In this study, the wild-type strain Chromobacterium violaceum ATCC 12472 and Pseudomonas aeruginosa PAO1 were used as biosensor strains for quorum sensing (QS) and motility assays, respectively. C. violaceum 12472 was a kind gift from Prof. Dr Robert J.C. McLean (University of Texas, TX, USA), and P. aeruginosa PAO1 was also a kind gift from Daniel Lopez, PhD (National Centre for Biotechnology (CNB), Autonomous University of Madrid, Madrid, Spain). The strains were routinely grown in Luria-Bertani (LB) broth (1% tryptone, 0.5% yeast extract and 1% NaCl (Sigma-Aldrich, Merck, St. Louis, MO, USA) medium; pH=7.0). Flasks with C. violaceum and P. aeruginosa PAO1 were incubated at 30 and 37 °C for 24 h, respectively.

The dietary phytochemicals (quercetin, curcumin, apigenin, pyrogallol and luteolin) were purchased from Sigma-Aldrich, Merck. Gallic acid and all solvents were from Merck (Darmstadt, Germany). The stock solutions of apigenin and luteolin were prepared in 10% dymethylsulphoxide (DMSO), those of curcumin and quercetin in 50% methanol, and pyrogallol and gallic acid were dissolved in distilled water. The final volume fraction of DMSO or methanol was adjusted to ≤0.5%, which did not have any discernible effect on the growth of C. violaceum 12472. The stock solutions were prepared daily and diluted to the desired volume fraction immediately prior to use. The control groups were 0.5% DMSO, 0.5% methanol and the bacteria cultured in LB broth.

Minimum inhibitory concentration (MIC) values of the dietary phytochemicals were determined by the broth microdilution test in 96-well plates as previously described (22) with some modifications. A volume of 100 μL of the stock solutions of the phytochemicals (50% V/V) was added to an equal volume of LB broth, and twofold serial dilutions (1500, 750, 375, 187.5, 93.75 and 46.87 µg/mL) were prepared in the microtiter plate. Overnight cultures of C. violaceum 12472 and P. aeruginosa PAO1 were adjusted to the absorbance value A_{600 nm}=0.4 using sterile broth, and 100 μL of each culture were added to the plates, after which the plates were incubated at 30 and 37 °C for 24 h, respectively. Inhibition of the bacterial growth in the wells containing the phytochemicals was assessed by a comparison with the growth in blank control wells. Every experiment included negative (medium, 0.5% DMSO, and 0.5% methanol) and positive control (medium including the inoculum). The MIC was recorded as the lowest concentration at which there was no visible growth of the bacteria. The MIC assay was repeated at least twice. The subminimal inhibitory concentrations (sub-MICs) of each phytochemical (46.87–750 µg/mL) were tested for the assessment of anti-QS and antibiofilm activity.

Biofilm formation in 96-well U-bottom polystyrene microtiter plates (Nunc™, Thermo Fisher Scientific, Waltham, MA, USA) was assayed via a previously described, slightly modified method (25). An overnight culture of C. violaceum 12472 was diluted 1:100 with LB broth and grown for another hour. After the addition of sub-MICs of the phytochemicals (46.87, 93.75, 187.5, 375 and 750 µg/mL), 100 µL of the culture were pipetted into the wells of the microtiter plates and the plates were incubated for 24 h at 30 °C. Then, the medium was removed and washed with 1×PBS buffer three times. The plates were dried at 65 °C in a universal oven (UNB 100; Memmert®, Schwabach, Germany) and then 100 µL of a 1% m/V aqueous solution of crystal violet were added. The stain was allowed to fix at room temperature for 20 min, after which the dye was removed from the wells by washing thoroughly with sterile water. For the quantification of the attached biomass, the bound dye was dissolved with 30% acetic acid solution, and the absorbance was determined at 595 nm. Inhibitor-mediated reduction of biofilm formation was assessed by comparing it to the positive control without phytochemicals. The biofilm assay was performed three times, and six wells per treatment were used each time.

