Memory of recent oxygen experience switches pheromone valence in Caenorhabditis elegans

Abstract

Animals adjust their behavioral priorities according to momentary needs and prior experience. We show that Caenorhabditis elegans changes how it processes sensory information according to the oxygen environment it experienced recently. C. elegans acclimated to 7% O₂ are aroused by CO₂ and repelled by pheromones that attract animals acclimated to 21% O₂ This behavioral plasticity arises from prolonged activity differences in a circuit that continuously signals O₂ levels. A sustained change in the activity of O₂-sensing neurons reprograms the properties of their postsynaptic partners, the RMG hub interneurons. RMG is gap-junctionally coupled to the ASK and ADL pheromone sensors that respectively drive pheromone attraction and repulsion. Prior O₂ experience has opposite effects on the pheromone responsiveness of these neurons. These circuit changes provide a physiological correlate of altered pheromone valence. Our results suggest C. elegans stores a memory of recent O₂ experience in the RMG circuit and illustrate how a circuit is flexibly sculpted to guide behavioral decisions in a context-dependent manner.

Keywords: acclimation; experience-dependent plasticity; neural circuit; oxygen sensing; tonic circuit.

Fig. S1.

Acclimation to different O₂ environments reprograms CO₂ responses in natural C. elegans isolates. (A–C) Wild strains modulate their CO₂ response according to recent O₂ experience. These strains encode NPR-1 215F, the natural low activity isoform of NPR-1. n = 116–171.

Fig. 1.

Recent O₂ experience regulates CO₂-evoked arousal. (A) N2 animals and npr-1 animals acclimated to 7% O₂ persistently increase their speed when CO₂ rises to 3%; npr-1 animals acclimated to 21% O₂ do not. n = 247–302 animals. ****P < 0.0001; ns, not significant; Wilcoxon signed-rank test. Animals were assayed on food at 7% O₂. In this and subsequent figures, animals were acclimated to 21% O₂ unless noted otherwise; solid lines indicate the mean and shaded areas, the SEM. Black bars here and throughout indicate intervals used for statistical comparisons; boxplots show the median and the 25th–75th percentiles; whiskers represent 10th–90th percentiles. (B) Selectively expressing NPR-1 215V in RMG, (C) disrupting gcy-35, or (D) RNAi knockdown of EGL-21 in RMG prevent npr-1 animals acclimated to 21% O₂ from suppressing CO₂-evoked arousal. n = 104–235.

Fig. S2.

Time line and reversibility of acclimation to different O₂ levels. (A) N2 are strongly aroused by a 3% CO₂ stimulus, irrespective of whether they have been acclimated at 21% or 7% O_2. n = 462–518 animals. ****P < 0.0001; Wilcoxon signed-rank test. (B and B′) Mean speed of npr-1 animals at 3% CO₂, plotted against the time animals were previously exposed to 7% or atmospheric (∼21%) O₂. (B) Acclimation to 7% happens gradually and animals continue to increase their speed over many hours. n = 151–171. (C) Acclimation is reversed rapidly, and after ≤3 h, animals behave like siblings grown at 21% O₂. n = 138–188. Error bars (B and C) and shaded regions (B′ and C′) represent SEM. ****P < 0.0001; ***P < 0.001; ns, not significant; Kruskal–Wallis ANOVA with Dunn’s multiple comparisons test.

Fig. 2.

Pheromone valence changes with prior O₂ experience. (A) Quadrant assay for pheromone preference (as in ref. 36). (B) Behavioral responses to an equimolar 10 nM mix of C3, C6, and C9 ascaroside pheromones. npr-1 animals acclimated to 21% O₂ are attracted to the pheromone, whereas siblings acclimated to 7% O₂ avoid it. N2 avoid pheromones irrespective of whether they have been acclimated to 7% or 21% O_2. The soluble guanylate cyclase GCY-35 is required for normal O₂ responses and pheromone attraction in npr-1 animals acclimated at 21% O₂. **P < 0.01; ns, not significant; one-way ANOVA with Tukey’s multiple comparisons test. n = 8 assays each. (C) Previous O₂ experience sculpts pheromone responses in ASK sensory neurons. Acclimation to 7% O₂ reduces pheromone-evoked Ca²⁺ responses in ASK, consistent with altered behavioral preference. The gray shading in this and subsequent figures indicates addition of pheromone. (D) Quantification of data shown in C. Heat maps in this and all subsequent figures show individual Ca²⁺ responses. n = 35–36 animals. ****P < 0.0001; **P < 0.01; *P < 0.05; ns, not significant; Mann–Whitney U test.

Fig. 3.

