Fatty Acid Desaturation Links Germ Cell Loss to Longevity Through NHR-80/HNF4 in C. elegans

Lifespan extension induced by germline ablation in C. elegans is regulated by the nuclear hormone receptor NHR-80 in a process that requires the production of oleic acid by activation of the lipid desaturase FAT-6/SCD1.


Introduction
Removing the germ line of Caenorhabditis elegans extends its lifespan by approximately 60% [1].Eliminating germ cells also increases the lifespan of Drosophila, suggesting that a conserved mechanism links the germ line to longevity [2].In C. elegans, removal of the germ line can be achieved either by laser ablation of germline precursor cells at early developmental stages or through mutations that impair the proliferation of germline stem cells (GSCs) [3].The glp-1(e2141ts) and mes-1(bn7) alleles deplete the germ line by either blocking proliferative signals for GSC or inhibiting cell division in the P lineage at early embryonic stages.As a result, animals carrying these alleles are long-lived [3].Longevity is not merely caused by sterility because animals lacking both germ cells and the somatic gonad are sterile but not long lived.The germ line and the somatic gonad have been suggested to have opposite effects on longevity [1], but the molecular basis of germline-mediated longevity remains poorly understood.
Hsin and Kenyon showed that the presence of daf-16/FOXO or daf-12/VDR is required for extending the lifespan of animals whose germ line had been ablated [1].DAF-16/FOXO is a forkhead transcription factor that translocates into intestinal nuclei and promotes transcription when GSCs stop proliferating [4].daf-16/FOXO is a key downstream component of the insulin/IGF1 signaling (IIS) pathway that also regulates longevity [5,6].Although inhibiting GSC proliferation and down-regulating the activity of the IIS pathway both result in lifespan extension and translocation of DAF-16 into intestinal nuclei, several experiments show that these manipulations are not equivalent.First, GSC removal extends the lifespan of daf-2 mutants that are already long-lived due to a constitutively down-regulated IIS pathway (hypomorphic allele of the sole insulin receptor) [1].Nuclear translocation of DAF-16 requires the intestinal protein KRI-1 (an ankyrin-repeat protein) when it is provoked by a GSC proliferation arrest, but not by daf-2 mutations [4].Finally, the transcription elongation factor, TCER-1, promotes the transcriptional activity of daf-16/FOXO when GSCs stop dividing but not in long-lived IIS mutants [7].Taken together, these data indicate that IIS and GSCs affect longevity though distinct mechanisms although they are both mediated by DAF-16.
The second pathway required for germline longevity is the DAF-12 lipophilic-hormone signaling pathway.In response to the loss of germ cells, the cytochrome P450, DAF-9 [8], and the Rieske protein, DAF-36 [9], use cholesterol to produce a steroid hormone (dafachronic acids) that activates the nuclear hormone receptor, DAF-12/VDR [10].DAF-12/VDR is homologous to the vertebrate vitamin D receptor, and its presence in its activated form is required to extend lifespan though depletion of the germ line [1].The interactions of KRI-1/DAF-16/TCER-1 and the DAF-9/DAF-36/DAF-12 pathways are still unclear.
Similar to KRI-1, DAF-9 and DAF-12 facilitate the nuclear translocation of DAF-16 triggered by germline removal [4], suggesting that the lipophilic-hormone signaling pathway may act upstream of DAF-16.However, recent work showed that DAF-12 and DAF-16 also function separately.First, germline-less animals in which DAF-16 is forced into intestinal nuclei still require daf-12 to be long-lived [4].Second DAF-12 and DAF-16 promote the expression of different gene sets [11].The Kenyon lab showed that sod-3 and cdr-6 are DAF-16 and DAF-12 targets, respectively [11].The K04A8.5 lipase is induced in glp-1(e2141ts) mutant animals in a daf-16 dependent manner but is not affected by daf-12 [12].Since the K04A8.5 lipase is also required for lifespan extension, these results suggest that the KRI-1/DAF-16/K04A8.5 pathway can promote longevity independently of daf-12 and that DAF-16 dependent transcription does not strictly require daf-12.Finally, it has also been shown that the DAF-12 lipophilic-hormone signaling pathway can mediate longevity in response to the somatic reproductive tissues, but this also requires the presence of DAF-16 [11].Thus, it is still unclear whether DAF-12 can promote longevity in the absence of daf-16.
In the present study, we searched for new nuclear receptors that can mediate longevity of C. elegans through depletion of the germ line using an RNAi-based genetic screen.We report that nhr-80/HNF4 is required for extending lifespan through germline removal, although it does not affect the lifespan of wild type animals.We show that NHR-80 is specific to this pathway since other longevity paradigms are not affected by a loss-of-function mutation of nhr-80.Moreover, the levels of NHR-80 increase in intestinal cells when germ cells are depleted.This increase is physiologically relevant because (1) overexpressing nhr-80 further extends the lifespan of germline-less animals and (2) germline ablation leads to the nhr-80 dependent up-regulation of the stearoyl-CoA desaturase (SCD), fat-6, that produces oleic acid (OA) from stearic acid and (3) increased fatty acid desaturation and OA production are necessary to extend the lifespan of germline-less animals.A link between fat metabolism and germline-mediated longevity has already been reported by the Ruvkun lab.In this recent report, the authors reported that the triacylglyceride (TAG) lipase (K04A8.5) is required for germline longevity [12].Both fat-6 and K04A8.5 are induced in germline-less animals, and their inactivation by RNAi fully suppresses lifespan extension by depletion of the germ line.However, in contrast to the K04A8.5 lipase that acts downstream of the KRI-1/DAF-16 pathway and independently of the DAF-9/ DAF-36/DAF-12 lipophilic-hormone signaling pathway, we show that the NHR-80/FAT-6/OA pathway does not require the presence of daf-16 but necessitates the presence of daf-12.Taken together, our data and that of Wang et al. are consistent with the conclusion that the lifespan benefits triggered by inhibiting GSC proliferation require a important modification of the metabolism of fat since the NHR-80/FAT-6/OA and the KRI-1/DAF-16/ K04A8.4 pathways are activated independently to promote longevity though the activation of fat remodeling enzymes.

