Propolis Affects Pseudomonas aeruginosa Growth, Biofilm Formation, eDNA Release and Phenazine Production: Potential Involvement of Polyphenols

Pseudomonas aeruginosa (P. aeruginosa) is an opportunistic pathogen responsible for a wide range of clinical conditions, from mild infections to life-threatening nosocomial biofilm-associated diseases, which are particularly severe in susceptible individuals. The aim of this in vitro study was to assess the effects of an Albanian propolis on several virulence-related factors of P. aeruginosa, such as growth ability, biofilm formation, extracellular DNA (eDNA) release and phenazine production. To this end, propolis was processed using three different solvents and the extracted polyphenolic compounds were identified by means of high performance liquid chromatography coupled to electrospray ionization mass spectrometry (HPLC-ESI-MS) analysis. As assessed by a bioluminescence-based assay, among the three propolis extracts, the ethanol (EtOH) extract was the most effective in inhibiting both microbial growth and biofilm formation, followed by propylene glycol (PG) and polyethylene glycol 400 (PEG 400) propolis extracts. Furthermore, Pseudomonas exposure to propolis EtOH extract caused a decrease in eDNA release and phenazine production. Finally, caffeic acid phenethyl ester (CAPE) and quercetin decreased upon propolis EtOH extract exposure to bacteria. Overall, our data add new insights on the anti-microbial properties of a natural compound, such as propolis against P. aeruginosa. The potential implications of these findings will be discussed.


Introduction
In nature, many microbial species use a cell-to-cell signaling system, named quorum sensing (QS), to form biofilms on both biotic and abiotic surfaces. Microorganisms embedded in a biofilm

Microbial Strain
The bioluminescent P. aeruginosa strain (P1242) (BLI-Pseudomonas) was maintained and handled as previously described [11]. This strain had been engineered to express the luciferase gene and luciferase substrate under the control of a constitutive P1 integron promoter. Thus, live cells constitutively produced a detectable bioluminescent signal [30]. Bacteria from −80 • C glycerol stocks were initially seeded onto Tryptic Soy Agar (TSA) (OXOID, Milan, Italy) plates and incubated overnight at 37 • C; then, isolated colonies were collected, added to 10 mL of Tryptic Soy Broth (TSB) (OXOID, Milan, Italy) and allowed to grow overnight at 37 • C with gentle shaking. Bacterial concentrations were then measured by the McFarland standard curve and appropriate dilutions prepared, according to the protocol.

Propolis
Crude propolis, produced by Apis mellifera L., was collected on June 2018, in north Albania (Puka, Thethi and Velë forest areas), far from asphalted urban or interurban roads. The woodland vegetation in this area, mainly consists of Picea abies L., Populus alba L., Aesculus hippocastanum L., Castanea sativa Miller, Arbutus unedo L., Quercus pubescens Wild., Prunus dulcis (Mill.) D.A. Webb, Prunus avium L., Helianthus annuus L., Cucumis sativus L., Ulmus L., Thymus vulgaris L., Cornus mas L., Salvia officinalis L., Acacia penninervis DC, Salix L., Cytisus scoparius L. & Link, Trifolium L. and Carpinus L. The crude propolis appeared as a mild amorphous mass, with an aromatic balsamic smell, in dark yellow to light brown color, with a bitter and burning taste numbing the oral mucosa. It was delivered at room temperature (RT) in the dark and kept at −80 • C until the extraction procedure.

Extraction of Phenolic Compounds from Crude Propolis
One gram of an Albanian frozen propolis (kept at −80 • C) was grinded in a mortar and reduced to uniform particle size powder. The extraction was carried out by dynamic maceration with 10 mL of solvent (i.e., EtOH, PG and PEG 400) under the dark for 24 h, at RT. The extracts were centrifuged for 5 min at 4000 rpm. The supernatant solutions were filtered in a vacuum into a 10 mL volumetric flask and the solvents were added to the final volume. The extraction procedure was repeated twice for each solvent tested.

Spectrophotometric Analysis of Total Phenolics
The total phenolic content was determined by using the Folin-Ciocâlteu colorimetric assay with some modifications [31]. A solution of gallic acid in water, at different concentration (2-20 µg/mL), was used as the reference. The total phenolic content was assessed by using 50 µL of each extract, previously diluted 1:1 with H 2 O. Fifty µL of pure solvent (EtOH, PG and PEG 400) were used as the blank, respectively. Then, 500 µL of the Folin-Ciocâlteu reagent and 1 mL of sodium carbonate (Na 2 CO 3 ) saturated solution were added. The solution was then adjusted to the final volume of 10 mL with H 2 O. Afterwards, the solutions were incubated at RT in the dark and, after 2 h, the absorbance Microorganisms 2020, 8, 243 4 of 16 was evaluated at 760 nm wavelength. The phenolic content was determined from the equation of the regression curve and expressed as mg of gallic acid equivalents for mL of propolis extract (mg GAE/mL).