The swimming and swarming motility assays were performed using a previously described, slightly modified method (26). In swimming assay, 5 µL of overnight culture of the Pseudomonas aeruginosa PAO1 (A_{600 nm}=0.4) were point inoculated at the centre of an agar medium consisting of 1% tryptone, 0.5% NaCl and 0.3% agar with the sub-MIC concentration of the phytochemicals (93.75 µg/mL). For swarming assays, the agar medium comprised 1% peptone, 0.5% NaCl, 0.5% agar and 0.5% filter-sterilized d-glucose with the same sub-MIC concentration of phytochemicals (93.75 µg/mL). The plates were then incubated at 37 °C in an upright position for 16 h. The reduction in swimming and swarming migration was recorded by measuring the swimming and swarming zones of the bacterial cells after 16 h compared to the negative controls.

To confirm the anti-QS activity of the dietary phytochemicals, the growth curve assay of C. violaceum 12472 cultivated in the presence and absence of phytochemicals was performed. Overnight culture of the bacteria (1%; A_{600 nm}=0.4) was inoculated in a 250-mL Erlenmeyer flask containing 50 mL of LB broth supplemented with the sub-MIC concentrations (375 and 750 µg/mL) of each phytochemical. The flasks were incubated at 30 °C and 180 rpm in a rotary shaker (KS300; IKA®-Werke GmbH & Co. KG, Staufen, Germany). The cell density was measured by UV-Vis spectrophotometer (UV-1800; Shimadzu) at 1-hour intervals up to 20 h. The control was the bacteria without the treatment with phytochemicals (19).

The results were analysed using SPSS (Statistical Package for the Social Sciences) v. 20.0 (27) and expressed as mean value±standard deviation (S.D.). The differences between the control and test samples were analysed using t-test and one--way ANOVA. Differences at p<0.05 were considered statistically significant.

MIC was determined for each of the dietary phytochemicals at the concentrations ranging from 1500 to 46.87 µg/mL against Chromobacterium violaceum 12472. The MIC value of gallic acid was found to be ≥1500 µg/mL. This result corroborates well the findings of Borges et al. (12), who reported the MIC value of gallic acid of >1000 μg/mL for the same strain. The effect of phenolic acids on the physicochemical properties of bacterial cell surface has shown that these compounds, in particular gallic acid, change bacterial hydrophobicity. As known, phenolic acids alter the polar, nonpolar and electron acceptor components of bacterial cells (28).

In addition to these results, the MIC values of curcumin and pyrogallol were found to be 1500 µg/mL and the MIC values of quercetin, luteolin and apigenin were found to be 750 µg/mL. Gopu et al. (6) found that the MIC value of quercetin was 120 μg/mL against C. violaceum CV026. This result differs from our data, which may be due to the use of different methods employed to detect the MIC values, variations in the preparation of the solutions of phytochemicals and/or due to the use of different biosensor strains. In our study, we believe that dietary phytochemicals affect bacterial biochemical activities responsible for bacterial growth.

Phytochemicals are known to have strong antimicrobial effects, which mainly cause structural or functional damage to the bacterial cell membrane (29). They could also show different target mechanisms of antimicrobial activity on bacterial cells, such as the degradation of the cell wall, the leakage of the cell contents, the depletion of the proton motive force, or cytoplasmic protein coagulation or inhibition (30). Ohemeng et al. (31) and Mirzoeva et al. (32) reported that quercetin inhibited DNA gyrase and disrupted the bacterial membrane potential. Sorrentino et al. (33) also reported that gallic acid caused irreversible changes in permeability profile, rupture and pore formation of the bacterial cell membranes. In our study, the dietary phytochemicals may use one of the target mechanisms to inhibit the growth of C. violaceum 12472.

