Introduction

Otopetrin 1 (OTOP1) is a multi-transmembrane protein highly conserved across vertebrates, arthropods, and nematodes1,2. OTOP1 plays a crucial role in various physiological processes, particularly in the formation of otoliths in zebrafish3. Depletion of zebrafish Otop1 results in the absence of otoliths without affecting the sensory epithelium or other inner ear structures3. OTOP1 is implicated in regulating cellular calcium levels in vestibular supporting cells, responding to purinergic agonists such as ATP to elevate intracellular calcium concentrations during otoconia mineralization4. In mice lacking functional otoliths, otolith sensor input enhances compensation after unilateral labyrinthectomy5. Additionally, OTOP1 reduces inflammation in adipose tissue and helps maintain metabolic balance in individuals with obesity6. Recent studies have highlighted OTOP1’s role in cellular and neural responses to sour stimuli, underscoring its importance in taste cell and gustatory nerve responses to acids7,8. OTOP1 also functions as a sensor for ammonium chloride and alkali, further emphasizing its role in taste perception9,10. Further investigations have delved into the molecular mechanisms of the OTOP1 proton channel, particularly focusing on the roles of extracellular loops in proton sensing and permeation10,11.

Although a partial understanding of the structure and function of OTOP1 exists, this knowledge remains incomplete. For instance, uncertainty surrounds the operational state of the channel-whether it is closed, open, or desensitized-along with the mechanisms by which it is gated and desensitized, and the potential for additional functionalities12. Pharmacological agents capable of activating or inhibiting ion channels have been extensively utilized to clarify the mechanisms of channel gating and their significance in physiological processes13,14. While protons (H+) are known to potently activate the OTOP1 channel, they can also pass through the pore region of the channel during activation, resulting in proton accumulation and channel desensitization2,15. Furthermore, zinc ion (Zn2+) was reported to mildly potentiate the OTOP1 channel in a time-dependent manner13. However, it is important to note that Zn2+ is also a potent inhibitor of the OTOP1 channel2. Consequently, Zn2+ may be an inappropriate pharmacological tool for probing the structure of OTOP1 or investigating its physiological functions in vivo.

Carvacrol, a phenolic compound found in essential oils like oregano and thyme, is recognized for its pungent, warm odor and taste16. While some describe it as bitter and unpleasant, others find it to have a spicy, herbal, and woody taste. In this study, we established that carvacrol activates the OTOP1 channel, independent of extracellular protons. Using chimeric channels and point mutations, we further identified essential residues within the TM3 segment and TM5-6 linker, which are critical determinants for the carvacrol activation. Overall, our findings suggest that these regions in the N domain are key to channel gating and constitute the primary sites for carvacrol activation, highlighting their potential as promising sites for pharmacological intervention.

Results

Carvacrol is a novel OTOP1 channel agonist

To identify novel modulators, we conducted a primary screening of chemical libraries against the OTOP1 channel. The OTOP1 channels were heterologously expressed in HEK293T cells and examined using whole-cell patch clamp recordings. As shown in Fig. 1a, mOTOP1 facilitated inward currents when exposed to a pH 6.0 stimulus, which was further potentiated by 1 mM carvacrol at the same pH, indicating carvacrol potentiates mOTOP1 currents in the presence of protons. The potentiating effects of carvacrol were dose-dependent (0.1–1 mM), without evidence of saturation (Fig. 1b). To investigate whether carvacrol directly potentiates or activates mOTOP1, we explored the current–voltage (I–V) relationship of the mOTOP1 current in response to carvacrol using ramp depolarizations at pH 7.4 (Fig. 1c). After excluding proton interference, the mOTOP1 currents increased monotonically with the increase in the extracellular concentration of carvacrol (Fig. 1d). These results suggest that carvacrol can directly activate mOTOP1 without the assistance of protons. Carvacrol dose-dependently elicited mOTOP1 currents, yielding a half-maximal effective concentration (EC50) of 692.27 ± 28.86 μM (Fig. 1e, f). Furthermore, the application of carvacrol within a concentration range of 10 μM-1 mM did not alter the reversal potential (Erev) of the OTOP1 channel, in alignment with expectations due to the relatively consistent proton concentrations inside and outside of the cell.

