Targeting mitochondria by α-tocopheryl succinate overcomes hypoxia-mediated tumor cell resistance to treatment

Abstract

Rapidly proliferating tumor cells easily become hypoxic. This results in acquired stability towards treatment with anticancer drugs. Here, we show that cells grown at 0.1 % oxygen are more resistant towards treatment with the conventionally used anticancer drugs doxorubicin and cisplatin. The stimulation of apoptosis, as assessed by the number of cells in the SubG1 fraction of the cell cycle, release of cytochrome c into the cytosol, activation of caspase-3, and cleavage of PARP, was markedly suppressed under low oxygen content or when hypoxia was mimicked by deferoxamine. Hypoxia or deferoxamine treatment was accompanied by stabilization of the hypoxia-inducible factor (HIF-1). The downregulation of HIF-1 using siRNA technique restored cell sensitivity to treatment under hypoxic conditions to the levels detected under normoxic conditions. In contrast to cisplatin or doxorubicin, α-tocopheryl succinate (α-TOS), a compound that targets mitochondria, stimulated cell death irrespective of the oxygen concentration. Moreover, under hypoxic condition cell death induced by α-TOS was even enhanced. Thus, α-TOS can successfully overcome resistance to treatment caused by hypoxia, which makes α-TOS an attractive candidate for antitumor therapy via mitochondrial targeting.

Introduction

Apoptosis and cancer are regarded as antagonistic processes in cell physiology [1]. Apoptosis may be involved in spontaneous regression of tumors, whereas defects in apoptosis programs may contribute to tumor progression and resistance to treatment. Stimulation of apoptosis is a powerful tool in tumor cell elimination [2]. Different signals trigger distinct cell death pathways, which often merge at the mitochondria, a common “regulator” of this multistep process. Permeabilization of the outer mitochondrial membrane (OMM) and release of proteins from the intermembrane space represents a crucial event in many models of apoptosis [3].

Despite the heterogeneity of tumors, which dictates an individual approach to anticancer treatment, almost all tumor cells demonstrate enhanced uptake and utilization of glucose even under aerobic conditions, a phenomenon known as the Warburg effect. This is dictated by the fast proliferation of tumor cells, which easily makes rapidly growing tumors hypoxic due to the inability of the local vasculature to supply an adequate amount of oxygen. Owing to the inability of the mitochondria to provide enough ATP for cell survival under hypoxic conditions, tumor cells must upregulate the glycolytic pathway [4]. This is mediated by the induction of HIF-1 [5], which stimulates the key steps of glycolysis and regulates genes that control angiogenesis, cell survival, and invasion.

The upregulation of glycolysis under hypoxic conditions causes the suppression of mitochondrial activity—the “silencing” of mitochondria [4]. Due to the glycolytic shift, the contribution of mitochondria to ATP supply in the majority of tumors is suppressed. Surprisingly, however, tumor cells remain glycolytic also after restoration of the oxygen supply. In fact, the amount of glucose taken up by cancer cells exceeds their bioenergetic demand. It has been suggested that the excessive glycolysis in tumors is required to support cell growth [6]. One of the most important consequences of mitochondrial silencing in tumors is the attenuation of mitochondrial pathways in apoptosis, which largely contributes to tumor resistance to treatment. Considering that mitochondria are a key participant in apoptosis, two different therapeutic approaches can be suggested for the revival of mitochondria-dependent apoptotic pathways: (a) to suppress glycolysis and stimulate the mitochondria, in order to restore metabolic pathways characteristic of non-malignant cells; and (b) to weaken the mitochondria, causing OMM permeabilization and stimulation of mitochondria-dependent cell death pathways.

Here, we analyze the effect of hypoxia on the sensitivity of the cell to treatment with conventionally used anticancer drugs. We also evaluate how the protective effect of hypoxia can be overcome by mitochondrial targeting.

