What is LongevityOne?

LongevityOne is a global platform dedicated to curing human aging. We consolidate advanced research, emerging technology, and unified data to accelerate innovation in the longevity space. Our mission is to serve scientists, bio-hackers, and advocates with trustworthy resources and collaborative tools.

Who founded LongevityOne?

LongevityOne was founded by Michael Peres, a serial entrepreneur and biomedical researcher with a background in Mathematics, Computer Science, and Jewish Studies. He is currently pursuing Biomedical Engineering to strengthen the technical foundation of LongevityOne’s work. Learn more about him at michaelperes.com.

What makes LongevityOne different?

We don’t just report on breakthroughs — we unify them. LongevityOne acts as a central repository where research, news, tools, and contributors converge. By removing silos in the aging research ecosystem, we accelerate progress toward reversing age-related decline.

Who is LongevityOne for?

Our platform is built for:Researchers in biotechnology, genetics, and agingMedical professionals and institutionsBio-hackers and health optimizersInvestors in life-extension technologiesPublic advocates for healthy aging

What are your long-term goals?

Our core objective is to contribute to the development of a foundational cure for aging. We aim to reshape the human experience of aging by supporting breakthrough research, increasing public understanding, and fostering international collaboration over the next 15 years.

How do I get involved?

Whether you’re a researcher, journalist, developer, or longevity enthusiast, we welcome your participation. Email us at contact@longevityone.org to explore partnerships, media opportunities, or to join our contributor network.

A TFEB–TGFβ axis systemically regulates diapause, stem cell resilience and protects against a senescence-like state

Main

During food scarcity or environmental stress, organisms throughout the tree of life can persist in long-lived states of quiescence, such as diapause and torpor, that enable them to outlast adversity until conditions improve^1,2. Diapause is an extreme form of prolonged arrest typically resistant to stress and geared toward survival, and can serve as a model of quiescence and resilience from cellular to organismal levels³. Indeed, its study has led to fundamental insights into aging, cellular endurance and disease. Even animals without explicit states of diapause, such as humans, globally remodel metabolism in response to fasting–refeeding, often with health benefits^4,5,6. Stem cells also persist through periods of quiescence followed by activation⁷. We can thus be thought of as cellular mosaics comprising dormant diapause-like cells and actively growing or dividing non-diapause cells^4,8, whose balance ensures stem cell longevity and organismal lifespan^4,9,10. Notably, such mechanisms can be co-opted by tumor cells that enter diapause-like states to escape immune surveillance and resist chemotherapy^11,12. Additionally, failure to exit cell cycle pausing may result in irreversible arrest, leading to a loss of tissue homeostasis, often described as cellular senescence¹³. Recent studies have identified only a handful of regulators of vertebrate diapause, and little is known about the mechanisms governing diapause entry, maintenance or exit and possible relationship to senescence¹⁴.

Results

HLH-30 protects against a senescence-like state, enabling dormancy and stem cell longevity

The nematode Caenorhabditis elegans is a facile model for the study of diapause. Aside from the well-characterized developmental dauer diapause¹, the worm can enter several other quiescent states, including the ARD, a largely unstudied state of dormancy triggered in response to late larval starvation^9,15. Such animals exhibit adult features and live over 2 months¹⁶. Upon refeeding, animals reactivate germline stem cells (GSCs) and regenerate somatic tissues to reproduce and live normal lifespans, revealing extraordinary survivorship and rejuvenation in the adult animal. We recently identified HLH-30/TFEB as a master regulator of ARD, whose mutation drastically reduces survivorship to 8–10 days and results in a failure to recover (Fig. 1a,b)¹⁶.

To better understand the role of HLH-30 during ARD and exit, we performed an in-depth phenotypic characterization of hlh-30 mutants (Fig. 1c–j) with a focus on the adult GSCs. GSCs are the only known stem cell compartment active in the adult worm, and their activity and differentiation are regulated by the distal tip cell (DTC) niche¹⁷. During our studies, we observed that hlh-30 mutants in ARD and recovery arrest in a senescence-like state. This state, not seen before in C. elegans, is characterized by several features of mammalian cellular senescence (Supplementary Table 1).

Mammalian cellular senescence is triggered by DNA damage, oncogene activation, mitochondrial dysfunction and other types of cellular damage¹⁸. Although senescent cells are heterogeneous, they are typified by enlarged cellular morphology, nucleolar expansion, cell cycle arrest, increased mitochondrial reactive oxygen species (ROS), senescence-associated β-galactosidase (SA-β-gal) activity and the senescence-associated secretory phenotype, as well as other features^19,20.

To examine GSCs in hlh-30 and N2 wild-type (WT) animals, we induced worms into ARD for only 2–4 days, to avoid the onset of death, examining the gonads by microscopy. Additionally, we investigated hlh-30 refeeding response by keeping worms in ARD for only 2 days, followed by 1–2 days of refeeding. While WT GSCs were compact during ARD, hlh-30 GSCs were swollen and harbored enlarged nucleoli (Figs. 1c and 4f). Upon refeeding, WT GSCs readily entered the cell cycle (Figs. 1d and 4g), while hlh-30 GSCs failed to do so. Importantly, irreversible cell cycle arrest and enlarged nucleoli were seen in WT animals kept long term (40 days) in ARD, suggesting WT GSCs experience similar senescence-like phenotypes with age (Extended Data Fig. 1a–c).

A common cause of cell cycle arrest is DNA damage and double-strand breaks²¹. To assess DNA damage in the GSCs, we stained the extruded gonads of animals with anti-RAD-51 antibodies and visualized foci proximal to the DTC niche by using a co-staining for the stem cell marker SYGL-1. In control WT animals maintained under ad libitum conditions and exposed to ionizing radiation, DNA damage foci were readily apparent in GSCs (Extended Data Fig. 1d). In contrast, no such foci were visible in WT experiencing ARD recovery without radiation (Figs. 1e and 4h). Interestingly, in hlh-30 mutants, we observed an elevated fraction of animals harboring GSCs with RAD-51 foci during ARD recovery, suggesting that hlh-30 mutants experience more DNA damage or have less efficient DNA repair.

Under physiological conditions, mitochondria dynamically switch between fusion and fission according to cellular metabolic demand²². With cellular senescence or aging, mitochondria can lose their dynamic flexibility and often appear fused, leading to increased ROS species²³. We, therefore, examined mitochondrial morphology in GSCs during ARD and refeeding and ROS production during ARD. Whereas WT worm GSCs went from fragmented mitochondria (higher sphericity) during ARD to fused (lower sphericity) upon refeeding (Fig. 1f and Extended Data Fig. 1e), hlh-30 mutant GSCs exhibited more fused mitochondria during ARD compared to WT (Fig. 1f). Upon refeeding, fusion was induced but to a lesser extent than WT. In line with this, we observed that hlh-30 mutants showed elevated levels of oxidative stress as measured by MitoTracker CM-H₂TMRos (Fig. 1g and Extended Data Fig. 4f), suggesting mitochondrial dysfunction similar to senescent cells in culture²⁴.

On an organismal level, we found increased SA-β-gal activity in hlh-30 mutants in ARD (Fig. 1h and Extended Data Fig. 1f), a hallmark of cellular senescence reflecting lysosomal dysfunction²⁰. Importantly, we did not see changes in SA-β-gal expression under ad libitum conditions (Extended Data Fig. 1g). Alongside this, we observed a number of phenotypes consistent with organismal aging and functional decline, including decreased pharyngeal pumping activity and muscle motility (Fig. 1i,j), as well as acceleration of transcriptomic age (Fig. 4i) as measured by the BiT age algorithm²⁵.

TFEB is considered a major regulator of autophagy, and dysregulation of this process accelerates cellular senescence¹³. To understand the potential role of autophagy during ARD, we examined 14 representative mutants of the autophagy cycle and lysosomal ATPase function. Mutants were kept in ARD for 10 days, refed, and the fraction of recovered reproductive animals and brood size measured. Generally, we found that most mutations (12/14) had no significant effect on the percentage of reproductive animals compared to WT, except for epg-8/ATG-14L and atg-2/ATG-2A mutations, which decreased this value (Fig. 1k). Additionally, the percentage change in brood size compared to ad libitum controls was minimal, except for epg-8/ATG-14L, whose mutation significantly decreased brood size (Extended Data Fig. 1h). Conceivably, this exceptional behavior is due to noncanonical roles of EPG-8/ATG-14L. We conclude that canonical autophagy is not essential for recovery during the time frame in which HLH-30 is required. Hence, other HLH-30/TFEB target processes must contribute to ARD stem cell longevity and recovery.

Selection against ARD senescence-like state reveals longevity and stem cell resilience pathways

To identify epistatic mechanisms acting downstream of hlh-30, we carried out ethyl methanesulfonate (EMS) mutagenesis of hlh-30(tm1978) mutants and screened for F2 suppressors that live at least 20 days, and which, upon refeeding, recover and reproduce (Fig. 2a). Whole-genome sequencing revealed that eight mutants harbored lesions in transforming growth factor beta (TGFβ) signaling (daf-1/TGFβR, daf-3/SMAD4), insulin–insulin-like growth factor (IGF) signaling (pdk-1/PDPK1, akt-1/AKT kinase, daf-2/Insulin/IGF receptor) and cGMP signaling (tax-4/CNGA1/CNGA2; Fig. 2b), the latter regulating both TGFβ and insulin-like peptide production in ciliated neurons^1,26. Mutations in these genes are known to affect the third larval stage dauer diapause¹ (that is, Daf mutants), suggesting an overlap of mechanisms with ARD. The remaining seven mutants were not identified because they did not map cleanly to a unique locus.

Identified candidate genes were validated for hlh-30 ARD rescue by using independent reference alleles (daf-1(m40), pdk-1(sa680), tax-4(p678), daf-2(e1368 and e1370); Fig. 2c, Extended Data Fig. 2a–c and Supplementary Table 2). We also tested the daf-7(e1372)/TGFβ mutant and saw similar rescue (Extended Data Fig. 2d). Notably, relative fold increase in mean lifespan of hlh-30 daf-1 double mutants compared to hlh-30 was usually greater (>3.2–4.1-fold) than that of daf-1 single mutants compared to WT (1.1–1.2-fold; Fig. 2c and Supplementary Table 2), suggesting a more specific interaction of TGFβ signaling with hlh-30 within ARD. These mutations also rescued other hlh-30-associated traits, namely ARD body size (Fig. 2d and Extended Data Fig. 2e) and reproductive ARD recovery (Fig. 2e). Given the robust rescue of reproductive capacity as an indirect readout for GSC health, we focused mainly on TGFβ signaling for further study.

