Artichoke Polyphenols Sensitize Human Breast Cancer Cells to Chemotherapeutic Drugs via a ROS-Mediated Downregulation of Flap Endonuclease 1

Combined treatment of several natural polyphenols and chemotherapeutic agents is more effective comparing to the drug alone in inhibiting cancer cell growth. Polyphenolic artichoke extracts (AEs) have been shown to have anticancer properties by triggering apoptosis or reactive oxygen species- (ROS-) mediated senescence when used at high or low doses, respectively. Our aim was to explore the chemosensitizing potential of AEs in order to enhance the efficacy of conventional chemotherapy in breast cancer cells. We employed breast cancer cell lines to assess the potential synergistic effect of a combined treatment of AEs/paclitaxel (PTX) or AEs/adriamycin (ADR) and to determine the underlying mechanisms correlated to this potential therapeutic approach. Our data shows that AEs/PTX reduced cell proliferation by increasing DNA damage response (DDR) mediated by Flap endonuclease 1 (FEN1) downregulation that results into enhanced breast cancer cell sensitivity to chemotherapeutic drugs. We demonstrated that ROS/Nrf2 and p-ERK pathways are two molecular mechanisms involved in the synergistic effect of AEs plus PTX treatment. To highlight the role of ROS herein, we report that the addition of antioxidant N-acetylcysteine (NAC) significantly decreased the antiproliferative effect of the combined treatment. A combined therapy could be able to reduce the dose of chemotherapeutic drugs, minimizing toxicity and side effects. Our results suggest the use of artichoke polyphenols as ROS-mediated sensitizers of chemotherapy paving the way for innovative and promising natural compound-based therapeutic strategies in oncology.


Introduction
Breast cancer is the most common malignancy in women around the world [1] and is a heterogeneous disease with high degree of diversity between and within tumors and among individual patients [2][3][4]. Of the various factors involved in breast carcinogenesis, oestrogen receptors (ER) play a major role and are considered an important therapeutic target. ER-positive tumors are further subtyped into low proliferation rate luminal A and higher proliferation rate luminal B tumors. Patients with the triple negative breast cancer (TNBC) subtype, characterized by the absence of ER, progesterone receptor (PR), and human epidermal growth factor receptor-2/neu receptors (HER2/neu) have a poor prognosis [5,6] also due to the few clinical treatments available. Considerable effort has gone into identifying new therapeutic agents, with multiple targeting abilities, able to circumvent the limitation of current conventional therapy.
Combined cancer therapy utilizes two or more agents and may improve the therapeutic efficacy of the single drug through a synergistic effect, leading to a potentially reduced drug resistance [7].
Many epidemiological studies suggest that phytochemicals, present at high levels in vegetables and fruits, have anticarcinogenic properties [8][9][10][11] and, triggering apoptosis, may be an effective treatment in cancer.
There is considerable interest in identifying bioactive compounds which, by increasing the sensitivity to conventional chemotherapeutic agents, could improve the patient's quality of life by reducing the side effects of therapy [12][13][14][15][16][17]. It has been recently demonstrated that combined treatment of natural polyphenols and chemotherapeutic agents are more effective than the drug alone in hindering the growth of cancer cells [18,19] and in promoting chemosensitivity in multidrug resistance (MDR) cancer cell lines [20].
Growing interest in dietary phytochemicals has led to renewed attention being paid to the artichoke, because of its high content in polyphenols. Artichoke polyphenols are mainly glycoside forms of flavonoid, such as apigenin and luteolin in the leaves and hydroxycinnamic acid derivatives in the edible part, mainly represented by mono-and dicaffeoylquinic acids. Many in vitro and in vivo experiments have shown that artichoke has diuretic, hepatoprotective, hypocholesterolemic, and antioxidant properties [21][22][23][24] and, more recently, antitumoral activities [24][25][26]. Our previous findings indicate that AEs protect hepatocytes from oxidative stress and show cancer chemopreventive properties by triggering apoptosis in human hepatoma cells [24] and in human breast cancer cell lines without any toxicity in the nontumorigenic MCF10A cells [25]. We have also provided evidence that low doses and chronic AE treatments exert anticancer activity through induction of premature senescence in MDA-MB231, a triple negative and highly aggressive breast cancer cell line [27]. Furthermore, the bioavailability of metabolites of hydroxycinnamic acids, after ingestion of cooked artichoke, has also been demonstrated in human subjects [28].
Taxanes are a family of chemotherapeutic drugs employed for the treatment of many tumors including breast cancer in both early and metastatic stages [29]. One of these, PTX, is a microtubule-stabilizing drug [30] which, because of its effect on mitotic spindle dynamics, may lead to cell cycle arrest and apoptosis [31]. More recently, it has been suggested that many anticancer drugs, including taxanes, have the ability to induce oxidative stress [32], which indicates an additional antitumoral mechanism.
FEN1 is a key member of the endonuclease family involved in cellular DNA replication and repair [33]. As a structure-specific nuclease, FEN1 stimulates Okazaki fragment maturation during DNA repair and efficient removal of 5′-flaps during long-patch base excision repair [34]. FEN1 is also reported to be linked to apoptosis-induced DNA fragmentation in response to apoptotic stimuli [34,35], and its expression is closely associated with cell proliferation and correlated with increased tumor grade and aggressiveness [36].
Oxidative stress is a result of a cellular imbalance in the production of ROS and the activity of the endogenous antioxidant protective system [37]. Notably, ROS-induced oxidative stress plays an important role in cancer development and progression. Among ROS-inducing agents, many phytochemicals, including curcumin [38,39], resveratrol [40,41], and epigallocatechin-3-gallate [42], have been shown to enhance the anticancer properties of chemotherapeutic agents. We have previously demonstrated a dual role of AEs, as prooxidant in breast cancer cells [25] and as antioxidant in normal hepatocyte [24] showing an inhibitory effect on growth of tumor cells with little or no toxicity on normal cells based on their differential redox status.
It is well known that induction of phase II enzymes, counteracting reactive electrophiles including ROS, plays an important role in response to many anticancer agents including dietary compounds. Upon cellular stimulation by oxidative stressor molecules Nrf2, the main transcription factor involved in the regulation of phase II and antioxidant gene expression moves to the nucleus where it interacts with antioxidant response elements (AREs) present in the promoter region of many phase 2 genes [43][44][45].
The roles of growth factors and mitogens in regulating gene expression, apoptosis, and differentiation have been reported to be mediated by the Ras/MEK/ERK signaling cascade [46]. This pathway has been shown to be frequently activated in breast cancer, and the MAPK pathway is a well-explored target of therapeutic intervention. Therefore, specific inhibitors targeting Ras, Raf, MEK, and other downstream proteins have been tested in clinical trials [47]. Since p-ERK expression sensitizes activation of DNA damageinduced checkpoints, inhibition of p-ERK may enhance the genotoxic effect of chemotherapy, probably as a consequence of the accumulation of DNA lesions due to compromised checkpoint activation [48].
To improve our knowledge for designing new therapeutic interventions, we have characterized the biological role of the combined treatment involved in the synergistic antitumor effect of AE/PTX approach.

