Isolimonic acid interferes with Escherichia coli O157:H7 biofilm and TTSS in QseBC and QseA dependent fashion

Background E. coli O157:H7 (EHEC) is an important human pathogen. The antibiotic treatment of EHEC reportedly results in release of Shiga toxin and is therefore discouraged. Consequently, alternative preventive or therapeutic strategies for EHEC are required. The objective of the current study was to investigate the effect of citrus limonoids on cell-cell signaling, biofilm formation and type III secretion system in EHEC. Results Isolimonic acid and ichangin were the most potent inhibitors of EHEC biofilm (IC25=19.7 and 28.3 μM, respectively) and adhesion to Caco-2 cells. The qPCR analysis revealed that isolimonic acid and ichangin repressed LEE encoded genes by ≈3 to 12 fold. In addition, flhDC was repressed by the two limonoids (≈3 to 7 fold). Further studies suggested that isolimonic acid interferes with AI-3/epinephrine activated cell-cell signaling pathway. Loss of biofilm inhibitory activity of isolimonic acid in ΔqseBC mutant, which could be restored upon complementation, suggested a dependence on functional QseBC. Additionally, overexpression of qseBC in wild type EHEC abated the inhibitory effect of isolimonic acid. Furthermore, the isolimonic acid failed to differentially regulate ler in ΔqseA mutant, while plasmid borne expression of qseA in ΔqseA background restored the repressive effect of isolimonic acid. Conclusions Altogether, results of study seem to suggest that isolimonic acid and ichangin are potent inhibitors of EHEC biofilm and TTSS. Furthermore, isolimonic acid appears to interfere with AI-3/epinephrine pathway in QseBC and QseA dependent fashion.


Background
Enterohaemorrhagic Escherichia coli (EHEC) is a major foodborne pathogen associated with frequent outbreaks of diarrheal disease. Most individuals develop watery diarrhea and recover. However, about 15-20% cases may develop life-threatening bloody diarrhea and hemolytic uremic syndrome (HUS) [1,2]. Dissemination and contact of humans with EHEC from multiple sources such as undercooked meats, raw fruits and vegetables, physical contact with EHEC harboring animals further contribute to increased frequency of illness [2,3].
EHEC is usually ingested through contaminated food products. Once inside the host, EHEC traverses to colon and establishes itself in the distal ileum or large bowel. Inside the colon, EHEC is thought to use guided motility, provided by flagellar motion, to reach its preferred site of attachment [4]. Autoinducer molecules (AI-2/AI-3) and hormones (epinephrine/norepinephrine) induce various virulence factors and are speculated to help in attachment and subsequent infection process [5]. A two-component system QseBC [6] induces flagellar operon in response to hormones and AI-2/AI-3, resulting in increased and guided motility [4] towards epithelial cell layer. Upon encountering the epithelial cell layer, the flagella and other surface structures such as type 1 pili and hemorrhagic coli pilus help EHEC to attach to the surface [7][8][9]. Multiple environmental and genetic factors such as pH, hormones, signaling molecules as well as quorum sensing (QS) regulate the expression of Locus of enterocyte effacement (LEE) and flagellar operons [10][11][12][13]. The hormones and AI-3 also induce type III secretion system (TTSS) in EHEC through QseEF and QseAD [14,15]. TTSS is encoded in LEE, which is organized in five operons LEE1-LEE5. LEE1-encoded regulator (Ler) is the first gene on LEE1 operon and subject to modulation by various regulators. In turn, Ler activates the transcription of the five operons [13,15,16].
The TTSS penetrates the host cell membrane and serves as conduit for injecting effector proteins. These effector proteins manipulate the host machinery including actin cytoskeleton, resulting in attaching and effacing lesions. Some of the secreted effectors disrupt the tight junction leading to higher secretion of chloride ions and ultimately developing in diarrhea [17]. The phage encoded Shiga toxin is the main virulence factor of EHEC and other Shiga toxin producing E. coli. The Shiga toxin disrupts the protein synthesis in host epithelial cells causing necrosis and cell death [17]. Additionally, Shiga toxin travels to kidney through blood stream and damages renal endothelial cells inciting renal inflammation, potentially leading to HUS [2,18]. Along with the direct injury to epithelial cells, biofilms formed by pathogenic E. coli strains can pose serious health problems such as prostatitis, biliary tract infections, and urinary catheter cystitis [19].
Antibiotics and antidiarrheal drug therapy of EHEC activates the stress response resulting in induction of phage lytic cycle and subsequent release of Shiga toxin. The release of Shiga toxin is directly correlated with increase in HUS incidence [2,18]. At present, CDC recommends preventive measures such as washing hands and thorough cooking of meats etc. to control EHEC infections. However, these preventive measures need to be supported with alternative strategies for prevention and control of EHEC infections. A promising strategy is to identify anti-virulence agents, which may be used alone or in conjunction with antibiotic therapy [20]. Antivirulence agents target bacterial virulence determinants including toxin production, adhesion to host cells, specialized secretion systems such as TTSS [21]. Application of anti-virulence agents is speculated to allow host immune system to prevent or clear the bacterial infection. Several synthetic and natural molecules with antivirulence properties have been discovered [20,21] and at least one molecule, LED209, was shown to be effective in animal models [20]. However, none of the molecules have entered wide-scale clinical trial as of yet, owing to various concerns such as their toxicity and safety. Therefore, there is an urgent need to identify a more diverse pool of molecules with anti-virulence activities. Availability of such a pool will ensure better drug designing strategies, to combat bacterial infections like EHEC. Secondary metabolites produced by plants present very diverse scaffolds, which have been used for designing novel drugs including antimicrobials. In nature, secondary metabolites contribute to systemic and induced plant defense system against insect, bacterial and fungal infestation [22]. Several secondary metabolites belonging to classes such as coumarins, flavonoids, terpenoids and alkaloids demonstrate inhibitory properties against numerous microorganisms. Recently our group and others identified QS inhibitory properties of several plant secondary metabolites and extracts rich in phytochemicals [23][24][25][26][27][28].

