APL-1, the Alzheimer’s Amyloid Precursor Protein in Caenorhabditis elegans, Modulates Multiple Metabolic Pathways Throughout Development
Collin Y. Ewald, Daniel A. Raps, Chris Li

Abstract

Mutations in the amyloid precursor protein (APP) gene or in genes that process APP are correlated with familial Alzheimer’s disease (AD). The biological function of APP remains unclear. APP is a transmembrane protein that can be sequentially cleaved by different secretases to yield multiple fragments, which can potentially act as signaling molecules. Caenorhabditis elegans encodes one APP-related protein, APL-1, which is essential for viability. Here, we show that APL-1 signaling is dependent on the activity of the FOXO transcription factor DAF-16 and the nuclear hormone receptor DAF-12 and influences metabolic pathways such as developmental progression, body size, and egg-laying rate. Furthermore, apl-1(yn5) mutants, which produce high levels of the extracellular APL-1 fragment, show an incompletely penetrant temperature-sensitive embryonic lethality. In a genetic screen to isolate mutants in which the apl-1(yn5) lethality rate is modified, we identified a suppressor mutation in MOA-1/R155.2, a receptor-protein tyrosine phosphatase, and an enhancer mutation in MOA-2/B0495.6, a protein involved in receptor-mediated endocytosis. Knockdown of apl-1 in an apl-1(yn5) background caused lethality and molting defects at all larval stages, suggesting that apl-1 is required for each transitional molt. We suggest that signaling of the released APL-1 fragment modulates multiple metabolic states and that APL-1 is required throughout development.

THE cause of Alzheimer’s disease (AD) remains unknown. Mutations in several genes, including the amyloid precursor protein (APP), are correlated with inherited forms of AD. Furthermore, a defining feature of AD is large numbers of senile plaques in the brain, and the plaques’ major component is a cleavage byproduct of APP. The normal function of APP and its cleavage products is still unclear. Here, we report that the Caenorhabditis elegans APP-related protein APL-1 has multiple functions during development, including modulating the insulin pathway. These results indicate that human APP may similarly regulate metabolic processes, such as the insulin pathway.

AD is a neurodegenerative disorder that leads to cognitive decline (Alzheimer’s Association 2010). One postmortem criterion in the diagnosis of AD is the presence of senile plaques in AD patients (Kidd 1964; Luse and Smith 1964; Terry et al. 1964; Krigman et al. 1965). The major component of the senile plaques is the β-amyloid peptide, which is a cleavage fragment of APP (Kang et al. 1987). Mutations and duplications of APP have been correlated with familial Alzheimer’s disease (Chartier-Harlin et al. 1991; Goate et al. 1991; Murrell et al. 1991; Cabrejo et al. 2006; Rovelet-Lecrux et al. 2006; Sleegers et al. 2006). APP is a single pass transmembrane domain protein (Kang et al. 1987), which can be cleaved by either an α- or a β-secretase to release a large extracellular fragment (sAPPα or sAPPβ, respectively); the remaining transmembrane fragment is subsequently cleaved by the γ-secretase to release a small intracellular fragment (APP intracellular domain, AICD) and, in the case of a previous β-secretase cleavage, the β-amyloid peptide (reviewed in Gralle and Ferreira 2007). The biological functions of the cleaved APP fragments, sAPPα/β and AICD, remain unclear. Crystal structures of sAPP revealed a growth-factor–like domain that is conserved and present in all mammalian APP-family members as well in C. elegans and Drosophila orthologs (Rossjohn et al. 1999), consistent with a growth factor role reported in vitro (reviewed in Mattson 1997; Schmitz et al. 2002). Conversely, fragments of sAPPβ can act as a ligand that directly binds death receptor 6 (DR6) to initiate neurodegeneration (Nikolaev et al. 2009; Kuester et al. 2011). In vivo sAPP can act as a co-factor to promote cell proliferation of ventricular zone cells (Caille et al. 2004). However, determining the function of APP in mammals is complicated by two functionally redundant proteins, APLP1 and APLP2. In mice, knockout of APP leads to mild deficits (Zheng et al. 1995), while double knockouts of APP and APLP2 or triple knockouts of APP, APLP1, and APLP2 lead to postnatal lethality (Heber et al. 2000; Herms et al. 2004). The nematode C. elegans encodes only one APP-related gene, apl-1 (Daigle and Li 1993). Like the APP family in mice, apl-1 has an essential function: knockout of apl-1 results in larval lethality due to a molting defect during the first to the second larval transition (Hornsten et al. 2007; Wiese et al. 2010). The molting defect of apl-1 knockouts is rescued not only by reintroducing an apl-1 genomic fragment, but also by reintroducing a fragment containing only the APL-1 extracellular domain (Hornsten et al. 2007). These results suggest that sAPL-1 acts during early development and is sufficient for viability. Whether sAPL-1 is necessary later in development is unclear.

In C. elegans postembryonic developmental programs and progression through larval transitions are influenced by environmental conditions (Tennessen et al. 2010; Monsalve et al. 2011; for review see Resnick et al. 2010). Under favorable conditions, C. elegans eggs hatch and develop through four larval stages (L1–L4) before reaching adulthood (Sulston and Horvitz 1977). If no food is present when the eggs hatch, the first larval stage animals halt development and go into an L1 arrest until food becomes present (Baugh and Sternberg 2006). However, if food is limited during the first and second larval stages, L2 worms enter an alternate stage called dauer (Cassada and Russell 1975). Dauer animals can survive in a harsh environment for >3 months and are resistant to heat and various noxious chemicals (Cassada and Russell 1975; Klass and Hirsh 1976; Larsen 1993; Lithgow et al. 1995). The activity of the DAF-2 insulin/IGF-1 receptor regulates both L1 arrest and dauer formation (Riddle et al. 1981; Vowels and Thomas 1992; Larsen et al. 1995; Gems et al. 1998). Complete loss of daf-2 function leads to L1 arrest and lethality (Gems et al. 1998), whereas reduced DAF-2 activity can keep newly hatched eggs in L1 arrest, even when food is present (Gems et al. 1998; Baugh and Sternberg 2006). For L1 arrest either by starvation or reduced daf-2 activity, activity of the DAF-16 FOXO transcription factor is required (Baugh and Sternberg 2006). DAF-2 insulin/IGF-1 receptor signaling negatively regulates DAF-16 FOXO activity by phosphorylation of DAF-16, thereby limiting its localization to the cytoplasm (Lin et al. 1997; Ogg et al. 1997).

Developmental programs activated by environmental conditions are integrated with a complex regulatory pathway of heterochronic genes that control the timing of stage-specific developmental programs to allow smooth transitioning into the different larval stages or adulthood (Monsalve et al. 2011; for review see Resnick et al. 2010). For the last molt of L4 to adulthood, for instance, the microRNA (miRNA) let-7 binds the 3′-UTR of mRNA from the heterochronic genes lin-41 (Slack et al. 2000) and hbl-1 (Abrahante et al. 2003) and the daf-12 nuclear hormone receptor (NHR) gene (Antebi et al. 2000; Grosshans et al. 2005) to prevent their translation. The let-7 targets, daf-12 and hbl-1, in turn negatively feed back to regulate let-7 (Bethke et al. 2009; Hammell et al. 2009; Roush and Slack 2009). During late L4 development, let-7 also regulates expression of apl-1 via hbl-1, lin-41, lin-42 heterochronic, and nhr-25 NHR genes (Niwa et al. 2008; Hada et al. 2010). RNAi of apl-1 rescues let-7 mutant phenotypes of vulva bursting, extra molts, and lethality (Niwa et al. 2008), suggesting that apl-1 is negatively regulated by let-7 during the L4-to-adult transition. Although no let-7 binding sites are present in the 3′-UTR of apl-1, other miRNA binding sites have been found in the 3′-UTR of apl-1 (Figure 1) (Niwa et al. 2008).