nhr-80 Is Required for Lifespan Extension through Depletion of the Germ Line
To identify new genes required for lifespan extension triggered by germline ablation, we screened for genes encoding for nuclear receptors (NHRs) using RNAi by feeding.We sought genes whose inactivation could suppress the lifespan of glp-1(e2141ts) mutant C. elegans without affecting the lifespan of wild type animals.Therefore, we compared the proportion of dead animals in glp-1(e2141ts) mutant and wild type animals after 20 d of RNAi treatment (starting at day 1 of adulthood).At 20 d, 50% to 60% of glp-1(e2141ts) mutants were alive, compared to less than 30% of wild type animals.In our screen, successful candidates lowered the survival of glp-1(e2141ts) mutants significantly but did not affect that of wild type animals at 20 d of adulthood.Of the 195 NHRs present in the Ahringer library (70% of all NHRs present in C. elegans), only one, nhr-80, in addition to our positive control (daf-12), reduced the lifespan of glp-1(e2141ts) mutants without affecting wild type lifespan.In wild type fertile animals, NHR-80 promotes fatty acid (FA) desaturation without affecting lifespan [13].

nhr-80 Specifically Promotes Germline-Mediated Longevity
To determine whether nhr-80 is specific to germline-mediated longevity, we knocked out nhr-80 in several other longevity

Author Summary
Reproduction and aging are two processes that seem to be closely intertwined.Experiments in Caenorhabditis elegans and Drosophila have shown that depletion of the germ line increases lifespan and that this process depends on insulin and lipophilic-hormone signaling.Recently, it was demonstrated that when germline stem cells (GSCs) cease to proliferate, fat metabolism is altered and this affects longevity.In this study, we have identified a nuclear hormone receptor, NHR-80, that mediates longevity through depletion of the germ line by promoting fatty acid desaturation.The nhr-80 gene is up-regulated at the mRNA and protein levels in germline-less animals, leading to the transcription of the gene, fat-6, and the production of oleic acid (OA).Our experiments also show that the NHR-80/FAT-6/OA pathway does not require the presence of DAF-16 but instead, depends fully on the presence of DAF-12, a steroid receptor that affects lifespan.We provide evidence that other NHR-80 targets must be present concomitantly.Our results reinforce the notion that fat metabolism is profoundly altered in response to GSC proliferation, and the data contribute to a better understanding of the molecular relationship between reproduction, fat metabolism, and aging.
The nhr-80 mRNA and Protein Levels Increase in glp-