Materials
Previously purified isolimonic acid, ichangin, isoobacunoic acid, IOAG and DNAG were used in the present study [36]. Purity of the individual limonoids was calculated from percent peak area using high performance liquid chromatography (HPLC) analysis [38]. A stock solution was prepared by dissolving 20 mg of each purified limonoid in 1 ml of dimethyl sulfoxide (DMSO).

Bacterial strains and plasmids
Bacterial strains and plasmids used in the study are listed in Table 1. Unless otherwise specified, bacterial cultures were grown at 37°C in Luria-Bertani (LB) medium supplemented with 0.5% glucose. When appropriate, medium was supplemented with 10 μg of chloramphenicol or 100 μg of ampicillin per ml. Biofilm studies were carried out in colony forming antigen (CFA) medium [39,40]. Plasmids pVS150 (qseA in pACYC177) and pVS178 (qseBC in pBAD33) were purified from strains VS151 and VS179 respectively, using Qiagen Plasmid Purification Kit (Qiagen) and electroporated into EHEC ATCC 43895. The transformed strains were designated as AV43 (EHEC containing pVS178) and AV45 (EHEC containing pVS150). In addition, pVS150 was electroporated into strain TEVS232 and resulting strain were designated as AV46. Furthermore, qseB and qseC were amplified from EHEC genomic DNA, using primers qseB (cloning) and qseC (cloning) . The primers were designed by altering one base to create restriction sites for the respective enzymes. Amplified fragment of qseC was digested with SacI and SalI and cloned into pBAD33, generating plasmid pAV11. The qseB fragment was digested with SacI and HindIII and cloned into pBAD33, generating plasmid pAV12. Plasmids pAV11 and pAV12 were subsequently electroporated into EHEC ATCC 43895 and strains were designated as AV48 and AV49, respectively.

Growth and metabolic activity
The growth and metabolic activity of EHEC was measured as previously described [36]. Briefly, overnight cultures of EHEC were diluted 100 fold in LB media. Two hundred microliters of diluted cultures was placed in each well of 96-well plates and grown for 16 h at 37°C in presence of 6.25, 12.5, 50, or 100 μg/ml limonoids or equivalent volume of DMSO. The plates were constantly shaken at medium speed in Synergy™ HT Multi-Mode Microplate Reader (BioTek, Instruments, Winooski, VT). OD 600 was recorded every 15 min. Metabolic activity of EHEC was measured by adding AlamarBlue (25 μl/well) and absorption at 570 and 600 nm was monitored in similar fashion as growth curve.