Figure 1 

apl-1(yn5) mutants have increased levels of the APL-1 extracellular fragment. (A) Schematic of genomic organization of apl-1 and corresponding APL-1 protein compared with human APP. The 3′-UTR of apl-1 has three possible polyadenylation sites (A), all of which are used in vivo (Daigle and Li 1993), and several possible microRNA (miR) binding sites. The yn5 allele deletes the region encoding the transmembrane domain (gray vertical line) and the entire cytoplasmic domain, and a large part of the 3′-UTR, including two of the three polyadenylation sites. S, signal peptide; KPI, Kunitz-type protease inhibitor sequence (not present in APP695 or APL-1); Aβ, amyloid peptide; the extracellular domain contains the conserved E1 and E2 domains, which share 46 and 49% similarity between C. elegans APL-1 and human APP695, respectively (Daigle and Li 1993). The E1 domain contains a putative growth-factor–like domain, copper, zinc and heparin binding domains, and 12 conserved cysteines followed by an acidic rich region (aqua box) (Daigle and Li 1993). The E2 domain contains an N-glycosylation site (N-Glyc) and heparin-binding domain. The crystal structure of the APL-1 E2 domain is similar to the human APP E2 domain (Hoopes et al. 2010). The cytoplasmic domain (indicated by C) shows 71% similarity between C. elegans APL-1 and human APP695, has a conserved YENPTY internalization signal (yellow), a consensus sequence for Go binding, and four of the seven conserved phosphorylation sites of APP. (B) Western blot of mixed stage wild-type (WT) animals probed with an antibody against the extracellular domain of APL-1 (αAPL-1EXT) show proteins corresponding to glycosylated and unmodified full-length APL-1 (double bars) and one protein corresponding to the cleaved extracellular APL-1, sAPL-1. apl-1(yn5) mutants show high levels of APL-1EXT, which is slightly larger than sAPL-1. (C) Expression pattern of translational fusion of APL-1 protein with GFP (ynIs79 [APL-1::GFP]) of first larval stage C. elegans. APL-1 is expressed in head neurons (lower left), arcade cells (a), and ventral cord neurons (arrowheads). Bar, 20 μm. (D) Pharyngeal pumping rates of heterozygous apl-1(yn10) null mutants, homozygous apl-1(yn5) mutants, or transgenic animals overexpressing full-length APL-1 (apl-1(yn10) [APL-1]) or the extracellular domain of APL-1 (ynIs71 [APL-1EXT]) were similar to the pharyngeal pumping rate of wild-type animals during all four larval stages (L1–L4; N > 30 for each larval stage). Because complete loss of apl-1 leads to a molting defect during the first to second larval stage transition, the pharyngeal pumping rate for homozygous apl-1(yn10) mutants could only be scored for the first larval stage (L1).

The only viable apl-1 mutant isolated thus far, apl-1(yn5), contains a deletion mutation that removes the coding region for the transmembrane and cytoplasmic portions of the APL-1 protein and a large portion of the 3′-UTR and produces only the extracellular domain of APL-1, again demonstrating that the APL-1 extracellular domain is sufficient for viability (Figure 1) (Hornsten et al. 2007). apl-1(yn5) mutants show several phenotypes, including a slowed development compared to wild-type animals (Hornsten et al. 2007). In this study, we investigate the function of the extracellular domain of APL-1 during development. In addition to the slowed development, apl-1(yn5) mutants have several other phenotypes, including a temperature-sensitive embryonic lethality. We determined that several of the apl-1(yn5) phenotypes can be suppressed by daf-16 FOXO and daf-12 NHR mutations. Furthermore, we performed a small scale modifier screen to isolate enhancers and suppressors of the temperature-sensitive lethality of apl-1(yn5) mutants.

Materials and Methods

Strains

C. elegans strains were grown and maintained on MYOB plates (Church et al. 1995) containing OP50 Escherichia coli bacteria at 20° using methods as described (Brenner 1974), unless noted. All mutations used are described in WormBase (www.wormbase.org) and include: LGI: daf-16(mu86) (Lin et al. 1997); LGII: moa-2/B0495.6(yn39); LGIII: moa-1/R155.2(yn38), daf-2(e1370) (Kimura et al. 1997); LGIV: flp-1(ok2781); and LGX: daf-12(m20) (Larsen et al. 1995), apl-1(yn5 and yn10) (Hornsten et al. 2007). Construction of transgenes to rescue modifiers of the temperature-sensitive apl-1(yn5) lethality: phusion high-fidelity polymerase (Finnzymes) was used to amplify the genomic region, including the promoter and 3′-UTR, corresponding to genes of interest from wild-type (N2 var. Bristol) genomic DNA. The following primer pairs (5′ to 3′) were used: tag-235, Ptag235F (ggaacgagtgatgtgaggcag)/3tag235R (cggtgcctgttggagattcg); dnj-24, Pdnj24F3 (gccaaactctcggccaactc)/dnj24R (tacgtgcctcatggctctcc); moa-1/R155.2, P-R155-F1 (gctctggaaccggcttatgg)/3-R155-R1 (cgtaggccgcttccaaacaac); and moa-2/B0495.6, 3-operon-KpnI (agagggtaccgaaaggacgtgcgggaaagc; restriction site underlined)/5-operon2 (gcggcaagggaatagtcagag) and 5-B0495-KpnI (agagggtaccggagtacggtggacaagtacg; restriction site underlined)/B0495-3o (gagcattccacggttgtcgtc). The moa-2/B0495.6 products were digested with KpnI and ligated together. Amplified fragments were gel purified and inserted into a TOPO Blunt vector (Invitrogen), which was used for microinjection into yn38; apl-1(yn5) or yn39; apl-1(yn5) animals. Extrachromosomal transgenic lines generated were (transgenes are indicated in brackets; unless otherwise noted, the promoter used was that of the indicated gene and the transgene contained a wild-type copy of the gene): ynEx201 [tag-235; sur-5::GFP], ynEx202 [tag-235; sur-5::GFP], ynEx207 [dnj-24; tag-235; sur-5::GFP], ynEx208 [dnj-24; sur-5::GFP], ynEx203 [R155.2; sur-5::GFP], ynEx204 [R155.2; sur-5::GFP], ynEx205 [R155.2; sur-5::GFP], ynEx206 [R155.2; sur-5::GFP], ynEx209 [B0495.6; sur-5::GFP], ynEx210 [B0495.6; sur-5::GFP], and ynEx211 [B0495.6; sur-5::GFP]. Construction of the APL-1 transgenes and the resulting transgenic lines are described (Hornsten et al. 2007). Integrated apl-1 transgenic lines used were (transgenes are indicated in brackets; unless otherwise noted, the promoter used was the apl-1 promoter): ynIs106 [apl-1(yn32 yn5), Pmyo-2::GFP]; LGIV: zIs356 [daf-16::GFP, pRF4 rol-6(su1006gf)] (Henderson and Johnson 2001); LGV: ynIs71 [apl-1(yn5), sur-5::GFP], ynIs79 [apl-1::GFP], ynIs100 [apl-1(yn32)::GFP, pRF4 rol-6(su1006gf)]; and LGX: ynIs86 [apl-1, sur-5::GFP], and ynIs107 [apl-1(yn32/D342C/S362C)::GFP, Pmyo-2::GFP] (Hoopes et al. 2010). For simplicity in the text, we will only indicate the protein expressed by the transgenes, including GFP-tagged proteins, in brackets; unless otherwise indicated, the apl-1 promoter was used to express the transgene.

