Antimicrobial Testing of Schinus molle (L.) Leaf Extracts and Fractions Followed by GC-MS Investigation of Biological Active Fractions.

Schinus molle (L.) is a dioecious plant of the Anacardiaceae family, originating in South America and currently widespread in many regions throughout the world. In this work leaf extracts and derived low-pressure column chromatography (LPCC) fractions of S. molle L. male and female plants were investigated for the antimicrobial activity. Leaf extracts were tested on microbes Escherichia coli, Pseudomonas aeruginosa, Staphylococcus aureus, Enterococcus faecalis, Candida albicans and Bacillus subtilis. Furthermore, the extracts showing antimicrobial activity were fractionated by LPCC and the obtained fractions tested on the same microorganism strains. Positive fractions were investigated by gas-chromatography/mass spectrometry (GC-MS) and were seen to be rich in sesquiterpenes, sesquiterpenoids and other terpens. The obtained effects highlighted the antimicrobial properties of S. molle (L.) leaf compounds and revealed their importance as a source of bioactive molecules of potential pharmaceutical interest. To our knowledge, this is the first paper reporting investigations on the chemical composition of the extracts and derived positive fractions from Schinus molle (L.) plants grown in central Italy


Introduction
Over the decades, the extensive use of antibiotics, especially prophylaxis, has led to the development of resistant pathogens [1]. Disease control to combat emerging and re-emerging pathogen resistance can be counteracted by modifying existing antibiotics [2] and searching for new antibiotics from natural products, which can provide a range of molecules to be tested for this purpose.
From this perspective, the literature of the last 20 years has grown in quantity and quality and many studies have been carried out to test the antimicrobial activity of extracts from plant matrices. Gomes and colleagues evaluated antibacterial activity against multidrug-resistant strains of hospital origin and standard strains using extracts exploiting a plant matrix of waste from the Brazilian pepper tree (Schinus terebinthifolia Raddi) processing industry chain [3]. Sharma et al. evaluated the potential antibacterial and antibiofilm capacity of bioactive compounds extracted from onions at different stages of age on gram-negative Escherichia coli and Pseudomonas aeruginosa and gram-positive Staphylococcus aureus and Bacillus cereus [4]. Several studies have shown the antibacterial activity of extracts from plants such as essential oils [5][6][7][8][9]. Industrial hemp was used to obtain essential oil rich in terpenic compounds and this oil was tested against pathogenic and spoilage microorganisms to evaluate the antimicrobial activity [10]. Lavandin essential oil (liquid and vapour phase) showed an antimicrobial activity against gram-negative (E. coli, Acinetobacter bohemicus, and Pseudomonas fluorescens) and gram-positive (B. cereus and Kocuria marina) bacteria [11].
For the first time, in this work the antibacterial activity of extracts and their fractions of leaves from S. molle (L.) grown in central Italy was evaluated on clinically relevant bacterial strains and the chemical composition of biological active fractions investigated.

Results
The investigation of antibacterial activity was carried out on female and male leaf extracts of S. molle (L.) using four solvents with increasing polarity with this crescent order: petroleum ether, diethyl ether, acetone and distilled water.

Testing of Antimicrobial Activity
The antibacterial assay of the S. molle (L.) leaf extracts was carried out using the disk diffusion test. Treatments with petroleum ether (SM♀1) and diethyl ether (SM♀2) on extracts from the S. molle (L.) female plant (Table 1) and with petroleum ether (SM♂L1) and diethyl ether (SM♂L2) on the S. molle (L.) male plant (Table 2) showed inhibition zones between 8 and 17 mm on Staphylococcus aureus, Enterococcus faecalis, Candida albicans and Bacillus subtilis. A minimal inhibitory concentration (MIC) assay of the tested extracts with a positive response in the disk diffusion test (DDT) recorded values between 25 µg/mL and 400 µg/mL. The DTT of the leaf extracts from the female S. molle (L.)-SM♀1, SM♀2, SM♀3 (acetone extract), SM♀4 (water extract)-and the male-SM♂L1, SM♂L2, SM♂L3, SM♂L4-produced no effects on E. coli and P. aeruginosa. The extracts SM♀3, SM♀4, SM♂L3 and SM♂L4 had no effects on any of the strains used (Tables 1 and 2).
The extracts with antimicrobial activity (SM♀1, L2 and SM♂L1, L2) were fractioned by the low-pressure column chromatography (LPCC) technique using solvents with increasing polarity in this order: petroleum ether 100%, petroleum ether/ethyl acetate 95%/5%, 90%/10%, 80%/20%, 70%/30% and 0%/100% and methanol 100%. The obtained fractions have been investigated by disk diffusion and microdilution tests. Concerning the size of the inhibition halo and the MIC values, as shown in Tables 1 and 2, an increase in the antibacterial activity occurred for many of the tested fractions.

