To assess flies’ avoidance behavior toward chemicals, we employed a two-choice positional preference assay (Supplementary Figure 1A; Sanchez-Alcaniz et al., 2017). Starved adult males were released onto an assay plate coated with agarose containing sucrose. The plate was divided into four sections, with two diagonal quadrants containing a test chemical and the remaining two quadrants without the chemical. A group of approximately 50 flies was allowed to explore and select a location using their chemosensory abilities. Thus, the number of flies in each quadrant at a given time could reflect their decision, either preference or avoidance of the test chemicals. Although the preference index (PI) exhibited considerable fluctuations every minute throughout the experimental period (Supplementary Figure 1B), we observed consistent trends of PI decrease after calculating the moving averages over a 21-min interval (see Methods).

When we tested quinine, known to elicit aversive responses in flies through bitter taste neurons (Wang et al., 2004), the flies displayed an increasing aversion to quinine over time (Figure 1A; Supplementary Figure 2A). In contrast, when quinine was absent from the agarose, flies distributed evenly across all quadrants. This result indicated that the assay could detect chemical avoidance in a population of flies. In contrast to starved flies, satiated flies did not exhibit avoidance of the same concentration of quinine (Figure 1A; Supplementary Figure 2A). Moreover, flies did not develop an aversion to quinine when sucrose was excluded from agarose (Figure 1A; Supplementary Figure 2A). These results suggested that quinine avoidance detected by this positional choice assay was primarily dependent on feeding behavior.

Next, we investigated 2MT, a volatile compound that induces innate fear responses in mice (Isosaka et al., 2015), and its derivatives are present in some fruits and vegetables (Servili et al., 2000; Fernandez et al., 2001; Takahashi and Shibamoto, 2008). Flies immediately and consistently displayed avoidance of 1 mM 2MT (Figure 1B; Supplementary Figure 2B). Unlike quinine, the avoidance of 2MT was observed regardless of the inclusion of sucrose in the agarose, and it was also evident in satiated flies (Figure 1B; Supplementary Figure 2B). In subsequent experiments, we conducted the positional choice assay using 24-hour-starved flies in the presence of 2 mM sucrose. Collectively, these findings suggest that 2MT avoidance is regulated not only by taste but also by other sensory pathways.

To determine which sensory molecules are involved in 2MT avoidance, we examined mutants of TRP channels known to be associated with chemical avoidance in adult flies (Al-Anzi et al., 2006; Kang et al., 2010; Kim et al., 2010; Kwon et al., 2010; Badsha et al., 2012; Zhang et al., 2013; Mandel et al., 2018; Li et al., 2020; Boonen et al., 2021; Melo et al., 2021). Null mutation of TrpA1 nearly abolished avoidance responses to 1 mM 2MT (Figure 2A; Supplementary Figure 3A). In contrast, other TRP channel mutants, including TRPA subfamily (painless; pain and water witch; wtrw) and TRPC subfamily (Trp, Trp-like; Trpl, and Trpγ) showed minimal or no reduction in avoidance behavior to the same concentration of 2MT (Figures 2B, C; Supplementary Figures 3B, C). These results demonstrate the essential role of TrpA1 in detecting 2MT in flies.

Next, as 2MT is a volatile compound, we investigated whether olfaction was necessary for 2MT avoidance. We observed that flies carrying a null mutation of Odorant receptor co-receptor (Orco), resulting in olfactory dysfunction (Larsson et al., 2004), exhibited a partial but significant impairment in the avoidance response to 1 mM 2MT at all tested time points (30, 60, and 90 min) compared to the control (Figure 2A; Supplementary Figure 3A). Orco mutant flies did not avoid low concentrations of 2MT (30, 100, and 300 μM). Notably, TrpA1 mutant flies displayed avoidance behavior indistinguishable from control flies at these concentrations (Figures 2D, E; Supplementary Figures 3D, 4). Therefore, we conclude that TrpA1 and ORs are involved in the contact chemosensory- and olfactory-dependent detection of 2MT, respectively, with their contributions depending on the concentration of the chemical.

