Plant extracts preparation

Fresh plants of P. odoratissimum, P. graveolens, and P. zonale were purchased in October 2022 from two traditional markets of the Quito Metropolitan District in Ecuador: Santa Clara and Iñaquito. Identification and validation of the plants were carried out by specialists from the Botanical Garden of Quito, Ecuador. Table 1 shows the common Spanish names, scientific names, and parts of the plants used in this study. The parts of the selected plants were mixed to prepare one extract for each species. The only extract containing flowers in the preparation was P. zonale, as it is commonly sold with flowers to prepare infusions or drinks. Therefore, we decided to keep the flowers in the preparation of P. zonale extract to maintain the traditional medicinal uses.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Plants and parts of the plants used in this study.

https://doi.org/10.1371/journal.pone.0306637.t001

The plant extracts were prepared according to our previous studies [12,18,19]. Initially, 0.5 kg of fresh plants were dried and ground to fineness using liquid nitrogen and stored at -80°C. To prepare the methanolic extract, 2 g of the ground sample was extracted with 20 mL of a methanol-water solution (80:20, v/v) (methanol 99% v/v; Sigma-Aldrich, St. Louis, MO, USA) and stirred for 2 h in the dark at room temperature (around 21°C). The samples were then centrifuged twice at 1500 rpm for 10 min and filtered through a cellulose filter paper (0.2 μm; Whatman, Merck & Co., Inc, USA). The supernatant was concentrated under vacuum using a Büchi Rotavapor R-210 set (BÜCHI Labortechnik, Switzerland) at 30°C in a water bath and cooling water at 0°C until most of the solvent was removed. The concentrate was then dried in a miVAC centrifugal concentrator (GeneVacTM, Thermo Fisher Scientific, Sweden) at 30°C for 5 h. Dried extracts were stored at -80°C for further use. Stock solutions were prepared by resuspension of the dry sample in 1 mL of methanol-water solution (80:20, v/v).

Phytochemical screening

The hydroalcoholic extracts of P. odoratissimum, P. graveolens, and P. zonale were tested for the presence of phenolics, flavonoids, tannins, anthocyanins, and amino acids. Total phenolic content (TPC) was determined using the Folin-Ciocalteu colourimetric method [20]. For this, an aliquot of 100 mL of the resuspended extract, standard or blank, was mixed with 500 mL of 1:10 (v/v) Folin-Ciocalteu reagent (2N, Sigma-Aldrich, USA). The mixture was incubated for 5 min, and then 400 mL of 0.7 M sodium carbonate solution (PanReac AppliChem, Spain) was added and incubated for 2 h at room temperature in a dark room. The absorbance was then measured in a plate reader (Synergy HT, Bio-Tek Instruments, Inc., USA) at 760 nm against a blank of Milli-Q water. Gallic acid (Loba Chemie Pvt. Ltd., India) was used to construct the standard curve (0.5–1.2 mM). Results were expressed in milligrams of gallic acid equivalents (GAE) per gram of dry extract (DE) as mg GAE per g DE.

Total flavonoid content (TFC) was determined by the aluminium chloride colourimetric method [21]. For this, 80 mL of the suspended extract, standard or blank, was mixed with 400 mL of Milli-Q water and then added 24 mL of 5% sodium nitrite solution (Sigma-Aldrich, USA). After 6 min, 48 mL of a methanolic solution of aluminium chloride hexahydrate (10%, Loba Chemie Pvt. Ltd., India) was added to the mixture and incubated for a further 5 min. The mixture was then treated with 160 mL of 1 M sodium hydroxide solution and 88 mL of Milli-Q water and incubated for 15 min at room temperature in a dark room. The final mixture was analysed by measuring the absorbance in a plate reader at 510 nm with Milli-Q water as a blank. Catechin (Cayman Chemical, USA) was used to construct the standard curve (0.0625 mM to 1.0 mM), and results were expressed as milligrams of catechin equivalents (CE) per gram of dry extract (DE) (mg CE/g DE).

The total tannin content (TTA) was determined according to the method outlined in ISO 9648:19880 “Sorghum—Determination of tannin content”, available at https://www.iso.org/standard/17487.html (International Organization for Standardization, 1988). For this, an aliquot of 100 μL of the extract, standard or blank, was mixed with 100 μL of N-N-dimethylformamide 75% reagent (Sigma-Aldrich, USA). The mixture was incubated for 20 min at room temperature with stirring and then centrifuged at 3000 (RCF) for 10 min. The supernatant was collected and mixed with either solution A or B. For solution A, 75 μL of supernatant was mixed with 450 μL distilled water, and 75 μL ammonia solution was then stirred for a few seconds. For solution B, 75 μL of supernatant was mixed with 375 μL of distilled water and 75 μL of ferric ammonium citrate, stirred for a few seconds, and finally mixed with 75 μL of ammonia solution. The mixtures were incubated for 10 min in a dark room. The absorbance was then measured using a plate reader at 525 nm against a blank of distilled water and the solutions A or B. The TTA is the difference between absorbances B-A. The results were expressed as milligrams of tannic acid (TAEq) per gram of dry extract (DE) (TAEq/g DE).

Total monomeric anthocyanin pigment content (ACY) was determined using the pH differential method [22]. For this, a buffer pH of 1.0 (potassium chloride, 0.025M) and a buffer pH of 4.5 (sodium acetate, 0.4M) were used. A stock of pelargonidin solution (Sigma-Aldrich, USA) was used to generate a standard curve (10 mg to 150 mg). For each sample, 100 μL of standard or extract was added to either 900 μL of buffer pH 1.0 or 900 μL of buffer pH 4.5. For the blanks, 100 μL of distilled water was added to 900 μL of buffer pH 1.0 or buffer pH 4.5. The absorbance was read at 500 and 700 nm for both buffer solutions. Anthocyanin content was expressed as mg pelargonidin equivalents (mg PgEq) per mg of dry extract (DE) (mg PgEq/g DE).

The modified colourimetric ninhydrin method determined the total free amino acid content (FAA) [23]. For this, a working solution was prepared with a mixture of ninhydrin (Loba Chemie Pvt. Ltd., India), ethanol 99.5%, acetic acid, and CdCl₂ (Sigma-Aldrich, USA). L- leucine (Sigma-Aldrich, USA) solution was used to generate a standard curve (2.4 mg/L to 60 mg/L). Then, 400 μL standard or extract was mixed with 800 μL of working solution. For the blank, 400 μL of distilled water and 800 μL of working solution were used. The solutions were heated in a water bath (Memmert-Ghb14) at 84°C for 5 min, and after cooling in a cold block, the absorbance was read at 507 nm using a plate reader. The results were expressed as mg of Leucine equivalents (LEeq) per g of dry extract (DE) (LEeq/g DE).

