Strains were cultured on nematode growth media (NGM) plates seeded with Escherichia coli OP50 bacteria as per standard protocols. The following strains were used: N2 (Bristol wild type), FY928 grIs17 [Psra-6::GCaMP3] (wild type expressing GCaMP3 in ASH), VC224 octr-1(ok371) (octr-1 null mutants), FY991 octr-1(ok371); grIs17 [Psra-6::GCaMP3] (octr-1 null mutants expressing GCaMP3 in ASH), FY1030 ser-3(ad1774); grIs17 [Psra-6::GCaMP3] (ser-3 null mutants expressing GCaMP3 in ASH), FY1053 irk-1(n4895); grIs17 [Psra-6::GCaMP3] (irk-1 null mutants expressing GCaMP3 in ASH), FY1056 irk-2(n4896); grIs17 [Psra-6::GCaMP3] (irk-2 null mutants expressing GCaMP3 in ASH), FY1061 irk-3(n5049); grIs17 [Psra-6::GCaMP3] (irk-3 null mutants expressing GCaMP3 in ASH), FY1063 goa-1(n1134); grIs17 [Psra-6::GCaMP3] (goa-1 null mutants expressing GCaMP3 in ASH), MT15934 irk-1(n4895) (irk-1 null mutants), FQ295 irk-2(n4896) (irk-2 null mutants), MT17360 irk-3(n5049) (irk-3 null mutants), and LX1751 lin-15(n765) X; vsEx700 (ASH::PTX worms, expressing pertussis toxin (PTX) in ASH under the control of the osm-10 promoter). The Psra-6::GCaMP3 transgene was obtained as the extrachromosomal array kyEx2865 in strain CX10979 (Kato et al., 2014). The Posm-10::PTX transgene was obtained as extrachromosomal array vsEx700 in strain LX1751 (Hofler and Koelle, 2011). MT15934, FQ295, and MT17360 were kindly supplied by N. Ringstad (Skirball Institute, New York, NY, United States), LX1751 was a kind gift from M. Koelle (Yale University, New Haven, CT, United States), and CX10979 was generously supplied by C. Bargmann.

Calcium imaging experiments were performed as previously described (Mills et al., 2012; Zahratka et al., 2015; Williams et al., 2018b). Worms were glued down using WormGlu cyanoacrylate glue (GluStitch, Delta, BC, Canada), on a 15 mm diameter round coverslip coated with Sylgard (Dow Corning, Midland, MI, United States), immersed in external solution (see below). The coverslips were then placed in a laminar flow chamber (Warner RC26G, Warner Instruments, Hamden, CT, United States) which was continuously perfused with external solution; 150 mM NaCl, 5 mM KCl, 5 mM CaCl₂, 1 mM MgCl₂, 10 mM glucose, 15 mM HEPES, pH 7.30, 327–333 mOsm. Saturated 1-oct (2.37 mM in external solution) was delivered under gravity feed through Luer valves using a perfusion pencil (AutoMate Scientific, Berkeley, CA, United States) or homemade equivalent. Fluorescent tracer sulforhodamine 101 (SR101, 1 μM, Thermo Fisher Scientific) was used in the 1-oct solution for visual inspection of flow and to check for staining of the worm’s nose, indicative of successful stimulus application. All solutions were delivered using the perfusion pencil as described above, mounted on a SF77B Perfusion Fast Step device (step size 200 μM, Warner Instruments, Hamden, CT, United States) to provide precise computer control of pipette position. After each exposure, we visually examined the animals for SR101 staining. No response was observed in ASHs to external solution containing 1 μM SR101 alone. The sra-6 promoter is active in ASH and ASI neurons, which are distinguishable based on their relative positions in the head. ASI neurons did not respond to 1-oct.