The production of the purple pigment violacein in C. violaceum is controlled by QS (4). Loss of the violacein is a hallmark of QS inhibition in C. violaceum 12472 by the dietary phytochemicals. The phytochemicals showed promising anti-QS activity, and a white opaque zone of inhibition was observed in the biosensor plate containing the reference strain C. violaceum 12472 (Fig. 1). The sub-MICs of quercetin, luteolin and apigenin were found to be <750 μg/mL, and of curcumin, pyrogallol and gallic acid <1500 µg/mL. The sub-MICs of each tested phytochemical ranged from 46.87 to 750 µg/mL. As shown in Table 1, the anti-QS activity was concentration-dependent, showing an increase in the diameter of the QS inhibition zones with increasing concentrations of the phytochemicals (p<0.05). Among all the phytochemicals screened for the QS inhibition, curcumin, quercetin, luteolin and apigenin exhibited the anti-QS activity, but pyrogallol and gallic acid did not. Al-Hussaini and Mahasneh (34) reported the QS inhibition zones for different herbal extracts with diameters 8–10.5, 13 and 18 mm, corresponding to weak, moderate and strong anti-QS activity, respectively. In our study, the QS inhibition zones (Table 1) were found for curcumin ((7.0±0.3)–(13.3±0.1) mm), quercetin ((7.5±0.1)–(13.3±0.5) mm), luteolin ((9.1±0.1)–(14.2±0.1) mm) and apigenin ((9.3±0.7)–(14.3±0.0) mm) at the concentration range of 46.87–750 µg/mL. At low concentrations (46.87 and 93.75 µg/mL), these phytochemicals had weak anti-QS activities. In addition, curcumin ((9.7±0.1) mm) and quercetin ((9.5±0.7) mm) exhibited weaker anti-QS activities than luteolin ((12.3±0.6) mm) and apigenin ((11.5±0.5) mm) at 187.5 µg/mL. However, curcumin ((12.5±0.3) mm), luteolin ((13.6±0.5) mm) and apigenin ((13.0±0.1) mm) showed stronger anti-QS activities than quercetin ((12.0±0.2) mm) at 375 µg/mL (p<0.05). At the highest sub-MIC (750 µg/mL), curcumin demonstrated the highest QS inhibition (the diameter of zone of inhibition being (13.3±0.1) mm; Table 1). The concentration-dependent anti-QS activity of the phytochemicals observed in this study is in agreement with some of the findings reported by Borges et al. (12), Husain et al. (15), and Vasavi et al. (20).

To confirm the QS inhibitory activity of the dietary phytochemicals, the extraction and the quantification of the violacein from C. violaceum 12472 culture was also performed in the presence and absence of the phytochemicals (curcumin, quercetin, pyrogallol, luteolin, gallic acid and apigenin) at the same sub-MIC levels (46.87–750 µg/mL). The results showed a concentration-dependent inhibition of the violacein production by curcumin, quercetin, luteolin and apigenin. Compared to the control, all concentration ranges of curcumin, quercetin, luteolin and apigenin (46.87–750 µg/mL) demonstrated a significant drop in the violacein content of C. violaceum 12472, without the inhibition of bacterial growth (p<0.05). As shown in Fig. 2, curcumin, quercetin, luteolin and apigenin inhibited the violacein production ((14.0±0.3)–(88.2±0.1) %, (11.1±0.1)–(55.03±0.07) %, (37.49±0.08)–(59.48±0.08) % and (36.3±0.2)–(58.95±0.08) %) in a concentration-dependent manner. Among all the phytochemicals, curcumin exhibited the most powerful QS inhibitory effect ((54.7±0.3), (72.2±0.2) and (88.2±0.1) %) at 187.5, 375 and 750 µg/mL. At all concentrations, luteolin and apigenin demonstrated stronger violacein inhibition than quercetin (Fig. 2). The results revealed that the percentage of inhibition of QS by apigenin and luteolin was similar at the tested concentrations. This may be because apigenin and luteolin are the flavonoid compounds with similar structures. Thus, the QS inhibitory effect of the phytochemicals was found to be in the following order: curcumin>apigenin>luteolin>quercetin.