1(e2141ts) Mutants
We analyzed the localization of NHR-80, using a functional protein fused to a GFP tag (see Materials and Methods and Figure 3B).In contrast to previous reports [16], we found that NHR-80 is localized in the nucleus and that it is expressed in the intestine and in neurons (some head and tail neurons, as well as the ventral cord; Figure 3A and B).This discrepancy is likely due to the fact that we fused a GFP tag to the full-length NHR-80 sequence driven by its own promoter, while Miyabayashi et al. fused the tag to the nhr-80 promoter and may have missed the nuclear import signal [16].We found that NHR-80 nuclear localization is constitutive and independent of the presence of the germ line (i.e., glp-1(e2141ts);nhr-80(tm1011);NHR-80::GFP mutant animals at restrictive versus permissive temperature) (Figure 3A  and B).However, the intensity of the NHR-80::GFP in intestinal cells is increased by 60% when glp-1(e2141ts);nhr-80(tm1011) mutant animals are shifted to the restrictive temperature at the L1 stage while no changes are noted in neuronal cells (Figure 3C  and D).When glp-1(e2141ts) mutant animals were shifted to restrictive temperature at later stages (L4 and day 1 of adulthood), the induction of intestinal NHR-80 still occurs, but less dramatically (L4 and day 1 of adulthood; Figure S1).Taken together with previous data showing that late shifts extend lifespan to a lesser extent [3], our data suggest that NHR-80 induction correlates with the extent to which lifespan is extended.To confirm this increase, we measured the overall expression level of nhr-80 by qRT-PCR and found that it is induced 5.6-fold in glp-1(e2141ts) mutant animals (Figure 3E).Thus, in glp-1(e2141ts) mutant animals, overall nhr-80 mRNA levels and the intensity of NHR-80::GFP in the intestine are increased.This strongly suggests that NHR-80 promotes longevity in the intestine.Supporting this notion, we found that nhr-80 RNAi also suppresses the lifespan of glp-1(e2141ts) mutant animals, although neurons are refractory to RNAi (Figure S2, Table S1).
Overexpressing nhr-80 Increases the Lifespan of glp-1(e2141ts) Mutant Animals But Not That of Wild Type Because nhr-80 is a positive longevity regulator that is induced in glp-1(e2141ts) mutant animals, we examined whether overexpressing nhr-80 could recapitulate germline-mediated longevity in a wild type context.Surprisingly, the nhr-80 transgene, which fully restores the longevity of glp-1(e2141ts);nhr-80(tm1011) mutant animals (Figure S3, Table S1), fails to extend the lifespan of wild type animals (Figure 4A, Table S1) but increases the mean lifespan of glp-1(e2141ts) mutant animals by 80% (Figure 4B, Table S1).It is remarkable that a loss-of-function mutation and an overexpression of nhr-80 have opposite effects on lifespan in the absence of proliferating GSCs only (Figures 1A and 4B, Table S1).The mechanism through which this is achieved remains to be determined.Our data do not allow discrimination between activation by binding of NHR-80 to a ligand, post-translational modifications, or interaction with a partner.To gain further insights into NHR-80 functions, we examined the interaction of nhr-80 with other longevity determinants of germline-less animals.

nhr-80 Does Not Require daf-16 to Promote Longevity
To examine mechanisms through which NHR-80 may promote the longevity of germline-depleted animals, we first tested whether nhr-80 longevity function depends on DAF-16.To address this question, we assessed whether NHR-80 may promote DAF-16 nuclear localization, similar to KRI-1 (orthologous to the human disease gene KRIT1; [4]).We found that down-regulating nhr-80 by injection of double-stranded RNA does not affect the localization of DAF-16 (Figure 5A).Next, we tested whether daf-16 was required for the increase of nhr-80 mRNA levels triggered by the depletion of the germ line or the longevity extension provoked by nhr-80 overexpression.We found that, in daf-16(mu86);glp-1(e2141ts) mutant animals, the nhr-80 mRNA levels are increased relative to wild type (Figure 5B; 3.7-fold; p = 0.002) and that overexpressing nhr-80 increases lifespan by 40% (Figure 5C, Table S1), while nhr-80 RNAi decreases lifespan by 58% (Figure S4, Table S1).However, the transcriptional induction of nhr-80 and the lifespan extension triggered by nhr-80 overexpression are decreased in the absence of daf-16 relative to glp-1(e2141ts) mutants, suggesting that DAF-16 can modulate the transcriptional induction of nhr-80.Thus, DAF-16 is not strictly required for nhr-80 function, but DAF-16 can modulate nhr-80 mRNA levels.
Next, we tested the effect of OA on the lifespan of glp-1(e2141ts);daf-12(rh61rh411) and glp-1(e2141ts);daf-9(rh50) mutants and found that it fails to extend lifespan in these two backgrounds (Figure S12, Table S1).Furthermore, we found that fat-6 mRNA levels are still induced in glp-1(e2141ts);daf-12(rh61rh411) mutant relative to wild type animals (Figure S13).Taken together, our data exclude the possibility that the FAT-6/ OA branch is a common target of DAF-12 and NHR-80.Thus, we conclude that DAF-12 does not interact with NHR-80 by co-promoting FAT-6.

