Serotonin and the neuropeptide PDF initiate and extend opposing behavioral states in C. elegans

Abstract

Foraging animals have distinct exploration and exploitation behaviors that are organized into discrete behavioral states. Here, we characterize a neuromodulatory circuit that generates long-lasting roaming and dwelling states in Caenorhabditis elegans. We find that two opposing neuromodulators, serotonin and the neuropeptide pigment dispersing factor (PDF), each initiate and extend one behavioral state. Serotonin promotes dwelling states through the MOD-1 serotonin-gated chloride channel. The spontaneous activity of serotonergic neurons correlates with dwelling behavior, and optogenetic modulation of the critical MOD-1-expressing targets induces prolonged dwelling states. PDF promotes roaming states through a Gαs-coupled PDF receptor; optogenetic activation of cAMP production in PDF receptor-expressing cells induces prolonged roaming states. The neurons that produce and respond to each neuromodulator form a distributed circuit orthogonal to the classical wiring diagram, with several essential neurons that express each molecule. The slow temporal dynamics of this neuromodulatory circuit supplement fast motor circuits to organize long-lasting behavioral states.

Figure 1. Serotonin affects exploration behavior

(A) Locomotion of an adult wild-type animal on an E. coli lawn. Locomotion speed and angular speed (turning rate) differ in two distinct behavioral states, roaming and dwelling. (B) Simplified assay for measuring exploration behavior, based on movement across a 35 mm bacterial lawn. The grid has 86 squares. The exact number of squares entered can vary from day to day, so all genetic manipulations are compared to controls tested in parallel. (C) Exploration behavior of 57 mutant strains normalized to wild-type controls. Asterisks indicate statistical significance at FDR<0.05. For detailed results, see Table S1. (D) Rescue of mod-1 mutant phenotype by a mod-1::mod-1::GFP transgene. (E) Rescue of tph-1 mutant phenotype by a genomic tph-1 transgene. For (D-E), asterisks indicate p<0.05 by ANOVA and Bonferroni-Dunn post-hoc test. Data are shown as means ± SEM. See also Fig. S1 and Table S1.

Figure 2. Behavioral state defects in serotonergic signaling mutants

(A) Complementary cumulative distribution function (ccdf) for dwelling state durations in wild-type animals (dots) fit to a single exponential (pink line). (B) Ccdf for roaming state durations in wild-type animals (dots) fit to a double exponential (pink line). The two exponentials are individually displayed as dashed black lines, and w₁/w₂ are the weights of each exponential, i.e. 85% of roaming states have a mean lifetime of 75 sec. (C,D) Behaviors of wild-type and mutant animals. (C₁,D₁) Fraction of time spent in roaming and dwelling states. (C₂,D₂) Dwelling state durations and (C₃,D₃) Roaming state durations, expressed as means of individual animal state duration means ± SEM, not as event distributions as in A,B. Asterisks indicate p<0.01, t-test. See also Fig. S2.

Figure 3. A distributed serotonergic circuit controls exploration behavior

(A,B) Cell-specific deletion of tph-1 using a Cre/Lox strategy. (A) Schematic depicting genotypes. (B) Exploration behavior. Asterisks indicate p<0.01 (versus wild type) by ANOVA with Dunnett test. (C-E) Intersectional cell-specific rescue of mod-1 using an inverted Cre-Lox strategy. (C) Schematic depicting transgenes. (D) Rescue of mod-1 by expression in AIY, RID and ASI. (E) Rescue of mod-1 by expression in RIF. For (D-E), asterisks indicate p<0.05 by ANOVA with Bonferroni-Dunn post-hoc test. (F) Laser ablations of individual mod-1-expressing neurons. RID, PVN and RIF were ablated in mod-1 mutants; AIY and ASI were ablated in wild-type animals. Asterisks indicate p<0.05, t-test. (G) Serotonin promotes dwelling by inhibiting MOD-1-expressing neurons that promote roaming. All data are shown as means ± SEM. See also Fig. S3.

Figure 4. Changes in NSM and AIY calcium levels correlate with behavioral transitions

(A) Representative image from a video recording of a freely-moving NSM::GCaMP5 transgenic animal on a bacterial lawn. Asterisk indicates the position of the NSM neuron; gut autofluorescence is also visible. (B) NSM calcium imaging in a freely-moving animal as in (A). Arrows mark calcium peaks (see Extended Experimental Procedures for identification criteria). Roaming states are abbreviated in the calcium imaging device due to the small viewing field and bacterial lawn; nevertheless, we observe clusters of forward runs that are related to roaming states. (C-F) Averaged NSM calcium levels, speeds, and forward runs in wild-type and mod-1 animals. (C,E) Event-triggered average aligning NSM calcium peaks with locomotion speed and forward runs in wild-type animals (C, n=112; p<0.001 for calcium levels and speed, before vs. during peak, paired t-test) and mod-1 mutants (E, n=61; p<0.001 for speed during calcium peak in wild type vs. mod-1, t-test on values normalized to pre-event baseline; p<0.001 for fraction of animals in forward run, wild-type vs. mod-1, chi-squared test). (D,F) Event-triggered average aligning forward runs with NSM calcium signal in wild-type animals (D, n=51; p<0.01 for NSM calcium during pre-event baseline vs. forward run or -60 to -20 sec, paired t-test) and mod-1 mutants (F, n=32; p<0.01 for NSM calcium during pre-event baseline vs. forward run, paired t-test, but no significant difference at -60 to -20 sec). (G) Event-triggered average aligning forward runs with AIY calcium signal in wild-type animals (n=27; p<0.01 for AIY calcium during pre-event baseline vs. forward run, paired t-test). For (C-G), horizontal dashed lines indicate pre-event baseline calcium signals and speed, and data are shown as means ± SEM. See also Fig. S4.