Materials and Methods
2.1. Artichoke Extract Preparation. The edible part (head) of fresh artichoke is used for extract preparation, and the analysis of polyphenols contained in the extracts was performed by HPLC as previously described [25].

Cell Lines and Cultured
Conditions. The human breast cell lines were maintained in a humidified incubator with 5% CO 2 and 95% air at 37°C. MCF7 and MDA-MB231 cells, respectively, luminal A oestrogen receptor positive and basal B triple negative receptor subtypes were grown in RPMI and DMEM, respectively (Invitrogen, Life Technologies, Monza, Italy) and supplemented with10% FBS, 10 IU/ml of penicillin and 10 μg/ml of streptomycin.
PCR conditions were 50°C for 2 min, 95°C for 2 min, followed by 40 cycles of 95°C/15 s, annealing at 56°C/30 s and 72°C/30 s. All reactions were performed in triplicate. Data was normalized to GAPDH and the fold change in gene expression relative to normal was calculated using the comparative Ct method [49].
2.9. ROS Detection 2.9.1. Fluorescence Microscopy. MDA-MB231 cells were treated with vehicle or PTX plus and minus AEs and GOx as positive control (as indicated). After 4 h, the oxidationsensitive fluorescent probe DHE was used to assess the production of cytosolic superoxide anions. Briefly, after exposure, cells were incubated with 5 μM DHE for 40 minutes at 37°C in the dark and then rinsed twice with PBS. The cellpermeant DHE entered the cells, was oxidized by superoxide anions to form ethidium (ETH) which binds to DNA, and produced fluorescent ETH-DNA. The fluorescent signals were obtained by the cultured cells at λ ex 300 nm and λ em 610 nm. Cells were visualized and images were captured using a fluorescence microscope apparatus (Olympus IX71-Olympus Tokyo, Japan) equipped with a digital camera (Tucsen Photonics Co., Ltd., Fuzhou, Fujian, China). For every histogram, a minimum of 20,000 events were counted. The mean fluorescence intensity was detected and expressed as a percentage of relative ROS level versus control cells.

