A steroid receptor-microRNA switch that regulates lifespan in response to signals from the gonad

Abstract

While the gonad primarily functions in procreation, it also affects animal lifespan. Here we show that removal of the C. elegans germline triggers a switch in the regulatory state of the organism to promote longevity, co-opting components involved in larval developmental timing circuits. These include the DAF-12 steroid receptor, which upregulates members of the let-7 microRNA family, involved in L2/L3 transitions. The microRNAs target an early larval nuclear factor, lin-14, and akt-1/kinase, thereby stimulating DAF-16/FOXO signaling to extend life. These studies suggest that metazoan lifespan is coupled to the gonad through elements of a developmental timer.

When germline stem cells (GSCs) are eliminated from the C. elegans gonad, animals live up to 60% longer (1, 2). This longevity represents more than a simple tradeoff with fertility as it depends on the presence of the somatic gonad, suggesting a model whereby GSCs antagonize life-lengthening signals from the somatic gonad. Importantly, work from D. melanogaster suggests that gonadal longevity is evolutionarily conserved, since flies lacking GSCs live longer than controls (3). Although the life-shortening signals from the germline remain elusive, life-lengthening signals from the somatic gonad include bile acid-like steroids, called dafachronic acids (DAs), which activate the steroid hormone receptor DAF-12, a homolog of vertebrate liver-X, farnesoid-X, and vitamin-D-receptors (4, 5). How the DAs themselves are regulated and activate downstream targets remains unclear. Evidence indicates that DA/DAF-12 signaling regulates genes important for longevity, and also converges on the DAF-16/FOXO transcription factor, by potentiating nuclear localization and augmenting transcriptional activity on longevity promoting genes (6, 7), yet the mechanisms coupling these pathways are unknown. DAF-16/FOXO is also stimulated independently by decreased insulin/IGF receptor (IR) signaling, since the longevity of daf-2/IR and germlineless glp-1 mutants is additive (1).

To illuminate how germline loss stimulates longevity, we first asked whether it affects regulation of DA signaling. When we examined mRNA levels of DA signaling components by qPCR, no differences were observed between germlineless glp-1 mutants and gonad-intact wild-type animals (WT) at the third larval (L3) stage. However, by L4 and day 1 of adulthood (D1), the hormone biosynthetic gene daf-36/Rieske-oxygenase was significantly upregulated in glp-1 (Fig. 1A, Fig. S1D) but downregulated in WT (Fig. 1A). Other DA-biosynthetic genes including daf-9/CYP27A1 and dhs-16/HSD were less affected (8) (Fig. S1A). Expressed mainly in the intestine, daf-36 catalyzes the first step in Δ⁷-DA biosynthesis, converting cholesterol to 7-dehydrocholesterol (9, 10). Accordingly, 7-dehydrocholesterol and Δ⁷-DA, were increased 4-5-fold in glp-1 animals as measured by GC-MS-MS (Fig. 1B–C). In D1 adults, daf-36 upregulation was largely independent of daf-12 and daf-16 (Fig. S1B–C). These data suggest that a regulatory switch governs DA signaling in response to signals from the reproductive system, and reveal that germline loss stimulates the DA signaling pathway.

Fig. 1. Ablation of the germline upregulates DA/DAF-12 signaling.

(A) Timecourse of daf-36 mRNA level in WT and germlineless animals (glp-1).

Error bars: SEM.

(B–C) Δ⁷-DA (C) and 7-dehydrocholestrol (B) are elevated in germlineless animals. t-test, **P<0.01, *** P<0.001.

(D) Timecourse of mir-84 and mir-241 levels in WT and glp-1 animals.

(E) In glp-1 animals, mir-84p::gfp and mir-241p::gfp expression are upregulated in seam cells (arrowheads) and intestine (arrows) respectively.

(F–G) Germline absence results in upregulation of mir-84 and mir-241 in a DAF-12/DA -dependent manner. Strains were grown on plates supplemented with ethanol (EtOH) or Δ⁴-dafachronic acid (DA). ANOVA, vs glp-1 EtOH, ns: non-significant, * P<0.05, ** P<0.01, *** P<0.001.

