Background
Lung cancer is one of the most commonly diagnosed cancer worldwide, where it also ranks first as a leading cause of cancer-related deaths among both men and women [
1]. In the clinical practice, lung carcinomas are divided into two main histological types: small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC), where the other accounts for 75–85 % of total lung cancer cases [
2]. Since a predominant number of NSCLC cases fail to be detected at the surgically resectable early stage of disease, therefore chemotherapy and radiotherapy remain the mainstay of the treatment for inoperable NSCLC patients. However, while there have been numerous drugs approved for use, they have often suffered from the limited clinical applicability due to the development of resistance by tumor cells and non-specific toxicity toward normal cells [
3,
4]. One such a drug, paclitaxel (PTX), a taxane plant product derived from the bark of the pacific yew tree
Taxus brevifolia, has demonstrated a clinically significant activity against a broad variety of tumor types and has become a first-line treatment for NSCLC [
5]. Nevertheless, the administration of this drug at optimum doses and for a prolonged time has been hampered by the prevalence of serious side effects, such as neutropenia, neuropathy, and the acquisition of clinical resistance as well [
6]. Recently, a considerable attention has been given to the identification of new therapeutic agents with synergistic effects with paclitaxel and other conventional cytostatic drugs as a promising direction to overcome the above mentioned drug drawbacks. In general, the combination therapies have proven to be more potent than monotherapy in the treatment of cancers. They not only potentiate the therapeutic efficacy of each agent alone and/or enable the use of reduced doses of a single drug but also decrease the possibility of the development of drug resistance [
7,
8].
In this context, plant polyphenols, especially these from dietary sources, have recently received an increased scientific attention as the appropriate contenders to serve as a partner for traditional chemotherapeutic drugs. The dietary phytochemicals are commonly perceived as non-toxic, well-tolerated, easily available, inexpensive compounds that can target multiple cellular pathways [
9,
10]. Indeed, there is evidence that these agents may potentiate the cytotoxic effects of chemotherapy and radiotherapy, protect normal cells from therapy-associated toxicity, increase a systemic bioavailability of cytostatic agents, and in some cases, even overcome chemoresistance [
11,
12]. Despite the large number of the dietary polyphenols, not all of them share the same anticancer activity, hence a considerable attention of researchers has focused on the selected groups, including flavonoids. In several comparative studies, fisetin (3,3′,4′,7-tetrahydroxyflavone; FIS), a naturally occurring diet-based flavonoid, has stood alone as an effective anticancer agent against a broad spectrum of tumor cell lines, with some of the antitumor effects being achieved at physiologically relevant concentrations, and without affecting normal cells, including human bronchial epithelial (NHBE) cells [
3,
13‐
18]. Therefore, the aim of this study was to investigate whether fisetin, at physiologically attainable concentrations, could act synergistically with clinically achievable doses of paclitaxel to produce growth inhibitory and/or pro-death effects on A549 non-small cell lung cancer cells, and if it does, what mechanisms might be involved. Here, we report the first experimental evidence on such synergistic action, which was, at least partially, ascribed to the induction of mitotic catastrophe and the autophagic cell death.
Materials and methods
Cell culture
The human non-small cell lung cancer cell line A549 and the human colon adenocarcinoma cell line LoVo were kindly provided by Dr. P. Kopinski (Department of Gene Therapy, Collegium Medicum in Bydgoszcz, Nicolaus Copernicus University, Poland) and Prof. P. Dziegiel (Department of Histology and Embryology, Wroclaw Medical University), respectively and both cell lines were cultured in Dulbecco’s Modified Eagle Medium (DMEM; Lonza; Verviers, Belgium). The human breast adenocarcinoma cell line MCF-7 was purchased from American Type Culture Collection (ATCC; HTB-22) and maintained in Minimum Essential Medium (MEM, Lonza; Verviers, Belgium) supplemented with 1 % non-essential amino acids (Sigma-Aldrich; St. Louis, MO, USA). The human non-small cell lung cancer cell line H1299 was bought from American Type Culture Collection (ATCC; CRL-5803) and these cells were grown in Roswell Park Memorial 1640 medium (RPMI 1640; Sigma-Aldrich; St. Louis, MO, USA). All media were supplemented with 10 % fetal bovine serum (PAA; Pasching, Austria) and 50 µg/ml gentamycin (Sigma-Aldrich; St. Louis, MO, USA) and all cultures were carried out in a humidified atmosphere of 95 % air and 5 % CO2 at 37 °C. After reaching 70–80 % confluence during the exponential growth, the cells were harvested with trypsin–EDTA solution (Sigma-Aldrich; St. Louis, MO, USA) and subcultured on 12- or 6-well plates (at the density of 0.11 × 106 cells/well and 0.3 × 106 cells/well, respectively) (BD Falcon; Franklin Lakes, NJ) for further experiments.
