interspecies conjugation ability of PAO1 lasI and PAO1 rhlI (Figure 1C). What's more, we counted the amount of donor SM10λπ after co-culture with PAO1, PAO1 lasI or PAO1 rhlI, and found that there is no difference among the three groups ( Figure S3), indicted that the observed effect of quorum sensing on conjugation efficiency was not due to the growth suppressive effect on SM10λπ. These data suggested that the QS system may negatively regulate SM10λπ-PAO1 conjugation.
The Quorum Sensing System of P. aeruginosa Inhibits Conjugation by Activating SdiA of E. coli It is well recognized that AHLsq regulate gene transcription via binding to their receptor proteins (LuxR-like proteins). In this conjugation model, in contrast to the recipient cells PAO1, the donor E. coli SM10λπ cells lack AHL synthase and therefore do not produce AHLs; however, these cells produce a LuxR homolog known as SdiA that can bind AHLs produced by other bacterial species to regulate gene transcription. Given that the conjugative apparatus exist in donor cells, we speculated that P. aeruginosa-released AHLs may act on SdiA of E. coli. To assess whether SdiA of E. coli is involved in the ability of P. aeruginosa's AHLs to inhibit E. coli-P. aeruginosa conjugation, we constructed the sdiA deficient mutant SM10λπ sdiA. As expected, deficiency of sdiA in SM10λπ significantly enhanced E. coli-P. aeruginosa conjugation, whereas overexpression of SdiA reversed the phenotype (Figure 2A). However, when using a AHLs-non-producing E. coli strain EC600 as the recipient cell, SM10λπ sdiA did not increase conjugation ability compared to the WT strain ( Figure 2B). More importantly, the conjugation efficiency of SM10λπ and EC600 significantly decreased in the presence of exogenous 3-oxo-C12-HSL and C4-HSL, while sdiA deletion in SM10λπ abrogated the effects of AHLs on conjugation (Figure 2B), suggesting the inhibitory effect of SdiA on E. coli-P. aeruginosa or SM10λπ-EC600 conjugation is dependent on the presence of AHLs.
In addition, growth curves demonstrated that deficiency of sdiA in E. coli had no influence on cell proliferation ( Figures  S4, S5), confirming that the regulatory function of SdiA in conjugation in this model was not due to an altered growth rate.
Collectively, these data imply that AHLs produced by PAO1 may repress SM10λπ-PAO1 conjugation through binding to SdiA of E. coli.
The Interaction between P. aeruginosa HSL and E. coli SdiA Inhibits the Expression of traI in E. coli Mechanisms behind transcription regulation function of SdiA is being disclosed, it seems that genes with specific DNA sequences (SdiA-box) 5 ′ -AAAAG(N8)GAAAA-3 ′ in the promoter region may be the potential targets of SdiA (Yamamoto et al., 2001). In view of the presence of SdiA-box in the promoter of many SdiAregulated genes in our bioinformatics analysis (Table S2), we computationally mapped the DNA sequence in the RP4 plasmid to search for conjugation-related genes potentially regulated by SdiA. An SdiA-box sequence (5 ′ -AAGAGcgattgagGAAAA-3 ′ ) was identified −317 bp upstream of the traI start codon ( Figure  S6). Subsequently, EMSA assays confirmed the interaction between SdiA and the predicted SdiA-box of the traI promoter in vitro (Figures 3A,B). We therefore further evaluated the role of SdiA in the regulation of traI transcription. DNA fragments of traI promoter carrying the predicted SdiA-box was cloned upstream of the β -galactosidase gene in the pQF50-promoter reporter. When transformed into BW25113 (another E. coli strain without endogenous β-galactosidase compared to SM10λπ), the β-galactosidase activity of pQF50-traI was greatly elevated, compared to that of the control, while addition of 3-oxo-C12-HSL and C4-HSL impaired this activity, which was severely attenuated when the sdiA was deleted ( Figure 3B). Intriguing, when AHLs was absent, deletion of sdiA hardly affected β-galactosidase activity of pQF50-traI ( Figure 3B), this is consistent with the phenotype shown in Figure 2. Compared with the WT strain, SM10λπ sdiA showed higher mRNA expression of traI when cultured with PAO1 ( Figure 3C). On the other hand, in the SM10λπ-PAO1 co-culture system, deficiency of lasI or rhlI in PAO1 also led to enhanced expression of traI (Figure 3D), while supplementation with exogenous 3-oxo-C12-HSL and C4-HSL significantly repressed traI expression (Figure 3E). These results suggest that repressing traI expression in the donor cells may be a critical mechanism behind the inhibitory effect of the AHLs on conjugation.