Fig. 1: Carvacrol dose-dependently activates the mOTOP1 channel.
figure 1

a Representative traces showing mOTOP1 currents elicited in response to pH 6.0 stimulus or 1 mM carvacrol at pH 6.0 (pHi = 7.3, membrane potential Vm = 0 mV). b Fold potentiation as a function of different concentrations of carvacrol (Icarvacrol/IpH6), mean ± standard error of the mean (SEM), n = 6. The current amplitude refers to the stabilized current level after stimulation. c The current–voltage (I–V) relationships of currents in mOTOP1-expressing HEK293T cells were recorded using different concentrations of carvacrol at pH 7.4 extracellular solution, voltage ramps from −100 mV to +80 mV. d Whole-cell patch clamp ramp recordings of mOTOP1 currents in response to carvacrol at different concentrations, currents recorded at −100 mV (pHi = 7.3). e Change in currents under different concentrations of carvacrol was measured as outlined in d. mean ± SEM, n = 3–6. f Hill equation fitting of dose-dependent activation of mOTOP1 channel by carvacrol with EC50 of 692.27 ± 28.86 μM. Hill coefficient, 1.68. Current was measured at −100 mV, data points represent mean ± SEM, n = 6.

Full size image

Carvacrol potentiates pH fluorescence signals in OTOP1-expressing cells

To validate the aforementioned findings, we assessed the efficacy of carvacrol using intracellular pH imaging. Consistent with previous studies2,10, lowering extracellular pH (pHo) from 7.4 to 5.2 resulted in the entry of protons into the cell cytosol, as substantiated using fluorogenic pH-dependent and membrane-permeant dye pHrodo Red (Fig. 2a). Notably, a significantly greater increase in the emission of pHrodo Red was observed upon 1 mM carvacrol application (pH 5.2: 0.55 ± 0.04; pH 5.2 + 1 mM carvacrol: 1.79 ± 0.11, n = 20) in OTOP1-expressing HEK293T cells, whereas no significant emission change was observed in mock-transfected cells (Fig. 2a, b). Additionally, the experimental series were repeated with two different samples for pH 5.2 and pH 5.2 + carvacrol, respectively. Consistently, carvacrol application significantly enhances acid-induced pH fluorescence signals in mOTOP1-transfected cells (Supplementary Fig. 1a, b). These findings align with the patch-clamp observations, indicating that mOTOP1 exhibits greater proton flow in response to carvacrol within an extracellular acidic environment.

Fig. 2: Carvacrol-activated OTOP1 channel selectively permeates protons.
figure 2

a Fluorescence intensity of pH indicator pHrodo in response to different stimuli were measured, in HEK293T cells expressing mOTOP1(n = 20) and sham transfected cells (n = 13), F/F0 is relative value of pH fluorescence intensity, shown as mean ± SEM based on the data from traces for mOTOP1 and mock transfected cells, the initial fluorescence intensity of cells was defined as F0. b Changes in intracellular fluorescence intensity were statistically analysed by Two-way ANOVA, Dunnett’s multiple comparisons test, ****p < 0.0001, ΔF = F − F0, mOTOP1, n = 20, pH 5.2: ΔF = 0.549 ± 0.04, pH 5.2 + 1 mM carvacrol: ΔF = 1.791 ± 0.11 ; sham cells, n = 13, mean ± SEM. The F value is measured as marked with an arrow in Fig. 2a. c mOTOP1 currents were evoked in response to 500 μM carvacrol at pH 5.5, with Na+, K+, Li+, Cs+ (160 mM each), or Ca2+ (40 mM) replacing NMDG+ in the extracellular solution as indicated (Vm = 0 mV). d Percentage change of currents for each ion replacement respectively. Na+, n = 5, 0.011 ± 0.007; K+, n = 8, −0.010 ± 0.009; Li+, n = 6, 0.008 ± 0.005; Cs+, n = 7, −0.002 ± 0.013 and Ca2+, n = 11, −0.030 ± 0.008, mean ± SEM. e I–V relationships of mOTOP1 channels measured under 130 mM Li+ replacing NMDG+ in the pH 7.3 extracellular solution as indicated, NMDG+ or Li+-containing solution both dissolve 500 μM carvacrol. f Ionic permeability ratio of Na+, K+, Li+, and Cs+ to H+ when mOTOP1 channels were activated by 500 μM carvacrol at pH 7.3 extracellular solution.

Full size image

Carvacrol-activated OTOP1 channel selectively permeable to protons

OTOP1 exhibits selective proton permeation upon extracellular acid exposure2. To investigate the ion selectivity of OTOP1 in carvacrol-induced currents, we activated OTOP1 currents by applying 500 μM carvacrol at pH 5.5, then evaluated the impact of exchanging NMDG+ in the extracellular solution with equimolar concentrations of K+, Na+, Li+, Cs+, or isosmotic concentrations of Ca2+ (Fig. 2c). In all five scenarios, the observed change in current magnitude was less than 4%, indicating that the OTOP1 channel does not demonstrate significant permeability to these ions in carvacrol-induced currents (Fig. 2d). Additionally, the Goldman–Hodgkin–Katz (GHK) equation was utilized to determine the ion selectivity of the carvacrol-enhanced OTOP1 channel. Selectivity ratios for specific ions relative to H+ were calculated by measuring Erev following the replacement of NMDG+ by Na+, K+, Li+ and Cs+ in the extracellular solution (Fig. 2e, f). The calculated the selectivity ratios of carvacrol-enhanced OTOP1 for specific ions relative to H+, which were 9.05 × 10−8 (PNa+/PH+), 4.48 × 10−8 (Pk+/PH+), 5.17 × 10−8 (PLi+/PH+), and 6.80 × 10−8 (PCs+/PH+), respectively (Fig. 2f).