Figure 5. Optogenetic manipulations of a serotonergic neural circuit

(A) ChR2-mediated activation of serotonergic neurons in animals that were roaming (red) prior to LED illumination. (B) ARCH-mediated silencing of serotonergic neurons in animals that were dwelling (blue) prior to LED illumination. (C) ARCH-mediated silencing of mod-1-expressing neurons in animals that were roaming prior to LED illumination. (D) ChR2-mediated activation of mod-1-expressing neurons in animals that were dwelling prior to LED illumination. For (A-D), asterisks indicate p<0.001 compared to control strain, chi-squared test, and controls (grey lines) were identically treated wild-type animals. Blue light delivery was followed by 60s of green light delivery to inactivate ChR2(C128S). (E) Long-lasting effects of mod-1::ARCH activation. Left: Experimental design. Right: Durations of control dwelling states and dwelling states that overlapped with ARCH activation. Data are shown as medians ± 95% confidence intervals. Asterisk indicates p<0.05, Wilcoxon Rank Sum test. See also Fig. S5.

Figure 6. PDF signaling controls exploration behavior

(A) Exploration behavior of PDF signaling mutants. **p<0.01 and ***p<0.001 by ANOVA and Bonferroni-Dunn post-hoc test. (B-C) Fraction of time spent in roaming and dwelling states, and dwelling and roaming state durations in (B) pdfr-1 and (C) pdf-1; pdf-2 mutants, shown as in Fig. 2C,D. Asterisks indicate p<0.01, t-test. (D,E) Cell-specific deletion of pdf-1 using a Cre/Lox strategy. (D) Schematic depicting transgenes. (E) Exploration behavior. *p<0.05, **p<0.01, ***p<0.001 by ANOVA with Dunnett test. (F-H) Intersectional cell-specific rescue of pdfr-1 using an inverted Cre-Lox strategy. (F) Schematic depicting transgenes. (G) Rescue of pdfr-1 by expression in all pdfr-1-expressing neurons. (H) Partial rescue of pdfr-1 by expression in AIY, RIM and RIA. For (G-H), asterisks indicate p<0.01 by ANOVA and Bonferroni-Dunn post-hoc test. All data are shown as means ± SEM. (I) Neural circuit for exploration behavior. Synapses and gap junctions in the C. elegans wiring diagram are in black; neuromodulatory connections defined here are in red (serotonin) and green (PDF). NSM neurons are implicated in feeding and HSN neurons in egg-laying. AIY and RIM neurons regulate reversal frequencies. ASI neurons are sensory neurons (triangles) that sense food, pheromones, and peptide cues to regulate dauer larva development. RIA interneurons regulate head curving during locomotion. AVB interneurons are forward command neurons in the motor circuit. PVP, SIAV (not diagrammed here), and RIF interneurons do not have other known functions. See also Fig. S6.

Figure 7. pdfr-1acts through cAMP signaling, and relationship of serotonin and PDF signaling pathways

(A) Roaming and dwelling states in pdfr-1::acy-1(P260Sgf) animals, shown as in Fig. 2C. (A₁) Fraction of time spent in roaming and dwelling states. (A₂) Dwelling state durations. (A₃) Roaming state durations. Asterisks indicate p<0.01, t-test. (B) Exploration behavior of pdfr-1::acy-1(P260Sgf) animals. Asterisks indicate p<0.01, ANOVA and Bonferroni-Dunn post-hoc test (versus wild type). (C) Optogenetic strategy for mimicking acute PDFR-1 activation. (D) BlaC activation in pdfr-1-expressing neurons in animals that were dwelling prior to LED illumination. Controls (grey line) were identically treated wild-type animals. Asterisks indicate p<0.05, chi-squared test. (E) Roaming and dwelling in single and double mutants. (E₁) Fraction of time spent in roaming and dwelling states. (E₂) Dwelling state durations. (E₃) Roaming state durations. Asterisks indicate p<0.05, ANOVA and Bonferroni-Dunn post-hoc test (for (E₃), each genotype compared to wild type). (F) pdfr-1::BlaC activation induces roaming behavior in mod-1; pdfr-1 double mutants that were dwelling at the time of LED illumination. (G) mod-1::ARCH activation induces dwelling behavior in mod-1; pdfr-1 double mutants that were roaming at the time of LED illumination. Controls (gray lines) were matched mutant animals with no LED illumination. Asterisks indicate p<0.01, chi-squared test. All data are shown as means ± SEM. See also Fig. S7.