Western blot analysis

Preparations of animal lysates and Western blots were performed as described; protein levels were normalized to levels of actin (Hornsten et al. 2007). Roughly the same extract amount for each strain was electrophoresed and transferred. After transferring, the blot was cut in two: one blot was probed with an antiserum against the extracellular domain of APL-1 (1:2000) (Hornsten et al. 2007) and one blot was probed with an actin monoclonal antibody (JLA20 at 1:500; Developmental Studies Hybridoma Bank); secondary antibodies were used at 1:4000 to 1:2000. Relative protein levels were determined by relative intensity to wild type (N2) using National Institutes of Health ImageJ Gel analyzer.

Pharyngeal pumping rate assays

To synchronize worm populations ∼15 gravid adult worms were placed into a bleach solution to release the eggs. Hatched worms were raised at 20°. Developmental stage was measured by gonadal development. Pharyngeal pumping rate was measure by visually counting the movement of the pharyngeal grinder on a stereomicroscope for a period of 20 sec. Only worms on the bacterial food source and with constant pharyngeal pumping were scored.

Developmental timing and egg-laying rate assays

To synchronize worm populations ∼15 gravid adult worms were placed into a bleach solution to release the eggs. Hatched worms were raised at 20°. Four days later [or 5 days for slower developing strains such as apl-1(yn5) and zIs356(DAF-16::GFP)], 10 synchronized adults were placed onto a fresh plate and allowed to lay eggs for 1–1.5 hr at room temperature (22–24°). Eggs were counted to determine an egg-laying rate; F1 progeny were placed at 20° to allow development. After 70–72 hr, the developmental stages of the animals at 20° were scored according to their gonadal development and body size. For development at 25°, animals were scored after 48 hr. Each individual trial was performed with at least three plates of synchronized eggs for each strain and always included wild-type animals as a control. For statistical analysis for the egg-laying rate, one-way ANOVAs with Tukey’s post-test (95% confidence intervals) were performed to assess similarity between groups and for developmental timing, a χ2 test was performed using Prism 4.0a software (GraphPad).

Body length measurements

For each individual trial, 30 L4 animals (10 L4 per plate) for each strain were picked and allowed to develop for 3 days at room temperature (22–24°). Animals were mounted onto 2% agar pads containing a drop of 10 mM NaN3 and pictures of the animals were taken at ×100 magnification on a confocal microscope (Zeiss LSM 510 confocal laser scanning system). The lengths of the worms were determined by drawing a line along the midline of the animals from the tip of the mouth to the tail. For statistical analysis one-way ANOVAs with Tukey’s post-test (95% confidence intervals) were performed to assess similarity between groups using Prism 4.0a software (GraphPad).

Critical period assays

Ten synchronized gravid adults were placed onto a fresh plate and allowed to lay eggs for 0.5, 1, or a maximum of 1.5 hr at 15°. Eggs were counted and adult Po were killed; F1 progeny were either shifted to 27° or placed back at 15° to allow development, so that eggs in intervals of 30 min up to 6 hr were shifted to 27°. After 44 hr at 27°, the developmental stages and number of surviving animals were scored. Each individual trial was performed with at least three plates of synchronized eggs for each strain and always included wild type as a control. For statistical analysis one-way ANOVAs with Tukey’s post-test (95% confidence intervals) were performed to assess similarity between groups using Prism 4.0a software (GraphPad).

RNA interference assays

RNAi by feeding: day −1, a single RNAi clone [HT115 bacteria, which are maintained at −80° on Luria broth medium (LB) agar plates with 25 μg/ml carbenicillin and 12.5 μg/ml tetracycline] was picked from the Ahringer library (Kamath et al. 2001) (Geneservice) and incubated in 1 ml LB containing 100 μg/ml ampicillin (at 37°, 280 rpm) overnight; day 0, the 1-ml bacterial culture was transferred into 10 ml LB containing 100 μg/ml ampicillin and incubated for another 4–6 hr at 37° at 280 rpm. A total of 450 μl of the bacteria culture was spread onto MYOB plates (Church et al. 1995) containing 400 mM of βD-isothiogalactopyranoside (IPTG) and 50 μg/ml ampicillin, and these RNAi plates were placed in 37° overnight; day 1, eight L4 animals (Po) were placed on bacteria lawn that express double-stranded RNA (dsRNA) of target gene or empty vector control (L4440) to knock down expression levels of the targeted gene; day 4, ∼50 F1 L4 animals were transferred onto new plates containing the same dsRNA-expressing bacteria; day 5, ∼10 F1 1-day-old adults were transferred onto new plates containing the same dsRNA-expressing bacteria to lay eggs for 1–1.5 hr at 20°; F2 eggs were placed at either 20° or 27°; and day 8, the F2 population was scored for developmental progression and survival.

DAF-16::GFP nuclear translocation assays

L4 animals were placed onto new plates and grown at 20°. One day later, animals were placed at 35° and scored at different times (T = 0, 30, 60, 90, 120, and 150 min) by mounting the animals onto 2% agar pads containing a drop of M9 physiological buffer (Brenner 1974) and looking at the worms at ×100 magnification on a Zeiss Axioplan microscope. To ensure exact timing, the individual strains were placed at 35° in 5-min intervals. Upon 35° heat stress, DAF-16::GFP translocates from the cytoplasm into the nucleus (Henderson and Johnson 2001). The rate of DAF-16::GFP nuclear translocation was scored as described (Curran and Ruvkun 2007): category 0: all DAF-16::GFP showing diffuse localization in the cytoplasm; category 1: more DAF-16::GFP localized in cytoplasm than in nucleus; category 2: more DAF-16::GFP localized in nucleus than in cytoplasm; category 3: almost all DAF-16::GFP localized in nucleus. Statistical significance was determined by a χ2 test using Prism 4.0a software (GraphPad).

Mutagenesis screen and mapping

Worms were mutagenized with 50 mM EMS as described (Brenner 1974). Mutagenized L4 animals (Po) were singly plated and placed at 20° to develop until F1 animals reached adulthood. Ten F1 adults were allowed to lay F2 eggs, which were then shifted to 27°. Plates on which the number of F2 progeny was greater or smaller than the number of progeny from nonmutagenized apl-1(yn5) mutants were selected for further analysis. A total of 200 haploid genomes were screened and five mutants were isolated. The strongest suppressor (yn38) and enhancer (yn39) of the apl-1(yn5) lethality at 27° were selected for further characterization. yn38 was mapped by conventional methods (Brenner 1974) to chromosome III and yn39 was mapped using SNPs to chromosome II as described (Davis et al. 2005). The DNA from both mutants was isolated and used for whole genome deep sequencing as described (Sarin et al. 2010).