We investigated the impact of 2MT on another behavioral response, oviposition behavior, in females. Female flies use their chemosensory abilities to avoid laying eggs in the presence of harmful or aversive chemicals (Stensmyr et al., 2012; Melo et al., 2021). We conducted an oviposition assay similar to the two-choice positional preference assay by releasing adult females onto an agarose-coated plate divided into four sections. After 6 h, we counted the number of eggs in each quadrant and calculated the oviposition index (OI). Control and Orco mutant flies exhibited strong avoidance of laying eggs on agarose containing 1 mM 2MT (Figure 2F). In contrast, TrpA1 mutant flies did not show this pronounced avoidance response (Figure 2F). These results indicate that 2MT inhibits egg laying in female flies via TrpA1. Thus, the action of 2MT through TrpA1 elicits general aversive responses in adult flies, as observed in both male flies (two-choice positional preference assay) and female flies (oviposition assay).

In mice, exposure to 2MT leads to freezing behavior and suppression of locomotor activity (Isosaka et al., 2015; Matsuo et al., 2021b). To assess the locomotor activities of flies during 2MT exposure, we recorded their spontaneous walking and measured the total distance moved during 10-min test intervals using a video-tracking system. Flies were exposed to 100 μM to 10 mM 2MT, and their locomotor activity was compared to that of flies not exposed to the chemical. In control flies, there were no differences in locomotor activity when exposed to any concentration of 2MT during all test intervals (Supplementary Figure 5A). TrpA1 mutant flies appeared to exhibit increased movement compared to control flies in the absence or presence of 2MT up to 1 mM. However, their locomotor activity was reduced in the presence of 10 mM 2MT for unknown reasons (Supplementary Figure 5B). These results highlight the difference in the physiological effects of 2MT between mice and flies.

We also investigated the aversive responses induced by other thiazoline-related compounds, considering that some of these compounds can activate mouse TRPA1 (Matsuo et al., 2021a). 2-Methyl-4-ethylthiazoline (4E2MT) elicited avoidance behavior in control flies, which was partially reduced in TrpA1 mutant flies (Figure 3A; Supplementary Figure 6A). Unlike 2MT, the avoidance response to 4E2MT was not affected in Orco mutant flies, although the chemical is volatile. Thiomorpholine (TMO) induced avoidance behavior in both control and TrpA1 mutant flies (Figure 3B; Supplementary Figure 6B), suggesting that it acts on unknown molecules. 2-Methyl-2-oxazoline (2MO), which does not activate mouse TRPA1 (Wang Y. et al., 2018; Matsuo et al., 2021a), did not elicit clear behavioral responses in either control or TrpA1 mutant flies (Figure 3C; Supplementary Figure 6C). In summary, among the tested thiazoline-related compounds, 2MT and 4E2MT exerted their aversive effects primarily through TrpA1, whereas only 2MT induced avoidance behavior through olfaction.

TrpA1 is expressed in bitter taste neurons (Kang et al., 2010, 2011; Kim et al., 2010; Leung et al., 2020), OSNs (Kwon et al., 2010), and nociceptive sensory neurons (Khuong et al., 2019) in adult flies. To determine which sensory neurons are involved in 2MT avoidance, we performed knockdown experiments targeting TrpA1 in specific sensory neurons using the GAL4/UAS binary expression system and RNA interference (RNAi). Knockdown of TrpA1 in bitter taste neurons using Gr66a-GAL4 or in nociceptive neurons using ppk-GAL4 significantly reduced 2MT avoidance (Figures 4A, B; Supplementary Figure 7A), whereas knockdown in the OSNs using Orco-GAL4 did not have a significant effect (Figure 4C; Supplementary Figure 7A).

Next, we inhibited neural activities in the OSNs by expressing Kir2.1, because the avoidance of 2MT was altered in Orco mutant flies, particularly at low concentrations (Figures 2A, D, E). Electrical silencing of the OSNs by Orco-GAL4 abolished the avoidance of 100 μM 2MT and partially impaired the avoidance of 1 mM 2MT (Figures 4D, E; Supplementary Figure 7B), consistent with the results obtained from Orco mutant flies (Figures 2A, D). These results indicate that adult flies detect 2MT through TrpA1 expressed in bitter taste neurons and nociceptive neurons, in conjunction with ORs expressed in OSNs.

