|Home | About | Journals | Submit | Contact Us | Français|
Increasing evidence suggests that cancer cells (relative to normal cells) have altered mitochondrial electron transport chains (ETC) that are more likely to form reactive oxygen species (ROS; i.e., O2•- and H2O2) resulting in a condition of chronic metabolic oxidative stress, that maybe compensated for by increasing glucose and hydroperoxide metabolism. In the current study, the ability of an inhibitor of glucose metabolism, 2-deoxy-D-glucose (2DG), combined with mitochondrial electron transport chain blockers (ETCBs) to enhance oxidative stress and cytotoxicity was determined in human colon cancer cells. Treatment of HT29 and HCT116 cancer cells with Antimycin A (Ant A) or rotenone (Rot) increased carboxy-dichlorodihydrofluorescein diacetate (H2DCFDA) and dihydroethidine (DHE) oxidation, caused the accumulation of glutathione disulfide and enhanced 2DG-induced cell killing. In contrast, Rot did not enhance the toxicity of 2DG in normal human fibroblasts supporting the hypotheses that cancer cells are more susceptible to inhibition of glucose metabolism in the presence of ETCBs. In addition, 2-methoxy-antimycin A (Meth A; an analog of Ant A that does not have ETCB activity) did not enhance 2DG-induced DHE oxidation or cytotoxicity in cancer cells. Finally, in HT29 tumor bearing mice treated with the combination of 2DG (500 mg/kg) + Rot (2 mg/kg) the average rate of tumor growth was significantly slower when compared to control or either drug alone. These results show that 2DG-induced cytotoxicity and oxidative stress can be significantly enhanced by ETCBs in human colon cancer cells both in vitro and in vivo.
Colorectal cancer is the third most common cancer in both men and women, accounting for 10% of all cancer deaths in the United States. The National Cancer Institute has made it a research priority to define the biomolecular and cellular changes that occur during the development of colon cancer as well as determine if this information can be used for drug development and prediction of sensitivity to anti-cancer agents.1 It has been hypothesized, with increasing supportive evidence, that cancer cells may have alterations in mitochondrial electron transport chains (ETC) that are more likely to cause univalent reduction of O2 to form reactive oxygen species (ROS; i.e., O2 •- and H2O2) resulting in a chronic condition of metabolic oxidative stress.2-8 In support of this idea, mitochondria from human colorectal cells have been shown to have higher levels of ROS as well as having increased oxidative damage, as measured by TBARS and protein carbonyl content compared with mitochondria from normal adjacent colon tissue.9 If disruptions in electron flow through cancer cell mitochondrial ETCs contribute to increased steady-state levels of ROS, then it logically follows that cancer cells may be more vulnerable to interventions that perturb the ETC system selectively killing cancer vs. normal cells via ROS-mediated oxidative stress.
Malignant cells also have significantly increased rates of glucose metabolism when compared with normal cells and it has been hypothesized that glucose provides reducing equivalents necessary for H2O2 detoxification via the formation of pyruvate (that scavenges hydrogen peroxide directly)10,11 as well as through the regeneration of the reduced nicotinamide adenine dinucleotide phosphate (NADPH) as the source of electrons for glutathione and thioredoxin dependent peroxidase enzymes.12-14 We and other investigators have shown that glucose deprivation selectively induces oxidative stress and cell death in cancer cells vs. normal cells in a variety of cell lines including colon cancer.2,6,7,15,16 Consistent with the idea that mitochondrial ROS might contribute to the increased susceptibility of cancer cells to glucose deprivation, we and others have previously shown that electron transport chain blockers (ETCBs) increased production of H2O2 and O2•- as well as enhanced cytotoxicity with glucose deprivation in human colon and breast carcinoma cell lines7,15 as well as in glioblastoma cells6 vs. their normal cell counterparts.
