Biochemical Journal

Research article

Pyruvate fuels mitochondrial respiration and proliferation of breast cancer cells: effect of monocarboxylate transporter inhibition

Anne R. Diers , Katarzyna A. Broniowska , Ching-Fang Chang , Neil Hogg

Abstract

Recent studies have highlighted the fact that cancer cells have an altered metabolic phenotype, and this metabolic reprogramming is required to drive the biosynthesis pathways necessary for rapid replication and proliferation. Specifically, the importance of citric acid cycle-generated intermediates in the regulation of cancer cell proliferation has been recently appreciated. One function of MCTs (monocarboxylate transporters) is to transport the citric acid cycle substrate pyruvate across the plasma membrane and into mitochondria, and inhibition of MCTs has been proposed as a therapeutic strategy to target metabolic pathways in cancer. In the present paper, we examined the effect of different metabolic substrates (glucose and pyruvate) on mitochondrial function and proliferation in breast cancer cells. We demonstrated that cancer cells proliferate more rapidly in the presence of exogenous pyruvate when compared with lactate. Pyruvate supplementation fuelled mitochondrial oxygen consumption and the reserve respiratory capacity, and this increase in mitochondrial function correlated with proliferative potential. In addition, inhibition of cellular pyruvate uptake using the MCT inhibitor α-cyano-4-hydroxycinnamic acid impaired mitochondrial respiration and decreased cell growth. These data demonstrate the importance of mitochondrial metabolism in proliferative responses and highlight a novel mechanism of action for MCT inhibitors through suppression of pyruvate-fuelled mitochondrial respiration.

  • α-cyano-4-hydroxycinnamic acid
  • extracellular flux technology
  • mitochondrion
  • reserve capacity
  • Warburg effect

INTRODUCTION

Otto Warburg [1] first described the increased utilization of anaerobic metabolism in the presence of adequate oxygen by cancer cells compared with their normal counterparts: termed the ‘Warburg effect’. Since these initial observations, it is now clear that classic oncogene activity not only regulates proliferation, but also leads to alterations in metabolic pathways (e.g. glutaminolysis, glycolysis and mitochondrial function) which may play a causative role in tumour development [2]. Recent studies have highlighted the fact that metabolic reprogramming of cancer cells is required to drive biosynthesis pathways which enables rapid replication and proliferation (reviewed in [2]). As such, the targeting of metabolic pathways is emerging as a novel strategy in the treatment of many malignancies.

An emerging concept in the field of cancer metabolism is the importance of mitochondrial metabolism, particularly citric acid cycle activity, in providing intermediates required for the biosynthesis of cellular macromolecules (e.g. fatty acids and non-essential amino acids). It is now clear that metabolism of mitochondrial substrates such as glutamine and pyruvate is necessary to support the rapid proliferation of multiple cancer cell types (e.g. colon cancer and glioblastoma), and a functional link between mitochondrial respiration and proliferative capacity has been established [3,4].

In the present study, we have examined the role of MCTs (monocarboxylate transporters) because of their critical role in transportation of multiple monocarboxylate molecules, in particular pyruvate, across cell membranes [5,6]. In normal physiology, MCTs play a crucial role in ‘lactate shuttles’ where they function to transport lactate between cells (e.g. whiteglycolytic and red-oxidative fibres in working muscle) or between intracellular compartments (e.g. lactate uptake into mitochondria) [7]. At least 14 members of this transporter family have been identified and they have unique tissue expression patterns and kinetic properties; however, in the context of cancer, the best-characterized have been the expression of MCT1 and MCT4. Expression of both MCT1 and MCT4 has been shown to be elevated in several tumour types when compared with matched normal tissue (e.g. the breast, prostate, ovarian, cervix and gastrointestinal tract), and high levels of these proteins often correlate with poor prognosis and disease progression [812]. MCT expression is also thought to underpin aspects of the Warburg effect. Since highly glycolytic cancer cells produce increased levels of lactate, MCT-dependent lactate efflux from cells is required to maintain intracellular pH and avoid cytotoxic accumulation of lactate. A central role for MCTs in cancer is further supported by clinical evidence which demonstrates that tumour-produced lactate correlates with poor prognosis and resistance to radiotherapy [1315]. For these reasons, inhibition of MCTs has been proposed as a therapeutic strategy targeting the metabolic pathways in cancer. Preclinical studies have demonstrated that the compound CHC (α-cyano-4-hydroxycinnamic acid), an inhibitor of plasma membrane MCTs and the mitochondrial pyruvate carrier, decreases tumour size and sensitizes hypoxic tumour regions to radiotherapy [16].

The effects of MCT inhibitors on cancer cell growth have largely been attributed to their ability to block lactate efflux; however, since MCTs also transport other monocarboxylates, such as pyruvate, MCT inhibition will probably result in changes in the influx and/or efflux of other metabolically important molecules. Increasing evidence shows that energy substrates metabolized through mitochondria (e.g. glutamine and pyruvate) are required for biosynthesis of macromolecules in rapidly dividing cells [17]; thus, in the present study, we examined the effect of metabolic substrates (glucose and pyruvate) on breast cancer cell proliferation and mitochondrial function. We demonstrated that cancer cells proliferate more rapidly when provided with pyruvate when compared with glucose, but this effect was not observed with lactate. In addition, inhibition of cellular pyruvate uptake using the MCT inhibitor CHC decreases cell growth. Pyruvate supplementation fuelled mitochondrial oxygen consumption and altered mitochondrial function correlated with proliferative potential. These data demonstrate that metabolism through mitochondria drives cancer cell proliferation and highlight a novel mechanism of action for MCT inhibitors in cancer.

