Differential adiponectin signalling couples ER stress with lipid metabolism to modulate ageing in C. elegans

The metabolic and endocrine functions of adipose tissue and the ability of organisms to cope with cellular stress have a direct impact on physiological ageing and the aetiology of various diseases such as obesity-related pathologies and cancer. The endocrine effects of adipose tissue are mediated by secreted adipokines, which modulate metabolic processes and influence related maladies. Although a plethora of molecules and signaling pathways associate ageing with proteotoxic stress and cellular metabolism, our understanding of how these pathways interconnect to coordinate organismal physiology remains limited. We dissected the mechanisms linking adiponectin signalling pathways and endoplasmic reticulum (ER) proteotoxic stress responses that individually or synergistically affect longevity in C. elegans. Animals deficient for the adiponectin receptor PAQR-1 respond to ER stress, by rapidly activating the canonical ER unfolded protein response (UPRER) pathway, which is primed in these animals under physiological conditions by specific stress defence transcription factors. PAQR-1 loss enhances survival and promotes longevity under ER stress and reduced insulin/IGF-1 signalling. PAQR-1 engages UPRER, autophagy and lipase activity to modulate lipid metabolism during ageing. Our findings demonstrate that moderating adiponectin receptor -1 activity extends lifespan under stress, and directly implicate adiponectin signalling as a coupler between proteostasis and lipid metabolism during ageing.

Ageing is characterized by deterioration of physiological regulatory functions, and progresses towards frailty and age-associated pathology. Cellular adaptation to extrinsic and intrinsic stress stimuli relies on a wide range of cellular processes that are tightly controlled. Failure of organisms to quickly and efficiently respond to stress has a direct negative impact on homeostasis, leading to feeble healthspan and reduced survival [1][2][3][4] . Although a large number of stress responses have been associated with ageing, the intersecting regulatory networks coordinating these responses remain enigmatic.
Perturbed protein homeostasis (proteostasis) is one of the hallmarks of ageing 1,3,5,6 . Several compensatory stress response mechanisms, acting at the cellular or organelle level, function to preserve the stability and functionality of the proteome 1, 2 . The heat shock response (HSR) is one of the best characterized responses aimed at maintaining cytoplasmic proteome integrity 2,5 . Organelle-specific response pathways are regulated by the endoplasmic reticulum (ER) unfolded protein response (UPR ER ), the less well characterized mitochondrial unfolded protein response (UPR mt ), the peroxisomal quality control system and autophagy 2,7,8 . Deterioration of these homeostatic mechanisms is a general feature of ageing and also occurs in several age-associated diseases. However, it remains unclear whether deterioration is simply a consequence of the ageing process or a significant causative factor.
Perturbations of ER proteostasis results in ER stress and initiation of the UPR ER aimed at ameliorating accumulation of unfolded proteins. ER stress and impaired UPR ER signalling have been implicated in the aetiology of age-related maladies including metabolic diseases 9 . ER stress aggravates metabolic dysfunction, thereby contributing to enhanced lipogenesis, obesity and insulin resistance. In a reciprocal manner, obesity, caused by

Results
The C. elegans adiponectin receptors differentially influence resistance to ER Stress and animal survival. To investigate the impact of the C. elegans adiponectin receptors PAQR-1, PAQR-2 and PAQR-3 on the sensitivity to ER stress, we exposed wild-type (WT) and paqr-1(tm3262), paqr-2(tm3410) and paqr-3(ok2229) mutant animals to the ER stress inducers tunicamycin and thapsigargin, which inhibit N-linked glycosylation and sarco/endoplasmic reticulum Ca 2+ -ATPase (SERCA), respectively. Compared to WT animals, paqr-1 mutants were more resistant, whereas paqr-2 and paqr-3 mutants were more sensitive, to conditions of ER stress ( Fig. 1A and Supplementary Fig. S1). paqr-1 silencing also increases ER stress resistance ( Supplementary Fig. S2A), however the relative improvement is weaker than the one found in paqr-1 mutant animals. This is likely due to partial loss of PAQR-1 by RNAi. Next, and to determine whether adiponectin receptors also impact general proteostatic mechanisms initiated by HSR, WT and mutant animals were exposed to non-permissive temperatures. No significant differences in survival between the different mutants under heat stress conditions were observed ( Supplementary Fig. S3), suggesting that the observed effects are specific to the ER stress response. Notably, and despite the fact that exposure to ER stress, from early in development, was fatal for most animals and their progeny, animals lacking a functional PAQR-1 exhibited substantial adaptation to adverse conditions. We observed that the small percentage of paqr-1 mutants that developed into adults during the ER stress test remained viable more than 8 days after hatching on plates containing tunicamycin, whereas WT animals, paqr-2 or paqr-3 mutants that reached adulthood under the same conditions perished in less than 3 days. Moreover, the paqr-1 mutants that survived under harsh experimental conditions, in contrast to WT, paqr-2 or paqr-3 mutant animals, were able to lay viable progeny; however, the progeny never reached adulthood ( Fig. 1B and Supplementary Fig. S4, images for paqr-1 and WT shown).