NHR-80 and Oleic Acid Promote Germline-Mediated Longevity in Concert
Because our data suggest that OA and overexpressing nhr-80 are distinct interventions, we next verified whether NHR-80 and OA must work in concert to mediate germline longevity.To address this, we first investigated whether OA could extend lifespan in the absence of nhr-80.Convincingly, the lifespan of glp-1(e2141ts);nhr-80(tm1011) mutant animals is not affected by OA (Figure 8A, Table S1).Conversely, we overexpressed nhr-80 in animals that could no longer produce OA (glp-1(e2141ts);fat-6(tm331);fat-7(wa36)) and found that it fails to increase the lifespan of these animals (Figure 8B, Table S1).In this experiment, we noted that overexpressing nhr-80 does not increase the lifespan of glp-1(e2141ts);fat-6(tm331);fat-7(wa36) mutants complemented with OA.This is surprising because nhr-80 overexpression increases the lifespan of glp-1(e2141)ts mutants (Figure 4B, Table S1).The reason for this is unclear but may be explained by the fact that OA is not as efficient when it is externally provided since it is prone to oxidation.Finally, we showed that the effect of OA on the lifespan of glp-1(e2141ts);fat-6(tm331);fat-7(wa36) mutant animals is no longer significant when nhr-80 is knocked down by RNAi (Figure 8C, Table S1).Taken together, our data suggest that OA and NHR-80 promote longevity in concert.This can be simply explained by the assumption that the OA producing pathway is not the only longevity-promoting branch downstream of NHR-80 (Figure 9).

