Co-option of neurotransmitter signaling for inter-organismal communication in C. elegans

Biogenic amine neurotransmitters play a central role in metazoan biology, and both their chemical structures and cognate receptors are evolutionarily conserved. Their primary roles are in cell-to-cell signaling, as biogenic amines are not normally recruited for communication between separate individuals. Here, we show that in the nematode C. elegans, a neurotransmitter-sensing G protein-coupled receptor, TYRA-2, is required for avoidance responses to osas#9, an ascaroside pheromone that incorporates the neurotransmitter, octopamine. Neuronal ablation, cell-specific genetic rescue, and calcium imaging show that tyra-2 expression in the nociceptive neuron, ASH, is necessary and sufficient to induce osas#9 avoidance. Ectopic expression in the AWA neuron, which is generally associated with attractive responses, reverses the response to osas#9, resulting in attraction instead of avoidance behavior, confirming that TYRA-2 partakes in the sensing of osas#9. The TYRA-2/osas#9 signaling system represents an inter-organismal communication channel that evolved via co-option of a neurotransmitter and its cognate receptor.


Introduction
Inter-organismal communication occurs in several forms across the animal kingdom, both within and between species: prairie dogs use audio alarm calls to signal danger to conspecifics (1), birds display ornate visual cues and dances to attract mates (2), and honeybees dance to signal food location (3). Less apparent, though ancient and ubiquitous across all kingdoms of life, is chemical communication, which underlies social responses driven by chemosensation (4)(5)(6)(7). Social chemical communication requires both intra-and inter-organismal signaling. First, a chemical cue is released into the environment by one organism that is then detected by specific receptors in another organism. Upon sensation, intra-organismal signaling pathways, e.g. neurotransmitter signaling, are activated that ultimately coordinate a social response.
Neurotransmitter monoamines such as dopamine, serotonin, tyramine and octopamine serve diverse functions across kingdoms (8). The associated signaling pathways often rely on highly regulated compound biosynthesis, translocation, either by way of diffusion or through active transport, and finally perception by dedicated chemoreceptors. Many neurotransmitters are perceived via G protein-coupled receptors (GPCRs); in fact, there appears to be a close relationship between GPCR diversification and neurotransmitter synthesis in shaping neuronal systems (9). Notably, the most common neurotransmitters share similar behavioral functions across phyla, for example, serotonin is commonly involved in regulating food responses (10)(11)(12). Other neurotransmitters, such as tyramine and octopamine, are only found in trace amounts in vertebrates, and in invertebrates act as adrenergic signaling compounds (13)(14)(15).
The nematode Caenorhabditis elegans affords many advantages for studying social chemical communication and neuronal signaling, namely, the animal's tractability, well-characterized nervous system, and social behavioral responses to pheromones (16,17). C. elegans secretes a class of small molecules, the ascaroside pheromones, which serve diverse functions in inter-organismal chemical signaling (18)(19)(20). As a core feature, these molecules include an ascarylose sugar attached to a fatty acid-derived side chain that can be optionally decorated with building blocks from other primary metabolic pathways (21). Ascaroside production, and thus the profile of relayed chemical messages, is strongly dependent on the animal's sex, life stage, environment, and physiological state (22)(23)(24)(25). Depending on their specific chemical structures and concentration, the effects of ascaroside signaling vary from social (e.g. attraction to icas#3) to developmental (e.g. induction of dauer by ascr#8) (Fig. 1A) (25)(26)(27)(28).
Recently, an ascaroside, named osas#9, that incorporates the neurotransmitter octopamine was identified (22). Osas#9 is produced in large quantities specifically by starved L1 larvae and elicits aversive responses in starved, but not well fed conspecifics (22). The dependency on starvation of both its production and elicited response suggests osas#9 relays information on physiological status and unfavorable foraging conditions. However, it is unknown how osas#9 is perceived and drives starvation-dependent behavioral responses. Based on the unusual incorporation of a monoamine neurotransmitter building block in osas#9, we asked whether other components of monoamine signaling pathways have been recruited for inter-organismal signaling via osas#9. Here, we show that TYRA-2, an endogenous trace amine receptor, is required for the perception of osas#9, demonstrating co-option of a neurotransmitter and a neurotransmitter receptor for inter-organismal communication.