Brackman et al. (35) found that curcumin inhibited violacein production by (18±12) % in C. violaceum CV026 at a concentration of 184 µg/mL. In our findings, curcumin inhibited violacein production by (54.7±0.3) % at 187.5 µg/mL. In addition, in the present study, the violacein inhibition by 750 µg/mL curcumin ((88.2±0.1) %) is comparable with that of Packiavathy et al. (19), who reported 89% violacein inhibition in C. violaceum CV026 when treated with 100 µg/mL curcumin. Moreover, the violacein inhibition by quercetin at the maximum sub-MIC of 375 µg/mL ((39.9±0.1) %) is comparable with that of Gopu et al. (6), who reported that quercetin exhibited maximum violacein inhibition of 83.23% in C. violaceum CV026 at a concentration of 80 μg/mL. Similarly, 50% inhibition of the violacein production in C. violaceum 12472 by quercetin (50 µg/mL) was reported by Vasavi et al. (20) using a different method.

Apigenin and luteolin inhibited violacein ((58.4±0.2) and (55.13±0.04) %) at 375 µg/mL. Vandeputte et al. (36) reported that apigenin and luteolin at the concentration of 4 mM (approx. 1.2 mg/mL) exhibit no QS inhibiton, but a bactericidal or bacteriostatic activity on C. violaceum CV026. Our study showed that apigenin and luteolin at the sub-MICs of 375, 187.5, 93.75 and 46.87 µg/mL showed anti-QS activity on C. violaceum 12472 without inhibition of bacterial growth (Fig. S1, Fig. 1a, Fig. 1d and Fig. 2).

Pyrogallol and gallic acid did not exhibit the anti-QS activity when evaluated by the disk diffusion method. Therefore, the percentage of the violacein inhibition by these phytochemicals was evaluated only at the highest concentration (1500 µg/mL), and the obtained results were (76.95±0.02) and (58.76±0.05) % for pyrogallol and gallic acid respectively (data not shown). Considering that no QS inhibition zones were detected on the agar, these results were observed solely due to the inhibition of bacterial growth since gallic acid and pyrogallol only inhibited microbial growth and not the violacein synthesis. The obtained results are also in agreement with those of Borges et al. (12), who reported that gallic acid exhibited no anti-QS activity on C. violaceum biosensor systems.

Previous research of Borges et al. (12, 37) also revealed that as a sustainable source of new broad-spectrum antimicrobial products, gallic acid can cause irreversible changes in some Gram-negative and Gram-positive bacteria. In addition, Tinh et al. (38) reported that pyrogallol also exhibited the antibacterial activity against Vibrio parahaemolyticus. Although pyrogallol is known as the phytochemical with anti-QS activity and the ability to inhibit autoinducer-2 (AI-2)-mediated QS in Vibrio harveyi (39, 40), it has been revealed in further investigations that this AI-2-mediated QS inhibition is a side effect of the peroxide-producing activity of this compound rather than true QS inhibition (41). Our findings also show that pyrogallol does not inhibit QS system of the C. violaceum 12472 biosensor strain.

In the present study, the biofilm inhibition potential of all the phytochemicals (curcumin, quercetin, pyrogallol, luteolin, gallic acid and apigenin) was evaluated at the same sub- -MIC range (46.87–750 µg/mL) (Fig. 3). All the phytochemicals exhibited a significant concentration-dependent antibiofilm activity (p<0.05). Gallic acid showed the weakest biofilm inhibition (from (1.38±0.08) to (9.57±0.06) %) at 46.87–750 µg/mL, although it did not possess antimicrobial activity against C. violaceum 12472 at <1500 µg/mL. Compared to the other compounds (pyrogallol, apigenin, quercetin and luteolin), curcumin showed the weakest biofilm inhibition ((11.0±0.5) and (26.7±0.3) % at 46.87 and 93.75 µg/mL, respectively). Our results are comparable with the results of Packiavathy et al. (18), who found that curcumin (>100 μg/mL) inhibited the biofilm formation of Vibrio harveyi (69%), Vibrio parahaemolyticus (56%) and Vibrio vulnificus (79%), without affecting their planktonic growth. Packiavathy et al. (19) also reported that curcumin (100 µg/mL) efficiently inhibited the biofilm biomass growth of some uropathogens, including Escherichia coli 
(52%), Pseudomonas aeruginosa PAO1 (89%), Proteus mirabilis (52%) and Serratia marcescens (76%). In addition, the combined effects of curcumin and honey, as a traditional medicine, and epigallocatechin gallate, as a green tea polyphenol, were reported to enhance significantly the inhibition of biofilm formation in wastewater bacteria (52 to 99%) and P. aeruginosa PAO1 (20 to 94.6%) (21, 42). Therefore, the antibiofilm activity of curcumin may vary depending on the type of strain, the method used and/or its combination with some other constituent. Our results (Fig. 3) showed that although curcumin exhibited a weak biofilm inhibition potential at concentrations of 46.87 and 93.75 µg/mL, at higher sub-MICs of 187.5, 375 and 750 µg/mL, it had a strong concentration-dependent activity ((57.7±0.4), (75.6±1.0) and (84.2±0.2) %; p<0.05). Fig. 3 shows that the biofilm inhibitory potentials of the phytochemicals at concentrations of 46.87 and 93.75 µg/mL increased in the following order: gallic acid<curcumin< pyrogallol<quercetin<apigenin<luteolin. At these concentrations, luteolin had the maximum biofilm inhibition values of (46.94±0.05) and (60.0±0.2) %, respectively. Quercetin, apigenin and pyrogallol also significantly reduced (p<0.05) the biofilm formation of C. violaceum 12472 (Fig. 3).