Biofilm assay
EHEC biofilms were grown in polystyrene 96-well plates by plating 200 μl/well of 100 fold diluted overnight cultures in presence of 6.25, 12.5, 50, or 100 μg/ml of limonoids at 26°C for 24 h without shaking [23,39]. For VS138 (ΔqseC) and VS179 (VS138 + qseBC) biofilms were quantified after 48 h growth in 96-well plates. The biofilms were quantified by staining with 0.3% crystal violet (Fisher, Hanover Park, IL) for 20 min. Extra stain was washed with phosphate buffer (0.1 M, pH 7.4) and dye associated with attached biofilm was dissolved with DMSO. An absorbance at 570 nm was used to quantify the total biofilm mass.

In vitro adhesion assay
Human epithelial Caco-2 cells were purchased from ATCC (Manassas, VA) and maintained in Dulbecco's Minimal Essential Medium (DMEM) with nonessential amino acids and 10% fetal bovine serum without antibiotics. Caco-2 cells were seeded at 1 × 10 5 cells/well in 6well plates and infected with approximately 5 × 10 6 cells/well of freshly grown EHEC ATCC 43895 in presence or absence of 100 μg/ml isolimonic acid, ichangin, isoobacunoic acid, IOAG and DNAG. The plates were incubated for 3 h at 37°C in 5% CO 2 environment. After completion of incubation, plates were washed 3× with sterile PBS to remove any loosely attached cells. Caco-2 cells were lysed with 0.1% Triton-X in PBS to release the bacteria and serial dilutions were plated on LB-agar and incubated at 37°C for 24 h. Colonies were counted after incubation period and presented as log 10 CFU/ml.

Quantitative PCR
Relative transcript amount of selected genes ( Table 2) was measured by qRT-PCR as described [23]. Briefly, overnight cultures of EHEC ATCC 43895 were diluted 100 fold with fresh LB medium or DMEM+10% FBS (referred as DMEM henceforth), treated with limonoids (100μg/ml) or DMSO and grown further at 37°C, 200 rpm. Bacterial cells were collected at OD 600 ≈1.0. RNA was extracted using RNeasy minikit (Qiagen Inc., Valencia CA) and converted to cDNA using MuLV reverse transcriptase enzyme and random hexamer in a Reverse-Transcriptase polymerase chain reaction (RT-PCR) [43] at 42°C for 1 h. PCR products were purified with QIAquick PCR-purification kit (Qiagen Inc.). Twenty five nanogram cDNA from each sample was amplified with 10 pmol target primers using SYBR Green PCR master mix (Life Technologies Corporation, Carlsbad, CA) for 40 amplification cycles. After completion of 40 PCR cycles, melt curve data was generated. All the measurements were done on three biological replicates consisting of three technical replicates each. Amplification of target sequences was done on ABI-Prism 7000 HT (Applied Biosystems, Foster City, CA). The C t values for primers were normalized against that of 16S rRNA. Fold change in the gene expression was calculated by 2 (−ΔΔCt) [44] and expressed as fold change ±SD.

AI-3 reporter assay
Preconditioned media (PM) was prepared as described [41]. Overnight cultures of TEVS232, TEVS21 and AV45 (EHEC ATCC 43895 harboring pVS150) were diluted 100 fold in LB medium and grown till OD600 ≈0.2. The cells were collected by centrifugation at 2500 × g for 10 min and resuspended in either fresh LB media supplemented with 50 μM epinephrine or PM and treated with 100 μg/ml isolimonic acid or equivalent amount of DMSO. The β-galactosidase activity was measured after 30 min incubation at 37°C using o-nitrophenyl β-Dgalactopyranoside as previously described [45] and reported as mean ± SD of three replicates.