Materials and methods

Cell culture and hypoxia treatment

Human colon carcinoma HCT116 cells and human osteosarcoma U2OS cells were cultured in DMEM supplemented with 10 % heat-inactivated fetal bovine serum, penicillin/streptomycin (100 U/ml), and 0.11 mg/ml pyruvate at 37 °C under a 20 % O₂ humidified atmosphere. Hypoxia was established in a New Brunswick Galaxy 48R CO₂ Incubator (New Brunswick Scientific Co., Inc, Edison, NJ, USA), which allowed the maintenance of a 0.1 % O₂ level. Control cells were incubated under normal culture conditions. Cells were maintained in a logarithmic growth phase for all experiments. In some experiments, hypoxia was followed by re-oxygenation at 20 % O₂ for 6, 24, and 48 h. Apoptosis was induced by 1 μM doxorubicin, or 20 μg/ml cisplatin, or 100 μM α-TOS under hypoxic or normoxic conditions. Cells were harvested after 48-h (cisplatin and doxorubicin) or 24-h treatment (α-TOS).

Western blotting and fractionation

Analysis of the level of the proteins of interest was performed by Western blotting (WB) using V3 Western Workflow Complete System. Briefly, samples were trypsinized, incubated in RIPA buffer for 20 min, and centrifuged at 16,000 × g for 15 min at 4 °C. Protein concentration was measured by the Bradford Assay Kit (Thermo Scientific, Waltham, MA, USA) according to the manufacturer’s instructions. Samples were mixed with Laemmli’s loading buffer, boiled for 5 min, and subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE, 12 % gel) at 100 mV, followed by blotting onto nitrocellulose membranes for 30 min at 25 V using the Trans Blot Turbo Transfer System (Bio-Rad, Hercules, CA, USA). Membranes were blocked for 1 h with 5 % nonfat milk in phosphate-buffered saline (PBS) at room temperature and subsequently probed overnight with appropriate antibodies. The following antibodies were used: anti-HIF-1α (Cell Signaling, Danvers, MA, USA), anti-PARP (BD Biosciences, San Jose, CA, USA); anti-cytochrome c (BD Biosciences). The membranes were rinsed and incubated with a horseradish peroxidase-conjugated secondary antibody (1:5,000) and visualized with ECL (Amersham Biosciences, Piscataway, NJ, USA) and Chemi-Doc (Bio-Rad, Hercules, CA, USA). For the cytochrome c release assay, control cells and cells treated with anticancer drugs were spun down (200 × g, 5 min) and resuspended in fractionation buffer (pH 7.0–7.2) containing 150 mM KCl, 1 mM MgCl₂, 5 mM Tris, 0.5 mM EGTA, and 0.01 % digitonin for plasma membrane permeabilization. After 10 min of incubation at room temperature, the cells were vortexed and spun down for 5 min at 10,000 × g in order to obtain cytosolic and membrane fractions. The supernatant was gently removed and the pellet containing mitochondria was resuspended in the same amount of the buffer. The samples were mixed with Laemmli buffer and used for Western-blot analysis.

Assessment of hypoxia response element activity

The reporter construct, pGL3PGK6TKp, containing six copies of the hypoxia response element (HRE), was a kind gift from Prof. Maria Alfonsina Desiderio (University of Milan, Milan, Italy). Subconfluent HCT116 cells were transiently transfected using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) at a ratio recommended by the manufacturer, in Opti-MEM without antibiotic/antimicotic. The Renilla luciferase vector was co-transfected with each construct as a control for transfection efficiency. After overnight incubation at 37 °C, the cells were treated with cell death inducers and placed under hypoxic conditions for 48 h. At the end of the treatments, the cells were harvested in Passive Lysis Buffer from the Dual-Luciferase Reporter Assay Kit (Promega, Fitchburg, WI, USA). Luciferase activities were measured using the Zenyth-3100 Microplate Multimode Detector (Anthos Labtec Instruments GmbH, Salzburg, Austria) according to Promega’s protocol. Then, the ratio of firefly/Renilla luciferase activity was calculated.

RNA interference

siGENOME human HIF-1α siRNA duplex specific to human HIF-1α was purchased from Thermo Scientific. HCT116 and U2OS cells were transfected with 20 pmol (final concentration) HIF-1α siRNA duplex using Lipofectamine RNAi MAX reagent (Invitrogen); 4 h after transfection, the cells were subjected to hypoxic conditions and incubated for the indicated time periods with or without cell death inducers.