To determine if a clinically relevant inhibitor of glucose metabolism, 2-deoxy-D-glucose (2DG), is capable of mimicking the effect of glucose deprivation, the ability of 2DG combined with ETCBs to cause selective cytotoxicity and oxidative stress in colon cancer vs. normal cells was tested. In this study mitochondrial ETCBs (Rot and Ant A) were found to increase intracellular ROS, disrupt glutathione metabolism causing increases in % glutathione disulfide (%GSSG), as well as significantly enhancing the cytotoxicity of 2DG in the human colon cancer cell lines (HCT116 and HT29) but not in normal human fibroblasts in vitro. These results were confirmed in vivo when inhibition of colon cancer cell growth in nude mice was found to be maximal with 2DG + Rot vs. either agent alone. These results support the hypothesis that agents which increase mitochondrial production of ROS can enhance the anti-tumor activities of clinically significant inhibitors of glucose metabolism in colon cancer.
To test the hypothesis that treatment with 2DG and mitochondrial ETCBs enhances steady-state levels of intracellular prooxidants (presumably hydroperoxides), human colon carcinoma cells HT29 and HCT116 were treated with 2DG for 24 h and the last 2 h of treatment co-incubated with ETCB, 10 μM Ant A. At the end of treatment the cells were labeled with the oxidation sensitive dye H2DCFDA that detects changes in intracellular prooxidants. In both cell lines 2DG treatment alone resulted in a slight increase in H2DCFDA oxidation [as determined by mean fluorescence intensity (MFI)] compared to control cells (Fig. 1A and B, p < 0.05 HT29 only). Ant A (an ETC complex III inhibitor) treatment alone also resulted in a significant increase in MFI in both cell lines (Fig. 1A and B, p < 0.05). In addition, 2DG combined with Ant A treatment in HT29 cells resulted in a 6-fold increase in H2DCFDA oxidation when compared to control and at least a three-fold increase when compared to 2DG or Ant A treatment (Fig. 1A, p < 0.001). HCT116 cells treated with 2DG + Ant A also resulted in a significant, 4-fold, increase in H2DCFDA oxidation when compared with control or 2DG alone, however the increase was not greater than that seen with Ant A alone (Fig. 1B). To ensure that the changes in fluorescence seen in the presence of 2DG and/or ETCBs were due to alterations in H2DCFDA oxidation (and not changes in uptake, ester cleavage or efflux of the probe) similarly treated cells were labeled with the already oxidized analog of H2DCFDA (DCFDA). When this was done no significant changes in fluorescence were noted in the presence or absence of Ant A ± 2DG (data not shown) confirming that the changes in MFI seen in the presence of Ant A are indicative of changes in quantities of the dye being oxidized by the cells. Overall, the results in Figure 1A and B indicated that the combination of a complex III inhibitor ± 2DG caused increases in steady-state levels of intracellular prooxidants in human colon cancer cells.
In order to determine if intracellular O2•- was contributing to the increase in steady-state levels of prooxidants detected using the non-specific probe H2DCFDA, the experiments shown in Figure 1A and B were repeated measuring oxidation of dihydroethidine (DHE), where fluorescence intensity is thought to correlate with relative specificity to intracellular steady state O2•- levels.7 HT29 and HCT116 were treated with 2DG for 24 h with 10 μM Ant A or 50 μM Rot (an ETC complex I inhibitor) added during the last 2 h of incubation, labeled with DHE and analyzed by flow cytometry (Fig. 1C and D). In both cell lines 2DG treatment resulted in a slight but non-significant increase in DHE oxidation when compared to control cells (Fig. 1C and D). Treatment with ETCBs (Ant A and Rot) alone or combined with 2DG resulted in 3- to 6-fold increases in DHE oxidation when compared to control (p < 0.01, Fig. 1C and D). Overall, the data in Figure 1 are consistent with the hypothesis that 2DG in combination with either a complex I or complex III ETCBs increase steady-state levels of prooxidants including superoxide in human colon cancer cells.
Ant A has long been known as a potent inhibitor of the mitochondrial respiratory chain which binds to the quinone reduction site of the cytochrome bc1 of complex III.17 However recently a cytotoxic mechanism of action other than electron transport chain blockade has been suggested for Ant A. Hockenberry et al. reported Ant A binds to the hydrophobic groove of the anti-apototic proteins Bcl-2 and Bcl-xL thereby inducing apoptosis.18 2-Methoxy-antimycin A (Meth A) is an analog of Ant A which maintains Bcl-2/Bcl-xL binding activity but does not inhibit electron transport.19 When colon cancer cells were treated with 10 or 25 μM Meth A in the presence or absence of 2DG, no significant changes in DHE oxidation were noted in either cell line (Fig. 1C and D and data not shown). These results clearly show that the ETC blocking activity of Ant A and not the Bcl-2/Bcl-xL binding activity caused the increase in steady-state levels of O2•- in colon cancer cells.
