Figure 1. Metformin increases healthspan without a requirement for function of the insulin signaling pathway.

A. Survival curves of wild-type (N2) animals raised at 20°C on nematode growth media plates containing either no metformin or final concentrations of 1 mM, 10 mM, or 50 mM metformin. (Note that C. elegans has a highly protective cuticle and intestinal lining that generally limit drug uptake such that it is not unusual for polar drugs to be applied at a concentration 1000 fold higher than their predicted affinity for the target [94], [95]; physiological levels of drug in the animals are anticipated to be much lower). For the trial presented here, median survival for animals on 0 mM, 1 mM, 10 mM, and 50 mM metformin plates was 15, 15, 17, and 21 days, respectively. The survival curves of the nematodes raised on 0 mM, 1 mM, and 10 mM metformin plates are not significantly different, while the survival curve for animals raised on 50 mM metformin is significantly different than the 0 mM metformin control (P<0.0001 by the Log-rank (Mantel-Cox) test; benefits are lost at 100 mM metformin, data not shown). The pooled data for all trials show an approximately 27% increase in median lifespan with 50 mM metformin treatment and a significantly right-shifted survival curve (P<0.0001 by the Log-rank test), see Table S1A. All studies documented here involved life-long metformin exposure, but a preliminary trial in which metformin was introduced at the last larval stage suggests that treatment in adult life is sufficient to confer some lifespan benefit (data not shown). B. Swimming rates for wild-type animals raised from eggs on plates (time 0) containing either no metformin or 50 mM metformin at 20°C. We recorded the number of body bends/30 seconds for individual animals placed in liquid media at early adulthood (5 days), mid-life (10 days), and late life (15 days). We present the averages of three independent trials. Metformin-treated animals swim at similar rates in young adulthood, supporting that metformin does not induce hyperactive swimming or confer developmental defects, but the swimming rates of animals raised on 50 mM metformin are significantly higher than those of the control group on days 10 and 15 (P = 0.0058 and P = 0.0346, respectively, by unpaired t tests), indicating extended mid-life and late-life locomotory ability consequent to metformin treatment. C. Survival curves of null mutant daf-16(mgDf50) raised on 0 mM and 50 mM metformin plates at 20°C. The median survival is 11 and 15 days for animals raised on 0 mM and 50 mM metformin, respectively, and the survival curves of the two groups are significantly different (P = 0.0111 by the Log-rank test). 1 mM and 10 mM metformin do not significantly change the survival curves of daf-16(mgDf50) animals (data not shown). We performed this experiment a total of four times with similar results (pooled data show a 15% median lifespan increase and a significantly right-shifted survival curve (P<0.0001, Log-rank) for animals treated with 50 mM metformin, see Table S1B for additional data). D. Survival curves of age-1(hx546) mutants raised on 0 mM and 50 mM metformin plates at 20°C. Median survival for animals on 0 mM and 50 mM metformin is 24 and 31 days, respectively, and the survival curves are significantly different by the Log-rank test (P = 0.0014). We found similar median survival extension in a single repeat of this experiment (pooled data for the two trials show an approximate 36% median lifespan increase and significantly different survival curves (P<0.0001, Log-rank), see Table S1C).