Sample Preparation for HPLC Analysis
An aliquot of 200 µL of each propolis extract was properly diluted with 1 mL of EtOH in a volumetric flask, filtered through a 0.45 µm PTFE filter into a HPLC vial and injected in the HPLC system. All sample preparations were carried out in duplicate.

HPLC Analysis of Phenolics in Propolis Extracts
Chromatography was performed using an Agilent Technologies (Waldbronn, Germany) modular model 1100 system, consisting of a vacuum degasser, a quaternary pump, an autosampler, a thermostated column compartment and a diode array detector (DAD). The chromatograms were recorded using an Agilent ChemStation for LC and LC-MS systems (Rev. B.01.03).
HPLC coupled with electrospray ionization mass spectrometry (ESI-MS) analyses were performed using an Agilent Technologies modular 1200 system, equipped with a vacuum degasser, a binary pump, an autosampler, a thermostatted column compartment and a 6310 A ion trap mass analyzer with an ESI ion source. The HPLC column and the applied chromatographic conditions were the same as reported for the HPLC-DAD system. The flow-rate was split 6:1 before the ESI source. For ESI-MS 2 , the parameters were set as follows: the capillary voltage was 3.5 kV, the nebulizer (N 2 ) pressure was 20 psi, the drying gas (N 2 ) temperature was 350 • C, the drying gas flow was 9 L/min and the skimmer voltage was 40 V. Data were acquired by Agilent 6300 Series Ion Trap LC/MS system software (version 6.2). The mass analyzer was used in the full-scan positive and negative ion modes in the m/z range 100-1000. MS 2 spectra were automatically performed with helium as the collision gas by using the SmartFrag function.

Minimal Inhibitory Concentration (MIC) Assay
The MIC assay was performed by the microbroth dilution method according to the Clinical and Laboratory Standards Institute/National Committee for Clinical Laboratory Standard (CLSI/NCCLS M7-A6) [32]. According to the experimental protocol, each propolis extract was tested at final dilutions, ranging from 500 µg GAE/mL to 1.9 µg GAE/mL. In parallel, each solvent (at the corresponding dilutions) and gentamicin (2 mg/mL) were included as negative and positive controls, respectively. A bacterial cell suspension (5 × 10 5 cells/mL in Mueller Hinton plus 2% sucrose, obtained from overnight cultures) was seeded (100 µL/well) in a 96 U-bottom microtiter-plate; then, the bacterial cells were added with medium (100 µL/well) or treated with scalar doses of propolis extracts or their respective solvent (100 µL/well). Therefore, the plate was incubated at 37 • C for 24 h. The MIC of each extract was defined as the lowest concentration that inhibited visible Pseudomonas growth and in which the relative concentration of the solvent showed the minimal toxicity.

Assessment of Propolis Effects on Microbial Growth and Early Biofilm Formation
BLI-Pseudomonas is known to produce biofilm, as detailed elsewhere [11]. In order to monitor the total microbial growth under different experimental conditions, 100 µL of overnight cultures of BLI-Pseudomonas (5 × 10 5 /mL) in TSB plus 2% sucrose were seeded in a 96 black well-plate, containing 100 µL of TSB 2% sucrose (untreated) or propolis extract (treated) or solvent (control). The plates were then incubated at 35 • C for 16 h, into the Fluoroskan reader and the bioluminescence was detected at every hour. The values, collected in real time as a bioluminescence signal and expressed as relative luminescence units (RLU)/s, indicated the total microbial growth; based on an internal reference curve, such values could be converted in colony forming units (CFU)/mL.
In order to measure biofilm production by BLI-Pseudomonas, the samples were incubated for 16 h, washed twice with phosphate buffered saline (PBS) at RT to remove the planktonic cells and then the bioluminescence signal was measured. Such values were referred to the amounts of early biofilm formed onto plate surfaces, in treated or untreated samples.

