Intestinal lipid metabolism controls molecular and behavioral immunity through NHR-68 and gut–brain signaling in Caenorhabditis elegans

Abstract

Animals must allocate limited energetic resources across competing defense programs in response to infection. Here, we show that the conserved nuclear hormone receptor NHR-68 integrates fatty acid metabolism with the neural control of molecular and behavioral immunity in Caenorhabditis elegans. Acting in parallel with NHR-10, NHR-68 controls genes involved in polyunsaturated fatty acid (PUFA) metabolism. Loss of NHR-68 disrupts linoleic acid (LA) homeostasis, impairing pathogen avoidance behavior. Supplementation with LA restores avoidance, and fat-3 inhibition, which elevates LA, enhances pathogen avoidance, whereas loss of LA synthesis by fat-2 inhibition diminishes this behavior, indicating that LA promotes behavioral immunity. We further show that NHR-68 acts in the intestine to regulate linoleic acid homeostasis, and that changes in intestinal lipid metabolism influence an AWC-dependent pathogen-avoidance circuit through intestine-to-neuron communication. NHR-68 suppresses activation of the PMK-1/p38 MAPK and DAF-16/FOXO pathways, which mediate molecular immune responses. These findings identify a gut-brain transcriptional circuit that connects intestinal lipid metabolism to neural and immune outputs, revealing a mechanism by which the metabolic state coordinates behavioral and molecular defenses to optimize host protection.

Mounting an immune response is energetically costly, and so host defenses are often influenced by internal metabolic states. This study identifies a gut-brain signaling mechanism in C. elegans wherein intestinal NHR-68 coordinates fatty acid homeostasis to shape pathogen avoidance and molecular immune responses.

Introduction

Animals must continuously balance the allocation of limited resources among competing physiological needs, including growth, metabolism, and defense. Mounting immune responses is energetically costly, and excessive or prolonged activation can compromise fitness. As a result, host defense systems have evolved strategies that integrate immune signaling with metabolic state and environmental cues to optimize protection while preserving energy homeostasis. The coordination of these regulatory layers across tissues remains poorly understood.

Nuclear hormone receptors (NHRs) are a family of transcription factors that couple metabolic state to gene-expression programs controlling development, metabolism, reproduction, and stress responses [1]. These receptors act by binding to specific signaling molecules, such as steroid hormones, retinoids, and thyroid hormones, which modulate gene expression to control a variety of physiological processes [2]. In C. elegans, NHRs are involved in key biological functions such as responses to oxidative stress [3], starvation [4,5], hypoxia [6], and pathogen infection [7–12]. The C. elegans genome encodes over 280 NHRs, more than in humans or mice, and only a small subset has been functionally characterized [13].

In C. elegans, the intestine is a major metabolic and immune organ that also influences behavior through inter-tissue communication [14–17]. In addition to molecular immune responses, C. elegans relies on pathogen avoidance as an important behavioral defense strategy. Animals can detect pathogenic bacteria and undergo aversive learning to reduce subsequent exposure, and this process is regulated by multiple neuronal and signaling pathways [18–20]. Fatty-acid metabolism is a central readout of intestinal metabolic state governed by networks of NHRs [21,22]. C. elegans has evolved an array of conserved desaturase (fat) and elongase (elo) genes that can synthesize a broad range of fatty acids de novo from acetyl-CoA [23]. In addition to their roles in nervous system function [24], both monounsaturated fatty acids (MUFAs) and polyunsaturated fatty acids (PUFAs) can exhibit antibacterial functions [25]. The MUFA oleate is necessary for innate immune gene expression and protects C. elegans from Pseudomonas aeruginosa infection [26,27]. The PUFA eicosapentaenoic acid (EPA) also enhances the C. elegans immune response and functions as an effector that can disrupt or inhibit biofilm formation, helping to regulate host immunity against Candida albicans [28]. Genetic studies have shown that fatty-acid desaturation and elongation pathways, including fat-2 and fat-3, regulate polyunsaturated fatty acid composition and affect susceptibility to P. aeruginosa infection [27,29]. In addition, fatty-acid supplementation can modulate host immune responses and survival during bacterial infection, supporting a functional link between lipid metabolism and innate immunity [26,30]. Despite these links, the specific intestinal metabolic changes that occur during bacterial infection and how they are integrated with neural control remain unclear.

Here, we show that NHR-68, a member of the expanded C. elegans nuclear hormone receptor family with similarities to mammalian hepatocyte nuclear factor 4-alpha (HNF4α), acts in parallel with NHR-10 to coordinate fatty acid metabolism, innate immune signaling, and pathogen avoidance. NHR-68 regulates genes involved in PUFA biosynthesis, and its loss reduces linoleic acid (LA), a fatty acid that promotes avoidance behavior. We found that NHR-68 acts in the intestine and signals through the AWC chemosensory neuron to drive behavioral immunity, while suppressing the PMK-1/p38 MAPK and DAF-16/FOXO pathways that mediate molecular immune responses. Together, these findings identify a transcriptional mechanism that links intestinal metabolism to neural and immune outputs. Finally, our results suggest a general principle for cross-tissue coordination of defense. By aligning intestinal lipid metabolism with neural avoidance and modulation of immune activation, nuclear receptors provide a way to balance energetic costs and benefits during infection. This principle may be conserved given the broad conservation of nuclear receptor signaling and lipid-derived cues.

Results

NHR-10 and NHR-68 inhibit innate immune activation

Nuclear hormone receptors (NHRs) are well-characterized regulators of metabolism, development, and reproduction [1], but their involvement in innate immune regulation remains poorly understood. To identify NHRs that control immune activation, we performed a reverse genetic screen using 236 nhr genes from the Ahringer C. elegans RNAi feeding library [31]. As a readout of intestinal immune activation, we used a transcriptional reporter for clec-60, a C-type lectin gene whose expression is strongly induced in response to infection and other immune triggers [32–34]. The clec-60p::gfp reporter is widely used as a readout for the activation of protective innate immune pathways in the C. elegans intestine [32–34].

We identified nine nhr genes (nhr-77, nhr-10, nhr-130, nhr-222, nhr-162, nhr-42, nhr-140, nhr-68, and nhr-54) whose knockdown resulted in pronounced increased clec-60p::gfp expression (Fig 1A; Fig A in S1 Appendix; S1 Table). Most of these NHRs, including nhr-77, nhr-10, nhr-42, nhr-68, and nhr-54, have been identified in genome-wide transcription factor profiling studies as components of regulatory networks associated with metabolic and stress-responsive processes [35], although their roles in intestinal immune regulation remain largely uncharacterized. Notably, nhr-10 and nhr-68 were of particular interest, as they are known to function together as a persistence detection module in propionate metabolism [36]. Given their shared roles in metabolic regulation and in the expression of immune gene clec-60, we focused our analysis on nhr-10 and nhr-68. Animals treated with double RNAi targeting both nhr-10 and nhr-68 showed a further enhancement of the clec-60p::gfp signal compared to either single knockdown (Fig 1A; Fig A in S1 Appendix). This was confirmed by qRT-PCR, which revealed an ~8–10-fold increase in clec-60 transcript levels in single mutants and a ~20-fold increase in the double mutant (Fig 1B), indicating that the two NHRs have additive effects in suppressing immune activation.

Fig 1. NHR-10 and NHR-68 inhibit innate immune response.