MATERIALS AND METHODS

Materials

All chemicals were of analytical grade and purchased from Sigma–Aldrich unless otherwise noted. DMSO (0.1%) was used as a vehicle control for CHC in all experiments.

Cell culture

MDA-MB231 (MB231) and T-47D human mammary adenocarcinoma cells were obtained from the A.T.C.C., and MCF7 human mammary adenocarcinoma cells were a gift from Dr Balaraman Kalyanaraman (Medical College of Wisconsin, Milwaukee, WI, U.S.A.). All of the cells were maintained in low-glucose (5.56 mM) DMEM (Dulbecco's modified Eagle's medium) containing sodium pyruvate (1 mM) and glutamine (4 mM) supplemented with 10% FBS (fetal bovine serum; Invitrogen), 200 units/ml penicillin and 200 μg/ml streptomycin. Cultures were incubated at 37°C in a humidified atmosphere of 5% CO2 and 95% air.

In some experiments, medium containing different metabolic substrates was used. In these experiments the base medium was DMEM powder (CellGro), and metabolic substrates were supplemented to the final concentrations noted. Sodium bicarbonate (44 mM) was added to media for all assays except extracellular flux analysis. FBS was added to media for all assays except extracellular flux analysis and glucose/pyruvate uptake assays.

MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide] assay

Cells were seeded at a density of 5000 cells/well in 48-well cluster plates and allowed to attach and grow for 24 h. Cells were then washed with sterile PBS, and DMEM supplemented with 10% FBS and specified metabolic substrates (glucose, sodium pyruvate and lactate) were added. Media were replaced daily to ensure metabolic substrates did not become limiting. Cell proliferation was monitored every 24 h for growth kinetic assays or after 72 h treatment by MTT assay as described previously [18]. Treatment media were replaced with complete culture medium containing 0.4 mg/ml Thiazoyl Blue Tetrazolium, and the cells were incubated at 37°C for 2 h. The medium was then removed, the resulting formazan crystals were solubilized in DMSO and the absorbance was read at 590 nm with background reference at 620 nm. MTT assay absorbance values correlated with total number of cells per well (results not shown).

Trypan Blue exclusion assay

After the indicated treatment, cells were trypsin-harvested and diluted 1:1 with 0.4% Trypan Blue. The Countess Automated cell counter (Invitrogen) was then used to score the percentage of cells which excluded Trypan Blue.

Colony formation assay

For acute treatment studies, colony formation was measured after treatment in the indicated metabolic-substrate-containing media. Cells from experimental dishes were trypsinized and collected. All cells from each dish were then centrifuged (600 g for 5 min at room temperature), resuspended in fresh medium and counted using a Countess Automated cell counter. Cells were then plated in six-well cluster plates at low density (100–200 cells per well), and clones were allowed to grow for 7 days in complete culture medium in the presence of 0.1% gentamicin. Cells were subsequently fixed with 70% ethanol and stained with Coomassie Brilliant Blue for analysis of colony formation as described previously [19]. For chronic treatment studies, routinely cultured cells were collected and counted as described above and then 200–400 cells per well were plated into six-well cluster plates containing DMEM supplemented with 10% FBS, 0.1% gentamicin and the indicated metabolic substrates (5.56 mM glucose, 1 mM sodium pyruvate or both). Clones were allowed to grow for 14 days prior to fixation, staining and scoring of colony formation.

Extracellular flux technology

To measure mitochondrial function in intact breast cancer cells, a Seahorse Bioscience XF24 Extracellular Flux Analyzer was used [2023]. This instrument allows for the sensitive measurement of OCRs (oxygen consumption rates) from adherent intact cultured cells. The mitochondrial function assay employed in the present study used sequential injection of oligomycin (1 μg/ml) and FCCP (carbonyl cyanide p-trifluoromethoxyphenylhydrazone; 3 μM) to define the mitochondrial function parameters basal OCR, maximal OCR and reserve respiratory capacity as described previously [24]. All assays were conducted using a seeding density of 20000 cells/well in unbuffered DMEM supplemented with specified metabolic substrates (5.56 mM glucose, 1 mM lactate and/or 0.1–5 mM sodium pyruvate), and the AKOS algorithm was used [21]. The cells were switched to assay media 1 h prior to the beginning of the assay and maintained at 37°C. Values were normalized to the total protein per well after the completion of the XF assay by the Bradford protein assay (Bio-Rad Laboratories).

Glucose and pyruvate consumption assays

The consumption of glucose and pyruvate by cells was monitored by assaying media levels of these metabolic substrates before and after 4 h incubation with cells using commercially available kits (BioVision). Substrate consumption was then normalized to total cell number. For analysis of glucose consumption and pyruvate consumption, cells were incubated in 0.75 ml of DMEM without FBS supplemented with 5.56 mM glucose or 1 mM sodium pyruvate respectively. Medium lactate levels were measured using a commercially available kit (BioVision).

