Transcriptional regulation of Caenorhabditis elegansFOXO/DAF-16 modulates lifespan

daf-16 functions as a central regulator of multiple biological processes including lifespan, development, fat storage, and stress resistance [1‐4]. The C. elegans genome resource, Wormbase (http://​www.​wormbase.​org), predicts eight putative isoforms of daf-16, with each isoform designated by a lowercase letter following the cosmid name for daf-16 (R13H8.1a-h). To verify expression of the daf-16 isoforms, we first analyzed the Gurdon Institute C. elegans expressed sequence tag (EST) database (http://​genomics.​nimr.​mrc.​ac.​uk/​online/​worm-fl-db.​html) [35, 36]. As shown in Additional file 1: Table S1, ESTs were identified that map to daf-16f (R13H8.1f), daf-16d (R13H8.1d), daf-16a (R13H8.1b/c), and daf-16b (R13H8.1a). In addition to ESTs, we examined published capped RNA (Cap-Seq) (Gu et al.[37]), total mRNA, and ribosome-associated mRNA deep sequencing datasets (Stadler et al.[38]). These datasets support the existence of daf-16f, daf-16d, daf-16a1, daf-16a2, and daf-16b. In particular, the Cap-Seq data specifically identified four spliced leader (SL1) sites corresponding to the capped 5’ ends of daf-16f, daf-16d, daf-16a1/a2, and daf-16b, as indicated in green in Figure 1. Therefore, our findings verify the existence of five isoforms in contrast to eight that were predicted initially for DAF-16 - (DAF-f (R13H8.1f), DAF-16d (R13H8.1d), DAF-16a1 (R13H8.1b), DAF-16a2 (R13H8.1c), and DAF-16b (R13H81.a).

Previous RNAi and transgenic studies examined the function of different daf-16 isoforms (daf-16a, b, d/f) [19‐21], which revealed that daf-16b functions in dauer formation, while daf-16a and daf-16d/f contribute to lifespan regulation. According to our previous studies [21], daf-16a constructs could not rescue the lifespan phenotype of a complete loss of daf-16, and daf-16d transgenes were unstable and did not consistently rescue daf-16 loss-of-function. However, transgenes expressing the daf-16f isoform or daf-16d/f showed similar expression patterns and consistently rescued the complete loss of daf-16[21] (Figure 2). Moreover, based on the CapSeq data, both daf-16f and daf-16d have the same transcription start site. Therefore, we continue to refer to these isoforms as daf-16d/f in this study.

Dissecting the relative contributions of DAF-16d/f and DAF-16a in the regulation of longevity has been challenging due to incomplete knockdown by isoform-specific RNAi [21] and the lack of isoform-specific mutants. This is further complicated by the existence of a dose-dependent effect of daf-16a and daf-16d/f transgenes [21]. To expand our analysis of the daf-16 dosage effect on lifespan, we first generated a new low-copy daf-16a transgene by microparticle bombardment, which we refer to as daf-16a:gfp ^HT . In addition, we chose three commonly used high-copy daf-16a transgenes, referred to as daf-16a::gfp ^CF derived from strain CF1407 [20], daf-16a::gfp ^TJ from TJ356 [39], and daf-16a::gfp ^GR from RX86 [40]. For all the analyses that follow, we first placed each daf-16 transgene into the same genetic background which is a null allele of the endogenous daf-16 locus [daf-16(mgDf50)].

To compare the levels of daf-16a in the different transgenic strains, we measured the level of daf-16a expression from each transgene in the daf-16(mgDf50); daf-2(e1370) background and compared it to the endogenous level of daf-16a in daf-2(e1370) mutants (Figure 2A, Additional file 1: Table S2). The level of daf-16a expressed from the newly generated daf-16a ^HT transgenic strain was approximately two-fold higher than the endogenous daf-16a in the daf-2(e1370) mutant, while the commonly used daf-16a transgenes were expressed at 10- to 400-fold higher levels than endogenous daf-16a. These results show the following order for levels of the daf-16a transcript: endogenous daf-16a < HT < CF < TJ < < GR. Consistent with the mRNA expression data, the level of DAF-16a protein from the transgenic strains also increases (Additional file 1: Figure S1). The GFP intensity in the four transgenic strains correlated with the level of expression of DAF-16a (data not shown).