Antioxidant activity analysis

The total antioxidant capacity (TAC) was analysed by two methods: the ferric reducing antioxidant power (FRAP) assay [24] and the 2,2-diphenyl-1-picrylhydrazyl (DPPH) free radical method [25]. For both assays, Trolox (Sigma-Aldrich, USA) was used to generate a standard curve (50 to 500 μM), and results were expressed as μmol Trolox equivalents (TEq) per gram of dry extract (μmol TEq/g DE). For the FRAP assay, a fresh FRAP reagent was prepared daily by combining sodium acetate trihydrate buffer (300 mM, Fisher Chemical, USA), 2,4,6-tri(2-pyridyl)-1,3,5-triazine (TPTZ) (10 mM in 40 mM HCl, Sigma-Aldrich, USA) and FeCl₃6H₂O solution (20 mM, Loba Chemie Pvt. Ltd, India) in a ratio of 10:1:1 and incubated at room temperature for 2 h. Then, 20 μL of extract, standard or blank, was mixed with 180 μL of FRAP solution, and the absorbance of the resulting mixture was measured at 593 nm using a plate reader.

For the DPPH assay, 50 μL of the extract was mixed with 400 μL of DPPH solution (0.2 mM, Alpha Aesar, USA) in absolute methanol and 550 μL of ethanol (70%). Two blank samples, B1 and B2, were prepared as follows: B1 consisted of 400 μL of DPPH solution, 50 μL of absolute methanol, and 550 μL of ethanol (70%), while B2 was prepared with 450 μL of absolute methanol, and 550 μL of ethanol (70%). After incubation for 15 min at room temperature, the absorbance of the extracts was determined at 517 nm using a plate reader. DPPH scavenging activity was calculated according to previous studies [12,26].

HPLC-MS/MS analysis of P. odoratissimum, P. graveolens, and P. zonale

A total of 100 mg of dry extract was resuspended in 1 mL of methanol-water solution (80:20, v/v), filtered, and used for HPLC-MS/MS analysis. The system used for HPLC analysis was a Vanquish (Thermo Fisher Scientific, Massachusetts, USA) equipped with a dual pump and diode array detection (DAD) coupled to an LTQ-XL (Thermo Fisher Scientific, Massachusetts, USA) and controlled by Xcalibur software. Separation was performed on an Accucore Vanquish C18 column (1.5 μm, 100 × 2.1 mm; Merck KGaA, Darmstadt, Germany) at a constant temperature of 35°C. The mobile phase consisted of a solution of 0.1% formic acid (A) and acetonitrile (B). The elution gradient was set as follows: 2% B, 0–4 min; 4% B, 4–22 min; 40% B, 22–32 min; 70% B, 32–40 min; 2% B, 40–45 min, and the column was re-balanced to the initial solvent conditions. The injection volume was 50 μL, and the flow rate was set to 0.2 mL/min. The DAD performed double line detection at 220, 280, 330, and 370 nm, while the mass spectrometer (MS) was operated in the positive ion mode. Recorded spectra ranged from 50 to 2000 m/z. In positive ionisation mode, the electrospray ionisation (ESI) conditions were capillary temperature of 275°C, source voltage of 4.5 kV, capillary voltage of 18 V, and tube lens of 70. The tentative identification of the parent ions in positive ion mode was based on the fragmentation patterns. The patterns were compared with the MzCloud database (https://www.mzcloud.org/), and those with a match score of more than 70% were selected. In addition, fragmentation patterns were identified by a literature search and the GNPS database [27]. Retention time was also considered for the tentative identification in cases where ions with the same fragmentation pattern were detected in two or three Pelargonium species.

Fractioning and chemical characterisation of P. zonale

The fractioning process of P. zonale extract was performed following the procedure described by Grkovic et al. [28] with some modifications. Briefly, 200 mg of dried extract was resuspended in 2 mL of a methanol-water solution (80:20, v/v) in a cotton swab (23 mm x 80 mm) stored in a solid phase extraction (SPE) column Bond Elut C18 (Agilent, USA), followed by a lyophilisation process to remove the solvent. The SPE column was equilibrated with 10 mL of methanol-water (5:95, v/v). The following eluents were used to collect the fractions: a) methanol-water (5:95, v/v), b) methanol-water solution (20:80, v/v), c) methanol-water solution (40:60, v/v), d) methanol-water solution (60:40, v/v), e) methanol-water solution (80:20, v/v) f) methanol and f) acetonitrile-methanol (50:50, v/v). Of each eluent, 2 mL was added to the column, and 1 mL was collected separately in 2 mL Eppendorf tubes that had been previously weighed. A total of 14 fractions were obtained, which were ultra-cooled to -80°C and then lyophilised. After lyophilisation, the fractions were weighed and re-suspended in a methanol-water solution (20:80, v/v) for drug susceptibility assays.

The HPLC-MS/MS analysis from fractions 1 to 10 was carried out with ZORBAX Eclipse Plus C18 column (5 μm, 150 × 4.6 mm; Agilen USA). The mobile phase consisted of 0.1% formic acid (A) and acetonitrile (B) solutions. The elution gradient was set up in the following way: 20% B, 0–5 min; 25% B, 5–10 min; 40% B, 10–20 min; 55% B, 20–30 min; 65% B, 30–50 min and the column was re-balanced to the initial solvent conditions. The mass spectrometry conditions and further analysis of identification are the same as explained above for the three species.

Bacterial strains, cells, media, and growth conditions

The bacterial and yeast strains Enterococcus faecium ATCC 27270, Enterococcus faecalis ATCC 29212, Staphylococcus aureus ATCC 25923, Acinetobacter baumannii ATCC 19606, Pseudomonas aeruginosa ATCC 27853, Enterobacter cloacae ATCC 23355, Escherichia coli ATCC25922, Candida albicans ATCC 10231, and Candida tropicalis ATCC 1369 were obtained from the American Type Culture Collection (ATCC) [12]. The strains Methicillin-resistant S. aureus 333 previously isolated from nasal samples [29] were obtained from the Research Laboratories of the Universidad de Las Américas (UDLA); while Vancomycin-resistant E. faecalis INSPI 032 were donated by the National Institute for Research in Public Health (INSPI) (http://www.investigacionsalud.gob.ec/webs/ram/evaluacion-calidad/) to the Institute of Microbiology of the Universidad San Francisco de Quito (IM-USFQ). Bacterial strains were initially grown on Nutrient Agar (NA) at 37°C for 24 h and yeast strains on Sabouraud Dextrose Agar (SDA, both from Becton, Dickinson and Company, USA) at 35°C for 48 h. Raw 264.7 murine macrophages (ATCC) were cultured in Dulbecco′s Modified Eagle′s Medium (DMEM, Sigma-Aldrich, USA) containing 25 mM Glucose (4500 mg/L), 4 mM L-glutamine, 1 mM sodium pyruvate, 1500 mg/L sodium bicarbonate and supplemented with 10% heat inactivated fetal bovine serum (FBS, Invitrogen) at 37°C in a 5% CO₂.

