A hallmark of aging is immunosenescence, a decline in immune functions, which appeared to be inevitable in living organisms, including Caenorhabditis elegans. Here, we show that genetic inhibition of the DAF-2/insulin/IGF-1 receptor drastically enhances immunocompetence in old age in C. elegans. We demonstrate that longevity-promoting DAF-16/FOXO and heat-shock transcription factor 1 (HSF-1) increase immunocompetence in old daf-2(−) animals. In contrast, p38 mitogen-activated protein kinase 1 (PMK-1), a key determinant of immunity, is only partially required for this rejuvenated immunity. The up-regulation of DAF-16/FOXO and HSF-1 decreases the expression of the zip-10/bZIP transcription factor, which in turn down-regulates INS-7, an agonistic insulin-like peptide, resulting in further reduction of insulin/IGF-1 signaling (IIS). Thus, reduced IIS prevents immune aging via the up-regulation of anti-aging transcription factors that modulate an endocrine insulin-like peptide through a feedforward mechanism. Because many functions of IIS are conserved across phyla, our study may lead to the development of strategies against immune aging in humans.
An increase in mortality after infection is a key feature of aging. This age-dependent decline in immunity (i.e., immunosenescence or immune aging) is nearly universal across phyla (Goronzy and Weyand, 2013; Müller et al., 2013; Fulop et al., 2018). Old mice and humans have increased proinflammatory cytokine levels, termed “inflamm-aging,” because of the chronic activation of innate immune systems (Baylis et al., 2013; Franceschi and Campisi, 2014). However, the complexity of the immune systems and cell types in vertebrates impedes the dissection of the molecular mechanisms underlying immunosenescence.
Caenorhabditis elegans is an excellent model for research on aging and immunity, phenomena which are closely related. Many long-lived C. elegans mutants display enhanced innate immunity at a young age (Garsin et al., 2003; Alper et al., 2010; Yunger et al., 2017; Tiku et al., 2018). Aged C. elegans display an increased susceptibility to pathogen infection, indicating that C. elegans can be used as a model of immune aging (Kurz and Tan, 2004; Laws et al., 2004; Youngman et al., 2011). However, the mechanisms by which longevity affects immune aging remain largely unknown.
Insulin/insulin-like growth factor 1 (IGF-1) signaling (IIS) plays important roles in aging and immunity in various species. In C. elegans, inhibition of IIS, such as mutations in the daf-2/insulin/IGF-1 receptor or its agonist, ins-7, increase lifespan and survival on pathogens at a young age (Kenyon et al., 1993; Garsin et al., 2003; Murphy et al., 2003; Evans et al., 2008a; Evans et al., 2008b; Kawli and Tan, 2008). In addition, mutations in daf-2 confer resistance to colonization of the gut by dietary bacteria, which contributes to extended survival of old worms (Podshivalova et al., 2017). Reduced IIS activates several transcription factors, such as DAF-16/FOXO, heat-shock transcription factor 1 (HSF-1), and SKN-1/NRF, which up-regulate the expression of various genes that contribute to longevity, stress resistance, and immunity (Kenyon, 2010; Murphy and Hu, 2013; Riera et al., 2016). Despite extensive research on the roles of IIS in aging and immunity, its role in the regulation of immunosenescence remains unknown.
In this study, we investigated whether longevity interventions affect the rate of immunosenescence in C. elegans. Among the various longevity regimens tested here, genetic inhibition of daf-2 substantially improved immunocompetence in old (day 9) worms. We found that temporal inhibition of daf-2 starting from middle age (day 4) rejuvenated immunity in day 8 adult animals. We then found that DAF-16/FOXO and HSF-1 increased immunocompetence in day 9 daf-2 mutants by regulating the expression of various target genes. In particular, DAF-16/FOXO and HSF-1 down-regulated an agonistic insulin-like peptide, INS-7, by decreasing the expression of a basic leucine zipper (bZIP) transcription factor, ZIP-10, which in turn reduced IIS further and enhanced immunity at day 9. Thus, daf-2 mutants appear to elude normal immunosenescence via a positive feedback feedforward endocrine circuit that consists of the DAF-16/FOXO, HSF-1, and ZIP-10 transcription factors, together with a DAF-2 agonist, INS-7.
Results and discussion
We examined whether the rates of immune aging were proportionally affected by genetic mutations that promote longevity. Similar to the WT counterparts (Fig. 1, A and B; Laws et al., 2004; Youngman et al., 2011), long-lived mutants such as sensory-defective osm-5, dietary restriction mimetic eat-2, mitochondrial respiration–impaired isp-1, and germline-deficient glp-1 mutants exhibited an age-dependent increase in susceptibility to Pseudomonas aeruginosa strain, PA14, infection (Fig. 1 A and Fig. S1, A–E; and Table S1).
