Advertisement

TRPA1 Is Differentially Modulated by the Amphipathic Molecules Trinitrophenol and Chlorpromazine*

  • Kerstin Hill
    Affiliations
    Institut für Pharmakologie, Charité-Universitätsmedizin Berlin, Campus Benjamin Franklin, Thielallee 67-73, 14195 Berlin, Germany
    Search for articles by this author
  • Michael Schaefer
    Correspondence
    To whom correspondence should be addressed. Tel.: 49-30-84451863; Fax: 49-30-84451818
    Affiliations
    Institut für Pharmakologie, Charité-Universitätsmedizin Berlin, Campus Benjamin Franklin, Thielallee 67-73, 14195 Berlin, Germany
    Search for articles by this author
  • Author Footnotes
    * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Open AccessPublished:January 11, 2007DOI:https://doi.org/10.1074/jbc.M609600200
      TRPA1, a poorly selective Ca2+-permeable cation channel, is expressed in peripheral sensory neurons, where it is considered to contribute to a variety of sensory processes such as the detection of painful stimuli. Furthermore, TRPA1 was also identified in hair cells of the inner ear, but its involvement in sensing mechanical forces is still being controversially discussed. Amphipathic molecules such as trinitrophenol and chlorpromazine have been shown to provide useful tools to study mechanosensitive channels. Depending on their charge, they partition in the inner or outer sheets of the lipid bilayer, causing a curvature of the membrane, which has been demonstrated to activate or inhibit mechanosensitive ion channels. In the present study, we investigated the effect of these molecules on TRPA1 gating. TRPA1 was robustly activated by the anionic amphipathic molecule trinitrophenol. The whole-cell and single channel properties resemble those previously described for TRPA1. Moreover, we could show that the toxin GsMTx-4 acts on TRPA1. In addition to its recently described role as an inhibitor of stretch-activated ion channels, it serves as a potent activator of TRPA1 channels. On the other hand, the positively charged drug chlorpromazine modulates activated TRPA1 currents in a voltage-dependent way. The exposure of activated TRPA1 channels to chlorpromazine led to a block at positive potentials and an increased open probability at negative potentials. The variability in the shape of the I-V curve gives a first indication that native mechanically activated TRPA1 currents must not necessarily exhibit the same biophysical properties as ligand-activated TRPA1 currents.
      TRPA1 (also known as ANKTM1) is a calcium-permeant nonselective cation channel that belongs to the superfamily of the TRP
      The abbreviations used are: TRP, transient receptor potential; AITC, allylisothiocyanate; HEK293, human embryonic kidney 293; YFP, yellow fluorescent protein; eYFP, enhanced YFP; CFP, cyan fluorescent protein; PKCα, protein kinase C α; TNP, trinitrophenol; OHC, outer hair cell.
      2The abbreviations used are: TRP, transient receptor potential; AITC, allylisothiocyanate; HEK293, human embryonic kidney 293; YFP, yellow fluorescent protein; eYFP, enhanced YFP; CFP, cyan fluorescent protein; PKCα, protein kinase C α; TNP, trinitrophenol; OHC, outer hair cell.
      ion channels. Besides its transmembrane domain structure, which is characteristic of all TRP channels, TRPA1 possesses 17 ankyrin repeats in its cytoplasmic N-terminal domain (
      • Corey D.P.
      • Garcia-Anoveros J.
      • Holt J.R.
      • Kwan K.Y.
      • Lin S.Y.
      • Vollrath M.A.
      • Amalfitano A.
      • Cheung E.L.
      • Derfler B.H.
      • Duggan A.
      • Geleoc G.S.
      • Gray P.A.
      • Hoffman M.P.
      • Rehm H.L.
      • Tamasauskas D.
      • Zhang D.S.
      ), which could function as a gating spring (
      • Sotomayor M.
      • Corey D.P.
      • Schulten K.
      ), thus making TRPA1 an attractive candidate as a potential mechanosensor.
      The activation mechanisms of TRPA1 are diverse. It can be activated by environmental irritants derived from plants such as isothiocyanates (
      • Jordt S.E.
      • Bautista D.M.
      • Chuang H.H.
      • McKemy D.D.
      • Zygmunt P.M.
      • Hogestatt E.D.
      • Meng I.D.
      • Julius D.
      ), icilin (
      • Story G.M.
      • Peier A.M.
      • Reeve A.J.
      • Eid S.R.
      • Mosbacher J.
      • Hricik T.R.
      • Earley T.J.
      • Hergarden A.C.
      • Andersson D.A.
      • Hwang S.W.
      • McIntyre P.
      • Jegla T.
      • Bevan S.
      • Patapoutian A.
      ), and allicin (
      • Bautista D.M.
      • Movahed P.
      • Hinman A.
      • Axelsson H.E.
      • Sterner O.
      • Hogestatt E.D.
      • Julius D.
      • Jordt S.E.
      • Zygmunt P.M.
      ) and has been proposed to contribute to a variety of sensory processes, including detection of noxious cold, mechanosensation, and inflammatory hyperalgesia (
      • Story G.M.
      • Peier A.M.
      • Reeve A.J.
      • Eid S.R.
      • Mosbacher J.
      • Hricik T.R.
      • Earley T.J.
      • Hergarden A.C.
      • Andersson D.A.
      • Hwang S.W.
      • McIntyre P.
      • Jegla T.
      • Bevan S.
      • Patapoutian A.
      ,
      • Bandell M.
      • Story G.M.
      • Hwang S.W.
      • Viswanath V.
      • Eid S.R.
      • Petrus M.J.
      • Earley T.J.
      • Patapoutian A.
      ). Although an involvement in neurogenic inflammation and nociceptive hypersensitivity has been clearly demonstrated (
      • Kwan K.Y.
      • Allchorne A.J.
      • Vollrath M.A.
      • Christensen A.P.
      • Zhang D.S.
      • Woolf C.J.
      • Corey D.P.
      ,
      • Bautista D.M.
      • Jordt S.E.
      • Nikai T.
      • Tsuruda P.R.
      • Read A.J.
      • Poblete J.
      • Yamoah E.N.
      • Basbaum A.I.
      • Julius D.
      ), the contribution of TRPA1 to the transduction of noxious cold and mechanical stimuli remains controversial. Corey et al. (
      • Corey D.P.
      • Garcia-Anoveros J.
      • Holt J.R.
      • Kwan K.Y.
      • Lin S.Y.
      • Vollrath M.A.
      • Amalfitano A.
      • Cheung E.L.
      • Derfler B.H.
      • Duggan A.
      • Geleoc G.S.
      • Gray P.A.
      • Hoffman M.P.
      • Rehm H.L.
      • Tamasauskas D.
      • Zhang D.S.
      ) provided evidence that TRPA1 is part of the hair cell mechanotransduction process; however, two independently generated TRPA1-deficient mice failed to show vestibular or auditory defects, raising doubts about the contribution of TRPA1 to the function of the hair-cell transduction channel (
      • Kwan K.Y.
