Rhus coriaria L. was purchased from Al Alim Ltd. (Medicinal Herb Center, Zippori, Israel), and the fruits were ground to yield a reddish-purple powder. MG (analytical standard) was purchased from Sigma-Aldrich; Merck KGaA (Darmstadt, Germany).

The air-dried, powdered fruit of sumac plant (25 g) was packed in an Erlenmeyer flask and extracted with 250 ml methanol (MeOH), sonicated for 2 h at 45°C, then left in a dark glass bottle at room temperature for 24 h for complete extraction. The methanolic extract was filtered and evaporated to dryness with a rotary vacuum evaporator.

A Waters Alliance e2695 separations module, 2998 photo diode array (PDA) and Empower version 3 software was used (Waters, Eschborn, Germany). The preparative HPLC system consisted of a 3535-quaternary gradient module and a 996 PDA detector.

¹H-NMR and ¹³C-NMR measurements of isolated MG from sumac was carried out on a Bruker Avance II 500 spectrometer, which was equipped with a 5 mm indirect detection probe with Z gradient.

The UHPLC system (Accela; Thermo Fisher Scientific, Inc., Waltham, MA, USA) was equipped with an XBridge ODS-column (150×2.1 mm i.d.), 3.5-µm particle size (Waters, Milford, MA, USA) and a mobile phase containing 0.1% formic acid (FA) as (eluent A) and MeOH containing 0.1% FA as (eluent B). Linear gradient elution was used starting with 95% A and continuously increased to 100% B in 20 min. The UHPLC system was coupled to LTQ Orbitrap XL system (Thermo Fisher Scientific, Inc.) equipped with an electrospray ionization (ESI) source. The positive ionization mode was used at a scan range of m/z 100–1000.

MG quantitation was run on a Waters HPLC ODS Column (XBridge, 4.6 ID ×150 mm, 5 µm) with a guard column (Xbridge ODS, 20×4.6 mm ID, 5 µm). The mobile phase consisted of water (labeled A) and acetonitrile (labeled B) solvent mixture. The gradient was as follows: 95% A and 5% B at 0 min, held there for 2 min, then raised to 50% A and 50% B over 13 min, then to 10% A and 90% B over 1 min, held there for 3 min, and finally returned to 95% A, 5% B over 1 min. All of the samples were filtered with a 0.45 µm micro-porous filter. The PDA wavelengths ranged from 210 to 500 nm, and the monitoring wavelength of MG was 272 nm. The flow rate was 1 ml/min. The injection volume was 10 µl, and the column temperature was at room temperature.

The UHPLC was equipped with an XBridge ODS-column (150×2.1 mm i.d.), 3.5 µm particle size (Waters) and a mobile phase containing 0.1% formic acid (FA) as (eluent A) and MeOH containing 0.1% FA as (eluent B). Linear gradient elution was used starting with 95% A and continuously increased to 100% B in 20 min.

The prep-HPLC experiments were run on an ODS column (Agilent PrepHT C18, 22.2×250 mm, 10 µm). The linear gradient started at 98% A, where it stayed for 3 min; it was then raised to 5% A over 15 min, where it remained for another 4 min and was then returned to 98% A over 2 min. The flow rate used was 15 ml/min, the injection volume was 1,000 µl.

Five different concentration levels of standards for MG were prepared: 10, 50, 100, 250 and 500 ppm in methanol diluents. These standards were used to construct a calibration curve to quantitate MG in sumac.

Crude sumac was prepared by dissolving ~3,000 mg in 15 ml of methanol, sonicated for 3 min and then filtered via a 0.45 µm membrane filter before injection. The concentration of the final red solution was 200 mg/ml. This solution (1 ml) was injected into the preparative HPLC Chromatograph and six fractions were collected.

Streptococcus mutans UA159 (700610; American Type Culture Collection, Manassas, VA, USA) was selected for this study because it preferably colonizes humans. Stocks of this strain were maintained in 10% skimmed milk (Difco; BD BioSciences, Franklin Lakes, NJ, USA) at −70°C until use. Prior to experiments, S. mutans was cultured in Bacto™ Todd Hewitt broth (THB; BD BioSciences) under anaerobic conditions (95% N₂ and 5% CO₂) at 37°C for 18 h. Purity of the culture was checked on Mitis Salivarius agar (Difco; BD BioSciences) and Columbia agar with 7% sheep blood (E&O Laboratories, Bonnybridge, Scotland).

