Abstract
Animals integrate internal states to guide survival-critical decisions, but whether and how intestinal bacteria influence this process by interacting with host metabolic cues remains unclear. Here we show that the intestinal pathogen P. aeruginosa overrides host decision-making in fasted C. elegans by modulating central serotonin (5-HT) signaling. Fasting promotes risk-taking by activating an intestinal energy-sensing pathway that induces the chemoreceptor SRI-36 in ADF 5-HT neurons, sensitizing ADF to food odors and triggering moderate 5-HT release that drives food attraction despite risk. In contrast, intestinal P. aeruginosa reverses this strategy by activating a distinct immune-brain axis that further amplifies SRI-36 expression and ADF sensitivity, leading to excessive 5-HT release that suppresses food attraction and prioritizes safety. These findings reveal a gut-to-brain mechanism by which metabolic and immune signals converge on central 5-HT to reshape behavioral strategies.
Similar content being viewed by others
Introduction
To survive in dynamic environments filled with both threats and rewards, animals must detect and integrate external sensory cues to make adaptive behavioral choices1,2. This process, known as multisensory threat-reward decision-making, is strongly shaped by internal states, which optimize behavioral responses to meet survival needs. For example, hunger often biases decisions toward risk-taking, prioritizing food acquisition despite potential threats3,4,5,6. One effective way internal states influence decision-making is through sensory modulation. Feeding states can alter sensory neuron sensitivity, rapidly shifting sensory perception and behavioral preferences7,8,9. While such adaptive mechanisms promote survival, maladaptive changes in sensory processing and impaired threat-reward decision-making are hallmarks of neuropsychiatric disorders such as autism, schizophrenia, anxiety, and depression2,10,11. Yet, how internal states regulate sensory processing to guide decision-making remains poorly understood.
Among internal cues, intestinal bacteria are increasingly recognized as potent modulators of brain function and behavior12,13. Intestinal pathogens, in particular, can elicit profound neural changes, including shifts in sensory preferences and the development of anhedonia14,15, a core symptom of neuropsychiatric disorders characterized by the inability to experience reward16. Dysbiosis of the gut microbiota is associated with multiple neuropsychiatric conditions involving impaired decision-making13,17,18. Despite these associations, the mechanisms by which intestinal bacteria influence host decision-making remain elusive. In particular, whether they interact with metabolic states to regulate sensory processing and behavioral choices is unknown.
The nematode C. elegans provides a powerful genetic model for dissecting host-microbe interactions and their impact on neural function and behavior19,20. Its simple intestinal anatomy and compact, well-defined nervous system facilitate mechanistic studies using monoxenic bacterial cultures19,21. C. elegans also exhibits multisensory decision-making behaviors22. For example, when facing a hyperosmotic glycerol barrier placed before the attractive food odor diacetyl, worms must decide whether to risk crossing the potentially lethal barrier to reach the reward6,23. This simple form of decision-making is state-dependent, as hunger enhances risk-taking6,24. Glycerol and diacetyl are detected by the primary sensory neurons ASH and AWA, respectively6,25, and downstream interneurons integrate these inputs to drive behavior26. However, whether intestinal pathogens alter these processes and how they may interact with metabolic states to regulate behavioral choices remains unknown.
Here, we show that the intestinal pathogen P. aeruginosa overrides hunger-driven decision-making in C. elegans by modulating central serotonin (5-HT) signaling through an immune-brain axis. Mechanistically, fasting promotes risk-taking by activating the intestinal energy sensor AAK-2/AMPK, which promotes DAF-16/FOXO-dependent release of the insulin-like peptide (ILP) INS-37 from intestinal epithelial cells (IECs). INS-37 then engages the canonical DAF-2-DAF-16 insulin/IGF-1 signaling (IIS) cascade in ADF 5-HT neurons. This FOXO-to-FOXO signaling induces expression of the chemoreceptor SRI-36, conferring diacetyl sensitivity to ADF, which are otherwise insensitive in the well-fed state. Diacetyl-evoked ADF activation triggers moderate 5-HT release that inhibits AIB and activates AUA interneurons via the 5-HT receptors SER-4 and LGC-50, respectively, promoting diacetyl attraction despite the associated risk. In contrast, intestinal P. aeruginosa subverts this hunger-driven decision-making by activating the intestinal PMK-2/p38 MAPK immune pathway, which induces a distinct ILP, INS-7, to further amplify SRI-36 expression and ADF responsiveness through the same FOXO-to-FOXO axis. This leads to excessive 5-HT release that predominantly inhibits AIA and AIY interneurons via the 5-HT receptor MOD-1, suppressing diacetyl attraction and prioritizing safety.
Together, these findings establish P. aeruginosa as an intestinal pathogen capable of overriding host decision-making through immune modulation of central 5-HT signaling, and identify central 5-HT as a convergence target for intestinal metabolic and immune cues. Our study provides a mechanistic framework for how intestinal microbes and internal physiological states dynamically modulate brain function and behavioral decisions.
Results
Intestinal pathogen P. aeruginosa reverses hunger-driven decision-making via ADF 5-HT neurons in C. elegans
To investigate how intestinal pathogens influence host decision-making, we employed a multisensory behavioral choice assay in C. elegans6,23. In this assay, worms are attracted to the food odor diacetyl but must decide whether to cross a hyperosmotic glycerol barrier to reach it (Supplementary Fig. 1a). We tested three paradigmatic human pathogenic bacteria, Pseudomonas aeruginosa (P. aeruginosa), Staphylococcus aureus (S. aureus), and enteropathogenic E. coli (EPEC). However, none of these pathogens altered the behavioral choice under well-fed conditions, as escape percentages remained comparable to those of control worms fed nonpathogenic E. coli (Supplementary Fig. 1b).
Because hunger is known to promote risk-taking3,4,5,6, we adapted the assay by fasting worms before the choice test (Fig. 1a). Fasting increased escape probability in a time-dependent manner across all glycerol concentrations tested, with the strongest effect observed after 6 h of fasting at 3 M glycerol (Supplementary Fig. 1c), which was selected for subsequent experiments. Under these conditions, P. aeruginosa infection significantly reduced escape probability compared to E. coli, returning it to levels similar to well-fed controls, whereas S. aureus and EPEC had no effect (Fig. 1b). In contrast, P. aeruginosa did not affect unisensory decision-making, as chemotaxis toward diacetyl (in the absence of glycerol) and escape from glycerol (in the absence of diacetyl) remained unchanged (Supplementary Fig. 1d, e). Locomotor activity was also unaffected (Supplementary Fig. 1f). These results suggest that P. aeruginosa selectively reverses hunger-driven risky decision-making in C. elegans. Similar escape suppression occurred after 4 or 24 h of infection (Supplementary Fig. 1g), with 4 h chosen for subsequent experiments.
a Schematic of the multisensory behavioral choice assay for fasted C. elegans. b Escape percentages of well-fed or fasted worms. Well-fed worms were maintained on nonpathogenic E. coli, and fasted worms were fed E. coli or pathogenic bacteria (P. aeruginosa, S. aureus, or EPEC) before fasting. p = 0, 0, 0, and <0.0001. c Escape percentages of fasted worms fed E. coli, P. aeruginosa, or P. aeruginosa components, including supernatant, odors, heat-killed cells, the nonpathogenic gacA(−) mutant, or live cells only. p = 0, 0.953, 0.9809, 0.9997, 0.981, and 0. d Escape percentages of fasted worms fed E. coli or P. aeruginosa with 0.35% peptone to enhance virulence. p = 0, 0.0365, and 0. e Intestinal colonization of PA14-dsRed in the indicated genotypes. Shown are representative images and comparisons of relative fluorescence normalized to wild type (WT). Intestine-specific RNAi was performed using Pges-1. Scale bar, 100 μm. p < 0.0001. f–i Escape percentages of fasted (fed E. coli before fasting) or P. aeruginosa-infected (fed P. aeruginosa before fasting) worms with the indicated genotypes. NSM- and ADF-specific rescue was performed using Pceh-2 and Psrh-142, respectively. p = 0.5963 and <0.0001 (f), 0.6745 and 0.7903 (g), 0, 0, 0.948, <0.0001, <0.0001, and 0.9677 (h), and 0.0035 and 0.0036 (i). j–l Escape percentages of fasted or P. aeruginosa-infected worms with or without 5 mM histamine (Hist) treatment. HisCl1 was expressed in ADF (j), NSM (k), and HSN (l) using Psrh-142, Pceh-2, and Pclh-3, respectively. p = 0.9117, 0.9914, <0.0001, 0.9691, 0.9791, and <0.0001 (j), 0.9691, 0.9156, 0.7535, 0.9251, 1, and 0.9661 (k), and 0.9887, 0.9979, 0.9977, 0.9894, 0.9928, and 1 (l). *p < 0.05, **p < 0.01, and ***p < 0.001 (one-way ANOVA with Tukey’s post hoc test for b–d, h, j–l; two-sided unpaired t-test for e–g, i). Brackets indicate the number of independent assays (b–d, f–l) or animals tested (e). Data are shown as means ± SEM. Source data are provided as a Source Data file.
Previous studies have suggested that P. aeruginosa metabolites may modulate C. elegans behavior via chemosensory neurons27,28. However, exposure to P. aeruginosa supernatant, odors, heat-killed cells, or a nonpathogenic mutant gacA(−) did not affect fasting-induced escape (Fig. 1c). In contrast, infection with live cells alone was sufficient (Fig. 1c), and enhancing virulence by supplementing the medium with 0.35% peptone29 increased P. aeruginosa colonization and escape suppression (Fig. 1d and Supplementary Fig. 1h). Knockdown of nol-6, which reduces intestinal colonization of P. aeruginosa30 (Fig. 1e), attenuated its effect on escape (Fig. 1f). Consistent with the similar escape behavior observed at 4 h and 24 h of infection, intestinal P. aeruginosa colonization did not differ significantly between these time points (Supplementary Fig. 1g, i). These results suggest that intestinal colonization of virulent P. aeruginosa is required to suppress hunger-driven decision-making.
