Background
Glioblastoma multiforme (GBM), also called glioblastoma, is the most common and aggressive primary brain tumor in adults. It is classified as a Grade IV brain tumor according to the World Health Organization (WHO) classification. Due to its aggressive biological behavior, diffuse infiltrative growth and central location, it has become one of the most challenging cancers of the central nervous system (CNS) [
1,
2]. The therapeutic approach to glioblastoma includes maximal safe resection surgery followed by radiation therapy plus concomitant and adjuvant chemotherapy [
1,
3]. Glioblastoma often show little to no response to conventional anti-cancer drugs such as temozolomide (TMZ) and it becomes resistant to apoptosis after a short period of treatment. This is especially true for invasive malignant glioma cells that are resistant to pro-apoptotic chemotherapy and radiotherapy [
4‐
6]. The current 5-year overall survival of grade IV GBM patients using radiotherapy with concomitant TMZ treatment is only 9.8%. The resistance to chemotherapy remains a critical issue in the failure of successful treatment of cancer, especially in GBM patients.
Tumor-induced hypoxic barriers, existence of cancer stem cells, enhanced membrane transporter activities and other mechanisms may be important factors in drug resistance [
7]. One way in which cancer cells can achieve resistance to anti-cancer drugs is by up-regulating the ATP-binding cassette transporter proteins which are responsible for the efflux of anti-cancer molecules from the intracellular compartment [
8]. Another mechanism of resistance to chemotherapy involves the hypoxic conditions in the central portions of the tumor and the resultant over-expression of HIF-1α that enhances a cell’s tolerance to insults including anti-cancer drugs. Furthermore, hypoxic cells may be less proliferative and thus less responsive to anti-cancer drugs that target rapidly proliferating cells [
9]. We hypothesize that a new therapeutic approach that can simultaneously trigger more than one cell death program/mechanism may have a better chance of overcoming the drug resistance of glioblastoma cells.
Na
+/K
+-ATPase, also known as the Na
+ pump or more accurately the Na
+/K
+ pump, is a ubiquitously expressed transmembrane transporter composed of tetramers of alpha and beta subunits. A normal activity of Na
+/K
+-ATPase is essential for maintaining ionic homeostasis, cellular pH, and cell volume [
10]. The catalytic alpha subunit is a large polypeptide of ~1,000 amino acid residues, which catalyzes the ion-dependent ATPase activity and carries the binding sites for ATP and the specific inhibitor ouabain. The beta subunit is a smaller polypeptide of about 300 residues, which regulates conformational stability and activity of the alpha subunit. The Na
+/K
+ pump is critical in maintaining high extracellular Na
+ (~145 mM) and high intracellular K
+ (~150 mM) by pumping Na
+ ions out of the cell and importing K
+ ions into the cell [
11]. By doing so, these Na
+/K
+ pumps maintain a physiological electrochemical gradient that is essential for cell survival and for many cellular activities. Consistent with its pro-life role, Na
+/K
+-ATPase is highly expressed in cancer cells including glioblastoma cells [
12‐
14]. The Na
+/K
+ pump activity increases during the course of malignant cell transformation [
15]. This increased expression and elevated activity suggest that Na
+/K
+-ATPase may serve as a biological marker and a therapeutic target of cancer cells. Along with the identification of its high expression in cancer cells and its critical roles in cell survival, proliferation, adhesion and migration, the clinical potential of Na
+/K
+-ATPase modulators such as cardiotonic steroids or digitalis in oncology has drawn increasing attention in recent years [
12,
16]. Several cardenolides have been shown to display
in vitro antitumor activities against various types of cancer cells [
17‐
21], including glioma cells [
22,
23].