Surprisingly, day 9 adult daf-2(e1370) (a hypomorphic mutation in insulin/IGF-1 receptor gene [hereafter daf-2(–) for simplicity]) mutants displayed substantially enhanced survival upon PA14 infection compared with day 1 adult daf-2(–) mutants and WT worms (Fig. 1, A–C). Similarly, day 9 daf-2(RNAi) worms also displayed enhanced PA14 resistance (Fig. 1 D), without defects in feeding rates (Fig. S1 F), which is different from daf-2(–) mutants that displayed defects in feeding (Fig. S1 G; also see Fig. S1 legends for discussion; Kenyon et al., 1993; Gems et al., 1998; Dillon et al., 2016; Podshivalova et al., 2017; Wu et al., 2019). Remarkably, daf-2(–) mutant and daf-2(RNAi) animals both maintained enhanced PA14 resistance at day 15 and day 30, respectively, compared with that of young (day 1) adult animals (Fig. 1, E and F). Day 9 daf-2(–) mutants also survived longer than young daf-2(–) worms after infection with PAO1, a moderately virulent P. aeruginosa strain (Fig. S1, I and J). We next asked whether temporal knockdown of daf-2 was sufficient to suppress immune aging. We found that knockdown of daf-2 during development was sufficient for increasing immunocompetence against PA14 at day 9 adulthood (Fig. S1 K; see also Fig. S1 legends for discussion). Furthermore, temporal genetic inhibition of daf-2 (Fig. 1 G) from middle age (day 4) was sufficient to enhance immunocompetence against PA14 at day 8 adulthood (Fig. 1 H). In contrast, the temporal daf-2 RNAi from day 4 did not prevent the age-dependent decline in resistance against oxidative stress or heat stress (Fig. S1, P and Q). Together, these data suggest that genetic inhibition of DAF-2 confers sustained resistance against pathogenic bacteria, P. aeruginosa, during aging.
WT animals display an age-dependent increase in the intestinal accumulation of PA14 (Youngman et al., 2011). daf-2 mutants displayed substantially reduced PA14 levels in the intestine in both day 1 (Evans et al., 2008a) and day 9 adults (Fig. 1 I). In contrast, we observed an age-dependent increase in the intestinal accumulation of PA14 in daf-2 RNAi–treated animals (Fig. 1 J and Fig. S1 R). As both day 9 daf-2 mutant and daf-2(RNAi) adults survived longer on PA14 than did day 1 adults (Fig. 1, A and C–F), genetic inhibition of daf-2 appears to maintain immunocompetence in relatively old (day 9) age, at least in part, independently of PA14 intake and/or clearance.
We sought to identify downstream factors that mediate the delayed immune aging in daf-2 mutants. DAF-16/FOXO, HSF-1, and SKN-1/NRF are required for increased pathogen resistance in young daf-2 mutants (Garsin et al., 2003; Singh and Aballay, 2006; Miyata et al., 2008; Papp et al., 2012); however, it remains unknown whether these transcription factors play a role in daf-2(–)–mediated delay in immune aging. Importantly, we found that daf-16 mutations fully suppressed the PA14 resistance of daf-2 mutants at all tested ages and unveiled immune aging in daf-2 mutants (Fig. 2 A; and Fig. S2, A and B; and Table S2). hsf-1 RNAi also largely suppressed the enhanced immunocompetence of day 6 and day 9 daf-2(–) adults (Fig. 2 B; and Fig. S2, C and D). In contrast, RNAi targeting skn-1 did not affect the enhanced immunocompetence of daf-2 mutants (Fig. 2 C; and Fig. S2, E and F). In addition, PMK-1/p38 MAPK, a major immune-regulatory factor, whose reduction is the critical determinant of immune aging in WT animals (Youngman et al., 2011; Kim et al., 2002) and suppresses immunity in young daf-2(–) mutants (Troemel et al., 2006), was not required for the enhanced immunity of day 9 daf-2(–) adults (Fig. 2 D; and Fig. S2, I and J). Mutations in each of nsy-1/MAPKKK and sek-1/MAPKK, which encode upstream kinases of PMK-1 (Irazoqui and Ausubel, 2010; Kim and Ewbank, 2018), were only partially required for the enhanced immunity of day 9 daf-2(–) adults (Fig. 2, E and F; and Fig. S2, K–N). Together, these data suggest that DAF-16/FOXO and HSF-1 are mainly responsible for drastically delayed immune aging in daf-2 mutants.
