The cellular energy metabolism shift, characterized by the inhibition of oxidative phosphorylation (OXPHOS) and enhancement of glycolysis, is involved in nickel-induced neurotoxicity. MicroRNA-210 (miR-210) is regulated by hypoxia-inducible transcription factor-1α (HIF-1α) under hypoxic conditions and controls mitochondrial energy metabolism by repressing the iron–sulfur cluster assembly protein (ISCU1/2). ISCU1/2 facilitates the assembly of iron–sulfur clusters (ISCs), the prosthetic groups that are critical for mitochondrial oxidation-reduction reactions. This study aimed to investigate whether miR-210 modulates alterations in energy metabolism after nickel exposure through suppressing ISCU1/2 and inactivating ISCs-containing metabolic enzymes. We determined that NiCl2 exposure leads to a significant accumulation of HIF-1α, rather than HIF-1β, in Neuro-2a cells. The miR-210 overexpression and ISCU1/2 downregulation was observed in a dose- and time-dependent manner. The gain-of-function and loss-of-dysfunction assays revealed that miR-210 mediated the ISCU1/2 suppression, energy metabolism alterations, and ISC-containing metabolic enzyme inactivation after nickel exposure. In addition, the impact of miR-210 on ISC-containing metabolic enzymes was independent from cellular iron regulation. Overall, these data suggest that repression of miR-210 on ISCU1/2 may contribute to HIF-1α-triggered alterations in energy metabolism after nickel exposure. A better understanding of how nickel impacts cellular energy metabolism may facilitate the elucidation of the mechanisms by which nickel affects the human health.
Nickel (Ni) compounds are widely distributed environmental pollutants and are hazardous to public health.1 Although the exact mechanism of Ni-induced toxicity is unknown, increasing evidence indicates that the hypoxia-mimic responses, which are primarily mediated by hypoxia-inducible factor-1α (HIF-1α) stabilization, may contribute to the adverse effects of Ni on the human body.2, 3 The activation of HIF-1α target genes, including plasminogen activator inhibitor (PAI-1), vascular endothelial growth factor (VEGF), and the CXC chemokine, are involved in Ni dust-induced fibrotic lung disorders.4, 5 HIF-1α-dependent upregulation of IL-6 in human primary endothelial cells occurs in Ni contact allergic reactions.6 Moreover, HIF-1α-regulated angiogenesis and gene silencing may contribute to Ni-induced carcinogenesis.2 Additional investigations of Ni-induced hypoxia-mimic responses may elucidate the mechanisms of Ni toxicity in humans.
The alterations of energy metabolism, characterized by the repression of the tricarboxylic acid (TCA) cycle, mitochondrial electron transport, and oxidative phosphorylation (OXPHOS), but the enhancement of glycolysis, are typical adaptive responses under hypoxia-induced stress.7 Interestingly, these alterations are HIF-1α dependent and arise under normoxic conditions if cellular HIF-1α is stabilized by chemical agents.8, 9 Although the metabolic shift favors resistance and survival of cells under hypoxic conditions,10 this shift contributes to the harmful effects of Ni. Chen and Costa11 reported that Ni decreases the activity of cellular iron-containing enzymes, inhibits OXPHOS, and increases cellular glycolytic activity. The inducible adaptability of cells under hypoxic conditions facilitates their malignant transformation after chronic exposure to Ni.2 In addition, energy supply reduction leads to the dysfunction of tissues with high energy demand, including the nervous system. In our previous study, ingestion of soluble Ni compounds in mice resulted in the disturbance of aerobic metabolism accompanied by neurobehavioral changes.12 Therefore, investigation of the relationship between HIF-1α stabilization and energy metabolism alteration is an ideal strategy for elucidating the mechanism of Ni toxicity.
