Suppression of autophagic activity by Rubicon is a signature of aging

Abstract

Autophagy, an evolutionarily conserved cytoplasmic degradation system, has been implicated as a convergent mechanism in various longevity pathways. Autophagic activity decreases with age in several organisms, but the underlying mechanism is unclear. Here, we show that the expression of Rubicon, a negative regulator of autophagy, increases in aged worm, fly and mouse tissues at transcript and/or protein levels, suggesting that an age-dependent increase in Rubicon impairs autophagy over time, and thereby curtails animal healthspan. Consistent with this idea, knockdown of Rubicon extends worm and fly lifespan and ameliorates several age-associated phenotypes. Tissue-specific experiments reveal that Rubicon knockdown in neurons has the greatest effect on lifespan. Rubicon knockout mice exhibits reductions in interstitial fibrosis in kidney and reduced α-synuclein accumulation in the brain. Rubicon is suppressed in several long-lived worms and calorie restricted mice. Taken together, our results suggest that suppression of autophagic activity by Rubicon is one of signatures of aging.

Fig. 1

C. elegans Rubicon regulates lifespan via modulating autophagy. a Autophagosomes labelled with GFP::LGG-1 (open arrowheads) were more abundant in rub-1 knockdown, but concomitant knockdown of bec-1/Beclin 1 abolished the increment. Anterior intestines of GFP::LGG-1 transgenic worms at L4 stage were shown. Each RNAi is indicated. Knockdown was conducted from egg onward. b Quantification of GFP::LGG-1 puncta in anterior intestines under each knockdown condition. Values represent means ± s.e.m. (n = 3). P value (**P < 0.01) was determined by one-way ANOVA with Tukey’s test. c No. of GFP::LGG-1 is more increased by rub-1 knockdown 2 h after BafA injection compared to control knockdown, indicating the autophagic flux is increased by rub-1 knockdown. Values represent means ± s.e.m. (n = 3). P value (*P < 0.05, **P < 0.01) was determined by one-way ANOVA with Tukey’s test. The representative picture of DMSO or BafA injected GFP::LGG-1 worms were shown in Supplementary Fig. 2e. d Representative pictures of transgenic worm expressing RUB-1::EGFP showing that the fluorescence in anterior region including neuronal cells (brackets) were increased at day 5 compared to day 1. All images were taken together with N2 showing autofluorescence at same exposure times. e qRT-PCR analysis showing rub-1 expression at day 1, 3, 5 and 7 in wild-type worms, indicating that rub-1 is upregulated from day3 onward. Mean ± s.e.m. from three independent experiments are depicted and are normalised to wild-type N2 day1 samples. P value (*P < 0.05, **P < 0.01, ***P < 0.001) was determined by one-way ANOVA with Tukey’s test. f qRT-PCR analysis showing rub-1 expression in several long-lived strains including daf-2(e1370), eat-2(ad465), and glp-1(e2141) at day1 adult stage. Mean ± s.e.m. from three independent experiments are depicted and are normalised to wild-type N2 day1 samples. Germline-less phenotype of glp-1 was induced by exposure to elevated temperature (25 °C) for 2 days. P value (*P < 0.05, **P < 0.01) was determined by one-way ANOVA with Tukey’s test. g Knockdown of rub-1 extends wild-type lifespan. Longevity conferred by rub-1 was abolished by concomitant knockdown of bec-1/Beclin1. Knockdown was conducted from egg onward. Log-rank test was conducted for statistical analysis. See Supplementary Data 1 for details and repeats. Scale bars, 20 μm (a); 50 μm (d)

Fig. 2

Rubicon knockdown ameliorates age-dependent phenotypes in worms. a Transgenic worms expressing polyQ35::YFP in body wall muscle exhibited age-dependent accumulation of polyQ inclusions (open arrowheads) at day5 stage. rub-1 knockdown decreased the accumulation of inclusions, and concomitant knockdown of bec-1 reverted the phenotype. Each RNAi is indicated. Knockdown was conducted from egg onward. b Number of polyQ inclusions per worm in a. More than 20 worms were analysed in each experiment, and the experiments were repeated three times. Values represent means ± s.e.m. (n = 3). P value (*P < 0.05, **P < 0.01) was determined by t-test. c Locomotion speed, calculated from multi-worm tracking analysis at day5. rub-1 knockdown worms maintain high locomotion activity. Knockdown was conducted from egg onward. Values represent means ± s.e.m. (control, n = 12; rub-1, n = 8; atg-18, n = 8; rub-1 + atg-18, n = 13). P value (*P < 0.05) was determined by t-test. The representative picture of the tracking is shown in supplementary fig. 6a. d rub-1 knockdown increased the oxidative stress resistance in an autophagy dependent manner. Survivorship of wild-type day 1 worms subjected to indicated RNAi after 4.4 mM H₂O₂ treatment for 2 h. Knockdown was conducted from egg onward. Values represent means ± s.e.m. (n = 3). P value (**P < 0.01) was determined by t-test. Scale bars 50 μm (a)