Screening of antibacterial activity

The antimicrobial activity of P. graveolens, P. odoratissimum, and P. zonale methanolic extracts was initially assessed using a microdilution assay. Briefly, single bacterial colonies were selected to prepare an inoculum corresponding to 1.5 × 10⁸ colony-forming units (CFU)/mL using a 0.5 McFarland standard turbidity. Bacterial suspensions were adjusted to a final microorganism density of 5 × 10⁵ CFU/mL. For yeast, single colonies were selected to prepare an inoculum of an optical density between 0.12 to 0.15 at 530 nm, corresponding to 1 × 10⁶–5 × 10⁶ CFU/mL. Yeast suspensions were adjusted to a final microorganism density of 2 × 10³ CFU/mL. Working solutions of crude plant extracts were prepared by diluting the stock solutions with Mueller Hinton Broth (MHB; Becton, Dickinson and Company, USA) to a final concentration of 2000 μg/mL. A 96-well U-shaped plate (Tecan Group Ltd., Switzerland) was filled with 100 μL of working solution and allowed methanol evaporation for 30 min. Then, 100 μL of bacterial or yeast suspension was added to each well. The microplate was incubated at 37°C between 16 to 20 h for bacteria and at 35°C for 48 h for yeast. The final concentration of plant crude extracts was 1 mg/mL. The negative control was a solution with a concentration of methanol-water (80:20, v/v) containing the inoculum. A well with MHB plus inoculum was used as a positive control to discard any influence of methanol in the inhibitory activity of the crude extracts. A medium with no inoculum was applied to control sterility, and a well with a working solution (crude extract) without inoculum was used as a crude extract control. The inhibitory activity of the crude extracts was measured by the unaided eye by comparing the results obtained with the negative control. Inhibition was considered when a clear well was observed by the unaided eye.

Screening of the P. zonale fractions against S. aureus ATCC 25923 was performed as explained above with some modifications. A 96-well U-shaped plate (Tecan Group Ltd., Switzerland) was filled with 50 μL of working solution and allowed methanol evaporation for 30 min. The final reaction volume was 100 μL and the final concentration of the fractions was 800 μg/mL. The negative control was a solution with a concentration of methanol-water (20:80, v/v). The inhibitory activity of the crude extracts was measured by plating 20 μL of samples from clear wells onto NA plates without plant extract. The agar plates were incubated at 37°C for 24 h. Results are reported as colony-forming units per milliliter (CFU/mL).

Drug susceptibility assays

Minimum inhibitory concentration (MIC) and minimal bactericidal concentration (MBC) of P. graveolens, P. odoratissimum, and P. zonale methanolic extracts were performed using the broth microdilution technique according to the Clinical and Laboratory Standards Institute (https://clsi.org/) [30] recommendations with some modifications. Bacterial suspensions were prepared as explained above. Working solutions of crude plant extracts were prepared by diluting the stock solutions with MHB to a final concentration of 4000 μg/mL. A 96-well U-shaped plate was filled in the first wells with 100 μL of working solution and allowed methanol evaporation for 30 min. Then, 100 μL of MHB was added to the first well, mixed, and 100 μL was taken from the first well to do 2-fold serial dilutions. After that, 100 μL of bacterial suspension (5 × 10⁵ CFU/mL) was added to each well. The microplate was incubated at 37°C between 16 to 20 h. The final concentrations of plant crude extracts were between 15.6 to 1000 μg/mL. The same controls as explained above were prepared with the addition of ciprofloxacin (Sigma-Aldrich, USA) used as a positive control with concentrations ranging from 0.09 to 3 μg/mL. The MIC was measured by the unaided eye by comparing the results obtained with the negative control. MBC was determined following the MIC assay, by plating 3 μL of samples from clear wells onto NA plates without plant extract. The agar plates were incubated at 37°C for 24 h. MBC was estimated as the least sample concentration where no visible microbial growth was observed.

Time-kill assays

The methanolic extracts of P. graveolens, P. odoratissimum, and P. zonale were tested at concentrations corresponding to 4X MIC to evaluate their bactericidal or bacteriostatic effect according to the Clinical and Laboratory Standards Institute guidelines [31] with some modifications. The positive control ciprofloxacin was set at 4X MIC for the susceptible strains E. faecalis ATCC 29212 (3 μg/mL) and S. aureus ATCC 25923 (1.5 μg/mL). The negative control was a solution with a concentration of methanol-water (80:20, v/v). The experiment was performed in a 96-well round-bottom plate following the same conditions as explained above and incubated at 37°C for 24 h. At defined time intervals (4, 6, 8, 16, and 24 h), 20 μL of serial 10-fold dilutions were plated on NA (without plant extract) and incubated at 37°C for 24 h. Killing curves were constructed by plotting the Log₁₀ CFU/mL versus time over 24 h. Bactericidal activity was defined as a decrease of more than 1000 times (3 Log₁₀) of viable bacteria compared to the original inoculum.

Biofilm inhibition and eradication assays

The biofilm inhibition and eradication activity of P. graveolens, P. odoratissimum, and P. zonale methanolic extracts were then evaluated as previously described [12]. Briefly, in both assays, an overnight culture growth was used to prepare a bacterial suspension with a final concentration of 5 × 10⁷ CFU/mL in MHB. In the inhibition assays, 190 μL of the inoculum was added to 96-well flat bottom plates and supplemented with 10 μL of crude extracts to achieve 1X and 2X MIC concentrations in each well. Plates were incubated for 24 h at 37°C. Regarding the eradication assays, the same bacterial suspension in MHB was prepared in 96-well plates incubating them for 48 h at 37°C. Subsequently, the MHB was removed from the wells, and a wash with PBS was performed. The mature biofilms were supplemented with crude extracts at 1X and 2X MIC concentrations in flesh media. Plates were incubated for 24 h at the same conditions. At the end of inhibition or eradication assays, three washes were performed in all plates with PBS, and each biofilm sample was fixed with 200 μL of methanol for 15 min. Crystal violet staining (0.1% w/v; Sigma-Aldrich, St. Louis, MO, USA) was performed for 10 min, followed by four washing steps with distilled water. After ethanol (95% v/v; Sigma-Aldrich, St. Louis, MO, USA) resuspension, each biofilm sample was measured at 570 nm in the ELISA Elx808 spectrophotometer (BioTek, Winooski, GU, USA) comparing with positive controls (bacteria growth in media only). Finally, the biofilm inhibition/eradication percentage was calculated as previously described [12]. A negative control was a solution of methanol-water (80:20, v/v), and a well with merely medium was applied as a sterility control. All the biological assays were set up in triplicates in at least two independent experiments.

