Bioenergetic properties of human sarcoma cells help define sensitivity to metabolic inhibitors

Abstract

Sarcomas represent a diverse group of malignancies with distinct molecular and pathological features. A better understanding of the alterations associated with specific sarcoma subtypes is critically important to improve sarcoma treatment. Renewed interest in the metabolic properties of cancer cells has led to an exploration of targeting metabolic dependencies as a therapeutic strategy. In this study, we have characterized key bioenergetic properties of human sarcoma cells in order to identify metabolic vulnerabilities between sarcoma subtypes. We have also investigated the effects of compounds that inhibit glycolysis or mitochondrial respiration, either alone or in combination, and examined relationships between bioenergetic parameters and sensitivity to metabolic inhibitors. Using 2-deoxy-D-glucose (2-DG), a competitive inhibitor of glycolysis, oligomycin, an inhibitor of mitochondrial ATP synthase, and metformin, a widely used anti-diabetes drug and inhibitor of complex I of the mitochondrial respiratory chain, we evaluated the effects of metabolic inhibition on sarcoma cell growth and bioenergetic function. Inhibition of glycolysis by 2-DG effectively reduced the viability of alveolar rhabdomyosarcoma cells vs. embryonal rhabdomyosarcoma, osteosarcoma, and normal cells. Interestingly, inhibitors of mitochondrial respiration did not significantly affect viability, but were able to increase sensitivity of sarcomas to inhibition of glycolysis. Additionally, inhibition of glycolysis significantly reduced intracellular ATP levels, and sensitivity to 2-DG-induced growth inhibition was related to respiratory rates and glycolytic dependency. Our findings demonstrate novel relationships between sarcoma bioenergetics and sensitivity to metabolic inhibitors, and suggest that inhibition of metabolic pathways in sarcomas should be further investigated as a potential therapeutic strategy.

Introduction

Sarcomas represent a diverse group of mesenchymal malignancies that arise from connective and soft tissues, including bone, muscle, and cartilage. Sarcomas affect approximately 200 000 individuals worldwide each year and represent a higher percentage of overall cancer morbidity and mortality in children and adolescents than in adults.^1,2 Research identifying alterations associated with specific histological subtypes of sarcoma has indicated that previous classifications based on the site of tumor (bone or soft tissue) are less important than the tumor molecular and pathological features.¹ Thus, a better understanding of the genetic and molecular alterations present in specific sarcoma subtypes is critically important to improve sarcoma treatment and develop new therapeutic approaches.

Rhabdomyosarcoma (RMS) is the most common soft tissue sarcoma of childhood and adolescence, with approximately 350 new cases diagnosed each year in the United States.³ Like other sarcomas, RMS tumors are molecularly diverse.⁴ There are 2 major histologic subtypes of RMS, embryonal (eRMS) and alveolar (aRMS), each having distinct underlying genetic alterations that participate in pathogenesis.³ Patients with eRMS generally have a more favorable prognosis than patients with aRMS, which has a 5-year survival rate of <50%.⁴ aRMS is further characterized by chromosomal translocations, resulting in fusion genes such as PAX3-FOXO1. PAX3-FOXO1-positive aRMS patients, with metastatic disease survive in <10% of cases.⁴ Current treatment for both types of RMS includes surgery, chemotherapy (vincristine, actinomycin-D, and cyclophosphamide), and radiation, with an overall survival rate of about 70%. Despite this aggressive multimodal treatment, patients with metastatic disease have an overall survival rate of <20%.^3-5

Osteosarcoma is the most common primary tumor of bone and mainly affects adolescents and young adults.^6,7 Osteosarcoma is the eighth most common malignancy of childhood, with approximately 400 new cases diagnosed in children and young adults in the United States each year.⁷ Like other sarcomas, osteosarcoma tumors are molecularly diverse.⁶ Despite treatment protocols that combine chemotherapy, surgery, and sometimes radiotherapy, the 5-year survival rate for patients diagnosed with osteosarcoma is 60–70%.^6,8 The long-term survival rate for patients with metastatic disease detectable at presentation is 30–35%.⁷ Current treatments for osteosarcoma are associated with significant morbidity, and there have been no significant improvements in prognosis in the last 2 decades, hence there is a significant need for improved therapies for osteosarcoma.^6,7