(A) Representative fluorescence micrographs of clec-60p::gfp animals grown on E. coli HT115 carrying empty vector or RNAi against nhr-77, nhr-10, nhr-130, nhr-222, nhr-162, nhr-42, nhr-140, nhr-68, nhr-54, or nhr-10 and nhr-68 (n = 8 animals per condition per experiment; N = 3 independent experiments). Fluorescence images were captured with 100-ms exposures. Scale bar, 200 µm. (B) qRT–PCR analysis of the immune gene clec-60 in wild-type, nhr-10(tm4695), nhr-68(gk708) and nhr-10(tm4695); nhr-68(gk708) animals fed on E. coli. Data are presented as the mean ± SD from three independent biological replicates (N = 3; n = 200 animals per condition per biological replicate). ***p < 0.001, one-way ANOVA with Tukey’s multiple-comparison test. (C) Survival of WT, nhr-10(tm4695), nhr-68(gk708) or nhr-10(tm4695);nhr-68(gk708) animals grown on P. aeruginosa at 25 °C. Mean survival times were 2.06, 2.36, 2.82, and 3.13 days, respectively. Data are presented from three independent experiments, with 90 animals per condition per experiment (N = 3; n = 90 animals per condition per experiment). The Kaplan–Meier method was used to calculate the survival fractions, and statistical significance between survival curves was determined using the log-rank test. WT vs. nhr-10(tm4695), p < 0.01; WT vs. nhr-68(gk708), p < 0.001; WT vs. nhr-10(tm4695); nhr-68(gk708), p < 0.001; nhr-10(tm4695) vs. nhr-10(tm4695); nhr-68(gk708), p < 0.001; nhr-68(gk708) vs. nhr-10(tm4695); nhr-68(gk708), p = 0.0435. (D) Occupancy index for WT, nhr-10(tm4695), nhr-68(gk708) or nhr-10(tm4695);nhr-68(gk708) animals on P. aeruginosa at 8 h and 12 h. The occupancy index was calculated as (N_on lawn/N_total). Data are presented as mean ± SD from three independent experiments (N = 3). The data underlying this Figure can be found in S1 Data.

We then evaluated the nhr-10(tm4695), nhr-68(gk708), two deletion alleles predicted to cause loss of function, as well as nhr-10(tm4695);nhr-68(gk708) animals for susceptibility to P. aeruginosa. Both single mutants survived longer than wild-type animals, indicating increased resistance, and nhr-10(tm4695);nhr-68(gk708) animals survived slightly longer than either single mutant, consistent with an additive effect of the two genes (Fig 1C). C. elegans uses behavioral avoidance as a defense against pathogens [37,38]. We found that nhr-10(tm4695) and nhr-68(gk708) animals showed significantly reduced bacterial avoidance behavior, with a modest further reduction in the nhr-10(tm4695);nhr-68(gk708) animals (Fig 1D; Fig BA in S1 Appendix). Reversal and locomotion assays confirmed that the reduced avoidance phenotype was not due to impaired locomotor ability in these mutants (Fig BB and BC in S1 Appendix). These findings indicate that NHR-10 and NHR-68 concurrently suppress molecular immunity linked to pathogen resistance while promoting pathogen avoidance behavior.

Because nhr-68(gk708) animals displayed a pathogen-resistant phenotype, reduced bacterial avoidance, and increased clec-60 expression, we next asked whether the expression of nhr-68 could reverse these phenotypes. To address this, we constructed an nhr-68p::nhr-68::gfp translational reporter expressing GFP-tagged NHR-68 under its native promoter to rescue expression in the mutant background. As shown in Fig 2A, nhr-68::gfp was expressed in the hypodermis, intestine, and head neurons, consistent with a previous study [39]. Expression of nhr-68 also rescued the increased clec-60 mRNA levels in nhr-68(gk708) animals (Fig 2B) and partially rescued pathogen susceptibility (Fig 2C) and bacterial avoidance (Fig 2D; Fig BD in S1 Appendix). These results indicate that nhr-68 expression compensates for the mutant phenotype and that nhr-10 and nhr-68 act in a parallel fashion to control molecular and behavioral immune responses to P. aeruginosa.

Fig 2. The expression of nhr-68 restored pathogen susceptibility.

(A) Representative fluorescence micrographs of nhr-68(gk708);nhr-68p::nhr-68::gfp animals. n = 30; data are representative of four independent experiments. Fluorescence images were captured using 400-ms exposures. Scale bars indicate 100 µm in the left two panels and 50 µm in the right two panels. (B) qRT–PCR analysis of the immune gene clec-60 in wild-type, nhr-68(gk708) or nhr-68(gk708);nhr-68p::nhr-68 animals fed on E. coli. Data are presented as the mean ± SD from three independent biological experiments. N = 3; n = 200 animals per condition per biological replicate. ***p < 0.001, one-way ANOVA with Tukey’s multiple-comparison test. (C) Survival of wild-type N2, nhr-68(gk708) or nhr-68(gk708);nhr-68p::nhr-68 animals on P. aeruginosa at 25 °C. Mean survival times were 2.07, 2.83, and 2.35 days, respectively. Data are presented from three independent experiments, with 90 animals per condition per experiment (N = 3; n = 90 animals per condition per experiment). Wild-type vs. nhr-68(gk708), p < 0.0001; Wild-type vs. nhr-68(gk708); nhr-68p::nhr-68, p = 0.0002; nhr-68(gk708) vs. nhr-68(gk708); nhr-68p::nhr-68, p = 0.0202. (D) P. aeruginosa occupancy indexes for WT, nhr-68(gk708) and nhr-68(gk708);nhr-68p::nhr-68 animals at 8 h and 12 h after seeding. The occupancy index was calculated as (N_on lawn/N_total). Data are presented as mean ± SD from three independent experiments (N = 3). The data underlying this Figure can be found in S1 Data.

Previous studies showed that NHR-10 and NHR-68 regulate genes linked to vitamin B12 metabolism and act together as a persistence detector for propionate metabolism, enabling metabolic network rewiring in response to dietary inputs [36]. To assess whether vitamin B12 supplementation affects the NHR-68-regulated immune response, we measured expression of the immune gene clec-60 with or without vitamin B12. Vitamin B12 did not alter clec-60 expression in nhr-68(gk708) animals (Fig CA in S1 Appendix), suggesting that the NHR-68-mediated regulation of clec-60 is unlikely to be directly linked to vitamin B12 metabolism under these conditions.

We showed that nhr-68(gk708) animals exhibit increased pathogen resistance against P. aeruginosa (Figs 1C and 2C). To test whether vitamin B12 influences pathogen resistance, we examined resistance using E. faecalis, which, unlike P. aeruginosa [40], is vitamin B12 auxotrophic and relies on salvage [41]. As shown in Fig CB in S1 Appendix, nhr-68(gk708) animals displayed increased resistance to E. faecalis, and this phenotype was unaffected by vitamin B12 supplementation. Together with the additive effects observed for NHR-10 and NHR-68 in controlling clec-60 expression and pathogen resistance (Fig 1C and 1D), these results support a model in which NHR-10 and NHR-68 both contribute to the regulation of molecular and behavioral immune responses.

NHR-68 inhibits the PMK-1 and DAF-16 pathways during immune activation

To investigate the mechanism underlying the increased survival of nhr-68(gk708) animals upon P. aeruginosa (Fig 1C), we performed an RNA-seq analysis of the changes in gene expression profiles of these animals after 12 h infection with P. aeruginosa. Principal component analysis (PCA) of the transcriptomes revealed clear separation between nhr-68(gk708) and wild-type animals, indicating distinct gene expression patterns (Fig DA in S1 Appendix). To reduce false positives, only genes with a fold change greater than two and p < 0.05 were considered differentially expressed (S2 Table).

An unbiased gene enrichment analysis performed using the Database for Annotation, Visualization, and Integrated Discovery (DAVID) (https://david.ncifcrf.gov/) [42] revealed that the innate immune response Gene Ontology (GO) cluster had the second-highest number of hits, following signal transduction. Other enriched GO clusters corresponded to transmembrane transport and proteolysis (Fig 3A and S3 Table).

Fig 3. NHR-68 inhibits immune pathways.

