|Home | About | Journals | Submit | Contact Us | Français|
We developed a model system to investigate apoptotic resistance in T cells using osmotic stress (OS) to drive selection of death-resistant cells. Exposure of S49 (Neo) T cells to multiple rounds of OS followed by recovery of surviving cells resulted in the selection of a population of T cells (S49 (OS 4–25)) that failed to die in response to a variety of intrinsic apoptotic stimuli including acute OS, but remained sensitive to extrinsic apoptotic initiators. Genome-wide microarray analysis comparing the S49 (OS 4–25) with the parent S49 (Neo) cells revealed over 8500 differentially regulated genes, with almost 90% of those identified being repressed. Surprisingly, our data revealed that apoptotic resistance is not associated with expected changes in pro- or antiapoptotic Bcl-2 family member genes. Rather, these cells lack several characteristics associated with the initial signaling or activation of the intrinsic apoptosis pathway, including failure to increase mitochondrial-derived reactive oxygen species, failure to increase intracellular calcium, failure to deplete glutathione, failure to release cytochrome c from the mitochondria, along with a lack of induced caspase activity. The S49 (OS 4–25) cells exhibit metabolic characteristics indicative of the Warburg effect, and, despite numerous changes in mitochondria gene expression, the mitochondria have a normal metabolic capacity. Interestingly, the S49 (OS 4–25) cells have developed a complete dependence on glucose for survival, and glucose withdrawal results in cell death with many of the essential characteristics of apoptosis. Furthermore, we show that other dietary sugars such as galactose support the viability of the S49 (OS 4–25) cells in the absence of glucose; however, this carbon source sensitizes these cells to die. Our findings suggest that carbon substrate reprogramming for energy production in the S49 (OS 4–25) cells results in stimulus-specific recognition defects in the activation of intrinsic apoptotic pathways.
Cell death plays an essential function in organismal life by balancing cell proliferation to preserve the natural homeostatic physiological processes in our bodies.1, 2, 3, 4 Apoptosis is a physiological mode of cell death that permits the removal of unwanted cells from the body at a specific time or in response to a given signal. All the molecular components cells need to carry out this cell death process are present in normal healthy cells, and require only activation for apoptosis to ensue. However, in various human disease states, including neurodegeneration, autoimmunity, and cancer, a deregulation or malfunction of this inherent program occurs and cells may develop resistance to apoptosis.5, 6
Apoptosis occurs through two main signaling pathways: an extrinsic pathway that utilizes a diversified group of cell surface death receptors;7, 8, 9, 10, 11, 12, 13, 14 and an intrinsic pathway that utilizes various intracellular organelles to execute the programmed cell death machinery.15, 16, 17, 18, 19, 20 An important and well-studied point of control for both the extrinsic and intrinsic apoptotic pathways is the Bcl-2 family of proteins that comprise both pro- and antiapoptotic members and regulate the apoptotic program through a tightly controlled series of checks and balances.21, 22
Resistance to cell death is a common feature in many disease states that impedes both therapy and treatment. The mechanisms of resistance to apoptosis are poorly understood and can vary from cell to cell. Apoptotic resistance has been reported to result from a variety of mechanisms including overexpression of antiapoptotic proteins, inhibition of proapoptotic proteins, direct inhibition of the apoptotic machinery,23, 24 and cell volume regulation.25, 26 As nutrient, oxygen, and energy supplies have a critical role in signaling a cell to live or die, recently it has been recognized that cross talk between cell metabolism and cell death machinery may have a major contribution in cellular life and death decisions.27
Most apoptotic stimuli result in a very asynchronous cell death, with cells dying over a period of hours to even days. In contrast, T cells induced to die by hyperosmotic stress undergo a rapid and synchronous apoptotic response, with over 90% of the cells dead by 4h.28 Using this system as a cell death model, we developed T cells resistant to osmotic stress (OS). These S49 (OS 4–25) cells are similarly resistant to various intrinsic apoptotic stimuli, lack the initial signaling associated with programmed cell death, but remain sensitive to extrinsic apoptotic stimuli. Our analysis shows the resistance to apoptosis can be associated with major changes in the gene expression profile and altered cell death signaling not related to changes in the Bcl-2 family of proteins. Interestingly, the S49 (OS 4–25) cells exhibit essential signs of the Warburg effect and are completely dependent on glucose as a carbon source. Glucose withdrawal leads to cell death with multiple characteristics of apoptosis. Moreover, we show that a change in carbon substrate utilization is critical for resistance to cell death and can sensitize cells to apoptosis.