To see if germline loss stimulates DAF-12 transcriptional activity, we focused on let-7-related microRNAs, mir-84 and mir-241, which are direct DAF-12 targets in larval developmental timing pathways during L2/L3 transitions (11, 12). Indeed, microRNA expression increased in glp-1 mutants by the L4 stage, and peaked at 3-4-fold by D1 (Fig. 1D–E). MicroRNA upregulation was DA and daf-12 dependent, whereas daf-16 or nhr-80, an HNF4-like nuclear receptor regulating gonadal longevity (13), had little effect (Fig. 1F–G, Fig. S2F). Consistently, mir-84p::gfp and mir-241p::gfp promoter constructs also exhibited transcriptional upregulation in glp-1 mutants, particularly in epidermal seam cells (mir-84), and intestine (mir-241), revealing complex tissue-specific regulation (Fig. 1E, Fig. S2E). Other members of this microRNA family, let-7 and mir-48, were also upregulated, but showed no clear DA/DAF-12 dependence. By contrast, an unrelated microRNA, mir-228, decreased during the same time frame (Fig. S2). We conclude that germline absence upregulates DA signaling, accompanied by transcriptional activation of let-7 family members including the DAF-12 target genes, mir-84 and mir-241.

To test if the microRNAs function in the gonadal longevity pathway, we removed GSCs by laser microsurgery in WT and mir-84;mir-241 mutants. As expected, lifespan was extended in germline-ablated WT compared to mock-ablated controls. Whereas gonad-intact mir-84;mir-241 controls resembled WT, lifespan extension was strikingly abolished in germline-ablated mir-84;mir-241 double mutants (Fig. 2A),. Similarly, microRNA loss suppressed longevity and stress resistance in glp-1 mutants (Fig. S3A–C, Table S1). By contrast, mir-84;mir-241 mutation had little effect on longevity caused by reduced mitochondrial function (cco-1RNAi) or IR signaling (daf-2RNAi) (Fig. 2B). mir-84 and mir-241 transgenes driven by endogenous promoters restored stress resistance and longevity in mir-84;mir-241;glp-1 triple mutants, but did not significantly extend lifespan in gonad-intact animals (Fig. S3B–C, Table S1). Thus, mir-84 and mir-241 are specifically required but not sufficient for life extension in the gonadal pathway. mir-48 mutants also significantly decreased glp-1 longevity, but affected WT as well (Table S1). let-7 mutants were not analyzed because of their severe developmental defects. Therefore we focused on mir-84 and mir-241 for further analysis.

Fig. 2. DAF-12 target microRNAs are required for gonadal longevity through DAF-16/FOXO.

(A) mir-84 and mir-241 are required for gonadal longevity. Laser microsurgery was performed to ablate the germline.

(B) mir-84 and mir-241 do not affect the longevity by reduced IR signaling (daf-2RNAi) or itochondrial function (cco-1RNAi).

(C) Lifespan analysis of indicated strains on daf-16RNAi.

(D) qRT-PCR of DAF-16/FOXO target genes (T08B1.1 and btb-14) in worms of indicated genotypes. ANOVA, * P<0.05, ** P<0.01, *** P<0.001.

