Chemical characterization, antioxidant, genotoxic and in vitro cytotoxic activity assessment of Juniperus communis var. saxatilis

Chemical composition and antioxidative, genotoxic and cytotoxic potential of essential oil (EO) and post-distillation waste (PDW) of Serbian Juniperus communis L. var. saxatilis Pall. was studied in human lung carcinoma (A549) and normal lung fibroblast (MRC-5) cells. GC-MS analysis identified 93.95% of total EO content and determined α-pinen as a dominant component (23.61%). LC-MS/MS analysis of PDW pointed at rutin (12.2 mg g-1) and quinic acid (11.1 mg g-1) as the most abundant. Antioxidativity of PDW was strong in DPPH (IC50 was 5.27 μg mL-1), and moderate in TBA and FRAP assays. Both substances were more cytotoxic to A549 than to MRC-5 cells. Obtained IC50 values were 69.4 μg mL-1 and 120 μg mL-1 for EO, and 1.27 mg mL-1 and 2.86 mg mL-1 for PDW, respectively. PDW was genotoxic (0.3 mg mL-1 and 1 mg mL-1 in A549 and MRC-5 cells, respectively) and induced apoptosis and arrested cell cycle in G2/M phase in A549 cells (0.3 mg mL-1). In mixtures with doxorubicin cytotoxicity of EO and PDW increased, and combination index values (0.12-0.18) revealed clear synergistic effect, stronger in cancer cells. This indicates that J. communis var. saxatilis could decrease the chemotherapeutic doses of doxorubicin, potentially reducing its side effects.


Introduction
Juniperus communis L. is a conifer plant widely distributed in the Northern Hemisphere. Its essential oil (EO), distilled dominantly from berries (female cones) is used as a flavoring in food and alcoholic beverage industry. Juniper berries are used as a spice for meat dishes in European cuisines and also give distinguishing flavor to alcoholic beverage gin (Lim, 2012). Specific type of brandy 'Klekovača', containing juniper berries, is very popular alcoholic drink in Serbia, famous for unique aroma and known for appetite stimulation (Lesjak et al., 2013). Juniperus plants are also well-known in traditional medicine; needles and especially cones are widely used as folk remedies for digestive and gynecological disorders, cold and headache, and also known as potent diuretic agents. They are known as particularly useful against respiratory diseases, cough, bronchitis and asthma (Leporatti and Ivancheva, 2003;Tucakov, 1996). Numerous reports indicate antioxidant, anti-inflammatory and antimicrobial activities of different Juniperus species (Carpenter et al., 2012;Glišić et al., 2007;Lesjak et al., 2013;Orhan et al., 2011), justifying traditional use of Juniperus plants against respiratory disorders. In addition, nephroprotective and hepatoprotective effects of juniper leaves extract have been determined by Al-Attar et al. (2016, 2017. Furthermore, cytotoxic effects of some M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT 2 properties of both EO and PDW. Antioxidant activity was measured by DPPH and TBA assays; FRAP assay was additionally performed for PDW. Cytotoxicity was evaluated in human lung adenocarcinoma epithelial cells (A549) and normal lung fibroblasts (MRC-5) by MTT assay, while genotoxicity was determined in comet assay in the same cell lines. In addition to EO and PDW applied individually, we examined cytotoxic effect of binary combinations of EO, PDW and conventional cytostatic doxorubicin (DOX). Finally, we determined the effect of EO and PDW on apoptosis and cell cycle arrest in A549 cells.