      • Allchorne A.J.
      • Vollrath M.A.
      • Christensen A.P.
      • Zhang D.S.
      • Woolf C.J.
      • Corey D.P.
      ,
      • Bautista D.M.
      • Jordt S.E.
      • Nikai T.
      • Tsuruda P.R.
      • Read A.J.
      • Poblete J.
      • Yamoah E.N.
      • Basbaum A.I.
      • Julius D.
      ). In addition to its expression in inner and outer hair cells, TRPA1 can also be found in somatosensory neurons, where it might play a role in decoding sensory or painful mechanical stimuli. Indeed, TRPA1-deficient mice have been reported to display a higher threshold in sensing noxious mechanical stimuli when compared with wild-type mice and had reduced responses to a series of suprathreshold stimuli (
      • Kwan K.Y.
      • Allchorne A.J.
      • Vollrath M.A.
      • Christensen A.P.
      • Zhang D.S.
      • Woolf C.J.
      • Corey D.P.
      ). These findings could not be validated by a second group (
      • Bautista D.M.
      • Jordt S.E.
      • Nikai T.
      • Tsuruda P.R.
      • Read A.J.
      • Poblete J.
      • Yamoah E.N.
      • Basbaum A.I.
      • Julius D.
      ), and it remains to be confirmed that TRPA1 indeed contributes to a high threshold mechanotransduction complex localized in the peripheral sensory system. A direct activation of TRPA1 by mechanical forces has not been described so far. TRPA1 channels are not sensitive to applying either positive or negative pressure via the recording pipette in patch clamp experiments.
      Here we have used an alternative approach to examine a potential mechanical gating of TRPA1. We imparted mechanical stress to the lipid bilayer using two amphipathic molecules, trinitrophenol and chlorpromazine, which have been previously shown to influence gating of mechanosensitive channels (
      • Martinac B.
      • Adler J.
      • Kung C.
      ,
      • Qi Z.
      • Chi S.
      • Su X.
      • Naruse K.
      • Sokabe M.
      ,
      • Maingret F.
      • Fosset M.
      • Lesage F.
      • Lazdunski M.
      • Honore E.
      ).
      Trinitrophenol is a negatively charged amphiphilic compound, which accumulates in the outer leaflet of the plasma membrane, inducing a crenation of the membrane (
      • Sheetz M.P.
      • Singer S.J.
      ). In contrast, the positively charged phenothiazine drug chlorpromazine incorporates preferentially into the inner leaflet of the membrane inducing a so-called cup formation (
      • Sheetz M.P.
      • Singer S.J.
      ). We analyzed the effect of these compounds on gating characteristics of TRPA1 in calcium imaging and electrophysiological experiments. Both molecules modulate TRPA1 gating but do so in different ways. The application of trinitrophenol led to a robust and reversible activation of TRPA1 channels in both the whole-cell and cell-attached configurations, similar to an activation by AITC. In contrast, chlorpromazine inhibited TRPA1 at positive holding potentials but markedly increased the open probability of TRPA1 at negative holding potentials. The presented data give a first indication that TRPA1 gating may be sensitive to membrane deformation and provide additional tools for identifying TRPA1-mediated cation permeabilities.

      EXPERIMENTAL PROCEDURES

      Cell Culture and Reagents—HEK293 cells, stably transfected with human TRPA1 cDNA (HEK293TRPA1), were kindly provided by Dr. Natalie Tigue and Dr. Steve Moore (GlaxoSmithKline, Harlow, UK). Parental HEK293 cells (for transient transfection) or HEK293TRPA1 cells were maintained on tissue culture plastic in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and 2 mm glutamine at 5% CO2 and 37 °C. Cells were plated onto glass coverslips 24–48 h prior to the experiments. GsMTx-4 peptide (PeptaNova, Sandhausen, Germany) was dissolved at a concentration of 100 μm in distilled water, stored at –20 °C, and was freshly diluted into extracellular recording solution before use.
      Transient Transfection of HEK293 Cells—HEK293 cells were seeded onto 35-mm coverslips (105 cells/well) and grown for 18–24 h prior to transfection. Transient transfections were performed using a FuGENE 6 reagent (Roche Diagnostics), 1.5 μg of human TRPA1 together with 100 ng of eYFP as a transfection marker, 1.5 μg of TRPV1-YFP, 1.5 μg of TRPM8-YFP, or 1.5 μg of TRPV4-CFP. For confocal imaging experiments of PKCα translocation, 0.5 μg of an eYFP-tagged PKCα construct was co-transfected with TRPA1. For control experiments, either non-transfected cells from the same coverslip were used (internal controls) or 1.5 μg of pcDNA3 together with 100 ng of eYFP were used in control transfections. Transfected cells were utilized for patch clamp recordings 24–48 h after transfection.
      Electrophysiology—Cells were transferred to a continuously perfused recording chamber (500-μl volume) mounted on the stage of an inverted microscope. For whole-cell patch clamp experiments, the standard bath solution consisted of (in mm): 140 NaCl, 5 CsCl, 2 MgCl2, 10 HEPES, 10 d-glucose pH 7.4 (NaOH). The pipette solution for whole-cell experiments contained (in mm) 140 CsCl, 4 MgCl2, 5 EGTA, 10 HEPES, pH 7.2 (CsOH). For cell-attached recordings, the standard bath solution was used in the patch pipette. All experiments were performed at room temperature using a HEKA EPC-9 amplifier (HEKA Electronics, Lambrecht/Pfalz, Germany). Patch pipettes of 3–5 megaohms were fabricated from borosilicate glass capillaries. Experiments were carried out and analyzed under the control of the Pulse and Pulsefit software (HEKA Electronics). Series resistances were <10 megaohms and were compensated by 75–85%. Stated membrane potentials are corrected for liquid junction potentials. Whole-cell currents were filtered at 3 kHz (four-pole Bessel filter) and sampled continuously at 5 kHz. Single channel currents were sampled at 10 kHz. Data are presented as mean ± S.E. Stated membrane potentials always refer to the physiological inner side of the membrane.
      Data Analysis—Single channel analysis was performed using the Pclamp software (Axon Instruments). In single channel experiments, transitions between open and closed states were detected using a half-amplitude threshold criterion and a minimum event width of 0.2 ms. For recordings with more than one active channel, the number of available channels in each membrane patch was estimated by the maximum number of channels simultaneously open observed in recordings of at least 3 min. For open time analysis, only recordings with a single active channel were used. The open time distributions were fitted to monoexponential functions. Mean channel open times were determined as arithmetic means of the dwell time data.