Cell cycle analysis

For cell cycle analysis, cells were harvested at the indicated time points and fixed with 70 % ethanol (final concentration) for at least 60 min on ice, rinsed in PBS, and stained in 500 μl of a solution containing 50 μg/ml propidium iodide (Sigma, St. Louis, MO, USA) and 25 μg/ml RNase A (Sigma) for 15 min. Data were acquired using a BD FACS CantoII flow cytometer (BD Biosciences) and analyzed using FACS Diva software.

Measurement of caspase activity

Cleavage of the fluorogenic peptide substrate Ac-DNLD-AMC (PeptaNova, Sandhausen, Germany) was measured using a fluorometric assay. Cells were harvested and washed with PBS. After centrifugation, they were resuspended in PBS at a concentration of 2 × 10⁶ cells/100 μl; 25 μl of the suspension was added to a 96-well plate and mixed with the appropriate peptide substrate dissolved in a standard reaction buffer (100 mM HEPES, 10 % sucrose, 5 mM DTT, 0.001 % NP-40, and 0.1 % CHAPS, pH 7.2). Cleavage of the fluorogenic peptide substrate was monitored by AMC liberation in a VarioScan Flash multimode detector (Thermo Scientific) using 380-nm excitation and 460-nm emission wavelengths.

Measurement of ATP level

ATP content was measured using an ATP determination kit (Sigma) according to the manufacturer’s protocol. Briefly, cells were harvested, counted, and lysed in ATP-releasing buffer (Sigma). The lysates were mixed with an equal volume of ATP Assay Mix containing luciferase and luciferin. After mixing, the samples were immediately placed in a Zenyth-3100 Microplate Multimode Detector (Anthos Labtec Instruments GmbH, Salzburg, Austria) to assess the luminescence. As a standard, a known concentration of ATP was used.

Analysis of oxygen consumption

Cellular respiration was monitored using an oxygen electrode (Hansatech Instruments, Norfolk, UK) and analyzed using OxygraphPlus software (Hansatech Instruments, Norfolk, UK). Briefly, cells were harvested, counted, and centrifuged. The pellet, containing five million cells, was resuspended in 300 μl of the medium in which the cells were growing, and transferred into the Oxygraph chamber. The chamber was closed and basal respiration was measured for 3–4 min. To assess the maximum capacity of the respiratory chain, mitochondria were uncoupled by 5 μM carbonyl cyanide 3-chlorophenylhydrazone (CCCP).

Statistical analysis

Results are presented as mean ± standard deviation (SD). Where appropriate, comparisons between experimental results were made using the Mann–Whitney U test at p ≤ 0.05.

Results

The incubation of cells under hypoxic conditions (0.1 % oxygen) for 24 h suppressed basal cellular respiration by almost fourfold, and respiration stimulated by CCCP by 2- to 2.5-fold (Fig. 1a). Subsequent reoxygenation caused a time-dependent restoration of mitochondrial oxygen consumption. This is indicative of reversible alteration of metabolic pathways during hypoxia and subsequent reperfusion. Analysis of ATP content in these cells revealed that the inhibition of respiration attenuated ATP level by 40–45 % (Fig. 1b). The alteration of metabolic pathways under hypoxic conditions is, in many instances, mediated by stabilization of HIF-1. Under normoxic conditions, the HIF-1α subunit undergoes proteosomal degradation. Hypoxia suppresses prolyl hydroxylases, oxygen- and iron-dependent enzymes that normally regulate proteosomal degradation of HIF-1α, stabilizing this transcriptional factor [7]. Indeed, hypoxic conditions caused time-dependent accumulation of HIF-1 in cells (Fig. 1c), as detected by WB with an antibody specific to HIF-1. Maximal level of this protein was detected between 4 and 8 h; at 24 h, the level of HIF-1 decreased to the control level. Two bands appeared on the blot; the slower migrating line is a phosphorylated form of HIF-1α. Analysis of HIF-2α did not reveal any expression in HCT116 cells, confirming that HCT116 cells express predominantly HIF-1α [8].