In summary, we disclosed the cooperative effect of AHLs produced by recipient P. aeruginosa cells and SdiA of donor E. coli cells in the conjugation regulation. These findings indicate that QS may inhibit conjugation and prevent the excessive dissemination of plasmid.

DISCUSSION
Most recent publications in this field have focused on the regulatory function of QS in virulence and biofilm formation. Here, using E. coli (SM10λπ) as donor cells and AHLs-producing P. aeruginosa (PAO1) or non-AHLs producing E. coli (EC600) as recipient cells, we identified a conjugation-inhibitory effect for FIGURE 1 | The quorum sensing system of P. aeruginosa inhibits conjugation between P. aeruginosa and E. coli. (A) Deficiency of the AHLs synthase genes rhlI or lasI in P. aeruginosa (PAO1) resulted in the absence of C4-HSL or both 3-oxo-C12-HSL and C4-HSL, respectively. The 3-oxo-C12-HSL and C4-HSL in the cell-free supernatants were extracted with ethyl acetate and redissolved in methanol, followed by HPLC-MS/MS analysis. (B) Deficiency of lasI or rhlI in P. aeruginosa abolished production of the downstream toxin of rhlI system pyocyanin. PAO1, PAO1 lasI and PAO1 rhlI were cultured in the presence or absence of 3-oxo-C12-HSL or C4-HSL as indicated for 30 h. (C) Deficiency of lasI or rhlI in P. aeruginosa significantly promoted SM10λπ-PAO1 conjugation; this effect could be abrogated by supplementation with exogenous 3-oxo-C12-HSL or C4-HSL. SM10λπ and PAO1 (10 7 CFU/mL each) were mated at 37 • C for 6 h in the presence or absence of 40 µM of C4-HSL or 3-oxo-C12-HSL. Ctrl, control; C12, 3-oxo-C12-HSL; C4, C4-HSL. Values are mean ± SEM of at least three independent experiments; ***P < 0.001. QS based on the following evidence. First, for SM10λπ and PAO1 co-culture system in which AHLs is normally self-sustained, deficiency of the AHLs-producing genes lasI or rhlI in PAO1 or the solo AHLs receptor SdiA in SM10λπ promoted SM10λπ-PAO1 conjugation, while supplementation with exogenous 3oxo-C12-HSL or C4-HSL abrogated the enhanced conjugation ability of PAO1 lasI and PAO1 rhlI. On the other hand, for both non-AHLs producing SM10λπ and EC600 mixed cultures, stimulation with exogenous 3-oxo-C12-HSL and C4-HSL inhibited conjugation, while deletion of sdiA in SM10λπ attenuated this effect. Conventionally, conjugation is considered to be mainly regulated by the self-transmissible plasmids. While our results indicate that QS system of donor and recipient cells may play a role in conjugation regulation.
to new circumstances (Norman et al., 2009). Therefore, the ability to inhibit conjugation may be a potentially efficacious strategy for avoiding the spread of resistance traits. Here, we demonstrate that AHL-SdiA is capable of suppressing conjugation. Most SdiAexpression bacteria, such as Escherichia, Salmonella and Shigella are enterobacteria, while many biological evidences suggest a lack of HSLs in the normal mammalian intestine (Swearingen et al., 2013), despite the presence of AHLs in bovine rumen (Hughes et al., 2010). Thus, although P. aeruginosa could be detected in stool sample in our clinical microbiology laboratory, future studies are needed to illuminate the role of AHL-SdiA signaling in pathogenic bacteria communities within the gastrointestinal tract.
