An mTORC2-Lipid Signaling Axis Controls Stress-Induced Organismal Death
Gang Wu, Wenjing Qi, Clara Essmann, Ralf Baumeister, Thomas Heimbucher
TL;DR
The study reveals how mTORC2 signaling, through lipid metabolism and a specific signaling molecule, controls whether an organism survives or dies under cold stress.
Contribution
The paper identifies a novel lipid signaling axis downstream of mTORC2 that coordinates organismal survival or death in response to stress.
Findings
mTORC2 signaling disrupts somatic lipid homeostasis and promotes organismal death through apoptosis during cold stress.
Sphingosine-1-phosphate (S1P) acts as a cross-tissue signal from lipid stores to regulate survival.
PPARα/NHR-49 repression of ASM-3 is critical for survival during cold stress.
Abstract
mTORC2 signaling plays a central role in regulating growth and survival under both physiological and stress conditions. Unlike mTORC1, however, the mechanisms by which mTORC2 integrates external nutrition or stress signals to coordinate internal metabolic homeostasis with organismal growth and survival remain poorly understood. Here, we find that mTORC2 signaling induces a decline in somatic lipid homeostasis, which in turn signals through a lipid/nuclear hormone receptor pathway that determines organismal survival or death following a severe cold stress (CS). CS disrupts somatic lipid homeostasis and induces rapid organismal death through apoptosis, a process we found to be promoted by mTORC2 and its downstream kinase SGK-1. Our study further identifies the sphingolipid metabolite sphingosine-1-phosphate (S1P) as a signal mediating cross-tissue communication from lipid stores. S1P…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsSphingolipid Metabolism and Signaling · PI3K/AKT/mTOR signaling in cancer · Phagocytosis and Immune Regulation
Introduction
The mTORC2 signalling pathway is an essential regulator of lipid homeostasis^1,2^. Compromising mTORC2 signalling leads to an imbalance in lipogenesis and lipid catabolism in tissues for lipid production and storage. A tissue-specific disturbance of lipid homeostasis can induce cross-tissue communication to ensure systemic metabolic adjustments in an organism^3^. Such inter-organ communication pathways are mediated by a variety of bioactive factors including lipid-derived signalling molecules^4^. Bioactive lipid molecules serve as metabolic messengers for predominantly communicating the energy status of metabolically active tissues to coordinate physiological adaptation in response to metabolic or external stress. The inter-organ cross-talks and cellular effectors of bioactive lipids are poorly understood and subject to intensive research.
PPARs (peroxisome proliferator-activated receptors) are transcription factors whose activities are typically modulated by lipid-derived ligands. As such, they are considered as key lipid sensors and master regulatorsof lipid metabolism^5^. For instance, PPARα is induced during fasting to promote lipid catabolism and plays an essential role in hibernating animals during seasonal cold exposure^5–7^. PPARα is activated by fatty acids (FA) and their derivatives, including oleoylethanolamide (OEA), palmitoylethanolamide (PEA) and phosphocholines (PCs)^4,8^. In addition to its metabolic role in liver and adipose tissues, PPARα is expressed in the brain and mediates protective functions in stress response and neurodegeneration^9^. However, how stress promotes neuronal protection by PPARα is largely unknown.
PPARα and hepatocyte nuclear factor 4 alpha (HNF4α) are functionally represented by NHR-49 in C. elegans. NHR-49 acts as a master regulator for lipid metabolism in C. elegans, promoting expression of genes involved in processes such as fatty acid beta oxidation and fatty acid desaturation^10^, which are critical for survival during CS conditions. NHR-49 is implicated in regulating diverse physiological responses under various conditions and stresses, suggesting that precise modulation of its activity is essential. Similar to PPARα, NHR-49 may be activated by specific secreted fatty acids and their derivatives^3,11–13^. Among potential NHR-49 ligands, the sphingolipid metabolite sphingosin-1-phospate (S1P) has previously been proposed, based on an in silico docking approach, as a putative ligand that may bind to the ligand-binding domain of NHR-49^14^. However, further functional validation is required to confirm this interaction. Moreover, although several candidate ligands for NHR-49 have been identified, their physiological roles in mediating stress responses remain uncharacterized.
The ceramide/sphingolipid pathway and its metabolites are major mediators of cellular stress responses and balance cell fate either towards cell survival or cell death^15–17^, including organismal death in C. elegans in temperature-stressed animals^18^. The sphingolipid metabolite S1P not only contributes to membrane fluidity in C. elegans and mice^19^, but is also distributed across tissues and has been suggested to mediate neuro-protective effects in both parents and progeny^20^. Acid sphingomyelinase (ASM), which hydrolyses sphingomyelin (SM) to generate ceramide^25^, can be secreted to the cell surface to activate neuronal apoptosis upon irradiation. This process requires the phosphorylation of ASM by the protein kinase C-delta (PKCδ) on a specific serine residue (S508)^21,22^. ASM mediated ceramide generation on the cell surface can induce apoptosis in endothelial cells of the central nervous system following radiation-induced stress^23,24^.
The target-of-rapamycin (TOR) pathway forms two distinct complexes, mTORC1 and mTORC2 which are evolutionarily conserved and play crucial roles in controlling multiple biochemical processes in response to nutrient availability and stress^25,26^. In C. elegans, mTORC2 controls a soma-to-germline lipid transport through activating the currently only bona-fide target of mTORC2, the serum-and-glucocorticoid-inducible kinase 1 SGK-1^27^, which in turn promotes the expressing vitellogenin lipid transporter genes^28^. Sequence and pathway conservations suggest that SGK-1, like its yeast and mammalian homologs, requires a second activating phosphorylation by PDK-1, indicating that SGK-1 serves as a logical AND gate that is only fully active when it receives dual inputs from both insulin/IGF and mTORC2 signaling^29^. Under normal growth conditions, SGK-1 controls vitellogenesis, a process that is also critical for lipid transfer from mothers to progeny during severe stress conditions. It has been reported that when severe CS is applied during early adulthood, C. elegans exhibits degeneration in multiple tissues, including neurons, the gonad and the intestine^30–32^. Following a short recovery phase following CS, a lipid reallocation from somatic tissues to developing oocytes was observed which improves survival of the offspring at the expense of maternal depletion of lipid storage and a rapid maternal death during the early recovery period^30–32^. Given the role of mTORC2/SGK-1 in regulating vitellogenesis and stress responses, we hypothesized that mTORC2/SGK-1 may also play a critical role in modulating the organismal response to severe CS.
Many organisms face trade-offs between parental and offspring survival, depending on the severity of stressors. Mild stress usually promotes somatic lipid consumption to maintain cellular functions at the cost of a temporary reduction in reproduction, whereas severe stress often triggers lipid transfer to progeny, risking maternal energy depletion and organismal death^33–37^. Details of how stress levels are sensed and evaluated, and most of the signaling molecules, downstream effectors, and mechanisms that link stress to maintaining or sacrificing organismal integrity remain enigmatic. Lipid stores are a vital energy source for surviving the period after CS, and in many species key resources are allocated to offspring under extreme conditions^36,37^. Moreover, lipid-derived metabolites generated during lipid catabolism also can function as signaling molecules, coordinating the metabolic status across tissues and making them strong candidates as systemic stress response mediators. Although several lipid signaling pathways have been identified in C. elegans, the physiological conditions that activate these pathways remain poorly understood^3,11,20,38^.
In this study, we found that compromising SGK-1 or the mTORC2 component RICT-1 in intestinal cells results in an activation of the nuclear receptor NHR-49 in both neurons and the intestine to promote survival following CS. Improved post-CS survival requires the signalling sphingolipid S1P that accumulates in sgk-1(lf) mutants and activates NHR-49 both cell-autonomously (in the intestine) and cell-nonautonomously (in neurons). Additionally, we identified NHR-49 as a transcriptional repressor for the expression of ASM-3, which can be secreted from intestinal cells to induce neuronal beading and germline apoptosis. In cold-shocked wild-type animals, SGK-1 activity alters lipid metabolism, affecting inter-organ signalling that induces cellular damage and ultimately organismal death.
Results
mTORC2/SGK-1 regulates post CS recovery through controlling NHR-49 activity
Recovery of C. elegans from an extended cold period induces the reallocation of lipid resources from the intestine, the main tissue of lipid metabolism and storage, to the germline oocytes, comprising the next generation. This ultimately improves the survival of progeny at the expense of that of the mother^31^. Such lipid transfer requires the activity of vitellogenins, apolipoprotein-like lipid transporters. Vitellogenin expression in C. elegans during standard growth conditions is regulated by mTORC2/SGK-1 signaling^28,39^. We wondered whether loss of this signaling control might prevent lipid reallocation also after CS and, as a consequence, would improve maternal survival. To test this hypothesis, we exposed wild-type and sgk-1(lf) animals to a 2°C cold stress (CS) for 6 hours, followed by a recovery period at 20°C. Compared to wild-type, sgk-1(lf) mutant survived considerably longer (Fig. 1a), and survival correlated with the retention of somatic lipids during recovery stages, as indicated by ORO staining (Fig. 1b,c). Since sgk-1 and rict-1 loss-of-function mutants displayed a similar phenotype during larval development and in lipid metabolism, indicating an epistatic relationship of both factors^40,41^, we assumed the loss of mTORC2 signaling results in similar phenotype of sgk-1(lf) following a CS. rict-1(lf), like sgk-1(lf), also improved adult survival after CS profoundly (Fig. 1d). Therefore, in agreement with a recent report^31^, we conclude that depletion of somatic lipids after CS is correlated with a reduced survival, and this may be mediated by the activity of mTORC2/SGK-1 signaling. How maintenance of somatic lipid depots contributes to a better post-CS recover, however, was left unanswered.
Our own previous studies had suggested that lipid retention in sgk-1(lf) mutants might promote recovery of adult animals after CS^6^. We had speculated that retention of bulk lipids could be used as energy source via ß-oxidation immediately after termination of CS, since CS at 2°C is probably a biologically inactive state and recovery post CS is an energy consuming process^6^. Alternatively, there is precedence that specific fatty acids might also serve as signalling molecules in the adult body and trigger yet to be identified recovery pathways^38^. To distinguish these two situations, we first inactivated by RNA interference the expression of acs-2, encoding a rate-limiting fatty acid CoA synthetase, and the ech-1.1 enoyl-CoA hydratase gene, two enzymes involved in ß-oxidation. Knock downs of both genes only partially reduced the post-CS survival of sgk-1(lf) (Fig. 1e,f), which suggests that retained lipids in adult animals may fulfil functions other than serving only as energy resources. To mimic and identify individual lipid molecules contributing to extended survival of sgk-1(lf), we supplemented the food of wild-type with various fatty acids. We found that, surprisingly, stearic acid (SA) improved post-CS survival of wild-type animals to an extent similar to that observed by sgk-1(lf); however, SA did not further extend the already long survival of sgk-1(lf) (Fig. 1g). Conversely, its dehydrogenation product, oleic acid (OA), did not improve post CS recovery in wild type (Fig. S1a). This result was unexpected and somewhat counter-intuitive because, in addition to acting as an energy source, unsaturated fatty acids like OA, but not saturated fatty acids like SA, had previously been suggested to promote survival at low temperature due to their ability to increase membrane fluidity^42^. We conclude that, based on these results, SA may more likely have a regulatory role in CS recovery, rather than simply serving as an energy resource.