Cytotoxicity assays

MTT (3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide) assay was performed as previously described [32]. Briefly, Raw 264.7 murine macrophages (ATCC) were seeded in 96-well flat-bottom culture plates (Corning, USA) at a density of 8 × 10³ in 200 μL of DMEM containing 10% FBS and cultured overnight. The following day media was removed and fresh DMEM was added supplemented with methanolic extracts at concentrations of 0, 125, 250, 500, and 1000 μg/mL. After 24 h, cells were washed once with Dulbecco’s phosphate buffered saline (PBS, pH 7.4; Sigma-Aldrich, St. Louis, MO, USA), and 200 μL of fresh DMEM was added. At 48 h, media was removed and changed for 200 μL of fresh DMEM with 1% FBS. Cell viability was assessed by adding 50 μL of filter sterilised MTT (5 mg/mL in PBS, Sigma-Aldrich, USA) to each well, followed by a 2 h incubation period at 37°C in 5% CO₂. Media was removed and the blue formazan crystals trapped in cells were dissolved by adding 200 μL of DMSO and 25 μL of Sorensen’s glycine buffer. The absorbance at 570 nm was measured in a spectrophotometer. The percentage of viability was calculated relative to the control (DMEM media with the dissolvent methanol-water 80:20, v/v).

Data processing and statistical analysis

MzMine V3.9.1 was used for data visualisation, mass detection, chromatogram builder, local minimum identification, MS1/MS2 building, isotope grouping, alignment, filtering, and gap filling [33] for the comparison of the P. zonale fractions. The resulting comma-separated values (cvs.) and mascot generic format (mgf.) files were used to compare differences between fractions. The metabolite area was used as an indicator of concentration. The metabolite areas in each fraction were obtained and then scaled using the Pareto method [34] with the following formula: Where: and to compare the different fractions.

Hierarchical clustering was done using Ward´s method as previously described [35]. For the comparison between chemical compounds, we used structural and physicochemical similarities computed with ChemMine Tools [36]. For pairwise comparison between control and treated samples in biofilm inhibition and eradication assays the Wilcoxon non-parametric test was used through R studio version 4.0 (https://www.rstudio.com/products/rstudio/download/) using several R packages ("ggpubr", "rstatixs", "openxlsx", and "tidyverse") [37]. For analysis of phytochemical screening and antioxidant activity, the IBM SPSS Statistics software for Windows version 25.0 (SPSS Inc., Chicago, IL, USA) was used and statistical analysis between different groups was performed using one-way ANOVA and Bonferroni post hoc tests. All p-values <0.05 were considered significant.

Phytochemical profile and antioxidant activity of P. odoratissimum, P. graveolens, and P. zonale

The hydroalcoholic extracts of P. odoratissimum, P. graveolens, and P. zonale were analysed qualitatively and quantitatively for their phytochemical profile (Table 2). Overall, the total phenolic content (TPC) for the three Pelargonium species was higher than 200.0 mg GAE per g DE, with P. zonale showing the highest values of total phenolic compounds (350.2 mg GAE per g DE) and total tannin content (TTA) (142.3 TAEq per g DE) when compared to P. odoratissimum and P. graveolens (p-value < 0.05). P. graveolens extract showed higher TPC values (210.2 mg GAE per g DE) than those reported in previous studies when using methanolic extracts (145.0 mg GAE per g DE [38] and 84.2 mg GAE per g DE [39] and essential oils (102.4 mg GAE per g DE) [40]. In the case of P. odoratissimum extract, the TPC value correlates with our previous report [12] and is currently the only data available for comparison. The data available for P. zonale regarding TPC are not comparable with our methodology [9,17]. On the other hand, the total flavonoid content (TFC) of P. graveolens extract was the highest (49.2 mg EC per g DE) from our study set (Table 2), showing significant differences when compared to P. zonale and P. odoratissimum (p-value < 0.01).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Phytochemical composition and total antioxidant capacity of P. graveolens, P. odoratissimum, and P. zonale hydroalcoholic extracts.

https://doi.org/10.1371/journal.pone.0306637.t002

Only P. zonale methanolic extract showed a detectable anthocyanin content (ACY) of 142.1 mg PgEq per g DE (Table 2), likely due to the presence of its characteristic red flowers in the extract preparation (Table 1), as others have previously reported [41]. No anthocyanin content was detected in the case of P. graveolens and P. odoratissimum methanolic extracts, probably due to the lack of flowers in the extract preparation (Table 1). Other studies have also shown the presence of anthocyanins in dried aerial parts (also containing flowers) and flower extracts of P. graveolens (0.28–1.01 g cyanine chlorine/100 g herbal product) [42,43] and in the pink-white flowers of P. odoratissimum (0.48–0.65 mg of malvin chloride/g fresh flowers) [44]. Thus, the anthocyanin content could be associated with the presence of flowers in the extract preparation. In the case of free amino acid content (FAA), the highest values were found in P. odoratissimum (43.2 mg LEeq per g DE), followed by P. graveolens and P. zonale (Table 2). No data on FAA is available on these three Pelargonium species being this study the first report.

The three methanolic extracts showed strong antioxidant activity as measured by DPPH (>1000 mmol TEq per g DE) (Table 2), with P. zonale showing the highest free radical-scavenging activity (3538.9 mmol TEq per g DE). In contrast, P. zonale had the lowest FRAP value (756.0 mmol TEq per g DE), but all three extracts showed medium FRAP values (756 to 937 mmol TEq per g DE). Significant differences (p-value < 0.01) were observed in both assays when comparing P. zonale with P. graveolens and P. odoratissimum.

Other authors have reported the antioxidant activity of P. graveolens [3] and P. odoratissimum [11,12] extracts. However, we cannot make comparisons due to differences in how the results were reported. To the best of our knowledge, there is no information available on the antioxidant activity of P. zonale. Overall, the three species have notable antioxidant activity regardless of the presence of flowers in the extract preparation. Considering that the total phenolic content (TPC) values in the three Pelargonium species were 4–10 times higher than the total flavonoid content (TFC), the antioxidant capacity of these methanolic extracts may have resulted from the high content of total phenolic compounds.

Phenolic antioxidants are highly efficient and typically required in very small amounts to neutralise many free radicals [45]. According to literature reports [45], the phenolic antioxidants identified in this study for P. graveolens include quercetin and kaempferol derivatives and a type of epi-gallocatechin and myricetin derivatives (Table 3). For P. zonale, quercetin and kaempferol derivatives, along with coumaroylquinic acid and anthocyanins derivatives, were identified. Lastly, for P. odoratissimum, quercetin and kaempferol derivatives and a type of epi-gallocatechin were identified.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Tentative identification of the most abundant masses in Pelargonium species hydroalcoholic extracts.

https://doi.org/10.1371/journal.pone.0306637.t003

Chemical characterisation of P. graveolens, P. odoratissimum, and P. zonale

The chemical characterization of the most abundant compounds from the three Pelargonium species was analysed by high-performance liquid chromatography-diode array detector-tandem mass spectrometry (HPLC-DAD-MS/MS). The tentative identification of the parent ions in positive ion mode is shown in Table 3 and a detailed discussion about these identifications is presented in S1 Table.

