Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 24 (19), 2238-46

MicroRNAs Mediate Dietary-Restriction-Induced Longevity Through PHA-4/FOXA and SKN-1/Nrf Transcription Factors

Affiliations

MicroRNAs Mediate Dietary-Restriction-Induced Longevity Through PHA-4/FOXA and SKN-1/Nrf Transcription Factors

Thalyana Smith-Vikos et al. Curr Biol.

Abstract

Background: Dietary restriction (DR) has been shown to prolong longevity across diverse taxa, yet the mechanistic relationship between DR and longevity remains unclear. MicroRNAs (miRNAs) control aging-related functions such as metabolism and lifespan through regulation of genes in insulin signaling, mitochondrial respiration, and protein homeostasis.

Results: We have conducted a network analysis of aging-associated miRNAs connected to transcription factors PHA-4/FOXA and SKN-1/Nrf, which are both necessary for DR-induced lifespan extension in Caenorhabditis elegans. Our network analysis has revealed extensive regulatory interactions between PHA-4, SKN-1, and miRNAs and points to two aging-associated miRNAs, miR-71 and miR-228, as key nodes of this network. We show that miR-71 and miR-228 are critical for the response to DR in C. elegans. DR induces the expression of miR-71 and miR-228, and the regulation of these miRNAs depends on PHA-4 and SKN-1. In turn, we show that PHA-4 and SKN-1 are negatively regulated by miR-228, whereas miR-71 represses PHA-4.

Conclusions: Based on our findings, we have discovered new links in an important pathway connecting DR to aging. By interacting with PHA-4 and SKN-1, miRNAs transduce the effect of dietary-restriction-mediated lifespan extension in C. elegans. Given the conservation of miRNAs, PHA-4, and SKN-1 across phylogeny, these interactions are likely to be conserved in more-complex species.

Conflict of interest statement

Competing Interests

The authors declare that they have no competing financial interests.

The authors declare no conflicts of interest.