To date, many SdiA regulon members have been described (Kanamaru et al., 2000;Wei et al., 2001;Dyszel et al., 2010;Sabag-Daigle et al., 2015). Here we report the identification of SdiA-regulated and AHL-responsive gene traI in the plasmid RP4. TraI is reported to function as a relaxase enzyme that creates a nick at the oriT of conjugative plasmids, which is required to initiate conjugation (Furuya and Komano, 2000). We discovered a DNA motif recognized by SdiA in the promoter region of the traI gene in the plasmid RP4, and the interaction between SdiA and the predicted SdiA-box was validated in vitro using an EMSA. However, some SdiA-regulated genes do not have this particular SdiA-box (Dyszel et al., 2010;Swearingen et al., 2013;Abed et al., 2014;Nguyen et al., 2015), there (B) Activity of β-galactosidase reporters containing the predicted SdiA-box in genomic regions upstream of traI under various conditions. E.coli BW25113 or E.coli BW25113 sdiA were cultured with SM10λπ in the presence of DMSO (Ctrl) or 40 µM C4-HSL and 3-oxo-C12-HSL for 6 h, followed by β-galactosidase activity analysis. (C) Deficiency of sdiA promoted traI expression in E. coli. (D) Deficiency of lasI/rhlI promoted traI expression in E. coli. For (C,D), 10 7 CFU/mL (each) of the indicated E. coli donor and recipient P. aeruginosa cells were mated at 37 • C for 6 h, followed by real-time PCR analysis. The rpoD gene of E. coli was used as an internal control. (E) Exogenous 3-oxo-C12-HSL or C4-HSL inhibited traI expression in E. coli. PAO1 lasI or PAO1 rhlI were cultured with SM10λπ in the presence of DMSO (Ctrl) or 40 µM C4-HSL and 3-oxo-C12-HSL for 6 h, followed by real-time PCR analysis. The rpoD gene of E. coli was used as an internal control. Ctrl, control; C12, 3-oxo-C12-HSL; C4, C4-HSL. Values are mean ± SEM of at least three independent experiments; *P < 0.05; **P < 0.01; ***P < 0.001. may be other conjugation-related genes repressed by AHL-SdiA. Moreover, the EMSA was performed without addition of AHLs, so it seems that high concentration of SdiA could bind to traI promoter in the absence of AHLs in vitro (Figure 3A), however, the reporter system ( Figure 3B) and conjugation experiment ( Figure 2B) showed that in the absence of AHLs, deletion of sdiA hardly affected the promoter activity of traI, as well as conjugation frequency in vivo. Thus, we proposed AHLs may increase both SdiA protein stability and traI promoter-binding affinity to repress traI expression.
Despite the advantages of conjugation for bacteria, the introduction of novel genes into the pre-existing, well-tuned genetic background is a source of genetic conflict, and possession of the conjugation-associated machinery also places a burden on the host arising from the energy expended to create and maintain the conjugative apparatus and its associated features (Zatyka and Thomas, 1998;Baltrus, 2013;San Millan et al., 2015). This raises the question of how host bacteria minimize the metabolic cost while obtaining the benefits provided by conjugation. In this study, we found that under normal conditions, when mobilizable plasmid containing a resistance gene was not required by PAO1 (Table S3), conjugation between SM10λπ and PAO1 was inhibited via the LasI/RhlI-HSL-SdiA pathway. These findings reveal that QS system may play a role in protecting host cells against external conjugative plasmids.
Utilizing ecological data from 2801 samples, Freilich et al. explored the ubiquitous competitive and cooperative interactions between the bacteria within natural communities (Freilich et al., 2011). Nonetheless, revealing more detail of the strategies bacteria adopt for survival in mixed cultures remains a major challenge. The E. coli-P. aeruginosa conjugation model has been widely used in studies of bacterial conjugation, and the most prevalent donor strain is E. coli SMλπ in which the RP4 plasmid is chromosomally-integrated. Thus, conjugation-associated genes, such as traI initially only exist in and are expressed by the E. coli (SM10λπ) cells, similarly rhlI and lasI are only expressed by PAO1. This makes it easy to detect the expression of these genes in E. coli (SM10λπ) and PAO1, specifically in mixed-cultures. Using this co-culture system, we found that LasI/RhlI and SdiA jointly repressed traI expression in E. coli and inhibited SM10λπ-PAO1 conjugation, indicating that the QS system may provide a mechanism of cooperative regulation between bacteria.
In conclusion, the findings of this study highlight the regulatory role for the QS system in conjugation, and expand our understanding of the bacterial communication and defense systems of P. aeruginosa.

AUTHOR CONTRIBUTIONS
YL, JZ, XH, BH, and CC designed research; YL, JZ, BW, RC, and NZ performed research; SE and YQL contributed new reagents/analytic tools; YL, JZ, and BW analyzed data; YL, JZ, and CC wrote the paper.