Evaluation of the Cytotoxicity of Structurally Correlated p-Menthane Derivatives

Compounds isolated from essential oils play an important role in the prevention and treatment of cancer. Monoterpenes are natural products, and the principal constituents of many essential oils. The aim of this study was to investigate the cytotoxic potential of p-menthane derivatives. Additionally, analogues of perillyl alcohol, a monoterpene with known anticancer activity, were evaluated to identify the molecular characteristics which contribute to their cytotoxicity, which was tested against OVCAR-8, HCT-116, and SF-295 human tumor cell lines, using the MTT assay. The results of this study showed that (−)-perillaldehyde 8,9-epoxide exhibited the highest percentage inhibition of cell proliferation (GI = 96.32%–99.89%). Perillyl alcohol exhibited high cytotoxic activity (90.92%–95.82%), while (+)-limonene 1,2-epoxide (GI = 58.48%–93.10%), (−)-perillaldehyde (GI = 59.28%–83.03%), and (−)-8-hydroxycarvotanacetone (GI = 61.59%–94.01%) showed intermediate activity. All of the compounds tested were less cytotoxic than perillyl alcohol, except (−)-perillaldehyde 8,9-epoxide (IC50 = 1.75–1.03 µL/mg). In general, replacement of C-C double bonds by epoxide groups in addition to the aldehyde group increases cytotoxicity. Furthermore, stereochemistry seems to play an important role in cytotoxicity. We have demonstrated the cytotoxic influence of chemical substituents on the p-menthane structure, and analogues of perillyl alcohol.


Introduction
Cancer is a disease in which certain cells in the body grow in an uncontrolled way. It is one of the world's most serious illnesses, and together with heart attacks, it kills more people than any other disease. Its hallmarks constitute an organizing principle for rationalizing the complexities of neoplastic diseases. They include sustained proliferative signaling, evading growth suppressors, resisting cell death, enabling replicative immortality, inducing angiogenesis, and activating invasion and metastasis [1].
Nature is an important source of new candidates for therapeutic compounds; a large chemical diversity is found in several species of plants, animals, and microorganisms [2]. Prakash and collaborators [3] have reported that natural products, since ancient times, have been used for health purposes by all cultures, as well as being a source of medicines. It has been estimated that about 80%-85% of the global population relies on traditional medicines for their primary health care needs, and it is assumed that a major part of traditional therapy involves the use of plant extracts or their active principles [4][5][6]. Although many recent investigations and advancements in the treatment and control of cancer progression have been carried out, significant work and room for improvement remain. The main disadvantage of synthetic drugs is their associated side effects. However, natural therapies, that use plants or plant-derived natural products are being found beneficial to combat cancer [3]. The search for antitumor agents from plant sources started in the 1950s with the discovery and development of the vinca alkaloids (vinblastin and vincristine), and the isolation of other cytotoxic compounds [7].
Some antitumor essential oils have a high percentage of monoterpenes in their chemical compositions. These substances often contribute to the pharmacological activity of these essential oils [16]. Monoterpenes are found in the essential oils of many plants including fruits, vegetables, and herbs. They have been shown to have a large number of diverse cellular and molecular effects both in vitro and in vivo, in addition to preventing the process of carcinogenesis at both the initiation and the promotion/progression stages [16,17]. They are effective in treating early and advanced cancers. Because of results like these, about 74% of drugs in the anticancer area are either natural products or derived from natural products [18].
Compounds such as thymol, perillic acid, and perillyl alcohol, among others, have been shown to prevent mammary, liver, lung, and other cancers [16,19]. Perillyl alcohol, a naturally occurring monoterpene found in lavender, cherries, and mint, has been suggested as an effective agent against a variety of tumors [20,21]. Perillyl alcohol has shown antitumor activity against pancreatic carcinomas [22,23], anti-metastatic activity in a chorioallantoic membrane model [24], and it inhibits the proliferation of human adenocarcinoma (A549), squamous cell carcinoma (H520) [21], and in vitro cultured A549 and BroTo cells [25]. Among the monoterpenes perillyl alcohol is thus one of the most promising anticancer agents. As a chemotherapeutic agent it has advanced to phase II clinical trials in cancer patients, and the preliminary results indicate that it is well tolerated [26,27].
According to Sobral and collaborators [16], many studies evaluate the cytotoxic activity of essential oils, and the isolation of their main constituents, but do not correlate their chemical structures with biological activity. Therefore considering the anticancer bioactivity of perillyl alcohol, this study aims to evaluate the cytotoxic activity of 18 p-menthane derivatives structurally correlated with perillyl alcohol, and to investigate their structure-activity relationships against human tumor cell lines.