Previous results already had indicated that small lipid molecules, including SA, may function as signaling molecules during stress, and can trigger stress response pathways in several species^38^. By binding to LBP-5, SA may also activate the nuclear hormone receptor NHR-49^12^ that also triggers responses to various stress conditions, including infection, mitochondrial, and oxidative stress^10^. Therefore, we speculated that NHR-49 activation may contribute to post-CS survival of animals with compromised mTORC2/SGK-1 signalling. In agreement with this hypothesis, the improved post-CS survival of sgk-1(lf), rict-1(lf) and SA-treated wild-type was substantially suppressed in an nhr-49(lf) or nhr-49 RNAi background (Fig. 1h–j). Our data suggest that nhr-49 suppression by rict-1 and sgk-1 to promote CS recovery is mediated by an alteration in the lipid homeostasis of the intestine. We noticed that inactivation of nhr-49 also moderately reduced the survival of wild-type animals after CS (Fig. 1i), suggesting that during recovery of wild-type, NHR-49 may retain a basal level activity. To examine the possibility that loss of nhr-49 induces sickness, we also tested the contribution of nhr-49 to extended survival of another mutant, daf-2(lf). nhr-49 RNAi only moderately reduced the improved survival of daf-2(lf) mutants, suggesting that loss of nhr-49 does not induce sickness and may act in parallel to the insulin/IGF effector DAF-16 (Fig. S1b). In contrast to loss of nhr-49, the nhr-49(et8) gain-of-function allele that does not increase lifespan^14^, nevertheless strongly improved CS recovery and mimicked the effect of sgk-1(lf), but did not further improve recovery of sgk-1(lf) (Fig. 1k). These results further suggest that NHR-49 plays a dedicated role during CS recovery when mTORC2/SGK-1 signalling is compromised.
To monitor if NHR-49 transcriptional activity is altered in an sgk-1(lf) mutant, we inspected worms harbouring acs-2p::gfp, a transcriptional reporter for NHR-49 activity^43^. GFP was induced in sgk-1(lf) mutants, and suppressed after knocking down nhr-49 (Fig. S1c,d). Additionally, we quantified mRNA levels of several NHR-49 target genes and found that their transcript levels were increased in sgk-1(lf) mutants in an nhr-49 dependent manner (Fig. 1l). We conclude that loss of sgk-1 results in a transcriptional activation of NHR-49 which in turn is beneficial for recovery after CS. NHR-49 may partner with additional NHRs, including NHR-80, to regulate target gene expressions^42^. Similar to nhr-49(lf), loss of nhr-80 also abolished the improved post-CS survival of sgk-1(RNAi) animals (Fig. 1m). In conclusion, our results suggest that during CS recovery, mTORC2/SGK-1 may be active to negatively regulate survival of mothers and to boost progeny survival^31^, whereas inactivation of mTORC2/SGK-1 signaling promotes post-CS survival of the mothers via induction of NHR-49 and its interaction partner NHR-80.
SGK-1 signals to NHR-49 in both a cell autonomous and a cell non-autonomous manner
Both nervous system and intestine in C. elegans play crucial roles in sensing environmental stressors and facilitating lipid mobilization under stress. Whereas NHR-49 is expressed ubiquitously, SGK-1 can be mainly detected in neurons and intestine where SGK-1 is differently regulated, suggesting distinct functions^27^. To identify critical tissues for CS recovery, we tested sgk-1 through tissue specific RNAi. Results indicated that neuronal knock down of sgk-1 did not influence the post CS recovery, while intestinal knock down of sgk-1 phenocopied the effect of the sgk-1(lf) mutant (Fig. 2a,b). Next, we performed tissue-specific rescue experiments for nhr-49. Both neuronally and intestinally expressed nhr-49 rescued the post CS survival of nhr-49(lf); sgk-1(RNAi) animals (Fig. 2c–e), whereas hypodermal, as well as muscular or endogenous nhr-49 expression, failed to rescue short CS survival of the nhr-49(lf) single mutant (Fig. 2c,f–h). Intestinal NHR-49 additionally rescued the reduced lipid levels of nhr-49(lf); sgk-1(lf) double mutants, whereas neuronally expressed NHR-49 reduced lipid levels already in the absence of CS both in wild-type and sgk-1(lf) backgrounds (Fig. 2i,j, Fig. S1e,f).
As shown above, reducing ß-oxidation only moderately reduced sgk-1(lf) CS recovery (Fig. 1e,f), indicating that both lipid catabolism and other effects contribute to sgk-1(lf) survival after CS. The intestine is considered the prime tissue for lipid storage and mobilization, and both aspects are regulated by nhr-49. We therefore hypothesized that down-regulating ß-oxidation genes should have stronger effects in reducing the long CS recovery of sgk-1(lf) that only express nhr-49 in intestine cells, since NHR-49 activity in the neurons would be responsible for the unidentified beneficial effects other than lipid homeostasis. However, knock down of either acs-2 or ech-1.1 only partially reduced the post-CS survival of nhr-49(lf); sgk-1(lf) double mutants in which nhr-49 expression had been reconstituted only in the intestine (Fig. 2k,l). This suggests that NHR-49 activation in the intestine of sgk-1(lf) might also induce unidentified target genes that are not directly involved in maintenance of lipid homeostasis.
Loss of sgk-1 retains a putative NHR-49 ligand in the intestine, which can activate NHR-49
Although many NHR-49 target genes have been identified in various stress scenarios, little is known about how NHR-49 itself is regulated^10^. There is however currently no indication that the kinase SGK-1 directly phosphorylates NHR-49 since, unlike other putative targets DAF-16, SKN-1, or PQM-1, NHR-49 does not harbour a conserved AKT/SGK phosphorylation motif, R-X-R-X-X-S/T (Fig. S2a). Furthermore, NHR proteins seem to require ligands for activation, and several small signalling lipids have been described as potential NHR-49 ligands, including OEA, DGLA, SA and S1P^3,11,12,19^. We reasoned that retained lipids in sgk-1(lf) may serve as ligands to induce NHR-49 in maternal somatic tissues, including the intestine and neurons. We therefore next test whether either neuronal or intestinal NHR-49 could be induced by SA supplemented diet, and we reasoned that this could improve post-CS recovery. Dietary SA supplements improved the post-CS survival of wild-type animals that expressed nhr-49 only in the intestine, but had no effect on animals with only neuronal nhr-49 expression (Fig. 3a–c). We therefore suggest that SA accumulation may induce NHR-49 activity in the intestine, whereas other NHR-49 ligands seem to be required in the nervous system. Neither OEA nor DGLA food supplements had effects on wild-type CS recovery (Fig. S2b). Consistently, knock down of either lbp-3 or lbp-8 encoding OEA and DGLA binding proteins only moderately diminished the extended CS recovery of sgk-1(lf) animals in which nhr-49 is considered active (Fig. S2c,d). This suggests that neither LBP-3, LBP-8, nor their target lipids OEA and DGLA are likely involved in neuronal NHR-49 activation under CS conditions.
Sphingosine-1 phosphate (S1P) has been shown to activate the human PPARγ and regulates NHR-49/PPARα subnuclear localization in C. elegans^19^. S1P levels may well be affected by SGK-1 function, since the SGK-1 yeast homologue Ypk1 has been reported as a regulator of sphingolipid metabolism^44^. Furthermore, S1P shuttles via vitellogenin lipid transporters from the parental intestine to the germline oocytes to confer a transgenerational protective effect on neurons^45^. Dietary S1P significantly extended the post-CS survival of wild-type, and this required NHR-49, since it is suppressed in nhr-49(lf) (Fig. 3d). In addition, S1P supplementation did not further improve the post-CS survival of sgk-1(lf) animals suggesting an already increased S1P level in sgk-1(lf) animals (Fig. 3e). To our surprise, dietary S1P substantially improved the CS recovery of nhr-49(lf); sgk-1(+) strains that either intestinally or neuronally express nhr-49 (Fig. 3f,g). We therefore suggest that S1P, rather than SA, may boost NHR-49 activity in both tissues and, to do so, overwrites inhibition by sgk-1. Next, we quantified S1P levels from sgk-1(lf) animals using an S1P antibody-based ELISA^46^ and detected a 3.3-fold increase in sgk-1(lf) mutants compared to wild-type (Fig. 3h). Consistently, transgenic expression of the lipid transporter gene vit-2 was sufficient to reduce S1P levels in sgk-1(lf) animals (Fig. 3h), indicating that S1P accumulation is at least partially due to reduced vitellogenin expression in sgk-1(lf). To further examine whether S1P activates NHR-49, we used acs-2p::GFP reporter. Like in sgk-1(lf) animals, dietary supplementation with S1P also induced acs-2p::gfp in an otherwise wild-type background, and this induction is dependent on nhr-49 (Fig. 3i,j). Since our results indicate an accumulation of S1P in somatic cells rather than in oocytes upon the loss of sgk-1 we used a visually traceable S1P–fluorescein to directly monitor the distribution of S1P in somatic and germline tissues^20^. Consistent with the previous report^20^, S1P–fluorescein could be detected in oocytes of wild-type animals, indicating a cross-tissue signalling activity of S1P, since the prime tissue to resorb dietary S1P–fluorescein is the intestine. Compared to wild-type animals, loss of sgk-1 diminished the S1P–fluorescein signal in the oocytes of the proximal gonad (Fig. 3k,l), but elevated S1P–fluorescein in the maternal intestine (Fig. 3m,n). Although S1P-fluorescein signals were too weak to be detected in neurons, the activation of neuronal NHR-49 by ingested S1P (Fig. 3g) suggests that it may also be transported from the intestine to neurons. Although we consider it less likely, we cannot exclude at this moment a more indirect mode of activation of neuronal NHR-49 by intestinal S1P.