Results

Expression of the extracellular domain of APL-1 is sufficient to slow developmental progression

Wild-type C. elegans have a very stereotyped pattern of cell divisions. Synchronized eggs develop into fourth stage (L4) larva within 65 hr and into adults by 72 hr at 20° (Ailion and Thomas 2000). By contrast, apl-1(yn5) mutants were found mostly in L4 or earlier larval stages at 72 hr (Figure 2; Table 1) (Hornsten et al. 2007). As predicted from the mutation (Figure 1), apl-1(yn5) mutants contained high levels of only the extracellular fragment of APL-1 (APL-1EXT), which is slightly larger than the cleaved sAPL-1 produced in wild-type animals (Figure 1) (Hornsten et al. 2007). The slowed developmental progression of apl-1(yn5) mutants can be phenocopied in wild-type animals by microinjection of an APL-1EXT fragment (ynIs71) (Table 1). As a control, we generated an APL-1EXT transgene that contained a mutation corresponding to the apl-1(yn32) null mutation [APL-1EXT(yn32)]; transgenic animals carrying this transgene (ynIs106) did not rescue apl-1 knockouts and these transgenic animals developed at the same rate as wild-type animals (Table 1). These results indicate that overexpression of APL-1EXT and presumably sAPL-1 is sufficient to slow developmental progression.

Figure 2 

apl-1(yn5) mutants show a slowed developmental progression that is enhanced by decreased daf-2 activity. After 48 hr at 25°, all eggs from wild-type animals developed into L4 or adult animals. By contrast, apl-1(yn5) mutants developed mostly into L3 or L4 animals and daf-2(e1370) mutants mainly transitioned into an alternative dauer stage. Most daf-2(e1370); apl-1(yn5) double mutants, however, arrest in the L1 stage at 25°. N = number of animals, T = number of independent trials. Table 1 shows detailed statistical analysis and results of additional strains.

View this table:
Table 1  The delayed development of apl-1(yn5) mutants requires daf-12 and daf-16 activity

Although the apl-1(yn5) mutation appears to delay developmental progression by one larval stage (Table 1), in actuality the delay is more significant. After fertilization, wild-type eggs develop and are laid when they reach about the 30-cell stage of division (Sulston et al. 1983). apl-1(yn5) mutants retain eggs in the uterus longer than wild-type animals; eggs laid by apl-1(yn5) mutants, therefore, are chronologically older than eggs laid by wild-type animals. For instance, at 15° most wild-type eggs were at the 30- to 100-cell stage after 30 min or at the comma stage after 6 hr after being laid. By contrast, most apl-1(yn5) eggs were already in the comma stage after 30 min or the threefold (pretzel) stage 6 hr after being laid at 15°; the time to progress from the comma stage to the pretzel stage is about half the time to progress from the 30- to 100-cell stage to the comma stage (see Figure 4 below). Hence, the apl-1(yn5) mutation causes a severe delay in developmental progression.

One possible explanation for the slowed development of apl-1(yn5) animals could be a lower feeding rate of the animals, thereby leading to a slower metabolic rate and developmental delay. We tested whether modulating apl-1 levels affected feeding by measuring pharyngeal pumping rates. Heterozygous apl-1(yn10) animals, which carry an apl-1(yn10) null allele, homozygous apl-1(yn5) mutants, or transgenic animals carrying the APL-1EXT transgene showed pumping rates similar to wild-type animals during all four larval stages (L1–L4; Figure 1D). Because homozygous apl-1(yn10) mutants die during the L1-to-L2 transition (Hornsten et al. 2007; Wiese et al. 2010), we were only able to measure the pumping rates of homozygous apl-1(yn10) animals during L1; the homozygous apl-1(yn10) L1 animals also had pumping rates similar to wild type (Figure 1D). These results suggest that neither overexpression of APL-1EXT nor decreased apl-1 activity have any affect on feeding rates during development. These results are in contrast to the increased pharyngeal pumping rates when apl-1 was knocked down by RNAi through microinjection (Zambrano et al. 2002).

apl-1(yn5) enhances daf-2-induced L1 arrest but not dauer formation

Different environmental conditions result in the activation of multiple parallel pathways that determine the animal’s developmental progression and adjust the animal’s metabolic processes accordingly (reviewed in Fielenbach and Antebi 2008). apl-1(yn5) mutants could show a slowed development because of altered metabolic processes. We examined the effects of altering the insulin signaling pathway, a key metabolic pathway in C. elegans as well as mammals, in apl-1(yn5) mutants.

Activation of the insulin pathway is necessary for reproductive growth (Gems et al. 1998), whereas unfavorable environmental conditions lead to decreased insulin signaling in C. elegans (Henderson and Johnson 2001). Even in the presence of food, strongly reducing daf-2 insulin/IGF-1 receptor activity induces L1 arrest (Gems et al. 1998; Baugh and Sternberg 2006), while slightly reducing daf-2 activity induces dauer formation (Kimura et al. 1997). The L1 arrest due to decreased daf-2 activity requires the activity of daf-16 FOXO (Table 1; Baugh and Sternberg 2006). Animals with a weak temperature-sensitive daf-2(e1370) mutation have a slowed progression through all larval stages at the 20° permissive temperature (Table 1); in addition, ∼3% of the animals (23 dauers/748 total) enter the dauer stage compared to none of the wild-type animals (0 dauers/3572 total) or apl-1(yn5) mutants (0 dauers/943 total). At the nonpermissive temperature of 25°, all wild-type animals developed into L4 or adult animals after 48 hr (Figure 1) (Kimura et al. 1997; Gems et al. 1998), whereas 97% (528/545 total) of daf-2(e1370) mutants enter the dauer stage and 3% enter L1 arrest (Figure 2; Table 1) (Kimura et al. 1997; Gems et al. 1998).

To determine whether apl-1 signaling modulates the insulin signaling pathway, we examined daf-2(e1370); apl-1(yn5) double mutants. daf-2(e1370); apl-1(yn5) double mutants showed an even slower developmental progression at 20° than daf-2(e1370) or apl-1(yn5) single mutants, but the rate of dauer formation was not increased (Table 1). Furthermore, the percentage of daf-2(e1370); apl-1(yn5) double mutants entering L1 arrest at 25° was greatly enhanced: 92% of daf-2(e1370); apl-1(yn5) double mutants entered L1 arrest, while the remaining 8% became dauer animals (Figure 2; Table 1). Thus, apl-1(yn5) activity either reduces the activity of the insulin/IGF-1 signaling pathway or acts in parallel to the insulin pathway to enhance L1 arrest and delay development.

The slowed development of apl-1(yn5) mutants requires the activity of daf-16 and daf-12

Signaling through the daf-2 insulin/IGF-1 receptor decreases daf-16 activity. daf-16(mu86) mutants showed a similar developmental progression pattern as wild-type animals, whereas overexpression of daf-16 with a functional translational fusion of DAF-16 with green fluorescent protein (DAF-16::GFP) slows developmental progression, such that most animals are in the L4 stage after 72 hr (Table 1) (Henderson and Johnson 2001). To determine whether the development of apl-1(yn5) mutants is affected by daf-16 activity, we made daf-16(mu86); apl-1(yn5) double mutants. daf-16(mu86); apl-1(yn5) double mutants showed a similar developmental progression as daf-16(mu86) or wild-type animals (Table 1), indicating that the slowed development of apl-1(yn5) mutants requires daf-16 activity. Moreover, the increased rate of L1 arrest in daf-2(e1370); apl-1(yn5) double mutants was suppressed by loss of daf-16 activity (Table 1). At 25°, 73% of the daf-2(e1370); daf-16(mu86); apl-1(yn5) triple mutants developed into L4 (73% in L4; 8% in L3, 2% in L2, and 17% in L1; Table 1). By contrast, the apl-1(yn5) mutation was additive to the effects of DAF-16 overexpression: after 72 hr DAF-16::GFP; apl-1(yn5) animals are mostly found in the L2–L3 stage. These results suggest that APL-1EXT, and presumably sAPL-1, signal to modulate the insulin pathway, thereby increasing daf-16 activity to affect developmental progression.