Assessment of Phenazines and Propolis Polyphenols in Cell-Free Supernatants
In order to evaluate the amount of phenazines and to determine the levels of propolis polyphenols in supernatants from Pseudomonas exposed or not to propolis, a suitable HPLC-ESI-MS analysis was performed. To do this, all supernatants were filtered by Amicon Ultra-0.5 10 K centrifugal filter devices and diluted 1:5 (v/v) with 5% MeOH-0.2% HCOOH in H 2 O. The HPLC-ESI-MS instrument used was an UltiMate 3000 system, consisting of an online degasser, a binary pump HPG 3400RS, a well plate autosampler WPS 3000RS and a thermostatted column compartment TCC 3000RS coupled to a Q Exactive hybrid quadrupole-orbitrap mass analyzer via a HESI-II heated electrospray ion source (Thermo Scientific). Chromatographic separation of a 5 µL sample injection was performed on a Zorbax SB-C 18 RRHT (50 × 2.1 mm I.D., 1.8 µm) column (Agilent) at 25 • C and a 0.3 mL/min flow rate. A linear gradient elution scheme was used with mobile phase components, being 0.1% HCOOH in H 2 O (A) and MeOH (B). The gradient started at 2% B, which was maintained for 0.5 min, then raised up to 42% B for 30 min and up again to 95% B for 4 min. The column was kept at 95% B for 4.4 min; then, the starting conditions were restored in 0.1 min and maintained for 11 min pending a successive injection. Electrospray ionization was operated in positive ion mode, using N 2 as the sheath gas (40 arbitrary units), auxiliary gas (290 • C, 30 arbitrary units) and sweep gas (two arbitrary units). The sprayer voltage was kept at 3.5 kV and the transfer capillary temperature was set at 320 • C. The Q-Exactive was operated in Full MS/dd-MS 2 mode. The full MS scan range was set from m/z 150 to 1000 at 70,000 full width at half maximum (FWHM) resolution (m/z 200). The automatic gain control (AGC) target was set at 1.0 × 10 6 with a maximum injection time (IT) of 200 ms. Data-dependent MS 2 (dd-MS 2 ) acquisitions at 17,500 FWHM resolution (m/z 200) were triggered for the Top 3 precursor ions following each full MS scan. The intensity threshold for precursor ion selection was set to 1.0 × 10 5 then dynamic exclusion was active for 20.0 s. AGC target and maximum IT for the MS 2 experiments were set to 5.0 × 10 5 and 80 ms.

Assessment of Propolis Effects on eDNA Release
For the analysis of eDNA, 100 µL of overnight cultures of BLI-Pseudomonas (5 × 10 5 /mL) in TSB plus 2% sucrose were seeded in 96 well-plate, containing 100 µL of medium (untreated) or propolis extract (treated) or solvent (control). The plates were then incubated at 35 • C for 16 h. The supernatants were collected and centrifuged twice at 14,000 rpm for 15 min and filtered by Amicon Ultra-0.5 10 K centrifugal filter devices in order to remove any remaining bacteria. To exclude residual viable bacteria, 50 µL of the supernatants were seeded onto TSA plates and incubated for 48 h at 37 • C under aerobic conditions; no bacterial CFU on TSA plates were ever observed. To quantify eDNA concentration in the cell-free supernatants, 100 µL of each sample were incubated with Propidium Iodide (PI) (1 µg/mL) for 15 min at 37 • C; then, the fluorescence emission was quantified by Fluoroskan reader (excitation/emission: 584/612).

Statistical Analysis
Quantitative variables were tested for normal distribution. Statistical differences between propolis treated and untreated (solvent) samples were analyzed according to Mann-Whitney test by using GraphPad prism 8. Values of p < 0.05 were considered significant.

Total Phenolic Compounds in Propolis Extracts
Initially, the total polyphenol content of the three different propolis extracts (obtained by EtOH, PG and PEG 400) was determined by means of the Folin-Ciocâlteu colorimetric method. In particular, the PEG 400 extract showed the highest polyphenol content (5.8 ± 0.2 mg GAE/mL), while PG and EtOH extracts showed values of 4.8 ± 0.5 and 4.1 ± 0.4 mg GAE/mL, respectively.

HPLC Analysis of Polyphenols in Propolis Extracts
In order to investigate the presence of polyphenols on the three propolis extracts, a HPLC-UV/DAD analysis was performed and representative chromatograms are shown in Figure 1.
The compounds present in propolis extracts were identified by comparing the retention times of each peak with those of the standards, and by UV/Vis, MS and MS 2 data [33]. The list of the polyphenols detected in propolis is shown in Table 1. The composition of the Albanian propolis extracts appeared qualitatively similar, irrespectively of the solvent used, likely because of the similar extraction properties of the solvents applied in this work. Quantitative variables were tested for normal distribution. Statistical differences between propolis treated and untreated (solvent) samples were analyzed according to Mann-Whitney test by using GraphPad prism 8. Values of p < 0.05 were considered significant.