Cell treatment
Stock solutions of fisetin and paclitaxel (Sigma-Aldrich; St. Louis, MO, USA) at concentration of 100 mM were prepared in 100 % dimethyl sulfoxide (Sigma-Aldrich; St. Louis, MO, USA), stored at 25 °C and serially diluted in complete growth medium immediately before use. After a 24-h incubation to allow cell attachment, paclitaxel (at concentrations of 0.1, 0.2, 0.3, 0.4, 0.5 µM for MTT assays and 0.1 µM for other experiments) or fisetin (at concentrations of 10, 20, 30, 40, 50 µM for MTT assays and 10 µM for other experiments) were added to cells for the indicated times as either single or combined agents at a fixed concentration ratio of 1:100. Preliminary screening was carried out to ascertain the incubation time and the treatment regimen—a sequential versus combined treatment (data not shown). The final concentrations of DMSO did not affect the cell viability (data not shown). Control cells were cultured under identical conditions, but without the addition of the tested agents.
MTT assay
The cell viability was assessed using MTT colorimetric assay. The cells were seeded in 12-well plates in complete grown medium and 24 h later, the cells were treated with paclitaxel at doses from 0.1 to 0.5 μM and with fisetin at concentrations ranging from 10 to 50 μM, either alone or in a fixed ratio of 1:100, for the next 24 h. The MTT stock solution was made by dissolving 5 mg of thiazolyl blue tetrazolium bromide (MTT; Sigma-Aldrich; St. Louis, MO, USA) in 1 mL of phosphate-buffered saline (PBS) and sterilized by passage through a Whatman filter (Florham Park, NJ) with a pore size of 0.2 µm. After the drug treatment, the cells were once washed with PBS and incubated for 3 h (37 °C, 5 % CO2, 95 % air atmosphere) in a working solution of MTT, prepared by diluting a stock solution with DMEM without phenol red (Lonza; Verviers, Belgium) in the ratio 1:9. The surviving cells converted MTT to formazan that generated a blue-purple color when dissolved in acidic isopropanol. Dye absorbance was measured at 570 nm using spectrophotometer (Spectra Academy, K-MAC, Korea). The experiment was repeated six times and the cell viability was calculated as the percentage of MTT reduction, assuming the absorbance of control cells as 100 %. Under the used conditions, MTT assay allowed to estimate the loss in cell viability resulting from the inhibition of the cell proliferation, the increase in cell death or the sum of both processes.
Drug interaction analysis
To evaluate whether the antitumor effects of the combination of fisetin and paclitaxel were synergistic, additive or antagonistic, the drug interactions were analyzed based on the combination index method of Chou and Talalay [
19]. Using data obtained from MTT assays and CompuSyn software [
20], the dose–effect curves for single agents and their combinations were generated and the combination index (CI) values for each dose and the corresponding effect level, referred to as the fraction affected (f
a; the fraction of cells inhibited after the drug exposure, e.g. 0.5 when cell growth is inhibited by 50 %), were calculated. The resulting combination index offers a quantitative definition for an additive effect (CI = 1), synergism (CI < 1), and antagonism (CI > 1) in drug combinations [
21]. Then, to provide a visual illustration of drug interactions, the F
a–CI plot was constructed by simulating CI values over a range of f
a levels from 0.1 to 0.95.
Annexin V/propidium iodide (PI) binding assay
To assess the extent of apoptosis, the Tali Apoptosis kit—Annexin V Alexa Fluor 488 and Propidium Iodide (Invitrogen/Life Technologies, Carlsbad, CA, USA) was used according to the manufacturer’s instructions. In short, after the treatment, the cells were collected from 6-well plates using trypsin–EDTA solution, centrifuged at 300×g for 8 min, resuspended in annexin binding buffer (ABB) and incubated with Annexin V Alexa Fluor 488 at room temperature (RT) in the dark for 20 min. Following the centrifugation at 300×g for 5 min, the cells were again resuspended in ABB and incubated with propidium iodide at RT in the dark for 5 min. The cells were analyzed using Tali image-based cytometer (Invitrogen/Life Technologies, Carlsbad, CA, USA). The data were quantified by FCS Express Research Edition software (version 4.03; De Novo Software, New Jersey, NJ, USA) and expressed as the percentage of cells in each population (viable Annexin V−/PI−; early apoptotic Annexin V+/PI−; late apoptotic Annexin V+/PI+; necrotic Annexin V−/PI+). The sum of the early and late apoptotic cells represented the total apoptosis.