Statistical analysis
Percent inhibition of biofilm formation was calculated from three experiments consisting of three replicate wells using the formula 100-[(OD 570 of sample well/ OD 570 of positive control) × 100]. Effects of different limonoids for each activity were analyzed using analysis of variance (ANOVA) followed by Tukey's pairwise  16.0 (SPSS Inc., Chicago, IL, USA). The effect was considered significant at p <0.05. The data for EHEC biofilm was fitted to a 3-parameter sigmoid models y= a/(1+exp(−(x-x0)/b)) using SIGMAPLOT 11.0 (Systat Software, Inc.). In order to conduct the analysis, concentration of each limonoids was converted to Log 10 μM and plotted against percent inhibition values.

Effect of citrus limonoids on EHEC growth and biofilm formation
The purity of all tested limonoids was >95% ( Figure 1). Furthermore, limonoids in the concentration range of 6.25-100 μg/ml, did not affect EHEC growth (Table 3) and viability (Additional file 1: Figure S1). All the five limonoids inhibit biofilm formation in concentration dependent manner ( Figure 2). Biofilm inhibitory activities of limonoids were compared by calculating IC 25

Effect of limonoids on adhesion of EHEC to Caco-2 cells
To further understand the effect of limonoids, adherence of EHEC to colon epithelial Caco-2 cells was studied. Isolimonic acid and ichangin (100 μg/ml) treatment significantly (p<0.05) reduced the number of EHEC cells attached to Caco-2 cells by 0.66 and 0.59 Log 10 cfu/ml, respectively ( Figure 3A). Isoobacunoic acid, IOAG and DNAG did not affect the number of EHEC cells adhering to Caco-2 cells. To determine, if the observed reduction in adhesion of EHEC was due to reduced cell viability of Caco-2 cells, survival of Caco-2 in presence of 100 μg/ml limonoids at 6 h was assayed by measuring extracellular LDH. Survival of Caco-2 cells in presence of 100 μg/ml limonoids was similar to solvent control ( Figure 3B).

Citrus limonoids repress the LEE, flagellar and stx2 genes
Adherence of EHEC to epithelial cells is facilitated by several factors including locus of enterocyte effacement (LEE) encoded TTSS, flagella, effector proteins and intimin [46][47][48]. To determine the probable mode of action, effect of limonoids on expression of six LEE encoded genes ler, escU, escR (LEE1 encoded), escJ, sepZ and cesD (LEE2 encoded), flagellar master regulators flhDC and stx2 was studied. Isolimonic acid and ichangin exerted the strongest effect on the LEE in EHEC grown to OD 600 ≈ 1.0 in LB media. The transcriptional regulator of LEE, the ler, was repressed 5 fold by isolimonic acid, while other LEE encoded genes were down-regulated by 6-10 fold (Table 4). Ichangin treatment resulted in ≈ 2.5-6 fold repression of LEE encoded genes. IOAG repressed the escU, escR, escJ and cesD by 3.2, 2.5, 3.7 and 2.6 fold, respectively while aglycone, isoobacunoic acid did not seem to affect the expression of LEE encoded genes under investigation (Table 4). Similarly, DNAG treatment did not resulted in differential expression of any genes. Furthermore, isolimonic acid repressed the flhC and flhD by 4.5 and 6.9 fold, respectively (Table 4), while ichangin exposure resulted in 2.8 fold repression of flhC and flhD. IOAG repressed flhC and flhD by 2.1 and 2.3 folds, respectively. Isoobacunoic acid and DNAG treatment did not seem to modulate the expression of flhDC (Table 4). Shiga toxin produced by EHEC is responsible for HUS [2]. We were further interested in learning if any of the limonoids modulate expression of stx2. Isolimonic acid and ichangin (100 μg/ml) repressed the stx2 by 4.9 and 2.5 fold, respectively (Table 4), while IOAG, isoobacunoic acid and DNAG did not seem to affect the expression of stx2.
The culture of EHEC in DMEM was reported to activate LEE expression [41]. To determine, if isolimonic acid represses LEE under DMEM growth conditions, expression of ler, stx2, escJ and sepZ were measured. Isolimonic acid treatment repressed ler, stx2, escJ and sepZ in DMEM media by >5, 7, 8 and 10 fold whereas, expression of rpoA was unaffected ( Figure 4). The escJ and sepZ, which are coded as a polycistronic message, demonstrated differing levels of regulation in presence of isolimonic acid ( Figure 4). However, differential degradation and processing of genes encoded as polycistronic mRNA is well documented [49,50], and could potentially be the reason of different levels of mRNA transcripts recorded for escJ and sepZ.