Many signaling pathways that are affected by genetic events in cancer, as well as the tumor microenvironment, significantly alter cellular bioenergetics to support growth and survival.⁹ Cancer cells exhibit a metabolic phenotype known as aerobic glycolysis, or the Warburg effect, which is characterized by increased glycolysis regardless of oxygen availability.^9,10 Alterations in cellular metabolism are recognized as a crucial hallmark of cancer.¹¹ Exploiting the fundamental differences between cancer and normal cell metabolism may provide an opportunity for therapeutic intervention through selective targeting of the metabolic dependencies of cancer cells. Despite a lack of drugs that target specific metabolic enzymes, several readily available compounds, including 2-deoxy-D-glucose (2-DG), a competitive inhibitor of glycolysis, and metformin, a widely used diabetes drug that can inhibit complex I of the mitochondrial respiratory chain, have been described as having effects on cancer cell growth and cellular bioenergetics.⁹ Additionally, several recent studies have shown that simultaneous inhibition of glycolysis and mitochondrial respiration can act synergistically to reduce tumor cell survival in vitro and in vivo.^12-14 With respect to sarcomas, 2-DG induces apoptosis in human aRMS, but not eRMS, cell lines and inhibits osteosarcoma tumor growth (in combination with doxorubicin) and metastasis, although the mechanism and contribution of effects on cellular bioenergetics are not well-characterized.^5,15,16

Here, we describe the cellular bioenergetics of a panel of human RMS and osteosarcoma cell lines and examine the relationships between bioenergetic properties and sensitivity to metabolic inhibitors. We show that sarcoma cells, like other cancer cells, are differentially sensitive to inhibition of glycolysis and are even more sensitive to combined inhibition of glycolysis and mitochondrial respiration. We also show that bioenergetic parameters are related to sensitivity to inhibition of glycolysis and mitochondrial respiration. These findings suggest that targeting cellular bioenergetics may represent a novel therapeutic strategy for the treatment of sarcomas.

Results

Bioenergetic profiling of human sarcoma cells

Sarcomas, including RMS and osteosarcoma, represent a diverse group of malignancies with distinct molecular and pathological features. To examine and characterize the bioenergetic characteristics of human sarcoma cells, we performed extracellular flux analysis on a panel of human RMS and osteosarcoma cell lines using a Seahorse XF24 Extracellular Flux Analyzer to determine oxygen consumption rates (OCR) and extracellular acidification rates (ECAR) as readouts of mitochondrial respiration and glycolysis, respectively.¹⁷ We first examined mitochondrial bioenergetics by taking baseline OCR measurements, then measuring real-time changes in OCR following sequential addition of the metabolic modulators oligomycin, FCCP, and rotenone, which inhibit mitochondrial ATP synthase, uncouple mitochondrial oxidative phosphorylation and induce maximal respiration, and inhibit complex I of the mitochondrial respiratory chain, respectively. These compounds help generate mitochondrial bioenergetic profiles (Fig. 1A), which are used to determine key parameters of mitochondrial function, as previously described.^13,18-23

Figure 1. Mitochondrial bioenergetic profiles of human sarcoma cell lines. (A) Mitochondrial bioenergetic profiles were generated using the Seahorse XF24 Analyzer. Data represent the mean ± SEM of 3 independent experiments. (B and C) Basal and ATP-linked respiration rates were determined from the bioenergetic profiles and are represented as the mean ± SEM of 3 independent experiments. *P < 0.05.

Figure 1A shows the resulting mitochondrial bioenergetic profiles for aRMS, eRMS, and osteosarcoma cell lines. Basal and ATP-linked respiration rates were determined from the traces in Figure 1A, as previously described,^18-23 and are shown in Figure 1B and C. Notably, basal and ATP-linked respiration rates were significantly lower in aRMS cells than osteosarcoma cells (Fig. 1B and C). Among RMS lines, the eRMS RD line had a significantly higher basal, ATP-linked, and maximal OCR (Fig. 1A–C), suggesting that this cell line may utilize mitochondrial respiration to a greater extent than the other RMS lines. Among osteosarcoma lines, the OHS cell line had the lowest basal and ATP-linked OCR (Fig. 1B and C), suggesting that this cell line may be less dependent on mitochondrial ATP production than the other osteosarcoma lines.