The glutathione (GSH)/glutathione disulfide (GSSG) redox couple is the most abundant soluble thiol redox buffer in the cell and the ratio of GSH to GSSG is considered a good indicator of intracellular redox status as well as providing a source of reducing equivalents that protects cells from oxidative stress.20 In HT29 cells, total glutathione content was significantly decreased in cells treated with Rot or 2DG alone along with 2DG in combination with either Ant A or Rot (Fig. 2A). The percent glutathione disulfide was significantly increased when HT29 were treated with 2DG in combination with either ETCB compared with any of the agents alone (Fig. 2C). In HCT116 cells, total glutathione was significantly decreased when cells were treated with Rot alone (Fig. 2B), however the %GSSG was significantly increased with Rot or 2DG alone or in combination with an ETCBs (Fig. 2D). Taken together the data in Figure 2 suggests that 2DG in combination with ETCBs cause disruption in intercellular thiol metabolism indicative of changes in redox status consistent with cells undergoing oxidative stress.
We have previously shown that ETCBs enhance glucose deprivation-induced cytotoxicity in several human cancer cell lines.7,15 To extend these observations to a clinically relevant drug that mimics glucose deprivation, HT29 and HCT116 cells were treated with 20 mM 2DG for 24 h, with 10 μM Ant A, 50 μM Rot or 25 μM Meth A added during the last 2 h (Fig. 3A and B). Treatment with either Ant A or Rot alone did not result in significant cell killing (as determined by clonogenic survival) in either of the two cell lines (Fig. 3A and B). Treatment with 2DG alone caused modest but significant (~20%, p < 0.05) cell killing in both HT29 (Fig. 3A) and HCT116 (Fig. 3B) cells, relative to control. The addition of Rot or Ant A to 2DG treatment caused much more pronounced cell killing [70% in HT29 and 65% in HCT116 (p < 0.01)], relative to control (Fig. 3A and B). These results show that the combination of ETCBs with 2DG caused greater than additive cytotoxicity in both human colon cancer cell types tested. In contrast, Meth A did not enhance 2DG-induced cell killing, clearly indicating that the ETCB ability of Ant A (and not the Bcl protein binding activity) is responsible for the enhanced cell killing (Fig. 3) and increased prooxidant production (Fig. 1) noted in colon cancer cells. These results provide strong support for the hypothesis that treatments with mitochondrial ETCBs that increase prooxidant production significantly enhance the cytotoxicity of a clinically relevant inhibitor of glucose metabolism in colon cancer cells.
Recent studies have suggested that cancer cells demonstrate fundamental defects in mitochondrial oxidative metabolism that render them more vulnerable (relative to normal cells) to agents that induce oxidative stress and inhibit glucose metabolism.5-7 Consistent with these previous findings, two normal human fibroblast cell isolates (GM00038C and GM02037C) did not demonstrate enhanced 2DG-induced cytotoxicity when treated with 2DG + 50 μM Rot (Fig. 3C and data not shown). These results support the hypothesis that 2DG and agents that increase mitochondrial prooxidant production (i.e., ETCBs) can be combined to selectively inhibit tumor growth in vitro.
The LD50 in mice for Ant A has been reported to be 1.5 mg/kg when administered i.p.21 and this excessive toxicity preclude its use in animal studies. Previous reports indicated two hours after acute i.p. administration of 2 mg/kg Rot, liver levels of 0.05 ppm (~127 nM) were achieved, and well-tolerated in mice.22 In order to determine the minimum dose of Rot required to achieve maximal enhancement of 2DG-induced cancer cell killing, Rot doses of 1 nM to 10 μM were tested in vitro using the clonogenic survival assay (Fig. 4A and B). HT29 and HCT 116 cells treated with 2DG for 24 h as previously described and with 100 nM Rot added for the last hour, enhanced 2DG-induced cell killing (Fig. 4A and B; see Fig. 3 for 2DG and Rot alone data). To test if treatment with this lower dose and exposure time of Rot is effective at increasing steady-state levels of intracellular prooxidants we repeated the H2DCFDA experiment. When HCT116 cells were treated with 100 nM Rot for 1 h, a slight increase in H2DCFDA oxidation was noted that was significantly enhanced when combined with 24 h of 2DG treatment (p < 0.05; Fig. 4C). These results demonstrate that 2 mg/kg Rot in mice (which is expected to result in ~127 nM Rot) should also be effective at enhancing oxidative stress and 2DG-induced tumor cell killing in our model system.