Statistical
Analysis. Data is presented as mean ± standard deviation (SD). Statistical analysis of the results was performed using Student's t-test, with GraphPad Prism v5.01 for Windows (GraphPad Software, San Diego, CA, USA). For all statistical tests, a two-tailed p value < 0.05 was considered significant. All data reported were verified in at least three independent experiments and expressed as mean ± SD.

Synergy between AEs and Chemotherapeutic Drugs
Induces Loss of Cancer Cell Viability. In order to highlight a potential effect of compound combination between AEs and the most active and widely used cancer drugs in clinical management (taxanes, anthracyclines, and platinum complexes), we used two breast cancer cell lines, MCF7 and MDA-MB231, as experimental models. In both cell lines, PTX, ADR, or CDDP was employed at fixed concentrations (20 nM, 2.0 μM, and 20 μM, respectively). These treatments for 24 h lead to a reduction of cell viability of 30% (IC 30). By using the algorithm described by Fransson et al. [51] to calculate the combination index (CI), we analyzed the effect of AEs plus PTX, ADR, and CDDP compared with that of the single agents. The addition of AEs, from 12.5 μM up to 50 μM, to PTX or ADR in MDA-MB231 (Figures 1(a) and 1(b)) and in MCF7 (Figured 1(c) and 1(d)) cells yielded a decrease in cell viability which could be ascribed to a chemosensitizing effect of AEs to these drugs. The analysis of this data showed that the combined treatment, as demonstrated by the CI values in Figure 1, enhances cytotoxicity in a synergistic manner. Conversely, the      Oxidative Medicine and Cellular Longevity presence of AEs did not increase the number of dead cells caused by CDDP in a significant manner in both MCF7 and MDA-MB231 cells (Fig. 1s).
Since triple negative breast cancers have a more aggressive phenotype and a poorer prognosis due to the high propensity for metastatic progression and absence of specific hormonal-based targeted treatment, novel therapeutic strategies are required. To this end, we focused on MDA-MB231 as a TNBC cellular model to best describe a synergistic effect of AEs in combination with low doses of PTX, a first-line chemotherapeutic agent in breast cancer.

The Combination of AEs and PTX Inhibits Human Breast
Cancer Cell Proliferation. In view of the aforementioned effects on cell viability, we thoroughly explored the mechanisms involved in the synergistic effect of the combined treatment on breast cancer cells.
To further characterize the decreased cell viability induced by cotreatment, we evaluated the expression of cleaved PARP (c-PARP) and LC3 as potential molecular signs of regulated cell death. No significant modulation of these proteins was detected compared to PTX treatment alone, which is well known to cause marked cell death [52,53]. These slight effects that were detected demonstrate that both apoptotic and autophagic cell deaths are not relevant for the synergistic response of the combined treatment (Fig. 2s).
These findings prompted us to detect DNA synthesis ratio in treated cells. As shown in Figure 2, AEs strongly decrease 3 H-thymidine incorporation in a dose-dependent manner in MDA-MB231 cells treated with 20 nM PTX. The highest concentration tested (25 μM AEs) inhibited cell proliferation by 70% compared to PTX alone.
In order to investigate additional features of cellular response to cotreatment, we evaluated the clonogenic ability of breast cancer cells. Since MDA-MB231 cells do not aggregate well and form very dispersed colonies [54], we switched to MCF7 cell model with a similar sensitivity to AEs/PTX (Figure 1) to test the effect of cotreatment on the colony for-mation ability. Our results demonstrate that this property was poorly affected (Fig. 3s). [27,[55][56][57] and on the ability of PXT to promote intracellular ROS formation [32], we evaluated the oxidative pathway as a potential mechanism involved in AE/PTX-induced cell proliferation inhibition. In agreement with Chikara et al. data [58], which shows that several polyphenols strengthen the anticancer properties of chemotherapeutic drugs by elevating ROS levels, we report (Figure 3(a)) increased numbers of bright red fluorescent cells indicating enhanced levels of superoxide anions.