QseBC dependent inhibition of biofilm by isolimonic acid
QseBC is a two component system, which detects AI-3 and epinephrine and modulates biofilm formation and flagellar expression [6]. As isolimonic acid seems to interfere with AI-3/epinephrine induced pathway, it was possible that this interference is dependent on QseBC.
To determine if isolimonic acid inhibits EHEC biofilm formation by affecting QseBC, biofilm formation in The mean ± SD of three replicates are presented. Figure 2 Three parameter models of biofilm formation inhibition by citrus limonoids. Line curves at 50% and 25% represent the IC 50 and IC 25 values for compounds. Biofilms were grown in 96-well plates and quantified using crystal violet. Percent inhibition over solvent control (DMSO) was calculated. To generate 3-parameter models, concentrations were changed to Log10 μM and plotted against percent inhibition.
EHEC 86-24, QseC deletion mutant (VS138) and complemented strain VS179 [6] was studied. Since ΔqseBC strain (VS138) did not form appreciable biofilm at 24 h, the biofilms were grown up to 48 h. The biofilm formation in ΔqseBC at 48 h was similar between solvent control (DMSO) and isolimonic acid (p>0.05) ( Figure 6A). In contrast, isolimonic acid reduced the biofilm formation by 61.33% in complemented strain VS179. To further understand the role of QseBC in wild type strain ATCC 43895, plasmid pVS178 (carrying qseBC), was purified from VS179 and introduced into wild type strain. In addition, qseB and qseC were amplified from EHEC genomic DNA, cloned into pBAD33 vector and introduced into EHEC strain ATCC 43895. The expression of qseBC/qseB/qseC was induced by addition of 0.2% arabinose in the media. Overexpression of qseBC/qseC/qseB formed significantly more biofilm, when compared to EHEC wild type carrying vector alone ( Figure 6B). We further measured the effect of isolimonic acid on the biofilm formation in strains overexpressing qseBC/qseC/qseB ( Figure 6C). The isolimonic acid treatment did not significantly affect the biofilm formation, measured after 24 h of growth, in EHEC strains overexpressing qseBC/qseC/ qseB ( Figure 6C). Furthermore, it was possible that isolimonic acid modulates the expression of qseBC leading to inhibition of biofilm. To determine the effect of isolimonic acid, expression of qseB and qseC was measured by qRT-PCR. The results indicate that isolimonic acid do   The EHEC ATCC 43895 was grown to OD600≈1.0, RNA was extracted using RNeasy kit and converted to cDNA as described in text. Target genes were amplified from three biological samples. Fold change was calculated using 2 (−ΔΔCt) method and presented as mean ± SD of three replicates.
not regulate the expression of qseB and qseC ( Figure 6C). Altogether, finding of these experiments seem to suggest that isolimonic acid affects the QseBC activity but not the expression to inhibit biofilm formation.

QseA dependent inhibition of ler by isolimonic acid
Repression of LEE and interference of AI-3/epinephrine mediated signaling by isolimonic acid prompted us to investigate the role of QseA. To determine the contribution of QseA, change in ler expression was monitored in qseA deletion (VS145) and complemented (VS151) strains.
Isolimonic acid (100 μg/ml) treated cultures demonstrated a <2 fold change in ler expression in qseA deletion mutant. In comparison, isolimonic acid repressed the ler by 7.4 fold in complemented strain VS151 ( Figure 7A). To further confirm the role of QseA, qseA was overexpressed by introducing the plasmid pVS150, harboring qseA, into reporter strain TEVS232 and expression of chromosomal fusion LEE1:LacZ (β-galactosidase activity) was measured.
Overexpression of qseA from a multicopy plasmid negated the inhibitory activity of isolimonic acid ( Figure 7B). Furthermore, the possibility of transcriptional regulation of qseA by isolimonic acid was determined by assessing the  qseA expression. A < 2 fold change in the transcript levels of qseA indicated that isolimonic acid do not regulate the expression of qseA ( Figure 7C). Altogether, the isolimonic acid appears to repress ler expression and possibly LEE by modulating QseA activity.