Drosophila melanogaster TrpA1 has five different splicing variants: TrpA1-A, TrpA1-B, TrpA1-C, TrpA1-D, and TrpA1-E (Zhong et al., 2012; Gu et al., 2019). While all isoforms except TrpA1-E respond to the wasabi ingredient AITC when expressed in tissue culture cells, they have different contributions to thermal and chemical sensation in larvae (Zhong et al., 2012; Gu et al., 2019) and adult flies (Kang et al., 2011; Du et al., 2015; Leung et al., 2020). To determine which TrpA1 isoforms are essential for 2MT avoidance, we used TrpA1 isoform-specific knock-in (KI) alleles (Gu et al., 2019), which allow the expression of a single TrpA1 isoform by mutating exons specific to other isoforms. Heterozygotes of the KI or KO alleles were generated by crossing with the TrpA1¹ null mutant to ensure the viability of some unhealthy homozygous KI alleles. Heterozygotes of TrpA1-CKI and TrpA1-DKI displayed strong avoidance behavior to 2MT, similar to TrpA1-KI heterozygote that expresses all isoforms (Figure 5A; Supplementary Figure 8A). Conversely, TrpA1-AKI, TrpA1-BKI, and TrpA1-EKI heterozygous flies did not exhibit avoidance of 2MT at the same concentration, resembling the response of TrpA1-KO/TrpA1¹ transheterozygotes (Figure 5B; Supplementary Figure 8B). These results indicate that the expression of either the TrpA1-C or TrpA1-D isoform is sufficient to respond to 2MT in adult flies.

To determine the expression pattern of TrpA1-C and TrpA1-D isoforms in chemosensory and nociceptive neurons, we examined the expression of membrane-tethered green fluorescent protein (mCD8::GFP) induced in TrpA1 isoform-specific KI lines. These KI lines carried a T2A-GAL4 sequence just upstream of the stop codon (Gu et al., 2019), allowing GAL4 expression in isoform-specific cells. Since the expression patterns in the proboscis and maxillary palps have been previously described (Gu et al., 2019), we investigated the expression patterns of TrpA1-C and TrpA1-D in sensory neurons in the legs and labellum.

First, we observed TrpA1-T2A-GAL4 (TrpA1-KI), which expresses all isoforms, to determine the overall TrpA1 expression pattern. In the forelegs, two large cell bodies located in the fifth segments of protarsus were clearly labeled, similar to the pattern of Gr66a- and ppk-positive neurons (Figures 6A, D, E; Supplementary Figure 9A). This expression was also observed in TrpA1-CKI-T2A-GAL4 and TrpA1-BKI-T2A-GAL4 flies but not in TrpA1-DKI-T2A-GAL4 and TrpA1-AKI-T2A-GAL4 flies (Figures 6B, C; Supplementary Figures 9B, C). TrpA1-KI flies also exhibited weak GFP signals in neurons with small cell bodies located in the proximal segments of the tarsus in the forelegs, midlegs, and hindlegs. The position and shape resembled those of ppk-GAL4-expressing nociceptive neurons (Figures 6A, E; Supplementary Figure 9A). This ppk-like expression pattern was observed in both TrpA1-CKI-T2A-GAL4 and TrpA1-DKI-T2A-GAL4 flies (Figures 6B, C), but not in TrpA1-AKI-T2A-GAL4 or TrpA1-BKI-T2A-GAL4 flies (Supplementary Figures 9B, C). To confirm TrpA1 expression in gustatory and nociceptive neurons, we crossed TrpA1-T2A-LexA, which carries a red fluorescent protein (mCherry) expressed in all isoform-expressing cells, with Gr66a-GAL4 or ppk-GAL4, expressing mCD8::GFP in these neurons (Figure 6F). As anticipated from individual labeling (Figures 6A, D, E), TrpA1 co-localized with Gr66a and Ppk in all legs (Figure 6F). We also examined the expression patterns of Gr66a- and ppk-GAL4 in the labellum alongside TrpA1-T2A-LexA. TrpA1 showed widespread expression in gustatory neurons and co-localization within a subset of Gr66a neurons (Supplementary Figure 10A). However, co-expression of Ppk and TrpA1 was not observed in the labellum (Supplementary Figure 10B). Based on these results and the behavioral outputs in Figure 4, we suggest that bitter taste neurons expressing TrpA1-C and nociceptive neurons expressing TrpA1-C and TrpA1-D contribute to aversive responses to 2MT.