We next aimed to reduce S1P levels through interfering with S1P production. Sphingosine kinase 1, encoded by sphk-1, is currently the only known enzyme producing S1P in C. elegans. Loss of sphk-1 significantly reduced the post-CS survival of sgk-1(lf) animals. In contrast, it did not reduce recovery of wild-type animals, in which somatic levels of S1P, as shown above, are relatively low (Fig. 3o). In summary, our results suggest that one of the consequences of lipid retention in sgk-1(lf) is the accumulation of S1P in maternal tissues that may contribute to NHR-49 activation in both intestine and neurons to promote post-CS survival.
asm-3 repression by NHR-49 contributes to cell non-autonomous protection of neuron integrity in response to CS
C. elegans becomes temporarily and reversibly immobilized and paralyzed during extreme cold periods. Extended CS can also induce neuronal abnormalities, including dendritic protrusions and neuronal beading/blebbing, which can easily be visualized in multidendritic neurons like the PVDs^32^. PVDs are a pair of nociceptors that respond to harsh touch and reduced temperature (15°C)^47^. We noticed that sgk-1(lf) mutants during CS recovery regained mobility faster than WT, and that this recovery also depends on nhr-49 (Fig. 4a). We reasoned that one neuronal function of NHR-49 might therefore be to protect neurons from CS-induced functional, and maybe structural, damage. To examine this hypothesis, we monitored PVD dendrite morphology immediately after CS. PVD morphology was better preserved in sgk-1(lf) mutants than in wild-type, which indicates that neurons in the sgk-1(lf) mutants are somehow protected from detrimental consequences of CS (Fig. 4b,c). Loss of nhr-49 eliminated this protection (Fig. 4d), but surprisingly intestinal, but not neuronal, nhr-49 expression rescued PVD defects in nhr-49(lf); sgk-1(RNAi) background (Fig. 4e,f). From this data we conclude that NHR-49 activity in the intestine may activate the transcription of a signal to protect neurons from the detrimental consequences of CS. This signal is likely different from S1P, which we previously proposed to be activated in neurons of sgk-1(lf) animals and to act downstream of NHR-49.
We next analysed the transcriptome changes in sgk-1(lf) animals by performing RNA-seq. Cold stressed wt, sgk-1(lf), and nhr-49(lf); sgk-1(lf) animals were used for the analysis. After 6 hours CS and 2 hours recovery 342 genes were up- and 147 were significantly down-regulated in sgk-1(lf) (Fig. 4g,h) (Table S1). A subset of those genes was suppressed in nhr-49(lf); sgk-1(lf) double mutants, which we defined as potential nhr-49 targets (Fig. 4g,h). For these candidates a GO analysis was performed (Fig. S3a) (Table S1). For their functional contribution to PVD neuronal beading we performed an RNAi screen. Since C. elegans neurons are refractory to RNAi, our RNAi screen preferentially targeted genes that are expressed in non-neuronal tissues to regulate neuronal integrity. One candidate we obtained was asm-3. asm-3, an ortholog of human SMPD1, encodes one of the three acid sphingomyelin diesterases in C. elegans which converts sphingomyelin to ceramide. asm-3 expression was repressed in sgk-1(lf) mutants, and repression was relieved in nhr-49 (Fig. 4i), suggesting that low ASM-3 levels correlate with protection of PVD morphology after CS. Correspondingly, RNAi knock-down of asm-3 expression in wild-type had a pronounced protective effect on PVD neurons after CS (Fig. 4j). To further distinguish between cell autonomous and non-autonomous ASM-3 functions to protect PVD morphology, we performed tissue specific knock-down of asm-3. Neuron-specific asm-3 RNAi was not sufficient to protect PVD neurons from CS-induced beading (Fig. 4k), while an intestine-specific knock down of asm-3 conferred a protective effect in cold stressed animals (Fig. 4l). Our data argue in favour of intestinal asm-3 regulating neuronal beading of PVD neurons in CS, suggesting a cell non-autonomous function of ASM-3.
Since the RNAi experiments above cannot fully exclude a contribution of neuronal asm-3 for survival, we designed a complementary experimental approach. For this, we artificially expressed asm-3 in sgk-1(lf) animals by driving its expression from the heterologous (SGK-1 and NHR-49 independent) intestinal promoter gly-19. Intestine-specific expression of asm-3 was also sufficient to suppress PVD protection of sgk-1(lf) animals (Fig. 4m), again indicating that intestinal ASM-3 is detrimental for cold-stressed PVD neurons. Knock-down of asm-3 was sufficient to increase the post-CS survival of WT animals without further improving survival of sgk-1(lf) mutants. Conversely, transgenic asm-3 expression only in the intestine reduced the post CS recovery of sgk-1(lf) mutants (Fig. 4n,o). Our data support a role of ASM-3 downstream of the SGK-1/NHR-49 signalling axis to neuronal integrity which correlates with improved CS recovery. Since both nhr-49 and asm-3 intestinal expression suffices to impact on neuronal integrity, we suggest that either ASM-3 is secreted from the intestine to signal across tissues, or ASM-3 triggers the release of a yet to be identified molecule to stimulate neuronal beading.
ASM-3 functions in the intestine to induce neuronal beading via apoptosis
Studies from mammalian systems indicate that, in addition to cell-autonomous functions, ASM variants can indeed be secreted. As a component of the sphingolipid pathway, secretory ASM (S-ASM) contributes to increased ceramide levels at the outer leaflet of the plasma membrane that may induce cell death of neuronal cells and other tissue after irradiation and pathogenic infection^48,49^. Since neuronal beading has been recognized as an intermediate and reversable state of neuronal cell death^50^, we hypothesized that ASM-3-induced neuronal beading/blebbing in C. elegans may indicate an initial state of neuronal cell death as a consequence of CS. If this were the case, then blocking programmed cell death should stall or even inhibit neuronal beading. To test this, we performed systemic (which we consider to be preferentially active in non-neuronal cells) and neuron-specific knock down of ced-4, an orthologue of Apaf-1 and critical component of the apoptotic pathway. Systemic ced-4(RNAi) had no effect on PVD beading after CS, while neuron-specific knock down of ced-4 expression protected PVD neurons after CS (Fig. 5a–c). Next, we reasoned that PVD beading induced by intestinal ASM-3 might activate a neuronal apoptosis pathway, involving ced-4. Since in sgk-1(lf) animals asm-3 expression is shut off, we re-expressed transgenic asm-3 only in the intestine of sgk-1(lf), and then performed neuron-specific ced-4(RNAi). In this genetic setting, ASM-3 induced PVD beading was efficiently reduced by CED-4 inactivation (Fig. 5d, Fig. S4a). We conclude that after CS, intestinally expressed ASM-3 is sufficient to induce PVD neuronal damage via induction of a CED-4 pathway, whereas in sgk-1(lf) animals are protected due to NHR-49 repression of intestinal asm-3.
Speculating that ASM-3 might be secreted from intestinal cells, we reasoned that in principle it should also be able to target tissues other than neurons in response to CS. In C. elegans, germ cells are prone to apoptosis, both in response to developmental regulation and, particularly, upon stress^51^. Increased apoptotic germ cells were detected in wild-type animals after CS, and this increase was suppressed in sgk-1(lf) (Fig. 5e,f). This suggests that inactivation of sgk-1 is beneficial for the survival of germline nuclei. Both pan-organismal and intestinal knock-down of asm-3 significantly reduced germ cell apoptosis (Fig. 5g,h), whereas neuron-specific asm-3(RNAi) was ineffective (Fig. 5i). Transgenic expression of asm-3 in the intestine partially suppressed the anti-apoptotic effect of sgk-1(lf) in germ cells in a ced-4 dependent manner (Fig. 5j,k). Since a reduction in the steady-state levels of apoptotic germ cells could also be the consequence of accelerated clearance of cell corpses, in an alternative approach we quantified apoptosis in mutants in which engulfment is blocked. CED-1 and CED-5 are downstream effectors of two independently working pathways involved in clearance of cell corpses in C. elegans^52^. Thus, in a ced-1(lf); ced-5(lf) double mutant cell corpse engulfment is strongly reduced^53,54^. Interestingly, the ced-1(lf); ced-5(lf); sgk-1(lf) triple mutant did not display an increased number of apoptotic germ cells. We conclude that levels of apoptosis induction are already low in sgk-1(lf) (Fig. 5l). Taken together, in response to CS, SGK-1 maintains ASM-3 expression that may act cell non-autonomously to induce CED-4 and cell death in various tissues. Inactivating sgk-1 results in loss of asm-3 expression, protecting both neurons and germ cells from detrimental consequences of CS. In summary, the CS induced induction of CED-4 might contribute to a pan-organismal CS vulnerability, enhancing the likelihood of death during recovery. In support of this hypothesis, knock-down of ced-4 strongly reduced the extremely high death rates of wild-type during the initial two days of CS recovery (Fig. 5m). In stark contrast, knock down of ced-4 had no effect in an sgk-1(lf) background (Fig. 5m). Transgenic sgk-1(lf) animals expressing asm-3 only in the intestine showed reduced survival post-CS, and this reduction was dependent on ced-4, suggesting that intestinal ASM-3 in this setting induced apoptosis after CS (Fig. 5n,o). Our results point to an important role of sgk-1 in modulating intestinal ASM-3 expression for a cell non-autonomous induction of apoptosis in response to CS.
CS results in ASM-3 secretion to trigger neuronal beading, germ cell apoptosis and organismal death
Finally, we tested the hypothesis that ASM-3 itself may act as a secreted signal to trigger apoptosis by cell non-autonomous mechanisms. This model is based on mammalian studies, in which it was shown that ASM can be secreted in response to irradiation and infection to induce apoptosis^55^. Phosphorylation of S508 in ASM is required for its translocation to the cell surface and secretion, and is conserved in the C. elegans ASM-3 (Fig. S5a). Mutating this site to alanine reduced ASM secretion without influencing its basal level of enzymatic activity^48^. We constructed ASM-3::EGFP and expressed it from an intestine-specific promoter to monitor first whether we could detect the fusion protein either in tissues other than intestine, or in the extracellular space. In the absence of CS, ASM-3::EGFP levels were barely detectable in the intestine cells or in coelomocytes, scavenger cells that endocytose secreted molecules from the pseudocoelom. However, after 6 hours CS and 3 hours recovery, a strong ASM-3::EGFP signal was observed in the coelomocytes indicating their secretion from intestine (Fig. 6a,b). To experimentally interfere with ASM-3 secretion, we mutated two adjacent S/T sites corresponding to S508 in mammalian ASM to alanines, to generate ASM-3(AA). These mutations eliminated CS induced GFP signals in coelomocytes, suggesting that this ASM-3 variant is no longer capable of being either secreted and/or endocytosed (Fig. 6c,d). Next, we verified in vitro that the acid sphingomyelinase activity of purified ASM-3(AA)::EGFP was comparable to that of wild-type (Fig. 6e). We then examined whether the non-secreted ASM-3(AA) variant still induces neuronal beading/blebbing and germ cell apoptosis. Intestinal expression of ASM-3(AA), in comparison to ASM-3(wt), displayed reduced PVD dendritic beading in wild-type worms and sgk-1(lf) mutants under CS (Fig. 6f), suggesting that its secretion is required for the neuronal phenotype. Consistently, transgenic ASM-3(AA) is also incapable of inducing germline apoptosis, in contrast to ASM-3(wt) (Fig. 6g). Noticeably, in wild-type background in which we considered asm-3 to be active, transgenic ASM-3(AA) even reduced germline apoptosis to some extent (Fig. 6g). We at last looked at how non-secretory ASM-3(AA) influence post-CS survival. Unlike ASM-3(wt) that can be secreted, the non-secretory ASM-3(AA) derivative moderately increased the post-CS survival of sgk-1(lf) mutants, and also significantly extended the post-CS survival of wild-type animals (Fig. 6h). We consider the beneficial effect of ASM-3(AA) on germ cell apoptosis and post-CS survival a consequence of a potential competition of ASM-3(AA) with endogenous ASM-3 activity. In summary, intestinal ASM-3 may be secreted to affect PVD neuronal beading and apoptosis, ultimately together modulating organismal death after a severe CS.
