All components for cell culture were obtained from Invitrogen (Thermo Fisher Scientific, Inc., Waltham, MA, USA). 2′7′-Dichlorodihydrofluorescein diacetate (H₂DCFDA) and propidium iodide (PI) were obtained from Molecular Probes (Thermo Fisher Scientific, Inc., Eugene, OR, USA). The protein kinase inhibitors SB203580, PD98059 and SP600125, and the rabbit polyclonal antibodies (pAbs) phospho-p38-MAPK (Thr180/Tyr182) (catalog #9211), phospho-Akt (Ser473) (catalog #4060), and phospho-p44/42 MAPK (Erk1/2) (Thr202/Tyr204) (catalog #4370), were obtained from Cell Signaling Technology (Danvers, MA, USA). Peroxidase-conjugated immunoglobulin G antibodies were obtained from Dako Diagnostics, S.A. (Barcelona, Spain). N-acetyl cysteine (NAC), GA and PTX were obtained from Sigma-Aldrich Quimica SL (Madrid, Spain). GA and PTX were dissolved in deuterated dimethylsulfoxide (DMSO-D6) at 20 mM and the final concentration was obtained by successive dilutions with culture medium.

The human ovarian carcinoma A2780 cells (drug-sensitive) and A2780AD cells (multi-drug-resistant ovarian cancer) included in this study were a generous gift from Dr P. Giannakakou, Weill Cornell Medical College (New York, NY, USA). Cell lines were maintained in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, 40 µg/ml gentamicin and penicillin-streptomycin (100 U/ml penicillin and 100 µg/ml streptomycin) at 37°C in a humidified atmosphere of 5% CO₂.

To determine the cytotoxic effect of the compounds on cell lines A2780 and A2780AD, 10,000 cells/well were seeded in a 96-well cell culture plate, and treated for 48 h with the desired concentration of GA and PTX, alone or in combination. At the end of treatment, 20 µl MTT (2.5 mg/ml) was added to each well and the plate was incubated for 4 h at 37°C. The reaction was then stopped by adding 100 µl of MTT solubilizer 10% SDS (sodium dodecyl sulfate) and 45% DMF (N,N-dimethylformamide) pH 5.5 (20). The plate was incubated at 37°C overnight to dissolve the formazan precipitates, and the absorbance of each well was measured at a wavelength of 595/690 nm in an Appliskan (Thermo Electro-corporation, Vantaa, Finland) plate reader. The targets used were wells without cells and the growth controls were wells with cells containing the same proportion of DMSO present in the wells with the treatments. The results were analyzed in the GraphPad Prism 5 software (GraphPad Software, Inc., La Jolla, CA, USA) and were expressed as the mean ± standard error of several independent experiments.

To measure the drug effects on the relative intracellular ROS accumulation, at the end of the treatments (24 h), the cells were washed twice with PBS and incubated for 30 min at 37°C with RPMI-1640 medium containing 5 µM H₂DCFDA, a non-specific ROS-sensitive probe. H₂DCFDA in the cell is cleaved by intracellular esterases, producing the membrane-impermeable product H₂DCF, which accumulates into the cell. H₂DCF is not a fluorescent molecule, but is oxidized by intracellular ROS to give the fluorescent product DCF (21). The resulting fluorescence was then measured in a fluorimeter FluoroMax-2 (Horiba Scientific, Minami-ku Kyoto, Japan).

The effects of the drug on cell cycle progression were determined via flow cytometry. Following treatments (24 h), the cells were washed with PBS and fixed in 70% ethanol at 4°C for at least 1 h. The cells were then washed twice with PBS, resuspended in 500 µl PBS containing 60 µl/ml DNase-free RNase A and 50 µl/ml PI, and the fluorescence was measured using a Coulter Epics XL flow cytometer (Beckman Coulter Inc., Brea, CA, USA), as previously described (22). The resulting flow cytometry histograms were analyzed with the FlowLogic 7.0 program (Iniwai, Victoria, Canada), to obtain the relative percentages of cells at the G₁, S and G₂/M phases of the cell cycle. In addition to the typical cell cycle phases, the flow cytometry histograms showed a sub-G₁ population. This suggests reduction and damage of DNA related to cell death (23).