We partially dissected the animals to expose the ASH to external solution for Ca²⁺ imaging experiments with high K⁺ stimulation and in experiments where OA was acutely applied to the ASHs. Dissection was performed according to previously described procedure (Goodman et al., 1998). The ASH was exposed by slitting the cuticle with a glass patch pipette (TW150-3, World Precision Instruments) that had been melted, drawn to a fine point and broken back to create a sharp-ended cutter using a Narishige MF-83 microforge (Narishige, Setagaya-ku, Tokyo, Japan). Cutters were mounted on a micromanipulator (Sutter MP285, Sutter Instruments, Novato, CA, United States) to slit open the cuticle, exposing one of the two ASHs to the external bath solution (Williams et al., 2018b). High K⁺ external solution was composed of 120 mM NaCl, 30 mM KCl, 1 mM MgCl₂, 15 mM HEPES, 5 mM CaCl₂, 10 mM glucose, with pH 7.3, 327–333 mOsm. High K⁺ containing external solution was applied using a four-barrel glass puffer with barrel cross section of 300 μm, mounted on a SF77B Perfusion Fast Step device. Solutions were delivered using a syringe pump (KD Scientific, Holliston, MA, United States). Recordings were performed on an Axioskop 2 FS Plus upright compound microscope (40× Achroplan water immersion objective, GFP filter set # 38) (Zeiss, Germany), fitted with an Orca ER CCD camera (Hamamatsu, Skokie, IL, United States) and an automated shutter (Uniblitz, Vincent Associates, Rochester, NY, United States). Minimal illumination intensity was used to prevent GCaMP3 photobleaching, and we did not observe differential photobleaching rates between different genotypes and treatment groups. Exposure times were 50 ms with 4× binning and frame rate of 15 frames/s. Monoamine (OA/5-HT/OA + 5-HT, Sigma-Aldrich, St. Louis, MO, United States) exposure was performed by either incubating the animals on NGM plates containing monoamine(s) at 4 mM each, or by direct application to ASH neurons in partially dissected worms. For plate incubation, each monoamine was added to the NGM agar at 4 mM, and worms were incubated for 30–45 min. OA, 5-HT, OA + 5-HT plates were prepared fresh on each day of recording. Direct application of OA was achieved using a two-opening theta (θ) glass tube that had been heated and drawn to a fine caliber by hand. A PE10 polyethylene tube (0.61 OD × 0.28 ID) (Warner Instruments, Hamden, CT, United States) was inserted into the back of each barrel and sealed using Sylgard to supply external (control) and OA-containing solutions. Solutions flow was regulated using nylon three-way Luer Lock stopcocks; and delivered, using a syringe pump (KD Scientific, Holliston, MA, United States), such that only one barrel of the theta tubing (i.e., buffer or OA) was flowing at a time. Either stream delivered from the dual-chambered pipette was able to shield the exposed ASH from the 1-oct stimulus; inadvertent exposure of exposed ASHs was immediately obvious due to SR101 staining, and such recordings were discarded. Fluorescent images were acquired via MetaVue 7.6.5 (MDS Analytical Technologies, Sunnyvale, CA, United States). For quantifying fluorescence changes in the soma, a square region of interest was drawn, such that the entire soma was the only fluorescent object present. For ASH amphids, a square region was drawn around the ciliated endings located at the worm’s nose, excluding the dendrite. The acquired images were analyzed using Jmalyze software (Rex Kerr). Roughly circular or oval regions of interest limited to the cell soma or amphid were tracked using Jmalyze. Fluorescence amplitude was calculated as ΔF/F_o, where the fluorescence at a given time point (F) minus the fluorescence immediately before stimulus application (F_o) was divided by F_o [i.e., (F − F_o)/F_o]. For quantitative comparisons, the maximal ΔF/F_o over the time course of the stimulus was plotted in the bar graphs and used for statistical analysis. Wild-type non-drug-treated controls were routinely run in parallel with experimental samples.

For patch-clamp analysis, animals were glued and placed in the recording chamber as described above. ASH cell bodies (identified by GCaMP3 expression) were exposed for whole-cell recordings by partially dissecting the animals as described above, taking care to distinguish ASH neurons and ASI neurons. Dissection quality was validated by ensuring that the dendrite of the exposed ASH had not been severed. Whole cell recordings were performed as previously described (Zahratka et al., 2015). Briefly, we used pressure-polished patch pipettes (Goodman and Lockery, 2000) with 12–30 MΩ resistance containing low Cl^– internal solution (15 mM KCl, 115 mM K gluconate, 10 mM HEPES, 5 mM MgCl₂, 0.25 mM CaCl₂, 5 mM EGTA, 20 mM sucrose, 5 mM MgATP, 0.25 mM NaGTP, pH 7.20, 315 mOsm). OA was directly applied on the exposed ASH using a perfusion pencil (AutoMate Scientific, Berkeley, CA, United States) as described above and delivered using a syringe pump (KD Scientific, Holliston, MA, United States). 1-oct was delivered as described above and the cells were observed after each 1-oct application for SR101 staining to ensure there had not been inadvertent exposure of the cell soma to the 1-oct solution. Signals were recorded with an Axopatch 200B amplifier (Molecular Devices, Sunnyvale, CA, United States) in current-clamp mode (0 pA injected current, 10 kHz sampling, 2 kHz filtering), digitized with a Digidata 1440A digitizer, and analyzed using pCLAMP10 software (Molecular Devices, Sunnyvale, CA, United States).