DAF-16/FOXO is essential for longevity in the gonadal pathway. In germlineless animals, it accumulates in intestinal nuclei, where it regulates genes important for lifespan extension (1, 6). Both DAF-12 and DAF-36 promote DAF-16 nuclear localization (4, 6), and DAF-12 and DAF-16 share transcriptional responsibilities for longevity (6). To investigate whether the microRNAs interact with DAF-16, we analyzed lifespan upon daf-16RNAi (daf-16i). If microRNA deficiency reduces glp-1 longevity by a mechanism independent of daf-16, then mir-84;mir-241;glp-1;daf-16i animals should live even shorter than glp-1;daf-16i animals. Instead, mir-84;mir-241 deletion does not further reduce life span of glp-1 upon daf-16i, (Fig. 2C), suggesting that the microRNAs and daf-16 work in the same pathway. To test this hypothesis, we examined the effect of the microRNAs on DAF-16 localization and activity. mir-84;mir-241 deficiency modestly diminished DAF-16::GFP nuclear localization, with little effect on overall expression levels (Fig. S4A–C). Consistent with a role in regulating DAF-16 activity via DA/DAF-12 signaling, microRNA mutation significantly reduced expression levels of 6 genes commonly regulated by DAF-12 and DAF-16 (Fig. 2D, Fig. S5A–B) (7). mir-84;mir-241 also impacted other DAF-16 targets, including superoxide dismutase sod-3 and the lipase lipl-4 (Fig. S5B–C) required for gonadal longevity (14). On daf-16i, mir-84;mir-241 deletion did not further decrease daf-16/FOXO-regulated genes (Fig. S5D), suggesting that the microRNAs control these genes’ expression through daf-16/FOXO but not another transcription factor. By contrast, the microRNAs had little influence on the expression of daf-12-regulated genes, fard-1 and cdr-6 (Fig. S6A–C)(7), nor on the transcriptional output of other major regulators of gonadal longevity, including PHA-4/FOXA targets, lgg-1 and unc-51, involved in autophagy (15), and the NHR-80 target, fat-6, involved in fatty acid desaturation (13)(Fig. S6D–F). These results reveal that mir-84 and mir-241 specifically stimulate DAF-16 localization and transcriptional activity.

MicroRNAs downregulate gene expression by binding to the 3′ UTR of target mRNAs and reducing stability and translation. To explore targets of mir-84;mir-241 and how they stimulate DAF-16/FOXO activity, we searched for DAF-16/FOXO inhibitors with potential mir-84;mir-241 binding sites in their 3′ UTRs, using the bioinformatic algorithm mirWIP (16). Among predicted targets were components of IR signaling, including the PDK/AKT kinase cascade which inhibits DAF-16/FOXO (17).

Consistent with microRNA-mediated inhibition, mir-84 and mir-241 significantly repressed luciferase reporters containing the 3′ UTRs of akt-1 and pdk-1 in cell culture (Fig. S7A, G–H). In C. elegans, akt-1-3′UTR dual reporter (DR) constructs were upregulated in mir-84;mir-241 double mutants relative to WT, while an unc-1-3′UTR-DR control reporter remained unchanged (Fig. 3A–B, Fig. S7C), indicating an inhibition of the akt-1 3′UTR by mir-84 and mir-241 in vivo. Moreover, the akt-1-3′UTR-DR construct was downregulated 40–50% in glp-1 animals compared to WT (Fig. 3C–D). In contrast, a pdk-1-3′UTR-DR construct was not downregulated in glp-1 animals (Fig. 7I). These results indicate that mir-84;mir-241 downregulate akt-1 through its 3′UTR, in response to gonadal signals.

Fig. 3. akt-1 kinase is a microRNA target influencing gonadal longevity.

(A–B) mir-84 and mir-241 target the 3′UTR of akt-1 but not unc-1 in vivo. While GFP driven by the akt-1 promoter in the DR construct remains similar, RFP controlled by the akt-1 3′UTR is significantly elevated in mir-84;mir-241.

(C–D) akt-1-3′UTR-DR is downregulated in germlineless animals. 17 WT and 29 glp-1 worms at D1 were examined from microscopic photos.

t-test, *** P<0.001.

(E) akt-1RNAi increases the lifespan of mir-84;mir-241;glp-1 and glp-1 to the same absolute extent.

When we examined functional interactions between akt-1 and the microRNAs, we observed that, akt-1i extended the lifespan of WT and glp-1 mutants by 27% and 23% respectively (Fig. 3E, Table S1), consistent with previous results (18). Strikingly, akt-1i enhanced longevity in mir-84;mir-241;glp-1 triple mutants to the same absolute extent as in glp-1 (Fig. 3E), suggesting akt-1-knockdown bypasses the requirement for the microRNAs. Correlatively, akt-1i restored expression of daf-16 target genes, sod-3 and lipl-4, in mir-84;mir-241;glp-1 animals (Fig. 4F, Fig. S7F). These results argue that akt-1 acts in the gonadal pathway, where the microRNAs normally antagonize akt-1 to promote longevity, as well as independently through canonical insulin/IGF signaling (18). Because daf-2/IR and gonadal longevity are additive, akt-1 could be regulated by inputs other than daf-2 upon germline ablation and serve as a general regulator of daf-16/FOXO.

