A secreted microRNA disrupts autophagy in distinct tissues of Caenorhabditis elegans upon ageing

Abstract

Macroautophagy, a key player in protein quality control, is proposed to be systematically impaired in distinct tissues and causes coordinated disruption of protein homeostasis and ageing throughout the body. Although tissue-specific changes in autophagy and ageing have been extensively explored, the mechanism underlying the inter-tissue regulation of autophagy with ageing is poorly understood. Here, we show that a secreted microRNA, mir-83/miR-29, controls the age-related decrease in macroautophagy across tissues in Caenorhabditis elegans. Upregulated in the intestine by hsf-1/HSF1 with age, mir-83 is transported across tissues potentially via extracellular vesicles and disrupts macroautophagy by suppressing CUP-5/MCOLN, a vital autophagy regulator, autonomously in the intestine as well as non-autonomously in body wall muscle. Mutating mir-83 thereby enhances macroautophagy in different tissues, promoting protein homeostasis and longevity. These findings thus identify a microRNA-based mechanism to coordinate the decreasing macroautophagy in various tissues with age.

Fig. 1

Ageing upregulates mir-83 in the intestine through hsf-1. a qRT-PCR of miR-83 in the wild type (WT) worms of indicated ages. The levels of miR-83 were normalized against the one at day 1. n = 7, 4, 3, 3, and 3 independent experiments for samples of day 1, 4, 7, 14, and 21, respectively. b Immunoblots of mir-83p::GFP at indicated ages. Blots against α-tubulin and day 1 sample serve as controls for loading and normalization respectively. n = 3 independent experiments. c Representative images of mir-83p::gfp transgenic animals at indicated ages. n: neuron, i: intestine. d GFP intensity of mir-83p::gfp in the intestine and head neurons at day 1, day 4, and day 7 of adulthood. Samples of day 1 serve as controls for normalization. n = 3 independent experiments containing at least 41 worms. e qRT-PCR of miR-83 in WT worms treated with luc2 or hsf-1 RNAi. RNAi treatment was performed from hatching and worms were examined at indicated ages. Samples of day 1 serve as controls for normalization. n = 3 independent experiments. f Expression of mir-83p::gfp under luc2 or hsf-1 RNAi. Worms were treated with RNAi from hatching and examined for GFP intensity at indicated ages. day 1 worms subjected to luc2 RNAi serve as control for normalization. n = 3 independent experiments containing at least 39 worms. Scale bars: 100 μm. Statistical significance was calculated by One-way ANOVA in (a and b), or Two-way ANOVA in d–f. ns: non-significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Source data are provided as a Source Data file

Fig. 2

mir-83 inhibits cup-5 by binding to its 3′-UTR. a cel-mir-83-3p mimic (mir-83) but not a control oligo (Ctrl) suppresses a luciferase reporter with cup-5 3′-UTR (WT) in HEK293T cells. Ctrl transfected cells serve as controls for normalization. Mutating the mir-83 binding site in cup-5 3′-UTR (Mut) blocked this interaction. n = 3 independent experiments. b A depiction of the dual fluorescence reporter to test the interaction of mir-83 and its targets in vivo. c Fluorescent signals and immunoblots of the dual fluorescence reporter of cup-5 3′-UTR in WT worms and mir-83(-) mutants at day 1 of adulthood. Quantification is from the western blots. GFP blots and WT worms serve as controls for loading and normalization respectively. Scale bar: 100 μm. n = 3 independent experiments. d Fluorescent signals of the cup-5 translational reporter (mCherry::CUP-5) in the posterior intestine (dashed lines) and nearby BWM (yellow lines) of WT worms and mir-83(-) mutants at indicated ages. WT signals at day 1 serve as controls for normalization. Scale bar: 10 μm. n = 3 independent experiments. e The protein level of the endogenously tagged 3xFLAG::WrmScarlet::CUP-5 is increased upon mutating mir-83 at indicated ages. α-tubulin and WT blots at day 1 serve as controls for loading and normalization respectively. n = 5 independent experiments. Statistical significance was calculated by unpaired t-test in a, c, e, or two-way ANOVA in d. ns: non-significant, *p < 0.05, **p < 0.01, ***p < 0.001. Source data are provided as a Source Data file