To evaluate the effect of treatments, S. mutans biofilm formation was performed on polystyrene and glass surfaces in separate experiments. Prior to each experiment, the optical density (OD) of the bacterial culture was adjusted to 0.2 at 630 nm, using a microplate-reader spectrophotometer. For biofilm formation on the polysterene surface, 24-well, flat-bottomed, polystyrene cell culture plates (Sarstedt, Nümbrecht, Germany) were filled with THB containing 1% sucrose, and then solutions of methanolic sumac extract (MSE), prepared in sterile distilled water (Milli-Q water), were added to appropriate wells in the plates at final concentrations of 4, 5 and 6 mg/ml. Solutions of MG, also prepared in Milli-Q water, were added to appropriate wells in the plates at final concentrations of 0.55, 0.7, 0.85 and 1 mg/ml. Afterwards, S. mutans bacteria was added to the wells at a final dilution of 1:100, and all of the plates were incubated anaerobically at 37°C for 24 h. Quantification of the formed biofilm (biomass) was performed using a colorimetric assay. The same experimental procedures were used for biofilm formation on the glass surface, except that sterile glass slides of 1-mm thickness, cut from standard microscope slides (76×26 mm; Thermo Fisher Scientific, Inc.), were inserted vertically into wells containing only MG at the above indicated concentrations prior to inoculation of bacteria. Quantitative assessment of the developed biofilm was further performed applying an optical profilometry assay. In these experiments, plate wells without bacterial cells were used as blank controls, and the untreated bacteria served as experimental controls.

After 24 h of incubation, THB was discarded from plates, wells were rinsed with distilled water to remove loosely bound bacterial cells, and then adherent bacteria were fixed with 95% ethanol. For quantification of biofilm biomass, the fixed and air-dried S. mutans biofilm in plate wells was stained with 1 ml/well of 0.01% crystal violet solution (Merck KGaA) for 15 min, and then the bound dye was extracted using 1 ml/well of 33% acetic acid solution (Merck KGaA) for 30 min. Afterwards, 200 µl extracted dye solution from each well was transferred to appropriate wells in an optically clear, flat-bottomed 96-well microplate. The OD of the samples was measured at a wavelength of 595 nm with a microplate-reader spectrophotometer. Background staining was corrected for by subtracting amount of the staining in blank wells.

Following 24 h of incubation, glass slides with adherent S. mutans biofilm were removed from plate wells, air-dried and further analyzed using a non-contact optical imaging profilometer Sensofar PLµ 2300 system (Terrassa, Spain), using a ×50 confocal objective with a view field of 253×190 µm. Primarily, six regions of the glass slide were scanned seeking to evaluate surface roughness per slide halfway from bottom to top of visible biofilm. Afterwards, a vertical scratch was artificially made down the glass surface by a scalpel in the middle of every slide covered with biofilm, and then five regions of the glass slide were scanned to assess the biofilm thickness per slide halfway from the bottom to the top of the visible biofilm. The bottom of the scratch served as a reference point for accurate measurement of the biofilm thickness. All images were captured in a vertical scanning mode, and data of the collected images were further processed using Gwyddion software (version 2.40; http://gwyddion.net) to calculate the surface roughness and biofilm thickness parameters. A median filter (10 pixels or 3 µm) was selected to remove the errors of form and waviness. Root mean square roughness (R_q), as the most critical parameter, was calculated to evaluate quantitatively the slide surface roughness, which indicated adherence of bacteria. The R_q parameter is an average of the measured height deviations taken within the evaluation length and measured from the mean line. Indeed, R_q represents standard deviation of the surface profile heights, and it is calculated according to ISO 4287/1-1997 standard by the following formula:

Where N is the number of points within a sampling length, and rj is the height value at point j. To measure the biofilm thickness, which indicated the maturity of biofilm, the height of the artificially produced vertical scratch on each slide with adherent bacteria was used. Calculation of the biofilm thickness involved generation of the height distribution graph curve from the entire area of the scanned region containing the scratch, followed by Gaussian function fitting as defined here:

Where y₀ is the peak height, a is the amplitude (height) distribution, x₀ is the peak position, and b is the standard deviation. Background for the parameters of surface roughness and biofilm thickness was corrected for by subtracting the R_q and thickness values of a blank glass slide.

S. mutans biofilm formation and treatments were performed using the same procedures as described above. After 24 h of incubation, the biofilm growth medium (THB) was collected from the wells of all plates and transferred to 1.5 ml microcentrifuge tubes. The pH of S. mutans biofilm growth medium collected in 1.5 ml microcentrifuge tubes was measured with a microelectrode connected to a benchtop pH meter (Knick 766 Calimatic) at room temperature. The microelectrode was calibrated using standard pH buffers (pH 4.01 and 7.00) prior to and following each measurement.

Data were analyzed using SPSS version 23.0 (IBM Corp., Armonk, NY, USA). Differences between the control (untreated) and treatment groups were evaluated applying one-way analysis of variance followed by a post hoc least significant difference test for multiple comparisons. Data are presented as the mean ± standard error. P<0.05 was considered to indicate a statistically significant difference.

The primary assessment of crude MSE efficiency for the inhibition of S. mutans biofilm formation showed its capacity to reduce significantly the biofilm biomass on the polystyrene surface in a dose-dependent manner (Fig. 1). The greatest inhibitory effect of crude extract was achieved using a concentration of 6 mg/ml, which decreased S. mutans biofilm formation by 77%, compared to the formation of untreated bacteria (P<0.05) cultured in THB with 1% sucrose.

Sumac is very rich in phenolic and anthocyanin compounds. These classes of compounds are known to exhibit a range of biological activities. In the attempt to isolate the active ingredients from sumac, several analytical HPLC runs were performed using different mobile phases. Upon scaling up using preparative HPLC, separations were carried out by collecting six fractions, which were subjected to antibacterial assays. A typical preparative HPLC chromatogram of the crude methanol mixture and its corresponding overlaid UV-Vis spectra, in the range of 210–500 nm, are presented in Figs. 2 and 3, respectively.