Endogenous microRNA molecules (miRNAs) are important mediators of numerous cellular processes, including the response to hypoxia.13 MicroRNA-210 (miR-210), which is induced by HIF-1α during hypoxia, is one of the most hypoxia-sensitive miRNAs and is an ideal factor modulating the hypoxia response.14 Recent studies have investigated the repression of miR-210 on the iron–sulfur cluster (ISC) assembly proteins, ISCU1 and ISCU2, whose 3′-untranslated region (UTR) is predicted to contain a highly conserved binding site for miR-210.15, 16 In mammalian cells, ISCU1 is located in the cytosol, whereas ISCU2 is located in the mitochondria. ISCU1/2 is a key chaperone for the assembly of cellular ISCs and transportation to enzymes that are responsible for mitochondrial respiration and energy production.17 These enzymes include aconitase, which is integral to the TCA cycle, and the mitochondrial respiratory complexes (complexes I, II, and III), which mediate mitochondrial electron transport. Therefore, the suppression of ISCU1/2 leads to the inactivation of ISC-containing enzymes by ISC metabolic disorder, and ultimately inhibits mitochondrial respiration. The regulation axis comprising HIF-1α, miR-210, and ISCU1/2 mediates the mitochondrial energy metabolism processes that respond to hypoxia in different cell types.18, 19, 20, 21 Compared with hypoxia, Ni exposure can disturb cellular iron homeostasis that is essential for the metabolism of ISCs.22 The evaluation of whether the modification of cellular iron states is involved in Ni-mediated metabolic events is of interest.
Therefore, the present study aimed to investigate whether the repression of miRNA-210 on ISCU1/2 underlay a Ni-induced energy metabolism shift. Ni-mediated miR-210 upregulation repressed the expression of ISCU1/2, inhibited the typical ISC enzymes, and shifted downstream mitochondrial metabolism. Our results indicate that miR-210 is essential and sufficient to modulate the cellular energy metabolism alterations induced by Ni.
Ni induces HIF-1α stabilization and miR-210 expression
The expressions of HIF and miR-210 induced by NiCl2 treatments were investigated in Neuro-2a cells. Cells were exposed to different concentrations of NiCl2 (0, 0.25, 0.5, and 1 mM) for 4 h or 1 mM of NiCl2 for different time periods (0, 2, 4, and 8 h). Western blot analysis revealed that Ni treatment resulted in obvious stabilization of HIF-1α but did not change the expression of HIF-1β (Figures 1a and b). Ni-induced overexpression of miR-210, identified by qRT-PCR analysis, was observed in dose- and time-dependent manners (Figures 1c and d). In addition, the established agents that stabilized HIF-1α, deferoxamine (DFX) and CoCl2, increased miR-210 expression after 4 h of treatment in Neuro-2a cells, as presented in Figure 1e.
Ni-induced miR-210 expression represses ISCU1/2 protein levels
The protein and transcript levels of ISCU1/2 in cells with different NiCl2 treatments were investigated. The protein levels of ISCU1/2 decreased in dose- and time-dependent manners in cells that were exposed to NiCl2 (Figures 2a and b). The Ni-induced activation of HIF-1α/miR210/ISCU regulation axis, which embodied as a stabilization of HIF-1α, an overexpression of miR210, and a repression of ISCU, was identified in different cells (Supplement Figures 1, 2 and 3). The transcript levels of ISCU1/2 were not significantly different among cells with different NiCl2 treatments (Figures 2c and d). To determine whether ISCU1/2 was regulated by miR-210, the wild-type and mutant (71–77 ACGCACA mutated to AGCGAGA) forms of the predicted target site in the 3′-UTR of ISCU1/2 (http://www.targetscan.org/mmu_61/) were cloned into luciferase reporter gene vectors. The transfection of miR-210 mimic repressed the activity of luciferase that interacted with the wild-type miR-210 predicted target site. However, the miR-210-mimic transfection did not alter the activity of luciferase that interacted with mutant target site (Figure 2e). These results verified that miR-210 regulates the expression of ISCU1/2 by binding to the predicted target site in the 3′-UTR of ISCU1/2. The analysis of protein level in cells that were transfected with the miR-210 mimic or inhibitor was performed to further establish that ISCU1/2 was the downstream target of miR-210. Under non-Ni exposure conditions, the transfection of miR-210 mimic (denoted as M) suppressed the ISCU1/2 protein level compared with that of the mimic-control (MC). Furthermore, under Ni exposure conditions (1 mM for 4 h), the transfection of the miR-210 inhibitor (I) attenuated the Ni-induced ISCU1/2 downregulation compared with that of the inhibitor-control (IC, Figure 2f).
