*2.12. Anti-Proliferative Effect of OOPs on Different Cancer Cell Lines*

The ATP and MTT assays utilized to assess cell viability in the experiments described above do not provide information about the OOPs' mechanism(s) of action. To distinguish if the observed reduction in cell numbers upon treatment with OOPs was due to a proliferation arrest or to a cytotoxic effect, the levels of DNA replication arrest were evaluated after treatment with each of the five OOPs (i.e., oleocanthal (**1**), oleacein (**2**), oleuropein aglycone

(**3a,b**), ligstroside aglycone (**4a,b**) and oleomissional (**6a,b,c**)). DNA replication, as a key determinant of chromosome segregation and stability in eukaryotes, is directly related to cell proliferation [91].

For this purpose, ten cancer cell lines (i.e., MDA-MB 231, SK-BR-3, MCF-7, A2058, SK-MEL-28, AGS, HepG-2, PANC-1, H1299 and Hela) originating from breast, melanoma, stomach, hepatic, pancreatic, lung and cervical tumors were analyzed in parallel with control samples (i.e., the same cells treated only with 0.2% (*v*/*v*) DMSO). They were treated for 24 h with OOPs having EC50 values ≤ 50 μM at concentrations equal to the EC50 of each one. Subsequently, the treated cells were allowed to incorporate 5-ethynyl-2- deoxyuridine (EdU) into replicating DNA according to the protocol described in detail in the Experimental Section. EdU, a thymidine analog, can be incorporated into DNA in vivo and detected later by using copper-catalyzed azide–alkyne cycloaddition (click reaction) without prior DNA denaturation [91]. The pool of cells in the S phase can be then easily detected by fluorescence (FL) microscopy or by flow cytometry by analyzing the incorporation of EdU in replicating DNA of single cells [92]. In this study, after EdU incorporation, the nuclei of the entire cell population were stained with Hoechst and imaged by confocal microscopy (Experimental Section). EdU-positive nuclei, marked by green FL, as well as the total number of nuclei, marked by blue FL, were enumerated by applying the Icy image analysis algorithm on the digital images acquired by confocal microscopy. Treatment of the cell lines analyzed for 24 h with OOPs as single compounds (i.e., oleocanthal (**1**), oleacein (**2**), oleuropein aglycone (**3a,b**), ligstroside aglycone (**4a,b**) or oleomissional (**6a,b,c**)) resulted in inhibition of proliferation ranging from 4.7–47.8% (Figure 7 and Figure S5, Tables 2 and 3). Oleacein (**2**), oleuropein aglycone (**3a,b**) and oleomissional (**6a,b,c**) appeared as the most effective OOPs in the SK-MEL-28 melanoma cells (20.3–27.6% inhibition of proliferation), with (**3a,b**) and (**6a,b,c**) demonstrating very significant (*p* < 0.0001) inhibition of cell proliferation (Tables 2 and 3). Oleocanthal (**1**), oleacein (**2**) and the two aglycones (**3a,b** and **4a,b**) were the strongest proliferation inhibitors in the AGS stomach cancer cells (28.9–31.6% inhibition of proliferation), giving statistically significant results (*p* < 0.01) as compared to control samples (Tables 2 and 3). The strongest antiproliferative effect (47.0–47.8% inhibition) was observed in the treatment of H1299 lung cancer cells with oleocanthal (**1**) and oleacein (**2**), while in the PANC-1 pancreatic cancer cells, oleacein (**2**), oleuropein aglycone (**3a,b**), ligstroside aglycone (**4a,b**) and oleomissional (**6a,b,c**) had a similar effect (21.9–23.7% inhibition of DNA replication) (Tables 2 and 3).