Following treatments (24 h), the cells were collected and lysed in lysis buffer (Tris-HCl 20 mM, pH 7.5, glycerol 10%, NaCl 137 mM, NP40 1%) containing a protease inhibitor cocktail (Thermo Fisher Scientific, Inc.), and the proteins were quantified. Equal amounts of protein were dissolved in 2X Laemmli buffer with β-mercaptoethanol, heated to 95°C for 3 min and resolved via 10% SDS-PAGE. After electrophoretic separation, the proteins were transferred to a PDVF membrane using the Trans-Blot^® Turbo™ Transfer System (Bio-Rad, Hercules, CA, USA), following which the membranes were blocked with 5% milk in TBS-Tween 20 for 1 h, and then incubated overnight at 4°C with the following primary antibodies: phospho-p38-MAPK (Thr180/Tyr182), 1:100 dilution; phospho-Akt (Ser473), 1:200 dilution; phospho-p44/42 MAPK (Erk1/2) Thr202/Tyr204, dilution 1:1,000. The membrane was then washed twice in TBS-Tween, and incubated for 1 h at room temperature with the corresponding secondary antibody (anti-rabbit IgG, HRP-linked, dilution 1:2,000). The proteins were visualized using the ChemiDoc Touch Imaging System (Bio-Rad Laboratories, Inc.), using NZY Supreme ECL HRP Substrate (NZYTech, Lda., Lisbon, Portugal) as a developer. β-actin was used as the loading control.

Data were analyzed using the GraphPad Prism 5.0 statistical program (GraphPad Software, Inc.), and statistical differences between groups were evaluated using one-way analysis of variance followed by the Dunnett's test. P<0.05 was considered to indicate a statistically significant difference.

As an initial approach, we aimed to comparatively examine the effect of GA, in a concentration range of 10–200 µM, on total cell proliferation activity, as measured by the number of viable cells in sensitive (A2780) and multi-drug-resistant (A2780AD) ovarian carcinoma cell cultures. As indicated in Fig. 1A, GA caused a concentration-dependent reduction in the percentage of viable cells, an effect that was of lower intensity in the resistant cell line. The approximate reduction rates were 25% with 50 µM GA in A2780 cells, and 20% with 100 µM GA in A2780AD cells. These concentrations were selected for the combination treatment assays with the respective cell lines in the following determinations of ROS generation, cell cycle and protein kinases.

Subsequently, we comparatively analyzed the effect of PTX, at concentrations ranging from 10–400 nM, alone and in combination with the previously indicated concentrations of GA. As depicted in Fig. 1B and C, PTX caused a drug-dependent decrease in the number of viable cells, which was less pronounced in the resistant cell line. Thus, treatment with 50 nM PTX caused an ~30% decrease in the number of A2780 cells, whereas 200 nM caused only an ~20% decrease in the number of A2780AD cells. In addition, we observed that PTX and GA cooperated in more than an additive manner to inhibit cell viability, even in the resistant A2780AD cell line. For example, 50 nM PTX plus 50 µM GA caused an almost 60% decrease in the number of viable cells in the A2780 cell cultures; and the same result was obtained using 200 nM PTX plus 100 µM GA in the A2780AD cell cultures. Therefore, the concentrations of 50 and 200 nM PTX were adopted for further experiments with the A2780 and A2780AD cell lines, respectively.

