Loci identification of a N-acyl homoserine lactone type quorum sensing system and a new LysR-type transcriptional regulator associated with antimicrobial activity and swarming in Burkholderia gladioli UAPS07070

2.1 Bacterial strains, plasmids and culture conditions

The plasmids and strains of this study are shown in Table 1. B. gladioli UAPS07070 and the mutants were grown on Luria Bertani (LB) agar plates or liquid medium at 30oC or 37oC, depending on the experiment. E. coli DH5α was grown on LB plates or in LB broth at 37oC. Kanamycin (Km) was supplemented when required (80 μg/ml).

2.2 Mutagenesis with Himar1

A randomly mutated library of B. gladioli UAPS07070 was obtained containing the transposon Himar1, an efficient genetic tool for B. gladioli and other Burkholderia species [59,60]. Additionally, Himar1 possesses an origin or replication that allows for the recovery of sequences next to the transposon insertion even without cloning them into another vector. The suicide plasmid pHBurk3 was introduced into B. gladioli UAPS07070 by electroporation [61]. Electrocompetent cells were obtained as follows; cells were inoculated in 1 ml of LB broth and incubated at 30oC under agitation for 24 h. The bacterial cells were washed twice with 300 mM sucrose at room temperature and the pellet was resuspended in 100 μl of the same solution. The conditions of the electrical pulse were 25 μF; 200 Ω; 2.5 kV on a Bio-Rad GenePulserXcell. The transformants were selected in plates of LB kanamycin (80 μg/ml), incubated at 37oC until isolated colonies appeared. The mutants were preserved at -70oC in 50% glycerol (v/v).

2.3 Screening of mutants with reduced antimicrobial activity

Inhibition-impaired mutants against the sensitive bacterium Acinetobacter sp. UAPS0169 were initially screened with a multiple antagonism assay as described below. The mutants were grown with agitation in LB broth at 30oC for 24 h. 20 μl of this culture was plated on LB agar and incubated for 24 h at 30oC. Thereafter, 20 μl of culture of the sensitive strain in stationary phase was placed in the vicinity of the antagonistic strain growth without touching it. The cells were further incubated at 30oC for 24 h. Candidates for loss of antagonism were detected by the growth of the sensitive strain into the radius of one centimeter starting in the edge of the cell mass of the mutant. Double layer assay [Modified from 62] corroborated the phenotype. All antagonistic strains were grown in LB broth at 30oC for 24 h to reach stationary phase. The cultures were adjusted to an OD₆₂₀ of 0.05 in LB broth, 6 μl was inoculated onto solid LB agar plates and incubated for 24 h at 30oC. Subsequently, the bacterial growth was removed with a sterile glass slide, and the remaining cells were inactivated by exposure to chloroform vapors for 1 h. The plates were left opened in a laminar flow cabinet until the residual chloroform had been evaporated. Acinetobacter sp. UAPS0169 was grown and its density was adjusted in a similar manner to the mutants. 200 μl of the adjusted Acinetobacter culture was spread onto the LB agar plates where the mutant clone had been previously grown. The plates were incubated for 24 h at 30oC and inhibition halos were measured. For statistical analysis, the double layer assay was performed with twelve repeats for each mutant clone. The statistical significance was tested with ANOVA (p < 0.05).

2.4 Identification of transposition location

The transposon-chromosomal junction in clones with decreased antagonistic activity was rescued according to the methodology described by Rholl et al. [59]. Briefly, DNA was extracted and purified with Wizard Genomic DNA Purification kit (Promega Co). The DNA was digested with NotI (Thermo Fisher Scientific) and purified with the High Pure PCR Cleanup Micro Kit (Roche), and self-ligated with the Rapid DNA ligation kit (Roche). The DNA was transformed into E. coli DH5α, and plated on LB agar plates containing Km (30 μg/ml) at 37oC. Plasmid extraction was conducted with the Pure Yield Plasmid Miniprep System kit (Roche). All kits were used following manufacturer’s instructions. The mutated sites were sequenced with the transposon specific primers 1670 (5`-TCGGGTATCGCTCTTGAAGGG-3`) and 1829 (5`-GCATTTAATACTAGCGACGCC-3`). Blastx (http://blast.ncbi.nlm.nih.gov/Blast.cgi?PROGRAM=blastx&PAGE_TYPE=BlastSearch&LINK_LOC=blasthome) was used to compare the sequences. Multiple sequence alignment with ClustalW was used to calculate the percentage of identity of the translated orfs [63]. The DNA sequence was deposited in NCBI GenBank with the acc. numbers KP123645 and MF325937. The orfs were searched with the Interproscan platform (http://www.ebi.ac.uk/interpro/search/sequence-search) to identify conserved domains in TofI, TofR and LysR.