**Table 2.** OOPs' antiproliferative effect. S-phase cells (i.e., % EdU +ve) are presented for each OOP treatment in each cell line. The effect of the most effective OOPs (EC50 ≤ 50 μM) on cell proliferation was evaluated on all the breast cancer (i.e., MDA-MB 231, SK-BR-3 and MCF-7) and skin melanoma (SK-MEL-28, A2058) cell lines. Moreover, the antiproliferative effect of OOPs was tested on the most resistant cell line of the other tissue origins (i.e., AGS, HepG-2, PANC-1, H1299 and Hela cells). For this, the Cell proliferation kit III (EdU-488; FM) was used after 24 h treatments with the OOPs. The doubling times for each cell line are presented as well. The results are means ± SE from two (or three) independent experiments (total no. of cells ≥ 300). \* *p* < 0.05, \*\* *p* ≤ 0.01, \*\*\* *p* ≤ 0.001, \*\*\*\* *p* ≤ 0.0001 (*t*-test) compared to the corresponding control samples (i.e., cells treated with 0.2% (*v*/*v*) DMSO).



**Table 2.** *Cont.*

<sup>a</sup> Results from three independent experiments. The rest of values are from two independent experiments.

**Figure 7.** OOPs' effect on cell proliferation. (**A**) Confocal microscopy images of SK-MEL-28 cells evaluated for proliferation 24 h after treatment with oleuropein aglycone (**3a,b**) and oleomissional (**6a,b,c**) using concentrations equal to EC50 values (μM) for each cell line. S-phase cell nuclei were stained with EdU (green fluorescence) and all nuclei with Hoechst (Blue fluorescence, grey pseudocolor). Bar graphs represent the % of cells in S Phase (proliferating) calculated by the quantification of the EdU-positive cells divided by the number of Hoechst-positive cells. (**B**) Cell proliferation was determined after 24 h treatment with the EC50 values (μM) of oleocanthal (**1**), oleacein (**2**), oleuropein aglycone (**3a,b**), ligstroside aglycone (**4a,b**) and/or oleomissional (**6a,b,c**) on MDA-MB 231, SK-BR-3, SK-MEL-28, AGS, H1299 and HepG-2. Graphs represent the quantification of EdU incorporation by counting the number of EdU +ve cells/Hoechst +ve cells. The results are means ± SE from two or three independent experiments (total no. of cells ≥ 300). \* *p* < 0.05; \*\* *p* ≤ 0.01; \*\*\* *p* ≤ 0.001; \*\*\*\* *p* ≤ 0.0001 (*t*-test) compared to the corresponding control samples (i.e., cells treated with 0.2% *v*/*v* DMSO).


**Table 3.** OOPs' antiproliferative effect. The data from Table 2 expressed as levels of % inhibition of cell proliferation after normalization with the control. The doubling times for each cell line are presented as well. The results are means ± SE from two (or three) independent experiments (total no. of cells ≥ 300).

<sup>a</sup> Results from three independent experiments. The rest of values are from two independent experiments.

In summary, all OOPs analyzed in this study with EC50 ≤ 50 μM exert antiproliferative effect already detectable at 24 h treatment in all cell lines tested. Interestingly, each OOP caused different levels of cessation in DNA replication in each cell line.

Antiproliferative effect in vitro has mainly been reported for oleocanthal (**1**) and less for oleacein (**2**) and oleuropein aglycone (**3a,b**). Oleocanthal (**1**) was shown to suppress breast cancer cell proliferation detected by G0/G1 cell cycle arrest via inhibition of HGFinduced phosphorylation of c-Met and by modulating Ca2+ entry through TRPC6 [20,28]. Moreover, oleocanthal (**1**) was described to act as a dual inhibitor of c-MET and COX-2 on lung cancer cells [31]. As for melanoma and hepatocellular carcinoma cells it has been reported that oleocanthal (**1**) suppressed cell growth by inhibiting the phosphorylation of STAT3 (signal transducer and activator of transcription 3) [22,29]. On the other hand, oleacein (**2**) treatment induced G1/S phase arrest and downregulated the expression of pro-proliferative proteins (i.e., c-KIT, K-RAS, PIK3R3, mTOR) [26]. Moreover, oleacein (**2**) was found to suppress the proliferation of neuroblastoma cells by blocking the cell cycle in the S phase [19]. With respect to breast cancer cell lines, oleuropein aglycone (**3a,b**) induced cell cycle arrest in the G0/G1 phase and reduction of cells in the S phase as well as a significant down-regulation of cyclin D1 and cyclin E expression [27]. No reports were retrieved on the mechanisms by which ligstroside aglycone (**4a,b**) or oleomissional (**6a,b,c**) exert antiproliferative action.