Total Phenolic Compounds in Propolis Extracts
Initially, the total polyphenol content of the three different propolis extracts (obtained by EtOH, PG and PEG 400) was determined by means of the Folin-Ciocâlteu colorimetric method. In particular, the PEG 400 extract showed the highest polyphenol content (5.8 ± 0.2 mg GAE/mL), while PG and EtOH extracts showed values of 4.8 ± 0.5 and 4.1 ± 0.4 mg GAE/mL, respectively.

HPLC Analysis of Polyphenols in Propolis Extracts
In order to investigate the presence of polyphenols on the three propolis extracts, a HPLC-UV/DAD analysis was performed and representative chromatograms are shown in Figure 1.  Table 1.
The compounds present in propolis extracts were identified by comparing the retention times of each peak with those of the standards, and by UV/Vis, MS and MS 2 data [33]. The list of the polyphenols detected in propolis is shown in Table 1. The composition of the Albanian propolis extracts appeared qualitatively similar, irrespectively of the solvent used, likely because of the similar extraction properties of the solvents applied in this work.  Table 1.

Antibacterial Activity of Propolis
The antimicrobial activity of the three propolis extracts was evaluated in vitro, according to the standardized CLSI/NCCLS method [32]. In detail, nine different dilutions of each extract and their corresponding solvent dilutions were assessed by microbroth dilution. The MIC values obtained for both propolis EtOH and PG extracts were 15.6 µg/mL. Differently, the MIC obtained for propolis PEG 400 extract was as high as 62.5 µg/mL. Based on these results, a kinetic analysis of Pseudomonas growth upon exposure to each of the three propolis extracts was carried out, by means of a BLI-based assay known to provide direct and real time assessment of viable cells [11]. For each propolis extract, the MIC values and their corresponding diluted solvent or the medium alone were used ( Figure 2). As depicted in Figure 2A, the RLU/s observed in the two controls (medium and EtOH) were similar, although an appreciable anticipation of the curve was observed with EtOH; in contrast, minimal or no RLU/s were ever detected in propolis extract treated samples. These differences were statistically significant, when comparing propolis treated-to solvent treated-P. aeruginosa, within the 12 to 16 h time frame. As shown in Figure 2B, in the medium and PG solvent controls, the RLU/sec had comparable time-related trends, with a slight anticipation by the latter. Propolis PG extract also significantly affected bacterial growth; in particular, a detectable signal occurred at 9 h reaching the highest value (30.47 RLU/s) at 15 h. Statistical significance was achieved at 12-14 h when comparing propolis extract with its solvent. Then, a slight decrease occurred in the luminescence signal down to 26.5 RLU/s at 16 h. When P. aeruginosa was grown in the presence or absence of PEG 400 propolis extract, a major toxicity of the solvent per se was evident ( Figure 2C). In fact, all the RLU/s were consistently lower than the corresponding medium values. As for PG, the PEG 400 propolis extract did not completely affect bacterial growth. Statistical significance was reached within the 13-15 h time frame, when comparing propolis extract with solvent. When in parallel groups, gentamicin was used as the positive control, a complete inhibition of BLI-Pseudomonas growth was observed, as shown by the little or no luminescence signal detected (Panels A, B and C). Moreover, as depicted in Figure 2 (right panels), the conversion of the RLU in CFU/mL at 16 h allowed to underline that the most consistent inhibitory effects were indeed observed upon exposure to propolis EtOH extract (approximately 2.5 log decrease). used ( Figure 2). As depicted in Figure 2A, the RLU/s observed in the two controls (medium and EtOH) were similar, although an appreciable anticipation of the curve was observed with EtOH; in contrast, minimal or no RLU/s were ever detected in propolis extract treated samples. These differences were statistically significant, when comparing propolis treated-to solvent treated-P. aeruginosa, within the 12 to 16 h time frame. As shown in Figure 2B, in the medium and PG solvent controls, the RLU/sec had comparable time-related trends, with a slight anticipation by the latter. Propolis PG extract also significantly affected bacterial growth; in particular, a detectable signal occurred at 9 h reaching the highest value (30.47 RLU/s) at 15 h. Statistical significance was achieved at 12-14 h when comparing propolis extract with its solvent. Then, a slight decrease occurred in the luminescence signal down to 26.5 RLU/s at 16 h. When P. aeruginosa was grown in the presence or absence of PEG 400 propolis extract, a major toxicity of the solvent per se was evident ( Figure 2C). In fact, all the RLU/s were consistently lower than the corresponding medium values. As for PG, the PEG 400 propolis extract did not completely affect bacterial growth. Statistical significance was reached within the 13-15 h time frame, when comparing propolis extract with solvent. When in parallel groups, gentamicin was used as the positive control, a complete inhibition of BLI-Pseudomonas growth was observed, as shown by the little or no luminescence signal detected (Panels A, B and C). Moreover, as depicted in Figure 2 (right panels), the conversion of the RLU in CFU/mL at 16 h allowed to underline that the most consistent inhibitory effects were indeed observed upon exposure to propolis EtOH extract (approximately 2.5 log decrease). show the total growth of Pseudomonas exposed to EtOH (A), PG (B) and PEG 400 (C) propolis extracts at their MICs, as measured kinetically by the BLI-based assay; in the right panels, the data are shown as total CFU/mL at 16 h. The respective solvents at the same dilutions were tested as negative controls; gentamicin (2 mg/mL) was used as the positive control. The results were expressed as mean ± SEM of the RLU/sec of 6-8 replicate samples obtained in two independent experiments. An internal calibration curve was used to convert to the RLU/s in total CFU/mL detectable at time 16 h. n.d.: not detectable. Statistical analysis was performed according to Mann Whitney test. * p < 0.05 propolis treated vs. solvent. Figure 2. P. aeruginosa growth in the presence or absence of the three propolis extracts. The left panels show the total growth of Pseudomonas exposed to EtOH (A), PG (B) and PEG 400 (C) propolis extracts at their MICs, as measured kinetically by the BLI-based assay; in the right panels, the data are shown as total CFU/mL at 16 h. The respective solvents at the same dilutions were tested as negative controls; gentamicin (2 mg/mL) was used as the positive control. The results were expressed as mean ± SEM of the RLU/sec of 6-8 replicate samples obtained in two independent experiments. An internal calibration curve was used to convert to the RLU/s in total CFU/mL detectable at time 16 h. n.d.: not detectable. Statistical analysis was performed according to Mann Whitney test. * p < 0.05 propolis treated vs. solvent.

