ABSTRACT
The renin-angiotensin-aldosterone system (RAAS) plays a well-characterized role regulating blood pressure in mammals. Pharmacological and genetic manipulation of the RAAS has been shown to extend lifespan in Caenorhabditis elegans, Drosophila and rodents, but its mechanism is not well defined. Here, we investigate the angiotensin-converting enzyme (ACE) inhibitor drug captopril, which extends lifespan in worms and mice. To investigate the mechanism, we performed a forward genetic screen for captopril-hypersensitive mutants. We identified a missense mutation that causes a partial loss of function of the daf-2 receptor tyrosine kinase gene, a powerful regulator of aging. The homologous mutation in the human insulin receptor causes Donohue syndrome, establishing these mutant worms as an invertebrate model of this disease. Captopril functions in C. elegans by inhibiting ACN-1, the worm homolog of ACE. Reducing the activity of acn-1 via captopril or RNA interference promoted dauer larvae formation, suggesting that acn-1 is a daf gene. Captopril-mediated lifespan extension was abrogated by daf-16(lf) and daf-12(lf) mutations. Our results indicate that captopril and acn-1 influence lifespan by modulating dauer formation pathways. We speculate that this represents a conserved mechanism of lifespan control.
INTRODUCTION
Aging is characterized by progressive degeneration of tissue structure and function that leads inexorably to death. Determining the mechanisms of age-related degeneration is a major goal, because this information could promote the development of interventions to extend health span and lifespan. The renin-angiotensin-aldosterone system (RAAS) in mammals has been characterized extensively for its role in blood pressure regulation (Basso and Terragno, 2001; Igic, 2009). More recently, this pathway has been demonstrated to influence aging; Egan et al. (2022) reviewed emerging evidence that this pathway plays a conserved role in longevity control in worms, flies and rodents. However, the mechanism of longevity control has not been well defined in any animal.
Kumar et al. (2016) reported that captopril, an angiotensin-converting enzyme (ACE) inhibitor, extended Caenorhabditis elegans mean lifespan by ∼23% and maximum lifespan by ∼18%. The first of what is now a large class of ACE inhibitors, captopril is an oligopeptide derivative developed in 1975 based on a peptide found in pit viper venom (Ondetti et al., 1977; Cushman and Ondetti, 1991). ACE inhibitors modulate the RAAS, a mechanism by which the body adapts to hypotension (Skeggs et al., 1956; Peach, 1977; Reid et al., 1978; Basso and Terragno, 2001; Fyhrquist and Saijonmaa, 2008; Igic, 2009; Zehnder et al., 2009). In response to a decline in blood pressure, the kidney releases renin, which cleaves angiotensinogen to angiotensin I. ACE converts angiotensin I to angiotensin II, and angiotensin II binds two transmembrane receptors – the primary effects are stimulating aldosterone secretion and promoting vasoconstriction to increase blood pressure. By inhibiting ACE and preventing conversion of angiotensin I to angiotensin II, captopril lowers blood pressure (Fig. 1A).
Captopril causes dose-dependent effects in C. elegans. (A) Summary of the effects of captopril on the renin-angiotensin system. Captopril inhibits ACE (angiotensin-converting enzyme) in mammals and ACN-1 (ACE-like non-metallopeptidase) in C. elegans. ACE2, angiotensin-converting enzyme 2; AGT, angiotensinogen; AGT1R, angiotensin II type-I receptor; Ang 1-7, angiotensin 1-7; Ang I, angiotensin I; Ang II, angiotensin II; MasR, Mas receptor. (B) Comparison of captopril concentration in the agar medium used to culture worms (‘Media Captopril’) and concentration in whole-animal lysates measured by HPLC-MS (‘Internal captopril’). Numbers above bars are the ratio of internal captopril to media captopril, an indication of the ability of animals to exclude captopril. Bars represent average and s.d., n=4 biological replicates. (C,D) Effects of different concentrations of captopril on survival of wild-type worms. Wild-type animals were cultured on NGM dishes seeded with E. coli bacteria and the indicated concentration of captopril in the medium. (C) Values indicate the fraction of the starting population that remains alive. (D) Bars represent the average change in mean lifespan compared with 0 mM control; positive and negative values indicate lifespan extension and reduction, respectively (see Table S1 for statistics). ns, non-significant (P≥0.05); *P<0.05; **P<0.01; ***P<0.001.
Captopril causes dose-dependent effects in C. elegans. (A) Summary of the effects of captopril on the renin-angiotensin system. Captopril inhibits ACE (angiotensin-converting enzyme) in mammals and ACN-1 (ACE-like non-metallopeptidase) in C. elegans. ACE2, angiotensin-converting enzyme 2; AGT, angiotensinogen; AGT1R, angiotensin II type-I receptor; Ang 1-7, angiotensin 1-7; Ang I, angiotensin I; Ang II, angiotensin II; MasR, Mas receptor. (B) Comparison of captopril concentration in the agar medium used to culture worms (‘Media Captopril’) and concentration in whole-animal lysates measured by HPLC-MS (‘Internal captopril’). Numbers above bars are the ratio of internal captopril to media captopril, an indication of the ability of animals to exclude captopril. Bars represent average and s.d., n=4 biological replicates. (C,D) Effects of different concentrations of captopril on survival of wild-type worms. Wild-type animals were cultured on NGM dishes seeded with E. coli bacteria and the indicated concentration of captopril in the medium. (C) Values indicate the fraction of the starting population that remains alive. (D) Bars represent the average change in mean lifespan compared with 0 mM control; positive and negative values indicate lifespan extension and reduction, respectively (see Table S1 for statistics). ns, non-significant (P≥0.05); *P<0.05; **P<0.01; ***P<0.001.
In Drosophila, drugs that inhibit the ACE pathway and mutations in ACE pathway genes can extend lifespan (Ederer et al., 2018; Gabrawy et al., 2019, 2022). In normotensive mice and rats, the ACE inhibitor enalapril and the angiotensin II receptor antagonist losartan can extend lifespan and delay age-related degeneration of tissue structure and function in the kidney, cardiovascular system, liver and brain (Ferder et al., 1993, 1994, 2002; Inserra et al., 1995; González Bosc et al., 2000; de Cavanagh et al., 2003; Inserra, 2003; Carter et al., 2004; Basso et al., 2005, 2007; de Cavanagh et al., 2008; Inserra et al., 2009; Santos et al., 2009; Carter et al., 2011; de Cavanagh et al., 2011; Keller et al., 2019). Based on findings in C. elegans (Kumar et al., 2016), the Interventions Testing Program (ITP) of the National Institute on Aging tested captopril in normotensive mice; captopril significantly increased the median and maximum lifespan of female and male mice (Strong et al., 2022). Genetic studies provide important support for these pharmacology studies; disruption of the angiotensin II type I receptor (AGT1R) promotes longevity in normotensive mice (Benigni et al., 2009; Mattson and Maudsley, 2009; Nishiyama et al., 2009; Benigni et al., 2010; Cassis et al., 2010; Yabumoto et al., 2015). These results may be relevant to humans, as polymorphisms in the angiotensin II type I receptor gene are associated with extreme human longevity (Rigat et al., 1990; Benigni et al., 2013; Garatachea et al., 2013; Revelas et al., 2018).
Isaac and colleagues reported that C. elegans acn-1 (ACE-like non-metallopeptidase) encodes the homolog of mammalian ACE (Brooks et al., 2003). acn-1 RNA interference (RNAi) caused larvae to arrest at the L2 stage with unshed L1 cuticle, indicating a defect in molting. Delivering acn-1 RNAi at later stages revealed cuticle defects in L3/L4 larvae and adults, indicating that acn-1 functions to promote molting at multiple stages of larval development. Furthermore, seam cells displayed fusion defects, indicating that acn-1 plays a role in establishing cell fates (Brooks et al., 2003). Consistent with this analysis, Ruvkun and colleagues identified acn-1 in a genome-wide RNAi screen for molting defects (Frand et al., 2005). Analysis of ACN-1 transcriptional and translation reporters revealed that ACN-1 is expressed in hypodermal seam and excretory gland cells in embryos and larvae, as well as in the developing vulva and male tail (Brooks et al., 2003; Frand et al., 2005). Slack and colleagues reported that acn-1 interacts with heterochronic pathways (Metheetrairut et al., 2017). acn-1 RNAi suppresses let-7(lf) lethality and seam cell defects and enhances the precocious phenotype of hbl-1(lf). Loss of apl-1 causes a similar phenotype, and loss of both apl-1 and acn-1 causes additive or synergistic defects. Reducing acn-1 activity reduces expression of apl-1, suggesting that acn-1 functions upstream to control apl-1 expression. These results are consistent with two reports of global analysis that identified acn-1 among a set of genes with cyclic larval expression (Kim et al., 2013; Hendriks et al., 2014). These studies led to the model that acn-1 is a heterochronic gene that promotes larval seam cell fates by functioning downstream of let-7 and upstream of apl-1.
To test the hypothesis that acn-1 regulates lifespan, Kumar et al. (2016) used RNAi to reduce acn-1 activity. Wild-type animals cultured with acn-1 RNAi starting at the embryonic stage displayed significantly extended lifespan. Furthermore, acn-1 functions in adults to extend lifespan, as animals cultured with RNAi bacteria starting at the L4 stage also display an extended lifespan; this indicates that, in addition to its regulation of molting and development in larvae, acn-1 continues to be expressed in adults to control longevity. acn-1 RNAi delayed age-related declines of body movement and pharyngeal pumping and increased resistance to heat and oxidative stress, indicating that acn-1 activity accelerates aging and reduces stress resistance (Kumar et al., 2016). Captopril did not further extend the life span of animals treated with acn-1 RNAi, indicating that captopril treatment and acn-1 RNAi extend lifespan by a similar mechanism (Kumar et al., 2016). These findings support the hypothesis that acn-1 controls lifespan, and captopril extends lifespan by inhibiting ACN-1 activity (Fig. 1A). Captopril and other ACE inhibitors have been studied intensively in recent years as one of its targets, ACE2, is the cellular point-of-entry for the SARS-CoV-2 coronavirus responsible for the COVID-19 pandemic (Wrapp et al., 2020).