      Fluorescence Imaging—Measurements of the intracellular Ca2+ concentration ([Ca2+]i) in single cells were carried out using the fluorescent indicator fura-2. In brief, transiently transfected HEK293 cells were seeded on coverslips and loaded with 2 μm fura-2/AM for 30 min in a Hepes-buffered solution consisting of (in mm) 135 NaCl, 6 KCl, 1 MgCl2, 1 CaCl2, 5.5 glucose, 10 HEPES, pH 7.4 (NaOH). This solution was supplemented with 0.2% (w/v) bovine serum albumin. Coverslips were mounted on a monochromator-equipped (Polychrome UV, TILL-Photonics, Martinsried, Germany) inverted microscope (Axiovert 100, Carl Zeiss, Oberkochen, Germany). The fluorescence of fura-2 and of eYFP was sequentially excited at 340, 380, and 475 nm, filtered through a 512-nm long-pass filter, and recorded with a 12-bit cooled CCD camera (IMAGO, TILL-Photonics). The calibration of the calcium concentration was performed using a regression-based spectral unmixing method as described by Ref.
      • Lenz J.C.
      • Reusch H.P.
      • Albrecht N.
      • Schultz G.
      • Schaefer M.
      . For confocal imaging of the PKCα-YFP translocation, experiments were performed with an LSM510 inverted laser scanning microscope and a ×63/1.4 Plan-Apochromat objective (Carl Zeiss MicroImaging, Inc.). CFP was excited with the 458-nm line of an argon laser, and emitted light was collected with a 470–500-nm band pass filter. YFP was excited with the 488-nm laser line, and emission was recorded with a 530–560-nm band pass filter. To obtain a measure for the membrane translocation of PKCα-YFP, regions of interest were defined over the cytosol of the cell (Fcyt). To detect agonist-induced changes in the plasma membrane association, the resulting ratios were normalized to the initial values. In each imaging experiment, 20–40 individual cells were analyzed and averaged. Means of the indicated number of individual experiments were calculated and expressed as mean ± S.E.

      RESULTS

      The biophysical and pharmacological properties of TRPA1 were investigated in a heterologous expression system. TRPA1 was either transiently expressed in HEK293 cells or stably expressed in a HEK293TRPA1 cell line. There were no differences in TRPA1 characteristics detectable apart from a higher TRPA1 current density in the stable cell line when compared with transiently transfected HEK293 cells.
      Trinitrophenol Induces Ca2+ Influx through Recombinant Human TRPA1—We initially used ratiometric single cell calcium imaging to investigate the effect of trinitrophenol on TRPA1 gating. The basal level of the intracellular calcium concentration ([Ca2+]i) and the basal Mn2+ influx (data not shown) were higher in HEK293 cells transiently expressing TRPA1 when compared with non-transfected cells on the same coverslip or mock-transfected HEK293 control cells, indicating a basal activity of TRPA1 (Fig. 1A). The addition of 500 μm trinitrophenol induced a rapid and sustained increase in [Ca2+]i similar to responses elicited by 10 μm AITC (Fig. 1, A and B). The rise in [Ca2+]i could be almost completely reversed by the addition of 10 μm ruthenium red, a broad spectrum TRPV and TRPA1 channel blocker. A trinitrophenol-induced calcium entry was not present in non-transfected control cells (Fig. 1, A and B) or in HEK293 cells transiently expressing the sensory TRP channels TRPM8 or TRPV1 (Fig. 1B). One should mention that calcium imaging experiments were hampered by the fact that, upon the addition of trinitrophenol, UV illumination at 340 and 380 nm led to a decrease in fura-2 fluorescence. To verify TRPA1 activation by trinitrophenol without UV illumination of the probe, we used a PKCα-YFP fusion protein bearing a Ca2+-sensitive C2 domain as a cellular read-out sensor for intracellular calcium that is excited at longer wavelengths. Under resting conditions of the cell, PKCα is located in the cytoplasm and is targeted to the plasma membrane in response to a rise in [Ca2+]i (
      • Cannell M.B.
      • Cheng H.
      • Lederer W.J.
      ). In unstimulated TRPA1-expressing HEK293 cells, PKCα-YFP exhibited a homogenous cytoplasmic fluorescence. Within 20 s after the addition of trinitrophenol, a PKCα-YFP translocation to the cell membrane was visible (Fig. 2A), which was absent in Ca2+-free bath solutions and, thus, most likely reflects a Ca2+-influx through activated TRPA1 channels. Translocation of PKCα-YFP could be partially reversed by the addition of ruthenium red (Fig. 2, A and C). When TRPA1-expressing cells were preincubated with ruthenium red, the PKCα-YFP distribution was unaltered by the addition of trinitrophenol (data not shown). This was also the case in control experiments with HEK293 cells expressing PKCα-YFP alone but no TRPA1, which did not show any PKCα-YFP translocation in response to trinitrophenol addition (Fig. 2D). The activation of TRPA1 by noxious cold remains a controversial topic. When we exposed the cells to 8 °C cold Hepes-buffered solution, we did not observe a translocation of PKCα-YFP to the plasma membrane (n = 4, data not shown), indicating no significant cold activation of TRPA1 in our hands.
      Figure thumbnail gr1
      FIGURE 1Trinitrophenol leads to an increase in [Ca2+]i in TRPA1-expressing HEK293 cells. A, time-lapse analysis of [Ca2+]i during the sequential application of 500 μm trinitrophenol (TNP) and 10 μm ruthenium red (RR). Cells were loaded with the Ca2+ indicator fura-2, washed, and kept in Hepes-buffered solution supplemented with 1 mm CaCl2. Trinitrophenol and ruthenium red were added to the extracellular medium as indicated by the bars. Twenty to 40 cells were recorded from each coverslip. Averaged data from three individual coverslips for HEK293 cells transiently expressing TRPA1 (filled symbols) and from five coverslips for control cells transfected with the empty pcDNA3 expression vector (open symbols) are shown. B, statistical analysis of experiments such as in A, carried out in pcDNA3-transfected control cells and in HEK293 cells transiently expressing TRPA1, TRPM8, or TRPV1. Open bars represent [Ca2+]i prior to activation of the respective channels; black bars represent [Ca2+]i after the addition of 500 μm TNP; and gray bars represent [Ca2+]i after the addition of the following activating agonists: 10 μm capsaicin (for TRPV1), 10 μm AITC (for TRPA1) or 10 μm icilin (for TRPM8). Data represent means ± S.E. of 3–6 independent experiments. mock, mock-transfected.
      Figure thumbnail gr2
      FIGURE 2Trinitrophenol-induced activation of TRPA1 causes translocation of PKCα-YFP to the plasma membrane. A and B, HEK293 cells were transiently transfected with plasmids encoding TRPA1 (1 μg) and PKCα-YFP (0. 5 μg) (A) or TRPV4 (1 μg) and PKCα-YFP (0.5 μg) (B). Confocal images were taken after the sequential addition of 100 μl of solvent (control), TNP (500 μm), and ruthenium red (30 μm RR); for B, TRPV4 cells were sequentially stimulated with trinitrophenol and 4α-phorbol didecanoate (1 μm 4α-PDD). Representative examples of confocal microscopy images are shown. C, time-lapse experiment showing the cytoplasmic fluorescence intensity of PKCα-YFP in four individual TRPA1-expressing cells during the addition of TNP and ruthenium red as indicated by the bars. D, statistical analysis of experiments such as in C. Cytoplasmic fluorescence was normalized to the initial cytoplasmic fluorescence (averaged over the first 5 s after starting the experiment for each individual cell). Filled circles represent single cells expressing TRPA1 and PKCα-YFP, and open circles represent control cells expressing PKCα-YFP only.