Cell cycle analysis
For DNA content analysis, the Tali Cell Cycle Kit (Invitrogen/Life Technologies, Carlsbad, CA, USA) was used according to the manufacturer’s instructions. Briefly, the treated cells were harvested from 6-well plates by trypsinization, rinsed with PBS, fixed in ice-cold 70 % ethanol at 4 °C, and left at −25 °C overnight. The next day, the cells were centrifuged at 1000×g for 5 min at 4 °C and washed with PBS. After the centrifugation at 500×g for 10 min at 4 °C, the cells were resuspended in the Tali Cell Cycle Solution containing propidium iodide (PI), RNase A, and Triton X-100. Following 30-min incubation at RT in the dark, the cells were analyzed using Tali image-based cytometer (Invitrogen/Life Technologies, Carlsbad, CA, USA) and the percentage of cells in each phase of the cell cycle was determined using FCS Express Research Edition software (version 4.03; De Novo Software, New Jersey, NJ, USA).
Fluorescent staining of β-tubulin and cell nuclei
For spindle morphology assessment, the cells were seeded on glass cover slides in 12-well plates, permitted to adhere overnight and then treated with fisetin and/or paclitaxel. After the prefixation step with 1 mM bifunctional protein cross-linking reagent 3,30-dithiodipropionic acid (DTSP; Sigma-Aldrich, St. Louis, MO, USA) in Hank’s balanced salt solution (HBSS; Sigma-Aldrich, St. Louis, MO, USA) for 10 min, the cells were extracted in TsB (0.5 % Triton X-100; Serva, Heidelberg, Germany) in microtubule stabilizing buffer (MTSB; 1 mM EGTA, 4 % poly(ethylene glycol), 10 mM PIPES; Sigma-Aldrich, St. Louis, MO, USA) containing DTSP (dilution 1:50) for 10 min and rinsed with TsB for 5 min. Following the fixation of the cells with 4 % paraformaldehyde (Serva, Heidelberg, Germany) in MTSB for 15 min and three washing steps with PBS, the cells were incubated with 1 % bovine serum albumin (BSA; Sigma-Aldrich, St. Louis, MO, USA) diluted in tris-buffered saline (TBS; Sigma-Aldrich, St. Louis, MO, USA) for 15 min. β-tubulin was labeled using the mouse monoclonal antibody specific for β-tubulin (Sigma-Aldrich, St. Louis, MO, USA) diluted 1:65 in 1 % BSA-PBS (1 h, a moist chamber). This was followed by rinsing the cells three times in 1 % BSA-PBS and incubation with the goat anti-mouse IgG-TRITC secondary antibody (Sigma-Aldrich, St. Louis, MO, USA), diluted 1:85 in PBS (45 min, a moist chamber, in the dark). To assess the nuclear morphology, the slides were incubated with 4′,6-diamino-2-phenylindole (DAPI, diluted 1:20,000 in distilled water; Sigma-Aldrich; St. Louis, MO) for 10 min in the dark. Finally, the slides were rinsed three times with distilled water, mounted with Aqua-Poly/Mount (Polysciences; Warrington, PA) and analyzed using Nikon Eclipse E800 fluorescence microscope and NIS-Elements 4.0 software (all from Nikon; Tokyo, Japan). At least 100 mitotic cells from three independent experiments were counted to determine the percentage of cells with the bipolar, multipolar or monopolar spindles.
Transmission electron microscopy
To detect the presence of autophagic vacuoles in transmission electron microscope (TEM), the treated cells were harvested from 6-well plates, fixed with 3.6 % (v/v) glutaraldehyde (Polysciences; Warrington, PA) in 0.1 M sodium cacodylate buffer (pH 7.4; Roth; Karlsruhe, Germany) for 30 min at RT and washed three times in 0.1 M sodium cacodylate buffer. Then, the cells were entrapped within fibrin clots, which were formed by the vigorous mixing of cell pellets with the equal volumes of the fibrinogen (3 mg/ml; Sigma-Aldrich; St. Louis, MO, USA) and thrombin solution (50 U/ml; Biomed-Lublin; Lublin, Poland). Afterwards, the samples were post-fixed with 1 % (w/v) osmium tetroxide (Serva; Heidelberg, Germany) in 0.1 M sodium cacodylate buffer for 1 h at RT, rinsed three times with 0.1 M sodium cacodylate buffer, dehydrated through a graded ethanol (30–90 %; POCH S.A., Gliwice, Poland) and acetone (90–100 %; POCH S.A., Gliwice, Poland) series, infiltrated with increasing ratios of epoxy resin (Epon 812; Roth; Karlsruhe, Germany) with hardeners (DBA and MNA; Roth; Karlsruhe, Germany): 100 % acetone and finally embedded in gelatin capsules (Ted Pella, Inc.; Redding, CA) filled with pure epoxy resin with DMP-30 (Roth; Karlsruhe, Germany). The polymerization of the resin occurred at 37 °C for 24 h, and then at 65 °C for 120 h. The selected parts of the material were cut into ultra-thin sections using Reichert Om U3 ultramicrotome, placed on copper grids (Sigma-Aldrich; St. Louis, MO, USA), stained with 1 % uranyl acetate (Ted Pella, Inc.; Redding, CA) and examined with JEM-100CX transmission electron microscope (JEOL; Tokyo, Japan).