5-HT and dopamine are key neuromodulators that regulate cognitive processes, including threat and reward processing22,31. Loss of the dopamine biosynthetic gene cat-232 had no effect on escape in the fasted or infected state (Fig. 1g). However, loss of the 5-HT biosynthetic gene tph-132 reduced fasting-induced escape and reversed the P. aeruginosa-induced suppression of escape (Fig. 1h). Additionally, mod-5(n822) mutants, which have elevated extracellular 5-HT due to impaired reuptake33, showed exaggerated behavioral responses to both hunger and infection (Fig. 1i). These results suggest that both hunger and P. aeruginosa regulate decision-making through 5-HT signaling.
To identify the relevant 5-HT neurons, we silenced tph-1-expressing neurons NSM, ADF, or HSN34 using the histamine-gated chloride channel HisCl135. Silencing NSM or HSN had no effect (Fig. 1k, l), but silencing ADF significantly impaired both fasting-induced escape and P. aeruginosa-mediated suppression (Fig. 1j). Consistently, expressing wild-type tph-1 in ADF, but not NSM, fully rescued the effects of both fasting and P. aeruginosa (Fig. 1h). Together, these results suggest that hunger promotes risky decision-making, while P. aeruginosa overrides this decision, both via ADF 5-HT neurons.
In addition to 5-HT, ADF neurons are also cholinergic36. However, ADF-specific unc-17 RNAi, which reduces vesicular acetylcholine release37, did not significantly affect escape in either the fasted or infected state (Supplementary Fig. 1j). These results suggest that acetylcholine release from ADF is not required for the state-dependent regulation of decision-making, although it may contribute to other ADF-mediated functions.
ADF neurons act as primary olfactory sensory neurons for diacetyl in fasted and infected states
We next investigate how hunger and P. aeruginosa infection influence decision-making via 5-HT signaling. ASH and AWA sensory neurons, which detect glycerol and diacetyl respectively6,25, form direct synaptic connections with ADF38 (Supplementary Fig. 2a). We hypothesized that changes in these upstream neurons might influence ADF responses under different internal states. Indeed, fasting reduced glycerol-evoked calcium responses in ASH and enhanced diacetyl-evoked responses in AWA (Supplementary Fig. 2b, c). However, P. aeruginosa infection did not further alter ASH or AWA responses in fasted worms (Supplementary Fig. 2b, c). Silencing AWA with HisCl1 reduced escape in both well-fed and fasted states, consistent with its role in sensing diacetyl, but fasted worms still exhibited substantial escape probability compared to well-fed controls (Supplementary Fig. 2d). These results suggest that while ASH and AWA sensory responses are modulated by hunger, they are not the primary sites through which P. aeruginosa infection overrides hunger-driven decision-making.
ADF neurons are 5-HT sensory neurons that respond to noxious stimuli in well-fed worms39 and to food odors under dietary restriction40. In a transgenic strain expressing GCaMP6 in ADF, which exhibited normal escape behavior compared to wild type (Supplementary Fig. 2e), ADF showed no response to glycerol but were robustly activated by diacetyl (Fig. 2a, b), with responses enhanced by fasting and further amplified by P. aeruginosa infection (Fig. 2b). In unc-13(e51) mutants, which are defective in neurotransmitter release41, diacetyl-evoked ADF responses were absent in well-fed worms but persisted in fasted and infected states, albeit at a reduced amplitude (Fig. 2c). In contrast, unc-31(e169) mutants, which are defective in neuropeptide release42,43, showed no such defect (Fig. 2c), Moreover, in odr-7(ky4) mutants, where AWA neurons were eliminated44, and in odr-7(ky4);unc-13(e51) double mutants, ADF responses to diacetyl were abolished in the well-fed state but retained in both fasted and infected states (Fig. 2d). These results suggest that while ADF neurons receive synaptic input from AWA in response to diacetyl stimulation, they function as primary olfactory neurons for diacetyl during fasting and P. aeruginosa infection, independent of synaptic input.
Glycerol-evoked (a) and diacetyl-evoked (b) GCaMP6 responses in ADF neurons of well-fed, fasted, or P. aeruginosa-infected worms. Left, averaged GCaMP6 signals (solid lines, mean; shaded regions, SEM). Right, comparisons of mean responses during the 30-s stimulation period. GCaMP6 was expressed in ADF using Psrh-142. p = 0.9152 and 0.7084 (a), and 0.007, 0.0001, and 0 (b). c–e Diacetyl-evoked GCaMP6 responses in ADF neurons of worms with the indicated genotypes and conditions. ADF-specific rescue was performed using Psrh-142. WT wild type. p = <0.0001, 0.9999, 0.0001, 0.7778, 0.0043, and 0.8963 (c), 0.0898, 0.8251, and 0.9601 (d), and <0.0001, 1, <0.0001, 0.9989, <0.0001, 0.9943, 0, 1, 0, 0.999, 0, and 0.9989 (e). **p < 0.01 and ***p < 0.001 (one-way ANOVA with Tukey’s post hoc test for (a–c, e); two-sided unpaired t-test for (d). Brackets indicate the number of animals tested. Data are shown as means ± SEM. Source data are provided as a Source Data file.
Olfactory detection in C. elegans is mediated by G-protein-coupled receptors (GPCRs) and downstream transduction channels45. ODR-3/Gα mediates food odor sensation in ADF neurons40, and the TRPV channel subunits OCR-2 and OSM-9, which are coexpressed in ADF, likely function downstream of GPCR signaling45. In odr-7(ky4) mutants, mutations in odr-3, ocr-2, or osm-9 abolished ADF responses to diacetyl in both fasted and infected worms, and ADF-specific expression of the corresponding wild-type genes rescued the defect (Fig. 2e). These results suggest that an unidentified GPCR, acting upstream of the ODR-3-OCR-2/OSM-9 signaling pathway, mediates ADF responses to diacetyl during fasting and P. aeruginosa infection.
Hunger- and infection-induced SRI-36 acts as a diacetyl receptor in ADF neurons
To identify the GPCR responsible for ADF diacetyl sensitivity, we first tested ODR-10, the only known diacetyl receptor, which is exclusively expressed in AWA neurons of well-fed worms46. However, ADF responses to diacetyl remained comparable between unc-13(e51) and unc-13(e51);odr-10(ky32) mutants across all states (Fig. 3a), indicating that ODR-10 is not required for ADF sensitivity in the absence of synaptic input.
a Diacetyl-evoked GCaMP6 responses in ADF neurons of unc-13(e51) or unc-13(e51);odr-10(ky32) mutants under the indicated conditions. Left, averaged GCaMP6 signals (solid lines, mean; shaded regions, SEM). Right, comparisons of mean responses during the 30-s stimulation period. p = 0.4287, 0.9069, and 0.9945. b Transcriptomic analysis of published RNA-sequencing datasets. A Venn diagram shows 126 putative chemosensory GPCR genes upregulated during fasting and 1667 genes upregulated during P. aeruginosa infection. Four genes were common: sri-36, srr-6, str-163, and srap-1. c–e Diacetyl-evoked GCaMP6 responses in ADF neurons of worms with the indicated genotypes and conditions. ADF-specific RNAi and rescue were performed using Psrh-142. p = 0, 0, 1, 0.9985, and 0.9662 (c), 0.0001 (d), and <0.0001, 1, 0.9906, 0, 0, and 0.9945 (e). f Escape percentages of fasted or P. aeruginosa-infected worms with the indicated genotypes. WT wild type. p = <0.0001, 0.9903, 0.8333, 0, <0.0001, <0.0001, 0.5497, and 0. g Psri-36::GFP expression in well-fed, fasted, or P. aeruginosa-infected worms. Shown are representative images and comparisons of GFP fluorescence intensity. ADF neurons were labeled with Psrh-142::NLS::mCherry. AU arbitrary units. Scale bar, 20 μm. p = <0.0001, 0.0018, and 0. h Subcellular localization of the SRI-36::GFP translational fusion protein in ADF neurons. The inset shows GFP signals in the ADF cilium (region with dashed outline). Scale bar, 20 μm. A similar pattern was observed in 16 animals. **p < 0.01 and ***p < 0.001 (one-way ANOVA with Tukey’s post hoc test for c, e–g; two-sided unpaired t-test for a, d). Brackets indicate the number of animals tested (a, c–e, g) or independent assays (f). Data are shown as means ± SEM. Source data are provided as a Source Data file.
We next analyzed published RNA-sequencing datasets for chemosensory GPCRs upregulated during fasting or P. aeruginosa infection. We identified 126 candidate GPCRs induced by fasting (Fasted vs. Well-fed)47 and 1,667 genes upregulated by P. aeruginosa exposure (P. aeruginosa vs. E. coli)48, with four GPCRs, sri-36, srr-6, str-163, and srap-1, shared between comparisons (Fig. 3b). Screening these genes in P. aeruginosa-infected odr-7(ky4) mutants revealed that ADF-specific RNAi of sri-36, but not the other candidates, abolished ADF calcium responses to diacetyl (Fig. 3c). This effect was confirmed in fasted odr-7(ky4) mutants (Fig. 3d).
To further validate the role of SRI-36, we generated a CRISPR deletion allele, sri-36(plc921) (Supplementary Fig. 3a), which eliminated ADF responses to diacetyl in both fasted and infected odr-7(ky4) mutants (Fig. 3e). This defect was rescued by ADF-specific expression of wild-type sri-36 (Fig. 3e). The deletion also impaired escape in both fasted and infected worms, which was rescued by ADF-specific expression of wild-type sri-36 and recapitulated by ADF-specific RNAi (Fig. 3f). A Psri-36::GFP transcriptional reporter was undetectable in well-fed worms but strongly induced in ADF during fasting and further enhanced by P. aeruginosa infection (Fig. 3g and Supplementary Fig. 3b). Consistently, a CRISPR knock-in reporter, SRI-36::mNeonGreen, with mNeonGreen inserted at the C-terminus of the endogenous locus, showed a similar ADF-specific induction pattern (Supplementary Fig. 3c). To test whether receptor abundance modulates ADF sensitivity, we expressed SRI-36 at two different levels in sri-36(plc921) mutants by injecting 5 ng/µl (low) or 75 ng/µl (high) of Psrh-142::sri-36::mStrawberry plasmid (Supplementary Fig. 3d). In the fasted state, low-level expression restored ADF diacetyl responses and escape to control levels (Fig. 3e, f). In contrast, high-level expression enhanced ADF diacetyl responses and suppressed escape, phenocopying the behavioral pattern observed during P. aeruginosa infection (Fig. 3e, f). These results suggest that SRI-36 is induced in ADF by hunger and infection and is required for diacetyl sensation, and that its expression level tunes ADF sensitivity and behavioral output.