Aversive responses to osas#9 require the GPCR TYRA-2
Previous work showed that production of the ascaroside osas#9 ( Fig. 1A) is starkly increased in starved L1 larvae and elicits avoidance behavior in starved young adult hermaphrodites using a behavioral drop test assay (Fig. 1B) (22). This starvation dependent response is reversible: when animals are starved for an hour, and then reintroduced to food for two hours, no avoidance behavior is observed (Fig. S1A). For the current study we tested a broader range of conditions. We found that osas#9 elicits avoidance regardless of sex or developmental stage of animals (Fig. 1C), and that osas#9 is active over a broad range of concentrations (fM -µM) (Fig. S1B). 1 µM osas#9 was used for the remainder of this study unless otherwise noted (Fig. 1D). Ascarosides such as the male attractant ascr#3 and aggregation ascaroside icas#3 show activity profiles that are similarly broad as that of osas#9, whereas others, such as the mating cue ascr#8, are active only within more narrow concentration ranges (26,34,35).
TYRA-2 is a G protein-coupled receptor (GPCR) that has been shown to bind tyramine with high affinity and octopamine to a lesser extent (38). To exclude the possibility that tyra-2 is necessary for avoidance behaviors in general, we subjected tyra-2 lof animals to three well-studied chemical deterrents, SDS, copper chloride (CuCl2), and glycerol. No defects were found in the animals' ability to respond aversively to these deterrents (Fig. 2C). This indicates that tyra-2 is specifically required for osas#9 avoidance and is not part of a generalized unisensory avoidance response circuit. Since the response to osas#9 is dependent on physiological state, we examined whether tyra-2 transcript levels changed under starved versus fed conditions using RT-qPCR. Starved animals exhibited a nearly two-fold increase in tyra-2 expression (Fig. S2C).
We then asked whether tyramine signaling is required for the osas-9 avoidance response as tyra-2 is known to bind to tyramine (38). We assayed two tdc-1 lof mutants, which lack the ability to synthesize tyramine (44). We observed that the behavioral response to osas#9 was unaltered in animals lacking tyramine biosynthesis (Fig. 2D).
This demonstrates that the function of TYRA-2 in osas#9 avoidance is independent of tyramine, suggesting that TYRA-2 may be involved in perception of a ligand other than tyramine to promote aversive response to osas#9.

tyra-2 is required in the ASH sensory neuron for physiological osas#9 response
We next asked where tyra-2 is acting in the osas#9 aversion pathway. To determine the site of action of tyra-2 in osas#9 avoidance, we designed a tyra-2 translational fusion construct consisting of the entire genomic locus, including 2kb upstream, fused to GFP (ptyra-2::TYRA-2::GFP). We observed TYRA-2 expression in four sensory neurons: ASH, ASE, ASG, and ASI (Fig. 3A). These results are in agreement with previous expression studies on tyra-2 localization (38) (Fig. 3A). We laser-ablated individual amphid sensory neurons to determine if a tyra-2 expressing sensory neuron is required for the response.
This revealed that ASH neurons are required for osas#9 response, whereas ablation of other neurons did not have a strong effect (Fig. 3B). We observed a slight reduction in the magnitude of the osas#9 aversive response in ASE-and ASI-laser-ablated animals ( Fig. 3B); however, ASH/ASE and ASH/ASI double ablated animals did not differ in response from animals with ASH ablated alone, and ASE/ASI ablated animals did not differ from ASE or ASI alone (Fig. 3B). We then tested ASH, ASE, and ASI genetic ablation lines (45)(46)(47)(48) and observed that at all tested concentrations, only ASH genetic ablation line resulted in complete abolishment of osas#9 avoidance (Fig. S3A,B,C). As with the laser ablation studies, we observed a slight decrease in osas#9 avoidance in ASE and ASI ablated animals (Fig. S3A,B,C) consistent with the findings for laser-ablated animals. Neurons not expressing tyra-2 showed no defect in response to osas#9 (Fig.   S3D). Our data implies that osas#9 is primarily sensed by ASH sensory neurons and that the ASE and ASI sensory neurons can potentially contribute by sensitizing ASH sensory neurons or by regulating downstream interneurons within the osas#9 response circuit.
To further elucidate the role of the ASH sensory neurons and TYRA-2 in osas#9 sensation, we utilized a microfluidic olfactory imaging chip that enables detection of calcium transients in sensory neurons (49,50). We observed that, upon exposure to 1 µM osas#9, wildtype animals expressing GCaMP3 in the ASH sensory neurons exhibit robust increase in fluorescence upon stimulus exposure (Fig. 3C,D and Supplementary Video 1). Animals lacking tyra-2 displayed no changes in fluorescence upon osas#9 exposure (Fig. 3C,D). These findings imply that tyra-2 activity is necessary in ASH sensory neurons to sense and elicit osas#9 physiological responses.
Given that tyramine and octopamine are known ligands of TYRA-2, we also tested whether these neurotransmitters elicit aversive responses in C. elegans (38). Previous studies have shown that both tyramine and octopamine inhibit serotonin food-dependent increases in aversive responses to dilute octanol via specific G protein-coupled receptors (40). Both biogenic amines exhibited aversive behaviors at non-physiological concentrations much higher than required for osas#9, 1 mM for tyramine and octopamine compared to 1 µM for osas#9 ( Fig. S4A,B, S1B). Similarly, high concentrations of tyramine (1mM) elicited calcium transients in ASH::GCaMP3 but lower concentrations (1 µM) did not show calcium changes (Fig. S4C,D). Worms exposed to 1 mM octopamine displayed minimal change in calcium transients (Fig.S4C,D). These data show that the TYRA-2 receptor in the ASH sensory neurons is specifically involved in the avoidance response to osas#9. Tyramine or octopamine do not appear to be participating in the avoidance response, in agreement with the finding that tyramine biosynthesis is not required for avoidance to osas#9 (Fig. 2D).