      To investigate whether other TRP channels are also sensitive to trinitrophenol in this assay, we conducted experiments as described above but using heterologously expressed TRPV4. TRPV4 can be activated by hypotonic cell swelling (
      • Strotmann R.
      • Harteneck C.
      • Nunnenmacher K.
      • Schultz G.
      • Plant T.D.
      ) and is thus another candidate for a mechanosensitive channel (
      • O'Neil R.G.
      • Heller S.
      ). In HEK293 cells transiently expressing both PKCα-YFP and TRPV4, trinitrophenol failed to induce a translocation of PKCα-YFP to the plasma membrane, indicating an absence of trinitrophenol activation of TRPV4. As a control for an intact TRPV4 function in our assay, we subsequently activated the channel using 1 μm 4α-phorbol didecanoate (Fig. 2B, 1 μm 4α-PDD), which led to an immediate PKCα-YFP translocation (Fig. 2B). These data, together with the absence of trinitrophenol-induced TRPM8 and TRPV1 activation in calcium imaging experiments, suggest that trinitrophenol is not an unspecific activator of TRP channels.
      Electrophysiological Characterization of the Trinitrophenol-evoked TRPA1 Current—The addition of trinitrophenol clearly led to an elevation of [Ca2+]i in TRPA1-expressing cells. We next wanted to investigate the electrophysiological properties of the currents evoked by trinitrophenol. All experiments shown were carried out in the absence of extracellular calcium to prevent time-dependent loss of channel activity (
      • Nagata K.
      • Duggan A.
      • Kumar G.
      • Garcia-Anoveros J.
      ). As illustrated in Fig. 3A, the application of 500 μm trinitrophenol to HEK293TRPA1 cells resulted in significant increases in whole-cell currents, which were absent in control parental HEK293 cells (Fig. 3B). We tested different concentrations of trinitrophenol for their ability to activate TRPA1. A concentration of 100 μm trinitrophenol failed to activate any significant TRPA1 currents, whereas the addition of 200 μm led to a small increase in channel activity. We could not obtain a full dose-response curve because of solubility problems of trinitrophenol and an increased instability of the patch at trinitrophenol concentrations above 1 mm. The activation of TRPA1 by trinitrophenol was fully reversible, and the reapplication of trinitrophenol repeatedly induced current activation. To confirm that trinitrophenol-activated currents are indeed mediated by TRPA1, we characterized the current-voltage relationship of the evoked current. Voltage ramps from –100 mV to +100 mV confirmed the marked outward rectification and demonstrated a reversal potential close to 0 mV (Fig. 3C) and, thus, were similar to AITC-evoked TRPA1 currents. In accordance with the calcium imaging experiments, trinitrophenol failed to activate currents in TRPV1- and TRPM8-expressing HEK293 cells (Fig. 3C).
      Figure thumbnail gr3
      FIGURE 3Trinitrophenol activates TRPA1 in whole-cell recordings. A, example recording of the development of currents that can reversibly and repeatedly be elicited by 500 μm trinitrophenol in HEK293TRPA1 cells. Data are extracted from voltage ramps and depict the current at +50 mV (filled symbols) and at –50 mV (open symbols); the zero current level is indicated by the dashed line. B, statistical analysis of 4–5 independent experiments such as in A. Peak currents at a holding potential of +50 mV are illustrated before the addition of trinitrophenol (–), after the addition of 500 μm TNP, and after washout of trinitrophenol (w). C, current recordings of HEK293 cells transiently expressing TRPA1, TRPM8, or TRPV1 in response to slow voltage ramps. Cells were stimulated with 500 μm trinitrophenol. For TRPM8- and TRPV1-expressing cells, the current traces for untreated and trinitrophenol-treated cells are overlapping. To control for the presence of functional channels, cells were stimulated with 10 μm menthol (acting on TRPM8) or with 10 μm capsaicin (to activate TRPV1). D, current recordings of HEK293TRPA1 cells in response to voltage steps (–160 mV to +160 mV) starting from a holding potential of +160 mV (upper graph) or –160 mV (lower graph). TRPA1 was activated by 500 μm trinitrophenol. E, statistical analysis of seven independent recordings as shown in the upper panel of D. Open square symbols represent peak currents (tail currents at negative potentials), and filled circles represent plateau currents (average current of last 20 ms). For each cell, the recorded current is normalized to the current obtained at +100 mV.
      We used step protocols to determine gating characteristics of the channel (Fig. 3D). A prepulse of +160 mV was used to fully activate TRPA1 channels in the presence of trinitrophenol. Voltage steps to negative holding potentials then led to an inactivation of the channel, which resulted in distinct tail currents that could be described with a monoexponential function (Fig. 3D, upper graph). The time constant for inactivation varied with the potential applied, from 3.5 ± 0.6 ms (n = 5) for a voltage step from +160 mV to –160 mV to 10.5 ± 1.6 ms (n = 5) for a voltage step from +160 mV to –20 mV. In contrast, a prepulse of –160 mV led to an almost complete inactivation of the trinitrophenol-activated TRPA1 channels and was followed by a slow reactivation after steps to positive holding potentials (Fig. 3D, lower graph).
      We next wanted to know whether trinitrophenol-induced TRPA1 currents exhibit the same single channel characteristics as when activated by AITC. Fig. 4A illustrates that the basal single channel activity in the on-cell configuration was low and that after the addition of trinitrophenol to the bath solution, channel activity was clearly elevated (Fig. 4B). The activation usually occurred with a delay of about 20 s. Note the strong voltage dependence of the open probability, which was higher at positive potentials. A unitary conductance of 106 pS ± 5.5 pS (n = 6) was calculated from I-V curves as shown in Fig. 4D based on amplitude histograms (like that in Fig. 4C) compiled for channel activity recorded at a range of different voltages. The calculated single channel conductance is very similar to the one that has been previously reported for AITC activation of TRPA1 (98 ± 16 pS, (
      • Nagata K.
      • Duggan A.
      • Kumar G.
      • Garcia-Anoveros J.
      )). As the I-V relationship of the unitary currents was linear with a reversal potential close to 0 mV, the outward rectification of the whole-cell current is caused by the higher open probability at positive potentials. We can exclude a possible role of the cytoskeleton in the trinitrophenol-induced activation of TRPA1 as the evoked currents were not abolished after incubating the cells for 2–4 h in 2 μm cytochalasin D (data not shown).