Discussion
In the present study, we show that, when germ cells are removed, nhr-80 mRNA and protein levels increase.This promotes the mono-desaturation of stearic acid to OA by inducing the transcription of the stearoyl-CoA-desaturase, fat-6/SCD1.This cascade is physiologically relevant to longevity since both nhr-80 and the SCD activity are required to augment the lifespan of germline-depleted animals.Furthermore, the lack of SCD activity can be bypassed by addition of exogenous OA in the medium, confirming the pivotal role of this metabolite.
Our data also indicate that the FAT-6/OA branch is required, but not sufficient, to promote longevity in response to depletion of the germ line downstream of NHR-80.This is evidenced by the fact that overexpressing nhr-80 extends the lifespan of both glp-1(e2141ts) and daf-16(mu86);glp-1(e2141ts) mutants but OA does not, suggesting that these two interventions are not equivalent (Figures 4B, 5C, 7A, and S13A, Table S1).Supporting this view, we showed that OA and NHR-80 must act in concert to support the lifespan extension conferred by germ cell loss.Indeed, OA does not increase the lifespan of glp-1(e2141ts);nhr-80(tm1011) mutants and overexpressing nhr-80 is inefficient when the SCD genes are deleted (OA producing genes).Moreover, while OA restores the lifespan of glp-1(e2141ts);fat-6(tm331);fat-7(wa36) mutant animals, it essentially fails to do so when nhr-80 is deactivated by RNAi (Figure 8C, Table S1).Thus, our data support the notion that the OA production pathway is not the only longevity-promoting branch downstream of NHR-80.Finding other lifespan promoting NHR-80 targets will be an important goal in the future.
Our data are also compatible with the non-exclusive hypothesis that OA may act as a NHR-80 ligand.Although we do not provide biochemical evidence for this interaction in C. elegans, several articles have shown that long chain free fatty acids act as a ligand for the NHR-80 homolog in Drosophila and mammals, HNF4 [19][20][21][22][23].It will be interesting to critically test this possibility in the future by either performing structural studies or transactivation assays to test the binding of OA to the NHR-80 ligand binding domain.
The NHR-80/FAT-6/OA and the KRI-1/DAF-16/K04A8.5 Pathways Are Independent Our results suggest that lipid metabolism and, in particular, fatty acid desaturation links signals from the germ line to longevity.This confirms previous findings suggesting a link between fat metabolism and longevity in germline-less animals through the K04A8.5 lipase [12].However, results presented here argue that the NHR-80/FAT-6/OA and the KRI-1/DAF-16/K04A8.5 pathways can act independently.First, we observed that the increase in nhr-80 mRNA levels observed in glp-1(e2141ts) mutant animals still occurs in the absence of daf-16 (Figure 5B).Second, the translocation of DAF-16 into intestinal nuclei occurs in the absence of nhr-80 (Figure 5A).Third, we show that NHR-80 and DAF-16 have distinct transcriptional targets, although some overlap between the two transcription factors exists.The fat-6 is one of its targets and it encodes for a Stearoyl Co-A D9 Desaturase that produces OA.OA production is required, but not sufficient to promote longevity in the absence of proliferating GSCs.Thus, the FAT-6/OA acts in concert with other NHR-80 critical targets (Crit.targets).The DAF-16/ K04A5.8 and the NHR-80 pathways can act independently, but DAF-12 is required for NHR-80 function.DAF-12 and NHR-80 do not interact at a transcriptional level and we propose that DAF-12 and NHR-80 targets interact to promote longevity (fat-6 is not a DAF-12 target).The critical targets could be shared by DAF-12 and NHR-80 or distinct.Alternatively, DAF-12 may physically interact directly with NHR-80 (grey double head arrow).doi:10.1371/journal.pbio.1000599.g009transcription of fat-6 is elevated in glp-1(e2141ts) mutant animals in a NHR-80 dependent manner (Figure 6B) and K04A8.5 mRNA levels are increased in a DAF-16 dependent way (Figure 6D; [12]).fat-5 is also increased, but its induction relies on both DAF-16 and NHR-80 (Figure 6A).Fourth, the overexpression of nhr-80 increases the lifespan of daf-16(mu86);glp-1(e2141ts) mutant animals (Figure 5C, Table S1), demonstrating that NHR-80 signaling does not require the presence of daf-16.Finally, OA does not increase the lifespan of daf-16(mu86);glp-1(e2141ts) mutants, suggesting that, similar to glp-1(e2141ts) mutants, the SCD activity is already elevated in these animals (Figure S12A, Table S1).
Thus, our data confirm that germline ablation leads to an alteration of fat metabolism that is required for extending lifespan but challenge the view that this is centered on insulin signaling.Germ cell removal extends longevity in response to two independent fat modifying pathways: the NHR-80/FAT-6/OA and the KRI-1/DAF-16/K04A8.5 pathways.
The NHR-80/FAT-6/OA Pathways Requires the Presence of DAF-12 DAF-12 is also required for longevity extension by depletion of the germ line [1], and it seems to act upstream of DAF-16 since its presence is partially required for DAF-16 translocation into intestinal nuclei in response to germline ablation [4].However, recent observations suggest that DAF-12 can also act in parallel to DAF-16 [11,12].Our data show that the NHR-80/FAT-6/OA and DAF-12 act in concert, independently of DAF-16.Indeed, overexpressing nhr-80 and providing exogenous OA do not affect the lifespan of glp-1(e2141ts);daf-12(rh61rh411) mutant animals (Figures 5D, S13B, Table S1), indicating that the NHR-80/FAT-6/OA pathway requires the presence of DAF-12.The expression levels of daf-12 are not affected in glp-1(e2141ts);nhr-80(tm1011) mutants and the induction of nhr-80 mRNA levels occurs in the absence of daf-12, but it is slightly decreased relative to glp-1(e2141ts) mutants, suggesting that, similar to DAF-16, DAF-12 modulates the NHR-80 response (Figure 5F).Our data therefore indicate that the DAF-12/NHR-80 interaction is not strictly transcriptional.Rather, we propose that DAF-12 and NHR-809s targets act together to promote longevity or that NHR-80 and DAF-12 share some critical targets for longevity.Since fat-6 mRNA levels are normally induced in glp-1(e2141ts);daf-12(rh61rh411) double mutants relative to wild type, we can exclude the possibility that the FAT-6/OA branch acts downstream of DAF-12.Thus, other targets must be involved.We propose a model where DAF-12 and NHR-80 targets cooperate to promote longevity in concert with the FAT-6/OA branch.However, it is also possible that DAF-12 interacts directly with NHR-80.Whether DAF-12 and NHR-80 only cooperate through their critical targets or whether they physically interact remains to be determined.
The NHR-80/FAT-6/OA Pathway Does Not Require the Activation of Lipophilic-Hormone Signaling Pathway by Dafachronic Acid We were surprised to find that overexpressing nhr-80 extends the lifespan of glp-1(e2141ts);daf-9(rh50) double mutants in the presence or in the absence of D7 dafachronic acid, the DAF-12 ligand produced by DAF-9 ( [10]; Figures 5E and S5, Table S1).This suggests that NHR-80 interacts with DAF-12 independently of DAF-9 and D7 dafachronic acid and explains why treating wild type animals overexpressing nhr-80 with D7 dafachronic acid (DAF-12 ligand) fails to recapitulate germline longevity (unpublished data).We note that the Kenyon lab had already suggested that DAF-12 might function in a DAF-9 independent manner [4].Indeed, mutations causing DAF-16 to be constitutively nuclear extends the lifespan of germline-less animals lacking daf-9, but not of animals lacking daf-12 [4].
It is possible that DAF-12 can also be activated by other ligands or that it can interact with NHR-80 under its unliganded form.DAF-12 might also be activated by an unknown cofactor or by a heterodimeric partner nuclear receptor to interact with NHR-80.The finding that the overexpression of nhr-80 fails to extend the lifespan of wild type animals (where DAF-12 is not activated; Figure 4A, Table S1) suggests indeed that other ligand(s) or coactivators may only be present in germline-less animals.However, we cannot exclude that other, yet unidentified, modulators preclude lifespan extension through nhr-80 overexpression.In the future, it will be interesting to explore these non-exclusive possibilities.

