Cytotoxicity and Genotoxicity Assessment of Sandalwood Essential Oil in Human Breast Cell Lines MCF-7 and MCF-10A

Sandalwood essential oil (SEO) is extracted from Santalum trees. Although α-santalol, a main constituent of SEO, has been studied as a chemopreventive agent, the genotoxic activity of the whole oil in human breast cell lines is still unknown. The main objective of this study was to assess the cytotoxic and genotoxic effects of SEO in breast adenocarcinoma (MCF-7) and nontumorigenic breast epithelial (MCF-10A) cells. Proteins associated with SEO genotoxicity were identified using a proteomics approach. Commercially available, high-purity, GC/MS characterized SEO was used to perform the experiments. The main constituents reported in the oil were (Z)-α-santalol (25.34%), (Z)-nuciferol (18.34%), (E)-β-santalol (10.97%), and (E)-nuciferol (10.46%). Upon exposure to SEO (2–8 μg/mL) for 24 hours, cell proliferation was determined by the MTT assay. Alkaline and neutral comet assays were used to assess genotoxicity. SEO exposure induced single- and double-strand breaks selectively in the DNA of MCF-7 cells. Quantitative LC/MS-based proteomics allowed identification of candidate proteins involved in this response: Ku70 (p = 1.37E − 2), Ku80 (p = 5.8E − 3), EPHX1 (p = 3.3E − 3), and 14-3-3ζ (p = 4.0E − 4). These results provide the first evidence that SEO is genotoxic and capable of inducing DNA single- and double-strand breaks in MCF-7 cells.


Introduction
Sandalwood essential oil (SEO) is extracted from trees from the Santalum genus. Among various species of sandalwood, the most common are the Indian sandalwood (Santalum album) and Australian sandalwood (S. spicatum) followed by the species found in Hawaii (S. ellipticum), New Caledonia (S. austrocaledonicum), and French Polynesia (S. insulare) [1]. The chemical composition of SEO has been studied in detail. At least 300 chemical constituents have been identified, of which (Z)--santalol and (Z)--santalol are the most abundant [2]. SEO is used in the food industry as a flavoring ingredient and also in the cosmetic and perfume industry [3]. SEO has also been studied as a chemopreventive agent for skin papillomas in mice [4][5][6] and in vitro in human epidermoid carcinoma cells [7]. Matsuo and Mimaki studied the cytotoxicity of various -santalol derivatives and found that some of them present tumor-specific cytotoxicity [8]. This group also found that -santalol induces DNA fragmentation as a result of apoptosis [7]. Even though previous studies have provided valuable data on the pharmacological properties of individual constituents of SEO, there is no information on the activity of whole SEO in human breast cells, or on its DNA-damaging potential. Due to the increased popularity of essential oils (EOs) for massage and aromatherapy, people are frequently exposed to SEO through various routes of administration including the skin. EOs can easily be absorbed through the human skin due to their lipid solubility and extremely low molecular size and the lipophilic nature of the skin itself [9].
DNA damage can be induced by a wide variety of factors such as radiation and chemical substances [10]. While some chemicals can cause DNA damage, mutations are only produced when the DNA repair system of the cell malfunctions 2 Evidence-Based Complementary and Alternative Medicine or during replication of the damaged DNA [11]. DNA-damaging chemicals are considered genotoxic [12]. Genotoxic compounds can cause mutations in somatic cells that can lead to chromosomal alterations, insertions, deletions, or translocations [13]. Since EOs are present in many household products and are used in folk medicine, an assessment of the potential of these substances to induce damage to DNA is needed. Some groups have begun to evaluate the capacity of EOs to induce DNA damage. For example, Péres and coworkers found that the EO extracted from pariparoba (Piper gaudichaudianum), a herb commonly used in Brazilian folk medicine, is cytotoxic and genotoxic against V79 cells [14]. Genotoxicity of pariparoba EO was assessed using the comet and micronucleus assays. This group found that this EO has the capacity to induce DNA damage in a dose-dependent manner. Moreover, they found that lipid peroxidation could be the potential mechanism for its cytotoxic and genotoxic effects [14]. Although the genotoxicity of some EOs has been studied and possible mechanisms have been proposed, there are large gaps in the knowledge of the DNA-damaging potential of EOs in human breast cells. Moreover, there are even fewer studies that elucidate the mechanisms or identify key proteins involved in the induction of DNA damage. Therefore, the objective of this study is to assess the cytotoxic and genotoxic effects of SEO in human breast cell lines MCF-7 and MCF-10A. We also aimed to identify proteins associated with genotoxicity by means of quantitative tandem mass spectrometry-based microwave and magnetic (M 2 ) proteomics [15,16]. This technique allowed us to establish a correlation between relative protein expression and the genotoxic effects of SEO at different exposure times in human breast cancer cells.