Behavioral responses to 1-oct were assayed as previously described (Chao et al., 2004; Harris et al., 2009). L4 animals were picked the night before the assay onto fresh NGM plates seeded with OP50. Plates containing 4 mM 5-HT, 4 mM OA, or 4 mM 5-HT + 4 mM OA were prepared fresh, 2 h before the experiment. 1-oct was diluted to 30% in 100% ethanol (v/v) and was presented to a forward moving animal via a glass capillary. For assays in the absence of any monoamine, animals were transferred from their regular stock plate to a food free intermediate plate for 1 min to remove any OP50 and then transferred to a food free assay plate and tested 10 min later. In contrast, for assays in the presence of monoamines, animals were transferred to the respective monoamine containing assay plate after the intermediate plate and tested 30 min later. All reagents were obtained from Fisher Scientific (Waltham, MA, United States) or Sigma-Aldrich (St. Louis, MO, United States).

For behavioral experiments, a minimum of 20 young adult hermaphrodites were analyzed for each treatment/mutation. For Ca²⁺ imaging and electrophysiology experiments, minimum of five young adult hermaphrodites were recorded each day, where practical. All experiments were performed between 19°C and 23°C. All reagents were prepared fresh before the experiment. Statistical analysis was performed using unpaired, two-tailed Student’s t-tests or one-way ANOVA with Tukey post-test, using GraphPad Prism software (San Diego, CA, United States) with data presented as mean ± SD.

5-HT decreases the amplitude of 1-oct-dependent Ca²⁺ transients in the soma of ASH neurons while increasing depolarization amplitude and potentiating 1-oct-stimulated avoidance behavior, dependent on SER-5/Gα_q signaling within ASHs (Harris et al., 2009; Zahratka et al., 2015; Williams et al., 2018b). Like 5-HT, OA (4 mM) also reduces 1-oct-stimulated Ca²⁺ transients in the ASH soma (Figures 1A,B; baseline Ca²⁺ levels were not affected by the monoamines, alone or in combination, Supplementary Figure S1). OA does not affect 30% 1-oct avoidance behavior on its own, but does antagonize 5-HT potentiation, dependent on the cell autonomous effects of the OCTR-1 OA receptor in ASHs (Wragg et al., 2007). We confirmed the OA antagonism of 5-HT potentiation (Figure 1C). Despite reversing the 5-HT effect on behavior, OA does not reverse the effect of 5-HT on 1-oct-dependent Ca²⁺ transients in the ASH soma when worms are co-incubated with both monoamines (Figures 1A,B). These results imply that 5-HT and OA activate different signaling pathways which converge on the optically measurable somal Ca²⁺ transient, and indicate that, in terms of behavior, the OA pathway dominates the 5-HT pathway when both are active simultaneously. These results also reinforce the idea that somal Ca²⁺ transient modulation may not be a reliable predictor of a neuron’s output to the downstream locomotory circuitry.