(A) Gene Ontology analysis of genes upregulated in nhr-68(gk708) vs. WT animals after 12 h infection with P. aeruginosa. The results were filtered to include only terms significantly enriched in nhr-68(gk708) with a q value <0.1. (B) The representation factors of the identified upregulated genes controlled by the PMK-1, DAF-16, SKN-1, and ELT-2 pathways. (C) qRT–PCR analysis of PMK-1- and DAF-16-dependent genes in nhr-68(gk708) compared with WT animals. (N = 3 biological replicates). (D) Survival of wild-type and nhr-68(gk708) animals fed empty vector control or pmk-1 RNAi followed by exposure to P. aeruginosa. Mean survival times were 2.40 and 3.01 days for empty vector control RNAi and 1.88 and 1.92 days for pmk-1 RNAi, respectively. Wild-type vs. nhr-68(gk708), p < 0.0001 and p = 0.1121 for empty vector control and pmk-1 RNAi, respectively. Data are presented from three independent experiments, with 90 animals per condition per experiment (N = 3; n = 90 animals per condition per experiment). The Kaplan–Meier method was used to calculate the survival fractions, and statistical significance between survival curves was determined using the log-rank test. (E) Survival of wild-type and nhr-68(gk708) animals fed empty vector control or daf-16 RNAi followed by exposure to P. aeruginosa. Mean survival times were 2.40 and 3.01 days for empty vector control RNAi and 2.14 and 2.43 days for daf-16 RNAi, respectively. Wild-type vs. nhr-68(gk708), p < 0.0001 and p = 0.0130 for empty vector control and daf-16 RNAi, respectively. Data are presented from three independent experiments, with 90 animals per condition per experiment (N = 3; n = 90 animals per condition per experiment). The Kaplan–Meier method was used to calculate the survival fractions, and statistical significance between survival curves was determined using the log-rank test. The data underlying this Figure can be found in S1 Data.

To further explore immune activation, we compared genes upregulated in nhr-68(gk708) animals to previously identified gene sets regulated by immune pathways in C. elegans (S4 Table). Gene sets controlled by PMK-1 and DAF-16 showed significant overlap with genes whose expression increased in nhr-68(gk708) animals infected with P. aeruginosa (Fig 3B). qRT-PCR validation of selected genes confirmed the RNA-seq data (Fig 3C), supporting the conclusion that loss of nhr-68 leads to activation of these immune genes.

To identify NHR-68-controlled genes in the absence of infection, we also performed RNA-seq on nhr-68(gk708) and wild-type animals fed only on E. coli OP50. PCA analysis showed clear separation between the two groups, indicating distinct expression profiles (Fig DB in S1 Appendix). Comparison of nhr-68 mutants and wild-type animals under these basal conditions revealed genes directly regulated by NHR-68 (S2 Table). Gene Ontology analysis revealed that genes involved in the innate immune response were highly enriched (Fig EA in S1 Appendix and S3 Table). Our analysis identified immune genes that overlap with the NHR-68-regulated genes from previously published studies [36,43], as well as additional immune-related genes uniquely detected in our study (S3 Table), despite differences in experimental conditions highlighted in S3 Table. Many of these NHR-68-dependent immune genes overlapped with immune-related gene sets previously associated with the PMK-1 and DAF-16 pathways, as well as with the ELT-2 and SKN-1 pathways (Fig EB in S1 Appendix and S4 Table), consistent with the overlap patterns observed in P. aeruginosa-infected samples. These results suggest that NHR-68 represses a subset of immune genes even in the absence of pathogen exposure and functions as a negative regulator of multiple immune signaling pathways.

We hypothesized that if PMK-1 and DAF-16 activation drives the enhanced survival of nhr-68(gk708) animals during P. aeruginosa infection, then impairing these pathways should reduce the phenotype. Indeed, RNAi knockdown of pmk-1 or daf-16 reduced the magnitude of the increased survival in nhr-68(gk708) animals infected with P. aeruginosa (Fig 3D and 3E). These results indicate that the activation of the PMK-1 and DAF-16 immune pathways contributes to the increased pathogen resistance observed in nhr-68(gk708) animals.

Having shown that NHR-68 inhibits the PMK-1 and DAF-16 pathways, we next compared NHR-68-dependent gene sets with P. aeruginosa-regulated transcriptomes (S2 and S3 Tables). This analysis revealed overlaps with both infection-induced and infection-repressed gene categories, as well as with immune and metabolic regulatory transcriptional programs, including multiple nuclear hormone receptors. These shared transcriptional features further support a role for NHR-68 in coordinating immune and metabolic gene regulation during pathogen response. Consistent with this transcriptional response, qRT-PCR analysis showed that both nhr-68 and nhr-10 transcript levels were modestly increased upon exposure to P. aeruginosa in wild-type animals (Fig FA in S1 Appendix), indicating that pathogen infection also modulates NHR expression.

To further examine the relationship between NHR-10 and NHR-68 in immune gene regulation, we quantified additional NHR-68-regulated immune genes in nhr-10(tm4695) animals (Fig FB in S1 Appendix). Analysis of shared target genes revealed that most were upregulated, whereas a subset was unaffected, indicating partial overlap between NHR-10- and NHR-68-mediated immune gene regulation.

NHR-68 regulates avoidance behavior through polyunsaturated fatty acid metabolism

Pathogen avoidance can be strongly induced by disrupting the defecation motor program, and aex-5 RNAi animals, which undergo intestinal distension, robustly avoid P. aeruginosa [19,20]. A gas chromatography–mass spectrometry (GC–MS)-based metabolomic profiling we had carried out on these strong-avoiding animals indicated that linoleic acid (LA), an omega-6 polyunsaturated fatty acid, was markedly elevated relative to controls (Fig 4A and S5 Table).

Fig 4. LA accumulation enhances pathogen avoidance.

(A) Metabolomic analysis. Animals fed on empty vector or aex-5 RNAi were harvested for metabolomic analysis using GC–MS. The top-ranked differentially abundant metabolites were identified, and their relative expression was plotted in a heatmap. (B) Relative linoleic acid content measured by LC–HRMS in wild-type and nhr-68(gk708) mutant animals is shown as a heatmap; each column represents one of three biological replicates. (C) P. aeruginosa occupancy index in wild-type and nhr-68(gk708) animals treated with vehicle or LA at 8 h and 12 h. The occupancy index was calculated as (N_on lawn/N_total). LA, linoleic acid. The bars represent the means, while the error bars indicate the SDs of three independent experiments (N = 3; n = 90 animals per condition per experiment); ***p < 0.001. (D) Schematic of the PUFA synthesis pathway in C. elegans, adapted from [29]. The diagram depicts all known steps in PUFA synthesis as well as the elongase and desaturase enzymes involved. Abbreviations: PA, palmitic acid; PLA, palmitoleic acid; VA, vaccenic acid; SA, stearic acid; OA, oleic acid; LA, linoleic acid; ALA, alpha-linolenic acid; GLA, gamma-linolenic acid; STA, stearidonic acid; DGLA, dihommo gamma-linolenic acid; ETA, eicosatrienoic acid; AA, arachidonic acid; EPA, eicosapentaenoic acid. (E) P. aeruginosa occupancy index in wild-type and nhr-68(gk708) animals treated with vehicle or VA at 8 h and 12 h. The occupancy index was calculated as (N_on lawn/N_total). VA, vaccenic acid. The bars represent the means, while the error bars indicate the SDs of three independent experiments (N = 3; n = 90 animals per condition per experiment). (F) qRT–PCR analysis of fat-1, fat-2, fat-3 and fat-4 in animals fed on E. coli. Data are presented as the mean ± SD from three independent experiments (N = 3; n = 200 animals per condition per biological replicate). ***p < 0.001, t test. (G) GC–MS analysis of linoleic acid abundance in fat-3 RNAi animals. Animals fed empty vector or fat-3 RNAi were harvested under E. coli or P. aeruginosa conditions and subjected to GC–MS-based metabolomic profiling. (H) Quantification of linoleic acid abundance in control and P. aeruginosa-exposed animals. Data are presented as mean ± SD. One-way ANOVA was used to determine statistical significance. N = 3; n = 2,000 animals per condition per experiment. (I) P. aeruginosa occupancy index of wild-type and nhr-68(gk708) animals fed empty vector control or fat-3 RNAi at 8 h and 12 h. (J) P. aeruginosa occupancy index of wild-type and nhr-68(gk708) animals fed empty vector control or fat-2 RNAi at 8 h and 12 h. The occupancy index was calculated as (N_on lawn/N_total). Data are presented as mean ± SD from three independent experiments (N = 3; n = 90 animals per condition per experiment). The data underlying this Figure can be found in S1 Data.