To investigate the responsiveness of Drosophila TrpA1-C and TrpA1-D isoforms to 2MT, we conducted in vitro Ca²⁺ imaging in S2R+ cells expressing either TrpA1-C or TrpA1-D. The application of 2MT resulted in dose-dependent increases in intracellular Ca²⁺ ([Ca²⁺]_i) levels, specifically in cells expressing TrpA1-C or TrpA1-D, whereas control cells did not exhibit such responses (Figures 7A–C). Notably, TrpA1-D showed a more potent increase in [Ca²⁺]_i than TrpA1-C. We also observed [Ca²⁺]_i increases in TrpA1-C and TrpA1-D-expressing cells upon exposure to 4E2MT, although non-specific effects of the compound became evident at higher concentrations, and [Ca²⁺]_i increases were indistinguishable between TrpA1-expressing cells and control cells (Figures 7D–F).

Previous studies have demonstrated that 2MT activated mouse TRPA1 by covalently modifying six cysteine residues (Wang Y. et al., 2018). In Drosophila TrpA1, two cysteine residues are conserved: C480 and C729 in TrpA1-D (corresponding to C415 and C666 in mice, respectively), and C480 in TrpA1-C (corresponding to C415 in mice) (Figure 7G; Supplementary Figure 11). To assess the role of these conserved cysteine residues in Drosophila TrpA1, we substituted each residue with serine in TrpA1-C (C480S) and TrpA1-D (C480S, C729S, and C480S/C729S). Mutating single or multiple cysteine residues in these isoforms resulted in a significant reduction in [Ca²⁺]_i responses to 2MT, almost reaching the background level (Figure 7H). Importantly, we confirmed that the double mutant in TrpA1-D retained responsiveness to citronellal, which activates the channel through non-covalent mechanisms (Du et al., 2015; Boonen et al., 2021), indicating that the mutated channels remained functional (Supplementary Figure 12). These results strongly suggest that 2MT activates Drosophila TrpA1 by specifically interacting with these conserved cysteine residues.

Flies were reared on glucose-yeast-cornmeal media: 2,500 ml of reverse osmosis water, 180 g of cornmeal (Oriental Yeast, Tokyo), 100 g of dry brewer’s yeast Ebios (#128-297405, Mitsubishi Tanabe Pharma, Osaka), 19 g of agar (#RSU-AL01, RIKAKEN, Tokyo), 250 g of glucose (#TDH, San-ei Sucrochemical, Aichi), 24 ml of methyl 4-hydroxybenzoate (10% in 70% ethanol; #H5501, Sigma-Aldrich, MA), and 8 ml of propionic acid (#81910, Sigma-Aldrich, MA). Flies were raised in vials or bottles at 25°C under 12-h light/12-h dark cycle. w¹¹¹⁸ was outcrossed for ten generations to the wild-type Canton-S (CS) genetic background and used as a control. The fly lines used in this study are as follows: TrpA1¹ [Bloomington stock center (BL, IN) #26504], Orco² (BL #23130), wtrw² (BL #59038), Trp^MB03672 (BL #23636), Trpl^MB10553 (BL #29134), Trpγ^G4 (BL # 64313), Orco-GAL4 (BL #26818), UAS-dicer2 (BL #24650), UAS-TrpA1 RNAi (BL #36780), UAS-Kir2.1::GFP (KYOTO Drosophila Stock Center, Kyoto #108846), UAS-mCD8::GFP (BL #5137 and #32195). The following stocks were provided by the indicated investigators: TrpA1-T2A-GAL4, TrpA1-KO, TrpA1-AKI-T2A-GAL4, TrpA1-BKI-T2A-GAL4, TrpA1-CKI-T2A-GAL4, TrpA1-DKI-T2A-GAL4, TrpA1-EKI-T2A-GAL4, and lexAop2-mCherry; TrpA1-T2A-LexA (Dr. Y. Xiang) (Gu et al., 2019), Gr66a-GAL4 (Dr. H. Amlein), ppk-GAL4 B-3207d (Dr. D. N. Cox), and pain⁴ (Dr. C. Montell) (Liu et al., 2023). All lines except the UAS-dicer2, UAS-TrpA1 RNAi, UAS-mCD8::GFP, TrpA1-T2A-GAL4, TrpA1-KO, TrpA1-KI-T2A-GAL4, and lexAop2-mCherry; TrpA1-T2A-LexA lines were outcrossed to w¹¹¹⁸ flies on the CS genetic background for five generations.