Under physiological conditions, lifespan is not significantly affected by the paqr-1(tm3262) or paqr-3(ok2229) mutations, whereas it is reduced by the paqr-2(tm3410) mutation 15 . We could recapitulate these findings ( Supplementary Fig. S5). However, since paqr-1(tm3262) animals are more resistant to ER stress, we queried whether their survival under conditions of ER stress is improved. Lifespan assays performed following challenging 1-day old animals with tunicamycin showed that indeed paqr-1(tm3262) animals survived substantially longer than the WT animals, or paqr-2(tm3410) and paqr-3(ok2229) mutants ( Fig. 1C left panel). In a similar setting WT animals subjected to paqr-1(RNAi) survived also substantially better than the control (EV) animals (Supplementary Fig. S2B left panel). WT and paqr-3(ok2229) animals showed comparable sensitivity, whereas paqr-2(tm3410) animals were less resistant ( Fig. 1C left panel). Interestingly, when 7-day old animals were challenged with tunicamycin the survival advantage of paqr-1(tm3262) animals or WT animals subjected to paqr-1(R-NAi) was no longer evident ( Fig. 1C right panel and Supplementary Fig. S2B right panel). Given that the ability to activate the UPR ER and initiate protective mechanisms against ER stress declines with age 16,17 , our findings indicate that UPR ER activity is important for the survival advantage of paqr-1(tm3262) mutants under ER stress.
PAQR-1 mitigates resistance to ER stress by impeding the canonical UPR ER pathway. The evolutionary conserved IRE-1/XBP-1 pathway is mainly orchestrating the transcriptional response to ER stress in C. elegans. In addition, two other branches of UPR ER , are mediated by PERK-1/PEK-1 and ATF-6. To investigate whether activation of the canonical UPR ER pathway is required for increased ER stress resistance upon PAQR-1 depletion, we examined ER stress-sensitive xbp-1, pek-1 and atf-6 mutants that also carry the paqr-1(tm3262) lesion. We found that loss of PAQR-1 fully restored ER stress resistance in atf-6 mutants and slightly increased ER stress resistance of xbp-1 and pek-1 mutants ( Fig. 2A). A two-way ANOVA analysis showed that paqr-1 plays a significant role on its own, which however, is highly dependent on the genetic background. These findings suggest a synergistic effect mediated at least by two arms of the UPR response. Upon ER stress, transcription of hsp-4 (homolog of mammalian chaperone BiP/Grp78) is upregulated in an IRE-1/XBP-1 dependent manner. Therefore, we additionally examined UPR ER activation in paqr-1(tm3262) mutants using a p hsp-4 GFP reporter for UPR ER activation. Basal activation of hsp-4 was induced by PAQR-1 deficiency (Fig. 2B left panel). Exposure of wild-type animals to tunicamycin for either 4h or 24h induced hsp-4 expression and this response was amplified in paqr-1(tm3262) mutants, or WT animals subjected to paqr-1(RNAi) (Fig. 2B middle and right panels and Supplementary Fig. S2C).
In C. elegans, ER stress responses decline early during adulthood and are almost abolished by the age of 7 days 17 . We queried whether loss of PAQR-1 confers enhanced UPR ER during ageing. Tunicamycin treatment of wild-type animals at ages ranging from day 1 to day 12 of adulthood potently induced hsp-4 expression early during adulthood, while induction markedly declined with age (Fig. 2C). PAQR-1 depletion amplified tunicamycin-triggered UPR ER as manifested by monitoring hsp-4 expression (Fig. 2C).