Because NHRs play an important role in regulating lipid metabolism and maintaining fatty acid homeostasis [21,22,44], we measured LA levels in nhr-68(gk708) animals. Liquid chromatography–high-resolution mass spectrometry (LC–HRMS) analysis showed a significant decrease in LA content in nhr-68(gk708) animals (Fig 4B and S5 Table). To assess whether elevated LA levels contribute to pathogen avoidance, we supplemented wild-type animals with LA and observed enhanced avoidance behavior toward P. aeruginosa (Fig 4C and Fig GA in S1 Appendix). We ruled out a direct aversive effect of LA using bacterial choice assays (Fig GB in S1 Appendix). To test whether LA is required for NHR-68-mediated avoidance, we supplemented nhr-68(gk708) animals with LA. While these nhr-68(gk708) animals typically exhibit reduced avoidance behavior, LA treatment fully rescued this defect and further enhanced avoidance behavior. The enhanced avoidance observed in nhr-68(gk708) animals was comparable to that of wild-type animals supplemented with LA (Fig 4C and Fig GA in S1 Appendix). Importantly, despite its strong effects on pathogen avoidance, LA supplementation did not alter the increased survival phenotype of nhr-68(gk708) animals (Figs HA in S1 Appendix), indicating that LA-dependent behavioral changes are not sufficient to modify survival outcomes. To examine whether this effect was specific to LA, we tested vaccenic acid (VA; 18:1 n-7), a monounsaturated fatty acid structurally similar to LA but outside the LA metabolic pathway (Fig 4D). VA supplementation did not affect general avoidance behavior or NHR-68-regulated avoidance behavior (Fig 4E and Fig GC in S1 Appendix).

To determine if nhr-68 influences LA metabolism, we measured the expression of genes involved in LA biosynthesis. In nhr-68(gk708) animals, we observed elevated expression of fat-3 and fat-4, which are involved in downstream desaturation steps, while fat-1 and fat-2 showed no significant changes (Fig 4F). This expression pattern implies a disruption in LA accumulation in the absence of functional NHR-68.

To test whether LA is controlled by the desaturase pathway downstream of NHR-68, we examined LA in fat-3 RNAi animals. Based on the PUFA biosynthetic pathway (Fig 4D), fat-3 knockdown is predicted to cause LA accumulation due to reduced downstream desaturation. GC–MS profiling revealed a significant increase in LA levels upon fat-3 knockdown (Fig 4G and 4H; S5 Table), consistent with FAT-3 functioning in LA utilization downstream of NHR-68.

Behaviorally, fat-3 RNAi, which increases LA levels (Fig 4G and 4H; S5 Table), also enhanced pathogen avoidance in wild-type animals (Fig 4I and Fig GD in S1 Appendix), linking LA accumulation to avoidance behavior. Importantly, fat-3 RNAi did not increase the survival of nhr-68(gk708) animals (Fig HB in S1 Appendix), further supporting that LA-associated behavioral regulation is separable from survival regulation in this context.

To examine whether FAT-3 also mediates pathogen-associated changes in LA homeostasis, we quantified LA levels by GC-MS in vector control and fat-3 RNAi animals under both E. coli and P. aeruginosa conditions. P. aeruginosa exposure increased LA abundance in vector control animals, whereas no further increase was observed in fat-3 RNAi animals, which already displayed elevated basal LA levels (Fig 4G and 4H; S5 Table). These data are consistent with a role for FAT-3 in regulating LA levels during infection.

Additionally, reduction of LA synthesis by fat-2 RNAi significantly decreased pathogen avoidance, phenocopying the defect observed in nhr-68(gk708) animals (Fig 4J and Fig GE in S1 Appendix). Together, these findings identify LA as a key metabolic component of the NHR-68-regulated fatty-acid pathway that promotes pathogen-avoidance behavior and link intestinal lipid homeostasis to behavioral immune regulation.

NHR-68 controls pathogen avoidance via the chemosensory neuron AWC

Previous studies have shown that chemosensory neurons AWB, AWC, and ASI are required for P. aeruginosa-induced avoidance behavior [20]. Consistent with these findings, genetic ablation of AWC, AWB, or ASI significantly reduced avoidance behavior in the corresponding control animals (Fig 5A–5C and Fig IA–IC in S1 Appendix, empty vector groups). To identify the specific neurons involved in NHR-68-regulated avoidance of P. aeruginosa, we used animals with genetic ablation of those individual chemosensory neurons and knocked down nhr-68 by RNAi in each background. ASE(−) animals were used as a negative control, as ASE neurons are required for E. faecalis-induced avoidance but not P. aeruginosa-induced avoidance [20]. Among the neuron-ablated strains, only AWC(−) animals failed to exhibit an additional reduction in avoidance behavior upon nhr-68 knockdown (Fig 5A and Fig IA in S1 Appendix). Ablation of AWB, ASI, or ASE neurons did not abolish the effect of nhr-68 knockdown (Fig 5B–5D and Fig IB–ID in S1 Appendix), indicating that the effect of nhr-68 RNAi is retained in those backgrounds. These results indicate that AWC neurons are required for the NHR-68-dependent component of behavioral immunity against P. aeruginosa.

Fig 5. NHR-68 RNAi inhibits pathogen avoidance behavior via the chemosensory neuron AWC.

(A) P. aeruginosa occupancy indexes of WT and AWC(−) animals fed empty vector control or nhr-68 RNAi at 8 h and 12 h. (B) P. aeruginosa occupancy index for WT and AWB(−) animals fed empty vector control or nhr-68 RNAi at 8 h and 12 h. (C) P. aeruginosa occupancy index of WT and ASI(−) animals fed empty vector control or nhr-68 RNAi at 8 h and 12 h. (D) P. aeruginosa occupancy index of WT and ASE(−) animals fed empty vector control or nhr-68 RNAi at 8 hr and 12 hr. (E–H) P. aeruginosa occupancy indexes at 8 h and 12 h for WT and sensory neuron-ablated animals treated with vehicle or LA. Each panel includes WT animals and one neuron-ablated strain: (E) AWC(−), (F) AWB(−), (G) ASI(−), or (H) ASE(−). LA, linoleic acid. The means and SDs of three independent experiments are shown; the bars, indicated by solid black horizontal lines within the violin plots, represent the means, and the error bars, indicated by thin horizontal lines within the violin plots, represent the SD. Data are presented as mean ± SD from three independent experiments (N = 3; n = 90 animals per condition per experiment). *p < 0.05, **p < 0.01, ***p < 0.001, and ns = not significant. The data underlying this Figure can be found in S1 Data.

AWC-mediated olfactory adaptation is known to depend on PUFA signaling, and animals with mutations in fat genes display defects in AWC function [45]. Since our findings indicate that LA enhances avoidance behavior (Fig 4C and Fig GA in S1 Appendix), we next asked whether this effect is also dependent on specific sensory neurons. LA supplementation failed to rescue the avoidance defect in AWC(−) animals, which remained significantly impaired relative to wild-type controls (Fig 5E and Fig IE in S1 Appendix), whereas it enhanced avoidance behavior in AWB(−), ASI(−), and ASE(−) animals (Fig 5F–5H and Fig IF–IH in S1 Appendix). Together, these data suggest that AWC neurons are required for both NHR-68-regulated and PUFA-induced avoidance of P. aeruginosa, indicating that AWC functions as a consistent component of the behavioral circuit in this response.