Next, we determined the nuclear:cytosolic ratio of DAF-16a in the transgenic strains expressing different levels of DAF-16a (Additional file 1: Figure S2) in daf-2(e1370) background. DAF-16a was almost entirely in the nucleus when expressed at low levels from the daf-16a::gfp ^HT transgene. By contrast, DAF-16a was mainly cytosolic in the daf-2(e1370); daf-16a ^GR strain, which shows the highest level of DAF-16a expression, while the daf-2(e1370); daf-16a::gfp ^CF and daf-2(e1370); daf-16a::gfp ^TJ strains showed an intermediate nuclear localization of DAF-16a. Importantly, since these different transgenic strains were generated with different methods and may be integrated at different locations, this indicates that it is the level of mRNA and protein of DAF-16a that determines its regulatory ability.

We reasoned that if a daf-16a transgene is expressed at close to endogenous levels, the daf-16a transgenic animals should not live longer than wild-type animals expressing endogenous levels of all three DAF-16 isoforms. Interestingly, our low-copy daf-16a::gfp ^HT transgenic worms lived shorter than wild-type (Figure 2B, Additional file 1: Table S3), suggesting that low level expression of daf-16a alone cannot replace loss of the endogenous daf-16 for lifespan regulation. However, the three high-copy daf-16a transgenic strains lived longer than wild type (Figure 2B, Additional file 1: Table S3). Therefore, these data suggest that the level of DAF-16a expressed from the daf-16a::gfp ^CF , daf-16a::gfp ^TJ , and daf-16a::gfp ^GR transgenes exceeds the inhibitory capacity of the IIS pathway. This gene dosage effect was even more pronounced when each daf-16a transgene was analyzed in the daf-16(mgDf50); daf-2(e1370) double mutant background (Figure 2C, Additional file 1: Table S3). As shown in Figure 2B and consistent with previous findings [19‐21], when expressed at low levels from a daf-16a::gfp ^HT transgene, DAF-16a cannot fully compensate for a complete loss of DAF-16 function (Figure 2B) [21]. An increase in lifespan was observed with increased expression of DAF-16a (daf-16a::gfp ^CF , daf-16a::gfp ^TJ ). However, as the dosage of DAF-16a continued to increase (for example, daf-16a::gfp ^GR ), a reduction in lifespan was seen (Figure 2C, Additional file 1: Table S3). Consistent with this finding, unlike the other daf-16a transgenic strains, the daf-16a::gfp ^GR showed severely delayed development and reduced fecundity, suggesting that extremely high levels of DAF-16a are toxic (Additional file 1: Figure S3). Together, these findings suggest that the three commonly used strains for analyses of DAF-16:GFP do not accurately reflect the endogenous function of DAF-16a in lifespan regulation. Furthermore, these results indicate that low-copy daf-16a strains are more relevant for functional analyses as well as for DAF-16 localization assays (Additional file 1: Figures S2).

Next, we examined the low-copy daf-16d/f::gfp ^HT transgene [21], which had approximately four-fold higher transcript level than endogenous daf-16d/f levels in the daf-2(e1370) mutant background (Additional file 1: Table S2). Unlike high-copy daf-16a and daf-16d/f transgenic worms [21], these daf-16(mgDf50); daf-16d/f::gfp ^HT transgenic worms did not live longer than wild-type (Figure 2B, Additional file 1: Table S3), suggesting that DAF-16d/f activity in the low-copy daf-16d/f transgenic strain is regulated within the inhibitory capacity of the IIS pathway. Even though the strains were generated by different methods, we suggest that it is the level of the DAF-16 isoform protein that leads to an environment that does not accurately reflect endogenous levels of DAF-16.

The low-copy daf-16 isoform transgenes were crossed into the background of another widely used daf-2 allele, daf-2(e1368), which exhibits weaker longevity and dauer arrest phenotypes than daf-2(e1370)[5, 41]. Based on our lifespan data in the daf-16(mgDf50); daf-2(e1370) background, we expected that the daf-16d/f::gfp ^HT transgene would be sufficient to fully complement the lifespan of a daf-16(mgDf50); daf-2(e1368) double mutant. Yet, to our surprise, daf-16(mgDf50); daf-2(e1368); daf-16d/f::gfp ^HT transgenic worms lived significantly shorter than the daf-2(e1368) single mutant (Figure 3A, Additional file 1: Figure S4, Additional file 1: Table S3, Additional file 1: Table S4). In contrast, the daf-16a::gfp ^HT transgene rescued lifespan to similar levels in both daf-2(e1368) and daf-2(e1370) mutant backgrounds (Additional file 1: Figure S4, Additional file 1: Table S3, Additional file 1: Table S4).

