The flavonoid 4,4'-dimethoxychalcone promotes autophagy-dependent longevity across species

Abstract

Ageing constitutes the most important risk factor for all major chronic ailments, including malignant, cardiovascular and neurodegenerative diseases. However, behavioural and pharmacological interventions with feasible potential to promote health upon ageing remain rare. Here we report the identification of the flavonoid 4,4'-dimethoxychalcone (DMC) as a natural compound with anti-ageing properties. External DMC administration extends the lifespan of yeast, worms and flies, decelerates senescence of human cell cultures, and protects mice from prolonged myocardial ischaemia. Concomitantly, DMC induces autophagy, which is essential for its cytoprotective effects from yeast to mice. This pro-autophagic response induces a conserved systemic change in metabolism, operates independently of TORC1 signalling and depends on specific GATA transcription factors. Notably, we identify DMC in the plant Angelica keiskei koidzumi, to which longevity- and health-promoting effects are ascribed in Asian traditional medicine. In summary, we have identified and mechanistically characterised the conserved longevity-promoting effects of a natural anti-ageing drug.

Fig. 1

4,4′-dimethoxychalcone promotes longevity in yeast, nematodes, flies and human cells. a Screening procedure for anti-ageing flavonoids in a yeast chronological ageing model. b, c Z-scores of AUCs obtained for each flavonoid during the yeast screen and for each assay performed: PI staining, outgrowth capacity (b), and DHE to ethidium conversion (c); data obtained in 1–2 independent runs with three replicates each. Each data point represents one flavonoid. The data point and the structure of 4,4′-dimethoxychalcone (DMC) are shown in red. d–f Phosphatidylserine externalisation and membrane dysintegrity (d), ROS production (e) and survival (f) of DMC-treated yeast cells (100 µM) at indicated time points of chronological ageing using AnnexinV/PI costaining (d), ***P < 0.001, n = 18, DHE to ethidium conversion, n = 6 (e) and clonogenicity, n = 8 (f) independent biological replicates; P-values represent T, treatment; A, age; T ×A, interaction. g, h Survival of C. elegans (g) or D. melanogaster (h) during ageing with supplementation of food with DMC (41.6 µM for worms, 200 µM for flies). For other ageing replicates, see Supplementary Figure 2 (yeast) and 3 (nematodes, flies). i, j Replicative viability of DMC-treated (50 µM) U2OS cells. Representative images (i) and quantification (j) are shown, **P = 0.0069, n = 11 independent biological replicates. Data in (d) represent means ± SEM; box plots in (j) represent IQR (line at median) and whiskers 10–90 percentile. Significance in (d, j) was assessed by two-sided Student’s t-test, in (e, f) by repeated measures two-way ANOVA. Source data for (b–f, j) are provided as a Source Data file

Fig. 2

4,4′-dimethoxychalcone induces autophagy across species. a–d Autophagic flux in DMC-treated (100 µM) yeast indicated by the vacuolar accumulation of GFP-Atg8 (green); a representative micrograph is shown in (a), where propidium iodide (PI) counterstaining served to visualise dead cells (magenta). Corresponding quantification was performed by analysing the free GFP/GAPDH ratio (b). **P = 0.0024, n = 19 (Ctrl.), 18 (DMC) independent biological replicates; c representative immunoblot. Autophagy measured via alkaline phosphatase (ALP) activity of Pho8ΔN60 strains (d). *P = 0.0020, n = 11 independent biological replicates. See Supplementary Figure 5d for other time points. e, f Quantification (e) and representative pictures (f) of mCherry-foci in the intestine cells of DMC-fed (41.6 µM) young adult worms expressing Pnhx-2::mCherry::lgg-1. ***P < 0.0001, n = 61 (Ctrl.), 65 (DMC) animals. Scale bar: 50 µm. g, h Immunofluorescence analysis (g) and representative pictures (h) of ref. P-marked protein aggregates in fly brains after 30 days DMC feeding (200 µM). **P = 0.0026, n = 12 (Ctrl.), 11 (DMC) animals. Scale bar: 50 μm. Significance in (b, e, g) was assessed by two-sided Student’s t-tests, in (d) by ANOVA/Bonferroni. i, j Videomicroscopic analysis (i) and representative pictures (j) of DMC-treated (50 µM) U2OS cells expressing GFP-LC3 with or without chloroquine, cell nuclei were stained with Hoechst 33342. Comparisons by two-way ANOVA (T, treatment; C, chloroquine; T × C, interaction) followed by Bonferroni-corrected simple main effects, ***P < 0.0001, n = 22 (Ctrl.-Chlq), 21 (DMC-Chlq), 18 (Ctrl. + Chlq), 30 (DMC + Chlq) independent biological replicates. Scale bar: 10 μm. Box plots represent IQR (line at median) and whiskers 10–90 percentile. Source data for (b, d, e, g, i) are provided as a Source Data file