Reduced potentiation of carvacrol on OTOP1 at lower pH

To clarify the gating and permeation of OTOP1 channels, we initially examined their functional traits using extracellular solutions ranging in pH from 6.0 to 4.0. In mOTOP1-expressing cells, the application of extracellular acidification induced mOTOP1 currents, which increased monotonically as pHo was lowered (Supplementary Fig. 2a and b). Strikingly, carvacrol-induced current activation was highly influenced by extracellular pH (Supplementary Fig. 2c and d). Specifically, when 1 mM carvacrol was applied at pHo 6.0, 5.5, 5.0, 4.5, and 4.0, the current amplitudes increased to 5.52 ± 0.30 (n = 5), 2.12 ± 0.22 (n = 5), 1.52 ± 0.08 (n = 5), 1.40 ± 0.08 (n = 5), and 1.37 ± 0.08 (n = 6) times the initial current, respectively (Supplementary Fig. 2e). The potentiation ratio of carvacrol decreased consistently with decreasing extracellular pH in the mOTOP1 channel (Supplementary Fig. 2e). These findings imply that the potentiation effect of carvacrol on OTOP1 channel might decrease at lower pH. Nonetheless, additional factors, such as accumulation of intracellular protons, channel desensitization and others, could also influence this process. For example, lower pH levels, could lead to an increased accumulation of protons within the cell, and the channel may reach higher open probability at these lower pH values.

Sensitivity to carvacrol is relatively conserved across species variants of OTOP1

Given that OTOP2 and OTOP3 channels also function as proton channels, we investigated whether carvacrol can also activate these channels under neutral pH. We found that carvacrol failed to elicit significant currents in mOTOP2 and mOTOP3 channels at concentrations between 0.1 and 1 mM (Fig. 3a–c), although these channels showed activity upon stimulation at pH 5.0. We also tested the carvacrol activity in acidic pH (Supplementary Fig. 3a–d). Align with previous observations 500 μM carvacrol did not induce significant currents in mOTOP2 (Supplementary Fig. 3b). Interestingly, at the same concentration, carvacrol attenuated the pH 5.0-induced currents in mOTOP3 (Supplementary Fig. 3c). Our findings show that carvacrol specifically activates the OTOP1 channel at neutral pH but inhibits the acid-induced activation of the mOTOP3 channel at acidic pH. Given the high conservation of OTOP1 in vertebrates, we investigated whether sensitivity to carvacrol is also conserved among different vertebrate species variants of OTOP1 channels. In brief, OTOP1 was identified from representative species across various taxa, encompassing fish, amphibians, reptiles, birds, and mammals, for our analysis of carvacrol sensitivity, followed by the construction of OTOP1 eukaryotic expression plasmids (Fig. 3d–i). Similar to mOTOP1, OTOP1 channels from Oryx dammah (OdOTOP1), Gallus gallus (GgOTOP1), and Bufo gargarizans (BgOTOP1) were activated by carvacrol. Conversely, Danio rerio OTOP1 (DrOTOP1) and Chelonia mydas OTOP1 (CmOTOP1) did not evoke remarkable currents upon 500 μM carvacrol stimulation (Fig. 3g-i). These results suggest a broadly conserved sensitivity to carvacrol among OTOP1 variants, albeit with notable exceptions in the turtle and fish OTOP1 channels.

Fig. 3: Selectivity of carvacrol in activating OTOP1 channels.
figure 3

a, b I–V relationships of mOTOP2 or mOTOP3 channels in response to different concentrations of carvacrol or pH 5.0. Carvacrol at different concentrations was prepared using pH 7.4 extracellular solutions. Voltage ramps from −100 mV to +80 mV. c Plot of normalized currents (Icarvacrol/IpH5, %) versus carvacrol in mOTOP1, mOTOP2, or mOTOP3 channels (mean ± SEM, n = 3–5 cells). currents recorded at −100 mV. d–h Representative I–V curves of five vertebrate OTOP1 channels in response to 500 μM carvacrol at pH 7.4, voltage ramps from −100 mV to +80 mV. i Statistics of activity under 500 μM carvacrol on OTOP1 channels in various vertebrates, currents recorded at −100 mV (pHo = 7.4). Animal icons: Oryx dammah adapted from “Oryx gazella”, credited to Jan A. Venter, Herbert H. T. Prins, David A. Balfour & Rob Slotow (vectorized by T. Michael Keesey). Gallus gallus reprinted from “Gallus gallus domesticus”, credited to Soledad Miranda-Rottmann. Both “Oryx gazella” and “Gallus gallus domesticus” were retrieved from the Attribution 3.0 Unported license: https://creativecommons.org/licenses/by/3.0/. Bufo gargarizans, Chelonia mydas and Danio rerio adapted from public domain, retrieved from CC0 1.0 Universal Public Domain Dedication license: https://creativecommons.org/publicdomain/zero/1.0/.