To unravel how TGFβ signaling impacts hlh-30 ARD collapse, we first examined genetic epistasis interactions. Under ad libitum-fed conditions, DAF-7/TGFβ ligand is secreted mainly from the ASI sensory neurons and acts through TGFβ type I and II receptors, DAF-1 and DAF-4, throughout the body. This in turn activates the positive arm DAF-8 and DAF-14/r-SMADs, which inhibit the negative arm DAF-3/SMAD4 and DAF-5/SNO-SKI to promote the growth state (Fig. 2f). Under fasted conditions, TGFβ signaling is low, and DAF-3/DAF-5 become derepressed to promote the quiescent state²⁷. As expected, we observed a clear genetic dependence of daf-1-induced lifespan rescue of hlh-30 on downstream transcription factors daf-3/SMAD4, daf-5/SNO-SKI and daf-16/FOXO, another mediator of reduced TGFβ signaling (Fig. 2g and Supplementary Table 2)²⁸. In contrast to the daf-3(e1376) null mutation, which abolished daf-1-dependent survivorship and is dauer formation defective (Daf-d), the daf-3(syb2718) allele obtained from our genetic screen behaved oppositely; it restored hlh-30 survivorship and reproductive competence (Fig. 2h and Extended Data Fig. 2g), and on its own caused a dauer constitutive (Daf-c) phenotype, acting as a semi-dominant gain-of-function mutation (Extended Data Fig. 2h). Interestingly, the changed amino acid (p.Glu853Lys), falls within the MH2 domain (Extended Data Fig. 2f), and is predicted to disrupt inhibitory interactions with r-SMADs, uncoupling DAF-3/SMAD4 activity from upstream inputs²⁹. Last, we tested daf-12/VDR, a major downstream transcriptional mediator of bile-acid-like steroid signaling essential for dauer formation³⁰, but found that it was not required, consistent with both common and distinct regulators of resilience during larval dauer and ARD (Extended Data Fig. 2i).

HLH-30 regulates TGFβ signaling in response to nutrient cues

The genetic epistasis experiments described above suggest that HLH-30(+) could act to downregulate TGFβ signaling under ARD. To directly address this idea, we measured HLH-30 nuclear localization and expression of daf-7/TGFβ ligand in the ASI neurons during ARD and refeeding. These neurons secrete TGFβ, serving as an endocrine source to regulate TGFβ receptor signaling in target tissues throughout the body^27,31. We observed that HLH-30::mNeonGreen rapidly entered the nucleus of ASI neurons within 2 h of ARD induction, persisted during early ARD and promptly exited upon refeeding (Fig. 3a and Extended Data Fig. 3a). In line with nutrient-dependent regulation, daf-7p::GFP was down during WT ARD, but up upon refeeding (Fig. 3b,c). In hlh-30 mutants, daf-7p::GFP dynamic regulation was the opposite: up during ARD and no response to refeeding (Fig. 3b,c). Interestingly, we also observed a low-level daf-7p::GFP expression in outer labial OLQ neurons under ARD, not seen under ad libitum conditions, and whose regulation was also perturbed by hlh-30 (Extended Data Fig. 3b), suggesting ARD-specific regulation within these neurons. Together, our results indicate that HLH-30(+) partially downregulates daf-7 expression in a cell-autonomous manner in the ASI neurons in response to nutrient cues.

We observed a similar regulatory pattern in WT whole-body daf-1p::GFP expression (Fig. 3d). In WT, global daf-1p::GFP was expressed at low levels during ARD, and became elevated upon refeeding. In hlh-30 mutants, daf-1p::GFP was expressed at low levels similar to WT under ARD, but failed to upregulate upon refeeding, suggesting an inability to adapt to food cues (Fig. 3d). To examine tissue-specific regulation, we then monitored daf-1p::GFP within the DTC³². In WT, daf-1p::GFP expression in the DTC decreased during ARD and increased during refeeding (Fig. 3e,f). In hlh-30 mutants, by contrast, mean daf-1p::GFP expression was elevated by 27% compared to WT during ARD, and did not respond to refeeding, showing a misregulation of daf-1 receptor expression within the stem cell niche (Fig. 3e,f).

To test whether HLH-30 transcriptionally regulates TGFβ signaling components directly, we performed chromatin immunoprecipitation coupled with quantitative PCR (ChIP–qPCR) during ARD, pulling down HLH-30::FLAG and testing for candidate promoter enrichment on predicted HLH-30 binding sites by qPCR (Extended Data Fig. 3c,d). We found significantly enriched HLH-30 occupancy at daf-14 and daf-5 promoters (Fig. 3g). Accordingly, we saw that the hlh-30 mutation caused significant upregulation of daf-14 and downregulation of daf-5 mRNAs under ARD (Fig. 3h), confirming that HLH-30 directly inhibits TGFβ signaling under ARD by both repressing the positive arm and promoting the negative arm of the pathway. Our findings are also supported by existing yeast one-hybrid data³³ and hlh-30 ChIP–seq in the daf-2(e1370) background³⁴. We did not observe enrichment at daf-7 or daf-1 promoters (Fig. 3g). This observation is presumably because regulation mainly takes place in single cells where occupancy may fall below the limit of detection, binding might occur outside the tested region or regulation is indirect. The prominent rescue of hlh-30 ARD lifespan collapse and other features by daf-7/TGFβ or daf-1/TGFβRI mutation, together with the deregulated transgene expression, suggest an integral role for HLH-30 regulating the TGFβ pathway at multiple points, starting from the ASI neuron and signaling throughout the entire body via a systemic mechanism.

Regulation of growth signaling protects against a senescence-like state

C. elegans GSCs are regulated primarily by Notch signaling, which signals from the DTC niche to the stem cell pool. The DTC itself forms a cap-like structure around the GSCs and extends long processes to expand the zone of activation^35,36 (Fig. 4a). Under ad libitum conditions, the Delta/Serate/LAG-2 (DSL) ligand expressed in the DTC niche binds to the GLP-1/Notch receptor in adjacent GSCs, to promote expression of target genes including sygl-1, resulting in stem cell activation and mitosis, and preventing premature meiosis^37,38,39.

Given that TFEB–TGFβ impacts progeny production, we wondered whether it correspondingly affects Notch signaling and gonadal architecture and, thereby, GSC function during ARD and recovery. In particular, we hypothesized that in hlh-30 mutants, increased TGFβ signaling from neurons to the DTC might result in derepression of Notch signaling during ARD and drive GSCs into a senescence-like state. If true, then daf-1/TGFβ mutation should result in a prominent rescue. We measured three parameters: lag-2p::mKate2::PH expression level in the DTC (niche), DTC long process extension (niche matrix remodeling) and sygl-1::3xFlag abundance (GSCs) as proxy measures of Notch activity. In WT, we observed that all three parameters were reduced under ARD (Fig. 4b–d), while DTC extension and sygl-1::3xFlag expression increased upon recovery, consistent with downregulation of Notch signaling in quiescence and upregulation during activation, respectively (Fig. 4c,d and Extended Data Fig. 4a,b). By contrast, hlh-30 mutants exhibited elevated lag-2p::mKate2::PH and sygl-1::3xFlag expression under ARD compared to WT, yet during recovery, failed to extend DTC processes, and limited sygl-1::3xFlag expression, suggesting hlh-30 loss dysregulates Notch signaling during both quiescence and activation in niche and stem cells (Fig. 4b–d and Extended Data Fig. 4a,b). Remarkably, daf-1 mutation reversed features of hlh-30-induced Notch dysregulation (Fig. 4b,c,e and Extended Data Fig. 4a), implying that HLH-30 regulates Notch signaling via TGFβ to facilitate GSC resilience.

Of note, elevated Notch and TGFβ signaling are both considered hallmarks of senescent cells^40,41, suggesting that their dysregulation in hlh-30 mutants could contribute to a senescence-like state. In particular, we wondered if the daf-1 mutation could protect against other hlh-30-induced senescence-like phenotypes, as described in Fig. 1. Indeed, daf-1 loss was sufficient to restore GSC nucleolar area (Fig. 4f), mitosis (Fig. 4g), DNA damage status (Fig. 4h), transcriptomic age (Fig. 4i), SA-β-gal expression (Extended Data Fig. 4c), pumping activity, muscle health (Extended Data Fig. 4d,e) and mitochondrial ROS (Extended Data Fig. 4f) of hlh-30 mutants.

Aside from physiological changes, we observed vast transcriptomic changes in hlh-30 mutants compared to WT or hlh-30 daf-1 during ARD and refeeding. Reasoning that genes downregulated in hlh-30/WT might be upregulated in hlh-30 daf-1/hlh-30 and vice versa, we found a large fraction of such differentially expressed genes (DEGs) negatively correlated in ARD and refeeding (REF) (Fig. 4j and Extended Data Fig. 5a). We found mainly metabolic and growth processes misregulated in hlh-30 and reversed in hlh-30 daf-1 mutants (Extended Data Fig. 5b–e), highlighting the importance of the TFEB–TGFβ axis in metabolic remodeling during ARD. Notably, ARD DEGs strongly upregulated in hlh-30/WT and rescued by daf-1 included the innate immune response, a feature similar to increased inflammatory signaling in senescence (Extended Data Fig. 5c). In accordance with increased RAD-51 foci observed in GSCs from hlh-30 mutants, we could detect a signature of DNA damage response in our transcriptomic dataset (Supplementary Table 3). Interestingly, we found that hlh-30 upregulated refeeding DEGs overlapped significantly with the SenMayo gene set for senescent cells (16/32 orthologs, P = 0.00068 by hypergeometric test; Supplementary Table 4), which are core genes associated with senescence⁴². Altogether, these findings suggest that the hlh-30 mutation could trigger genomic instability, which, together with dysregulated metabolism, growth signaling and immune response, leads to a senescence-like state. Remarkably, this state can be prevented by downregulation of TGFβ signaling.

In summary, our results are consistent with the idea that ARD induces an aberrant cellular and organismal quiescence in hlh-30 mutants that resemble cellular senescence, dependent on TGFβ signaling. Building on earlier work^27,32,39, we incorporate HLH-30 and growth regulatory pathways into a working model: In response to ARD entry, HLH-30(+) normally downregulates TGFβ within ASI sensory neurons and TGFβ receptor in the DTC niche, resulting in lower TGFβ signaling and activation of the DAF-5–DAF-3 repressor complex (Fig. 4k). HLH-30 also directly downregulates daf-14/SMAD3 and upregulates daf-5/Sno-Ski, reinforcing the inhibition of the pathway. Consequently, downregulated lag-2/DSL in the niche reduces Notch signaling, resulting in reproductive quiescence. The proper regulation of this signaling cascade ensures stem cell resilience during dormancy and restoration of the stem cell pool upon refeeding. In hlh-30 mutants, there is a misalignment between nutrient cues, and growth signaling regulation, TGFβ and Notch are improperly elevated during diapause, triggering a state reminiscent of mammalian senescence.

TFEB controls resilience in mammalian diapause models

We next wondered whether the role of TFEB was conserved in mammalian diapause models, including mouse embryonic diapause and human cancer dormancy. These diapause states can be mimicked in vitro using INK128 a dual inhibitor of mTOR–PI3K^11,43.