The insulin pathway converges with a parallel pathway that signals through the DAF-12 nuclear hormone receptor (NHR) during the developmental decision to enter reproductive growth or dauer formation. While L1 arrest requires daf-16 activity (Baugh and Sternberg 2006), dauer formation (Lin et al. 1997; Ogg et al. 1997) is regulated by DAF-12 NHR as well as DAF-16 FOXO activity (Henderson and Johnson 2001; Lee et al. 2001; Lin et al. 2001). daf-12 NHR and daf-16 FOXO are expressed ubiquitously (Antebi et al. 2000; Henderson and Johnson 2001; Lee et al. 2001; Lin et al. 2001) and daf-12(m20) and daf-16(mu86) null mutants are dauer defective (Vowels and Thomas 1992; Larsen et al. 1995; Antebi et al. 2000). In addition, daf-12 NHR also plays an important role in developmental timing by forming a feedback loop with the let-7 miRNA family of heterochronic genes (Hammell et al. 2009) and by acting in a complex genetic network with the lin-42 period gene (Monsalve et al. 2011).

To determine whether daf-12 activity is necessary for the slowed development in apl-1(yn5) mutants, we knocked down daf-12 activity in apl-1(yn5) mutants by RNAi. The slowed development of apl-1(yn5) mutants was suppressed by daf-12 NHR knockdown, whereas daf-12 RNAi had no effect on developmental progression of wild-type or daf-12(m20) mutant animals (Table 1). Similarly, the slowed development of ynIs71 [APL-1EXT] overexpression animals was rescued in a daf-12(m20) mutant background (Table 1), excluding an RNAi off-target effect of daf-12 RNAi. Signaling of the extracellular domain of APL-1, therefore, requires DAF-12 NHR activity to delay development.

apl-1(yn5) slows DAF-16 nuclear localization under heat-shock conditions

Because APL-1EXT is a released fragment, APL-1EXT signaling will influence DAF-16 activity indirectly. Since C. elegans is transparent, DAF-16 localization can be monitored by using DAF-16::GFP. Under well-fed, noncrowded, and unstressed laboratory conditions, DAF-16::GFP is predominantly found diffused in the cytoplasm of all cells in wild-type (Henderson and Johnson 2001) and apl-1(yn5) animals (Supporting Information, Table S1). Translocation of DAF-16::GFP from the cytoplasm to the nucleus can be visualized in intestinal cells by putting animals under a heat stress (Henderson and Johnson 2001). When DAF-16::GFP animals were shifted from 20° to 35°, DAF-16::GFP translocated into the nucleus within 3 hr (Table S1) (Henderson and Johnson 2001). DAF-16::GFP; apl-1(yn5) animals showed a delayed DAF-16::GFP nuclear translocation compared to DAF-16::GFP animals at 35° (Table S1). As a control, nonfunctional APL-1EXT [APL-1EXT(yn32)] did not alter the timing of DAF-16::GFP nuclear translocation (DAF-16::GFP; ynIs106; Table S1). These results indicate that APL-1EXT activity slows intestinal DAF-16 nuclear translocation in response to stress. Thus, apl-1 acts in multiple pathways to affect DAF-16 FOXO activity.

Several apl-1(yn5)–induced phenotypes require daf-16 FOXO and daf-12 NHR activity

Our results suggest that apl-1(yn5) acts in multiple pathways that converge on daf-16 FOXO and daf-12 NHR. Alterations in the TGFβ signaling pathway can lead to DAF-16 nuclear localization (Lee et al. 2001; Shaw et al. 2007; Jeong et al. 2010) and there is an extensive cross-talk between the insulin/IGF-1 and TGFβ pathways for multiple processes (Narasimhan et al. 2011). Consequently, we examined whether daf-16 FOXO and daf-12 NHR activity mediate other apl-1(yn5) phenotypes. Body size in C. elegans is regulated by genetic and environmental factors through the insulin (So et al. 2011) and TGFβ pathways (Savage-Dunn et al. 2003). Wild-type adult animals are 1225 ± 6.6 μm (n = 172) in length. apl-1(yn5) mutants and transgenic animals carrying an APL-1EXT transgene (ynIs71) were 15% (1047 ± 11.6 μm, n = 63) or 27% (894 ± 28.2 μm, n = 33) shorter, respectively, than wild-type animals (Figure 2A; Table S2), indicating that the shortened body length is due to high levels of APL-1EXT and not due to loss of the APL-1 intracellular domain (C. elegans AICD). Similarly, animals that overexpress full-length APL-1 (ynIs86 and ynIs79) were 12–20% shorter than wild-type animals, whereas animals that carry a transgene with a mutated apl-1 (ynIs100) were wild type in length (Figure 3A; Table S2). Thus, apl-1 activity, and specifically the activity of sAPL-1, affects body length.

Figure 3 

Decreased DAF-16 FOXO and DAF-12 NHR activity are required for APL-1 signaling to modulate body size and egg-laying rate. (A) Overexpression of APL-1 or APL-1EXT shortened the body length of the animal (white bars) compared to wild type (black bar). This shortened body length of apl-1(yn5) mutants and ynIs79 [APL-1::GFP] transgenic animals was abolished in a daf-16 null background and ynIs71 [APL-1EXT] transgenic animals in a daf-12(m20) mutant background. Decreased daf-2 activity enhances the shortened body length of apl-1(yn5) mutants. (B) Overexpression of APL-1 or APL-1EXT caused a reduction in the egg-laying rate (white bars); an egg-laying rate higher than wild type was restored in a daf-16 null background or in a daf-12(m20) mutant background. Table S2 and Table S3 show detailed statistical analyses. Protein expressed by transgene in the transgenic lines (the ynIs strains) is indicated above the corresponding bars; daf-16 null background is indicated in red, daf-12(m20) mutant background in yellow, and daf-2(e1370) mutant background in orange; Is, integrated transgene. ***P < 0.001 determined by one-way ANOVA with Tukey’s post-test.

daf-2(e1370) mutants are also slightly shorter than wild type (Figure 3A). The shortened body length of apl-1(yn5) mutants was enhanced when daf-2 activity was decreased (Figure 3A), suggesting that, as with developmental progression, apl-1(yn5) activity either reduces the activity of the insulin/IGF-1 signaling pathway or acts in parallel to the insulin pathway to affect body size. daf-16(mu86) mutants were slightly (3%) longer and DAF-16::GFP animals were slightly shorter than wild type, although both not significantly; similarly, daf-12(m20) mutants were similar in length to wild-type animals (Figure 3A; Table S2). In a daf-16(mu86) background, apl-1(yn5) mutants and transgenic APL-1 overexpression lines were the same length as daf-16(mu86) mutants (Figure 3A; Table S2), suggesting that the shorter body length of apl-1(yn5) mutants requires daf-16 activity. Moreover, the shortened body size of the daf-2(e1370); apl-1(yn5) double mutants was rescued to wild-type length in a daf-16(mu86) background (Figure 3A). Similarly, in a daf-12(m20) background, transgenic APL-1 overexpression lines were the same length as daf-12(m20) mutants (Figure 3A; Table S2). Furthermore, RNAi knockdown of daf-12 in apl-1(yn5) mutants was sufficient to rescue the apl-1(yn5) shortened body length (Table S2). Hence, both daf-16 FOXO and daf-12 NHR activity are required for the shortened body length of apl-1(yn5) animals. By contrast, DAF-16 overexpression enhanced the shortened body length of transgenic APL-1 overexpression lines (Table S2). Collectively, these results suggest that apl-1 activity modulates the insulin pathway to increase daf-16 activity to affect body size.