To better understand how OA signaling affects ASH responses to 1-oct, we further investigated the OA intracellular signaling cascade. We investigated octr-1 null mutants to test the potential role for the OCTR-1 receptor, based on previous behavioral results (Wragg et al., 2007). OCTR-1 is a GPCR in C. elegans homologous to the mammalian α2A adrenoceptor (Sellegounder et al., 2018). OCTR-1 binds OA with high affinity and tyramine with lower affinity (Wragg et al., 2007), and activates G_o signaling when expressed in Xenopus oocytes (Mills et al., 2012). The octr-1 promoter drives expression in ASHs as well as several other neurons and non-neuronal cells (Wragg et al., 2007). OCTR-1 is involved in innate immunity and the unfolded protein response in C. elegans (Liu et al., 2016; Sellegounder et al., 2018), and plays a key role in the OA modulation of 1-oct aversive responses (Wragg et al., 2007; Mills et al., 2012). First, we confirmed that OCTR-1 was essential for OA to counteract 5-HT potentiation of 1-oct avoidance behavior in our hands (Figure 2A, compare to Figure 1C), which had been reported previously (Wragg et al., 2007). We then showed the OA-dependent reduction in 1-oct-stimulated Ca²⁺ transients in the ASH soma was absent in octr-1 mutants (Figure 2B). These findings suggest that OCTR-1 initiates a signaling cascade which inhibits both ASH somal Ca²⁺ transients and avoidance behavior. OCTR-1 couples to G_o in Xenopus oocytes (Mills et al., 2012) and G_o signaling is required in ASHs for OA signaling based on neuron-specific RNAi experiments (Harris et al., 2010). To provide further evidence for G_o signaling involvement, we examined worms expressing PTX under the control of the osm-10 promoter, which expresses strongly in ASH and weakly in ASI (Hofler and Koelle, 2011). PTX inactivates G_o signaling through ADP-ribosylation of Gα_o (Gierschik, 1992; Darby and Falkow, 2001). OA inhibition of 5-HT potentiation was defective in these transgenic strains, while 5-HT potentiation itself, and OA inhibition in wild type controls tested in the same experiment, were both normal (Figure 2C), confirming a role for G_o signaling downstream of OCTR-1 in OA modulation of aversive behavior. Unfortunately, we could not test OA modulation of ASH somal Ca²⁺ transients in the ASH::PTX transgenic strains because they also expressed a strong GFP marker in ASH which interfered with GCaMP imaging. Instead, we performed Ca²⁺ imaging in ASHs of goa-1 mutants, which lack Gα_o and are therefore defective in G_o signaling (Ségalat et al., 1995). OA failed to inhibit 1-oct induced somal ASH Ca²⁺ transients in the goa-1 background (Figure 2D), demonstrating a role for G_o signaling downstream of OCTR-1 in OA modulation of somal Ca²⁺ transient amplitudes as well. Taken together, these results support a model in which OA activates OCTR-1 and G_o signaling in ASHs, leading to reduced somal Ca²⁺ transient amplitudes in ASHs and behavioral inhibition.

How does OCTR-1/G_o signaling modulate ASH responses to 1-oct? Gα_o can inhibit olfactory transduction in mammalian olfactory receptor cells (Corey et al., 2021), suggesting that OA could be inhibiting primary sensory transduction at the ASH cilia, at the amphid openings. Ca²⁺ influxes in the ASH cilia are due to activation of OSM-9/OCR-2 TRP channel, downstream of odorant binding to its receptor (Ferkey et al., 2021). Therefore, these influxes can be interpreted as a marker of the efficiency of primary sensory transduction, upstream of any potential effect on cellular excitability (Williams et al., 2018b). 1-oct exposure caused a sharp rise in the Ca²⁺ signal in the amphid region which was not significantly decreased by OA incubation (Figures 3A–C), suggesting that OA did not affect the primary olfactory transduction cascade.

An alternative mechanism for OA to reduce ASH Ca²⁺ transients could be direct inhibition of L-type VGCCs by OA signaling. We previously observed this phenomenon in the ASH soma, where 5-HT signaling through SER-5/Gα_q induced the release of Ca²⁺ from intracellular stores, leading to calcineurin (CaN)-dependent dephosphorylation and inactivation of the L-VGCC (Williams et al., 2018b). G_o signaling can also directly inhibit L-VGCC activation through Gβγ subunits, which are released upon dissociation of activated heterotrimeric G proteins (Waterman, 2000). To test whether these mechanisms were operating, we depolarized neurons with high K⁺-containing external solution after partially dissecting worms to expose the soma of one ASH neuron to the bath. Exposure to solutions containing 30 mM K⁺ reliably depolarizes neurons leading to activation of L-VGCCs, which allows direct inhibition of L-VGCCs by monoamine signaling to be detected (Williams et al., 2018b). Importantly, this approach bypasses any upstream steps in the neuronal activation pathway, so any effect of monoamine signaling would necessarily be due to modulation of L-VGCC open channel probabilities (Williams et al., 2018b). Switching from normal external solution to high K⁺ external solution led to robust Ca²⁺ transients in the soma of bath-exposed ASH neurons (Figures 3D,E). These transients were strongly inhibited by 5-HT treatment, as previously reported (Williams et al., 2018b), but were not significantly reduced by OA treatment (Figures 3F,G). These results reinforce our conclusion that while 5-HT and OA have similar quantitative effects on somal Ca²⁺ transients in ASH neurons, the underlying signal transduction pathways of the two monoamines are very different.