Human cell lines
The human cell lines used in cytotoxicity assay were fetal lung fibroblasts MRC-5 (ECACC 84101801) and human lung adenocarcinoma epithelial cells A549 (ATCC CCL-185). Cells were cultured in DMEM supplemented with 10 % fetal bovine serum at 37 ˚C in 5 % CO 2 . The cells were sub-cultured at 90 % confluence, using 0.1 % trypsin, every 2-3 days.
Adams (2014) and Jovanović (1970) by Nemanja Rajčević (PhD, botanist), and deposited at the Herbarium of University of Belgrade, Faculty of Biology, Institute of Botany and Botanical Garden "Jevremovac" (BEOU Herbarium). Air-dried and finely ground plant material (300 g of seed cones) was added to 1200 mL of dH 2 O and distilled in Clevenger-type apparatus for 4 h, followed by removal of recipient solvent (hexane) under reduced pressure. EO was dissolved in hexane (1:2000) for GC-MS analysis, in methanol for antioxidant activity assays, and in DMSO for MTT assay and flow-cytometric analysis. In order to prepare PDW extract, aqueous solution remained after distillation was evaporated in vacuum at 45 ˚C, dissolved in hot, distilled water (1 g mL -1 ), exhaustively washed with petrol ether (fraction 40-60 ˚C) to remove non-polar compounds and dried under vacuum. For LC-MS/MS analysis dried PDW extract was dissolved in the mixture of 0.05 % aqueous formic acid and methanol (ratio 7:3) to obtain 2 % (w/v) stock solution. For DPPH and FRAP assays PDW extract was dissolved in 80 % aqueous methanol, while dH 2 O was used as solvent for TBA and MTT assays, as well as for the flow-cytometric analysis.

Chemical analysis
Chemical characterization of EO was determined by an GC-MS method described by Lesjak et al. (2013) using Agilent Technologies 6890N gas chromatograph coupled with Agilent Technologies 5975B electron ionization MS detector and controlled by Agilent Technologies MSD ChemStation software (revision E01.01.335) combined with AMDIS (ver. 2.64) and NIST MS Search (ver. 2.0d).
The content of 45 selected secondary biomolecules in PDW extract was determined by an LC-MS/MS method described by Orčić et al. (2014), using Agilent Technologies 1200 Series high-performance liquid chromatograph coupled with Agilent Technologies 6410A Triple-Quad tandem mass spectrometer with electrospray ion source, and controlled by Agilent Technologies MassHunter Workstation software (ver. B.03.01). The content of total flavonoids in PDW was determined according to the method described by Lesjak et al. (2011). The concentration of total flavonoids, expressed as milligrams of quercetin equivalents (QE) per gram of dry weight, was calculated according to the standard calibration curve.

Antioxidant activity
The DPPH radical neutralization effect of PDW and EO was determined by method previously described by Orčić et al. (2011). Briefly, 10µL of the dissolved EO and PDW (serial two-fold dilutions resulting in final concentrations ranging 230-1500 µg mL -1 and 0.52-66.67 µg mL -1 , respectively) was added to mixture of 100 µL of DPPH solution in methanol (66 µmol L -1 ) and 190 µL of methanol. Both negative controls and corrections were included. Radical neutralization activity was estimated by measuring absorbance at 515 nm (Multiskan Spectrum, Thermo Scientific) and calculating the concentrations needed to decrease initial DPPH concentration by 50% (IC 50 ).
Reducing power of PDW was additionally determined in FRAP assay according to the method previously described by Orčić et al. (2011). Briefly, 10 µL of PDW or standard (ascorbic acid) was mixed with 300 µL of FRAP reagent (obtained by mixing 0.3 mol L -1 acetate buffer pH = 3.6, 20 mmol L -1 aqueous solution of FeCl 3 , and 10 mol L -1 2,4,6-tripyridil-s-triazine in 40 mmol L -1 HCl, in 10:1:1 ratio). Correction (absorbance of the untreated PDW) and control (absorbance of the FRAP reagent) were also measured. After 6 min of incubation, the absorbance at 593 nm was measured. The reducing power, expressed as mg of ascorbic acid equivalents (AA) per gram of dw, was calculated according to the standard calibration curve. All reactions in antioxidant assays were carried out in triplicates.

Cytotoxicity and drug synergism analysis
The cytotoxic effect of EO, PDW and DOX was measured by MTT assay, as described by Nikolić et al. (2015). Cell viability was determined by measuring absorbance at 570 nm, using a micro-plate reading spectrophotometer (Multiskan FC, Thermo Scientific). For each test substance or its combinations, three independent experiments with eight replicates for every concentration were performed. To evaluate the nature of interaction between test substances in binary mixtures we used combination index (CI) analysis, providing quantitative definition for additive effect (CI=1), synergism (CI<1), and antagonism (CI>1) in drug combinations (Zhao et al., 2004). The CI was calculated for IC 50 values of the mixtures, using the formula: CI=D 1 /Dx 1 +D 2 /Dx 2 , where D 1 is the concentration of the first test substance in the binary mixture; Dx 1 is the concentration of the first test substance alone; D 2 is the concentration of the second test substance in the binary mixture; Dx 2 is the concentration of the second test substance alone.