Fig. 4. The lin-14 heterochronic locus is a microRNA target regulating gonadal longevity.

(A) lin-14i restores longevity to mir-84;mir-241;glp-1.

(B–C) DR construct with the 3′UTR of lin-14 but not of unc-1 is upregulated in mir-84;mir-241 animals compared to WT. Arrowheads: intestinal nuclei.

(D–E) lin-14-3′UTR-DR is downregulated in glp-1. Arrowheads: intestinal nuclei. For each genotype, 60 intestinal cells from 20 worms were analyzed. t-test, *** P<0.001.

(F) lin-14i or akt-1i increases lipl-4 expression in mir-84;mir-241;glp-1. ANOVA, * P<0.05, ** P<0.01.

The let-7 family of microRNAs target several genes in the heterochronic pathway, a circuit that controls larval developmental timing (19). These include the zinc-finger protein hbl-1/hunchback and the ring-finger protein lin-41/trim71. By mirWIP, other heterochronic genes, including nuclear protein lin-14 and the let-7 binding protein lin-28 are predicted targets. If these genes are microRNA targets in the gonadal pathway, then their downregulation should restore longevity to mir-84;mir-241;glp-1 triple mutants. RNAi treatment from L4 onwards revealed that only lin-14i restored lifespan extension to the triple mutants (Fig. 4A, Table S1).

lin-14 is an intriguing candidate since Slack had previously shown that lin-14 loss-of function extends lifespan in a daf-16/FOXO dependent manner, and lin-14 gain-of-function mutations shorten lifespan (20). During development lin-14 governs L1/L2 transitions, but its context in aging is unclear. We found that lin-14i extended lifespan in WT as reported, but did not further extend that of glp-1 mutants (Fig. 4A). To examine its relationship with daf-16/FOXO, we tested whether lin-14 knockdown influenced daf-16 target gene expression. Similar to aging experiments where longevity was restored, lin-14i also significantly restored daf-16 expression of sod-3 and lipl-4 to mir-84;mir-241;glp-1 (Fig. 4F, Fig. S7F). Altogether these observations suggest that the microRNAs downregulate lin-14 and promote longevity via daf-16.

To test this hypothesis, we examined regulation of lin-14 by mir-84;mir-241. As above, a luciferase reporter with the lin-14-3′UTR was downregulated by microRNAs (Fig. S7B). Similarly, a lin-14-3′UTR-DR construct was upregulated in mir-84;mir-241 double mutants relative to WT, whereas the unc-1-3′UTR-DR controls were unchanged (Fig. 4B–C). Consistent with a role in the gonadal pathway, the lin-14-3′UTR-DR construct as well as full-length lin-14::gfp were downregulated in intestinal nuclei of germline-less animals relative to gonad-intact controls (Fig. 4D–E, Fig. S7D–E). Collectively these results reveal that lin-14, a core component of the developmental clock, functions in the gonadal longevity circuit, where it is downregulated by microRNAs upon germline removal.

In this work we show that components of an early-life developmental timing switch (i.e., the steroid receptor DAF-12, its ligands, its target microRNAs of the let-7 family, and LIN-14, the microRNAs’ target) are used to regulate adult lifespan in response to signals from the gonad. We propose a model in which they work as part of a hormone- regulated switch between reproductive and survival modes at larval/adult stage commitments (Fig. S8). When GSC proliferation is prevented, unknown signals upregulate daf-36 and DA production by the L4/young adult stage, subsequently activating DAF-12 and its microRNA targets, mir-84 and mir-241. These microRNAs in turn downregulate akt-1, lin-14, and possibly other targets, which stimulate DAF-16/FOXO transcriptional activity, extending survival and lifespan. Because microRNA deletion does not fully abolish DAF-16 activity, other signals from either gonad or DAF-12 may also prompt gonadal longevity. Conversely, when GSC proliferation ensues, DA signaling is downregulated, microRNA expression is low, and lin-14 and akt-1 expression are high, resulting in normal lifespan. This switch could provide a critical link between development and longevity, serving as a checkpoint monitoring the state of the germline. For example, germline absence could mimic endogenous stress signals induced by germline quiescence or proliferative arrest in response to nutrient deprivation, infection or damage. As components of a developmental timer, the hormone-microRNA axis could ensure coordinate metabolism, maturation, and the relative timing of events between the reproductive system and the soma, with ultimate effects on lifespan. These findings extend the role of let-7 family members beyond developmental timing and differentiation to the regulation of insulin/IGF signaling and metabolism, similar to recent studies in mammals (21). Because let-7 family members and other components of this circuitry are evolutionarily conserved, it will be interesting to see if similar pathways impact longevity in vertebrates.