Several bioactive compounds were identified in the three Pelargonium species, mainly phenolic acids, flavonoids, tannins, and chalcones. A previous study also described the presence of myricetin, quercetin, and kaempferol flavonoids, or their derivatives, in several Pelargonium species [57]. The flavonoids identified in P. graveolens are consistent with those reported by Boukhris et al. (2013) [39]. Moreover, among the three species studied, P. graveolens stands out due to the various types of myricetin derivatives that were not detected in P. odoratissimum and P. zonale. On the other hand, P. odoratissimum showed a chemical composition similar to the one reported in our previous work [12], comprising flavonoids, gallocatechin and epigallocatechin tannins, and gallic acid and derivatives. Gallocatechin-derived tannins, which are known for their antimicrobial activity [58], were detected in P. odoratissimum and P. zonale. It should be noted that no in-depth study has been conducted on the chemical characterisation of P. zonale hydroalcoholic extracts, and the existing information is based only on the results of HPLC analyses [9,17]. In addition to the reported compounds in P. zonale, four anthocyanins were identified in the present work, more exactly cyanidin 3,5-diglucoside, pelargonidin-galactoside, pelargonidin 3,5-diglucoside, and peonin. The latter two compounds are the most described constituents in the Pelargonium genus [59] and are particularly reported in species with red flowers [60].

Screening of the antimicrobial activity of P. graveolens, P. odoratissimum, and P. zonale

Next, we investigated the antimicrobial activity of P. graveolens, P. odoratissimum, and P. zonale methanolic extracts using a microdilution broth assay against a set of seven susceptible bacterial strains and two susceptible yeast strains (Table 4). These strains were chosen as representatives of some of the most common microbial pathogens that are critical to human health. The initial screening was carried out at a maximum single-point concentration of 1000 μg/mL. Overall, these three methanolic extracts showed visible growth inhibition only towards Gram-positive bacteria (Table 4). We did not observe antimicrobial activity from any of the three Pelargonium species towards the Gram-negative bacteria tested nor for Candida species at the maximum concentration of 1000 μg/mL.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Screening of the antimicrobial activity of P. graveolens, P. odoratissimum, and P. zonale methanolic extracts against bacterial and yeast strains of clinical importance using a microdilution assay.

Methanolic extracts are at a concentration of 1000 μg/mL.

https://doi.org/10.1371/journal.pone.0306637.t004

However, other authors reported antimicrobial activity against the Gram-negative bacteria Klebsiella pneumoniae (MIC = 312 μg/mL) [7] and E. coli [5] when using dried leaves of P. graveolens dissolved in methanol and ethanol, respectively. In addition, Lewtak and colleagues reported the antifungal activity of the flower stalks of P. zonale hydroalcoholic extract against C. albicans (some activity observed at 37.5 μg/mL) [8]. In our study, we used leaves and stalks to prepare P. graveolens extract and flowers, stalks, and leaves were used to prepare P. zonale extract, both dissolved with methanol. Therefore, the differences observed in the antimicrobial activity may be attributed to the alteration in quantity and quality of secondary products of plants by different factors such as changes in cultivation, origin of the plant, growing seasons, time of collection, temperature, soil water, light, part of the plant used for the extraction or infusion preparations, among others [61]. In addition, it is well-known that the differences in cell membrane structure and composition between Gram-negative and Gram-positive bacteria play a critical role in the susceptibility to plant phenolics [62].

To this point, we also evaluated if the antibacterial activity observed against the susceptible strains E. faecalis ATCC 29212 and S. aureus ATCC 25923 was also detected in the Vancomycin-resistant E. faecalis INSPI 032 and the Methicillin-resistant S. aureus (MRSA) 333 clinical strains (Table 4). Overall, P. graveolens methanolic extract showed complete visible growth inhibition against E. faecalis ATCC 29212, S. aureus ATCC 25923, and the Vancomycin-resistant E. faecalis INSPI 032 strains, while P. zonale inhibited the growth of all these three strains plus the MRSA 333 strain. In contrast, P. odoratissimum only showed complete growth inhibition against S. aureus ATCC 25923, in agreement with our previous report [12].

Drug susceptibility assays

Confirmation of the hits from the initial screening was performed by calculating the minimum inhibitory concentration (MIC) (Table 5). The main difference between the MIC values of these three crude extracts was observed in the MIC of P. zonale towards S. aureus ATCC 25923, which was four times lower (MIC = 250 μg/mL) than the MIC values shown by P. graveolens and P. odoratissimum (MIC = 1000 μg/mL). In addition, among the three Pelargonium species, only P. zonale inhibited the growth of the MRSA 333 strain (MIC = 250 μg/mL).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Growing inhibitory parameters (MIC, and MBC) of P. graveolens, P. odoratissimum and P. zonale methanolic extracts against bacteria of clinical importance using broth dilution format.

https://doi.org/10.1371/journal.pone.0306637.t005

Hsouna and Hamdi reported a MIC value of 2500 μg/mL against S. aureus ATCC 25923 using leaves of P. graveolens to prepare a methanolic extract, but no activity towards E. faecalis ATCC 29212 was observed [7]. This is in contrast with our results displaying MIC values of 1000 μg/mL for both strains. In the case of P. zonale and P. odoratissimum, studies on antimicrobial activities are scarce and mainly reported using essential oils [10,13,14], ethanolic extracts [8] or in combination with zinc oxide (ZnO) nanoparticles using P. odoratissimum aqueous leaf extract [11]. It should be noted that the antibacterial activity of P. odoratissimum methanolic extracts has only been previously reported by us [12], showing MIC values against S. aureus ATCC 25923 similar to this current report.

Overall, P. zonale’s antibacterial activity against Gram-positive bacteria was shown to be superior to the activity presented by P. graveolens and P. odoratissimum (Table 5). This greater activity may be related to the differences in the quantity of the phenolic content, especially of anthocyanins, which are only present in P. zonale (Table 2). For instance, anthocyanins such as cyanidin and pelargonidin detected in pomegranate fruit were found to be active against a broad range of Gram-positive bacteria, including S. aureus and E. faecalis [63]. However, Cisowska and colleagues have pointed out that the antimicrobial activity of anthocyanins is likely to be caused by multiple mechanisms and synergies involving a variety of different chemical compounds [64].

Likewise, there are reports of polyphenolic extracts with antibacterial activity against S. aureus and MRSA strains where the inhibitory activity was attributed to the presence of compounds such as myricetin, quercetin and kaempferol derivatives, (-)-epi-gallocatechin, and coumaroylquinic acid, among others [65]. The same type of compounds were also identified in this study.

Bactericidal activity of P. graveolens, P. odoratissimum, and P. zonale

Next, to assess if P. graveolens, P. odoratissimum, and P. zonale methanolic extracts exhibit bactericidal or bacteriostatic activity, we determined the minimal bactericidal concentration (MBC) (Table 5). The crude extracts showed an MBC/MIC ratio value lower or equal to four, indicating bactericidal activity [31]. Then, to further validate the effects of the methanolic extracts on the growth kinetics of E. faecalis ATCC 29212, S. aureus ATCC 25923, the Vancomycin-resistant E. faecalis INSPI 032, and MRSA 333 strains, time-kill assays were performed (S1 Fig).