Apoptosis and cell cycle analysis
Apoptotic cell death and analysis of the cell cycle phase distribution were analyzed using a fluorescence activated sorting cells (FACS) Calibur Becton Dickinson flow cytometer and Cell Quest computer software. Apoptotic or necrotic cell death was assessed using the annexin V-fluorescein isothiocyanate /propidium iodide (annexin V-FITC/PI) kit, as described by Srdic-Rajic et al. (2016).
Samples were prepared according to the manufacturer's instructions. Briefly, annexin V binds to the exposed phosphatidylserine of the early apoptotic cells, while PI labels the late apoptotic/necrotic cells, containing damaged membrane. The numbers of viable (annexin − PI − ), early apoptotic (annexin + PI − ) and late apoptotic/necrotic (annexin + PI + ) cells were determined. Quantitative analysis of the proportion of cells in different cell cycle phases, including the hypodiploid cells with fragmented DNA (sub-G 0 /G 1 ), was performed after staining with PI.

Comet assay
The genotoxic effect of EO, PDW and DOX was measured by Comet assay, as described by Nikolić et al. (2015). The comets were visualized using fluorescence microscope (Leica, DMLS, Austria) with an excitation filter 510-560 nm, barrier filter 590 nm, at 400x magnification. Image analysis software (Comet Assay IV, Perceptive Instruments, UK) was used for comet analysis. Fifty nuclei per experimental point in each of the three independent experiments were analyzed; the tail intensity was scored as a reflection of DNA damage.

Statistical analyses
The one-way ANOVA with Mann-Whitney U test was employed for the results of the comet assay, while for all the rest assays the one-way ANOVA with Tukey's post hoc test was used. The difference was considered significant when p<0.05.