To further test the effect of DAF-16d/f in the daf-2(e1370) and daf-2(e1368) mutant backgrounds, we examined dauer formation as an additional output of DAF-16 activity. At the restrictive temperature of 25°C, both daf-2(e1370) and daf-2(1368) form 100% dauers whereas daf-16(mgDf50); daf-2(e1370) and daf-16(mgDf50); daf-2(e1368) double mutants do not form any dauers. However, once again, the DAF-16d/f transgene had different effects in the two daf-2 alleles: daf-16(mgDf50); daf-2(e1368); daf-16d/f::gfp ^HT transgenic worms formed approximately 39% dauers, whereas daf-16(mgDf50); daf-2(e1370); daf-16d/f::gfp ^HT worms formed approximately 98% dauers at 25°C (Additional file 1: Figure S5A). At the semi-permissive temperature of 20°C, a similar trend was observed. This is in contrast to daf-16a transgenic worms where dauer formation was comparable in both daf-2 mutant backgrounds (Figure 3B, Additional file 1: Figure S5).

Examining the DAF-16 nuclear/cytosolic localization, DAF-16d/f was enriched in the nucleus in the daf-2(e1370) background, but more cytosolic in the daf-2(e1368) background (Figure 3C,D). In contrast, DAF-16a was primarily enriched in the nucleus in both daf-2(e1368) and daf-2(e1370) backgrounds (Figure 3C,D). Thus, the DAF-16 localization data correlated with both functional outputs - lifespan and dauer formation. Whereas DAF-16a is activated to a similar level in both alleles, DAF-16d/f is less active in the daf-2(e1368) mutant than in the daf-2(e1370) mutant. Therefore, these data show that DAF-16a and DAF-16d/f exhibit different thresholds for inhibition by the upstream IIS pathway.

Although the spatial expression patterns of daf-16 have been extensively investigated [19, 20, 39], the temporal expression of daf-16 has not been studied. We monitored the expression of the daf-16 isoform::gfp strains throughout development and for the first few days of adulthood. Interestingly, as shown in Additional file 1: Figure S6, the GFP intensity of the daf-16d/f::gfp transgenic worms dramatically increased in the adult stage compared to larval stages.

Next, we asked if the increase in GFP intensity in adult daf-16d/f::gfp transgenic animals correlates with the level of the endogenous transcript. Using qRT-PCR, we found that each of the daf-16 isoforms was expressed at a constant level throughout larval development. However, as worms aged, the daf-16d/f transcript dramatically increased in both wild-type and long-lived daf-2(e1370) mutants, while the daf-16a transcript increased only slightly (Figure 4A,B, Additional file 1: Figure S7-S9). To further confirm this result, we tested the upregulation of DAF-16a vs DAF-16d/f at the protein level. Since it is not possible to distinguish differences in endogenous levels of DAF-16a and DAF-16d/f protein, we were limited to examining changes in DAF-16 protein levels in strains that only bear one DAF-16 isoform (daf-2;daf-16;daf-16a::gfp or daf-2;daf-16;daf-16d/f::gfp). Using these transgenic strains, consistent with our mRNA data, the age-dependent increase in the levels of DAF-16d/f was also observed at the protein level (Figure 4C,D).

The qRT-PCR assay is a read-out of the steady-state level of mRNA, but the DAF-16::GFP fusion fluorescence is the product of both transcriptional and post-transcriptional regulation. Therefore, to measure the transcriptional promoter activity of each daf-16 isoform, we generated daf-16 promoter::GFP fusions (Pdaf-16::gfp). The promoter of each daf-16 isoform was placed upstream of the GFP open reading frame fused to the unc-54 3′UTR, which is known to be absent of temporal regulatory elements [42]. We reasoned that if increased transcription is responsible for the elevated mRNA level in young adults, then the GFP intensity should become brighter in the promoter::gfp transgenic worms as the worms age. In the Pdaf-16a::gfp worms, the GFP signal throughout the body increased slightly with age. In contrast, in Pdaf-16d/f::gfp worms, GFP expression increased dramatically throughout the intestine (data not shown Figure 4E). Therefore, across multiple independent experiments we found that transcription of daf-16, particularly daf-16d/f, was regulated in an age-dependent manner.