Discussion
Our study reveals that the mTORC2/SGK-1/NHR-49 axis relays an intestine-to-neuron cross-talk to coordinate a systemic CS response. As a key downstream effector in this pathway, NHR-49 senses somatic lipid availability and determines maternal fate following CS by regulating ASM-3 expression and apoptosis (Fig. 6i).
In a natural habitat, famine and predators are among the main causes of early death. Upon limited food availability, animals exposed to severe stresses have evolved two major strategies to maximize reproductive success. One strategy prioritizes parental recovery and survival, at the cost of temporarily reducing reproduction^36,37^. The alternative strategy favours progeny survival, even if it requires depleting parental energy stores and ultimately results in parental death^56,57^. The molecular programs that assess stress severity and determine the appropriate reproductive strategy remain poorly understood. Under severe circumstances like CS, C. elegans animals seemingly prioritize lipid transport from the maternal organism to the progeny, to protect the offspring at the expense of the parent^31^. Using this CS model, we identified the mTORC2/SGK-1/NHR-49 signaling axis as a potential evaluator of both stress intensity and maternal lipid availability. In turn, this pathway determines maternal fate during the early recovery phase following CS.
In C. elegans, lipid transfer from somatic parental tissues to germline oocytes is mediated by vitellogenins. Under non-stress conditions, the production of these lipid transporters is regulated by SGK-1. Our data suggest that the same regulatory cascade also operates during the recovery phase following periods of severe cold. SGK-1, like its yeast homolog Ypk1, functions as a logical AND-gate that integrates binary phosphorylation signals from two upstream kinases: PDK-1, a key component of insulin/IGF signaling, and the mTORC2/RICT-1 complex^27,41^, both of which are central to sensing and responding to nutrient availability and cellular stress. Simultaneous activity of both insulin/IGF signaling and the mTORC2 pathway is therefore a prerequisite for maintaining SGK-1 in an active state. This positions SGK-1 as a promising candidate for the critical switch governing different reproductive strategies after stress. Supporting this notion, our data indicate that in cold-stressed wild-type animals, SGK-1 is likely active and promotes lipid depletion from somatic cells, likely through lipid transport towards embryos^31^. In contrast, compromised upstream activation—either through rict-1 mutations or SGK-1 inactivity—preserves maternal lipid stores. Although SGK-1 also acts in insulin/IGF signaling, the improved CS survival of a daf-2(lf) insulin receptor mutant most likely signals through an sgk-1 independent branch of this pathway, most likely involving AKT-1 and AKT-2, since it does not require NHR-49.
Our data suggest that lipid retention under compromised mTORC2/SGK-1 signaling benefits the mother in two key ways: it ensures sufficient energy supply for the metabolically demanding recovery process and may promote the production of lipids with potential signaling functions, such as SA and S1P. Members of the nuclear hormone receptor family are known to respond to various lipid ligands^58 Chemical reviews^, and NHR-49 has previously been shown to be activated by several small lipid molecules^10^. Our results suggest that NHR-49 is activated by both SA and S1P, likely acting primarily in intestinal cells, the predominant tissue synthesizing and storing these lipid ligands^3,59^. Together, nutrient surplus and NHR-49 induction contribute to maternal recovery and enhanced survival following cold stress^38^.
SGK-1 functions in intestinal cells to negatively regulate NHR-49 activity in both neurons and intestine, which suggests both cell-autonomous and non-autonomous signaling mechanisms. While non-autonomous regulation likely occurs indirectly via the production or distribution of lipid mediators such as S1P and SA, cell-autonomous regulation may involve direct molecular interactions between SGK-1 and NHR-49. Although NHR-49 does not contain a canonical phosphorylation motif recognized by AGC kinases, such as those found in SKN-1 or DAF-16, we cannot rule out the possibility that NHR-49 is a direct phosphorylation target of SGK-1.
Our tissue-specific analyses revealed that SGK-1, as an integrator of nutritional and stress cues, functions primarily in intestinal cells—not neurons—to regulate lipid distribution and survival following cold stress. This suggests that extreme cold may be sensed predominantly by non-neuronal tissues such as the intestine. Notably, the molecular sensors that detect near-freezing temperatures, which C. elegans likely encounters in its natural environment, remain largely unexplored. Most studies on “cold” sensors in C. elegans have focused on temperatures within the lower physiological range for reproduction, typically around 15 °C. TRPA-1, a Ca^2+^-sensitive TRP channel, GLR-3, a kainate-type glutamate ionotropic receptor, and OCTR-1, an octopamine receptor, have all been implicated in responses to cooling to 15 °C^60–62^. However, whether these sensors are also capable of detecting more extreme cold, such as the 2 °C conditions used in our study, remains unclear. Studies on these proteins consistently point to a neuronal origin of function, implying that cold-induced intestinal responses via the mTORC2/SGK-1 pathway would require signal relay from the nervous system^63^. However, cell-type specific RNA-seq analyses have revealed that trpa-1 is also expressed in intestinal cells, in addition to PVD neurons^64,65^, positioning it as a promising candidate for sensing severe cold directly in the intestine. In addition to TRPA-1, another potential cold sensor that might functions in non-neuronal tissues to perceive cold is PAQR-2, a homologue of the mammalian adiponectin receptor. PAQR-2 has been identified as a membrane fluidity sensor and is expressed in multiple tissues in C. elegans. As membrane fluidity declines markedly at lower temperatures, PAQR-2 also acts as a sensor of membrane stress at 15°C^42,66^. Under such conditions, PAQR-2 activates NHR-49 to promote the expression of fatty acid desaturase genes, thereby restoring membrane fluidity^42^. Since our results suggest that mTORC2/SGK-1 also functions upstream of NHR-49, it will be of interest to determine whether PAQR-2 acts upstream of, or in parallel with, mTORC2/SGK-1 in the intestine to mediate responses to severe cold.
We identified ASM-3 as a critical downstream mediator of NHR-49 in the intestine, regulating a neuronal beading/blebbing phenotype non-cell autonomously through its secretion. Additionally, ASM-3 induces a systemic apoptotic response after a severe CS. This finding aligns with mammalian studies demonstrating that ASM is secreted in response to stress^67^. In humans, secreted ASM contributes to the biological effects of tumor necrosis factor-α (TNF-α), including the induction of apoptosis and programmed cell death^68^.
Under unfavourable conditions, particularly when somatic lipid stores are depleted during recovery from cold stress, the absence or unavailability of ligands such as SA or S1P results in failure to activate NHR-49, which in turn favours the expression and secretion of ASM-3 by the intestine and potentially other somatic tissues. As demonstrated in our study, secreted ASM-3 acts on distant tissues—including the nervous system and germline—to trigger programmed cell death. These findings lend support to the hypothesis that, in human therapeutic contexts, secretion of ASM from endothelial cells may underlie the apoptotic cell death observed in central nervous system following a single high-dose radiation treatment^69,70^.
What might be the evolutionary advantage of a stress-induced suicide program in an organism? One possibility is that selective apoptosis of individual cells facilitates the release of nutrients via autophagy, thereby supporting residual parental investment in the progeny. Alternatively, widespread apoptosis leading to organismal death may enable cannibalization by the subsequent generation, ultimately benefitting the progeny by providing additional nutritional resources that enhance their chances of survival.
Our results indicate that SGK-1 promotes, while NHR-49 inhibits apoptosis, through controlling expression of ASM-3. Supporting this model, previous studies in mice have demonstrated that either disruption of ASM or treatment with S1P can suppress oocyte apoptosis, highlighting potential strategies to mitigate oocyte loss during anticancer therapies^71^. Consistent with the reduced apoptosis observed in sgk-1(lf) animals under CS conditions, earlier work showed that functional SGK-1 also promotes radiation-induced germ cell apoptosis^72^. These findings suggest that SGK-1 governs germ cell apoptosis not only during CS, but also in broader developmental and stress contexts, and that NHR-49-mediated ASM-3 suppression may represent a general mechanism of stress resilience.
Our data show that cold stress (CS), followed by ASM-3 secretion, induces neuronal beading and blebbing, a hallmark of early neurodegeneration. Ceramide, the enzymatic product of ASM, is notably elevated in several neurodegenerative disorders, including Alzheimer’s disease (AD), Parkinson’s disease (PD), and amyotrophic lateral sclerosis (ALS)^73^. ASM activity has been implicated in AD progression^74^, although the mechanisms by which it contributes to neurodegeneration remain incompletely understood. Our findings indicate that ASM-3 is secreted in response to CS and promotes apoptosis in peripheral tissues, suggesting that dysregulated ASM trafficking may contribute to neuronal cell death in neurodegenerative contexts. Moreover, calcium influx, a known driver of neuronal beading and blebbing during excitotoxicity, may provide a more direct link between neuronal death and SGK-1 function. As SGK-1 regulates ion channel homeostasis^75^, it may play a direct role in the ionic imbalance characteristic of various neurodegenerative diseases^76^.
In addition to NHR-49, several transcription factors have been reported to act downstream of mTORC2/SGK-1, including DAF-16, PQM-1, PHA-4, and SKN-1. Notably, SGK-1 functions as a negative regulator of SKN-1 by inhibiting its nuclear translocation^77,78^. A gain-of-function mutation in skn-1 has been shown to suppress lipid reallocation following cold stress (CS)^31^. Thus, activated SKN-1 may promote maternal survival and preserve lipid stores post-CS, phenocopying the effects observed in sgk-1 loss-of-function mutants.
