We report that the anticancer activity of the widely used diabetic drug metformin is strongly potentiated by syrosingopine. Synthetic lethality elicited by combining the two drugs is synergistic and specific to transformed cells. This effect is unrelated to syrosingopine’s known role as an inhibitor of the vesicular monoamine transporters. Syrosingopine binds to the glycolytic enzyme α-enolase in vitro, and the expression of the γ-enolase isoform correlates with nonresponsiveness to the drug combination. Syrosingopine sensitized cancer cells to metformin and its more potent derivative phenformin far below the individual toxic threshold of each compound. Thus, combining syrosingopine and codrugs is a promising therapeutic strategy for clinical application for the treatment of cancer.
Metformin, an oral antidiabetic of the biguanide class, is the most widely prescribed drug for type 2 diabetes. It acts as a mild mitochondrial inhibitor, and its impact on key metabolic organs, such as the liver and muscle, results in reduced blood glucose levels and restoration of insulin responsiveness at the whole-body level. Metformin is a well-tolerated drug that can be taken for years with few adverse effects, such as rare cases of lactic acidosis.
There is increasing evidence that metformin, in addition to its antidiabetic effect, has anticancer properties. Epidemiological studies have shown that the incidence of cancer is significantly reduced in diabetic patients on metformin (1–5). Metformin prescription as a neoadjuvant also results in an improved clinical response after operation (6, 7). These observations are corroborated by in vitro studies in which metformin inhibits growth of cancer cell lines from a wide variety of tissue types (8), and in vivo studies where metformin is active in mouse xenograft models (9). Metformin is believed to exert its anticancer effect by an indirect systemic and direct cellular effect (10). The systemic effect of metformin lowers blood glucose levels and, hence, glucose availability to glucose-hungry cancer cells. At the cellular level, metformin inhibits mitochondrial respiration in cancer cells that are already experiencing a shortfall in glucose uptake, leading to lower adenosine 5′-triphosphate (ATP) levels. The resulting activation of AMPK leads to inhibition of mammalian target of rapamycin complex 1 (mTORC1), a major signaling hub for cell growth, translation, and metabolism. Several clinical trials investigating metformin’s antineoplastic activity are currently in progress [www.clinicaltrials.gov, which is reviewed by Pollak and Chae et al. (11, 12)]. Despite this generally favorable picture of metformin as a potential anticancer agent, there is a persistent background of studies where metformin shows little to no effect in cancer (13–16) or no benefit in comparison with other antidiabetics (17, 18).
A resolution to the discrepancy between the studies noted above is that the concentration of metformin used to obtain antineoplastic activity in vitro greatly exceeds the in vivo serum concentration that can be attained with standard antidiabetic prescription (19). Thus, there may be a clinical gray zone, where metformin efficacy is dependent on factors that may vary between different study groups. This suggests that, for the development of metformin as a clinically efficacious anticancer agent, it would be desirable to lower its therapeutic threshold to within a reasonably attainable range in vivo. With this objective, we performed a screening to identify compounds that interact with a low, sublethal concentration of metformin to induce cancer cell killing. We show that the antihypertensive syrosingopine, a derivative of reserpine, is synthetically lethal with metformin and that both compounds interact synergistically to kill a broad variety of cancer cell types while demonstrating no activity against nontransformed cells. Synthetic lethality with syrosingopine was also observed with every inhibitor of the mitochondrial electron transport chain (ETC) tested. The synthetic lethal interaction evoked by syrosingopine and mitochondrial inhibitors occurs at concentrations substantially below the toxic threshold for either compound alone. These results suggest that a combination of syrosingopine with mitochondrial inhibitors represents a novel option for both classes of compounds in cancer therapy. Lowering the therapeutic threshold of metformin, and the related biguanide phenformin, may allow their clinical application as anticancer agents.
Syrosingopine potentiates cell killing by metformin
We screened a drug library (1120 compounds) using murine 6.5 mast cells to identify drugs that are cytotoxic only in the presence of metformin. These are Pten-null, vH-Ras–expressing, and mTORC2-addicted cells that display a transformed phenotype (20). Parallel screens were performed in the absence and presence of metformin. The concentration of metformin used in the screen was determined by a prior titration with metformin; eventually, 4 mM metformin was chosen because it has minimal effect on cell growth (fig. S1A). Syrosingopine was the only compound that fulfilled our selection criteria (>80% growth in the absence of metformin and <10% growth in the presence of metformin; Fig. 1A). Syrosingopine is a semisynthetic derivative of reserpine (fig. S1B), both of which are clinically approved antihypertensives (21).