daf-16 expression in the intestine is critical for lifespan regulation in C. elegans[21, 43], and our temporal and spatial expression data indicate that daf-16d/f mRNA is dramatically upregulated in the intestine as worms age. To determine the importance of DAF-16 upregulation in aging worms, we used RNAi to maintain daf-16d/f expression in the adult stage at a level comparable to that in the L4 stage, thereby preventing the upregulation of DAF-16d/f during aging. To achieve the correct knockdown level, we prepared serial-dilutions of daf-16 RNAi bacteria (from no dilution to 64-fold dilution) and fed daf-16(mgDf50); daf-2(e1370); daf-16d/f::gfp ^HT worms from the L4 stage to day 2 adults. After 2 days of growth on the daf-16 RNAi plates, worms were transferred to control RNAi plates. We found that worms grown on four-fold to 16-fold diluted daf-16 RNAi bacteria maintained daf-16d/f expression at levels similar to that of the L4 worms (Additional file 1: Figure S10). Importantly, the lifespan of daf-16(mgDf50); daf-2(e1370); daf-16d/f::gfp ^HT worms fed on the four-fold to 16-fold diluted daf-16 RNAi bacteria was reduced as compared to animals grown on control bacteria (Figure 4F, Additional file 1: Table S5). Therefore, the upregulation of daf-16d/f gene expression at the young adult stage is an important determinant of C. elegans lifespan.

Our findings suggest that transcriptional regulators expressed in the intestine activate daf-16d/f transcription in an age-dependent manner. To identify regulator(s) of daf-16d/f gene expression, we performed an RNAi screen of 892 transcriptional regulators using Pdaf-16d/f::gfp transgenic worms. RNAi of two transcriptional regulators reduced Pdaf-16d/f::gfp expression: elt-2, a GATA transcription factor; and swsn-1, a core component of the SWI/SNF chromatin remodeling complex (Figure 5A, B,C). elt-2 is essential for intestinal development [31, 32, 44, 45] as well as innate immune responses to bacterial and fungal infection [46] and directly regulates the expression of >80% of intestinal genes, including genes that are downstream of daf-16 (dod gene) [32].

Because worms fed with elt-2 RNAi from hatching develop an abnormal intestinal structure and die within a few days, we used post-developmental RNAi to silence elt-2 beginning at the L4 stage when intestinal development is complete. Indeed, post-developmental elt-2 RNAi treatment also reduced Pdaf-16d/f::gfp expression (Figure 5B). When we examined the endogenous daf-16 transcripts, we found that post-developmental elt-2 RNAi treatment reduced the expression of endogenous daf-16d/f compared to control RNAi (Figure 5C). In contrast, the Pdaf-16a::gfp was only marginally downregulated on elt-2 RNAi at day 2 and endogenous daf-16a expression was not changed, likely due to the fact that its intestinal expression is only a minor portion of total daf-16a (Additional file 1: Figure S11) [21]. Accordingly and consistent with a recent report [47], elt-2 RNAi reduced the lifespan of the daf-2(e1370) single mutant (expressing endogenous daf-16), as well as daf-16(mgDf50); daf-2(e1370); daf-16a::gfp ^HT and daf-16(mgDf50); daf-2(e1370); daf-16d/f::gfp ^HT transgenic worms (Figure 6A, C, D, Additional file 1: Table S6). Relative to the control RNAi, elt-2 RNAi did not shorten the lifespan of a daf-16(mgDf50); daf-2(e1370) mutant (Figure 6B), suggesting that elt-2 RNAi shortens lifespan by reducing intestinal daf-16 gene expression and not by a daf-16-independent pathway.

We analyzed several additional GATA transcription factors for a role in daf-16 transcriptional regulation and longevity, including elt-4 and elt-7, which are expressed in the intestine [44], and elt-3, elt-5, and elt-6, which were previously implicated in lifespan regulation [48]. Only elt-4 RNAi reduced the expression of the Pdaf-16d/f::gfp transcriptional reporter (Additional file 1: Figure S13), though less dramatically than elt-2 RNAi. Furthermore, only elt-4 RNAi reduced the lifespan of both daf-16(mgDf50); daf-2(e1370); daf-16a::gfp ^HT and daf-16(mgDf50); daf-2(e1370); daf-16d/f::gfp ^HT transgenic worms (Figure 6E–H, Additional file 1: Table S6). Neither Pdaf-16d/f::gfp expression nor lifespan was affected by elt-7, elt-3, elt-5, or elt-6 RNAi (Additional file 1: Figure S13 and Additional file 1: Table S6). These data are consistent with a recent study [49] suggesting that the previously reported role of ELT-3 in lifespan regulation [48] should be re-evaluated. Our findings indicate that the ELT-2 and ELT-4 GATA factors promote longevity by regulating intestinal daf-16 expression.