We finally examined whether differential activation of UPR ER by tunicamycin is due to alterations in either uptake or intestinal release of the stressor. Assessment of ultradian rhythms (pharyngeal pumping, defecation), with or without tunicamycin, showed no differences between WT and paqr-1(tm3262) mutant animals ( Supplementary Fig. S6). Collectively, our findings indicate that loss of PAQR-1 function potentiates ER stress responses. Animals bearing paqr-1 lesions mount a rapid and enhanced response to ER stress, revealing a novel role for PAQR-1 in fine-tuning the canonical UPR ER pathway.
PAQR-1 depletion abrogates the effects of low insulin/IGF-1 signalling on survival and stress resistance. Reduced insulin/IGF-1 signalling increases resistance to a variety of insults and stressors, including heat, hypoxia, paraquat, DNA damaging agents, pathogens, tunicamycin and dithiothreitol (DTT), in C. elegans and other metazoans 18,19 . Impairment of the insulin/IGF-1 signalling pathway also extends lifespan by promoting proteostasis, among others 18 . ER stress response pathways are important mediators of longevity in C. elegans DAF-2 (insulin/IGF-1 receptor) deficient animals. The IRE-1/XBP-1 branch of the UPR ER pathway increases longevity and ER stress resistance of daf-2 mutants by setting a lower UPR ER threshold, through mechanisms that involve activation of the DAF-16/FOXO transcription factor 18 . Given that PAQR-1 is a negative regulator of UPR ER , we examined the effects of that paqr-1 knockdown on the lifespan and ER stress resistance of animals with reduced insulin/IGF-1 signalling. We find that paqr-1(tm3262) mutants subjected to daf-2 RNAi display enhanced survival compared to WT animals (Fig. 3A). daf-2 silencing increases ER stress resistance in both WT animals and paqr-1(tm3262) mutants (Fig. 3B). paqr-1 expression is not altered during ageing or under Eggs derived from WT, paqr-1(tm3262), paqr-2(tm3410) and paqr-3(ok2229) strains were laid on plates in the absence (white bars) or presence of 5 μg/ml tunicamycin (black bars). Percentage of eggs that developed into mature adults was scored. Each strain was scored on four independent plates and each experiment was repeated independently at least three times. Error bars represent SEM for repeat plates within the experiment. ***P < 0.001 vs. WT + tunicamycin. (B) Survival of eggs that developed into mature adults after 8 days on tunicamycin (5 μg/ ml). Representative images illustrating survival of WT and paqr-1(tm3262) strains are shown. Differences on bacterial lawn thickness at the time of observation indicate changes in mobility and viability. Scale bar, 250 μm. White arrowhead indicates a dead worm; white arrows indicate eggs that never developed into mature adults; red arrowheads indicate live worms; red arrows indicate newly hatched eggs. (C) WT (black), paqr-1(tm3262) (red), paqr-2(tm3410) (blue), and paqr-3(ok2229) (green) animals were transferred at day 1 or day 7 onto plates containing 50 μg/ml tunicamycin and survival was monitored. Lifespan values are given in Table S1.
ER stress ( Supplementary Fig. S7). The relative improvement in ER stress resistance upon daf-2 RNAi silencing (≈ 1.6-fold above vector control) is weaker than that reported for daf-2 mutant animals (≈ 4-fold above WT) 18 . This is likely due to partial inactivation of daf-2 by RNAi. We find that PAQR-1 deficiency restores UPR ER as manifested by HSP-4 expression in daf-2(RNAi) animals, and further induces HSP-4 expression both, under control and ER stress conditions (Fig. 3C, D). Notably, the effects of DAF-2 deficiency on UPR ER induction were fully bypassed by PAQR-1 loss.
reporter, was not altered upon DAF-16 and/or SKN-1 depletion in WT animals. By contrast, hsp-4 expression was significantly increased by skn-1(RNAi) and reduced by simultaneous DAF-16 and/or SKN-1 depletion in paqr-1(tm3262) mutants ( Fig. 4F and Supplementary Fig. S8). Together, these findings indicate that SKN-1 is capable of regulating PAQR-1-dependent ER homeostasis, although its relative contribution depend on the state of stress experienced by the organism.