Interestingly, as already noted previously^31^, this contrasts sharply with SKN-1 activity in unstressed adult animals. A skn-1 gain-of-function allele promotes somatic lipid depletion, a phenotype termed Adsf (adult depletion of somatic fat), which resembles lipid loss observed under oxidative stress or adult starvation conditions^79,80^. The mechanisms by which SKN-1 exerts these opposing effects on lipid stores under normal versus stress conditions remain to be elucidated^31^.
A detailed analysis of the combinatorial activities within the transcription factor network downstream of mTORC2/RICT-1/SGK-1 may provide insights into this regulatory complexity. Both NHR-49 and the metabolic regulator PQM-1 (ParaQuat/Methylviologen-responsive), acting downstream of mTORC2/SGK-1, are recognized as key modulators of lipid metabolism, orchestrating lipid synthesis and lipid catabolism. These transcription factors exert significant influence on lipid homeostasis under both physiological and stress conditions^10,57^. In an independent, parallel study, we recently discovered that PQM-1 functions downstream of mTORC2/SGK-1 signaling independently of NHR-49 to mediate post-CS responses [Gupta et al., pers. commun.]. Whether PQM-1 and NHR-49 directly interact or modulate each other remains unclear; however, such cross-talk could provide a mechanistic basis for fine-tuning responses under varying physiological conditions. Notably, CEH-60 and UNC-62, well-characterized PQM-1 interactors^28,39^, represent additional promising candidates for components of this regulatory transcription factor network.
In summary, our study identifies mTORC2/SGK-1 signaling as a central regulatory pathway of lipid distribution in response to CS, with its downstream effector NHR-49 acting as a key sensor of lipid availability and a decision-maker in responding to a severe CS condition. We further demonstrate a cross-tissue communication axis from intestine to neurons that coordinates the systemic CS response, and we uncover an apoptosis-driven neurodegeneration mechanism triggered by cold exposure. These findings underscore the critical role of lipid metabolism in stress adaptation, coupling lipid-derived signals to NHR activation and a systemic fate decision between organismal survival and death. Future work will determine whether other stressors converge on this regulatory network.
Methods
C. elegans strains
C. elegans strains were cultured and maintained at 20 °C with NGM plates seeded with Escherichia coli (E. coli), using standard protocols^81^. The N2 Bristol strain was used as the WT strain. Transgenic worm strains with extrachromosomal arrays were generated using microinjection. Genotypes of all the strains used in this study are shown in Supplementary Table 1.
Cloning and constructs for transgenic worm strains
Transgenic worm strains were constructed using microinjection^82 EMBO J^. To construct recombinant plasmid for asm-3::egfp driven by intestine specific promoter, we first amplified the genomic region from the 5’ UTR of asm-3 (mRNA isoform W03G1.7a.1) to the end of coding region without stop codon, and the fragments of asm-3 3’ UTR. Then “5’UTR::asm-3” was inserted to the 5’ and “3’ UTR” was inserted to the 3’ end of egfp of a pEGFP-N1 plasmid. The recombinant plasmid generated from this process was named as “5’UTR::asm-3::egfp::3’UTR”. Then we amplified the promoter region (2k upstream of the transcription start site) of gly-19 (mRNA isoform F22D6.12.2) and inserted it into the 5’ end of 5’UTR of the “5’UTR::asm-3::egfp::3’UTR” recombinant plasmid. Then a mutation generating kit was used to generate “5’UTR::asm-3(AA)::egfp::3’UTR” recombinant plasmid from “5’UTR::asm-3::egfp::3’UTR”. The recombinant plasmids are sequenced using Sanger sequencing for identification and confirmation.
To do microinjection, 20 ng/μl pCFJ[myo-2p::mCherry] was used as a coinjection marker and injected together with 100 ng/μl recombinant “5’UTR::asm-3::egfp::3’UTR” or “5’UTR::asm-3(AA)::egfp::3’UTR” plasmid. mCherry positive F1 progeny were isolated to new plates, with one worm on one plate. When F2 generation grows up, stable lines containing heritable extrachromosomal transgene were selected by screening for a positive coinjection marker signal in the F2 generation.
RNAi treatment
RNAi experiments were carried out using OP50(xu363)^83^ bacteria strains containing either empty L4440 vector or recombinant L4440 plasmid inserted with homologous regions of targeting genes for knocking down. Recombinant L4440 plasmids were extracted from HT115 RNAi bacteria of either C. elegans ORFeome or Ahringer RNAi library, and transformed into OP50(xu363) competent cells. RNAi was initiated from hatching and worms were kept on RNAi plates throughout all the experiments using RNAi. For neuronal and intestinal specific RNAi, TU3401 worm strain: sid-1(pk3321) V uIs69 V [pCFJ90 (myo-2p::mCherry) + unc-119p::sid-1] and IG1839 worm strain: frSi17 [mtl- 2p::rde-1 3’UTR] II; frIs7 [nlp-29p::GFP + col-12p::DsRed] IV; rde-1(ne300) V were used.
Dietary supplementation of small lipid molecules
SA (Cay10011298–1, cayman), OA (Cay90260–100, cayman), DGLA (Cay90230–50, cayman), OEA (Cay90265–100, cayman), S1P (860492P-1MG, Merck) and S1P-fluorescin (Echelon Biosciences) were all dissolved in 100% ethanol to make stock solutions. Concentrations for stock solutions are: SA 100 mM, OA 1.77 M, DGLA 100 mM, OEA 10 mM, S1P 2 mM, and S1P-fluorescin 2 mM. Working concentrations are: SA 0.8 mM, OA 14.16 mM, DGLA 100 μM, OEA 10 μM, S1P 2 μM, and S1P-fluorescin 2 μM. One day before use, stock solutions were added onto the surface of NGM plates that were already seeded with OP50. Plates were dried in air and kept in room temperature. S1P-fluorescin plates were stored in dark before use, and unnecessary light exposure was avoided during experiments. Worms were treated with small lipid molecules from hatching. For survival assays using small lipid molecule supplementations, after reproduction period, worms were transferred to freshly prepared plates every one week till all the worms are dead.
Post CS survival assay and statistics
To avoid potential transgenerational effects of temperature shifts on post CS survival, worms were maintained on 20 °C for at least three generations before they were used for a post CS survival assay. Synchronized L4 stage larvae were transferred to new plates to avoid stage specific effects on post CS survival. 18–20 hours after L4 synchronization, worms are transferred to 2 °C for 6 hours in a Binder KB E6 incubator. During the CS and early recovery period post CS, the lid of the plate container box was removed and plates were placed in one layer without staggering. During the CS period, the door of the incubator was not opened again to avoid temperature disturbance. During the CS and early recovery period post CS, the temperature of the NGM agar was recorded using a NIST traceable TMP117 digital temperature sensor (Texas Instruments). After cold shock, worms were immediately transferred to 20 °C. After 1-hour recovery, dead, bagged, missing and exploded worms are censored and removed. Alive worms (more than 95% of the worms initially used for CS for all genotypes) were used for recording the post CS recovery survival. Alive and dead worms were scored everyday till all the worms were dead. Bagged, missing and exploded worms were censored. For the first 6–8 days, alive worms were transferred to freshly seeded plates to avoid progeny contamination. Data was analysed using SPSS software, Kaplan-Meier survival method. Statistics was done using the Log-rank (Mantel–Cox) test.
Temperature monitoring for CS
A Binder KB E6 incubator was used for all CS experiments, and the temperature was recorded using a NIST traceable TMP117 digital temperature sensor. The temperature sensing chip was embedded into the NGM agar of a non-seeded plate. The temperature was recorded every 5 minutes. To maximize the accuracy of temperature recording, the sensor was operated with 64 times internal averaging performed in one second for each measurement. During the interval between two measurements, the sensor was kept in low-power shutdown mode to reduce self-warming. The expected accuracy of the temperature measurement is ±0.1°C.
Oil red O (ORO) staining
ORO staining was used to detect neutral lipids as previously described^84^. Synchronized adult day 1 animals were used for CS and ORO staining. To prepare for the staining experiments, one day before synchronizing L4 worms for CS, 0.5% Oil Red O stock solution was prepared by dissolving ORO powder in 100% isopropanol and kept in dark and on a shaker at room temperature for overnight. In the next day, at 12 pm, L4 worms were synchronized (for CS group) and Oil Red O stock was diluted into working concentration using water, with 60:40 (stock:MQ water) and kept in dark and on a shaker for overnight. Diluted solution was filtered using 0.45 μm filter before use. 6 hours after L4 synchronization, a second L4 synchronization was done and used as non-CS controls. 13 hours after the first L4 synchronization, the CS group worms were transferred to 2 °C incubator for 6 hours CS. After CS, worms were immediately transferred to 20 °C for recovery. After 10 hours recovery, when a pronounced somatic pigmentation loss was observed in wild-type animals, both CS and non-CS group worms were rinsed off the plates and washed for 3 more times using 500 μl M9 buffer. After washing, 500 μl 60% isopropanol was added to fix the samples at room temperature for 2 mins. Worms were collected by centrifuging at 2000 rpm and supernatant is removed, then 500 μl filtered ORO working solution was added. During staining process, worms were kept in dark and incubated in a Thermomixer, at 20 °C with shaking at 550 rpm for 18–20 hours. After staining, worms were washed with 500 μl 0.001% Triton X-100 in M9 buffer for three times. After the last washing, washing buffer was kept to store the samples until they were used to prepare slides for imaging. Imaging was performed using a Zeiss axioplan 2 imaging microscope and a AxioCam HRc camera.
Image was analyzed using ImageJ. After splitting three RGB channels, green channel was used for intensity quantification. ORO intensity within a worm body region was calculated and adjusted against the background intensity taken nearby the worm body. Relative intensity was calculated by comparing to non-CS wild-type animals in each independent replicates. 20–25 worms were used for each condition in each replicate. Three replicates were performed for each experiment.
Total RNA extraction and quantitative RT-PCR
Synchronized day 1 adult worms were collected and stored at −80 °C until all the conditions and replicates were ready. Three biological replicates were used to perform statistics analysis. To extract total RNA, worms were thawed and washed three times using M9 buffer. Then worms were lysed using Precellys 24 tissue homogenizer, using 6500 rpm X 3 times. Total RNA extraction was performed using RNeasy Mini Kit (QIAGEN) according to manufacturer’s instruction. Total RNA was digested with DNAase to remove DNA contaminations. RNA concentration was quantified using a Nanodropt spectrophotometer (NanoPhotometer NP80, IMPLEN).
Around 1 μg total RNA was used for reverse transcription using Invitrogen^™^ SuperScript^™^ III Reverse Transcriptase kit. Luna Universal master mix (New England Biolabs) and LC480 LightCycler (Roche Life Science) were used to perform quantitative PCR. pmp-3 was used as internal reference control. Primers for targeting genes and internal reference gene were listed in Supplementary Table 2. Data was analyzed using ΔΔC_t_ method.
