A monocot, C. laevigatus, growing luxuriantly in the wet coastal areas of New-port, Bhavnagar, India (Latitude N 21° 45.124″, Longitude E 72° 13.579″), was collected. Bacteria were isolated from rhizosphere using a standard method, and axenic cultures were made for each isolate. Isolated axenic cultures were subjected to the screening of anti-quorum sensing activity using the reference strain Chromobacterium violaceum (CV026), cinnamaldehyde (Sigma-Aldrich, USA) as a positive control and methanol as a negative control in a plate-based bioassay (Singh et al., 2013). Bacterial isolates showing quorum sensing inhibition (QSI) activity were selected and checked further for antibacterial activity on Mueller-Hinton agar (MHA), along with tobramycin, which used as a positive control (Choo et al., 2006). A bacterial isolate showing promising positive QSI and negative anti-bacterial activities was selected further. The QSI and anti-bacterial activities of the selected isolate were repeated five times independently.

Genomic DNA of selected bacteria was isolated, and the 16S rRNA gene amplified with universal primers fD1-5′-AGA GTT TGA TCC TGG CTC AG-3′ and rP2-5′-ACG GCT ACC TTG TTA CGA CTT-3′ (Weisburg et al., 1991) and optimized PCR conditions (Keshri et al., 2013, 2015). The PCR product was purified, sequenced (M/s Macrogen Inc., South Korea) and subjected to BLAST analysis. Phylogenetic analysis was performed using MEGA (Molecular Evolutionary Genetics Analysis) version 6.0 software (Tamura et al., 2013). The phylogenetic tree was reconstructed using neighbor-joining methods (Saitou and Nei, 1987), bootstrap analysis was performed (Felsenstein, 1985), and evolutionary distances were determined using maximum composite likelihood algorithms (Tamura et al., 2004). The bacterial isolate was identified as D. tsuruhatensis strain SJ01, and the 16S rRNA gene sequence was deposited in the NCBI GenBank (KX130769).

Fatty acid methyl ester (FAME) profiling of identified bacteria was performed using Microbial Identification System (MIDI; Microbial ID) coupled with gas chromatography (GC system-6850, Agilent Technologies, USA). For whole cell fatty acid methyl ester profiling, the bacteria were grown on tryptic soy yeast agar for 24 h at 30°C, and fatty acid methyl esters were prepared according to the instruction manual of the Microbial Identification System (MIDI; Microbial ID). Peaks were identified and matched with RTSBA6 6.10 database (Jha et al., 2015).

Bacterial culture (D. tsuruhatensis strain SJ01, 500 ml in nutrient broth, NB), grown for 48 h, 180 rpm at 30°C was centrifuged for 15 min. at 10,000 × g, 4°C, and the supernatant was collected in a flask. The supernatant was filtered through 0.45 and 0.22 μm vacuum filters for the complete removal of bacterial cells. The filtrate was extracted twice with an equal volume of ethyl acetate. Ethyl acetate extract was evaporated to dryness under vacuum in a rotary evaporator (Büchi, Switzerland) and dissolved in methanol for further studies (Nithya et al., 2010).

The anti-quorum sensing activity of a methanolic extract of bacteria was tested by quantifying violacein (Choo et al., 2006). In brief, 1 ml of the freshly grown (OD_600nm 0.7) reference strain C. violaceum (CV026) was added to 20 ml NB Hi-veg media (Hi-media, India) containing hexonyl homoserine lactone (0.0625 μg/ml) and different concentrations of bacterial extract (0.01, 0.02, 0.03, 0.04, 0.05, 0.075, or 0.1 mg/ml). Cultures without extract and with methanol were considered the control and negative control, respectively. All cultures (controls and experimental) were incubated for 24 h at 30°C and 180 rpm (Choo et al., 2006). One milliliter of overnight grown culture from each flask was centrifuged 16,000 × g for 10 min, and the pellet containing violacein (produced by CV026) was suspended in 1 ml of dimethylsulfoxide (DMSO). The solution was centrifuged at 16,000 × g for 10 min to remove cell debris and absorbance was read at 585 nm in a microplate reader (Spectra Max Plus, USA).