      Figure thumbnail gr4
      FIGURE 4Trinitrophenol elicits single channel activity of TRPA1 in the cell-attached configuration. A and B, example sweeps from a cell-attached recording at different voltages as indicated, prior to (A) and after (B) the addition of 500 μm trinitrophenol to the bath solution. C, histogram of a current recording at a patch potential of +100 mV. D, plot of the mean single channel current amplitude (obtained from fitting Gaussian curves to amplitude histograms such as in C) against the patch potential for a representative experiment. Linear regression analysis of the data yields a slope conductance of 109 pS. For the analysis, only the main conductance level was used, whereas infrequently appearing subconductance levels were unaccounted for.
      Modulation of AITC-induced TRPA1 Currents by Chlorpromazine—We next investigated the influence of chlorpromazine on TRPA1 gating. Chlorpromazine is positively charged and has previously been shown to induce a cup-forming of the plasma membrane and might therefore gate TRPA1 channels via exhibiting mechanical stress to the lipid bilayer. Chlorpromazine alone showed little or no effect on TRPA1 gating. In cells that exhibited a constitutive TRPA1 activity, however, an unexpected modification of the gating characteristics was observed upon addition of chlorpromazine (data not shown). These effects were more clearly discernible when TRPA1 was previously activated by AITC. AITC-induced TRPA1 currents are usually characterized by an outwardly rectifying I-V curve. Fig. 5, A and B, depict TRPA1 gating in response to 10 μm AITC. Tail currents were recorded after a prepulse to either –160 mV or +160 mV. Chlorpromazine affected the tail currents in a complex manner. (i) The inactivation at negative potentials was less pronounced, resulting in an increased current density at negative holding potentials, and (ii) chlorpromazine caused a voltage-dependent block of the current at holding potentials between +20 mV and +80 mV (Fig. 5, C–E). The reversal potential of the elicited current remained unchanged. This modified current is most likely carried by TRPA1 as it is absent in control cells and can be blocked by ruthenium red (Fig. 5F). Again, these changes were unaffected by elimination of the actin cytoskeleton with cytochalasin D.
      Figure thumbnail gr5
      FIGURE 5Modulation of AITC-induced TRPA1 whole-cell currents by chlorpromazine (CPZ). A and B, whole-cell current in HEK293TRPA1 cells in response to voltage step protocols as indicated. TRPA1 currents were evoked by 10 μm AITC. C and D, AITC-stimulated cells were additionally exposed to 50 μm chlorpromazine. E, pooled data of 5 independent experiments such as in A and B. A step protocol with a prepulse to +160 mV was used. Depicted are TRPA1 plateau currents (average current of last 20 ms after voltage step) after stimulation with 10 μm AITC (open symbols) and upon co-stimulation with 10 μm AITC and 50 μm chlorpromazine (filled symbols). The currents were normalized to the plateau current at +160 mV. F, statistical analysis of 4 independent experiments; the current density at +50 mV (open bars) and at –50 mV (filled bars) is shown. Data were obtained from slow voltage ramps from –100 mV to +100 mV. RR, ruthenium red.
      To further study the nature of the chlorpromazine-dependent modulation of the AITC-induced TRPA1 current, we conducted single channel recordings in the cell-attached configuration. The unitary conductance was unaffected by additional chlorpromazine application (103 ± 2.9 pS; n = 4 before the addition of chlorpromazine versus 104 ± 6.2 pS; n = 3 following chlorpromazine application), but the gating behavior of the single channels changed substantially. Channels exhibited shorter but more frequent opening events at positive holding potentials and an increased open probability at negative holding potentials. This is evident in the example traces (Fig. 6, A and B) in the open time histograms at different voltages (Fig. 6C) and in the open probability of single channels (Fig. 6D). We determined the mean open times of AITC-activated TRPA1 channels at positive and negative holding potentials with topen = 20.9 ± 3.9 ms (n = 5) at +80 mV and topen = 2.9 ± 0.33 ms (n = 5) at –80 mV. The values are in a good agreement with a previous study (topen = 4.4 ± 5.9 ms at –80 mV) (
      • Nagata K.
      • Duggan A.
      • Kumar G.
      • Garcia-Anoveros J.
      ). Additional exposure of the AITC-activated TRPA1 cells to chlorpromazine significantly lowered topen to 1.9 ± 0.1 ms (n = 5) at +80 mV holding potential, whereas mean open times at –80 mV remained almost unchanged (topen = 3.9 ± 0.8; n = 5). The closed times of the TRPA1 channel changed also but can only be approximated as the absence of multiple openings during long continuous recordings does not ensure that only one active channel is present in the patch. The acute application of chlorpromazine to cell-attached patches exhibiting AITC-evoked single channel activity led to a decrease in the closed time of the channel by a factor of 9 at both positive and negative holding potentials. We examined the voltage-dependent open probability (Po) of AITC-activated TRPA1 channels in the absence and presence of chlorpromazine (Fig. 6D). At negative and low positive holding potentials, Po was increased about 5-fold in the presence of 50 μm chlorpromazine and was only slightly decreased at positive holding potentials. Thus, under physiological conditions at negative membrane potentials, an exposure of TRPA1-expressing cells to chlorpromazine would lead to an increased excitability of these cells.
      Figure thumbnail gr6
      FIGURE 6Modulation of AITC-induced TRPA1 single channel currents by chlorpromazine (CPZ). A and B, cell-attached recordings of HEK293TRPA1 cells at different voltages as indicated. Single channels were activated by the addition of 10 μm AITC to the bath solution. Current traces (5 s per sweep) were recorded immediately prior to additional exposure to chlorpromazine (A) and after the addition of 50 μm chlorpromazine (B). Level of 0 pA is indicated by a dashed line. C, open time analysis for TRPA1 in the AITC-activated state and upon gating modification by chlorpromazine. Currents were evoked with 10 μm AITC and recorded at +80 mV and –80 mV for 30 s prior to and after modulation by 50 μm chlorpromazine. Data were gathered from patches with one active channel. Pooled data of 5 independent experiments are displayed for each graph. D, statistical analysis of the open probability of single channels depending on the holding potential applied. For patches with more than one channel, the open probability was divided by the number of channels available in the patch. The graph illustrates a summary of 6 independent experiments.
      Activation of TRPA1 Currents by the Tarantula Toxin GsMTx-4—The tarantula toxin GsMTx-4 has been shown to block a range of stretch-activated ion channels (
      • Suchyna T.M.
      • Johnson J.H.
      • Hamer K.
      • Leykam J.F.
      • Gage D.A.
      • Clemo H.F.
      • Baumgarten C.M.
      • Sachs F.
      ,
      • Spassova M.A.
      • Hewavitharana T.
      • Xu W.
      • Soboloff J.
      • Gill D.L.
      ). The proposed mechanism of the block presumably involves partitioning of the amphipathic peptide into the outer leaflet of the plasma membrane (
      • Jung H.J.
      • Kim P.I.