Propolis Effects on P. aeruginosa Early Biofilm Formation
Bacterial cells, exposed or not to each propolis extract (used at its own MIC, i.e., 15.6 µg GAE/mL for propolis EtOH and PG extracts and 62.5 µg GAE/mL for propolis PEG 400 extract), were allowed to form a 16 h-old biofilm. Then, the wells were washed to remove the planktonic cells and the residual bioluminescent signal was evaluated to measure the early biofilm produced under the different experimental conditions (Figure 3). Propolis EtOH extract greatly prevented biofilm formation, as indicated by the low bioluminescent signal (panel A, pink column: 3.4 RLU/s). As shown in the right insert of Figure 3A, this drop corresponded to a significant inhibition (81%) in biofilm formation, when propolis extract was compared with the solvent. Differently, Pseudomonas cells exposed to EtOH alone produced a biofilm comparable to that developed in the presence of the medium alone (6% inhibition). When BLI-Pseudomonas was exposed to propolis PG extract, a still significant biofilm reduction (39%) was observed when propolis extract was compared with the solvent (right insert of Figure 3B), while the solvent per se had irrelevant effect (5% inhibition only). When PEG 400 was employed, an inhibitory effect by the solvent per se was observed (31.7%, with respect to medium); a further decrease occurred upon propolis PEG 400 extract treatment (38%, propolis vs. solvent) as shown in the right insert of Figure 3C.

Propolis effects on P. aeruginosa Early Biofilm Formation
Bacterial cells, exposed or not to each propolis extract (used at its own MIC, i.e., 15.6 μg GAE/mL for propolis EtOH and PG extracts and 62.5 μg GAE/mL for propolis PEG 400 extract), were allowed to form a 16 h-old biofilm. Then, the wells were washed to remove the planktonic cells and the residual bioluminescent signal was evaluated to measure the early biofilm produced under the different experimental conditions (Figure 3). Propolis EtOH extract greatly prevented biofilm formation, as indicated by the low bioluminescent signal (panel A, pink column: 3.4 RLU/s). As shown in the right insert of Figure 3A, this drop corresponded to a significant inhibition (81%) in biofilm formation, when propolis extract was compared with the solvent. Differently, Pseudomonas cells exposed to EtOH alone produced a biofilm comparable to that developed in the presence of the medium alone (6% inhibition). When BLI-Pseudomonas was exposed to propolis PG extract, a still significant biofilm reduction (39%) was observed when propolis extract was compared with the solvent (right insert of Figure 3B), while the solvent per se had irrelevant effect (5% inhibition only). When PEG 400 was employed, an inhibitory effect by the solvent per se was observed (31.7%, with respect to medium); a further decrease occurred upon propolis PEG 400 extract treatment (38%, propolis vs. solvent) as shown in the right insert of Figure 3C.