Detection and quantification of acidic vesicular organelles with acridine orange staining
To detect the development of acidic vesicular organelles (AVOs), which are the hallmark of autophagy, the vital staining of A549 cells with acridine orange (AO; Sigma-Aldrich; St. Louis, MO, USA) was performed. The cells were seeded on coverslips in 12-well plates and after the attachment, they were incubated with fisetin and/or paclitaxel. Then, the cells were stained with medium containing 1 μg/ml AO for 15 min at 37 °C, washed twice in PBS and immediately examined with Nikon Eclipse E800 fluorescence microscope and NIS-Elements 4.0 software (all from Nikon; Tokyo, Japan). The cytoplasm and nucleus of AO-stained cells fluoresced bright green, whereas the acidic autophagic vacuoles fluoresced bright red, as described previously [
22]. To quantify the development of AVOs, the cell pellets were stained with AO (1 μg/mL) for 15 min at 37 °C, washed twice with PBS and instantly analyzed with Tali Image-Based Cytometer (Invitrogen/Life Technologies, Carlsbad, CA, USA) and FCS Express Research Edition software (version 4.03; De Novo Software, New Jersey, NJ, USA). Approximately equal amounts of cells were measured for each sample and the geometric mean of red fluorescence intensity was used to quantify the responses. To inhibit autophagy, the cells were pretreated with 100 nM bafilomycin A1 (Baf A1) for 4 h, followed by washing with PBS and subsequent incubation in the absence or presence of the tested compounds for the indicated period of time.
Immunofluorescent staining of LC3-II
To examine the intensity and the pattern of LC3-II immunostaining, the cells were seeded on glass cover slides in 12-well plates and left to attach overnight. After the treatment with fisetin and/or paclitaxel, the cells were fixed with 4 % paraformaldehyde (Serva; Heidelberg, Germany) in PBS for 30 min, washed three times with PBS, permeabilized with 0.25 % Triton X-100 (Serva; Heidelberg, Germany) in PBS for 5 min and blocked with 1 % BSA (Sigma-Aldrich; St. Louis, MO, USA) in PBS (BSA–PBS pH 7.6) for 20 min. The staining of LC3-II was performed using the rabbit anti-LC3-II antibody (Thermo Scientific; Rockford, USA) diluted 1:2000 in 1 % BSA–PBS (1 h, RT, a moist chamber). After rinsing three times with 1 % BSA–PBS, the cells were incubated with Alexa Fluor 488-labeled goat anti-rabbit secondary antibody (Life Technologies Corp.; Carlsbad, CA, USA) diluted 1:100 in PBS (60 min, RT, a moist chamber in the dark). Following three washing steps with PBS, the cell nuclei were counterstained with DAPI (diluted 1:20,000 in distilled water; Sigma-Aldrich; St. Louis, MO, USA) for 10 min. Finally, the slides were rinsed three times with distilled water, mounted with Aqua-Poly/Mount (Polysciences; Warrington, PA) and examined using Nikon Eclipse E800 fluorescence microscope and NIS-Elements 4.0 software (all from Nikon; Tokyo, Japan).