Pyrogallol showed weak biofilm inhibition with (34.2±0.2), (36.4±0.2) and (38.5±0.0) % at higher sub-MICs of 187.5, 375 and 750 µg/mL, respectively (Fig. 3). Thus, the biofilm inhibitory potentials of all the phytochemicals at higher sub-MICs of 187.5 and 375 µg/mL increased in the following order: gallic acid<pyrogallol<quercetin<apigenin<curcu-min<luteolin. Luteolin had the maximum biofilm inhibition ((46.94±0.05)–(83.9±0.2) %) at the sub-MIC values of 46.87 to 375 µg/mL. Our results are in accordance with the findings of Vikram et al. (43), who reported that quercetin and apigenin reduced the biofilm formation in V. harveyi and E. coli O157:H7. In addition, Shehzad et al. (44) and Lee et al. (45) also reported that curcumin, pyrogallol and apigenin were effective polyphenols against the biofilm formation of Candida albicans.

In this study, the motility (swimming and swarming behaviour) of P. aeruginosa PAO1 in the presence and absence of the dietary phytochemicals at the sub-MIC of 93.75 µg/mL was evaluated using agar plates. Our results demonstrate that all the phytochemicals significantly (p<0.05) blocked the swimming and swarming motility of P. aeruginosa PAO1 without impairing its growth capacity. Its swimming and swarming motilities after the treatment with the phytochemicals were observed to be significantly poor (p<0.05; Table 2). Compared to the control, curcumin exhibited the maximum reduction in the swimming and swarming motility assays. These results are in accordance with the previous reports of Packiavathy et al. (19) and Jadaun et al. (42), who observed a remarkable decrease in the swimming and swarming motility of P. aeruginosa PAO1 when treated with curcumin. Table 2 also indicates the inhibition of swimming and swarming behaviour of P. aeruginosa PAO1 by quercetin.

Our findings are consistent with the reports of Vasavi et al. (20), who found that the flavonoid fraction of Psidium guajava leaves included an active compound, quercetin-3-O-arabinoside, inhibits the swarming motility of P. aeruginosa PAO1. In addition, Gopu et al. (6) reported that quercetin significantly reduced the swimming and swarming behaviour of foodborne isolates of Pseudomonas aeruginosa PUFSTb04 at the concentration of 80 μg/mL. Table 2 shows that the inhibition potential of the swimming and swarming motilities of the bacterial strain by the investigated phytochemicals decreased in the following order: curcumin>quercetin>pyrogallol>apigenin>luteolin>gallic acid.

The obtained results clearly indicate that curcumin and quercetin exhibited the most powerful inhibition of the motility. The results also suggest a correlation between the inhibition of the swimming and swarming motility for each phytochemical.

The bacterial growth curve revealed that the sub-MIC concentrations of dietary phytochemicals used in this study (375 and 750 µg/mL) did not have a growth inhibitory effect on C. violaceum 12472 (Fig. S1).