To further characterize the bioenergetic properties of the sarcoma lines, we measured ECAR under specific conditions and generated glycolytic profiles.¹³ Following incubation in glucose-free assay media for 1 h, ECAR was measured. After measurement of the glucose-deprived ECAR, real-time changes in ECAR following the sequential addition of glucose and oligomycin were measured. Figure 2A shows the resulting glycolytic profiles. Following incubation in glucose-free media, the addition of glucose (25 mM) resulted in a robust increase in ECAR in all cell lines (Fig. 2A). Subsequent addition of oligomycin (1 μM) to inhibit mitochondrial ATP production and induce a compensatory increase in glycolysis¹³ further stimulated ECAR in most lines. The difference in ECAR between the glucose-stimulated and oligomycin-treated conditions represents the spare glycolytic capacity, or glycolytic reserve (Fig. 2B). Cells with a low spare glycolytic capacity are likely to be more dependent on glycolysis, whereas cells with a higher glycolytic reserve capacity should be more tolerant of metabolic stress.¹³ Notably, aRMS cells had a significantly lower spare glycolytic capacity than eRMS and osteosarcoma cells (Fig. 2B), suggesting that aRMS cells may be more dependent on glycolysis than eRMS and osteosarcoma cells. Among the osteosarcoma lines, OHS cells had the highest glucose-stimulated ECAR (Fig. 2A) and the lowest spare glycolytic capacity (Fig. 2B), indicating that OHS cells may be more dependent on glycolysis than the other osteosarcoma lines tested.

Figure 2. Glycolytic profile of human sarcoma cell lines. (A) Glycolytic profiles were generated using the Seahorse XF24 Analyzer. Data represent the mean ± SEM of 3 independent experiments. (B) Spare glycolytic capacities were determined from the glycolytic profiles and are represented as the mean ± SEM of 3 independent experiments. *P < 0.05. (C) Correlation between spare glycolytic capacity and ATP-linked respiration of sarcoma cell lines.

We also examined whether the experimentally determined mitochondrial bioenergetic parameters were related to glycolytic properties. Notably, spare glycolytic capacity positively correlated with ATP-linked respiration (Fig. 2C), with a correlation coefficient of 0.78, P < 0.05, indicating that cells that are more dependent on glycolysis have lower ATP-linked respiration rates.

Sensitivity of human sarcoma cells to glycolysis inhibition

To examine the consequences of metabolic stress on sarcoma cell growth, we analyzed the effects of inhibiting glycolysis with 2-DG in the sarcoma lines as well as 2 normal cell types, normal human skeletal muscle cells (SKMC) and dermal fibroblasts (NHDF). We hypothesized that metabolic inhibition would affect cell growth differentially depending upon the bioenergetic characteristics of the cells. To investigate this hypothesis, we treated the sarcoma and normal cell lines with 2-DG for 48 h and evaluated cell viability. As shown in Figure 3A, the RMS lines were more sensitive to 2-DG-induced growth inhibition than the osteosarcoma lines (with the exception of OHS osteosarcoma cells). Among the RMS lines, Rh30 and Rh41 aRMS cells were significantly more sensitive to 2-DG than RD and A-204 eRMS cells. Notably, aRMS cells were significantly more sensitive to 2-DG-induced growth inhibition than normal cells (Fig. 3A). Similar effects were seen over a broad range of 2-DG concentrations (2.5–40 mM, Fig. S1A). The difference in sensitivity to 2-DG between aRMS and eRMS cells is consistent with a previous study.⁵

Figure 3. Sensitivity of human sarcoma cell lines to inhibition of glycolysis. (A) Cells were treated with 10 mM 2-DG for 48 h, and cell viability was determined by AlamarBlue assay. Data represent the mean ± SD of 4 independent experiments. *P < 0.05. (B) Immunoblot analysis of PARP cleavage following treatment with 10 mM 2-DG for the indicated amounts of time. Actin was used as a loading control. (C) Cell cycle analysis of RD eRMS cells following treatment with 10 mM 2-DG for 24 h. (D) Cells were treated with 10 mM 2-DG for 6 h, and intracellular ATP was measured by CellTiter-Glo assay. Data represent the mean ± SD of 2 independent experiments.

