Oleuropein-Enriched Olive Leaf Extract Affects Calcium Dynamics and Impairs Viability of Malignant Mesothelioma Cells

Malignant mesothelioma is a poor prognosis cancer in urgent need of alternative therapies. Oleuropein, the major phenolic of olive tree (Olea europaea L.), is believed to have therapeutic potentials for various diseases, including tumors. We obtained an oleuropein-enriched fraction, consisting of 60% w/w oleuropein, from olive leaves, and assessed its effects on intracellular Ca2+ and cell viability in mesothelioma cells. Effects of the oleuropein-enriched fraction on Ca2+ dynamics and cell viability were studied in the REN mesothelioma cell line, using fura-2 microspectrofluorimetry and MTT assay, respectively. Fura-2-loaded cells, transiently exposed to the oleuropein-enriched fraction, showed dose-dependent transient elevations of cytosolic Ca2+ concentration ([Ca2+]i). Application of standard oleuropein and hydroxytyrosol, and of the inhibitor of low-voltage T-type Ca2+ channels NNC-55-0396, suggested that the effect is mainly due to oleuropein acting through its hydroxytyrosol moiety on T-type Ca2+ channels. The oleuropein-enriched fraction and standard oleuropein displayed a significant antiproliferative effect, as measured on REN cells by MTT cell viability assay, with IC50 of 22 μg/mL oleuropein. Data suggest that our oleuropein-enriched fraction from olive leaf extract could have pharmacological application in malignant mesothelioma anticancer therapy, possibly by targeting T-type Ca2+ channels and thereby dysregulating intracellular Ca2+ dynamics.


Introduction
Malignant mesothelioma is a poor prognosis cancer with worldwide increasing insurgence, arising from the pleura and other mesothelial tissue and showing close association with asbestos exposure. An effective therapeutic approach for the disease is currently lacking, while alternative treatments are urgently needed [1]. Among alternative remedies for cancer treatment, there is a growing interest in the anticancer action of natural substances, some of which are present in large amounts in byproducts from agrofood chains.
Olive (Olea europaea L.) is a main temperate fruit crop, mostly destined to oil production [2]. Olive oil is reputed to be an important health promoting factor in the Mediterranean diet, especially because of its alleged prevention of cardiovascular problems, metabolic syndrome, cancer, alleviation of inflammatory and autoimmune conditions, and wound healing [3]. However, olive tree leaves have a several-centurylong tradition in the folk medicines of the Mediterranean basin and are currently contemplated in the Pharmacopoeia Ph. Eur. 5 [4].
Leaves and other materials from olive tree pruning accumulate in yearly amounts as high as about 25 kg per tree [5]. This material is generally disposed by olive tree growers, representing an overhead cost of olive and oil production, while the possibility of using it as biomass for thermal energy 2 Evidence-Based Complementary and Alternative Medicine is under study [6]. Olive leaves are used in the form of herbal extracts, tea, and powder for nutritional supplement and cosmetics [4], but these products represent a minimal portion with respect to the huge amount of disposed materials.
The major leaf secondary metabolite is the secoiridoid oleuropein, a glycosylated ester of elenolic acid with hydroxytyrosol (2-(3,4-dihydroxyphenyl)ethanol). Oleuropein leaf content has been reported to vary about tenfold among different olive varieties, rating on average between 10 and 100 mg/g of dry matter [7,8]. Many other biologically active phenolics are present in olive leaves, including hydroxytyrosol itself [9].
Oleuropein is deemed to have great potential as an antioxidant and food additive, but also as a possible therapeutic tool. A wide range of studies on oleuropein have been carried out using in vitro assays, animal models of disease, or human volunteers, in order to explore possible beneficial effects for human health [10]. The reported findings mainly include antioxidant and anti-inflammatory effects, hepatoprotection, neuroprotection, hypoglycemic and hypolipidemic activities, and cardiovascular protection [11,12]. Antiproliferative activities on cancer cell lines, antitumor effects in animals, and antimicrobial and antiviral effects have also been shown [13].
Hydroxytyrosol in olive leaves, fruits, and extracts is believed to derive from oleuropein hydrolysis, both by specific enzymes and as an extraction artifact [9]. In vitro and in vivo studies suggest that this compound shares remarkable antioxidant and anti-inflammatory power with oleuropein, allegedly related to antiatherogenic, antithrombotic, cardioprotective, and anticancer effects [14,15].
The value of olive leaf extract for possible use in health products and medical food has been previously assessed in terms of antioxidant and antimicrobial power [16]. However, the profitable reutilization of agrofood chain byproducts needs to overcome several problems, including the costs of transport and processing and the seasonal availability of these materials. In this study, we have developed a rapid method for obtaining olive leaf extract enriched in oleuropein, involving raw extraction, extract partitioning, and chromatographic separation. The material has been tested in pharmacologically relevant, intracellular Ca 2+ mobilization assay and antiproliferative assay on mesothelioma cells. In these experiments, standard oleuropein has been used as control, while hydroxytyrosol has been also used in Ca 2+ assays to provide insight into the Ca 2+ mobilizing properties of the oleuropein molecule.