Total lipid extraction
Adult day 1 animals were rinsed off the plate and washed three times using M9 buffer. After removing supernatant, around 100 μl packed worms from each condition were stored at −80 °C freezer. When three biological replicates were collected, frozen worms were thawed and mixed with 500 μl extraction buffer which is freshly prepared by mixing chloroform with methanol at 2 : 1 volume proportion. Then pre-cooled plastic beads were added and worms were lysed with Precellys 24 tissue homogenizer at 6500 rpm for 3 times. Worm lysate was kept on ice during the 5 mins interval between two shakings. The total worm lysate was transferred to new tube and added with 100 μl water and incubated on an overhead shaker for 20 mins, at 4 °C, 20 rpm. After centrifugation at 4000 g for 10 mins at 4 °C, water phase (top layer), protein phase (middle layer) and lipid phase (bottom layer) were separated from each other. The water phase was discarded and lipid phase was transferred to a new tube. The total protein (middle layer) was kept and re-dissolved with 2% SDS for quantification and normalization. Total lipid was used for S1P quantification. For this purpose, total lipid extraction was air-dried at room temperature for 1 h and redissolved using 400 μl methanol.
S1P quantification using ELISA
96-well high binding ELISA plate (MSEHNFX40, Sigma) was coated with 200 μl S1P standard/total lipid extract diluted in methonal, by evaporating at room temperature for 1 h. Concentrations of S1P standards are: 0 μM, 0.1 μM, 0.2 μM, 0.5 μM, 1 μM, 2 μM. Three technical replicates for each standard/total lipid sample, and three biological replicates were used. Coated ELISA plate was washed with PBST (0.05% Tween 20) for 3 X 5 mins, and blocked with 1% BSA in PBS for 1 hr on a shaker, at room temperature. Then 0.1 μg/ml anti-S1P mAb (Z-P300, Echelon Biosciences) in PBS was added to each well and plate was incubated for 1 hour on a shaker, at room temperature. For the first time to dissolve antibody, use 50%PBS with 50% glycerol and store in −80 °C freezer. The stock concentration is 0.1 μg/μl. Then the wells were washed with PBST (0.05% Tween 20) for 3 X 5 mins. For secondary antibody incubation, wells were added with 100 μl per well of 0.1 μg/ml HRP-conjugated goat anti-mouse secondary antibody and incubated for 1 hr at on shaker, at room temperature. After washing with PBST (0.05% Tween 20) for 3 X 5 mins, plate was then incubated with tetramethylbenzidine for 30 mins at 37 °C. Then 100 μl stopping buffer (1 M H_2_SO_4_) was added and plate was shaken gently to stop the reaction. Optical density was measured at 450 nm using a spectrophotometer (Infinite M200 Pro Multi Mode Microplate Reader, Tecan). Background absorption was extracted. Data was analysed using excel software to calculate S1P concentration of each sample by comparing to the standard curve. Normalization was performed using total protein levels quantified using the protein layer kept from lipid extraction step.
Protein quantification using micro BCA kit
Total protein levels of worm lysate for lipid extraction were quantified using the micro BCA kit (23235, Thermo Fisher Scientific Biosciences). A serial of dilutions of protein solution samples were prepared with 2% SDS. Then Working Reagent (WR) and protein standards were prepared according to the instruction of the micro BCA kit. Then 150 μl of each standard or diluted protein samples were added into microplate wells. Three technical replicates for each condition and each protein standard were performed. Three biological replicates were used. Then 150μl of the WR was added to each well and plate was thoroughly mixed on a shaker for 30 seconds. Then plate was incubated at 37°C for 2 hours. After the plate was cooled to room temperature, absorbance at 562 nm was measured on a spectrophotometer (Infinite M200 Pro Multi Mode Microplate Reader, Tecan). Background absorption was extracted. Data was analysed using excel software to calculate protein concentration of each sample by comparing to the standard curve.
Protein immunoprecipitation (IP)
Synchronized day 1 adult worms were collected and stored at −80 °C before use. To perform immunoprecipitation, around 100 μl packed frozen worms were thawed and added with cold NP40 lysis buffer (150mM NaCl, 50mM Tris HCl pH 8.0, 0.1% NP-40) containing protease inhibitors (cOmplete^™^ Proteasehemmer-Cocktail, Sigma-Aldrich); then worms were lysed with Precellys 24 tissue homogenizer using 6500 rpm X 3 times. Worm lysate was kept on ice during the 5 mins interval between two shakings. Total worm lysate was centrifuged at 12,000 rpm at 4 °C for 20 mins. Then supernatant was transferred to a new tube and added with 25 μl GFP Nanobody bead slurry (ChromoTek GFP-Trap^®^ Magnetic Agarose) and incubated at 4 °C on an overhead shaker at 20 rpm for overnight. Then GFP Nanobody beads were washed for 3 times with washing buffer (PBS, 0.5% Triton X-100, 1mM EDTA). To perform aSMase activity assay using the IP product, 5 μl GFP Nanobody beads were resuspended with 50 μl Substrate Buffer from the aSMase activity assay kit (EA KIT (K-3200), Echelon Biosciences), and used for the aSMase activity detection. To perform western blot with the IP product, 1 μl GFP Nanobody beads were resuspended in 40 μl of 1× Laemmli Sample buffer and denatured at 99 °C for 15 mins. 10 μl of the denatured IP product was used for SDS-PAGE and WB.
Acid sphingomyelinase (aSMase) activity assay
The aSMase activity assay was performed using the aSMase activity assay kit (Echelon Biosciences, K-3200). A series of aSMase standards were prepared from diluting a 32.8 μM stock using the Substrate Buffer, both of which are included in the kit. 50 μl standards or purified ASM-3::GFP, bound with GFP Nanobody beads were loaded into the wells of a 96-well plate. Two technical replicates and three biological replicates were used. Then the aSMase Substrate (provided in the kit) was quickly thawed at 70 °C for 2 mins with a heat block. aSMase Substrate was diluted with the Substrate Buffer at 1:40 ratio. 50 μl diluted aSMase Substrate was added into each well containing either standards or samples. Then the plate was incubated at 37 °C on a shaker, for 3 hours. Then 50 μl Stop Buffer was added to each well and the plate was incubated at room temperature for 30 mins on a shaker, to stop the reactions. During all the incubation processes, the plate was protected from light. The fluorescence was measured using a spectrophotometer (Infinite M200 Pro Multi Mode Microplate Reader, Tecan) at 360 nm excitation and 460 nm emission. Background absorption was extracted. The relative aSMase activity was normalized to the protein levels added to each well which was quantified by western blot.
Western blot (WB)
The total worm lysate or IP products were resolved by 10% SDS-PAGE gel and transferred to a PVDF membrane (Bio-Rad). Then the PVDF membrane was blocked with 5% skim milk dissolved in 1X TBST, on a shaker at room temperature for 1 hour. Then the membrane was incubated with primary antibody, anti-GFP (11814460001, Roche) antibody dissolved in the blocking buffer. Primary antibody incubation was done at 4 °C on a shaker for overnight. After 3 X 5 mins washing using 1 X TBST, the membrane was incubated with secondary antibody which was dissolved in blocking buffer, for 1 hour on a shaker and at room temperature. After 3 X 5 mins washing using 1 X TBST, the membrane was developed with the Western Blot Substrate (34096, Thermo Scientific^™^) and imaged with FUJIFILM LAS 4000 Imager. Quantification was performed using ImageJ and data was analysed using excel software to calculate relative protein levels.
Acridine orange (AO) staining
20 mg/ml AO stock was prepared using M9 buffer and can be stored at −20 °C for up to 3 months. To start an AO staining assay, AO stock was freshly diluted with M9 buffer into 20 μg/ml as working concentration. Synchronized adult day 1 CS and non-CS worms were stained with AO working solution on seeded NGM plates, by adding 500 μl AO working solution onto bacteria lawn. Plates were kept in dark and shaken for 1 hour at room temperature. Then worms were transferred to new seeded plates (no AO) and kept in dark for 1 hour at room temperature, to destain/remove gut lumen signals. After destaining, worms were mounted on slides containing 3% agarose pad using levamisole, and slides were analyzed using a Zeiss Z1 compound microscope equipped with an AxioCam MRm3 CCD camera, under a 60X objective lens. For quantification, apoptotic germ cells per gonad arm were counted and 20–30 worms per condition per replicate were used. Three independent replicates were performed for each experiment. Statistics was done using un-paired t-test.
S1P- fluorescin tracing
S1P- fluorescin containing plates was prepared as described above. Worms were hatched on S1P- fluorescin plates and a synchronization was performed at L4 stage. Adult day 1 worms were used for imaging analysis. Adult day 1 animals were mounted to 3% agarose pad and levamisole was used to immobilize the worms. Slides were analyzed using a Zeiss Z1 compound microscope with an AxioCam MRm3 CCD camera, under a 60X objective lens. S1P- fluorescin gives a green emission signal at 508 nm, upon excitation at 488nm.
Quantification was done using ImageJ. After splitting RGB channels, red channel was used for S1P- fluorescin intensity quantification. The S1P- fluorescin intensity in both intestine cells and oocytes were quantified and adjusted by extracting background intensity measured in a small region near the worm body. 20–25 worms were used for each condition in each replicate. Three biological replicates were performed for each experiment. Relative intensity was calculated by comparing to wild-type animals in each independent replicates.
Pacs-2::GFP reporter analysis
Synchronized adult day 1 *Pacs-2::*GFP expressing animals were used for imaging analysis. For RNAi and S1P treatment, animals were hatched on corresponding treatment plates, and afterward synchronized at L4 stage. Adult day 1 animals were mounted to 3% agarose pad and levamisole was used for immobilization. Slides were analyzed using a Zeiss Z1 compound microscope with an AxioCam MRm3 CCD camera, under a 10X objective lens. ImageJ was used as described above to quantify GFP intensity of each worm. Red channel was chosen for quantification after channel splitting. GFP intensity was quantified and adjusted by extracting background intensity measured in a small region near the worm body. 20–25 worms were used for quantification for each condition of each replicate. Three biological replicates were performed for each experiment.