2.5 Acyl homoserine lactone assay

The secretion of acyl homoserine lactones was detected with the quorum sensing indicator strain C. violaceum CV026 [64] which exhibits a purple pigmentation in the presence of AHLs. Cells were grown in LB broth to the stationary phase, a layer of C. violaceum CV026 (approximately 200 μl) was distributed over a PSUC plate (4.2 mM succinic acid, 1% casein peptone and 1.6% bacteriological agar). The plates were left opened until the suspension had dried up. 10 μl of the bacterial suspension of the tested cells was placed onto each plate and incubated for 24 h at 30oC. A positive reaction was indicated by the presence of a violet halo around the colony (violacein production).

2.6 Swarming assay

Bacterial cultures were grown for 24 h at 30oC in LB broth. One ml of bacterial culture, with an OD₆₂₀ of 0.01, was washed twice with sterile LB broth. A volume of 1 μl of bacterial cells was placed in the center of an LB plate and incubated either at 30oC or 37oC for 48 h. There were a total of five replicates for each sample. A dendritic pattern formation was regarded as positive for swarming.

2.7 Toxoflavin quantification

The toxoflavin production was quantified in solid medium. B. gladioli UAPS07070 or derivate mutants were grown to stationary phase in LB broth with an OD₆₂₀ of 0.05. 200 μl of suspension was spread over LB plates and incubated at 30oC for 24 h. The cell mass was removed mechanically with a glass slide and the agar surface was cleaned with a swab soaked with 70% ethanol to remove the remaining cells. An agar fragment (500 mg) where the cells had been grown, was sliced up and resuspended in 500 ml of chloroform. The chloroform fraction was transferred to fresh polypropylene tubes and the solvent was evaporated. The residue was dissolved in 80% methanol and the absorbance was measured at 393 nm [65]. The absorbance lectures were compared with ANOVA (p < 0.05) using six replicates.

2.8 End-point RT-PCR analysis oftoxAgene

The transcription of toxA gene by B. gladioli UAPS07070, by QS and by LysR mutants were analyzed by RT-PCR. Bacterial cells were grown at 30oC in LB broth until reaching the stationary phase with an OD₆₂₀ = 0.05. 30 μl of this bacterial suspension was plated on LB plates and incubated for 24 h at 30oC. The cells were recovered in 1 ml of RNase free water and centrifuged at 4oC and 8,000 x g for 3 min. The simple phenol method [66] was used for total RNA extraction and were as follows: 100 μl of RNase free water was added to the pellet and the cells were vortexed for 3 min, 100 μl of acid phenol-chloroform (1:1) was added and the tube was vortexed for 1 min. Following this, tubes were incubated at 70oC for 30 min, vortexing each for 5 min. The sample was centrifuged at 12,000 x g for 10 min, 100 μl of aqueous phase was transferred to a clean tube and 200 μl of isopropanol was added. The tube was vortexed for 3 min and centrifuged at 12,000 x g for 10 min, the RNA pellet was washed twice with 200 μl of 70% ethanol and centrifuged at 12,000 x g for 5 min. The pellet was air dried and RNA was solubilized in 25 μl of RNase free water. The integrity of the RNA was qualitatively evaluated in 1% agarose gel electrophoresis. The concentration and purity were measured in a NanoDrop spectrophotometer (Thermo Scientific). RNA (4 mg) was treated with DNaseI (Thermo Fisher Scientific) according to manufacturer’s instructions. For synthesis of cDNA, 2 mg of DNA-free RNA was used. cDNA was generated with the Transcriptor High Fidelity cDNA Synthesis Kit (Roche) according to manufacturer’s instructions. The toxA gene expression was evaluated by PCR amplification of a fragment of 181 bp using the oligonucleotides RTP3 (5`-GTT CAG CTT CTA CCG CTG GA-3`) [45] and TOXA2 (5`-TCA AGG CTT GCA GAC CAG-3`) [48]. A 412 bp fragment of 16S rDNA gene was amplified as control of constitutive expression with the oligonucleotides Fwd (5`-GTG CCA GCM GCC GCG GTA ATA C-3`) and Rev (5`-CCG TCA ATT CCT TTG AGT TT-3`) [67]. 1 μl of cDNA was used for amplification of both, toxA and 16S rDNA with the following conditions: 96oC for 2 min, 30 cycles at 96oC for 1 min, 50oC for 1 min for toxA; and, for 16 rDNA, 58oC for 1 min and a final extension at 72oC for 30 s.