Sandalwood Essential Oil.
Commercially available 100% pure SEO was acquired from Mountain Rose Herbs Co. (Eugene, OR). Their oils comply with high standards of quality and are certified to be grown using organic components, thus avoiding any interference from pesticides. These oils are certified as organic (through the Oregon Tilth) and are characterized by the company using GC/MS (Table 1).

Liposome Encapsulation for Essential Oil Delivery.
EOs, in general, are hydrophobic and biologically unstable as are many other plant products [17]. They have poor solubility in water and are distributed poorly to target sites [18]. Due to these characteristics, we decided to use liposomal encapsulation for delivery to improve the stability and bioavailability of SEO across cell membranes as presented by Shoji and Nakashima (2004) [19]. SEO was encapsulated into liposomes by Ingredient Innovative International (3i) Solutions Company (Wooster, OH). Each liposome is composed of 15% SEO, 78.5% water, 4% enzyme modified lecithin, and 2.5% polysorbate.

MTT Assay.
Cell proliferation was measured using the MTT assay as described by Mosmann (1983) [20,21]. MTT reduction is a measure of mitochondrial activity based on the enzymatic reduction of a tetrazolium salt by the mitochondrial dehydrogenase of viable cells. This provides an estimate of cell viability. MCF-7 and MCF-10A cells were seeded in 96-well plates to a volume of 1,000 cells per (2) class 2, with tail shorter than the diameter of the head (nucleus); (3) class 3, with tail as long as 1-2x the diameter of the head; and (4) class 4, with tail longer than 2x the diameter of the head ( Figure 1). The DNA damage index (DI) was calculated as described by Sastre et al. (2005), using a modified weighted average formula: where refers to the damage category (from 1 to 4) and is the number of cells belonging to each damage category [22]. Kobayashi et al. (1995) showed that the manual microscopic analysis was less time-consuming and had equal or better sensitivity than the computerized image analysis [23].