Quantitative real-time PCR analysis
To determine the expression level of LC3-II, Bax, Bcl-2 and caspase-3, SYBR green-based quantitative real-time PCR was performed using LightCycler 2.0 Instrument (Roche Applied Science; Mannheim, Germany) and LightCycler Software Version 4.0. Total RNA from the A549 cells was prepared by using the Total RNA kit (A&A biotechnology; Gdynia, Poland) according to the manufacturer’s protocol. The reverse transcription and quantitative PCR reactions were performed in a single 20-μl LightCycler capillary (Roche Applied Science; Mannheim, Germany) as a one-step real-time qRT-PCR with TranScriba reverse transcriptase and Master Mix SYBR (TranScriba-qPCR Master Mix SYBR kit; A&A biotechnology; Gdynia, Poland) as specified by the manufacturer. The total reaction mixture (20 µl) contained 65 ng of RNA and 0.2 μM of each primer in addition to the TranScriba-qPCR Master Mix SYBR kit components. The sequences of primers were as follows: LC3-II forward 5′-CGGTGATAATAGAACGATACAAGG-3′; LC3-II reverse 5′-CTGAGATTGGTGTGGAGACG-3′; Bax forward 5′-AGATGTGGTCTATAATGCGTTTTCC-3′, Bax reverse 5′-CAGAAGGCACTAATCAAGTCAAGGT-3′; Bcl-2 forward 5′-AACATCGCCCTGTGGATGAC-3′, Bcl-2 reverse 5′-AGAGTCTTCAGAGACAGCCAGGAG-3′; caspase-3 forward 5′-TGGTTCATCCAGTCGCTTTG-3′, caspase 3 reverse 5′-CATTCTGTTGCCACCTTTCG-3′. One cycle of reverse transcription was carried out for 10 min at 50 °C, one cycle of denaturation for 3 min at 95 °C, and 40 cycles of denaturation for 15 s at 95 °C, followed by annealing and elongation for 30 s at 57–60 °C (depending on the primer’s melting temperature). Relative mRNA expression levels of LC3-II, caspase-3, Bax, and Bcl-2 were quantified using the ΔΔCt method (2
−ΔΔCt method) [
23] and the results were normalized to the expression of the housekeeping gene glucose 6-phosphate dehydrogenase (G6PD) or glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and presented as a fold difference relative to a calibrator sample (untreated A549 cells).
Statistical analysis
Statistical analysis was performed with GraphPad Prism software (GraphPad Software, San Diego, CA). The nonparametric Mann–Whitney U test was used to compare the differences between experimental points, and the significance level was set to p < 0.05. Data are presented as mean ± standard deviation (SD).
Discussion
In the present paper, we reported the first experimental evidence for the existence of synergism between fisetin and paclitaxel in the in vitro model of NSCLC. This synergism was quantified by the combination index method of Chou and Talalay [
19], which is based on the multiple drug effect equation derived from the median-effect principle of the mass-action law [
21]. Such preclinical drug combination studies in vitro and/or in animals are necessary to obtain the basis and rationale for the drug combination clinical trials, since from scientific, practical and ethical reasons it is impossible to determine synergism in humans [
46]. However, to avoid a lack of reproducibility of the in vitro tests in the clinical trials, as it is often seen in the clinical settings, an important issue for the in vitro testing of agents is the use of clinically relevant concentrations, on condition that the pharmacokinetic data are available [
47]. With this in mind, our intention was to use the in vivo attainable concentrations of fisetin (≤20 µM) [
3,
24] to find its synergistic combinations with clinically achievable doses of cytostatic drugs, such as paclitaxel, mitoxantrone, methotrexate or arsenic trioxide. According to the literature reports, the doses of PTX below or equal to 1 µM [
25,
26], MIT below or equal to 1 µM [
48], MTX below or equal to 1 µM [
49], and ATO below or equal to 2 µM [
50] are achievable in vivo and clinically relevant. Therefore, although our studies were carried out in vitro, all drugs were tested at the plasma concentrations attainable in vivo, providing a high hope for their easy reproducibility into the clinical settings.
In the present studies, the CI analysis demonstrated that the combination effect of FIS and PTX was synergistic in the A549 non-small lung cancer cells, in all concentrations tested and over a wide range of effect levels (f
a). The highest degree of synergism was found when 10 μM FIS was combined with 0.1 μM PTX (CI = 0.15), therefore we selected these doses for further drug combination studies on mechanistic aspects. In accordance with the previous report of Liao et al., we have shown that 10 μM FIS itself was not toxic to A549 [
51]. However, the contrary results have been obtained by Khan et al., who have revealed that the treatment with 10 µM fisetin for 24 h decreased the viability of A549 cells by 25 %, as measured using MTT assay [
3]. Although the reasons for such a significant discrepancy are unclear, the differences in the experimental methodology, e.g. in the MTT assay and the cell treatment (at different degrees of confluence) may be, at least partially, responsible. In the case of paclitaxel, we noticed that 0.1 μM concentration of this drug failed to substantially decrease the viability of A549 cells, allowing ~83 % of cells to remain viable, the results of which are in agreement with earlier reports from this cell line (for example, the IC
50 values of PTX in A549 cells was determined to be 5 μM [
27], 22.5 μM [
28] or 8.20 μM [
29].