The following chemicals were purchased from Tokyo Chemical Industry (Tokyo): quinine hydrochloride dihydrate (#Q0030), 2-methylthiazoline (2MT; #M0285), 2-methyl-4-ethylthiazoline (4E2MT; #M0689), thiomorpholine (TMO; #T1007), and 2-methyl-2-oxazoline (2MO; #M0857). N-methylmaleimide (NMM; #389412) and citronellal (#373753) were purchased from Sigma-Aldrich (MA). Quinine was dissolved in water and kept at −20°C. 2MT, TMO, and 2MO were dissolved in water just before use for behavioral assays. 4E2MT was dissolved in dimethyl sulfoxide (DMSO) just before use for behavioral assays. 2MT and 4E2MT were dissolved in a bath solution and sonicated for 1 min in an ultrasonic cleaner (#MCS-2P, AS ONE Corporation, Osaka) just before use (see below for the composition of the bath solution). NMM was dissolved in DMSO and kept at −20°C.

The assay was conducted following a similar protocol as previously described (Sanchez-Alcaniz et al., 2017). Male flies (3–8 days old) were used for all behavioral assays. A total of fifty adult males were collected and kept in vials containing standard food until the onset of starvation. To induce starvation, flies were deprived of food and maintained in vials containing 1 ml of 1% agarose for 24 h prior to the assay. Non-starved flies were transferred to vials containing food 24 h before the experiment.

For the assay plates, we used 100-mm diameter four-section petri dishes (#25384-308, VWR, PA) along with custom-made quadrant acrylic blocks. Each quadrant of the petri dish was fitted with the acrylic block, and a small amount of 1% agarose was added to fill the gap. After the agarose has solidified, the surface of the acrylic block in each quadrant was covered with 0.5% agarose containing either a test chemical or a solvent (1.2 ml per quadrant). Unless otherwise stated, all quadrants contained 2 mM sucrose. The test chemicals were alternately present or absent in each quadrant. If the chemical was dissolved in DMSO, the final concentration of DMSO was 0.1%. Flies in vials were gently anesthetized with ice for up to 70 s and immediately transferred to an assay plate. The assay plates were covered with lids and placed upside down on an acrylic stage, allowing the flies come into contact with the agarose through negative geotaxis. The acrylic stage was elevated 5 cm from an LED light tracer (6800 lux; SV531A, Sinkosha, Tokyo), with the LED light covered with a red film (#105, LEE Filters, Hampshire). The assay chamber was covered with a light shielding curtain. Still images of the assay plates were captured from the top every 1 min using a digital camera (FDR-AX60, Sony, Tokyo) for a duration of up to 120 min. All behavioral experiments were initiated in the morning [zeitgeber time (ZT) 1.5–2]. See Supplementary Figure 1 for a schematic of the setup.

