Beyond Warburg effect--dual metabolic nature of cancer cells

Abstract

Warburg effect is a dominant phenotype of most cancer cells. Here we show that this phenotype depends on its environment. When cancer cells are under regular culture condition, they show Warburg effect; whereas under lactic acidosis, they show a nonglycolytic phenotype, characterized by a high ratio of oxygen consumption rate over glycolytic rate, negligible lactate production and efficient incorporation of glucose carbon(s) into cellular mass. These two metabolic modes are intimately interrelated, for Warburg effect generates lactic acidosis that promotes a transition to a nonglycolytic mode. This dual metabolic nature confers growth advantage to cancer cells adapting to ever changing microenvironment.

Figure 1. The effect of lactic acidosis on oxygen consumption, glycolysis, and fate of glucose metabolism in 4T1 cells.

(a) The effect of lactic acidosis on oxygen consumption rate (OCR), extracellular acidification rate (ECAR), and OCR/ECAR ratio in 4T1 cells. Vertical lines indicate time points of the administration of corresponding inhibitors. When oligomycin is added, the ATP synthase complex is inhibited such that the respiratory chain associated oxygen consumption is inhibited. When the ATP synthesis uncoupler FCCP is added, oxygen consumption by the respiratory chain is resumed; when rotenone (complex I inhibitor) and antimycin A (complex III inhibitor) are added, the electron transfer as well as oxygen consumption by the respiratory chain is blocked. (b) ATP levels in 4T1 cells with or without lactic acidosis. (c) 4T1 cells are incubated in culture medium with or without lactic acidosis (20 mM lactate, pH 6.7) for 48 hours traced by [¹⁴C]-6-D-glucose. The following parameters are determined as described in Methods: glucose consumption, lactate generation, cell counts and incorporation of ¹⁴C into DNA and RNA (μmole glucose/μg DNA or RNA/per μmole glucose consumed, see Methods). The data are mean ± SD, n = 3, except n = 6 in Seahorse experiment, and are confirmed by 2 independent experiments. * p < 0.05, ** p < 0.01, compared with control.

Figure 2. Effect of lactic acidosis on glucose consumption and lactate generation by 4T1 cells.

(a) A titration of exogenous lactic acid on glucose consumption, net lactate production and cell division. 4T1 cells were incubated in RPMI-1640 medium supplemented with lactic acid to desired lactate concentrations as indicated, incubated for 24 hours, followed by cell counting and glucose consumption and net lactate generation determinations. L/G ratio: lactate generated over glucose consumption; C/G ratio, cells produced by consumption of 1 μ mole glucose. (b) Lactic acidosis generated by 4T1 cancer cells transit Warburg effect to a nonglycolytic phenotype. 4T1 cells were incubated in the RPMI-1640 medium containing 18 mM glucose; at indicated hours, medium glucose and lactate levels, medium pH and L/G ratios were monitored. A given L/G ratio denotes μmoles lactate generated from consuming one μmole glucose within an incubation period, e.g., the L/G ratio at 64 hours means the lactate generated between 48 and 64 hours divided by the glucose consumed in the same period. Note that the L/G ratio is correlated with the medium pH and inversely correlated with the lactate concentration. Data are mean ± SD, n = 3, and were confirmed by 2 independent experiments. * p < 0.05, ** p < 0.01, compared with control.

Figure 3. The effect of lactic acidosis on oxygen consumption and glycolysis in Bcap 37, Hela and A549 cells.

The effect of lactic acidosis on oxygen consumption rate (OCR), extracellular acidification rate (ECAR), and OCR/ECAR ratio in Bcap37, Hela, and A549 cells. The data are mean ± SD, n = 6, and are confirmed by 2 independent experiments. * p < 0.05, ** p < 0.01, compared with control.

Figure 4. The conversion between pyruvate + NADH and lactate + NAD⁺ in 4T1 cells under varying lactic acidosis conditions.

(a–c) Intracellular concentrations of pyruvate and lactate, and lactate/pyruvate ratios under varying lactic acidosis conditions (culture medium supplemented with pure lactic acid as indicated). The data represent mean ± SD, n = 3, and are confirmed by 2 independent experiments. * p < 0.05, ** p < 0.01, compared with control. (d & e) Glucose consumption and lactate generation by cells under varying lactic acidosis conditions. (f) The estimated mass action ratio Q ([lactate][NAD]/[pyruvate][NADH]) values in 4T1 cells cultured under varying lactic acid conditions. (g) Changes of Gibbs free energy (ΔG) for the conversion between pyruvate + NADH and lactate + NAD⁺ in 4T1 cells, according to the equation in Methods.