The second positive clone we obtained from our screen was swsn-1, which encodes a core component of the highly conserved SWI/SNF nucleosome-remodeling complex [50]. In C. elegans, SWI/SNF is required for asymmetric cell division [51] and differentiation [52]. Prior to the L4 stage, Pdaf-16d/f::gfp expression was unaffected by swsn-1 RNAi (Figure 5A,B). In adult worms, however, the upregulation of GFP in the mid-intestine is severely reduced by swsn-1 RNAi (Figure 5B). Expression of endogenous daf-16d/f was also reduced in animals exposed to swsn-1 RNAi (Figure 5C, Additional file 1: Figure S12B). Interestingly, swsn-1 RNAi did not significantly change the expression of Pdaf-16a::gfp (Additional file 1: Figure S11C), suggesting that SWSN-1 regulates the temporal and spatial expression of daf-16d/f, specifically. RNAi targeting snfc-5 and swsn-2.1, core and accessory subunits of SWI/SNF, respectively, also reduced the temporal upregulation of daf-16d/f expression with age (Additional file 1: Figure S12) but did not change daf-16a expression, indicating that the SWI/SNF complex specifically controls daf-16d/f expression.

We next asked if SWI/SNF plays a role in lifespan regulation. Both swsn-1 and snfc-5 are essential genes [51] and hatched larvae grown on swsn-1 or snfc-5 RNAi develop into sick adults with pleiotropic defects [51, 53]. To avoid these complications in lifespan analysis, we used post-developmental (L4) RNAi to silence swsn-1. However, Pdaf-16d/f::gfp expression was largely unaffected by post-developmental swsn-1 RNAi (data not shown) and we observed only a marginal reduction of lifespan in daf-16(mgDf50); daf-2(e1370); daf-16d/f::gfp ^HT transgenic worms (Figure 6D, Additional file 1: Table S6), suggesting that post-developmental (L4) RNAi does not efficiently silence swsn-1. Worms exposed to swsn-2.1 RNAi from hatching, however, were much healthier than animals grown on swsn-1 or snfc-5 RNAi, which allowed us to assess the effect of SWI/SNF on lifespan. As shown in Figure 6H and Additional file 1: Table S6, swsn-2.1 RNAi shortened the lifespan of daf-16(mgDf50); daf-2(e1370); daf-16d/f::gfp ^HT worms, but only modestly reduced the lifespan of daf-2(e1370) or daf-16(mgDf50); daf-2(e1370); daf-16a::gfp ^HT animals (Figure 6F,G, Additional file 1: Table S6). Interestingly, a very recent paper found that SWI/SNF interacts with DAF-16 [54]. These studies revealed that a SWI/SNF-DAF-16 complex co-localizes at the promoters of DAF-16 direct targets and activates genes important for lifespan, dauer formation, and stress resistance [54]. Therefore, these results support a biochemical interaction of DAF-16 and SWI/SNF. Taken together, our data are consistent with a model in which the SWI/SNF complex directly interacts with DAF-16 to specifically promote longevity by regulating daf-16d/f expression.

All strains were maintained and handled as described [74]. Animals were grown on standard NGM plates at 15°C using standard C. elegans techniques, unless otherwise indicated. Mutants used in this study included, LGI: daf-16(mgDf50); LGIII: daf-2(e1368, e1370); unc-119(ed3). All transgenic worms have an unc-119(ed3) mutation rescued by an unc-119 ⁺ co-transformation marker. Transgenic strains generated and/or used in this study are listed in Additional file 1: Table S8.

Both genes map to Chromosome II but are 13 map units apart. daf-2(e1368) males were mated to unc-119(ed3) hermaphrodites. Approximately 20 F1 progeny were picked onto individual plates and allowed to produce progeny at 25°C. F2 daf-2(e1368) dauers were selected from each plate and allowed to recover at 15°C. From the F3 progeny, Unc worms were selected, generating daf-2(e1368); unc-119(ed3).

ii) daf-16(mgDf50) ; daf-2(e1368) unc-119(ed3) mutants

daf-16(mgDf50); daf-2(e1368) males were mated to daf-2(e1368); unc-119(ed3) hermaphrodites. Approximately 20 F1 progeny were picked onto individual plates and allowed to have progeny at 25°C. From the F2 progeny on each plate, Unc Non-dauer worms were selected and tested for the daf-16(mgDf50) mutation by PCR. Primer sequences are listed in Additional file 1: Table S7. daf-16(mgDf50); daf-2(e1370); unc-119(ed3) mutants were used for microparticle bombardment [75] to generate daf-16 isoform transgenic worms.