Propolis Effects on Phenazines Release by P. aeruginosa
Phenazines are relevant virulence factors of P. aeruginosa. They are essential for adhesion and biofilm formation; they are also involved in oxidative stress, causing cell injury and death [9,10]. For this reason, the levels of three phenazines (PCA, PYO and 1-OH-PHZ) were assessed by HPLC-ESI-MS analysis, using P. aeruginosa cell-free supernatants from bacteria exposed to propolis EtOH extract, solvent or medium for 16 h. The results showed that propolis extract influenced the release of phenazines to a different extent, depending on the dose used. In particular, as shown in Table 2, the peak areas of PCA, PYO and 1-OH-PHZ decreased and this decrease (expressed as reduction percentage) ranged between 55% and 92.2%, depending on the propolis concentration. In all the cases, a relevant reduction was due to the solvent per se, when used at the condition corresponding to the highest propolis concentration. The supernatants of BLI-Pseudomonas exposed to the medium, propolis EtOH extract or solvent for 16 h were collected and tested for phenazines levels by HPLC-ESI-MS analysis. The peak area values of the three phenazines (PCA, PYO and 1-OH-PHZ) in their specific chromatographic runs were used for semiquantitative evaluation. The percentage reduction was expressed with respect to the medium alone or to the solvent. The results shown are from a representative experiment out of two performed.

Propolis Effects on eDNA release by P. aeruginosa
It is known that eDNA is a relevant component of P. aeruginosa biofilm, essential for the initial adhesion and stability of the sessile community [2,8]. In order to assess the capacity of P. aeruginosa to release eDNA, the bacteria were allowed to produce biofilm in the presence or absence of propolis extract for 16 h; then, cell-free supernatants were tested for the presence of eDNA, as detailed in the Materials and Methods. As shown in Table 3, eDNA release by P. aeruginosa was affected by propolis; in particular, the eDNA release decreased in a dose-dependent fashion, to 24.8% and 43.8% when using 15.6 and 31.2 µg/mL propolis, respectively. A partial reduction was also observed by the solvent per se (0.081 vs. 0.121 RFU; 33.1% decrease), when used at the condition corresponding to the propolis dose of 31.2 µg/mL. Moreover, a reduction of eDNA release was observed when comparing propolis vs. solvent; in this case, the RFU% reduction was 27.8 and 16.1, at 15.6 and 31.2 µg/mL respectively. The eDNA content was determined in 16 h cell-free supernatants from P. aeruginosa, exposed or not to propolis. PI was added before fluorescence reading, as detailed in the Materials and Methods. The results were expressed as mean fluorescence values (RFU) of triplicate samples. Standard deviations values < 5% were omitted. The reduction (%) was calculated with respect to the medium and the solvent. These values are from a representative experiment out of two performed.

Polyphenol Content in Propolis Exposed or Not to P. aeruginosa
Propolis is a complex mixture of components with a broad spectrum of activities, including antimicrobial, antioxidant, anti-inflammatory, anti-proliferative and anti-angiogenic effects [14][15][16]. A selected group of compounds, occurring in propolis and known for their biological activities, were analyzed by means of HPLC-ESI-MS. In particular, the peak areas observed in the propolis extract (15.6 µg/mL) exposed or not to Pseudomonas for 16 h were compared. The overlapped chromatographic peaks of eight polyphenols are shown in Figure 4. A reduction in the peak areas of quercetin and CAPE was observed upon propolis exposure to P. aeruginosa (black lines) with respect to the controls (propolis extract alone; red lines). A slight decrease was also observed for the pinobanksin-3-O-acetate and pinobanksin-3-O-butyrate peak areas, while that of chrysin moderately increased.
The eDNA content was determined in 16 h cell-free supernatants from P. aeruginosa, exposed or not to propolis. PI was added before fluorescence reading, as detailed in the Materials and Methods. The results were expressed as mean fluorescence values (RFU) of triplicate samples. Standard deviations values < 5% were omitted. The reduction (%) was calculated with respect to the medium and the solvent. These values are from a representative experiment out of two performed.