HPLC analysis of nucleotides

Adenine nucleotides (ATP and ADP) and NAD+ were extracted using perchloric acid precipitation as described previously [25]. Solvent A [75 μl; 0.1 mol/l potassium phosphate, 4 mmol/l tetrabutylammonium bisulfate (pH 6.0) diluted v/v in water 64:36] was added to supernatants, and these were filtered prior to HPLC analysis. Protein pellets were resuspended in 0.5 N NaOH (200 μl) and the protein concentration was determined using the Bradford assay. HPLC analysis of nucleotides based on a previously described method [26] was performed on Kinetex C-18 column (2.6 μm, 100 mm×4.6 mm internal diameter) using solvent A and solvent B [0.1 mol/l potassium phosphate, 4 mmol/l tetrabutylammonium bisulfate (pH 6.0) diluted v/v in methanol 64:36] with a flow rate of 1 ml/min. For NADH measurements, samples were analysed using the same chromatographic conditions, but the cells were harvested under alkaline conditions (0.5 M KOH/Hanks balanced salt solution, 3:1). The pH of lysates was adjusted to ~8 using 6 M HCl and ammonium acetate (1 M, pH 4.7), and samples were filtered prior to HPLC analysis. ATP, ADP, NAD+ and NADH peaks levels were measured for each sample, compared with the standard and expressed in nanomoles per milligram of protein.

Statistical analysis

Results are reported as means±S.E.M. for n≥3 as indicated in the Figure legends. Statistical significance was evaluated by Student's t test. The minimum level of significance was set at P<0.05.

RESULTS

Effect of metabolic substrates on breast cancer cell proliferation

To determine the effect of metabolic substrates on breast cancer cell proliferation, MCF7, MB231 and T-47D breast cancer cells were supplied with either glucose-only (5.56 mM), pyruvate-only (1 mM), or both glucose and pyruvate (referred to as ‘complete media’) in DMEM, and cell proliferation was determined after 72 h (Figure 1A). In all three of the cell lines, more rapid proliferation was observed for cells provided with pyruvate-only when compared with glucose-only. In addition, the growth of the cells on pyruvate-only was comparable with that of the cells provided complete media. We also examined the effect of CHC, an inhibitor of plasma membrane MCTs and the mitochondrial pyruvate carrier, on cell growth. Treatment with CHC (500 μM, a concentration 3-fold higher than the IC50 value [5]) in complete media resulted in a significant decrease in the total number of cells after 72 h incubation for MCF7 and MB231 cells, but not for T-47D cells. The growth kinetics of MCF7 cells cultured under similar conditions are shown in Figure 1(B). Over 72 h, cells proliferated most rapidly when supplied either pyruvate-only or complete media when compared with the glucose-only condition. CHC also significantly decreased proliferation of MCF7 cells at the 48 and 72 h time points. In addition, cell proliferation in the presence of pyruvate was concentration-dependent (Figure 1C). Light micrographs of MCF7 cells cultured in different metabolic substrates for 72 h are shown in Figure 1(D), and marked morphological changes were not observed between any of the conditions.

Figure 1 Effect of metabolic substrates on breast cancer cell proliferation

(A) MCF7, MB231 and T-47D cells were cultured in medium containing glucose-only (5.56 mM), pyruvate-only (1 mM), glucose and pyruvate (complete media), or complete media containing CHC (500 μM) for 72 h. Total cell number was then assessed using the MTT assay. (B) MCF7 cells were exposed to the same medium conditions for 24–72 h, and total cell number was assessed using the MTT assay. (C) Medium containing glucose (5.56 mM) and increasing concentrations of pyruvate (0.1–10 mM) was used to culture MCF7 cells for 72 h prior to the assessment of total cell number. (AC) Results are means±S.E.M., n=6. *P<0.05 compared with complete media. #P<0.05 compared with 0 mM sodium pyruvate. (D) Light micrographs of cells cultured in medium containing metabolic substrates specified for 72 h.

Since supplementation of pyruvate clearly promoted proliferation, we examined whether another physiological monocarboxylate, lactate, could also promote cell growth. MCF7 cells were cultured for 72 h in medium containing glucose (5.56 mM) and increasing concentrations (0.1–10 mM) of lactate; however, lactate supplementation had no effect on proliferative responses (Supplementary Figure S1A at http://www.BiochemJ.org/bj/444/bj4440561add.htm). To further understand the difference between lactate and pyruvate supplementation, we examined whether there were limitations in the ability to convert lactate into pyruvate in the cell through LDH (lactate dehydrogenase). We measured both the activity of LDH and levels of NAD+ and NADH in MCF7 cells provided different metabolic substrates for 72 h. LDH catalyses the conversion of lactate into pyruvate using NAD+ as a cofactor. There was no difference in LDH activity or levels of NAD+ and NADH in cells cultured in glucose and lactate or glucose and pyruvate medium (Supplementary Figures S1B and S1C). Moreover, NAD+ levels were always significantly higher than NADH, a condition which should promote the conversion of lactate into pyruvate.