S49 (Neo) cells are exceptionally sensitive to programmed cell death induced with a variety of agents.28 We exploited this property of S49 (Neo) cells by exposing them to multiple rounds of hyperosmotic stress and recovery to select cells that were resistant to apoptosis. After 25 generations of 4-h hyperosmotic treatment (500mOsm mannitol) followed by several days/weeks of recovery of the viable population of cells, we developed a cell line designated S49 (OS 4–25). Microscopic examination of the S49 (OS 4–25) cells showed they were inherently larger in cell size (Figure 1a), and more than doubled their cell volume when compared with the S49 (Neo) cells (Figure 1b). The S49 (OS 4–25) cells were resistant to acute OS (Figure 1c), and gained an inherent regulatory volume increase (RVI) response under hyperosmotic conditions (Figure 1d). Remarkably, these cells were also resistant to a variety of intrinsic apoptotic stimuli, as illustrated by dexamethasone (Figure 1e) and other apoptotic agents (Supplementary Figure 1). However, they remain sensitive to extrinsic apoptotic stimuli such as Fas ligand (Figure 1f).
Apoptotic resistance is often associated with over- or underexpression of anti- or proapoptotic genes, respectively. We examined whether changes in the gene expression were associated with intrinsic apoptotic resistance by performing a genome-wide microarray analysis. Comparing S49 (OS 4–25) cells with the parental S49 (Neo) cells showed 8560 genes being differentially expressed; 7605 (89%) genes had a decrease in expression, whereas only 955 (11%) genes were increased (Figure 2a). Supplementary Tables 1 and 2 show the top 100 genes with increased or decreased expression, respectively. We verified several genes that were differentially expressed in the microarray via real time PCR (RT-PCR) analysis (Supplementary Figure 2). Many of the genes with increased expression, including granzyme B (GZMB),29 a calcium binding protein (S100A6),30 a gap junction protein (GJA1),31 and a cyclin-dependent kinase inhibitor (CDKN2A),32 have been reported to have a role in cell death and/or apoptosis. Several genes showing decreased expression, including the inhibitory receptor of peripheral mononuclear cells (LAIR1),33 a conserved cysteine-rich protein (PLAC8),34 an apoptosis-associated speck-like protein (PYCARD),35 and a receptor tyrosine kinase (ERBB3),36 have also been reported to have a role in cell death. CARD11, a caspase recruitment domain-containing protein, has decreased expression and been reported to interact with the proapoptotic BCL10 protein.37
Ingenuity pathway analysis (IPA) was employed to examine this gene set in regard to molecular and cellular functions. Table 1 shows the top 10 molecular and cellular functions (based on their P-value), along with the number of genes altered for each condition. The S49 (OS 4–25) cells showed many changes related to posttranscriptional and/or posttranslational modifications, gene expression, cell growth, and proliferation, along with cell death and survival. Network analysis using IPA showed of the 36 genes specific for the canonical apoptotic pathway, 29 genes had reduced expression in the S49 (OS 4–25) cells, whereas only 7 genes had increased expression (Figure 2b, Table 2), suggesting a global repression of the cell death response. Intriguingly, of the seven apoptotic genes with increased mRNA expression, five genes (lamin A, Fas receptor, NFkB, MRAS, and KRAS) can be considered proapoptotic in nature, despite the stimulus-specific resistance to cell death in the S49 (OS 4–25) cells.