The number of flies In each quadrant were quantified using a custom macro in ImageJ (Sanchez-Alcaniz et al., 2017) with some modifications. The preference index (PI) was calculated using the following formula: PI (%) = (N_chemical – N_control)/(N_chemical + N_control) × 100, where N_chemical and N_control represent the number of flies in the chemical-containing quadrants and control quadrants, respectively. PIs were calculated over time using 21-min moving averages.

We used female flies aged 4–8 days old. A total of 30–40 adult females and ten adult males were collected and kept in vials containing standard food. Four days prior to the assay, the flies were transferred to food vials containing yeast paste. The preparation of assay plates followed a similar procedure as described above for the two-choice positional preference assay. To enhance egg production, we coated the surface of the acrylic blocks in quadrants with soft agarose (0.25%) containing 10 mM sucrose. Five to eight female flies were gently transferred onto each assay plate using an aspirator. The assay plates were then placed in a dark box with water-soaked paper. Females were allowed to lay eggs from the morning (ZT 2–2.5) to the evening (ZT 8–8.5) at a temperature of 25°C.

The number of eggs in each quadrant was manually counted. Oviposition index (OI) was calculated using the following formula: OI (%) = (E_chemical – E_control)/(E_chemical + E_control) × 100, where E_chemical and E_control represent the number of eggs in the quadrants with the chemical and control quadrants, respectively. Trials in which the total number of eggs laid was fewer than twenty were excluded from the analyses.

Assay plates were prepared using a 96-well microplate (#1-1601-02, VIOLAMO, Osaka). We applied 180 μl of 1% agarose with or without test chemicals to each well. Once the agarose has solidified, we sealed the plate with plate seal (#547-SBS-PET, WATSON, Tokyo) and cut a cross slit over each well. Male flies (3–8 days old) were individually transferred into each well using an aspirator. The flies were allowed to move freely within the wells and were recorded using a digital camera (FDR-AX60, Sony, Tokyo) in the assay chamber, similar to the two-choice positional preference assay as described above. The recordings were captured at a frame rate of 30 frames per second for 10 min at 0, 30, and 60 min after placing the assay plate in the chamber. All behavioral experiments were conducted in the afternoon (ZT 5.5–6.5). We measured the total distance moved (mm) as an index of locomotor activity (Sakai et al., 2009), which was calculated using the video tracking software Move-tr/2D (version 8.4, Library, Tokyo).

For observation of the expression pattern of TrpA1 isoforms, UAS-mCD8::GFP was crossed with TrpA1-T2A-GAL4, TrpA1-AKI-T2A-GAL4, TrpA1-BKI-T2A-GAL4, TrpA1-CKI-T2A-GAL4, TrpA1-DKI-T2A-GAL4, Gr66a-GAL4, and ppk-GAL4 B-3207d. For observation of the co-expression of TrpA1 with Gr66a- and ppk-GAL4, lexAop2-mCherry; TrpA1-T2A-LexA was crossed with Gr66a-GAL4; UAS-mCD8::GFP and UAS-mCD8::GFP; ppk-GAL4. We dissected forelegs, midlegs, hindlegs, and labellum from male flies (4–10 days old) in phosphate-buffered saline (PBS), and mounted the samples in 50% glycerol/PBS. GFP and mCherry fluorescence and brightfield images were acquired using a confocal laser scanning microscope (FV1200, Olympus, Tokyo) and a 30 × /1.05 UPLSAPO30XSIR dry objective lens. Z-sections were taken at a 1-μm interval. Images were analyzed using FLUOVIEW (version 4.2c, Olympus, Tokyo) and Fiji software (Schindelin et al., 2012).