2.9 Assay for virulence in onion and lettuce

B. gladioli UAPS07070, BG1232, BG87 and BG79 were grown until reaching the stationary phase at 30oC in Luria-Bertani broth. Cultures were washed twice with sterile culture medium and adjusted to OD₆₂₀ = 0.05 (ca. 10⁷ cfu/ ml). For the onion assay a modification of the method of Jacobs et al. [68] was used. Fragments of ‘Yellow Globe’ onion bulbs (ca. 10 cm²) were inoculated with 5 ml of bacterial suspension inside of the inner surface from a wound made with a micropipette tip. The onion bulbs were incubated in a wet chamber at 30oC for 48 h and virulence was demonstrated by tissue maceration. For the assays in lettuce plants three bright leaves of four weeks old plants were inoculated in the midribs with 10³ washed CFU in 10 ml. The control was inoculated with 10 ml of 10 mM MgSO₄. The plant were maintained in humid conditions and the appearance was observed daily for 5 days [Modified from 69].

2.10 Complementation strategy

The AHL synthase defective mutant BG1232 was complemented in trans with the plasmid pAHL-7 (Table 1 and Fig. S1) harboring 615 bases of tofI plus 386 nucleotides upstream and 343 bases downstream. The plasmid pAHL-7 was constructed by ligating the 1,344-bp SalI fragment of plasmid p87-2 (Fig. S2) into the corresponding restriction site of backbone vector pBBR1MCS-5 [70] (see Table1). The mutant BG87 was complemented with 723 bases of tofR plus 374 bases upstream and 170 bases downstream. A fragment of 1,267-bp containing tofR was obtained from the plasmid p1232 digested with SalI-XhoI (Thermo Fisher Scientific) (Fig. S3) and cloned in the XhoI site of pBBR1MCS-5, yielding plasmid pQSR-15 (Fig. S4). The plasmid pQSR-15 was transformed in BG87. Both inserts were confirmed by sequencing.

Ethical approval: The conducted research is not related to either human or animals use.