We then investigated which transcriptional changes were associated with and responsible for the enhanced immunocompetence observed in aged daf-2 mutants. We conducted an mRNA sequencing (RNA-seq) analysis using day 1 and day 9 daf-2(–) and WT adults. Hierarchical clustering indicated a clear separation of transcriptomes in accordance with age and genotype (Fig. 3 A), and a principal-component analysis revealed that the effect of age was larger than that of genotype (Fig. S3 A). Subsequently, we found that the age-dependent gene expression changes detected in WT worms displayed a positive correlation with those observed in daf-2(–) mutants (Fig. 3 B). We then showed that the mRNA levels of four representative DAF-16 and HSF-1 common targets, mtl-1, lys-7, dod-6, and hsp-16.1/11 (Hesp et al., 2015; Hsu et al., 2003; Murphy et al., 2003; Barsyte et al., 2001; Sural et al., 2019), increased age dependently in daf-2(–) animals under Escherichia coli OP50-fed and PA14-exposed conditions (Fig. 3, C–F; and Fig. S3, B and D–G). In contrast, the mRNA levels of sod-3, a DAF-16/FOXO target, or any of three selected PMK-1 targets, T24B8.5, C17H12.8, and K08D8.5 (Troemel et al., 2006; Shivers et al., 2010; Youngman et al., 2011; Wu et al., 2019; Hsu et al., 2003; Murphy et al., 2003), did not increase with age (Fig. 3, G–I; and Fig. S3 C and H–L). These data raise the possibility that common targets of DAF-16/FOXO and HSF-1 contribute to age-dependent increases in immunity observed in daf-2(–) adults.
We then asked which specific genes, whose expression was changed in an age-dependent manner, were responsible for the delayed immune aging of daf-2(–) mutants. Aged daf-2(–) mutants displayed increased immunity, which is the opposite of the impaired immunity observed in aged WT worms. Therefore, we focused our RNA-seq analysis on genes whose expression levels were altered in an opposite manner between WT and daf-2(–) worms during aging (Table S3). We found that the mRNA levels of 72 genes increased with age in WT worms but decreased with age in daf-2(–) animals (Fig. 3 B, red dots; fold change >1.5; Benjamini and Hochberg [BH]–adjusted P value < 0.1 with 19,327 genes). Conversely, the expression of 14 genes decreased with age in WT worms and increased in daf-2(–) animals (Fig. 3 B, blue dots; fold change >1.5; BH-adjusted P value < 0.1). A gene set enrichment analysis (GSEA) revealed that “immune/defense response” genes were significantly down-regulated during aging in daf-2(–) worms but were marginally down-regulated in WT animals (Fig. 3, J and K; q value [BH-adjusted P value] < 0.1); thus, these genes may underlie the distinct immunity phenotypes observed between aged WT and aged daf-2(–) animals.
Next, we performed immune aging assays using mutants that carried mutations in each of the 13 genes that were selected among the 72 candidate genes (Fig. 4 A), whose expression was specifically increased during WT aging and decreased during daf-2(–) aging (Fig. 3 B and Table S3 A). We found that the pmp-1/peroxisomal ABC transporter ABCD3, ins-7/an insulin-like peptide, zip-10/bZIP transcription factor, valv-1/cysteine-rich intestinal protein, and lips-10/lipase loss-of-function mutants displayed enhanced immunocompetence at day 6 of adulthood compared with their WT counterparts (Fig. 4 A; see Fig. S3 legends for further discussion). Among them, ins-7 (Fig. 4, B and C), an agonist of DAF-2 that acts as a positive feedback regulator of IIS (Murphy et al., 2003; Murphy et al., 2007), was negatively regulated by both DAF-16/FOXO (Fig. 4 D; Murphy et al., 2003; Murphy et al., 2007; Lee et al., 2009) and HSF-1 (Fig. 4 E); we confirmed this result by using quantitative reverse transcription polymerase chain reaction (qRT-PCR) assays (Fig. 4, F and G). Notably, mutations in ins-7 enhanced immunocompetence in day 6 and day 9 adults (Fig. 4 H; and Fig. S3, N, P, and Q; and Table S4) as well as in day 1 worms (Fig. S3, M and O; Evans et al., 2008b; Kawli and Tan, 2008). These data suggest that the up-regulation of DAF-16/FOXO and HSF-1 decreases ins-7 expression and enhances immunocompetence during aging in daf-2 mutants, containing a hypomorphic e1370 allele.