Ni-induced miR-210 expression modulates the shift of energy metabolism
The Ni-induced shift of energy metabolism, characterized as the inhibition of OXPHOS and increase in glycolysis, plays important roles in Ni-mediated neurotoxicity and carcinogenesis. In our study, the viability of Neuro-2a cells that were treated with 1 mM NiCl2, 10 mM 2-deoxy-D-glucose (2-DG), or 25 μM bromopyruvic acid (BrPA) was equivalent to control levels. However, Ni exposure combined with 2-DG or BrPA, both of which are inhibitors of glycolysis, resulted in significant cytotoxicity in Neuro-2a cells (Figure 3a). The cells exposed to 1 mM NiCl2 for 4 h also manifested a low oxygen consumption rate (OCR) (Figure 3b). To confirm that the Ni-induced energy metabolism shift was modulated by miR-210 expression, we investigated the energy metabolism states of cells that were transfected with miR-210 mimic or inhibitor and treated with NiCl2. The quantifications of OCR revealed that oxygen consumption was inhibited in cells that were transfected with miR-210 mimic or treated with NiCl2. The transfection of miR-210 inhibitor eased off the suppression of Ni on oxygen consumption (Figure 3c). The transfection of miR-210 mimic reduced the concentration of cellular adenosine-5'-triphosphate (ATP) under non-Ni conditions, whereas inhibition of miR-210 attenuated the ATP depletion caused by Ni (Figure 3d). Hexokinase (HK) is the key enzyme in glycolysis, and an increase in HK activity indicates the enhancement of glycolysis. The transfection of miR-210 mimic enhanced the activity of HK under non-NiCl2 conditions, whereas inhibition of miR-210 mitigated the HK activity increase caused by Ni (Figure 3e). The transfection of miR-210 mimic increased the content of cellular lactic acid, the end product of glycolysis, whereas the inhibition of miR-210 attenuated the lactic acid accumulation caused by Ni (Figure 3f).
Ni-induced miR-210 expression inhibits the ISC-containing metabolic enzymes
The activity of two typical ISC-dependent metabolic enzymes, aconitase and complex I, in cells transfected with miR-210 mimic and inhibitor was measured. The transfection of miR-210 mimic suppressed the activity of these two enzymes under non-Ni conditions, whereas inhibition of miR-210 attenuated the suppression of enzymes caused by Ni. Furthermore, the transfection or treatment did not change the protein level of the representative component of aconitase (ACO2) or complex I (NDUFA9), as depicted in Figures 4c and d.
Treatment with 1 mM NiCl2 for 4 h does not disturb cellular iron homeostasis
To determine whether the Ni-mediated ISC-containing enzyme suppression relates to the Ni-caused iron homeostasis disturbance, the iron metabolism states of cells exposed to 1 mM NiCl2 for different times were examined. A significant decrease of cellular iron ions is only found in the 8 h exposure groups (Figure 5a). The levels of total ferritin protein, the major form of stored iron in cells, were analyzed by western blot after Ni treatment. NiCl2 exposure for 2 or 4 h did not reduce cellular ferritin levels. However, the level of ferritin in cells that were exposed to Ni for 8 h was reduced (Figure 5b). The overexpression of the transferrin receptor (TfR), an important cellular response under iron-deficient conditions, was only observed in cells exposed to NiCl2 for 8 h (Figure 5c).
In this study, treatment with 1 mM NiCl2 for 4 h induced HIF-1α stabilization, miR-210 overexpression, ISCU1/2 protein reduction, energy metabolism alterations, and ISC-containing metabolic enzyme suppression in Neuro-2a cells. Although the Ni levels in our study are unlikely to be achieved in vivo, such levels of Ni exposure are commonly applied in in vitro study to investigate the mechanism of Ni toxicity.23, 24, 25, 26 Our previous study revealed that 1 mM NiCl2 exposure for more than 12 h produced obvious cytotoxicity and mitochondrial dysfunction in Neuro-2a cells.27 Because energy metabolism plays an important role in physiological processes of the brain, investigating energy metabolism alterations before cell death occurs may contribute to the elucidation of the mechanisms of Ni-induced neurotoxicity.