3.1 Screening of antimicrobial activity impaired mutants

We employed random mutagenesis for identifying loci associated with antimicrobial activity of B. gladioli UAPS07070. A library of 3,500 random mutants of B. gladioli UAPS07070 was analyzed. To show antimicrobial activity to B. gladioli UAPS07070, the gram-negative strain Acinetobacter sp. UAPS0169 was used. This strain exhibits fast and homogenous growth that allows for the clear detection of inhibition halos. The mutants, BG1232, BG87 and BG79 exhibited remarkable reduction of antimicrobial activity against Acinetobacter sp. UAPS0169 (Fig. 1B). The Himar1 insertions in BG1232 and in BG87 were located close to each other allowing for the ligation of a 2,098 bp DNA segment (Fig. 2A). The sequence showed high similarity to tofI, tofM and tofR sequences (Table 2), which is related to the Acyl homoserine lactone quorum sensing system (AHL QS). The insertion in BG1232 interrupted a sequence corresponding to the putative Acyl-homoserine lactone synthase tofI (Table 2), and in the mutant BG87, affected the putative QS regulator tofR (Table 2, Fig. 2A). The QS circuitry that regulates pathogenicity of both B. gladioli, including UAPS07070 and B. glumae corresponds to the topological arrangement of M1 described in Proteobacteria [71,72]. The arrangement in the group M1 is characterized by the divergent transcription of tofI and tofR. Besides, tofM, a negative regulator homolog to rsaM, is found in the intergenic region between tofI and tofR. Both tofM and tofI are transcribed in tandem, and tofR is transcribed divergently to tofMI [72]. Domain search of TofI of B. gladioli UAPS07070 with Interproscan identified domains related to acyl transferase activity, autoinductor synthase activity and autoinductor binding (Fig. S5). The same analysis of TofR showed the presence of four characteristic domains of LuxR (Fig. S6).

Fig. 1

Role of Quorum-sensing and LysR in antagonism, toxoflavin synthesis, motility, and virulence. A, Acyl homoserine lactones detection. a, UAPS07070; b, BG1232; c, BG1232-pAHL-7; d, BG1232-pBBR1MCS-5; e, CV026; f, BG87; g, BG87-pQSR-15; h, BG87-pBBR1MCS-5. Positive reaction is by the presence of a violet halo around of the colony (marked with a white arrow). An aliquot of 10 μl of the suspension of the tested cells was placed onto a dry layer of C. violaceum CV026 distributed over a PSUC plate. B, Antagonism assay against Acinetobacter sp. UAPS0169. Inhibition halos produced by UAPS07070, BG1232, BG87, BG79, BG1232-pAHL-7 and BG87-pQSR-15. Double layer assay in LB plates incubated for 24 h. S.D. Standard deviation of five replicates. C, Toxoflavin detection. Toxoflavin level detected to 393 nm of absorbance with a standard deviation of six repeats of the strain UAPS07070, BG1232, BG87, BG79, BG1232-pAHL-7 and BG87-pQSR-15. D, Motility assay. a and e, UAPS07070; b, BG1232; c, BG1232-pAHL-7; d, BG1232-pBBR1MCS-5; f, BG87; g, BG87-pQSR-15; h, BG87-pBBR1MCS-5. Motility was determined in LB agar, 0.4 %, after incubation for 48 h at either 30 or 37°C. E Virulence lettuce assay, the inoculated strains are indicated in the figure. Observations was determined after incubation for 5 days at 30°C. The arrows indicate control inoculation.

Fig. 1

Role of Quorum-sensing and LysR in antagonism, toxoflavin synthesis, motility, and virulence. A, Acyl homoserine lactones detection. a, UAPS07070; b, BG1232; c, BG1232-pAHL-7; d, BG1232-pBBR1MCS-5; e, CV026; f, BG87; g, BG87-pQSR-15; h, BG87-pBBR1MCS-5. Positive reaction is by the presence of a violet halo around of the colony (marked with a white arrow). An aliquot of 10 μl of the suspension of the tested cells was placed onto a dry layer of C. violaceum CV026 distributed over a PSUC plate. B, Antagonism assay against Acinetobacter sp. UAPS0169. Inhibition halos produced by UAPS07070, BG1232, BG87, BG79, BG1232-pAHL-7 and BG87-pQSR-15. Double layer assay in LB plates incubated for 24 h. S.D. Standard deviation of five replicates. C, Toxoflavin detection. Toxoflavin level detected to 393 nm of absorbance with a standard deviation of six repeats of the strain UAPS07070, BG1232, BG87, BG79, BG1232-pAHL-7 and BG87-pQSR-15. D, Motility assay. a and e, UAPS07070; b, BG1232; c, BG1232-pAHL-7; d, BG1232-pBBR1MCS-5; f, BG87; g, BG87-pQSR-15; h, BG87-pBBR1MCS-5. Motility was determined in LB agar, 0.4 %, after incubation for 48 h at either 30 or 37°C. E Virulence lettuce assay, the inoculated strains are indicated in the figure. Observations was determined after incubation for 5 days at 30°C. The arrows indicate control inoculation.