Results and Discussion
The chemical characterization of EO by GC-MS analysis identified 93.95 % of total components and revealed 20 peaks exceeding 1 % and no single peak exceeding 25 % (Fig. 1).
Unidentified 6.05 % of the EO belongs to the sesquiterpenes that could not be identified due to insufficient selectivity (existence of several compounds with similar spectra and retention), lack of corresponding spectra in used libraries, and/or insufficient quality of experimental spectra due to their low concentrations. Identified constituents are exclusively monoterpene and sesquiterpene hydrocarbons; all of them were unsaturated (1-3 double bonds), and all except β-myrcene were cyclic (64.2 % bicyclic, 24.8 % monocyclic, 3.6 % tricyclic. Among others, α-pinene (23.61 %), δ-cadinene (10.71 %), sabinene (9.53 %), α-muurolene (6.58 %) and γ-cadinene (5.87 %) were the most dominant (Table 1). The high abundance of α-pinene and sabinene is consistent with previous studies, while observed significant quantitative differences in terpenoid profiles could be attributed to intraspecific variability, which is generally high in Juniperus species (Adams, 2014).
Out of the 45 investigated secondary biomolecules, only 25 were detected in PDW. LC-MS/MS analysis identified 3.2 % of its total content (Table 2). Among the quantified constituents, rutin (12.2 mg g -1 ) and quinic acid (11.1 mg g -1 ) were the most abundant, followed by catechin (5.53 mg g -1 ) and epicatechin (1.74 mg g -1 ). The content of common phenolic acids was low, with total M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT 7 hydroxybenzoic acids amounting up to 0.34 mg g -1 and total hydroxycinnamic acids up to 0.26 mg g -1 .
Total content of flavonoids determined by two methods -LC-MS/MS and spectrophotometric -was in a good agreement (20.5 mg g -1 and 19.1 mg g -1 , respectively).
Antioxidant activity of EO and PDW was measured by DPPH and TBA assays, with BHT used as a positive control. Additionally, rutin and quercetin were used for comparison in DPPH assay.
The results showed strong radical neutralization activity of PDW, which was close to that of BHT (IC 50 values were 5.27 µg mL -1 and 4.9 µg mL -1 , respectively), but lower than the activity of rutin (IC 50 value was 1.8 µg mL -1 ) and especially of quercetin (IC 50 value was 0.4 µg mL -1 ). The activity of EO was multifold lower, with IC 50 value determined at 1.88 mg mL -1 . The efficiency of lipid peroxidation inhibition was lower than the efficiency of DPPH radical neutralization. Obtained IC 50 values were 0.54 mg mL -1 , 2.44 mg mL -1 and 20.35 µg mL -1 for PDW, EO and BHT, respectively.
Additionally, FRAP test was used for PDW and it demonstrated moderate reducing capacity towards Fe 3+ -TPTZ complex, with 78.77 mg of ascorbic acid equivalents per g of dry weight. Observed antioxidativity of PDW can be partially attributed to the dominant identified compounds rutin, catechin and epicatechin, since they are ubiquitous plant phenolics with well-known antioxidant activities (Azevedo et al., 2013;Cruz-González et al., 2016). Furthermore, strong antioxidant potential of PDW is in accordance with the literature data concerning antioxidant potential of Juniperus extracts (Lesjak et al., 2013;Orhan et al., 2011). However, low antioxidant effect of EO is not in line with current literature data (Lesjak et al., 2013).
Considering that lung cancer is one of the most common cancers with high mortality and frequent development of cytostatic resistance (Chang, 2011), in further work we examined the cytotoxic potential of EO and PDW against lung adenocarcinoma A549 cells and normal fetal lung fibroblasts MRC-5. The conventional cytostatic DOX, used to treat numerous malignances including lung cancer (Tacar et al., 2013), was used as a positive control. MTT assay revealed that both EO and PDW induced cytotoxicity in a dose-dependent manner. Although cytotoxicity of EO and especially of PDW was considerably lower comparing to DOX, preferable feature was that both substances exhibited higher selectivity towards cancer A549 cells (Supplementary Fig. 1). Taking into account M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT 8 the EO composition, its cytotoxicity could be attributed to lipophilic terpenoid components which disrupt and permeabilize cell membranes, especially mitochondrial, leading to reactive oxygen species (ROS) release (Bakkali et al., 2008). Higher cytotoxicity of EO against A549 cells could be attributed to additional cytotoxic mechanisms, recognized specifically in cancer cells, such as topoisomerase inhibition, modulation of p53, bcl-2, AMPK and MAPK/ERK pathways and inhibition of isoprenylation of p21 ras . These mechanisms have been described for limonene, β-elemene, βcaryophyllene and α-humulene (Lesgards et al., 2014), constituting 12.45 % of J. communis EO. On the other hand, cytotoxicity of PDW could be assigned to its polyphenols. Although polyphenols are recognized as naturally occurring antioxidants, their high concentrations possess pro-oxidant properties and therefore they could induce cytotoxicity (Babich et al., 2011). Indeed, obtained results show that effective concentrations of PDW were extremely high, indicating that cytotoxicity could be attributed to pro-oxidative effect of its polyphenols.