Chemical Investigation of Positive Fractions
The fractions derived from female and male S. molle (L.) leaf extracts that produced an inhibition halo equal to or greater than 10 mm were further investigated with the gas-chromatography/mass spectrometry. The identification of the components separated by GC/MS was carried out by comparing the mass spectra for each compound with that reported in mass spectrometry (MS) libraries and by calculating their linear retention indices (LRIs). Table 3 shows the chemical composition of petroleum ether and diethyl ether fractions from female S. molle (L.) leaves; 40 identified molecules are reported. Table 4 describes the chemical composition of petroleum ether and diethyl ether fractions from male S. molle (L.) leaves; 12 identified molecules are listed. The composition of the biologically active fractions obtained by LPCC from the male and female extracts in petroleum ether showed a particular abundance of sesquiterpenes and monoterpene hydrocarbons. The fractions obtained from the male and female diethyl ether extracts were rich in sesquiterpenes and alcohol terpenes; fatty acids, phenols, esters and hydrocarbons were also present.  Table 4. Chemical composition (%) of petroleum ether and diethyl ether fractions from male S. molle (L.) leaves.

Discussion
Plant secondary metabolites possess numerous biological activities and are useful in the treatment of human and animal health problems. Antibacterial activities have been well defined in different classes of natural compounds and, due to the high incidence of antibiotic resistance, the search for new antibiotics is ongoing.
The antibacterial activity of the SM♀1-F7 fraction against S. aureus showed an increase in the inhibition halo, from 10.3 ± 1.5 mm to 20 ± 1 mm, and a reduction in the MIC, from 400 µg/mL to 128 µg/mL, compared to the extract SM♀1. The biological activity could be related to the presence of β-eudesmol (67.9%), together with cyclotetradecene (15.7%), γand α-eudesmol (6.4% and 3.9%, respectively), dehydroxyisocalamendiol (3.4%) and elemol (2.7%). Salem and colleagues [31] demonstrated the antimicrobial activity of the water extract of wood branches of S. molle (L.), which has a high percentage of β-eudesmol; this work supports the relationship between the chemical content found and the biological activity observed. Furthermore, the lowest MIC against S. aureus was obtained in the fraction SM♀1-F4, containing elemol (38.1%), β-terpinene (33.8%)-known for its antibacterial activity against S. aureus [6], β-eudesmol (10.2%), dehydroxyisocalamendiol (8.5%) and γand α-eudesmol (5.2% and 4,1%, respectively). The effects observed against against S. aureus led us to consider a synergism between the molecules present in the fraction and the antimicrobial activity of this strain.
Among fractions obtained from the extract SM♀2, only SM♀2-F4 showed a modest activity against C. albicans. A reduction in its inhibition halo was observed, from 14.3 ± 2.1 mm to 9.7 ± 0.6 mm, and a decrease was seen in the MIC from 50 µg/mL to 32 µg/mL.
The fraction SM♀2-F4 was the only one to show activity in response to E. faecalis, with an increase in the inhibition halo from 8.3 ± 0.6 mm to 14.7 ± 1.2 mm and a reduction in the MIC from 200 µg/mL to 128 µg/mL.
As shown in Table 2, in the SM♂L1-F1 fraction the increase in antibacterial activity in response to B. subtilis was related exclusively to the disk diffusion test (not to the MIC). Indeed, the inhibition diameter reached 26.3 ± 0.6 mm, compared to 13.7 ± 0.6 mm for the SM♂L1 extract. This phenomenon could be due to the presence of molecules already known for their antibacterial activity, such as germacrene-D [41,42], δ-cadinene [52], elixene [53,54] and β-caryophyllene [39,41].
As already reported for Piper nigrum (L.) fruits [40], there could be relationship between the presence of isocalamendiol and the antibacterial activity against S. aureus. In this study, the SM♂L1-F6 fraction showed a good antibacterial activity against this strain, with a halo of 16.7 ± 0.6 mm and an MIC of 8 µg/mL. In this fraction, isocalamendiol (84.8%), β-eudesmol (12.9%) and spathulenol (2.3%) were present. The antibacterial activity of β-eudesmol has been evaluated through the analysis of essential oils of Amazonian species of the genus Guatteriopsis [36] and a low activity against S. aureus was reported, with an MIC of over 1000 µg/mL. Our results supported the hypothesis of an antibacterial activity against S. aureus, exerted by isocalamendiol or by a synergistic action of the three molecules. The inhibitory activity against S. aureus could be associated with the high presence of β-eudesmol, also found in the fractions SM♂L1-F10 (91.7%) and SM♀1-F7 (67.9%), which have a DDT = 15.3 ± 1.2 mm/MIC = 128 µg/mL and a DDT = 20 ± 1 mm/MIC = 128 µg/mL, respectively.
Some papers have reported the experimental data antibacterial activity against S. aureus of isoaromadendrene, epoxide [55]. The ineffectiveness of squalene was reported by Sharma et al. [56] in a study of the ethanolic extract from leaves of Syzygium jambos L. (Alston), in which no antibacterial activity occurred during the testing of squalene.
Most of the fractions derived from SM♂L2 exhibited inhibitory activity against the fungus C. albicans, mainly expressed by the fraction SM♂L2-F6, with a similar DDT result (11.7 ± 1.5 mm in the fraction compared to 12.7 ± 0.6 mm in the total extract) and enhanced MIC activity of 400 µg/mL to 64 µg/mL. As shown by GC-MS investigations, the fraction consisted of hexadecanoic acid. Biological activities of hexadecenoic acid reported in literature, such as larvicidal [57], anti-inflammatory [58] and antimicrobial [35,36]. In this work, we highlighted the antifungal and antibacterial activity of the fatty acid found in the fraction. The antibacterial activity of hexadecanoic acid was confirmed against B. subtilis, with an increase in the inhibition halo from 11.3 ± 1.5 mm in the total extract up to 28.3 ± 1.2 mm in the fraction SM♂L2-F6. This last fraction, in line with the cited literature, also showed an increase in antibacterial activity with the broth microdilution method, with an MIC reduced from 100 µg/mL to 16 µg/mL.