G_o signaling can also hyperpolarize neuronal membranes through the activation of GIRK channels by G_βγ subunits (Kano et al., 2019). OCTR-1, heterologously expressed in Xenopus oocytes, is capable of signaling in this manner (Mills et al., 2012). The C. elegans genome contains three GIRK genes (irk-1, irk-2, and irk-3), which are widely expressed in the nervous system (Emtage et al., 2012; Hobert, 2013; WormBase, 2022). Therefore, we hypothesized that these IRK proteins could play a role downstream of OCTR-1/G_o. To test the potential involvement of GIRK channels in OA signaling, we performed calcium imaging in the three different GIRK null mutants (irk-1, irk-2, and irk-3). OA did not significantly inhibit ASH somal Ca²⁺ transients in the irk-2 and irk-3 null backgrounds but produced normal inhibition in the irk-1 background (Figure 4). This result indicates that IRK-2 and IRK-3 are required downstream of OCTR-1/G_o signaling and suggests that the OA inhibition of ASH somal Ca²⁺ transients is due to reduction of ASH excitability. Behavioral analysis revealed a role for all three GIRK channels in OA modulation of 1-oct avoidance, but the results did not cleanly correspond to the circumscribed role for these channels predicted by our model for OA signaling in ASH (see Discussion), possibly due to the expression of IRK channels in neurons other than ASHs, and the involvement of additional circuits and OA signaling pathways controlling 1-oct aversive responses (Supplementary Figure S2).

To determine whether OA reduces ASH excitability, we performed direct electrophysiological recordings on the ASHs. These recordings are performed on partially dissected worms in which an ASH soma is exposed to the bath, taking care not to sever the dendritic process connecting the soma to the amphid opening (Zahratka et al., 2015; Williams et al., 2018b). Dissection and proper placement of flow pipettes generally takes longer than the perdurance of OA signaling when worms are pre-incubated on OA plates, which was the approach used in the above experiments (see section “Materials and Methods”). Therefore, it was necessary to treat ASHs with OA after they had been dissected. Incubation of C. elegans with monoamines on plates requires very high concentrations because of low cuticular permeability to polar compounds, which prevents accurate determination of effective internal monoamine concentrations (Williams et al., 2018a). Therefore, it was first necessary to determine a minimal effective concentration of OA for these electrophysiological experiments. Using Ca²⁺ imaging in the exposed ASH soma as a readout, we performed paired recordings on individual neurons before and after exposure to OA at 1, 10, and 100 nM for 60 s (Figure 5A). 1-oct responses were unaltered at 1 nM OA, but significantly decreased at 10 and 100 nM (Figures 5B,C), establishing 10 nM as an appropriate OA concentration, comparable to the minimal effective concentrations observed for 5-HT receptors from C. elegans and other species (Adolph and Tuan, 1972; Bunin and Wightman, 1998). Significantly, octr-1 mutants did not show reduced 1-oct dependent Ca²⁺ transients after 10 nM OA treatment (Figures 5B,C), confirming that OCTR-1 is the relevant receptor for OA inhibition of ASH somal Ca²⁺ transients under these conditions. A second OA receptor, SER-3, also acts in ASHs at higher OA concentrations (Mills et al., 2012), but ser-3 mutants still showed inhibition of 1-oct Ca²⁺ transients in the presence of 10 nM OA, confirming the central role for OCTR-1 in OA modulation of somal Ca²⁺ transients in ASH.

Patch clamp recordings demonstrated that OA treatment reduced the amplitude of 1-oct-induced depolarization in ASHs. In control samples, exposing the tip of the worm’s nose to 2.2 mM 1-oct in external solution (a saturating concentration) resulted in a slowly developing +15 mV depolarization which returned to baseline following 1-oct removal (Figures 6A,C), comparable to previously published results (Zahratka et al., 2015; Williams et al., 2018b). The placement of flow pipettes provided a constant stream of clean external solution to the ASH soma, shielding it from direct 1-oct exposure (Figure 6A), so we can confidently attribute the observed depolarization to olfactory signaling at the ASH sensory ending (Zahratka et al., 2015), not the direct exposure of the ASH soma to 1-oct. Sixty second pre-treatment of the ASH soma with 10 nM OA significantly reduced this 1-oct response (Figures 6B,C). While OA signaling inhibited 1-oct-dependent depolarization, it did not appear to hyperpolarize ASH neurons at rest, since somal Ca²⁺ levels in individual bath-exposed ASHs were not decreased by 10 nM OA application (Figure 6D). This result implies that the membrane potential of unstimulated ASH neurons may be at or near the K⁺ reversal potential. Based on these results, we conclude that OA inhibition of 1-oct-stimulated Ca²⁺ transients in the soma of ASH neurons reflects G_o-dependent activation of GIRK channels leading to reduced neuronal depolarization, which stands in contrast to 5-HT inhibition of ASH somal Ca²⁺ transients, which reflects Gα_q-dependent inhibition of L-VGCCs leading to enhanced neuronal depolarization.