Antiproliferative Effect of Compounds
Based on the excellent cytotoxic activity of perillyl alcohol and several monoterpenes found in essential oils, we chose 18 structurally correlated compounds ( Figure 1) to evaluate their cytotoxic activity against tumor cell lines and establish the corresponding SAR.
The cytotoxic activity of the eighteen monoterpenes was evaluated against several human tumor cell lines: HCT-116 (colon), OVCAR-8 (ovarian), and SF-295 (brain). In the screening program for the discovery and development of potential anticancer of compounds, the criteria of the American National Cancer Institute were adopted for the selection of cells [2,28,29]. The results were assessed by comparing the cell growth inhibition percentage (GI%) values, expressed in percentage. (GI%) values are presented as the mean ± SD of three replicates measured by MTT assay after 72 h of incubation, as summarized in Table 1. The results of the cytotoxicity assays revealed that all of the test compounds exhibit cytotoxic activity against these tumor cell lines, as shown in Table 1. The results were evaluated using an intensity scale for each tested cell line as follows: samples with weak cytotoxic activity (GI < 50%), intermediate activity (GI 51%-75%), and high activity (GI > 75%) [30]. The compounds 1, 2, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 17 and 18 demonstrated GI < 50% for the three cell lines studied, being classified as compounds with weak cytotoxic activity.
Compound 16 exhibited high cytotoxicity when compared with the other monoterpenes, with a GI value ranging from 96.32%-99.89%, inducing almost 100% mortality in the cells, at a concentration of 25 µg/mL. A comparison of perillyl alcohol and its analogues showed that all tested analogues ( Figure 1) were found to have less potent cytotoxic activities than perillyl alcohol itself (Table 1) Table 1. In in vitro cell culture systems, cytotoxic compounds interfere with cellular attachment, resulting in significant alteration in morphology, adversely affecting cell growth rate, or causing cell death [31]. Therefore, the cell proliferation assays were performed with structurally correlated compounds. This experimental model is a well-characterized colorimetric assay that is based on the enzymatic reduction of the tetrazolium salt MTT in living, metabolically active cells, but not in dead cells. It has been widely used to determine the cytostatic/cytotoxic potential of medicinal agents in screening programs [28,29,32]. Among the five structurally related compounds with cytotoxic activity ranging from intermediate to high, the compounds 12 [33], 14 [21], and 15 [25] exhibit cytotoxic activity and are already described in the literature. Therefore, we determined the median inhibitory concentration able to induce 50% of maximal effect (IC50) of the compounds 3 and 16, which have no cytotoxic activity described in the literature. Since compound 16 was more potent, it was subjected to additional evaluation of its cytotoxic activity against HL-60, with the intention of assessing the cell death process. The HL-60 cell is more sensitive to chemotherapeutic agents, with well-defined methods for checking cellular death processes [34][35][36]. Evaluating the IC50 of compound 16, besides presenting the highest percentage of inhibition of cell proliferation, 16 was more potent than 3, showing high cytotoxicity in all three cell lines, with IC50 values ranging from 1.03-1.75 µg/mL, as shown in Table 2.

Hemolytic Assay
Since compound 16 showed high cytotoxicity in tumor cells, it was tested for its ability to induce lysis in mouse erythrocytes (data not shown). The mechanical stability of the erythrocyte membrane is a good indicator of in vitro damage in cytotoxicity assays, since many drugs can alter this delicate structure [37]. Red blood cells provide a preliminary model to study protective effects, substance toxicity, (or conditions associated with oxidative stress), where they are a possible indicator of such damage [38][39][40][41]. However, compound 16 was not hemolytic even at the highest tested concentration (500 μg/mL). This suggests that the cytotoxicity mechanism is probably related to a more specific pathway.

(−)-Perillaldehyde 8,9-epoxide Inhibits the Proliferation of Human Leukemia in HL-60 Cells
Three concentrations of compound 16, ½ IC50 (0.32 µg/mL), IC50 (0.64 µg/mL) and 2 × IC50 (1.28 µg/mL) were chosen against HL-60. The anti-proliferative effects of 16 were demonstrated through the trypan blue dye exclusion test ( Figure 2). It showed a reduction in the number of viable cells, and increases in the number of non-viable cells in a concentration-dependent manner after 24 h of incubation. The quantitative decrease in viable cells is common in cytotoxic compounds, as demonstrated by de Barros and collaborators [42], with a thiazacridine derivative, which decreases the number of viable HCT-8 cells in a concentration and time dependent manner.