Cardiac glycosides including digoxin, marinobufagenin, telocinobufagin and ouabain, represent a group of compounds isolated from plants and animals [
24]. Endogenous ouabain-like substances were also identified as a hormone or stress signal that responds to exogenous and endogenous stimuli such as physical exercise, stress, hypertension, hypoxia/ischemia, among many others [
24]. These cardiac glycosides have been used in clinical therapies of heart failure and atrial arrhythmia for many years [
19,
24]. Meanwhile, digoxin acts as a specific neuroblastoma growth inhibitor in mice grafted with the neuroblastoma cell lines SH-SY5Y and Neuro-2a [
25]. Blocking Na
+/K
+-ATPase using the exogenous cardiac glycoside ouabain is cytotoxic to a variety of cancer and non-cancerous cells; the sensitivity depends on the expression level of the functional Na
+/K
+ pump and dosage used [
26‐
29]. Ouabain and the specific knockdown of the Na
+/K
+-ATPase alpha subunit inhibits cancer cell proliferation and migration [
13,
22], sensitizes resistant cancer cells to anoikis and decreases tumor metastasis [
30]. However, the cellular/molecular mechanisms underlying the cytotoxic effect of cardiac glycosides in tumor cells have been poorly defined. We noticed that blocking Na
+/K
+-ATPase has two direct and marked impacts on the cellular ionic homeostasis: increased intracellular Na
+ concentration and decreased intracellular K
+ concentration. The majority of previous studies have been focused on the intracellular Na
+ increase and the consequent intracellular Ca
2+ increases due to the enhanced reversal operation of the Na
+-Ca
2+ exchanger [
31‐
33]. On the other hand, increasing evidence from our groups and other’s have demonstrated that, in many noncancerous neuronal and non-neuronal cells, depletion of intracellular K
+ is a prerequisite for apoptotic cell shrinkage, activation of caspases and initiation of apoptotic programing [
34‐
36]. Consistently, attenuating the outward K
+ current with tetraethylammonium or elevating extracellular K
+ prevented apoptosis while treatment with the K
+ ionophore valinomycin induced apoptosis [
37,
38], There is also evidence that cytosolic Ca
2+ levels may not directly regulate apoptotic cell death [
11,
39]. Therefore, besides the regulation by a series of apoptotic genes, apoptosis is regulated by an ionic mechanism closely associated with K
+ homeostasis [
11,
39,
40]. Up to now, little attention has been paid to the intracellular K
+ loss in cancer cells.
We previously demonstrated in different noncancerous cells that inhibition of Na
+/K
+-ATPase induced a mixed form of cell death composed of concurrent necrotic and apoptotic components in the same cells, which we named hybrid death [
41]. Specifically, the increases in intracellular Na
+ and Ca
2+ are associated with necrosis and K
+ depletion is linked to apoptosis. These events may take place simultaneously and trigger activation of multiple signaling pathways. The identification of hybrid cell death was also based on cellular/sub-cellular morphological changes, gene expression, and alterations in intracellular signaling pathways [
11,
41].
In this investigation, we tested the main hypothesis that inhibition of Na+/K+-ATPase could disrupt K+ and Na+/Ca2+ homeostasis and subsequently induce hybrid death in human glioblastoma cells. Ouabain was tested because of its high selectivity in blocking NA+/K+-ATPase. We also tested whether inhibition of Na+/K+-ATPase or deletion of its specific subunit could enhance the sensitivity of glioblastoma cells to TMZ in the drug-resistant T98G glioblastoma cells.
Methods
Cultures of human glioblastoma cells
Human glioblastoma cell lines LN229 and T98G (kindly supplied by Dr. Erwin G. Van Meir, Emory University, Winship Cancer Institute) were maintained in Dulbecco’s modified Eagle’s media supplemented with 10% fetal bovine serum (FBS).
Ethics statement
LN229 and T98G cells are established cell lines from glioblastoma of anonymous patients and are commercially available. These cells have been extensively used in cancer research and related information is publically available. Therefore, their use was not classified as human subject research, and no Institutional Review Board approval was needed.
Cell viability assay by MTT spectrophotometry
Cells were cultured at a density of 3000 cells/well in 96-well plates at 5% CO2 and 37 °C. At 70% confluence, cells were treated with either ouabain or other drugs. At selected time points, 3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyltetrazolium (MTT) was added at a final concentration of 0.5 mg/mL. After 4 hrs incubation, the reaction was stopped by adding a solubilization buffer (10% SDS, 10 μM HCl). After the mixture was incubated at 37°C for 2 hrs, the relative optical density for each well was determined at 570 nm by a microplate spectrophotometer (Bio-Tek, Winooski, Vermont).