Similar to hypoxia, Ni also suppresses the activity of prolyl hydroxylase and disrupts the degradation of HIF-1α. Several studies revealed that brief (2–6 h) exposures to Ni caused the accumulation of HIF-1α in various types of cell and triggered the HIF-1α-modulating hypoxia-mimic responses.5, 28, 29 Therefore, the evidence of cellular responses to hypoxia is applied to explain the health effects of Ni. The elucidation of miR-210, a HIF-1α-controlled microRNA, substantially clarifies the nature of the varied cellular responses to hypoxia.9, 30 MiR-210 targets the receptor tyrosine kinase ligand EphrinA3 (EFNA3), enhancing the differentiation of human umbilical vein endothelial cells and promoting angiogenesis in hypoxia.31, 32 In addition, hypoxia-induced miR-210 participates in cell cycle regulation by modulating its target genes, MNT, a known MYC antagonist, and E2F3, a key protein in cell cycle regulation.33, 34 Under oxygen/glucose deprivation conditions, the overexpressed miR-210 represses the expression of Bcl-2 and amplifies apoptosis of Neuro-2a cells.35 In our study, HIF-1α stabilization and miR-210 overexpression induced by NiCl2 were observed in dose- and time-dependent manners. Moreover, CoCl2 and DFX, the well-established hypoxia-mimic compounds, increased the expression of miR-210. Therefore, miR-210 may contribute to the chemical hypoxia responses, in addition to the low oxygen content.
The energy metabolism shift is an important cellular response to hypoxia. When oxygen is abundant, cellular ATP is primarily generated through OXPHOS that generates 32 moles of ATP from 1 mole of glucose. Under hypoxic conditions, OXPHOS is suppressed and cellular energy production shifts to the glycolytic pathway that only generates 2 moles of ATP from 1 mole of glucose. Although glycolysis is not economic for ATP generation, it provides the minimum energy supplement for cell survival under hypoxic conditions.36 Recently, the energy metabolic shift regulated by the HIF-1α-miR-210-ISCU1/2 pathway has been revealed to participate in multiple hypoxia-related physiological and pathological processes.18, 21 In our study, Ni exposure resulted in the downregulation of ISCU1/2 protein levels, which was connected to the changes in HIF-1α and miR-210. The results of the reporter gene assay for the predicted miR-210 bonding site on ISCU1/2 and the western blot analysis of ISCU1/2 protein levels after miR-210 transfection further supported that ISCU1/2 was regulated by miR-210 in Ni exposure conditions. The lack of change in ISCU1/2 mRNA levels after Ni exposure revealed the post-transcriptional regulatory effects of miR-210 on ISCU1/2.15 Moreover, the activation of HIF-1α/miR-210/ISCU1/2 regulation pathway was universal in different Ni-treated cells. The downstream effects of this regulation pathway may underlie the Ni-mediated hypoxia-mimic response.
The energy metabolism shift, characterized by glycolysis enhancement under normoxic conditions, is considered to be involved in the toxicity of Ni. In the central nervous system, Ni-mediated inhibition of OXPHOS resulted in energy supplement stress.12 Inhibition of glycolysis with 2-DG or BrPA increased the cytotoxicity of NiCl2 in Neuro-2a cells in the present study. In addition, the cellular phenotype of preferential glycolytic energy generation, known as the Warburg effect, is implicated in the carcinogenesis of Ni.2 The elucidation of the impacts of Ni on cellular energy metabolism may be important for studying Ni toxicity. We found that miR-210 was an essential modulator of Ni-induced energy metabolism shift, resulting in oxygen consumption decrease, ATP deficit, alteration of HK activity, and accumulation of lactic acid. The key enzymes for OXPHOS, aconitase for the Krebs cycle and complex I for mitochondrial electron transfer, were suppressed by miR-210 without a protein level change of their component elements. It means that the inactivation of these enzymes accounts for the ISC deficit, rather than the inhibition of protein expression. Therefore, the repression of miR-210 on ISCU1/2 and downstream ISC-depended metabolic enzymes provides a logical explanation for the HIF-1α stabilization and energy metabolism shift, both of which are observed under different Ni exposure conditions.11, 29 Moreover, deficiencies in ISC-containing metabolic enzymes may attribute to the generation of reactive oxygen species (ROS) and mitochondria dysfunction, through which Ni induced significant cytotoxicity in nerve cells, as previously reported.27, 37 In addition to Ni, other heavy metals, including cobalt, manganese, and vanadium, induce HIF-1α stabilization.28 Hence, the functional study of miR-210 may be valuable for the elucidation of heavy metal toxicity.