To determine whether 2-DG was inhibiting sarcoma cell growth by apoptosis, we examined cleavage of the caspase substrate PARP. Exposure to 2-DG induced robust PARP cleavage in Rh30 aRMS cells, but not RD eRMS cells (Fig. 3B). 2-DG treatment also induced PARP cleavage in OHS, but not HOS (TE85) osteosarcoma cells (Fig. 3B). These findings are consistent with the changes in viability seen in Figure 3A. Because RD cells were growth-inhibited by 2-DG exposure (Fig. 3A), but not apoptotic (Fig. 3B), we examined cell cycle effects of 2-DG on these eRMS cells. As shown in Figure 3C, 2-DG treatment for 24 h led to G₀/G₁ cell cycle arrest in RD cells. Thus, 2-DG inhibits cell growth by inducing apoptosis in aRMS cells and cell cycle arrest in eRMS cells, which is consistent with previous findings.⁵

Several studies have shown that exposure of cancer cells to 2-DG leads to a reduction in intracellular ATP levels.^5,12-14,17 To investigate this effect in the sarcoma lines, we measured intracellular ATP levels following treatment with 2-DG. After 6 h, 2-DG exposure resulted in a differential decrease in intracellular ATP levels (Fig. 3D). Similar results were seen over a broad range of 2-DG concentrations (Fig. S1B). These data are consistent with previous findings that 2-DG treatment decreases intracellular ATP levels, and suggest that sarcoma cells, like other cancer cell lines, are differentially dependent on glycolysis-derived energy for growth.^5,12-14,17

Sensitivity to 2-DG is correlated with bioenergetics

We next examined whether 2-DG-induced growth inhibition of sarcoma cells was related to the bioenergetic parameters in Figures 1 and 2. Notably, basal and ATP-linked respiration rates were positively correlated with viability following 2-DG treatment (Fig. 4A and B), with correlation coefficients of 0.80 and 0.77, respectively, P < 0.05, indicating that cells with a higher basal or ATP-linked respiration were less sensitive to 2-DG-induced growth inhibition. Consistent with this observation, Rh30, Rh41, and OHS cells had the lowest respiration rates (Fig. 1B and C) and were most sensitive to 2-DG-induced growth inhibition (Fig. 3A), while HOS (TE85), KHOS/NP, and U2OS cells had the highest respiration rates (Fig. 1B and C) and were least sensitive to 2-DG-induced growth inhibition (Fig. 3A). Positive correlations between respiration rates and cell viability following 2-DG exposure were also seen at lower 2-DG concentrations (2.5–5 mM, data not shown).

Figure 4. Sensitivity to 2-DG is correlated with cellular bioenergetics. (A–C) Correlation between basal respiration, ATP-linked respiration or spare glycolytic capacity and viability of the indicated sarcoma cell lines.

Additionally, spare glycolytic capacity was positively correlated with viability following 2-DG exposure (Fig. 4C), with a correlation coefficient of 0.76, P < 0.05, indicating that cells with a higher spare glycolytic capacity (glycolytic reserve) were less sensitive to 2-DG-induced growth inhibition. Consistent with this observation, Rh30, Rh41, and OHS cells had the lowest spare glycolytic capacities (Fig. 2B) and were the most sensitive to 2-DG treatment (Fig. 3A), while HOS (TE85), KHOS/NP, and U2OS cells had higher spare glycolytic capacities (Fig. 2B) and were least sensitive to 2-DG-induced growth inhibition (Fig. 3A). Positive correlations between spare glycolytic capacity and cell viability following 2-DG treatment were also seen at lower 2-DG concentrations (2.5–5 mM, data not shown). Overall, these findings indicate that sensitivity to 2-DG is related to cellular bioenergetics, and that cells that are more dependent on glycolysis (lower glycolytic reserve capacity), such as aRMS cells, are more sensitive to 2-DG, while cells that have higher respiratory rates, such as osteosarcomas, are less sensitive to 2-DG-induced growth inhibition.