(−)-Perillaldehyde 8,9-epoxide Induces Apoptosis and Necrosis in Human Leukemia HL-60 Cells
Attempting to ascertain the cellular death process (as induced by 16), in cancer cells, two tests were performed, fluorescence microscopy using acridine orange/ethidium, and hematoxylin-eosin coloration analyzed by optical microscopy. After 24-h incubation, the effects of 16 were evaluated based on cell morphology using hematoxylin-eosin. All concentrations and showed severe drug-mediated changes. The hematoxylin-eosin stained HL-60 cells treated with 16 presented a morphology consistent with apoptosis, including a reduction in cell volume, chromatin condensation, and nuclei fragmentation, as well as, necrosis with membrane disruption, cell swelling and rupture leading to inflammation. The acridine orange/ethidium bromide stained and treated cells also displayed a morphology consistent with apoptosis and necrosis, in a concentration-dependent manner (p < 0.05) (Figure 3). Compound 16 caused cell death by apoptosis in the three concentrations tested, as shown in Figures 3  and 4. The induction of apoptosis is an important target in cancer therapy, achieved by compounds such as the vinca alkaloids, taxanes and colcichine [43]. It is the most well-known mechanism of cell death which functions to maintain homeostasis of the cells [44]. It is characterized by phosphatidylserine exposure, loss of mitochondrial membrane potential, caspase activation, chromatin condensation, nuclear fragmentation, and results in phagocytosis of membrane bound apoptotic bodies [45].
Besides the apoptotic process in doses of 0.64 and 1.28, compound 16 presented necrotic cell death simultaneously. This type of cell death is characterized by swelling of cellular organelles, cell membrane rupture, cell lysis (unlike apoptosis), and the core becomes distended yet substantially intact. Necrotic death is typically followed by inflammatory reactions [46]. This appears to be a limiting factor as the concentration of 16 is increased. Various chemical agents such as β-lapachone described by Li and collaborators [47], induce cellular death by necrosis concomitant with apoptosis in a variety of cancer cells such as ovarian, colon, lung, prostate and breast cancer. Furthermore, some forms of treatment such as alkylating agents of DNA and photodynamic treatment are also able to induce cell death, activating necrosis [48]. Thus we might consider 16 as a possible candidate for in vivo tests, even with the two processes of cell death responsible for its in vitro cytotoxicity.

Structure-Activity Relationships (SAR)
This evaluation is important due to the fact that perillyl alcohol and its analogues are structurally similar to many of the chemical constituents of essential oils and other natural products. Therefore the results obtained in this study may well provide a reference for the development of novel cytotoxic agents having an appropriate biological profile.
The monoterpenes perillyl alcohol (14) and (−)-perillaldehyde (15) differ from each other only by a functional group. Compound 14 has a hydroxyl group whereas 15 presents an aldehyde function, both located at the same position in the p-menthane skeleton (C-7). Perillyl alcohol (14) (GI = 90.92%-95.82%) exhibited higher cytotoxicity compared to (−)-perillaldehyde (15) (GI = 59.28%-83.03%), suggesting that a hydroxyl group at C-7 position produces in a more potent inhibition of cell proliferation than an aldehyde group at the same position.
Replacement of a carbon-carbon double bond, conjugating the carbonyl with an epoxide group resulted in a modest enhancement of anti-proliferative effect. Improvement of the biological potency was observed thru comparison of (−)-carvone (1)  To ascertain the influence of the disposition of the epoxide and ketone groups in cytotoxicity, we compared (−)-carvone epoxide (5) with (R)-pulegone oxide (10). Compounds 5 and 10 present a ketone and epoxide group in different positions in the p-menthane skeleton, and different biological effects for these monoterpenes was observed. (R)-Pulegone oxide (10) was more bioactive (GI = 16.02%-43.21%) than (−)-carvone epoxide (5) (GI = 8.21%-29.24%), suggesting that the position of these groups in the p-menthanic structure influence the cytotoxic activity.
To evaluate the influence of chirality on cytotoxicity, we compared the pairs of enantiomers (−)-carvone (1)
At the end of incubation, the plates were centrifuged, and the medium was replaced by fresh medium (150 μL) containing 0.5 mg/mL MTT. Three hours later, the formazan product was dissolved in 150 μL DMSO, and absorbance was measured using a multiplate reader (DTX 880 Multimode Detector, Beckman Coulter Inc., Packard, ON, Canada). The treatment effects were expressed as the percentage of control absorbance of reduced dye at 595 nm. All absorbance values were converted into a cell growth inhibition percentage (GI%) by the following formula: where C is the absorbance for the negative control, and T is the absorbance in the presence of the tested compound.
The compounds that caused greater degree of cell growth inhibition and which have not been described in the literature for cytotoxic activity against tumor cell lines were evaluated for their median inhibitory concentration able to induce 50% of maximal effect (IC50). The same protocol for the same cells was used to determine the IC50, varying only the concentration of the compound tested from 0 to 25 μg/mL, to verify the most potent compound. In addition, the IC50 was determined against HL-60 only for the most cytotoxic compound 16. Compounds with an IC50 value lower than 4 μg/mL were considered promising for the search for new anticancer agents [62].