From a genome-wide CRISPR–Cas9 screen performed in mouse embryonic stem (mES) cells, we identified TFEB as a major determinant of diapause survival⁴⁴. In particular, we used a doxycycline-inducible Cas9 system, which was activated in either proliferating control mES cells or in INK128-diapause mES cells, both carrying a genome-wide library of short-guide RNAs (sgRNAs)⁴⁵. The relative abundance of sgRNAs was calculated as the ratio of before and after Cas9 induction and was analyzed for both diapause mES cells and proliferating mES cells 3 days after sgRNA editing. TFEB emerged as a gene depleted explicitly in the diapause population after Cas9 activation, suggesting a possible role in diapause survival and resilience (Fig. 5a)⁴⁴. To confirm these findings under non-screening conditions, we targeted TFEB by short interfering RNA (siRNA)-mediated knockdown. Consistently, TFEB siRNA significantly reduced the survival of mES cells by 39.4% (Fig. 5b). These data indicate that TFEB is essential for the survival of mES cells in diapause but not for proliferating mES cells. Similarly, hlh-30 mutation in C. elegans shows little effect on WT survival under ad libitum conditions but is essential for ARD^16,46.

**Fig. 5: TFEB controls resilience in mammalian diapause models.**

Having found evidence for a conserved role of TFEB in mammalian embryonic diapause, we asked if its function could be extended to the recently reported human cancer cell diapause-like state^11,12. Cancer diapause is a resilience mechanism that enables tumor cells to evade chemotherapy and immune surveillance, only to emerge again and relapse. For these experiments, we used human melanoma SK-Mel-147 cells that undergo a robust and reversible diapause-like arrest in the presence of INK128 (ref. ⁴⁴). Notably, siRNA-mediated downregulation of TFEB reduced survival of resistant cancer diapause-like cells by almost 50%, but did not affect survival of proliferating cancer cells (Fig. 5c). We wondered whether the remaining 50% of SK-Mel-147 cells that survived TFEB siRNA during the diapause state transitioned to senescence or continued to proliferate upon INK128 withdrawal. To test this, we investigated cellular morphology after 0 h and 72 h of drug removal after siRNA treatment. However, we did not detect senescence of SK-Mel-147 cells in the siTFEB or control cells; instead, the remaining viable cells proliferated slowly (Extended Data Fig. 6a). It is possible that the decision between senescence and cell death depends on cell type, media conditions or presence of niche cells. Interestingly, RNA sequencing (RNA-seq) revealed an upregulation in the levels of TFEB mRNA during diapause (Fig. 5d), and gene-set enrichment analysis revealed altered TGFβ signaling (Fig. 5e), further supporting the similarity of diapause-like states in human cancer cells and worms. It is important to note that elevated TGFβ signaling generally inhibits proliferation in mammalian cells⁴⁷ but promotes proliferation in worms. Hence, in both systems, proliferation and growth are inhibited during diapause: in mammalian cells by activating TGFβ and in worms by inhibiting TGFβ.

Discussion

In this work, we showcase ARD as a tractable model to illuminate the molecular architecture of stem cell longevity, regeneration, and senescence in vivo. Further, we develop a platform to decipher conserved resilience pathways governing diapause entry, maintenance and exit through rapid genetic screening.

We find that fasting-induced ARD entry downregulates growth signaling and fosters metabolic remodeling and reproductive quiescence, while refeeding during ARD exit reverses these processes, showing an adaptive response to nutrient cues that ensures stem cell and organismal resilience. This adaptive response is mediated by HLH-30/TFEB as its mutation causes a misalignment between nutrient cues and growth signaling, thereby inducing a senescence-like state that abrogates GSC and organismal longevity. In line with our results, TFEB overexpression in the mouse brain seems to mitigate senescence and memory deficits⁴⁸. We speculate that mammalian senescence can also arise from a similar misalignment between nutrient supply, metabolism and growth signaling. Further study of hlh-30-induced dysfunction in non-proliferating cells, such as neurons, could provide a unique opportunity to define features of senescence in postmitotic cells.

Importantly, our studies establish a causal relationship between TFEB–TGFβ signaling axis regulation with respect to diapause and senescence. As part of this adaptive mechanism, we discovered that HLH-30 potently inhibits TGFβ signaling by acting at several points in the pathway. In the absence of HLH-30 during ARD, reinstated TGFβ regulation restores resilience during quiescence, protects against senescence and promotes rejuvenation during recovery, linking regulation and physiology in a causal chain. Interestingly, secreted mammalian TGFβ is often an integral component of the senescence-associated secretory phenotype ⁴⁹. Future studies should illuminate the detailed cellular foci, regulatory circuity and time of action of HLH-30 function to convey cellular and organismal resilience. The fact that TGFβ receptor mutation alone leads to an ARD lifespan extension of WT (1.1–1.2-fold) could be because TGFβ growth signaling during ARD is downregulated but not off (Fig. 3). Additionally, daf-1 loss might further attenuate ARD aging-induced cellular senescence phenotypes, thus extending worm life.

Notably, although HLH-30/TFEB is considered a master regulator of autophagy, surprisingly, we saw little specific impact of genetically disabling this process during ARD. Under prolonged fasting and hypometabolism, such as ARD, autophagy might be curtailed to prevent autolysis and ration molecular resources.

Besides TGFβ signaling, our screens also identified insulin–IGF signaling components and tax-4/CNGA1/CNGA2 as mediators of ARD survivorship and recovery. Both act within ciliated sensory neurons to regulate TGFβ²⁶, and TAX-4 can additionally induce insulin-like peptide production⁵⁰. In line with this, hlh-30 loss causes dysregulation of numerous insulin-like peptides under ARD and refeeding (Extended Data Fig. 5f). Therefore, it seems likely that insulin–IGF signaling is also downregulated during ARD and upregulated during recovery, thus contributing to resilience and regeneration. Interestingly, reduced TGFβ, IIS and TAX-4 converge on DAF-16/FOXO to various extents, and HLH-30 and DAF-16 have overlapping target genes³⁴. Further, DAF-16/FOXO is required for ARD survival¹⁶, and a critical downstream mediator of daf-1 hlh-30 suppression (Fig. 2g), suggesting complementary roles of HLH-30/TFEB and DAF-16/FOXO in maintaining quiescence. Our study concurs with others linking TGFβ and insulin/IGF signaling to regulating Notch signaling within the GSC and niche^27,32,39.

Extending our studies to the mammalian system, we find that TFEB knockdown alone was sufficient to drastically reduce the viability of dormant mES cells and resistant cancer cells in a diapause-like state by up to 50%. In both worm and mammalian cancer diapause models, TFEB and TGFβ components are highly regulated. In support of the role of TFEB in mammalian diapause states, recent work revealed that TFE3 (a TFEB family member) is recruited to dormancy loci during embryonic diapause⁵¹. Interestingly, studies in cell culture have revealed that TFEB regulates fat metabolism⁵² and preferentially binds FOXO1 (ref. ³⁴), pointing toward a potential involvement of HLH-30/TFEB in lipid metabolism during diapause. Indeed, lipid metabolic pathways were prominently deregulated in our transcriptomics during hlh-30 ARD (Extended Data Fig. 5b).

It is important to also point out that we observed clear differences between worms and the cancer diapause model. Whereas hlh-30 mutants enter a senescence-like state upon refeeding, diapaused cells upon TFEB knockdown do not. Instead, they lose viability in diapause and have reduced regrowth after drug washout (Extended Data Fig. 6a). These differences in fate could reflect cell-type specificity or culture conditions because cancer diapause is induced with INK128 in nutrient-replete conditions, while worm diapause is induced under nutrient deprivation. Further, in the worm, stem or cancer cells are controlled by surrounding cells, a regulatory layer that is missing in vitro.

Turquoise killifish enter an obligate embryonic diapause as a resilience mechanism to outlast the dry season. Interestingly, tfeb, smad4, smad2 and smad3 are upregulated during killifish embryonic diapause based on existent transcriptome data (Extended Data Fig. 6b,c)^53,54. We find that these same factors are also upregulated in the adipose tissue of fasted young adult fish and downregulated upon refeeding (Extended Data Fig. 6d). Some of these factors become misregulated in older fish, consistent with age-related perturbation of the fasting–feeding response⁵⁵. It seems reasonable to speculate that loss of plasticity in older fish could reflect the accumulation of cells in a senescent state. Thus, rather than viewing ARD as a C. elegans-specific phenomenon, we argue that a core nutrient-sensing network is regulated and responsible for diapause and quiescent states across taxa, intimately linked to senescence.

The systemic regulation observed in C. elegans during ARD–refeeding and killifish fasting–refeeding paradigms may conceptually resemble heterochronic parabiosis experiments in mice, where circulating factors (for example, TGFβ relatives such as GDF15 and myostatin) modulate stem cell function across different physiological states^56,57. While heterochronic parabiosis is often framed in terms of ‘youthful’ and ‘aged’ molecular environments, many of these systemic factors function through differential expression levels rather than an intrinsic aging signature. Similarly, plasma from exercised aged mice has been shown to transfer rejuvenating effects to sedentary aged mice, enhancing neurogenesis and cognitive function through liver-derived circulating factors such as Gpld1 (ref. ⁵⁸). These findings suggest that systemic shifts in signaling—whether driven by age, exercise or metabolic states such as fasting–refeeding—can regulate resilience and regeneration across taxa.

In sum, our studies reveal the TFEB–TGFβ axis as a primordial circuit systemically regulating resilience, stem cell dynamics and senescence. Further study of ARD may provide unique molecular insights into the dynamics of senescence and its reversal.

Methods

ARD induction was performed as previously described¹⁶. Briefly, worms were grown to the mid-L3 larval stage. Larvae were collected in M9 buffer and washed four times with M9 buffer. In total, 600 worms were pipetted onto 3-cm plates containing 4 ml nematode growth medium (NGM) with UltraPure agarose (Thermo Fisher Scientific) and 50 μg ml⁻¹ ampicillin. Worms were maintained for their whole life on one plate at 20 °C, unless noted otherwise.

Materials availability

Further information and requests for resources and reagents should be directed to and will be fulfilled by the corresponding author. This study did not generate unique reagents or new plasmids. Strains generated in this study are available upon request.

Experimental model and subject details

Nematodes were cultured using standard techniques at 20 °C on NGM agar plates with the Escherichia coli strain OP50, unless otherwise noted. All strains used in this study are listed in the Table 1. Mutant strains were obtained from the Caenorhabditis Genetics Center (CGC) or the National BioResource Project and outcrossed to our N2 WT (C. elegans variant Bristol) as indicated in Table 1. CRISPR–Cas9 mutants were generated by SunyBiotech (https://www.sunybiotech.com/). ARD conditions are described below.

Table 1 Reagents

Full size table

ARD induction

ARD induction was performed as previously described¹⁶. Briefly, worms were grown to the mid-L3 larval stage. Larvae were collected in M9 buffer and washed four times with M9 buffer. In total, 600 worms were pipetted onto 3-cm plates containing 4 ml NGM with UltraPure agarose (Thermo Fisher Scientific) and 50 μg ml⁻¹ ampicillin. Worms were maintained for their whole life on one plate at 20 °C, unless noted otherwise.

Recovery from ARD

ARD worms were transferred to NGM plates seeded with E. coli OP50. Successful exit and recovery from ARD were determined by visual inspection (body size, motility) and the ability to produce progeny, indicating germline reconstitution. For total brood size measurement, individual ARD worms were picked to OP50-seeded 3-cm NGM plates, and the progeny number per worm was counted. While producing progeny, the worms were transferred regularly until reproduction had ceased.

