The Caenorhabditis elegans Germ Line Regulates Distinct Signaling Pathways to Control Lifespan and Innate Immunity*

The relationship between the mechanisms that control an organism's lifespan and its ability to respond to environmental challenges are poorly understood. In Caenorhabditis elegans, an insulin-like signaling pathway modulates lifespan and the innate immune response to bacterial pathogens via a common mechanism involving transcriptional regulation by the DAF-16/FOXO transcription factor. The C. elegans germ line also modulates lifespan in a daf-16-dependent manner. Here, we show that the germ line controls the innate immune response of C. elegans somatic cells to two different Gram-negative bacteria. In contrast to the insulin-like signaling pathway, the germ line acts via distinct signaling pathways to control lifespan and innate immunity. Under standard nematode culture conditions, the germ line regulates innate immunity in parallel to a known p38 MAPK signaling pathway, via a daf-16-independent pathway. Our findings indicate that a complex regulatory network integrates inputs from insulin-like signaling, p38 MAPK signaling, and germ line stem cells to control innate immunity in C. elegans. We also confirm that innate immunity and lifespan in C. elegans are distinct processes, as nonoverlapping regulatory networks control survival in the presence of pathogenic and nonpathogenic bacteria. Finally, we demonstrate that the p38 MAPK pathway in C. elegans is activated to a similar extent by both pathogenic and nonpathogenic bacteria, suggesting that both can induce the nematode innate immune response.

Here, we show that the C. elegans germ line controls nematode resistance to two different Gram-negative bacterial pathogens, Pseudomonas aeruginosa and Serratia marcescens. However, unlike the case with daf-2, the effects of the germ line on lifespan and pathogen resistance are mediated by separate signaling pathways. While the FOXO family transcription factor DAF-16 is necessary for germ line-induced changes in lifespan (1,2), daf-16 does not affect germ line-induced changes in innate immunity under standard nematode culture conditions. Our findings also indicate that the germ line acts in parallel to a p38 MAPK 2 pathway to induce changes in innate immunity, suggesting that there are additional regulatory components of C. elegans host defense. Surprisingly, the extent of p38 MAPK phosphorylation was not different in nematodes exposed to either nonpathogenic or pathogenic bacteria, suggesting that either both bacteria induce the innate immune response to a similar extent or that the p38 pathway in C. elegans may act permissively rather than actively in controlling the nematode innate immune response.
Antimicrobial gene expression was monitored by quantitative PCR as described previously (45). In these experiments, eggs from either wild-type (N2), glp-1, or vhp-1 mutant animals were collected and synchronized by bleaching (37), allowed to hatch at 15°C, and then shifted 12 h later as young L1s to 25°C. Nematodes were collected for RNA preparation as young adults. For the P. aeruginosa exposures, nematodes were shifted from E. coli to P. aeruginosa 12 h prior to nematode collection.
To monitor phosphorylation of the p38 MAPK PMK-1, eggs from wild-type, glp-1 mutant, or glp-4 mutant animals were collected and synchronized by bleaching (37), allowed to hatch, and were then grown and exposed as described above for the quantitative PCR experiments. Mixed stage nematodes were collected for the wild-type versus nsy-1 mutant nematode comparison. For the arsenic exposure, mixed stage nematodes were exposed to 5 mM sodium arsenite for 30 min as described in Ref. 33. Nematodes were collected by centrifugation, and pellets were then rapidly frozen in liquid nitrogen. Protein was isolated, and Western blot analysis was performed as described (33,46). A BCA protein assay (Thermo Scientific, Rockford, IL) was then performed to quantify the protein in the lysate. Each sample was analyzed by Western blotting using three different antisera: anti-phospho-p38 (Thr 180 /Tyr 182 ) antibody (Cell Signaling Technology), rabbit polyclonal antiserum specific to PMK-1 protein (gift of Kuni Matsumoto), and anti-␤-tubulin monoclonal (E7, Developmental Studies Hybridoma Bank, Iowa City, IA). Western blots were developed using ECL Plus (GE Healthcare), and the STORM 860 Molecular Imager (Amersham Biosciences) was used to capture images. Image-Quant software (GE Healthcare) was used to quantify bands on the Western blots. PMK-1 and phosopho-PMK-1 quantities were normalized using the anti-␤-tubulin blots to control for protein concentration in the samples.