      • Lee S.K.
      • Lee C.W.
      • Eu Y.J.
      • Lee D.G.
      • Earm Y.E.
      • Kim J.I.
      ). To test whether GsMTx-4 shows a similar effect on TRPA1 currents, we have exposed HEK293TRPA1 cells first to trinitrophenol to elicit TRPA1 currents and then, additionally, to 1 μm GsMTx-4 to study a possible gating modification. Surprisingly, GsMTx-4 did not lead to an inhibition of TRPA1 currents but to a massive increase in whole-cell currents (Fig. 7A), exhibiting similar properties as TRPA1 with a reversal potential close to zero and marked outward rectification. We next wanted to know whether GsMTx-4 can activate TRPA1 channels also in the absence of another stimulus. Indeed, 1 μm GsMTx-4 alone was capable of activating large TRPA1-like currents in HEK293TRPA1 cells, which were absent in untransfected HEK293 control cells (Fig. 7C), thus making GsMTx-4 a novel and interesting tool for identifying TRPA1 currents in native cells.
      Figure thumbnail gr7
      FIGURE 7Activation of TRPA1 currents by the tarantula toxin GsMTx-4. A, whole-cell current of a representative HEK293TRPA1 cell in response to 500 μm TNP and additional exposure to 1 μm GsMTx-4. Data were extracted from voltage ramps and depict the current at +70 mV and –70 mV. The inset shows 500-ms voltage ramps from –80 mV to +80 mV applied at the time as marked in the current trace. B, similar experiment as described as in A but without prestimulation with trinitrophenol. C, pooled data of 5 independent experiments such as in A and B. The graph shows the current densities at +70 mV (open bars) and –70 mV (filled bars). Data were obtained from 500-ms voltage ramps from –80 mV to +80 mV. Untransfected HEK293 cells serve as control.

      DISCUSSION

      The highly asymmetric lipid composition of the membrane leads to charge asymmetries within the lipid bilayer. In mammalian cells, the inner leaflet of the plasma membrane is about four times more negatively charged when compared with the outer leaflet (
      • Zhang P.C.
      • Keleshian A.M.
      • Sachs F.
      ). These charge asymmetries cause cationic amphipathic molecules such as chlorpromazine to accumulate preferentially within the inner leaflet of the plasma membrane, whereas anionic amphipathic molecules such as trinitrophenol are expected to remain in the less negatively charged outer leaflet of the lipid bilayer. At high concentrations, the differential insertion of the amphipathic molecule into the outer and inner leaflets then induces a curvature of the membrane (analogue to cell shrinkage) and a crenation (analogue to cell swelling), respectively. In this context, trinitrophenol has previously been shown to expand the exterior leaflet of the lipid bilayer and induce erythrocytes to crenate (
      • Sheetz M.P.
      • Singer S.J.
      ). Hereby arising transversal forces can influence mechanosensitive channels, as has already been shown for trinitrophenol-induced activation of bacterial mechanosensitive channels (
      • Martinac B.
      • Adler J.
      • Kung C.
      ) and mammalian neuronal mechanosensitive K+ channels (
      • Maingret F.
      • Fosset M.
      • Lesage F.
      • Lazdunski M.
      • Honore E.
      ). Chlorpromazine can activate and inhibit mechanosensitive channels via a cup-forming of the membrane, and an activation has been shown for mechanosensitive channels in Escherichia coli (
      • Martinac B.
      • Adler J.
      • Kung C.
      ), whereas hTREK-1 channels are inhibited by chlorpromazine (
      • Miller P.
      • Kemp P.J.
      • Lewis A.
      • Chapman C.G.
      • Meadows H.J.
      • Peers C.
      ).
      Bautista et al. (
      • Bautista D.M.
      • Jordt S.E.
      • Nikai T.
      • Tsuruda P.R.
      • Read A.J.
      • Poblete J.
      • Yamoah E.N.
      • Basbaum A.I.
      • Julius D.
      ) showed that acrolein (2-propenal) and related compounds such as 2-pentenal stimulate TRPA1 currents. Considering the increasing number of TRPA1 agonists, which are at least partly structurally unrelated, one should consider the possibility that for some of the substances, an activation might occur not via a direct binding to the channel protein itself. Acrolein, for example, has also been shown to induce dramatic changes in cell morphology in PC12 cells (
      • Liu-Snyder P.
      • McNally H.
      • Shi R.
      • Borgens R.B.
      ), indicating that a TRPA1 activation might also be caused by mechanical stress.
      In our study, we could clearly demonstrate an activation of TRPA1 by the negatively charged trinitrophenol in calcium imaging experiments as well as in electrophysiological recordings. On the other hand, the positively charged phenothiazine drug chlorpromazine modified TRPA1 gating in a voltage-dependent manner; the open probability at positive potentials was slightly affected, whereas the open probability at negative holding potentials was significantly increased. This overall effect on the open probability is caused by a shortened closed time of the single channels at both positive and negative holding potentials together with a reduced open time only at positive holding potentials. We suggest that the results of this study may give a first indication of a direct modulation of TRPA1 by local mechanical forces, as has been already described for other mechanosensitive K+ and Cl channels where chlorpromazine and trinitrophenol were used in a similar concentration range (
      • Martinac B.
      • Adler J.
      • Kung C.
      ,
      • Qi Z.
      • Chi S.
      • Su X.
      • Naruse K.
      • Sokabe M.
      ,
      • Buckler K.J.
      • Honore E.
      ,
      • Tseng G.N.
      ). Despite strong indices for a mechanically induced modulation of TRPA1 by those molecules, other explanations for the observed gating modifications of TRPA1 have to be taken into account. (i) The incorporation of charged amphipathic molecules could change the surface charge distribution within the membrane, which in consequence might influence a voltage-dependent activation of TRPA1. Incorporation of negatively charged anionic trinitrophenol into the outer leaflet or of positively charged chlorpromazine into the inner leaflet of the plasma membrane would be expected to influence TRPA1 gating in a similar manner. However, we think that it is unlikely that this effect alone is responsible for the observed changes in channel gating as both molecules, chlorpromazine and trinitrophenol, modify TRPA1 gating in different ways. Moreover, TRPM8 and TRPV1 channel opening should be facilitated in a similar manner as they exhibit a comparable outward rectification, but we failed to detect any activation of these channels by trinitrophenol in calcium imaging experiments or electrophysiological recordings. (ii) Conflicting results have been reported concerning the activation of TRPA1 by an elevation of [Ca2+]i. A study by Jordt et al. (
      • Jordt S.E.
      • Bautista D.M.
      • Chuang H.H.
      • McKemy D.D.
      • Zygmunt P.M.
      • Hogestatt E.D.
      • Meng I.D.
      • Julius D.