Microwave and Magnetic ( 2 ) Sample Preparation.
Sample preparation for isobaric labeling was performed as described by Raphael et al. (2014) [24]. C8 magnetic beads (BcMg, Bioclone Inc., CA) were suspended in 1 mL of 50% methanol. Around 100 L of the beads was washed with equilibration buffer (200 mM NaCl and 0.1% trifluoroacetic acid (TFA)). Lysate from breast cancer cell lines untreated (1 126 ) and treated with 6 g/mL SEO for 5 minutes (1 127 ), 1 hour (1 128 ), 5 hours (1 129 ), and 24 hours (1 130 ); and a pooled reference of all the samples (1 131 ) (25-100 g at 1 g/ L) was mixed with preequilibrated beads and 1/3rd sample binding buffer (800 mM NaCl and 0.4% TFA) by volume. The beads were washed twice with 40 mM triethylammonium bicarbonate (TEAB), and 10 mM dithiothreitol (DTT) was added followed by microwave heating for 10 s. After removing the DTT solution, 50 mM iodoacetamide (IAA) was added followed by microwave heating for 10 s. Beads were washed with 40 mM TEAB and resuspended in 150 L of 40 mM TEAB. In vitro proteolysis was performed with 4 L of trypsin in a 1 : 25 trypsin-to-protein ratio (stock = 1 g/ L in 50 mM acetic acid) and microwave-heated for 20 s in triplicate. The supernatant was transferred to a new tube for immediate use or stored at −80 ∘ C. Released tryptic peptides from digested lysates, including the reference material described above, were modified at the N-terminus and at lysine residues with the tandem mass tagging (TMT)-6plex isobaric labeling reagents (Thermo Scientific, CA). Each sample was encoded with one of the TMT-126-130 reagents, while reference material was encoded with the TMT-131 reagent. Then, 41 L of anhydrous acetonitrile was added to 0.8 mg of TMT labeling reagent and 25 g of lysate was added and microwave-heated for 10 s. To quench the reaction, 8 L of 5% hydroxylamine was added to the sample at room temperature. To normalize across all SEO-treated samples, TMT-encoded lysates from individual samples, labeled with the TMT-126-130 reagents, respectively, were mixed with the reference material encoded with the TMT-131 reagent in a 1 126 :  were fragmented by data-dependent high-energy collisioninduced dissociation (HCD Upon centrifugation, total cellular proteins were collected for quantification using the Quick Start Bradford Protein Assay (Bio-Rad, CA). 15 g of the total cellular proteins from each sample was treated with ME (5% by volume) prior to boiling for 5 minutes and separating proteins on 10% SDS-PAGE gels. Separated proteins were transferred to a 0.

Cytotoxicity Assay.
After treatment of the MCF-7 and MCF-10A cell lines with eight concentrations of SEO, there was a decrease in cell viability to less than 20% in both cell lines ( Figure 2). From these data, we calculated IC 50 , which is the concentration of SEO required to reduce cell viability by 50%. For the MCF-7 cell line, IC 50 was 8.03 g/mL, while for the MCF-10A cell line it was 12.3 g/mL.

DNA Single-Strand Breaks Caused by Sandalwood
Essential Oil. The alkaline version of the comet assay allowed us to determine the capacity of SEO of inducing DNA singlestrand breaks (SSBs) in both cell lines. Our results show that increasing concentration of SEO reduces the incidence of class 1 comets to almost 10% and increases the frequency of class 2 comets in MCF-7 cells (Figures 3(a) and 4(a)-4(e)). In Figure 3(c), it can be seen that the DI increases in a dose-dependent manner. The highest DIs were observed at 6 and 8 g/mL with values of 1.81 and 2.11, respectively. This effect was statistically significant for both concentrations ( = 0.039 and 0.030, resp.) (Figure 3(c)). For the human breast nontumorigenic cell line MCF-10A, this effect was not observed. Even with increasing concentration of SEO, the incidence of class 1 comets remained consistent and the appearance of class 2, class 3, or class 4 comets was rare (Figures 3(b) and 4(g)-4(k)). In Figure 3(c), the DI remains almost unchanged even at the highest concentration of SEO. When comparing the DI in the two cell lines at the same concentrations, the increase in DNA damage is more evident at 6 and 8 g/mL of SEO. At 6 g/mL, the DI for MCF-10A cells was 1.14 whereas for MCF-7 cells it was 1.81 ( = 0.01). At 8 g/mL, the DI for MCF-10A cells was 1.11, while for MCF-7 cells it was 2.11 ( = 0.05).