To investigate the mechanism of captopril in longevity control, we performed dose-response experiments and exploited this information to design a forward genetic screen for mutants that are hypersensitive to captopril. We identified am326, a novel missense allele of the daf-2 gene, which encodes the C. elegans homolog of the insulin/IGF-1 receptor. The daf-2(am326) mutation results in an alanine-to-valine substitution at amino acid 261, a residue in the extracellular L1 ligand-binding domain of the DAF-2 protein. This alanine residue is conserved in the human insulin receptor (INSR) at amino acid position 119, and humans with the identical alanine-to-valine substitution have Donohue syndrome, a disease characterized by growth restriction, morphological abnormalities, reduction of glucose homeostasis, and early death (<1 year of life) (Donohue, 1948; Longo et al., 2002). daf-2 mutations cause an extended lifespan and a dauer constitutive (Daf-c) phenotype; mutant animals enter the dauer stage under conditions where wild-type animals undergo reproductive development (Cassada and Russell, 1975; Klass and Hirsh, 1976; Klass, 1977; Kenyon et al., 1993; Kimura et al., 1997). Our results indicate that captopril treatment and reducing the activity of acn-1 by RNAi also cause a Daf-c phenotype, indicating that acn-1 controls both dauer formation and adult longevity. The lifespan effects of captopril require the daf-16 and daf-12 genes, providing further evidence that acn-1 interacts with the dauer pathway to promote longevity. Together, our results advance understanding of the mechanism of action of the ACE pathway in longevity control.
RESULTS
High-dose captopril causes developmental arrest
The optimal captopril dose for extending longevity in C. elegans, 2.5 mM dissolved in nematode growth medium (NGM), increased lifespan more than 30% (Fig. 1C,D, Table S1). It has also been shown to delay the age-related decline of pharyngeal pumping and body movement (Kumar et al., 2016). This dose also increases heat and oxidative stress resistance (Kumar et al., 2016). The longevity effect is dose dependent; reducing the concentration to 0.3 mM or increasing the concentration to 5.7 mM abrogated lifespan extension (Fig. 1C,D, Table S1). At even higher doses, captopril delayed larval development, reduced animal length at adulthood, and reduced lifespan compared with untreated animals, indicating dose-dependent toxicity (Fig. 1C,D, Fig. S1A,B, Table S1). On 16 mM captopril, most wild-type worms grew to adulthood and produced progeny by day 3 post-hatch (Fig. 2B, Fig. S1B). However, at a higher dose of 19 mM captopril, wild-type worms displayed high rates of lethality and failed to reach adulthood (Fig. 2B).
A forward-genetic screen for mutants hypersensitive to high-dose captopril. (A) Diagram of a forward-genetic screen for mutant animals that are hypersensitive to high-dose captopril. Wild-type (WT) L4 hermaphrodites were mutagenized with ethyl methanesulfonate (EMS); clonal populations derived from individual F2 self-progeny were analyzed for failure to grow to adulthood on 16 mM captopril. (B) Brightfield images showing WT and am326 mutant animals cultured with the indicated concentration of captopril in the medium. WT animals mature to adulthood and produce progeny when cultured with 0, 10 and 16 mM captopril (black arrows) but display larval arrest when cultured with 19 mM captopril (yellow arrow). am326 mutant animals mature to adulthood and produce progeny when cultured with 0 and 10 mM captopril (black arrows) but develop as dauer larvae when cultured with 16 and 19 mM captopril (red arrows). Some am326 animals enter dauer when cultured with 10 mM captopril. Scale bars: 0.5 mm.
A forward-genetic screen for mutants hypersensitive to high-dose captopril. (A) Diagram of a forward-genetic screen for mutant animals that are hypersensitive to high-dose captopril. Wild-type (WT) L4 hermaphrodites were mutagenized with ethyl methanesulfonate (EMS); clonal populations derived from individual F2 self-progeny were analyzed for failure to grow to adulthood on 16 mM captopril. (B) Brightfield images showing WT and am326 mutant animals cultured with the indicated concentration of captopril in the medium. WT animals mature to adulthood and produce progeny when cultured with 0, 10 and 16 mM captopril (black arrows) but display larval arrest when cultured with 19 mM captopril (yellow arrow). am326 mutant animals mature to adulthood and produce progeny when cultured with 0 and 10 mM captopril (black arrows) but develop as dauer larvae when cultured with 16 and 19 mM captopril (red arrows). Some am326 animals enter dauer when cultured with 10 mM captopril. Scale bars: 0.5 mm.
Determination of internal captopril concentration
C. elegans dwells in a chemically complex soil environment and is adept at excluding xenobiotic compounds. Thus, the drug concentration inside worms is typically much lower than in the environment, often two to four orders of magnitude lower (Lindblom and Dodd, 2006; Burns et al., 2010; Xiong et al., 2017). The concentration of captopril in the medium that is optimal for lifespan extension was established, but the concentration inside the worms was unknown. To determine this value, we cultured a large population of animals with 2.5 and 10 mM captopril in the medium, collected a whole-worm lysate, and used high performance liquid chromatography/mass spectroscopy (HPLC-MS) to measure the internal concentration of captopril. When the external concentration was 2.5 mM, the concentration in worm lysate was ∼0.02 mM; this value is 107-fold lower than the external concentration, demonstrating that these animals effectively exclude captopril (Fig. 1B). Compared with the serum concentration in humans treated with captopril (∼0.002-0.003 mM), this value is higher by less than one order of magnitude, indicating that the lifespan-extending concentration in nematodes is similar to the therapeutic concentration in humans (Bahmaei et al., 1997; Huang et al., 2006). Increasing the external concentration to 10 mM resulted in worm lysate with ∼0.06 mM captopril, 169-fold lower than the external concentration. These results indicate that the internal concentration of captopril is dose dependent (Fig. 1B).
A forward-genetic screen for captopril-hypersensitive mutants
To investigate the mechanism of lifespan extension caused by captopril, we performed a genetic screen for mutants with an abnormal dose response to the drug. The aging-related phenotypes caused by captopril, such as increased lifespan or delayed decline of body movement rate, are laborious to score and not suitable for such a genetic screen. Instead, we chose to screen for mutations that caused hypersensitivity to captopril larval arrest/lethality, as this phenotype can be readily scored. The highest dose in the medium at which the majority of wild-type animals survived to adulthood and produced progeny was 16 mM captopril (Fig. 2B, Fig. S1B); thus, we performed a clonal screen for populations of mutant animals that failed to survive and reproduce at this dose (Fig. 2A).
Wild-type L4 hermaphrodites were exposed to a chemical mutagen, 50 mM ethylmethanesulfanate (EMS), as described by Brenner (1974). These P0 animals were allowed to recover and reproduce, and F1 self-progeny were moved to individual dishes. F2 self-progeny were moved to individual dishes and cultured until F3 self-progeny reached adulthood. F3 adults were treated with alkaline hypochlorite to yield eggs, and populations of F4 eggs were cultured on dishes containing 16 mM captopril. After 4-5 days, dishes with populations that lacked fertile adults were identified using a dissecting microscope.
To control for mutants that fail to thrive under non-specific stress conditions, we incorporated control dishes with a metal stress environment into the genetic screen. F4 populations that were challenged with 16 mM captopril were also cultured on dishes containing high-dose manganese, high-dose zinc, or zinc deficiency caused by a zinc chelator (TPEN). If a mutant strain failed to reproduce in multiple stress conditions, the mutant was discarded. After screening ∼1950 haploid mutant genomes, we identified one mutant strain that displayed specific hypersensitivity to captopril; we designated this mutation am326 (Fig. 2B). This strain was backcrossed three times to the wild-type N2 strain, and the resulting strain (WU1939) was used for all future experiments.
am326, a mutation that causes hypersensitivity to captopril, also causes lifespan extension and promotes dauer development
When cultured with high-dose captopril, am326 mutant animals failed to mature into adults and produce progeny after 5 days (Fig. 2B). We noticed that these mutant animals formed dauer larvae, as determined by morphology observed with a dissecting microscope and survival following treatment with sodium dodecyl sulfate (SDS) (Cassada and Russell, 1975; Riddle et al., 1981; Swanson and Riddle, 1981; Vowels and Thomas, 1992; Hu, 2018). Dauer is an alternative L3 larval stage that C. elegans can enter when conditions are not optimal for reproductive development (Cassada and Russell, 1975; Klass and Hirsh, 1976; Klass, 1977). These conditions include high population density, low resource availability, and elevated temperature. Dauer animals are stress resistant and extremely long-lived (Larsen, 1993). am326 mutants displayed high rates of dauer formation when cultured with 16 mM captopril; by contrast, wild-type animals did not form dauer larvae under these conditions (Figs 2B, 3A, Table S2). The percentage of dauers in a population increased as the captopril concentration increased, indicating that dauer formation is dose dependent (Fig. 3A, Table S2).