Bactericidal activity is defined as 99.9% killing and can be calculated from time-kill curves when reaching a decrease in bacterial growth ≥ 3 Log₁₀ [31]. Overall, P. graveolens, P. odoratissimum, and P. zonale at 4X MIC showed a decrease of 3 Log₁₀ between 6 to 8 h, while the positive control ciprofloxacin at 4X MIC exhibited bactericidal activity at 4 h. Thus, the time-kill values support the MBC data and showed that P. graveolens, P. odoratissimum, and P. zonale exhibited bactericidal activity against E. faecalis and S. aureus susceptible and resistant strains.

Biofilm inhibition and eradication assays

The inhibition and eradication effect of P. graveolens, P. odoratissimum, and P. zonale extracts was also studied against biofilms of S. aureus ATCC 25923, MRSA 333, E. faecalis ATCC 29212, and the Vancomycin-Resistant E. faecalis INSPI 032 (Figs 1 and 2, S2 Table). P. graveolens, P. odoratissimum, and P. zonale were applied at 1X and 2X MIC concentrations against the evaluated bacterial pathogens (Figs 1 and 2). When compared to the positive control (bacteria and media only), the range of the overall inhibition percentage values of the extracts against the evaluated bacterial pathogens was between 20.3 and 51.3% (Fig 1), while the eradication values were between 8.7 and 64.8% (Fig 2). All the negative methanol-water controls (80:20, v/v) showed statistical differences when compared with the positive control, indicating a biofilm inhibition or eradication by itself. It is important to mention that the methanol evaporation step performed in all previous drug susceptibility assays was not done as it interferes with the procedure in both biofilm inhibition and eradication assays. The lack of methanol evaporation clearly showed a greater inhibitory/eradication activity of the negative controls when compared with the positive control and overestimated biofilm inhibition and eradication effects on the evaluation of the plant extracts. Hence, when counteracted with the methanol-water control, no inhibition effects were observed in S. aureus ATCC 25923 and MRSA 333 biofilms when using P. graveolens, P. odoratissimum, and P. zonale extracts (Fig 1A–1D).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Representative illustration of the biofilm inhibition by P. graveolens, P. odoratissimum and P. zonale against bacteria of clinical importance.

The plots illustrated the biofilm inhibition of S. aureus ATCC 25923 by (A) P. graveolens, (B) P. odoratissimum, and (C) P. zonale. Biofilm inhibition of S. aureus MRSA 333 by (D) P. zonale, followed by the biofilm inhibition of (E) E. faecalis ATCC 29212 and (F) E. faecalis INSPI 032 by P. graveolens. Finally, biofilm inhibition of (G) E. faecalis ATCC 29212 and E. faecalis INSPI 032 (H) by P. zonale. Biofilm formation values of methanol-water controls and samples were calculated as the percentage of bacteria biofilm formation through the optical density comparison between methanol-water controls/samples and bacterial growth in only medium culture (positive control). Error bars indicate ± SD of two independent experiments. Statistical significance was analysed using a non-parametric Wilcoxon test (95% confidence interval) for comparison between biofilm formation values where: * p-values < 0.05; ** p-values < 0.01; *** p-values < 0.001; and ns: Non-significant p-value.

https://doi.org/10.1371/journal.pone.0306637.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Representative illustration of the biofilm eradication by P. graveolens, P. odoratissimum and P. zonale against bacteria of clinical importance.

The plots illustrated the biofilm eradication of S. aureus ATCC 25923 by (A) P. graveolens, (B) P. odoratissimum, and (C) P. zonale. Biofilm eradication of S. aureus MRSA 333 by (D) P. zonale, followed by the biofilm eradication of (E) E. faecalis ATCC 29212 and (F) E. faecalis INSPI 032 by P. graveolens. Finally, biofilm eradication of (G) E. faecalis ATCC 29212 and E. faecalis INSPI 032 (H) by P. zonale. Biofilm formation values of methanol-water controls and samples were calculated as the percentage of bacteria biofilm formation through the optical density comparison between methanol-water controls/samples and bacterial growth in only medium culture (positive control). Error bars indicate ± SD of two independent experiments. Statistical significance was analysed using a non-parametric Wilcoxon test (95% confidence interval) for comparison between biofilm formation values where: * p-values < 0.05; ** p-values < 0.01; *** p-values < 0.001; and ns: Non-significant p-value.

https://doi.org/10.1371/journal.pone.0306637.g002

On the other hand, P. graveolens and P. zonale extracts evidenced significant biofilm inhibition values against E. faecalis ATCC 29212 and the Vancomycin-resistant E. faecalis INSPI 032 strains (p-values < 0.05), independently of the presence of methanol, demonstrating higher inhibition values at 2X MIC concentration (Fig 1E–1H). More specifically, at 2X MIC (2000 μg/mL), P. graveolens and P. zonale slightly inhibited biofilm formation by 7.9 and 6.3% against E. faecalis ATCC 29212 (Fig 1E and 1G), respectively, while the Vancomycin-Resistant E. faecalis INSPI 032 showed 32.9 and 22.5% biofilm inhibition when counteracted by the methanol-water control (Fig 1F and 1H). Reports of P. graveolens antibiofilm activity have been found using ethanolic extracts and essential oils. For instance, P. graveolens ethanolic extract showed antibiofilm activity against Streptococcus pyogenes ATCC 19615, ATCC 49399, and one clinical isolate with MIC and MBC values superior to 1000 μg/mL [66]. To the best of the authors’ knowledge, no antibiofilm activity by P. zonale was previously reported.

Concerning the four anthocyanins (cyanidin 3,5-diglucoside, pelargonidin-galactoside, pelargonidin 3,5-diglucoside, and peonin) identified in P. zonale in this study, Jeyaraj and colleagues demonstrated that the aqueous anthocyanin-rich fraction of Clitoria ternatea flowers of Asia was responsible for the biofilm-inhibiting activity (around 64.0%) and also significantly reduced bacterial attachment of Pseudomonas aeruginosa ATCC 9027 [67]. However, the study did not identify individual anthocyanins from the anthocyanin-rich fraction that could be the active molecules responsible for the observed antibiofilm activity. Meanwhile, when compared to the present work, Zhang and colleagues identified similar anthocyanins (cyanidin, pelargonidin, and peonidin) in purple highland barley bran (PHBB) from the Tibetan Plateau in China, evidencing their antibiofilm activity against P. aeruginosa PAO1 and Salmonella enterica ATCC 10398 [68]. Furthermore, Pejin and colleagues evaluated the individual contribution of three anthocyanidins (pelargonidin, cyanidin, and delphinidin), demonstrating their antibiofilm effects on P. aeruginosa PAO1 [69]. Therefore, anthocyanins seem to contribute to the antibiofilm activity of P. zonale in the present study, but further studies are needed to analyse their individual antibiofilm potential and future applications.