Interestingly, reduced avoidance in nhr-68-deficient animals did not compromise host defense against P. aeruginosa. Instead, nhr-68(gk708) animals showed increased survival despite their avoidance defect (Fig 1C). To test whether the neuronal circuitry controlling avoidance also contributes to survival, we examined nhr-68(gk708) animals in backgrounds lacking individual chemosensory neurons. Ablation of AWC, AWB, ASI, or ASE neurons did not alter the enhanced survival phenotype of nhr-68(gk708) animals on P. aeruginosa (Fig JA–JD in S1 Appendix). Under each condition, nhr-68 RNAi animals remained more resistant than controls. These results indicate that the increased pathogen resistance conferred by loss of nhr-68 is independent of the chemosensory neurons required for avoidance behavior. Together, these findings suggest that NHR-68 regulates behavioral immunity via AWC neurons, whereas its effect on host survival is mediated through a distinct mechanism.

The intestinal function of NHR-68 is required for avoidance behavior against P. aeruginosa

To identify the tissues in which nhr-68 functions to regulate avoidance behavior, we performed tissue-specific RNAi knockdown of nhr-68. Intestinal knockdown using the intestinal-specific RNAi strain MGH171 significantly decreased avoidance behavior (Fig 6A and 6B). In contrast, neuron-specific knockdown using strain MAH677, which expresses sid-1 pan-neuronally and has been previously validated for efficient neuronal RNAi [46,47], had no effect on avoidance behavior (Fig 6C), indicating that neuronal NHR-68 is not required for behavioral immunity.

Fig 6. The intestinal function of NHR-68 is required for pathogen avoidance behavior.

(A) P. aeruginosa occupancy index of wild-type animals fed empty vector control or nhr-68 RNAi at 8 h and 12 h. (B) P. aeruginosa occupancy index of MGH171 animals fed empty vector control or nhr-68 RNAi at 8 h and 12 h. (C) P. aeruginosa occupancy index of MAH677 animals fed empty vector control or nhr-68 RNAi at 8 hr and 12 hr. (D) P. aeruginosa occupancy index for WT, nhr-68(gk708) and nhr-68(gk708);vha-6p::nhr-68 animals at 8 h and 12 h after seeding. The occupancy index was calculated as (N_on lawn/N_total). Data are presented as mean ± SD from three independent experiments (N = 3; n = 90 animals per condition per experiment). (E) Survival of wild-type animals fed empty vector control or nhr-68 RNAi on P. aeruginosa at 25 °C. Mean survival times were 2.24 and 2.89 days, respectively. Empty vector vs. nhr-68 RNAi, p < 0.0001. (F) Survival of MGH171 animals fed empty vector control or nhr-68 RNAi on P. aeruginosa at 25 °C. Mean survival times were 2.23 and 3.02 days, respectively. Empty vector vs. nhr-68 RNAi, p < 0.0001. (G) Survival of MAH677 animals fed empty vector control or nhr-68 RNAi on P. aeruginosa at 25 °C. Mean survival times were 2.27 and 2.28 days, respectively. Empty vector vs. nhr-68 RNAi, p = ns. (H) Survival of wild-type N2, nhr-68(gk708), or nhr-68(gk708);vha-6p::nhr-68 animals on P. aeruginosa at 25 °C. Mean survival times were 2.17, 2.81, and 2.40 days, respectively. Wild-type vs. nhr-68(gk708), p < 0.0001; Wild-type vs. nhr-68(gk708);vha-6p::nhr-68, p = 0.0145; nhr-68(gk708) vs. nhr-68(gk708);vha-6p::nhr-68, p = 0.0017. Data are presented from three independent experiments, with 90 animals per condition per experiment (N = 3; n = 90 animals per condition per experiment). The Kaplan-Meier method was used to calculate the survival fractions, and statistical significance between survival curves was determined using the log-rank test. The data underlying this Figure can be found in S1 Data.

Since intestinal knockdown of nhr-68 impaired avoidance behavior, we next tested whether intestine-specific expression of nhr-68 is sufficient to rescue the reduced avoidance behavior of nhr-68(gk708) animals. To drive expression specifically in the intestine, we placed nhr-68 under the control of the vha-6 promoter, which is active in intestinal cells [48]. Intestinal expression of nhr-68 rescued the reduced avoidance behavior of nhr-68(gk708) animals (Fig 6D).

Consistent with these findings, intestine-specific knockdown of nhr-68 also increased survival on P. aeruginosa, whereas neuron-specific knockdown had no detectable effect (Fig 6E–6G), indicating that intestinal NHR-68 is required for both behavioral and survival responses during infection. In addition, intestinal expression of nhr-68 partially rescued the pathogen susceptibility of nhr-68(gk708) animals (Fig 6H), further supporting a major role of intestinal NHR-68 in host defense.

Together, these findings indicate NHR-68 functions in the intestine to regulate avoidance behavior against P. aeruginosa.

Discussion

In this study, we identify NHR-68 as a transcriptional regulator that coordinates metabolism and immune responses in C. elegans (Fig 7). Our findings show that NHR-68 controls LA metabolism in the intestine and that this regulation influences an AWC-dependent pathogen-avoidance behavior. These results define a transcriptional circuit linking intestinal lipid metabolism to the neural control of behavioral immunity, expanding our understanding of how metabolic state influences host defense.

Fig 7. NHR-68 integrates lipid metabolism with behavioral and molecular immunity.

NHR-68 acts in the intestine to regulate linoleic acid homeostasis through control of fat-3 expression. Changes in intestinal lipid metabolism influence an AWC-dependent pathogen-avoidance circuit through currently unidentified intestine-to-neuron signaling mechanisms, enhancing survival during P. aeruginosa infection. In parallel, NHR-68 suppresses PMK-1/p38 MAPK and DAF-16/FOXO immune pathways, coordinating behavioral and molecular defenses to optimize host protection. Created in BioRender. Sang, Y. (2026). https://BioRender.com/pdbx63j.

By regulating PUFA levels, NHR-68 modulates the organism’s ability to mount behavioral responses to pathogens. Fatty acids, particularly PUFAs, are integral to the regulation of various biological processes, including cellular signaling, immune modulation, and membrane dynamics [28,49]. Our data reveal that alterations in intestinal PUFA metabolism directly affect the ability of C. elegans to detect and avoid pathogenic bacteria. This finding is consistent with prior studies showing that metabolic changes can influence behavioral responses to environmental stressors, including infection [50]. Although fat-3 expression is elevated in nhr-68(gk708) animals, it remains unclear whether NHR-68 regulates fat-3 through direct promoter binding or indirectly through additional transcriptional regulators. Future studies will be required to define the molecular mechanisms underlying this regulation.

We demonstrate that AWC neurons are required for both NHR-68-regulated and LA-enhanced avoidance behavior. However, our data do not identify the molecular signal that links intestinal lipid metabolism to AWC-dependent behavioral responses. While LA supplementation and fatty-acid desaturase perturbations strongly implicate intestinal PUFA homeostasis in this process, future studies will be required to determine whether the relevant signal is LA itself, a lipid-derived metabolite, a neuroendocrine factor, or another intestine-to-neuron signaling mechanism. This connection supports the idea that neural circuits can act as interfaces between physiological states and immune responses. Recent studies have further highlighted the importance of peripheral-to-neuron signaling in coordinating metabolism and behavior. Emerging evidence suggests that metabolic and physiological states in peripheral tissues, including the intestine, can modulate neuronal activity and behavioral outputs through diverse signaling mechanisms [51–53]. In particular, lipid-derived signals and metabolic intermediates have been shown to influence sensory neuron function and behavioral decision-making [54]. These findings provide a conceptual framework for our observations and support a model in which intestinal lipid metabolism, regulated by NHR-68, can shape AWC-dependent pathogen avoidance behavior through periphery-to-neuron communication. PUFAs have been shown to influence chemosensory and olfactory neuron function, including AWC-mediated adaption [45]. Previous work demonstrated that FAT-3 functions cell autonomously in AWC neurons to regulate olfactory adaptation [45], suggesting that PUFA metabolism can directly influence sensory neuron function. Our data further support an additional non-cell-autonomous mechanism in which changes in intestinal lipid metabolic state influence an AWC-dependent behavioral circuit through intestine-to-neuron communication.