To test whether SRI-36 functions as a diacetyl receptor, we performed multiple receptor validation assays29,49. First, both an SRI-36::GFP translational fusion and the CRISPR-generated endogenous SRI-36::mNeonGreen fusion protein localized to the sensory cilia of ADF (Fig. 3h and Supplementary Fig. 3c), supporting its role in detecting environmental odors. Second, ectopic expression of sri-36 in ASH nociceptive neurons, which normally do not respond to the attractive food odor diacetyl, conferred robust diacetyl-evoked calcium responses in ASH (Supplementary Fig. 3e), demonstrating that SRI-36 is sufficient to mediate diacetyl sensitivity. Third, because ASH activation drives avoidance25, we assessed behavioral consequences using a two-choice chemotaxis quadrant assay29. Ectopic sri-36 expression in ASH significantly reduced diacetyl attraction (Supplementary Fig. 3f). Notably, in odr-7(ky4) mutants, ASH-expressed sri-36 even converted diacetyl attraction to aversion (Supplementary Fig. 3f), further supporting the role of SRI-36 in mediating diacetyl sensation. Together, these results establish SRI-36 as a diacetyl receptor or essential receptor subunit in ADF neurons.
Hunger- and infection-induced SRI-36 expression in ADF requires intestinal AAK-2/AMPK and PMK-2/p38 MAPK pathways
To investigate how fasting and P. aeruginosa infection regulate ADF diacetyl sensitivity, we examined the roles of the AAK-2/AMPK energy sensor and the PMK-1/p38 MAPK immune pathway. AMPK is a key metabolic sensor that responds to nutrient deprivation40,50. In C. elegans, AAK-2/AMPK phosphorylates and activates the transcription factor DAF-16/FOXO, driving DAF-16-dependent gene expression40,51. In fasted odr-7(ky4) mutants, aak-2 deletion abolished diacetyl-evoked calcium responses in ADF, which were restored by intestinal expression of wild-type aak-2 and recapitulated by intestine-specific RNAi of aak-2 (Fig. 4a). Similarly, intestine-specific RNAi of daf-16 in odr-7(ky4) mutants reduced ADF responses to levels comparable to odr-7(ky4);aak-2(ok524) mutants (Fig. 4a). Fasting-induced nuclear translocation of DAF-16 was also reduced by intestine-specific RNAi of aak-2 (Fig. 4b), whereas intestinal overexpression of a constitutively nuclear-localized DAF-16 (DAF-16aAM)52 restored ADF responses in fasted odr-7(ky4);aak-2(ok524) mutants (Fig. 4a). The intestine-specific RNAi of aak-2 and daf-16 produced similar results in the systemic RNAi-defective sid-1(qt9) mutant background53 (Fig. 4a, b), confirming tissue specificity. These results suggest that AAK-2 activates DAF-16 in IECs to drive ADF diacetyl sensitivity during fasting.
a, c Diacetyl-evoked GCaMP6 responses in ADF neurons of worms with the indicated genotypes and conditions. Left, averaged GCaMP6 signals (solid lines, mean; shaded regions, SEM). Right, comparisons of mean responses during the 30-s stimulation period. Intestine-specific expression and RNAi were performed using Pges-1. The systemic RNAi-defective sid-1(qt9) background was used to abolish dsRNA spreading and validate tissue specificity. p = 1, <0.0001, 0.3912, <0.0001, <0.0001, <0.0001, <0.0001, and 0.9025 (a), and 0.9979, <0.0001, <0.0001, <0.0001, 0, <0.0001, <0.0001, 1, <0.0001, <0.0001, 1, 0, 0, and <0.0001 (c). b Nuclear localization of DAF-16a/b::GFP in fasted or P. aeruginosa-infected worms with the indicated genotypes. Top, representative images. Bottom: comparisons of the percentage of GFP-labeled IEC nuclei per animal. WT wild type. Scale bar, 50 μm. p = 1, <0.0001, <0.0001, <0.0001, <0.0001, 0.9635, and 0.9635. Psri-36::GFP expression in fasted (d) or P. aeruginosa-infected (f) worms with the indicated genotypes. Shown are representative images and comparisons of fluorescence intensity. ADF neurons were labeled with Psrh-142::NLS::mCherry. AU arbitrary units. Scale bars, 20 μm. p = 0.9985, 0, 0, 0, and 0 (d), and 0.5826, <0.0001, <0.0001, and <0.0001 (f). Escape percentages of fasted (e) or P. aeruginosa-infected (g) worms with the indicated genotypes. p = 0.9807, 0.0013, <0.0001, 0.0001, 0.0008, <0.0001, and <0.0001 (e), and 0.9962, <0.0001, 0, 0, <0.0001, and <0.0001 (g). **p < 0.01 and ***p < 0.001 (one-way ANOVA with Tukey’s post hoc test). Brackets indicate the number of animals tested (a–d, f) or independent assays (e, g). Data are shown as means ± SEM. Source data are provided as a Source Data file.
We next examined the role of the PMK-1/p38 MAPK immune pathway, which cooperates with DAF-16 in epithelial immunity against P. aeruginosa48,54. Given that p38 MAPK can activate AKT and antagonize DAF-16 nuclear translocation55, we hypothesized that P. aeruginosa modulates ADF diacetyl sensitivity via p38-mediated inhibition of DAF-16. Indeed, in infected odr-7(ky4) mutants, ADF responses were abolished by mutations or intestine-specific RNAi of p38 MAPK pathway genes nsy-1 or sek-1 (Fig. 4c). Unexpectedly, deletion of pmk-1 had no effect (Fig. 4c). The C. elegans genome encodes three p38 homologs, PMK-1, PMK-2, and PMK-3. Intestine-specific RNAi of pmk-2 eliminated ADF responses in odr-7(ky4) mutants, whereas deletion of pmk-3 had no effect (Fig. 4c). As previously reported54, P. aeruginosa infection significantly inhibited DAF-16 nuclear translocation (Fig. 4b). This inhibition was reversed by intestine-specific RNAi of pmk-2 (Fig. 4b). Notably, intestinal DAF-16aAM overexpression suppressed ADF responses in odr-7(ky4) mutants, with no additive effect when combined with intestine-specific RNAi of pmk-2 (Fig. 4c). Comparable intestine-specific RNAi effects were observed in the sid-1(qt9) background (Fig. 4b, c), confirming tissue specificity. These results suggest that PMK-2 inhibits DAF-16 in IECs to suppress ADF diacetyl sensitivity during P. aeruginosa infection.
We then asked whether these pathways regulate sri-36 expression in ADF and influence decision-making. Intestine-specific RNAi of aak-2 or daf-16 similarly reduced Psri-36::GFP fluorescence and escape in fasted worms (Fig. 4d, e). Simultaneous RNAi of both genes produced no additive effect on escape (Fig. 4e). In infected worms, intestine-specific RNAi of pmk-2 or overexpression of DAF-16aAM similarly reduced Psri-36::GFP expression and impaired escape (Fig. 4f, g). RNAi of pmk-2 in the presence of DAF-16aAM also produced no additive effect on escape (Fig. 4g). Comparable intestine-specific RNAi effects were observed in the sid-1(qt9) background (Fig. 4d–g), confirming tissue specificity. These results suggest that both the AAK-2/AMPK energy-sensing pathway and the PMK-2/p38 MAPK immune pathway remotely regulate ADF sri-36 expression via DAF-16/FOXO from the intestine.
Intestine-released ILPs regulate SRI-36 expression via the IIS pathway in ADF
To investigate how intestinal signals regulate SRI-36 expression in ADF, we considered endocrine signaling, a conserved mechanism for gut-to-brain communication24,56. The C. elegans genome encodes three major neuropeptide families, FLPs, NLPs, and ILPs57. Intestine-specific RNAi of hid-1, a gene required for intestinal neuropeptide sorting and secretion58, or deletion of egl-3, encoding a proprotein convertase essential for neuropeptide maturation57, abolished ADF responses in both fasted and infected odr-7(ky4) mutants (Fig. 5a). Comparable intestine-specific RNAi effects were observed in the sid-1(qt9) background (Fig. 5a), confirming tissue specificity. Similarly, ADF-specific RNAi of daf-2, encoding the sole ILP receptor in C. elegans59, eliminated ADF responses (Fig. 5b). These results suggest that intestinal ILPs act via DAF-2 in ADF to promote diacetyl sensitivity.
a–d Diacetyl-evoked GCaMP6 responses in ADF neurons of worms with the indicated genotypes and conditions. Left, averaged GCaMP6 signals (solid lines, mean; shaded regions, SEM). Right, comparisons of mean responses during the 30-s stimulation period. Intestine- and ADF-specific RNAi were performed using Pges-1 and Psrh-142, respectively. The sid-1(qt9) background was used to validate tissue specificity. p = 0.8812, 0, 0, <0.0001, 0.8016, 0, 0, and 0 (a), <0.0001 and <0.0001 (b), 0.3917, <0.0001, and 0 (c), and 0.6576, 0, 1, 0, and 0 (d). e, f Pins-37::GFP expression in worms with the indicated genotypes and conditions. Shown are representative images and comparisons of fluorescence intensity. Intestine-specific RNAi was performed using Pges-1. WT wild type. AU arbitrary units. Scale bar, 100 μm. p = <0.0001, 0, and 0.7061 (e), and 0.8978, <0.0001, <0.0001, <0.0001, <0.0001, <0.0001, and <0.0001 (f). g Pins-7::GFP expression in fasted or P. aeruginosa-infected worms with the indicated genotypes. Shown are representative images and comparisons of fluorescence intensity. Scale bar, 100 μm. p = <0.0001, <0.0001, 1, 1, 0.9973, 1, and 1. h, j Comparisons of Psri-36::GFP fluorescence intensity in worms with the indicated genotypes and conditions. p = 0, 0, 0, <0.0001, <0.0001, and <0.0001 (h), and <0.0001, <0.0001, and <0.0001 (j). i, k Escape percentages of worms with the indicated genotypes and conditions. p = <0.0001, <0.0001, <0.0001, 0.0011, 0.0025, and 0.0004 (i), and <0.0001 and <0.0001 (k). **p < 0.01 and ***p < 0.001 (one-way ANOVA with Tukey’s post hoc test for a, c–i; two-sided unpaired t-test for (b, j, k). Brackets indicate the number of animals tested (a–h, j) or independent assays (i, k). Data are shown as means ± SEM. Source data are provided as a Source Data file.