Full size image

Key regions necessary and sufficient to confer sensitivity to carvacrol

Given the pronounced disparity in carvacrol sensitivity between mOTOP1 and CmOTOP1, we speculated that creating chimeric channels incorporating elements of both could help elucidate the structural basis for this variance. Specifically, we exchanged various components, including the 12 transmembrane domains and N and C termini, resulting in 13 chimeric channels, with seven derived from the mOTOP1 channel and six derived from the CmOTOP1 channel. After exposure to 500 μM carvacrol, each chimera underwent potentiation testing (Fig. 4). Notably, incorporation of the Met1-Thr151 and Leu152-Lys301 regions from CmOTOP1 into mOTOP1 led to a reduction in carvacrol-induced currents (Fig. 4b, d). Additionally, the construction of these chimeras had minimal impact on inward current intensity or kinetics following acid stimulus (Supplementary Fig. 4). Replacing the Phe76-Thr151 or Val225-Lys301 regions from CmOTOP1 had a significant effect on carvacrol sensitivity, whereas replacing the Val302-Ile450 region had no obvious effect (Fig. 4c, e, f). Subsequently, we examined the CmOTOP1 channel with incorporated elements from mOTOP1. Remarkably, inserting the Phe67-Thr143 or Val216-Arg294 regions from mOTOP1 onto CmOTOP1 induced carvacrol sensitivity (Fig. 4i, k). However, the other two chimeras containing a CmOTOP1 backbone remained insensitive to carvacrol activation (Supplementary Fig. 5d, e).

Fig. 4: Phe67-Thr143 and Val216-Arg294 regions are sufficient for carvacrol activation in mOTOP1.
figure 4

a–e Representative currents recorded from HEK293T cells expressing either wild-type mOTOP1 channels or chimeric channels (substituting homologous fragments of CmOTOP1 for mOTOP1) in response to 500 μM carvacrol. Voltage ramps from −100 mV to +80 mV. f Current amplitude changes of mOTOP1 and its chimeras, with recorded at −100 mV. ΔI= Icarvacrol - IpH 7.4, (mean ± SEM, n = 3–6 cells), Cm(451-620)/M(443-600) chimera was observed to be non-conductive. g–k Representative currents recorded from HEK293T cells expressing either wild-type CmOTOP1 channels or chimeric channels (substituting homologous fragments of mOTOP1 for CmOTOP1) in response to 500 μM carvacrol. Voltage ramps from −100 mV to +80 mV. l Current amplitude changes of CmOTOP1 and its chimeras under 500 uM carvacrol, currents recorded at −100 mV. ΔI = Icarvacrol−IpH 7.4 (mean ± SEM, n = 3–6 cells).

Full size image

As Zn2+ is a potent inhibitor of the OTOP1 channel2,7, we next explored its inhibitory effects on carvacrol-induced OTOP1 currents. As expected, 500 μM Zn2+ completely inhibited wildtype mOTOP1 currents (Supplementary Fig. 6a), and effectively suppressed the Cm(Met1-Thr151)/M(Met1-Thr143) and Cm(Leu152-Lys301)/M(Val144-Arg294) chimeras, despite a reduction in current amplitude resulting from chimera construction (Supplementary Fig. 6b, c). As anticipated, no inhibitory effects were observed in CmOTOP1 channels, given that carvacrol does not affect CmOTOP1 functionality. Intriguingly, Zn2+ fully inhibited the carvacrol-induced currents in both the M(Met1-Thr143)/Cm(Met1-Thr151) and M(Val144-Arg294)/Cm(Leu152-Lys301) chimeras, akin to the wild-type mOTOP1 channels (Supplementary Fig. 6d, e). In conclusion, these findings suggest that the Phe67-Thr143 and Val216-Arg294 regions are both essential and sufficient for conferring sensitivity to carvacrol activation.