Regarding the biofilm eradication assays (Fig 2), the plant extracts showed no eradication ability against S. aureus ATCC 25923, E. faecalis ATCC 29212 and the Vancomycin-resistant E. faecalis INSPI 032 biofilms except for MRSA 333 biofilms, which suffered a slight eradication of 7.2% by P. zonale at 2X MIC (500 μg/mL) when counteracted with the methanol-water control (Fig 2D). However, no statistical difference was observed. Overall, both P. graveolens and P. zonale extracts in the present study revealed biofilm inhibition against the Vancomycin-resistant E. faecalis INSPI 032 despite the presence of methanol, with P. zonale showing a slight biofilm eradication activity towards the MRSA 333 strain.

Influence of P. graveolens, P. odoratissimum, and P. zonale on cell viability

Next, we set to investigate the cytotoxicity effect of the three Pelargonium species in a dose-dependent manner (Fig 3). Macrophage viability with P. graveolens and P. odoratissimum at a concentration of up to 500 μg/mL was between 85 and 100%. A study evaluating P. graveolens acetone extract showed non-toxic effects at a concentration of 83 μg/mL [70]. However, P. graveolens and P. odoratissimum methanolic extracts at a concentration of 1000 μg/mL were toxic, with macrophage viability between 30 and 40% (Fig 3A and 3B). In contrast, P. zonale showed high cytotoxicity (macrophage viability of 50%) at the lowest concentration tested of 125 μg/mL, while no macrophage viability was observed at concentrations higher than 500 μg/mL (Fig 3C). This overall cytotoxicity effect might be related to the presence of tannins in P. graveolens and P. odoratissimum, as well as tannins and flavonoids (like anthocyanins) in P. zonale (Table 2). Tannins and flavonoids are reported as potential anticancer agents for their high levels of cytotoxicity in tumour cells [71]. Thus, the high cytotoxicity of P. zonale suggests that its antibacterial activity might be ascribed to its general cytotoxic effect. Indeed, P. zonale showed the best antibacterial activity among P. graveolens and P. odoratissimum methanolic extracts (Table 5). To the best of our knowledge, no data about the cytotoxicity of P. odoratissimum and P. zonale has been reported.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Raw 264.7 macrophage viability at 48 h upon treatment with the methanolic extracts.

(A) P. graveolens, (B) P. odoratissimum and (C) P. zonale at concentrations of 0, 125, 250, 500 and 1000 μg/mL. The percentage of viability was calculated relative to the control (DMEM media with the dissolvent methanol-water 80:20, v/v). Error bars indicate ± SD of two independent experiments done in quadruplicates.

https://doi.org/10.1371/journal.pone.0306637.g003

Antibacterial activity of P. zonale fractions

The best antibacterial activity was encountered in P. zonale methanolic extract when compared to P. graveolens and P. odoratissimum (Table 5), leading us to identify the active fractions. For this, P. zonale methanolic extract was fractionated with seven eluents using an SPE column (see methods section for details) to determine the tentative compounds exhibiting antimicrobial activity. A total of 14 fractions were obtained, from which fractions 1 to 10 were tested against S. aureus ATCC 25923 at a maximum single-point concentration of 800 μg/mL. The other fractions, 11 to 14, were not evaluated due to low solubility in methanol-water (20:80, v/v). S. aureus ATCC 25923 was selected as a representative of the Gram-positive bacteria tested in this study. Among all the fractions tested, fractions 1 to 4 showed the best antibacterial activity with colony-forming units (CFU) counts of less than 200 CFU/mL when compared with the other fractions (Fig 4A). Fraction 3 showed a slightly better antimicrobial activity (56 CFU/mL, Fig 4B) compared with fractions 1, 2, and 4 (150, 93, and 112. CFU/mL, respectively). Thus, the four fractions were further analysed by HPLC-DAD-MS/MS in positive mode.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Antibacterial effect of P. zonale fractions in S. aureus ATCC 25923 growth.

Fractions 1 to 10 (A) and fractions 1 to 4 (B) were evaluated against S. aureus ATCC 25923 at 800 μg/mL. Growth inhibition of the 14 fractions is expressed as CFU/mL. Error bars indicate ± SD of two independent experiments done in duplicates.

https://doi.org/10.1371/journal.pone.0306637.g004

Chemical characterisation of P. zonale fractions

A chromatographic analysis of fractions 1 to 4 (S2A Fig) revealed a similar chemical profile among them, which was also confirmed by a dendrogram (Fig 5A). The dendrogram is based on the rescaled areas of the metabolites present in a fraction, where areas are used as an indicator of metabolites’ relative concentration. The dendrogram’s elements represent individual fractions, with the distance between them reflecting similarities or differences in metabolite profiles. In Fig 5A, we can see two distinct clusters. The first cluster comprises fractions 1 to 4, and the second cluster comprises fractions 5 to 10. This clustering pattern may indicate a variation in the chemical composition of the analysed fractions, more exactly beyond the fourth fraction.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Overall evaluation of the chemical profile among fractions.

(A) Dendrogram computed from fractions 1 to 10 post solid-phase extraction of P. zonale based on rescaled metabolite areas. Two clusters are observed comprising fractions 1–4 (cluster 1) and comprising fractions 5–10 (cluster 2). (B) Dendrogram based on tentatively identified compounds from fractions 1 to 4. Distances were computed from their structural and physicochemical similarities using ChemMine Tools. Four clusters are observed and are represented in blue (cluster 1), light blue (cluster 2), green (cluster 3) and orange (cluster 4). (C) The relative abundance of each chemical cluster in the fractions was determined by calculating the mean of the normalised areas of all compounds in the cluster. It should be noted that Cluster 5 (yellow), which corresponds to unidentified compounds, has been included.

https://doi.org/10.1371/journal.pone.0306637.g005

Considering the antibacterial activity (Fig 4), fractions 1 to 4 were more relevant and, consequently, the focus of further analysis. Several peaks were found in these fractions, and their tentative identifications are presented in Table 6. The chromatograms and a complete discussion about the tentative identification of compounds are shown in S2B Fig and S3 Table. Several compounds listed in Table 6 were already presented in Table 3 for P. zonale.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 6. Tentative identification of metabolites identified in fractions 1–4 in P. zonale in positive ionization mode.

https://doi.org/10.1371/journal.pone.0306637.t006

In addition, the chemical analysis of the four fractions revealed the presence of phenolic acids (IDs 14, 16, and 17 in Table 6), numerous flavonoids (mostly quercetin derivatives, IDs 18, 20, 23, and 27, and kaempferol derivatives, IDs 21, 25, 28, 32, 33, 35, and 39), and gallotannins (IDs 15, 19, 26, 29, and 30). The clustering of the identified compounds indicated four major groups (Fig 5B): The first group (cluster 1; Fig 5B in blue) is composed mainly of monoglycosylated quercetin (IDs 20 and 27) and kaempferol (IDs 28, 32, 33, and 35) derivatives. The second group (cluster 2; Fig 5B in light blue) is comprised of di-glycosylated derivatives of quercetin (ID18) and kaempferol (IDs 21 and 25), as well as 3-[6’’-(3-hydroxy-3-methylglutaryl)galactoside] quercetin (ID23) and di-glycosylated pelargonidin (ID8). The third group (cluster 3; Fig 5B in green) consists of amino acids (ID10), sugar (ID2), kaempferol (ID39), and phenolic acids (IDs 14, 16, and 17). The fourth group (cluster 4; Fig 5B in orange) has metabolites with a complex structure, consisting mainly of gallotannins (IDs 15, 19, 26, 29, and 30).