Polyphenol Content in Propolis Exposed or Not to P. aeruginosa
Propolis is a complex mixture of components with a broad spectrum of activities, including antimicrobial, antioxidant, anti-inflammatory, anti-proliferative and anti-angiogenic effects [14][15][16]. A selected group of compounds, occurring in propolis and known for their biological activities, were analyzed by means of HPLC-ESI-MS. In particular, the peak areas observed in the propolis extract (15.6 μg/mL) exposed or not to Pseudomonas for 16 h were compared. The overlapped chromatographic peaks of eight polyphenols are shown in Figure 4. A reduction in the peak areas of quercetin and CAPE was observed upon propolis exposure to P. aeruginosa (black lines) with respect to the controls (propolis extract alone; red lines). A slight decrease was also observed for the pinobanksin-3-O-acetate and pinobanksin-3-O-butyrate peak areas, while that of chrysin moderately increased.

Discussion
Propolis is known to exert antimicrobial activity more efficaciously against Gram-positive than Gram-negative bacteria [34,35]; this is likely due to the peculiar structure of the latter as well as to their ability of producing a wide range of hydrolytic enzymes, which in turn likely break down the active compounds of propolis [34,35]. On these bases, our study focused on Pseudomonas, as a prototype of a hardly attackable pathogen, capable to express numerous virulence factors; its susceptibly to propolis has been evaluated in terms of variation in growth ability, biofilm formation, production of phenazines and eDNA release. In particular, an Albanian propolis, previously described for its therapeutic efficacy in dentistry settings [22], has been extracted using three different solvents, EtOH, PG and PEG 400, and then analyzed by HPLC-UV/DAD and HPLC-ESI-MS. The three profiles obtained, which happen to be comparable irrespectively of the solvent used, show that the most abundant flavonoids are chrysin, galangin, pinocembrin and pinobanksin (and its esters). Moreover, among phenolic acids, caffeic acid derivatives, such as CAPE, also occur in high amounts. Overall, the main components identified in the Albanian propolis closely recall those previously described for propolis samples of Italian origin and, more generically, of European origin [33]. Hereafter, the rough extracts have directly been tested (i.e., without solvent removal) against Pseudomonas to avoid loss of any volatile compounds, likely mediating antimicrobial activities [33].
When expressed in terms of MIC, the anti-Pseudomonas activity of each propolis extract varies from 15.6 µg/mL for EtOH and PG extracts to a four-fold higher value (62.5 µg/mL) for the PEG 400 extract. To better investigate the phenomenon, a highly sensitive bioluminescence-based model has been used [11], in order to real-time monitor both total microbial growth and viable cells organized as biofilm. Preliminary results, aimed at testing serial dilutions in the sub-MIC range, provided evidence for inconsistent inhibitory effects (data not shown), thus orienting the focus of our efforts on the MIC condition. As detailed in the Results section, relevant differences among the three extracts have been observed. In particular, the propolis EtOH extract happens to be the most efficient in decreasing the total microbial growth (the solvent per se showing no toxic effects). In contrast, the efficacy of the other two extracts is partially hampered by the toxicity of both PG and PEG 400 solvents. The anti-Pseudomonas effect of propolis EtOH extract, demonstrated by the persistently low BLI signal, is further highlighted when converting the bioluminescence values in CFU/mL at 16 h. Thus, propolis EtOH extract appears as the most interesting anti-Pseudomonas among the three extracts tested. Further analysis has revealed that EtOH propolis significantly reduces (81% decrease) the amounts of viable-metabolically active cells, capable of producing biofilm; in contrast, only a partial reduction (26%) of the total biofilm mass occurs under that same condition, as assessed by crystal-violet staining (data not shown), thus suggesting that the amounts of viable cells and, to a lesser extent, the polymeric extracellular matrix accumulation are indeed affected by propolis.
The BLI-based assay allowed us to demonstrate a significant reduction (39%), also when using propolis PG extract. Differently, the propolis PEG 400 extract inhibitory effects are somehow hampered by the action due to the solvent per se (31.7% inhibitory effect). Thus, by a highly sensitive BLI-based system, these findings provide novel evidence on the anti-Pseudomonas activity of propolis extracts, underlying that its efficacy also depends on the solvent used. Whether the anti-Pseudomonas effects of propolis are mainly due to a direct antibacterial activity or rather to a specific antibiofilm effect remains to be elucidated. Further experiments aimed at evaluating propolis effects on pre-formed Pseudomonas biofilm have revealed a consistent lack of activity independently on the propolis extract considered (data not shown), thus implying that propolis acts as an anti-Pseudomonas agent when the pathogen is in a planktonic stage, while an already structured sessile microbial community appears to be insensitive.