Fig. 1.

Effect of hypoxia on cellular bioenergetics and Hypoxia Inducible Factor (HIF-1α) stabilization. Hypoxia (a) or deferoxamine, DFO, (f) suppresses cellular respiration and ATP content (b, g), and stabilizes HIF-1α (c, d). e The HIF-1-responsive element, assessed as relative luciferase activity (fold increase). Experiments were performed at least three times, and representative blots (c, d) are shown. *p < 0.05, **p < 0.01 versus control

Similar results were obtained when cells were incubated in the presence of 100 μM deferoxamine (DFO), an iron chelator that is widely used as a hypoxia-mimicking agent [9]. DFO induced time-dependent stabilization of HIF-1 (Fig. 1d). Stimulation of HIF-1 expression induced by hypoxia or DFO was also confirmed by measuring HIF-1-regulated signaling pathway activity using a dual-luciferase reporter assay system (Fig. 1e). Similar to hypoxia, DFO suppressed respiration (Fig. 1f) and attenuated cellular ATP content (Fig. 1g).

It should be noted that hypoxic conditions themselves increased cell number in the SubG1 phase to certain extent; under normoxic conditions, the number of cells in the SubG1 phase was around 1 %, whereas after exposure to hypoxia it increased to 6–7 %. Comparison of cell death induced by conventional anticancer drugs, doxorubicin or cisplatin, under hypoxic and normoxic conditions revealed that hypoxia attenuated cell sensitivity to treatment. Thus, under normoxic conditions, the exposure of cells to 1 μM doxorubicin or 20 μg/ml cisplatin increased the number of cells in the SubG1 phase to 12–13 % and 30–35 %, respectively. Under hypoxic conditions, the increase in cell number induced by 1 μM doxorubicin or 20 μg/ml cisplatin was much more modest: 1–2 % and 10–12 %, respectively (Fig. 2a).

Fig. 2.

Hypoxia suppresses apoptosis induced by doxorubicin (Dox), or cisplatin (Cis) after 48-h incubation. a Assessment of number of cells treated with 1 μM doxorubicin or 20 μg/ml cisplatin for 48 h in the SubG1 subpopulation; b release of cytochrome c from cells treated with 1 μM doxorubicin or 20 μg/ml cisplatin. Cytosolic fraction and membrane fraction containing mitochondria were obtained after plasma membrane permeabilization with digitonin and separation by centrifugation as described in “Materials and methods”; c analysis of caspase-3-like activity in cells treated with 1 μM doxorubicin or 20 μg/ml cisplatin; d cleavage of PARP induced by 1 μM doxorubicin or 20 μg/ml cisplatin. Experiments were performed at least three times, and representative blots (b, d) are shown. *p < 0.05 versus doxorubicin or cisplatin alone

One of the crucial events in apoptosis is permeabilization of the OMM, which makes the release of cytochrome c from the mitochondrial intermembrane space an important event in apoptosis [10]. Treatment with doxorubicin or cisplatin caused the release of cytochrome c into the cytosol. Hypoxia attenuated doxorubicin- or cisplatin-induced cytochrome c release (Fig. 2b). Once released, cytochrome c activates the caspase cascade, resulting in the processing and activation of caspase-3, which is responsible for the manifestation of various characteristic features of apoptosis [11]. Doxorubicin stimulated caspase-3-like activity under normoxic conditions; under hypoxic conditions, this stimulation was markedly lower (Fig. 2c). One of the targets of caspase-3 is PARP [12, 13]. Stimulation of apoptosis by doxorubicin or cisplatin caused PARP cleavage (Fig. 2d), which was suppressed when the cells were treated under hypoxic conditions. Similar results were obtained when hypoxia was mimicked by DFO, which also suppressed the apoptotic manifestations induced by cisplatin or doxorubicin (Fig. 3a–d).

Fig. 3.