We then functionally characterized the role of the ZIP-10/bZIP transcription factor, a pathogen-responsive gene (Shapira et al., 2006), whose expression pattern was similar to that of ins-7 in aged WT versus daf-2(–) worms (Fig. 3 B). Like ins-7, zip-10 expression was increased by daf-16 mutations and hsf-1 RNAi in daf-2(–) mutants (Fig. 5, A and B). We found that zip-10 mutations suppressed the age-dependent increase in ins-7p::gfp and ins-7 mRNA expression (Fig. 5, C–E), suggesting that ZIP-10 induces ins-7. We also showed that zip-10 mutation increased the PA14 resistance at day 9 (Fig. 5 F), but not at day 1, of adulthood (Fig. S3 R). In addition, hsf-1 RNAi or daf-16 RNAi suppressed the extended survival of day 9 daf-2(–); zip-10(–) animals on PA14 (Fig. S3, S and T). We also found that mutations in isy-1/ISY splicing factor homolog, which cause up-regulation of zip-10 (Jiang et al., 2018), significantly reduced immunocompetence in day 9, but not in day 1, adults (Fig. 5 G and Fig. S3 Y). Moreover, we found that the isy-1 mutation reduced enhanced immunocompetence in day 9 daf-2(–) adults (Fig. 5 H; and Fig. S3, Z–A′). Altogether, these data suggest that a hypomorphic daf-2(e1370) mutation enhances immunocompetence in old age by down-regulating ZIP-10, subsequently decreasing INS-7 expression, which underlies immune aging in WT worms (Fig. 5 I).
Despite the close association between aging and immunity, it remains largely unexplored whether and how longevity interventions affect immune aging. Here, we determined a relationship between longevity and immune aging using C. elegans. We found that reduction of IIS greatly delayed and even rejuvenated immunity at some point during aging. This retarded immune aging conferred by inhibition of the DAF-2/insulin/IGF-1 receptor was dependent on DAF-16/FOXO and HSF-1. We also showed that daf-2 mutations suppressed the age-dependent up-regulation of the ZIP-10/bZIP transcription factor, which led to the down-regulation of INS-7, an agonist of DAF-2 (Murphy et al., 2003). Down-regulation of INS-7, in turn, appears to contribute to a feedforward circuit to enhance immunity further in daf-2 mutants during aging. We propose that daf-2 mutations can uncouple aging and the age-dependent declines in immune function by increasing the activities of DAF-16/FOXO and HSF-1, but not those of p38 MAPK signaling, via a feedback amplification mechanism during aging.
Previous studies have shown that IIS regulates innate immunity in various species. FOXO, a conserved transcription factor that acts downstream of IIS, promotes innate immunity directly in Drosophilamelanogaster and mammalian cells (Becker et al., 2010). Reduced IGF-1 signaling promotes the renewal of hematopoietic stem cells, which are crucial for the generation of cells important for proper immune functions (Cheng et al., 2014). In addition, inhibition of target of rapamycin, which intersects IIS, increases vaccination response in elderly people and enhances the self-renewal of hematopoietic stem cells in mice (Chen et al., 2009; Mannick et al., 2014). In the current study, we identified systemic positive feedback mechanisms by which reduced IIS rejuvenates immunity in C. elegans. Because the functions of IIS in various physiological processes, including aging and immunity, are conserved across phyla, the findings reported here for C. elegans may lead to the development of therapeutic strategies against immunosenescence in humans.
Materials and methods
C. elegans strains and maintenance
All strains were maintained as previously described (Stiernagle, 2006). Some strains were obtained from the Caenorhabditis Genetics Center, which is funded by the National Institutes of Health National Center for Resources (p40 OD010440), or the National Bio-Resource Project, Japan. Strains used in this study are as follows: N2 WT, CF1041 daf-2(e1370) III, CF2553 osm-5(p813) X, CF1903 glp-1(e2141) III, CF2172 isp-1(qm150) IV, IJ173 eat-2(ad1116) II obtained by outcrossing DA1116 four times to Lee-laboratory N2, IJ130 pmk-1(km25) IV obtained by outcrossing KU25 four times to Lee-laboratory N2, IJ1147 daf-2(e1370) III; pmk-1(km25) IV, IJ128 sek-1(km4) X obtained by outcrossing KU4 four times to Lee-laboratory N2, IJ1299 daf-2(e1370) III; sek-1(km4) X, IJ131 nsy-1(ok593) II obtained by outcrossing VC390 four times to Lee-laboratory N2, IJ1298 nsy-1(ok593) II; daf-2(e1370) III, CF1042 daf-16(mu86) I, CF1085 daf-16(mu86) I; daf-2(e1370) III, IJ1733 ins-7(tm1907) obtained by outcrossing FX01907 four times to Lee-laboratory N2, IJ1010 ins-7(tm2001) obtained by outcrossing QL126 four times to Lee-laboratory N2, ZC1436 yxIs13[ins-7p::gfp(70 ng/ul); unc-122p::dsred] a gift from Yun Zhang laboratory, IJ1871 zip-10(ok3462) obtained by outcrossing RB2499 five times to Lee-laboratory N2, IJ1919 isy-1(dma50) obtained by outcrossing DMS175, a gift from Dengke Ma laboratory. IJ1936 daf-2(e1370) III; isy-1(dma50) V, RB908 pmp-1(ok773), RB1262 cpr-1(ok1344), RB1415 capt-3(ok1612), RB2452 acs-14(ok3391), RB2473 cpr-4(ok3413), RB2499 zip-10(ok3462), RB686 fmo-3(ok354), FX06726 valv-1(tm6726), FX07601 lips-10(tm7601), FX01907 ins-7(tm1907), CF2078 sgk-1(ok538), VC2249 dod-19(ok2679), and VC2200 grl-22(ok2704). IJ1875 daf-2(1370) III; zip-10(ok3462) V. IJ2066 isy-1(dma50) V; yxIs13[ins-7p::gfp(70 ng/ul); unc-122p::dsred]. All strains except daf-2(e1370); sek-1(km4) were maintained on nematode growth media (NGM) plates seeded with E. coli OP50 strain as a food source at 20°C. daf-2(e1370); sek-1(km4) was maintained and synchronized at 15°C because of a semi-sterility phenotype at 20°C.