Hemolytic Assay
The test was performed in 96-well plates using a 2% mouse erythrocyte suspension in 0.85% NaCl containing 10 mm CaCl2, following the method as described by Jimenez and collaborators [63]. Various concentrations of compound 16 (0-500 µg/mL) were added to the suspension of red blood cells obtained from mice according to methodology adapted from Kang and collaborators [64], Pita and collaborators [65], and Bezerra and collaborators [66]. The tubes with the compound 16 erythrocyte mixtures were incubated on a mixer for 60 min and then centrifuged at 3000 rpm for 5 min. Mixtures were incubated on a mixer for 60 min and then centrifuged at 3000 rpm for 5 min. After incubation at room temperature for 30 min and centrifugation, the supernatant was removed and the liberated hemoglobin was measured spectrophotometrically as the absorbance at 540 nm.

Cell viability-Trypan Blue Dye Exclusion Test
The cell viability was determined by the trypan blue dye exclusion test. HL-60 cells were seeded (0.3 × 10 6 cells/mL) in absence or presence of different concentrations of compound 16 (0.32, 0.64, and 1.28 µg/mL). After the incubation period of 24 h, 90 µL were removed from the cell suspension and added to 10 µL of trypan blue. Viable and non-viable cells were differentiated and counted in a Neubauer chamber [67].

Morphological Analyses Using a Fluorescence Microscope
HL-60 cells were pelleted and re-suspended in 20 µL of PBS. Then, 1 µL of aqueous acridine orange/ethidium bromide solution (AO/EB, 100 µg/mL) was added, and the cells were observed under a fluorescence microscope. Three hundred cells were counted per sample and classified as viable, apoptotic or necrotic [68].

Morphological Analysis with Hematoxylin-Eosin Staining
For morphological cell analysis, we used hematoxylin-eosin and examined using light microscopy. HL-60 cells were seeded (0.3 × 10 6 cells/mL) in the absence or presence of different concentrations of compound 16 (0.32, 0.64 and 1.28 µg/mL). After an incubation period of 24 h, 50 µL of cell suspension were transferred to cytospin slides, fixed with metanol for 60 s, and stained with hematoxylin-eosin [43].

Statistical Analysis
Data are presented as mean ± SEM (or SD) or IC50 values, and their 95% confidence intervals (CI 95%) were obtained by nonlinear regression. Differences among the experimental groups were compared by one-way variance analysis (ANOVA), followed by Newman-Keuls test (p < 0.05). All analyses were carried out using the Graphpad program (Intuitive Software for Science, San Diego, CA, USA).

Conclusions
Based on the data of this study, we can conclude that among the 18 derivative compounds of perillyl alcohol tested, compound 16 was the p-menthane derivative with the highest cytotoxic activity against the cancer cells lines tested. The process induced apoptotic and necrotic cell death. In addition the analysis of structure-activity relationships demonstrated that the greater anti-proliferative activity of 16 was determined by chemical aspects, such as presence of functional groups and their positions on p-menthane skeleton, suggesting that using appropriate structural modifications in these compounds, it may be possible to develop new cytotoxic agents. Therefore, compound 16 is a potential anticancer drug, yet requiring further tests to determine its in vivo antitumor activity.