HT29 tumor bearing athymic nude mice were divided into four groups and treated with 2 mg/kg Rot, 500 mg/kg 2DG, or the combination, 5 d/w for 2 w. None of the 23 mice treated demonstrated acute toxicity as measured by weight loss (no trend in weight change in any of the four treatment groups p = 0.793) and the mice appeared healthy for the duration of the experiment. Treatment with 2DG alone did not result in a significant difference in the rate of tumor growth when compared to control mice (Fig. 5A and B). However, Rot alone did result in a slight but significant decrease in tumor growth rate when compared to control (Table 1 and Fig. 5A and C). Most importantly, the combination of 2DG + Rot resulted in a highly significant decrease in tumor growth rate when compared to all three other groups (Table 1 and Fig. 5A—D). These data strongly support the conclusion that the combination of inhibitors of glucose metabolism (i.e., 2DG) with mitochondrial ETCBs that increase oxidative stress (i.e., Rot) are able to enhance tumor growth inhibition in vivo without overt signs of morbidity and mortality.
During oxidative phosphorylation within the inner mitochondrial membrane, electrons are shuttled down electron transport chain complexes I-IV resulting in the production of a transmembrane proton potential gradient which is coupled to the production of ATP through the ATP synthase in complex V. In normal cells, as much as 1% of the electrons flowing through the ETCs are thought to undergo one-electron reductions of O2 to form O2•- which can undergo further chemical and enzymatic reactions to become H2O2 and other ROS.7,23-25 Cellular antioxidant detoxification pathways (i.e., glutathione/glutathione peroxidase and thioredoxin/peroxiredoxin pathways) scavenge ROS and prevent them from causing intracellular damage including lipid peroxidation, DNA damage and protein oxidation.14,20,26,27 However, in cancer cells, as opposed to normal cells, it has been hypothesized that alterations in ETC structure and function could result in increased steady-state levels of ROS (i.e., superoxide and hydrogen peroxide), resulting in cancer cells existing in a chronic condition of metabolic oxidative stress that is compensated for by increasing hydroperoxide metabolism using reductants derived from glucose.7,15,28-30
This chronic condition of oxidative stress in cancer cells (relative to normal cells) has been suggested to contribute to the high rates of mitochondrial DNA mutations31,32 as well as significant variability in levels of glutathione and thioredoxin dependent hydroperoxide detoxifying systems.33 This heterogeneity in cancer cell lines also results in variability in measurements of steady-state levels of prooxidants using oxidation sensitive probes. Consistent with this idea of tumor cell heterogeneity in redox biology, HCT116 (reviewed in ref. 7) and SW480,7 colon cancer cell baseline levels of O2•- and hydroperoxides measured using DHE and H2DCFDA were found to be 3-fold higher than HT29 colon cancer cells. However, more importantly, all colon cancer cell lines tested (HT29, HCT116, SW480) demonstrated at least three- to five-fold greater steady-state levels of prooxidants, relative to normal colon epithelial and fibroblast cells, that significantly contributed to the differential susceptibility of these cancer cells to glucose deprivation-induced oxidative stress.7
Glucose metabolism is known to play a major role in the detoxification of peroxides via the formation of pyruvate, which can scavenge peroxides directly through a deacetylation reaction11,27 and through regeneration of NADPH which acts as a co-factor in both the glutathione/glutathione peroxidase and thioredoxin/peroxiredoxin pathways.12-14,26 2DG is converted by hexokinase to 2-deoxy-glucose-6-phosphate, a compound that is not a substrate for further steps in glycolysis effectively restricting the production of pyruvate by glycolysis. However, 2-deoxy-glucose-6-phosphate (2DG-6-P) can enter the pentose cycle to be acted upon by glucose-6-phosphate dehydrogenase (G6PD) to regenerate one molecule of NADPH, however, the product of that reaction 2DG-6-phosphoglaconalactone is not a substrate for the second enzyme in the pentose cycle. This theoretically reduces by 50% the number of molecules of NADPH that can be regenerated in the pentose cycle from one molecule of 2DG (relative to glucose). Previous investigations have shown that treatment with 2DG results in depletion of NADPH, disruptions in glutathione metabolism, as well as increases in steady-state levels of hydroperoxides, rendering cancer cells more susceptible to oxidative stress by agents that are known to induce free radicals including ionizing radiation and chemotherapy.7,34-37 Consistent with these previous reports in other cancer cells, the data in the current report shows treatment of human colon cancer cells with 2DG (resulting in ~20% clonogenic cell killing) was able to cause a slight increase in steady-state levels of intracellular prooxidants as well as a disruption in the thiol redox balance (as demonstrated by a decrease in GSH in HT29 and an increase in %GSSG in HCT116).
