A modular library of small molecule signals regulates social behaviors in Caenorhabditis elegans

Abstract

The nematode C. elegans is an important model for the study of social behaviors. Recent investigations have shown that a family of small molecule signals, the ascarosides, controls population density sensing and mating behavior. However, despite extensive studies of C. elegans aggregation behaviors, no intraspecific signals promoting attraction or aggregation of wild-type hermaphrodites have been identified. Using comparative metabolomics, we show that the known ascarosides are accompanied by a series of derivatives featuring a tryptophan-derived indole moiety. Behavioral assays demonstrate that these indole ascarosides serve as potent intraspecific attraction and aggregation signals for hermaphrodites, in contrast to ascarosides lacking the indole group, which are repulsive. Hermaphrodite attraction to indole ascarosides depends on the ASK amphid sensory neurons. Downstream of the ASK sensory neuron, the interneuron AIA is required for mediating attraction to indole ascarosides instead of the RMG interneurons, which previous studies have shown to integrate attraction and aggregation signals from ASK and other sensory neurons. The role of the RMG interneuron in mediating aggregation and attraction is thought to depend on the neuropeptide Y-like receptor NPR-1, because solitary and social C. elegans strains are distinguished by different npr-1 variants. We show that indole ascarosides promote attraction and aggregation in both solitary and social C. elegans strains. The identification of indole ascarosides as aggregation signals reveals unexpected complexity of social signaling in C. elegans, which appears to be based on a modular library of ascarosides integrating building blocks derived from lipid β-oxidation and amino-acid metabolism. Variation of modules results in strongly altered signaling content, as addition of a tryptophan-derived indole unit to repellent ascarosides produces strongly attractive indole ascarosides. Our findings show that the library of ascarosides represents a highly developed chemical language integrating different neurophysiological pathways to mediate social communication in C. elegans.

Figure 1. Identification of indole ascarosides as daf-22-dependent metabolites.

(A) Chemical structures of important ascarosides . (B) Schematic representation of Differential Analysis via 2D-NMR spectroscopy (DANS). Comparison of wild-type NMR spectra with daf-22 mutant NMR spectra enabled detection of spectroscopic signals that may represent daf-22-dependent signaling molecules. (C) Small section of the actual wild-type and daf-22 NMR spectra used for DANS. Signals of indole carboxylic acid are present in both spectra (green box), whereas another indole-derived signal (red box) is only present in the wild-type, but not the daf-22 spectrum. (D) HPLC-MS-based comparison of the wild-type and daf-22 metabolomes. Ion chromatograms obtained for wild-type show peaks for the molecular ions of five different indole ascarosides which are absent from the daf-22 chromatograms. (E) Structures of identified indole ascarosides. (F) Structures of additional non-indole ascarosides identified in this study. (G) Relative amounts of indole ascarosides icas#3 and icas#9 and non-indole ascarosides ascr#3 and ascr#9 secreted by C. elegans N2 grown in liquid culture, as determined by HPLC-MS analyses of media extracts (normalized to concentration of ascr#3; n = 5, ±SEM). For mass spectrometric quantification of indole and non-indole ascarosides, standard mixtures of authentic reference compounds were used.

Figure 2. Indole ascarosides attract C. elegans hermaphrodites and males.

(A) Schematic representation of the bioassay used to measure attraction behavior in worms. Zone A is the region where the sample or control solution is applied. The red X denotes the initial position of the assayed worms. (B) Schematic representation of a quadrant chemotaxis assay. A red X denotes the spot where washed worms are placed at the beginning of the assay. The shaded regions of the quadrant plate indicate the agar containing the chemical, whereas the white regions denote control agar. The number of animals in each quadrant was counted after 30 min and a chemotaxis index was computed (see Materials and Methods). The chemotaxis index for the schematic is 0.84. (C) icas#1, icas#3, and icas#9 are attractive to both C. elegans sexes. All three compounds were assayed at 1 pmol using N2 hermaphrodites and him-5 males. Open bars: no compound (solvent vehicle) controls. (D) Dose dependence of icas#3 response for N2 hermaphrodites and him-5 males in the spot attraction assay (*p<0.01, **p<0.001, ***p<0.0001, unpaired t test with Welch's correction). (E) Dose dependence of icas#3 attraction for N2 hermaphrodites in the quadrant chemotaxis assay (one-factor ANOVA with Dunnett's post-test, **p<0.01).