Analysis of am326 phenotypes. (A) Wild-type (WT) (black line) and am326 mutant (red line) embryos were cultured on dishes containing the indicated concentration of captopril at 20°C. Values are the average percentage of animals in the dauer stage and s.d. (see Table S2 for statistics). (B) WT and am326 mutant animals were cultured with 0 or 2.5 mM captopril (+ Cap). Values indicate the fraction of the starting population that remains alive (see Table S3 for statistics). (C) Movement of WT and am326 mutant animals was measured by counting body turns in 15 s intervals using a dissecting microscope throughout adulthood. Values are body bends per minute and s.d. n=25-40 animals, one-way ANOVA with Tukey's post-hoc HSD. (D) Pharyngeal pumping rate of WT and am326 mutant animals was measured in 15 s intervals using a dissecting microscope throughout adulthood. Values are average pharynx pumps per minute and s.d. n=27-30 animals, one-way ANOVA with Tukey's post-hoc HSD. (E) Thermotolerance was measured by culturing WT and am326 mutant embryos at 20°C for 72 h (until ∼adult day 0), shifting animals to 35°C, and measuring survival every 12 h. Values indicate the fraction of the starting population that remains alive (see Table S3 for statistics). (F) Resistance to oxidative stress was measured by culturing WT and am326 embryos on standard medium for 72 h (until ∼adult day 0), transferring animals to medium containing 40 mM paraquat, and measuring survival every 12 h. In this experiment, animals that died from matricidal hatching were not censored because matricidal hatching was the primary cause of death for paraquat-treated animals. Values indicate the fraction of the starting population that remains alive (see Table S3 for statistics). (G) Freshly laid eggs from WT and am326 hermaphrodites were cultured for 65 h; resulting self-fertile adults were observed hourly to determine the time until first egg lay. Values are average time and s.d. n=15-16 animals, two-tailed, unpaired Student's t-test. (H,I) Total brood size and daily progeny production of self-fertile hermaphrodites were measured by counting the number of viable self-progeny produced by WT and am326 hermaphrodites throughout their reproductive period. Values are average number of progeny and s.d. n=14 animals. (H) Two-tailed, unpaired Student's t-test. (I) One-way ANOVA with Tukey's post-hoc HSD. ns, non-significant (P≥0.05); *P<0.05; **P<0.01; ***P<0.001.
Analysis of am326 phenotypes. (A) Wild-type (WT) (black line) and am326 mutant (red line) embryos were cultured on dishes containing the indicated concentration of captopril at 20°C. Values are the average percentage of animals in the dauer stage and s.d. (see Table S2 for statistics). (B) WT and am326 mutant animals were cultured with 0 or 2.5 mM captopril (+ Cap). Values indicate the fraction of the starting population that remains alive (see Table S3 for statistics). (C) Movement of WT and am326 mutant animals was measured by counting body turns in 15 s intervals using a dissecting microscope throughout adulthood. Values are body bends per minute and s.d. n=25-40 animals, one-way ANOVA with Tukey's post-hoc HSD. (D) Pharyngeal pumping rate of WT and am326 mutant animals was measured in 15 s intervals using a dissecting microscope throughout adulthood. Values are average pharynx pumps per minute and s.d. n=27-30 animals, one-way ANOVA with Tukey's post-hoc HSD. (E) Thermotolerance was measured by culturing WT and am326 mutant embryos at 20°C for 72 h (until ∼adult day 0), shifting animals to 35°C, and measuring survival every 12 h. Values indicate the fraction of the starting population that remains alive (see Table S3 for statistics). (F) Resistance to oxidative stress was measured by culturing WT and am326 embryos on standard medium for 72 h (until ∼adult day 0), transferring animals to medium containing 40 mM paraquat, and measuring survival every 12 h. In this experiment, animals that died from matricidal hatching were not censored because matricidal hatching was the primary cause of death for paraquat-treated animals. Values indicate the fraction of the starting population that remains alive (see Table S3 for statistics). (G) Freshly laid eggs from WT and am326 hermaphrodites were cultured for 65 h; resulting self-fertile adults were observed hourly to determine the time until first egg lay. Values are average time and s.d. n=15-16 animals, two-tailed, unpaired Student's t-test. (H,I) Total brood size and daily progeny production of self-fertile hermaphrodites were measured by counting the number of viable self-progeny produced by WT and am326 hermaphrodites throughout their reproductive period. Values are average number of progeny and s.d. n=14 animals. (H) Two-tailed, unpaired Student's t-test. (I) One-way ANOVA with Tukey's post-hoc HSD. ns, non-significant (P≥0.05); *P<0.05; **P<0.01; ***P<0.001.
To investigate the role of the am326 mutation in aging, we measured age-related phenotypes. Homozygous am326 mutant animals displayed an increased lifespan of 74.5% compared with wild type; captopril treatment further extended the lifespan of am326 mutants, indicating that the mutation does not abrogate the effect of the drug (Fig. 3B, Table S3). To analyze health span, we measured age-related declines of body movement and pharynx pumping. Homozygous am326 mutant animals displayed reduced age-related declines in body movement and pharyngeal pumping; these phenotypes are only apparent after mid-life (Fig. 3C,D). Some mutations that extend lifespan also increase stress resistance (Lithgow et al., 1995; Johnson et al., 2000). Homozygous am326 mutant animals displayed increased heat and oxidative stress resistance (Fig. 3E,F, Table S3). Some mutations that extend lifespan also delay development, reduce fecundity, and/or extend the reproductive period (Huang et al., 2004; Hughes et al., 2011; Luo and Murphy, 2011). Homozygous am326 mutants did not display delayed development compared with wild type, as determined by measuring the duration from hatching to egg-laying (Fig. 3G). Homozygous am326 mutant hermaphrodites displayed a 35% reduction in self-progeny number compared with wild type (Fig. 3H) without an extension of the reproductive period (Fig. 3I). Thus, the am326 mutation causes a large-magnitude lifespan extension, extension of neuromuscular health span, increased stress resistance, and a moderate reduction in brood size, but does not cause a delay in development or reproduction.
am326 is a previously uncharacterized missense mutation in the daf-2 receptor-tyrosine kinase gene
To identify the molecular lesion in the am326 strain, we utilized an EMS density-mapping strategy (Zuryn et al., 2010). We performed multiple independent backcrosses to the wild-type N2 strain using the temperature-dependent dauer formation phenotype to homozygose am326/am326 animals. Using whole-genome sequencing, we generated a list of 45 base pair variants that were present in all backcrossed am326/am326 strains but were absent from wild type, indicating that they were caused by the EMS mutagenesis and are strongly linked to am326 (Table S4). Thirty-six variants were clustered in a region of chromosome III between 2.5 and 5.6 Mbp, suggesting that one of the variants in this region was likely the causative mutation. The remaining nine variants were dispersed on other chromosomes (Fig. 4A, Table S4). We predicted the effect of each variant on gene activity; variants outside coding regions (introns, 5′ untranslated regions, intergenic regions, or non-coding RNA) were excluded. Eight variants affected exons, including one missense change in the daf-2 gene. This mutation is a single nucleotide change (cytosine to thymine) at position 782, resulting in an alanine (GCG) to valine (GTG) substitution. The lesion occurs in exon 7, which encodes the extracellular L1 ligand-binding domain of the DAF-2 protein (Fig. 4C).
am326 is an allele of daf-2. (A) The x-axis represents the location on chromosome III in megabases (Mb). Bars indicate the number of base pair variants in the am326 strain in 500,000 base pair bins. Red indicates the location of am326, an allele of daf-2, at position 3,015,386 bp (see Table S4 for details). (B) WT and daf-2(syb4952A261V), a CRISPR-generated mutant strain, were analyzed for survival. Values indicate fraction of the starting population that remains alive (***P<0.001; see Table S5 for statistics). (C) Left: Diagram of the predicted DAF-2 protein as a dimer with domains colored and labeled. The am326 mutation causes an alanine-to-valine substitution at amino acid 261 that affects the L1 extracellular ligand-binding domain (L1BD). Right: Structure of the DAF-2 protein showing the location of the alanine-to-valine substitution in the L1BD and the location of other amino acid substitutions caused by previously characterized mutations of daf-2. Orange boxes highlight daf-2 mutations used in this study. Adapted from Patel et al. (2008). (D) Multiple sequence alignment of DAF-2 and homologs in human, mouse and Drosophila. Conserved amino acids are red; alanine 261 is boxed. Alignment was performed using COBALT (Papadopoulos and Agarwala, 2007).
am326 is an allele of daf-2. (A) The x-axis represents the location on chromosome III in megabases (Mb). Bars indicate the number of base pair variants in the am326 strain in 500,000 base pair bins. Red indicates the location of am326, an allele of daf-2, at position 3,015,386 bp (see Table S4 for details). (B) WT and daf-2(syb4952A261V), a CRISPR-generated mutant strain, were analyzed for survival. Values indicate fraction of the starting population that remains alive (***P<0.001; see Table S5 for statistics). (C) Left: Diagram of the predicted DAF-2 protein as a dimer with domains colored and labeled. The am326 mutation causes an alanine-to-valine substitution at amino acid 261 that affects the L1 extracellular ligand-binding domain (L1BD). Right: Structure of the DAF-2 protein showing the location of the alanine-to-valine substitution in the L1BD and the location of other amino acid substitutions caused by previously characterized mutations of daf-2. Orange boxes highlight daf-2 mutations used in this study. Adapted from Patel et al. (2008). (D) Multiple sequence alignment of DAF-2 and homologs in human, mouse and Drosophila. Conserved amino acids are red; alanine 261 is boxed. Alignment was performed using COBALT (Papadopoulos and Agarwala, 2007).
DAF-2 is a well-studied member of the insulin/IGF-1 signaling (IIS) pathway; numerous partial loss-of-function mutations in this gene have been shown to affect aging, development, and dauer entry (Kenyon et al., 1993; Gems et al., 1998; Patel et al., 2008). We hypothesized that the missense mutation in daf-2 causes the am326 mutant phenotype, given that daf-2 mutations have been demonstrated to cause phenotypes including Daf-c, extended lifespan, stress resistance, and reduced brood size (Kenyon et al., 1993; Dorman et al., 1995; Kimura et al., 1997) (reviewed by Gems et al., 1998; Patel et al., 2008). To test this hypothesis, we performed complementation assays. Male P0am326 animals were crossed to hermaphrodite P0 animals with daf-2(lf) mutations, as well as hermaphrodite P0 animals with an age-1(lf) mutation as a control; age-1(lf) mutants also display temperature-dependent dauer formation (Friedman and Johnson, 1988a,b; Klass, 1983). Heterozygous F1 larvae were temperature-shifted to 25°C for 48 h and observed for dauer formation. Table 1 shows that am326 failed to complement daf-2(lf) mutations but complemented an age-1(lf) mutation, indicating that am326 is an allele of daf-2.