Blocking electron movement down ETCs, using compounds such as Ant A and Rot, has long been known to increase mitochondrial production of O2•- and H2O2 23,25. More recently, Pelicano et al. demonstrated that partial inhibition of mitochondrial respiration with Rot enhanced univalent reduction of O2 from electron transport chains, leading to increases in O2•- and sensitization of leukemia cells to anticancer agents whose action involve free radical generation.5 Furthermore, Liu et al. found that 2DG effectively sensitized osteosarcoma cells to clonogenic cell killing with inhibitors of oxidative phosphorylation including Ant A and Rot.38 If the probability of O2•- formation is greater from tumor cell ETCs, relative to normal cells, it follows that ETC blockers might selectively enhance oxidative stress and cell killing by inhibitors of glucose metabolism in cancer cells, relative to normal cells.
Consistent with these ideas, the current study shows Ant A and Rot increased steady-state levels of intracellular pro-oxidants and GSSG in HT29 and HCT116 human colon cancer cell lines. Furthermore, 2DG was able to enhance the steady-state levels of H2DCFDA oxidation when added with 10 μM Ant A for 2 h in HT29, but not in HCT116 (Fig. 1). It is possible that high levels of H2DCFDA oxidation seen with Ant A in HCT116 (four-fold greater than untreated HCT116) made it difficult to detect any further increase of H2DCFDA oxidation in the presence of 2DG. Consistent this idea, when HCT116 cells were treated with 100 nM Rot for 1 h, H2DCFDA oxidation was significantly enhanced when combined with 2DG treatment (p < 0.05; Fig. 4C). More importantly, both HCT116 and HT29 cells were dramatically sensitized to clonogenic cell killing in the presence of 2DG by either Ant A or Rot (Figs. (Figs.33 and and4)4) and this cell killing did not occur in normal fibroblasts treated with 2DG + Rot (Fig. 3C). These results support the hypothesis that mitochondrial dysfunction in cancer cells (relative to normal cells) may selectively enhance oxidative stress and cell killing mediated by exposure to inhibitors of glucose metabolism combined with ETC blockers.
Ant A binds to the quinone reduction site of cytochrome bc1 inhibiting the transfer of electrons to coenzyme Q on the matrix side of the inner mitochondrial membrane which is thought to result in an increase in O2•- production from ubisemiquinone (Q•). In addition to blocking electron transport, Ant A has also been reported to bind the hydrophobic groove of Bcl-2 and Bcl-xL thereby inducing apotosis.18 Meth A is an analog of Ant A which maintains binding to Bcl-xL 19 but unlike Ant A, does not inhibit electron transport.39,40 Since HCT116 and HT29 express Bcl-2 and Bcl-xL41 in the current study Meth A was compared to Ant A to determine if ETC blocking activity was essential for the observed biological effects. When HCT116 and HT29 were exposed to 2DG in the presence and absence of Ant A and Meth A, only exposure to Ant A + 2DG resulted in cytotoxicity or increases in steady-state levels of O2•-, supporting the hypothesis that the ETC blocking activity of Ant A was critical to enhancing oxidative stress and cell killing by inhibitors of glucose metabolism.