Effect of metabolic substrates on MCF7 cell viability and colony formation

The results described above suggest that pyruvate may be a more preferable metabolic substrate than glucose for promoting cell proliferation, and the largest difference in glucose- compared with pyruvate-fuelled proliferation was observed in MCF7 cells; however, it is also possible that alterations in metabolic substrates may be cytotoxic. To examine this, MCF7 cells were incubated in glucose-only, pyruvate-only or complete media for 48 h, and then cell viability was assessed by Trypan Blue exclusion assay. There was no significant difference in the number of cells which excluded Trypan Blue (scored as viable) in any condition (Figure 2A). Exposure of cells to CHC (500 μM) in complete media (48 h) also had no effect on cell viability.

Figure 2 Effect of metabolic substrates on MCF7 cell viability and colony formation

MCF7 cells were cultured in medium containing glucose-only (5.56 mM), pyruvate-only (1 mM), glucose and pyruvate (complete media), or complete media containing CHC (500 μM) for 48 h. The number of cells which exclude Trypan Blue (A) and colony formation (B) were quantified. Results are means±S.E.M., n=6. *P<0.05 compared with complete media. N.S. denotes no significant difference between groups. (C) MCF7 cells were also seeded at low density in medium containing glucose-only (5.56 mM), pyruvate-only (1 mM) or glucose and pyruvate (complete media), and the number of colonies formed was measured. Representative images of the cloning wells and quantification are shown.

To confirm these observations, we exposed MCF7 cells to the different metabolic substrate-containing media, and their ability to form colonies was assessed. Following incubation in glucose-only, pyruvate-only, complete media, or complete media and CHC (500 μM) for 48 h, MCF7 cells were seeded at low density (100–200 cells/well) and allowed to form colonies for 7 days in DMEM (described in Materials and methods section). Pre-exposure (48 h) to different metabolic substrates or CHC had no effect on colony formation (Figure 2B). This result further supports the conclusion that alterations in metabolic substrate supply do not cause overt cell death.

To examine the substrate-dependence of the proliferative capacity of cancer cells, MCF7 cells were seeded at low density in glucose-only, pyruvate-only or complete media, and colony formation was quantified after 14 days. Figure 2(C) shows both representative images from cloning plates as well as quantification of these data. MCF7 cells were unable to form colonies in glucose-only medium; however, colony formation was observed in both the pyruvate-only medium and complete media conditions. Surprisingly, significantly more colonies were formed in pyruvate-only medium when compared with cloning in complete media. Taken together, these data indicate that metabolic substrates regulate cancer cell proliferation and do not cause overt cell death. Moreover, pyruvate appears to be a key metabolic substrate to support cell proliferation.

MCT expression and effects of CHC on metabolic substrate transport into cells

In the present study, CHC is used as an inhibitor of plasma membrane MCTs and the mitochondrial pyruvate carrier which function, in part, to transport lactate and/or pyruvate across cellular membranes [5]. To confirm that CHC does in fact inhibit pyruvate transport into cells without interfering with glucose transport into cells, we treated MCF7 cells with increasing concentrations of CHC (50–500 μM) for 4 h, and assayed medium glucose and pyruvate concentrations before and after this incubation. As expected, there was no effect of CHC on glucose consumption by MCF7 cells after 4 h (Figure 3A); however, treatment with CHC caused a concentration-dependent inhibition of pyruvate consumption by these cells (Figure 3B). In the concentration range examined in the present study, CHC was not capable of blocking lactate excretion from these cells (Figure 3C). In fact, treatment with 500 μM CHC for 4 h resulted in a small, but significant, increase in lactate levels in the medium. These results confirm that CHC is an effective inhibitor of cellular pyruvate consumption presumably through the inhibition of plasma membrane MCTs and/or the mitochondrial pyruvate carrier; however, lactate export from cells was not blocked.

Figure 3 Effects of CHC on metabolic substrate transport into cells

Cellular consumption of glucose (A) and pyruvate (B) from the medium after exposure to increasing concentrations of CHC (50–500 μM) was determined in MCF7 cells. Lactate production from cells was assessed under the same experimental conditions (C). Glucose, pyruvate and lactate levels in the medium were determined before and after 4 h incubation with cells. Results were normalized to total cell number/well and are means±S.E.M., n=3–5. *P<0.05 compared with the control. N.S. denotes no significant difference between groups.

Regulation of mitochondrial function by metabolic substrate supply

Although a large literature supports the integral role of glucose metabolism in cancer (reviewed in [27]), the results of the present study indicate that pyruvate may also be important for rapid proliferation of cancer cells. Since pyruvate is a key substrate for mitochondrial function, we next examined the effects of pyruvate and glucose supplementation on mitochondrial function in MCF7 cells. We used novel extracellular flux technology [20] to assess multiple parameters of mitochondrial function: (i) basal OCR, (ii) maximal OCR, (iii) ATP-linked respiration, (iv) proton leak, (v) reserve capacity and (vi) oxygen consumption independent of Complex IV (termed ‘non-mitochondrial’) in MCF7 cells. These parameters can be assessed using a protocol of OCR measurement after the sequential injection of oligomycin (ATP synthase inhibitor), FCCP (uncoupling agent) and Antimycin A (Complex III inhibitor). A typical OCR trace from this mitochondrial function assay and how each parameter is derived is shown in Figure 4(A). Using this protocol, MCF7 cells were assayed in medium supplemented with glucose (5.56 mM) and increasing concentrations of pyruvate (0.1–5 mM). Representative OCR traces from cells assayed in medium containing 0, 0.1 and 1 mM pyruvate are shown in Figure 4(B). The results from the entire concentration dependency are shown in Figure 5.