We analyzed various members of the Bcl-2 family for their gene and/or protein expression in the S49 (Neo) and S49 (OS 4–25) cells. Antiapoptotic family members Bcl2L1 (also known as Bcl-xL) and Bcl-2 showed reduced expression via rtPCR in the S49 (OS 4–25) cells, despite the microarray suggesting that Bcl-2 might have increased expression (Figure 3a, top). We examined the protein expression of these genes and observed no significant difference in protein expression between the S49 (Neo) and S49 (OS 4–25) cells (Figure 3a, bottom), suggesting the observed apoptotic resistance does not reflect overexpression of these antiapoptotic genes. In addition, we examined several proapoptotic Bcl-2 family members including Bax, Bad, and Bid for changes in gene or protein expression. Figure 3b showed that none of these proapoptotic family members had a significant change in either gene or protein expression. Furthermore, other Bcl-2 family members including Bim, PUMA, Mcl-1, and NOXA showed no significant change in gene expression between the two cell lines (Figure 3c). Our results suggest that the stimulus-specific resistance to apoptosis in the S49 (OS 4–25) cells occurs independent of significant changes in either pro- or antiapoptotic Bcl-2 family members.
To analyze the molecular basis for this apoptotic resistance in S49 (OS 4–25) cells, we evaluated these cells for well-known components of the apoptotic process. Figure 4 shows that S49 (Neo) cells treated with OS resulted in an increase in mitochondria-derived reactive oxygen species (ROS; a), an increase in intracellular calcium (b), a loss of glutathione (c), caspase activity (d), and release cytochrome c (e), all well-recognized characteristics of cells undergoing apoptosis. In contrast, S49 (OS 4–25) cells did not show any of these apoptotic-associated changes in regards to OS. The role of reactive oxygen species and caspases in the cell death process was further examined using the ROS scavenger N-acetylcysteine (NAC) and the pan-caspase inhibitor (Q-VD-OPh). A significant decrease in the extent of cell death was observed in the osmotically stressed S49 (Neo) cells in the presence of increasing concentrations of NAC, although the highest NAC concentration failed to fully protect these cells from death (Supplementary Figure 3). In contrast, caspase inhibition resulted in a clear concentration-dependent protection from cell death (Supplementary Figure 3), suggesting that ROS production has a substantial role in the cell death process; however, caspases has a critical role in inducing apoptosis in these cells. We examined the S49 (OS 4–25) cells for their response to the BH3 mimetic ABT737, known to activate the intrinsic apoptotic pathway. The addition of ABT737 resulted in sensitizing the S49 (OS 4–25) cells to undergo cell death, resulting in the release of cytochrome c (Figure 4f), indicating that the intrinsic cell death pathway remains intact. These data suggest a defect in signaling of the cell death program as a major cause of the selective apoptotic resistance in S49 (OS 4–25) cells.
An interesting phenotypic alteration was observed in the S49 (OS 4–25) cells, as after 48h of culturing S49 (Neo) and S49 (OS 4–25) cells from an initial density of 1 × 105cells/ml the culture media for the S49 (OS 4–25) cells appeared disproportionately acidic (Figure 5a). This phenomenon can occur after excess growth; however, the S49 (OS 4–25) cells had significantly less number of cells when compared with the S49 (Neo) cells (Figure 5b). This acidic environment observed for the S49 (OS 4–25) cells is characteristic of the Warburg effect, a high rate of glycolysis in the presence of oxygen that results in lactic acid fermentation. Our analysis showed a fivefold increase in intracellular lactate in the S49 (OS 4–25) cells compared with the S49 (Neo) cells (Figure 5c), suggesting metabolic reprogramming in the S49 (OS 4–25).
Lymphoid cells in general use glucose and glutamine as primary substrates to generate cellular energy. We examined the S49 (Neo) and S49 (OS 4–25) cells for their utilization of these cellular energy sources by culturing the cells in the absence of extracellular glucose or glutamine for 48h. Our data showed a 30–40% increase in dead cells for the S49 (Neo) cells under both conditions, whereas the S49 (OS 4–25) cells survived in the absence of glutamine, but had greater than 80% cell death in the absence of glucose (Figure 5d). These data indicate that the S49 (OS 4–25) cells have developed a unique and absolute requirement for glucose for survival, whereas the parental S49 (Neo) cells required both glucose and glutamine to maintain cell viability.