RESULTS
To examine the role of the C. elegans germ line in innate immunity, we evaluated survival of nematodes harboring muta-FIGURE 1. Mutations that prevent germ line proliferation in C. elegans increase resistance to pathogenic bacteria. Depicted are representative survival plots for the indicated C. elegans wild-type or mutant strains, which were exposed to either P. aeruginosa (A-D) or S. marcescens (E-G). Survival studies were initiated with nematodes in the late L4 larval stage. The tested alleles and further statistical data (medians, n, and p values) are listed in supplemental Table 1. The glp-1(e2124) allele was used in all experiments except for B and F, where the glp-1(q224) allele was used. mes-1 F refers to fertile nematodes; mes-1 S refers to sterile nematodes. Some wild-type (w.t.) curves in individual panels are from the same experiment as in other panels. All survival studies were carried out at least three times; the data is representative of these multiple experiments. tions in any of several genes required for germ line proliferation. These mutations are known to increase nematode lifespan in the presence of E. coli (1) (the standard experimental food source, considered nonpathogenic to C. elegans). As observed previously (47), mutation of glp-4, a gene required for germ line proliferation (29), enhanced nematode survival in the presence of the pathogen Pseudomonas aeruginosa ( Fig. 1A; alleles used and statistical information for all survival plots are presented in supplemental Table 1). To verify that this effect is due to loss of the germ line and not due to some other potential function of glp-4, we demonstrated that two different mutations in glp-1, which is also required for germ line proliferation (27,48), both extended nematode survival in the presence of P. aeruginosa (Fig. 1, B and C). The mes-1 gene controls initial development of the nematode germ line; some mes-1 mutant animals are completely fertile, whereas others are completely sterile (32,49,50). mes-1 mutant nematodes that are fertile survive as long as wildtype when exposed to P. aeruginosa, whereas mes-1 mutant nematodes that are sterile are much more resistant (Fig. 1D). Together, these findings indicate that germ line proliferation normally increases C. elegans susceptibility to P. aeruginosa as nematodes lacking a germ line are more resistant.
The effect of the germ line on pathogen resistance was not unique to P. aeruginosa, as removal of the germ line by mutation of either glp-4 or either of two glp-1 alleles also increased survival when C. elegans was exposed to a second Gram-negative bacterial pathogen, S. marcescens (Fig. 1, E-G). Thus, nematodes lacking a germ line are resistant to at least two different Gram-negative bacterial pathogens. Because we observed that mutant nematodes that lack a germ line exhibit enhanced host defense, we can conclude that the germ line actively inhibits host defense in wild-type nematodes.
The longer lifespan of germ linedeficient nematodes in the presence of the nonpathogenic bacterium E. coli depends on the activity of the FOXO family transcription factor DAF-16 (Fig. 2, A and B) (1). In contrast, we find that the enhanced resistance to P. aeruginosa caused by mutation of either glp-1 or mes-1 is completely independent of the activity of DAF-16. daf-16(mu86);mes-1 sterile double mutant nematodes live as long as mes-1 sterile single mutant nematodes in the presence of P. aeruginosa (Fig. 2C). Likewise, daf-16(mu86);glp-1 double mutant nematodes live as long as (or even slightly longer than) glp-1 single mutant nematodes in the presence of pathogen (Fig. 2D). To confirm that the daf-16 mutation could still exert an effect in these strains, we verified that the presence of the daf-16(mu86) mutation did diminish the lifespan of either mes-1 or glp-1 mutant nematodes grown in the presence of E. coli (Fig. 2, A and B). Thus, daf-16 is required for lifespan FIGURE 2. The resistance of sterile nematodes to P. aeruginosa is independent of daf-16 under standard nematode culture conditions. Depicted are representative survival plots for the indicated C. elegans wildtype (w.t.) or mutant strains, which were exposed to either E. coli (A and B) or P. aeruginosa (C-H). Survival studies were carried out as described in the legend to Fig. 1, and the alleles tested and further statistical data (medians, n, and p values) are listed in supplemental Table 1. The mu86 (A-D and G) and mgDf47 (E-H) alleles of daf-16 were used. In G and H, survival analysis was carried out under slightly different conditions (modified NGM and different bacterial growth temperature in G, standard NGM and different bacterial growth temperature in H) as described under "Experimental Procedures" and "Results." extension in the presence of sterilizing mutations but not for the sterility-induced increase in pathogen resistance. This is in contrast to the role of daf-16 in the daf-2 insulin-like signaling pathway, where daf-16 is required for the long lifespan and pathogen resistance of daf-2 (Fig. 2E) (13).