2DG has also been shown to have anticancer effects and has been used safely in both animal models and humans.42-44 Rot, a naturally occurring plant toxin which works by inhibiting NAD+ linked oxidation in complex I of the ETC, along with deguelin (a similar rotenoid) have been shown to have cancer chemo preventative activity.45-47 Recently, Kim et al. demonstrated that administration of deguelin resulted in profound inhibition of tumor growth and angiogenesis when combined with radiation in a mouse model of non-small cell lung cancer.48 In order to determine if inhibitors of ETCBs could be used to enhance the antitumor activity of 2DG in vivo, tumor bearing mice were treated with 0.5 mg/kg 2DG combined with 2 mg/kg Rot 5 d/w for 2 w. Consistent with the in vitro data, the combination of 2DG + Rot demonstrate significantly increased tumor growth inhibition in HT29 human colon cancer xenografts grown in nude mice. Furthermore this drug combination was well tolerated by the mice as determined by a lack of significant weight loss. Overall the results in the current report provide strong support for the hypothesis that agents which increase mitochondrial production of ROS via ETC blocking activity do enhance oxidative stress and the anti-tumor activities of a clinically significant inhibitor of glucose metabolism (2DG) and suggest that this biochemical rationale could be used to selectively kill colon cancer vs. normal cells.
HCT116 and HT29 human colon carcinoma cells were obtained from American Type Culture Collection (ATCC) and maintained in RPMI 1640 media with 10% fetal bovine serum (HyClone). Cultures were maintained in 5% CO2 and humidified in a 37°C incubator. Normal non-fetal skin fibroblasts cells (GM00038C and GM02037C) from Coriell Institute were maintained in Eagles modified MEM with Earls salt media supplemented with 10 ml non-essential amino-acids (100X MEM-NEAA solution GIBCO 11140), 20 mL essential amino-acids (50X MEM-AA solution GIBCO 11130), 10 mL vitamins (100X MEM-vitamin solution Sigma M6895) and 15% fetal bovine serum with the pH maintained with sodium bicarbonate.
2-deoxy-D-glucose (2DG), Rotenone (Rot) and Antimycin A (Ant A) were obtained from Sigma Chemical Co., (St. Louis, MO). 2-methoxyantimycin A (Meth A) was obtained from BIOMOL Research Labs, Inc., (Plymouth Meeting, PA). All drugs were used without further purification. Drugs were added to cells at a final concentration of 20 mM 2DG, 10 μM Ant A, 0.1–50 μM Rot and 10–25 μM Meth A. All stock solutions were dissolved in dimethylsufoxide (DMSO) except 2DG was dissolved in water, and the required volume added directly to complete cell culture media on cells to achieve the desired final concentrations. Cells were treated with or without 2DG for 24 h while Rot, Ant A and Meth A were added for the last 1 or 2 h of the experiment as indicated. Final concentration of DMSO in the culture dish did not exceed 0.1%. All cells were placed in a 37°C incubator and harvested at the time points indicated.
1.5 × 105 HCT116, 2.0 × 105 HT29, 1.2 × 105 GM00038C and GM02037C cells were plated in 60 mm dishes and allowed to grow in their respective stock culture media for 48 h. To ensure physiological concentrations of glucose at the time of exposure to drugs all treatments were done in RPMI (11 mM glucose) with 10% FBS. Floating cells in medium were collected and combined with the attached cells from the same dish that were trypsinized with 1 mL trypsin-EDTA (CellGro, Herndon, VA). Samples were centrifuged and cells were counted using a Beckman Coulter Counter. Cells were plated at low density (150–400 cells per dish), and clones were allowed to grow for 14 days in maintenance media with 0.1% gentamycin added. Cells were then fixed with 70% ethanol and stained with coomassie blue for analysis of clonogenic survival as previously described.49 Individual assay colony counts were normalized to that of control (DMSO alone) with at least four cloning dishes per condition, repeated in at least three separate experiments. For the survival analysis, the number of cells per dish following exposure to combinations of 2DG with Ant A, Rot or Meth A was normalized to the number of cells present in 2DG alone treatment dishes immediately before the addition of ETCBs to account for the loss of cell sized particles caused by the sever toxicity of drug combination.