The shift in the S49 (OS 4–25) cells to glycolytic energy dependence suggests a change in the signaling mechanism(s) controlling the metabolism in these cells. Our analysis of various metabolic genes showed no significant change in gene expression in the S49 (OS 4–25) cells compared with the S49 (Neo) cells (Supplementary Figure 4). As AKT has been suggested to be an important driver of the glycolytic phenotype and is linked to many cellular processes including apoptosis and cell growth, we examined the S49 (Neo) and S49 (OS 4–25) cells for acute changes in the phosphorylation of AKT. Upon OS, an initial high level of activated phospho-AKT was lost in the S49 (Neo) cells, but sustained in the S49 (OS 4–25) cells (Figure 6a). As AKT is known to signal through mTOR, we examined the S49 (OS 4–25) cells for their response to inhibition of this kinase. Treatment of S49 (OS 4–25) cells with the mTOR inhibitor rapamycin resulted in the sensitization of these cells to apoptotic stimuli (Figure 6b), suggesting an alteration in the AKT-mTOR signaling pathway has a role in the observed apoptotic resistance.
The findings of a complete dependence on glucose for S49 (OS 4–25) cell survival led us to examine the function of the mitochondria in these cells. Glycolytic metabolism of galactose yields no net ATP, forcing the cells to rely solely on oxidative phosphorylation for energy;38 thus, cells with dysfunctional mitochondria would die. Substitution of galactose for glucose resulted in cell survival and high viability of the S49 (OS 4–25) cells (Figure 7a), suggesting functional mitochondria. Methyl pyruvate as a surrogate for glucose also resulted in cell survival in the S49 (OS 4–25) cells (Figure 7a). Examination of both the S49 (Neo) and S49 (OS 4–25) cells for mitochondrial function using a Seahorse mitochondrial stress analysis (Figure 7b) showed similar oxygen consumption rate (OCR; left) and extracellular acidification rate (ECAR; right) profiles. The overall increase in ECAR in the S49 (OS 4–25) cells reflects the increased level of lactic acid in these cells (Figure 5c); however, the addition of various mitochondrial stress reagents showed no difference in mitochondrial function between the two cell lines. As the S49 (OS 4–25) cells were viable in galactose and showed no mitochondrial dysfunction, we determined if galactose could sensitize these cells to undergo apoptosis. When S49 (OS 4–25) cells were cultured in galactose for 48h before exposure to OS, ~40% of the cells died within 6h, compared with cells cultured in glucose (Figure 7c). These data suggest that a change in carbon source (galactose) that permits predominantly oxidative phosphorylation as the primary mechanism for energy production can sensitize cells to undergo cell death.
We tested the hypothesis that heightened glycolysis can result in apoptotic resistance by using primary T cells isolated from the mouse spleen as a model system. Naïve T cells generate ATP via oxidative phosphorylation, whereas activated T cells predominately rely on glycolysis as the major mode of energy generation even in the presence of sufficient oxygen (Warburg effect). The naïve T cells were significantly more sensitive to apoptotic stimulation than the anti-CD3/CD28-activated T cells (Figure 7d), suggesting that activated T cells, similar to our S49 (OS 4–25) cells that predominately use glycolysis as a source of energy, have increased resistance to cell death.
We analyzed the mode of cell death in the S49 (OS 4–25) cells after 30–36h of glucose withdrawal for changes in various core characteristics associated with apoptosis. As shown in Figure 8a, glucose withdrawal from the S49 (OS 4–25) cells resulted in propidium iodide-positive (PI+) cells with a slight decrease in forward-angle light scatter indicative of a reduced cell size. The loss of cell volume in the S49 (OS 4–25) is most likely opposed by the gain of a RVI response in these cells (Figure 1d), restricting the extent of this parameter. To directly address this issue of cell volume changes, we measured the mean cell volume (MCV) via electronic sizing in the viable population of cells in the presence and absence of glucose, observing a 13.3±3.9% decrease in MCV in the S49 (OS 4–25) cells grown without glucose. These glucose-starved cells also show an increase in caspase-3/7 activity, ROS production, and intracellular calcium, coupled with a loss of glutathione and a well-defined internucleosomal DNA cleavage pattern (Figure 8b and f), all characteristic of cells undergoing apoptosis. Both examination of these cells for changes in their membrane lipid symmetry (externalization of phosphatidylserine) and release of cytochrome c, showed a significant increase in the S49 (OS 4–25) cells in the absence of glucose (Figure 8g and h, respectively). Thus, the absence of glucose from the S49 (OS 4–25) cells results in the occurrence of several fundamental cell death characteristics reflective of essential apoptosis.