The hypoxia mimetic deferoxamine (DFO) suppresses apoptosis induced by 1 μM doxorubicin or 20 μg/ml cisplatin after 48-h incubation. a Assessment of number of cells treated 1 μM doxorubicin or 20 μg/ml cisplatin in the SubG1 subpopulation; b release of cytochrome c from cells treated with 1 μM doxorubicin or 20 μg/ml cisplatin. Cytosolic fraction and membrane fraction containing mitochondria were obtained after plasma membrane permeabilization with digitonin and separation by centrifugation as described in “Materials and methods”; c analysis of caspase-3-like activity in cells treated with 1 μM doxorubicin or 20 μg/ml cisplatin; d cleavage of PARP induced by 1 μM doxorubicin or 20 μg/ml cisplatin. Experiments were performed at least three times, and results from one representative experiment (b, d) are shown. *p < 0.05 versus doxorubicin or cisplatin alone

In order to estimate the impact of HIF-1 on cells’ resistance to treatment under hypoxic conditions, downregulation of HIF-1 using siRNA-mediated gene silencing was performed. Transfection of cells with siRNA to HIF-1α for 24 and 48 h prevented stabilization of this protein under hypoxic conditions (Fig. 4a). The efficiency of silencing was confirmed by the luciferase reporter assay (Fig. 4b). Analysis of cell death revealed that HIF-1 downregulation reversed the effect of hypoxia and abolished the difference in sensitivity to treatment between hypoxic and normoxic conditions. The size of the cell subpopulation in the SubG1 fraction was similar under normoxic and hypoxic conditions after treatment with doxorubicin or cisplatin (Fig. 4c). Additionally, the same extent of PARP cleavage was observed (Fig. 4d). These results clearly show that the stabilization of HIF-1 is involved in the inhibition of cell death under hypoxic conditions.

Fig. 4.

Silencing of HIF-1α using siRNA technique restores cellular sensitivity to cisplatin or doxorubicin under hypoxic conditions. siRNA prevents HIF-1 stabilization under hypoxic conditions, as assessed by Western blot (a) or relative luciferase activity (fold increase) (b); assessment of apoptosis induced by 1 μM doxorubicin or 20 μg/ml cisplatin under normoxic or hypoxic conditions after 48-h treatment by assessing cell number in the SubG1 fraction (c) or by cleavage of PARP (d). Experiments were performed at least three times, and representative blots (a, d) are shown. **p < 0.01 versus DFO

Using mitochondria as a target for apoptosis induction in tumor cells represents a promising strategy in fighting cancer. In order to answer the question of whether mitochondrial deterioration can overcome cells’ resistance to treatment under hypoxic conditions, we used a redox-silent analog of α-tocopherol, α-TOS. In contrast to doxorubicin or cisplatin, incubation under hypoxic conditions or mimicking hypoxia by DFO considerably stimulated cell death induced by α-TOS (Fig. 5). Hypoxia or DFO enhanced the SubG1 subpopulation of cells treated with 100 μM α-TOS by four- and seven-fold, respectively (Fig. 5a). α-TOS-induced cytochrome c release (Fig. 5b), caspase-3-like activity (Fig. 5c), and subsequent cleavage of PARP (Fig. 5d) were also markedly enhanced under hypoxic conditions or in the presence of DFO.

Fig. 5.

α-Tocopheryl succinate (α-TOS) kills cells irrespective of oxygen content in the air. Assessment of apoptosis by cell number in the SubG1 fraction (a); cytochrome c release (b); caspase-3-like activity (c); and cleavage of PARP (d). For the assessment of cytochrome release cytosolic fraction and membrane fraction containing mitochondria were obtained after plasma membrane permeabilization with digitonin and separation by centrifugation as described in “Materials and Methods”. Experiments were performed at least three times, and representative blots (b, d) are shown. *p < 0.05 versus α-TOS alone

Discussion

The resistance of tumor cells to treatment represents one of the most important problems in antitumor therapy. Cancer cells growing under hypoxic conditions acquire resistance to conventionally used anticancer drugs. This resistance is, to a large extent, determined by the silencing of mitochondrial apoptotic pathways. Hence, steps that lead to the sensitization of cells to treatment via mitochondrial targeting should considerably contribute to tumor cell elimination.