Figure 4 Assessment of mitochondrial function using extracellular flux technology

(A) Schematic representation of the mitochondrial function assay. After establishment of baseline OCR, sequential injection of oligomycin (O), FCCP (F) and Antimycin A (A) allows for the determination of multiple mitochondrial function parameters including basal OCR, maximal OCR, ATP-linked OCR (ATP), proton leak (leak), reserve capacity and OCR independent of Complex IV (non-mito). (B) MCF7 cells were seeded in specialized microplates and cultured for 24 h. Cells were then switched to unbuffered DMEM supplemented with glucose (5.56 mM) and increasingly concentrations of pyruvate (0.1–5 mM) and mitochondrial function was assessed. OCRs were normalized to total protein/well after completion of the assay, and representative OCR traces are shown. Results are means±S.E.M., n=3–5.

Figure 5 Regulation of mitochondrial function by pyruvate

MCF7 cells were seeded in specialized microplates and cultured for 24 h. Cells were then switched to unbuffered DMEM supplemented with glucose (5.56 mM) and increasing concentrations of pyruvate (0.1–5 mM) and mitochondrial function was assessed using sequential injection of oligomycin, FCCP and Antimycin A. Basal and maximal OCR (A), ATP-linked OCR (B), proton leak (C), reserve capacity (D) and non-mitochondrial OCR (Non-Mito) (E) are shown. (F) ECAR was measured concomitantly. OCR and ECAR were normalized to total protein/well after completion of the assay. Results are means±S.E.M., n=3–5. *P<0.05 compared with 0 mM sodium pyruvate. N.S. denotes no significant difference between groups.

Interestingly, although the addition of pyruvate had little effect on basal OCR, there was a concentration-dependent increase in the maximal OCR with increasing concentrations of pyruvate (Figure 5A). This stimulation of maximal OCR by pyruvate also translates to a concentration-dependent increase in the reserve capacity (Figure 5D). Pyruvate had little effect on the mitochondrial function parameters ATP-linked OCR, proton leak and non-mitochondrial OCR (Figures 5B, 5C and 5E). ECAR (extracellular acidification rate), an indicator of glycolytic flux, was measured under the same conditions, and a slight decrease was observed with pyruvate supplementation (Figure 5F).

We next determined the effect of culture in glucose-only, pyruvate-only and complete media on these mitochondrial function parameters. After a 1 h incubation, basal and maximal OCR were significantly lower in cells provided only glucose in the medium; however, there was little difference in either of these parameters for cells provided only pyruvate or complete media (Figure 6A). Treatment with CHC, the plasma membrane MCT and mitochondrial pyruvate carrier inhibitor, also caused a significant decrease in the basal and maximal OCR when compared with complete media. Calculation of the reserve capacity from these data showed that cells provided complete media have the largest reserve capacity (Figure 6B). Cells provided only pyruvate have a significantly smaller reserve capacity than complete media, and cells provided only glucose or complete media and CHC have no reserve capacity. The time-dependent OCR traces from this experiment, as well as the effects on other mitochondrial function parameters, are shown in Supplementary Figure S2 (at http://www.BiochemJ.org/bj/444/bj4440561add.htm), and the effects on ATP-linked OCR largely mirror the effects on basal and maximal OCR (Figure 6A). Not surprisingly, ECAR was also higher under conditions where glucose was present in the assay medium.

Figure 6 Regulation of mitochondrial function by metabolic substrate supply

MCF7 cells were seeded in specialized microplates and cultured for 24 h. Cells were then switched to unbuffered DMEM containing glucose-only (5.5.6 mM), pyruvate-only (1 mM), glucose and pyruvate (complete media), or complete media with CHC (500 μM) 1 h prior to measuring mitochondrial function. Basal and maximal OCR (A) and the reserve capacity (B) were measured using sequential injection of oligomycin, FCCP and Antimycin A. OCRs were normalized to total protein/well after completion of the assay. Results are means±S.E.M., n=3–5. *P<0.05 compared with complete media.

In additional studies, we compared the effect of pyruvate and lactate on mitochondrial function. Using a similar approach as described above, MCF7 cells that were provided pyruvate in the assay medium had a high reserve capacity both in the presence and absence of glucose (Supplementary Figure S3 at http://www.BiochemJ.org/bj/444/bj4440561add.htm). In contrast, lactate was not able to support the reserve capacity independent of the presence of glucose in the assay medium. Moreover, to relate our mitochondrial function results to cellular energy balance, we assessed ATP and ADP levels after culture in different metabolic substrates for 72 h. Interestingly, steady-state levels of these adenine nucleotides were unchanged by supplementation with different substrates (Supplementary Figure S4 at http://www.BiochemJ.org/bj/444/bj4440561add.htm). Thus the changes in mitochondrial function observed in the present study do not equate with changes in the steady-state levels of adenine nucleotide pools.

These data indicate that pyruvate uptake via pyruvate transporters (MCTs) contributes to mitochondrial respiration and is essential if cells need to access their respiratory reserve, as may be required for robust proliferation. Moreover, lactate, unlike pyruvate, does not support the reserve capacity in MCF7 cells.