This study shows that repetitive OS and recovery of lymphoid cells results in permanent global changes at both the genetic and metabolic levels, which has consequences leading to aberrant cell signaling and activation of intrinsic apoptosis. These cellular transformations involved a change in cell size, a loss of essential characteristics associated with intrinsic apoptosis, and a complete dependence on glucose as an energy source, while maintaining sensitivity to extrinsic stimuli. Interestingly, glucose withdrawal from the S49 (OS 4–25) cells resulted in cell death with many classical characteristic of apoptosis, including cell shrinkage, increased intracellular calcium, increased caspase activity, loss of cytochrome c, externalization of phosphatidylserine, and internucleosomal DNA cleavage. Galactose as an alternative carbon source for glucose resulted in sensitizing these cells to apoptosis. These data suggest that the global cell death machinery remains intact in the S49 (OS 4–25) cells, and the resistance to apoptosis is associated with alterations in signal transduction pathways that activate the apoptotic program.
We previously showed that S49 (Neo) cells are devoid of a RVI response and undergo a very rapid and synchronous apoptosis.39 In an earlier generation of stressed and recovered cells, we demonstrated that OS resistance imparts acquired antiapoptotic mechanisms through an acquired RVI response and AKT signaling that combats the ionic changes associated with cell shrinkage.28 Our current study uses cells that were carried through an additional 10 generations of OS and recovery, resulting in a more stable population of cells. We show here that the activation of the AKT signaling is retained in the S49 (OS 4–25) cells and extended this finding, showing that mTOR also has a critical role in the intrinsic apoptotic resistance. In addition, the discord between the intrinsic apoptotic resistance of S49 (OS 4–25) cells and their ability to die upon glucose deprivation or their sensitization to die in the presence of galactose suggests that an appropriate carbon source for energy is a critical requirement to initiate the cell death program. Therefore, the signaling or activation of the S49 (OS 4–25) cells to apoptosis is compromised in both a stimulus- and/or metabolic-specific manner.
Analysis of genome-wide changes in gene expression showed 28.5% of all genes in the S49 (OS 4–25) cells were significantly altered with 89% of these genes having reduced gene expression, suggesting an overall suppression of the transcriptional activity or increased RNA turnover in the S49 (OS 4–25) cells. Pathway analysis of the S49 (OS 4–25) cells affected several important molecular and cellular functions. The reduced expression of genes is particularly evident in the apoptosis pathway, where 29 of 36 S49 (OS 4–25) genes (81% Table 2) had diminished expression. Of note, the Fas cell surface death receptor showed increased gene expression in the S49 (OS 4–25) cells, supporting our finding that these cells remain sensitive to extrinsic apoptotic stimuli and the core cell death machinery remains intact. In-depth analysis of both pro- and antiapoptotic Bcl-2 family members showed no marked changes in either mRNA or protein that could account for the selective cell death resistance in the S49 (OS 4–25) cells. Overall, the genetic adaptations observed in the S49 (OS 4–25) cells indicate a tendency to promote cell survival under adverse conditions and resistance to apoptosis in this paradigm likely reflects numerous changes in cellular gene expression.