S134 and T247 are crucial for carvacrol sensitivity

Given the insensitivity of Danio rerio and Chelonia mydas OTOP1 channels to carvacrol, in contrast to the sensitivity observed in the mouse OTOP1, we performed amino acid sequence alignment of zebrafish, green sea turtle, and mouse OTOP1 in the Phe67-Thr143 and Val216-Arg294 regions (Supplementary Fig. 5f, g). This alignment highlighted differences in OTOP1 amino acid residues among these species, enabling the replacement of corresponding amino acids in mouse OTOP1 with homologous ones in DrOTOP1 and CmOTOP1. To investigate the role of these residues in the sensitivity of OTOP1 to carvacrol, we generated a series of single and double amino acid mutations (Supplementary Fig. 5h). Strikingly, single mutations of S134G and T247A significantly attenuated carvacrol-activated currents (Fig. 5a, b). Furthermore, mutating both residues simultaneously (S134A/T247A) abolished carvacrol activation (Fig. 5c, d). Consistently, inserting serine and threonine (G134S and A247T) into equivalent positions in CmOTOP1 was sufficient to trigger carvacrol activation (Fig. 5e–g).

Fig. 5: S134 and T247 are essential for carvacrol activation.
figure 5

Representative traces of mOTOP1_S134G (a), mOTOP1_T247A (b), and mOTOP1_S134G_T247A (c) currents in response to 500 μM carvacrol or pH 4.5 extracellular solutions. d Normalized currents for fold potentiation measured from single and double mutations of mOTOP1. mean ± SEM, n = 3–5. Statistical significance compared to wild-type was determined using one-way ANOVA followed by Bonferroni’s post hoc test, ***p < 0.001, ****p < 0.0001. Representative traces of CmOTOP1_G134S (e) and CmOTOP1_A247T (f) currents in response to 500 μM carvacrol or pH 4.5 extracellular solutions. g Normalized currents for fold potentiation measured from single mutations of CmOTOP1. mean ± SEM, n = 3. Statistical significance compared to wild-type was determined using one-way ANOVA followed by Bonferroni’s post hoc test, *p < 0.05, **p < 0.01. h Representative traces of mOTOP1 currents in response to 500 μM p-Cymene. i Images were generated using the AlphaFold predicted structure of mOTOP1. Left panel shows sideview. Each zoom-in highlights residues abrogates the activatory effect of carvacrol.

Full size image

To understand the specific interactions between carvacrol and mOTOP1, we utilized p-Cymene, a carvacrol analog lacking the phenolic hydroxyl group (Fig. 5h), to interact with the mOTOP1 channel. Based on patch-clamp recordings, p-Cymene had no effect on the mOTOP1 channel, suggesting that the phenolic hydroxyl group may be essential for mOTOP1 activation. Supporting this hypothesis, thymol, an isomer of carvacrol, also demonstrated activation of the mOTOP1 channel (Supplementary Fig. 7a, b). In the absence of an experimentally resolved structure for the mOTOP1 channel, we analyzed the predicted structures of mOTOP1 generated using Alphafold17 to gain a structural understanding of how specific residues may contribute to carvacrol activation. Inspection of the predicted OTOP1 structure revealed that both S134 and T247 reside in the N domain of the OTOP1 channel (Fig. 5i). Notably, S134 is located in the N domain vestibule, while T247 is situated at the intrasubunit interface. Thus, these findings indicate that activation of OTOP1 channels may be modulated by interactions between carvacrol and residues located within the N domain vestibule and intrasubunit interface.

Discussion

Otopetrin 1 is a highly conserved, multi-transmembrane domain protein found in vertebrates, with no homology to any known transporters, channels, exchangers, or receptors1. Functional profile analysis has revealed that OTOP1 is required for the formation of otoconia in the inner ear of zebrafish and mouse3,18,19. Significantly, Tu et al.2 uncovered that Otop1 encodes for a previously unidentified proton channel in taste cells, a finding further validated by subsequent studies confirming its pivotal role in the perception of sour taste7,8,20,21. The revelation of the zebrafish OTOP1 structure at near-atom resolution15 has shed light on the structure-function dynamics of OTOP1, marking a significant advancement in our understanding. Despite these advancements, several critical questions remain unresolved, such as the open state of the OTOP1 channel, the mechanism of its gating by protons, and the precise location of the proton permeation pathway.

Pharmacological probes that modulate channel activity have been instrumental in elucidating the gating and ion permeation processes, with specific compounds targeting ligand- and voltage-gated ion channels to explore their conformational changes22,23. However, aside from H+, Zn2+, OH−, and NH4Cl2,9,10,13, no agonists have yet been reported. The mechanism by which protons initiate gating and subsequent ion conduction in the OTOP1 channel remains largely elusive2. Therefore, the exploration of novel modulators could significantly enhance our understanding of the roles of OTOP1. In this study, we provide the first evidence that carvacrol acts as an agonist of the OTOP1 channel, demonstrating that carvacrol activates OTOP1 channels in a dose-dependent manner, irrespective of extracellular acidification.