DNA Double-Strand Breaks Caused by Sandalwood
Essential Oil. The neutral comet assay allows for determination of DNA double-strand breaks (DSBs). In these experiments, a similar trend was observed as in the alkaline comet assay (Figures 3 and 4). As presented in Figures 5(a) and 6(a)-6(e), SEO decreased the appearance of class 1 comets to almost 30% with increasing concentration in MCF-7 cells, therefore, increasing the appearance of higher class comets 3 and 4 ( Figure 5(a)). At the highest concentration studied (8 g/mL SEO), the frequency of comets was almost the same for all four classes, therefore, yielding a higher DI ( Figure 5(a)). While control (untreated) MCF-7 cells had a DI of 1.38, as the concentration of SEO increased the DI also increased. Similar to the results of the alkaline comet assay, at 6 and 8 g/mL, the DIs were 2.42 and 2.35, respectively ( = 0.02, = 0.05) ( Figure 5(c)). For MCF-10A cells, the results were similar to the ones obtained through the alkaline comet assay. The frequency of class 1 comets remained almost unchanged with increasing concentration of SEO, while the appearance of higher class comets was rare ( Figure 5(b)). As a result of this, the DI remained similar independent of the concentration of SEO used ( Figure 5(c)). When comparing the two cells lines, the greatest difference in DI was at the concentrations of 4, 6, and 8 g/mL. At 4 g/mL, the DI for MCF-10A cells was 1.32, while for MCF-7 cells it was 2.09 ( = 0.008). At 6 g/mL, the DI for MCF-10A cells was 1.23, while for MCF-7 cells it was 2.42 ( = 0.03). At the highest concentration studied (8 g/mL SEO), the DI for MCF-10A cells was 1.38, while for MCF-7 cells it was 2.35 ( = 0.04) ( Figure 5(c)).  Table 2). The expression of selected candidate proteins was confirmed with Western blotting. Figure 7 shows the results obtained for this validation. Almost all of the proteins studied were induced after 1 hour of exposure to SEO. However, their expression profile varies depending on the exposure time. For example, Ku70 shows a slight increase in expression at 1 hour of exposure but reaches its peak at 24 hours (Figure 7(b)). However, this increase was not statistically significant ( = 0.290). In contrast, Figure 7(c) shows that Ku80 has its greater expression after 1 hour of exposure to SEO ( = 0.024). After this time point, the expression of this protein becomes reduced but it is still greater than that in the untreated sample. EPHX1 has an expression profile similar to the one of Ku80 with the highest expression after 1 hour of exposure to SEO ( = 0.050) and decreases in expression at 5 and 24 hours ( = 0.050 and 0.025, resp.) (Figure 7(d)). 14-3-3 also becomes induced at 1 hour of exposure to the oil ( = 0.030) (Figure 7(e)).