In addition to modulating behavior, NHR-68 regulates the activation of the p38/PMK-1 and FOXO/DAF-16 pathways, two conserved regulators of innate immunity. The enhanced expression of immune effectors in nhr-68 mutants, coupled with the suppression of pathogen resistance by pmk-1 or daf-16 RNAi, indicates that NHR-68 functions upstream of these signaling modules. This regulatory role suggests that NHR-68 coordinates metabolic state with immune activation. Beyond PMK-1 and DAF-16, innate immune responses in C. elegans are also shaped by additional conserved pathways, including the insulin/IGF-1 signaling (IIS) axis and the SKN-1/Nrf2 oxidative stress response pathway [55,56], which together integrate metabolic and stress cues to fine-tune host defense. By tempering overactive immune responses, NHR-68 may help balance resource allocation during infection—a principle potentially conserved across species. Such a regulatory balance may prevent the detrimental effects of chronic immune activation, especially in nutrient-limited environments.

In exploring immune gene regulation, we found that NHR-10 and NHR-68 act additively to repress target genes. Our qRT-PCR and reporter analyses show that the double knockdown induces higher clec-60 expression than either single knockdown, demonstrating that both NHRs independently contribute to immune gene repression. This additive effect supports their joint contribution to the regulation of molecular and behavioral immune responses. This behavior likely reflects a context-dependent mechanism, where NHR-10 and NHR-68 respond to distinct upstream signals and may target partially overlapping genes with different transcriptional cofactors, highlighting the versatility of nuclear hormone receptor networks in coordinating immune responses.

NHR-10 and NHR-68 have been previously described as a transcriptional feed-forward module in the context of vitamin B12 and propionate metabolism, where NHR-10 acts upstream of NHR-68 to form an AND-logic regulatory circuit required for metabolic gene regulation [36]. In this study, we extend the functional relevance of this conserved module to immune regulation. NHR-10 and NHR-68 jointly regulate pathogen avoidance behavior and immune gene clec-60 expression. Notably, although NHR-10 functions upstream of NHR-68 at the transcriptional level, their effects on immune outputs are additive, suggesting that feed-forward regulatory wiring can generate graded or combinatorial transcriptional outputs depending on physiological context. This framework provides a mechanism by which conserved nuclear receptor modules integrate metabolic signals and immune regulation, while allowing flexible tuning of downstream functional responses such as behavioral immunity and pathogen resistance.

In addition to NHR-10 and NHR-68, our RNAi screen also identified several other NHRs (e.g., nhr-77, nhr-130, nhr-222, nhr-162, nhr-42, nhr-140, and nhr-54) whose knockdown enhanced immune reporter expression. The majority of these receptors remain poorly characterized at the functional level. However, as members of the expanded nuclear hormone receptor family in C. elegans, they are likely to participate in broader metabolic and stress-responsive regulatory networks. Their identification in our screen raises the possibility that additional, previously uncharacterized NHRs may contribute to coordinating metabolic state with intestinal immune responses. While genome-wide transcriptomic comparison between nhr-10 and nhr-68 mutants remains an interesting direction for future studies, our current genetic and targeted transcriptional analyses indicate that NHR-10 and NHR-68 have partially overlapping roles in coordinating metabolic state and immune regulation.

In wild-type animals, NHR-10 and NHR-68 likely function as metabolic sensors, responding to propionate accumulation and vitamin B12 availability. While their specific ligands remain unidentified, their behavior is consistent with ligand-dependent transcription factors. Upon infection, metabolic stress or altered nutrient absorption could modulate NHR activity, potentially shifting the balance of propionate detoxification and lipid metabolism. Further studies are needed to determine whether infection directly influences NHR-10/NHR-68 activity or ligand availability.

This intersection between lipid metabolism, immune regulation, and behavior is especially notable given the well-established role of PUFAs in inflammation and immunity [57,58]. Previous studies have also shown that dietary PUFA supplementation can reshape fatty-acid composition and influence host physiology in C. elegans, further supporting the idea that changes in PUFA availability can have broad functional consequences [30,59]. While PUFAs have been studied in the context of cytokine signaling and membrane composition, our findings highlight their importance in shaping behavioral immune responses. That avoidance behavior can be modulated by a defined metabolic product and a transcriptional regulator opens new avenues for dissecting how host physiology integrates metabolic cues to tailor immune and behavioral defenses.

The gut-brain axis has gained increasing attention in recent years as a crucial link between the microbiome, metabolism, and behavior [60–62]. Our results identify NHR-68 as a transcriptional regulator that couples intestinal lipid metabolism to immune and neural outputs in C. elegans. By tuning polyunsaturated-fatty-acid balance, NHR-68 shapes pathogen-avoidance behavior through AWC neurons while restraining PMK-1 and DAF-16-dependent immune activation. This coupling provides a mechanistic basis for coordinating behavioral and molecular defenses with the energetic demands of infection. Because nuclear receptors and lipid-derived signals are conserved, similar gut-to-brain mechanisms may integrate metabolic and immune states in other animals. Together, these findings reveal a general principle in which nuclear receptors align metabolic regulation with adaptive host-defense strategies.

Materials and methods

Strains

The C. elegans strains were cultured under standard conditions and fed E. coli OP50. All strains were maintained on nematode growth medium (NGM) seeded with E. coli (OP50). The C. elegans strains used include wild-type N2 Bristol, JIN810 agIs26 [clec-60p::gfp + myo-2p::mCherry], RB969 fat-2(ok873), VC1527 nhr-68(gk708), MGH171 alxIs9 [vha-6p::sid-1::SL2::GFP] and MAH677 sid-1(qt9) V; sqIs71 [rgef-1p::GFP + rgef-1p::sid-1], PY7502 oyIs85 [ceh36p::TU#813 + ceh36p::TU#814 + srtx1p::GFP + unc122p::DsRed] [AWC(−)], JN1715 peIs1715 [str1p::mCasp-1 + unc122p::GFP] AWB(−), PY7505 oyIs84 [gpa-4p::TU#813 + gcy-27p::GFP + unc-122p::DsRed] [ASI(−)], PR680 che-1(p680) [ASE(−)] were obtained from the Caenorhabditis Genetics Center (University of Minnesota, Minneapolis, MN). nhr-10(tm4695) was obtained from the National Bioresource Project (NBRP), Japan. Detailed information about the strains used is provided in S6 Table.

The following bacterial strains were used: Escherichia coli OP50, E. coli HT115(DE3), Pseudomonas aeruginosa PA14, Enterococcus faecalis OG1RF. The E. coli OP50, E. coli HT115(DE3), and P. aeruginosa PA14 cultures were grown at 37 °C in Luria–Bertani (LB) broth. The E. faecalis cultures were grown at 37 °C in brain heart infusion (BHI) broth.

Feeding RNAi NHR screen

RNA interference (RNAi) was employed to induce loss-of-function RNAi phenotypes by feeding nematodes with E. coli HT115(DE3) expressing double-stranded RNA (dsRNA) homologous to the indicated target gene [31,63]. E. coli with the appropriate vectors was grown in LB broth containing ampicillin (100 mg/mL) and tetracycline (12.5 mg/mL) at 37 °C overnight and seeded onto NGM plates containing 100 mg/mL ampicillin and 3 mM isopropyl b-D-thiogalactoside (IPTG) (RNAi plates). RNAi-expressing bacteria were allowed to grow overnight at 37 °C. Gravid adults were transferred to RNAi-expressing bacterial lawns and allowed to lay eggs for 2 hours. The gravid adults were removed, and the eggs were allowed to develop into young adults for downstream assays. unc-22 RNAi was included as a positive control to account for RNAi efficiency. All RNAi clones were obtained from the Ahringer RNAi library.

Fluorescence imaging

Animals were anesthetized in M9 salt solution containing 40 mM sodium azide and mounted onto 2% agar pads. The animals were then visualized using a Leica M165 FC fluorescence stereomicroscope. The GFP signal from eight animals per condition was quantified via Image J software. The fluorescence of an entire animal was calculated with the following equation: corrected whole animal fluorescence = integrated density − (area of selected animal × mean fluorescence of background readings).