As aforementioned, phenolic agents often provoke oxidative stress in cultured tumor cells, which may be a determinant of proliferation inhibition and cell death. The generation of oxidative stress by GA and PTX, evaluated alone and in combination, was determined by measuring ROS production, using an H₂DCFDA ROS-sensitive probe. The results are indicated in Fig. 2A and B. Treatment with PTX alone did not significantly affect ROS production, while treatment with GA, either alone or in combination with PTX, caused an approximate 2-fold increase. At the drug concentrations used, the results were approximately the same in the two cell lines.

In order to investigate the role of ROS overproduction in the inhibition of cell proliferation, we used the antioxidant/ROS-scavenging agent NAC. The results presented in Fig. 3 indicate that NAC markedly attenuated the decrease in viability caused by the combination of PTX plus GA in the A2780AD cell cultures, in such a manner that the decrease in viability caused by the NAC + PTX + GA triple combination was similar to that obtained with PTX alone. This seems congruent with the observation that GA causes ROS overproduction, while PTX does not. Taken together, the results in Figs. 2 and 3 indicate that ROS overproduction is an essential determinant of the capacity of GA to potentiate PTX toxicity in drug-resistant ovarian cancer cells.

PTX is a potent microtubule-targeting agent known to cause mitotic cell cycle arrest (3). After establishing the inhibitory action of PTX and GA on total cell proliferation activity, as well as the protective action of NAC, we aimed to comparatively analyze the effects of these agents on cell-cycle progression in drug-resistant A2780AD cells. The flow cytometry assay results are indicated in Fig. 4. As hypothesized, treatment with PTX increased the proportion of cells in the G₂/M phase, a response that was potentiated by co-treatment with GA. Of note, G₂/M phase accumulation was suppressed by the presence of NAC.

In addition to discerning the distribution of the typical cell cycle phases (G₁, S and G₂/M), a flow cytometry assay can also reveal a subpopulation (sub-G₁) of cells with reduced DNA content, normally interpreted as dead cells (23). In the present study, this subpopulation was negligible in cells treated with PTX and GA alone; but it reached a modest, albeit significant, value upon treatment with GA plus PTX. As in the case of G₂/M, the sub-G₁ fraction was suppressed by co-treatment with NAC. Taken together, these results suggest that the decrease in total cell proliferation in the PTX plus GA-treated cultures (Fig. 1B and C) is a consequence of both cell cycle blockade and induced cell death.

It is known that Akt and ERK normally function as defensive protein kinases, which prevent apoptotic cell death while favoring cell proliferation/cell cycle progression (24–26). In addition, MAPK p38 is also activated in response to oxidative stress (27,28), and may behave as either a positive or negative regulator of apoptosis and the cell cycle, depending on the cell line and the experimental conditions (29,30). Thus, western blot assays were carried out in A2780AD cells in order to investigate the phosphorylation/activation of these kinases in response to GA and PTX, alone and in combination, and in the absence or the presence of NAC. The results depicted in Fig. 5 indicate that these treatments did not significantly affect Akt and p38 phosphorylation. In contrast, we observed that PTX stimulated ERK phosphorylation and that GA (which increased ROS production; Fig. 2) decreased the basal phosphorylation level of ERK and prevented the aforementioned PTX-induced increase. We also observed that the inhibitory action of GA in combination therapy was suppressed by the presence of NAC. Taken together, these results suggest that ROS negatively regulated ERK activation under experimental conditions used in the present study.

Finally, the functional importance of ERK activation as a regulator of cell proliferation and cell cycle progression was examined using the MEK/ERK inhibitor PD98059 (20 µM). It was observed that PD98059, which was innocuous per se, potentiated the PTX-induced decrease in cell viability, alone and in combination with GA (Fig. 6A). Accordingly, PD98059 potentiated the G₂/M arrest caused by GA plus PTX, although it did not potentiate, and even slightly reduced, the size of the sub-G₁ cell fraction (Fig. 6B). Taken together, these results indicate that ERK activation exerts a defensive role, attenuating the excessive cytostatic action induced by PTX in A2780AD cells.