In order to further investigate obtained anticancer activity of EO and PDW, their pro-apoptotic potential was monitored using flow cytometric analysis of A549 cells. Cells were initially treated with IC 50 values of EO (69.4 µg mL -1 ) or PDW (1.27 mg mL -1 ) for 24 h. However, this concentration of PDW induced high percent of the late apoptotic and necrotic cells (data not shown) and subsequently we applied it in lower concentration, inducing only 30% lethality (0.3 mg mL -1 ). As shown in Fig. 2, EO showed weak apoptotic effect, which could be seen as a slight increase of the cell numbers in early (annexin + /PI − ) and late apoptotic phase (annexin + /PI + ), but did not significantly affect the cell cycle progression. On the contrary, PDW induced significant increase in both early and late apoptotic cells and caused a cell cycle arrest in G 2 /M phase that was associated with an increased number of apoptotic hypodiploid cells with fragmented DNA (sub-G 0 /G 1 phase). This demonstrates that apoptotic potential of PDW could account for its cytotoxicity. Moreover, when we examined genotoxicity of EO and PDW in comet assay, no genotoxicity of EO was found, while PDW induced significant genotoxic effect, stronger in A549 cells (Fig. 3). This shows that pro-apoptotic effect of PDW could be mediated by its damaging effect on cellular DNA, which could be induced by its prooxidant activity. It is known that DNA damage initiates p53-mediated signaling cascades that can lead M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT 9 to cell cycle arrest and apoptosis (Vogelstein et al., 2000). Literature data point that among identified PDW components rutin can induce DNA damage and modulate a wide range of intracellular signaling cascades leading to apoptosis of cancer cells (Perk et al., 2014;Marcarini et al., 2011).
Taking into account that clinical use of DOX is limited due to systemic toxicity, especially to liver, heart and kidney, a continuous search for auxiliary substances which could decrease its therapeutic doses is encouraged (Wang et al., 2012). For that reason, cytotoxic potential of binary combinations of EO/PDW with DOX was also monitored. While DOX was applied in concentrations sub-lethal for normal MRC-5 cells (0.312 µg mL -1 and 0.625µg mL -1 , resulting in about 70 % of cells survival), the concentrations of EO and PDW were variable, but also in the range inducing low cytotoxicity (survival was higher than 70 %). Comparison of results obtained in different cell lines indicated that both EO and PDW highly sensitized cancerous cell line to DOX, while the effect in normal cells was less pronounced (Figs. 4 and 5).
Cytotoxicity was also monitored for binary combinations of EO and PDW, both applied at concentrations which were sub-lethal for normal MRC-5 cells (survival was higher than 70 %). The results showed that increased cytotoxicity was obtained for all applied combinations in A549 cells, while in MRC-5 cells it was observed only at the highest tested concentration of PDW (Fig. 6). In Table 3 we summarized cytotoxic effect of tested compounds and their binary mixtures, specified by estimated IC 50 values. It clearly shows that concentrations required to induce 50 % lethality in both cell lines were considerably lower in combinations than if substances were applied individually.
To monitor the mode of interactions between EO, PDW and DOX in binary mixtures, the combination index (CI) was calculated for the IC 50 concentrations (Zhao et al., 2004). As shown in Table 4, all determined CI values were lower than 1 indicating synergism. Considering that DOX suppresses cell division by intercalating into DNA which inhibits topoisomerase II, but also by extensive production of ROS which lead to oxidative stress (Mizutani et al., 2005), we propose that pro-oxidative activity of PDW and ROS release caused by EO could increase DOX-induced oxidative stress and enhance its cytotoxic effect. Additionally, DNA damage caused by PDW could contribute M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT 10 to DOX induced genotoxicity. Increased permeability of cell membranes caused by EO could also improve the uptake of DOX and its accumulation in nuclei, enhancing its cytotoxic effect (Ambrož et al., 2016). Stronger synergism obtained in A549 compared to MRC-5 cells is in line with previously reported data indicating that some EOs and polyphenols can activate additional pro-apoptotic pathways specific for cancer cells (Lesgards et al., 2014;Mahbubet al., 2015;Wang et al., 2004;Wang et al., 2012).

M A N U S C R I P T
A C C E P T E D      Cells were treated with EO+DOX mixtures for 24h at 37°C in 5% CO 2 and 100% humidity. Experiments were performed in two independent experiments with eight replicates.
Effect of combinations containing 0.312 µg mL -1 of DOX and shown concentrations of EO against A549 cells (A) and

MRC-5 cells (B).
Effect of combinations containing 0.625 µg mL -1 of DOX and shown concentrations of EO against A549 cells (C) and

MRC-5 cells (D).
Statistically significant difference (p<0.05) comparing to solvent control (*), to DOX alone (a), and between binary combination (EO+DOX) and EO alone (b). Cells were treated with PDW+DOX mixtures for 24h at 37°C in 5% CO 2 and 100% humidity. Experiments were performed in two independent experiments with eight replicates.
Effect of combinations containing 0.312 µg mL -1 of DOX and shown concentrations of PDW against A549 cells (A) and

MRC-5 cells (B).
Effect of combinations containing 0.625 µg mL -1 of DOX and shown concentrations of PDW against A549 cells (C) and