tyra-2 expression confers the ability to sense osas#9
Since expression of tyra-2 in the ASH sensory neurons is required for calcium transients in response to osas#9, we asked whether tyra-2 expression in the ASH neurons is sufficient to rescue the osas#9 behavioral response in tyra-2 lof animals. Expression of tyra-2 under the nhr-79 promoter, which is expressed in the ASH and ADL neurons, fully restored osas#9 avoidance (Fig. 4A,B) (51). To test whether expression of tyra-2 in the ADL neurons is required for the phenotypic rescue, we ablated the ADL neurons in the transgenic animals. Ablation of the ADL neurons did not affect avoidance to osas#9 (Fig.   4C). Additionally, injection of the tyra-2 translational reporter into tyra-2 lof animals displayed sub-cellular localization in the ASH sensory cilia (Fig. 4D) and was observed to be functional as osas#9 aversion is rescued in these animals (Fig. 4E). These results affirm that the aversive behavioral response to osas#9 is dependent on tyra-2 expression in the ASH neurons.
Previous studies in C. elegans indicate that behavioral responses (such as aversion or attraction) elicited by an odorant are specified by the olfactory neuron in which the receptor is activated in, rather than by the olfactory receptor itself (31, 52). Therefore, we asked whether expression of TYRA-2 in AWA neurons, which are generally involved in attractive responses to chemical cues (53,54) would switch the behavioral valence of osas#9, resulting in attraction to osas#9, instead of aversion. Misexpression of tyra-2 in the AWA sensory neurons in a tyra-2 lof background did not result in avoidance of osas#9, in contrast to expression of tyra-2 in the ASH neurons (Fig. 5A). We then performed a leaving assay to test for attraction to osas#9 in the worms expressing tyra-2 in the AWA neurons. This assay involves the placement of animals into the center of a NGM agar plate where osas#9 is present and measuring the distance of animals from the origin in one-minute intervals (Fig. 5B). tyra-2 lof animals displayed osas#9 leaving rates equal to the solvent control (Fig. 5C, S5), whereas worms misexpressing tyra-2 in the AWA neurons displayed osas#9 leaving rates lower than that for solvent controls, indicating attraction ( Fig. 5C, S5). Furthermore, worms misexpressing tyra-2 in the AWA neurons stayed significantly closer to the origin than either wildtype or tyra-2 lof animals when exposed to osas#9 (Fig. 5C, S5). We confirmed that ectopic expression of tyra-2 in AWA sensory neurons did not alter the native chemosensory parameters of AWA neurons (Fig.   S6A,B). Hence misexpression of tyra-2 in AWA neurons resulted in reprogramming of these nematodes, promoting attraction to the normally aversive compound osas#9.
Finally, we tested whether ectopic expression of tyra-2 in the ADL neurons, which have been shown to detect aversive stimuli (55)(56)(57)(58), results in a behavioral response to osas#9. For this purpose, we ablated the ASH neurons in the pnhr-79::tyra-2 strain, in which tyra-2 is expressed in the ASH and ADL neurons. We found that these ASH ablated animals still avoid osas#9, similar to ADL ablated worms from this rescue line (Fig. 5D).
Ablation of both the ASH and ADL neurons in this strain abolished the avoidance response (Fig. 5D). This implies that mis-expression of tyra-2 in the ADL neurons confers the ability of this neuron to drive avoidance to osas#9. Taken together, results from both misexpression experiments (AWA and ADL neurons) demonstrate that TYRA-2 is necessary and sufficient to elicit osas#9-dependent behaviors.