### *2.13. Pro-Apoptotic Activity of OOPs on Different Cancer Cell Lines*

Oleocanthal (**1**) has been shown to cause apoptosis in several cancer cell lines [21,28,29,38,48,93]. Moreover, a few studies have also reported the pro-apoptotic effect of oleacein (**2**), ligstroside aglycone (**4a,b**) and oleuropein aglycone (**3a,b**) [19,27,53].

To discern whether cell death triggered by OOPs treatment occurred via apoptosis or necrosis under the experimental conditions applied in this study, live cells treated with OOPs were stained simultaneously with FITC-conjugated annexin V and propidium iodide (PI). Annexin V binds to phosphatidylserine (PS) translocating from the inner to the outer leaflet of the plasma membrane in apoptotic cells. Therefore, annexin V-FITC binding to PS labels live apoptotic cells with green FL. PI stains DNA only in late apoptotic or necrotic cells since it does not permeate the intact membrane of live cells [94]. PS exposure to the extracellular space and cell membrane permeability were analyzed by flow cytometry as described in detail in the Experimental Section. Single OOPs with EC50s ≤ 50 μM were used to treat cells from nine cancer cell lines for 48 h (i.e., SK-BR-3, MDA-MB 231, MCF-7, SK-MEL-28, A2058, AGS, HT-29, PANC-1 and H1299) originating from breast, melanoma, stomach, pancreatic and lung tumors. OOP concentrations used were equal to the EC50 values for each cell line (Table 1). annexin V and PI staining discriminated between early- (i.e., annexin V +ve and PI −ve) and late-apoptotic cells (i.e., annexin V +ve and PI +ve), as well as between necrotic (i.e., annexin V −ve and PI +ve) and live (i.e., annexin V −ve and PI −ve) cells (Table 4) [95].

**Table 4.** OOPs' apoptotic effect on a panel of cancer cell lines. Flow cytometry analysis of apoptotic cells after 24 and 48 h treatment with specific OOPs (EC50 values). Live cells were labelled with annexin V-FITC and PI as described in the Experimental Section. Control samples (i.e., treated only with 0.2% (*v*/*v*) DMSO) were analyzed in parallel. The percentage cell population of annexin V +ve cells (early apoptotic) and that of both annexin V and PI +ve cells (late apoptotic) over the whole cell population were determined using the FlowJo software. The results are the means% ± SE from two or three independent experiments. L.A. = late apoptotic.



**Table 4.** *Cont.*

<sup>a</sup> Results from one experiment.

As shown in Figure 8A for the breast cancer cell lines, the apoptotic events increased as a result of treatment with OOPs, a result reflected in the decrease in live cell numbers (Table 4). In more detail, treatment of the breast cancer cell lines SK-BR-3, MDA-MB-231 and MCF-7 with oleuropein aglycone (**3a,b**) for 48 h resulted in similar levels of apoptotic events in all three cell lines (Table 4). Oleocanthal (**1**) was most effective in the SK-BR-3 and MCF-7 cells and triggered a similar percentage of apoptotic cells as oleuropein aglycone, while oleacein (**2**) had a low pro-apoptotic effect on SK-BR-3 and MDA-MB 231 cells (Table 4). In the melanoma cells, oleuropein aglycone (**3a,b**) was highlighted to induce higher levels of apoptosis than oleocanthal (**1**) at 48 h treatments. Most interestingly, oleomissional (**6a,b,c**) exerted a significant pro-apoptotic effect only in the SK-MEL-28 cells (*p* < 0.01, Table 4).