PAQR-1 regulates lipid droplet turnover during ageing in ER-stressed animals.
Adiponectin receptors have been implicated in several aspects of lipid metabolism 15,23,24 . Notably, disruption of the UPR ER in C. elegans causes severe defects in the intestine, the major lipid storage tissue of these animals 25 . Thus, we investigated the involvement of PAQR-1-mediated signalling on lipid metabolism under conditions inducing UPR ER . To this end, we measured lipid content in WT animals and paqr-1 mutants during ageing under ER stress. one-and five-day old paqr-1(tm3262) animals responded to tunicamycin by decreasing their lipid content (vs. control), but this response was absent in ten-day old animals (Fig. 5A). The decreased fat accumulation in stressed one-and five-day old paqr-1 mutants is not a consequence of altered food uptake or defecation rates, since these ultradian rhythms were not affected compared to WT animals ( Supplementary Fig. S6). Importantly, loss of lipid catabolic responses upon ER stress in ten day-old paqr-1(tm3262) mutants (Fig. 5A) coincides with loss of the capacity to activate UPR ER at this age (Fig. 2C), indicating that a UPR ER -mediated mechanisms might be involved in lipid clearance observed in PAQR-1-depleted animals. Given that adiponectin signalling suppresses lipolysis in mouse adipocytes 24 , we investigated whether impairment of PAQR-1 similarly induces lipid droplet clearance. We examined the consequences of paqr-1 silencing in WT C. elegans animals on lipid droplet size under ER stress conditions. The presence of tunicamycin significantly decreased droplet size upon silencing of paqr-1 (Fig. 5B and Supplementary Fig. S9). These findings indicate that both PAQR-1 and UPR ER modulate lipid droplet metabolism.
Next we investigated the mechanisms by which paqr-1 knockdown induces lipid catabolism. Lipolysis is defined as the hydrolysis of triacylglycerols (TGs) stored in cellular lipid droplets. The adipose triglyceride lipase (ATGL/ATGL-1) catalyses the initial hydrolysis step of the vast majority of TGs. Adiponectin binding to adiponectin receptors stimulates activation of AMP-activated protein kinase (AMPK) and the proliferator-activated receptor alpha (PPARα) 14 . In C. elegans the homolog of AMPK has been shown to inhibit atgl-1 26 . We find that tunicamycin does not affect ATGL-1 expression in control animals, whereas paqr-1(RNAi) appears to increase ATGL-1 protein expression under both control and ER stress conditions, to similar levels, without altering autofluorescence ( Fig. 6A and Supplementary Fig. S10). These observations suggest that in addition to the ATGL-1 lipase, other mechanisms mediate lipid droplet clearance observed during ER stress and ageing, upon PAQR-1 depletion.
A selective type of autophagy, specific for lipid droplets (also referred to as lipophagy) has been reported [27][28][29] . We examined the contribution of lipophagy on lipid droplet clearance, in PAQR-1 deficient animals. We developed a platform that allows monitoring of lipophagy in vivo by simultaneously expressing the lipid droplet-localized protein DGAT-2 fused with GFP and the autophagy specific protein LC3/LGG-1 fused with DsRed, in nematode cells and tissues of interest. Co-localization of these two proteins signifies induction of lipophagy. Using this system, we find that tunicamycin induces expression of LC3/LGG-1, confirming that general autophagy is elevated under ER stress. Notably, LC3/LGG-1 is localized in structures where DGAT-2 is also localized (Fig. 6B). Indeed, DGAT-2-positive lipid droplets were often fully surrounded by LC3/LGG-1 (Fig. 6B), indicating that selective lipophagy mediates lipid droplet clearance. These findings indicate that PAQR-1 adiponectin receptor signalling modulates lipid metabolism via lipophagy, under conditions of ER stress and during ageing. Autophagy is required for UPR ER activation while it is dispensable for enhanced survival upon PAQR-1 depletion during ER stress. Given that PAQR-1 depletion promotes ER stress-dependent survival and concomitant lipid metabolism through induction of lipophagy, we examined the requirement for autophagy in survival upon ER stress. We silenced two key autophagy components, LC3/LGG-1 and Beclin1/ BEC-1, in WT and paqr-1(tm3262) genetic backgrounds and assayed for lifespan and UPR ER induction (Fig. 7, for lgg-1(RNAi); Supplementary Fig. S11, for bec-1(RNAi)).