Reagents
All the solvents used for chemical analysis, extraction and fractionation purposes were analytical grade (Sigma-Aldrich, Darmstadt, Germany). The silica gel (60 Å, 0.04-0.063 mm) was from Macherey-Nagel (Düren, Germany). The microorganism media and reagents for antimicrobial tests were purchased from Sigma-Aldrich (Darmstadt, Germany).

Plant Materials and Extraction
The leaves of male and female S. molle (L.) plants were collected during the flowering period from the "Angelo Rambelli" botanical garden in Viterbo (Viterbo, Italy) and identified by Dr. Monica Fonck (Scientific Supervisor) with the number AS21 (male plant) and AS22 (female plant) in the Botanical Garden Catalogue. The leaves were washed with distilled water, air-dried in dark conditions at room temperature and stored at −80 • C until lyophilization. The frozen leaves were lyophilized for 3 days to eliminate traces of water, ground to obtain a powder (4 gr for female leaves and 4 gr for male leaves) and stored at 4 • C. In order to perform serial extractions, the powder was put in a cellulose thimble and processed in a soxhlet apparatus. Three solvents with increasing polarity-petroleum ether (PE), diethyl ether (DE) and acetone (AC)-were used. After the acetone extraction, the solid residue was macerated in distilled water overnight and the aqueous fraction filtered through Whatman paper.

Low-Pressure Column Chromatography
All the obtained extracts-SM♀1, SM♀2 and SM♂L1, SM♂L2-were fractionated by silica gel in the solid phase and solvents with increasing polarity in the mobile phase. The silica gel was packed using petroleum ether and helped by vibrations to improve the packing of the stationary phase. A layer of sand was put on top of the silica to avoid the dispersion of the solid phase and the creation of air bubbles [61]. The solvents used for the run were PE 100%, PE/Ethyl acetate (EA) 95%/5%, PE/EA 90%/10%, PE/EA 80%/20%, PE/EA 70%/30%, EA 100% and methanol 100%. The fractions were manually collected and those with a similar visible color were reunited, obtaining seven fractions from SM♀1, four fractions from SM♀2, nine fractions from SM♂L1 and three fractions from SM♂L2.