Sarcoma cells are sensitive to combined inhibition of glycolysis and mitochondrial respiration

Several recent studies have shown that simultaneous inhibition of glycolysis and mitochondrial respiration acts synergistically to reduce tumor cell survival in vitro and in vivo.^12-14 To examine whether this strategy would be effective in the sarcoma lines, we treated the sarcoma and normal cell lines with 2-DG and oligomycin, either alone or in combination, and evaluated cell viability after 48 h of exposure. As shown in Figure 5, the combination of 2-DG and oligomycin decreased cell viability more than either compound alone in all cell lines tested, and the decrease in growth of sarcoma lines was greater than that in the normal cells. Importantly, while oligomycin had minimal effects on cell growth as a single agent (Fig. 5; Fig. S1C), it increased 2-DG-induced growth inhibition of osteosarcoma cells, which were least sensitive to 2-DG alone (Fig. 3A). Similar results were seen with varying concentrations of 2-DG and oligomycin in combination (Fig. S1D). These findings are consistent with previous studies on other cancer types showing increased cell death by combined inhibition of glycolysis and mitochondrial respiration.^12-14

Figure 5. Combined inhibition of glycolysis and mitochondrial respiration inhibits the growth of sarcoma cells. Cells were treated for 48 h with the indicated concentrations of 2-DG and oligomycin, either alone or in combination, and cell viability was determined by AlamarBlue assay. Data represent the mean ± SD of 3 independent experiments.

Metformin sensitizes sarcomas to 2-DG

Recent studies have shown that metformin, a widely prescribed drug for treatment of type-2 diabetes that can activate AMPK and inhibit mitochondrial respiration, can synergize with 2-DG to induce cancer cell death and inhibit xenograft tumor growth.^9,12,14 Since oligomycin sensitized the sarcoma lines to 2-DG, we examined the growth-inhibitory effects of the more clinically used metformin as a single agent and in combination with 2-DG in the sarcoma and normal cells.

As shown in Figure 6A, 48-h treatment with a biologically active^12,14 metformin concentration (5 mM) alone had minimal effects on sarcoma cell growth, with no effect on the normal cell types. Similar results were obtained with a higher metformin concentration (10 mM, Fig. S2). Interestingly, the combination of 2-DG and metformin at previously tested concentrations^12,14 was more effective at inhibiting growth than either compound alone in all cell lines tested (Fig. 6A). Sarcoma cells were 2- to 5-fold more sensitive to the combination than normal cells (Fig. 6A). These findings are consistent with recent studies using the same combination in a variety of cancer cell types.^12,14

Figure 6. Metformin sensitizes sarcoma cells to inhibition of glycolysis. (A) Cells were treated for 48 h with the indicated concentrations of 2-DG and metformin, either alone or in combination, and cell viability was determined by AlamarBlue assay. Data represent the mean ± SD of 3 independent experiments. (B) Cells were treated for 48 h with the indicated concentrations of 2-DG, alone or in combination with 5 mM metformin, and cell viability was determined. Data represent the mean ± SD of 2 independent experiments. (C) Correlation between cell viability following treatment with 2-DG+oligomycin and 2-DG+metformin.

To further examine the effect of metformin on sensitizing sarcomas to 2-DG-induced growth inhibition, we selected 4 sarcoma lines with varying sensitivity to 2-DG and exposed them to a range of 2-DG concentrations, plus or minus metformin (5 mM) for 48 h, and measured cell growth. As shown in Figure 6B, metformin addition sensitized all 4 sarcoma cell lines to 2-DG-induced growth inhibition by varying degrees, consistent with the data in Figure 6A. Notably, the 2 lines that were less sensitive to 2-DG alone, RD and HOS (TE85), exhibited the greatest increase in sensitivity to 2-DG when metformin was added in combination, amounting to a 6-fold and >10-fold decrease in IC₅₀, respectively (Fig. 6B).

Based on these results, we compared sarcoma sensitivity to the combination of 2-DG and oligomycin, with the combination of 2-DG and metformin. As shown in Figure 6C, there was a strong positive correlation between cell viability following treatment with either 2-DG and oligomycin or 2-DG and metformin, with a correlation coefficient of 0.96, P < 0.05, suggesting that dual inhibition of metabolic pathways by either drug combination acts similarly to reduce sarcoma cell growth.