In contrast to low oxygen conditions, the hypoxia-mimic response induced by Ni may be accompanied by alterations of cellular iron levels. As previously reported, Ni may compete with iron for divalent metal transporter-1 (DMT1) and disturb cellular iron homeostasis.22, 38 Because cellular iron levels are directly linked to the assembly of ISCs and ISC-dependent enzymes activity,39 the influence of Ni on cellular iron levels should not be excluded. However, based on our results that cellular exposure to 1 mM NiCl2 for 4 h did not disturb cellular iron homeostasis, we concluded that miR-210-mediated suppression of ISC-containing metabolic enzymes was independent from cellular iron regulation in the time frame of our Ni treatments. Costa and colleagues38 have reported the effects of NiCl2 on cellular iron regulation. In their study, a series of significant alterations that reflect the activation of cellular iron regulation emerged at 6–8 h post Ni exposure, which was consistent with the results of our study. Only the change in cellular iron content after 4 h of Ni exposure was different between the two studies that may be attributed to the use of different cell lines or methods of iron content determination. It appears that miR-210 modulated the energy metabolism shift before cellular iron homeostasis imbalance. Further studies are needed to identify the specific contributions of these mechanisms to the toxicity of Ni.
In summary, our findings suggest that the HIF-1α-controlling microRNA, miR-210, modulated the cellular energy metabolism shift by repressing ISCU1/2 after Ni exposure. Because energy metabolism is critical in the central nervous system, our results provide insight into the understanding of Ni-mediated neurotoxicity. Furthermore, the ability of miR-210 to regulate hypoxia-mimic response by binding to target genes could be utilized to elucidate the health effects of a variety of heavy metals that induce HIF-1α stabilization.
Materials and Methods
Cell culture and treatments
Mouse neuroblastoma cell lines (Neuro-2a), obtained from the Institute of Biochemistry and Cell Biology (Chinese Academy of Science, Shanghai, China), were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Invitrogen, Carlsbad, CA, USA) supplemented with 10% heat-inactivated fetal bovine serum (FBS, HyClone, Logan, UT, USA) and 1% v/v penicillin/streptomycin (Sigma-Aldrich, St. Louis, MO, USA) in a 5% CO2-humidified atmosphere at 37°C. A 1 M stock solution of Ni chloride (NiCl2·6H2O, Sigma-Aldrich) was prepared with sterile H2O and diluted to the appropriate concentrations for cell treatment.
Western blot analysis
From whole-cell extracts, 40 μg of protein was resolved by SDS-PAGE and transferred onto a polyvinylidene fluoride membrane (Bio-Rad, Hercules, CA, USA). Membranes were probed for various proteins with the appropriate primary antibodies and visualized with the electrochemiluminescence system (Thermo Fisher Scientific, Waltham, MA, USA). The bands were imaged and analyzed using a ChemiDoc XRS+ System with Image Lab Software (Bio-Rad).
Mouse monoclonal antibody to HIF-1α was purchased from Novus Biologicals (Littleton, CO, USA). Rabbit monoclonal antibodies to HIF-1β and FTH1 were purchased from Cell Signaling Technology (Danvers, MA, USA). Rabbit polyclonal antibody to ISCU1/2 was purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Rabbit polyclonal antibody to Aconitase 2 and mouse monoclonal antibody to NDUFA9 were purchased from Abcam (Cambridge, UK).
The miRNA mimic (miR10000267) or inhibitor (miR20000267) for cell transfection were purchased from RiboBio (Guangzhou, China) and used according to the manufacturer’s instructions. Briefly, 24 h after cell seeding, the medium without penicillin/streptomycin was replaced with the transfection medium containing 100 nM miRNA mimic or inhibitor and 0.2% v/v Lipofectamine 2000 transfection reagent (Invitrogen). After 4 h, the transfection medium was replaced with fresh penicillin/streptomycin absent medium. The transiently transfected cells were collected after an additional 24-h incubation for further experiments.
The extraction of total RNA and quantification of ISCU, TfR1, TfR2, and miR-210 were performed following our previously published protocol.27 The bulge-loop miRNA qRT-PCR primer sets (one reverse transcription primer and a pair of quantitative PCR primers for each set, MQP-0102 and MQP-0201) were designed by RiboBio. U6 was utilized as an internal control for miRNAs, and β-actin was used as an internal control for mRNA. Primers used in this study were as follows: ISCU-forward, 5′-CTCTGCCTGCCGCCTGTG-3′ and reverse, 5′-CTGCTTCTCTGGCTCCTCCTTC-3′; TfR1-forward, 5′-GAGACTACTTCCGTGCTACTTC-3′ and reverse, 5′-TGGAGATACATAGGGCGACAG-3′; TfR2-forward, 5′-AGAGTGGTTGGAGGGCTAC-3′ and reverse, 5′-CAATGAGGCTGACGAGAAGG-3′; β-actin-forward, 5′-CCACACCCGCCACCAGTTC-3′ and reverse, 5′-CCTTCTGACCCATTCCCACCATC-3′.