LGG-1 or BEC-1 deficiency reduced survival under normal and ER stress conditions both in WT animals and paqr-1(tm3262) mutants. Under normal conditions, UPR ER induction monitored by p hsp-4 GFP, was not much affected by LGG-1 or BEC-1 silencing in either WT animals or paqr-1(tm3262) mutants. However, LGG-1 or BEC-1 deficiency decreased p hsp-4 GFP expression under ER stress in both WT and paqr-1(tm3262) animals. By contrast, autophagy is not required for UPR ER activation upon PAQR-1 depletion. These findings, coupled with observations described above, indicate a reciprocal, feed-forward loop between ER stress and autophagy, with ER stress promoting autophagy and autophagy promoting UPR ER . Nevertheless, mitigation of autophagy is largely not mediating paqr-1(tm3262)-dependent effects on UPR ER activation.

Discussion
Our study directly implicates adiponectin signalling in the maintenance of proteostasis under conditions of stress, and the regulation of lipid metabolism during ageing. Understanding how distinct pathways coordinate to influence metabolism and stress resistance across different tissues at the organismal level is a critical focal point of ageing research. Adipokines are prospective mediators of intercellular and inter-tissue communication by acting both locally, in an autocrine or paracrine fashion, or systemically in an endocrine fashion 14 . Thus, the adipose tissue can communicate with distant organs and the central nervous system (CNS) bidirectionally, to modulate the ageing process. We find that PAQR-1, the closest C. elegans homolog of human adiponectin receptor 1 (AdipoR1), is an important determinant of ER stress resistance and survival under conditions of stress. Notably, the effects of PAQR-1 on ER stress resistance and survival under stress oppose to those exerted by the two other adiponectin receptors, PAQR-2 and PAQR-3, in C. elegans. Interestingly studies on energy metabolism in adiponectin receptor knock-out (KO) mice have also revealed opposing effects of AdipoR1 and AdipoR2 30 .
Our findings indicate that PAQR-1 plays an important role in reprogramming ER homeostasis. paqr-1(tm3262) mutant animals exhibited elevated basal expression of the GRP78/BiP homolog, hsp-4, and enhanced expression under ER stress, suggesting that PAQR-1 functions to negatively regulate UPR ER . Exposure to mild ER stress has been shown to have a protective "preconditioning" effect that enables cells to better adapt to severe stress stimuli. This phenomenon, termed hormesis, has been described in various organisms including C. elegans and has attracted much attention, given that slight UPR ER activation may represent an effective strategy to alleviate age-associated diseases and augment lifespan [31][32][33][34] . For example, hormesis has been shown to protect brain and heart cells under ischemia 35,36 , and to ameliorate neurodegeneration in Drosophila and mouse models of Parkinson's disease, via mechanisms involving an interplay between ER stress responses and autophagy 32,37 .
Numerous additional examples of ER hormesis, relevant to a variety of human diseases have been described, raising the possibility that hormesis can be exploited for therapeutic interventions 38 . We found that basal UPR ER activation is elevated in animals lacking functional PAQR-1. Therefore, these mutants are experiencing mild ER stress, which hormetically enhances responsiveness to acute ER stress and promotes survival. Basal UPR ER induction requires activation of the SKN-1 transcription factor.
ER stress is a potent inducer of autophagy, and autophagy has been linked to ER stress resistance 39 . C. elegans mutants lacking HPL-2, the homolog of heterochromatin protein 1 (HP1), show elevated autophagic flux following ER stress, and increased survival, which has been attributed to hormesis 31 . We found that abrogation of general autophagy by lgg-1(RNAi) markedly reduces UPR ER activation and survival following ER stress, but does not alleviate the effects of PAQR-1 depletion. This suggests that elevated basal UPR ER in paqr-1(tm3262) mutants is sufficient to engage hormetic protection. However, the beneficial effects of paqr-1 lesions on ER stress resistance and survival decreased during ageing, in association with deterioration of UPR ER , indicating that inactivation of PAQR-1 per se is insufficient to reverse the detrimental consequences of ageing when UPR ER is already impaired. In this context, ectopic expression of the spliced form of XBP-1 in aged C. elegans was shown to constitutively activate the UPR ER , increase resistance to ER stress and prolong animal lifespan 17 .