Figure 5. Effect of pH on glycolytic flux.

(a) Intracellular pH (pHi) and extracellular pH (pHe), when 4T1 cells are cultured with or without lactic acidosis (culture medium with 20 mM lactate, pH 6.7). (b) A time course glycolysis assay containing 250 μg/ml 4T1 cell lysate along with all essential cofactors and 10 mM glucose, a titration glycolysis assay with indicated protein concentrations for 60 minutes, and pH-dependent inhibition of the glycolytic flux in a reaction containing 250 μg/ml protein of 4T1 cell lysate (60 minutes). Reaction mixtures containing boiled 4T1 cell lysates serve as controls. (c) pH-dependent inhibition of individual glycolytic enzymes and LDH in 4T1 cell lysate. The data are mean ± SD, n = 3, and are confirmed by 3 independent experiments. * p < 0.05, ** p < 0.01, compared with control.

Figure 6. Effect of lactic acidosis (20 mM/pH 6.7) on glucose uptake, intra/extracellular glucose levels, glucose consumption and lactate generation.

(a) Glut1 expression. 4T1 cells incubated in RPMI-1640 medium with lactic acidosis for indicated hours were lysed, total proteins resolved by SDS-polyacrylamide gel electrophoresis, and Glut1 expression analyzed by immuno-blot. (b) Lactic acidosis significantly inhibits [³H]-2DG uptake by 4T1 cells. (c & d) Intracellular and extracellular glucose levels of 4T1 cells incubated with or without lactic acidosis. (e) Lactate generation by 4T1 cells with or without lactic acidosis. Data represent mean ± SD, n = 3, and were confirmed by at least 2 independent experiments. * p < 0.05, ** p < 0.01, compared with control.

Figure 7. The effect of lactic acidosis on the expression and activities of glycolytic enzymes.

Cells were incubated in RPMI-1640 medium supplemented with lactic acidosis. At 0, 1, 2 or 3 day, individual enzyme activity and expression level in derived cell lysates were measured. (a) Enzyme activity for each glycolytic enzyme in the lysates of Bcap37 cells (upper panel) and 4T1 cells (lower panel) incubated with lactic acidosis over the indicated time course. Data represent mean ± SD, n = 3. * p < 0.05, ** p < 0.01, compared with control. Data were confirmed by 2 independent experiments. (b) Immuno-blots analyses for indicated glycolytic enzyme in the lysates of Bcap37 cells (left panel) or 4T1 cells (right panel) incubated at the same condition as in (a).

Figure 8. Impacts of lactic acidosis on cancer cell proliferation or survival in vitro and in vivo.

(a) Lactic acidosis allows 4T1 cells to grow under moderate glucose deprivation. Cells incubated in RPMI-1640 medium containing 0.5 mM glucose without lactic acidosis or with lactic acidosis (20 mM lactate and pH 6.7) are counted at indicated time points. Data represent mean ± SD, n = 3, and are confirmed by 2 independent experiments. (b) Lactic acidosis allows a significantly longer sustainable proliferation with limited supply of glucose. Cells incubated in RPMI-1640 medium containing 6 mM glucose without lactic acidosis or with lactic acidosis (20 mM lactate and pH 6.7) are counted at indicated time points. Data represent mean ± SD, n = 3, and are confirmed by 2 independent experiments. (c–i) Elevating intratumoral pH inhibits tumor growth and promotes tumor necrosis. Sodium bicarbonate administration surrounding tumor areas inhibits 4T1 tumor growth and increases their necrosis. (c–e) Intratumoral glucose and lactate concentration and pH; (f) Body weight curves; (g) Tumor growth curves; (h) Tumor weights on the day of sacrifice; (i) Necrosis in tumors. The n is 5 for each group at the starting point; however, one mouse in the control group dies on day 19. Hence, the images of 4 tumor samples from each group are shown. For detailed experimental procedures refer to Methods. The statistical data represent mean ± SEM. * p < 0.05, ** p < 0.01, *** p < 0.001, compared with control.