Apoptosis detection by flow cytometry
Phosphatidylserine (PS) membrane translocation and caspase-3 activation were determined by flow cytometry using FITC Annexin V Apoptosis Detection Kit (BD Pharmingen, San Diego, CA). Cells were treated with 1 μM ouabain or 10 μM valinomycin for selected time points and then washed twice with phosphate-buffered saline (PBS). Staining procedures followed the standard protocol provided by the manufacturer. Briefly, 1 × 106 cells were resuspended in 1 mL of binding buffer and then the 100 μL cell suspension was incubated with 1 μL Annexin-V-FITC and 1 μL propidium iodide (PI) for 15 min at room temperature in the dark. Propidium iodide was used as a marker of necrosis. The population of Annexin V-positive cells was evaluated by a BD Biosciences LSR II flow cytometer and analyzed by FlowJo Version 7.6 software (Tree Star, Ashland, OR).
Western blotting analysis
Cells were lysed in protein lysis buffer (25 mM Tris–HCl (pH 7.4), 150 mM NaCl, 5 mM EDTA, 0.1% SDS, 2 mM sodium orthovanadate, 100 mM NaF, 1% Triton, leupeptin, aprotinin, and pepstatin) containing protease inhibitor (Sigma, St Louis, MO). Protein concentration was determined using the Bicinchoninic Acid Assay (Sigma). 30 μg protein samples were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) in a Hoefer Mini-Gel system (Amersham Biosciences, Piscataway, NJ) and transferred onto a PVDF membrane (BioRad, Hercules, CA). The blotting membrane was incubated with primary antibodies overnight: Bcl-2 and cleaved Caspase-3 (1:1000, Cell Signaling, Danvers, MA), Cytochrome c and Caspase-9 (1:500, Millipore, Billerica, MA), β-actin (1:5000, Sigma). The blots were incubated for 1 hr at room temperature with anti-mouse or anti-rabbit alkaline phosphatase-conjugated IgG antibodies (1:2000, Promega, Madison, WI). The signals were detected by the addition of 5-bromo-4-chloro-3-indolylphosphate/nitroblue tetrazolium (BCIP/NBT) solution (Sigma) and quantified and analyzed by the NIH imaging software Image J (NIH, Bethesda, MD). The level of protein expression was normalized to β-actin controls. The value of protein levels was designed as 1 in the control group. The results were expressed as mean proportion of the control group values.
Immunocytochemistry staining
Cells were fixed with 4% paraformaldehyde and then treated with 0.2% Triton-X 100 for 5 min. After blocking with 1% fish gel for 1 hr, cells were incubated with primary anti-body AIF overnight (1:500, Millipore). Cells were then incubated with secondary antibody Cy3-conjugated anti-rabbit IgG (1:500, Invitrogen, Carlsbad, CA) for 1 hr at room temperature. Nuclei were stained with Hoechst 33342 (1:20000, Invitrogen). Staining was visualized by fluorescent and confocal microscopy (BX61; Olympus, Japan).
Fluorescent measurement of the mitochondrial membrane potential
Cells were treated with ouabain or valinomycin for 6 hrs and then loaded with 200 nM TMRM (Molecular Probes, Eugene, OR) for 30 min at 5% CO2 and 37°C in the dark. Prior to imaging, cells were washed with DMEM medium twice. Fluorescent images were captured by a fluorescent microscope (Leica DMIRB, Germany) and fluorescent intensity was measured by the NIH imaging software Image J.
Cellular ion measurements
Intracellular K+ content was measured using the cell permeant potassium indicator PBFI-AM (Invitrogen, Molecular Probes). Cells were washed with HBSS and then loaded with 5 μM PBFI and 10 μM F-127 for 40 min at 5% CO2 and 37°C in the dark. Cells were washed with HBSS three times before imaging. Measurements were made by exciting PBFI at 340 nm while monitoring emission at 500 nm using a fluorescent microscope (Leica DMIRB, Germany) and the fluorescence intensity was measured using the NIH imaging software Image J.
Intracellular Na+ content was measured using the cell permeant sodium indicator SBFI-AM (Invitrogen, Molecular Probes). Cells were washed with HBSS and then loaded with 5 μM SBFI and 10 μM F-127 for 40 min at 5% CO2 and 37 °C in the dark. After three HBSS washes, fluorescent imaging was carried out at room temperature using an inverted fluorescence microscope (Olympus IX81, Olympus America Inc., Center Valley, PA). Measurements were made by exciting SBFI at 340 nm while monitoring emission at 520 nm using a CCD camera. The imaging data were recorded with a digital camera Hamamatsu ORCA-ER (Hamamatsu Photonics K.K., Japan) and software Slidebook 4.1 for Windows (SciTech Pty Ltd., Australia).