Discussion
Our results suggest that SEO has selective genotoxic effects in MCF-7 cells when compared with noncancerous MCF-10A cells at noncytotoxic concentrations. Our findings provide evidence that SEO is capable of inducing single-and doublestrand DNA breaks in the human breast cancer cell line MCF-7. This study provides, to our knowledge, the first evidence that SEO has dose-dependent cytotoxic and genotoxic effects in the human breast adenocarcinoma cell line (MCF-7), whereas it was cytotoxic but not genotoxic to the MCF-10A cell line. Only a few studies have used the MCF-10A cell line as a model to study genotoxicity in nontumorigenic breast epithelial cells. A study by Stankevicins and coworkers studied the genotoxic effect of low dose radiation in this cell line using three doses of X-ray radiation including 12 and 48 mGy/28 kV and 5 Gy/30 kV. After radiation exposure, the cells were allowed to recover for 4 and 24 hours and the DNA damage was measured using the comet assay. This group found that although irradiation increased the amount of DNA lesions initially in MCF-10A cells, at 24 hours, the cells recovered their DNA integrity as was observed in the reduced levels of DNA damage measured with the comet assay [25]. This finding is also consistent with a study by Evidence-Based Complementary and Alternative Medicine Figure 4: Alkaline comet assay performed in MCF-7 ((a)-(f)) and MCF-10A ((g)-(l)) cells after exposure to sandalwood essential oil for 24 hours at concentrations of ((b), (h)) 2 g/mL, ((c), (i)) 4 g/mL, ((d), (j)) 6 g/mL, and ((e), (k)) 8 g/mL. EMS (12 mM) was used as a positive control ((f), (l)). Panels (a) and (g) show untreated cells.
Francisco and coworkers, in which they studied the induction and processing of DNA damage in breast cancer cells and the nontumorigenic cell line MCF-10A, upon exposure to radiotherapy-relevant -radiation doses. They assessed DNA damage using the comet assay to measure single-and doublestrand breaks. Their results show that breast cancer cells (MCF-7) have a tendency to accumulate more DNA lesions than MCF-10A after -radiation exposure [26]. These studies provide evidence on the decreased susceptibility of MCF-10A cells to DNA damage. These findings, previously reported in the literature, can partially explain the selective genotoxicity of SEO. Our study also examines the question of potential genotoxic effects caused by the whole SEO rather than on specific chemical components such as -santalol. When studying EOs, some biological effects are attributed to their main constituents; however, the possible synergy of all of the chemical components in the mixture working together must also be evaluated. In our study, we have shown that SEO induces DNA damage in the form of single-and double-strand breaks at nontoxic concentrations in MCF-7 cells. Since we assessed the genotoxic capacity of SEO through the comet assay, we decided to use quantitative LC/MS-based proteomics at multiple time points, incorporating rapid M 2 sample preparation, to identify specific proteins related to DNA double-strand breaks (DSBs), cell cycle control and regulation upon DNA damage, and cellular metabolism of genotoxic compounds.
In the proteomics data analysis, Ku70 and Ku80 were found to be differentially expressed upon SEO exposure. These proteins are involved in the process of repairing DNA DSBs; therefore, their differential expression correlates with the results of the alkaline and neutral comet assay. Upon induction of DNA DSBs, the cell can activate two types of repair: homologous recombination or nonhomologous end joining (NHEJ). NHEJ allows the ligation of two DNA ends without requiring sequence homology [27]. One of the key components of this DNA repair pathway is the Ku protein which is a heterodimeric complex composed of Ku70 (70 kDa) and Ku80 (80 kDa) subunits. This complex binds selectively to double-stranded DNA ends in a sequence independent manner. Ku70 and Ku80 initiate the repair process of DNA DSBs by activation of the DNA-dependent protein kinase after binding to the DNA DSBs [28]. Gu et al. (1997) showed that cells deficient in Ku70 expression have increased radiosensitivity and defects in DNA end-binding activity [29]. Moreover, Ku80 null mice have shown an increase in chromosomal aberrations and malignant transformations [30]. Upregulation of Ku70 occurs upon exposure to ionizing radiation via p53/ATM-dependent mechanism [31]. Various studies have suggested that the Ku complex recognizes DSBs and serves as an alignment factor that promotes end joining [32][33][34][35]. The first step for DSB repair is the recognition of the Evidence-Based Complementary and Alternative Medicine Figure 6: Neutral comet assay performed in MCF-7 ((a)-(f)) and MCF-10A ((g)-(l)) cells after exposure to sandalwood essential oil for 24 hours at concentrations of ((b), (h)) 2 g/mL, ((c), (i)) 4 g/mL, ((d), (j)) 6 g/mL, and ((e), (k)) 8 g/mL. EMS (12 mM) was used as a positive control ((f), (l)). Panels (a) and (g) show untreated cells.
damage by sensor proteins like the ATP-dependent helicase II (Ku70). This enzyme is found in increased levels 30 minutes after DSB induction [36]. Both of these proteins were found to be induced in samples treated with SEO, after 1 hour of exposure.
Our results also show a differential expression of EPHX1, or the human microsomal epoxide hydrolase (mEH). This protein is one of the many biotransformation enzymes which functions in the detoxification of chemical epoxide intermediates produced during phase I oxidation reactions [37].   Figure 7: Sandalwood essential oil induces protein expression of Ku70, Ku80, EPHX1, and 14-3-3 in MCF-7 cells. (a) Western blot analysis from 25 g of protein extracted from MCF-7 cells treated with SEO for 5 min, 1 hr, 5 hrs, and 24 hrs. GAPDH was used as loading control. Densitometric quantification of (b) Ku70 shows the highest induction of the protein at 24 hours, while (c) Ku80, (d) EPHX1, and (e) 14-3-3 achieve their highest induction at 1 hr of exposure. Asterisk ( * ) denotes statistical significance ≤ 0.05 when compared with the control (Student's -test). Each bar represents the mean of three independent experiments (mean ± SEM). mEH actively metabolizes potentially carcinogenic or genotoxic epoxides, such as those derived from the oxidation of polyaromatic hydrocarbons [38]. Epoxides are highly reactive compounds with an electrophilic functional group. This electrophilic group allows the epoxide to react with electronrich moieties in the DNA and produce DNA adducts or DNA strand breaks [39]. Styrene, for example, can be activated to a genotoxic intermediate in the human body. The genotoxic intermediate, an epoxide, becomes inactivated by the mEH [40]. Several alkene epoxides have genotoxic effects causing DNA damage when evaluated using the comet assay [41]. It is possible that some components of SEO might be causing the production of epoxides that also induce DNA damage. Upon a genotoxic insult, the cell needs to stop its replication to allow Evidence-Based Complementary and Alternative Medicine 11 the DNA repair enzymes to identify and repair the damage. The differential expression of 14-3-3 suggests that DNA repair is occurring after exposure to SEO. 14-3-3 proteins regulate cell division and play an important role in stopping cell cycle progression after the DNA damage checkpoints are activated [42]. Dirksen et al. (2006) studied protein expression in 14 human lymphoblast cell lines after induction of DSBs using bleomycin. 14-3-3 , a protein involved in cell cycle regulation, was found among the proteins that were expressed in the samples [36]. Upon induction of DNA damage, 14-3-3 binds to Cdc25 and removes it from the nucleus, halting the cell cycle [43]. Stopping cell cycle progression is crucial to prevent replication of damaged DNA and to activate the machinery needed for DNA repair.

