An AMPK-FOXO pathway mediates longevity induced by a novel method of dietary restriction in C. elegans

Abstract

Background: Dietary restriction (DR) is the most effective environmental intervention to extend lifespan in a wide range of species. However, the molecular mechanisms underlying the benefits of DR on longevity are still poorly characterized. AMP-activated protein kinase (AMPK) is activated by a decrease in energy levels, raising the possibility that AMPK might mediate lifespan extension by DR.

Results: By using a novel DR assay that we developed and validated in C. elegans, we find that AMPK is required for this DR method to extend lifespan and delay age-dependent decline. We find that AMPK exerts its effects in part via the FOXO transcription factor DAF-16. FOXO/DAF-16 is necessary for the beneficial effects of this DR method on lifespan. Expression of an active version of AMPK in worms increases stress resistance and extends longevity in a FOXO/DAF-16-dependent manner. Lastly, we find that AMPK activates FOXO/DAF-16-dependent transcription and phosphorylates FOXO/DAF-16 at previously unidentified sites, suggesting a possible direct mechanism of regulation of FOXO/DAF-16 by AMPK.

Conclusions: Our study shows that an energy-sensing AMPK-FOXO pathway mediates the lifespan extension induced by a novel method of dietary restriction in C. elegans.

Figure 1. sDR: a novel dietary restriction method in C. elegans

(A) Description of our dietary restriction method (sDR) in C. elegans. Populations of worms obtained by synchronized egg laying (EL) were scored and switched to freshly seeded plates every other day (*). sDR was initiated at day 4 of adulthood. (B) Representative pictures of worms fed different concentrations of bacteria expressing DsRed, a red fluorescent protein. Right panel: quantification of DsRed fluorescence of wild type (N2) worms incubated with different bacterial dilutions. Results represent the average and SEM of 3 independent experiments performed with 30 worms per condition at day 7. ***: p<0.001 in one way ANOVA. (C) Serial dilution of OP50-1 bacteria increases worm lifespan (5×10⁸ bacteria/ml= 28.9% increase over 5×10¹¹ bacteria/ml). Mean and standard errors for this experiment done in triplicate are presented in Table S1A. (D) Gompertz analysis shows that sDR decreases the worm mortality rate. (E) sDR treated worms displayed a 36% increase in survival compared with AL-fed worms at 200 mM Paraquat (p<0.0001). The mean and SEM for duplicate experiments done in sextuplicate is presented in Table S1B. (F) sDR delays the age-dependent decline in worm locomotor activity. Quantification of the length of worm tracks after one hour on fresh ad libitum lawns of OP50-1 bacteria was performed with Image J. Each point represents the average movement of at least 20 worms. *: p<0.05; **: p<0.01; ***: p<0.001 in one way ANOVA. (G) sDR induced by dilution of UV-killed bacteria results in a 15.9% increase in lifespan of N2 worms (statistically different from ad libitum fed worms with a p=0.0004). The mean and SEM of two independent experiments conducted in triplicate is presented in Table S1C.

Figure 2. AMPK is necessary for sDR’s beneficial effects on lifespan and age-dependent decline

(A) The AMP:ATP ratio, measured by reverse phase HPLC, increases when worms are dietarily restricted at day 7. The results presented are the mean and SEM of two independent experiments of 5 samples each. **, p=0.0079 by Mann Whitney test. (B) AMPK mediates sDR’s ability to extend lifespan. aak-2 mutant worms displayed a mild decrease in lifespan compared to WT (N2) worms in ad libitum (AL) conditions (5×10¹¹ bacteria/ml) (10.5%; p<0.0001). Under sDR conditions (5×10⁸ bacteria/ml), aak-2 mutant worms had a 48% shorter lifespan compared to WT (N2) worms (p<0.0001). Comparison of five independent experiments performed in triplicate is presented in Table S2. (C) sDR delays the age dependent decline in locomotor activity in an aak-2 dependent manner. The experiment was performed as described in the legend of Figure 1F. *: p<0.05; **: p<0.01; ***: p<0.001 in one way ANOVA. Note that N2 worms are also represented in Figure 1F.

Figure 3. FOXO/DAF-16 is necessary for sDR’s beneficial effects on lifespan and age-dependent decline

(A) DAF-16 is necessary for sDR to extend lifespan. sDR leads to a 30.6% increase (p<0.0001) in longevity in wild type (N2) worms but does not significantly affect the lifespan of daf-16 mutant worms. The mean and SEM of three independent experiments conducted in triplicates of 30 worms each is presented in Table S3A. (B) sDR delays the age dependent decline in locomotor activity in a daf-16 dependent manner. The experiment was performed as described in the legend of Figure 1F. *: p<0.05; **: p<0.01; ***: p<0.001 in one way ANOVA. Note that N2 worms are also represented in Figures 1F. (C) sDR is not dependent on the insulin receptor daf-2. sDR induced a 35.2% increase (p<0.0001) in wild type (N2) worms and a 31.0% increase (p<0.0001) in daf-2 mutant worms whereas sDR did not increase the lifespan of aak-2 mutant worms. The mean and SEM of two independent experiments conducted in triplicates of 30 worms each is presented in Table S3B. (D) sDR does not cause DAF-16 nuclear localization. DAF16:GFP expressing worms were placed on ad libitum or sDR diet for 6 hours at day 7 of life.