Intracellular free Ca2+ was measured using the cell permeate Ca2+ sensitive dye Fluo-4-AM (Invitrogen; 5 μM in 100 μl HEPES buffered solution) for 50 min at 5% CO2 and 37°C in the dark. Fluo-4 epifluorescence was excited at 480 nm light and images were obtained at 520 nm. The imaging data were collected by the same fluorescence microscopy system described for sodium imaging.
Cell volume assay
Cells were trypsinized after drug treatments. A 100 μL cell suspension of each sample was taken by Millipore Scepter™ Handheld Automated cell counter (Millipore). Cell volume was measured and analyzed by Scepter Software Pro 2.0.
Electron microscopic examination of ultrastructural changes
Cultures in 35 mm dishes were fixed in glutaraldehyde (1% glutaraldehyde, 0.1 M sodium cacodylate buffer, pH 7.4) for 30 min at 4 °C, washed with 0.1 M sodium cacodylate buffer, and post-fixed in 1.25% osmium tetroxide for 30 min. The staining and electron microscopy was performed at the Robert P. Apkarian Integrated Electron Microscopy Core (Emory University, Atlanta, GA).
Cytochrome c release assay
Cells were harvested by centrifugation at 200 g for 10 min at 4°C. Mitochondrial and cytoplasmic proteins were isolated using the Mitochondria Isolation Kit (Thermo Scientific, Rockford, IL) according to the kit’s instructions. Cytochrome c released from the mitochondria was detected by Western blot.
Knockdown of the Na+/K+-ATPase α3 subunit
Na+/K+-ATPase α3 stealth RNAi™ siRNA duplex oligoribonucleotides were synthesized by Invitrogen. The sequences of the siRNA duplex were designed by Invitrogen Block-iT RNAi Designer:
Forward: 5′-ACG ACA ACC GAU ACC UGC UGG UGA U-3′
Reverse: 5′-AUC ACC AGC AGG UAU CGG UUG UCG U-3′
The T98G cells were transfected with Na+/K+-ATPase α3 stealth RNAi™ siRNA or stealth RNAi™ siRNA negative control (Invitrogen) using Lipofectamine™ 2000 (Invitrogen) according to the manufacture’s instruction. Briefly, 0.5 × 105 T98G cells were plated in a 6-well plate and cultured overnight. 250 pmol siRNA duplex or siRNA negative control was mixed with 10 μL lipofectamine reagent in the serum free Opti-MEM medium and transfected the T98G cells for 6 hrs. 48 hrs later, the cells were harvested for the reverse transcriptase-polymerase chain reaction (PCR) to detect the expression of the α3 subunit.
Reverse transcriptase-polymerase chain reaction
Total RNA was extracted from human glioblastoma cells using the Trizolreagent (Invitrogen) according to the procedure suggested by the manufacturer. For cDNA synthesis, 1 μg of total RNA were reverse transcribed into cDNA using RNA to cDNA High Capacity kit (Invitrogen) and PCR was performed in a PTC-150 Minicycler (MJ Research Inc., Watertown, MA) with primer sets for target genes and a housekeeping gene, ribosomal protein large subunit 19 (RPL19) as an internal control for both cDNA quantity and quality. PCR primers, as listed below, were designed according to the sequences in a previous report [
42]. All the primers were designed to amplify products that covered one or more exons.
Na+/K+-ATPase á1 forward 5′-GAA AGA AGT TTC TAT GGA TG-3′ reverse 5′-ATG ATT ACA ACG GCT GAT AG-3′
Na+/K+-ATPase á2 forward 5′-AGA GAA TGG GGG CGG CAA GAA G-3′ reverse 5′-TGG TTC ATC CTC CAT GGC AGC C-3′
Na+/K+-ATPase á3 forward 5′-CCT CAC TCA GAA CCG CAT GAC-3′ reverse 5′-TTC ATC ACC AGC AGG TAT CGG-3′
RPL19 forward 5′-GAG TAT GCT CAG GCT TCA GA-3 reverse 5′-TTC CTT GGT CTT AGA CCT GC-3′
After an initial phase at 94°C for 2 min, amplification of α1 was run for 31 cycles and α 2 and α 3 for 40 cycles. The cycles consisted of denaturation at 94°C for 1 min, annealing at 50°C for 45 s for α 1 and 54°C for 1 min for α 2 and α 3, extension at 72°C for 1 min, and a final extension of 7 min at 72°C at the end of the program. The PCR products (25 μL) in TAE buffer were loaded onto 1.5% agarose gel and run at 36 V for 90 min. The Gel was scanned for quantitative analysis using the UnScan It program (Silk Scientific Inc., Orem, UT). The ratio of target gene to housekeeping gene, RPL19, was calculated.