C. elegans killing assays with Pseudomonas aeruginosa

The bacterial lawns used for the C. elegans killing assays, performed to assess resistance to P. aeruginosa, were prepared by spreading 20 µL of P. aeruginosa PA14 culture grown overnight at 37 °C on the full surface of modified NGM agar plates (0.35% peptone instead of 0.25% peptone) in 3.5-cm diameter plates. The plates were incubated at 37 °C for 16 hours and then cooled to room temperature for at least 1 hour before being seeded with synchronized young adult animals. The killing assays were performed at 25 °C, and live animals were transferred daily to fresh plates. Survival was evaluated at the times indicated; animals were considered dead if they failed to respond to touch. Each experiment was performed in triplicate (n = 90 animals).

P. aeruginosa avoidance assay

The bacterial lawns were prepared by inoculating individual bacterial colonies into 3 mL of LB broth as mentioned above and growing them for 24 hours on a shaker at 37 °C. Then, 20 µL of the culture was plated onto the center of a 3.5-cm modified NGM plate (3.5% instead of 2.5% peptone) and incubated at 37 °C for 24 hours. Thirty synchronized young gravid adult hermaphrodites grown on E. coli OP50 were transferred outside the bacterial lawns and incubated at 25 °C. Then, the number of animals on and off the lawns were counted at the indicated times for each experiment. The occupancy index was calculated as (N_on lawn/N_total). At least three independent experiments were performed, and three 3.5-cm plates were used per trial in each experiment.

Reversal assay

Young adult worms (N2, nhr-68(gk708), nhr-10(tm4695), and double mutants) were raised at 20 °C. Worms were first transferred to unseeded plates to remove residual bacteria and then placed on unseeded NGM plates. After 1-min acclimation, reversal events were counted over a 3-min period. A reversal was defined as at least two consecutive head bends during backward movement.

Locomotion analysis

Locomotion was assessed using an image-based assay in which C. elegans were transferred to plates containing a fresh bacterial lawn. A 15-s video was recorded immediately after transfer, and an image was captured at the 15-s time point. Movement distance was measured by tracing the worm’s movement track using NIH ImageJ, and locomotion was expressed as the ratio of movement distance to body length.

Choice assay with linoleic acid supplementation

A 10 mM linoleic acid (LA) stock was prepared by adding 1 µL of pure linoleic acid to 359 µL of 100% ethanol and vortexing thoroughly to ensure complete mixing. For the choice assay, a 200 µL bacterial suspension was prepared using 20× concentrated overnight E. coli or P. aeruginosa cultures, fatty-acid-free BSA (5 mg/mL stock), S basal or M9 buffer, and LA or vehicle control. For the LA condition, the mixture consisted of 20× concentrated bacteria (50%), BSA stock (50 mg/mL), buffer (29%), and LA stock (1%, final 100 µM). For the control condition, ethanol was added instead of LA at the same final concentration (1%), with all other components kept identical. Each mixture was vortexed briefly and equilibrated at room temperature before use. For the assay, 30 µL of each condition was spotted onto SK plates in symmetrical positions and allowed to dry prior to behavioral testing. Choice Index = (number of animals on E. coli + vehicle − number of animals on E. coli + LA)/(number of animals on E. coli + vehicle + number of animals on E. coli + LA). Choice indices for P. aeruginosa were calculated similarly.

Cloning and generation of transgenic C. elegans strains

For nhr-68 rescue, the recombinant plasmid pPD95.77_nhr-68p::nhr-68_SL2::gfp was constructed by cloning the nhr-68 gene and its 670-bp upstream sequence into the XbaI and SmaI sites of the pPD95.77_SL2::gfp vector [64]. For NHR-68 gut-specific rescue, the nhr-68-encoding sequence was amplified with the primers nhr-68_ATG_SaLI F and nhr-68_TAA_SmaI R. The amplified nhr-68 DNA was cloned under the vha-6 promoter in the plasmid pPD95.77_vha-6p_SL2 between the SalI and SmaI sites to generate the expression clone pPD95.77_vha-6p_nhr-68_SL2. Transgenic strains were created by injecting 25 ng/mL of the plasmids together with 50 ng/mL of the coinjection marker unc-122p::rfp.

C. elegans killing assays with Enterococcus faecalis

Twenty microliters of log-phase E. faecalis cultures were spread over the entire surface of 35 mm brain-heart infusion (BHI) agar plates, with or without 20 nM VB12. Plates were incubated overnight at 37 °C and then cooled to room temperature before seeding with synchronized young adult animals. Killing assays were performed at 25 °C, and live animals were transferred daily to fresh plates, with or without 20 nM VB12. Plates containing VB12 were protected from light. Survival was evaluated at the indicated time points; animals were considered dead if they failed to respond to touch. Each experiment was performed in triplicate (n = 90 animals).

RNA isolation and quantitative reverse transcription-PCR (qRT-PCR)

Young gravid adult wild-type N2 and nhr-68(gk708) animals were either infected for 12 h with P. aeruginosa PA14 or fed on E. coli OP50 at 25 °C. After the exposure period, animals were collected, washed three times with M9 buffer, and immediately frozen in QIAzol reagent (Qiagen, the Netherlands). Total RNA was extracted using the RNeasy Plus Universal Kit (Qiagen, the Netherlands). A total of 6 µg of total RNA was reverse transcribed with random primers using a high-capacity cDNA reverse transcription kit (Applied Biosystems, Foster City, CA).

Quantitative reverse transcription-PCR (qRT-PCR) was conducted using the Applied Biosystems one-step real-time PCR protocol with SYBR Green (Applied Biosystems) on a QuantStudio 5 Real-Time PCR System in 96-well plate format. Reaction mixtures of 25 µl were analyzed as outlined by the manufacturer (Applied Biosystems). The relative fold changes of the transcripts were calculated using the comparative cycle threshold (C_T) (2^−∆∆CT) method and normalized to pan-actin (act-1, act-3, and act-4). The cycle thresholds of the amplification were determined using QuantStudio 5 software (Applied Biosystems). All samples were run in triplicate. The primer sequences are listed in S7 Table.

RNA sequencing and gene expression analysis

RNA sequencing (RNA-seq) was performed for gene expression analysis and transcriptomic studies. Total RNA was isolated from wild-type N2 and nhr-68(gk708) animals infected for 12h with P. aeruginosa PA14 in three biological replicates. The sample integrity and purity were checked using an Agilent 2100 system. The RNA was sequenced on a HiSeq X sequencing platform with 150 bp paired-end reads. The RNA sequencing libraries were prepared via the NEBNextâ Ultra RNA Library Prep Kit for Illumina (NEB#E7530L). Library preparation and sequencing were performed at the Novogene Genomic Services & Solutions Company, USA.

The RNA sequence data were analyzed via Lasergene DNA star software. The RNA reads were aligned to the C. elegans genome (WS271) via the STAR aligner. Read counts were normalized for sequencing depth and RNA composition across all samples. Differential gene expression analysis was then performed on the normalized samples. Genes exhibiting at least a 2-fold change were considered differentially expressed. Gene enrichment of Gene Ontology (GO) biological processes was performed via the Database for Annotation, Visualization, and Integrated Discovery (DAVID) (https://david.ncifcrf.gov/). The overlap of the upregulated genes in previously defined pathways and gene sets, including DAF-16-, PMK-1-, and SKN-1-regulated genes, was calculated. The statistical significance of the overlap between two gene sets was calculated with nemates.org/MA/progs/overlap_stats.html. The representation factor represents the number of overlapping genes divided by the expected number of overlapping genes drawn from 2 independent groups.