The tofIMR genes are implicated in the regulation of toxoflavin biosynthesis and swarming movement in the plant-pathogenic bacteria, B. glumae BGR1, 336gr-1 and B. gladioli BSR3 [49, 50, 51, 52, 53,,73]. Toxoflavin is a virulence factor demonstrated by the phytopathogen B. glumae [49,50,74,75] and potentially involved in the phytopathogenicity of B. gladioli [53]. In both models, N-hexanoyl homoserine lactone (C6-HSL), and N-octanoyl homoserine lactone (C8-HSL), are synthesized by TofI, the ortholog of the AHL synthase LuxI [49,76]. This TofR-C8-HSL complex activates the regulatory cascade of toxoflavin synthesis and its transport in B. glumae [49]. Besides its virulence role in plant-pathogenic bacteria, toxoflavin shows strong antimicrobial activity. Recently, the genome analysis of 88 isolates of a B. gladioli collection revealed that the toxoflavin biosynthetic pathway is conserved across all genomes [77]. The genetic element interrupted by the transposon Himar1 in BG79 shows a 99% nucleotide identity with a gene of the lysR family, harbored in chromosome 2 of B. gladioli BSR3, nt 1581223 to nt 1582191 (GenBank CP002600.1). The gene is transcribed independently and encodes for a LysR-type protein of 322 a.a. (Table 2). The Himar1 insertion was located near the C-terminal domain (nucleotide 634) that is essential for binding the co-inducer (Fig. 2B). The antibiosis of BG79 against the sensitive strain was severely diminished, suggesting a positive regulation of genetic elements involved in the synthesis and/or transport of toxoflavin. All of the B. gladioli sequenced genomes, including BSR3, show this same region with identical gene context among them (Results not shown). In all of them, downstream to the lysR locus there is a hypothetical protein in the opposite transcription direction. Upstream to lysR there is one hypothetical OmpC family outer membrane protein and one hypothetical protein, both in the opposite transcription direction to the lysR gene. In the regulatory circuit of toxoflavin production by B. glumae BGR1, ToxR positively regulates the toxoflavin biosynthesis and transport genes. It is a LysR-type protein that recognizes toxoflavin as a co-inducer [49]. The sequence of lysR, is interrupted by Himar1 in BG79, and exhibits low sequence homology to the toxR sequences of B. glumae BGR1 and B. gladioli BSR3 (data not shown) suggesting that it could detect a toxoflavin alternative co-inducer.

We explored the production of a yellowish pigment, indicative of toxoflavin production [49] in B. gladioli UAPS07070. Normally, UAPS07070 produces toxoflavin on solid medium. However, the mutants BG1232, BG87, and BG79 do not (Fig. 1C). It has been reported that some B. gladioli phytopathogenic genotypes produces toxoflavin, and it is QS-dependent [52,53]. B. gladioli UAPS07070 is an endophytic bacterium isolated from inner tissues of pineapple. It exhibits strong antimicrobial activity against several bacteria and fungi [39]. It might also antagonize the phytopathogenic bacterium Tatumella ptyseos under natural conditions in pineapple plants [39].

While our results indicate that UAPS07070 synthesizes toxoflavin in culture media (Fig. 1C). We cannot discount the production of other antimicrobial metabolites. Kim et al. [52] observed that the expression of B. gladioli BSR3

polyketide genes related to the synthesis of the putative antibiotic bacillaene is QS-dependent [52]. Furthermore, other genotypes of B. gladioli synthesize antagonistic molecules like enacyloxin, bongkrekic acid, gladiolin, and a cyclic peptolide antibiotic [38,40,41,48,77], but the relationship to QS regulation remains to be elucidated. In addition to QS, the regulation of the toxoflavin synthesis in B. glumae includes an alternative pathway comprised of different loci, among them an orphan luxR [50,51,78]. The B. gladioli mutants BG87 and BG1232 exhibit slight antimicrobial activity against the sensitive strain Acinetobacter sp. UAPS0169 (Fig. 1B). This suggests alternative pathways of regulation in the residual production of inhibitory substances in UAPS07070 or the production of multiple antimicrobial molecules, as has been reported in Pantoea agglomerans Eh318 or Pseudomonas protegens Pf-5 [79,80].