A measure of 200 μl of overnight grown cultures (OD_600nm 0.1) of clinical isolates P. aeruginosa PAO1 (ATCC 15692) or P. aeruginosa PAH (by courtesy from Govt. Medical College, Bhavnagar; Goswami et al., 2011) was added to a 96-well microtiter plate with different concentrations of bacterial (strain SJ01) extracts (0.01, 0.02, 0.03, 0.04, 0.05, 0.075, and 0.1 mg/ml). The plate was incubated at 37°C, 100 rpm for 24 h, after which the growth of bacteria was measured at 600 nm and colony forming units (CFU) were also determined. Wells were washed after removing planktonic bacterial cells, dried and stained with 1% crystal violet. Excess dye was taken out after 20 min, wells were washed (with sterile distilled water), 200 μl ethanol (aqueous 96%) was added, and absorbance was measured at 590 nm (Andersson et al., 2009; Singh et al., 2013; Kavita et al., 2014). The experiments were performed thrice with five replicates each.

Cell viability within the biofilm was examined at different time points (24, 48, and 72 h) and compared with the control (Singh et al., 2013). Cells inhabiting the biofilm were stained with a fluorescent dye using the FilmTracer™Live/Dead® Biofilm Viability Kit (Invitrogen, USA) following manufacturer's instructions and visualized under an epi-fluorescence microscope (Axio Imager, Carl Zeiss AG, Germany).

The effect of bacterial extract (SJ01) on biofilm formation was visualized by scanning electron microscopy (SEM; Andersson et al., 2009; Singh et al., 2013). Biofilms of P. aeruginosa PAO1 and P. aeruginosa PAH, grown on glass coverslips (11 mm) submerged in nutrient broth with (0.1 mg/ml) or without bacterial extract were gently washed with 0.9% NaCl to remove planktonic cells. Samples were kept in 2.5% glutaraldehyde for 20 min followed by 4% OsO₄ in 0.1 M phosphate buffer for 30 min. Samples were dehydrated with a gradient ethanol series (10–95%) for 10 min. The dried biofilms were coated with gold and visualized under a scanning electron microscope (SEM, LEO series VP1430, Germany).

For atomic force microscopy (AFM), biofilms developed on glass coverslips were rinsed gently with phosphate buffer saline (pH 7.4) and kept in a desiccator for drying completely. The biofilm was scanned under AFM (NT-MDT, Russia) in a semi-contact mode at the speed of 1 Hz (Oh et al., 2009; Nithya et al., 2010). The surface bearing index (Sbi), core fluid retention index (Sci), valley fluid retention index (Svi), kernel roughness depth (Sk), reduced peak height (Spk), reduced valley depth (Svk), average roughness (Sa), root mean square (Sq), surface skewness (Ssk), coefficient of kurtosis (Ska), and surface area ratio (Sdr) were calculated.

Bacterial extract (SJ01) was tested on the swarming and swimming motility of P. aeruginosa. For the swarming motility assay, P. aeruginosa strains were spotted on a plate containing BM2 swarming medium (62 mM PBS at pH 7, 2 mM MgSO₄, 10 μM FeSO₄, 0.4% glucose, 0.1% casamino acids, and 0.5% agar) supplemented with (0.1 mg/ml) or without extract (Overhage et al., 2007). For the swimming motility assay, P. aeruginosa strains were spotted on a plate containing tryptone broth (10 g/l tryptone, 5 g/l NaCl, and 0.3% agar) supplemented with (0.1 mg/ml) or without extract (Rashid and Kornberg, 2000). Plates were analyzed after incubation of 24 h at 37°C.