For the cloning of TrpA1-C and TrpA1-D isoforms, we conducted RT-PCR using total RNA extracted from w¹¹¹⁸ third instar larvae following the manufacturer’s protocol. We prepared 13 μl reaction premix containing 50 μM Oligo(dT)₂₀, 10 mM dNTP mix, 5 mg of total RNA in DEPC-treated water in a 0.2-ml PCR tube. The premix was incubated at 65°C for 5 min, and chilled on ice for over 1 min. Subsequently, the incubated reaction premix was combined with 7 μl of mix containing 4 μl of 5X SuperScript IV RT buffer, 1 μl of 100 mM DTT, 1 μl of SuperScript™ IV Reverse Transcriptase (#18090010, Invitrogen, MA), and 1 μl of RNaseOUT™ Recombinant Ribonuclease Inhibitor (#10777019, Invitrogen, MA). The reaction mix was incubated at 55°C for 10 min followed by 80°C for 10 min. The RT product was stored in −20°C until further use.

To amplify TrpA1-C and TrpA1-D cDNAs, we performed PCR using a common forward primer containing the KOZAC sequence upstream of the start codon and a common reverse primer (Supplementary Table 1). Phusion High-Fidelity DNA Polymerase (#M0530S, Invitrogen, MA) was mixed with 2 μl (∼500 ng) of undiluted RT template in 50 μl reaction. The PCR protocol consisted of an initial denaturation step at 98°C for 30 s, followed by 35 cycles of denaturation at 98°C for 10 s, annealing at 66°C for 30 s, and extension at 72°C for 2 min and 40 s. The final extension was performed at 72°C for 2 min. The PCR product and pAc5.1-V5-His A vector (#V411020, Invitrogen, MA) were digested with NotI-HF [#R3189S, New England Biolabs (NEB), MA] and XbaI (#R0145S, NEB, MA). The digested PCR product and vector were ligated using Ligation High Ver.2 (#LGK-201, TOYOBO, Osaka) at 16°C for 30 min and then transformed into DH5α competent cells. E. coli culture were plated on LB agar (#20067-85, nacalai tesque, Kyoto) containing 100 μg/mL Ampicillin (#016-23301, Wako, Osaka) in a 10-cm petri dish (#SH90-15, IWAKI, Shizuoka) and incubated overnight at 37°C. The colonies were inoculated into Plusgrow II (#08202-75, nacalai tesque, Kyoto) containing 100 μg/mL Ampicillin and cultured overnight in a 37°C shaker. The plasmid was extracted using NucleoSpin Plasmid EasyPure (#U0727C, Takara, Shiga). The presence of dTrpA1-C or dTrpA1-D in the constructs was confirmed by digestion with SspI-HF (#R3132S, NEB, MA), resulted in the following fragments: TrpA1-C/pAc5.1-V5-His A: 4063, 2751, 992, 637, 321, and 284 bp; TrpA1-D/pAc5.1-V5-His A: 6817, 992, 637, 321, and 284 bp. The entire sequences were confirmed (Supplementary Figure 14). The plasmids were amplified using NucleoBond Xtra Midi (740410.50, Macherey-Nagel, Düren) and stored at −20°C.

We performed PCR to substitute cysteines at position 480 and/or 729 with serine in wild-type TrpA1-C/pAc5.1-V5-His A or TrpA1-D/pAc5.1-V5-His A. The cysteine codon in the region of interest was replaced with serine using a pair of overlapping oligonucleotides containing the designated mutation. The synthetic oligonucleotide primers carrying the mutations are listed in Supplementary Table 2.

A two-step PCR reaction was performed using Phusion High-Fidelity DNA Polymerase (#M0530S, NEB, MA). The reaction premix contained 200 μM dNTPs, 10 μM each of forward and reverse oligonucleotides, 200 pg of template DNA, 0.2 units of Phusion High-Fidelity DNA Polymerase dissolved in 5X Phusion buffer and MilliQ water. The PCR protocol consisted of an initial denaturation step at 98°C for 30 s, followed by 25 cycles of denaturation at 98°C for 10 s and annealing/extension at 72°C for 3 min. The final extension was performed at 72°C for 5 min. The amplified plasmid was subjected to demethylation using DpnI for 1 h at 37°C, and transformed into DH5α E. coli strain for overnight culture. The resulting colonies were purified using NucleoBond Xtra Midi (740410.50, Macherey-Nagel, Düren) for subsequent experiments.