Figure 3. Social and solitary wild-type hermaphrodites are attracted to icas#3, but not to non-indole ascarosides.

(A) Solitary and social wild-type hermaphrodites are not attracted to a physiological ascr#2,3,5 mixture in the spot attraction assays, in contrast to npr-1(ad609) mutant worms (***p<0.0001, unpaired t test with Welch's correction). (B) In the quadrant chemotaxis assay, hermaphrodites from all tested strains are attracted to 1 pM icas#3 and repelled by a physiological mixture of non-indole ascarosides (10 nM ascr#2,3,5), except for npr-1(ad609) mutant worms, which are also attracted to the ascr#2,3,5 blend (chemotaxis after 30 min; for chemotaxis indices at 15 min, see Figure S3C). (C) Dose-dependence of icas#3 attraction for social hermaphrodites in the quadrant chemotaxis assay (Figure 3B,C: *p<0.05, **p<0.01, one-factor ANOVA with Dunnett's post-test). (D) Social wild-type hermaphrodites and npr-1(ad609) mutant worms are attracted to icas#3 in the spot attraction assay (**p<0.001, ***p<0.0001, unpaired t test with Welch's correction).

Figure 4. Indole ascarosides mediate aggregation behavior in C. elegans.

(A) Aggregation behavior of solitary and social hermaphrodites at low worm densities (20 worms per plate) on different concentrations of icas#3. (B) Aggregation behavior of solitary and social hermaphrodites at high worm densities (∼120 worms per plate) on different concentrations of icas#3. (C) Aggregation of daf-22 hermaphrodites at low worm density on two different concentrations of icas#3. (D) Mean stopped duration of N2 hermaphrodites at different icas#3 concentrations (Figure 4A–D: *p<0.05, **p<0.01 one-factor ANOVA with Dunnett's post-test). (E) Aggregation (red arrow) of N2 hermaphrodites (20 worms per plate) on plates containing 10 pM of icas#3 compared to behavior on control plates. (F) Aggregation of N2 hermaphrodites (∼120 worms per plate) on plates containing 1 pM of icas#3 compared to behavior on control plates.

Figure 5. Response to icas#3 in N2 hermaphrodites is mediated by ASK sensory neurons and the downstream AIA interneurons.

(A) Schematic representation of the connectivity of the ASK sensory neuron to other neurons. The primary synaptic output of ASK is the AIA interneuron. (B) Attraction of hermaphrodites to icas#3 is dependent on the ASK sensory neurons and the AIA interneurons. Ablation of the RMG interneuron does not affect attraction of N2 or ncs-1::gfp hermaphrodites to icas#3 (*p<0.01, ***p<0.0001, unpaired Student's t test with Welch's correction). (C) Aggregation of N2 and ASK-ablated hermaphrodites at low worm density (20 worms per plate). ASK-ablated worms do not aggregate in response to icas#3. (D) icas#3 induces G-CaMP fluorescence signals in AIA interneurons. The colored traces represent fluorescence changes in the AIA neurons of individual animals upon exposure to 1 µM icas#3. The black traces represent fluorescent changes of individual animals upon exposure to buffer. The grey shading indicates presence of icas#3, n = 10 animals. (E) Average AIA fluorescence change in animals exposed to either buffer or icas#3 (**p<0.01, unpaired Student's t test with Welch's correction). Error bars indicate standard error of mean (S.E.M).

Figure 6. Emerging model for a modular language of signaling molecules.

(A) icas#3 and ascr#3 are competing signals for N2 hermaphrodites. Mixtures of 120 fmol ascr#3 and 10 fmol icas#3 (Condition 1) attract worms to zone A, whereas larger amounts of a mixture of the same ratio (Condition 2; 12 pmol ascr#3 and 1 pmol icas#3) deter worms from zone A and instead attract to the periphery (zones B and C). In experiments with Condition 2, only one worm entered the treated zone A, whereas 31 worms entered control zone A (***p<0.0001, unpaired Student's t test with Welch's correction). (B) Synergistic blends of non-indole ascarosides induce dauer at nanomolar to micromolar concentrations and function as a male attractant at picomolar to nanomolar concentrations, whereas indole ascarosides icas#3 and icas#9 act as hermaphrodite attractants and aggregation signals at femtomolar to picomolar concentrations. (C) Modular assembly of C. elegans signaling molecules, based on building blocks derived from tryptophan (green), fatty acids (black), p-aminobenzoic acid (PABA, red), and carbohydrate metabolism (blue).