Plant Material.
Branch specimens of Olea europaea L. cultivar "Taggiasca" [17] were obtained from olive groves located at Imperia, Italy, owned by the Pietro Isnardi S.r.l. food company. The material was taxonomically identified by one of us (LC), and a voucher specimen was deposited at DISTAV, University of Genova (cat.: GE-Olea europaea cultivar "Taggiasca"). Leaves were separated from twigs, airdried for one week, minced to <1 mm using a grinder, and then subjected to extraction procedure.
For LC/MS analysis, standard stock solutions of oleuropein and hydroxytyrosol (each at concentration of 100.0 mg L −1 ) were prepared in methanol and diluted as required with a buffer solution of ammonium acetate 10.0 mM brought to pH 4.0 for acetic acid. For in vitro bioassays, stock solutions of oleuropein and hydroxytyrosol (each at 100 mM) were prepared in DMSO and diluted as indicated with cell medium or loading buffer for Ca 2+ measurements. Stock solutions were preserved at −20 ∘ C in dark conditions and were stable for three months.

Olive Leaf Extraction
Procedure. The minced leaves were extracted with a mixture of 2-propanol : water of 9 : 1. Aliquots of 10 g of leaves were extracted twice with 200 mL of solvent each time. The extraction was carried out at RT for about 1.5 h. The resulting solutions were mixed and the solvent was eliminated under vacuum, yielding an amount of 2.35 g of dried material. This raw extract was then partitioned between a mixture of methanol : water of 3 : 1 and a mixture of toluene : petroleum ether of 2 : 1 (100 mL each phase). The hydrocarbon phase became deep green, indicating that it contained all the chlorophyll of the raw extract. The aqueous methanol phase was dried under vacuum, yielding 1.8 g of subextract. This material was then again partitioned between water and 2-butanone (100 mL each phase), to remove sugars and other water-soluble compounds. The organic phase was dried out, obtaining a partitioned subextract that amounted to 1.03 g (43% recovery from the first raw extract, 10% from the plant material). A control TLC showed that in the former partition no oleuropein passed in the hydrocarbon phase, while in the latter only a tiny amount of it was lost in the water phase.