Fat Composition, Not Fat Content, Correlates with Germline Longevity
Similar to daf-2 mutant animals, germline-depleted animals store more fat than wild type animals [24].This is not a systematic trait of long-lived animals since diet-restricted animals store less fat.Thus, higher fat content is not a general cause for life extension.However, it is still not clear whether high fat content extends lifespan when the germ line is depleted.First, the high fat content phenotype [24] is hard to reconcile with the finding that the K04A8.5 lipase is induced in these animals [12].Although RNAi against the K04A8.5 lipase increases Nile Red staining in glp-1(e2141ts) mutant animals, these data are difficult to interpret since it was shown that Nile Red does not stain fat reliably [24].It is possible that the K04A8.5 lipase changes the composition of fat rather than altering overall fat content by degrading a subset of TAGs only.Second, the nhr-80(tm1011) allele does not affect the overall fat content of glp-1(e2141ts) mutants (unpublished observation).This indicates that fat content is not the cause for longevity extension in germline-depleted animals since glp-1(e2141ts);nhr-80(tm1011) mutants are short-lived.Our data suggest that fatty acid desaturation, and therefore, fat composition, is altered in germline-less animals.This modification of fat composition does not directly impact fat content but correlates with longevity.Indeed, our data clearly establish that OA production is not sufficient to extend lifespan, but it is strictly required.Further work should aim at understanding all aspects of fat metabolism that matter to germline-mediated longevity.

A Conserved Mechanism?
Although nhr-80 is one of 269 genes encoding nuclear hormone receptors in C. elegans that are homologous to the mammalian HNF4 gene [25], there are interesting clues that suggest that the mechanism we describe in this work may be conserved.First, we report that NHR-80 and OA must act in concert to promote longevity.Although we failed to provide clear evidence that OA activates NHR-80, structural and biochemical data showing a direct interaction between long chain fatty acids and HNF4 in Drosophila and mice [19][20][21][22][23] suggest that a similar regulation may also occur in C. elegans and that OA may be an NHR-80 ligand.Second, we show that fat-6 is a target of NHR-80.Strikingly, SCD1, the mammalian homolog of FAT-6, is strongly upregulated in response to ovariectomy in mice [26] and is a known target for several nuclear receptors in mammals.Taken together, these data suggest that our work may be relevant in mammalian systems.Further experiments are underway to firmly identify the NHR-80 analog.

glp-1(e2141ts) Suppressor Screening
RNAi clones from the Ahringer's Library were grown overnight at 37uC in LB containing Ampicillin (50 mg/mL) and Tetracyclin (12.5 mg/mL).Each RNAi clone was spread on NGM plates supplemented with carbenicillin (25 mg/mL).RNAi expression was induced by the addition of 1 mM IPTG on top of seeded bacteria.About 150 glp-1(e2141ts) (CF1903) and wild type L1 larvae were added on each RNAi plate and incubated at 25uC until day 1 of adulthood.At day 20 the proportion of worm alive was visually inspected.Clones that led to a majority of dead worms at this time for both wild type and glp-1(e2141ts) mutant animals were selected for further analysis.

Plasmid Construction
The plasmid pJG01(nhr-80p::nhr-80::gfp) containing nhr-80 tagged with GFP driven by the endogenous nhr-80 promoter was constructed by amplifying genomic DNA from 1.4 kb upstream from the start codon until the end of the nhr-80 coding sequence without the stop codon.The 4kb XmaI/KpnI nhr-80 fragment generated was inserted upstream and in frame of the GFP sequence in the worm expression vector pPD95.75(Fire Lab Vector Kit-Addgene).Essential parts of the plasmid pJG01 were sequenced.

Lifespan Analyses
Lifespan assays were conducted according to standard protocols [29].All assays were performed at 20uC, starting from day 1 of adulthood.For glp-1(e2141ts) mutants, lifespan assays (and associated controls), animals were grown at 25uC from the L1 stage until the L4 stage to prevent germ cell proliferation.The rest of the assay was performed at 20uC.For cyc-1 RNAi experiments (and associated controls), larvae were left at 15uC for 24 h, shifted to 25uC for 24 h, and shifted back to 20uC at the L4 stage to avoid dauer formation or other larval arrest.All strains containing the glp-1(e2141ts) allele used in lifespan assays were completely sterile.Unless mentioned otherwise, lifespan assays of fertile strains were conducted on plates supplemented with 15 mM 5-Fluoro-Uracil in order to prevent progeny from hatching.Worms crawling off the plate, exploding, bagging, or contaminated were excluded.Plotting and statistical analysis were done using the Biopylife software.Biopylife was designed by students from INSA de Lyon using the following free softwares: R, MySQL, Python, MySQLdb, Qt4, PyQt4, and MacTeX.Biopylife allows easy plotting of lifespan assays and determines mean, maximal lifespan, and p values using log-rank (Mantel-Cox) statistics.

Oleic Acid Supplementation
NGM medium were prepared with the addition of 100 mM oleic acid (NuChek Prep) right before pooring plates.