Fig. 3

Tissue specific roles of Rubicon contributing to lifespan regulation in worms. a, b Neuronal knockdown of rub-1 extended lifespan efficiently. The longevity by rub-1 knockdown was abolished by concomitant atg-18 knockdown. TU3401(neuron specific RNAi sensitive strain) was used and knockdown was conducted from egg onward. See Supplementary Data 1 for details and repeats. c Neuron specific knockdown of rub-1 significantly increased no. of GFP::LGG-1 2 h after BafA injection compared to DMSO injection in neuron, while control knockdown did not, indicating autophagy flux was increased by rub-1 knockdown in neuron. Newly generated strains by crossing between MAH215 and TU3401 were used and images were taken at day 1 adult stage. Knockdown was conducted from egg onward. Arrowheads indicate GFP::LGG-1 puncta. d Quantification of GFP::LGG-1 puncta in neuron from c. Values represent means ± s.e.m. (n = 3). P value (*P < 0.05) was determined by t-test. Scale bars, 20 μm (c)

Fig. 4

dRubicon knockdown extends lifespan and ameliorates polyQ protein-induced toxicity in female fly. a dRubicon was increased in middle-aged female fly. b A quantification of dRubicon levels normalised by Actin in a. Values represent means ± s.e.m. (n = 6, each time points). P value (*P < 0.05) was determined by One-way ANOVA with Tukey’s test. c The autophagy assay in Kenyon cell layers of brains showing that dRubicon knockdown increased autophagosomes (AP, open white arrowheads) and autolysosomes (AL, open magenta arrowheads). d Quantification of mCherry-GFP-Atg8 puncta in c. The values represent the mean ± s.e.m. P values (**P < 0.01) were determined by t-test (n = 6 for control, n = 8 for dRub-IR). e Compound eye degeneration was attenuated by dRubicon knockdown in flies expressing the expanded polyQ protein MJDtrQ78s (GMR-GAL4_MJDtrQ78s flies) compared to control. Open arrowheads indicate necrotic patches. f A quantification of the numbers of necrotic patches in GMR-GAL4_MJDtr-Q78s fly eyes in e. More than 20 eyes were analysed for each genotype. P value (***P < 0.001) was determined by t-test. g Neuronal knockdown of dRubicon extends lifespan in female fly. Transgene expression of dRubicon-IR was induced by the pan-neuronal elav-GAL4 driver. See Supplementary Data 1 for details and repeats. h Neuronal knockdown of dRubicon decreased polyQ (MJDtrQ78w) inclusions in Kenyon cell layer of female fly brains. i Quantification of the polyQ protein intensity in h. 10 bilateral layers of Kenyon cell layers from 5 animals were analysed. P value (***P < 0.001) was determined by t-test. j dRubicon knockdown in neurons ameliorated the locomotor dysfunction by polyQ expression in female fly. Locomotor function was evaluated by the climbing assay. Data represents mean ± s.e.m. P values (*P < 0.05, **P < 0.01, ***P < 0.0001) were determined by two-way ANOVA with Tukey’s test (elav_MJDtrQ78w vs. elav_MJDtrQ78w/Rubicon-IR). MJDtrQ78w expression level did not change between elav_MJDtrQ78w and elav_MJDtrQ78w/Rubicon-IR (Supplementary Fig. 8g). Scale bars, 5 μm (c); 100 μm (e); 10 μm (h)

Fig. 5

Knockout of Rubicon ameliorates age-associated phenotypes in mice. a Rubicon protein level in kidney was elevated in 20-month-old mice relative to 2-month-old mice. Western Blot from WT mice and Rubicon KO (KO) mice shows the specific bands of Rubicon. b A quantification of Western Blot shown in a. Value represent means ± s.e.m (2 months, n = 3; 20 months, n = 3). P value (*P < 0.05) was determined by t-test. c Western blotting samples showing LC-3 and p62 in mouse kidney at 20 months of age. d Immunohistochemical images of 20-month-old kidney showing that collagen I–positive area was reduced in Rubicon-knockout (Rubicon KO) kidney relative to WT. P value (*P < 0.05) was determined by t-test. e A quantification of collagen I–positive area in d. Values represent means ± s.e.m (WT, n = 5; Rubicon KO, n = 6). P value (*P < 0.05) was determined by t-test. f qRT-PCR analysis of several fibrosis markers (TGF-b1, Col1a1) in 20-month-old kidney. Values represent means ± s.e.m (WT, n = 5; Rubicon KO, n = 6). P value (*P < 0.05) was determined by t-test. g Lewy body and Lewy neurite–like inclusions containing phosphorylated α-Syn (green punctae, indicated by open arrowheads) were less abundant in neuronal Rubicon-knockout (Rubicon^flox/flox: Nestin-Cre) mice than in controls (Rubicon^flox/+: Nestin-Cre), 10 months after the injection. h Percentages of mice having small (<100), moderate (100–200), or large (>200) numbers of phosphorylated α-Syn positive signals. P-value (*P < 0.05) was determined by χ² test (Rubicon^flox/+: Nestin-Cre, n = 7; Rubicon^flox/flox: Nestin-Cre, n = 6). Scale bars, 100 μm (d); 500 μm (g)