Lifespan experiments

ARD lifespans were determined by scoring a population of about 400–600 ARD worms every third day. In one experiment several plates per genotype were scored. Day 0 corresponds to the L3 stage¹⁶. Lifespan experiments under ad libitum conditions were performed as described⁵⁹. Day 0 corresponds to the L4 stage.

hlh-30 suppressor screen

hlh-30(tm1978) L4 larvae were mutagenized with 0.5% EMS in M9 buffer according to standard protocols⁶⁰. Each 10 mutagenized P0 adults were allowed to lay eggs overnight on a 10-cm plate. F1 adults were treated with hypochlorite to enrich eggs, and ARD was induced on mid-L3 worms of the F2 generation (see above). After 20 days in ARD, hlh-30 suppressors were transferred to OP50-seeded plates and recovery was monitored (‘Recovery from ARD’). Approximately 24,000 genomes were screened. Reproductive worms were singled, resulting in 15 strains after retesting: EMS-1 and EMS-7 (test mutagenesis), A1, I4, 37, 39, 40, 43, 51, 67, 70, 71, 87, 104 and 109 (mutagenesis 2).

Whole-genome sequencing and Galaxy MiModD analysis

For genomic sequencing, we prepared genomic DNA from strains listed above with QIAGEN Gentra PureGene tissue kit. Sequence libraries were created using the TruSeq DNA sample prep (Illumina). Libraries were sequenced on a HiSeq 2500 (Illumina) to generate 150-bp paired-end reads. Library preparation and sequencing were performed by the Max Planck Genome Center. Sequencing data were analyzed using Galaxy software. The WS220/ce10 C. elegans assembly was used as the reference genome for annotation.

Single nucleotide polymorphism-based mapping was performed on mutant strains A1, 39, 40, 43, 71 and 109 by crossing them to hlh-30(tm1978) in the Hawaiian strain CB4856 background. F1 generation was bleached and ARD was induced on the F2 generation. After 20 days, ARD worms were transferred to OP50-seeded plates, recovered worms were singled on 6-cm plates, and progeny were pooled for genomic DNA preparation. Pooled DNA was sequenced on an Illumina HiSeq platform (paired-end 150 nucleotide). Mutations were identified with MiModD software (https://celegans.biologie.uni-freiburg.de/?page_id=917/). WS220/ce10 C. elegans assembly was used as a reference genome for annotation. Causative mutations were confirmed by testing existing reference alleles or CRISPR–Cas9-designed base-pair change of identified genes (that is, daf-3(syb2718) in the hlh-30 background on ARD recovery).

Dauer assay

To determine dauer formation, synchronized worm populations were generated by allowing 20 worms to lay eggs for 4 h. Eggs were incubated at 25 °C for 48 h. Dauer characteristics such as a constricted pharynx, dark coloration and a long, thin body shape were used for identification. To study daf-3(syb2718) dominant or recessive dauer entry, we used a recessive dpy-1(e1) marker to score for heterozygosity. daf-3(syb2718);dpy-1 and control worms were crossed with WT males. F1 cross progeny eggs were shifted to 25 °C for 48 h, and heterozygous worms (non-dumpy phenotype) were scored for dauer entry.

Worm size measurements

Images of worms were taken with a Zeiss Axio Imager Z1 or a Leica M165 FC microscope. Body length was determined using ImageJ. At least 25 worms were analyzed per genotype.

RNA-seq C. elegans

For RNA-seq, total RNA was prepared from at least 3,000 worms per genotype, using the RNeasy Mini Kit (QIAGEN). Four independent biological replicates were prepared. polyA⁺ mRNA was isolated using NEBNext Poly(A) mRNA Magnetics Isolation Module (New England Biolabs). RNA-seq libraries were prepared with the NEBNext Ultra Directional RNA Library Prep Kit for Illumina (New England Biolabs). Libraries were quantified by fluorometry, immobilized and processed onto a flow cell with a cBot (Illumina) followed by sequencing-by-synthesis with TruSeq v3 chemistry on a HiSeq 2500 at the Max Planck Genome Center. Reads were quality trimmed with Flexbar v.2.5, then mapped to the reference genome (WBcel235.80) using HISAT2 v.2.0.4. Respective assemblies were merged with cuffmerge v.2.2.1, and differential gene expression analysis was performed with Cuffquant v.2.2.1 and Cuffdiff v.2.2.1. Gene Ontology annotation and enrichment was performed using DAVID bioinformatics resource database analysis via the Flaski web app for data analysis and visualization⁶¹ developed by the Bioinformatics Core Facility of the Max Planck Institute for Biology of Ageing.

RNA-seq killifish

All experiments were performed on adult (young 6–8 weeks, and old 18–20 weeks old) African turquoise killifish Nothobranchius furzeri laboratory strain GRZ-AD. Adult fish were single housed in 2.8 l tanks from the second week of life. Water parameters included a pH of 7–8, a kH of 3–5 and a temperature of 27.5 °C, 12 h of light and 12 h of darkness and fed with 10 mg of the dry pellet (BioMar INICIO Plus G) and Premium Artemia Coppens twice a day. The fish were either fasted for 72 h or fasted for the same amount of time and refed for 24 h and then euthanized. To reduce variability due to circadian rhythms, fish were euthanized all at once within 2 h in the early afternoon. Harvested tissues were snap frozen in liquid nitrogen and stored at −80 °C. RNA extraction of all samples was done at the same time. 1 μg of total RNA was used for library preparation. The sequencing was performed on the Illumina HiSeq 4000 sequencing system (∼50 million reads per sample) using a paired-end two × 100-nucleoide-long sequencing protocol. After removing rRNA and tRNAs, reads were pseudo-aligned to the reference genome (Nfu_20140520) using Kallisto (0.45.0). Pairwise differential gene expression was performed using DESeq2 (1.24.0). Animal experimentation was approved by ‘Landesamt für Natur, Umwelt und Verbraucherschutz Nordrhein-Westfalen’: 81-02.04.2019.A055.

Biological age prediction of worms

Biological age prediction of worms was carried out using a transcriptome-based aging clock, BiT age²⁵. Source code was downloaded from https://github.com/Meyer-DH/AgingClock/. Counts-per-million normalized RNA-seq reads for each genotype were used as input for BiT age analysis.

SenMayo gene analysis in C. elegans

We transferred the 125 SenMayo genes⁴² to C. elegans orthologs using OrthoList 2 (ref. ⁶²). If there were multiple orthologs, we chose the one with the most matches in homology databases. In total, we generated a list of 32 C. elegans genes (SenMayo C. elegans orthologs; Supplementary Table 4). We used FLASKI⁶¹ to perform hypergeometric testing (population size 20,000) comparing the SenMayo C. elegans gene set with gene sets up or down in hlh-30/N2 upon refeeding (REF).

Imaging and image analysis

For live imaging, worms were anesthetized, if not stated differently, in 0.1% sodium azide. Image analysis was performed on two-dimensional (2D) or 3D images using Fiji software⁶³. hlh-30::mNeonGreen nuclear localization in ASI neurons was determined at 2 h and 48 h of ARD, and after refeeding 48-h ARD worms for 1, 4, 6 or 24 h using a Zeiss Axio Imager Z1 microscope. Nuclear localization ratios were determined by measuring the fluorescence intensity in the nuclear and the cytosolic region of the ASI neurons. daf-7p::gfp expression was determined at 48, 72 and 96 h of ARD and 48 h of refeeding after a 48-h ARD period in one ASI neuron per worm using a Zeiss Axio Imager Z1 microscope. daf-7p::gfp expression in the OLQ was determined after 96 h of ARD. daf-1p::gfp expression in the DTC was determined at 48 h of ARD and 48 h of refeeding 48-h ARD worms using a Zeiss Axio Imager Z1 microscope. Whole-body daf-1p::gfp expression was monitored with the COPAS Biosorter (Union Biometrica, setting: green 450). lag-2p::mCherry-PH expression was determined in 48-h-old ARD worms. Worms were imaged with the Leica SP8 confocal microscope. z-stacks of whole DTCs (z-stack size, 0.3 µm) were captured. The sum projection of the DTC z-stacks was used to measure the total fluorescence intensity of the DTC cap structure. Maximal DTC length was determined by measuring the length of the longest DTC extension from distal to proximal using the lag-2p::mCherry-PH construct to color the DTC. GSC nucleolar area was determined from photos of the gonad taken with a Zeiss Axio Imager Z1 microscope (DIC contrast). Per genotype, 15 gonad arms from different animals were analyzed, scoring the area of each 3–5 nuclei located in the vicinity of the DTC. GSC mitochondria sphericity was assessed using tomm-20::mKate2 expression under a germline-specific promotor (pie-1) in 48-h-old ARD worms and worms refed from a 48-h ARD period for 48 h. Worms were anesthetized in levamisole and imaged with the Leica SP8-X confocal microscope. z-stacks of whole gonad arms (z-stack size, 0.35 µm) were captured. Image analysis was performed in 3D using the mitochondria analyzer plugin in Fiji⁶⁴ in the GSC (most distal 30 µm per gonad arm, thresholding method: mean). For different immunofluorescence analysis, germline dissections and staining were performed as described previously⁶⁵. Briefly, ARD and refed ARD worms were anesthetized in 200 mM levamisole, dissected, fixed in 2% formaldehyde and post-fixed in 100% methanol at −20 °C for 10 min. Fixates were blocked in 30% Goat serum (Cell Signaling, 5425S) or 1% BSA and stained with different primary antibody (Phospho-Histone H3, 1:150 dilution, Cell Signaling, 9706S; anti-FLAG, 1:1,000 dilution, Sigma-Aldrich, F3165, anti-RAD-51, 1:1,000 dilution, a gift from the Smolikove lab) overnight. Secondary antibody staining (Alexa Fluor 488, A-11008, Alexa Fluor 594, A-11005, 1:400 dilution) was followed by fixate embedding in Fluoromount-G (Invitrogen, 00-4959-52) containing DAPI. z-stacks of whole gonad arms (z-stack size, 0.3 µm) were captured using a Leica SP8-X or a Leica SP8-DLS confocal microscope. Imaging analysis was performed on z-projections in Fiji⁶³ and for the cell cycle analysis using the 3D image viewer of Imaris.

SA-β-gal assay

Equal amounts of 48-h-old ARD worms were freeze-cracked and stained using a senescence β-galactosidase staining kit (Cell Signaling, 9860) according to the manufacturer’s instructions. Image analysis was performed in Fiji⁶³, and color values were generated on RGB images.

CM-H₂TMROS staining

ARD plates contained 2 µM MitoTracker CM-H₂TMRos (Thermo Fisher Scientific). After 48 h in ARD, worms were transferred to a plate without the dye for 1 h to remove traces of air-oxidized dye. Worms were anesthetized using levamisole and imaged using the Zeiss Axio Imager Z1 microscope. Image analysis was performed in Fijij. The whole body was analyzed for orange fluorescence, excluding the head and tail regions.