Chromatographic Separation.
Liquid chromatography was performed by means of an MPLC system, consisting of an Alltech 426 HPLC pump equipped with a VWR LaPrep 3101 detector. Glass columns from Omnifit were used, homepacked with Fluka silica gel 100 RP-18 (15-35 m) stationary phase. TLC was performed with Merck F 254 glass plates (RP-18); the spots were detected using a UV lamp (254 and 366 nm) and additionally by spraying with sulfovanillin and heating at 100 ∘ C, obtaining an oleuropein coral pink spot on TLC.
The above subextract was loaded on a RP-18 preparative column and eluted with 80% water and 20% methanol. Fractioning was monitored through the UV detector ( = 250 nm) and with TLC. Finally, an oleuropein-enriched fraction was collected and subjected to further study.
Evidence-Based Complementary and Alternative Medicine 3 2.5. LC/MS Analysis 2.5.1. Apparatus. LC/MS analysis was performed on the above raw extract and subfractions by Nexera Liquid Chromatography Shimadzu (Kyoto, Japan) system equipped with a DGU-20A3R Degasser, two LC-30AD Pumps, a SIL-30AC Autosampler, a CTO-20AC column compartment, and a CMB-20A Lite system controller. The system was interfaced with a 3200 QTrap LC-MS/MS system (AB Sciex, Concord, Canada) by a Turbo V interface equipped with an ESI probe. The 3200 QTrap data were processed by Analyst 1.5.2 software (Toronto, Canada). The turbo ion spray (TIS) ionization was obtained using the Turbo V interface working in negative ion mode. The instrumental parameters were set as follows: curtain gas (N 2 ) at 30 psig, nebulizer gas GS1 and GS2 at 50 and 40 psig, respectively, desolvation temperature (TEM) at 550 ∘ C, collision activated dissociation gas (CAD) at 6 units of the arbitrary scale of the instrument, and ion spray voltage (IS) at −4500 V.

UHPLC-MS/MS
The 3200 QTrap was used in scheduled multiple reaction monitoring (sMRM) considering the transitions of each species at a prefixed retention time. Unit mass resolution was established and maintained in each mass-resolving quadrupole by keeping a full width at half maximum (FWHM) of about 0.7 u.
The analytes were previously subjected to MS/MS characterization study, to identify the fragmentation patterns taking place under increasing collisional energy. Characterization experiments were carried out for direct infusion of 1.0 mg L −1 standard solutions of each analyte connected through a T valve to the syringe pump (syringe flow-rate 20.0 L min −1 ; chromatographic pump flow-rate 200 L min −1 ). For each species, the most intense transition was used for the quantitative analysis and referred to as "quantifier" transition, while the second intense one (the "qualifier" transition) was employed in the identification step, as a confirmation. The "quantifier" and "qualifier" transitions and the instrumental potential values for each compound are reported in Table 1.
Calibration plots of five concentration levels (between LOQ value and 1000.0 g L −1 ), relating the peak area of the "quantifier" transition signal ( ) versus standard concentration ( ), were built for all the analytes. To avoid a possible influence of the experimental error on the execution order of analyses, the standard solutions were injected in randomized order. For all the analytes a linear regression fit was used with a weighting factor 1/ and for all the calibration plots a good linearity with regression coefficients ( 2 ) always greater than 0.9982 was obtained.
Detection limit values expressed as the concentration of the analyte that gives a signal equal to the average background ( blank ) plus 3 times the standard deviation of the blank (LOD = blank + 3 blank ) are 0.3 g L −1 for hydroxytyrosol and 0.4 g L −1 for oleuropein. The quantitation limits (LOQ), evaluated as LOQ = blank + 10 blank , are 1.0 g L −1 for hydroxytyrosol and 1.2 g L −1 for oleuropein.
2.6. In Vitro Cell Culture. In vitro experiments were carried out using the tumorigenic malignant mesothelioma REN cell line [1]. Cells were grown in DMEM, supplemented with 10% foetal bovine serum (FBS, Euroclone), at 37 ∘ C, in a 5% CO 2 , fully humidified atmosphere.

Cell Viability Assay.
Cell viability was determined on REN cells by the MTT assay. Cells were settled in 96-well plates for 24 h, exposed to various agents for 48 h as specified, incubated with 100 L/mL tetrazolium salt (MTT, 5 mg/mL in PBS) in cell culture medium without serum for 3 h at 37 ∘ C, and treated with a solution of 1 N HCl-isopropanol (1 : 24, v/v) followed by mixing to dissolve the dark-blue formazan crystals formed. After a few minutes at room temperature, the plates were read at 550 nm in a VMax microplate reader (Molecular Devices, Sunnyvale, CA).