Discussion
EHEC is an important gastrointestinal pathogen, prolific biofilm former and demonstrates resistance to various antimicrobials in biofilm mode of growth [51]. For successful colonization of gastrointestinal tract and initiation of infection, adhesion of EHEC to intestinal epithelium is an essential early event [47,48]. Additionally, several E. coli pathovars were reported to produce and live in biofilms inside the human body [19]. In order to counteract these maladies, an antivirulence molecule with anti-adhesion and/or anti-biofilm properties may be highly desirable. Research in our laboratory has identified several molecules with differing anti-virulence effects [23,28,36,37,52,53]. The current work examined the potential of five citrus limonoids-isolimonic acid, ichangin, isoobacunoic acid, IOAG and DNAG, to inhibit EHEC biofilm and TTSS. All the tested limonoids seem to interfere with the EHEC biofilm formation in a dose dependent fashion (Figure 2). Isolimonic acid was the most potent inhibitor of the EHEC biofilm and adhesion to Caco-2 cells. Moreover, the limonoids do not seem to affect growth of EHEC, suggesting that limonoids, especially isolimonic acid inhibits EHEC biofilm and adhesion without adversely affecting the growth or metabolic activity (Table 1, Additional file 1: Figure S1). In EHEC, the initial attachment to various surfaces such as epithelial cells and plastic surface is regulated by several factors including TTSS, flagella and fimbriae [47,48,54]. LEE encoded TTSS, effector proteins as well as flagella and intimin [47,48] play an important role in adhesion of EHEC to gastrointestinal tract surface, while flagella and fimbriae also contribute in biofilm formation. Results of the adhesion and biofilm assay indicated that one or more of above-mentioned factors may be affected by limonoids particularly by isolimonic acid. To investigate this hypothesis, expression of LEE encoded genes and flagellar master regulators flhDC was determined by qRT-PCR. Isolimonic acid and ichangin appear to exert their antivirulence and biofilm inhibitory effect by repressing TTSS carried on LEE, stx2, which encodes for Shiga toxin and flagellar master regulon flhDC (Table 4).
In EHEC, expression of LEE and flagellar operons are regulated by multiple environmental and genetic factors including QS [10][11][12][13]. In particular AI-2/AI-3/epinephrine mediated cell-cell signaling regulates the expression of both flagellar operon and LEE, which contribute to adhesion and biofilm formation. Furthermore, expression of stx2 is also regulated by QS [2,12,55,56]. Therefore, repression of TTSS, flagella and stx2 indicated a possibility that limonoids, especially isolimonic acid may interfere with EHEC QS. Isolimonic acid was chosen for further studies, as it demonstrated the most potent inhibition of biofilm formation, adhesion, LEE, flhDC and stx2. For determination of AI-3/epinephrine mediated QS in EHEC, reporter strains TEVS 232 and TEVS21 containing chromosomal fusions LEE1:LacZ and LEE2: LacZ were used. The analysis was confined to LEE1 and LEE2, because these two operons have been reported to be directly activated by AI-3/epinephrine mediated QS [15,41]. To test if the isolimonic acid acts as an QS inhibitor, PM/epinephrine stimulated activation of LEE1 and LEE2 in reporter strains was measured [41]. The PM, described earlier [41], was used as a source of AI-3 molecules as the purified AI-3 was not available. Repression of AI-3/epinephrine-induced ler, LEE1 and LEE2 ( Figure 5) indicated that isolimonic acid interferes with EHEC QS system.
The autoinducers and hormones reportedly increase the autophosphorylation levels of histidine kinase QseC, which then activates QseB to regulate motility and biofilm formation [57]. Furthermore, interaction of AI-3/ epinephrine with QseA activates LEE encoded genes [15,57]. It was possible that isolimonic acid interferes with EHEC QS in a mechanism involving QseBC and QseA. If activity of isolimonic acid depends upon functional QseBC, deletion of qseBC will eliminate the inhibitory effect. On the other hand, complementation of ΔqseBC with plasmid borne QseBC is likely to restore the inhibitory effect of isolimonic acid. Furthermore, overexpression of qseBC in wild type background (EHEC ATCC 43895) will result in higher levels of QseBC proteins in the cell and consequently will have a higher activity. This higher level of activity may compensate and relieve the inhibitory effect of isolimonic acid on biofilm formation. In order to verify QseBC dependent inhibition, biofilm formation in ΔqseBC strain (VS138) and complemented strain (VS179) [6] in presence of 100 μg/ ml of isolimonic acid was measured. As expected, isolimonic acid did not reduce the biofilm formation in VS138. In contrast, isolimonic acid exposure resulted in a significant decrease in VS179 (qseBC complemented strain) biofilm as measured by crystal violet (Figure 6A), indicating involvement of QseBC. Additionally, overexpression of qseBC, qseB and qseC in EHEC ATCC 43895, under the control of arabinose operon restored the inhibitory effect of isolimonic acid on EHEC biofilm formation ( Figure 6B). Taken together these results suggest that effect of isolimonic acid is dependent upon QseBC. Furthermore, the effects of isolimonic acid did not seem to arise from modulation of qseBC expression. However, based on the current data it was not possible to differentiate, if the effect is dependent solely upon qseB or qseC, as supplementation of EHEC by both qseB and qseC relieved the inhibitory effect. Further studies are required to precisely determine if the target of isolimonic acid is qseB or qseC.
To understand the role of QseA in isolimonic acid mediated repression of LEE, expression levels of transcriptional regulator ler were measured as QseA is reported to directly activate expression of ler [15]. Ler is the transcriptional regulator of the genes encoded in LEE and activates the genes encoded in LEE [15,21]. We hypothesized that if isolimonic acid affect ler via QseA, the ler expression will not change in ΔqseA strain (VS145) but complementation of qseA (strain VS151) from plasmid will restore the inhibitory effect. In addition, overexpression of qseA in wild type strain ATCC 43895 will negate the inhibitory effect of isolimonic acid. The hypothesis was tested by measuring the expression of ler using qRT-PCR in VS145 and VS151, grown in presence of 100 μg/ml isolimonic acid and compared with DMSO. The results demonstrated that expression of ler was not significantly altered in ΔqseA strain (VS145), whereas a 7.4 fold repression of ler ( Figure 7A) was observed in qseA complemented strain (VS179). Furthermore, overexpression of qseA from multicopy plasmid pVS150 in TEVS232 background (AV46) nullified the repressive effect ( Figure 7B) of isolimonic acid on LEE1 observed in Figure 5A. Collectively the data indicated that repression of LEE by isolimonic acid is dependent on QseA. However, isolimonic acid does not seem to transcriptionally modulate the expression of qseA. Thus the results of the study indicate towards a model where isolimonic acid modulates the biofilm and TTSS in QseBC and QseA dependent fashion, however without regulating the expression of these genes (Figure 8).