      ) provided evidence that a rapid rise of [Ca2+]i caused by the addition of thapsigargin is sufficient to activate TRPA1, a finding we could confirm in our experiments (data not shown). The addition of trinitrophenol and chlorpromazine might lead to changes in the intracellular calcium concentration. Indeed, for chlorpromazine, it has previously been shown that it can increase the release of calcium from intracellular stores in cultured human umbilical vein endothelial cells (
      • Chen K.H.
      • Chang B.H.
      • Younan P.
      • Shlykov S.G.
      • Sanborn B.M.
      • Chan L.
      ). To avoid calcium-dependent effects on TRPA1 gating, we conducted all experiments in the absence of extracellular calcium and with intracellular calcium buffered with 5 mm EGTA. Besides this, depletion of intracellular stores by 1 μm thapsigargin prior to stimulation of the cells with trinitrophenol and chlorpromazine did not abolish TRPA1 activation (data not shown), indicating that changes in the calcium concentration are not responsible for TRPA1 gating modification. (iii) Finally, we cannot rule out the possibility that the modulation of TRPA1 gating occurs via a direct binding of the molecules used in this study to the channel proteins. Chlorpromazine, for example, has previously been shown to affect various other ion channels. It inhibits the cardiac ether-a-go-go-related Kv1 1.1 channels (
      • Thomas D.
      • Wu K.
      • Kathofer S.
      • Katus H.A.
      • Schoels W.
      • Kiehn J.
      • Karle C.A.
      ), prevents store-operated Ca2+ entry in PC-12 cells (
      • Choi S.Y.
      • Kim Y.H.
      • Lee Y.K.
      • Kim K.T.
      ), blocks small conductance potassium channels in the brain (
      • Dreixler J.C.
      • Brian J.
      • Cao Y.
      • Roberts M.T.
      • Roizen J.D.
      • Houamed K.M.
      ), and inhibits P- and N-type calcium channels (
      • Dreixler J.C.
      • Brian J.
      • Cao Y.
      • Roberts M.T.
      • Roizen J.D.
      • Houamed K.M.
      ), and the interaction occurs presumably via direct binding to the channel itself. Hinman et al. (
      • Hinman A.
      • Chuang H.H.
      • Bautista D.M.
      • Julius D.
      ) could show that TRPA1 channel activation by some agonists can occur through covalent modification of cysteine residues within the cytoplasmic N terminus of the channel. However, we do not think that this is the case for the observed gating modification of TRPA1 in our study as covalent modification should lead to an irreversible modulation of the function of the channels.
      Chlorpromazine alters some biophysical features of TRPA1 (e.g. increasing the open probability at negative potentials and a voltage-dependent block at positive potentials). When looking for mechanically activated conductances in primary cells expressing TRPA1, one has to consider that the biophysical fingerprint might not be the same as for AITC-activated currents.
      Recently, a peptide toxin, GsMTx-4 from the Chilean rose tarantula Grammostola spatulata, has been isolated that provides the first specific tool for studying mechanosensitive ion channels (
      • Suchyna T.M.
      • Johnson J.H.
      • Hamer K.
      • Leykam J.F.
      • Gage D.A.
      • Clemo H.F.
      • Baumgarten C.M.
      • Sachs F.
      ). The amphipathic peptide GsMTx-4 blocks a range of stretch-activated channels, possibly not via binding to the channel proteins themselves but by accumulating in the outer sheet of the plasma membrane at the interface between the channel protein and boundary lipids (
      • Suchyna T.M.
      • Tape S.E.
      • Koeppe R.E.
      • Andersen O.S.
      • Sachs F.
      • Gottlieb P.A.
      ) where it is believed to relieve lipid stress, resulting in a shift of the activation curve toward stronger stimuli. In this context, an inhibition of TRPC1 and TRPC6 activity by GsMTx-4, acting as a gating modifier, has also been shown (
      • Spassova M.A.
      • Hewavitharana T.
      • Xu W.
      • Soboloff J.
      • Gill D.L.
      ,
      • Bowman C.L.
      • Gottlieb P.A.
      • Suchyna T.M.
      • Murphy Y.K.
      • Sachs F.
      ). Here, we present a contrary effect of GsMTx-4 on another candidate for a mechanosensitive channel, namely the activation of TRPA1. Incorporation of the toxin peptide into the outer leaflet of the plasma membrane might lead to a membrane deformation similar to the effect evoked by trinitrophenol. Also, an interaction of the toxin with a yet unlocated voltage sensor of TRPA1 might account for TRPA1 channel openings by GsMTX-4. For example, the inhibition of KV channels by hanatoxin, another cysteine knot G. spatulata toxin, is caused by binding of the toxin to the voltage sensor, stabilizing its resting conformation (
      • Lee H.C.
      • Wang J.M.
      • Swartz K.J.
      ). Perhaps binding of GsMTx-4 to a functionally equivalent region of TRPA1 will favor an active conformation of the channel, a mechanism that has already been suggested for activation of the capsaicin receptor by spider toxins (
      • Siemens J.
      • Zhou S.
      • Piskorowski R.
      • Nikai T.
      • Lumpkin E.A.
      • Basbaum A.I.
      • King D.
      • Julius D.
      ). Future studies will be required to fully elucidate the nature of the interaction of TRPA1 with membrane-perturbing agents, such as trinitrophenol, chlorpromazine, and GsMTx-4. Amphipathic molecules have been shown to influence the membrane motor of OHC in a way that they inhibit cochlear function. Chlorpromazine treatment of OHC, for example, leads to a shift of the membrane voltage at which they exhibit maximal electromotile gain toward depolarized potentials (
      • Lue A.J.
      • Zhao H.B.
      • Brownell W.E.
      ). In contrast, salicylate, an anionic amphipathic molecule, such as trinitrophenol, inhibits cochlear function by additionally reducing the magnitude of OHC electromotility (
      • Shehata W.E.
      • Brownell W.E.
      • Dieler R.
      ). Both drugs potentially exert their effect on OHC electromotility via membrane bending (
      • Lue A.J.
      • Zhao H.B.
      • Brownell W.E.
      ). A similar effect to the one reported for chlorpromazine has been assigned to GsMTx-4 (
      • Fang J.
      • Iwasa K.H.
      ). It is striking that TRPA1, which has been identified in hair cells of the inner ear, is modulated by similar compounds, pointing to a possible role of TRPA1 in modulating electromotility in OHC.

      Acknowledgments

      We thank Dr. N. Tigue and Dr. S. Moore for providing hTRPA1 stable HEK293 cells and Andy Randall (University of Bristol, UK) for helpful discussions.

      References

        • Corey D.P.
        • Garcia-Anoveros J.
        • Holt J.R.
        • Kwan K.Y.
        • Lin S.Y.
        • Vollrath M.A.
        • Amalfitano A.
        • Cheung E.L.
        • Derfler B.H.
        • Duggan A.
        • Geleoc G.S.
        • Gray P.A.
        • Hoffman M.P.
        • Rehm H.L.
        • Tamasauskas D.
        • Zhang D.S.
        Nature. 2004; 432: 723-730
        • Sotomayor M.