Hypoxia suppressed the efficiency of the conventionally used anticancer drugs cisplatin and doxorubicin. Cell death assessed by analyzing various characteristic features of apoptosis was markedly attenuated under hypoxic conditions (Fig. 2) or after treatment with DFO (Fig. 3), a compound known as a hypoxia-mimicking agent [9]. Our results clearly show that the observed resistance of cells to treatment was dependent on HIF-1 upregulation; knocking down this factor using siRNA technique restored the sensitivity of cells to treatment to the levels observed under normoxic conditions (Fig. 4).

Considering the importance of mitochondria for various modes of cell death, targeting these organelles represents a promising strategy for tumor cell elimination [14]. Recently, a class of compounds emerged with the common name of “mitocans” (originating from a combination of the words “MITOchondria” and “CANcer”). These compounds were shown to stimulate tumor cell death via a direct effect on mitochondria [15]. One of these compounds is α-TOS, which was used in our experiments. The ability of α-TOS to suppress cancer cell growth was reported by Prasad and Edwards-Prasad in 1982 [16]. They showed that α-TOS caused morphological changes and growth inhibition in mouse melanoma cells in culture. α-TOS was shown to inhibit the proliferation of avian reticuloendotheliosis virus-transformed lymphoblastoid cells in a dose-dependent manner, block the cells in the G2/M cell cycle phase, and induce apoptosis [17]. α-TOS was found to stimulate the production of reactive oxygen species (ROS) and eliminate malignant cells at concentrations non-toxic to normal cells and tissues [18]. In non-malignant cells, α-TOS is hydrolyzed by esterases [19]. As a result, gradually released α-tocopherol can even protect cells from oxidative stress. For example, α-TOS was shown to rescue cells from chemical-induced toxicity [20] and ionizing radiation [21]. However, in malignant cells, due to the lower esterase activity compared with non-malignant cells, the hydrolysis of α-TOS is suppressed [18, 22]. Thus, the inability of malignant cells to cleave α-TOS determines its ability to stimulate tumor cell death. The capacity of α-TOS to induce oxidative stress was explained by its interaction with Complex II of the mitochondrial respiratory chain, causing the leakage of electrons and formation of ROS [23] and leading to cell death. ROS can cause the oxidation of Bax and facilitate its translocation from the cytosol to the mitochondria, causing subsequent cytochrome c release [24]. In addition, it has been shown that α-TOS stimulates the rapid entry of Ca²⁺ into the cytosol, followed by the accumulation of this ion by mitochondria, which sensitizes them towards mitochondrial permeability transition (MPT) [25, 26], with rupture of the OMM and the release of cytochrome c. ROS, stimulated by α-TOS, represents an additional factor facilitating MPT. Paradoxically, the hypoxic environment of proliferating tumor tissue can facilitate ROS production [27]. Massive production of ROS is usually observed during reoxygenation of hypoxic tissue. However, the level of ROS may also increase by hypoxia alone, when electron transport complexes are in the reduced state. The main role in ROS production under hypoxic conditions belongs to Complex II of the mitochondrial respiratory chain, which has been implicated in the O₂-sensing pathway [28].

When cells are subjected to hypoxia, mitochondrial respiration is suppressed and, consequently, the capacity to build up and maintain the mitochondrial membrane potential substantially decreases [29]. This makes the mitochondria vulnerable to the induction of MPT, since the opening of the non-specific pore in the inner mitochondrial membrane is facilitated by low membrane potential [30]. Under such circumstances, hypoxia represents an additional destabilizing mitochondrial factor. Thus, in the case of treatment with α-TOS, instead of protecting cells against conventionally used anticancer drugs, hypoxia facilitates α-TOS-induced cell death. Previously, we have shown that α-TOS kills neuroblastoma cells irrespective of MycN oncogene expression, or p53 level [26]. Here, we demonstrate that α-TOS can also successfully overcome resistance to treatment caused by hypoxia, which makes α-TOS an attractive candidate for antitumor therapy via targeting of the mitochondria.