Here, we also found that the combination effect of FIS and PTX was cell line-specific. Synergy was demonstrated in the LoVo and H1299 cells but not in the MCF-7 cells, where the strong antagonism was observed. In the LoVo and H1299 cells, the synergism was seen in a narrow range of effect levels, which, in the case of the latter cells may be considered as irrelevant from a clinical perspective, because they represent only a minor growth inhibition (IC38). The genetic differences among the tested cell lines that contributed to the diversity in the obtained interaction patterns are not known for the moment. A simple explanation that a diverse sensitivity might arise from different p53 status of the used cell lines, has rather been dismissed, because A549 and MCF-7 cell lines, both having functional p53, exhibited an opposite interaction pattern. However, we cannot firmly rule out this account since A549 (wild-type p53) and H1299 (p53-deficient) cells, representing two histopathological subtypes of lung cancer, also exhibited a distinct interaction pattern, and simultaneously it is possible that in the MCF-7 cells, the antagonistic interactions between FIS and PTX might be determined by other factors. Future studies will have to address this issue since such information may determine the potential clinical utility of FIS/PTX in the treatment of cancer. A cell-type dependent response to drugs implies the necessity to explore genes/molecular pathways that determine the chemosensitivity of tested cell lines. From a clinical perspective, it means that careful consideration should be paid to each patient’s individual characteristics to choose the most beneficial drug combination.
To date, only several studies have evaluated the fisetin’s ability to potentiate the anticancer activity of classical chemotherapeutic agents. It has been shown that fisetin enhances the cytotoxicity of cisplatin, TNF and doxorubicin in the H1299 non-small cell lung cancer cells [
52]. Importantly, fisetin has also been reported to act synergistically with cisplatin in vivo, in the NT2/D1 mouse xenograft model [
37]. In the cited study, the extent of tumor regression achieved with the co-treatment of fisetin (1 mg/kg/day) and cisplatin (1.5 mg/kg/day) was significantly more than the monotherapeutic treatments, whereas no toxicity was detected. Furthermore, using Lewis lung carcinoma-bearing mice as an experimental model, Touil et al. have shown that when fisetin was combined with cyclophosphamide, a marked improvement in the anticancer and antiangiogenic activity was observed (92 % tumor growth inhibition and a significant reduction in the microvessel density), with a low systemic toxicity [
53]. However, to our knowledge, neither methotrexate nor mitoxantrone and arsenic trioxide have not been previously tested in the combination with fisetin. In the present study, we demonstrated that, while the combinations of FIS with MTX and especially with MIT do not deserve further attention (at least in the case of A549 cells), its strong synergistic action with ATO on the contrary does.
Here, we also provided some insights into the mechanism of the synergistic action of FIS and PTX in the A549 non-small cell lung cancer cells. Firstly, we proposed that mitotic catastrophe, rather than apoptosis, was one of the possible mechanism of the synergistic cytotoxicity between fisetin and paclitaxel. While some of the characteristic features of apoptosis were noticed, such as phosphatidylserine externalization, the increase in caspase-3 mRNA level, the nuclear shrinkage and fragmentation, the level of these changes was relatively low and could not account for the extent of cell death triggered by the combination of FIS and PTX. Instead, major cellular events associated with mitotic catastrophe, including G2/M arrest followed by the mitotic slippage and polyploidy/aneuploidy, as well as the chromosome misalignment and missegregation and the appearance of the enlarged mononucleated and multinucleated cells were predominantly observed in the cell populations treated with PTX alone and, to a greater extent, in the FIS/PTX co-treated cells. The polyploid/aneuploid cells were most likely the result of the multipolar spindle formation that led to either the cleavage failure or the multipolar cell division. The asymmetric cytokinesis following the multipolar mitosis resulted in the generation of three or more daughter cells with an abnormal DNA content, including the hypodiploid cells (DNA content lower than 2 N) that appeared in the sub-G1 peak of the DNA histogram. Indeed, it has previously been shown that the cells undergoing multipolar and asymmetric divisions may then have an average of 1.33N DNA instead of 2N DNA [