Conclusion
The present study demonstrates that the citrus limonoids, particularly isolimonic acid and ichangin are strong inhibitors of biofilm formation and attachment of EHEC to Caco-2 cells. Furthermore, isolimonic acid and ichangin seems to affect biofilm formation and TTSS by repressing LEE and flagellar operon. Isolimonic acid seems to exert its effect by inhibiting AI-3/epinephrine mediated cell-cell signaling in QseBC and QseA dependent manner. However, the mechanism by which isolimonic acid affects the QseBC and QseA remains to be elucidated. One possibility is that the isolimonic acid may interfere with the DNA binding activities of QseB and QseA. Another possible scenario will be that isolimonic acid interferes with phosphorylation events. However, further study is required to determine the target of isolimonic acid for the modulation of flhDC and ler. In addition, determination of the structure-activity relationship will provide further insights into the inhibitory action of isolimonic acid. In nutshell, isolimonic acid acts as an antivirulence agent in EHEC and may serve as lead compound for development of novel agents. Furthermore, the fact that isolimonic acid is present in citrus juices and do not demonstrate cytotoxic effect on normal human cell line [58], increases the desirability to develop it as antivirulence agent.

Additional file
Additional file 1: Figure S1. Metabolic activity of E. coli O157:H7 in presence of 100 μg/ml limonoids as measured by AlamarBlue reduction.