Fig. 3

mir-83 dysregulates autophagy in the intestine through cup-5. a, b GFP::LGG-1 puncta (arrow heads) in the posterior intestine cells (dashed lines) of the WT worms and mir-83(-) mutants at indicated ages. Compared to that of day 1 samples, the exposure of day 5 samples was increased to visualize GFP::LGG-1 puncta. Worms were treated with 5 mM chloroquine (CQ) or a mock control b. n = 3 independent experiments containing at least 22 worms in a and 25 worms in b. c The fluorescent signals of mCherry-GFP::LGG-1 in the posterior intestine cells (dashed lines) of WT worms and mir-83(-) mutants at indicated ages. White arrow heads denote the autolysosome (AL) puncta with only mCherry signals. Yellow arrows denote the autophagosome (AP) puncta with both GFP and mCherry signals. n = 3 independent experiments containing at least 19 worms. d APs (yellow arrows) and ALs (white arrow heads) in the posterior intestine cells (dashed lines) of WT worms and mir-83(-) mutants. Animals were treated with intestine specific RNAi against luc2 or cup-5 from hatching and examined at day 1 of adulthood. n = 3 independent experiments containing at least 16 worms. e The number of SQST-1::GFP puncta (arrow heads) is modestly reduced in the head neurons (dashed lines) of mir-83(-) mutants at the fourth larval stage (L4). n = 3 independent experiments containing at least 58 worms. f The dFP::LGG-1 reporter (mFP/dFP, mFP vs dFP::LGG-1) in neurons responds to the treatment of 5 mM CQ but not mir-83 mutation at day 1 of adulthood. WT samples with control treatment serve as controls for normalization. n = 3 independent experiments. g–h APs (yellow arrows) and ALs (white arrow heads) in the head neurons surrounding the posterior pharynx bulb (dashed lines). Worms of indicated genotypes were examined at day 1 and day 5 of adulthood. A CQ treatment (5 mM) was performed as indicated. n = 3 independent experiments containing at least 15 worms. Scale bars: 10 μm. Statistical significance was calculated by Poisson regression in a–d and g–h, unpaired t-test in e, or one-way ANOVA in f. ns: non-significant, *p < 0.05, ***p < 0.001. Source data are provided as a Source Data file

Fig. 4

mir-83 impairs autophagy in the body wall muscle with ageing. a GFP::LGG-1 puncta (arrow heads) in the body wall muscle (BWM) (dashed lines) of WT worms and mir-83(-) mutants at indicated ages. n = 3 independent experiments containing at least 25 worms. b mCherry::GFP::LGG-1 signals in the BWM (dashed lines) of WT worms and mir-83(-) mutants at indicated ages. White arrow heads denote the AL puncta and yellow arrows denote the AP puncta. n = 3 independent experiments containing at least 29 worms. c GFP::LGG-1 puncta (arrow heads) in the BWM (dashed lines) of indicated strains at day 1 of adulthood post the treatment of 5 mM chloroquine (CQ) or a mock treatment (Ctrl). n = 3 independent experiments containing at least 28 worms. d SQST-1::GFP puncta (arrow heads) in the BWM (dashed lines) of indicated strains. n = 5, 5, 4, and 4 independent experiments for samples of day 1, 3, 5, and 7, respectively. Each sample contains at least 106 worms. e The number of GFP::LGG-1 puncta (arrow heads) in WT worms and mir-83(-) mutants with or without an intestine-specific transgene of mir-83 (IntOE). Dashed lines denote BWM cells. n = 3 independent experiments containing at least 23 worms. f The intestine-specific expression of mir-83 abolishes the increased ALs (white arrow heads) in BWM cells (dashed lines) of mir-83(-) mutants. Note that APs (yellow arrows) were increased in WT worms due to a block of the autophagy flux. n = 4 independent experiments containing at least 19 worms. Scale bars: 10 μm. Statistical significance was calculated by Poisson regression. ns: non-significant, *p < 0.05, **p < 0.01, ***p < 0.001. Source data are provided as a Source Data file