It has been reported that under some conditions, daf-16 could suppress the increased pathogen resistance of sterile nematodes (12,14). We tested the daf-16 allele previously used (daf-16(mgDf47)) and found that daf-16 still failed to affect the pathogen resistance of glp-1 mutant animals (Fig. 2F). One key difference between our survival assays and those previously reported (12,14) is that our assays were carried out under standard nematode culture conditions using NGM (37). Evens et al. and Miyata et al. (12,14) both use a slightly modified growth medium supplemented with additional peptone (38) and grow bacterial lawns at 37°C for 24 h prior to adding nematodes. In the studies presented in this report, bacteria are grown at room temperature overnight prior to adding nematodes. When we carried out survival assays under conditions that were similar to these previous reports (12,14), we found that both daf-16(mu86) and daf-16(mgDf47) mutations did at least partially suppress the increased pathogen resistance of glp-1 (Fig. 2G). Thus, the growth conditions of the bacteria or the nematode host had profound consequences on C. elegans pathogen resistance. To differentiate between these two possibilities, we grew the bacteria at high temperature (37°C) on standard culture medium and then monitored nematode survival, conditions that could alter the pathogen but not the nematode. Under these conditions, we found that a daf-16(mgDf47) mutation did partially suppress the increased pathogen resistance of glp-1 (Fig. 2H), indicating that at least some of the differences in the effects of daf-16 under different conditions is due to changes in the pathogen and not the host.
As in mammals, a p38 MAPK pathway affects C. elegans host defense to numerous pathogens including P. aeruginosa (47). To investigate whether the p38 MAPK pathway might mediate the role of the germ line in host defense, we tested whether the pathogen resistance of glp-1 sterile mutant nematodes required the activity of genes in the p38 MAPK pathway. As observed previously (47), we found that loss of pmk-1 (p38 MAPK) or nsy-1 (MAPKKK) activity in otherwise wild-type nematodes shortened survival in the presence of P. aeruginosa (Fig. 3, A  and C). Mutation of either the MAPK pmk-1 or the MAPKKK nsy-1 also shortened survival of glp-1 mutant nematodes when exposed to P. aeruginosa (Fig. 3, A and C). However, the glp-1;pmk-1 or nsy-1;glp-1 double-mutant nematodes did not survive as short as the MAPK pathway single-mutant animals (Fig. 3, A and C). This indicates that the effect of the germ line on pathogen resistance may be due, at least in part, to a pathway other than the pmk-1 p38 MAPK pathway. These findings are consistent with previous studies showing that mutations in nsy-1 or the MAPKK sek-1 decrease the pathogen resistance of glp-4 (47). These results indicate that the role of the p38 MAPK pathway in regulating C. elegans host defense is additive to that the germ line and are consistent with models where the p38 MAPK pathway acts either downstream or in parallel to the germ line to regulate pathogen resistance.
In contrast to our findings with the pathogen P. aeruginosa, the addition of either a pmk-1 or a nsy-1 mutation to glp-1 mutant animals had little or no effect on nematode lifespan in the presence of nonpathogenic E. coli (Fig. 3, B and D). We note that pmk-1;glp-1 double mutant nematodes survived for a slightly shorter period of time than glp-1 mutant nematodes in the experiment depicted in Fig. 3D; however, in several other experiments, the lifespan of glp-1 and pmk-1;glp-1 mutant animals was indistinguishable (data not shown). Thus, in the absence of the nematode germ line, the daf-16 and p38 MAPK pathways act independently to control survival in the presence of nonpathogenic E. coli (lifespan) and pathogenic P. aeruginosa (innate immunity), respectively.