DNA sequencing confirmed the presence of the daf-2(am326A261V) allele in the originally isolated am326 mutant strain and in four serially backcrossed mutant strains (WU1975-WU1978); this allele was not present in the wild-type strain used for EMS mutagenesis, indicating a strong correlation between the daf-2(am326A261V) allele and the Daf-c phenotype. Because alanine and valine are similar amino acids, we wanted to confirm rigorously that this missense change causes the phenotype. CRISPR genome editing was used to insert this base change into the endogenous daf-2 locus of a wild-type N2 strain. The resulting daf-2(syb4952A261V) mutant animals displayed a 100% penetrant Daf-c phenotype (Table 1). Furthermore, the daf-2(syb4952) allele failed to complement the daf-2(am326) and daf-2(e1370) mutations for the Daf-c phenotype, demonstrating the A261V change caused a daf-2 loss-of-function phenotype (Table 1). As expected, daf-2(syb4952) mutant animals displayed temperature- and captopril-dependent dauer formation (Table 1, Table S2) and extended longevity (Fig. 4B, Table S5). To determine whether this amino acid is conserved, we performed amino acid sequence alignment with homologous proteins. Tyrosine 260, alanine 261 and leucine 262 are completely conserved in human and mouse INSR and the Drosophila insulin receptor (InR), indicating that this region has an important function (Fig. 4D).
Captopril treatment influences dauer development
Some daf-2 loss-of function mutations, such as daf-2(e1370), cause a temperature-sensitive Daf-c phenotype (Kenyon et al., 1993; Patel et al., 2008). To evaluate the temperature sensitivity of daf-2(am326), we cultured animals at 20, 22 and 25°C. The percentage of dauer larvae increased significantly from 0% at 20°C to 17% at 22°C and 90% at 25°C (Fig. 5A, Table S2). To investigate the interaction with captopril treatment, we cultured animals with 0, 4, 8, 12 or 16 mM captopril at these three temperatures. At 20 and 22°C, captopril treatment caused a dose-dependent increase in dauer formation (Fig. 5A, Table S2). At 25°C, the lowest dose of captopril (4 mM) resulted in constitutive dauer formation; at 16 mM captopril, the percentage of animals that entered dauer was reduced because some animals displayed larval arrest prior to the L2 stage. Thus, dauer formation caused by captopril is dose dependent and additive with high temperature.
Captopril promotes dauer formation. (A) daf-2(am326) embryos were cultured with the indicated concentration of captopril at 20°C (red), 22°C (orange) or 25°C (yellow). Values are average percentage of animals in the dauer stage and s.d. At 22 and 25°C, animals treated with 16 mM captopril displayed less than 100% dauer because some animals arrested at developmental stages prior to L2. (B) Wild-type (WT; black line) and daf-2(e1370) mutant (blue line) embryos were cultured on dishes containing the indicated concentration of captopril at 20°C. Values are average percentage of animals in the dauer stage and s.d. (C) WT embryos were cultured with either 0 (−) or 16 (+) mM captopril at the indicated temperatures. Values are average percentage of animals in the dauer stage and s.d. (see Table S2 for statistics for this figure). ns, non-significant (P≥0.05); *P<0.05; **P<0.01; ***P<0.001.
Captopril promotes dauer formation. (A) daf-2(am326) embryos were cultured with the indicated concentration of captopril at 20°C (red), 22°C (orange) or 25°C (yellow). Values are average percentage of animals in the dauer stage and s.d. At 22 and 25°C, animals treated with 16 mM captopril displayed less than 100% dauer because some animals arrested at developmental stages prior to L2. (B) Wild-type (WT; black line) and daf-2(e1370) mutant (blue line) embryos were cultured on dishes containing the indicated concentration of captopril at 20°C. Values are average percentage of animals in the dauer stage and s.d. (C) WT embryos were cultured with either 0 (−) or 16 (+) mM captopril at the indicated temperatures. Values are average percentage of animals in the dauer stage and s.d. (see Table S2 for statistics for this figure). ns, non-significant (P≥0.05); *P<0.05; **P<0.01; ***P<0.001.
To determine whether captopril-induced dauer formation was specific for the am326 allele or a general property of daf-2(lf) mutations, we measured the effect of captopril on daf-2(e1370). Captopril induced dauer formation in a dose-dependent manner, demonstrating that this effect is not specific to daf-2(am326) (Fig. 5B, Table S2).
To evaluate the chemical specificity of the captopril effect, we analyzed a different small molecule drug that extends lifespan and is toxic at high doses, the anticonvulsant ethosuximide (Evason et al., 2005; Collins et al., 2008). Ethosuximide is an FDA-approved medication previously shown to extend lifespan in C. elegans at a similar concentration to captopril (2-4 mM). We exposed daf-2(am326) mutants to ethosuximide at concentrations up to 32 mM; no significant dose-dependent effect on dauer formation was observed at 20, 22 or 25°C (Fig. S2A,B, Table S6). These results suggest that the dauer formation phenotype is a distinctive effect of captopril and not a general xenobiotic stress response.
To determine whether dauer formation caused by captopril requires a daf-2 mutation or also occurs in wild-type animals, we used high temperature to sensitize wild-type animals to dauer formation and examined the effect of captopril. Wild-type animals were cultured at 20, 22, 25 and 27°C, as 27°C was previously shown to induce dauer in HID (high-temperature dauer induction) mutants (Ailion and Thomas, 2003). At 22 and 25°C, 16 mM captopril caused a small increase of dauer formation to 0.8% and 3.9%, respectively, which was not statistically significant with this sample size. At 27°C, 16 mM captopril significantly increased dauer formation to 12.8% (Fig. 5C, Table S2). Thus, the dauer-promoting effect of captopril can be observed by sensitizing animals using a daf-2(lf) mutation or high temperature in wild-type animals.
The acn-1 gene promotes dauer formation
Based on our model that captopril functions by inhibiting the acn-1 gene (Fig. 1A), we hypothesized that reducing acn-1 activity would promote dauer formation. To test this hypothesis, we treated daf-2(am326) mutants with bacteria expressing acn-1 dsRNA. acn-1 RNAi failed to induce dauer in wild-type animals at 20, 22 or 25°C (Fig. 6A, Table S7), consistent with our observation that captopril failed to significantly induce dauer formation under similar conditions. However, acn-1 RNAi significantly increased the percentage of dauer-stage animals in am326 mutants from 13.8% to 31.5% at 22°C (Fig. 6A, Table S7). This finding suggests that reducing acn-1 activity through RNAi or inhibition with captopril promotes dauer formation under the sensitized conditions created by the daf-2(am326) mutation.
acn-1 is expressed in larvae and adults and influences dauer formation and adult lifespan. (A) Wild-type (WT) and daf-2(am326) embryos were cultured at 22°C with E. coli HT115 expressing either acn-1 dsRNA or a vector control. Values are average percentage of dauer larvae and s.d. (see Table S7 for statistics). (B) spe-9(hc88) animals arrested at the L1 stage were transferred to dishes with abundant E. coli and cultured at 25°C until the indicated day; total RNA was analyzed by RNA-seq. Values are acn-1 mRNA level in arbitrary units (AU) and s.d. Days −1 and 0 were larval stages; day 1 was defined as the first day of adulthood; at 25°C, day 1 occurred approximately 2 days after arrested L1 animals were transferred to food. (C) Diagram of the acn-1 gene showing the location of am314, a CRISPR-generated mutation. Green boxes, exons; black lines, introns; dark green box, 3′ UTR. (D) WT and acn-1(am314) animals (mixed genotype population) were analyzed for survival. Values indicate the fraction of the starting population remaining alive at that time point (see Table S5 for statistics). **P<0.01.
acn-1 is expressed in larvae and adults and influences dauer formation and adult lifespan. (A) Wild-type (WT) and daf-2(am326) embryos were cultured at 22°C with E. coli HT115 expressing either acn-1 dsRNA or a vector control. Values are average percentage of dauer larvae and s.d. (see Table S7 for statistics). (B) spe-9(hc88) animals arrested at the L1 stage were transferred to dishes with abundant E. coli and cultured at 25°C until the indicated day; total RNA was analyzed by RNA-seq. Values are acn-1 mRNA level in arbitrary units (AU) and s.d. Days −1 and 0 were larval stages; day 1 was defined as the first day of adulthood; at 25°C, day 1 occurred approximately 2 days after arrested L1 animals were transferred to food. (C) Diagram of the acn-1 gene showing the location of am314, a CRISPR-generated mutation. Green boxes, exons; black lines, introns; dark green box, 3′ UTR. (D) WT and acn-1(am314) animals (mixed genotype population) were analyzed for survival. Values indicate the fraction of the starting population remaining alive at that time point (see Table S5 for statistics). **P<0.01.
To elucidate how reducing acn-1 activity causes lifespan extension, we monitored acn-1 expression daily beginning in larvae and extending until ∼day 12 of adulthood. acn-1 mRNA levels were measured by RNA sequencing (RNA-seq) in a synchronized population of spe-9(lf) animals cultured at 25°C to prevent progeny production. acn-1 mRNA levels were high in larvae, consistent with previous reports (Fig. 6B) (Brooks et al., 2003; Frand et al., 2005; Oskouian et al., 2005; Metheetrairut et al., 2017). Interestingly, acn-1 mRNA levels declined in young adults but then remained relatively constant throughout adulthood (Fig. 6B). These results indicate acn-1 is expressed in young and old adults, consistent with an activity in adults that controls aging.