Materials and Methods

C. elegans strains and culture

Nematodes were grown with standard techniques at 20°C unless otherwise noted (22). All strains with glp-1(e2141) were maintained at 15°C and grown at 25°C to induce germlineless phenotype. NGM plates with ethanol or Δ⁴-dafachronic acid were prepared as reported (23). daf-16(mu86);muIs109(Pdaf-16::daf-16::gfp; Podr-1::rfp), pha-1(e2123);arEx1273(lin-14::gfp), and nhr-80(tm1011);glp-1(e2141) were kindly provided by Dr. Kenyon (UCSF), Dr. Greenwald (Columbia U.) and Dr. Aguilaniu (ENS de Lyon) respectively. Strains used are listed in Table S2.

Transgenes

Plasmids, akt-1p::DR::akt-1u, akt-1p::DR::unc-1u, pdk-1p::DR::pdk-1u, pdk-1p::DR::unc-1u, lin-14p::DR::lin-15u, and lin-14p::DR::unc-1u were injected into N2 to generate dhEx765, dhEx768, dhEx772, dhEx776, dhEx779, and dhEx781, respectively. For dhEx842, N2 was injected with mir-241p::mir-241, mir-84p::mir-84, and a co-transformation marker, myo-2::mCherry.

RNA interference

RNAi experiments were performed as described (24). For RNAi of cco-1, daf-2 and daf-16, synchronized eggs were seeded on HT115 plates expressing corresponding dsRNA. For RNAi of akt-1, lin-14, and other heterochronic genes, synchronized larvae were grown on OP50 plates until late L4 and transferred to corresponding HT115 plates for RNA interference.

Lifespan assays

Ablation of germline precursor cells by laser microsurgery and adult lifespan analysis were performed as previously reported (25). Worms undergoing internal hatching, bursting vulva, or crawling off the plates were censored. Statistical analysis was performed with the Mantel-Cox Log Rank method.

GC/MS/MS analysis

Worms were synchronized by bleaching and grown until day 1 of adulthood for analysis. Lipids were extracted from worms via the Bligh and Dyer method before derivitizing for GC-MS analysis (26). 7-dehydrocholesterol and dafachronic acid (DA) levels were analyzed by GC/MS/MS on a 7000A Triple Quadrupole GC/MS instrument (Agilent Technologies) equipped with an ESI source and an HP5-ms column. Whole-worm lipid extracts were spiked with cholesterol-d₇ and 5β-cholanic acid as internal standards and then derivitized with either Fluka III (Sigma) or trimethylsilyldiazomethane (Sigma) for 7-dehydrocholesterol or DA analysis, respectively. Compounds were analyzed in MRM mode using the following transitions: cholesterol-d₇ (m/z 465.4→360.3), 7-dehydrocholesterol (m/z 350.2→195.0), 5β-cholanic acid (m/z 374.3→264.0) and Δ⁷-DA (m/z 428.3→229). The results were from three independent experiments.