APL-1 is expressed in vulval hypodermal cells and vulval muscles cells, which regulate egg laying (Hornsten et al. 2007). apl-1(yn5) mutants, as well as transgenic APL-1 or APL-1EXT overexpression lines (data not shown), retain eggs and show a decreased egg-laying rate (Hornsten et al. 2007), suggesting that this phenotype is due to high levels of APL-1EXT and not due to loss of signaling through the C. elegans AICD fragment. Wild-type animals lay about seven eggs per hour, whereas apl-1(yn5) mutants lay about four eggs per hour (Figure 3B; Table S3) (Hornsten et al. 2007). Transgenic animals overexpressing either full-length APL-1 (ynIs86 and ynIs79) or APL-1EXT (ynIs71) laid about five eggs an hour (Figure 3B; Table S3) (Hornsten et al. 2007). daf-16(mu86) mutants laid significantly more eggs, about nine eggs per hour, than wild-type animals (Figure 3B; Table S3). Transgenic APL-1 overexpression lines and apl-1(yn5) mutants carrying the daf-16(mu86) mutation laid eggs at the same rate as daf-16(mu86) mutants (Figure 3B; Table S3), suggesting that the egg-laying defect requires daf-16 activity. Conversely, animals that overexpress DAF-16 showed a dramatic decrease in their egg-laying rate to about one to two eggs per hour (Table S3). Overexpression of full-length APL-1 (ynIs79) or the apl-1(yn5) mutation had no effect on the decreased egg-laying rate of DAF-16 overexpression animals (Table S3). daf-12(m20) mutants showed a similar egg-laying rate as wild-type animals (Figure 3B; Table S3). The decreased egg-laying rate of APL-1EXT overexpression animals (ynIs71) was completely rescued in a daf-12(m20) mutant background (Figure 3B; Table S3). These results suggest that APL-1EXT requires daf-16 FOXO and daf-12 NHR activity to decrease body length and egg-laying rate.

apl-1(yn5) mutants show a temperature-sensitive lethality and developmental arrest

As discussed above, at 20° and 25° wild-type animals hatch and develop into adults. At 27°, however, although all wild-type eggs hatch and animals survive (Table 2) (Ailion and Thomas 2000), ∼10% of developing animals enter the dauer life cycle (Ailion and Thomas 2000). At 20°, 86% of the apl-1(yn5) mutants survived and 14% remained either arrested in L1 or died (Table 2). This lethality was enhanced at slightly higher temperatures: 75 and 47% of the apl-1(yn5) mutants survived at 25° and 27°, respectively (Table 2). In addition, among the apl-1(yn5) mutants that survived, only a few developed into gravid adults and 46% remained in L1 arrest compared to only 1% of wild-type animals after 44 hr at 27° (Table 2); most of these apl-1(yn5) L1-arrested animals died within 5 days and showed morphological defects such that organs appeared detached from their underlying basal lamina and animals contained multiple vacuolar-like structures (data not shown). Thus, the apl-1(yn5) mutation causes a temperature-sensitive lethality and developmental progression block. To determine whether this lethality could be phenocopied, we examined transgenic APL-1EXT animals (ynIs71). A total of 82% of ynIs71 [APL-1EXT] animals survived at 20°, 64% at 25°, and 54% at 27° (Table 2). Animals carrying the mutated APL-1EXT transgene [APL-1EXT(yn32)] (ynIs106) survived at similar rates to wild-type animals at 20° and 27° (Table 2). Hence, high levels of APL-1EXT, and presumably sAPL-1 activity is sufficient to cause a temperature-sensitive lethality and L1 arrest.

View this table:
Table 2  APL-1 overexpression causes a temperature-sensitive lethality

The critical period for the temperature-sensitive apl-1(yn5) lethality is during embryogenesis

Because apl-1(yn5) mutants characteristically die after L1 arrest at 27°, we determined the critical time period of this temperature-sensitive lethality. Wild-type animals and apl-1(yn5) mutants were allowed to lay eggs at 15°. Eggs were shifted to 27° at 30-min intervals and scored for survival 44 hr later. All wild-type eggs hatched and all animals survived from the different times (Figure 4). By contrast, the fraction of apl-1(yn5) eggs that hatched and survived increased linearly. At 30 min, 30% of apl-1(yn5) mutants survived, whereas shifting the eggs to 27° at 6 hr resulted in ∼100% survival (Figure 4), indicating a critical time window of the apl-1(yn5)-induced lethality during embryogenesis. Specifically, apl-1(yn5) eggs that developed past the threefold (pretzel) embryonic stage at 15° survived the 27° shift (Figure 4), suggesting that the critical time period for APL-1 overexpression lethality is before the pretzel stage of embryonic development. These results would predict that all L1 APL-1 overexpression animals shifted to 27° should survive. Indeed, all apl-1(yn5) L1 mutants shifted to 27° survived (N = 562, T = 5), again restricting the APL-1–induced lethality to embryogenesis.

Figure 4 

The critical time period for the apl-1(yn5) lethality is during embryogenesis. Eggs were laid at 15° and shifted at intervals of 30 min to 27°; plotted is the survival rate of those eggs after 44 hr at 27°. All wild-type eggs survive the 27° shift, whereas after shifting at 30 min, only 30% of the apl-1(yn5) eggs survived into larval stages. However, by letting the apl-1(yn5) eggs develop at 15° for 6 hr, all eggs survived and developed into larval animals. The developmental stage of eggs (30–100 cell, comma, 1.5- to 2-fold embryo, 3-fold embryo, pretzel) was scored at various time points. apl-1(yn5) eggs were chronologically further developed when laid compared to wild type, because apl-1(yn5) animals retained eggs longer in the uterus than wild-type animals. At least three trials for each time point (N > 300).

Knockdown of apl-1 by RNAi on apl-1(yn5) mutants causes a molting defect and lethality

Since apl-1(yn5) mutants have high levels of APL-1EXT (Figure 1) (Hornsten et al. 2007), we hypothesized that apl-1 knockdown by RNAi could rescue the apl-1(yn5)-induced lethality. Surprisingly, feeding double stranded apl-1 RNA to L4 apl-1(yn5) mutants resulted in dead L1–L4 animals in the next generation (F1). These F1 animals showed severe molting defects, similar to those seen in apl-1(yn10) mutants, which show 100% lethality due to a molting defect during the first to second larval (L1/L2) stage transition. However, the RNAi apl-1–induced molting defect of apl-1(yn5) mutants occurred through all larval stages (N > 30 for each molting stage; 6 trials), suggesting that apl-1 is required for molting not only during the L1/L2 transition, but during all larval transitions. We speculate that feeding apl-1 RNAi to apl-1(yn5) mutants more efficiently knocks down apl-1, presumably because the yn5 mRNA is smaller than wild type. Consistent with our observation that apl-1 is needed for the molt in each larval transition, feeding RNAi of apl-1 to worms in an RNAi-sensitized background [rrf-3(pk1426)] resulted in molting defects during L3/L4 and L4/adult transitions (Wiese et al. 2010).