We previously used RNAi to investigate acn-1 function in adults. To extend this analysis, we used CRISPR/Cas9 genome editing to generate a chromosomal mutation, acn-1(am314). This allele contains a substitution in exon 1 (Fig. 6C). Animals containing the acn-1(am314) mutation were allowed to produce L4 progeny, which were then analyzed for adult lifespan. Mean lifespan was significantly extended by 15.1% compared with wild-type controls (Fig. 6D, Table S5). We interpret these results to indicate that partial reduction of acn-1 activity caused a lifespan extension, indicating that acn-1 is necessary to accelerate adult lifespan. This result is consistent with previous results that acn-1 RNAi, which also partially reduces gene activity, causes a lifespan extension (Kumar et al., 2016).
Extended longevity and dauer formation caused by captopril and reducing acn-1 activity are dependent on the daf-16 and daf-12 genes
daf-16 encodes a forkhead transcription factor that functions downstream of daf-2 (Larsen et al., 1995; Lin et al., 1997; Ogg et al., 1997). daf-16(lf) mutations cause a dauer defective (Daf-d) phenotype and suppress both the Daf-c and lifespan-extension phenotypes caused by daf-2(lf) mutants (Gottlieb and Ruvkun, 1994). To investigate the interaction with the new daf-2(am326) mutation, we constructed a double mutant with daf-16(mu86), a strong loss-of-function allele. Whereas captopril treatment extended lifespan and promoted dauer formation in daf-2(am326) single mutants, captopril failed to extend lifespan or induce dauer formation in these double-mutant animals (Table S5). Thus, daf-16 is necessary for the lifespan extension and Daf-c phenotype caused by daf-2(am326) and/or captopril. Kumar et al. (2016) reported that captopril and acn-1 RNAi do not extend the lifespan of daf-16(lf) mutants, indicating that the activity of daf-16 is necessary for these lifespan-extending treatments; here, we confirm this result for captopril treatment (Fig. 7A, Table S5). In addition, captopril treatment and acn-1 RNAi did not induce dauer formation in daf-16(lf) mutants (Tables S2, S7).
Captopril effects depend on daf-12 and daf-16. (A,B) daf-16(mu86) and daf-12(rh61rh411) are loss-of-function mutations. Mutant animals were cultured with 0 or 2.5 mM captopril (+ Cap). Values indicate the fraction of the starting population remaining alive at each time point. Captopril treatment caused a small yet significant reduction in mean lifespan in daf-12 mutants and no significant change in daf-16 mutants. ns, non-significant (P≥0.05); *P<0.05 (see Table S5 for statistics). (C) A model for captopril/ACN-1-mediated control of aging and dauer formation. The DAF-2 receptor-tyrosine kinase inhibits the DAF-16 transcription factor, thereby reducing lifespan, inhibiting dauer formation, and promoting larval development into reproductive adulthood. The DAF-12 transcription factor acts parallel to this pathway. ACN-1 functions to reduce adult lifespan, inhibit dauer formation, and promote larval development into reproductive adulthood; captopril functions by inhibiting ACN-1. We propose two models for ACN-1. Model I (blue): ACN-1 inhibits the activity of both the DAF-16 and DAF-12 transcription factors, reducing their ability to increase lifespan, promote dauer formation, and inhibit reproductive development. Model II, (red): ACN-1 acts parallel to DAF-16 and DAF-12 to affect these phenotypes via an independent mechanism.
Captopril effects depend on daf-12 and daf-16. (A,B) daf-16(mu86) and daf-12(rh61rh411) are loss-of-function mutations. Mutant animals were cultured with 0 or 2.5 mM captopril (+ Cap). Values indicate the fraction of the starting population remaining alive at each time point. Captopril treatment caused a small yet significant reduction in mean lifespan in daf-12 mutants and no significant change in daf-16 mutants. ns, non-significant (P≥0.05); *P<0.05 (see Table S5 for statistics). (C) A model for captopril/ACN-1-mediated control of aging and dauer formation. The DAF-2 receptor-tyrosine kinase inhibits the DAF-16 transcription factor, thereby reducing lifespan, inhibiting dauer formation, and promoting larval development into reproductive adulthood. The DAF-12 transcription factor acts parallel to this pathway. ACN-1 functions to reduce adult lifespan, inhibit dauer formation, and promote larval development into reproductive adulthood; captopril functions by inhibiting ACN-1. We propose two models for ACN-1. Model I (blue): ACN-1 inhibits the activity of both the DAF-16 and DAF-12 transcription factors, reducing their ability to increase lifespan, promote dauer formation, and inhibit reproductive development. Model II, (red): ACN-1 acts parallel to DAF-16 and DAF-12 to affect these phenotypes via an independent mechanism.
When activated, DAF-16 localizes to the nucleus and activates transcription of genes that control stress resistance and aging. However, previous research showed that captopril treatment does not induce nuclear localization of DAF-16, measured by observation of a DAF-16::GFP fusion protein (Kumar et al., 2016). Because this negative result may indicate a lack of sensitivity of the DAF-16::GFP localization assay, we analyzed expression of a daf-16 target gene, sod-3, using real-time quantitative polymerase chain reaction (RT-qPCR) in adult day 1 animals (Honda and Honda, 1999; Yanase et al., 2002; Wook Oh et al., 2006; Landis and Murphy, 2010). Captopril did not increase sod-3 expression levels (Fig. S3A), suggesting that captopril does not activate DAF-16 transcriptional activity. To determine whether the effect might be age dependent, we analyzed young (adult day 1) and middle-aged (adult day 5) animals using a fluorescent sod-3p::gfp transcriptional reporter. Captopril did not significantly increase sod-3 expression at either age (Fig. S3B). Thus, although daf-16(lf) mutations suppress the lifespan extension caused by captopril, drug treatment does not cause nuclear localization of DAF-16 or activation of the sod-3 target gene.
To further explore interactions with genes necessary for dauer larvae formation, we analyzed daf-12(lf) mutants. The daf-12 gene encodes a nuclear hormone receptor that functions in parallel with DAF-16 to promote dauer formation and transcription of longevity-inducing genes (Gottlieb and Ruvkun, 1994; Ludewig et al., 2004; Shostak et al., 2004; Fisher and Lithgow, 2006). daf-12(lf) mutations reduce lifespan and cause a Daf-d phenotype, similar to daf-16(lf) mutations (Riddle et al., 1981; Vowels and Thomas, 1992; Thomas et al., 1993; Antebi et al., 1998, 2000; Ludewig et al., 2004; Fisher and Lithgow, 2006). We first analyzed longevity: captopril treatment did not extend the lifespan of daf-12(rh61rh411) mutants, indicating that daf-12 is necessary for the captopril-induced lifespan extension (Fig. 7B, Table S5). We next analyzed dauer formation: neither captopril treatment nor acn-1 RNAi caused dauer formation in daf-12(rh61rh411) animals (Tables S2, S7). Thus, the activity of both daf-16 and daf-12 are necessary for the lifespan extension caused by captopril as well as dauer formation caused by captopril and acn-1.
DISCUSSION
The C. elegans DAF-2(A261V) mutant protein corresponds to the human INSR(A119V) mutant protein that causes Donohue syndrome
The IIS pathway controls aging across species: reducing signaling activity in C. elegans, Drosophila and mice extends lifespan and delays age-related degeneration (Kenyon et al., 1993; Clancy et al., 2001; Tatar et al., 2001; Taguchi and White, 2008; Bartke, 2008; Kenyon, 2010; Kannan and Fridell, 2013). In humans, reduced insulin signaling is typically associated with diabetes and reduced lifespan. However, humans with exceptional longevity often display reduced insulin resistance (Bonafè et al., 2003). In C. elegans, this pathway also regulates dauer entry. Using a forward genetic screen, we identified a novel daf-2 mutant and characterized a previously unobserved interaction between captopril treatment, acn-1 inhibition, and the dauer stage. Numerous daf-2 mutants have been categorized based on shared and distinct phenotypes (Gems et al., 1998; Patel et al., 2008); most mutations cause extended lifespan, ectopic dauer formation, thermotolerance, and fewer unfertilized oocytes compared with wild type. Some mutations cause larval arrest, gonad abnormalities, reduced motility and reduced fecundity. Our results suggest that daf-2(am326) is a class II allele, based on constitutive dauer larvae formation at 22 and 25°C, extended lifespan, thermotolerance, and reduced brood size. These animals are otherwise relatively healthy; we did not observe development delay, reduced motility, larval arrest, gonad abnormalities, or elevated matricidal hatching. Thus, daf-2(am326) represents a useful reagent for future research.
The daf-2(am326) mutant phenotype results from an alanine-to-valine substitution at position 261. This residue is located in the extracellular L1 ligand binding domain of the DAF-2 protein, where a variety of ligands bind to DAF-2 and activate its kinase activity (Kimura et al., 1997; Pierce et al., 2001; Kimura et al., 2011). Other missense mutations that affect this region affect DAF-2 function, including pe1230, gv51 and e979, suggesting that the function of this region is sensitive to single amino acid changes (Lewis and Hodgkin, 1977; Patel et al., 2008; Ohno et al., 2014; Bulger et al., 2017). Alanine 261 is conserved in human INSR at position 119. Interestingly, an alanine-to-valine substitution at position 119 in human INSR is associated with Donohue syndrome, a disease characterized by growth restriction, morphological abnormalities, reduction of glucose homeostasis, and early death (<1 year of life) (Longo et al., 2002). Our results contribute two insights regarding this A-to-V mutation that may be relevant to the human condition: (1) this mutation is recessive; (2) this mutation causes a strong loss of function in an intact animal. Consistent with these conclusions, cell culture experiments revealed that the A119V substitution severely reduces ligand-binding activity of the human INSR (Longo et al., 2002). Previous work by Krause and colleagues established C. elegans as a useful model for human insulin receptoropathies, although the INSR A119V mutation has not previously been examined in an animal model (Bulger et al., 2017). Thus, daf-2(am326) represents an opportunity to model a human INSR receptor pathology in the experimentally powerful C. elegans system.