The C. elegans genome encodes 40 ILPs57,59. To identify the specific ILPs involved, we performed a feeding RNAi screen using the intestine-specific RNAi strain VP303 and found that INS-37 and INS-7 selectively regulated ADF responses in fasted and infected worms, respectively (Supplementary Fig. 4). These effects were confirmed by intestine-specific RNAi of ins-37 or ins-7 via plasmid injection (Supplementary Fig. 4c, f). Additionally, intestine-specific RNAi of ins-37 and deletion of ins-7 abolished ADF responses in fasted and infected odr-7(ky4) mutants, respectively (Fig. 5c, d). The defect of ins-7(tm2001) was rescued by intestine-specific expression of wild-type ins-7 and recapitulated by intestine-specific RNAi (Fig. 5d). Furthermore, fasting-induced expression of the transcriptional reporter Pins-37::GFP, while P. aeruginosa infection suppressed it (Fig. 5e). Intestine-specific RNAi of aak-2, daf-16, or both genes reduced Pins-37::GFP expression similarly in fasted worms (Fig. 5f). In contrast, P. aeruginosa infection significantly increased Pins-7::GFP expression, which was blocked by intestine-specific RNAi of pmk-2 or overexpression of DAF-16aAM, with no additive effect when combined (Fig. 5g). Comparable intestine-specific RNAi effects were observed in the sid-1(qt9) background (Fig. 5c, d, f, g), confirming tissue specificity. Notably, Pins-7::GFP expression did not differ significantly between 4 h and 24 h of infection but was enhanced by 0.35% peptone, consistent with escape behavior (Supplementary Fig. 5a, b). Among other tested bacteria, although S. aureus also colonized the intestine, it did not induce Pins-7::GFP expression (Supplementary Fig. 5e, f). These results are consistent with a model in which hunger and P. aeruginosa infection induce intestinal INS-37 and INS-7 release, respectively, via AAK-2-DAF-16 and PMK-2-DAF-16 pathways, which in turn regulate ADF diacetyl sensitivity.
To test the specificity of these pathways, we performed reciprocal reporter assays. Fasting-induced Pins-37::GFP expression was not altered by intestinal pmk-2 RNAi (Supplementary Fig. 5c). In contrast, P. aeruginosa infection-induced Pins-7::GFP expression was further enhanced in aak-2 RNAi worms after 1 h and 2 h of infection, but not at 4 h, suggesting a transient peak of infection-induced ins-7 expression (Supplementary Fig. 5d). These results indicate that pmk-2 signaling is not required for fasting-induced ins-37 expression, whereas aak-2 normally acts to dampen ins-7 induction during infection. Thus, while aak-2 and pmk-2 serve as primary mediators of fasting- and infection-induced responses, respectively, these pathways also exhibit limited cross-talk.
To determine whether INS-37 and INS-7 regulate SRI-36 expression, we analyzed Psri-36::GFP fluorescence in ADF and escape behavior. In fasted worms, intestine-specific ins-37 RNAi or ADF-specific daf-2 RNAi similarly reduced Psri-36::GFP expression and escape, with no additive effect when combined (Fig. 5h, i). In infected worms, intestine-specific ins-7 RNAi or ADF-specific daf-2 RNAi similarly suppressed Psri-36::GFP expression and enhanced escape, again with no additive effect when combined (Fig. 5h, i). These results demonstrate that intestine-released INS-37 and INS-7 act through DAF-2 in ADF to regulate SRI-36 expression and decision-making. Consistently, P. aeruginosa also induced Pins-7::GFP expression and weak Psri-36::GFP expression in well-fed worms (Supplementary Fig. 5g, h). However, because P. aeruginosa did not affect escape in well-fed worms (Supplementary Fig. 5i), these results suggest that while P. aeruginosa activates intestinal immune signaling in both well-fed and fasted worms; its behavioral suppression is most pronounced under fasting, when hunger-driven risk-taking is elevated.
To dissect transcriptional mechanisms downstream of DAF-2, we examined known effectors of the IIS pathway. DAF-16/FOXO, HSF-1/HSF, SKN-1/Nrf2, PHA-4/FOXA, and DAF-12 are generally inhibited by DAF-2, while PQM-1 is activated by DAF-259,60. Interestingly, the upstream regulatory sequence of sri-36 contains predicted binding sites60 for both DAF-16 and PQM-1 (Supplementary Fig. 6a). In well-fed odr-7(ky4) mutants, ADF-specific RNAi of daf-16, but not other IIS effectors, induced robust diacetyl responses in ADF (Supplementary Fig. 6b). In infected worms, ADF-specific RNAi of pqm-1 had no effect (Supplementary Fig. 6c). These results suggest that DAF-16 represses sri-36 expression in ADF.
Consistently, ADF-specific daf-16 RNAi induced Psri-36::GFP expression in well-fed worms, while ADF-specific DAF-16aAM overexpression suppressed Psri-36::GFP expression and impaired escape in both fasted and infected worms (Fig. 5j, k). These results support a model in which DAF-16 acts cell-autonomously in ADF to inhibit sri-36 expression, and that ILP-DAF-2 signaling relieves this inhibition to drive sensory plasticity during hunger and infection.
AIB/AUA and AIA/AIY interneurons act downstream of ADF to regulate behavioral choices
We next sought to identify the interneurons that act downstream of ADF to mediate state-dependent decision-making. ADF neurons form synapses with multiple interneurons, including AIA, AIB, AIY, AIZ, AUA, and RIA38, with AIA, AIB, AIY, and AIZ serving as first-layer chemosensory interneurons26. In fasted worms, silencing AIB alone, AIB/AIZ, or AUA/URX using HisCl1 significantly reduced escape, whereas silencing AIY/AIZ, URX alone, or other interneurons had no effect (Fig. 6a). Notably, simultaneous silencing of AIB/AUA/URX further decreased escape compared to silencing AIB or AUA/URX alone (Fig. 6a). In infected worms, silencing AIA or AIY significantly increased escape, and the effect was amplified when both were silenced simultaneously (Fig. 6a). To investigate how ADF neurons regulate these interneurons, we performed calcium imaging. In fasted odr-7(ky4) mutants, diacetyl exposure inhibited AIB and activated AUA (Fig. 6b). In infected mutants, diacetyl inhibited both AIA and AIY (Fig. 6c). Silencing ADF using HisCl1 abolished diacetyl-evoked responses in these interneurons (Fig. 6b, c), whereas silencing AIB/AUA or AIA/AIY had no effect on ADF calcium responses (Supplementary Fig. 7a). These results suggest that ADF neurons regulate decision-making through distinct interneuron pathways depending on internal states. In fasted worms, ADF neurons promote risk-taking by inhibiting AIB and activating AUA, whereas in infected worms, they suppress risk-taking by inhibiting AIA and AIY. This is consistent with established roles of AIB in promoting avoidance, AUA in suppressing avoidance, and AIA/AIY in promoting attraction26,61.
a Escape percentages of fasted or P. aeruginosa-infected worms with or without 5 mM histamine (Hist) treatment. Neuron-specific HisCl1 expression was performed using Pgcy-28 (AIA), Pttx-3 (AIY), Pnpr-9 (AIB), Podr-2 (AIB/AIZ), Pser-2prom2 (AIY/AIZ), Pflp-8 (AUA/URX), Pgcy-36 (URX), and Pglr-3 (RIA). WT wild type. p = 1, 0.9958, 1, 1, 1, 1, 1, 1, 0.0137, 1, 0.0059, 1, 1, 1, 0.029, 1, 0, 1, 1, 1, 1, 1, 1, <0.0001, 1, 0, 1, <0.0001, 1, 1, 1, 1, 1, 0.0043, 1, 1, 1, 1, 1, 1, 1, and 1. Diacetyl-evoked GCaMP6 responses in AIB, AUA, AIA, and AIY neurons of fasted (b) or P. aeruginosa-infected (c) worms with the indicated genotypes. Left, averaged GCaMP6 signals (solid lines, mean; shaded regions, SEM). Right, comparisons of mean responses during the 30-s stimulation period. GCaMP6 was expressed in AIB, AUA, AIA, and AIY using Pnpr-9, Pflp-8, Pgcy-28, and Pttx-3, respectively. HisCl1 was expressed in ADF using Psrh-142. p = <0.0001 and <0.0001 (b), and <0.0001 and <0.0001 (c). d Escape percentages of fasted or P. aeruginosa-infected worms with the indicated genotypes. Neuron-specific rescue was performed using Pgcy-28 (AIA), Pttx-3 (AIY), Pnpr-9 (AIB), and Pflp-8 (AUA). p = 0, 0.8648, <0.0001, 0.4113, 0, and 0.0384. e, f 5-HT-evoked GCaMP6 responses in AIB, AUA, AIA, and AIY neurons of well-fed worms with the indicated genotypes. Neuron-specific RNAi was performed using Pgcy-28 (AIA), Pttx-3 (AIY), Pnpr-9 (AIB), and Pflp-8 (AUA). 5-HT was applied directly to the neuron cell body. p = <0.0001 and <0.0001 (e), and <0.0001 and <0.0001 (f). g Diacetyl-induced SNB-1::pHluorin fluorescence changes in ADF neurons of fasted or P. aeruginosa-infected worms with the indicated genotypes. Left, averaged pHluorin signals (solid lines, mean; shaded regions, SEM). Right, comparisons of mean responses during the 30-s stimulation period. SNB-1::pHluorin was expressed in ADF using Psrh-142. p = <0.0001, 0.0008, 0.9954, and <0.0001. Escape percentages of fasted or P. aeruginosa-infected worms with the indicated genotypes treated with capsaicin (h) or histamine (i). TRPV1 and HisCl1 were expressed in ADF using Psrh-142. p = 0.9303, 0.7527, 0.0044, <0.0001, <0.0001, <0.0001, and <0.0001 (h), and 0.5764, 0.8202, 0.0775, 0.0007, 0.0033, 0.5149, 0.82, 0.9005 (i). *p < 0.05, **p < 0.01, and ***p < 0.001 (one-way ANOVA with Tukey’s post hoc test for a, d, g; two-sided unpaired t-test for (b, c, e, f, h, i). ns no significance. Brackets indicate the number of independent assays (a, d, h, i) or animals tested (b, c, e–g). Data are shown as means ± SEM. Source data are provided as a Source Data file.