Identification of suppressors and enhancers of the apl-1(yn5) temperature-sensitive lethality

To identify genes in the pathway of apl-1, we performed a forward genetic screen for modifiers of the temperature-sensitive lethality of apl-1(yn5) mutants (Figure 5; Table 2). Mutagenized L4 animals were singly plated and 10 F1 adults were allowed to lay F2 eggs, which were then shifted to 27°. Plates on which the number of F2 progeny was greater or smaller than the number of progeny from nonmutagenized apl-1(yn5) mutants were selected for further analysis. In a screen of 200 haploid genomes, we isolated one mutation, yn39, that enhanced and one mutation, yn38, that suppressed the lethality rate. The survival rate of those mutants was determined after several generations. For the enhancer strain, only 18% of its progeny survived at 27°, while the suppressor strain showed a 96% survival rate (Figure 5; Table 2). These modifying effects were not temperature dependent. At 20° apl-1(yn5) animals have a lethality rate of 14%, whereas yn39; apl-1(yn5) and yn38; apl-1(yn5) double mutants showed a lethality rate of 31 and 4%, respectively (Table 2). To determine whether the yn39 enhancement of lethality is dependent on apl-1(yn5), we outcrossed yn39 from the apl-1(yn5) background, allowed F2 animals to lay eggs, shifted F3 eggs to 27°, and scored for survival after 44 hr. All eggs developed into L4 animals, suggesting that the yn39 mutation by itself does not cause lethality but rather enhances the apl-1(yn5) lethality. As a control, from the same cross, 40 F2 apl-1(yn5) heterozygous animals were also picked at 20°. Of the F3 progeny, ∼25% were homozygous for the yn39 mutation (9/40 = 0–33% survival), while the rest of the animals showed a similar survival rate (40–60%) as apl-1(yn5) animals, except for one where 100% F3 progeny survived, presumably due to recombination.

Figure 5 

Suppressor and enhancer of the temperature-sensitive apl-1(yn5) lethality. All wild-type eggs survived and developed into larval animals, whereas only 47% of apl-1(yn5) eggs survived and developed at 27°. From the mutagenesis screen, 96% of the yn38; apl-1(yn5) (dark blue bar) and 18% of the yn39; apl-1(yn5) (light blue bar) double mutant animals survived. The yn38; apl-1(yn5) and yn39; apl-1(yn5) survival rates were rescued to apl-1(yn5) rates in multiple independent transgenic lines carrying genomic fragments of MOA-1/R155.2 (dark blue hatched bars) and MOA-2/B0495.6 (light blue hatched bars), respectively. ***P < 0.001 to apl-1(yn5) determined with one-way ANOVA post hoc Tukey.

To determine whether the mutations suppressed or enhanced other apl-1(yn5) phenotypes, we examined L1 arrest. Both strains were outcrossed four times before phenotypic characterization and looked superficially wild type. The yn38 mutation partially suppressed the L1 arrest, while the yn39 enhanced the L1 arrest at 27°: apl-1(yn5) animals have an L1 arrest rate of 46%, whereas yn39; apl-1(yn5) and yn38; apl-1(yn5) double mutants showed L1 arrest rates of 69 and 23%, respectively (Figure S2; Table 2).

Both yn38 and yn39 are recessive alleles. yn38 was mapped by conventional methods (Brenner 1974) to chromosome III and fine mapped using SNPs; yn39 was mapped using SNPs to chromosome II. The DNA from both mutants was isolated and used for whole genome deep sequencing. Because both alleles were sequenced at the same time, we compared the DNA of each allele to wild type and to each other. For yn38, the deep sequencing revealed 41 variations on chromosome III compared to wild type. After subtracting those variations also found in the yn39 sequence data, six candidate variations remained. One variant, R155.2, mapped to where fine SNP mapping predicted the mutation; however, no RNAi clone was available in the Ahringer library for R155.2, which encodes a receptor protein tyrosine phosphatase (RPTP). RNAi knockdown of the remaining candidates revealed two candidates that phenocopied the increased survival of yn38; apl-1(yn5) animals (Figure S3). One candidate, tag-235, encodes a protein involved in endocytosis, and the second candidate, dnj-24, encodes a protein containing a DNA-J domain. To determine which candidate corresponds to yn38, wild-type DNA of these three candidates was microinjected into yn38; apl-1(yn5) mutants and tested for rescue by scoring the level of F1 survival at 27°; only one candidate, R155.2 RPTP, showed rescue; we have named R155.2 moa-1 (modifier of apl-1). All four independent transgenic lines of R155.2 reduced the 96% survival rates of yn38; apl-1(yn5) animals to an average of 52%, similar to the 47% survival rate of apl-1(yn5) mutants at 27° (Figure 5; Table S4). The missense mutation of yn38 causes a threonine-to-isoleucine transition at amino acid 111 (T111I) in the extracellular domain of R155.2 RPTP, raising the possibility that the substitution disrupts ligand binding.

The deep sequencing of yn39 revealed 42 variations on chromosome II compared to wild type. After subtracting those variations that overlapped with those of yn38, 15 candidate variations remained. RNAi knockdown of only 1 candidate, B0495.6, phenocopied the ∼10% survival rate of yn39; apl-1(yn5) animals (Figure S3). Wild-type B0495.6 DNA was introduced into yn39; apl-1(yn5) and the three independent transgenic lines showed partial rescue by returning survival rates to an average of 32%, slightly lower than the 47% survival rate of apl-1(yn5) mutants (Figure 5; Table S4). B0495.6, which we have named moa-2, encodes a protein of 87 amino acids; the yn39 mutation is a deletion of 13 nucleotides, which leads to a frameshift after the 18th amino acid. B0495.6 has a splice factor 3B subunit domain (amino acids 6–26) and has been potentially implicated in receptor-mediated endocytosis (Balklava et al. 2007) and larval development (Kamath et al. 2003; Simmer et al. 2003; Sönnichsen et al. 2005).

To determine whether the yn38 and yn39 mutations affect the levels of APL-1EXT expression, we performed Western blots on extracts from moa-1(yn38); apl-1(yn5) and moa-2(yn39); apl-1(yn5) mutants raised at 20° and 27°. The levels of APL-1EXT increased slightly to apl-1(yn5) levels when the mutants were raised at 27° (Figure 6). Consequently, although the overall levels of APL-1EXT in moa-1(yn38); apl-1(yn5) and moa-2(yn39); apl-1(yn5) mutants are less than that of apl-1(yn5) mutants at 20°, the APL-1EXT levels are similar at 27°, suggesting that suppression and enhancement of the yn5 temperature-sensitive lethality by the yn38 and yn39 mutations, respectively, are not simply due to modulating the levels of APL-1EXT.

Figure 6 

Protein levels of APL-1 are similar to apl-1(yn5) levels at 27°, but slightly decreased at 20° in the moa-1; yn5 and moa-2; yn5 mutants. (A and B) Western blots of extracts were probed with either an antiserum against the extracellular domain of APL-1 (αAPL-1EXT) or actin (α-actin). Animals were raised at the indicated temperature. (A) Double bar indicates the full-length glycosylated or unmodified APL-1, the circle indicates APL-1EXT, which corresponds to the entire extracellular domain of APL-1, and the arrow indicates sAPL-1, a cleavage product of APL-1 that is slightly smaller than APL-1EXT. (B) Relative intensities of APL-1 levels normalized to actin and levels in apl-1(yn5) extracts: yn5 20° 1; yn5 27° 0.88 ± 0.2; moa-1(yn38); apl-1(yn5) 20° 0.49 ± 0.1; moa-1(yn38); apl-1(yn5) 27° 0.96 ± 0.2; moa-2(yn39); apl-1(yn5) 20° 0.69 ± 0.1; and moa-2(yn39); apl-1(yn5) 27° 0.86 ± 0.2 (n = 2 trials; mean ± SEM are indicated). Molecular weight standards are indicated to the left.