Role of ROS in Synergistic Cytotoxicity in AE/PTX-Treated Cells. Based on the prooxidant activity of natural polyphenols in inducing cell growth inhibition
To further investigate the involvement of oxidative pathway in AE/PTX synergistic activity, the amount of ROS in particular peroxides was evaluated by flow cytometry. ROS production induced by the combined treatment takes place early, since it increases in relation to single agents after 2 h of exposure (Figures 3(b) and 3(b1)). This trend is less evident after 24 h treatment, probably as a result of ROS accumulation (Fig. 4s).
To further determine the role of ROS in AE/PTXinduced cell growth arrest, we sought to examine whether inhibition of ROS production by the well-known antioxidant NAC has any impact on synergy in breast cancer cell viability. As shown in Figure 3  Since we detected a partial cytotoxicity rescue by NAC exposure, in order to explore the AE/PTX effect further, we hypothesized that a different molecular mechanism might be involved in the cellular synergistic response.
Several studies have reported that the MAPK family members play crucial roles in cell proliferation, survival, and differentiation [62,63]. In particular, ERK kinases have been shown to play a part in DNA damage response (DDR) and that inhibition of p-ERK enhances the genotoxic effect of chemotherapeutic drugs [48]. Zou et al. reported that curcumin-treated breast cancer cells are more sensitive to cisplatin by downregulation of FEN1 achieved by reduction of p-ERK expression [19]. In agreement with published results, in our experimental settings, FEN1 and p-ERK expression levels decreased in combination treatment compared with   Figure 5(b)). In order to show a direct correlation between p-ERK and FEN1 expression, untreated MDA-MB231 cells were exposed to ERK inhibitor U0126 (20 μM). After 60 minutes of U0126 exposure, both ERK phosphorylation and FEN1 expression were clearly downregulated ( Figure 5(c)).
To test DNA damage level related to FEN1 downregulation, we looked at the extent of H 2 AX phosphorylation (γ-H 2 AX), a sensitive indicator of DNA double strand breaks ( Figure 5(d)). We detected a marked phosphorylation level of histone H 2 AX in the experimental cellular setting treated with 25 μM AEs plus PTX, compared to PTX alone.
This data is evidence that AEs strengthen the antitumor activity of PTX both through the ROS/Nrf2 pathway and via the downregulation of p-ERK, which result in the decrease in FEN1 expression.
These findings indicate that the synergistic response is a result of balance between the ROS/Nrf2 pathway and the impaired DNA damage response (DDR) triggered by FEN1 downregulation, suggesting the presence of at least two molecular mechanisms involved in the synergistic cytotoxicity elicited by our combined treatment.