We next identified the 5-HT receptors mediating ADF’s effects on these interneurons. Among the six known C. elegans 5-HT receptors62, deletion of ser-4 and lgc-50 significantly reduced escape in fasted worms, while deletion of mod-1 increased escape in infected worms (Supplementary Fig. 7b). ser-4, lgc-50, and mod-1 encode a Gi/o-coupled inhibitory receptor, a 5-HT-gated chloride channel, and a 5-HT-gated cation channel, respectively62. Based on the behavioral roles and activity patterns of AIB, AUA, AIA, and AIY (Fig. 6a–c), we hypothesized that SER-4, LGC-50, and MOD-1 function in AIB, AUA, and AIA/AIY, respectively. Indeed, neuron-specific expression of the corresponding wild-type receptor genes in these interneurons rescued the behavioral phenotypes of the respective mutants (Fig. 6d), consistent with their known expression patterns62. Furthermore, neuron-specific RNAi of ser-4, lgc-50, and mod-1 abolished exogenous 5-HT-evoked calcium responses in AIB, AUA, and AIA/AIY, respectively (Fig. 6e, f). These results identify SER-4, LGC-50, and MOD-1 as key 5-HT receptors that relay ADF 5-HT output to distinct interneurons.
Finally, we tested whether direct modulation of ADF activity is sufficient to shift decision-making. Because diacetyl evoked stronger ADF activation in infected worms than in fasted worms (Fig. 2), we hypothesized that P. aeruginosa infection increases ADF 5-HT release. Indeed, infected odr-7(ky4) mutants exhibited significantly stronger synapto-pHluorin (SNB-1::pHluorin, a synaptic vesicle release reporter63) signals in ADF upon diacetyl stimulation than fasted worms (Fig. 6g), demonstrating enhanced 5-HT release during infection. Consistent with bidirectional control by SRI-36, in fasted sri-36(plc921) mutants, low-level SRI-36 expression (5 ng/µl) restored moderate 5-HT release to a level similar to fasted controls, whereas high-level expression (75 ng/µl) enhanced 5-HT release to a level comparable to P. aeruginosa-infected controls (Fig. 6g). Chemogenetic manipulation of ADF activity further supported this bidirectional control. Enhancing ADF activity in fasted worms using TRPV1, a mammalian cation channel activated by capsaicin64, reduced escape in a dose-dependent manner, an effect abolished by AIA/AIY-specific mod-1 RNAi (Fig. 6h). Conversely, chemogenetically suppressing ADF activity in infected worms using HisCl1 increased escape in a dose-dependent fashion (Fig. 6i). Notable, histamine concentrations above 3 mM reversed this trend and again suppressed escape, consistent with ADF promoting risk-taking at moderate activity levels but suppressing it when excessively activated. This bidirectional effect was attenuated by RNAi of ser-4 in AIB and lgc-50 in AUA (Fig. 6i). Together, these results demonstrate that tuning ADF activity is sufficient to switch behavioral strategies, with moderate ADF activation promoting risk-taking through AIB/AUA and strong activation suppressing it via AIA/AIY.
Discussion
In this study, we investigate how intestinal bacteria and metabolic states intersect to regulate host decision-making in C. elegans. Our findings uncover a gut-to-brain mechanism whereby hunger promotes risk-taking while intestinal P. aeruginosa suppresses it, both by modulating central 5-HT signaling. Fasting and infection differentially regulate the chemoreceptor SRI-36 in ADF 5-HT neurons, altering their sensitivity to diacetyl and driving opposing behavioral strategies via distinct downstream interneurons. These sensory modulations are orchestrated by the intestinal AAK-2/AMPK energy-sensing and PMK-2/p38 MAPK immune pathways, which act through a shared FOXO-to-FOXO axis but engage different ILPs (Fig. 7). Together, these findings define a dynamic regulatory interface linking metabolism, immunity, and behavioral plasticity.
Hunger and intestinal pathogens differentially regulate expression of the chemoreceptor SRI-36 in ADF 5-HT neurons, altering their sensitivity to diacetyl and driving opposing behavioral strategies via distinct downstream interneurons. These effects are mediated by the intestinal AAK-2/AMPK energy-sensing and PMK-2/p38 MAPK immune pathways, both acting through a shared FOXO-to-FOXO signaling axis but engaging distinct intestine-released ILPs.
The state-dependent sensory modulation enables C. elegans to flexibly adapt behavior based on internal states. In natural environments, diacetyl attracts C. elegans to rotting fruit, a valuable but risky food source that often harbors pathogens65. Hunger increases diacetyl attraction, promoting foraging and survival3,4,5,6, whereas pathogen infection suppresses this attraction, minimizing exposure to harmful microbes. These bidirectional shifts demonstrate how internal states reshape sensory representations and behavioral priorities to support context-appropriate decision-making. Furthermore, our findings indicate that suppression of hunger-driven risk-taking is specific to P. aeruginosa, which activates the intestinal PMK-2/p38 MAPK pathway. In contrast, S. aureus colonizes the intestine, but neither activates this immune signaling nor alters behavior. These results suggest that the differential effects of pathogens reflect distinct activation of intestinal immune pathways rather than differences in colonization or persistence. Future studies will be important to test additional pathogens that both colonize and activate PMK-2 signaling, to determine whether this mechanism for overriding hunger-driven decision-making is broadly conserved across diverse microbial infections.
Behavioral outputs depend not only on sensory input but also on how neural circuits process and route that input. The same stimuli can elicit distinct behaviors depending on the physiological state of sensory neurons and the downstream circuits they engage22,66,67. While such flexibility is essential for adaptive decision-making, the underlying neural architectures have remained elusive. Our study identifies a circuit motif in which ADF 5-HT neurons, modulated by intestinal metabolic and immune cues, act as a state-sensitive switch that adjusts 5-HT output to engage distinct downstream interneuron pathways. This configuration enables the nervous system to switch behavioral strategies in response to changing internal demands. Notably, fasting still increased escape even when 5-HT synthesis, ADF activity, or sri-36 was disrupted (Supplementary Fig. 8), suggesting the presence of parallel, 5-HT-independent pathways. These may include other neuromodulators such as tyramine6, dopamine68, and additional signals69 that contribute to hunger-driven risk-taking. Future work will be needed to delineate these parallel mechanisms and determine how they interact with the 5-HT axis to shape adaptive decision-making.
Central 5-HT is a critical neuromodulator of threat and reward processing and plays a key role in neuropsychiatric disorders associated with impaired decision-making, including autism, schizophrenia, anxiety, and depression31,70,71,72,73. Despite extensive study, 5-HT’s function remains paradoxical. It facilitates aversive learning and threat detection but also modulates reward processing, with dysregulation contributing to anhedonia2,16,70,72,74,75,76. Clinically, selective serotonin reuptake inhibitors (SSRIs) remain first-line antidepressants, yet they often fail to relieve, and may even exacerbate, motivational symptoms and anhedonia77. The heterogeneous nature of 5-HT neurons, their widespread brain projections, and diverse receptors complicates efforts to resolve these opposing effects62,71,78. Our findings suggest a unifying model in which central 5-HT acts as a behavioral rheostat, promoting risk-taking at moderate levels and suppressing it when excessively released.
We further show that internal states regulate 5-HT output by modulating chemoreceptor expression in ADF neurons. This sensory plasticity provides a mechanism for flexible behavioral responses to the same external stimulus8,29,79. Building on prior work showing that fasting and infection can modulate chemoreceptor expression through insulin signaling29,47,80,81, our study identifies intestine-released ILPs as key molecular signals linking interoception to sensory gain control. Specifically, fasting-induced INS-37 and pathogen-induced INS-7 engage the canonical insulin signaling pathway to regulate SRI-36 expression in ADF. This chemoreceptor remodeling tunes ADF diacetyl sensitivity and, consequently, the level of 5-HT release, thereby driving divergent food-seeking behaviors. Notably, chemoreceptors are often encoded in large gene families in many animals46, potentially providing a flexible repertoire that can be selectively induced or suppressed by physiological state. We propose that such plasticity enables animals to integrate complex internal and environmental cues to generate coherent behavioral responses under dynamic conditions. In this context, SRI-36 modulation offers a model for how gut-derived cues reshape sensory input to support state-dependent behavioral strategies. These findings suggest that olfactory tuning may be a promising target for treating metabolic disorders82.
This ILP-based mechanism of sensory modulation may be evolutionarily conserved. In mammals, ILP receptors are widely expressed in the brain, including olfactory regions82,83, and gut-derived ILPs can cross the blood-brain barrier83, suggesting that ILPs may regulate neural function beyond their metabolic effects. Indeed, ILPs influence neurotransmitter release, receptor expression, and neuronal excitability83 and are critical regulators of feeding-related behaviors7,83. For example, central insulin administration improves odor discrimination and reduces feeding7,83, while in Drosophila, ILPs modulate olfactory neuron sensitivity and increase aversion to noxious food under starvation7,84. Moreover, subsets of IECs can detect pathogens via innate immune responses and release ILPs85. Although most mammalian ILPs remain uncharacterized, gut-derived ILPs such as INSL5 are upregulated by fasting and influenced by the microbiota86. These observations raise the possibility that gut-derived ILPs are conserved mediators of sensory plasticity and behavioral adaptation.
We also show how a single sensory neuron, ADF, can flexibly reconfigure behavior by engaging distinct downstream circuits. In the fasted state, moderate ADF activation biases behavior toward food attraction by inhibiting AIB and activating AUA. In the infected state, heightened ADF activation suppresses attraction by inhibiting AIA and AIY. These opposing effects are mediated by distinct 5-HT receptors, SER-4, LGC-50, and MOD-1, expressed in the respective interneurons, in line with their known behavioral roles26,61. This circuit architecture allows C. elegans to dynamically switch behavioral strategies based on internal state, highlighting how physiologically tuned sensory neurons coordinate antagonistic behavioral outputs. Future work should quantify 5-HT dose-response effects on downstream interneurons (e.g., changes in speed or turning dynamics across internal states) to further refine this circuit model.