Extraction of metabolites for GC–MS

C. elegans pellets of 2000 animals were lyophilized, and 1.0 mg of biomass was weighed and transferred to a microcentrifuge tube. The metabolites were extracted via the MPLEx protocol [65], in brief, by the addition of 470 µl of a mixture of methanol:water (1.0:1.4), followed by homogenization with a pellet pestle and the addition of 530 µl of ice-cold chloroform. The samples were vortexed for 1 min, placed in an ice block for 5 min, and vortexed again. Centrifugation was performed at 4 °C for 10 min at 7,500g. The upper and lower layers containing polar and nonpolar metabolites were collected and transferred to glass vials, dried under vacuum, and stored at −20 °C. For metabolomic analysis, aex-5 RNAi animals were grown on E. coli HT115 and collected without P. aeruginosa exposure.

Derivatization and GC–MS acquisition

The vials containing the dried metabolite extracts were stored at −20 °C and dried in vacuo for an additional 30 min immediately prior to derivatization to ensure the complete removal of residual moisture. The dried extracts were chemically derivatized by adding 20 µL of a 30 mg/ml methoxamine hydrochloride solution in pyridine, followed by vortexing and sonication for 30 s. The samples were then incubated for 90 min in a thermomixer (Eppendorf) at 37 °C with shaking at 1,000 rpm. After this step was completed, the samples were silylated by adding 80 µL of N-methyl-N-trimethylsilyl-trifluoroacetamide (MSTFA) with 1% trimethylchlorosilane and incubation in a thermomixer at 37 °C for 30 min with shaking at 1,000 rpm.

The samples were analyzed with an Agilent GC 7890A instrument equipped with an HP-5MS column (30 m × 0.25 mm × 0.25 μm; Agilent Technologies, Santa Clara, CA) coupled with a single quadrupole MSD 5975C (Agilent Technologies). The helium gas flow rate was determined by the Agilent Retention Time Locking (RTL) function on the basis of analysis of deuterated myristic acid (Agilent Technologies, Santa Clara, CA). One microliter of the derivatized sample was injected into a splitless port at a constant temperature of 250 °C. The GC temperature gradient started at 60 °C and was held at that temperature for 1 min after injection, followed by an increase to 325 °C at a rate of 10 °C/min and a 10-min hold at this temperature. The transfer line and quadrupole temperatures were 280 °C and 150 °C, respectively. The electron impact was set to 70 eV, and mass spectra were collected at a rate of 2.6 Hz over a scanning range of m/z 50–600. A fatty acid methyl ester (FAME) mixture (C8-28) (Sigma–Aldrich) was analyzed with each batch as a standard for retention time calibration.

GC–MS data analysis

Raw GC–MS data files were processed via Metabolite Detector software [66]. The retention indices (RIs) of the detected metabolites were calculated on the basis of the analysis of the FAME standards, followed by chromatographic alignment across all analyses after deconvolution. Metabolites were identified by matching experimental spectra to a PNNL-augmented version of the Agilent GC–MS Metabolomics Library, which contains mass spectra and validated RIs for over 1,200 metabolites. All metabolite identifications and quantification ions were manually confirmed to reduce deconvolution errors during automated data processing and to eliminate false identifications. The NIST20 and Wiley11 spectral libraries were also used to cross-validate the spectral matching scores obtained using the PNNL augmented library. The unknown peaks were subjected to further matching to the NIST20 and Wiley11 GC–MS libraries and reported as tentative identifications (denoted by an asterisk) if the MS score was high. The reported metabolite peak areas were normalized to the biomass used for extraction. A heatmap was generated using MetaboAnalyst 5.0 (https://www.metaboanalyst.ca/MetaboAnalyst/).

Analysis of fatty acids by LC–HRMS

Worm samples were homogenized in 300 µL of PBS. From each homogenate, 250 µL was mixed with 10 µL of internal standard (oleic acid-d17, 10 µg/mL) and saponified, followed by extraction with 4 mL of hexane. Extracts were vortexed for 10 min and centrifuged at 4,000g for 10 min at 4 °C. Supernatants were transferred to clean glass tubes and evaporated to dryness under nitrogen. Dried samples were reconstituted in 150 µL of isopropanol, and 5 µL was injected into a Thermo Vanquish LC system equipped with a Kinetex C18 column (2.1 × 100 mm, 2.6 µm). Mobile phase A was 0.1% formic acid in water, and mobile phase B was 80:20 acetonitrile:isopropanol. The flow rate was 300 µL/min at 35 °C, and fatty acids were separated over a 25-min gradient.

Data were acquired on a Thermo Orbitrap Lumos mass spectrometer in negative ESI mode at a resolution of 120,000 (scan range 100–500 m/z) and processed using Skyline-daily software.

Fatty acid treatment

Synchronized young adult animals were collected from NGM plates, washed twice with S basal buffer, and transferred to 6 cm dishes containing 5.8 mL of S basal supplemented with 0.02% NP-40 and 180 µL of an overnight E. coli OP50 culture. Linoleic acid (LA) or vaccenic acid (VA) was added to a final concentration of 100 µM from 100 mM ethanol stocks pre-mixed with 0.1% fatty acid-free bovine serum albumin (BSA) to enhance solubility and bioavailability [67]. Cultures were gently shaken at 60 rpm for 4 h at room temperature and protected from light. After incubation, animals were washed three times with S basal to remove residual fatty acids before subsequent behavioral assays. Control treatments contained equivalent amounts of ethanol and 0.1% BSA.

Quantification and statistical analysis

Statistical analysis was performed with GraphPad Prism 8 version 8.1.2 (GraphPad). All error bars represent the standard deviation (SD). Two-sample t tests were used for comparisons between two groups, while one-way ANOVA followed by Tukey’s multiple-comparison test was applied for comparisons among three or more groups. P values < 0.05 were considered statistically significant. In the figures, asterisks (*) denote statistical significance as follows: ns, not significant; * p < 0.05; ** p < 0.01; *** p < 0.001, compared with the appropriate controls. The Kaplan–Meier method was used to calculate the survival fractions, and the statistical significance of differences between survival curves was determined via the log-rank test. All experiments were performed at least three times.

Supporting information

S1 Appendix. Supplementary Figures A–J.

(DOCX)

S1 Table. Fluorescence Quantification of Screened NHR Genes.

(XLSX)

S2 Table. Upregulated and downregulated genes more than 2-fold (p < 0.05) in non-infected and P. aeruginosa-infected nhr-68(gk708) versus WT animals.

(XLSX)

S3 Table. GO enrichment analysis.

(XLSX)

S4 Table. Representation factors for upregulated genes in P. aeruginosa-infected nhr-68(gk708) animals.

(XLSX)

S5 Table. Metabolomic profiling and lipid quantification in NHR-68 pathway.

(XLSX)

S6 Table. Strains.

(XLSX)

S7 Table. Primers.

(XLSX)

S1 Data. Raw data.

(XLSX)

Abbreviations

BHI

brain-heart infusion

BSA

bovine serum albumin

CGC

Caenorhabditis Genetics Center

DAVID

Database for Annotation, Visualization, and Integrated Discovery

EPA

eicosapentaenoic acid

FAME

fatty acid methyl ester

GO

Gene Ontology

IIS

insulin/IGF-1 signaling

LA

linoleic acid

LB

Luria–Bertani

MUFAs

monounsaturated fatty acids

NBRP

National BioResource Project

NGM

nematode growth medium retention indices

NHRs

Nuclear hormone receptors

PCA

Principal component analysis

PUFA

polyunsaturated fatty acid

RIs

retention indices

RTL

Retention Time Locking

SD

standard deviation

VA

vaccenic acid

Data Availability

The raw RNA-seq data have been deposited in the NCBI Gene Expression Omnibus (GEO) database under accession number GSE344153. The raw GC-MS metabolomics data have been deposited in MassIVE under accession number: MSV000102834. The raw LC-HRMS metabolomics data have been deposited in the Metabolomics Workbench under study ID:ST005144. No custom code was generated for this study. All other underlying data for this study are available within the paper and its Supporting information files.

Funding Statement

This work was supported by the National Institutes of Health (GM070977 and AI117911 to AA). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.