Fig. 5

Intestinal mir-83 inhibits cup-5 in the body wall muscle. a SQST-1::GFP puncta (arrow heads) in the BWM cells (dashed lines) of indicated strains treated with BWM specific RNAi against luc2 or cup-5 from hatching. The GFP signals in nucleus were from the transgene enabling BWM specific RNAi. Worms were examined at day 1 of adulthood. n = 3 independent experiments containing at least 41 worms. b A diagram depicting an integrated transgene of a BWM-specific reporter of cup-5 3′-UTR. c Representative images of the BWM-specific cup-5 3′-UTR reporter in the indicated strains at day 5 of adulthood. IntOE: the intestine-specific overexpression of mir-83. Dashed lines denote myocytes. d The intestine-specific overexpression of mir-83 (IntOE) abolishes the upregulation of the BWM-specific cup-5 3′-UTR reporter in mir-83(-) mutants. Blots against α-tubulin and WT worms at day 1 serve as controls for loading and normalization respectively. n = 3 independent experiments. Scale bars: 10 μm. Statistical significance was calculated by Poisson regression in a, or unpaired t-test in d. ns: non-significant, *p < 0.05, **p < 0.01, ***p < 0.001. Source data are provided as a Source Data file

Fig. 6

mir-83 is transported from the intestine to the body wall muscle potentially through the extracellular vesicles in the pseudocoelomic fluidics. a The workflow of the single cell qPCR from the isolated BWM cells. Scale bars: 50 μm. b qRT-PCR of miR-83 and U18 in the isolated BWM cells and whole worms of indicated strains. Cells were collected from worms at day 1 of adulthood. Note that miR-83 is detected in the BWM cells of WT worms and the mir-83(-) mutants with an intestine specific transgene expressing mir-83 (mir-83(-);IntOE). Note that Ct anti-correlates with the expression level. Samples without qRT-PCR signal were set as Ct 40. ND: not detected. n = 5, 4, and 3 independent experiments for WT, mir-83(-), and mir-83(-);IntOE, respectively. c miR-83 is detected in the isolated coelomocytes by single cell qRT-PCR. Cells were isolated from worms at day 1 of adulthood. n = 3 independent experiments. d mir-83 mutation reduces the increased size of LMP-1::GFP-positive vacuoles in the aged coelomocytes. The largest vacuoles (asterisks) were measured. Scale bar: 10 μm. n = 3 independent experiments containing at least 26 worms. e The size of the largest LMP-1::GFP-positive vacuoles (asterisks) in the aged coelomocytes of WT worms and mir-83(-) mutants upon indicated RNAi treatments. Scale bar: 10 μm. n = 4 independent experiments containing at least 60 worms. f qRT-PCR of indicated genes in the purified extracellular vesicles (EVs) and worms of the indicated strains. n = 3 independent experiments. Statistical significance was calculated by Two-way ANOVA. ns: non-significant, *p < 0.05, **p < 0.01, ***p < 0.001. Source data are provided as a Source Data file

Fig. 7

mir-83 modulates ageing through cup-5 and autophagy. a, b PolyQ-YFP aggregates (arrow heads) in the intestine (a) and BWM (b) of WT worms and mir-83(-) mutants at indicated ages. n = 4 independent experiments containing at least 79 worms. c Survival curves of indicated strains. d, e RNAi against cup-5 increases the polyQ aggregates in the intestine (d) and BWM (e) of mir-83(-) mutants to WT levels. n = 3 independent experiments containing at least 44 worms. f cup-5 RNAi abolishes the extended lifespan of mir-83(-) mutants. g Blocking autophagy with 5 mM chloroquine (CQ) abolishes the extended lifespan of mir-83(-) mutants. h Ageing assays of the indicated strains. IntOE: an intestine-specific transgene expressing mir-83. i mir-83 is upregulated in the aged intestine by increased HSF-1 and transported across tissues to disrupt autophagy through inhibiting CUP-5. See Discussion for details. Scale bar: 100 μm. Statistical significance was calculated by unpaired t-test in a, b or Two-way ANOVA in d, e. ns: non-significant, *p < 0.05, **p < 0.01, ***p < 0.001. See Supplementary Data 1 for lifespan statistics (c and f–h). Source data are provided as a Source Data file