Identification of acn-1 as a regulator of dauer formation
The identification of compounds that delay aging is an important goal, and model organisms such as C. elegans are useful for screening potential compounds. Although many compounds can extend lifespan in C. elegans, only a few have been shown to extend lifespan in mammalian model organisms, and none has been documented in humans (reviewed by Collins et al., 2006; Lucanic et al., 2013; Partridge et al., 2020). Pharmacological and genetic inhibition of the RAAS pathway can extend lifespan in C. elegans, Drosophila melanogaster and rodents (Egan et al., 2022). However, the mechanism has not yet been well defined. An appealing model is that RAAS inhibition extends lifespan by lowering blood pressure. However, ACE inhibitors extend lifespan even in normotensive rodents (Santos et al., 2009). Furthermore, this model fails to explain the effect in Drosophila or C. elegans, animals which lack closed circulatory systems. The ACE protein evolved prior to the evolution of the closed circulatory system, implying that ACE has an ancestral function unrelated to blood pressure regulation (Simões-Costa et al., 2005; Burggren and Reiber, 2007). The observation that ACE inhibition extends lifespan in multiple species suggests that this ancestral function may control aging. Our results indicate that ACN-1 modulates IIS signaling pathways to control aging and development in C. elegans.
During early larval development, C. elegans assesses multiple external signals (food availability, population density, temperature, etc.) and integrates these inputs into a binary developmental output: either dauer or non-dauer. Here, we provide evidence that ACN-1 is part of this dauer decision machinery. The dauer stage is a diapause mechanism that promotes survival during resource deprivation. Dauer animals can survive in conditions that are fatal to non-dauers, but at the cost of delaying reproduction. The IIS pathway is involved in interpreting external signals and implementing changes in gene expression. The DAF-2 receptor interacts with extracellular ligands, which regulates its tyrosine kinase activity, resulting in a series of phosphorylation events that inhibit the DAF-16/FOXO transcription factor. In daf-2(lf) mutants, DAF-16 is activated, promoting dauer formation and expression of stress-resistance and longevity-inducing genes, such as sod-3 (Honda and Honda, 1999; Yanase et al., 2002; Landis and Murphy, 2010). ACN-1 may function in a similar fashion to integrate or modulate external signals into developmental decision-making.
Previous studies indicate that acn-1 regulates molting, development and aging (Brooks et al., 2003; Kumar et al., 2016; Metheetrairut et al., 2017); here, we demonstrate a previously uncharacterized activity: control of dauer entry. Our genetic epistasis experiments indicate that daf-16 and daf-12 are necessary for acn-1-mediated control of dauer entry and aging. These results are consistent with two models (Fig. 7C). In Model I, ACN-1 acts upstream in a linear genetic pathway to inhibit DAF-16 and DAF-12; the mechanism of regulation is likely indirect. Inhibition of ACN-1 by captopril or RNAi promotes the activity of DAF-16 and DAF-12, thereby increasing lifespan and dauer formation. In Model II, ACN-1 functions in a parallel genetic pathway to DAF-16 and DAF-12 to reduce lifespan and promote reproductive development. Although our results are consistent with either model, it is notable that acn-1 RNAi and/or captopril treatment did not cause DAF-16 nuclear localization (Kumar et al., 2016) or activation of the sod-3 target gene, results which favor Model II. Alternatively, daf-16 may interact with acn-1 by an alternative mechanism. Our experiments have focused primarily on sod-3, a canonical target of DAF-16; however, the expression of DAF-16 target genes is dependent on the specific isoform of DAF-16 (Lee et al., 2001; Lin et al., 2001; Kwon et al., 2010; Chen et al., 2015), the tissue (Kwon et al., 2010; Libina et al., 2003), upstream signaling (Lin et al., 2001; Oh et al., 2005; Robida-Stubbs et al., 2012; Tullet et al., 2014), and other factors that allow DAF-16 to preferentially regulate different target genes (reviewed by Landis and Murphy, 2010; Tissenbaum, 2018). Our results do not exclude the possibility that acn-1 functions upstream of daf-16 by mechanisms different from controlling DAF-16 nuclear localization or activation of canonical DAF-16 target genes.
Because captopril is a drug that is toxic at high concentrations in C. elegans, it raises the possibility that it modulates longevity and dauer formation by causing a general stress response. In this model, the daf-16 gene is necessary for the effects of captopril because DAF-16 is an integral part of the general stress response. Although our data do not formally exclude this possibility, several observations suggest it is unlikely. Kumar et al. (2016) demonstrated that acn-1 RNAi can extend lifespan, and captopril cannot further extend lifespan in this background. These results indicate that acn-1 is a specific target; if captopril causes a general stress response, it would be predicted to continue to do so in this background. Here, we show that acn-1 RNAI can also promote dauer formation, similar to captopril. These results suggest that captopril acts through acn-1 to influence dauer rather than triggering a general stress response.
To date, few compounds have been identified that induce dauer formation in C. elegans. Ascarosides, a structurally diverse group of small-molecule signaling pheromones, are excreted under conditions of high population density (Golden and Riddle, 1984; Ludewig and Schroeder, 2018). Dafachronic acids, a group of endogenously produced steroid hormones, bind to and remove the co-repressor from the DAF-12 nuclear hormone receptor (Ludewig et al., 2004; Motola et al., 2006). In addition to these endogenously produced compounds, the small molecule dafadine induces the dauer state by inhibiting DAF-9, a key effector of dafachronic acid synthesis and DAF-12-dependent dauer entry (Luciani et al., 2011). Captopril is a new addition to this list, as we demonstrated captopril can induce dauer formation in a concentration-dependent manner; our results suggest that the mechanism may be modulation of the daf pathway. Among these compounds, only captopril is known to be biologically active in humans and established to affect longevity across species. Captopril may prove useful as a tool for future screens of dauer-regulating genes.
MATERIALS AND METHODS
General methods and strains
C. elegans were cultured as described by Brenner (1974). Unless otherwise noted, strains were maintained at 20°C on Petri dishes containing NGM and seeded with 200 µl Escherichia coli OP50 bacteria. Nematode strains are listed in Table S8. Strains WU1975-WU1978 were created by independently backcrossing WU1939 to wild-type N2 (Bristol).
Captopril and ethosuximide culture conditions
Captopril {N-[(S)-3-mercapto-2-methylpropionyl]-L-proline; C4042, Sigma-Aldrich} was stored as a powder, dissolved in water at 75 mg/ml, and filter sterilized. Captopril-infused NGM dishes were prepared as previously described (Kumar et al., 2016): captopril was added to liquid agar at ∼50°C immediately prior to dispensing into dishes. Captopril-infused dishes were stored at 4°C for up to 1 month. For captopril dose-response experiments, embryos were placed onto E. coli-seeded captopril dishes and observed after 72 h. Adult and larval animals were scored based on morphological characteristics using a dissecting microscope.
Ethosuximide (2-ethyl-2-methylsuccinimide; E7138, Sigma-Aldrich) was similarly stored and prepared (based on previously described methods; Evason et al., 2005; Collins et al., 2008).
Measurement of internal captopril concentration
HPLC-MS was performed on whole-animal lysates. Twenty to thirty wild-type adult hermaphrodites were cultured on medium containing 2.5 mM or 10 mM captopril and allowed to lay eggs for 2 h; the adults were removed, and the eggs were cultured at 20°C for 72 h. To remove captopril that was outside the animals, animals were washed off dishes using 1 ml water, pelleted by centrifugation [3000 rpm (1750 g) for 5 min], and the supernatant removed. Three successive wash steps were performed, and then the worms were incubated in 1 ml water on the benchtop for 1 h to allow excess captopril in the intestinal lumen to be secreted. A sample of this liquid was retained for analysis, and three further wash steps were performed. The size of the worm pellet was estimated to be 2.5 µl based on the criteria and technique established by Evason et al. (2005). The animals were lysed with proteinase K (S-1000-1, EZ BioResearch), resuspended into a final volume of 100 µl, centrifuged at high speed (12,000 g for 15 min) to pellet debris, and the supernatant was submitted for HPLC-MS.
As the internal standard for quantification, 100 ng of deuterated phenylalanine d8 (Cayman Chemical Company) was spiked in 100 µl of the extract. At the same time, four-point calibration samples (0.05 ng/100 µl, 0.5 ng/100 µl, 10 ng/100 µl and 100 ng/100 µl) were run containing 100 ng of phenylalanine d8 for absolute quantification. An HPLC-MS system equipped with a Shimadzu autosampler (20X), Shimadzu HPLC (20A), and an Applied Biosystem Sciex API-4500 Qtrap mass spectrometer were utilized for analysis operating in positive ion multiple reaction monitoring (MRM) mode. The transitions of Q1/Q3 ions for captopril and phenylalanine d8 detections were set at 218.1/116.1 and 174.1/128.1, respectively. HPLC solvent (A: 10 mM ammonium acetate in 7:3 water/ACN; B: 10 mM ammonium acetate in 1:1 isopropanol/methanol) gradient was 75% B to 30% in 4 min with a flow rate of 1 ml/min. A HILIC LC column (Waters Atlantis 2×100 mm, 3 µm) was used. The SCIEX data system (Analyst 1.52 v) was used for instrument control and data analysis (quantification). Each sample was injected twice and data were averaged.