Gα protein gpa-6 is necessary in ASH sensory neurons for osas#9 avoidance
Since expression of the tyra-2 GPCR is required in ASH neurons for osas#9 response, we sought to identify the Gα subunit necessary for osas#9 avoidance. Eight of the 21 Gα proteins are expressed in subsets of neurons that include the ASH sensory pair (gpa-1, gpa-3, gpa-6, gpa-11, gpa-13, gpa-14, gpa-15, and odr-3) (59-61). We tested mutants for each of those eight Gα subunits for their response to osas#9, (Fig. 6A) and found that gpa-6 lof animals do not avoid osas#9 (Fig. 6A). To determine whether gpa-6 is necessary in ASH sensory neurons to mediate osas#9 responses, we expressed gpa-6 using pnhr-79 in the ASH neurons in a gpa-6 lof background. These animals displayed wildtype behavior when tested for osas#9 avoidance (Fig. 6B). To characterize cellular and subcellular localization of the gpa-6 Gα subunit, we created a full length RFP translational fusion of the entire gpa-6 locus including 4kb upstream. We detected gpa-6 expression in the soma of AWA and ASH sensory neurons (Fig. 6C), in agreement with previous studies (60). However in addition to ASH soma localization, the translational fusion revealed presence of gpa-6 in ASH cilia (Fig. 6C). Behavioral rescue by gpa-6 expression specifically in the ASH neurons and its ciliary localization, support that this Gα subunit functions in mediating osas#9 avoidance.

Discussion
How does a worm survive in changing environmental and physiological conditions? Given C. elegans' complex ecology and a boom and bust lifestyle, worms need to make frequent adaptive developmental and physiological choices (62). The octopamine-derived pheromone osas#9, secreted in large quantities by L1 larvae under starvation conditions, appears to promote dispersal away from unfavorable conditions (Fig. 7). Here we show that this pheromone is detected by the GPCR tyra-2, a canonical neurotransmitter receptor that is expressed in the ASH sensory neurons. To our knowledge this is the first instance in which a "repurposed internal receptor" partakes in pheromone perception.
However, by ectopically expressing tyra-2 in ADL and AWA neurons , we were able to elicit responses characteristic to each neuron (Fig. 5). These data show that the response to osas#9 depends on the neuron tyra-2 is expressed in, providing additional support for direct involvement of TYRA-2 in chemosensation of osas#9. Alternatively, a different receptor that directly interacts with TYRA-2 and is expressed in the ASH, ADL, and AWA neurons could bind osas#9.
Our data suggests that ASE and ASI sensory neurons may regulate ASH sensitivity during osas#9 avoidance serving as modulators at the sensory level, similar to previously observed cross inhibition of ASI and ASH neuronal activity in avoidance to copper, and decision making based on physiological state (73,74). Alternatively, these neurons could be interacting with ASH neuronal targets in the osas#9 response, strengthening or dampening the relayed signal, possibly through peptidergic or aminergic signaling to establish the functional circuit. Recent studies have shown that tyra-2 is necessary for binding tyramine in a RIM-ASH feedback loop in multisensory decision making (75). Animals lacking TYRA-2, or the tyramine biosynthetic enzyme TDC-1, crossed a 3M fructose barrier towards an attractant, diacetyl, faster than wildtype C.
elegans. This demonstrated the endogenous role of tyramine binding to TYRA-2 increasing avoidance in multisensory threat tolerance (75); however, our results show that tyramine signaling is not involved in the response to osas#9. It will be interesting to elucidate the role other neurons or tissues and neuromodulatory signaling have in shaping the osas#9 response. Such modulation of the osas#9 response circuitry remains to be investigated.
Our findings demonstrate that TYRA-2, a member of a well conserved family of neurotransmitter receptors, functions in chemosensation of osas#9, a neurotransmitterderived inter-organismal signal. Typically, neurotransmitter signaling is intra-organismal, facilitating cell-to-cell communication. This involves the highly regulated biosynthesis of specific chemical compounds, e.g. biogenic amines, their translocation (either by way of diffusion or through active transport), and, finally, perception by dedicated chemoreceptors (76). This mode of communication is strikingly similar to pheromone communication between organisms, as it involves highly specific production and reception of ligands for communication. As evolution is opportunistic, it stands to reason that some machinery from intra-cellular signaling would be utilized for inter-organismal signaling. Indeed, co-option has been hinted at before, in both the trace amine associated receptor (TAAR) and formyl peptide receptor-like (FPRL) receptor classes, both of which are involved in inter-organismal signaling (77)(78)(79)(80). Of the TAARs, only TAAR1 and TAAR2 have been found to have endogenous roles: TAAR1 in mammalian CNS, and both TAAR1 and TAAR2 in leukocyte migration (78,81). Additionally, TAAR2 mRNA has been detected in mouse olfactory epithelium, suggesting it may be involved in both intra-and inter-organismal signaling (77). However, no odor molecules have been linked to TAAR2 in the olfactory epithelium.
How key innovations in metazoan complexity could have evolved from pre-existing machineries is of great interest (82). Our findings demonstrate that the tyramine receptor TYRA-2 functions in chemosensation of osas#9, a neurotransmitter-derived interorganismal signal, thus revealing involvement of both neurotransmitter biosynthesis and neurotransmitter reception in intra-and inter-organismal signaling. Therefore, evolution of an inter-organismal communication channel co-opted both a small molecule, octopamine, and the related receptor TYRA-2, for mediating starvation-dependent dispersal in C. elegans (Fig. 7), suggesting that such co-option may represent one mechanism for the emergence of new inter-organismal communication pathways.

