Metabolic Capability to Induce the Pasteur Effect Mediates Sensitivity of Human Leukemic Cells to Metformin

Abstract

2601 An emerging hallmark of cancer cells is the reprogramming of intermediary and energy metabolism these cells undergo. Several epidemiological studies have shown that metformin, widely used to treat patients with type 2 diabetes, may reduce their risk of cancer. Despite several reports of anti-neoplastic activity of metformin, the mechanisms responsible for this activity have not been fully elucidated in cancer or leukemic cells. We hypothesized that metformin elicits a metabolic reprogramming driven by alterations in mitochondrial function and signaling, which induces apoptosis in leukemic cells, and that metabolic flexibility determines the variation(s) of the cytotoxic response to metformin among different leukemic cell lines. We first demonstrated that metformin markedly decreased oxygen consumption of six leukemic cell lines in a concentration-dependent manner. We also observed that the cytotoxic effect of metformin varies between cell lines reflecting their energetic capacity to compensate for the mitochondrial inhibition induced by metformin (eg. to induce the Pasteur effect). Importantly, metformin-insensitive leukemic cells did not exhibit a Pasteur effect in response to metformin. All leukemic cells exhibited high basal conversion of glucose to lactate (eg. aerobic glycolysis) and specific expression of key metabolic genes as compared to normal mononuclear cells. Despite dependence on glucose catabolism, metformin sensitivity was associated with relative resistance to glucose starvation. Metformin effects in drug-resistant cells were potentiated by the addition of a glycolytic inhibitor, but not by inhibitors of the pentose phosphate pathway or glutaminolysis. Leukemic cells with broad metabolic capacities to utilize other energetic substrates in response to diverse nutrient starvation showed insensitivity to metformin. Metformin induced a significant decrease in metabolites of the upper segment of glycolysis and the oxidative branch of the pentose phosphate pathway as well as a clear increase of PRPP and IMP biosynthesis. Energy charge, the nucleotide phosphate pool and lactate/glucose ratio remained stable after metformin treatment. Furthermore, our results showed that basal glucose uptake/consumption and the activity of the lower segment of the glycolytic pathway are key determinants of a cytotoxic response to metformin. In addition, high glutathione, malate, IMP and orotate content were observed in metformin-insensitive leukemic cells. Moreover, the cytotoxic effect of metformin was independent of AMPK/LKB1 status of the leukemic cells while p53 expression abrogated this effect. The presence of wild-type p53 appears to partially protect tumor cells from glucose starvation and metformin cytotoxicity and prevents the induction of the Pasteur effect. Finally, we demonstrated that metformin increased the cytotoxicity of chemotherapy agent, cytarabine, on all leukemic cell lines in vitro and significantly reduced leukemic colony-forming units (CFU-L) from six primary AML patient samples in a concentration-dependent manner. Additional experiments on metabolic and signaling pathways as well as in vivo studies are in progress to better understand the cytotoxic response of metformin in both AML cell lines and primary AML patient specimens that impact the therapeutic potential of metformin in vivo. Disclosures: Carroll: Agios Pharmaceuticals: Research Funding; TetraLogic Pharmaceuticals: Research Funding; Sanofi Aventis Corporation: Research Funding; Glaxo Smith Kline, Inc.: Research Funding.