PVD neuronal beading/blebing monitoring and quantification
A PVD neuronal mCherry reporter strain which expresses a membrane associated mCherry in PVD neurons was used to analyse the PVD morphology changes during CS. The genotype of the reporter strain, TV15916 is: wyIs581[ser2prom3::myr-mCherry::unc-54 3’UTR (pOL036), odr-1p::GFP]. Synchronized adult day 1 TV15916 animals (20 hours post L4) were cold stressed at 2 °C for around 24 hours. Age matched non-CS controls were prepared 24 hours after synchronization of CS worms. After CS worms were mounted to 3% agarose. To avoid neuronal toxicity of chemical anesthetics like levamisole or sodium azide, 0.1 μm diameter polystyrene microspheres (Polysciences 00876–15) were used to immobilize the worms. PVD neuronal morphology was analysed using a Zeiss Z1 compound microscope and an AxioCam MRm3 CCD camera under 20 X lens. The middle region of PVD dendrite that covers about half of the length of the whole worm was used for imaging. This region is also flanked by the neuron bodies expressing mCherry under the ser2prom3 promoter’s activity. For quantification, the total beading number in the imaging region of each worm was counted and 20–30 worms were used for quantification for each condition in each replicate. Three biological replicates were performed for each experiment. To perform neuronal specific RNAi, strain BR9222: sid-1(pk3321) uIs69[pCFJ90 (myo-2p::mCherry) + unc-119p::sid-1] is crossed into the PVD neuronal GFP reporter background.
Detecting ASM-3 secretion using confocal microscope
Transgenic worms expressing ASM-3::GFP or ASM-3(AA)::GFP driven by intestine specific promoter were used to analyse the ASM-3 secretion after CS. Synchronized transgenic adult day 1 worms (20 hours post L4 synchronization) were cold stressed at 2 °C for 6 hours, and then recovered at 20 °C for 3 hours. Age matched non-CS control worms were prepared 6 hours after CS animals’ synchronization. Worms were mounted on slides containing 3% agarose pad and immobilized using levamisole. The ASM-3::GFP or ASM-3(AA)::GFP signal in coelomocytes were detected by Nikon eclipse Ti A1 confocal microscope, to measure ASM-3 secretion. One pair of coelomocytes were used for one worm. 20 worms in total were used for quantification for each condition of each replicate. Three biological replicates were performed.
Quantification was performed using ImageJ as described above. Red channel was chosen for quantification after splitting channels. GFP intensity within coelomocytes were quantified and adjusted by extracting background intensity measured in a small region near the worm body. Relative intensity was calculated by comparing to non-transgenic siblings under non-CS condition.
RNAseq sample preparation and data analysis
Synchronized adult day 1 worms were applied for a CS at 2 °C for 6 hours and then recovered at 20 °C for 2 hours. 2 hours is chosen to allow primary gene expression changes to occur and to avoid secondary gene expression changes. Then worms were harvested and washed with M9 buffer for three times. After removing supernatant worms were frozen at −80 °C until all the biological replicates were collected. Two biological replicates were prepared for each condition. After collecting all the biological replicates, frozen worms were thawed and total RNA was extracted as described above. RNA qualities were evaluated by running a 1% agarose gel. Library construction and sequencing were performed in the Eurofins company.
Quality control and data processing of the RNA sequencing results were performed in Galaxy (https://usegalaxy.eu) using an RNAseq analysis tutorial from Galaxy (https://training.galaxyproject.org/training-material/topics/transcriptomics/tutorials/ref-based/tutorial.html). Briefly, FastQC, RNA STAR, featureCounts, DESeq2 and goseq were used for reads quality control, mapping, counting, differential gene expression analysis and GO enrichment analysis. Parameters were set according to the tutorial. The C. elegans reference genome WBcel235.51 was used for mapping and WBcel235.96.gtf was used for annotation. Mis-regulated genes in sgk-1(lf) comparing to WT after CS are defined as log_2_FC (sgk-1_CS vs. WT_CS) > 1 & FDR < 0.01, or log_2_FC (sgk-1(lf)_CS vs. WT_CS) < −1 & FDR < 0.01. Among these genes, NHR-49 dependent genes are defined as −1 < log_2_FC (nhr-49(lf); sgk-1(lf)_CS vs. WT_CS) < 1.
Mobility recovery assay
30 synchronized AD1 worms (20 hours after L4 synchronization) were cold stressed for 6 hours at 2 °C. The fours strains were transferred to CS incubator in a special order with 1 min interval, from N2, sgk-1(lf), nhr-49(lf), and nhr-49(lf); sgk-1(lf) double mutant. After CS, the worms were transferred to 20 °C incubator in the same order, also with 1 min interval. After 5 mins post the transfer of the first plate, the N2 plate was taken out for recording the mobility of the worms on the plate, and was quickly put back to the incubator within 1 min. Then sgk-1(lf), nhr-49(lf) and nhr-49(lf); sgk-1(lf) plates were taken out for recording at 6th, 7th, and 8th mins after the CS of the N2 plate. Then all the plates were recorded for the mobility recovery for every 5 mins. To record mobility recovery, plates were gently shaken and mobile worms which can crawl or mildly move their bodies were recorded and removed from the plates. This assay stopped when all the worms regain mobility. For the worms that are still immobile after 35 mins post CS, they are considered as dead and censored.
Statistical analysis
All data in this study were shown as mean ± standard deviation (SD) unless otherwise stated. Survival assays and mobility recovery assays were analyzed using SPSS software with Kaplan-Meier survival method; and P value was calculated using log-rank test. Otherwise, P value was calculated using an unpaired two-tailed Student t-test to compare two different groups. Confidence intervals of 95% were chosen for unpaired two-tailed Student t-test. P values below 0.05 were regarded as significant (* P < 0.05, ** P < 0.01, *** P < 0.001, ns P > 0.05). P values were shown in each experiment and representative results were shown. For all the experiments, at least three replicates were performed.
Strain information
BR7946 N2 (wild-type);
BR4774 (sgk-1(ok358) X);
BR9062 (nhr-49(et7) I);
BR9063 (nhr-49((et8) I);
BR8919 (nhr-49(nr2041) I);
BR5611 (rict-1(mg360) II);
BR9222 (sid-1(pk3321) V uIs69 V [pCFJ90 (myo-2p::mCherry) + unc-119p::sid-1]);
BR9210 (frSi17[mtl-2p::rde-1 3’UTR] II; frIs7[nlp-29p::GFP + col-12p::DsRed] IV; rde-1(ne300) V);
BR8924 (glmEx5[nhr-49p::nhr-49::gfp + myo-2p::mCherry]);
BR9109 (glmEx9[gly-19p::nhr-49::GFP + myo-2p::mCherry]);
BR9110 (glmEx11[col-12p::nhr-49::GFP + myo-2p::mCherry]);
BR9111 (glmEx20[unc-119p::nhr-49::GFP + myo-2p::mCherry]);
BR8925 (nhr-49(nr2041); glmEx5 [nhr-49p::nhr-49::GFP + myo-2p::mCherry]);
BR9108 (nhr-49(nr2041); glmEx7[myo-3p::nhr-49::GFP + myo-2p::mCherry]);
BR9114 (nhr-49(nr2041); glmEx9 [gly-19p::nhr-49::GFP + myo-2p::mCherry]);
BR9161 (nhr-49(nr2041); glmEx11[col-12p::nhr-49::GFP + myo-2p::mCherry]);
BR9115 (nhr-49(nr2014);glmEx20[unc-119p::nhr-49::GFP + myo-2p::mCherry]);
BR9169 (nhr-49(nr2041); sgk-1(ok538); glmEx9[gly-19p::nhr-49::GFP + myo-2p::mCherry]);
BR9171 (nhr-49(nr2014); sgk-1(ok538); glmEx20[unc-119p::nhr-49::GFP + myo-2p::mCherry]);
BR8941 (nhr-49(nr2041); sgk-1(ok538));
BR6940 (pwIs23[vit-2::GFP]);
BR6982 (sgk-1(ok538) X; pwIs23[vit-2::GFP]);
BR8991 (wbmEx57[acs-2p::GFP + rol-6(su1006)]);
BR9226 (sphk-1(ok1097) II);
BR9337 (sphk-1(ok1097); sgk-1(ok538));
BR9188 (wyIs581[ser2prom3::myr-mCherry::unc-54 3’UTR (pOL036), odr-1p::GFP]);
BR9211 (sgk-1(ok538); wyIs581[ser2prom3::myr-mCherry::unc-54 3’UTR (pOL036), odr-1p::GFP]);