The effect of bacterial extracts (SJ01; 0.1 mg/ml) was studied on the production of virulence factors of reference P. aeruginosa strains by quantifying pyocyanin and rhamnolipid, and analyzing elastase and protease activities. Briefly, P. aeruginosa PAO1 and P. aeruginosa PAH were grown overnight in 5 ml of PB medium (20 g/l peptone, 1.4 g/l MgCl₂ and 10 g/l K₂SO₄) supplemented with extract of strain SJ01 (0.1 mg/ml) and without extract (control) at 37°C (180 rpm). The culture was centrifuged at 10,000 × g for 10 min, and pyocyanin was extracted first from the supernatant in 3 ml of chloroform, followed by 1 ml of 0.2 N HCl. The absorbance was measured spectrophotometrically at 520 nm (Essar et al., 1990).

For rhamnolipid, reference strains (P. aeruginosa) were grown in nutrient broth supplemented with bacterial extract (SJ01; 0.1 mg/ml) or without extract (control). The culture was centrifuged at 10,000 × g for 10 min, supernatants were collected, acidified with HCl (to pH 2) and absorbance was measured at 570 nm (McClure and Schiller, 1992). Supernatants (750 μl) of overnight grown (with 0.1 mg/ml or without extract of strain SJ01) P. aeruginosa were incubated with 250 μl elastin Congo-red solution (5 mg/ml in 0.1 M tris-HCl pH 8; 1 mM CaCl₂) at 37°C, 180 rpm for 16 h. After incubation, the mixture was centrifuged at 3,000 × g for 10 min, and absorbance was measured at 490 nm for elastase activity (Zhu et al., 2002). For protease activity, supernatant (400 μl) was incubated with an equal volume of 2% azocasein solution (prepared in 50 mM phosphate buffer saline, pH 7) at 37°C for 1 h. The reaction was stopped by adding 500 μl of 10% trichloroacetic acid (TCA), and reaction mix was centrifuged at 8,000 g for 5 min to remove residual azocasein. The absorbance of the supernatant was read at 400 nm (Adonizio et al., 2008).

Bacterial extract (D. tsuruhatensis SJ01) was fractionated by the solid phase extraction (SPE) method using different cartridges (non-polar C18, polar SI, anion exchanger DAE and cation mixed Plexa PCX) and each fraction was screened for anti-quorum sensing activity. The positive fraction was further analyzed, and an active compound was identified by GC-MS. Briefly, crude bacterial extract (1 ml) was loaded to the preconditioned (by 5 ml methanol, 10 ml water and 5 ml acidified water pH 2.0) SPE cartridges (Agilent, USA). The elution was performed with a different concentration of 1 ml methanol (20, 40, 60, 80, and 100% v/v in water) and different fractions were collected (Singh et al., 2013). Each fraction was screened for plate based anti-quorum sensing activity (as described above) using the reference strain C. violaceum (CV026). The positive fraction was subjected to GC-MS (GC-2010, Shimadzu, Japan) and the identification of compounds was done by comparing the mass spectra with the reference mass spectra library. The mass of the fractionated compound identified was further confirmed by electrospray ionization mass spectrometry (ESI-MS; Q-Tof micro TM, Micromass, UK), performed in a positive mode.

Differential expression of regulatory genes of reference strain P. aeruginosa PAO1, involved in the quorum sensing was analyzed using microarray. Total RNA was isolated from reference strain P. aeruginosa PAO1, grown with or without bacterial extracts (0.1 mg/ml) using TRI reagent (Sigma, USA). Total RNA was quantified, and 10 μg RNA was converted to cDNA, befor being fragmented and labeled by following the GeneChip® P. aeruginosa PAO1 genome array user manual (Affymetrix, USA). Labeled cDNAs were hybridized with the P. aeruginosa genome array gene chip (containing total 5,886 gene probes), and then washed and stained (Singh et al., 2016a). Hybridized chips were scanned (Scanner 3000 7G, Affymetrix, USA), processed and analyzed using the expression console and the transcriptome analysis console (Affymetrix, USA). Microarray analysis was performed in duplicate (n = 2) and genes exhibiting significant fold expression (ANOVA p < 0.05) were considered for the study. All microarray data are available with Array-Express accession number E-MTAB-5693. For expression profiling, key regulatory genes (LasI, LasR, RhlI, and RhlR) were selected. Total RNA was extracted from control and treated P. aeruginosa (PAO1 and PAH strains) converted to cDNA and then quantitative real-time PCR was performed (Wang et al., 2005). A melt curve analysis was also done for the validation of specificity of the qRT-PCR reaction, and the relative fold expression change was calculated using the CT method (Livak and Schmittgen, 2001). The 16S rRNA gene was used as a reference gene (Wang et al., 2005).