T cells meet their basal energy demands by utilizing an aerobic process to metabolize glucose to pyruvate, which in turn enters the mitochondria to generate ATP via the tricarboxylic acid cycle and oxidative phosphorylation.40 Activated T cells change their metabolic state by utilizing aerobic glycolysis.41, 42, 43 The high rate of glycolysis in activated T cells bypasses the normal oxidation of pyruvate in the mitochondria, resulting in lactic acid fermentation, whereas minimal lactic acid production occurs in resting cells.44 In our study, we show that CD3/CD28-activated T cells from mouse spleen exhibited significant resistance to apoptosis compared with their naïve/resting counterparts, suggesting that heightened glycolysis as reported in activated cells, and observed in the S49 (OS 4–25) cells, can have a key role in the cell death program.
A key finding in our study was the absolute requirement for glucose as an energy source in the S49 (OS 4–25) cells. Even though energy depletion has been classically associated with necrosis, the S49 (OS 4–25) cells in the absence of glucose display many characteristics of apoptosis. Recently, cross talk between cell death and the ability of cells to utilize a variety of carbon substrates has begun to be established,27 and the switch from one substrate to another can have protective effects on a cell fate.45, 46 The generation of S49 (OS 4–25) cells has resulted in a carbon substrate reprogramming characteristic of the Warburg effect, which we show offers a selective survival advantage to these cells and has a critical role in whether a cell lives or dies. Collectively, our data suggest that repetitive stress and recovery results in genetic changes that are not restricted to commonly studied pro- or antiapoptotic genes, and these genetic changes result in a modification of carbon substrate utilization for energy production that impinges on the cell's ability to activate or signal apoptosis.
S49 Neo cells are S49.1 mouse lymphoma cells stably infected with a recombinant amphotropic retrovirus carrying a G418 antibiotic resistance gene.20 Cells were maintained in RPMI-1640 supplemented with 10% heat-inactivated fetal calf serum, 4.8mM glutamine, 100μg/ml streptomycin, and 100U/ml penicillin at 37°C, in 7% CO2 atmosphere. Standard RPMI-1640 was made hyperosmotic by the addition of solid mannitol before the addition of the supplemental components. RPMI-1640 media without glucose or glutamine was purchased through Life Technologies (Grand Island, NY, USA). Dexamethasone was purchased through Steraloids (Wilton, NH, USA), and Fas ligand was purchased through Kamiya Biomedical (Seattle, WA, USA).
S49 Neo cells were exposed to RPMI-1640 media containing 500mM mannitol for 4h at 37°C, in 7% CO2 atmosphere. After this time, the cells were centrifuged at 3000 r.p.m., then returned to normal RPMI-1640 media, and incubated at 37°C, in 7% CO2 atmosphere. Over a period of 7–10 days of culture, the surviving cells regenerated to a viable population of cells in the presence of normal RPMI-1640 media. This protocol was repeated to generate cells that were repetitively exposed to 500mM mannitol for 4h, then recovered for 25 generations.
Cell volume was determined via electronic sizing using a Cell Lab Quanta flow cytometer (Beckman Coulter, Indianapolis, IN, USA) equipped with a 488-nm laser. The electronic volume (EV) channel was calibrated using Flow-Check Fluorospheres (10.4μm; Beckman Coulter). Cells at a density of 1 × 106cells/ml were centrifuged and resuspended in normal or hypertonic RPMI-1640 media and examined as described in the results. An EV gate was used to eliminate cell debris and to exclude portions of cells (apoptotic bodies) that may be formed and released during apoptosis. The percent change in MCV was determined by subtracting the hypertonic MCV value from the normal control MCV value (x100) for each individual time point.