To generate daf-16 isoform::gfp constructs, each cDNA encoding daf-16a, daf-16d/f was cloned into the pGEM-T vector. The clones were then verified by sequencing. For the daf-16a isoform promoter construct, a SalI/BamHI fragment containing 6.0 kb upstream of daf-16a was cloned into pPD95.75 (GFP containing plasmid). For the daf-16d/f isoform promoter construct, a SphI/BamHI fragment containing 4.1 kb upstream of daf-16d/f was cloned into pPD95.75 (GFP containing plasmid). We then generated full-length daf-16a1 cDNAs using mutagenic primers with BamHI/SmaI restriction enzyme sites, which were subcloned into pPD95.75 with the upstream promoter fragment. For daf-16d/f, since we were unable to amplify the daf-16d/f specific cDNA, we generated a construct which could generate both DAF-16d1 and DAF-16d/f by subcloning the 4.1 kb upstream genomic DNA next to the 5’ putative start codon of daf-16d1 cDNA in frame [21].

To allow the constructs to be compatible with ballistic transformation, the unc-119 ⁺ gene was introduced into each vector using PCR redirected DNA recombination method [76]. To generate daf-16 isoform specific RNAi vector, NheI/HindIII fragments covering isoform specific/overlapping cDNA were cloned into L4440 vector [21]. The primers used for the PCR analysis are listed in Additional file 1: Table S7.

All strains were allowed to grow at 15°C. For Additional file 1: Figure S1, approximately 100 L4 stage worms were collected for each of the transgenic strain in 10 to 15 mL of M9 buffer. Worms were then washed once with M9. An equal amount of SDS containing loading buffer was added to the worm pellets and samples were immediately boiled to lyse the worms. Samples were cooled, centrifuged briefly, and the supernatant was loaded onto the gel. For the western blot analyses in Figure 4C and D, equal amounts (10 mL) of worms were used. For each experiment, the lysates were resolved on a 10% polyacrylamide gel by SDS-PAGE. Proteins were transferred to a nitrocellulose membrane and probed with antibody against DAF-16 [78] (1:2,500 dilution). The membrane was then reprobed with anti -α-tubulin (1:8,000 dilution).

All the strains were grown at 15°C and five L4s were picked onto plates at 15°C. The plates were left undisturbed for 7 days and photos of each of the plates were taken using Nikon Coolpix995 camera to compare the growth of the strains.

All lifespan analyses were performed at 20°C. Strains were semi-synchronized by allowing gravid adults to lay eggs overnight and then removing the adult worms. Worms were grown for several days until they reached the young adult stage at 15°C. Approximately 150 young adult worms were transferred to five freshly seeded plates containing FuDR to a final concentration of 0.1 mg/mL [77]. Worms were then scored as dead or alive by tapping them with a platinum wire every 2 to 3 days. Worms that died from vulval bursting were censored. Day 1 of the lifespan was when worms were transferred to the FuDR plate. Statistical analyses were done using the standard chi-squared-based log rank test. Lifespans were also verified on non-FuDR NGM plates (Additional file 1: Figure S15).

For measuring the nuclear:cytosolic ratio (Additional file 1: Figure S2 and Figure 3), worms were grown at 15°C until they reached young adult stage. Then, worms were mounted on glass slides in 50 mM sodium azide and visualized using Zeiss Axioscope 2+ microscope. Fluorescent images were taken using OpenLab.3.1.7 software with a Hamamatsu camera. For the quantification, only the region above intestinal cells in the pharynx was considered. Using ImageJ software, the fluorescence was quantified by measuring the pixel intensity in the nuclear versus cytosolic region. The ratio between the respective intensities was then calculated and plotted. This was repeated twice with approximately 14 to 15 worms in each assay.

RNAi knockdown by feeding was performed essentially as described in Timmons et al. [79] using a library of clones representing 892 of the predicted 937 C. elegans transcription factors (MacNeil and Walhout, unpublished). Briefly, 50 μL of an overnight culture of HT115 bacteria carrying RNAi clones was used to inoculate 1 mL LB supplemented with 50 μg/mL ampicillin. Cultures were grown, with shaking, for 6 h at 37°C. Bacteria were pelleted and resuspended in 10% of the original volume. NGM plates containing 5 mM IPTG were seeded with concentrated bacteria and allowed to dry. Synchronized L1 larvae were used to seed prepared RNAi plates. Animals were visually examined for changes in GFP expression during the adult stage.