Together, our study advances understanding of how gut microbes influence brain function, an area where mechanisms remain largely undefined12,13,14,15. We demonstrate that P. aeruginosa activates an intestinal immune pathway that reprograms sensory function and overrides hunger-driven decision-making via central 5-HT signaling, pointing to an immune-mind connection. Similar principles may operate in mammals.
A limitation of this study is the use of RNAi to infer cell-specific gene function. Previous studies have shown that dsRNA can spread between tissues87, and therefore, low-level silencing in unintended cells cannot be excluded. However, RNAi-based results were interpreted in conjunction with complementary approaches, including genetic mutants, tissue-specific rescue, CRISPR knock-in reporters, and functional imaging and behavioral analyses.
Method
Bacterial strains and culture
E. coli OP50, P. aeruginosa PA14, S. aureus NCTC8325, and EPEC were cultured under conditions optimized for their growth and virulence88,89. E. coli, P. aeruginosa, and EPEC were grown in LB medium, while S. aureus was cultured in tryptic soy broth (Hopebio) supplemented with 10 μg/ml nalidixic acid (Macklin) to optimize growth and prevent overgrowth of contaminants. P. aeruginosa and S. aureus cultures were maintained under gentle shaking. All bacteria were grown overnight at 37 °C before seeding onto agar plates. E. coli was plated on nematode growth medium (NGM), P. aeruginosa on NGM supplemented with 0.25% peptone (Aobox) to support robust virulence factor production while maintaining animal viability for behavioral assays. S. aureus cultures were diluted 1:5 before plating on tryptic soy agar (Hopebio) containing 10 μg/ml nalidixic acid (Macklin). EPEC was plated on modified NGM containing 2 mg/ml tryptophan (Aladdin) to enhance its virulence89. Plates seeded with E. coli and EPEC were incubated at 37 °C for 20 h, P. aeruginosa plates at 37 °C for 24 h followed by 6 h at 25 °C, and S. aureus plates at 37 °C for 6–8 h. All plates were equilibrated at 25 °C for 1–2 h before use. The strains used in this study are listed in Supplementary Data 1.
C. elegans strains and culture
C. elegans strains were raised on NGM plates seeded with E. coli OP50 at 22 °C. Transgenic lines were generated by microinjection of plasmid DNA, and at least two independent lines were tested for confirmation. Integrated transgenic strains were created using gamma irradiation and outcrossed at least three times before use. Unless otherwise noted, all experiments used young adult hermaphrodites. Investigators were not blinded to genotype or condition. The strains used in this study are listed in Supplementary Data 1.
Mutant rescue and RNAi
Mutant rescue was achieved by expressing wild-type cDNAs under cell-specific promoters. RNAi knockdown was performed by co-injecting two plasmids encoding complementary sense and antisense mRNA fragments of the target gene. To confirm tissue specificity and exclude systemic RNAi effects, intestine-specific RNAi experiments (aak-2, daf-16, nsy-1, sek-1, pmk-2, hid-1, ins-37, and ins-7) were repeated in the systemic RNAi-defective sid-1(qt9) mutant background, in which dsRNA spreading is abolished53. Promoters used included Pges-1 (IECs)49, Pceh-2 (NSM)90, Psrh-142 (ADF)90, Pclh-3 (HSN)91, Pgcy-28 (AIA)92, Pttx-3 (AIY)92, Pnpr-9 (AIB)92, Podr-2 (AIB/AIZ)28, Pser-2prom2 (AIY/AIZ)62, Pflp-8 (AUA/URX)93, Pgcy-36 (URX)93, Pglr-3 (RIA)92, Psra-6 (ASH)94, and Pgpa-6 (AWA)95. All cDNAs and RNAi target sequences were cloned from a Bristol N2 cDNA library. Primers (Tsingke Biotechnology Co., Ltd.) are listed in Supplementary Table 1.
CRISPR-Cas9 genome editing
To generate the sri-36(plc921) deletion allele, two sgRNAs (5′ region: AGTACCTGATAAGCCAGACA; 3′ region: TATAATACCAAGAGTGGCCC) were cloned into the pDD162 (Peft-3::Cas9 + empty sgRNA) vector. The resulting plasmids (20 ng/μl each) were co-injected into young adult hermaphrodites together with Pmyo-2::mStrawberry (25 ng/μl) as a co-injection marker. F1 progeny were screened by PCR for deletion events, and homozygous mutants were confirmed by sequencing. To generate the endogenous sri-36::mNeonGreen knock-in reporter, three sgRNAs (GAAATTTGTAATCAGAAAGT; TAACTTTTACAATCCCATAT; ACTCCATTCGTCTGTCAATG) targeting the sri-36 C-terminus were used together with a donor repair plasmid containing the mNeonGreen coding sequence flanked by ~1 kb homology arms. The injection mixture contained 5 ng/μl sgRNA-Cas9 plasmids, 25 ng/μl donor plasmid, and 25 ng/μl Pmyo-2::mCherry co-injection marker. Correct knock-in events were identified by PCR and verified by sequencing. All CRISPR-edited strains were outcrossed three times before use.
RNAi screen of intestine-released ILPs
Intestine-specific RNAi of ILP genes was performed by feeding C. elegans strain VP303 with E. coli HT115 carrying L4440 plasmids containing target gene fragments96. Briefly, HT115 was cultured overnight in LB with 100 μg/ml ampicillin (Biosharp) at 37 °C, seeded onto NGM RNAi plates supplemented with 100 μg/ml ampicillin and 5 mM IPTG (Biotopped), and incubated at 37 °C for 12–14 h. Three L4-stage VP303 worms were transferred to each plate at 20 °C until their progeny reached young adulthood. rab-5 RNAi was used as a positive control.
Behavioral analysis
All behavioral assays were conducted at 22 °C. Spontaneous locomotion was recorded and analyzed using the Track-A-Worm automated tracking system (version 2.0)23. A single young adult was transferred to the center of a 6-cm NGM plate, allowed to recover for 30 s, and recorded for 1 min at 15fps.
For behavioral choice assays23, 6-cm choice plates were prepared by drawing a 1-cm glycerol ring (10 μl, 3 M) and placing two diacetyl spots (1 μl of 1:1000 diacetyl mixed with 1 μl of 0.1 M NaN3) near the plate edges. Synchronized worms were cultured on E. coli OP50 until young adulthood and then transferred to plates seeded with different bacterial strains. After 4 h, worms were washed with M9 buffer, and 10–15 worms were placed at the center of each choice plate. The number of worms at the diacetyl spots was counted after 15 min. For fasting assays, worms were deprived of food on unseeded NGM plates for the indicated time before the choice test. NGM plates containing histamine (Sigma-Aldrich) or capsaicin (Aladdin) were prepared by adding the chemicals at their final concentrations, and then seeded with 200 μl of bacteria pre-mixed with the same concentration of histamine or capsaicin.
Unisensory avoidance assays were performed using glycerol-only plates, and worms outside the glycerol ring were counted after 15 min23. Unisensory attraction assays were performed by dividing plates into four equal quadrants (A–D)23. Diacetyl (1 μl, 1:1000) and control water (1 μl) were applied to opposite edges along with NaN3. Worms were transferred to the center and counted after 15 min. Chemotaxis index = (worms in A − worms in D)/total worms × 100%.
For two-choice chemotaxis quadrant assays29, 9-cm quadrant plates were divided into four sections, with opposite quadrants containing either diacetyl (20 μl, 1:1000) or control water (20 μl). 100–150 M9-washed worms were placed in the center. After 15 min, plates were frozen at −20 °C for 10 min to stop movement, and worms in diacetyl and control quadrants were counted. Chemotaxis index = (worms in diacetyl − worms in control)/total worms × 100%. A positive index indicates attraction to diacetyl.
Imaging
Images of the SRI-36::GFP translational fusion and the CRISPR knock-in SRI-36::mNeonGreen fusion protein were acquired using an Olympus IXplore SpinSR confocal microscope with a 100×/1.45 NA oil objective. All other fluorescence images were captured with a Nikon Ts2R inverted microscope equipped with an Mshot MD60 CCD camera and analyzed using the Mshot Image Analysis System (version 1.1). Acquisition settings were kept constant within each experiment. For DAF-16 subcellular localization analysis, strains co-expressing Pdaf-16::daf-16a/b::GFP and the IEC nuclear marker Pges-1::NLS::mCherry were used. The percentage of GFP-labeled IEC nuclei was counted per animal. For calcium and pHluorin imaging, young adults were immobilized on coverslips using Vetbond Tissue Adhesive (3M Company) in bath solution (140 mM NaCl, 5 mM KCl, 5 mM CaCl2, 5 mM MgCl2, 11 mM dextrose, 5 mM HEPES, pH 7.2, 320 mOsm) and imaged at 1fps for 2 min. Diacetyl or glycerol was pressure-ejected (2 psi, 30 s) near the nose using an Eppendorf FemtoJet 4i microinjector. For exogenous 5-HT stimulation, because the targeted interneurons (AIA, AIY, AIB, and AUA) are located deep within the head and inaccessible to bath-applied 5-HT, a longitudinal incision was made along the glued region to expose neurons. 5-HT was then applied directly to neuron somas using pressure ejections (2 psi, 30 s), ensuring spatial precision, reproducibility, and consistent stimulation of the relevant interneurons.
Quantification of intestinal bacteria accumulation
To quantify the intestinal accumulation of PA14-dsRed97, full-lawn plates were prepared by spreading an overnight culture of PA14-dsRed. Young adult hermaphrodites grown on E. coli OP50 were transferred onto these plates and exposed for 4 h, then deprived of food on unseeded NGM plates for 6 h. Worms were washed with M9 buffer to remove external bacteria and imaged using the Nikon Ts2R inverted microscope.
Intestinal accumulation of E. coli, S. aureus, EPEC, and PA14-gacA(−) was quantified using colony-forming unit (CFU) assays48. Worms were similarly exposed to bacterial lawns and food-deprived conditions as above. Fifty worms were then collected into M9 containing 0.1 M sodium azide for paralysis, washed three times with M9 containing 1 mg/ml ampicillin (Biosharp) and 1 mg/ml gentamicin (Biosharp), and incubated for 30 min in antibiotic M9 to remove external bacteria. Worms were then lysed with a motorized pestle, serially diluted, and spread on the appropriate agar plates for each bacterial species. Plates were incubated overnight at 37 °C to determine CFUs.