Caspase activity for caspase-3/7-like enzymes was accomplished using a CaspaTag in situ assay kit (Chemicon, Temecula, CA, USA) according to the manufacturer's instructions. Briefly, 1h before cytometric analysis, 300μl of cells were added to 10μl of a 30x CaspaTag reagent working stock and incubation was continued. Immediately before cytometric analysis, the cells were washed in 2ml of CaspaTag wash buffer, resuspended in 500μl of 1x phosphate-buffered saline (PBS), and 2μl of PI (supplied in the kit) was added to each sample. Intracellular calcium was determined by adding Fluo-3 (1μM final; Life Technologies) to each sample for 30min at 37°C, 7% CO2 atmosphere before examination. Caspase inhibitor Q-VD-OPh was purchased from Sigma (St Louis, MO, USA). Analysis of mitochondrial ROS was done by adding MitoSox Red (5μM final; Life Technologies) to each sample for 30min at 37°C, in 7% CO2 atmosphere before examination. Analysis of intracellular glutathione was done by adding monochlorobimane (mBCl; 10μM final; Life Technologies) to each sample for 10min at 37°C, in 7% CO2 atmosphere before examination. PI (10μg/ml; Sigma) or Sytox Blue (5μM; Life Technologies) was used as vital dye and was added immediately before flow cytometric examination. Phosphatidylserine symmetry was determined using Annexin-V conjugated to fluorescein isothiocyanate (Trevigen, Gaithersburg, MD, USA) according to the manufacturer's instructions. All samples were analyzed using a Becton Dickinson LSRII (San Jose, CA, USA) equipped with FACSDiVa software for acquisition and analysis. For each sample, cells were excited with a 488-nm laser and detected at 530nm (CaspaTag, Fluo-3, Annexin-V), a 561-nm laser and detected at 585nm (PI, MitoSox), and/or a 405-nm laser and detected at 440nm (mBCl, Sytox Blue). Only cells that did not lose their membrane integrity (PI negative) were included in the analysis for caspase activity, calcium, ROS, and glutathione.
Cells were initially stained with MitoTracker Deep Red (200nM final; Life Technologies) for 30min at 37°C, in 7% CO2 atmosphere before harvest. The cells were washed once in 1x PBS, resuspended in 500 CytoFix (BD Biosciences, San Jose, CA, USA) while vortexing, and held at room temperature for 30min, after which 1ml of Perm Wash buffer (BD Biosciences) was added. The cells were centrifuged, resuspended in 100μl of Perm Wash buffer, and stored overnight at 4°C. The next day, 2μl of a 1:10 dilution of an Alexa Fluor 488 anti-cytochrome c antibody (BioLegend) and DAPI (50ng/ml final) was added to each sample, incubated at room temperature for 60min, after which 500μl of Perm Wash buffer was added. The cells were centrifuged, resuspended in 200μl of Perm Wash buffer, and examined by flow cytometry using a Becton Dickinson LSRII equipped with FACSDiVa software for acquisition and analysis. For each sample, cells were excited with a 488-nm laser and detected at 530nm (AF488-cytochrome c), a 56-nm laser and detected at 660nm (MitoTracker Deep Red), and a 405-nm laser and detected at 440nm (DAPI).
A total of 120000 cells/well in DMEM XF assay media supplemented with 10mM glucose and 1mM sodium pyruvate were plated in a XF 24-well plate (Seahorse Bioscience, North Billerica, MA, USA), which was initially coated with 34μg/ml Corning Cell Tak (Fisher Scientific, Houston, TX, USA) prepared using 0.1M NaHCO3 (pH 8.0). OCR were determined in situ using a Seahorse Extracellular Flux Analyzer (Seahorse Bioscience). Baseline oxygen consumption was measured three times for 3min following a 3-min mix and 2-min wait period. Following determination of the basal OCR, oligomycin (1μM final), 2,4, DNP (150μM final), and rotenone/antimycin A (0.5μM final) were added consecutively using the same three cycle – 3min – 2min–3min mix, wait, and measure protocol.
Intracellular lactate was determined using an l-lactate assay kit (Fluoro Lactate, Cell Technology Inc., Mountain View, CA, USA) according to the manufacturer's instructions. A lactate standard curve was generated according to the manufacturer's instructions. Samples were read using a POLARstar omega microplate reader (BMG Labtech, Cary, NC, USA) with an excitation/emission of 544/590nm, respectively.
DNA from the S49 (Neo) and S49 (OS 4–25) cells was isolated using DNeasy Blood & Tissue Kit (Qiagen, Germantown, MD, USA) according to the manufacturer's instructions. Samples were quantified using a spectrophotometer (Nanodrop-ND-1000; Thermo Scientific, Waltham, MA, USA), and purity was analyzed by the 260/280nm absorbance ratios. Five micrograms of DNA was analyzed on a 1.8% agarose gel using 1x TAE (tris-acetate-EDTA) running buffer. Before casting the gel, GelRed Nucleic Acid Stain (Phenix Research Products, Candler, NC, USA) was added at a 1:10000 dilution.