Previous studies have revealed that the OTOP1 channel is selectively permeable to protons2,10. Consistent with these findings, we found that OTOP1 selectively facilitated the conduction of protons in carvacrol-potentiated currents. Zn2+ is considered a potent OTOP1 channel inhibitor and exhibits pH-dependent inhibitory effects on proton currents7,24. Similarly, we found that activation of carvacrol was influenced by changes in pH, suggesting a pH sensitivity in its mechanism of action. Among the sensitivity analysis of representative species across various taxa, all orthologs from terrestrial species are sensitive to carvacrol, whereas aquatic species are not. This observation suggests that OTOP1 activation by carvacrol might potentially serve a natural deterrent role against animal feeding for terrestrial plants rich in carvacrol. Notably, mOTOP1 was gated by carvacrol, and thus selectively conducted protons triggered by voltage ramps, while CmOTOP1 was insensitive to carvacrol modulation. The remarkably distinct functional properties of the two channels, despite their otherwise similar architecture, enabled us to identify motifs involved in gating by employing a chimeric approach.

Interestingly, we identified two small amino acid sections within the Phe67-Thr143 and Val216-Arg294 regions necessary for carvacrol activation in mOTOP1 and sufficient to confer carvacrol-sensitive gating on CmOTOP1. Within these sections, two mutations, S134 and T247, significantly attenuated carvacrol-sensitive gating on mOTOP1, while double mutations (S134A/T247A) almost completely abolished carvacrol-induced currents in mOTOP1. It is noteworthy that two residues are located in completely different channel regions, making direct interaction of a single carvacrol molecule with both sites improbable. Meanwhile, the Hill coefficient of 1.68 observed during carvacrol activation of OTOP1 indicates positive cooperativity in ligand binding. This implies potential allosteric coupling between the two sites upon carvacrol application. Interestingly, the S5-S6 linker plays a crucial role in the proton activation and Zn2+ potentiation11,13 in the mOTOP1 channel. Remarkably, we consistently identified T247 in the S5-S6 linker as a key residue responsible for carvacrol activation. Furthermore, the side chain of S134 extended to the intrasubunit interface, a potential pore of the OTOP1 channel25. These results suggest that both the N domain vestibule and intrasubunit interface may be involved in carvacrol-mediated proton flow.

While few regulators of OTOP1 channels have been identified, the proton and alkali binding sites have been notably characterized10,11,26. For instance, H229 in the S5–S6 linker is a crucial factor for sensing protons in the human OTOP1 channel11. Mutations in K221 and R554 near the S5-S6 and S11-S12 linkers result in a marked decrease in alkali affinity, while not affecting acid activation in mOTOP110. Notably, key residues in the S5-S6 linker played a crucial role in proton, alkali, and carvacrol sensitivity, indicating that this linker may be an important gating apparatus for ligand-gated activation of the OTOP1 channel (Supplementary Fig. 7c). Unlike proton and alkali, the key residue accounting for carvacrol sensitivity in the S5-S6 linker is threonine, not charged amino acids. Interestingly, the carvacrol -OH group could be hydrogen bonding with the oxygen group that is shared in both the S134 and T247 side chains.

In summary, our findings demonstrate the activation of the mOTOP1 channel by carvacrol and shed light on the molecular basis of OTOP1 activation. Notably, utilizing carvacrol as a lead molecule may enable the creation of agonists with greater specificity and potency. These findings not only contribute to a deeper understanding of the physiological roles of OTOP1 but also suggest potential treatments for conditions resulting from OTOP1 inhibition.

Methods

Transfection of HEK293T cells

HEK293T cells were obtained from the Kunming Cell Bank, Kunming Institute of Zoology, Chinese Academy of Sciences. Cells were cultured in Dulbecco’s modified Eagle’s medium containing 10% premium fetal bovine serum (Corning, 35-081-CV) and 1% penicillin/streptomycin (10000 U/mL penicillin, 10000 μg/mL streptomycin, Gibco) and incubated at 37 °C with 5% CO2. mOTOP1, mOTOP2, mOTOP3, or other plasmids were co-transfected with enhanced green fluorescent protein (5:1) into the HEK293T cells using a Lipofectamine® 3000 transfection Kit (Invitrogen, L3000-015). Cells were digested with 0.25% trypsin-EDTA after 36 h of transfection and plated on coverslips for pH imaging or patch clamp recordings.

Gene synthesis and mutagenesis

The gene sequences of OTOP1 orthologs with their corresponding NCBI code including, Mus musculus OTOP1 (NCBI Gene ID: 21906), Chelonia mydas OTOP1 (Gene ID: 102932521), Mus musculus OTOP2 (Gene ID: 237987), Mus musculus OTOP3 (Gene ID: 69602), Danio rerio OTOP1 (Gene ID: 322893), Columba livia OTOP1 (Gene ID: 102088863), Oryx dammah OTOP1 (Gene ID: 120862214), Gallus gallus OTOP1 (Gene ID: 2024404040), Equus caballus OTOP1 (Gene ID: 100069620), Sorex araneus Linnaeus OTOP1 (Gene ID: 101539020), and Bufo gargarizans OTOP1 (Gene ID: 122924553) were synthesized by human expression system optimization and cloned into the pCDNA3.1 eukaryotic expression vector by Sangon Biotech (Shanghai). All OTOP1 chimeras and point mutants were constructed using a Mut Express® II Fast Mutagenesis Kit V2 (C214, Vazyme) following the provided instructions. Chimeras and point mutants were verified by sequencing.

pH imaging

HEK293T cells were cultured in cell culture dishes (35 mm) and transfected with 5 μg of mOTOP1 plasmid about 36 h. Following enzymatic digestion, the cells were plated onto coverslips and incubated at 37 °C. After two hours, then subjected to pH fluorescence experiments (Fig. 2a). Wash cells with Live Cell imaging Solution (LCIS), Mix 10 μL of pHrodoTM Red AM with 100 μL of PowerLoadTM concentrate and add to 10 mL of LCIS, Add the pHrodoTM Red AM/ PowerLoadTM/LCIS (Life Technologies, P35372) mix to cells and incubate at 37 °C for 15-30 min, wash cells using OTOP1 channel extracellular solution(160 mM NMDG-Cl, and 10 mM HEPES, pH 7.4 adjusted with HCl), pHrodoTM Red fluorescence intensity was measured for each cell in response to pH 5.2 solutions (160 mM NMDG-Cl, and 10 mM HomoPIPES, pH adjusted with HCl) or pH 5.2 solutions added 1 mM carvacrol.

pH imaging optics and data acquisition were conducted with a Hamamatsu C14440-20UP digital camera on an Olympus SpinSR10 microscope equipped with a B617/73 filter cube. Emissions at 630 nm were detected after excitation with 560 nm light. Fluorescence intensity of pHrodo Red in cells was normalized with pH 7.4 solution (F0). OlyVIA software from Olympus was used for processing and analyzing fluorescence images. The last time segment of stimulation was selected to calculate the average fluorescence intensity of each cell for statistical analysis.

Patch-clamp recordings

Current recordings were performed using a HEKA EPC10 amplifier controlled by PatchMaster software (HEKA). Micropipette pullers (SUTTER P-97) pull borosilicate glass (WPI) into patch pipettes and fire polished to a resistance of 4–6 MΩ using a polisher (NARISHIGETM MF-830). Patch pipettes for whole-cell patch-clamp recordings of proton currents were filled with 160 mM NMDG-Cl and 10 mM HEPES (for pH 7.4). The extracellular solutions contained 160 mM NMDG-Cl and either 10 mM HEPES (for pH 7.4), 10 mM MES (for pH 6-5.5), or 10 mM HomoPIPES (for pH 5-4), pH adjusted with HCl. In some experiments, for gap-free recording, the membrane potential was held at 0 mV. For voltage ramp recording, voltage ramp from −100 mV to +80 mV were applied to monitor channel currents.

A gravity-driven system (RSC-200, Bio-Logic) was used to perfuse either bath or carvacrol solutions directly onto the cell membrane. This setup included tubes filled with bath or carvacrol solutions in different concentrations, with a pipette positioned before the perfusion tube outlet. Channel currents were recorded for each cell following the application of these solutions, with normalization and analytical methods detailed in the figure legends.

In the ion selectivity experiments2 (Fig. 2c), the pipette solution contained 160 mM NMDG-Cl and 10 mM HEPES (pH adjusted to 7.4 with HCl), The extracellular solutions contained 160 mM NMDG-Cl and 10 mM HEPES (for pH 7.4). Solutions of activated mOTOP1 channel containted: 160 mM NMDG-Cl, 10 mM MES (for pH 5.5), and 500 µM carvacrol, pH adjusted with HCl. For assays involving different ion solutions, 160 mM NMDG-Cl was substituted with equimolar concentrations of KCl, NaCl, LiCl, or CsCl, with 200 µM amiloride added to inhibit endogenous acid-sensing ion channels. For the Ca2+ selectivity experiment, 60 mM NMDG-Cl was replaced by 40 mM CaCl2 to maintain consistent osmolarity, 100 µM DIDS was added to recording solutions to block endogenous Cl− currents.

For measurement of permeability ratio, (Fig. 2e, f), the pipette solution contained: 130 mM TMA-methane sulfonate, 5 mM TEA-Cl, 2 mM Mg-ATP, 5 mM EGTA, 2.4 mM CaCl2 and 80 mM MES, pH adjusted with TMA-OH. NMDG+ - containing OTOP1 channel extracellular solution in the perfusion system contained: 130 mM NMDG-methane sulfonate, 2 mM CaCl2, 100 mM HEPES (pH 7.3), pH adjusted with NMDG-OH. Na+- containing OTOP1 channel extracellular solution in the perfusion system contained: 130 mM Na-methane sulfonate, 2 mM CaCl2, 100 mM HEPES (pH 7.3), pH adjusted with NaOH. For substitution experiments involving K+, Li+, and Cs+, refer to Na+, 130 mM NMDG+ was replaced by equimolar concentrations of K+, Li+, or Cs+ and 50 µM ruthenium red was added to block endogenous transient receptor potential channel, 100 µM DIDS was added to block endogenous Cl− channels, 200 µM amiloride added to inhibit endogenous acid-sensing ion channels. Membrane potential was held at 0 mV, voltage ramp from −80 mV to +80 mV was applied to record channel currents and measured Erev. Erev was used to determine the relative permeability of monovalent cation X to H+ (PX/PH) according to the Goldman-Hodgkin-Katz (GHK) equation27,

$$\frac{{P}_{{X}^{+}}}{{P}_{{H}^{+}}}=\frac{(1-\exp (\frac{{zF}\Delta {E}_{{rev}}}{{RT}}))\times {[{H}^{+}]}_{o}}{{[{X}^{+}]}_{o}\times \exp ({zF}\Delta {E}_{{rev}}/{RT})}$$

where ΔErev, R, T, Z, and F are the reversal potential, the universal gas constant, absolute temperature, ion valency, and Faraday constant, respectively.

The activity experiment of carvacrol on mOTOP2 and mOTOP3 (Fig. 3a, b and Supplementary Fig. 3a–c), under neutral conditions, extracellular solutions contained 160 mM NMDG-Cl and 10 mM HEPES, pH adjusted with HCl to 7.4, carvacrol was directly added to extracellular solutions at different final concentrations. Membrane potential was held at 0 mV, voltage ramp from −100 mV to +80 mV was applied to record channel currents. For experiments in Supplementary Fig. 3a–c, the pipette solution contained: 160 mM NMDG-Cl, 10 mM MES and pH was adjusted with HCl to 6.0, extracellular solutions contained: 160 mM NMDG-Cl and 10 mM buffer (HEPES for pH 7.4, MES for pH 5.5), carvacrol was directly added to extracellular solutions at 500 µM final concentrations. Membrane potential was held at +30 mV, voltage ramp from −100 mV to +80 mV.

Drugs

We screened about 1000 bioactive molecules (Supplementary Data 1–3) and found two kinds of OTOP1 channel agonist, namely carvacrol and its isomer thymol. Carvacrol, thymol and p-Cymene were dissolved in dimethyl sulfoxide (DMSO), then separately packaged and stored at −20 °C. Prior to use in experiments, these solutions were diluted with various extracellular solutions.

The initial screening used the gap-free mode of patch clamp recording, membrane potential was held at 0 mV. The final concentration of each small molecule compound and peptide were 50 µM and 5 µM, respectively, with 10 compounds in each group, dissolved in a pH 5.5 extracellular solution and the pH of the mixtures were maintained at pH 5.5, the pipette solution contained: 160 mM NMDG-Cl, 10 mM HEPES (for pH 7.4), extracellular solutions contained: 160 mM NMDG-Cl and 10 mM buffer (HEPES for pH 7.4, MES for pH 5.5), after the cells had been clamped, perfusion was switched to pH 5.5 to stimulate the cells, and then quickly switched to a mixture of pH 5.5 and the compounds, while monitoring the changes in currents throughout the process, each group of compound mixtures was tested independently at least three times, the groups with significant changes in currents were further monitored for activity until it was determined whether a certain compound acted independently on the OTOP1 channel.

Statistical analysis and generation of AlphaFold models

All results are shown as means ± standard error of the mean (SEM), with n representing the number of independent biological replicates of the experiments. Statistical analyses (analysis of variance (ANOVA) or student’s t-test) were performed using GraphPad Prism9 (GraphPad Software Inc), with *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 indicating statistical significance. The structural models of mOTOP1 were predicted using AlphaFold, and visualizations of these models were generated with PyMOL. Illustrative current traces were obtained via IgorPro v6.04, and sequence alignments were conducted using MEGA-X software.

Statistics and reproducibility

Comparisons between groups were made using student’s t-test and one or two-way ANOVA followed by multiple comparison test, or Bonferroni’s post hoc test, as appropriate. All repetitions are performed independently, the legend in the figure shows the specific repeated values. Fig. 1b, 100 µM, n = 6; 200 µM, n = 6; 500 µM, n = 6; 1000 µM, n = 6, Fig. 1f, n = 6. Fig. 3c, mOTOP1, n = 5; mOTOP2, n = 4; mOTOP3, n = 3. The source data underlying Figs. 1b, e, f; 2a, b, d, f; 3c, I; 4f, l; and 5d, g can be found in supplementary data 4. All statistical results are presented as mean ± SEM.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.