These epistasis experiments indicate that the p38 MAPK pathway is acting either downstream or in parallel to the germ line in the control of innate immunity. To distinguish between these two possibilities, we used quantitative PCR to monitor the expression of four putative antimicrobial genes (lys-8, clec-85, dod-22, and K08D8.5) whose expression is regulated by the p38 MAPK signaling pathway. Transcription of these genes in the presence of either the nonpathogenic bacterium E. coli or the pathogenic bacterium P. aeruginosa requires the nsy-1 MAPKKK (45). As observed previously (45), expression of all four of these genes is induced by P. aeruginosa in wild-type nematodes (Fig. 4A). Similarly, expression of all four genes is induced by P. aeruginosa in nematodes harboring a mutation in glp-1 (Fig. 4A). However, the glp-1 mutation did not significantly alter expression of these four p38-dependent genes when the nematodes were exposed to either bacterium, suggesting that glp-1 does not alter p38 activity (Fig. 4A) (t test, p Ͼ 0.05 for all comparisons). This is consistent with the germ line and MAPK pathways acting in parallel rather than in serial fashion. To verify that expression of these genes could be induced when Depicted are representative survival plots for the indicated C. elegans wildtype (w.t.) or mutant strains, which were exposed to either P. aeruginosa (A and C), or E. coli (B and D). Survival studies were carried out as described in the legend to Figs. 1 and 2, and the alleles tested and further statistical data (medians, n, and p values) are listed in supplemental Table 1. p38 MAPK activity is increased, we also monitored expression of these genes in nematodes harboring a mutation in vhp-1, a negative regulator of the p38 pathway (51). The expression of lys-8 and dod-22 increased significantly when the MAPK pathway inhibitor vhp-1 was mutated (Fig. 4B) (t test, p ϭ 0.040 and  p ϭ 0.021, respectively).
To confirm that the germ line was not directly affecting the p38 MAPK pathway, we monitored phosphorylation of the MAPK PMK-1 using an antiserum specific to the phosphorylated form of the protein. We compared wild-type nematodes to those containing a mutation in either glp-1 or glp-4. The amount of total PMK-1 protein (phosphorylated and unphosphorylated) was not significantly different in either of the mutant lines tested (Fig. 5A), whether the nematodes were exposed to E. coli or P. aeruginosa. Nematodes containing a mutation in either glp-1 or glp-4 did not have substantially more phospho-PMK-1 than did wild-type nematodes, whether the nematodes were exposed to either E. coli or P. aeruginosa (Fig. 5B). This is consistent with the observation that a glp-1 mutation did not alter the expression of the MAPK-dependent antimicrobial genes (Fig. 4A) and suggests that the C. elegans germ line has little or no effect on p38 MAPK activity. As a control, we quantified phosphorylation of PMK-1 in a nys-1 MAPKKK mutant background. As expected, nematodes harboring a mutation in nsy-1 produced substantially less phosphorylated PMK-1 than wild-type animals (Fig. 5C). We also verified that we could detect an increase in PMK-1 phosphorylation by exposing nematodes to arsenic, which is known to increase PMK-1 phosphorylation (Fig. 5C) (33). Interestingly, the amount of phosphorylated PMK-1 did not increase when nematodes were exposed to P. aeruginosa, as compared with nematodes exposed to E. coli (Fig. 5B), even though P. aeruginosa induces the expression of many p38-dependent antimicrobial genes (Fig. 4) (45,52).