Total RNA from the S49 (Neo) and S49 (OS 4–25) cells was isolated using the QIAshredder and RNeasy mini-kit (Qiagen) according to the manufacturer's instructions. DNase treatment was performed on column using a ribonuclease-free DNase kit (Qiagen) according to the manufacturer's instructions. Samples were quantified using a spectrophotometer (Nanodrop-ND-1000; Thermo Scientific), and purity was analyzed by the 260/280nm absorbance ratios.
Quantitative RT-PCR was performed with a 7900HT sequence detection system using predesigned primer–probe sets (Life Technologies) according to the manufacturer's instructions. RT-PCR assays were performed using the two-step method: cDNAs were generated from 4μg of total RNA using reverse transcriptase reagents (Life Technologies), then RT-PCR was performed using the universal master mix (Life Technologies). The signal obtained from each gene primer–probe set was normalized to the housekeeping gene peptidylprolyl isomerase B. The delta–delta Ct method of analysis and RQ transformation were used to plot the data.
C57BL/6 mice were used and were cared for according to the guidelines of the Animal Care and Use Committee of the National Institute of Environmental Health Sciences, National Institutes of Health. Mouse naïve CD4+ T cells were isolated from the spleen using the EasySep negative selection isolation kit from Stem Cell Technologies (Vancouver, BC, Canada) according to the manufacturer's instructions. Naïve CD4+ T cells were stimulated using Dynabeads Mouse T-Activator CD3/CD28 from Life Technologies at a 1:1 bead-to-cell ratio according to the manufacturer's instructions for 48h. After 48h, Dynabeads were removed before subsequent experimentation.
Gene expression analysis was conducted using Agilent Whole Mouse Genome 4 × 44 multiplex format oligo arrays (Agilent Technologies, Englewood, CO, USA) following the Agilent 1-color microarray-based gene expression analysis protocol. Starting with 500ng of total RNA, Cy3-labeled cRNA was produced according to the manufacturer's protocol. For each sample, 1.65μg of Cy3-labeled cRNAs were fragmented and hybridized for 17h in a rotating hybridization oven. Slides were washed and then scanned with an Agilent Scanner. Data were obtained using the Agilent Feature Extraction software (v9.5), using the 1-color defaults for all parameters. The Agilent Feature Extraction Software performed error modeling, adjusting for additive and multiplicative noise. The resulting data were processed using the Rosetta Resolver system (version 7.2; Rosetta Biosoftware, Kirkland, WA, USA). Principal component analysis was performed on all samples and all probes to characterize the variability present in the data. Differentially expressed probes were identified using analysis of variance (ANOVA) to determine if there was a statistical difference between the means of groups. The number of false positives was reduced using a multiple test correction. Specifically, an error-weighted ANOVA and Benjamini–Hochberg multiple test correction with a P-value of <0.0 were performed with Rosetta Resolver (Rosetta Biosoftware). Significantly regulated genes were analyzed using IPA (Ingenuity Systems, Redwood City, CA, USA). The microarray data discussed in this manuscript have been deposited in the NCBI's Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/) and are accessible through GEO accession no. GSE62531.
One-way ANOVA followed by Tukey's multiple comparison tests or paired t-tests were used to evaluate the statistical relevance of control and experimental samples. A P-value of at least <0.05 was considered significant.
We thank Laura Wharey and Kevin Gerrish of the National Institute of Environmental Health Sciences microarray facility for the generation and initial analysis of the microarray data and Dr Karen Debalsi for her helpful discussion on the Seahorse Technology. This research was supported by the Intramural Research Program of the NIH, National Institute of Environmental Health Sciences.
The authors declare no conflict of interest.
Supplementary Information accompanies this paper on Cell Death and Differentiation website (http://www.nature.com/cdd)
Edited by DR Green