Avoidance drop test
In this assay, the tail end of a forward moving animal is subjected to a small drop (~5 nl) of solution, delivered through a hand-pulled 10 μl glass capillary tube. The solution, upon contact, is drawn up to the amphid sensory neurons via capillary action. In response, the animal either continues its forward motion (scored as "no avoidance response"), or displays an avoidance response within four seconds (83). The avoidance response is characterized by a reversal consisting of at least one half of a complete "head swing" followed by a change in direction of at least 90 degrees from the original vector. For quantitative analysis, an avoidance response is marked as a "1" and no response as a ASH::tyra-2 rescue: For ASH expression, a 3 kb nhr-79 promoter was isolated from genomic DNA using primers attB1gtgcaatgcatggaaaattg and attB5ratacacttcccacgcaccat.
Translational fusions: tyra-2::GFP fusions were created by PCR fusion using the following primers to isolate 2kb ptyra-2 with its entire genomic locus from genomic DNA: A) atgttttcacaagtttcaccaca, A nested) ttcacaagtttcaccacattacaa, and B with overhang) AGTCGACCTGCAGGCATGCAAGCT gacacgagaagttgagctgggtttc. GFP primers as described in WormBook (84). The construct was then co-injected with pelt-2::mCherry into both N2 and FX01846. We observed that RNAi worked best when animals were cultured at 15ºC. We used the nre-1(hd20);lin-15B(hd126) (VH624) strain for the RNAi studies as it has been previously shown to be sensitive to neuronal RNAi (42, 43).

Laser ablations
Laser ablations were carried out using DIC optics and the MicroPoint laser system following the procedures as outlined in Fang-Yen et al. 2012 (86,87). Ablated animals were assayed 72 hours later, at the young adult stage. All ablated animals were tested in parallel with control animals that were treated similarly as ablated animals but were not exposed to the laser microbeam. Calcium imaging was perfomed by using a modified olfactory chip as described in Reilly et al 2017 (49,50). A young adult animal was immobilized in a PDMS olfactory chip with its nose subject to a flowing solution. Animals were imaged at 40x magnification for 30 seconds, and experienced a 10 second pulse of osas#9 in between the solvent control.
Each animal was exposed to the stimulus three times. Soma fluorescence from GCaMP3 was measured using ImageJ. Background subtraction was performed for each frame to obtain the value ΔF. Change in fluorescence (ΔF/F0) was calculated by dividing the ΔF value of each frame by F0. F0 was calculated as the average ΔF of 10 frames prior to stimulus exposure (50).