Fig. 3

4,4′-dimethoxychalcone promotes autophagy and cardioprotection in mice. a, b Autophagy induction in mouse heart tissue (a) determined by LC3 lipidation with leupeptin or vehicle injection after intraperitoneal injection of DMC (100 mg/kg) or DMSO (Ctrl.). n = 9 (Ctrl. Vehicle), 10 (DMC Vehicle), 3 (Leup.) animals, *P = 0.0105, **P = 0.0028; b representative immunoblot. c, d Infarction area per area at risk (AAR) (c) and representative images of left ventricular myocardial sections (scale bar: 1 mm) (d) after DMC treatment as in (a–b) followed by 3 h prolonged ischaemia in wild type (WT) and cardiac-specific Atg7 knockout mice (Atg7cKO). *P = 0.0134, n = 5 (Ctrl.), 6 (DMC), 3 (Atg7cKO) animals. Significance in (a, c) was assessed by two-sided Student’s t-tests between Ctrl. and DMC. Box plots represent IQR (line at median) and whiskers 10–90 percentile. Source data for (a, c) are provided as a Source Data file

Fig. 4

Autophagy induction is essential for 4,4’-dimethoxychalcone-mediated protection. a Survival of DMC-treated (100 µM) yeast wildtype (WT), ATG5-deficient, and ATG7-deficient mutant strains at day 5 of chronological ageing measured by PI staining. Data are normalised to the WT Ctrl. at day 1. Comparisons by two-way ANOVA with treatment and strain as independent variables, followed by Bonferroni-corrected simple main effects. ***P < 0.001, n = 6 independent biological replicates. Box plots represent IQR (line at median) and whiskers 10–90 percentile. Refer to Supplementary Figure 7 for other time points and other replicate experiments. Source data for (a) are provided as a Source Data file. b, c Survival of DMC-fed (41.6 µM) control and autophagy-deficient Atg5 RNAi nematodes (b) or female wildtype (WT) and Atg7-deficient (Atg7^−/−) mutant flies (c) during ageing. Refer to Supplementary Figure 7 for other replicate experiments. P-values represent pairwise comparisons (Ctrl. vs. DMC) by log-rank analysis