Peptidergic feedback regulates RMG properties and pheromone preference. (A and B) Acclimation to 7% O₂, or knockdown of egl-21, similarly reduces the RMG Ca²⁺ responses evoked by a 7–21% O₂ stimulus. (B) Quantification of data in A. n = 20–21 animals. *P < 0.05; ns, not significant; Mann–Whitney U test. (C and D) Acclimation to 7% O₂ reduces URX Ca²⁺ responses evoked by a 7–21% O₂ stimulus. (D) Quantification of data in C. n = 38–39 animals. *P < 0.05; Mann–Whitney U test. (E and F) Knockdown of egl-21 in RMG diminishes pheromone-evoked Ca²⁺ responses in ASK. (F) Quantification of data in E. n = 20–21 animals. *P < 0.05; **P < 0.01; ****P < 0.0001; ns, not significant; Mann–Whitney U test. (G) RNAi knockdown of egl-21 in RMG prevents npr-1 animals acclimated to 21% O₂ from being attracted to pheromone. n = 12 assays. *P < 0.05; one-way ANOVA followed by Dunnett’s multiple comparisons test.

Fig. 4.

RMG hub neurons respond to pheromones and alter their response according to recent O₂ experience. (A) Circuit showing connections between RMG interneurons and O₂-sensing, nociceptive, and pheromone-sensing neurons. (B) An equimolar (100 nM) mix of C3, C6, and C9 ascarosides evokes a decrease in RMG Ca²⁺ in npr-1 animals acclimated to 21% O₂. n = 57 animals. (C and D) RMG shows robust pheromone-evoked Ca²⁺ responses in both N2 animals and npr-1 animals acclimated to 21% O₂. Acclimating npr-1 animals to 7% O₂ alters RMG properties and diminishes both ON and OFF responses to pheromone addition and removal. (D) Quantification of data shown in C. n = 36 animals each. ***P < 0.001; ****P < 0.0001; Mann–Whitney U test. (E–G) Acclimation to 7% O₂ enhances ADL pheromone responses and acute pheromone repulsion. npr-1 animals show decreased avoidance of the C9 ascaroside compared with N2 when grown under standard conditions (∼21% O₂), but not when acclimated to 7% O₂. Plotted are the avoidance index (E) and fraction of animals reversing (E′), in response to a drop of diluted C9 (10 nM) applied to the nose. n = 260–280 animals each. *P < 0.05; ***P < 0.001; ****P < 0.0001; ns, not significant; one-way ANOVA followed by Tukey’s multiple comparisons test. (F and G) The Ca²⁺ responses evoked in ADL by 10 nM C9 pheromone are larger in npr-1 animals acclimated to 7% O₂ compared with siblings acclimated at 21% O₂. (G) Quantification of data plotted in F. n = 23–24 animals. **P < 0.01; Mann–Whitney U test. (H) Model showing acclimation to different O₂ levels has opposite effects on antagonistic circuit elements promoting attractive and repulsive pheromone responses. This experience-dependent plasticity arises from prolonged changes in the activity state of RMG hub neurons, which receive tonic input from URX O₂-sensing neurons. Sustained peptide release from RMG at 21% O₂ feeds back to alter RMG properties. In animals kept at 7% O₂ peptide release from RMG is low, disrupting the feedback. The time required for animals to acclimate to 7% or 21% O₂ corresponds to hysteresis in the onset or decay of peptide signaling. See also Figs. S3 and S4.

Fig. S3.

ASK and ADL sensory neuron spokes respond to O₂. (A) O₂-evoked Ca²⁺ responses in ASK do not differ between N2 and npr-1 animals. (B) Quantification of data plotted in A. n = 21–24 animals; Mann–Whitney U test. Blue shading indicates a shift from 7% to 21% O₂. (C) ADL sensory neurons show robust responses to a 21% O₂ stimulus in npr-1 but not in N2 animals. (D) Quantification of data shown in C. n = 30 animals each. ****P < 0.0001; Mann–Whitney U test. Note different Ca²⁺ sensors were used to image ASK (YC3.60) and ADL (GCaMP3 var500).

Fig. S4.

ADL responses to 21% O₂ require GCY-35 signaling, but not the TRPV channel OCR-2. (A) The soluble guanylate cyclase GCY-35 is required for ADL responses to 21% O₂. The defect of gcy-35(ok769) mutants is fully rescued by expressing gcy-35 cDNA in URX using the flp-8 promoter. (B) Quantification of data shown in A. n = 25–27 animals each. ****P < 0.0001; ***P < 0.001; ns, not significant; Mann–Whitney U test. (C) OCR-2 is required cell autonomously for ADL response to the C9 ascaroside. (D) Quantification of data shown in C. n = 11–12 animals each. ****P < 0.0001; Mann–Whitney U test. Gray shading indicates stimulation with 10 nM C9. (E) OCR-2 is not required for ADL responses to a 21% O₂ stimulus, although Ca²⁺ rises less sharply in mutants. (F) Quantification of data plotted in E. n = 21–22 animals each; Mann–Whitney U test. Blue shading indicates a shift from 7% to 21% O₂.