3.75 × 105 HCT116 cells and 5.0 × 105 HT29 cells were plated in 100 mm dishes and treated as stated above. Immediately following treatment, media was collected and centrifuged from 100 mm dishes, then cells were scrape harvested in PBS on ice, centrifuged at 4°C, and the PBS was discarded. Cell pellets were transferred to 1.5 mL eppendorf tubes and stored at −20°C. Cell pellets were thawed and homogenized in 50 mM phosphate buffer pH 7.8 containing 1.34 mM diethylenetriamine-pentaacetic acid (DETAPAC). Total glutathione (GSH) content was determined by the method of Griffith.50,51 Glutathione disulfide (GSSG) was measured by adding 2 μL of a 1:1 mixture of 2-vinylpyridine and ethanol per 30 μL of sample and incubating for 2 h prior to assaying as described previously.51 GSH determinations were normalized to the protein content of whole homogenates using the Lowry method.52
Cells were grown and treated as stated above. Treatment media was removed, and cells were labeled with 10 μM dihydroethidium for 40 min (DHE, Molecular Probes, Eugene OR) or 10 μg/mL 5-(and 6)-carboxy-2,7-dichlorodihydrofluorescein diacetate for 15 min (H2DCFDA, Molecular Probes, Eugene, OR) in PBS with 0.1% DMSO at 37°C. In experiments measuring DHE oxidation pyruvate was added as a metabolic substrate. Ant A (10 μM) was used as a positive control.29 Cells were then trypsinized, suspended in EMEM media and centrifuged for collection. Cell pellets were then resuspended in 400 μL PBS and filtered through mesh. Samples were analyzed using a FACScan flowcytometer (Becton Dickinson Immunocytometry System, INC., Mountain View, CA) (excitation 488 nm, emission 585 nm for DHE and emission 530 nm H2DCFDA band-pass filters). The mean fluorescence intensity (MFI) of 10,000 cells was analyzed in each sample and corrected for auto-fluorescence from unlabeled cells. The MFI data was normalized to corresponding control group levels for each cell type.29
Six-week-old female athymic-nu/nu mice were purchased from Harlan Laboratories (Indianapolis, IN). All mice were housed in a pathogen-free barrier room in the Animal Care Facility at the University of Iowa and handled using aseptic procedures. All procedures were approved by the IACUC committee of the University of Iowa and conformed to the guidelines established by NIH. Mice were allowed a week to acclimate prior to beginning experimentation, and food and water were made freely available. Tumor cells were inoculated into nude mice by subcutaneous injection of 150 μL aliquots of phosphate buffered saline containing 3.5 × 106 HT29 cell suspension into the right flank using 26-gauge needles. Mice were weighed and evaluated daily. Tumor volumes were calculated using the formula: tumor volume = (length × width2)/2 where the length was the longest dimension and width was the dimension perpendicular to length. Four d after the cells were injected the tumors measured an average of 44 mm3 and the mice were started on drug treatment. The mice were divided into four groups (n = 5–6 mice/group). 2DG group: 2DG was dissolved in saline and administered i.p. 0.5 g/kg; Rot group: Rot was dissolved in a 1:1:6 Cremaphor EL®/ethanol/saline mixture and administered 2 mg/kg i.p; Combination group: 2DG + Rot mice were administered both drugs in succession at doses mentioned above; Control group: control mice were injected with 100 μl Cremaphor EL®, ethanol and saline mixture (Rot vehicle). Treatment for all groups was 5 d/w for 2 w. Mice were euthanized via CO2 gas asphyxiation when tumor length exceeded 1.5 cm in any dimension.
Statistical analysis of in vitro data was performed using GraphPad Prism 4 software. To determine differences between three or more means, one-way analysis of variance followed by a post-hoc Tukey test were performed. For analysis of the in vivo data, tumor measurements were serially recorded and their corresponding volumes calculated for each mouse. Linear mixed effects regression models were used to estimate and compare the group-specific change in weight and tumor growth curves. All tests were two-sided and carried out at the 5% level of significance. Analyses were performed with the SAS statistical software package.
Supported by, NIH RO1-CA100045, RO1-CA133114, T32-CA078586, P30-CA086862 and R21-CA139182A.