Data analysis
Fluorescence, calcium, and pHluorin imaging data were analyzed using ImageJ (FIJI, version 1.50i). For neuronal imaging, individual cells were designated as separate regions of interest (ROIs). For the CRISPR knock-in SRI-36::mNeonGreen imaging, the dendritic region was selected as the ROI. For pHluorin imaging, the axonal region in the nerve ring was selected as the ROI. For intestinal fluorescence analysis, a single worm was assigned as the ROI. Calcium and pHluorin traces were first plotted as absolute fluorescence intensity (F) over time and then converted to bleach-corrected ΔF/F0 values using a custom MATLAB module94.
Statistical analyses and data visualization were performed using Origin 2019 (OriginLab). All diagram images were created using Microsoft PowerPoint 2024. Statistical significance was determined using one-way ANOVA with Tukey’s post hoc test for multiple group comparisons and two-sided unpaired t-test for two-group comparisons. Detailed statistical methods, sample sizes (n), and p-values are provided in the figure legends. Statistical significance was set at p < 0.05. No data were excluded from analysis, and all data are shown as mean ± SEM.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
All data generated or analyzed during this study are included in this published article and its supplementary information files. Source data are provided with this paper.
References
Stein, B. E., Stanford, T. R. & Rowland, B. A. Development of multisensory integration from the perspective of the individual neuron. Nat. Rev. Neurosci. 15, 520–535 (2014).
Hu, H. Reward and aversion. Annu. Rev. Neurosci. 39, 297–324 (2016).
Liu, L. et al. Eating to dare - nutrition impacts human risky decision and related brain function. Neuroimage 233, 117951 (2021).
Padilla, S. L. et al. Agouti-related peptide neural circuits mediate adaptive behaviors in the starved state. Nat. Neurosci. 19, 734–741 (2016).
Inagaki, H. K., Panse, K. M. & Anderson, D. J. Independent, reciprocal neuromodulatory control of sweet and bitter taste sensitivity during starvation in Drosophila. Neuron 84, 806–820 (2014).
Ghosh, D. D. et al. Neural architecture of hunger-dependent multisensory decision making in C. elegans. Neuron 92, 1049–1062 (2016).
Root, C. M., Ko, K. I., Jafari, A. & Wang, J. W. Presynaptic facilitation by neuropeptide signaling mediates odor-driven food search. Cell 145, 133–144 (2011).
Wexler, L. R., Miller, R. M. & Portman, D. S. C. elegans males integrate food signals and biological sex to modulate state-dependent chemosensation and behavioral prioritization. Curr. Biol. 30, 2695–2706.e2694 (2020).
Sengupta, P. The belly rules the nose: feeding state-dependent modulation of peripheral chemosensory responses. Curr. Opin. Neurobiol. 23, 68–75 (2013).
Pedersen, A., Göder, R., Tomczyk, S. & Ohrmann, P. Risky decision-making under risk in schizophrenia: a deliberate choice?. J. Behav. Ther. Exp. Psychiatry 56, 57–64 (2017).
Thye, M. D., Bednarz, H. M., Herringshaw, A. J., Sartin, E. B. & Kana, R. K. The impact of atypical sensory processing on social impairments in autism spectrum disorder. Dev. Cogn. Neurosci. 29, 151–167 (2018).
Cryan, J. F. & Dinan, T. G. Mind-altering microorganisms: the impact of the gut microbiota on brain and behaviour. Nat. Rev. Neurosci. 13, 701–712 (2012).
Nagpal, J. & Cryan, J. F. Microbiota-brain interactions: moving toward mechanisms in model organisms. Neuron https://doi.org/10.1016/j.neuron.2021.09.036 (2021).
Dantzer, R., O’Connor, J. C., Freund, G. G., Johnson, R. W. & Kelley, K. W. From inflammation to sickness and depression: when the immune system subjugates the brain. Nat. Rev. Neurosci. 9, 46–56 (2008).
Perry, V. H., Cunningham, C. & Holmes, C. Systemic infections and inflammation affect chronic neurodegeneration. Nat. Rev. Immunol. 7, 161–167 (2007).
Der-Avakian, A. & Markou, A. The neurobiology of anhedonia and other reward-related deficits. Trends Neurosci. 35, 68–77 (2012).
Kang, W. K. et al. Vitamin B(12) produced by gut bacteria modulates cholinergic signalling. Nat. Cell Biol. https://doi.org/10.1038/s41556-023-01299-2 (2024).
Fülling, C., Dinan, T. G. & Cryan, J. F. Gut microbe to brain signaling: what happens in Vagus. Neuron 101, 998–1002 (2019).
Zhang, J., Holdorf, A. D. & Walhout, A. J. M. C. elegans and its bacterial diet as a model for systems-level understanding of host–microbiota interactions. Curr. Opin. Biotechnol. 46, 74–80 (2017).
Kim, D. H. & Flavell, S. W. Host-microbe interactions and the behavior of Caenorhabditis elegans. J. Neurogenet. 34, 500–509 (2020).
Matty, M. A. & Chalasani, S. H. Microbial mind control. Cell Host Microbe 28, 147–149 (2020).
Faumont, S., Lindsay, T. H. & Lockery, S. R. Neuronal microcircuits for decision making in C. elegans. Curr. Opin. Neurobiol. 22, 580–591 (2012).
Liu, P., Chen, B. & Wang, Z. W. GABAergic motor neurons bias locomotor decision-making in C. elegans. Nat. Commun. 11, 5076 (2020).
Matty, M. A. et al. Intestine-to-neuronal signaling alters risk-taking behaviors in food-deprived Caenorhabditis elegans. PLoS Genet. 18, e1010178 (2022).
Hilliard, M. A. et al. In vivo imaging of C. elegans ASH neurons: cellular response and adaptation to chemical repellents. EMBO J. 24, 63–72 (2005).
Metaxakis, A., Petratou, D. & Tavernarakis, N. Multimodal sensory processing in Caenorhabditis elegans. Open Biol. https://doi.org/10.1098/rsob.180049 (2018).
Meisel, J. D., Panda, O., Mahanti, P., Schroeder, F. C. & Kim, D. H. Chemosensation of bacterial secondary metabolites modulates neuroendocrine signaling and behavior of C. elegans. Cell 159, 267–280 (2014).
Ha, H. I. et al. Functional organization of a neural network for aversive olfactory learning in Caenorhabditis elegans. Neuron 68, 1173–1186 (2010).
Wu, T. et al. Pathogenic bacteria modulate pheromone response to promote mating. Nature https://doi.org/10.1038/s41586-022-05561-9 (2023).
Fuhrman, L. E., Goel, A. K., Smith, J., Shianna, K. V. & Aballay, A. Nucleolar proteins suppress Caenorhabditis elegans innate immunity by inhibiting p53/CEP-1. PLoS Genet. 5, e1000657 (2009).
Bang, D. et al. Sub-second dopamine and serotonin signaling in human striatum during perceptual decision-making. Neuron 108, 999–1010.e1016 (2020).
Chase, D. L. & Koelle, M. R. Biogenic amine neurotransmitters in C. elegans. WormBook https://doi.org/10.1895/wormbook.1.132.1 (2007).
Ranganathan, R., Sawin, E. R., Trent, C. & Horvitz, H. R. Mutations in the Caenorhabditis elegans serotonin reuptake transporter MOD-5 reveal serotonin-dependent and -independent activities of fluoxetine. J. Neurosci. 21, 5871–5884 (2001).
Sze, J. Y., Victor, M., Loer, C., Shi, Y. & Ruvkun, G. Food and metabolic signalling defects in a Caenorhabditis elegans serotonin-synthesis mutant. Nature 403, 560–564 (2000).
Pokala, N., Liu, Q., Gordus, A. & Bargmann, C. I. Inducible and titratable silencing of Caenorhabditis elegans neurons in vivo with histamine-gated chloride channels. Proc. Natl. Acad. Sci. USA 111, 2770–2775 (2014).
Pereira, L. et al. A cellular and regulatory map of the cholinergic nervous system of C. elegans. eLife https://doi.org/10.7554/eLife.12432 (2015).
Rand, J. B. Acetylcholine. WormBook https://doi.org/10.1895/wormbook.1.131.1 (2007).
Cook, S. J. et al. Whole-animal connectomes of both Caenorhabditis elegans sexes. Nature 571, 63–71 (2019).
Shao, J. et al. Serotonergic neuron ADF modulates avoidance behaviors by inhibiting sensory neurons in C. elegans. Pflugers Arch. 471, 357–363 (2019).
Zhang, B., Jun, H., Wu, J., Liu, J. & Xu, X. Z. S. Olfactory perception of food abundance regulates dietary restriction-mediated longevity via a brain-to-gut signal. Nat. Aging 1, 255–268 (2021).
Richmond, J. E., Davis, W. S. & Jorgensen, E. M. UNC-13 is required for synaptic vesicle fusion in C. elegans. Nat. Neurosci. 2, 959–964 (1999).
Speese, S. et al. UNC-31 (CAPS) is required for dense-core vesicle but not synaptic vesicle exocytosis in Caenorhabditis elegans. J. Neurosci. 27, 6150–6162 (2007).
Cornell, R., Cao, W., Liu, J. & Pocock, R. Conditional degradation of UNC-31/CAPS enables spatiotemporal analysis of neuropeptide function. J. Neurosci. 42, 8599–8607 (2022).
Sengupta, P., Colbert, H. A. & Bargmann, C. I. The C. elegans gene odr-7 encodes an olfactory-specific member of the nuclear receptor superfamily. Cell 79, 971–980 (1994).
Bargmann, C. I. Chemosensation in C. elegans. WormBook https://doi.org/10.1895/wormbook.1.123.1 (2006).
Robertson, H. M. & Thomas, J. H. The putative chemoreceptor families of C. elegans. WormBook https://doi.org/10.1895/wormbook.1.66.1 (2006).
McLachlan, I. G. et al. Diverse states and stimuli tune olfactory receptor expression levels to modulate food-seeking behavior. eLife https://doi.org/10.7554/eLife.79557 (2022).
Liu, J. et al. GABAergic signaling between enteric neurons and intestinal smooth muscle promotes innate immunity and gut defense in Caenorhabditis elegans. Immunity 56, 1515–1532.e1519 (2023).
Gong, J. et al. A cold-sensing receptor encoded by a glutamate receptor gene. Cell 178, 1375–1386.e1311 (2019).