While adiponectin signalling has been postulated to play a role in age-associated pathology, there is limited information directly linking adiponectin to ageing and the regulation of longevity. The lifespan of high-fat diet-fed AdipoR1-deficient and AdipoR2-deficient mice, as well as AdipoR1/AdipoR2-double knockout mice was found to be shorter, compared to wild type 40 . In C. elegans, we find that although loss of PAQR-1 depletion does not alter lifespan under physiological conditions, it significantly improves survival under ER stress and also enhanced longevity of animals with disrupted insulin/IGF-1 signalling. These findings indicate that the C. elegans PAQR-1 receptor is a negative regulator of longevity under conditions of ER stress.
We find that both ER stress-activated transcription factors (XBP-1, ATF-4) and general stress defence transcription factors (SKN-1 and to a lesser extend DAF-16) mediate paqr-1 lesion effects on lifespan, UPR ER and survival. Although DAF-16 and SKN-1 both influence ageing and are inhibited by the insulin/IGF-1 pathway, their requirement with respect to longevity and stress responses differs 21 . Indeed, the impact of SKN-1 and/or DAF-16 silencing on stress resistance and survival upon paqr-1 lesion varied according to growth conditions (normal vs. ER stress). This indicates that differential transcription regulation by DAF-16 and SKN-1 serves to fine tune proteostasis 41 . Notably, DAF-16 and SKN-1 function varies depending on tissue/cell type and genetic background 20,[42][43][44] . Cell non-autonomous UPR ER activation plays an important role in stress resistance and ageing 17,45 . DAF-16 predominantly exerts its effects in the nervous system and intestinal cells, whereas SKN-1 is constitutively localized to ASI neuron nuclei and accumulates in intestinal nuclei in daf-2 mutant animals. Thus, DAF-16, SKN-1 and the ER specific transcription factors regulate both distinct and overlapping sets of genes, in multiple tissues, depending on, environmental cues, physiological and genetic background, to modulate ageing.
Although, autophagy has been implicated in lipid storage and turnover 46 , the impact of this association on lifespan is not fully understood. Indeed, the role of the related, selective type of autophagy, lipophagy, the potential involvement of adiponectin receptors is unknown. The intestine, which is the main metabolic tissue and serves as the main site for lipid storage in C. elegans, is the tissue most responsive to ER stress and it is highly sensitive to ER proteostasis-perturbations 17,25,47 . Our findings indicate an association between UPR ER and lipid metabolism. This association is dependent on PAQR-1 and selective lipophagy. We find that induction of lipid catabolism and lower body fat enhances survival under conditions of ER stress. Notably, accumulating evidence link chronic ER stress and UPR ER defects to lipotoxicity, manifested as cellular abnormalities and cell death due to excessive lipid accumulation in non-adipose tissue. We hypothesize that a hormetic response triggered by loss of PAQR-1 protects from lipotoxicity, promoting longevity. Adiponectin KO mice show reduced liver lipid accumulation when fed a high-fat diet, with the effects attributed partially to the down-regulation of lipogenic genes 23 . In C. elegans, we find that PAQR-1 regulates expression of the lipolytic enzyme ATGL, suggesting that adiponectin signalling balances lipolytic and lipogenic pathways. PAQR-1 depletion appears to increase ATGL-1 expression, and reduces lipid content and droplet size, indicating that PAQR-1 is a central mediator of fat accretion by serving as a relay node for nutrient sensing pathways.