Effect of CHC on mitochondrial function in the presence of different metabolic substrates

The primary mechanism of action for MCT inhibitors in cancer is thought to be through blocking lactate efflux in highly glycolytic cancer cells; however, we have shown that MCT inhibition also limits pyruvate uptake (Figure 3). To disentangle whether CHC-dependent regulation of mitochondrial function occurs through deregulation of lactate efflux pathways or through limiting pyruvate uptake, we examined the effect of CHC on mitochondrial function in MCF7 cells. In this experiment, cells were supplied with glucose-only, pyruvate-only or complete media. We reasoned that inhibition of lactate efflux would be the predominant mechanism of action in cells cultured in glucose-only medium, whereas inhibition of pyruvate uptake would predominate in pyruvate-only medium. CHC (500 μM) significantly decreased basal and maximal OCR in cells provided pyruvate-only medium or complete media (Figures 7A and 7B). CHC treatment did not significantly decrease these parameters in cells provided glucose-only. A similar pattern was observed for the effect of CHC on the reserve capacity. CHC treatment inhibited the reserve capacity only in cells cultured under pyruvate-only medium or complete media conditions (Figure 7C). These results suggest that inhibition of pyruvate uptake by CHC regulates mitochondrial function in MCF7 cells. The time-dependent OCR traces from this experiment, as well as the effects on other mitochondrial function parameters, are shown in Supplementary Figure S5 (at http://www.BiochemJ.org/bj/444/bj4440561add.htm), and the effects on ATP-linked OCR largely mirror the effects on basal OCR (Figure 7A). Interestingly, ECAR was stimulated by CHC in conditions where glucose was present in the assay medium suggesting that at this concentration lactate efflux in not readily inhibited by CHC.

Figure 7 Effect of CHC on mitochondrial function in the presence of different metabolic substrates

MCF7 cells were seeded in specialized microplates and cultured for 24 h. At 1 h prior to the assessment of mitochondrial function, cells were switched to unbuffered DMEM containing glucose-only (5.56 mM), pyruvate-only (1 mM), or glucose and pyruvate (complete media) in the absence (closed bars) or presence (open bars) of CHC (500 μM). Basal OCR (A), maximal OCR stimulated with 3 μM FCCP (B) and reserve capacity (C) were measured. OCRs were normalized to total protein/well after completion of the assay. Results are means±S.E.M., n=3–4. *P<0.05 compared with media condition without CHC. N.S. denotes no significant difference between groups.

Effect of CHC on cell proliferation in the presence of different metabolic substrates

A similar approach was employed to examine whether inhibition of lactate efflux or pyruvate uptake by a MCT inhibitor is important in the regulation of cancer cell proliferation. MCF7, MB231 and T-47D breast cancer cells were treated with CHC (500 μM) in medium containing only glucose or only pyruvate for 72 h. CHC decreased the total cell number after treatment in all three cell types when cells were cultured in pyruvate-only medium (Figure 8A). Additionally, for both MCF7 and T-47D cells, but not MB231 cells, CHC treatment significantly decreased the total cell number after incubation in glucose-only medium.

Figure 8 Effect of CHC on cell proliferation in the presence of different metabolic substrates

(A) MCF7, MB231 and T-47D cells were cultured in medium containing glucose-only (5.56 mM) or pyruvate-only (1 mM) in the absence (closed bars) or presence (open bars) of CHC (500 μM) for 72 h. Total cell number was then assessed using the MTT assay. (B) MCF7 cells were exposed to the same media conditions for 24–72 h, and total cell number was assessed using the MTT assay. (C) MCF7 cells were seeded at low density in medium containing glucose-only (5.56 mM), pyruvate-only (1 mM), or glucose and pyruvate (complete media) in the absence (closed bars) or presence (open bars) of CHC (500 μM), and the number of colonies formed was measured (C). Results are means±S.E.M., n=3–6. *P<0.05 compared with media condition without CHC. N.S. denotes no significant difference between groups.

These findings were confirmed by examining the growth kinetics of MCF7 cells provided glucose or pyruvate with or without CHC (500 μM). CHC decreased cell proliferation in both medium conditions (Figure 8B). We also examined the effect of CHC on the ability of MCF7 cells to form colonies in different metabolic-substrate-containing media. There was significant inhibition of colony formation upon exposure to CHC in both pyruvate-only medium and complete media (Figure 8C). No colonies were observed in glucose-only medium with or without CHC (results not shown). In summary, these results demonstrate that MCT inhibition limits proliferation in the presence of only glucose or only pyruvate and imply that both lactate efflux and pyruvate uptake through MCTs are mechanistically linked to proliferative responses.

DISCUSSION

In the present study, we have examined the effect of different metabolic substrates on cancer cell proliferation and mitochondrial function. It is well-established that metabolic pathways are deregulated in cancer in a manner which supports rapid proliferation and provides a growth advantage over normal cells [28]. Since Otto Warburg's [1] early observations that cancer cells utilize glycolytic metabolism even in the presence of adequate oxygen, much of the research focus has been on the role of glycolysis in cancer. It is now becoming clear that, whereas glycolysis is a less efficient means to generate ATP when compared with mitochondrial OXPHOS (oxidative phosphorylation; 2 ATP molecules per glucose molecule using glycolysis compared with 38 ATP molecules using OXPHOS), glycolytic metabolism provides a growth advantage through several different mechanisms (reviewed in [28]). These include generating ATP at a faster rate than OXPHOS, providing biosynthetic substrates needed for rapid proliferation (e.g. NADPH and ribose 5-phosphate) [29,30], and supporting cell growth under hypoxic conditions [31]. Though less well-understood, recent literature also demonstrates the important role mitochondria play in cancer cell proliferation through the metabolism of glutamine in the mitochondrial citric acid cycle to supply additional biosynthetic substrates required for cell growth [17,29]. Because metabolic reprogramming is a nearly universal cancer phenotype, cancer cell metabolism is now thought to be an important therapeutic target in the treatment of many malignancies [2].