54]. In turn, the cytokinesis failure was manifested morphologically as the enlarged mononucleated or multinucleated cells that probably were represented by the G2/M peak (as a consequence of one round of aberrant mitosis) or the polyploid fraction (that arose from another round of aberrant mitosis) on the DNA histograms. In fact, the G2/M peak, consists of both the mitotic cells (mitotic arrest) and the postmitotic tetraploid cells, which escape the mitotic block without cytokinesis [
55]. Based on the fact that multinucleation is the hallmark of the mitotic slippage [
55,
56], we can presume that an increased accumulation of G2/M phase cells at 10 h after starting of the treatment with FIS plus PTX, was a result of the generation of postmitotic tetraploid cells (rather than mitotically arrested cells), since at this time point, we simultaneously observed a massive accumulation of the enlarged multinucleated cells. To determine the percentage of the mitotic cells, a dual-color flow cytometry analysis using for example PI and mitotic phosphoprotein monoclonal-2 (MPM-2) or phospho-histone H3 antibody should be performed in the future studies. The above-mentioned events, all of which are related to MC, have previously been observed in various PTX-treated cancer cell lines [
57,
58], including A549 cell line [
59], as well as in the response to other microtubule-stabilizing drugs [
60]. Although, the final outcome of mitotic catastrophe is the cell death, the time in which cells die, the ultimate cause of the cell death as well the type of a cell death pathway they follow to die, may vary in depending on the genetic background of cells and the type of drug as well as the dosage and the duration of treatment [
61]. At this point, it should be emphasized that in an international consensus mitotic catastrophe is defined as an oncosuppressive mechanism occurring during or after a faulty mitosis leading to the cell death, which takes place via apoptosis or necrosis, rather than cell death executioner pathway itself [
62,
63]. It has been shown that paclitaxel induces the activation of spindle assembly checkpoint through the suppression of microtubule dynamics, leading to a prolonged mitotic arrest before anaphase, followed by apoptosis [
64] or eventually the mitotic exit due to either the checkpoint adaptation [
65] or perturbation [
57]. The resultant multinucleated G1-tetraploid cells subsequently arrest in postmitotic G1, as a result of the activation of a p53-dependent G1 checkpoint, whereas the arrest at the metaphase–anaphase transition and the mitotic slippage are most likely not mediated by p53 [
66,
67]. Those cells either succumb directly to apoptosis (mitotic catastrophe followed by apoptosis) or continue another round of the cell cycle as a consequence of the G1 checkpoint failure. The latter event, leads to further polyploidization and aneuploidization and eventually to the cell death [
57,
59]. Importantly, herein we demonstrated that PTX-induced cellular events associated with mitotic catastrophe was significantly potentiated by fisetin. Bearing in mind that A549 cells have a wild type p53 gene, one can ask why the post-slippage A549 cells did not arrest in the G1 postmitotic checkpoint, and instead entered another round of the cell cycle to form the polyploid progeny and how to account for the reported potentiation of PTX-induced mitotic catastrophe by fisetin. One plausible explanation comes from the observation that fisetin may act as a strong inhibitor of the spindle checkpoint that induces a rapid escape from microtubule drug-induced mitotic arrest [
68]. In the cited studies, fisetin at concentration of 30 µM caused a forced mitotic exit from the mitotic block triggered by nocodazole, taxol and monastrol in various human cancer cell lines, including A549 cells. Furthermore, there is evidence that not only p53 but also a prolonged spindle checkpoint-mediated mitotic arrest is required for the postmitotic G1 checkpoint function [
67]. The duration of mitotic arrest has been shown to be critical for the stabilization and activation of p53 [
55,
67]. Indeed, Vogel et al. have demonstrated that in spindle checkpoint compromised cells, mitotic arrest is shortened, resulting in subsequent endoreduplication, whereas extending the time of mitotic block in these cells inhibited endoreduplication [
67]. Based on the above findings, we can speculate that the potentiation of PTX-induced mitotic catastrophe by fisetin may be associated with the perturbation of the spindle assembly checkpoint.
There are also several studies that reported on other than necrosis, non-apoptotic mechanisms leading to the cell death following mitotic catastrophe. In accordance with our results, Kuwahara et al. have suggested that the giant multinucleated cells may die through the autophagic cell death [
69]. In our studies, mitotic catastrophe was not followed by apoptosis at any time points examined. Instead, we observed that the cells with the mitotic catastrophe-like phenotype were filled abundantly with autolysosomes, what allows us to assume that these cells could be eliminated through autophagy.