González, A., Hall, M. N., Lin, S. C. & Hardie, D. G. AMPK and TOR: the Yin and Yang of cellular nutrient sensing and growth control. Cell Metab. 31, 472–492 (2020).
Greer, E. L. et al. An AMPK-FOXO pathway mediates longevity induced by a novel method of dietary restriction in C. elegans. Curr. Biol. 17, 1646–1656 (2007).
Lin, K., Hsin, H., Libina, N. & Kenyon, C. Regulation of the Caenorhabditis elegans longevity protein DAF-16 by insulin/IGF-1 and germline signaling. Nat. Genet. 28, 139–145 (2001).
Winston, W. M., Molodowitch, C. & Hunter, C. P. Systemic RNAi in C. elegans requires the putative transmembrane protein SID-1. Science 295, 2456–2459 (2002).
Evans, E. A., Kawli, T. & Tan, M. W. Pseudomonas aeruginosa suppresses host immunity by activating the DAF-2 insulin-like signaling pathway in Caenorhabditis elegans. PLoS Pathog. 4, e1000175 (2008).
Li, G., Gong, J., Lei, H., Liu, J. & Xu, X. Z. Promotion of behavior and neuronal function by reactive oxygen species in C. elegans. Nat. Commun. 7, 13234 (2016).
Lee, K. & Mylonakis, E. An intestine-derived neuropeptide controls avoidance behavior in Caenorhabditis elegans. Cell Rep. 20, 2501–2512 (2017).
Li, C. & Kim, K. Neuropeptides. WormBook https://doi.org/10.1895/wormbook.1.142.1 (2008).
De Rosa, M. J. et al. The flight response impairs cytoprotective mechanisms by activating the insulin pathway. Nature 573, 135–138 (2019).
Murphy, C. T. & Hu, P. J. Insulin/insulin-like growth factor signaling in C. elegans. WormBook https://doi.org/10.1895/wormbook.1.164.1 (2013).
Tepper, R. G. et al. PQM-1 complements DAF-16 as a key transcriptional regulator of DAF-2-mediated development and longevity. Cell 154, 676–690 (2013).
Ezcurra, M., Walker, D. S., Beets, I., Swoboda, P. & Schafer, W. R. Neuropeptidergic signaling and active feeding state inhibit nociception in Caenorhabditis elegans. J. Neurosci. 36, 3157–3169 (2016).
Dag, U. et al. Dissecting the functional organization of the C. elegans serotonergic system at whole-brain scale. Cell https://doi.org/10.1016/j.cell.2023.04.023 (2023).
Miesenböck, G., De Angelis, D. A. & Rothman, J. E. Visualizing secretion and synaptic transmission with pH-sensitive green fluorescent proteins. Nature 394, 192–195 (1998).
Tobin, D. M. et al. Combinatorial expression of TRPV channel proteins defines their sensory functions and subcellular localization in C. elegans Neurons. Neuron 35, 307–318 (2002).
Singh, A. & Luallen, R. J. Understanding the factors regulating host-microbiome interactions using Caenorhabditis elegans. Philos. Trans. R. Soc. Lond. B Biol. Sci. 379, 20230059 (2024).
Almoril-Porras, A. et al. Configuration of electrical synapses filters sensory information to drive behavioral choices. Cell 188, 89–103.e113 (2025).
Grunwald Kadow, I. C. State-dependent plasticity of innate behavior in fruit flies. Curr. Opin. Neurobiol. 54, 60–65 (2019).
Sawin, E. R., Ranganathan, R. & Horvitz, H. R. C. elegans locomotory rate is modulated by the environment through a dopaminergic pathway and by experience through a serotonergic pathway. Neuron 26, 619–631 (2000).
Flavell, S. tevenW. et al. Serotonin and the neuropeptide PDF initiate and extend opposing behavioral states in C. elegans. Cell 154, 1023–1035 (2013).
Liu, Z. et al. Dorsal raphe neurons signal reward through 5-HT and glutamate. Neuron 81, 1360–1374 (2014).
Homberg, J. R. Serotonin and decision making processes. Neurosci. Biobehav. Rev. 36, 218–236 (2012).
Seymour, B., Daw, N. D., Roiser, J. P., Dayan, P. & Dolan, R. Serotonin selectively modulates reward value in human decision-making. J. Neurosci. 32, 5833–5842 (2012).
Pollak Dorocic, I. et al. A whole-brain atlas of inputs to serotonergic neurons of the dorsal and median raphe nuclei. Neuron 83, 663–678 (2014).
Fonseca, M. S., Murakami, M. & Mainen, Z. F. Activation of dorsal raphe serotonergic neurons promotes waiting but is not reinforcing. Curr. Biol. 25, 306–315 (2015).
Li, Y. et al. Serotonin neurons in the dorsal raphe nucleus encode reward signals. Nat. Commun. 7, 10503 (2016).
Cohen, J. Y., Amoroso, M. W. & Uchida, N. Serotonergic neurons signal reward and punishment on multiple timescales. eLife https://doi.org/10.7554/eLife.06346 (2015).
Price, J., Cole, V. & Goodwin, G. M. Emotional side-effects of selective serotonin reuptake inhibitors: qualitative study. Br. J. Psychiatry 195, 211–217 (2009).
Okaty, B. W., Commons, K. G. & Dymecki, S. M. Embracing diversity in the 5-HT neuronal system. Nat. Rev. Neurosci. 20, 397–424 (2019).
Gruner, M. et al. Cell-Autonomous and non-cell-autonomous regulation of a feeding state-dependent chemoreceptor gene via MEF-2 and bHLH transcription factors. PLoS Genet. 12, e1006237 (2016).
van der Linden, A. M., Nolan, K. M. & Sengupta, P. KIN-29 SIK regulates chemoreceptor gene expression via an MEF2 transcription factor and a class II HDAC. EMBO J. 26, 358–370 (2007).
Kyani-Rogers, T. et al. Developmental history modulates adult olfactory behavioral preferences via regulation of chemoreceptor expression in Caenorhabditis elegans. Genetics https://doi.org/10.1093/genetics/iyac143 (2022).
Riera, C. E. et al. The sense of smell impacts metabolic health and obesity. Cell Metab. 26, 198–211.e195 (2017).
Fernandez, A. M. & Torres-Alemán, I. The many faces of insulin-like peptide signalling in the brain. Nat. Rev. Neurosci. 13, 225–239 (2012).
Wu, Q., Zhao, Z. & Shen, P. Regulation of aversion to noxious food by Drosophila neuropeptide Y- and insulin-like systems. Nat. Neurosci. 8, 1350–1355 (2005).
Kaelberer, M. M., Rupprecht, L. E., Liu, W. W., Weng, P. & Bohórquez, D. V. Neuropod cells: the emerging biology of gut-brain sensory transduction. Annu. Rev. Neurosci. 43, 337–353 (2020).
Lee, Y. S. et al. Insulin-like peptide 5 is a microbially regulated peptide that promotes hepatic glucose production. Mol. Metab. 5, 263–270 (2016).
Ravikumar, S., Devanapally, S. & Jose, A. M. Gene silencing by double-stranded RNA from C. elegans neurons reveals functional mosaicism of RNA interference. Nucleic Acids Res. 47, 10059–10071 (2019).
Powell, J. R. & Ausubel, F. M. in Innate Immunity (eds Ewbank, J. & Vivier, E.) 403–427 (Humana Press, 2008).
Anyanful, A., Easley, K. A., Benian, G. M. & Kalman, D. Conditioning protects C. elegans from lethal effects of enteropathogenic E. coli by activating genes that regulate lifespan and innate immunity. Cell Host Microbe 5, 450–462 (2009).
Cunningham, K. A. et al. AMP-activated kinase links serotonergic signaling to glutamate release for regulation of feeding behavior in C. elegans. Cell Metab. 16, 113–121 (2012).
Yue, X. M. et al. TMC proteins modulate egg laying and membrane excitability through a background leak conductance in C. elegans. Neuron 97, 571–585 (2018).
Li, Z., Liu, J., Zheng, M. & Xu, X. Z. Encoding of both analog- and digital-like behavioral outputs by one C. elegans interneuron. Cell 159, 751–765 (2014).
Noble, T., Stieglitz, J. & Srinivasan, S. An integrated serotonin and octopamine neuronal circuit directs the release of an endocrine signal to control C. elegans body fat. Cell Metab. 18, 672–684 (2013).
Zhan, X. et al. Locomotion modulates olfactory learning through proprioception in C. elegans. Nat. Commun. 14, 4534 (2023).
Liu, Q., Kidd, P. B., Dobosiewicz, M. & Bargmann, C. I. C. elegans AWA olfactory neurons fire calcium-mediated all-or-none action potentials. Cell 175, 57–70.e17 (2018).
Timmons, L. & Fire, A. Specific interference by ingested dsRNA. Nature 395, 854 (1998).
Pu, X. & Qi, B. Lysosomal dysfunction by inactivation of V-ATPase drives innate immune response in C. elegans. Cell Rep. 43, 114138 (2024).
Acknowledgements
We would like to thank Jianke Gong, Bin Qi, Anbing Shi, and Taihong Wu for providing bacterial strains; Jianke Gong, Long Lin, Anbing Shi, Haijun Tu, Wenxing Yang, and Donglei Zhang for plasmids; and Cornelia I. Bargmann, Jianke Gong, Long Lin, and Yun Zhang for C. elegans strains. We also acknowledge the Caenorhabditis Genetics Center (USA), which is funded by NIH Office of Research Infrastructure Programs (P40OD010440), for providing E. coli OP50 and C. elegans strains. This work was supported by grants from the National Natural Science Foundation of China (32571188 and 32171003 to P.L.) and the Interdisciplinary Research Program of HUST (5003170102 to P.L.).
Author information
Authors and Affiliations
Contributions
Y.L., C.C., and P.L. conceived and initiated the study. Y.L. and P.L. designed the experiments. Y.L., X.Z., M.X., Y.W., H.L., and J.Z. performed the experiments and analyzed the data. Y.L. and P.L. interpreted the results. P.L. wrote the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Lei, Y., Chen, C., Zhan, X. et al. Intestinal pathogens override hunger-driven decision-making via immune regulation of central serotonin signaling in C. elegans. Nat Commun 17, 3144 (2026). https://doi.org/10.1038/s41467-026-69924-w
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41467-026-69924-w