        • Corey D.P.
        • Schulten K.
        Structure (Lond.). 2005; 13: 669-682
        • Jordt S.E.
        • Bautista D.M.
        • Chuang H.H.
        • McKemy D.D.
        • Zygmunt P.M.
        • Hogestatt E.D.
        • Meng I.D.
        • Julius D.
        Nature. 2004; 427: 260-265
        • Story G.M.
        • Peier A.M.
        • Reeve A.J.
        • Eid S.R.
        • Mosbacher J.
        • Hricik T.R.
        • Earley T.J.
        • Hergarden A.C.
        • Andersson D.A.
        • Hwang S.W.
        • McIntyre P.
        • Jegla T.
        • Bevan S.
        • Patapoutian A.
        Cell. 2003; 112: 819-829
        • Bautista D.M.
        • Movahed P.
        • Hinman A.
        • Axelsson H.E.
        • Sterner O.
        • Hogestatt E.D.
        • Julius D.
        • Jordt S.E.
        • Zygmunt P.M.
        Proc. Natl. Acad. Sci. U. S. A. 2005; 102: 12248-12252
        • Bandell M.
        • Story G.M.
        • Hwang S.W.
        • Viswanath V.
        • Eid S.R.
        • Petrus M.J.
        • Earley T.J.
        • Patapoutian A.
        Neuron. 2004; 41: 849-857
        • Kwan K.Y.
        • Allchorne A.J.
        • Vollrath M.A.
        • Christensen A.P.
        • Zhang D.S.
        • Woolf C.J.
        • Corey D.P.
        Neuron. 2006; 50: 277-289
        • Bautista D.M.
        • Jordt S.E.
        • Nikai T.
        • Tsuruda P.R.
        • Read A.J.
        • Poblete J.
        • Yamoah E.N.
        • Basbaum A.I.
        • Julius D.
        Cell. 2006; 126: 1269-1282
        • Martinac B.
        • Adler J.
        • Kung C.
        Nature. 1990; 348: 261-263
        • Qi Z.
        • Chi S.
        • Su X.
        • Naruse K.
        • Sokabe M.
        Mol. Membr. Biol. 2005; 22: 519-527
        • Maingret F.
        • Fosset M.
        • Lesage F.
        • Lazdunski M.
        • Honore E.
        J. Biol. Chem. 1999; 274: 1381-1387
        • Sheetz M.P.
        • Singer S.J.
        Proc. Natl. Acad. Sci. U. S. A. 1974; 71: 4457-4461
        • Lenz J.C.
        • Reusch H.P.
        • Albrecht N.
        • Schultz G.
        • Schaefer M.
        J. Cell Biol. 2002; 159: 291-302
        • Cannell M.B.
        • Cheng H.
        • Lederer W.J.
        Biophys. J. 1994; 67: 1942-1956
        • Strotmann R.
        • Harteneck C.
        • Nunnenmacher K.
        • Schultz G.
        • Plant T.D.
        Nat. Cell Biol. 2000; 2: 695-702
        • O'Neil R.G.
        • Heller S.
        Pfluegers Arch. Eur. J. Physiol. 2005; 451: 193-203
        • Nagata K.
        • Duggan A.
        • Kumar G.
        • Garcia-Anoveros J.
        J. Neurosci. 2005; 25: 4052-4061
        • Suchyna T.M.
        • Johnson J.H.
        • Hamer K.
        • Leykam J.F.
        • Gage D.A.
        • Clemo H.F.
        • Baumgarten C.M.
        • Sachs F.
        J. Gen. Physiol. 2000; 115: 583-598
        • Spassova M.A.
        • Hewavitharana T.
        • Xu W.
        • Soboloff J.
        • Gill D.L.
        Proc. Natl. Acad. Sci. U. S. A. 2006; 103: 16586-16591
        • Jung H.J.
        • Kim P.I.
        • Lee S.K.
        • Lee C.W.
        • Eu Y.J.
        • Lee D.G.
        • Earm Y.E.
        • Kim J.I.
        Biochem. Biophys. Res. Commun. 2006; 340: 633-638
        • Zhang P.C.
        • Keleshian A.M.
        • Sachs F.
        Nature. 2001; 413: 428-432
        • Miller P.
        • Kemp P.J.
        • Lewis A.
        • Chapman C.G.
        • Meadows H.J.
        • Peers C.
        J. Physiol. (Lond.). 2003; 548: 31-37
        • Liu-Snyder P.
        • McNally H.
        • Shi R.
        • Borgens R.B.
        J. Neurosci. Res. 2006; 84: 209-218
        • Buckler K.J.
        • Honore E.
        J. Physiol. (Lond.). 2005; 562: 213-222
        • Tseng G.N.
        Am. J. Physiol. 1992; 262: C1056-C1068
        • Chen K.H.
        • Chang B.H.
        • Younan P.
        • Shlykov S.G.
        • Sanborn B.M.
        • Chan L.
        Atherosclerosis. 2002; 165: 5-13
        • Thomas D.
        • Wu K.
        • Kathofer S.
        • Katus H.A.
        • Schoels W.
        • Kiehn J.
        • Karle C.A.
        Br. J. Pharmacol. 2003; 139: 567-574
        • Choi S.Y.
        • Kim Y.H.
        • Lee Y.K.
        • Kim K.T.
        Br. J. Pharmacol. 2001; 2: 411-418
        • Dreixler J.C.
        • Brian J.
        • Cao Y.
        • Roberts M.T.
        • Roizen J.D.
        • Houamed K.M.
        Eur. J. Pharmacol. 2000; 401: 1-7
        • Hinman A.
        • Chuang H.H.
        • Bautista D.M.
        • Julius D.
        Proc. Natl. Acad. Sci. U. S. A. 2006; 103: 19564-19568
        • Suchyna T.M.
        • Tape S.E.
        • Koeppe R.E.
        • Andersen O.S.
        • Sachs F.
        • Gottlieb P.A.
        Nature. 2004; 430: 235-240
        • Bowman C.L.
        • Gottlieb P.A.
        • Suchyna T.M.
        • Murphy Y.K.
        • Sachs F.
        Toxicon. 2006; 49: 249-270
        • Lee H.C.
        • Wang J.M.
        • Swartz K.J.
        Neuron. 2003; 40: 527-536
        • Siemens J.
        • Zhou S.
        • Piskorowski R.
        • Nikai T.
        • Lumpkin E.A.
        • Basbaum A.I.
        • King D.
        • Julius D.
        Nature. 2006; 444: 208-212
        • Lue A.J.
        • Zhao H.B.
        • Brownell W.E.
        Otolaryngol. Head Neck Surg. 2001; 125: 71-76
        • Shehata W.E.
        • Brownell W.E.
        • Dieler R.
        Acta Oto-Laryngol. 1991; 111: 707-718
        • Fang J.
        • Iwasa K.H.
        Neurosci. Lett. 2006; 404: 213-216