The role of autophagy in cancer therapy raised a paradox wherein, on one hand, it can represent a protective mechanism that sustains tumor cell growth and survival in the face of the toxicity of the cytostatic drugs or radiation, but on the other hand it may constitute an alternative form of the programmed cell death, named the autophagic type II cell death [
70]. In the first scenario, autophagy contributes to the treatment failure, thus its inhibition can re-sensitize previously resistant cancer cells to the cytotoxic action of chemotherapy or radiotherapy, concurring to beneficial treatment outcome [
54]. In the latter case, autophagy may be therapeutically desired, as it mediates the cytotoxic effect of anticancer drugs, leading to tumor cell demise [
71]. The dual role of autophagy, either as pro-survival or pro-death mechanism, creates the need to carefully examine the functional status of autophagy before the administration of autophagy-induced therapy. In other words, from a therapeutic perspective, it is extremely important to determine whether the increase in the autophagy level is a sign of responsiveness or resistance to the treatment [
72]. Hence, in the current studies, having established that FIS and/or PTX trigger autophagy in A549 cells, we then asked whether FIS and/or PTX-mediated autophagy may have a cytoprotective or detrimental function. Recent studies have shown that the impact of PTX on autophagy may be cell type-specific [
73], and several reports have revealed that PTX-induced autophagy in the A549 cells represents a self-defense mechanism that protects these cells against PTX-mediated apoptosis [
45]. Our observations were also consistent with the premise that autophagy induced by paclitaxel in the A549 cells is cytoprotective. In turn, to our knowledge, there is only one published study that examined the functional significance of FIS-triggered autophagy. That study has revealed that fisetin promotes the autophagic cell death in the PC3 prostate cancer cells [
74]. On the contrary, we demonstrated that FIS-elicited autophagy provides a survival advantage to the A549 cells and protects them against apoptosis induced by this flavonoid. It should be emphasized that fisetin produced the protective autophagic response in the A549 cells at in vivo achievable concentration (10 μM), whereas the autophagic cell death in the PC3 cells was induced by much higher doses of fisetin (40–120 μM) [
74]. These results seem to support a more and more common opinion that a potential utility of the dietary polyphenols in anticancer therapy lies in the synergistic combinations rather than in monotherapy. Interestingly, the conversion of the autophagic function from the cytoprotective form with FIS alone or PTX alone to the cytotoxic form upon the exposure of the cells to the combination of these compounds was found to occur in our experimental conditions. Indeed, when FIS and PTX were applied simultaneously to the A549 cells, Baf A1 pre-treatment resulted in the restoration of the cell viability to the level similar to what was observed with PTX treatment alone. Among the criteria adopted to define the cell death by autophagy, the demonstration that a pharmacologic or genetic suppression of autophagy prevents cell death, is believed to be the critical one [
75,
76]. This type of “autophagic switch” has been first demonstrated by Wilson et al., who have shown that autophagy can actually have dual functions in the same experimental system (breast tumor cells), acting both as a cytoprotective mechanism for radiation alone and a cytotoxic mechanism when radiation is accompanied by vitamin D or 1.25 dihydroxy vitamin D3 [
77]. The cited authors have implicated the autophagic cell death in the mechanism underlying the radiosensitization by 1.25 dihydroxy vitamin D3. In another study, Gewirtz’s group have utilized vitamin D and its analogue in an effort to improve the effectiveness of radiation therapy in non-small cell lung cancer [
78]. As in the breast tumor cell studies, the switch between a cytoprotective and cytostatic autophagy appeared to mediated the sensitization to radiation. However, the specific signaling pathways mediating this dual role of radiation-induced autophagy have not been established so far [
77,
78]. Despite the fact that we also did not investigate the molecular mechanism governing the “autophagic switch” in our experimental conditions, we presume that the significant intensity of the autophagy level in the response to the combination treatment could be, at least in part, a cause for such conversion. This assumption was based on the previous suggestions that a basal enhanced level of autophagy in tumor cells contributes to therapy resistance but a prolonged and excess induction may result in cell death by cellular self-degradation [
79,
80].
The concept of “molecular switches” has also been used in the literature to describe the crosstalk between apoptosis and autophagy [
80]. Although the functional relationship between these two processes has not been fully clarified, it has been suggested that autophagy and apoptosis may occur independently of each other, cooperate or antagonize each other [
81]. As mentioned above, in this paper we revealed that autophagy induced by FIS alone protects the A549 cells against FIS-promoted apoptosis. In this case, the cytoprotective function of autophagy was mediated through a negative modulation of apoptosis. However, there are also reports showing that the inhibition of a cytotoxic autophagy switched cell death to apoptosis in cancer cells [
82]. Therefore, we were curious about whether such a cross-regulation between autophagy and apoptosis might also occur in the A549 cells exposed to the combination of FIS and PTX. We did not observe a significant increase in the percentage of the apoptotic cells following autophagy blockage in the A549 cells co-treated with FIS and PTX, thus we suppose that in the case of the combined treatment, the autophagic and apoptotic cell death are not related to each other.