RT-qPCR
RNA was isolated from individual animals, either freshly removed from food or after four hours of starvation using Proteinase K buffer as previously published (89 Both spots contained sodium azide for anesthetizing animals that entered the region.
After 45 minutes, the chemotaxis index was calculated by subtracting the number of animals in the solvent control from the number of animals in the solution of interest and divided by the total number of animals.

Leaving Assay
The leaving assay consisted of the use of 60 mm culture plates containing standard NGM agar. A transparency template that included a 6mm diameter circle in the center was attached to the underside of the NGM plate. One hour before running the assay, young adult animals were passed on to an unseeded plate and allowed to starve for one hour.
100 µl of E. coli OP50 liquid culture was spread onto a separate NGM assay plates. These plates were allowed to dry at 25 o C without a lid for one hour. After an hour of incubation, 4 µl of either solvent control or 10 pM osas#9 was pipetted onto the agar within the center circle outlined on the template. 10 animals were gently passed into the center circle and their movement was recorded. At one minute intervals, the distance the animals traveled from the origin was measured using ImageJ.

Statistical analysis
Statistical tests were run using Graphpad Prism. For all figures, when comparing multiple groups, ANOVAs were performed followed by Sidak's multiple comparison test. When only two groups were compared, a Student's t-test was used ( Figure 1D, S2C). When comparing different strains/conditions, normalized values of osas#9 avoidance index response relative to the respective solvent control were used. This was done to account for any differences in baseline response to solvent control for the respective genotypes, laser ablations, or physiological conditions. When normalizing fold change of osas#9 response to solvent control response for the avoidance assay within a strain/condition, data was first log transformed so a fold change could still be calculated for control plates that had a "0" value. For avoidance assays, statistical groups were based on the number of plates assayed, not the number of drops/animals. For calcium imaging, averages were calculated by obtaining the max peak value before and during exposure to the chemical of interest for each trial.    Screen for receptors required to mediate osas#9 avoidance. tyra-2 lof animals are defective in osas#9 avoidance response, n ≥ 4 trials. B) Two alleles of tyra-2 lof animals, tm1846 and tm1815, are defective in osas#9 avoidance behavior, n ≥ 4 trials. tyra-2(tm1846) lof animals were used for the remainder of data presented in this manuscript. C) tyra-2 lof mutants showed no significant differences when subjected to known chemical deterrents, n ≥ 3 trials. D) osas#9 avoidance response is not dependent on endogenous tyramine. Two different alleles of tdc-1 lof animals, n3419 and n3420, which lack tyramine biosynthesis, show normal response to osas#9, n ≥ 7 trials. Data presented as mean ± S.E.M; *P<0.05, **P<0.01, ***P<0.001, one factor ANOVA with Sidak's multiple comparison posttest. Asterisks displayed without bar depict compared osas#9 avoidance to respective solvent control within groups. '+' signs represent same p value as asterisks but representing difference between osas#9 avoidance of a strain/conditions in comparison to wildtype. Neurons expressing tyra-2 reporter were ablated using laser microbeam. ASH neuronal ablations resulted in abolished response to osas#9 that was indistinguishable from solvent control. ASE and ASI ablated animals showed a reduced avoidance, but not to the extent of ASH neurons, n ≥ 3 trials with at least 10 ablated animals for each condition.

C,D)
Calcium dynamics of ASH neurons upon osas#9 exposure in a microfluidic olfactory chip. C) ASH::GCaMP3 animals (black) display a change in calcium transients when exposed to osas#9. tyra-2 lof ASH::GCaMP3 animals (red) did not display a change in fluorescence upon stimulation with the chemical. Shaded blue region depicts time when animals were subjected to the stimulus, n = 10 animals, 30 pulses. D) Maximum fluorescence intensity before (solvent control) and during exposure to 1 µM osas#9. Data presented as mean ± S.E.M; *P<0.05, **P<0.01, ***P<0.001, one factor ANOVA with Sidak's multiple comparison posttest. Asterisks depict comparison between osas#9 and respective solvent control. '+' signs represent same p value as asterisks but representing difference between osas#9 avoidance of a strain/conditions in comparison to wildtype.           Video S1. Video of ASH::GCaMP3 animal being stimulated with 1 μM osas#9. osas#9 presented to animal when red dot appears on screen. Blue is low level of fluorescence and red is high fluorescence level.