DISCUSSION
The germ line in C. elegans, like the insulin signaling pathway, can control both lifespan in the presence of nonpathogenic bacteria and host defense in the presence of pathogenic bacteria. However, unlike the insulin signaling pathway, where daf-16 is critical to both the lifespan and host defense pathways (Fig. 6A), the germ line requires daf-16 to exert its effect on lifespan but acts independently of daf-16 in host defense (Fig.  6B). This observation demonstrates that innate immunity and lifespan are genetically separable and are therefore distinct processes in C. elegans even though both occur in the presence of bacteria and even though the timescale of infection and aging is much more similar in nematodes than in mammals. Our results are consistent with the prior observation that the germ line affects the response to oxidative damage and thermal stress independently of daf-16 (31,53). This raises an interesting paradox. The long lifespan of daf-2 mutant animals is thought to be due to the increased stress resistance of those animals. Sterile nematodes are likewise also long lived and stress resistant. However, while sterile, daf-16 double mutant nematodes are not long lived; these nematodes are still resistant to multiple stresses (thermal, oxidative, and pathogen stress), which raises the question, why are sterile nematodes long lived? It is possible that another stress is critical in lifespan regulation by the FIGURE 4. The glp-1 mutation does not alter the expression of genes that are regulated by the p38 MAPK pathway. A, wild-type (N2) or glp-1(e2141) mutant animals were exposed to E. coli or P. aeruginosa, RNA was collected, and antimicrobial gene expression was assayed by quantitative PCR using mlc-1 to normalize RNA concentration. Expression was measured relative to N2 grown in the presence of E. coli. Depicted are the results (ϮS.E.) of three independent experiments. B, as a control, wild-type and vhp-1(sa366) mutant nematodes were treated similarly, and antimicrobial gene expression was monitored. Expression levels that were significantly different from control (p Ͻ 0.05, t test) are indicated with an asterisk.  FIGURE 5. Mutations that prevent production of the C. elegans germ line do not increase phosphorylation of PMK-1. Production of PMK-1 (A) and phospho-PMK-1 (B and C) in the indicated wild-type (w.t.) or mutant strains and in the presence of the indicated bacteria was quantitated using antibodies specific to the protein and the phosphorylated form of the protein and normalized using antibodies specific for ␤-tubulin as a reference as described under "Experimental Procedures." Depicted are the results (ϮS.E.) of three independent experiments. Alleles used were: glp-1(e2141), glp-4(bn2), and nsy-1(ok593). As indicates wild-type nematodes that were treated with arsenic as described under "Experimental Procedures." Phospho-PMK-1 levels that were significantly different from wild-type (p Ͻ 0.05, t test) are indicated with an asterisk. germ line. Regardless, the link between lifespan and stress resistance may be more complicated than surmised.
The complexity of the relationship between lifespan and stress resistance is also supported by the observation that under some environmental conditions, the pathogen resistance of sterile nematodes requires daf-16 (Figs. 2G and 6B) (12,14). These modified conditions could either alter host defense in the nematode or virulence of the pathogen. We found that pathogen resistance of sterile nematodes also depended in part on daf-16 when the pathogen was grown at higher temperature on standard NGM (Fig. 2H), conditions that could alter bacterial virulence but not nematode host defense. Alterations in growth conditions are known to have a profound effect on P. aeruginosa virulence in C. elegans, affecting the mechanism and rate of killing (42). Perhaps a different set of virulence factors or toxins is produced by P. aeruginosa under these different conditions.
Our findings also indicate that the germ line acts in parallel to the p38 MAPK pathway to regulate host defense and point to the existence of a yet to be identified set of genes acting downstream of the germ line to control the C. elegans innate immune response (Fig. 6B). It will be of considerable interest to determine the identity of these genes, how they interface with the p38 MAPK pathway to affect innate immunity, and whether their role in the regulation of innate immunity and perhaps lifespan is evolutionarily conserved. In mammals, exposure to pathogen increases activity of the p38 MAPK pathway (54,55). The observation that the activity of the p38 MAPK pathway is not significantly different between nematodes exposed to nonpathogenic E. coli and pathogenic P. aeruginosa suggests two possibilities: First, E. coli could stimulate the C. elegans p38 MAPK pathway to the same level as P. aeruginosa. Consistent with this possibility, E. coli can be weakly pathogenic to C. elegans under some conditions (13,56,57). Alternatively, while the p38 MAPK pathway may be critical to antimicrobial gene expression, the p38 MAPK pathway may act passively in C. elegans, and a separate pathway may control the inducibility of the nematode pathogen response (45, 52, 58 -61). Distinguishing between these two possibilities in C. elegans will prove chal-lenging, given the difficulty of culturing C. elegans in the absence of bacteria.