In the present study, we show that breast cancer cells (MCF7, MB231 and T-47D) proliferate more rapidly when provided pyruvate as a metabolic substrate when compared with glucose (Figure 1), and this occurs in the absence of overt cell death (Figure 2). This result is somewhat surprising given that glycolysis is preferentially up-regulated in cancer cells when compared with their normal counterparts [1], and thus our data imply that glycolysis alone is not sufficient to support rapid proliferation. This concept is consistent with recent reports that indicate that supplementation of energy substrates which can be metabolized through the mitochondrial citric acid cycle and OXPHOS pathways (e.g. glutamine and pyruvate) support more rapid cancer cell growth when compared with glucose in several cell types (e.g. K-Ras-transformed HCT-116 colon carcinoma cells and glioblastoma) [3,32]. We also compared the proliferative responses after supplementation of pyruvate compared with lactate, another monocarboxylate that is transported through MCTs. Although pyruvate supplementation led to a concentration-dependent increase in proliferation, lactate supplementation was unable to support cell growth (Supplementary Figure S1). The different effects observed between these two monocarboxylates may result from alterations in their uptake into cells or their intracellular metabolism. Lactate can be converted into pyruvate by LDH using NAD+ as a cofactor. We found that LDH activity was not significantly different in cells cultured in lactate compared with pyruvate (Supplementary Figure S1). Moreover, there was no difference in the total levels of NAD+ or NADH in cells which were provided pyruvate or lactate, and NAD+ levels were higher than NADH in both substrate conditions which would allow for conversion of lactate into pyruvate (Supplementary Figure S1). Therefore neither of these findings clarifies why lactate was unable to promote proliferation, and further studies are warranted. Other possible explanations include that lactate is not readily transported into MCF7 cells or the rate at which lactate is metabolized to pyruvate is not sufficiently rapid to fuel mitochondrial respiration.

Previously, significant focus has been put on understanding the role of mitochondrial metabolism in cancer cells. New literature has highlighted the important ‘biosynthetic activity’ of mitochondria which generate substrates for lipid, protein and nucleic acid biosynthesis in rapidly dividing cells [29]. Since pyruvate is a substrate for the citric acid cycle, and thus drives mitochondrial function, we examined the effect of pyruvate on respiration in MCF7 cells. We showed that titration of pyruvate in the presence of glucose has little effect on the basal OCR; however, it clearly stimulated the maximal OCR elicited by the mitochondrial uncoupler FCCP (Figure 5). In addition, cells cultured in medium which contain pyruvate have high basal OCR, maximal OCR and reserve capacity than those which are cultured in the absence of pyruvate (Figure 6). Interestingly, there are no significant differences in the steady-state levels of ATP or ADP after culture in medium containing varied metabolic substrates (Supplementary Figure S4), so these results do not equate to general changes in cellular energy balance. However, we cannot rule out changes in total mitochondrial mass due to changes in mitochondrial biogenesis or mitophagy pathways. In fact, it has recently been shown that mitophagy is stimulated in response to nutrient deprivation in hepatocytes [33]. Nonetheless, our results clearly demonstrate that pyruvate is required for MCF7 cells to have a reserve capacity, and there is also a correlation between the inability of lactate to promote cell growth (Figure 1) and support the reserve capacity (Figures 5 and 6). These findings are particularly interesting since it is thought that the reserve capacity is critical for maintaining cellular function in response to multiple stressors [24,34]. In the context of cancer, loss of the reserve capacity in breast cancer cells correlates with activation of cell death through mitochondrial apoptotic signalling pathways [35]. Drug-resistant glioma cells have also been shown to have increased reserve capacity and are not sensitive to cell killing by multiple agents [36]. Taken together, these data indicate that supplying metabolic substrates which drive mitochondrial respiration and, specifically support the reserve capacity of MCF7, correlates with more rapid proliferation of these cells.

Given the important role of pyruvate in fuelling mitochondrial respiration and promoting cell growth, we also examined the effect of the inhibitor of pyruvate uptake into cells and mitochondria, CHC. Firstly, to demonstrate that this compound functions as expected in our system, we showed that treatment of MCF7 cells with CHC (50–500 μM) inhibits pyruvate transport into cells without affecting glucose uptake (Figure 3) in a concentration-dependent manner. Treatment of cells with CHC in complete media also slowed cancer cell growth in two of the three breast cancer cell lines examined (MCF7 and MB231; Figure 1) and significantly decreased basal OCR, maximal OCR and the reserve capacity (Figure 6). Together with the results from cells cultured with different metabolic substrates, this points to a link between mitochondrial function parameters and proliferative responses.