Immune aging assays
Immune aging assays were performed as described previously (Youngman et al., 2011) with minor modifications. Gravid adult worms were allowed to lay eggs overnight to synchronize progeny on NGM plates seeded with E. coli OP50 or RNAi bacteria. The progeny that reached L4 or the young prefertile adult stage were then transferred onto NGM plates treated with 50 µM 5-fluoro-2’-deoxyuridine (FUDR; Sigma-Aldrich). Worms were maintained on these plates at 20°C until they were used for experiments at day 1, 3, 6, 9, 12, 15, or 30 of adulthood. For the majority of immune aging assays, day 9 adult worms were used as aged worms to compare the survival of WT and mutant worms on PA14; after day 9 of adulthood, WT worms died very soon after treatment with PA14, and this caused difficulty in comparison analysis with mutants. For the immune aging assays using temperature-sensitive glp-1(e2141) mutants, WT and glp-1(e2141) worms were synchronized by allowing gravid adults to lay eggs on E. coli OP50-seeded plates for 12 h at 20°C and the progeny developed at 25°C until the worms reached L4 stage. For the immune aging assays using daf-2(e1370); sek-1(km4) mutants, which were semi-sterile at 20°C, worms were first cultured at 15°C to obtain a sufficient number of animals, transferred to 20°C, and allowed to age after they reached adulthood. Worms were transferred every other day until WT worms stopped producing progeny. Contaminated or E. coli–depleted plates were discarded. P. aeruginosa standard slow-killing assays were performed as previously described (Jeong et al., 2017). Briefly, 5 µl of overnight P. aeruginosa (PA14 or PAO1) liquid culture was seeded onto the center of 0.35% peptone NGM solid media. Plates were incubated at 37°C for 24 h and subsequently at room temperature for over 8 h before use. Worms of different ages simultaneously infected with P. aeruginosa on plates containing 50 µM FUDR at 25°C. After infection (time = 0 on survival curves), worms were scored more than once a day by gently prodding with a platinum wire to distinguish between live and dead worms. To screen 14 mutants that were chosen from RNA-seq analysis (Fig. 4 A), worms were synchronized on OP50-seeded NGM plates and transferred to 50 µM FUDR–treated NGM plates when they became young prefertile adults. After 6 d, the worms were infected with PA14. After 48 h of infection, live and dead worms were scored.
Temporal daf-2 RNAi experiments
WT worms were synchronized on control RNAi–treated plates and transferred onto new control RNAi plates when they became prefertile adults. Day 4 adult worms were transferred to control or daf-2 RNAi bacteria–containing plates, respectively, and maintained at 20°C for 4 d. For pathogen resistance assays, day 4 or day 8 adult worms were moved to plates with PA14. For oxidative stress resistance assays, day 4 or day 8 adult worms were transferred to OP50-seeded NGM plates with 7.5 mM tert-butyl hydroperoxide (Sigma-Aldrich) solution. For heat stress resistance assays, day 4 or day 8 adult worms on OP50-seeded plates were placed in a 35°C incubator. The number of live worms was scored over time, and worms that did not respond to a gentle touch with a platinum wire were counted as dead. OP50-seeded NGM plates were used for the heat and oxidative stress resistance assays with temporal daf-2 RNAi–treated animals to avoid continuous RNAi induction and match the RNAi conditions used for survival assays on PA14. For survival assays on PA14 using worms treated with daf-2 RNAi during development, WT worms were synchronized on daf-2 RNAi bacteria–seeded plates and transferred onto fresh dcr-1 RNAi plates at the prefertile young adult stage (Dillin et al., 2002; Durieux et al., 2011). Day 1 or day 9 adult worms were then exposed to PA14. Worms that reached young prefertile adulthood were treated with 50 µM FUDR until finishing the survival assays.