In addition to its insulin sensitizing function, adiponectin has a central role in energy homeostasis, and has been proposed to serve as a starvation signal invoked upon fat reserve insufficiency. In turn, adiponectin promotes nutrient intake, decreases energy expenditure and promotes fat accumulation. Our findings show that disruption of adiponectin signalling mimics perturbed or inadequate nutrient intake, which, coupled with UPR ER , triggers catabolic processes such as autophagy, including selective ER-phagy and lipophagy, to access and mobilize internal nutrient stores. Thus adiponectin receptor inactivation may represent a mechanism to promote survival and longevity under conditions of stress. Our findings further delineate an important role for PAQR-1, the C. elegans homolog of AdipoR1, in the regulation of metazoan lifespan under stress conditions, by linking cellular metabolism and stress responses. Combined, these observations indicate that impaired adiponectin signalling may underlie reduced nutrient sensing and metabolic adaptation during ageing and stress, leading to a range of age-and obesity-associated pathologies in humans.
Lifespan analysis. Lifespan analyses were performed at 20 °C, unless otherwise indicated. Eighty to one hundred and fifty synchronous animals were scored per condition. Lifespans were performed on E. coli OP50 or HT115. For lifespan assays that involved RNA silencing, worms were exposed to RNAi from the L4 stage. Animals were transferred to fresh plates every 2-4 days thereafter and examined every day for touch-provoked movement and pharyngeal pumping, until death. Worms that died owing to internally hatched eggs, an extruded gonad or desiccation due to crawling on the edge of the plates were censored and incorporated as such into the data set. Each survival assay was repeated at least twice and figures represent typical assays. Survival curves were created using the product-limit method of Kaplan and Meier.
ER Stress resistance assays. Gravid adult WT or mutant worms were placed on nematode growth medium (NGM) plates, seeded with OP50 or HT115 bacteria, containing tunicamycin (5 μg/ml) or trapsigargin (15 μM) (Sigma-Aldrich). Control animals were incubated in an equivalent dilution of DMSO in M9. Animals were allowed to lay eggs for up to 4h. Total number of eggs was counted and compared with the number of animals that reached the L4/adult stage within 72 to 96 h at 20 °C. For each strain approximately 100-200 eggs were placed on each of four independent plates.
Heat stress survival assays. To evaluate thermotolerance, adult WT or mutant animals were placed on prewarmed (37 °C) NGM plates and incubated at 37 °C. Worms were scored for motility and provoked movement after 4, 6, 9 and 11 h. Worms failing to display these traits were scored as dead.
Lipid droplet staining. For the staining of C. elegans lipid droplets a fluorescence-based fixative method was applied. Adult WT or mutant animals were washed three times in M9 buffer and fixed with 4% paraformaldehyde for 10 min at RT, followed by two freeze thaw cycles (5 minutes at −80 °C and 5 minutes at 40 °C). Worms were again washed three times and stained with C1-BODIPY-C12 500/510 (D3823) (1 µg/ml) (Thermos Scientific, Cat# D3823) for one hour at RT. Worms were washed and mounted on coverslips for microscopic examination.
Microscopy and quantification. Worms were immobilized with levamisole before mounting on coverslips for microscopic examination with a Zeiss AxioImager Z2 epifluorescence microscope or a Zeiss LSM 710 mutliphoton confocal microscope. Images were acquired under the same exposure conditions. Average pixel intensity values were calculated by sampling images of different animals. We calculated the mean pixel intensity for each animal in these images using the ImageJ software (http://rsb.info.nih.gov/ij/). For each experiment, at least 40 animals were examined for each strain/condition and each assay was repeated at least three times. For the quantification of lipid droplets, the ImageJ software was used to assess droplet surface. A 365/420 nm filter set was utilized for the detection of the lipofuscin signal. A Zeiss SteREO Lumar V12 Stereo Microscope was also used to acquire plate images.
Ultradian rhythm monitoring. Defecation cycles of synchronized animals were scored under a stereomicroscope. For each strain at least 30 animals were scored for three cycles each. Pharyngeal pumping rate was scored by measuring grinder movements under a stereomicroscope. Approximately 30 synchronized animals were monitored and grinder movement over a period of 20 seconds were counted. The feeding rate of each animal was calculated by averaging the three measures from each animal (pumps per 20 s) and subsequently by multiplying by 3.
Statistical analysis. All experimental series were repeated independently at least three times. Data are given as mean ± SEM. Prism 5 (GraphPad Software) software package was used for statistical analyses. The log-rank (Mantel-Cox) test was used to evaluate differences between survivals and determine P values. Differences among treatment groups were determined by Student's t test and two-way ANOVA analyses. A value of P < 0.05 was considered statistically significant.