Intracellular Ca 2+
Measurements. Cytosolic free Ca 2+ concentration was measured in fura-2-loaded REN cells using a microspectrophotometry fluorescence-ratio setup equipped with a perfusion system, as previously described [18]. Cells were incubated in loading buffer containing 10 M fura-2-AM at 37 ∘ C for 30 minutes, washed with buffer, and mounted on the stage of an inverted microscope (Axiovert Zeiss, Germany), where they were continually superfused with different solutions. Cells were illuminated by a xenon lamp through a wavelength selector monochromator; emission was observed through ×40 quartz objective and recorded by a photomultiplier. The ratio 340 / 380 was calculated every 40 msec to acquire a time-dependent internal Ca 2+ sensitive signal.
At the end of each experiment, cells were incubated with 10 M ionomycin in a Ca-free solution containing 2 mM EGTA for 20−40 min until the ratio reached a minimum value ( min ); then EGTA was washed out, normal external Ca 2+ (1 mM) was restored, and the ratio readily reached a maximum value ( max ). Finally, MnCl 2 (5 mM) was added to the bath to quench the fura-2 fluorescence and determine the background (cell autofluorescence) values. The fluorescence emissions relative to each excitation wavelength ( 340 and 380 , resp.) were corrected for this background signal before ratio = 340 / 380 determination. Internal Ca 2+ was calculated according to the Grynkiewicz equation [19].
2.9. Statistics. Statistics were obtained with the package, version 3.0.1 (http://www.r-project.org/foundation/), usingtest with Bonferroni's correction for multiple comparisons.  Cytotoxicity was determined using a logistic dose-response curve as reported in Ranzato et al. [20]. Analysis of data from Ca 2+ measurements and determination of EC 50 values were done by Sigmaplot 8 (Systat Software Inc., San Jose, CA).