Plasmid construction

To generate pDR, pMyo-LD17, a gift from Dr. Lo, was modified by first removing the myo-2 promoter and L-HDAg gene within, then replacing the original unc-54 3′UTR with unc-1 3′UTR. The Dual Reporter (DR) module in pDR consists of gfp::icr::rfp as reported (27). DR reporters were subsequently generated from pDR as follows:

akt-1p::DR::akt-1u and akt-1p::DR::unc-1u

The akt-1 promoter of 3.2 kb was amplified from genomic DNA by PCR and inserted in front of the DR module. For akt-1p::DR::akt-1u, a 444 bp akt-1 3′UTR was amplified from genomic DNA and cloned after the DR module.

pdk-1p::DR::pdk-1u and pdk-1p::DR::unc-1u

The pdk-1 promoter of 2.9 kb was amplified from genomic by PCR and inserted in front of the DR module. For pdk-1p::DR::pdk-1u, pdk-1 3′UTR of 492 bp was amplified from genomic DNA and cloned after the DR module.

lin-14p::DR::lin-14u and lin-14p::DR::unc-1u

The lin-14 promoter (5.2 kb) was amplified from genomic DNA by PCR and cloned before the DR module. For lin-14p::DR::lin-14u, a 1.5 kb lin-14 3′UTR was amplified from genomic DNA and cloned after the DR module.

To generate mir-241p::mir-241 and mir-84p::mir84, a 2.0 kb mir-241 promoter, a 2.8 kb mir-84 promoter, and the microRNA stem-loops together with ~200 bp of flanking sequence were PCR amplified from genomic DNA. PCR fragments were subsequently cloned into L3781 (Fire vector kit 1997).

For luciferase assays, 3′UTRs of interest were amplified from genomic DNA by PCR and cloned after the luciferase gene in pMIR-REPORT Luciferase (Ambion). For plasmids expressing mature microRNAs, the stem-loop of tested microRNA and ~200 bp of flanking sequence were amplified from genomic DNA by PCR and cloned into the 3′UTR of GFP in pEGFP-C1.

Cell culture and transfection

HEK293T cells were maintained in DMEM medium supplemented with 10% fetal bovine serum (Gibco). Transfections were performed with FuGENE HD (Roche) according to the manufacturer’s instruction.

Luciferase assay

Plasmids expressing microRNAs were co-transfected into HEK293T cells with firefly and renilla luciferase reporters. Luciferase activities were measured with Dual-Luciferase Reporter Assay System (Promega) 24 h post transfection. The 3′UTR of interest was fused with firefly luciferase and renilla luciferase serves as an internal control. At least three independent experiments were performed.

qRT-PCR

Synchronized worms were grown until day 1 of adulthood unless otherwise noted. Afterwards, worms were collected in TRIzol (Invitrogen) and frozen in liquid nitrogen. For mRNA, total RNA was prepared by RNeasy Mini kit (QIAGEN) and cDNA was subsequently generated by Superscript III First Strand Synthesis System with random hexamers (Invitrogen). For microRNA, miRNeasy Mini kit (QIAGEN) and TaqMan MicroRNA Reverse Transcription kit (Applied Biosystems) were used for total RNA and cDNA preparation respectively. qRT-PCR was performed with Power SYBR Green master mix (Applied Biosystems) on a 7900HT Fast Real-Time PCR System (Applied Biosystems). Three or four technical replicates were performed in each reaction. ama-1 and Sno-RNA U18 were used as internal controls for mRNA and microRNA qRT-PCR respectively. The results were from at least three biological replicates. Primer sequences are listed in Table S3.

Analysis of fluorescence intensity

Fluorescence intensity was measured by COPAS Biosort (Union Biometrica), or from microscopic pictures by Photoshop (Adobe) or ImageJ (http://rsbweb.nih.gov/ij/) as indicated. For microscopic pictures, background signal was substracted as reported (28). Fluorescence microscopy was performed on a Carl Zeiss Axio Imager Z1. For analysis of mir-84p::GFP and mir-241p::GFP, images of indicated tissues were calibrated with InSpeck Green Microscopy Image Intensity Calibration Beads (Molecular Probes).

Statistical analysis

Results are presented as mean±SD unless noted otherwise. Statistical tests were performed with either One-way ANOVA with Tukey test (ANOVA) or Student’s t-test (t-test) as indicated using GraphPad Prism (GraphPad software).

Abbreviations

DA

dafachronic acid

IR

Insulin/IGF receptor

GSC

germline stem cell

DR

dual reporter