BR9221 (nhr-49(nr2041); wyIs581[ser2prom3::myr-mCherry::unc-54 3’UTR (pOL036), odr-1p::GFP], glmEx9[gly-19p::nhr-49::GFP + myo-2p::mCherry]);
BR9213 (nhr-49(nr2041); sgk-1(ok538); wyIs581[ser2prom3::myr-mCherry::unc-54 3’UTR (pOL036), odr-1p::GFP], glmEx9[gly-19p::nhr-49::GFP + myo-2p::mCherry]);
BR9212 (nhr-49(nr2041); wyIs581[ser2prom3::myr-mCherry::unc-54 3’UTR (pOL036), odr-1p::GFP], glmEx20[unc-119p::nhr-49::GFP + myo-2p::mCherry]);
BR9219 (nhr-49(nr2014); sgk-1(ok538)X; wyIs581[ser2prom3::myr-mCherry::unc-54 3’UTR (pOL036), odr-1p::GFP], glmEx20[unc-119p::nhr-49::GFP + myo-2p::mCherry]);
BR9238 (sid-1(pk3321) V uIs69[pCFJ90 (myo-2p::mCherry) + unc-119p::sid-1] V; wyIs581[ser2prom3::myr-mCherry::unc-54 3’UTR (pOL036), odr-1p::GFP]);
BR9237 (sid-1(pk3321) V uIs69[pCFJ90 (myo-2p::mCherry) + unc-119p::sid-1] V; sgk-1(ok538); wyIs581[ser2prom3::myr-mCherry::unc-54 3’UTR (pOL036), odr-1p::GFP]);
BR9363 (frSi17[mtl-2p::rde-1 3’UTR] II; rde-1(ne300) V; wyIs581[ser2prom3::myr-mCherry::unc-54 3’UTR (pOL036), odr-1p::GFP]);
BR9327 (wyIs581[ser2prom3::myr-mCherry::unc-54 3’UTR (pOL036), odr-1p::GFP]; sgk-1(ok538), byEx1927[gly-19p::asm-3::egfp, myo-2p::mCherry]);
BR9328 (wyIs581[ser2prom3::myr-mCherry::unc-54 3’UTR (pOL036), odr-1p::GFP], byEx1927[gly-19p::asm-3::egfp, myo-2p::mCherry]);
BR9330 (wyIs581[ser2prom3::myr-mCherry::unc-54 3’UTR (pOL036), odr-1p::GFP]; sgk-1(ok538), byEx1924[gly-19p::asm-3(AA)::egfp, myo-2p::mCherry]);
BR9332 (wyIs581[ser2prom3::myr-mCherry::unc-54 3’UTR (pOL036), odr-1p::GFP], byEx1924[gly-19p::asm-3(AA)::egfp, myo-2p::mCherry]);
BR9365 (sid-1(pk3321) V uIs69[pCFJ90 (myo-2p::mCherry) + unc-119p::sid-1] V; sgk-1(ok538); wyIs581[ser2prom3::myr-mCherry::unc-54 3’UTR (pOL036), odr-1p::GFP], byEx1935[gly-19p::asm-3::egfp + myo-3p::egfp]);
BR9383 (sid-1(pk3321) V uIs69[pCFJ90 (myo-2p::mCherry) + unc-119p::sid-1] V; wyIs581[ser2prom3::myr-mCherry::unc-54 3’UTR (pOL036), odr-1p::GFP], byEx1939[gly-19p::asm-3::egfp + myo-3p::egfp]);
BR9317 (byEx1927[gly-19p::asm-3::egfp, myo-2p::mCherry]);
BR9329 (sgk-1(ok538), byEx1927[gly-19p::asm-3::egfp, myo-2p::mCherry]);
BR7967 (ced-1(n2089) I; ced-5(n1812) IV);
BR4201 (ced-1(n2089); ced-5(n1812); sgk-1(ok538));
BR9331 (sgk-1(ok538), byEx1924[gly-19p::asm-3(AA)::egfp, myo-2p::mCherry]);
BR9313 (byEx1924[gly-19p::asm-3(AA)::egfp, myo-2p::mCherry]);
BR927 (daf-2(e1370) III);
BR9162 (lbp-8(gk5151[loxP + myo-3p::GFP::unc-54 3 UTR + rps-27p::neoR::unc-54 3 UTR + loxP]) V);
BR9207 (lbp-8(gk5151[loxP + myo-3p::GFP::unc-54 3 UTR + rps-27p::neoR::unc-54 3 UTR + loxP]) V; sgk-1(ok538));
BR8988 (nhr-80(tm1011) III)
BR9052 (nhr-80(tm1011); sgk-1(ok538))
Primers for worm strains genotyping
Primer name sequence
asm-3(ok1744)-F AAAAAATCGCATTCAAATCTTGC
asm-3(ok1744)-delF AGTACCCCTATCGCAAAACTACC
asm-3(ok1744)-R ACACACCTTATTGGCTTCTCACG
ced-4(n1162)-F ATTTATCCTCTATATTACATGCCTG
ced-4(n1162)-R TTTTGGAGCTGTTCCACTATCTTTG
nhr-49(gf)-F GATATGGAAATATTTTGTTGTTGGC
nhr-49(gf)-R GTACATTATAGCTGGGCTCTGGACT
nhr-80(tm1011)-F ATAAGAAGGATGGATCAAGTAGGAA
nhr-80(tm1011)-delR GCCTCAACAACTACAAACTACAGTA
nhr-80(tm1011)-R ATCAGGGCTCAAGAGAAATAAGAAA
nhr-49(nr2041)-F TTCTTTCTTTCCTTTTCCTGTCC
nhr-49(nr2041)-delR GTTGCTGTTTCCGTTCTCGTA
nhr-49(nr2041)-R TCCATATCTTGTGGGTGCGTCAT
sgk-1(ok538)-F TTTCCTTGTTTCACTTCACTTTTCG
sgk-1(ok538)-delR GTGAGCACCAAGTAACCTTTATCGT
sgk-1(ok538)-R ATGTATCCATTGCGTTTTTCACTTT
rde-1(ne219)-F ATTTTTAGGGTATTTTCTTTGTAGT
rde-1(ne219)-R CATGTGATTTTTGTTGAAGTTGTCG
rde-1(ne300)-F GAACAACGACAATCGAGCACCA
rde-1(ne300)-R ATCTTGTGACCGAACTGTCC
frSi17-F AACAAACGTGGGATGTAACC
frSi17-R TCATACTCGTAGTATTCCCG
rict-1(mg360)-F AACCATCGAAGAATTCCGATGACTC
rict-1(mg360)-R ATTTGTTTGCAGCTCTGGGGAGCCA
sid-1(pk33321)-F CACCTGTCTTATCACTGCTTCTTGT
sid-1(pk33321)-R GTAACGCCTTTTTTTTGACTTTTCG
sphk-1(ok1097)-F AAATAAAATAAAATCTATACATCGG
sphk-1(ok1097)-delR AGATATTATAAACAAACGATTCTCG
sphk-1(ok1097)-R TCTCTTTCTCTAACACACTCTAACC
Primers for asm-3 transgenic worm strains
asm-3 promoter-F CGACGGTACCGCGGGCCCGGGAAGTAAAGACGTCTTCAGTTTC
asm-3 promoter-R GAAGCCGAGGTTTCCAGTAATAAAGATAAGAAATTTGGAGCGGGAG
gly-19 promoter-F CGACGGTACCGCGGGCCCGGCAAATATTCTCATTTCAAAATTTTCAG
gly-19 promoter-R GAAGCCGAGGTTTCCAGTAATCTGGAAATTTAAATTTAATTCTTTGG
unc-119 promoter-F CGACGGTACCGCGGGCCCGGAAGTGATTCGGAATAAGGAAATGAGG
unc-119 promoter-R GAAGCCGAGGTTTCCAGTAATAAAAAAACAGAATTTCAAATTTTTTG
asm-3 promoter + 5’UTR-F CCGGGATCCACCGGTCGCCACCGAAGTAAAGACGTCTTCAGTTTC
asm-3 promoter + 5’UTR-R CTCTACCCTGAAATGCGAGAGAAAGGCGGAGCCGTATCGTTTCTGATG
5’UTR + asm-3 (for asm-3 promoter) no stop codon-F GCTCCAAATTTCTTATCTTTATTACTGGAAACCTCGGCTTCTCCCCC
5’UTR + asm-3 (for asm-3 promoter) no stop codon-R CACCATGGTGGCGACCGGTGAATCTTGCACTCCTCCTTTCC
5’UTR + asm-3 (for gly-19 promoter) no stop codon-F CCAAAGAATTAAATTTAAATTTCCAGATTACTGGAAACCTCGGCTTCTCCCCC
5’UTR + asm-3 (for gly-19 promoter) no stop codon-R CACCATGGTGGCGACCGGTGAATCTTGCACTCCTCCTTTCC
5’UTR + asm-3 (for unc-119 promoter) no stop codon-F ATTTGAAATTCTGTTTTTTTATTACTGGAAACCTCGGCTTCTCCCCC
5’UTR + asm-3 (for unc-119 promoter) no stop codon-R CACCATGGTGGCGACCGGTGAATCTTGCACTCCTCCTTTCC
3’UTR oligo GACGAGCTGTACAAGTAAAGCGGCCGCATGCGATTTTTTTTTTTGAAATTTAAAATCCTACTAATCTCCGGGTCGTGTTGAATAAATTAATTTGTATTTGCGACTCTAGATCATAATC
identification primers
promoter 1 CGC AAA TGG GCG GTA GGC GTG
promoter 2_asm-3 AGAACTCACTGCGATTTGCATTAGG
promoter 2_unc-119 TCTAAGATTCTATTGAATTACCATC
promoter 2_gly-19 AGTCTCCACTGCCGCTTTTTAAGTC
promoter 3 AATAAAAACCGCAATTTCCATAACA
coding 1 AGGTTTTGTTGGTGTCGGGGTAGGC
coding 2 CGAGCGAATCGTATCCGACAAGTTT
coding 3 CCTACCATATTGTTCCCGTATCTCG
coding 4 AACCATTCCTCATAGTCGATTACTT
coding 5 GTC CAG CTC GAC CAG GAT G
3’UTR after EGFP GAT CAC TCT CGG CAT GGA C
qPCR primers
asm-3-F AACTTGTCGGATACGATTCGCTCGT
asm-3-R ATTGAAGGGTGCCATCTGGGTCAGT
pmp-3-F GAATGGAATTGTTTCACGGAATGC
pmp-3-R CTCTTCGTGAAGTTCCATAACACGATG
fat-7-F CAAGTCCAAGAGGAGAGCAAAAAAA
fat-7-R GAAGCCATCCCATGTGAGTGAAGAA
ech-1.1-F CTAGGACTAAAGCAAGGAAAACTCG
ech-1.1-R TCTTCAAGACAAAGAAGTGCCTCAT
ZK550.6-F GAGTGCCGAGCAGAGGCGATTCTAC
ZK550.6-F TGTGATCTTCTCCATTGCGGTCCAA
sodh-1-F AGCGTTTGCCCATTGGTTGGAGGAC
sodh-1-R GAATTGGAGCAGCAGCGGCGAGATT
acs-2-F AACGTCTATCCGACCGAAATCGA
acs-2-R GTAAAAAGCAAGGAAATCCCAGC
Supplementary Material
1
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Lamming D. W. & Sabatini D. M. A central role for m TOR in lipid homeostasis. Cell metabolism 18, 465–469 (2013).23973332 10.1016/j.cmet.2013.08.002PMC 3818790 · doi ↗ · pubmed ↗
- 2Szwed A., Kim E. & Jacinto E. Regulation and metabolic functions of m TORC 1 and m TORC 2. Physiol Rev 101, 1371–1426, doi:10.1152/physrev.00026.2020 (2021).33599151 PMC 8424549 · doi ↗ · pubmed ↗
- 3Folick A. Aging. Lysosomal signaling molecules regulate longevity in Caenorhabditis elegans. Science 347, 83–86, doi:10.1126/science.1258857347/6217/83 [pii] (2015).25554789 PMC 4425353 · doi ↗ · pubmed ↗
- 4Liu S., Alexander R. K. & Lee C.-H. Lipid metabolites as metabolic messengers in inter-organ communication. Trends in Endocrinology & Metabolism 25, 356–363 (2014).24895003 10.1016/j.tem.2014.05.002PMC 4077945 · doi ↗ · pubmed ↗
- 5Kersten S. Integrated physiology and systems biology of PPARα. Molecular metabolism 3, 354–371 (2014).24944896 10.1016/j.molmet.2014.02.002PMC 4060217 · doi ↗ · pubmed ↗
- 6Wu G., Baumeister R. & Heimbucher T. Molecular Mechanisms of Lipid-Based Metabolic Adaptation Strategies in Response to Cold. Cells 12, doi:10.3390/cells 12101353 (2023). · doi ↗
- 7Nelson C. J., Otis J. P. & Carey H. V. A role for nuclear receptors in mammalian hibernation. The Journal of physiology 587, 1863–1870 (2009).19289545 10.1113/jphysiol.2008.167692 PMC 2689326 · doi ↗ · pubmed ↗
- 8Wahli W. & Michalik L. PPA Rs at the crossroads of lipid signaling and inflammation. Trends in Endocrinology & Metabolism 23, 351–363 (2012).22704720 10.1016/j.tem.2012.05.001 · doi ↗ · pubmed ↗