Conclusion
The capacity of SEO to induce single-and double-strand breaks in human breast adenocarcinoma cells was confirmed by the alkaline and neutral comet assays. Although this assay does not provide information on specific DNA repair pathways involved in this process, the use of proteomics allowed us to define more precisely the possible DNA repair pathways and proteins that are being induced upon the DNA damage caused by SEO on breast cell lines.
Here, we present a possible mechanistic explanation for the genotoxic response of MCF-7 cells to SEO found in our experiments. Ku70/80 induction provides evidence that the DSB repair system becomes activated. However, the cell is not able to effectively repair the DSBs, possibly due to the amount of damage induced by SEO. We have also found evidence that some of the DSBs could be caused due to epoxide formation due to the induction of EPHX1. The cell cycle is halted for DSB repair due to the activity of the 14-3-3 family, mostly because of 14-3-3 activity. We provide evidence that, upon the genotoxic insult of SEO exposure, the cell is capable of activating several pathways to activate DNA repair. However, in the case of MCF-7 cells, this activation was not sufficient to mitigate the effects of SEO since although the proteins were induced, DSBs were still present as was revealed by the comet assay. Future studies will focus on the assessment of other genotoxicity endpoints such as chromosomal aberrations upon SEO exposure in MCF-7 cells using the micronucleus assay and investigating the status of these proteins in nontumorigenic breast epithelial cells to which SEO did not cause single-or double-strand breaks.
In conclusion, our findings on the genotoxic potential of SEO in breast cancer cells could lead to potential discoveries of molecules with specific anticancer activity that have a selective genotoxic effect to breast cancer cells. This project could be the first step in the process of finding alternative therapies with less toxicity to noncancerous cells.