Chemicals
The caspase inhibitor Z-VAD-FMK was purchased from Enzyme Systems Products [
42]. BAPTA-AM was from Tocris Bioscience (Bristol, UK). Ouabain and valinomycin were from Sigma Aldrich (St. Louis, MO) [
42].
Statistical analyses
One-way ANOVA followed by Tukey post-test was performed for multiple group comparisons. Two-way ANOVA followed by Bonfferoni post-tests was used for multiple groups with multiple time points. Data were shown as mean ± SEM. Changes were identified as significant if p value was less than 0.05.
Discussion
The present investigation shows for the first time in cancer cells that blocking or down regulation of Na+/K+-ATPase induces a cell death phenotype that has characteristics of both apoptosis and necrosis. We show that disruption of K+ homeostasis is a key factor in the induction of apoptosis in human glioblastoma cells. Contrary to what is widely believed that a cell may either die from apoptosis or necrosis, ouabain induced cell death does not have typical features of apoptosis or necrosis. Although strong apoptotic features such as phosphatidylserine translocation, caspase activation and Bcl-2 reduction were detected, ouabain-induced cell death in these cells exhibited necrotic features as well, including cell swelling, mitochondrial injury, [Ca2+]i increase, deteriorated cellular organelles and breakdown of the plasma membrane. Consistent with the multifaceted ionic changes, ultrastructural alterations include both necrotic and apoptotic features. Since much higher expression of the Na+/K+-ATPase α2/α3 subunits exists in drug-resistant glioblastoma cells compared to drug-sensitive and normal human glial cells, our data indicate that the α2 and/or α3 subunits are potential targets for anti-cancer treatments. This was demonstrated by the different ouabain induced dose-responses of TMZ-sensitive LN229 cells, TMZ-resistant T98G cells and normal human astrocytes. This principle was also specifically demonstrated in the subunit knockdown experiment. Furthermore, the hybrid cell death mechanism of multiple targets helped to overcome the TMZ resistance of glioblastoma cells. Taken together, this investigation provides a better understanding of the ionic and cellular mechanisms underlying ouabain-induced cell death in human glioblastoma cells and suggests a potential therapeutic target for glioblastoma treatment.
We noticed that ouabain-induced apoptotic changes in LN229 cells were not typical of those caused after valinomycin exposure. For example, valinomycin caused gradual and progressive cell volume shrinkage while ouabain did not show the same volume change. Instead of apoptotic cell shrinkage, ouabain causes an initial cell swelling followed by a gradual decrease in cell volume. According to our previous research in non-cancerous cells and the data in this investigation, the initial cell swelling is attributable to intracellular Na
+ and Ca
2+ accumulation, while the gradual volume decrease is associated with the slower process of intracellular K
+ depletion [
36]. The early cell swelling followed by cell volume reduction during ouabain exposure was analogous to the hybrid cell death model (concurrent apoptosis and necrosis in the same cell) that we reported before in cortical neurons [
41]. The acute increase in Ca
2+/Na
+ accompanied with a gradual K
+ depletion in glioblastoma cells are consistent with the unique time-dependent cell volume alterations. The ionic disruptions also can be linked to necrotic and apoptotic events in these cells.