Fig 5C shows the average relative abundance computed by the normalised peak areas of each of the identified compounds in each cluster. Additionally, we add the mean relative abundance of the unidentified peaks (defined as cluster 5). All average relative abundance of the clusters decreases after fraction 4 (as we would expect, based solely on the solubility gradient). However, in fraction 3, we can see the highest abundance of clusters 1, 2, 3, 5, and, to a lesser degree, 4. The relative abundance composition in fractions 1 and 2 is fairly similar, but an increment is noticed in fractions 3 and 4. The increment of relative abundance composition noticed in fraction 3 may be responsible for its slightly better inhibitory activity (Fig 4B) when compared with fractions 1, 2, and 4.

Numerous studies point towards the synergistic action of polyphenols in combating microbial growth [72–76]. This synergistic effect arises from the intricate interplay between various secondary metabolites within natural extracts. These interactions, influenced by both polyphenol structure and environmental factors, modulate the overall biological activity [72]. Notably, polyphenols possess the ability to affect multiple biological targets, including enzymes, transport proteins, membrane receptors, and lipid membranes, among others [73]. In fact, a study realised by Mokhar and colleagues demonstrated the antimicrobial activity of polyphenols present in pepper (Capsicum annuum L.) against S. aureus ATCC 6538, evidencing the individual efficacy of quercetin and kaempferol compounds. Moreover, when combined with caffeic acid, these compounds also showed a synergistic effect [74]. Other examples of synergistic effects with polyphenols have been reported, such as combinations of rutin with quercetin, morin, kaempferol, myricetin, and eriodictyol against Bacillus cereus, despite the fact that rutin alone does not possess antibacterial activity [75]. Therefore, the synergistic effects appear to depend on specific ratios and combinations of polyphenols [76].

In our particular case, the antibacterial activity noted on fractions 1 to 4 against S. aureus ATCC 25923 could be attributed mainly to the presence of flavonoids such as quercetin derivatives (e.g. quercetin hexoside, quercetin pentoside, rutin) and kaempferol and its derivatives (e.g. kaempferol-glucorhamnoside, trifolin, juglanin, and kaempferol-3-O-β-rhamnoside), and their synergistic interaction with gallotannins (e.g. 1,2,6-trigalloylglucose, tetragalloylglucopyranose, pentagalloylglucopyranose, hexagalloylglucopyranose, and heptagalloylglucopyranose) and anthocyanins (e.g. cyanidin 3,5-diglucoside, pelargonidin-galactoside, pelargonidin, 3,5-diglucoside and peonin). However, the increase of concentration in cluster 5 (unidentified molecules) in fractions 3 and 4 may indicate that some of the molecules in this cluster may also have a potential antibacterial effect or a synergistic effect in combination with the secondary metabolites mentioned above, which is enhanced by their increased concentration.

Overall, few studies evaluated the biological activity of pure compounds, and most of the reports focus on the antibacterial activity of crude extracts containing a mix of compounds. For instance, a study using pure quercetin compounds exhibited selective antibacterial properties towards S. aureus, MRSA, and S. epidermidis [77]. In addition, quercetin pentoside and other quercetin derivatives identified in walnut leaves were found to selectively inhibit the growth of Gram-positive bacteria (B. cereus, B. subtilis, S. aureus) [78]. Likewise, rutin (a quercetin derivative) exhibited MIC values of 512 μg/mL and 256 μg/mL towards S. aureus ATCC 25923 and E. faecalis ATCC 29212 [79] respectively.

In addition, pure kaempferol and kaempferol-containing extracts have exhibited antibacterial activity towards clinically important bacterial pathogens such as S. aureus and E. faecalis [80]. Antibacterial activity towards S. aureus and MRSA strains have also been reported in studies with different plant extracts containing kaempferol derivatives such as trifolin (kaempferol 3-O-β-D- galactopyranoside) in Terminalia petiolaris methanolic extract [81] and kaempferol-3-rhamnoside in the ethyl acetate extract of Zanthoxylum bungeanum leaves [82]. Nonetheless, other polyphenols were also identified in these plant extracts that could also be associated with the antimicrobial activity reported.

Finally, gallotannins are characterised by the presence of galloyl groups, which are reported to be responsible for gallotannins’ higher antibacterial activity compared to other hydrolysable tannins [58]. In the present study, we identified different gallotannins with a variety of galloyl groups, which may explain the best antibacterial activity encountered in P. zonale extract when compared to P. graveolens and P. odoratissimum.

Thus, based on this information, the antibacterial activity reported for P. zonale towards the Gram-positive bacteria S. aureus ATCC 25923, MRSA 333, E. faecalis ATCC 29212, and Vancomycin-resistant E. faecalis INSPI 032 (Table 5) may be attributed to the presence of these compounds. Although no specific metabolites were identified among the three species, P. zonale distinguishes itself notably for containing gallotannins and anthocyanins, which are unique to this species and may interact synergistically with other metabolites present.

In summary, our study evaluated the chemical profiles of three distinct species within the Pelargonium genus. Our findings revealed the presence of quercetin and kaempferol derivatives in P. graveolens, P. odoratissimum, and P. zonale. Notably, gallotannins and anthocyanins were uniquely identified in P. zonale. Likewise, P. graveolens stands out due to the various types of myricetin derivatives that were not detected in P. odoratissimum and P. zonale. Furthermore, P. zonale displayed superior antibacterial and antibiofilm activities in comparison to the other two species. The antibacterial activity of P. zonale was observed towards the clinically relevant strains of S. aureus ATCC 25923, MRSA 333, E. faecalis ATCC 29212, and Vancomycin-resistant E. faecalis INSPI 032. Fractionation analysis of P. zonale suggests that the antibacterial activity attributed to this plant is due to the presence of quercetin derivatives and kaempferol and its derivates, alongside their synergistic interaction with gallotannins and anthocyanins. Given the current challenges posed by the depletion of effective drug options against the resistant strains of S. aureus and E. faecalis, the secondary metabolites encountered in the fragmentation of P. zonale deserve further investigation.