Figures

Fig. 1
Fig. 1
Aging network of transcription factors and miRNAs (see also Fig. S1, Table S1). Pink lines: miRNA-mediated gene regulation, predicted by mirWIP; blue lines: transcription factor-mediated gene regulation, from modENCODE. Green nodes: transcription factors; orange nodes: miRNAs. Dotted line: PHA-4-mediated regulation; solid line: SKN-1-mediated regulation. Aging associated miRNAs that comprise feedback loops with either PHA-4 or SKN-1 are shown. Only mir-71 and mir-228 both target and are targeted by PHA-4 and SKN-1. Network visualization was performed in Cytoscape.
Fig. 2
Fig. 2
miRNAs function during dietary restriction of adult animals. (A–E) Error bars represent standard error mean. (a) miR-228 and miR-71 are required for dietary restriction (DR). E. coli (OP50) food was diluted 1:2.5, 1:5, 1:10 and 1:20 compared to the ad libitum (AL) concentration. Wild-type N2 animals fed diluted bacteria live significantly longer than N2 animals fed under ad libitum (* P < 0.01), as previously reported. By contrast, loss-of-function miRNA mutants mir-228 (n4382) (left panel) and mir-71 (n4115) (right panel) fail to exhibit this mean lifespan extension when fed diluted OP50, showing that miR-228 and miR-71 are required for DR response (see also Fig. S2a,b). (B) miRNAs are up-regulated during dietary restriction compared to ad libitum. Mature miRNA expression in day 3 adult animals was measured by qRT-PCR (normalized to U18). Left panel: When N2 animals are fed diluted OP50, expression of miR-228 is increased compared to miR-228 levels under ad libitum. Thus, miR-228 may be activated in response to dietary restriction. This was also confirmed by GFP analysis (see Fig. S2d). * = P < 0.05 (compared to ad libitum (AL)). Right panel: miR-71 is similarly up-regulated by dietary restriction induced by bacterial dilution. * = P < 0.05 (compared to ad libitum (AL)) (C) miRNAs are up-regulated under dietary restriction in the eat-2 mutant compared to wild-type as measured on day 3 of adulthood by qRT-PCR (normalized to U18). Similar to growing N2 animals on diluted OP50, mature levels of miR-228 and miR-71 are slightly higher in eat-2 (ad1113) mutants - a genetic model of DR- as compared to a wild-type background. By using an alternative model of DR, this also indicates that the miRNAs are activated under these conditions, as measured in day 3 adult animals via qRT-PCR. * = P < 0.05 (compared to N2 when measuring miR-71 and miR-228 levels) (D) Opposite regulation of miR-228 mature expression by pha-4 and skn-1 in adult animals, as measured on day 3 of adulthood by qRT-PCR (normalized to U18). miR-228 expression (measured by qRT-PCR) is also affected by pha-4 and skn-1, in which pha-4 promotes expression of miR-228, while skn-1 inhibits its expression, as compared to an empty vector RNAi control (L4440). Analysis of mir-228::GFP animals grown on pha-4 and skn-1 (RNAi) indicates that the PHA-4 and SKN-1 transcription factors may be directly targeting the mir-228 gene (see Fig. S2f). * = P < 0.05 (compared to L4440) (E) skn-1 is required for maximal expression of miR-71 in adult animals. In a skn-1 (RNAi) background, miR-71 is down-regulated, as compared to an empty vector RNAi control (L4440) at day 6 of adulthood (relative to RNA6B). * = P < 0.05 (compared to L4440)
Fig. 3
Fig. 3
miRNAs interact with PHA-4 and SKN-1 to affect longevity. (A) While mir-228 mutants grown on empty vector RNAi (L4440) are long-lived, mir-228 animals fed pha-4 and skn-1 (RNAi) during adulthood are short-lived (P < 0.05), similar to pha-4 and skn-1 (RNAi) alone. Therefore, mir-228 requires wild-type pha-4 and skn-1 to affect longevity and may act upstream of these transcription factors. (B) mir-71 overexpression, which causes animals to be long-lived, suppresses the short lifespan induced by skn-1 (RNAi) (P < 0.05). Thus, mir-71 may function downstream of skn-1. oeIs: overexpression integrated strain. (C–F) pha-4 and skn-1 mRNAs are targeted by miR-71 and miR-228. Error bars represent standard error mean. (C) pha-4 and skn-1 levels are increased in mir-228 mutant adults as measured by qRT-PCR, indicating that they may be targeted by this miRNA. This was true under both ad libitum and DR conditions (see Fig. S3a). qRT-PCR of 3-day adults, normalized to geometric mean of cdc-42, pmp-3 and Y45F10.* = P < 0.05 (compared to N2) (D) Using transgenic lines containing the endogenous 3′UTRs of pha-4 and skn-1, respectively, analysis of pha-4::GFP and skn-1::GFP expression validates that levels of these transcription factors are increased in mir-228 mutants. This held true under both ad libitum and DR conditions (see Fig. S3b). Fluorescence was compared between pha-4::GFP or skn-1::GFP in mir-228 mutant relative to N2 day 3 adults. * = P < 0.05 (compared to N2).(E) pha-4 levels are increased in mir-71 mutant adults as measured by qRT-PCR, indicating that pha-4 may be targeted by this miRNA. skn-1 levels are slightly down-regulated in mir-71 adults compared to N2 adults, perhaps through an indirect targeting mechanism. qRT-PCR of 3-day adults, normalized to geometric mean of cdc-42, pmp-3 and Y45F10. * = P < 0.05 (compared to N2) (F) Analysis of pha-4::GFP expression validates that levels of this transcription factor are increased in mir-71 mutants (see also Fig. S3c–h). Fluorescence was compared between pha-4:: in mir-71 mutant vs N2 day 2 adults. * = P < 0.05 (compared to N2)
Fig. 4
Fig. 4
miR-228 affects longevity and stress response in C. elegans (see also Fig. S4). (A) miR-228 antagonizes longevity in C. elegans, as mir-228 mutants are long-lived compared to N2 wild-type (P < 0.01). Three lines of mir-228 overexpressors co-injected with myo-3::GFP are shorter-lived than N2 or myo-3::GFP (P < 0.05). (B) mir-228 mutants are resistant to heat stress. When day 1 adult animals grown on solid media are subjected to a 4-hour heat shock at 35°C, the percentage of mir-228 animals surviving this stress is significantly higher than N2 (* P < 0.05). Conversely, the percentage of mir-228 overexpressor animals surviving this stress is significantly lower than N2 or myo-3::GFP (* P < 0.05). This indicates that the longevity and stress response phenotypes of this miRNA are highly correlated. Error bars represent standard error mean. (C,D) miR-228 is dynamically expressed during aging. (C) mir-228::GFP expression during aging (head region). GFP expression was quantified as a read-out of mir-228 levels, which peak in mid-adulthood. This same pattern holds true when quantifying mir-228 in the head region (amphid and excretory cells), where expression is brightest, or quantifying expression across the entire animal. Error bars represent standard error mean. (D) Expression of mature miR-71, miR-228, pha-4, and skn-1 in wild-type animals over multiple aging time points via qRT-PCR. Error bars represent standard error mean. pha-4 follows the same pattern as shown by Lund et al. [50], in which levels peak at about day 6 adulthood before decreasing; miR-228 exhibits the same expression profile, indicating that PHA-4 promoting expression of mir-228 is the stronger side of the feedback loop. As miR-71 levels are increasing, pha-4 levels are lower than miR-71, but when miR-71 levels decrease, pha-4 levels are higher; consistent with the model that miR-71 targets pha-4. Finally, skn-1 levels remain constant during all aging time points, indicating that the opposing interactions of miR-228 targeting and down-regulating skn-1 while SKN-1 targets mir-228 may effectively cancel each other out, so that skn-1 levels do not change.
Fig. 5
Fig. 5
A model for how miR-71 and miR-228 interact with PHA-4 and SKN-1 to mediate the effects of dietary restriction leading to increased lifespan. Dietary restriction promotes the expression of miR-71 and miR-228, which then function in feedback loops with PHA-4 and SKN-1 simultaneously. miR-228 targets both pha-4 and skn-1, and PHA-4 promotes mir-228 expression, while SKN-1 antagonizes mir-228 expression. Conversely, SKN-1 promotes mir-71 expression, and miR-71 targets pha-4. These feedback loops fine-tune the DR response in C. elegans to promote lifespan.

Similar articles

See all similar articles

Cited by 25 PubMed Central articles

See all "Cited by" articles

Publication types

MeSH terms

Feedback