ChIP–qPCR

ChIP–qPCR was performed after one day of ARD, as described previously⁶⁶. Briefly, at mid-L3 stage, 80,000–120,000 worms were induced into ARD and cultured on 10-cm ARD plates. Two genotypes were included: HLH-30::3xFLAG and WT. After 24 h of ARD, worms were collected in M9 buffer, snap frozen in liquid nitrogen, and stored at −80 °C.

Once two independent biological repeats were collected, samples were thawed and incubated with 2% paraformaldehyde solution for 20 min at room temperature. A total of four biological replicates were used. Worms were then washed with HLB buffer (50 mM HEPES-KOH, 150 mM NaCl, 1 mM EDTA, 0.1% sodium deoxycholate, 0.1% SDS and cOmplete Protease Cocktail (Roche)) and sonicated with the Bioruptor Plus (Diagenode) (three times, five cycles with 30 s on and 30 s off at 4 °C, low power). The total extract was centrifuged (20,000g, 10 min, 4 °C), and the supernatant transferred to a new microcentrifuge tube. Protein concentration was measured using a Bradford assay (Pierce Coomassie Plus, Bradford-Assay).

The total protein load was used for HLH-30::3xFLAG immunoprecipitation. Three input samples were saved and flash frozen in liquid nitrogen for later analysis. Total extract was incubated with Dynabeads Protein G magnetic beads (Thermo Fisher Scientific) for 1 h at 4 °C to remove unspecific binding proteins. 1 µl of anti-FLAG antibody (Sigma-Aldrich) was incubated with the total extract overnight. One ‘bait’ immunoprecipitation without antibody was performed for each biological repeat. Next, 50 µl magnetic beads were added to the lysate–antibody mixture for 2 h at 4 °C. Bead conjugates were washed four times (2× WB1 (50 mM HEPES-KOH, 150 mM NaCl, 1 mM EDTA, 0.1% Triton X-100, 0.1% SDS, 1 mM PMSF), 1× WB2 (50 mM HEPES-KOH, 500 mM NaCl, 1 mM EDTA, 0.1%Triton X-100, 0.1% SDS, 1 mM PMSF), 1× WB3 (50 mM HEPES-KOH, 0.25 mM LiCl, 1 mM EDTA, 0.5% sodium deoxycholate, 0.5% NP-40)). Washing solutions were incubated for 5 min at 4 °C on a rotator shaker. Antibody conjugates were eluted from the magnetic beads by adding the elution buffer (50 mM Tris-HCL, 1% SDS, 10 mM EDTA) and incubation for 20 min at 65 °C.

For the following steps, one input sample per biological replicate was thawed and treated equally to immunoprecipitation samples. Samples were treated with 1 µl of 10 mg ml⁻¹ RNase A (1 h, at 36 °C) and followed by 1 µl of 10 mg ml⁻¹ Proteinase K treatment (2 h, at 55 °C). Reverse crosslinking was performed overnight at 65 °C. DNA was purified with phenol–chloroform–isoamylalcohol extraction. Total input DNA and immunoprecipitation DNA was loaded for real-time qPCR. Primer, DNA, and Power SYBR Green Master Mix were pipetted using the JANUS automated workstation (PerkinElmer). Four technical replicates were transferred to the 384-well plate per sample. Before the experiment, all primer sets were validated using standard dilution curves (primer efficiencies were between 90% and 110%). The comprehensive primer list can be found in Table 1. Real-time qPCR was performed using the ViiA 7 Real-Time PCR system machine (Applied Biosystems). ∆c_t and ∆∆c_t values were calculated as described previously⁶⁶. ∆c_t = c_t (immunoprecipitation sample) − c_t (input sample); ∆∆c_t = ∆c_t (immunoprecipitation sample) − ∆c_t (control immunoprecipitation sample). The fold change was calculated using 2⁻^∆∆ct.

Western blotting

For western blotting, the following ChIP samples were used: input (worm lysis), first wash after immunoprecipitation, final wash and sample elution. Samples were directly added to 4× Laemmli buffer with 0.9% 2-mercaptoethanol and incubated for 5 min at 95 °C. Next, protein samples were loaded on a midi protein gel (Bio-Rad). The gel was blotted to a midi nitrocellulose membrane (Bio-Rad) using the Trans-Blot Turbo Transfer System (Bio-Rad). The membrane was blocked for 1 h in 5% skim milk and incubated with primary anti-FLAG antibody overnight at 4 °C. Next, the membrane was incubated with the secondary antibody for 2 h at room temperature and imaged using Western Lightning Plus Enhanced Chemiluminescence Substrate. The following antibodies were used: anti-FLAG, 1:1,000 dilution, anti-α-tubulin (Sigma, T9026; 1:10,000 dilution) Sigma-Aldrich, F3165, anti-rabbit horseradish peroxidase (HRP; Thermo Fisher, G-21234; 1:5,000 dilution), anti-rabbit HRP (Thermo Fisher, G-21234; 1:5,000 dilution) and anti-mouse HRP (Thermo Fisher, G-21040; 1:5,000 dilution).

Cell culture procedure

ES cells were grown on gelatin 1% with ES medium (EM): Dulbecco’s modified Eagle’s medium (DMEM, high glucose, Life Technologies) supplemented with 15% fetal bovine serum (FBS, Gibco), penicillin–streptomycin (100 U ml⁻¹), LIF (1,000 U ml⁻¹), 0.1 mM non-essential amino acids, 1% glutamax and 55 mM β-mercaptoethanol. Diapause was induced by 200 nM INK128 (S2811, SelleckChem). Human melanoma SK-Mel-147 cells were grown in DMEM (high glucose, Life Technologies) supplemented with 10% fetal bovine serum (FBS, Gibco) and penicillin–streptomycin (100 U ml⁻¹). The diapause-like state was induced by 100 nM INK128.

Genome-wide sgRNA screening

To perform the genome-wide screen, we used a CRISPR–Cas9 system previously described⁴⁵. ES cells were generated from mice ubiquitously expressing Cas9 under control of the endogenous Cola1a locus and a tetracycline responsive operator transgene; reverse tetracycline-controlled transactivator synthesis is under the control of the endogenous ROSA26 locus. ES cells from these mice were infected with a lentiviral library encoding sgRNAs targeting 19,150 mouse genes, with approximately five independent sgRNAs per gene. Addition of doxycycline to the ES cells causes inducible expression of the Cas9 enzyme, which induces editing of the sgRNA targets. Diapause was induced in ES cells through the mTOR–PI3K inhibitor INK128 at 200 nM. After 5 days, half of the proliferating cells and half of the diapaused cells were sequenced to know the initial diversity of the sgRNA library. Doxycycline was added to the remaining cells for 3 additional days to induce sgRNA editing before sequencing. Genomic DNA was isolated from cell pellets using a genomic DNA isolation kit (Blood & Cell Culture Midi kit, QIAGEN). After gDNA isolation, sgRNAs were amplified and barcoded by PCR as in ref. ⁶⁷, to amplify the DNA fragment containing sgRNA sequences. PCR products were sequenced on a HiSeq 4000 instrument (Illumina) at 50-bp reads to a depth of 30 million reads per sample. Reads were preprocessed by removing adaptors using Cutadapt (v.4.1)⁶⁸ with parameters ‘–e 0.2 –a GTTTTAGAGCTAGAAATAGCAAGTTAAAATA –m 18’. Next, sequences were trimmed to length 19 to match that of the probes. An artificial genome was created for alignment with Bowtie (v.0.12.9)⁶⁹ using the probes’ sequences and the function ‘bowtie-build’ with default parameters. Finally, reads were aligned to this genome with parameters –S –t –p 20 –n 1 –l 19. Read counts were imported into R (ref. ⁷⁰) by reading the sam files and counting the number of occurrences of each probe. The resulting count matrix was normalized using the rlog function from the DESeq2 v.1.34.0 R package⁷¹. A linear model with random and fixed effects was fitted to the normalized probe data for each gene. The condition was used as a fixed covariable while the guides were included as random effects whenever there was more than one. The model was fit with the function lmer from the lme4 package⁷¹ or with the native R lm function if there was only one probe for a given gene. Contrast coefficients and P values were computed using the glht function from the multcomp package⁷² without any P-value adjustment. Gene Ontology gene-set collections were downloaded from the Gene Ontology knowledgebase⁷³. Genes quantified in the microarray study were annotated according to the Broad Hallmark⁷⁴. Functional enrichment analyses were performed using a modification of ROAST⁷⁵, a rotation-based approach implemented in the Limma R package⁷⁶, which is especially suitable for small experiments. Such modifications were implemented to accommodate the proposed statistical restandardization⁷⁷ in the ROAST algorithm, which enables its use for competitive testing⁷⁸. The MaxMean statistic was used for testing gene-set enrichment of the different gene collections⁷⁷. For each gene, the most variable guide within each gene was used in these analyses (median absolute deviation). The results of these analyses were adjusted by multiple comparisons using the Benjamini–Hochberg False Discovery Rate method⁷⁹. All these was performed using the functions of the roastgsa R package⁸⁰.

siRNA treatment

siRNA for TFEB was purchased from siTOOLs. siRNAs (non-targeting (NT) and TFEB) were used at a final concentration of 3 nM. Both proliferating and diapause-like SK-Mel-147 were transfected with siRNAs for 5 days before performing viability assays. Lipofectamine reagent RNAiMAX (13778075) was used at 2 μl per ml to perform the transfection. Cell viability was measured using CellTiter-Glo Luminescent cell viability assay (Promega). Raw data were acquired by measuring luminescence in a VICTOR Multilabel Plate Reader (Pelkin Elmer). For the diapause exit experiment, SK-Mel-147 cells were either proliferating or treated with INK128 for 7 days and then transfected with siNT or siTFEB. We tested four different conditions, first proliferating cells, INK128-treated cells, cells washed out of INK128 at day 0 of siRNA transfection and cells washed out of INK128 3 days after siRNA infection. Images were taken 6 days after siRNA transfection using a Nikon Eclipse TS2 brightfield microscope (RRID: SCR_025716).

Statistics and reproducibility

The statistical tests performed in this study are indicated in figure legends and method details. Data are represented as the mean ± s.d. if not stated otherwise in the figure legends. Number of replicates and animals for each experiment are enclosed in their respective figure legends and or method details. No statistical methods were used to predetermine sample sizes, but our sample sizes are similar to those reported in previous publications¹⁶. Data collection and analysis were performed blind to the conditions of the experiments.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

All data generated in this study are available in the main text or the Supplementary Information. C. elegans RNA-seq data accession code: GSE291659; WBcel235.80 reference genome C. elegans RNA-seq; WS220/ce10 C. elegans assembly was used as reference genome mutagenesis; (Nfu_20140520) reference genome for killifish RNA-seq, RNA-seq data accession code: GSE296348; Killifish diapause transcriptomic data were obtained from ref. ⁵³ and ref. ⁵⁴.