Preparation of dauer-recovered worms
Dauer pheromone was used to induce dauer formation (Neal et al., 2013). Diluted crude dauer pheromone (6 µl crude dauer pheromone + 94 µl autoclaved water) was placed on each 35-mm plate. 3 ml dauer agar solution (0.3 g NaCl, 2 g Noble agar, 100 µl 1 M CaCl2, 100 µl 1 M MgSO4, and 2.5 ml 1 M KPO4 per 100 ml media) sterilized by autoclave was cooled down to 50–60°C and distributed into each of the 35-mm plates containing 100 µl diluted crude dauer pheromone. Because live bacteria interfere with the formation of dauer, heat-killed OP50 was used as a food source. Concentrated OP50 (10×) was resuspended by adding S-basal buffer (5.85 g NaCl, 1 g K2 HPO4, 6 g KH2PO4, and 1 ml cholesterol [5 mg/ml in ethanol] per 1 liter buffer). OP50 in S-basal buffer was heated at 95°C for 30 min with brief vortexing every 10 min. The heat-killed OP50 was cooled down at 4°C at least 30 min. WT worms were allowed to lay eggs at 25°C. After 72 h, worms that became dauer larvae were transferred onto OP50-seeded NGM plates at 20°C for recovery from dauer. When the recovered worms reached the L4 or young adult stage, they were transferred onto fresh OP50-seeded NGM plates containing 50 µM FUDR. Day 1 or day 9 adult worms were then infected with PA14 for survival assays.
Each RNAi clone expressing HT115 bacteria was cultured in Luria broth containing 50 µg/ml ampicillin (USB) overnight at 37°C. 100 μl of RNAi culture was seeded onto NGM containing 50 µg/ml ampicillin and incubated overnight at 37°C. 1 mM isopropylthiogalactoside (Gold Biotechnology) was added and incubated at room temperature for over 24 h before use.
Intestinal PA14-GFP accumulation assays
Intestinal PA14-GFP accumulation assays were performed as previously described (Jeong et al., 2017; Kim et al., 2002), with minor modifications. Day 1 and 9 adult WT and daf-2(e1370) worms were infected with PA14 that expresses GFP (PA14-GFP) at 25°C for 24 h and subsequently imaged using a Zeiss Axio Scope A1 compound microscope. For control and daf-2 RNAi treatments, worms were infected with PA14-GFP for 100 h.
Measurements of pharyngeal pumping
Pharyngeal pumping rates were measured as previously described (Cao and Aballay, 2016). PA14 bacterial lawns were prepared as described above. Aged animals were transferred onto PA14 lawn and incubated at 25°C. The number of contractions of the terminal bulb was counted for 15 s. The pumping rates for 10 adults were measured for each condition.
Worms were prepared at indicated ages and transferred on a 2% agarose pad with 100 mM sodium azide (Daejung) or 2 mM levamisole (tetramisole; Sigma-Aldrich). Images of worms were captured using an AxioCam HRc charge-coupled device digital camera (Zeiss) with a Zeiss Axio Scope A1 compound microscope. Carl Zeiss Microscopy AxioVision SE64 (Rel.4.9.1 SP2) software was used for imaging. ImageJ (https://imagej.nih.gov/ij/) was used for the quantification of fluorescent images. The fluorescence intensity of transgenic worms was obtained by subtracting that of WT worms to eliminate background fluorescence values.