Figure 4. FOXO/DAF-16 is necessary for AMPK induced increase in stress resistance and lifespan

(A) Constitutively active (CA) AMPK γ2 transgenic worms display an increased resistance to increasing concentrations of the reactive oxygen species generator Paraquat compared to control transgenic worms injected with an empty vector (EV) after 7 hours of incubation. Three independent lines of EV (1–3) and CA (1–3) were analyzed. (B) CA3 displayed an 86.0% increased resistance to 300 mM Paraquat over control EV1 (p<0.0001). The mean and SEM of two independent experiments conducted in sextuplicate is presented in Table S4A. (C) Stress resistance of CA AMPK γ2 transgenic worms is dependent on the presence of DAF-16. CA3 worms were fed bacteria expressing RNAi to daf-16 or control bacteria (ctl) and submitted to increased concentrations of Paraquat for 8 hours. (D) CA3 displayed a 48.9% increased stress resistance compared to EV1 (p=0.0005) but did not significantly increase stress resistance when treated with RNAi to daf-16. The mean and SEM of two independent experiments conducted in sextuplicate is presented in Table S4A. (E) CA AMPK γ2 transgenic worms (CA2 and CA3) displayed a statistically significant increase in median lifespan (9.5% and 17.7% respectively, p<0.001) compared to EV1 control worms. A comparison experiment done in triplicate is presented in Table S4B. (F) CA AMPK γ2 lifespan extension is DAF-16 dependent. CA AMPK γ2 transgenic worms (CA3) show a statistically significant increase in median lifespan (11%; p=0.0006) compared to control transgenic worms (EV1) but did not display a significantly lifespan extension compared to EV1 when treated with RNAi to DAF-16. Comparison of two independent experiments done in triplicate is presented in Table S4B.

Figure 5. AMPK activates FOXO/DAF-16 dependent transcription of sod-3

(A) sod-3 mRNA is increased in CA AMPK γ2 transgenic worms (CA3) compared with empty vector control transgenic line (EV1) and this increase is daf-16 dependent. The results presented correspond to the mean and SEM of three independent quantitative RT-PCR experiments. *: p<0.05; **: p<0.01; ***: p<0.001 in a one-way ANOVA statistical test. (B) sod-3 was decreased in aak-2 mutant worms compared with N2 wild type worms. The results presented correspond to the mean and SEM of two independent experiments: *: p<0.05 in a paired t-test.

Figure 6. AMPK directly phosphorylates FOXO/DAF-16 in vitro

(A) AMPK phosphorylates DAF-16 in an AMP-dependent manner. Results with human AMPK are presented but both human and worm AMPK phosphorylate DAF-16 in an AMP-dependent manner. (B) Stoichiometry of phosphorylation of DAF-16 by human AMPK. The graph represents the mean and standard deviation of normalized molar units of phosphate incorporated per mole of substrate of 3 independent experiments. Similar results were obtained with the worm AMPK. (C) Schematic of DAF-16 showing the location of the residues of DAF-16 phosphorylated by AMPK in vitro. DBD: DNA binding domain. Alignment of the AMPK consensus phosphorylation motif with phosphorylation sites in DAF-16. ϕ: hydrophobic residues; β: basic residues. The sequence in parenthesis represents an amphipathic helix. The sites in italics were identified as being phosphorylated in vitro by AMPK by tandem mass spectrometry. All the sites listed here were confirmed as being phosphorylated in vitro by AMPK by site-directed mutagenesis. (D) DAF-16 mutants display a significant reduction in phosphorylation by AMPK compared to WT DAF-16 in an in vitro kinase assay. 3A: T166A/S314A/S466A mutant; 6A: T166A/S202A/S314A/S321A/T463A/S466A.

Figure 7. Proposed model for the molecular networks integrating different DR methods

Different methods of dietary restriction may activate distinct signaling pathways or activate them to varying degrees. The table below the figure summarizes which pathways have been tested. Orange question marks indicate conflicting reports in the literature [41, 43, 50] and a red question mark indicates yet unconfirmed interactions in C. elegans. sDR: solid plate DR; bDR: liquid DR described in [28, 34]; lDR: liquid DR described in [27]; DD: dietary deprivation [24, 25]; axenic: absence of bacteria in liquid culture [26]; peptone: dilution of peptone in plates [36]; CDLM: chemically defined liquid media [35]. D: dependency; I: independency; NT: not tested. *: the daf-2 mutation is not a null mutation, which could affect the interpretation of the results. clk-1: mutant of a gene necessary for ubiquinone biosynthesis [29].