Among all the cancer cell lines studied, the highest amount of apoptotic cells was observed in the stomach cancer AGS, with the strongest pro-apoptotic effect (34.2% ± 2.7 cells of the total population) induced by oleocanthal (**1**) treatment (Figure 8A,B). The two aglycones (i.e., (**3a,b**) and (**4a,b**)) also caused the highest numbers of apoptotic events in AGS as compared to the other cell lines tested (Table 4, Figure 8A). Moreover, both aglycones had the most prominent pro-apoptotic action in the PANC-1 pancreatic cancer cells compared to the rest of the OOPs (Table 4). Observing the effect of oleocanthal (**1**) in the colon-originated HT-29 and the lung H1299 cancer cells (Table 4), it appeared that around 12% of the total cell population were apoptotic in both, including similar levels of early and late apoptotic events. Comparing the action of oleuropein aglycone (**3a,b**) between the two lines, a stronger apoptotic effect was observed in the H1299 cells than the effect of oleocanthal (**1**) and oleacein (**2**) (Table 4).

In summary, treatment with all five OOPs induced apoptotic events in all the cancer cells analyzed after 48 h treatment at concentrations equal to the EC50 of each compound for each cell line. Only the cases where OOPs had EC50s ≤ 50 μM were selected for this analysis. Not all OOPs were analyzed in each cell line and not all OOPs caused similar levels of apoptosis in the same cell line within the time window of analysis. The results presented in Table 4 confirmed the data already reported for the pro-apoptotic activity of oleocanthal (**1**), oleacein (**2**) and oleuropein aglycone (**3a,b**) but they additionally highlighted for the first time the significant pro-apoptotic activity of oleomissional (**6a,b,c**) in the SK-MEL-28 melanoma cells. Moreover, oleuropein aglycone (**3a,b**) appeared to have the strongest

pro-apoptotic effect in all cell lines tested in the time window within which the apoptotic events were analyzed. The pro-apoptotic effect was more pronounced in the AGS cells, and it was also confirmed by morphological alterations characteristic of apoptotic cells observed by BF brightfield microscopy (Figure S6). Cell changes at early apoptosis include membrane blebbing, cell shrinkage and pyknosis while necrotic cells appear as round or oval masses with nuclear fragmentation and chromatin condensation [94].

**Figure 8.** Treatment with OOPs generate apoptotic events in AGS stomach cancer cells. AGS cells were either left untreated (control) or treated with oleocanthal (**1**), oleuropein aglycone (**3a,b**), oleacein

(**2**) and ligstroside aglycone (**4a,b**) at concentrations of EC50 values for 48 h. At the end of the treatments, cells were stained with annexin V-FITC and propidium iodide (*Experimental Section*) and were analyzed by FACS. (**A**) The results are presented in bar diagrams as mean values of % annexin V-negative (−ve) for the viable cells and positive (+ve) for the apoptotic cells ± SE The results are from two or three independent experiments performed in duplicate. Differences compared to untreated cells were considered significant at *p* < 0.01 (\*\*) and *p* < 0.001 (\*\*\*). (**B**) Representative flow cytometry dot plots demonstrating the % of necrotic (upper left), late apoptotic (upper right), early apoptotic (lower right) or viable (lower left) cell populations in the respective quadrants.

It is worth noting that the scope of this study was to systematically investigate if the treatment of the cancer cell lines with the five OOPs under the experimental conditions used in this study could result in detectable apoptotic events. The mechanisms by which the different OOPs analyzed herein cause apoptosis were not the focus of this work.

Most reports on the pro-apoptotic effect of OOPs have mainly focused on oleocanthal (**1**) and breast cancer cell lines and less on cancer cell lines originating from other tissues. Several mechanisms have been proposed as the cause of apoptosis in the studied cell lines which in some cases coincide but in others differ [21,22,28,38,48,75]. Interestingly, in two reports by Legendre et al. and Goren et al., oleocanthal (**1**) induced lysosomal membrane permeabilization (LMP), thus compromising lysosomal integrity in a variety of cancer cells lines [96,97]. These results suggested that the apoptotic and necrotic events detected in oleocanthal-treated (**1**) cells are downstream of the observed LMP and depend on the corresponding degree of LMP [97]. The pro-apoptotic effects of oleacein (**2**) and oleuropein aglycone (**3a,b**) have been much less studied. Both OOPs were reported to trigger apoptosis by altering signaling events exerted by members of the BCL-2 protein family due to the upregulation of pro-apoptotic factors (i.e., BAX protein) and down-regulation of anti-apoptotic ones (i.e., BCL2 and MCL1) [19,26,27].

#### **3. Materials and Methods**