Although extensive research has been focused on Na
+/K
+-ATPase in tumor cells in the past few years, virtually all investigations assume apoptosis is the underlying mechanism for its anti-cancer effect. In the effort to identify therapeutic targets, many studies have focused on the α1 subunit while only very few reports have looked at the role of the α3 subunit [
50]. To understand the ionic mechanism mediating the anti-cancer property of cardiac glycosides, many research reports examined intracellular Na
+ accumulation and the Na
+-dependent Ca
2+ increases (e.g. Ca
2+ oscillations) via enhanced reverse operation of the Na
+/Ca
2+ exchanger [
31‐
33]. This research focus, however, overlooks the most abundant intracellular cation K
+ and disregards the K
+ role in apoptotic cell volume regulation and in the induction of apoptotic cascade. This was most likely due to the influence of many early investigations that simply linked K
+ gradient to membrane potential regulation and the consequent influence on Ca
2+ influx [
42]. Accumulating evidence in recent years, however, has endorsed that a pro-apoptotic K
+ efflux is an integrated cellular event in the early stage of apoptosis in non-cancerous neuronal, glial and peripheral cells [
34‐
36]. We now show new evidence that this K
+-mediated apoptotic mechanism similarly takes place in glioblastoma cells.
In the present investigation, low concentrations of ouabain and the induced hybrid cell death mechanism effectively sensitize drug-resistant glioblastoma cells to the anti-cancer effect of TMZ. It is important to note that, in this anti-cancer strategy, only a low concentration of ouabain is needed to sensitize the drug-resistant cancer cells. At the level of Na
+/K
+ pump down-regulation, survival of normal neuronal cells or even drug-sensitive glioblastoma cells are not affected. This is most likely due to the fact that expression of Na
+/K
+-ATPase is selectively and markedly higher in drug-resistant cells as shown in T98G cells. This selectivity supports a possible clinical significance of targeting Na
+/K
+-ATPase in glioblastoma treatment. Moreover, ouabain can pass through the blood brain barrier [
51], which facilitates clinical applications of ouabain-like drugs in the potential treatment for brain tumors.
We previously showed that the sublethal low concentration of ouabain (0.1 μM) could markedly enhance the vulnerability of neuronal cells to pathological insults of glutamate, ceramide and β-amyloid [
43]. The present study further suggests that ouabain and likely other cardiac glycosides can sensitize glioblastoma cells to conventional chemotherapy drugs via activation of multiple cell death mechanisms. While most cytotoxic anticancer drugs including alkylating agents such as TMZ have been shown to induce apoptosis [
52], cancer cells could potentially be targeted by other death mechanisms such as necrosis, autophagy, senescence, or mitotic catastrophe [
53]. To overcome the resistance of glioblastoma cells to TMZ, it has been postulated that it is possible to combine TMZ with other drugs to enhance glioblastoma cell response to TMZ cytotoxicity. In a randomized and double-blind trial for glioblastoma multiforme, adding chloroquine to conventional treatment including TMZ showed mediocre results, probably because of the small sample size [
54]. Other combination regimens include adding epigallocatechin gallate (EGCG) and oxygen-diffusion enhancing trans-sodium crocetinate (TSC); the later is now in phase I/II clinical trials [
55,
56].
In recent years, the mixed form of cell death of similar or partly overlapped definitions have been reported and are given different names such as aponecrosis and necroptosis [
57‐
59]. We prefer the term hybrid death since it covers different known and novel cell death mechanisms. While the results from this investigation are encouraging, we acknowledge the need for further experiments to fully characterize the nature of cell death and delineate the detail dose–response relationship of Na
+/K
+-ATPase activity and viability of glioblastoma cells. As an
in vitro investigation, our data provide some initial mechanistic information. The in vivo verification of the mechanism in an anti-cancer therapy will be essential. In addition to ouabain, application of other cardiac glycosides in the anti-cancer treatment should be tested since these clinical drugs have been used for the treatment of congestive heart failure and cardiac arrhythmia for many years. In addition to its effect on ionic homeostasis, Na
+/K
+-ATPase has been proposed to have a distinct function of directly regulating a number of intracellular signaling pathways [
60]. Regulation of these signals may have important impacts on cell viability status and contribute to the anti-cancer effect of Na
+/K
+-ATPase inhibition. For example, it will be interesting to investigate the role of the PI3K/Akt/mTOR pathway that is regulated by Na
+/K
+-ATPase and which plays a major role in cancer cell survival [
61].
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
DC performed cell culture procedures, cell death assays, flow cytomatry assay, and Western blot analysis. MS participated in cell culture and cell death experiments. OM participated in molecular biological experiments and helped to draft the manuscript. SY conceived of the study, developed the hypothesis and wrote the manuscript. All authors read and approved the final manuscript.