3.2 Search of putative lux-box and autoinducer signal detection

The mutated region was aligned with the cep-box consensus sequence of B. cenopacia and B. ambifaria [54,81]. These two lux-boxes are well characterized in species of Burkholderia. A putative box was detected between positions 63-80 nucleotides upstream of the first ATG in tofI (Table S1). While tofM exhibits a putative lux-box 36 bp upstream of the start codon (Table S1). There was no lux-box found upstream of tofR (Table S1). We also did not find any lux-box upstream to lysR gene of BG79 (data not shown). The synthesis of signaling molecules associated to the putative QS system of UAPS07070 was confirmed with the reporter strain C. violaceum CV026. Apparently BG1232 did not synthesize short chain N-acyl homoserine lactones (Fig. 1A). In B. glumae, TofR, the ortholog to LuxR, activates the transcription of itself and of tofI in the presence of C8-HSL [52]. Interestingly, the mutant in the regulator, B. gladioli BG87, was not affected in the synthesis of signal molecules detected by C. violaceum CV026 (Fig. 1A). In that mutant the transposon insertion is close to the codon of residue D70 of TofR (TraR nomenclature of Agrobacterium tumefaciens), which is essential for the activity of this regulator [82,83]. In B. glumae BGR1, the expression of tofI is completely dependent on the TofR-C8-HSL complex [49]. Thus, a tofR mutant derived of B. glumae 336gr-1 does not synthesize AHLs [50]. In contrast, a different regulation of tofI is exhibited by B. gladioli UAPS07070. The mutant BG87 (tofR::Himar1) still synthesized AHLs, demonstrating that

Fig. 2

A, Map of the quorum sensing region of B. gladioli UAPS07070. The site of insertion of the transposon is indicated with ‘Himar1 followed by the mutant name’. B, Map of the lysR gene region of B. gladioli UAPS07070. The site of the insertion of the transposon is indicated with ‘Himar1 followed by the mutant name’. The principal domains of LysR-type protein, DNA binding domain (HTH) and co-inducer binding domain (substrate binding) in the LysR-type protein of UASP07070 are indicated below.

UAPS07070 tofI can be transcribed regardless of the tofR mutation. An example of TofR-independent regulation of tofI is demonstrated by B. pseudomallei, in which two of three AHL synthases (bpsI2 and bpsI3) are expressed constitutively [84]. In UAPS07070 the constitutive expression of tofI might provoke a fast response to environmental changes before reaching high cell density. This aligns with the production of inhibitory molecules by UAPS07070 in the exponential phase [39]. Another possible explanation is that tofI might possess another regulatory sequence besides the lux box. Further research is required to clarify the self-regulation QS system in B. gladioli UAPS07070.

3.3 Detection of toxA expression

The wild type strain, B. gladioli UAPS07070, released a deep yellowish pigment into the agar which is indicative of toxoflavin synthesis. This in contrast to the mutants BG1232, BG87 and BG79 (Fig. 1C). Two of the genes for the synthesis of toxoflavin, toxA and toxB in B. gladioli UAPS07070 were sequenced and compared by BLAST (data not shown). The transcription of toxA under experimental condition was evaluated using end-point RT-PCR. This revealed undetectable transcription of toxA in the mutants BG1232, BG87 and BG79 (Fig. 3). In contrast, an intense amplicon of toxA from RNA was observed in B. gladioli UAPS07070 (Fig. 3). These results strongly suggest toxoflavin acting as an antimicrobial of B. gladioli UAPS07070 here.