D7 Dafachronic Acid Supplementation
D7 dafachronic acid was added on top of seeded bacteria to make a final NGM concentration of 100 nM.Worms were transferred on fresh plates every other day.

Dietary Restriction
DR was performed through bacterial deprivation from the fifth day of adulthood [30].

qRT-PCR
For each gene, analyses were performed on triplicate of at least five independent extracts.For each analysis, we present two statistical tests: the parametric unpaired two-tailed t student test and the non-parametric Wilcoxon rank test.While the former test assumes a Gaussian distribution of the samples, the latter does not.RNA extraction and purification.Total RNA was isolated from synchronized populations of day-2 adult worms (about 3,000 individuals per condition) using the following method.Worms were harvested and washed three times with M9 buffer and twice with DEPC water.TRIZOL reagent (MRC) was added to the worm pellet (Trizol/worm pellet ratio was 2/1) and the mixture was vigorously shaken for at least 1 min.The mixture was frozen at 280uC overnight or for a longer period prior to the next RNA extraction steps.
Frozen worms were then placed on ice, vortexed 5 min, and settled at room temperature.The trizol/worm mix was transferred to tubes and chloroform (200 mL per 1 mL trizol/worm mix) was added to each tube.Tubes were shaken for 15 s, incubated at room temperature 2 min, and centrifuged at 12,000 g for 5 min.The upper aqueous phase was transferred in a new tube.An equal volume of 70% ethanol (prepared in DEPC water) was added to the aqueous phase and tubes were gently mixed by inversion.
RNA extraction was then performed following the Rneasy (Qiagen) kit instructions, including the optional step of DNAseI digestion on column.RNAs were eluted in 50 mL RNAse free water.RNA concentrations were determined using a nanodrop spectrophotometer.RNA extracts were used directly or kept at 280uC.
RNA reverse transcription.cDNAs were created using the iScript cDNA Synthesis kit (Bio-Rad) using 1 ng RNA.
Quantitative real time PCR.SybrGreen real-time Q-PCR experiments were performed as described in the LightCycler FastStart DNA MasterPlus SYBR Green I manual.
Components for a reaction (slightly modified).(Water PCR grade 6 mL, PCR primers 1 mL, Master Mix 2 mL) Each pre-cooled capillary was filled with 9 mL of Q-PCR mix and 1 mL of cDNA template at a 1/10 dilution (or 1/10, 1/40, 1/160, 1/640, and 1/2,560 dilutions for standards).Quantitative qRT-PCR reactions were carried out on a Light Cycler 1.5 (Roche).Data were collected using RNA from at least three independent C. elegans cultures.Standard curve method was used to determine the relationship between mRNA abundance and PCR cycle number.Each primer sets were calibrated using serial dilutions of cDNA preparations.Primer sets were also calibrated by performing qRT-PCR reactions on serial dilutions of C. elegans genomic DNA.mRNA levels of the actin gene act-1 were used for normalization.

DAF-16 Nuclear Localization Assays
DAF-16 nuclear localization assays were performed as previously described [4].Briefly, daf-16(mu86);glp-1(e2141ts);muIs109 [Pdaf-16::gfp::daf-16] mutant animals were injected with water (negative control), kri-1 RNAi (control), or nhr-80 RNAi at 1 mg/mL and recovered at 15uC.The next day, animals were shifted to 20uC to lay eggs for several hours on OP50 plates.To obtain F1 animals with a glp-1(e2141ts) phenotype, eggs were shifted to 25uC, 24 h after being laid and shifted back to 20uC at the L4 stage.To obtain F1 animals without any glp-1(e2141ts) phenotype, eggs were left at 20uC during the whole course of the experiment.On day 1 of adulthood, F1 animals were assayed for DAF-16 nuclear localization in intestinal cells using a fluorescent microscope.Animals were scored as having nuclear-localized DAF-16 if the majority of intestinal cells displayed a distinct concentration of GFP in the nucleus.Twenty-five to 50 animals were analyzed for each condition.

NHR-80::GFP Quantification
The MetaMorph software was used for image processing and ImageJ was used for fluorescence quantification.Fluorescence microscopy was performed using an axioplan microscope (Zeiss), with a cooled charge-coupled device camera.GFP images of sterile and fertile worms (i.e., grown until L4 at permissive or restrictive temperature and then switched at 20uC) were collected at day 1 of adulthood.Regions of Interest (ROI) were manually designed within either the two first intestinal nuclei and head neuronal nuclei; the software (ImageJ) calculated the mean intensity value for pixel intensity within the ROIs.At least 10 animals were analyzed for each condition.Averages and errors are presented in our graphs.