Results
Multistep processing of olive leaves consisted of the following: (1) extraction of dried, minced leaves in 2-propanol : water (9 : 1), to obtain a raw extract; (2) partitioning between methanol : water (3 : 1) and toluene : petroleum ether (2 : 1), followed by partitioning of the dried aqueous methanol phase between water and 2-butanone, to obtain a subextract; (3) liquid (column) chromatography separation of the dried organic phase, to obtain an oleuropein-enriched fraction (see Methods). Oleuropein and hydroxytyrosol quantification in the raw extract and fractions was achieved by LC-MS technique. Data showed a progressive enrichment in oleuropein, up to a value of 60% w/w in the final fraction obtained through chromatographic separation (Table 2). In order to evaluate the pharmacological potentials of the oleuropein-enriched, leaf extract fraction, and of its major component oleuropein, we tested the ability of mobilizing cell Ca 2+ in mesothelioma REN cells. Intracellular Ca 2+ is a main regulator of cell growth, and therefore agents able to deregulate Ca 2+ dynamics can be possibly used to hinder cell proliferation.
Fura-2-loaded cells were transiently exposed by perfusion to the oleuropein fraction. This treatment induced [Ca 2+ ] i rises followed by a prompt recovery of the baseline level upon washout (Figure 1(a)). Data indicate that the observed [Ca 2+ ] i spikes are not due to cell membrane injury but rather depend on the influx of Ca 2+ through membrane Ca 2+ channels. Moreover, the [Ca 2+ ] i transient elevations were dependent on the nominal concentration of oleuropein (10-30 M) contained in the extract (Figure 1(b)).
To explore more in depth the mechanism of action, we also studied the Ca 2+ mobilizing property of standard oleuropein. Similar to the oleuropein fraction, increasing doses of standard oleuropein, in the range 10-100 M, induced dosedependent [Ca 2+ ] i spikes (Figures 1(c) and 1(d)). Removal of Ca 2+ from the external solution almost abolished the oleuropein effect (Figure 1(d)). In addition, the use of the T-type Ca 2+ channel inhibitor NNC-55-0396 (Sigma, ≥98%, HPLC) [21] reversibly reduced the Ca 2+ spike induced by the highest dose of 100 M oleuropein (Figures 1(c) and  1(d)), suggesting at least a partial involvement of T-type Ca 2+ channels.
We tried to provide some hint as to which oleuropein moiety could be active on cell Ca 2+ mobilization. Hydroxytyrosol is known to induce modulatory effects on [Ca 2+ ] i [22]. We therefore used standard hydroxytyrosol on fura-2loaded REN cells, observing the induction of dose-dependent [Ca 2+ ] i spikes that also in this case were inhibited by Ca 2+free medium. Moreover, NNC-55-0396 also hindered the hydroxytyrosol-induced [Ca 2+ ] i spikes, with 80% inhibition at 10 M hydroxytyrosol and 65% inhibition at 100 M hydroxytyrosol (Figures 1(e) and 1(f)).
As for comparative quantification of the effect on [Ca 2+ ] i , dose-response data allowed estimating an EC 50 of 12 M for hydroxytyrosol and of 53 M for oleuropein, showing that hydroxytyrosol is more powerful than oleuropein in inducing Ca 2+ spikes. No established drug is known to induce T-type Ca 2+ channel opening. We therefore used the natural phenolic epigallocatechin-3-gallate (EGCG) as positive control, since we had previously shown that this compound induces Ca 2+ rise in REN cells via T-type Ca 2+ channels [23]. Presently, EGCG was found to induce dosedependent [Ca 2+ ] i spikes with an estimated EC 50 of 69 M (Figures 1(e) and 1(f)).
The oleuropein-enriched fraction and standard oleuropein were separately used on REN cells at nominally equivalent oleuropein concentrations, in order to verify their antiproliferative activity. EGCG was used as positive control also in this case, since we had previously assessed its cytotoxicity on REN cells [24]. The cytotoxic effect of the oleuropein fraction (IC 50 = 22 g/mL expressed as oleuropein) was slightly, but not significantly, stronger than that of standard oleuropein (IC 50 = 25 g/mL). Both substances were slightly more cytotoxic than EGCG (IC 50 = 33 g/mL) (Figure 2).