RNA extraction and qRT-PCR
RNA extraction and qRT-PCR were performed as previously described, with minor modifications (Son et al., 2017). When worms that were fed with OP50 or RNAi bacteria reached the L4 or young prefertile adult stage, 50 µM FUDR was supplemented. After 1 or 9 d, worms were harvested by washing with M9 buffer two or three times to remove residual bacteria. Plates with bacterial or fungal contamination (or starved worms because of depletion of food) were discarded. If OP50 was almost depleted, then soaked freeze-dried OP50 (LabTIE) was added onto the plates to prevent starvation. To measure RNAi efficiency, gravid adults were allowed to lay eggs on RNAi bacteria–seeded NGM plates containing 100 µg/ml ampicillin. The worms were then cultured until reaching day 1 of adulthood and harvested by washing with M9 buffer twice. For temporal RNAi experiments, worms at the L4 or prefertile young adult stage were treated with 50 µM FUDR. After culturing for 4 d, control RNAi–treated worms were harvested with M9 buffer or transferred onto new control or daf-2 RNAi bacteria–seeded plates for an additional 4 d of culture before harvesting with M9 buffer for qRT-PCR analysis. To measure the knockdown efficiency of daf-2 RNAi during development, worms were grown on control or daf-2 RNAi bacteria until reaching the prefertile young adult stage and transferred to control or dcr-1 RNAi bacteria–seeded plates. Worms at day 1 or day 9 of adulthood were harvested by washing twice with M9 buffer. Total RNA in the animals was extracted using RNAiso plus (Takara), and cDNA was synthesized using the ImProm-II Reverse transcription kit (Promega). qRT-PCR was performed using the StepOne Real-Time PCR System (Applied Biosystems) with SYBR green dye (Applied Biosystems). Comparative CT method was used for quantitative PCR analysis. Two technical repeats were averaged per biological sample and analyzed for all datasets. For all biological datasets, ama-1 (an RNA polymerase II large subunit) and pmp-3 (a peroximal membrane protein) were used as reference genes for normalization (Hoogewijs et al., 2008; Taki and Zhang, 2013). Outliers were excluded to calculate statistics by using QuickCalcs of GraphPad (https://www.graphpad.com/quickcalcs/grubbs1/). Sequences of primers used for qRT-PCR are as follows: ama-1, 5′-TGGAACTCTGGAGTCACACC-3′ and 5′-CATCCTCCTTCATTGAACGG-3′; pmp-3, 5′-GTTCCCGTGTTCATCACTCAT-3′ and 5′-ACACCGTCGAGAAGCTGTAGA-3′; T24B8.5, 5′-TGTTAGACAATGCCATGATGAA-3′ and 5′-ATTGGCTGTGCAGTTGTACC-3′; C17H12.8, 5′-GAACAATAGTGTCAAGCCGATCTGC-3′ and 5′-TTCTGAATGATGAATGCATGTTTAC-3′; K08D8.5, 5′-CCATACATTTTCACGTCCCCAC-3′ and 5′-GGATATTTTGACCAGACAACCTTG-3′; mtl-1, 5′-GACTGCTGAAATTAAGAAATCATG-3′ and 5′-GTCTCCACTGCATTCACATTTGTC-3′; hsp-16.1/11, 5′-CTCATGAGAGATATGGCTCAG-3′ and 5′-CATTGTTAACAATCTCAGAAG-3′; sod-3, 5′-CTATCTTCTGGACCAACTTGG-3′ and 5′-GCAAGTTATCCAGGGAACCG-3′; ins-7, 5′-GTGGAAGAAGAATACATTCG-3′ and 5′-CTTCAGTATTCGATTCGCATG-3′; zip-10, 5′-TCGAGATGCTCTTCAACTG-3′ and 5′-CTAACTGCTTGCCGGAG-3′; dod-6, 5′-GTTCTCCCTCAAGACCGTCGC-3′ and 5′-GACCATCTTCAGCATCAGCGC-3′; lys-7, 5′-GTCTCCAGAGCCAGACAATCCGG-3′ and 5′-CGGTCGTGATCTGATTCCAGTCG-3′; daf-2 3′ UTR, 5′-CAACAGCCGTGACATTTTCAACG-3′ and 5′-GGAGATAATTTATAATTTAAGAGGC-3′; dcr-1, 5′-CCAGGTGGAACTTTTGGATAAAG-3′ and 5′-GAACTCCATATTCTTTCAGCAGTAG-3′; hsf-1, 5′-CATTTAGAGTTTAGTCATCCG-3′ and 5′-GTTCTTGCCGATTGCTTTCTC-3′; and skn-1, 5′-CGGAGATGTCATTAAGCGAG-3′ and 5′-GCAACCTTGTTCTTTCCGC-3′.
mRNA library preparation for RNA-seq analysis and data acquisition
Synchronized WT and daf-2(e1370) animals were treated with 50 µM FUDR when worms reached the L4 or young adult stage on OP50. At indicated ages (days 1 and 9 of adulthood), worms were harvested with M9 buffer, washed two or three times, and frozen in liquid nitrogen. Three biological repeats of the samples were used for all conditions. For RNA-seq analysis using hsf-1 RNAi, daf-2(e1370) mutants were synchronized on control or hsf-1 RNAi bacterial plates. Day 1 adult daf-2(e1370) animals on control or hsf-1 RNAi plates were harvested with M9 buffer. Two biological repeats of samples were used. RNA extraction for RNA-seq samples was performed as described above (see RNA extraction section). TruSeq (unstranded) mRNA libraries (Illumina) were constructed and paired-end sequencing of Illumina HiSeq2500 was performed by Macrogen.