References

Fielenbach, N. & Antebi, A. C. elegans Dauer formation and the molecular basis of plasticity. Genes Dev. 22, 2149–2165 (2008).
Article CAS PubMed PubMed Central Google Scholar
Lubzens, E. et al. (eds) Dormancy and Resistance in Harsh Environments Vol. 21 (Springer, 2010). https://doi.org/10.1007/978-3-642-12422-8
Riddle, D. L. & Albert, P. S. in C. elegans II (eds Riddle, D. L. et al.) 739–768 (Cold Spring Harbor, 1997).
Longo, V. D., Di Tano, M., Mattson, M. P. & Guidi, N. Intermittent and periodic fasting, longevity and disease. Nat Aging 1, 47–59 (2021).
Article PubMed PubMed Central Google Scholar
Sun, M. L. et al. Intermittent fasting and health outcomes: an umbrella review of systematic reviews and meta-analyses of randomised controlled trials. EClinicalMedicine 70, 102519 (2024).
Article PubMed PubMed Central Google Scholar
Margina, D. M. & Dragoi, C. M. Intermittent fasting on human health and disease. Nutrients 15, 4491 (2023).
Article PubMed PubMed Central Google Scholar
Cheung, T. H. & Rando, T. A. Molecular regulation of stem cell quiescence. Nat. Rev. Mol. Cell Biol. 14, 329–340 (2013).
Article CAS PubMed Google Scholar
de Morree, A. & Rando, T. A. Regulation of adult stem cell quiescence and its functions in the maintenance of tissue integrity. Nat. Rev. Mol. Cell Biol. 24, 334–354 (2023).
Article PubMed PubMed Central Google Scholar
Angelo, G. & Van Gilst, M. R. Starvation protects germline stem cells and extends reproductive longevity in C. elegans. Science 326, 954–958 (2009).
Article CAS PubMed Google Scholar
Brandhorst, S. et al. A periodic diet that mimics fasting promotes multi-system regeneration, enhanced cognitive performance, and healthspan. Cell Metab. 22, 86–99 (2015).
Article CAS PubMed PubMed Central Google Scholar
Rehman, S. K. et al. Colorectal cancer cells enter a Diapause-like DTP state to survive chemotherapy. Cell 184, 226–242 (2021).
Article CAS PubMed Google Scholar
Dhimolea, E. et al. An embryonic diapause-like adaptation with suppressed Myc activity enables tumor treatment persistence. Cancer Cell 39, 240–256 (2021).
Article CAS PubMed PubMed Central Google Scholar
Garcia-Prat, L. et al. Autophagy maintains stemness by preventing senescence. Nature 529, 37–42 (2016).
Article CAS PubMed Google Scholar
Hussein, A. M., Balachandar, N., Mathieu, J. & Ruohola-Baker, H. Molecular regulators of embryonic diapause and cancer diapause-like state. Cells 11, 2929 (2022).
Article CAS PubMed PubMed Central Google Scholar
Seidel, H. S. & Kimble, J. The oogenic germline starvation response in C. elegans. PLoS ONE 6, e28074 (2011).
Article CAS PubMed PubMed Central Google Scholar
Gerisch, B. et al. HLH-30/TFEB is a master regulator of reproductive quiescence. Dev. Cell 53, 316–329 (2020).
Article CAS PubMed Google Scholar
Crittenden, S. L., Leonhard, K. A., Byrd, D. T. & Kimble, J. Cellular analyses of the mitotic region in the Caenorhabditis elegans adult germ line. Mol. Biol. Cell 17, 3051–3061 (2006).
Article CAS PubMed PubMed Central Google Scholar
Di Micco, R., Krizhanovsky, V., Baker, D. & d’Adda di Fagagna, F. Cellular senescence in ageing: from mechanisms to therapeutic opportunities. Nat. Rev. Mol. Cell Biol. 22, 75–95 (2021).
Article PubMed Google Scholar
Gonzalez-Gualda, E., Baker, A. G., Fruk, L. & Munoz-Espin, D. A guide to assessing cellular senescence in vitro and in vivo. FEBS J. 288, 56–80 (2021).
Article CAS PubMed Google Scholar
Hernandez-Segura, A., Nehme, J. & Demaria, M. Hallmarks of cellular senescence. Trends Cell Biol. 28, 436–453 (2018).
Article CAS PubMed Google Scholar
Dulic, V. et al. p53-dependent inhibition of cyclin-dependent kinase activities in human fibroblasts during radiation-induced G1 arrest. Cell 76, 1013–1023 (1994).
Article CAS PubMed Google Scholar
Ren, L. et al. Mitochondrial dynamics: fission and fusion in fate determination of mesenchymal stem cells. Front. Cell Dev. Biol. 8, 580070 (2020).
Article PubMed PubMed Central Google Scholar
Lee, S. et al. Mitochondrial fission and fusion mediators, hFis1 and OPA1, modulate cellular senescence. J. Biol. Chem. 282, 22977–22983 (2007).
Article CAS PubMed Google Scholar
Martini, H. & Passos, J. F. Cellular senescence: all roads lead to mitochondria. FEBS J. 290, 1186–1202 (2023).
Article CAS PubMed Google Scholar
Meyer, D. H. & Schumacher, B. BiT age: a transcriptome-based aging clock near the theoretical limit of accuracy. Aging Cell 20, e13320 (2021).
Article CAS PubMed PubMed Central Google Scholar
Chang, A. J., Chronis, N., Karow, D. S., Marletta, M. A. & Bargmann, C. I. A distributed chemosensory circuit for oxygen preference in C. elegans. PLoS Biol. 4, e274 (2006).
Article PubMed PubMed Central Google Scholar
Ren, P. et al. Control of C. elegans larval development by neuronal expression of a TGF-beta homolog. Science 274, 1389–1391 (1996).
Article CAS PubMed Google Scholar
Shaw, W. M., Luo, S., Landis, J., Ashraf, J. & Murphy, C. T. The C. elegans TGF-beta dauer pathway regulates longevity via insulin signaling. Curr Biol 17, 1635–1645 (2007).
Article CAS PubMed PubMed Central Google Scholar
Yan, X. et al. Smad7 protein interacts with receptor-regulated Smads (R-Smads) to inhibit transforming growth factor-beta (TGF-beta)/Smad signaling. J. Biol. Chem. 291, 382–392 (2016).
Article CAS PubMed Google Scholar
Motola, D. L. et al. Identification of ligands for DAF-12 that govern dauer formation and reproduction in C. elegans. Cell 124, 1209–1223 (2006).
Article CAS PubMed Google Scholar
Gunther, C. V., Georgi, L. L. & Riddle, D. L. A Caenorhabditis elegans type I TGF beta receptor can function in the absence of type II kinase to promote larval development. Development 127, 3337–3347 (2000).
Article CAS PubMed Google Scholar
Dalfo, D., Michaelson, D. & Hubbard, E. J. Sensory regulation of the C. elegans germline through TGF-beta-dependent signaling in the niche. Curr. Biol. 22, 712–719 (2012).
Article CAS PubMed PubMed Central Google Scholar
Fuxman Bass, J. I. et al. Transcription factor binding to Caenorhabditis elegans first introns reveals lack of redundancy with gene promoters. Nucleic Acids Res. 42, 153–162 (2014).
Article CAS PubMed Google Scholar
Lin, X. X. et al. DAF-16/FOXO and HLH-30/TFEB function as combinatorial transcription factors to promote stress resistance and longevity. Nat. Commun. 9, 4400 (2018).
Article PubMed PubMed Central Google Scholar
Byrd, D. T., Knobel, K., Affeldt, K., Crittenden, S. L. & Kimble, J. A DTC niche plexus surrounds the germline stem cell pool in Caenorhabditis elegans. PLoS ONE 9, e88372 (2014).
Article PubMed PubMed Central Google Scholar
Gordon, K. L., Zussman, J. W., Li, X., Miller, C. & Sherwood, D. R. Stem cell niche exit in C. elegans via orientation and segregation of daughter cells by a cryptic cell outside the niche. Elife 9, e56383 (2020).
Article CAS PubMed PubMed Central Google Scholar
Henderson, S. T., Gao, D., Lambie, E. J. & Kimble, J. lag-2 may encode a signaling ligand for the GLP-1 and LIN-12 receptors of C. elegans. Development 120, 2913–2924 (1994).
Article CAS PubMed Google Scholar
Kershner, A. M., Shin, H., Hansen, T. J. & Kimble, J. Discovery of two GLP-1/Notch target genes that account for the role of GLP-1/Notch signaling in stem cell maintenance. Proc. Natl Acad. Sci. USA 111, 3739–3744 (2014).
Article CAS PubMed PubMed Central Google Scholar
Pekar, O. et al. Linking the environment, DAF-7/TGFbeta signaling and LAG-2/DSL ligand expression in the germline stem cell niche. Development 144, 2896–2906 (2017).
Article CAS PubMed PubMed Central Google Scholar
Hoare, M. et al. NOTCH1 mediates a switch between two distinct secretomes during senescence. Nat. Cell Biol. 18, 979–992 (2016).
Article CAS PubMed PubMed Central Google Scholar
Acosta, J. C. et al. A complex secretory program orchestrated by the inflammasome controls paracrine senescence. Nat. Cell Biol. 15, 978–990 (2013).
Article CAS PubMed PubMed Central Google Scholar
Saul, D. et al. A new gene set identifies senescent cells and predicts senescence-associated pathways across tissues. Nat. Commun. 13, 4827 (2022).
Article CAS PubMed PubMed Central Google Scholar
Bulut-Karslioglu, A. et al. Inhibition of mTOR induces a paused pluripotent state. Nature 540, 119–123 (2016).
Article CAS PubMed PubMed Central Google Scholar
Ramponi, V. et al. H4K20me3-mediated repression of inflammatory genes is a characteristic and targetable vulnerability of persister cancer cells. Cancer Res. 85, 32–51 (2024).
Article PubMed Central Google Scholar
Ruiz, S. et al. A genome-wide CRISPR screen identifies CDC25A as a determinant of sensitivity to ATR inhibitors. Mol. Cell 62, 307–313 (2016).
Article CAS PubMed PubMed Central Google Scholar
Lapierre, L. R. et al. The TFEB orthologue HLH-30 regulates autophagy and modulates longevity in Caenorhabditis elegans. Nat. Commun. 4, 2267 (2013).
Article PubMed Google Scholar
Siegel, P. M. & Massague, J. Cytostatic and apoptotic actions of TGF-beta in homeostasis and cancer. Nat. Rev. Cancer 3, 807–821 (2003).
Article CAS PubMed Google Scholar
Wang, H. et al. TFEB protein expression is reduced in aged brains and its overexpression mitigates senescence-associated biomarkers and memory deficits in mice. Neurobiol. Aging 106, 26–36 (2021).
Article CAS PubMed Google Scholar
Tominaga, K. & Suzuki, H. I. TGF-beta signaling in cellular senescence and aging-related pathology. Int. J. Mol. Sci. 20, 5002 (2019).
Article CAS PubMed PubMed Central Google Scholar
Li, W., Kennedy, S. G. & Ruvkun, G. daf-28 encodes a C. elegans insulin superfamily member that is regulated by environmental cues and acts in the DAF-2 signaling pathway. Genes Dev. 17, 844–858 (2003).
Article CAS PubMed PubMed Central Google Scholar
Stotzel, M. et al. TET activity safeguards pluripotency throughout embryonic dormancy. Nat. Struct. Mol. Biol. 31, 1625–1639 (2024).
Article PubMed PubMed Central Google Scholar
Settembre, C. et al. TFEB controls cellular lipid metabolism through a starvation-induced autoregulatory loop. Nat. Cell Biol. 15, 647–658 (2013).
Article CAS PubMed PubMed Central Google Scholar
Reichwald, K. et al. Insights into sex chromosome evolution and aging from the genome of a short-lived fish. Cell 163, 1527–1538 (2015).
Article CAS PubMed Google Scholar
Hu, C. K. et al. Vertebrate diapause preserves organisms long term through Polycomb complex members. Science 367, 870–874 (2020).
Article CAS PubMed PubMed Central Google Scholar
Ripa, R. et al. Refeeding-associated AMPK(gamma1) complex activity is a hallmark of health and longevity. Nat. Aging 3, 1544–1560 (2023).
Article CAS PubMed PubMed Central Google Scholar
Kim, D. H. et al. GDF-15 secreted from human umbilical cord blood mesenchymal stem cells delivered through the cerebrospinal fluid promotes hippocampal neurogenesis and synaptic activity in an Alzheimer’s disease model. Stem. Cells Dev. 24, 2378–2390 (2015).
Article CAS PubMed PubMed Central Google Scholar
Conboy, I. M. et al. Rejuvenation of aged progenitor cells by exposure to a young systemic environment. Nature 433, 760–764 (2005).
Article CAS PubMed Google Scholar
Horowitz, A. M. et al. Blood factors transfer beneficial effects of exercise on neurogenesis and cognition to the aged brain. Science 369, 167–173 (2020).
Article CAS PubMed PubMed Central Google Scholar
Gerisch, B. et al. A bile acid-like steroid modulates Caenorhabditis elegans lifespan through nuclear receptor signaling. Proc. Natl Acad. Sci. USA 104, 5014–5019 (2007).
Article CAS PubMed PubMed Central Google Scholar
Brenner, S. The genetics of Caenorhabditis elegans. Genetics 77, 71–94 (1974).
Article CAS PubMed PubMed Central Google Scholar
Iqbal, A., Duitama, C., Metge, F., Rosskopp, D. & Boucas, J. Flaski. Zenodo https://doi.org/10.5281/zenodo.4849515 (2021).
Kim, W., Underwood, R. S., Greenwald, I. & Shaye, D. D. OrthoList 2: a new comparative genomic analysis of human and Caenorhabditis elegans genes. Genetics 210, 445–461 (2018).
Article CAS PubMed PubMed Central Google Scholar
Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat. Methods 9, 676–682 (2012).
Article CAS PubMed Google Scholar
Chaudhry, A., Shi, R. & Luciani, D. S. A pipeline for multidimensional confocal analysis of mitochondrial morphology, function, and dynamics in pancreatic beta-cells. Am. J. Physiol. Endocrinol. Metab. 318, E87–E101 (2020).
Article CAS PubMed Google Scholar
Kocsisova, Z., Mohammad, A., Kornfeld, K. & Schedl, T. Cell cycle analysis in the C. elegans germline with the thymidine analog EdU. J. Vis. Exp. 22, 58339 (2018).
Google Scholar
Mukhopadhyay, A., Deplancke, B., Walhout, A. J. & Tissenbaum, H. A. Chromatin immunoprecipitation (ChIP) coupled to detection by quantitative real-time PCR to study transcription factor binding to DNA in Caenorhabditis elegans. Nat. Protoc. 3, 698–709 (2008).
Article CAS PubMed PubMed Central Google Scholar
Shalem, O. et al. Genome-scale CRISPR-Cas9 knockout screening in human cells. Science 343, 84–87 (2014).
Article CAS PubMed Google Scholar
Martin, M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J. 17, 10–12 (2011).
Langmead, B., Trapnell, C., Pop, M. & Salzberg, S. L. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 10, R25 (2009).
Article PubMed PubMed Central Google Scholar
R Core Team. R: a language and environment for statistical computing. R Project https://www.r-project.org/ (2022).
Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014).
Article PubMed PubMed Central Google Scholar
Hothorn, T., Bretz, F. & Westfall, P. Simultaneous inference in general parametric models. Biom. J. 50, 346–363 (2008).
Article PubMed Google Scholar
Gene Ontology Consortium; Aleksander et al. The Gene Ontology knowledgebase in 2023. Genetics 224, iyad031 (2023).
Article Google Scholar
Liberzon, A. et al. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst. 1, 417–425 (2015).
Article CAS PubMed PubMed Central Google Scholar
Wu, D. et al. ROAST: rotation gene set tests for complex microarray experiments. Bioinformatics 26, 2176–2182 (2010).
Article CAS PubMed PubMed Central Google Scholar
Ritchie, M. E. et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43, e47 (2015).
Article PubMed PubMed Central Google Scholar
Efron, B. & Tibshirani, R. On testing the significance of sets of genes. Ann. Appl. Stat. 1, 107–129 (2007).
Article Google Scholar
Goeman, J. J. & Buhlmann, P. Analyzing gene expression data in terms of gene sets: methodological issues. Bioinformatics 23, 980–987 (2007).
Article CAS PubMed Google Scholar
Benjamini, Y. & Hochberg, Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. Ser. B 57, 289–300 (1995).
Article Google Scholar
Caballé-Mestres, A., Berenguer-Llergo, A. & Stephan-Otto Attolini, C. Roastgsa: a comparison of rotation-based scores for gene set enrichment analysis. BMC Bioinform. 24, 408 (2023).
Article Google Scholar
Therneau, T. Package survival: a package for survival analysis in R v.2.3.8. CRAN https://doi.org/10.32614/CRAN.package.survival (2015).
Dodt, M., Roehr, J. T., Ahmed, R. & Dieterich, C. FLEXBAR-flexible barcode and adapter processing for next-generation sequencing platforms. Biology 1, 895–905 (2012).
Article PubMed PubMed Central Google Scholar
Kim, D., Langmead, B. & Salzberg, S. L. HISAT: a fast spliced aligner with low memory requirements. Nat. Methods 12, 357–360 (2015).
Article CAS PubMed PubMed Central Google Scholar
Pertea, M. et al. StringTie enables improved reconstruction of a transcriptome from RNA-seq reads. Nat. Biotechnol. 33, 290–295 (2015).
Article CAS PubMed PubMed Central Google Scholar
Trapnell, C. et al. Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nat. Protoc. 7, 562–578 (2012).
Article CAS PubMed PubMed Central Google Scholar
Venz, R. et al. In-vivo screening implicates endoribonuclease Regnase-1 in modulating senescence-associated lysosomal changes. Geroscience 46, 1499–1514 (2024).
Article CAS PubMed Google Scholar
Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021).
Article CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