Lifespan determination
Fifty to sixty hermaphrodites at the L4 larval stage were removed from gravid populations cultured in standard conditions onto new dishes unless otherwise noted. Beginning 1 day later (‘day 1’), animals were observed for spontaneous movement; if no movement was observed, they were gently stimulated with a piece of platinum wire. If no stimulated movement was observed, animals were considered dead and removed from the dish. Worms that died by matricidal hatching, extrusion of the gonad, or desiccation on the sides of the dish were censored from that day onward. Worms were moved to new dishes daily during the reproductive period, or as needed thereafter. Measurement continued until all worms had expired. Statistical analyses were performed using Kaplan–Meier survival function and log-rank (Mantel–Cox) test (Kaplan and Meier, 1958; Mantel, 1966).
Measurement of body movement and pharyngeal pumping
Twenty to thirty hermaphrodites at the L4 stage were removed from gravid populations cultured in standard conditions. Animals were evaluated daily, beginning one day later (‘day 1’). Body movement was measured by depositing a worm on a dish and immediately counting the number of sinusoidal movements completed in a 15 s span; one ‘body movement’ was defined as the movement of the head from its left-most position to its right-most position (or vice versa) while the animal moved either forward or backward. Pharynx pumping was counted in 15 s increments per animal. Animals were transferred to a new dish during the reproductive period or as needed thereafter. Assays were continued until all animals had expired. Living animals that displayed no body movement or pharyngeal pumping were scored as ‘0’ for the given phenotype, and dead animals were censored. Data were analyzed using one-way ANOVA with Tukey post-hoc HSD.
Measurement of heat and oxidative stress resistance
To measure resistance to acute heat stress, we transferred 20-30 adult hermaphrodites to standard NGM dishes for 1 h. Synchronized eggs laid during this hour were moved to new dishes, 30 per dish. The eggs were cultured at 20°C for 72 h, shifted to 35°C, and scored as alive or dead every 12 h until all expired. To measure resistance to oxidative stress, we prepared NGM dishes as above and added paraquat (N,N′-dimethyl-4,4′-bipyridium dichloride; 36541, Sigma-Aldrich) to the molten agar to a final concentration of 40 mM. Eggs obtained as described above were cultured at 20°C for 72 h, and then worms were transferred to paraquat-containing medium. Animals were scored as alive or dead every 12 h until all expired. We observed that matricidal hatching was the most common cause of death in animals exposed to paraquat; therefore, animals that died from matricidal hatching were not censored from the data analysis. Statistical analyses were performed using Kaplan–Meier survival function and log-rank (Mantel–Cox) test (Kaplan and Meier, 1958; Mantel, 1966).
Determination of dauer formation
Thirty to fifty embryos were picked from mixed stage populations cultured in standard conditions, placed on new dishes, and maintained at 20°C or shifted to 22, 25 or 27°C. After 2-3 days, dauer larvae were counted. The morphological criteria used to define dauer were: dark coloration, elongated and radially constricted body, lack of pharyngeal pumping, and presence of oral plug (Cassada and Russell, 1975; Riddle et al., 1981; Vowels and Thomas, 1992; Hu, 2018). Animals were observed until all had formed dauer, reached adulthood, or expired. The average percentage of dauer animals in a population was calculated by dividing the number of dauer larvae by the total number of animals. Data were analyzed using one-way ANOVA with Tukey post-hoc HSD.
Measurement of developmental rate and fecundity
To determine the percentage of embryos that can reach adulthood within 72 h, we allowed wild-type hermaphrodites to lay eggs on a standard NGM dish for 1 h. Approximately 30 eggs were picked onto several dishes containing 0, 4, 8, 12 or 16 mM captopril and cultured at 20°C for 72 h. Animals were scored as larva or adult using a dissecting microscope based on morphological criteria characteristic of late-stage larvae (size, presence of developing vulva, etc.) or adults (fully formed gonad with embryos, etc.).
To determine the time to first progeny, we placed 20-30 adults on a fresh dish for 1 h. Then, 20-30 of the synchronized eggs produced during that hour were picked onto individual dishes and cultured at 20°C for 65 h. The resulting adults were then observed each hour to determine whether they had laid an egg(s). The time when unhatched embryos were first observed was defined as the time to first progeny, which is equivalent to the duration necessary to complete the life cycle.
To measure total and daily progeny production, 20-30 L4 hermaphrodites (‘P0’) were transferred individually to fresh dishes and cultured with standard conditions. P0 animals were transferred to new dishes every 24 h; dishes containing eggs were cultured for 2-3 days to facilitate progeny identification, and the number of F1 progeny was counted. The experiment ended when self-fertile P0 hermaphrodites did not produce progeny for two consecutive days. Time to first progeny and total progeny were analyzed using two-tailed, unpaired Student's t-test; daily progeny production was analyzed using one-way ANOVA and Tukey post-hoc HSD.
Determination of worm length
Thirty to forty embryos were picked onto dishes with medium containing 16 mM captopril and cultured for 72 h. Animals were washed into Eppendorf tubes using 1 ml M9 and collected by centrifugation (6000 g for 1 min). After aspirating the supernatant, three further wash steps were performed using 1 ml M9 to remove excess bacteria. The animals were pipetted onto standard NGM dishes lacking a bacterial lawn. Excess liquid was allowed to dry, and the animals were allowed to crawl away from each other. Images of individual worms were obtained using a Leica Microsystems IC80 HD microscope and then converted to grayscale using Fiji (ImageJ, 1.53f51). Determination of length was performed using the WormSizer plugin (Moore et al., 2013), and data were analyzed using two-tailed, unpaired Student's t-test.
A forward genetic screen for captopril-hypersensitive mutants
To perform the chemical mutagenesis, we cultured large numbers of wild-type L4 hermaphrodites on dishes, washed the animals into 15 ml conical tubes using M9 buffer, pelleted animals by centrifugation at 3000 rpm (1750 g) for 3 min, and aspirated excess liquid to yield a volume of 4 ml. EMS (M0880, Sigma-Aldrich) was added to a final concentration of 50 mM, and animals were gently agitated on a shaker for 4 h at 20°C. To remove the EMS, we performed four wash steps: animals were pelleted, the supernatant was aspirated, and the animals were resuspended in 10 ml M9 buffer. Animals were transferred to NGM dishes seeded with E. coli OP50 and allowed to recover and lay eggs. F1 self-progeny embryos were picked onto individual dishes; one F2 self-progeny from each F1 hermaphrodite was picked onto an individual dish and cultured until a gravid population had formed. The F3+ embryos were extracted using alkaline hypochlorite treatment and divided equally between dishes that were untreated, or contained 16 mM captopril, 50 µM zinc chloride (229997, Sigma-Aldrich), 10 µM TPEN [N,N,N′,N′-tetrakis-(2-pyridylmethyl)ethylenediamine; 616394, Sigma-Aldrich] or 500 µM manganese (II) chloride (M8054, Sigma-Aldrich); dishes were seeded with 200 µl of 5× concentrated E. coli OP50 bacteria. Animals were allowed to grow for 3-5 days. If a strain failed to reach adulthood and produce progeny when cultured with 16 mM captopril, but did reach adulthood and produce progeny in all other stressful conditions, then it was pursued.
Whole-genome sequencing of daf-2(am326) mutants
The am326 mutation was identified with whole-genome sequencing using an EMS density-mapping strategy described by Zuryn et al. (2010). Using the temperature-dependent dauer phenotype to score am326 mutant animals, we performed three successive outcrosses to N2, generating WU1939. Next, four parallel populations of WU1939 were established, and each was serially outcrossed to N2 four times. These strains, WU1975, WU1976, WU1977 and WU1978, each contained homozygous mutations at the am326 locus and had been outcrossed a total of seven times.
To collect chromosomal DNA from these strains, we cultured mixed-stage populations on six to eight dishes. When all bacteria had been consumed, the animals were collected with 1 ml M9 buffer into 15 ml conical tubes, centrifuged at 3000 rpm (1750 g) for 5 min, and excess liquid aspirated. Four successive wash steps were performed with M9 to remove excess bacteria, and the worm pellet was snap-frozen using liquid nitrogen. Four further freeze-thaw cycles were performed to lyse the worms, and the lysate was treated with proteinase K in a 55°C water bath for 3 h. To remove excess worm debris, we centrifuged the lysate at 3000 rpm (1750 g) for 10 min and removed the supernatant from the pellet. Chromosomal DNA was extracted using the DNeasy Blood and Tissue Kit (69504, QIAGEN). For each sample, 4 µg of DNA were submitted to the Washington University Genome Technology Access Center (GTAC) for whole-genome sequencing. Libraries were constructed using the KAPA Hyper PCR-free Kit (KR0961, KAPA Biosystems), and sequencing was performed using a NovaSeq S4 XP (Illumina) at a coverage depth of ∼30X.
Analysis of whole-genome sequencing data
Analysis of whole-genome sequencing data was performed using an EMS-density mapping approach (Zuryn et al., 2010; Zuryn and Jarriault, 2013). First, we aligned the reads to the WS220.64/ce10 C. elegans reference genome originally derived by the Genome Institute at Washington University (WUSTL) and the Sanger Institute. Reads were aligned using the bowtie2 read-alignment tool (Langmead and Salzberg, 2012). Analysis of differences between strains was performed using the MiModD v.0.1.9 toolkit (https://sourceforge.net/projects/mimodd/). Loci in the mutant strains that differed from the N2 strain were examined. Based on the logic that the am326 mutation must be present in all backcrossed strains but absent in the N2 strain, we disregarded mutations present in some – but not all – backcrossed strains. We disregarded mutations present in the N2 strain (likely caused by genetic drift leading to differences between our lab N2 strain and the WS220.64 reference genome). This resulted in a list of 56 variants. Eleven variants, small insertions or deletions (‘indels’) in repetitive regions, were disregarded as likely sequencing errors, given that EMS mutagenesis typically creates base pair changes rather than indels. Each mutation was manually analyzed using the UCSC Genome Browser (UCSC Genomics Institute, Santa Cruz, CA, USA) to determine the genomic region affected. See Table S4 for detailed information on these variants. Multiple sequence alignment with human and mouse INSR and Drosophila InR was performed using COBALT (constraint-based multiple alignment tool) with C. elegans DAF-2 as the reference sequence (Papadopoulos and Agarwala, 2007).