GC-MS Analysis
The active fractions were analyzed with a Perkin Elmer GC/MS Clarus 500. The gas chromatograph was equipped with a Stabilwax fused-silica capillary column (Restek, Bellafonte, PA, USA) (60 m × 0.25 mm, 0.5 µm of film thickness).
The analytical conditions were set as follows: the injector was at 280 • C; the oven temperature was programmed from 60 • C to 220 • C at a rate of 5 • C/min and held for 30 min; the carrier gas was helium at 1.0 mL/min. One microliter of methanol was added to the extracts and 2 µL of each was injected directly with a split ratio of 1:20. The Clarus 500 Mass Spectrometer single quadrupole operated in the electron impact (EI) mode at 70 eV. The mass range was from 30 to 450 m/z. The relative percentages for quantification of the components were calculated by the electronic integration of the GC-FID (Gas Chromatograph-Ionization Flame Detector) peak areas using the normalization method. The identification of the components separated by GC/MS was performed by comparing the mass spectra for each compound with that reported in the MS library search (Wiley and Nist 02). Furthermore, the linear retention indices (LRI) of each compound were calculated according to van den Dool and Kratz [62], using a mixture of n-alkanes (C8-C30, Ultrasci) injected directly into the GC injector with the same operating conditions reported above. All the analyses were repeated twice. The percentage area values were obtained electronically from the GC-FID response without the use of an internal standard or correction factors.
The strains were maintained at 4 • C on slants of Tryptic Soy Agar (bioMerieux, Florence, Italy) and Malt Extract Agar for bacteria and molds, respectively. The bacterial and fungal strains were previously activated in Tryptic Soy Agar (TSA) at 37 • C/24 h and Sabouraud Dextrose Agar (SDA) at 25 • C/5 days.
All the inocula were prepared with fresh cultures plated the day before the test.

Disk Diffusion Test (DDT)
Sterile disks (6 mm diameter) impregnated with S. molle (L.) fractions were placed on TSA, having been previously seeded with the bacterial strain. The microorganisms were suspended in sterile Tryptic Soy broth with a turbidity from 0.5 to 1 McFarland (approximately 10 7 -10 8 CFU/mL). S. molle (L.) fractions which had growth inhibitory activity were highlighted by the absence of microorganism growth around spots and the haloes were measured using a Vernier calliper rule and expressed in mm [63,64]. Each strain was tested in triplicate and the results were reported in Tables 1 and 2. A sterilized physiological saline solution (5 µL) and dimethyl sulfoxide (DMSO) (5 µL) were used as a negative control sample.

Minimum Inhibitory Concentration (MIC)
The MIC values of S. molle (L.) fractions were determined against all strains by means of a microwell dilution method [65,66]. Dilutions of each fraction were prepared in DMSO. All 96 microplate wells were prepared by adding 95 µL of Tryptic Soy broth (TSB) and 5 µL of inoculum. Each well was inoculated with a different concentration of extracts (ranging from 1.56 to 3200 µg/mL) or fractions (ranging from 0.25 to 512 µg/mL) [67]. As a negative control, wells containing 195 µL of the nutrient broth and 5 µL of the bacterial strains without extracts were prepared. Each plate was mixed on the plate shaker at 300 rpm for 20 s and then incubated at 37 • C for 48 h. The optical density (OD) of the plates was measured at 570 nm using a microtiter plate reader. The MIC values were expressed as µg/mL, taking into account the density value of each sample. All the experiments were repeated three times [68].

Statistical Analysis
The results were expressed as means ± the standard deviation (SD). Data were analyzed with a one-way analysis of variance (ANOVA) using GraphPad Prism software (GraphPad Prism 5.0, GraphPad Software, Inc., San Diego, CA, USA), with p values of ≤ 0.05 considered statistically significant.

Conclusions
In conclusion, our findings confirm the antimicrobial properties of S. molle (L.) leaf extracts and derived fractions from S. molle (L.) male and female plants. In most cases, the obtained fractions showed improved antimicrobial activity compared to the in toto extracts.
The chemical investigations of positive fractions could suggest a possible correlation between molecules highlighted by GC-MS analysis and the antibacterial activity. Further studies are needed for a more detailed evaluation of the antibacterial properties of single identified molecules in extracts and their possible synergic effects.

Conflicts of Interest:
The authors declare no conflict of interest.