We thank the CGC (University of Minnesota), WormBase, the Japanese National Biosource Project for providing strains, Smolikove lab (University of Iowa) for sharing anti-RAD-51 antibodies and members of the laboratory of A.A. for helpful discussion. We thank all students who contributed during internships or rotations (D. Aidlen, L. Rance, M. Taglinger and M. Duncan), E. Ballhysa for assistance with gamma irradiation, the bioinformatics core and the imaging facility for assistance (MPI-AGE). Work in the laboratory of A.A. was supported by the Max Planck Society and Cologne Graduate School of Ageing. This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreement number 834259). V.R. is the recipient of a European Innovative Training Networks H2020-MSCA-ITN-2018 (Healthage – 812830) Marie Curie PhD Fellowship. Work in the laboratory of M.S. was funded by the IRB and ‘laCaixa’ Foundation, and by grants from the Spanish Ministry of Science co-funded by the European Regional Development Fund (ERDF) (SAF2017-82613-R), European Research Council (ERC-2014-AdG/669622) and Secretaria d’Universitats i Recerca del Departament d’Empresa i Coneixement of Catalonia (Grup de Recerca consolidat 2017 SGR 282). K.K. was supported by Japan Society for Promotion of Science Overseas fellowship (202260346).

Funding

Open access funding provided by Max Planck Society.

Author information

Author notes

These authors contributed equally: Tim J. Nonninger, Jennifer Mak, Birgit Gerisch.

Authors and Affiliations

Max Planck Institute for Biology of Ageing, Cologne, Germany
Tim J. Nonninger, Jennifer Mak, Birgit Gerisch, Kazuto Kawamura, Roberto Ripa, Klara Schilling, Christian Latza, Jonathan Kölschbach & Adam Antebi
Institute for Research in Biomedicine (IRB Barcelona), Barcelona Institute of Science and Technology (BIST), Barcelona, Spain
Valentina Ramponi & Manuel Serrano
Altos Labs, Cambridge Institute of Science; Granta Park, Cambridge, UK
Manuel Serrano
Cologne Excellence Cluster on Cellular Stress Responses in Aging-Associated Diseases (CECAD); University of Cologne, Cologne, Germany
Adam Antebi

Authors

Tim J. Nonninger
Jennifer Mak
Birgit Gerisch
Valentina Ramponi
Kazuto Kawamura
Roberto Ripa
Klara Schilling
Christian Latza
Jonathan Kölschbach
Manuel Serrano
Adam Antebi

Contributions

A.A., B.G., T.J.N. (worms) and M.S. V.R. (cells) conceived and designed the study. Investigation, J.M. (EMS mutagenesis screen, genomic and RNA-seq, lifespans), T.J.N. (all microscopy data, recovery), B.G. (initial idea and screen, lifespans, recovery), K.L. (daf-12 data), C.L. (technical assistance), K.K. (BiT age analysis), J.K. (single nucleotide polymorphism mapping), R.R. (killifish data) and V.R. (CRISPR cell screening). Writing—review and editing, A.A., B.G., T.J.N. and M.S.

Corresponding author

Correspondence to Adam Antebi.

Ethics declarations

Competing interests

M.S. is shareholder of Senolytic Therapeutics, Life Biosciences, Rejuveron Senescence Therapeutics and Altos Labs. In the past, M.S. has been a consultant (until the end of 2022) of Rejuveron Senescence Therapeutics and Altos Labs. The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript. The other authors declare no competing interests.

Peer review

Peer review information

Nature Aging thanks Masashi Narita and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 HLH-30 protects against cellular senescence, enabling dormancy and stem cell longevity.