P. aeruginosa produces a large amount of water-soluble blue-green phenazine pigments, known to exert antimicrobial [36] and antifungal [37] activities. Phenazines control colony size, favor bacterial adhesion and increase thickness and biomass of P. aeruginosa biofilm [38,39]. Moreover, PYO, one of the most studied phenazines, interacts with molecular oxygen to form reactive oxygen species, like H 2 O 2 , which modify the redox balance. By our study, we expanded the knowledge on Pseudomonas as a phenazine producer, demonstrating not only PYO but also PCA and 1-OH-PHZ release. Moreover, a semiquantitative analysis, performed by comparing the peak areas of each compound under different conditions, indicated that phenazine production occurs to a similar extent in untreated cells (i.e., grown in medium) and in cells exposed to EtOH solvent. Differently, when comparing controls vs. propolis treated groups, a dose-dependent reduction in the peak areas is observed, implying that the overall release of such virulence factors is affected (always >50% decrease) in the presence of propolis. A further analysis, carried out by normalizing phenazine production with respect to the area described under the cell growth curve within the period of 16 h (i.e., total viable cells), indicated an increase of several fold in phenazine production upon propolis treatment (see Table S1). This apparent discrepancy suggests that the single cell secretory activity of the surviving/persisting cells is indeed increased, opening to the hypothesis of a microbial reaction to propolis-mediated insult. Further in-depth studies are needed to elucidate this point.
eDNA, which has been demonstrated to support biofilm stability, is abundantly released by Pseudomonas [5,8]. In our model, eDNA levels detected following propolis treatment are impaired in a dose-dependent manner. This decrease is probably related to the reduced number of cells detected at the end of the treatment. In any case, in line with other studies [10], the reduction of eDNA is shown to parallel with the decreased production of PYO.
It is well-known that propolis composition may vary as a function of the geographical area of production, botanical sources, season of collection, etc. [15]. Its therapeutic efficacy is mainly related to its antimicrobial, antioxidant, antiproliferative effects. The capacity of propolis to modulate the immune system has also been described [40][41][42]. As main propolis components, polyphenols have been investigated to understand the mechanisms involved in propolis-mediated effects [43]. It has been reported that the antibacterial mechanism of quercetin probably depends on the disruption of target membrane and inactivation of extracellular proteins by forming irreversible complexes [44]. In addition, quercetin and its derivatives may reduce the expression of some inflammatory genes. The effects of this compound have been described in the murine RAW264.7 macrophage cell line, in terms of oxigenase-1 protein production, transduction of nuclear factor NFkB, decrease in Nrfk2 gene expression and inactivation of miR-155 [45]. Caffeic acid and CAPE have a significant role in cellular cycle and cancer cell apoptosis; bacterial replication also seems to be affected [46,47]. Chrysin and its phosphate ester exert a strong inhibitory effect on Enterovirus EV71 [48]. Galangin significantly suppresses growth of vancomycin-intermediate Staphylococcus aureus (VISA) strain Mu50 [49]. Veloz et al. [50] have shown that pinocembrin and apigenin are able to modify the architecture of S. mutans biofilm, reducing its thickness; antimicrobial activity against S. mutans has also been documented [50]. In our in vitro model, BLI-Pseudomonas exposure to propolis has been shown to reduce the levels of several phenolic compounds, as shown by a decrease in their corresponding peak areas; in particular, quercetin and CAPE are consistently affected, while pinobanksin-3-O-acetate and pinobanksin-3-O-butyrate appear to slightly decrease when compared to the control (propolis alone). On the contrary, chrysin slightly increases in the supernatants of bacteria treated with propolis. Given the antimicrobial properties of quercetin and CAPE, we favor the idea that these compounds can directly interact with bacteria, thus explaining the decrease observed upon propolis exposure to Pseudomonas. A similar phenomenon also occurs when considering pinobanksin esters, known for their antioxidant effects [51]. Whether these compounds may be involved in the inhibition or modulation of oxidation reactions remains to be investigated. Taken together, these data provide an initial evidence that Pseudomonas affects several polyphenols present in propolis; further in-depth studies are warranted to better address this issue.

Conclusions
Our results strengthen the relevance of propolis as a natural antimicrobial product against P. aeruginosa, a Gram-negative opportunistic pathogen, known to be highly refractory to disinfectants, antibiotics and host defense mechanisms due to its multiple virulence-factors. By a highly sensitive luminescence-based model, here, we provide the first evidence that Pseudomonas exposure to propolis impairs its growth ability, production of biofilm and capacity to release molecules, such as phenazines and eDNA. Being these peculiarities closely related to Pseudomonas pathogenic potential, we may envisage propolis as a precious source of natural compounds for the development of new therapeutic strategies, particularly against biofilm-related infections. In this perspective, it will be interesting to evaluate the antibacterial activity of specific propolis components, widening the panel of both polyphenolics and microbial agents as well.