Complementation assays
Adult male daf-2(am326) or daf-2(syb4952) animals were mated to hermaphrodites of a different genotype at 20°C (Table 1). F1 embryo cross-progeny were shifted to 25°C and examined for dauer formation based on morphological characteristics, as described above.
Time-series RNA-seq
Heat-sensitive sterile spe-9(hc88) mutant animals were synchronized by alkaline hypochlorite treatment and allowed to hatch overnight in S Basal medium (100 mM NaCl, 50 mM KH2PO4, 5 mg/ml cholesterol). Arrested L1 larvae were transferred to 10 cm NGM dishes seeded with 200 µl E. coli OP50 at a density of ∼1000 animals per dish and cultured at 25°C. One dish was harvested every 24 h by washing with M9 buffer, after which worms were lysed with TRIzol reagent (15596026, Invitrogen) and repeatedly freeze/thawed in liquid nitrogen. Total RNA was obtained by phenol chloroform extraction and purified according to the RNeasy Micro kit protocol (74004, QIAGEN). RNA quality was assessed using a 4200 TapeStation system (Agilent). Library preparation and sequencing were carried out by Novogene Co., Ltd. (Beijing, China).
RNAi
RNA interference was performed as described by Kamath et al. (2003). E. coli HT115 bacteria expressing either a control plasmid (L4440) or a plasmid encoding acn-1 dsRNA were obtained from the Ahringer library (Kamath et al., 2000). NGM dishes containing 50 µg/ml ampicillin (A-301-25, Gold Biotechnology) and 1 mM IPTG (isopropyl-β-D-thiogalactoside, I2481-C50, Gold Biotechnology) were poured and allowed to cool on the benchtop overnight. RNAi bacteria were grown overnight at 37°C in lysogeny broth (LB) medium [10 g/l NaCl, 10 g/l Bacto Tryptone (211701, Thermo Fisher Scientific), 5 g/l yeast extract (212750, Thermo Fisher Scientific), pH 7.0] supplemented with 50 µg/ml ampicillin; this starter culture was diluted 1:100 into LB media supplemented with 50 µg/ml ampicillin and grown for 6 h at 37°C. Then, 200 µl of this culture was seeded onto the NGM RNAi dishes. Dishes were allowed to dry on the benchtop at room temperature overnight, then stored at 4°C. Embryos were transferred to RNAi dishes and cultured at 20, 22 or 25°C until they either reached adulthood or formed dauer larvae.
CRISPR-mediated generation of the am314 mutation in the acn-1 locus
CRISPR-mediated modification of the acn-1 locus was performed using the Co-CRISPR methodology described by Arribere et al. (2014) and Kocsisova et al. (2018). Our goal was to introduce a change into the first exon of the endogenous acn-1 locus, thus reducing ACN-1 activity. We designed the acn-1 guide plasmids pANS2 and pANS4 with the sequences TTTGCTTCTCCTATTGCTTGTGG and TTTGATACTTCTCCTATTGCTTG, respectively, ligated into plasmid DR274 (MN3055). The following mixture was injected into the gonads of P0 wild-type hermaphrodites: 50 ng/µl Cas9-expressing PDD162 (gift from Mike Nonet, Washington University in St. Louis School of Medicine, MO, USA); 20 ng/µl dpy-10 guide plasmid pMN3153; 500 nM dpy-10(cn64) ssDNA repair template AFZF827; acn-1 guide plasmids pANS2 and pANS4, 40 ng/µl each; 600 nM ssDNA repair template TTTTCAGATGAAGTTTCATATACTTCTCCTATTGCTTGGCTAGCTGCCTGTTTGCCAGTATTCACTCAGGAAATCAAGCCAA. We selected F1 progeny displaying the Rol phenotype. acn-1(am314) WU1746 was identified using pyrosequencing to select F2 progeny. The mutation is a 42 bp substitution without a stop codon. Lifespan assays were performed on L4 hermaphrodites from this population as described above. The analyzed animals were a mixed population with respect to acn-1 genotype.
Determination of sod-3 gene expression levels by RT-qPCR
To measure relative gene expression, we collected whole-worm RNA, transcribed RNA into cDNA, and analyzed gene expression levels using RT-qPCR. To isolate whole-worm RNA, we picked approximately 400-500 unhatched embryos from dishes cultured in standard conditions onto 10 cm NGM dishes, which contained either 0 or 2.5 mM captopril and were seeded with 2 ml of E. coli OP50 bacteria. Worms were cultured at 20°C for 72 h and collected by washing four times with M9. Worms were compacted by centrifugation at 3000 rpm (1750 g) for 5 min, and excess liquid was removed by aspiration. Tubes containing the worms were frozen in liquid nitrogen and stored at −80°C. Worms were lysed by four successive freeze-thaw cycles in liquid nitrogen, then 500 µl Trizol and 100 µl chloroform were added to the lysed worm solution. The mixture was vortexed and then centrifuged at 12,000 g for 15 min at 4°C. The clear supernatant layer was extracted and mixed with 500 µl isopropanol, vortexed, and centrifuged at 12,000 g for 15 min at 4°C. The RNA pellet was washed with 1 ml of 70% ethanol and centrifuged as before. Excess ethanol was aspirated, and the pellet was dried in a 60°C heat block. The pellet was resuspended in 50 µl nuclease-free water, and the RNA concentration was measured by UV spectroscopy (NanoDrop One Microvolume UV-Vis Spectrophotometer, ND-ONE-W, Thermo Fisher Scientific). Reverse transcription was performed using the iScript cDNA Synthesis Kit (1708891, Bio-Rad) using 1 µg RNA (50 ng/µl). RT-qPCR was performed using iTaq Universal SYBR Green Supermix (1725121, Bio-Rad) using 200 ng cDNA. Relative fold change was calculated using the comparative CT method (Schmittgen and Livak, 2008); sod-3 mRNA level was normalized to the ama-1 control gene, and data were analyzed using two-tailed, unpaired Student's t-test.
Determination of sod-3 expression using a fluorescent reporter strain
Transgenic animals containing the sod-3p::gfp construct were cultured from the egg stage in medium containing 0 or 2.5 mM captopril for 3 days (adult day 1) or 7 days (adult day 5). To perform fluorescence microscopy, we prepared 3% agar slides, added a drop of levamisole (3 mM) (PHR1798, Sigma-Aldrich) to the agar, and added approximately ten worms per slide. As necessary, animals were oriented using an eye lash tool. Excess levamisole was removed, and a coverslip was applied and sealed with 3% agar. Brightfield and fluorescence images of animals were acquired using a Leica DMi8 inverted microscope using the same settings for control and drug-treated animals in each experiment. Fluorescence intensity of each worm was measured using Fiji (ImageJ, 1.53f51). The experiments were repeated at least three times independently, and data were analyzed using two-tailed, unpaired Student's t-test.
Acknowledgements
We thank the Caenorhabditis Genetics Center (CGC) at the University of Minnesota for providing nematode strains; the Washington University Genome Technology Access Center (GTAC) for whole-genome sequencing support; SunyBiotech for CRISPR/Cas9 genome editing; Michael Nonet for expertise and reagents; Heidi Tissenbaum, Gary Ruvkun, Tim Schedl, Abhinav Diwan, Heather True, Zachary Pincus, Michael Nonet and Roberta Faccio for useful discussion and expertise.
Footnotes
Author contributions
Conceptualization: B.M.E., F.P., A.S., M.M., S.K., H.F., F.-F.H., K.K.; Methodology: B.M.E., F.P., A.S., M.M., S.K., D.L.S., H.F., F.-F.H., K.K.; Validation: B.M.E., F.P., A.S., M.M., S.K., D.L.S., H.F., F.-F.H., K.K.; Formal analysis: B.M.E., F.P., X.A., S.C.W., I.G.A., P.H., Z.W., C.-H.C., A.S., M.M., S.K., D.L.S., H.F., F.-F.H.; Investigation: B.M.E., F.P., X.A., S.C.W., I.G.A., P.H., Z.W., C.-H.C., A.S., M.M., S.K., D.L.S., H.F., F.-F.H.; Resources: F.P., A.S., M.M., H.F., F.-F.H., K.K.; Data curation: M.M.; Writing - original draft: B.M.E.; Writing - review & editing: B.M.E., F.P., K.K.; Visualization: B.M.E., A.S.; Supervision: B.M.E., D.L.S., F.-F.H., K.K.; Project administration: B.M.E., K.K.; Funding acquisition: F.-F.H., K.K.
Funding
This work was supported by the National Institutes of Health [R01 AG02656106A1, R56 AG072169, R21 AG058037; R01 AG057748 to K.K.; GM103422 and P60-DK-20579 to F.-F.H.; 5T32GM007067-44 to B.M.E.], a Postdoctoral Fellow Seed of Independence Grant from the Department of Developmental Biology at Washington University School of Medicine in St. Louis (to F.P.), and the Irving Boime Graduate Student Fellowship (to B.M.E.). Deposited in PMC for release after 12 months.
Data availability
RNA-seq data have been deposited in Gene Expression Omnibus under accession number GSE254501.
Peer review history
The peer review history is available online at https://journals.biologists.com/dev/lookup/doi/10.1242/dev.202146.reviewer-comments.pdf
References
Competing interests
The authors declare no competing or financial interests.