The 3′-UTR regions of the ISCU containing the predicted binding sites for miR-210, wild or mutant (71–77 ACGCACA mutated to AGCGAGA), were cloned into pmiR-RB-REPORT vectors (GUR100509 and GUR100510; RiboBio). After co-transfection with the vectors and miR-210 mimic, the firefly and Renilla luciferase activities were measured using the Dual-luciferase Reporter Assay (Promega, Madison, WI, USA).
Cells were seeded into a 96-well plate and exposed to 1 mM NiCl2 for 2 h. Next, 2-DG (10 mM) or BrPA (25 μM) was added to the Ni-containing medium and incubated for an additional 2 h. Then, the medium in the 96-well plate was replaced with fresh medium containing 10% v/v CCK-8 solution (Dojindo, Kumamoto, Japan) and incubated for 2 h. The optical density (OD) of each well was determined at a wavelength of 450 nm with the Infinite M200 Microplate Reader (Tecan, Männedorf, Switzerland). Cell viability is reported as a percent of the control value.
The ATP content was determined using the ATP Determination Kit (Molecular Probes, Eugene, OR, USA). The luminescence of the sample was measured using an Infinite M200 Microplate Reader (Tecan). ATP standard curves were established, and the ATP concentrations were expressed as nmol/mg of protein.
Oxygen consumption rate
Cells (1 × 105 per well) were plated, transfected, and exposed to NiCl2 for 4 h before measurement. The rate of change of oxygen in media surrounding intact cells was measured by a XF96 Extracellular Flux Analyzer (Seahorse Bioscience, Billerica, MA, USA). The OCR of a well was determined by quantifying oxygen-dependent changes in fluorescence of a proprietary fluorescein complex (pmol/min/well).
HK activity assay
The activity of HK was measured using the Activity of Hexokinase Determination Kit (Nanjing Jiancheng, Nanjing, China) based on a coupled reaction of HK and 6-phosphoglucose dehydrogenase. The activity of hexokinase was determined by the generation of triphosphopyridine nucleotide (NADPH) and measured as the change in OD at 340 nm (ΔOD340) at 37°C over 3 min. The activity of hexokinase was expressed as ΔmOD340/min/mg protein (ΔmOD340=ΔOD340/1000).
The intracellular lactate content was measured using the Lactate Determination Kit (Nanjing Jiancheng) according to the manufacturer’s instructions. Standard curves were established, and the lactate concentrations were expressed as μmol/mg of protein.
Aconitase and complex I activity assays
After appropriate treatment, cells were collected in Eppendorf tubes, resuspended in a hypotonic buffer (120 mM KCl, 20 mM HEPES, 2 mM MgCl2, and 1 mM EGTA, pH 7.4) and disrupted by performing ten 1-s pulses with the Vibra-Cell ultrasonic processor (Sonics & Materials, Danbury, CT, USA). The subsequent assessments of the two typical ISC metabolic enzymes were performed according to the previously reported protocol.12
Cellular iron content
The total cellular iron (ferrous and ferric) ions were measured using the Iron Assay Kit (Abcam) following the manufacturer’s instructions. Cells were seeded into six-well plates and exposed to Ni for different time periods. Standard curves for the iron ions were established, and the cellular iron content was expressed as nmol per well.
Values are expressed as the mean±S.E.M. and were analyzed by SPSS (V18.0.0, SPSS Inc., Chicago, IL, USA). Data comparisons were completed using one-way ANOVA or paired t-test to compare the means of the different treatment groups. P<0.05 was considered statistically significant.
We thank Dr. Chunhai Chen and Zhujun Zhang for their hospitable and professional technical assistance. Grant support was provided by the National Natural Science Foundation of China (No. 81102108) and the Outstanding Youth Science Fund of Chongqing (No. CSTC2013JCYJJQ10002).
hypoxia-inducible transcription factor-1α
iron–sulfur cluster assembly protein
reactive oxygen species
About this article
Supplementary Information accompanies this paper on Cell Death and Disease website (http://www.nature.com/cddis)