To analyze the genetic interaction between moa-1 and moa-2, we used RNAi knockdown on yn38 and yn39 mutants. RNAi of either dnj-24, tag-235, or C08G5.1, which has off-target effects on R155.2, on moa-2(yn39); apl-1(yn5) animals did not alter the yn39; apl-1(yn5) lethality rate at 27° (Figure S3). Interestingly, feeding yn38; apl-1(yn5) double mutants with moa-2/B0495.6 RNAi resulted in an F2 synthetic lethality at 20° (three trials, number of F1 animals >60), suggesting that moa-1 and moa-2 either function in the same pathway or in partially redundant pathways to affect development.

Discussion

Complete loss of apl-1 causes a completely penetrant lethal molting defect during the first to second larval stage transition (Hornsten et al. 2007). We now show that apl-1 knockdown by RNAi leads to a molting defect during all four larval stages (Wiese et al. 2010), indicating that apl-1 is required for every larval molt. In addition, apl-1(yn5) mutants have a delayed development, suggesting that APL-1, and in particular sAPL-1, is involved in both developmental timing and the molting process, two processes that are intimately linked, but whose regulation can be genetically separated in C. elegans (Ruaud and Bessereau 2006; Monsalve et al. 2011), similar to the genetic separation of metamorphosis/molting and developmental timing in Drosophila (Thummel 2001). Interestingly, the metamorphosis/molting and developmental growth rate in Drosophila is regulated by the integrative action of insulin signaling via dFOXO and nuclear hormone receptor (ecdysone receptor) signaling (Colombani et al. 2005; Delanoue and Leopold 2010). In C. elegans, daf-2 insulin/IGF-1 receptor and daf-16 FOXO activity and TGFβ signaling pathway regulate expression of different collagen genes, which are necessary for stage-specific cuticles crucial during development (Yu and Larsen 2001; McElwee et al. 2004; Halaschek-Wiener et al. 2005; Ruzanov et al. 2007; Shaw et al. 2007;), but the pathways through which daf-2, daf-16, and daf-12, as well as apl-1 act to affect the molting process is unclear. However, although the exact molecular pathway by which APL-1 activity affects DAF-2, DAF-16, and DAF-12 activity is unknown, our results demonstrate that apl-1 activity requires the insulin, daf-16 FOXO, and daf-12 NHR pathways for multiple processes, including developmental progression, body length, and egg laying. We propose that after APL-1 cleavage, sAPL-1 signals to decrease insulin signaling to modulate DAF-16 function, thereby affecting developmental progression and metabolic functions regulating body length and reproduction. daf-12 NHR integrates hormonal signaling with developmental timing by its position in the heterochronic feedback loop of let-7 miRNA, which regulates late developmental progression from L4 to adults (Bethke et al. 2009; Hammell et al. 2009) and apl-1 expression in seam cells (Niwa et al. 2008). RNAi knockdown of daf-12 did not alter apl-1 expression in seam cells during the late L4 stage (Hada et al. 2010) and possible miRNA binding sites in the 3′-UTR were deleted by the apl-1(yn5) mutation. Nevertheless, daf-12 knockdown completely suppressed the developmental delay in apl-1(yn5) mutants. By contrast, the daf-16(mu86) null mutation did not completely rescue the slowed development of apl-1(yn5) mutants, since a low percentage of apl-1(yn5); daf-16(mu86) animals were still found in L1–L4 stages. Hence, apl-1 activity may modulate the daf-2 insulin/IGF-1 receptor pathway to affect or act in parallel with the daf-16 and daf-12 pathways. Interestingly, this (these) signaling pathway(s) is (are) also observed to regulate different modalities, such as egg-laying behavior and body size. Similarly, transgenic APP mice show impairments in behavior, are lighter, and show reduced body weight gain compared to their wild-type littermates (Pugh et al. 2007; Codita et al. 2010). Weight loss is also associated with AD patients, despite the fact that AD patients consume more calories than age-matched non-AD controls (reviewed in Aziz et al. 2008), suggesting that AD patients may have altered metabolic rates (Wang et al. 2004). Our results suggest that metabolic rate changes could be mediated by secreted sAPP, which alters hormonal and insulin signaling pathways.

The temperature-sensitive lethality of apl-1(yn5) animals is not dependent on daf-12 NHR or daf-16 FOXO activity. We were somewhat surprised that screening such a small number of haploid genomes as well as RNAi clones identified during deep sequencing could identify modifiers of apl-1 activity. These results suggest that APL-1 is involved in multiple pathways and/or that the apl-1 pathway involves many genes. Our finding that decreased activity of MOA-1/R155.2 RPTP suppresses the apl-1(yn5) lethality suggests that either MOA-1/R155.2 RPTP is a receptor for sAPL-1 or that MOA-1/R155.2 RPTP is activated as a result of sAPL-1 signaling. apl-1(yn5) mutants contain high levels of APL-1EXT, which presumably increases downstream signaling through MOA-1/R155.2 RPTP. This downstream signaling could be further increased at 27°, leading to lethality; this situation may mimic overexpression of APL-1, which can also lead to lethality (Hornsten et al. 2007). The yn38 mutation in moa-1 could decrease the apl-1(yn5) lethality by decreasing receptor signaling. Many human RPTPs have similar tyrosine phosphatase domains as MOA-1/R155.2, but we found no similarities in the extracellular domain of MOA-1/R155.2 among human RPTPs with our BLAST searches. Alternatively, a second mechanism to decrease APL-1EXT signaling is to endocytose the bound or unbound receptor without affecting the levels of APL-1EXT; disruption of these endocytic pathways could either increase, such as with MOA-2/B0495.6, or decrease, such as tag-235, the apl-1(yn5) lethality, respectively. While the identification of apl-1(yn5) modifiers might correspond to a special situation with respect to the normal physiological function of APL-1, as the apl-1(yn5) mutants never express the cytoplasmic domain of APL-1, nevertheless, the extracellular domain of mammalian sAPPβ has previously been shown to act as a ligand for death receptor 6 (DR6) to initiate neurodegeneration (Nikolaev et al. 2009). Our results suggest that mammalian sAPP may also bind different receptors to differentially activate a cell death or neuronal survival pathway.

Acknowledgments

We thank Cathy Savage-Dunn’s group for kindly providing their N2 strain; Alexander Boyanov together with Oliver Hobert’s lab for whole genome sequencing of yn38 and yn39 mutants; Sarah Tichelli and Casey Brander for help with body length measurements; Mboutidem Etokakpan for help in the genetic screen; Piali Sengupta and Chip Ferguson for comments or helpful discussions on the manuscript; Li lab members for helpful discussions; and the Caenorhabditis Genetics Center, which is funded by the National Institutes of Health (NIH) National Center for Research Resources, for providing daf-16, DAF-16::GFP, and daf-12 strains. This work was supported by grants from the Alzheimer’s Association (IIRG-05-14190), NIH (R21AG033912 and R01AG032042), the National Science Foundation (IOS08207) (C.L.), and a NIH Research Centers in Minority Institutions grant (G12-RR03060) to the City College of New York.

Footnotes

  • Communicating editor: K. Kemphues

  • Received January 18, 2012.
  • Accepted March 24, 2012.

Literature Cited

View Abstract