Fig. 3

End-point RT-PCR of toxA gene. A, Agarose gel electrophoresis of toxA amplified by approximately 181 bp: lane 1 UAPS07070 (gDNA); lane 2 UAPS07070; lane 3 BG1232; lane 4 BG87; lane 5 BG79, line M leader 100 bp (GeneCraft). B, End-point RT-PCR of 16S rDNA gene of approximately 412 bp: lane 1 UAPS07070 (gDNA); lane 2 UAPS07070; lane3 BG1232; lane 4 BG87; lane 5 BG79.

The transcription of toxA was undetectable in BG79 indicating a regulatory role of the LysR-type protein for the synthesis of toxoflavin. Our data revealed the participation of QS and of a LysR-type regulator in the toxoflavin production. However, the dependency between both regulation systems is unknown. To elucidate if both regulation systems are related, we explored the synthesis of acyl homoserine lactone by BG79. We found that the synthesis of autoinducer molecules was not hindered in BG79 (Fig. 1A), suggesting a higher hierarchical position of the system QS than the regulator LysR-type.

3.4 Swarming phenotype and virulence in onion and lettuce

Swarming is a multicellular movement over a solid or semisolid surface, and requires flagellae, cell-cell interactions and surfactant [55]. This is exhibited by B. gladioli UAPS07070 at 30°C with a dendritic pattern (Fig. 1D). A previous study in B. gladioli BSR3 showed QS-dependent swarming motility at 28°C with a movement pattern different than the dendritic one [52]. The interruption of tofI or tofR genes altered the swarming phenotype of B. gladioli UAPS07070 at 30°C (Fig. 1D). This may be related to the QS regulation of the synthesis of rhamnolipids, as what occurs in B. glumae [58]. In that model, a tofI mutant showed a significantly diminished synthesis of surfactant and decreased swarming activity. We show clearly that in B. gladioli UAPS07070, swarming motility is QS-dependent.

B. gladioli UAPS07070 inhabits pineapple as an avirulent bacterium [85]. In in vitro assays, we detected synthesis of toxoflavin, an important virulence factor involved in the pathogenicity of B. gladioli and B. glumae in rice and onion [53,75]. Although we have not detected pathogenic activity of B. gladioli UAPS07070 in pineapple. In in vitro assays in onions, this strain induces pathogenic symptoms. Peculiarly, the BG1232, BG87 and BG79 mutants seem to induce tissue maceration in onion bulb scales (Results not shown). Mutants BG1232 and BG79 showed less severe soft rotting in lettuce leaves (Fig. 1E). On the other hand, mutant in the tofR locus, BG87, did not show apparent decrease in virulence in lettuce (Fig. 1E). This is probably due to the AHL synthesis that could regulate toxoflavin-independent virulence mechanisms (Fig. 1A). QS regulatory networks of B. glumae, the close relative to B. gladioli are diverse as has been shown for strains BGR1 and PG1 [49,86]. Pathogenicity in B. glumae PG1 seem to be regulated by two different quorum sensing systems while in BGR1 is only regulated by one system [49,86]. B. gladioli BSR3 possess two distant quorum systems in different replicons (Acc. Nums. NC_015376 and NC_015382) [72]. It is unknown if UAPS07070 possesses more than one quorum loci and if one or both of them are related to pathogenicity. Lee et al. [53] reported that between B. gladioli and B. glumae there are differences in the participation of quorum sensing regulation. B. gladioli strain KACC11889 carries the genetic elements for the biosynthesis of toxoflavin but does not possess the quorum sensing genes for activating the synthesis of the phytotoxin. Nevertheless, B. gladioli KACC11889 still induces tissue maceration of onion bulb scales and exhibits swarming movement and cannot cause damage in rice. Thus, the regulation in this bacterium might have evolved differently to B. gladioli BSR3 and B. glumae BGR1, since in those strains the inactivation of QS system abolished the virulence in onion, toxoflavin production and the swarming motility [49,52,53,58]. Interestingly, two strains of the same species B. glumae, BGR1 and PG1, show different regulatory QS behavior [49,86]. It is known that B. glumae 237-5, an avirulent strain in rice does not produce toxoflavin and keeps its virulence in onion, indicating that the tissue maceration in onion is not attributed exclusively to toxoflavin [75].