(a) Number of M-phase cells in the distal gonad arms (progenitor zone) of wild type (N2) recovered for 24 hours from 4 or 40 days of ARD. The salmon color indicates worms without M-phase positive cells. Pooled data from 2 independent experiments. (b) Percentage of N2 gonad arms negative for M-phase positive cells (salmon color) upon recovery for 24 hours from 4 or 40 days of ARD. BR = 2. (c) Quantification of germ cell nucleolar areas of N2 worms at 4 or 40 days of ARD. BR = 3. One representative experiment. Each dot represents one nucleolus. (d) Photomicrographs of distal gonad arms of sygl-1::3xFlag worms in N2 background, stained with DAPI (nucleus), anti-3x FLAG (SYGL-1, GSC zone, yellow) and anti-RAD-51 (DNA damage foci, magenta) antibodies. Following 1 hour post 60 Gy irradiation, ad libitum worms were dissected and stained. Scale bar 10 μm. The distal end of the gonad is facing the left side of the image. (e) Representative images of mKate2 labeled mitochondria in GSCs at 48 hours of ARD and 48 hours of refeeding. Images show a single z-layer of the germline, fluorescence and DIC images. Genotypes pie-1p::tomm-20mKate in N2 and hlh-30(tm1978) background. Dashed lines mark the most distal GSC area (30 μm) used for image analysis. Arrowhead show the distal end of the gonad. Scale bar 10 μm (f) Representative photomicrographs of head regions of 48 hour ARD worms stained with SA-β-gal. Genotypes N2, hlh-30(tm1978). 4 worms per genotype. Scale bar 10 μm. (g) SA-β-gal RGB color intensity of head regions of young adult hlh-30(tm1978) worms normalized to N2. Staining control is RNAi again bgal-2⁸⁶. Each dot represents the SA-β-gal intensity of one worm. Pooled data of three biological repeats. One-way ANOVA followed by Turkey’s post-hoc test. (h) Precent mean brood size change of autophagy and lysosomal mutants under ad libitum (grey) or refeeding after 10 days in ARD (blue). Percent change was determined in respect to N2 or him-5 controls. Genotype alleles can be found in the reagent table. Each dots reflect one biological replicate. Kruskal-Wallis followed by Dunn’s post-hoc test. If not stated otherwise: Mean & s.d., Mann-Whitney test (two-sided).

Source data

Extended Data Fig. 2 Selection against ARD senescence reveals ARD longevity and stem cell resilience pathways.

(a), (b), (c) Insulin/IGF signaling mutants daf-2(e1370), daf-2(e1368) and pdk-1(sa680) extend ARD lifespan and partially rescue hlh-30(tm1978) ARD short survivorship. (d) TGFβ signaling mutant daf-7(e1372) extends ARD survival and partially rescues hlh-30(tm1978) ARD short survivorship. (e) Body length of ARD worms at 96 hours. Genotypes N2, hlh-30(tm1978), hlh-30 daf-1(m40), and hlh-30;daf-2(e1370). Each dot represents one animal. BR = 3, one representative experiment. One-way ANOVA followed by Turkey’s post-hoc test. (f) Predicted protein structure (Alphafold⁸⁷) of daf-3(syb2718) showing mutation E853K in the MH2 domain in red. Conserved Mad homology 1 and 2 domains (MH1 and MH2) in blue and yellow, respectively. (g) daf-3(syb2718) gain-of-function mutation restores progeny production of hlh-30 worms refed at days 10 ARD. daf-3(e1376) null mutant abolishes progeny production of hlh-30(tm1978) daf-1(m40) double mutants. Each dot represents the progeny number of one worm. One representative replicate. BR = 3. One-way ANOVA followed by Turkey’s post-hoc test. (h) daf-3(syb2718) exhibits a dominant gain-of-function dauer constitutive (Daf-c) phenotype at 25 °C. Genotypes N2, daf-3(syb2718), and daf-3(syb2718)/+ heterozygous. Each dot represents one experiment. One-way ANOVA followed by Turkey’s post-hoc test. (i) daf-12(rh61rh411) slightly enhances ARD survival of hlh-30(tm1978) daf-1(m40). Survival curves depict one representative experiment. Lifespan data and statistics (log-rank tests) are presented in Supplementary Table 2. If not stated otherwise: Mean & s.d.

Source data

Extended Data Fig. 3 HLH-30 regulates TGFβ signaling in response to nutrient cues.

(a) Nuclear/cytosolic hlh-30::mNeonGreen expression in ASI neurons under ARD (2 and 48 hours) and recovery from 48 h ARD (1, 4, 6, and 24 hours of refeeding). Each dot represents one experiment. BR = 3. Mean ± SEM (b) Quantification of daf-7p::GFP fluorescence in OLQ neurons of N2 and hlh-30(tm1078) at 96 hours of ARD. Exposure time 400 ms. Expression in OLQ is 5–10-fold lower than in ASI. Each dot represents the daf-7p::GFP expression in one OLQ neuron. BR = 3, one representative experiment. Mann-Whitney test (two-sided). (c) Schematic of ChIP-qPCR experimental design. (d) Control western blot ChIP-qPCR experiment showing specific elution of HLH-30::FLAG (BR = 2). ‘+’ indicates the use of an Anti-FLAG antibody during IP, ‘-‘ indicates no use of Anti-FLAG antibody (bait). Tubulin control staining was performed on the same gel after washing off the Anti-FLAG antibody. If not stated otherwise: Mean & s.d., One-way ANOVA followed by Turkey’s post-hoc test.

Source data

Extended Data Fig. 4 Regulation of growth signaling protects against senescence.

(a) Distal tip cell (DTC) extension measured from the cap to the end of the most distal process containing lag-2p::mKate2::PH. Each dot represents the mean DTC length of one BR. (b) Representative images of SYGL-1 stained and DAIP labeled germline stem cells (GSCs) at 48 hours of ARD and 24 hours of refeeding. Images show a single z-layer of the germline. Genotypes N2 and hlh-30(tm1978). Arrowhead show the distal end of the gonad. Scale bar 10 μm. (c) Mean SA-β-gal RGB color intensity in the head region of 48 h ARD worms normalized to N2. Genotypes N2, hlh-30(tm1978), hlh-30 daf-1(m40). Each dot represents the SA-β-gal mean intensity of one BR. (d) Mean pharyngeal pumping and (e) Mean body bending rates upon 48 hours of ARD. Genotypes: N2, hlh-30(tm1978), hlh-30 daf-1(m40). 30 second intervals. Each dot represents the mean pumping rate or mean body bends of one BR. (f) Quantification of mitochondrial ROS in N2, hlh-30(tm1078), hlh-30 daf-1(m40) at 48 hours of ARD. Pooled data from three BRs. Kruskal-Wallis followed by Dunn’s post-hoc test. If not stated otherwise: Mean & s.d., One-way ANOVA followed by Turkey’s post-hoc test.

Source data

Extended Data Fig. 5 Downregulation of growth signaling partially reverses senescence transcriptome.

(a) Correlation plot of differentially expressed genes (DEGs) of hlh-30/N2 and hlh-30 daf-1/hlh-30 at 12 hours of refeeding (REF). DEGs (padj < 0.05) highlighted in red (genes down in hlh-30 and reversed by daf-1, quadrant 2) or blue (genes up in hlh-30 and reversed by daf-1, quadrant 4). Other DEGs, gray. Simple linear regression line in gray. Equation Y = −0.35X – 0.06; R² = 0.27. (b) Gene ontology enrichment analysis of 2347 DEGs down in hlh-30(tm1978)/N2 ARD and up in hlh-30 daf-1(m40)/hlh-30 ARD (padj < 0.05, quadrant 2). Top 10 biological processes of the DAVID GO BP DIRECT database. Complete GO term list is shown in Supplementary Table 3. Modified Fisher Exact p value, EASE Score. (c) Gene ontology enrichment analysis of 2083 DEGs up in hlh-30/N2 ARD and down in hlh-30 daf-1/hlh-30 ARD (quadrant 4). Top 10 biological processes. Complete GO term list is shown in Supplementary Table 3. Modified Fisher Exact p value, EASE Score. (d) Gene ontology enrichment analysis of 1499 DEGs down in hlh-30/N2 and up in hlh-30 daf-1/hlh-30 (padj<0.05, quadrant 2) at 12 hours of refeeding (REF). Top 10 biological processes. Modified Fisher Exact p value, EASE Score. (e) Gene ontology enrichment analysis of 2173 DEGs up in hlh-30/N2 and down in hlh-30 daf-1/hlh-30 (padj<0.05, quadrant 4) at 12 hours of refeeding. Top 10 biological processes. Modified Fisher Exact p value, EASE Score. (f) Heat map of log2(FC) values of differentially expressed Insulin-like peptides in hlh-30/N2 at 48 hours of ARD and upon 12 hours of refeeding. padj< 0.05.

Source data

Extended Data Fig. 6 TFEB controls resilience in mammalian diapause models.

(a) Representative images of SK-Mel-147 cells treated with siTFEB or siNT control in diapause, or 6 days post INK128 wash out (w/o) either at day 0 of siRNA treatment or 72 hours post siRNA transfection. Scale bar: 100 µm (b) Heat map of wild type killifish tfeb, smad2, smad3b, and smad4b expression from published RNA seq dataset at 9 months of diapause or developing embryos 6 days post fertilization⁵³. z-score. BR = 4. (c) Heat map of wild type killifish tfeb, smad4(1 and 2 of 3) at 3 and 6 days and 1 months of diapause or developing embryos before and without diapause⁵⁴. z-score. BR = 3. (d) Heat map of wild type killifish tfeb, smad2, smad3b and smad4 expression from adipose tissues of young and old fasted and refed fish. RNA seq samples. z-score. BR = 4. ns = not significant.

Source data

Supplementary information

Reporting Summary

Supplementary Tables 1–4

Table 1. Hallmarks of cellular senescence. Table 2. Experimental data and statistics. Table 3. Overview of gene ontology biological processes (GOTERM_BP_DIRECT). Table 4. C. elegans SenMayo gene-set orthologs.

Source data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Nonninger, T.J., Mak, J., Gerisch, B. et al. A TFEB–TGFβ axis systemically regulates diapause, stem cell resilience and protects against a senescence-like state. Nat Aging (2025). https://doi.org/10.1038/s43587-025-00911-4

Download citation

Received: 30 January 2024
Accepted: 27 May 2025
Published: 30 June 2025
DOI: https://doi.org/10.1038/s43587-025-00911-4

A TFEB–TGFβ axis systemically regulates diapause, stem cell resilience and protects against a senescence-like state

Date

Source

Categories

Main

Results

HLH-30 protects against a senescence-like state, enabling dormancy and stem cell longevity

Selection against ARD senescence-like state reveals longevity and stem cell resilience pathways

HLH-30 regulates TGFβ signaling in response to nutrient cues

Regulation of growth signaling protects against a senescence-like state

TFEB controls resilience in mammalian diapause models

Discussion

Methods

Materials availability

Experimental model and subject details

ARD induction

Recovery from ARD

Lifespan experiments

hlh-30 suppressor screen

Whole-genome sequencing and Galaxy MiModD analysis

Dauer assay

Worm size measurements

RNA-seq C. elegans

RNA-seq killifish

Biological age prediction of worms

SenMayo gene analysis in C. elegans

Imaging and image analysis

SA-β-gal assay

CM-H2TMROS staining

ChIP–qPCR

Western blotting

Cell culture procedure

Genome-wide sgRNA screening

siRNA treatment

Statistics and reproducibility

Reporting summary

Data availability

References

Acknowledgements

Funding

Author information

Authors and Affiliations

Contributions

Corresponding author

Ethics declarations

Competing interests

Peer review

Peer review information

Additional information

Extended data

Supplementary information

Source data

Rights and permissions

About this article

Cite this article

Related Press Releases

FAQS

CM-H₂TMROS staining