Effects of metabolic inhibitors on cellular bioenergetics

We showed earlier that sensitivity to 2-DG is correlated with respiration rates and spare glycolytic capacity, and that exposure to oligomycin or metformin sensitizes sarcomas to 2-DG-induced growth inhibition. Based on these results, we hypothesized that oligomycin and metformin may be affecting respiration rates and spare glycolytic capacity. To test the effects of oligomycin and metformin on cellular bioenergetics, we treated cells with either oligomycin, metformin, or a vehicle control for 6 h, returned cells to normal culture media for 1 h, and then generated glycolytic or mitochondrial bioenergetic profiles. These profiles were used to determine relative changes in basal and ATP-linked respiration, as well as spare glycolytic capacity, in compound- vs. vehicle control-treated cells. We selected HOS (TE85) cells for these experiments because of their relatively high spare glycolytic capacity and respiration rates, as well as relative insensitivity to 2-DG exposure. Treatment with either oligomycin (0.3 μM) or metformin (5 mM) caused a >85% decrease in basal (Fig. S3A) and ATP-linked (Fig. S3B) respiration rates and an even larger decrease in spare glycolytic capacity (Fig. S3C), supporting the link between these bioenergetic parameters and sensitivity to growth inhibition.

Discussion

Sarcomas, including rhabdomyosarcoma and osteosarcoma, represent a diverse group of malignancies with distinct molecular and pathological features. A renewed interest in the metabolic properties of cancer cells, as well as the recognition of altered cellular metabolism as a critical hallmark of cancer, has stimulated an exploration of targeting metabolic vulnerabilities of cancer cells as a therapeutic strategy.^9,11 In this study, we have examined bioenergetic properties of human RMS and osteosarcoma cell lines and characterized relationships between bioenergetic parameters and sensitivity to metabolic inhibitors.

We show that aRMS cells have significantly lower respiration rates than osteosarcomas (with the exception of OHS osteosarcoma cells). Furthermore, aRMS cells have a significantly lower spare glycolytic capacity, or glycolytic reserve, than eRMS and osteosarcoma cells (with the exception of OHS cells). We also show that ATP-linked respiration is positively correlated with spare glycolytic capacity in these cell lines. These data suggest that aRMS cells (and OHS cells) may be more dependent on glycolysis than mitochondrial respiration for ATP production. The increased dependency of aRMS cells on glycolysis is consistent with our 2-DG-induced growth inhibition results and a recent study showing that aRMS cells are more sensitive than eRMS cells to 2-DG-induced apoptosis.⁵ Furthermore, we show that 2-DG-induced growth inhibition is inversely related to respiration rates and glycolytic reserve capacity, suggesting that differences in cellular bioenergetics may explain, at least in part, the differences in 2-DG sensitivity between aRMS and eRMS cells, as well as the relative insensitivity of osteosarcomas to 2-DG.

Recent studies have begun to explore the efficacy of simultaneously targeting glycolysis and mitochondrial respiration in prostate, breast, gastric, and esophageal cancer cells, using drug combinations, and showed that inhibition of glycolysis and mitochondrial respiration can act synergistically to reduce tumor cell survival in vitro and in vivo.^12-14 Given our findings that osteosarcoma and eRMS cells were relatively insensitive to 2-DG, we tested combinations of glycolysis and mitochondrial respiration inhibitors to sensitize these cells. We show that oligomycin, a potent inhibitor of mitochondrial ATP synthase, had minimal effects on sarcoma cell growth, but the combination of 2-DG and oligomycin decreased cell growth more effectively than either drug alone in all cell lines tested, consistent with previous studies showing increased effectiveness of combined inhibition of glycolysis and mitochondrial respiration.^12-14

Recent studies have shown that metformin, a widely prescribed and well-tolerated type-2 diabetes drug that can activate AMPK and inhibit mitochondrial respiration, can synergize with 2-DG to induce cancer cell death and inhibit xenograft tumor growth.^9,12,14 While not effective as a single agent, we show that metformin was able to enhance 2-DG cytotoxicity in all sarcoma lines tested. Importantly, combinations of 2-DG and either oligomycin or metformin were more cytotoxic toward all sarcoma lines tested than toward the normal cell types tested, suggesting that combined inhibition of glycolysis and mitochondrial respiration should be further evaluated as a potential novel therapeutic approach for the treatment of sarcomas.