Sequencing pairs were aligned to the C. elegans genome WBcel235 (ce11) and Ensembl transcriptome (release 97) using STAR (v.2.7.0e). Aligned pairs on genes or transcripts were quantified by using RSEM (v.1.3.1). Detailed parameters of alignment and quantification were calculated by following guidelines of ENCODE long RNA-seq processing pipeline. Differentially expressed genes (fold change >1.5 and adjusted P value < 0.1) were identified by using DESeq2 (v.1.22.2). Wald test P values are adjusted for multiple testing using the procedure of BH. Gene Ontology terms generally changed in a certain comparison were identified by GSEA (v.4.0.1) using read counts of all expressed genes with a parameter “Permutation type: gene_set.” Terms with q value (BH-adjusted P value) smaller than 0.1 were considered as significantly changed. R (v.3.6.1, http://www.r-project.org) was used for plotting results. Raw and processed data are available in Gene Expression Omnibus (accession no. GSE137861).
For quantification of imaging, qRT-PCR, and normalized survival on PA14, P values were calculated using a two-tailed Student’s t test. For semiquantification of PA14-GFP accumulation, P values were calculated using the χ2 test. Statistical analysis of survival data was performed using OASIS (http://sbi.postech.ac.kr/oasis), which calculates P values using the log-rank (Mantel–Cox method) test (Yang et al., 2011). All PA14 survival assays were performed at least twice independently. The mean survival on PA14 was calculated by pooling data from different experimental sets. For normalization of the survival screen on PA14 using mutants (Fig. 4 A), the proportion of live mutant worms were normalized by comparing the proportion of live day 6 WT as follows: percentage change of WT = 100% × [proportion of live mutants − proportion of live WT]/proportion of live WT.
Online supplemental material
Fig. S1 shows age-dependent changes in the survival of longevity mutants on PA14, and pharyngeal pumping rates and PA14 accumulation in animals with genetic inhibition of daf-2. Fig. S2 shows the role of DAF-16, HSF-1, SKN-1, PMK-1, NSY-1, and SEK-1 in the enhanced pathogen resistance of daf-2 mutants at different ages. Fig. S3 shows the role of ins-7, zip-10, and isy-1 in immune aging. Table S1 shows statistical analysis and additional repeats of immune aging and survival assays with longevity mutants or RNAi. Table S2 shows statistical analysis and additional repeats of immune aging assays (related to Figs. 2 and S2). Table S3 lists genes whose expression was oppositely regulated with age between WT and daf-2(−) animals. Table S4 shows statistical analysis and additional repeats of immune aging assays (related to Figs. S3, 4, and 5).
We thank Drs. Dennis H. Kim (Harvard Medical School, Boston, MA), You-Hee Cho (CHA University, Gyeonggi-do, South Korea), Emily Troemel (University of California, San Diego, San Diego, CA), Yun Zhang (Harvard University, Cambridge, MA), Dengke Ma (University of California, San Francisco, San Francisco, CA), and the Caenorhabditis Genetics Center for providing some bacteria and C. elegans strains. We also thank all Lee laboratory members and Dr. Kyuhyung Kim (Daegu Gyeongbuk Institute of Science and Technology, Daegu, South Korea) for providing dauer pheromone and discussion.
This study is supported by the Korean Government (Ministry of Science and Information and Communications Technology) through the National Research Foundation of Korea (grant NRF-2019R1A3B2067745 to S.-J.V. Lee).
C.T. Murphy is the Director of the Glenn Center for Aging Research at Princeton University. The authors declare no competing financial interests.
Author contributions: Y. Lee contributed to designing and performing the majority of experiments, data analysis, and writing the manuscript; Y. Jung contributed to survival assay experiments, qRT-PCR, fluorescence imaging, data analysis, figure preparation, and writing the manuscript; D.-E. Jeong, W. Hwang, H.-E.H. Park, and S. Kwon contributed to survival assay experiments; S. Ham contributed to analysis of RNA-seq data and writing the manuscript; J.M. Ashraf and C.T. Murphy contributed to transcriptome analysis; and S.-J.V. Lee contributed to designing all experiments, data analysis, and writing the manuscript.
A preprint of this paper was posted in bioRxiv on January 1, 2020.
Y. Lee and Y. Jung contributed equally to this paper.
D.-E. Jeong’s present address is Department of Pathology, Stanford University School of Medicine, Stanford, CA.
W. Hwang’s present address is SK Biopharmaceuticals, Gyeonggi-do, Korea.