Discussion
The oleuropein yield in our enriched fraction is very close to the highest concentrations present in commercially available olive leaf extract formulations [25], suggesting its possible direct use in medicinal preparations. Despite a vast array of  data on possible oleuropein cellular targets, our data about the effects on intracellular Ca 2+ are new. The activity of oleuropein was slightly lower than that of the oleuropeinenriched fraction, possibly due to the occurrence in the fraction of other secoiridoid esters known to be present in olive leaves, such as demethyloleuropein, ligstroside, and oleuroside [26], allegedly coeluting with oleuropein in the HPLC separation. We have previously found that low-voltage-activated Ttype Ca 2+ channels, specifically the isoform Cav3.2, mediate the Ca 2+ rise induced in REN cells by the green tea polyphenol epigallocatechin-3-gallate [23]. We have also shown that Cav3.2 is responsible for Ca 2+ spikes induced by epigallocatechin-3-gallate in MCF-7 breast cancer cells [20]. In the present study, an involvement of T-type Ca 2+ channels is called into question also for the Ca 2+ mobilizing effect of oleuropein by the use of Ca 2+ -free medium, indicating a Ca 2+ entry process, and of the specific T-type Ca 2+ channel blocker NNC-55-0396.
The stronger activity of hydroxytyrosol with respect to oleuropein and the inhibitory effect of NNC-55-0396 suggest that the hydroxytyrosol moiety could be the active site mediating the oleuropein effect on T-type Ca 2+ channels. The almost complete inhibition exerted by NNC-55-0396 on hydroxytyrosol at lower dose, compared with the partial inhibition at higher dose, seems to indicate that the effect of olive phenolics mostly involves a T-type Ca 2+ channeldependent mechanism at low doses (≤10 M). At higher doses, these phenolics could instead act through a more complex mechanism, possibly including other Ca 2+ mobilizing systems.
T-type Ca 2+ channels have been linked to various kinds of epilepsy [27], while they are thought to play a role in tumor cell cycle progression and proliferation [28,29]. Hence, much interest has been raised in their possible therapeutic targeting with more or less specific inhibitors [30,31]. However, the blockage of these channels does not seem to be an exclusive strategy in antitumor therapy. Antiproliferative effects obtained with mibefradil and pimozide, two T-type Ca 2+ channel inhibitors, have been partially ascribed to some unidentified mechanism [32]. Also, cell cycle progression needs a fine tuning of [Ca 2+ ] i oscillations, suggesting that intracellular Ca 2+ dysregulation determining transient [Ca 2+ ] i rise may also interfere with tumor development, possibly leading to cell death [33]. Accordingly, pharmacological agents like olive phenolics, able to abnormally activate Ttype Ca 2+ channel activity, could be promising tools for the inhibition of cancerous cell growth. In apparent contrast with the present data, it has been previously shown that an olive leaf extract exerts antagonistic effects on high-voltage-activated L-type Ca 2+ channel, thereby inducing blood pressure lowering [34,35]. A comparison between these previous reports and our study is difficult to make, owing to differences in both the extracts and the experimental models. However, such a complex of data advocates an intriguing scenario, entailing divergent effects of major olive phenolics on low-voltage and high-voltage Ca 2+ channels.
More insight into the possibility of using the oleuropein fraction in antitumor treatment for mesothelioma is provided by our cytotoxicity tests, indicating that the antiproliferative activity of the oleuropein fraction essentially depends on its major constituent oleuropein. The United States National Cancer Institute plant screening program has established that a plant extract can be deemed to have a cytotoxic effect if IC 50 for incubations of 48-72 h is 20 g/mL or lower [36]. Our oleuropein-enriched fraction has an oleuropein content of 60% and a cytotoxicity IC 50 of 22 (17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29) g/mL in terms of oleuropein. We can then estimate IC 50 of 36.7 (28.3-48.3) g/mL for the oleuropein-enriched fraction as a whole. Hence, our oleuropein fraction is close to the cytotoxicity threshold established by NCI and could therefore be appropriate for the development of synergistic treatments allowing reducing the doses of highly toxic conventional drugs, as previously suggested for EGCG [1]. Such an approach could overcome critical drawbacks in the treatment of chemoresistant tumors like malignant mesothelioma [37].
There is no established standard dose of oleuropein for humans, but in acute toxicity studies, oleuropein and hydroxytyrosol have shown no lethality or adverse effects in mice up to doses of 1000 and 2000 mg/kg, respectively. In addition, studies conducted on humans with olive extract or its polyphenolics have shown no undesired effects [38,39]. Pharmacokinetics of oral oleuropein and its metabolite hydroxytyrosol in human plasma indicates peaks of about 0.004 g/mL (0.01 M) for oleuropein, while its main metabolite hydroxytyrosol can reach values of about 0.15 g/mL (1 M) [40]. These values are close to the effective concentrations observed in our experiments for hydroxytyrosol, but intravenous delivery of oleuropein should allow reaching much higher blood peaks. Hence, pharmacokinetics data indicate that plasma concentrations compatible with dysregulating effects on tumor cell Ca 2+ could be easily achieved upon clinical administrations of 8 Evidence-Based Complementary and Alternative Medicine oleuropein, followed by its conversion into the more effective metabolite hydroxytyrosol.

Conclusions
We have shown a method to obtain oleuropein-enriched olive leaf extract fractions that could be of industrial interest, because extraction is generally competitive with chemical synthesis for complex molecules like oleuropein. In addition, we have highlighted a novel mechanism of action for oleuropein and its metabolite hydroxytyrosol, consisting in a dysregulation of Ca 2+ dynamics occurring via T-type Ca 2+ channels. This kind of effect suggests a possible use of the oleuropein fraction from olive leaves in the treatment of mesothelioma.