A total of 56 bacterial axenic cultures were obtained from the rhizosphere of C. laevigatus L., of which two axenic cultures showed anti-quorum sensing activity in a plate-based bioassay. The isolate SJ01 showed promising anti-quorum sensing activity and a clear white opaque zone of inhibition was observed in the biosensor plate containing reference strain C. violaceum CV026 (Figure 1). Furthermore, the bacterial crude extract also showed QSI, whereas the zone of inhibition was not detected with the negative control (methanol). The disc diffusion antibacterial assay confirmed that selected bacterial isolates did not show antibacterial activity against the reference strain C. violaceum CV026 (Figure S1).

The 16S rRNA gene sequence (accession no. KX130769) of the selected bacterial isolate showed 99% similarity to D. tsuruhatensis, with 100% query coverage; therefore, this was designated D. tsuruhatensis SJ01. The phylogenetic tree reconstructed using the neighbor-joining algorithm shows the taxonomic position of identified bacterium with other species (Figure S2). The whole cell fatty acid profiling of the bacterium D. tsuruhatensis SJ01 revealed the abundance of C_16:0 fatty acids (Figure S3).

The bacterium D. tsuruhatensis SJ01 and its methanolic extract showed anti-quorum sensing activity with the reference strain on a biosensor plate. Different concentrations of bacterial extract were used to quantify the inhibition of violacein, an indicator of quorum sensing activity (Figure 2). The violacein production decreased concomitantly with the increasing concentration of the extract, and about 98% inhibition was observed with 0.1 mg/ml extract.

The anti-biofilm activity of the extract (D. tsuruhatensis SJ01) was tested against the wild-type, widely used biofilm forming clinical isolate P. aeruginosa PAO1 and a local clinical isolate P. aeruginosa PAH. The biofilm formation decreased concurrently in both reference strains with increasing concentration of bacterial extracts (Figure 3). About 60–64% inhibition of the biofilm formation was observed with 0.1 mg/ml extract. The possibility of an inhibitory effect of D. tsuruhatensis SJ01 extract on the growth of reference strains (P. aeruginosa) was also analyzed (Figures S4, S5). No significant effect was observed on the planktonic growth of P. aeruginosa in the presence of different concentration of bacterial extracts (0.01–0.1 mg/ml). Further, the disc diffusion antibacterial assay performed with SJ01 extract confirmed that bacterial extract did not show antibacterial activity against the clinical isolates P. aeruginosa (Figure S6).

The effect of the bacterial extract on the viability of the reference strain in the biofilm (24–72 h) was studied with an epi-fluorescence microscope (Figure 4). The dead P. aeruginosa cells were labeled with propidium iodide whereas live cells stained with SYTO 9, which produced red and green fluorescence, respectively. Less attachment of P. aeruginosa cells to the surface was observed even up to 72 h in the treated biofilm compared to control, and an insignificant number of dead cells was detected in the biofilms.

The topology of the biofilm developed by P. aeruginosa and the effect of D. tsuruhatensis SJ01 extract on it was analyzed by SEM and AFM. A well-grown biofilm along with adhering bacterial cells was observed in controls (normal biofilm developed by P. aeruginosa) in the SEM analysis, whereas dispersed bacterial cells were observed in treated samples (Figure 5). Similarly, AFM clearly showed the disrupted surface topology and height distribution profile of the biofilm developed in the presence of D. tsuruhatensis SJ01 extract compared to the control biofilm (Figure 6). The surface bearing indices, roughness analysis, and functional parameters based on the linear material ratio curve showed alterations of the biofilm developed in treated samples (Table 1).