Fig. 5

Gln3 is a functional target of 4,4’-dimethoxychalcone. a DMC-rescuing effect determined as the AUC of PI-positive cells (day 1–5) upon DMC treatment (100 µM) of yeast deletion mutants involved in autophagic signalling, normalised to the rescuing effect in wildtype (WT) cells. Hits below the threshold (0.75) are depicted in blue. Comparisons by ANOVA/Bonferroni. **P < 0.01, ***P < 0.001, n = 6 independent biological replicates. b Cell death at day 3 of chronological ageing of DMC-treated (100 µM) yeast wildtype (WT) and GLN3-deficient mutant strains determined by PI staining normalised to Ctrl. at day 1 (b), ***P < 0.001, n = 6 independent biological replicates. (c, d) Autophagy induction of DMC-treated (100 µM) yeast wildtype (WT) and GLN3-deficient mutant strains indicated by the free GFP to GAPDH ratio normalised to WT Ctrl. (d), *P = 0.0242, n = 5 independent biological replicates; representative immunoblot in (c). e, f Changes in intracellular amino acid concentrations (e) and principal component analysis of yeast metabolites (f) 24 h after DMC-treatment (100 µM) in wildtype (WT) and GLN3-deficient yeast cells. g Gln3-dependent MEP2 expression using a P_MEP2-LacZ reporter at day 3 after DMC-treatment (100 µM) in wildtype (WT) and GLN3-deficient yeast cells. **P = 0.0056, n = 12 (WT) n = 9 (Δgln3) independent biological replicates; Comparisons in (b, d, g) by two-way ANOVA (T, treatment; S, strain; T × S, interaction) followed by Bonferroni-corrected simple main effects. Box plots represent IQR (line at median) and whiskers 10–90 percentile. Source data for (a, b, d–g) are provided as a Source Data file

Fig. 6

Rapamycin and 4,4′-dimethoxychalcone follow independent routes of cytoprotection. a–c Cell death assessed by PI staining and flow cytometry normalised to WT Ctrl. (a) and autophagy induction indicated by the free GFP to GAPDH ratio normalised to WT Ctrl. b, c of S. cerevisiae wildtype and GLN3-deficient mutant strains after treatment with 40 nM rapamycin (Rapa) at day 3 of chronological ageing. Comparisons in (a, b) by two-way ANOVA (T, treatment; S, strain; TxS, interaction) followed by Bonferroni-corrected simple main effects. *P < 0.0483, **P < 0.0055, ***P < 0.001, n = 4 (a), 5 (b) independent biological replicates. (d) Rps6 S232/S233 phosphorylation of S. cerevisiae wild-type cells after 6 h of DMC (100 µM) or 40 nM rapamycin (Rapa) treatment as determined by immunoblotting. Comparisons by ANOVA/Bonferroni. ***P < 0.001, n = 3 independent biological replicates. Box plots represent IQR (line at median) and whiskers 10–90 percentile. Source data for (a, b, d) are provided as a Source Data file

Fig. 7

GATA transcription factors are phylogenetically conserved effectors of 4,4′-dimethoxychalcone. a Survival of DMC-fed (41.6 µM) control and GATA transcription factor-deficient elt-1 RNAi nematodes. P-values represent pairwise comparisons by log-rank analysis. Refer to Supplementary Figure 12 for other replicate experiments. b, c Representative fluorescence pictures (b) and quantification (c) of mCherry-positive foci in the intestine cells of DMC-fed (41.6 µM) or untreated control or GATA-factor-deficient elt-1 (RNAi) nematodes expressing P_nhx-2::mCherry::lgg-1 reflective of autophagosome generation. Comparisons by two-way ANOVA (T, treatment; S, strain; TxS, interaction) followed by Bonferroni-corrected simple main effects. ***P < 0.001, n = 44 (WT Ctrl.), 53 (WT DMC), 57 (elt-1 Ctrl.), 51 (elt-1 DMC) animals. Scale bar: 50 µm. Box plots represent IQR (line at median) and whiskers 10–90 percentile. d, e Autophagy induction in DMC-treated (50 µM) human U2OS cells with siRNA against an unrelated sequence (UNR) or GATA transcription factors as determined via videomicroscopy of cells expressing GFP-LC3 (d). Cell nuclei were stained with Hoechst 33342. Data shown as means of three different siRNA constructs ± SEM. Comparisons by ANOVA with Tukey correction. *P < 0.05, **P < 0.001, ***P < 0.001, n = 4 (UNR), 3 (others) independent biological replicates. Representative images are shown in (e). Scale bar: 10 µm. Source data for (c, d) are provided as a Source Data file