Discussion
To explore the role of natural compounds as chemosensitizer agents in breast cancer, we have investigated the effect of combined treatment of artichoke polyphenolic extracts with paclitaxel, adriamycin, or cisplatin on breast cancer cell lines. We demonstrate that AEs synergized with PTX or ADR in hindering the growth of MDA-MB231 or MCF7 cells compared with drug alone. AEs enhanced breast cancer sensitivity to PTX by decreasing cell proliferation both through the ROS/Nrf2 pathway and via the downregulation of p-ERK, and these mechanisms resulted in the decrease in FEN1 expression leading to DNA damage accumulation and DNA replication reduction. These findings suggest the presence of at least two molecular mechanisms involved in the synergistic AE/PTX cytotoxicity in breast cancer cells.
Tumor cells exhibit excessive ROS production which is related to aberrant metabolism and continuous cell division; therefore, cancer cells appear to be more vulnerable to -    Oxidative Medicine and Cellular Longevity further oxidative insult compared to normal ones [64]. ROS cellular levels are thus crucial for designing advanced therapies and a future challenge in anticancer treatment [58]. There are many reports describing the prooxidant effect of polyphenols in sensitizing cancer cells to chemotherapeutic drugs through different pathways [38][39][40][41][65][66][67].
In light of our previous findings regarding the role of AEs as prooxidant players on breast cancer cells [27], this study has explored the ability of AEs to sensitize breast cancer cell to conventional chemotherapy through oxidative process. Based on literature data [38][39][40][41][42], our results strongly support the prooxidant role of AEs/PTX in hindering breast cancer cellular growth associated with ROS production. To further confirm the important role of ROS in enhancing the antitumor effect of the combined treatment, we demonstrate that the well-known antioxidant NAC attenuates the synergistic and antitumor effect on MDA-MB231 cells. Since we detected a partial cytotoxicity rescue by NAC exposure, we hypothesized that a further molecular mechanism was involved in the cellular synergistic response.
In agreement with published results, our data has shown that many anticancer drugs including ADR and PTX but not CDDP have the ability to induce oxidative stress [32]. We have shown that both ADR and PTX plus and minus AEs induce a synergistic effect on reducing cell viability, while AE/CDDP combined treatment did not modify cell death ratio. This finding suggests that AEs are able to produce ROS and synergize with the prooxidant activity of some chemotherapeutic drugs, which suggests that at least one of the mechanism involved in synergistic interaction is ROS production dependent.
Nrf2 is a main transcription factor in the regulation of many phase II and antioxidant genes [43,44]. It is well known that induction of phase II enzymes, counteracting reactive electrophiles including ROS, plays an important role in response to many chemotherapeutic agents. Nrf2 is located in the cytoplasm in a complex with the actinbinding protein Keap1. Upon oxidative stimulation, Nrf2 moves to the nucleus where it interacts with AREs present in the promoter region of many genes [45]. In our experimental model, after AE/PTX exposure for 24 h, the Nrf2 mRNA level, together with the nuclear protein level, significantly increased compared with PTX alone treatment. According to Kwak et al. [68], who demonstrated that Nrf2 has a short half-life, we hypothesized that the nuclear accumulation is due to de novo synthesis of the transcription factor. In agreement with Chen et al. [61], who showed that Nrf2 bounds to ARE-like sequence located in the FEN1 promoter region, we observed that Nrf2 downregulated the expression of FEN1, which suggests that Nrf2 may function as a repressor of this endonuclease gene.
Our data has shown that cotreatment significantly reduced the FEN1 RNA level together with a downregulation of the protein expression compared with PTX alone and suggests that such a mechanism is involved in the antiproliferative effect of AEs/PTX. The combined treatment results in DNA damage accumulation, confirmed by the phosphorylation of H 2 AX which leads to DNA replication reduction. This data suggests that downregulating the expression of FEN1, a potential biomarker and therapeutic target [36,69], may be a new mechanism by which AEs sensitize breast cancer cells to PTX.
It has been shown that polyphenols can modulate multiple signaling pathways including MAPK/ERK [19,38,70,71]. By inhibiting ERK phosphorylation, curcumin helps to sensitize the cells to cisplatin by targeting FEN1 [19]. In agreement with this data, our results showed that FEN1 and p-ERK expression levels decreased in combination treatment compared with PTX alone. Since the inhibition of ERK activation by the common MEK inhibitor U0126 induced a downregulation of FEN1, we envisage a marked correlation between this nuclease and the aforementioned kinase expression level in our experiments. As previously reported, aberrant activation of the ERK pathway is one of the most relevant events in human cancer as it stimulates cell proliferation [72]. However, the activation of the ERK pathway has an important role in DDR and has been associated with chemotherapeutic drugs commonly considered to be DNA damage inducers. For several tumors, inhibition of MAP-K/ERK cascade could enhance the genotoxic effect of chemotherapeutic agents. This is probably because of the accumulation of DNA lesions due to the impaired checkpoint activation ERK related [48]. Therefore, if a combinatory therapy is to be used, it is important to determine if the combination will lead to inefficient repair of DNA lesions due to downregulation of FEN1.
Our results summarized in supplementary materials (Fig.  5s) suggest that artichoke polyphenols can be used as sensitizers of chemotherapy paving the way for new combined treatment in oncology. Since polyphenols are widely used as dietary components and have shown no toxicity in humans, their use as adjuvant agents may be an innovative and promising natural compound-based therapeutic strategy. Our study adds a novel aspect of the underlying mechanisms of the anticancer properties of AEs to our previous findings. However, wide-ranging pharmacokinetic and metabolic studies of combined treatment on animal models are required to evaluate its efficacy in human clinical trials.

Conclusions
We envisage a combined AE/chemotherapeutic agent treatment that is able to reduce the dose of antitumor drugs minimizing toxicity and side effects of conventional cancer therapy.

Data Availability
The data used to support the findings of this study is included within the article.

Conflicts of Interest
The authors declare that there is no conflict of interest regarding the publication of this paper. 8 Oxidative Medicine and Cellular Longevity

Supplementary Materials
Supplementary materials integrate data described in the main document. Fig. 1s: effect of AEs on CDDP treated breast cancer cells. Cell viability assay: MDA-MB231 (a) and MCF7 (b) were treated with CDDP (20 μM) with or without AEs (from 12.5 to 50 μM) for 24 h. Histograms show cell viability. No significant differences are detected between combined treatment vs. CDDP alone. Fig. 2s: effect of AEs/PTX on regulated cell death. Apoptosis and autophagy marker analysis: cells were incubated with AEs or AEs/PTX and then assayed via western blot for c-PARP (a) and LC3 (b). Fig. 3s: cloning efficiency. Colony-forming analysis: MCF7 cells were exposed to AEs or AEs/PTX for 24 h and then allowed to grow and form colonies for subsequent 14 d. Cell colonies, after staining with crystal violet (a), were counted and the values reported as percent colony number in the histogram (b). C vs. PTX * * p = 0:0049, 12.5 μM AEs vs. 12.5 μM AEs/PTX * * * p < 0:0001, and 25 μM AEs vs. 25 μM AEs/PTX * * * p = 0:0005. There is no significant difference between colony numbers of cells treated with AEs/PTX or PTX alone.