Bacterial invasion is a prerequisite for biofilm formation. Therefore, the effect of bacterial extract (D. tsuruhatensis SJ01) was studied on the motility of biofilm forming P. aeruginosa bacterial cells. It was observed that bacterial extract (0.1 mg/ml) inhibits the swarming and swimming motility of P. aeruginosa strains in the plate assay (Figure 7). The extract reduced flagellum driven motility of P. aeruginosa in the treated sample compared to the control.

It was observed that bacterial extract (D. tsuruhatensis SJ01) reduced the production of virulence factors; pyocyanin and rhamnolipid (Figure 8). Pyocyanin production decreased about 70 and 55% in PAO1 and PAH strains, respectively with the treatment of 0.1 mg/ml bacterial extract. Similarly, rhamnolipid production was also decreased by 85 and 67% in PAO1 and PAH strains, respectively, in the presence of bacterial extract (0.1 mg/ml). The effect D. tsuruhatensis SJ01 extract on the elastase and protease activities of cell-free P. aeruginosa bacterial culture supernatant were also assessed (Figure 8). About 32–35% decrease in elastase activities was detected for both strains compared to the control. However, about 23–24% inhibition in the protease activity was found in both strains with 0.1 mg/ml bacterial extract compared to untreated samples.

In total, five fractions (in 20, 40, 60, 80, and 100% methanol) were collected through each SPE cartridge (non-polar C18, polar SI, anion exchanger DAE, and cation mixed Plexa PCX); all were screened individually for QSI using a biosensor plate containing C. violaceum CV026. Fraction (C18-100), collected through the C18 cartridge with 100% methanol, showed a maximum zone of QSI; therefore, this was selected for further characterization. Fraction C18-100 was subjected to GC-MS analysis, and the chromatogram showed a single peak at the retention time 16.518 min (Figure 9). The detected mass spectra showed some resemblance to 1,2-benzenedicarboxylic acid, diisooctyl ester, in the GC-MS library (NIST 27. LB). The calculated (theoretical) or expected molecular mass of compound 1,2-benzenedicarboxylic acid, diisooctyl ester (C₂₄H₃₈O₄) is 390.55_. The molecular mass of the active fraction (C18-100) was further confirmed by ESI-MS. A mass spectral peak, detected at m/z 397.1852, was considered the corresponding experimental mass of the active fraction (Figure 9).

Differential expression of quorum sensing regulatory genes of reference strain P. aeruginosa PAO1 treated with a bacterial fraction (C18-100) containing 1,2-benzenedicarboxylic acid, diisooctyl ester as a probable bioactive compound was analyzed using P. aeruginosa PAO1 genome array gene chip. Out of the 5,886 gene probe sets, 1,434 genes were differentially expressed (Table S1; Array-Express accession E-MTAB-5693) and showed at least 2-fold up- (>2) or down-(< −2) expression at p < 0.05 (Figure 10). Of these, 734 genes were up-regulated, whereas 700 genes were down-regulated. Some differentially expressed important genes (as observed in microarray analysis) involved in the quorum sensing and general metabolic pathways are listed in Table 2. The microarray scattered plot showed the differential expression of genes; up-regulation of genes was indicated by blue marks whereas green-colored dots represented down-regulation (Figure S7). The quantitative RT-PCR revealed that the genes LasI, LasR, RhlI, and RhlR were down-regulated in the treated P. aeruginosa compared to the control (Figure 10). About, 9.7-, 3.9-, 3-, and 5.9-fold down-regulation of the genes LasI, LasR, RhlI and RhlR, respectively, was observed in P. aeruginosa PAO1 strain. Similarly, 5.7-, 3.1-, 5.2-, and 4-fold decrease in gene expression was found in P. aeruginosa PAH strain.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.