The ultraviolet-visible spectra demonstrated many phenolics that possess typical absorption maximums between 268.8 and 277.1 nm. There was also a very strong absorption maximum at 519.5 nm, which is a typical anthocyanin compound that is responsible for the red pigment in sumac (Fig. 3).

Fraction III (Fig. 2) contains pure compound, which, upon reinjection using analytical HPLC-PDA, reached a peak at 7.88 min with a maximum wavelength of 272 nm (Fig. 4).

Accurate high-resolution mass spectroscopy (HR-MS) and ¹H-NMR and ¹³C-NMR disclosed the identity of the peak at 7.88 min to be MG (Fig. 5). The HR-QTOF-MS using the ESI in the positive mode gave a protonated molecular ion of [M+H]⁺ at m/z 185.0452 Da (calculated for C₈H₉O₅⁺, m/z 185.0450 Da). ¹H-NMR (CD₃OD) gave peaks at δ: 3.8 (3H s, -OCH₃), 7.05 (2H s, H-2/H-6). ¹³C-NMR (CD₃OD) gave peaks at δ: 52 (C-8), 109 (C-2/C-6), 120.5 (C-1), 139 (C-4), 145.3 (C-3/C-5), 168 (C-7).

Analytical HPLC was used to construct a calibration curve of MG at 5 levels (10, 25, 50, 100, 250 and 500 ppm) with a coefficient of determination (R²) of >0.999. The concentration of MG in the methanol extract was identified to be 4.85 ppm. Many other peaks that have a more lipophilic character were eluted between 18 and 22 mins (Fig. 6).

In an effort to evaluate anti-biofilm activity, the isolated MG (fraction III) was further tested on S. mutans strain UA159 because of its well-defined features to form biofilms on various solid surfaces (19,20). As presented in Fig. 7, all of the concentrations of MG significantly reduced S. mutans biofilm biomass on the polystyrene surface, compared with untreated bacteria (P<0.05) in THB containing 1% sucrose. The anti-biofilm activity of MG occurred in a dose-dependent manner, exhibiting its highest effect at concentrations of 0.7, 0.85 and 1 mg/ml, with biofilm inhibition at 68, 89 and 93%, respectively. These results confirmed the findings reported by Kang et al (16), where the researchers demonstrated using beaker-wire tests that 1 mg/ml MG significantly reduced the wet weight of S. mutans (strain Ingbritt) biofilm biomass in comparison with untreated bacteria, after 24 h of incubation in medium containing 5% sucrose.

The optical profilometry technique, applied in analysis of the surfaces of glass slides with S. mutans biofilm, was used to confirm the results of the colorimetric assay. Firstly, compared with bacteria incubated without MG and sucrose (Fig. 8A), where the R_q and thickness parameters for the untreated bacteria grown without sucrose were 0.04±0.01 and 0.02±0.01 µm, respectively, presence of 1% sucrose in THB induced adherence of the bacteria to glass slides and subsequent maturation of the biofilm (Fig. 8B). Secondly, in the THB with 1% sucrose, exposure of S. mutans to 0.55, 0.7, 0.85 and 1 mg/ml MG (Figs. 8C-F, respectively) decreased formation of the biofilm on glass surface in a dose-dependent manner. Quantification revealed that surface roughness parameter R_q of the biofilm (Fig. 9A) and biofilm thickness (Fig. 9B) were increased in control bacteria (Fig. 9); however, MG treatment inhibited this effect in a dose-dependent manner (P<0.05; Fig 9). In this respect, MG concentrations of 0.85 and 1 mg/ml exhibited the greatest effects for reducing the surface roughness parameter R_q of biofilm, by 94 and 99%, respectively (Fig. 9A). Furthermore, these concentrations of MG decreased S. mutans biofilm thickness by 97 and 99%, respectively (Fig. 9B).

The pH measurements of the collected S. mutans biofilm growth medium demonstrated that bacteria grown in THB with 1% sucrose produced organic acids from fermentation of this carbohydrate, leading to an ~1.8-fold decrease in pH, compared with the blank group (Tables I and II). This substantial decrease of pH indicated increased acidogenicity in the S. mutans biofilm. However, exposure of the bacteria to crude MSE led to augmentation of pH in a dose-dependent manner, though not up to the pH levels of the blank group (Table I). In contrast, treatment of S. mutans bacteria with the isolated MG significantly prevented a decrease in the pH level, compared with untreated bacteria grown in THB with 1% sucrose (P<0.05), by increasing the pH almost up to the pH levels of the blank group (Table II). This preventive effect of MG occurred in a dose-dependent manner, and MG concentrations from 0.7 to 1 mg/ml increased pH by 96–97%. Therefore, similarly to reduction of the biofilm biomass, crude MSE was unable to fully suppress the pH decrease in the S. mutans biofilm due to reduced concentration of the biologically active compound within it. However, the isolated MG was able to almost completely inhibit the acidogenicity of the S. mutans biofilm.