We show that sensitivity to 2-DG is correlated with respiration rates and spare glycolytic capacity, and that exposure to oligomycin or metformin sensitizes sarcoma cells to 2-DG-induced growth inhibition. Based on these results, we hypothesized that oligomycin and metformin may be affecting respiration rates and glycolytic reserve capacity. Indeed, treatment with either oligomycin or metformin reduced basal and ATP-linked respiration rates and glycolytic reserve capacity, consistent with increased sensitivity to 2-DG-induced growth inhibition. The effects of oligomycin and metformin on basal and ATP-linked respiration rates are likely due to inhibition of mitochondrial ATP synthase and complex I of the mitochondrial respiratory chain, respectively, resulting in reduced oxygen consumption and a compensatory increased dependency on glycolysis, thus reducing the glycolytic reserve capacity. A compensatory increase in glycolysis could explain why oligomycin and metformin alone, despite decreasing OCR, did not significantly affect sarcoma cell growth. We also showed that treatment with 2-DG led to a decrease in intracellular ATP levels, suggesting that sarcoma cells, like other cancer cells, are dependent on glycolysis-derived energy production for growth.^5,12-14,17 Taken together, our data suggest that 2-DG-induced growth inhibition is likely mediated by changes in cellular bioenergetics, with one consequence being reduced intracellular ATP levels, although other effects, such as ER stress due to altered protein glycosylation, have been reported.⁵

In summary, we have characterized the cellular bioenergetics of a panel of human RMS and osteosarcoma cell lines, and have identified relationships between bioenergetic properties and sensitivity to metabolic inhibitors. We have shown that sarcoma cells are differentially sensitive to inhibition of glycolysis, and are even more sensitive to combined inhibition of glycolysis and mitochondrial respiration by different drug combinations. We have also shown that bioenergetic parameters are related to drug sensitivity. These findings suggest that targeting cellular bioenergetics may represent a novel therapeutic strategy for the treatment of sarcomas.

Materials and Methods

Cell culture

Human sarcoma cell lines were purchased from ATCC, with the exception of OHS, which was obtained from the NCI repository (Frederick National Laboratory for Cancer Research), and Rh30 and Rh41, which were a gift of Dr Peter J Houghton (Nationwide Children’s Hospital). Human aRMS cell lines Rh30 and Rh41, and human eRMS cell lines RD and A-204, as well as human osteosarcoma cell lines HOS(TE85) and OHS were cultured in RPMI-1640 media supplemented with 1 mM L-glutamine and 10% fetal bovine serum (FBS). Human osteosarcoma cell lines KHOS/NP and U2OS were grown in EMEM and McCoy 5A Medium (Modified), respectively, supplemented with 10% FBS. Normal human dermal fibroblasts (NHDF) and skeletal muscle cells (SKMC) were purchased from Lonza and maintained in FGM-2 and SKGM BulletKit media (Lonza), respectively. All cell lines were maintained in a humidified incubator containing 5% CO₂ at 37 °C.

Preparation of compounds

Stock solutions of 5 mg/mL oligomycin, 10 mM carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (FCCP), and 1 mM rotenone were prepared in fresh molecular biology-grade DMSO. A 2 mM stock solution of ATP was prepared in water. All stock solutions were aliquoted and stored at −20 °C and diluted in appropriate culture media prior to use. Metformin and 2-DG solutions were prepared fresh in culture media just prior to use. All compounds were purchased from Sigma-Aldrich.

Measurement of oxygen consumption rate and extracellular acidification rate

Oxygen consumption rates (OCR) and extracellular acidification rates (ECAR) of live cells were measured in real time using a Seahorse Bioscience XF24 Extracellular Flux Analyzer as previously described.¹⁷ Briefly, cells were plated at 20 000 to 50 000 cells/well (~80% confluent) in XF24 V7 24-well cell culture plates (Seahorse Bioscience) and incubated overnight at 37 °C. Just prior to an experiment, growth media was removed from each well, cells were rinsed with freshly prepared Seahorse assay media (Seahorse Bioscience), and a final volume of 675 μL of assay media was added to each well. The plate was incubated in a 37 °C incubator lacking CO₂ for 45 to 60 min prior to the assay.