Schneider 2 R+ (S2R+) cells were grown in 60-mm dishes (#353002, Falcon, NY) at 25°C in 4 ml of Schneider’s medium (#21720-024, Gibco, MA) containing 10% inactivated fetal bovine serum (#10437-028, Gibco, MA) and 50 mg/mL penicillin/50 units/mL streptomycin (#15140-122, Gibco, MA). For a transient transfection, a mix was prepared with 1 μg of TrpA1-C, TrpA1-D, or the mutants of each isoform in a pAc5.1-V5-His expression vector and 0.1 μg of DsRed/pAc5.1-V5-His at the 10:1 ratio using X-tremeGENE9 Transfection Reagent (#06365787001, Roche, Basel) dissolved in 1X OPTI-MEM medium (#31985-070, Thermo Fisher, MA), following the manufacturer’s protocol. The transfection mix was incubated for 15 min at room temperature. S2R+ cells were reseeded onto 12-mm cover glasses (Matsunami, Osaka) in a 35-mm dishes (#1000-035, IWAKI, Shizuoka) in 2 ml of culture media and the transfection mix was added to the media. The cells were incubated for 48 h in a 25°C incubator. To load the cells with Fura-2 AM, we added 1 mL of the following mixture to the culture media and incubated the cells in a 25°C incubator for 1–3 h: 5 μM Fura-2 AM (#F-1201, Life Technologies, MA), 1 mM probenecid (#162-26112, Wako, Osaka), and 0.02% pluronic F-127 detergent (#P2443, Sigma-Aldrich, MA). The cover glasses were mounted in a chamber (#RC-26G; Warner Instruments, MA) filled with a bath solution containing 130 mM NaCl, 5 mM KCl, 2 mM MgCl₂, 2 mM CaCl₂, 30 mM sucrose, and 10 mM N-Tris(hydroxymethyl)methyl-2-aminoethanesulfonic acid (TES), after adjusting the pH to 7.2 with NaOH. 500 μM probenecid was dissolved in the bath solutions to retain Fura-2 in the cells. The chamber was connected to a gravity flow system to deliver various stimuli. A xenon lamp was used as an illumination source. To obtain fluorescent intensities of Ca²⁺-bound and Ca²⁺-free Fura-2, we excited the cells at 340 and 380 nm, respectively, and emission was monitored at 510 nm with a CCD camera, CoolSNAP DYNO (Photometrics, AZ) used with a fluorescent microscope (#TE-300, Nikon, Tokyo). Data was measured using the NIS element AR (Nikon, Tokyo).

We analyzed DsRed-positive cells and obtained a trace for each cell using the NIS element AR. The change in Ca²⁺_i was calculated as follows: Ca²⁺_i response = (F_res – F_min)/(F_max – F_min). To normalize the responses, the minimum values during the basal period (F_min) was subtracted from the responses every 3 s (F_res) and the maximum value obtained due to addition of the ionomycin (F_max) (#I0634, Sigma, MA). After the normalization, the maximum increase in Ca²⁺_i (Ca²⁺_i max) during the stimulation period was extracted for further analysis.

The data are presented as moving averages ± SEMs or means ± SEMs. The number of times each experiment was performed (N) is indicated in the figure legends. The normality of the data was assessed using the Shapiro-Wilk test. If the data were normally distributed, Student’s-t test or Welch’s-t test was conducted for pairwise comparisons, and one-way analysis of variance (ANOVA) with the Dunnett’s or Tukey’s post hoc analysis was performed for multiple pairwise comparisons. For non-normally distributed data, the Mann-Whitney U test was performed for pairwise comparisons, and non-parametric ANOVA (Kruskal-Wallis test) followed by post hoc analysis with the Steel test or Steel-Dwass test was conducted for multiple pairwise comparisons. All statistical analyses were performed using EZR (version 1.61; Saitama Medical Center, Jichi Medical University) (Kanda, 2013), which is a graphical user interface for R (The R Foundation for Statistical Computing). Statistical significance is indicated by asterisks, where *P < 0.05, **P < 0.01, and ***P < 0.001. NS denotes not significant.