Fatty Acid Methyl Ester Analysis
Each condition was analyzed in three independent extracts.For each extract, 4,000 to 5,000 adult worms at day 1 of adulthood grown at 25uC on ht115 were washed three times in M9 buffer.Worms were then homogenized in 2 ml of methanol/ 5 mM EGTA (2:1 v/v) with FAST-PREP (MP Biochemicals). 100 ml were evaporated and the dry pellets were dissolved in 0.2 ml of NaOH (0.1 M) overnight and proteins were measured with the Bio-Rad assay.Lipids corresponding to the total homogenate were extracted according to Bligh and Dyer [31] in chloroform/ methanol/water (2.5/2.5/2.1, v/v/v), in the presence of the internal standards glyceryl triheptadecanoate (5 g).The lipid extracts were hydrolysed in KOH (0.5 M in methanol) at 50uC for 30 min, and transmethylated in boron trifluoride methanol solution 14% (SIGMA, 1 ml) and hexane (1 ml) at 80uC for 1 h.After addition of water (1 ml) to the crude extract, FAMEs were extracted with hexane (3 ml), evaporated to dryness, and then dissolved in ethyl acetate (180 ml).FAME (1 ml) were analyzed by gas-liquid chromatography [32] on a Clarus 600 Perkin Elmer system using a Famewax RESTEK fused silica capillary columns (30 m60.32 mm i.d., 0.25 m film thickness).Oven temperature was programmed from 110uC to 220uuC at a rate of 2uC per min and the carrier gas was hydrogen (0.5 bar).The injector and the detector were at 225uC and 245uC, respectively.

Figure 1 .Figure 2 .
Figure 1.NHR-80 is required for lifespan extension caused by depletion of the germ line.(A) nhr-80 is required for glp-1(e2141ts) mediated longevity (p,0.0001;45% reduction in mean lifespan) but does not affect the lifespan of animals with a germ line (p = 0.55; mean lifespan of 15 d for wild type and nhr-80(tm1011)).All lifespan analyses were conducted under similar conditions, with a transient temperature shift to 25uC and without FU.(B) Lifespan of both sterile and fertile mes-1(bn7) mutants either subjected to nhr-80 RNAi or not.Mean lifespan of the sterile mes-1(bn7) mutant is 21 d and 14 d when submitted to the empty vector (ev) and nhr-80 RNAi, respectively (p,0.0001).Lifespan analyses were performed independently at least twice.The p values were calculated using the log rank (Mantel-Cox) analyses.doi:10.1371/journal.pbio.1000599.g001

Figure 3 .
Figure 3.In response to germline depletion, nhr-80 mRNA levels increase and NHR-80::GFP is induced in intestinal nuclei.(A and B) Using glp-1(e2141ts);lynEx[(nhr-80p::nhr-80::gfp) myo-2p::DsRed] mutant animals, we observed that NHR-80 is expressed in the nuclei of head and tail neurons (white arrows) and of the intestine (blue arrows).Because of the high intensity of the myo-2p::DsRed marker, the pharynx was artificially masked, outlined in red.NHR-80::GFP is nuclear at both permissive and restrictive temperature.(C) The intensity of the signal is increased almost 1.6-fold at restrictive temperature in intestinal nuclei (Wilcoxon rank-sum test p value ,0.001).(D) In contrast, when similar measurements were made in neuronal nuclei, no significant difference was detected (Wilcoxon rank-sum test p value is non-significant).(E) nhr-80 mRNA levels are increased in glp-1(e2141ts) mutants as measured by qRT-PCR (5.6-fold increase; Wilcoxon rank-sum test p value ,0.05 when compared with N2).doi:10.1371/journal.pbio.1000599.g003

Figure 9 .
Figure 9. Model for NHR-80 function.Proliferating GSCs negatively regulate the translocation of DAF-16 within intestinal nuclei, the activation of DAF-12, and the expression of nhr-80.DAF-12 is central because it participates in the translocation of DAF-16 into intestinal nuclei (grey arrow) and is required for NHR-809s longevity action.In response to depletion of the germ line, NHR-80 is up-regulated and becomes transcriptionally functional.fat-6 is one of its targets and it encodes for a Stearoyl Co-A D9 Desaturase that produces OA.OA production is required, but not sufficient to promote longevity in the absence of proliferating GSCs.Thus, the FAT-6/OA acts in concert with other NHR-80 critical targets (Crit.targets).The DAF-16/ K04A5.8 and the NHR-80 pathways can act independently, but DAF-12 is required for NHR-80 function.DAF-12 and NHR-80 do not interact at a transcriptional level and we propose that DAF-12 and NHR-80 targets interact to promote longevity (fat-6 is not a DAF-12 target).The critical targets could be shared by DAF-12 and NHR-80 or distinct.Alternatively, DAF-12 may physically interact directly with NHR-80 (grey double head arrow).doi:10.1371/journal.pbio.1000599.g009 Asterisk indicates the p value of the Wilcoxon rank-sum test as follows: *p,0.1, **p,0.05,***p,0.01.Standard deviations are displayed as error bars.