The primary mechanism of action for the regulation of cell growth by MCT inhibition is thought to be through blocking lactate efflux from highly glycolytic cancer cells. This inhibition of lactate efflux results in intracellular acidification and deregulation of cancer cell growth [37]. Our results suggest that pyruvate is an important metabolic substrate for cancer cell proliferation, and MCT inhibitors may work through blocking pyruvate metabolism in addition to their effects on lactate efflux pathways. To further investigate this concept, we examined the effects of CHC in cells provided glucose or pyruvate as metabolic substrates. We reasoned that, in cells provided only glucose, the inhibition of lactate efflux would be the predominant mechanism of action for CHC; however, in cells provided only pyruvate, inhibition of pyruvate uptake into cells would dominate. Examining the effects of CHC on mitochondrial function showed that CHC decreased basal OCR, maximal OCR and the reserve capacity of MCF7 cells supplied only pyruvate or complete media; however, there was not a significant effect of CHC on these parameters in cells provided only glucose (Figure 7). Consistent with this, in both MCF7 and T-47D cells, inhibition of proliferation was observed after treatment with CHC in medium containing either glucose or pyruvate. For MB231 cells, although there was a trend towards inhibition of growth with CHC in the glucose-only condition, statistically significant inhibition of growth was observed only when cells were provided pyruvate (Figure 8). CHC also attenuated colony formation when cloning medium contained only pyruvate.

Since CHC is known to be orders of magnitude more effective at inhibiting mitochondrial pyruvate transport than transport at the level of the plasma membrane [38], we believe the site of action for CHC in our model system is likely to be the mitochondrion. This is further supported by our results which indicate that the ECAR and lactate efflux are, in fact, increased with CHC treatment (Figure 3 and Supplementary Figure S5), not decreased as one would expect if the plasma membrane transporters were targeted. Volk et al. [39] were the first to demonstrate that CHC does not inhibit lactate efflux from glial cells, and the findings of the present study confirm their result.

In the present study, we have used CHC as a classical inhibitor of MCTs. Evidence suggests that CHC is selective for the MCT1 isoform and is an effective non-competitive inhibitor of lactate uptake in tumour cells at concentrations in the low micromolar range [IC50 for lactate (5 mM) uptake=166 μM] [5]. Although the bulk of enzyme kinetics studies using this compound have focused on the inhibitory action of CHC with respect to lactate uptake, since the Km values for lactate and pyruvate uptake in tumour cells are in the same range (Km for lactate=4.54 mM and Km for pyruvate=0.72 mM) [5], we believe they may be extrapolated to some extent to our findings. Recent studies have also investigated the intracellular distribution of MCT isoforms and have found differential expression patterns and differential distribution to the plasma and mitochondrial membranes in breast cancer cells [40]. Furthermore, MCT1 was reported to reside in the mitochondria (MitoCarta mitochondrial proteome list; [41]). Thus there are clearly multiple levels of regulation of monocarboxylate transport within cells.

We believe our findings may be particularly relevant in the context of the proposed ‘metabolic symbiosis’ model in the tumour microenvironment, first described by Sonveaux et al. [16]. They suggest that both glucose and lactate are metabolic substrates for cancer cells within solid tumours; however, their use as substrates is spatially distinct [16]. This process is somewhat analogous to the cell–cell lactate shuttle observed in the working muscle bed [42]. In the absence of adequate oxygen, hypoxic tumour cells rely on glucose as a metabolic substrate and export lactate through MCT4. This effluxed lactate is then transported into oxygenated cancer cells through MCT1 and is used to fuel oxidative metabolism. Inhibition of MCTs disrupts the ‘metabolic symbiosis’ and results in decreased tumour size in vivo.

Additional studies are certainly needed to determine the effects of MCT inhibitors on the metabolic pathways under hypoxic stress. Pyruvate-driven respiration is unlikely to provide a growth advantage under hypoxia. Nonetheless, the present study highlights the important role of mitochondrial function, and in particular the reserve capacity in cancer cell proliferation and describes an alternative mechanism of action for MCT inhibitors which may give renewed impetus for their development as chemotherapeutic agents.

AUTHOR CONTRIBUTION

Anne Diers performed experiments, analysed and interpreted the data, and wrote the paper. Katarzyna Broniowska and Ching-Fang Chang performed experiments, analysed and interpreted the data, and edited the paper prior to submission. Neil Hogg supervised experiments, performed data analysis and interpretation, and edited the paper prior to submission.

FUNDING

This work was supported by the Redox Biology Program at the Medical College of Wisconsin (to N.H.), an Interdisciplinary Cancer Research Post-Doctoral Fellowship from the Cancer Center of the Medical College of Wisconsin (to A.R.D.), and the National Institutes of Health [grant number R01-GM-55792 (to N.H.)].

Acknowledgments

The authors would like to acknowledge Dr Balaraman Kalyanaraman (Medical College of Wisconsin, Milwaukee, WI, U.S.A.) for important support for these studies.

Abbreviations: CHC, α-cyano-4-hydroxycinnamic acid; DMEM, Dulbecco's modified Eagle's medium; ECAR, extracellular acidification rate; FBS, fetal bovine serum; FCCP, carbonyl cyanide p-trifluoromethoxyphenylhydrazone; LDH, lactate dehydrogenase; MCT, monocarboxylate transporter; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide; OCR, oxygen consumption rate; OXPHOS, oxidative phosphorylation

References

View Abstract