Mitochondrial bioenergetic profiles were generated by taking baseline measurements of OCR then sequentially injecting oligomycin to inhibit ATP synthase (1 μg/mL final concentration), followed by FCCP to uncouple mitochondrial oxidative phosphorylation and induce maximal respiration (0.3 μM for Rh30, Rh41, RD, A-204, and U2OS cells; 0.1 μM for HOS (TE85) and OHS cells; 0.5 μM for KHOS/NP cells), followed by rotenone to inhibit mitochondrial complex I (0.1 μM final concentration). Following the addition of each compound, 3 measurements of OCR were taken. ATP-linked respiration rates were determined from the resulting mitochondrial bioenergetic profiles and represent the difference between basal and oligomycin-treated OCR, as previously described.^18-23

To generate glycolytic profiles, cells were plated as described above and rinsed with glucose-free Seahorse assay media. Following incubation in glucose-free assay media for 60 min, glycolytic profiles were generated by taking baseline measurements of ECAR then sequentially injecting Seahorse assay media containing 25 mM glucose, followed by oligomycin (1 μM final concentration). Three measurements of ECAR were taken following each injection. Spare glycolytic capacities were determined from the resulting glycolytic profiles and represent the difference between glucose-stimulated and oligomycin-treated ECAR, as previously described.¹³

For assays involving pre-treatment with oligomycin or metformin, cells were treated with 0.3 μM oligomycin or 5 mM metformin for 6 h prior to measurement of OCR and ECAR. For all experiments, compounds were prepared in Seahorse assay media.

At the conclusion of each Seahorse assay, media was removed from all the XF24 plate wells. Twenty μL of RIPA lysis buffer was added to each well, and the plate was stored at −80 °C. Following a freeze/thaw cycle, the protein content of each well was determined by Bradford assay. The measured OCR and ECAR for each experiment were normalized to protein content by dividing the OCR or ECAR by the protein content for each well.

Compound exposure and measurement of cell viability

For compound exposure experiments, cells were plated at 4500 to 7500 cells/well in 96-well plates and incubated at 37 °C for 16 to 24 h prior to exposure. All compounds were prepared in culture media just prior to addition. Cells were exposed to a vehicle control or the indicated compounds at the indicated concentrations for 48 h at 37 °C. After 48 h of compound exposure, AlamarBlue was added to 10% of the media volume and incubated at 37 °C for 3 to 6 h, during which time absorbance readings were taken every hour according to the manufacturer’s instructions, on a Tecan Infinite plate reader.

Immunoblotting

Following treatment, cells and media were scraped from 6-well tissue culture plates, transferred to 15 mL tubes, and centrifuged. Cell pellets were lysed in RIPA buffer supplemented with a protease/phosphatase inhibitor cocktail (Cell Signaling). Clarified total cellular lysates were immunoblotted with anti-PARP (Cell Signaling) and anti-actin (Millipore) antibodies using standard procedures.

Measurement of intracellular ATP concentrations

Intracellular ATP levels were determined using the CellTiter-Glo Luminescent Cell Viability Assay (Promega), following the manufacturer's instructions. Briefly, cells were plated at 12 500 to 15 000 cells/well in white-walled 96-well plates and incubated at 37 °C for 16 to 24 h prior to compound exposure. Cells were exposed to a vehicle control or the indicated compounds for 6 h at 37 °C before starting the ATP assay. As per the manufacturer's instructions, plates were equilibrated at room temperature for 30 min, during which time serial dilutions of ATP were made to generate an ATP standard curve. CellTiter-Glo reagent was added to each well, and plates were mixed for 2 min on an orbital shaker to induce cell lysis. Plates were incubated at room temperature for 10 min, and luminescence intensity was measured using a Tecan Infinite plate reader.

Cell cycle analysis

Following the indicated compound exposure, cell cycle analysis was performed on a Guava easyCyte flow cytometer (Millipore) using Guava Cell Cycle Reagent (Millipore) according to the manufacturer’s instructions.

Statistical analysis

Correlations between bioenergetic parameters and drug response were determined using Pearson correlation coefficients. Statistical significance was determined by Student t test. A P value of < 0.05 was considered significant.

Glossary

Abbreviations:

aRMS

alveolar rhabdomyosarcoma

eRMS

embryonal rhabdomyosarcoma

2-DG

2-deoxy-D-glucose

OCR

oxygen consumption rate

ECAR

extracellular acidification rate

SKMC

normal human skeletal muscle cells

NHDF

normal human dermal fibroblasts

ATP

adenosine triphosphate

FCCP

carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone