Ligand-gated ion channels (LGICs)¹ mediate the fast inhibitory and excitatory responses of neuronal and muscle cells to neurotransmitters. A universal feature of the type of “Cys-loop” class of LGIC is a common topology of four membrane- spanning segments (M1–M4) and a huge N-terminal extracellular domain with a hyperconservated cysteinebridge motive (1). In vertebrates this “Cys-bridge” family of phylogenetically related genes codes for cation channels activated by acetylcholine and serotonin or for anion channels activated by GABA and glycine (1). In addition, glutamateand serotonin-gated anion channel genes are known in invertebrates (2, 3). Recently, genes for histamine-gated chloride channels and GABA-gated cation channels were identified in invertebrates (4 –7). The molecular basis of further channel types like acetylcholine-gated chloride channels in invertebrates is, however, still unknown (8). Information from the Drosophila melanogaster genome sequencing project allows identifying all members of the superfamily of ligand-gated ion channels occurring in this species by bioinformatic analysis of new homologous genes. The summarized data obtained from several published bioinformatic analyses (5, 6, 9, 10) show that the group of ligand-gated “chloride” channels consists of 12 genes that are coding for GABA, histamine, and glutamate receptors or new, homologous ion channel types. Four members of this group cannot be directly assigned to the GABA, glutamate, or histamine branches and thus code for putative new types of ligand-gated chloride channels with yet unknown function. In a systematic expression approach of these predicted novel types of ion channels in Xenopus oocytes, it was found that none of the typical neurotransmitters activated these novel types of channels (6). Therefore, we extended the molecular biological analysis of the mRNA and found that the gene CG6112 encodes for transcripts that undergo extensive splicing. The functional expressions of these splice variants in Xenopus oocytes and Sf9 cells revealed a unique combination of pharmacological and biophysical properties.

EXPERIMENTAL PROCEDURES

Computer Analysis—The pHCl clones were sequenced using the LICOR 4200 laser fluorescent sequencing system (MWG Biotech, Ebersberg, Germany), Fluorescence Labeled Cycle Sequencing Kit (Amersham Biosciences), and infrared fluorescence-labeled primers (MWG) as described before (11). The sequences were analyzed with SAPS (12), ScanPROSITE, Prot-Param, and Predict Protein (13). The programs FASTA (14), BLITZ (15), BLASTN (16), and TBLASTN (17) were used to search EMBL and Swiss-Prot data bases. pHCl splice variants and similar sequences identified by FASTA were aligned by ClustalW (18).

Detection of pHCl Splice Variants in Different Drosophila Stages by RT-PCR—Total RNA of D. melanogaster imagos, pupae, larvae III, and eggs was isolated by the RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol for animal tissues. DNA was removed by on-column digestion during RNA purification with the Qiagen RNase-Free DNase Set, whereas the RNA was bound to the silica-gel membrane. The OneStep RT-PCR System (Qiagen) was used to transcribe 1µg of total RNA at 30 min, 50 °C and 15 min, 95 °C; subsequently, PCR amplification was performed for 35 cycles of 1 min, 94 °C; 1 min, 60 °C; and 2 min, 72 °C with final extension at 72 °C for 10 min employing splice variant-specific primer pairs (0.6 µM). As a negative control RT-PCR and PCR amplification was performed without template, and as a second control 1 µg of total RNA was used in a PCR amplification with gene-specific primers but without prior RTPCR step.

Detection of variants A–C were as follows: variant A, a 296-bp fragment with HT-F1 plus HT-R2 or 550 bp with HT-F2 plus HT-R2, variant B: 499 bp with HT-F2 plus HT-R2, variant C: 620 bp with HT-F1 plus HT-R2. (The locations of the primers are in Fig. 1A: HT-F1 and HT-F2; Š, HT-R1 and HT-R2). Primers used were: HT-R1, 5´-CTCCGATCCTGCTTAGTACTGCTGG-3´; HT-R2, 5´-TTACGAAAATCCTTTATAGTGTACAAAG-3´; HT-F1, 5´-GTTGCCTACAGGTCGAGTTGACA-3´; and HT-F2, 5´-CGTGCCAGCTAGATCGATGATAG-3´.

Preparation of Plasmid DNA for Xenopus laevis Oocyte Microinjection— The plasmids pCT19189A–C, which contain PCR products of the different pHCl splice variants generated with the primer pair AATTTGATGAGTCCAGTTCGGATAAGG and GCTTAATTTTACGAAAATCC, originated from the systematic expression screening approach described previously (6). The cDNAs were cloned into the blunt-ended XbaI site of the expression vector pSMyc (19). This vector construct facilitates expression of a fusion protein consisting of the N-terminal membrane import sequence of the guinea pig serotonin receptor (20) followed by a myc tag and then by the pHCl channel beginning at amino acid 39. Such constructs have been proven useful for the functional expression of ligand gated ion channels in heterologous systems and can substitute for the missing endogenous membrane import sequence (21). Plasmid DNA used for microinjection was prepared using an endotoxinfree Qiagen Maxiprep kit (Qiagen, Hilden, Germany) dissolved in water to yield 1µg/µl and frozen in aliquots until use for injection.

Whole Mount in Situ Hybridization with D. melanogaster Embryos— Antisense and sense RNA probes were labeled with digoxigenin-UTP (DIG-RNA-labeling kit SP6/T7, Roche Applied Science) by in vitro transcription using SP6 or T7 RNA-polymerase. The vector pCR BluntII TOPO (Invitrogen) containing the complete ORF of pHCl-A was linearized with HindIII (antisense) or EcoRV (sense) and served as the template. RNA probes were hydrolyzed at 60 °C under alkaline conditions (0.2 M sodium carbonate, pH 10.2) to yield probes with a length under 500 nucleotides. In situ hybridization on embryos was performed according to the method of Tautz and Pfeifle (22). Briefly, embryos were collected 9–12 h old. The embryos were then dechorionated for 3 min with 50% sodium hypochloride bleach (Sigma) and washed several times with 1 phosphate-buffered saline, fixed with a solution containing equal parts F-phosphate-buffered saline (4% formaldehyde in 1 phosphate-buffered saline, filtered), and heptane for 20 min with frequent shaking. Embryos were devitelinized with equal parts heptane and methanol for 2 min with vigorous shaking and allowing embryos to settle, followed by three times washing with 100% methanol and storage in 100% methanol at 4 °C. After rehydration for several times with 1 PBT (pH 7.4, 130 mM NaCl, 7 mM Na₂HPO₄, 3 mM NaH₂PO₄, 0.1% Tween 20 (v/v)) and first fixation with 1 F-PBT (4% paraformaldehyde in 1 PBT) for 20 min while shaking embryos were washed and shaken 3 5 min in 1 PBT. First fixation was followed by digestion with proteinase K (25µg/ml in 1xPBT) for 1 min. Reaction was stopped by inactivating proteinase K (ICN Biomedicals) with 2 mg/ml glycine in 1 PBT for 3 min at RT while shaking. After several washings (1x30 s, 2x5 min with 1xPBT) embryos were fixed for a second time (refixation) with 1xF-PBT for 20 min and followed by thorough washings while shaking (5x5 min with 1xPBT, 1x20 min with 1:1 PBT: prehybridization buffer). Embryos were incubated thereafter for 1 h at 50 °C in prehybridization buffer (50% formamide, 5xSSC, 50µg/ml heparin, 0.1% Tween 20) without shaking and then overnight incubation at 50 °C in hybridization buffer containing the linearized and freshly denaturized digoxigenin-labeled probe (50, 100, and 200 ng/ml). The embryos were washed several times: 5x15 min in prehybridization buffer at 50 °C without shaking, 1x20 min while shaking at room temperature with prehybridization buffer, 1x20 min while shaking at room temperature with 1:1 PBT:prehybridization buffer, 2x10 min while shaking at room temperature with 1xPBT. The embryos were treated for 1 h with 1% blocking solution (Roche Applied Science) in 1xPBT while rocking followed by incubation for 1–2 h with anti-digoxigenin- AP Fab fragments (1:2000, Roche Applied Science). Following eight 10-min washes in 1xPBT while shaking and two 10-min rinses in AP buffer (50 mM MgCl₂, 100 mM NaCl, 100 mM Tris, pH 9.5, 1 mM levamisole, 0.1% Tween 20), the nitro blue tetrazolium/5-bromo-4- chloro-3-indolyl phosphate color substrates (Roche Applied Science) were used to detect the hybridized probes. Reaction was stopped by several washes with 1xPBT while shaking. Embryos can be stored in 1xPBT buffer at 4 °C or be dehydrated step by step (1x5 min 40%, 70%, and 96% ethanol) while shaking and embedded in Canada balsam (Roth) or 100% glycerin (previous incubation in 70% glycerin in H₂O for 24 h).

Injection of cDNA into Xenopus Oocytes—Ovarian tissue was taken from anesthetized female Xenopus laevis (Nasco, Fort Atkinson, WI), and oocytes were released from the follicle tissue with collagenase (Sigma, 2 mg/ml). Stage V oocytes were selected by hand and plated individually into the conical wells of a 96-microtiter plate (Greiner, Frickenhausen, Germany) filled with modified Barth’s medium containing (in mM): NaCl 88, NaHCO₃ 2.4, KCl 1, Ca(NO₃)₂ 0.33, CaCl₂ 0.41, MgSO₄ 0.82, Tris/HCl 5 (pH 7.4, 200 mosmol/kg) (23). Oocytes were seeded with their animal (brown) pole facing up so that the nucleus is located just underneath the cell membrane (24). This facilitated intranuclear injection of cDNA that has been described before (25). We used a semi-automated system, the Roboocyte (Multi Channel Systems, Reutlingen, Germany) whose features have been described elsewhere (26). Varying concentrations of cDNA between 40 and 100 ng/µl gave rise to reproducible expression levels and channel properties. After injection, cells were then incubated for 2–5 days at 19 °C in Barth’s medium with gentamicin (50µg/ml), and functionally expressing cells were identified with the Roboocyte.

Two-electrode Voltage Clamp Experiments—Electrophysiological experiments on oocytes were carried out using the two-electrode voltage clamp method (27). The standard extracellular superfusion solution was normal frog Ringer’s solution containing (in mM): NaCl 115, KCl 2.5, CaCl₂ 1.8, HEPES 10 (pH 7.2, 240 mosmol/kg). Where stated the pH of the solutions was altered by addition of either NaOH or HCl and routinely checked before and during experiments. Functionally expressing oocytes were identified with the Roboocyte by clamping the oocytes to 80 mV and superfusion with a frog Ringer’s solution of pH 9. Further electrophysiological and pharmacological experiments were carried out on a manual set up. Cells were penetrated with two microelectrodes filled with 3 M KCl, usually clamped to 80 mV with a voltage clamp amplifier (TEC01/02, npi, Tamm, Germany), and the membrane currents were recorded. If not stated differently, recordings were performed at a holding potential of 80 mV and a sampling rate of 20 Hz. Substances were delivered from two reservoirs reaching the cell 7 s after valve opening and exchanging the solution in the recording chamber within 2 s. Data acquisition and analysis were performed with Pulse Pulsefit software (HEKA Elektronik GmbH, Lambrecht, Germany). All measurements were carried out at room temperature (23–28 °C) except those investigating temperature-dependent effects. For those, the glass-enclosed temperature sensor of a digital thermometer (Mawitherm, Germany) was positioned near the oocyte into the flowing stream of the extracellular solution.

Cell Culture and Transfection of Sf9 Cells for Patch-clamping Experiments— Sf9 cells were grown at 26 °C in Sf-900 II SFM (serum-free medium) (Invitrogen) supplemented with 10µg/ml gentamycin (Invitrogen). Semiconfluent cells were transfected in 24-well dishes (Nunc) on 12-mm glass coverslips by using the non-liposomal Fu- GENE 6 transfection reagent (Roche Applied Science) according to the manufacturer’s instructions. For the transfection, 1µg per dish of plasmids pIE1–3-pHCl-A, -B, or -C was used. Therefore the pCT19189A–C inserts were cloned SacII/NruI into the multicloning site of the insect expression vector pIE1–3 (Novagen). Efficiency of transfection, typically 20%, was checked by cotransfection of 0.5µg of pIE1–3-EGFP. For this purpose, EGFP was taken from pEGFP-N1 (BD Biosciences Clontech, Palo Alto, CA) and cloned into the SacII/ NotI site of pIE1–3. Electrophysiological experiments were done 24–48 h after transfection.

Whole Cell Voltage Clamp Experiments on Sf9 Cells—Membrane currents of EGFP-expressing Sf9 cells cotransfected with one of the pHCl splice variants were recorded in the whole cell configuration of the patch clamp technique (28). Application of test substances and bath solutions of various pH were applied using the U-tube-reversed-flow technique (29) with an application time of 1–2 s at intervals of 1 min. The perfusion chamber had a volume of 0.5 ml and was continuously perfused (flow rate 3 ml/min) with external bath solution driven by gravity. The standard external bath solution contained (in mM): 150 NaCl, 4 KCl, 2 MgCl2, 2 CaCl2, 10 Hepes (pH 7.3 adjusted with 1 N NaOH, 320 mosmol/kg). The pipette solution contained (in mM): 150 KCl, 10 K-EGTA, 10 Hepes (pH 7.3 adjusted with 1 N KOH, 320 mosmol/kg). For pH-variation experiments, the external bath solution was adjusted to pH 6.1 with 1 N HCl. In the case of pH 8.6 Tris buffer was used instead of Hepes buffer. For chloride exchange experiments, the pipette solution contained (in mM): 120 KF, 30 KCl, 10 K-EGTA, 10 Hepes (pH 7.3 adjusted with 1 N KOH, 320 mosmol/kg). Microelectrodes were pulled from borosilicate glass capillaries (external diameter 1.6 mm, Hilgenberg, Malsfeld, Germany) on a Zeitz Puller. The resistance of the fire-polished pipettes was 4–7 megohms using the internal and external solutions described above. All experiments were carried out at room temperature (22–25 °C). Currents were measured with an L/MEPC 7 patch clamp amplifier (HEKA Elektronik GmbH, Lambrecht, Germany). After the giga-seal formation, the EPC 7 circuitry was used to minimize the fast capacitance transients. No compensation was made for the series resistance after the whole cell configuration was obtained. The holding potential was ∼70 mV unless otherwise stated. The analog signals were low-pass (Bessel) filtered at 3.15 kHz (whole cell measurements) and digitized at 1 kHz. For recording and analysis the PClamp software (Axon Instruments, version 6.03) was used.

Drugs—Stock solution in Me₂SO were diluted to various concentrations into normal Ringer’s solution of the following compounds: picrotoxin (50 mM), capsaicin (100 mM), fipronil (100 and 10 mM), ivermectin (10 mM), avermectin B1a (major component of ivermectin, 1 mM), histamine (100 mM), dopamine (10 mM), octopamine (100 mM), and glycine (100 mM). 1 and 2% Me2SO did not have a significant effect on the membrane current of pHCl-injected oocytes or Sf9 cells at pH 7.2.

Statistics—Data are shown as mean S.D.

RESULTS

Sequence Analysis of the Putative Novel Invertebrate LGIC— The genomic region around the gene CG6112 that encodes a putative novel type of invertebrate ligand-gated ion channel was examined for coding regions and deduced transcripts homologous to known Drosophila ligand-gated ion channel subunit sequences. This analysis led to a postulated mRNA sequence that was experimentally proved to exist by RT-PCR and sequencing. The longest transcript identified experimentally in this way encompasses nearly the complete open reading frame of the postulated transcript except for a few nucleotides at the 5´ end.

The originally found cDNA was named pHCl according to the later identified features of the expressed channel (pH-sensitive chloride channel) has an open reading frame of 1464 nucleotides that predicts a protein of 487 amino acids (56 kDa). The extracellular N terminus consists of 277 amino acids in toto, starts with a signal peptide of 18 amino acids (30) followed by the conserved Cys-bridge (positions 195 and 209) and the four predicted transmembrane regions (M1-M4) conserved in ligand gated-chloride channels (Fig. 1A). A hydrophobicity plot detects three hydrophobic regions in the central part and one at the C-terminal part that fit to the location of M1–M4 in other LGICs (data not shown). The putative pore forming M2 region of pHCl is similar to the M2 region of other ligand-gated chloride channels, suggesting that the pHCl pore is chloride-selective also (Fig. 1B). As in other LGICs consensus sequences for putative N-glycosylation sites (positions 135, 180, 250, 263, and 336) and a protein kinase C-phosphorylation site (position 383) can also be detected (Fig. 1A). A putative orthologous gene exists in Anopheles gambiae; in addition to that, pHCl shows the greatest homology to invertebrate glutamate, teleosts, and mammalian glycine receptor subunits and exhibits a considerable amino acid identity with the D. melanogaster glutamategated (28%), the histamine-gated (23%), and the Rattus norvegicus  α3-glycine-gated chloride channels (24%), respectively (Fig. 1A). However, a tree constructed of the known and postulated amino acid sequences of Drosophila LGICs shows that the pHCl protein does not fit into the GABA, glutamate, or histamine groups of LGIC and forms a sub-branch of its own (Fig. 2).

Sequencing of the cloned cDNAs revealed the existence of several splice variants (pHCl-A, pHCl-B, and pHCl-C). We identified three sites of different splicing that can theoretically generate a variety of eight different splice variant combinations. In the N-terminal region (positions 68–92, Fig. 1A), a stretch of 25 amino acids is present (Variant 1, pHCl-A and pHCl-B) or lacking (pHCl-C) due to the presence or absence of an exon in the mRNA. In the region located at M1–M2, at the splicing site 2, pHCl-C differs at five positions due to the alternative use of an exon in the mRNA (Variant 2, Fig. 1A). In the cytoplasmic loop between M3 and M4, pHCl-A differs at a stretch of 17 amino at the splicing site 3. In the variants pHCl-B and pHCl-C, the stretch is absent due to the usage of different splice sites (positions 385–401, Fig. 1A).

Localization and Stage-specific Expression of Splice Variants— To test if the mRNAs for these splice variants are expressed in a stage-specific expression pattern, we performed RT-PCR with splice variant-specific primer pairs (Fig. 3). In all tested developmental stages (egg, larvae, pupae, and adult fly) of Drosophila, the expression of the different variants pHCl-A, -B, and -C was detected with apparently no variations depending on the developmental stage and the type of the splice variant. To locate the expression of pHCl in different tissues qualitatively, whole mount in situ hybridization with 9- to 12-h old embryos and larvae I was performed (Fig. 4). Drosophila embryos as well as larvae I (Fig. 4, A–D) showed a strong expression of pHCl in the neural cord and a weaker expression in the hindgut (Fig. 4, B and C).

Electrophysiological Characterization of pHCl Homomers— Oocytes injected with cDNA of one of the pHCl splice variants (pHCl-A, pHCl-B, or pHCl-C) exhibit pH-sensitive currents that are not found in non-injected controls. Changing the pH of the extracellular solution from pH 7.2 to 5.8 strongly reduced the membrane current, whereas changing it to a more basic pH of 9.0 evoked a non-desensitizing membrane current in pHCl- A-expressing oocytes (Fig. 5A). The splice variants pHCl-B and pHCl-C showed the same qualitative dependence of the current on the extracellular pH when expressed in oocytes (data not shown). The pHCl-A splice variant expressed most reliably in Xenopus oocytes and the electrophysiological characterization was therefore concentrated on this splice variant. All three pHCl-splice variants could also be functionally expressed in Sf9 cells, respectively, and were activated by basic and inhibited by acidic extracellular pH (Fig. 5B). The membrane current of non-transfected Sf9 cells showed no sensitivity to the pH of the extracellular solution. As Fig. 6 shows, the membrane current in pHCl-A-expressing oocytes was half-maximal at pH 7.33±0.16. In normal frog Ringer’s solution, the membrane current of pHCl-A-expressing oocytes was significantly higher than that of non-injected controls indicating that an additional conductance exists at pH 7.2 due to expression of pHCl-A ( -777± -594 nA (n=58) versus -136±-133 nA (n=28)). We also observed that oocytes kept in Barth’s solution of pH 6.0 remained longer viable than those kept at pH 7.2.

The current-voltage relationship of the additional membrane current in pHCl-A-expressing oocytes activated by enhancing the pH of the extracellular solution is slightly rectifying and has a reversal potential of -41±5 mV (n=16) in normal frog Ringer’s solution (Fig. 7). This is in the range of the reversal potential of 53 mV for chloride ions calculated by the Nernst equation assuming an intracellular chloride concentration of 15 mM. Reducing the extracellular chloride concentration to 36.3 and 12.1 mM shifts the reversal potential of the pH-induced current to more positive potentials ( 28±14 mV, n=9 and 12±17 mV, n=9). To maintain a constant offset potential at the bath electrode we used agar bridges for measurements with low extracellular chloride concentrations. The deviation of the measured from the calculated reversal potential for chloride ions ( -22 mV for 36.3 mM and +5 mV for 12.1 mM chloride extracellularly) may be attributed to the fact that the intracellular chloride concentration of the oocyte is not constant. Lowering the extracellular chloride concentration could induce leaking of chloride ions from the cytoplasm into the extracellular solution and result in a less positive reversal potential for chloride ions. The reversal potential of the pHinduced current did not depend on the extracellular pH, showing that no proton- or hydroxide-ion currents are involved (data not shown). To further support the finding, that chloride ions permeated through pHCl channels, we performed similar ion exchange experiments in Sf9 cells. Under nearly symmetrical chloride conditions, the current carried by the spontaneously open pHCl-A channels reversed at 6.2 mV, a reversal potential near zero, as expected for these ionic conditions ([Cl]_i 150 mM, [Cl]_o 162 mM; Fig. 8). Lowering of the intracellular chloride concentration shifted the reversal potential to 45.3 mV, in good agreement with the prediction by the Nernst equation for a chloride-selective ion channel ([Cl]_i 30 mM, [Cl]_o 162 mM E_Cl 43.3 mV at 25 °C; Fig. 8).

Pharmacological Properties of the Currents through pHCl Channels—The extracellular application of 10 µM ivermectin on pHCl-A-expressing oocytes activates a membrane current that desensitizes slowly in the presence of ivermectin. However, we mostly observed that after ivermectin application the membrane current did not return to its original value nor did it reach a stable plateau after 10-min superfusion with frog Ringer’s solution. The amplitude of the additional current stimulated by ivermectin depends strongly on the pH of the extracellular solution. After reduction of the membrane current due to application of pH 5.5 extracellularly, the membrane current was only slightly enhanced by subsequent addition of ivermectin at pH 5.5 (Fig. 9A). Activation occurred slowly, and the additional current did not desensitize over 40 s. In contrast, the effect of ivermectin at pH 8.5 was much more pronounced (Fig. 9A), and the membrane current activated and desensitized rapidly even in the presence of ivermectin.

In Sf9 cells, currents through pHCl-C could also be evoked by application of avermectin B1a, which is the major component of ivermectin. As in oocytes, currents through pHCl-A activate and inactivate rapidly in the presence of avermectin B1a applied at basic pH (8.6) (Fig. 9B). 10 µM ivermectin and 10 µM avermectin B1a, respectively, had no effect on the membrane currents of non-injected oocytes and nontransfected Sf9 cells. The application of high (1 mM) and low (1 µM) concentrations of various neurotransmitters (glutamate, GABA, glycine, histamine, acetylcholine, L-serine, Lalanine, taurine, -alanine, glycine, octopamine, dopamine, and N-methyl-D-aspartic acid) at pH 6 or 7.2 or 9 on either pHCl-expressing oocytes or Sf-9 cells did not induce any change in the membrane current compared with application of normal frog Ringer’s solution at the respective pH.

The chloride channel blockers niflumic acid, flufenamic acid, and phenylantranile acid did not block currents through pHCl-A channels at a concentration of 1 mM, respectively. Neither did the insecticide fipronil (10 µM) induce any change in the membrane current of pHCl-A-expressing oocytes at pH 7.2 or 9. 1 mM of picrotoxin, a plant-derived non-selective blocker of ligand-gated anion channels, led to a half-maximal inhibition of currents through pHCl-A channels induced by extracellular pH of 9.0. Compared with channels that are considered to be blocked by picrotoxin (31, 32), the pHCl-A channel is not sensitive to picrotoxin.

pHCl-A Channels Are Temperature-sensitive and Inhibited by Capsaicin—The membrane currents of pHCl-A-expressing cells are strongly modulated by the extracellular temperature. In contrast to non-injected oocytes where membrane currents exhibited only little temperature sensitivity, we found that reducing the temperature decreased and increasing the temperature stimulated currents through pHCl-A channels (Fig. 10A). Starting from a room temperature between 23 and 28 °C, oocytes were superfused first with cold frog Ringer’s solution that lowered temperature in the bath chamber to 13–20 °C as measured with a sensor right besides the oocyte. On average, membrane currents were reduced by 16±10 nA per degree Celsius (n=10) in pHCl-A-expressing oocytes compared with 1.8±4 nA per degree Celsius (n=8) in non-injected controls. Raising the temperature of the extracellular solution yielding 31–44 °C near the oocyte stimulated the membrane currents in pHCl-A-expressing oocytes by 46±26 nA per degree Celsius (n=10) and by 7±4 nA per degree Celsius (n=9) in non-injected oocytes. To further characterize the temperature dependence of the membrane currents, the temperature coefficients Q10 were determined in an Arrhenius plot in which the amplitude of the common logarithm of the current was plotted against the reciprocal of the absolute temperature (33, 34). The factor by which the peak membrane current decreased upon a 10 °C drop or increased upon a 10 °C rise in temperature were 1.7±0.24 (n=9) and 1.7±0.43 (n=13), respectively. The reversal potential of the additional membrane current evoked by the rise in the temperature of the extracellular solution was -44±9 mV (n=7) in pHCl-Aexpressing oocytes compared with -20±9 mV (n=3) in non-injected controls.

Capsaicin did not activate a membrane current in pHCl-Aexpressing oocytes but slightly reduced the membrane current at 7.2 where approximately half of the pHCl-A channels are open. Further quantification of this effect showed that capsaicin blocked the additional current evoked by basic extracellular pH with an IC₅₀ of 51±13µM (n=5–7 per data point, Hill coefficient of 1.4±0.26) (Fig. 10B).

DISCUSSION

A systematic analysis of the Drosophila genome data reveals the existence of a novel branch of ligand-gated ion channel (LGIC) subunits. It was named pHCl according to the properties of the expressed ion channel that is sensitive to pH and permeable for chloride ions. The overall structure clearly classifies pHCl unequivocally as a member of the superfamily of the cys-bridge type of LGICs. pHCl shows nearly identical similarity to glutamate-, glycine-, and histamine-gated ion channels; however, it does not belong to any of these ion channel types. It seems possible that pHCl shares the same common ancestral gene as postulated for the ligand-gated chloride channels (35) but separated early and evolved independently from the later ion channel subunits. This new branch of ion channels is possibly unique for insects (or maybe arthropods). A putative orthologue gene is present in Anopheles, but in the genomic data of nematodes, teleosts and mammals, no such gene can be found. pHCl encodes for a variety of ion channel subunits due to splicing of the mRNA at, at least, three different positions. The three different splicing sites can generate a variety of eight different ion channels subunits. An RT-PCR analysis suggests that all combinations are expressed in all developmental stages of Drosophila. To elucidate, if the splice variants are responsible ion channel subunits with different properties, we chose the pHCl-A, pHCl-B, and pHCl-C subunits that are different at all tree positions for a detailed functional characterization.

We found that all characterized splice variants of pHCl are strongly sensitive to external pH, activated by the insecticidal compounds ivermectin and avermectin B1a, respectively, in a pH-dependent manner and modulated by extracellular temperature. Thus, the described chloride channels most likely encode pH-sensitive ion channels, and compounds affecting these channels have the potential to provide novel strategies in agriculture and public health.

Various LGICs are sensitive to protons. In invertebrates, a modulation by external pH has been reported for a GABAgated Cl^- conductance in the crayfish leg opener muscle fiber (36). However, in contrast to the pHCl channels, this conductance was inhibited by raising the extracellular pH. In vertebrates, protons were found to differentially regulate neuronal GABA_A receptors, resulting in potentiation, inhibition, or no effect (37–41). An inhibition of single channel currents of GABA_A receptors by H⁺ in outside-out granular cell patches of early developmental stages was shown to result from an increase in the long shut times (39). In primary hippocampal neurons the major effect of protons on GABA_A receptors was revealed to be an enhancement of the desensitization and binding rates by decreasing proton concentration (42). Specifically, it was shown that even variations of a few tenths of a pH unit can have major effects on the amplitudes and kinetics of GABA_A receptors. Also, modulation of GABA receptors by external pH was shown to be dependent on the receptor subunit composition (39, 43). Lately, a single histidine residue in the ion channel domain of the β-subunit was identified to be solely responsible for proton regulation of α₁β_i heteromers (44). The pHCl-A channel shows moderate similarity to the vertebrate ρ1 GABA receptors that were also shown to be sensitive for protons (45, 46). GABA-mediated currents through receptors composed of the rat ρ1 subunit show a similar sensitivity to protons over a wide range of acidic and alkaline pH like the pHCl-A. However, on ρ1-oligomers, protons act as a modulator of GABAevoked currents but do not activate channels by themselves as they do on the pHCl-A channel.

From our data, we cannot decide if the ion channel is gated by hydroxyl ions and protonation/deprotonation of amino acids in a putative ligand binding site (model 1) or if the protons modulate the open probability of the channel by acting on an amino acid in the pore region (model 2). The fact that LGICs have a considerably open probability in the unliganded state has been reported for several type of ion channels (5, 47). According to model 2 the current evoked in pHCl-A at an elevated pH would then represent the unliganded form of the receptor that exhibits a considerable open probability in this state. Ivermectin or an as yet unidentified endogenous ligand would then gate this channel. The open ligand-bound state is then equally modulated by the pH, resulting in a small activation at an acidic pH and an enhanced activation at a basic pH. So, in model 2, the pH modulation of the pore and the gating of the channel (by ivermectin) would then be two separable processes and ivermectin would act as an agonist. In model 1, the hydroxyl ion would really “gate ” this channel; e.g. by deprotonating an amino acid side chain in a potential ligand binding region and ivermectin would act as an allosteric modulator as described for glutamate gated channels in Caenorhabditis elegans and Drosophila (48, 49) and the nicotinic acetylcholine receptor (50).

It is tempting to speculate whether the pHCl-channel is a modulatory subunit that can confer pH sensitivity to a heteromultimeric ligand-gated anion channel. The expression of pHCl in the neural cord implicates a role of pHCl in synaptic transmission. Transient changes in the extracellular pH are produced by excitatory and inhibitory neuronal activity (51–53). Variations of the interstitial pH are also induced by the release of vesicles with acidic contents into the synaptic cleft (54) as well as the reuptake of neurotransmitter (55). The local proton concentration near the plasma membrane is influenced by the passive and active ion transport across the membrane. Specifically, for the GABA_A receptor, the permeation of HCO₃ ions influences the pH in the immediate environment of the ion channel (37, 56). If located postsynaptically, pHCl could modulate the strength of GABA-mediated inhibitory synaptic transmission. At an extrasynaptically located receptor, acidification or alkalization of the extracellular space would strongly modulate inhibitory neurotransmission in general, and this would have physiological implications on the neuronal activity of the total Drosophila.

The temperature sensitivity of the pHCl-A splice variant was quantified by dependence of the membrane currents of pHCl- A-expressing oocytes upon a 10 °C decrease or increase in extracellular temperature. The Arrhenius plot was linear over the whole temperature range yielding a Q₁₀ value of 1.7 for both hot and cold temperatures. This relatively low Q₁₀ value together with the linearity of the Arrhenius plot indicates that the pHCl channel is not gated by temperature but merely modulated as many ligand- or voltage-gated ion channels. Similar temperature sensitivities of current amplitudes have been reported for GABA-induced chloride currents in sensory frog neurons (57) and for the mean open time of nicotinic acetylcholine receptor channels in BC3H-1 cells (58).

Some members of the family of transient receptor potential channels that are activated by temperature are also activated by capsaicin, the main active component of red hot chili peppers. We therefore tested whether pHCl-A is sensitive to capsaicin and found that pHCl-A-mediated currents in oocytes are inhibited by capsaicin with an IC₅₀ of 51 µM. This IC₅₀ for the inhibitory effect of capsaicin is 100-fold higher than that for the activation of the vanilloid receptor (59) but is of the same order of magnitude reported for the inhibition of various voltagegated cation channels (60–66). In a more recent report (67), capsaicin was found to inhibit voltage-gated sodium, calcium, and potassium currents with IC₅₀ values between 9 and 40 µM in the nodose ganglion, the primary sensory ganglion of the vagus. This nonspecific block of voltage-gated ion channels seemed to be independent of the action of capsaicin on the VR-1 receptor also expressed in the majority of these cells. Storeoperated calcium channels were shown to be inhibited by micromolar concentrations of capsaicin (68, 69). To our knowledge, no anion channel has been reported to be inhibited by capsaicin. However, capsaicin has been shown to potentiate cAMP-stimulated CFTR currents with an apparent EC₅₀ of 48 µM (70).

The expression of the pHCl-splice variants in the hindgut may imply a role in the water and/or salt resorption from the urine and the feces. For most insects, the pH of the gut lumen varies with a general trend to neutral to acidic hindguts (71). Possibly, chloride ions are taken up through pHCl channels in the epithelium, creating a hyperosmotic intracellular environment. As a result, water is also reabsorbed and flows toward the hemolymph. However, it might also be possible that the pHCl channel is expressed in the enervating neurons rather than in the epithelium of the hindgut which cannot be decided from the in situ hybridization.

Acknowledgments—We acknowledge the excellent technical assistance of A. Stoeck, S. Seil, and M. Kuester.

REFERENCES

1. Karlin, A., and Akabas, M. H. (1995) Neuron 15, 1231–1244

2. Cully, D. F., Vassilatis, D. K., Liu, K. K., Paress, P. S., Van der Ploeg, L. H., Schaeffer, J. M., and Arena, J. P. (1994) Nature 371, 707–711

3. Ranganathan, R., Cannon, S. C., and Horvitz, H. R. (2000) Nature 408, 470–475

4. Gisselmann, G., Plonka, J., Pusch, H., and Hatt, H. (2004) Br. J. Pharmacol. 142, 409–413

5. Zheng, Y., Hirschberg, B., Yuan, J., Wang, A. P., Hunt, D. C., Ludmerer, S. W., Schmatz, D. M., and Cully, D. F. (2002) J. Biol. Chem. 277, 2000–2005

6. Gisselmann, G., Pusch, H., Hovemann, B. T., and Hatt, H. (2002) Nat. Neurosci. 5, 11–12

7. Beg, A. A., and Jorgensen, E. M. (2003) Nat. Neurosci. 6, 1145–1152

8. Kehoe, J., and McIntosh, J. M. (1998) J. Neurosci. 18, 8198–8213

9. Xue, H. (1998) J. Mol. Evol. 47, 323–333

10. Witte, I., Kreienkamp, H. J., Gewecke, M., and Roeder, T. (2002) J. Neurochem. 83, 504–514

11. Saeger, B., Schmitt-Wrede, H. P., Dehnhardt, M., Benten, W. P., Krucken, J., Harder, A., Samson-Himmelstjerna, G., Wiegand, H., and Wunderlich, F. (2001) FASEB J. 15, 1332–1334

12. Brendel, V., Bucher, P., Nourbakhsh, I. R., Blaisdell, B. E., and Karlin, S. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 2002–2006

13. Appel, R. D., Bairoch, A., and Hochstrasser, D. F. (1994) Trends Biochem. Sci. 19, 258–260

14. Pearson, W. R., and Lipman, D. J. (1988) Proc. Natl. Acad. Sci. U. S. A. 85, 2444–2448

15. Smith, T. F., and Waterman, M. S. (1981) J. Mol. Biol. 147, 195–197

16. Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. J. (1990) J. Mol. Biol. 215, 403–410

17. Altschul, S. F., Madden, T. L., Schaffer, A. A., Zhang, J., Zhang, Z., Miller, W., and Lipman, D. J. (1997) Nucleic Acids Res. 25, 3389–3402

18. Thompson, J. D., Higgins, D. G., and Gibson, T. J. (1994) Nucleic Acids Res. 22, 4673–4680

19. Wellerdieck, C., Oles, M., Pott, L., Korsching, S., Gisselmann, G., and Hatt, H. (1997) Chem. Senses 22, 467–476

20. Lankiewicz, S., Lobitz, N., Wetzel, C. H. R., Rupprecht, R., Gisselmann, G., and Hatt, H. (1998) Mol. Pharmacol. 53, 202–212

21. Gisselmann, G., Galler, A., Friedrich, F., Hatt, H., and Bormann, J. (2002) Eur.J. Neurosci. 16, 69–80

22. Tautz, D., and Pfeifle, C. (1989) Chromosoma 98, 81–85

23. Gurdon, J. B., and Wickens, M. P. (1983) Methods Enzymol. 101, 370–386

24. Dumont, J. N. (1972) J. Morphol. 136, 153–179

25. Ballivet, M., Nef, P., Couturier, S., Rungger, D., Bader, C. R., Bertrand, D., and Cooper, E. (1988) Neuron 1, 847–852

26. Schnizler, K., Kuster, M., Methfessel, C., and Fejtl, M. (2003) Receptors Channels 9, 41–48

27. Stuhmer, W. (1992) Methods Enzymol. 207, 319–339

28. Hamill, O. P., Marty, A., Neher, E., Sakmann, B., and Sigworth, F. J. (1981) Pflugers Arch. 391, 85–100

29. Fenwick, E. M., Marty, A., and Neher, E. (1982) J. Physiol. 331, 577–597

30. Nielsen, H., Engelbrecht, J., Brunak, S., and Von Heijne, G. (1997) Protein Eng. 10, 1–6

31. Etter, A., Cully, D. F., Liu, K. K., Reiss, B., Vassilatis, D. K., Schaeffer, J. M., and Arena, J. P. (1999) J. Neurochem. 72, 318–326

32. Zhang, D., Pan, Z. H., Zhang, X., Brideau, A. D., and Lipton, S. A. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 11756–11760

33. Vyklicky, L., Vlachova, V., Vitaskova, Z., Dittert, I., Kabat, M., and Orkand, R. K. (1999) J. Physiol. 517, 181–192

34. Hille, B. (2001) in Ionic Channels in Excitable Membranes, Third Ed. (Hille, B., ed) pp. 51, Sinauer Assoc., Sunderland, MA

35. Vassilatis, D. K., Elliston, K. O., Paress, P. S., Hamelin, M., Arena, J. P., Schaeffer, J. M., Van der Ploeg, L. H., and Cully, D. F. (1997) J. Mol. Evol. 44, 501–508

36. Pasternack, M., Bountra, C., Voipio, J., and Kaila, K. (1992) Neuroscience 47, 921–929

37. Kaila, K. (1994) Prog. Neurobiol. 42, 489–537

38. Zhai, J., Peoples, R. W., and Li, C. (1998) Pflugers Arch. 435, 539–545

39. Krishek, B. J., and Smart, T. G. (2001) J. Physiol. 530, 219–233

40. Robello, M., Baldelli, P., and Cupello, A. (1994) Neuroscience 61, 833–837

41. Pasternack, M., Smirnov, S., and Kaila, K. (1996) Neuropharmacology 35, 1279–1288

42. Mozrzymas, J. W., Zarnowska, E. D., Pytel, M., and Mercik, K. (2003) J. Neurosci. 23, 7981–7992

43. Krishek, B. J., Amato, A., Connolly, C. N., Moss, S. J., and Smart, T. G. (1996) J. Physiol. 492, 431–443

44. Wilkins, M. E., Hosie, A. M., and Smart, T. G. (2002) J. Neurosci. 22, 5328–5333

45. Wegelius, K., Reeben, M., Rivera, C., Kaila, K., Saarma, M., and Pasternack, M. (1996) Neuroreport 7, 2005–2009

46. Rivera, C., Wegelius, K., Reeben, M., Kaila, K., and Michael, P. (2000) Neuropharmacology 39, 977–989

47. Maksay, G., Thompson, S. A., and Wafford, K. A. (2003) Neuropharmacology 44, 994–1002

48. Arena, J. P., Liu, K. K., Paress, P. S., Schaeffer, J. M., and Cully, D. F. (1992) Brain Res. Mol. Brain Res. 15, 339–348

49. Cully, D. F., Paress, P. S., Liu, K. K., Schaeffer, J. M., and Arena, J. P. (1996) J. Biol. Chem. 271, 20187–20191

50. Krause, R. M., Buisson, B., Bertrand, S., Corringer, P. J., Galzi, J. L., Changeux, J. P., and Bertrand, D. (1998) Mol. Pharmacol. 53, 283–294

51. Kraig, R. P., Ferreira-Filho, C. R., and Nicholson, C. (1983) J. Neurophysiol. 49, 831–850

52. Chesler, M. (1990) Prog. Neurobiol. 34, 401–427

53. Chesler, M., and Kaila, K. (1992) Trends Neurosci. 15, 396–402

54. Krishtal, O. A., Osipchuk, Y. V., Shelest, T. N., and Smirnoff, S. V. (1987) Brain Res. 436, 352–356

55. Bouvier, M., Szatkowski, M., Amato, A., and Attwell, D. (1992) Nature 360, 471–474

56. Kaila, K., Paalasmaa, P., Taira, T., and Voipio, J. (1992) Neuroreport 3, 105–108

57. ffrench-Mullen, J. M., Tokutomi, N., and Akaike, N. (1988) Br.J. Pharmacol. 95, 753–762

58. Dilger, J. P., Brett, R. S., Poppers, D. M., and Liu, Y. (1991) Biochim. Biophys. Acta 1063, 253–258

59. Caterina, M. J., Schumacher, M. A., Tominaga, M., Rosen, T. A., Levine, J. D., and Julius, D. (1997) Nature 389, 816–824

60. Petersen, M., Pierau, F. K., and Weyrich, M. (1987) Pflugers Arch. 409, 403–410

61. Castle, N. A. (1992) Cardiovasc. Res. 26, 1137–1144

62. Akins, P. T., and McCleskey, E. W. (1993) Neuroscience 56, 759–769

63. Kehl, S. J. (1994) Br. J. Pharmacol. 112, 616–624

64. Baker, M. D., and Ritchie, J. M. (1994) Proc. R Soc. Lond. B Biol. Sci. 255, 259–265

65. Grissmer, S., Nguyen, A. N., Aiyar, J., Hanson, D. C., Mather, R. J., Gutman, G. A., Karmilowicz, M. J., Auperin, D. D., and Chandy, K. G. (1994) Mol. Pharmacol. 45, 1227–1234

66. Kuenzi, F. M., and Dale, N. (1996) Br. J. Pharmacol. 119, 81–90

67. Bielefeldt, K. (2000) Neuroscience 101, 727–736

68. Fischer, B. S., Qin, D., Kim, K., and McDonald, T. V. (2001) J. Pharmacol. Exp. Ther. 299, 238–246

69. Flemming, R., Xu, S. Z., and Beech, D. J. (2003) Br. J. Pharmacol. 139, 955–965

70. Ai, T., Bompadre, S. G., Wang, X., Hu, S., Li, M., and Hwang, T. C. (2004) Mol. Pharmacol. 65, 1415–1426

71. Harrison, J. F. (2001) Annu. Rev. Entomol. 46, 221–250

FIG. 1. Comparison of pHCl with sequences of homologue LGICs. A, comparison of the amino acid sequence of pHCl-A (487 amino acids), Drosophila GluCl (456 amino acids, AF297500), HisCl-α₂ (426 amino acids, AF435470), and R. norwegicus GlyCl-α₃ (464 amino acids, NM_052724). The location of the three splice variants is marked in bold letters and dots. M1–M4 segments are boxed, and the putative signal peptide (positions 1–19) of pHCl is in bold letters. N-Glycosylation sites (positions 135, 180, 250, 263, and 336) and the protein kinase C phosphorylation site (positions 383) are doubly underlined. Further important amino acids mentioned in the text are typed in bold letters for easier identification. Conserved positions with a minimum of three identical amino acids are shaded in gray. RT-PCR Primers are indicated by ‹ (forward primer) and Š (reverse primer). B, alignment of the M2 regions of pHCl with the ligand-gated chloride channels HisCl-α₁ (AF435469), RDL (M69057), GluCl (AF297500), and Gly-α₁ (NM_013133) and the serotoninand acetylcholine-gated cation channels gp-5HT₃As (AF006462) and ratnAChR-αM (X03986). The putative poreforming M2 regions of pHCl are more similar to the M2 regions of other ligandgated chloride channels from Drosophila, such as Rdl and GluCl, than to those of ligand-gated cation channels, suggesting that the pHCl pore may also be chloride-selective.

FIG. 2. A dendrogram showing the relationship of pHCl to the different groups of LGICs. The dendrogram was constructed by MegAlign (Lasergene) on the basis of the ClustalW algorithm (18). pHCl clearly forms a separate branch in the group of LGICs that further contains the families of GABA-, glutamate-, and histaminegated channels.

FIG. 3. Detection of pHCl splice variants in different Drosophila stages by RT-PCR. RT-PCR with splice variant-specific primer pairs detected expression of pHCl in all tested stages of Drosophila (adult, pupae, larvae III, and eggs). pHCl-A: a 296-bp fragment with primers HT-F1 plus HT-R2; pHCl-A/B: 550 bp (for pHCl-A), 499 bp (for pHCl-B), with HT-F2 plus HT-R2; pHCl-C: 620 bp with HT-F1 plus HT-R2. (+: cDNA; -: mRNA prior reverse transcription).

FIG. 4. Whole mount in situ hybridization with D. melanogaster embryos and larvae I. Antisense pHCl-A RNA probes labeled with digoxigenin-UTP detected high expression levels in the central nervous system/neural cord (A–D) together with a particular hybridization signal in the hindgut (B and C). The negative controls, sense pHCl-A RNA probes, showed no hybridization signals at all (E and F). Antisense probes: A, embryo lateral view; B, embryo dorsal; C, larvae I lateral; D, embryo ventral. Sense probes: E, embryo ventral; F, embryo lateral. Scale bar: 100 µm; nt: neural tube; hg: hindgut.

FIG. 5. Influence of the external pH on pHCl mediated currents. A, membrane currents of a pHCl-A-expressing oocyte and a non-injected oocyte controlled by pH upon superfusion with normal frog Ringer’s solution of acidic (5.8) and basic (9.0) pH. B, membrane currents of a pHCl-C-expressing Sf9 cell and a non-transfected cell controlled by pH. Short application of pH 6.1 completely blocked the membrane current of a pHCl-C-expressing Sf9 cell under control conditions (pH=7.3), whereas changing the pH from 7.3 to pH 8.6 doubled the membrane current.

FIG. 6. Dependence of the membrane current of pHCl-A-expressing oocytes on the pH of the extracellular solution. Current amplitudes were referred to the amplitude induced by changing the pH from 7.2 (normal frog Ringer’s) to pH 8 after subtraction of the current measured at pH 5.83. Data points were fitted by I/I_max 1/(1+(EC₅₀/ x)ⁿ) yielding a half-maximal activation at a H⁺ concentration of 4.71x10^-8 7.58x10^-9 mol, which corresponds to a pH of 7.33±0.16 and a Hill coefficient of 1.07±0.18 (n=3–5 per data point).

FIG. 7. Current-voltage relationship of pHCl-A mediated current in oocytes. Current-voltage relationship of the additional current activated by changing the extracellular pH from 7.2 to 9.0 in a pHCl- A-expressing oocyte and non-injected oocyte. From a holding potential of -80 mV the membrane potential was clamped for 400 ms from -100 mV to 60 mV in 20-mV steps, total pulse duration was 500 ms. Current values are mean amplitudes of the last 100 ms of the voltage step.

FIG. 8. Dependence of the reversal potential of the pHCl-A carried current on the intracellular chloride concentration. Under physiological ([Cl]_i 30 mM, [Cl]_o 162 mM) or almost symmetrical ([Cl]_i 150 mM, [Cl]_o 162 mM) chloride concentrations the current through pHCl-A channels expressed in Sf9 cells reverses at -45 mV and 6 mV respectively, which is in good agreement with the Nernst equation that yields -43 mV and 0 mV. At a given holding potential extracellular solution of pH 6.1 was applied for 2 s. Current values are the blockable membrane current by application of the pH 6.1 solution for 2 s.

FIG. 9. Potentiation of pHCl-induced currents by avermectins. A, potentiation of activation of pHCl-A expressed in oocytes at basic pH through ivermectin. After strong reduction of the membrane current by changing the pH of the extracellular solution from 7.2 to 5.5 ivermectin enhanced the membrane current only slightly. In basic extracellular solution ivermectin potentiated the stimulation of an additional membrane current that quickly activates and desensitized. B, effect of avermectin B1a on a pHCl-C-expressing Sf9 cell at basic pH (8.6). Application of 10 µM avermectin B1a induced a fast activating and desensitizing membrane current additional to the typical membrane current at basic pH (8.6).

FIG. 10. Temperature dependence and capsaicin sensitivity of the pHCl-A-mediated current. A, modulation of the membrane current of pHCl-A-expressing oocyte by temperature. Current was recorded as tempered extracellular solution was superfused at a holding potential of -80 mV. B, blockade of pH 9-induced currents by capsaicin in oocytes that express pHCl-A. The percentage of the current blocked by capsaicin applied at pH 9 was fitted using I/I_max 1/(1+(IC₅₀/x)ⁿ) yielding an IC₅₀ of 51±13 M and a Hill coefficient of 1.4±0.25. Holding potential was -80 mV.