|
Advertisement | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
J. Biol. Chem., Vol. 279, Issue 33, 34553-34561, August 13, 2004
TRPV1 Acts as Proton Channel to Induce Acidification in Nociceptive Neurons*![]() ![]() ![]() ![]() ![]() ¶
From the
Received for publication, March 17, 2004 , and in revised form, May 28, 2004.
The low extracellular pH of inflamed or ischemic tissues enhances painful sensations by sensitizing and activating the vanilloid receptor 1 (TRPV1). We report here that activation of TRPV1 results in a marked intracellular acidification in nociceptive dorsal root ganglion neurons and in a heterologous expression system. A characterization of the underlying mechanisms revealed a Ca2+-dependent intracellular acidification operating at neutral pH and an additional as yet unrecognized direct proton conductance through the poorly selective TRPV1 pore operating in acidic extracellular media. Large organic cations permeate through the activated TRPV1 pore even in the presence of physiological concentrations of Na+, Mg2+, and Ca2+. The wide pore and the unexpectedly high proton permeability of TRPV1 point to a proton hopping permeation mechanism along the water-filled channel pore. In acidic media, the high relative proton permeability through TRPV1 defines a novel proton entry mechanism in nociceptive neurons.
During ischemia or inflammation, pain sensation is augmented by the acidic extracellular pH (pHext).1 A - and C-fiber neurons sense extracellular acid by means of two different classes of cation channels, namely TRPV1, the founding member of the vanilloid receptor-like transient receptor potential channel family (1, 2), and the acid-sensing ion channels (ASIC) (3, 4). Both channel types are expressed in small diameter dorsal root ganglion (DRG) neurons, and their role in mediating inflammatory hyperalgesia has been proven by gene deletion techniques (57). TRPV1, a poorly selective cation channel, integrates multiple pain-inducing stimuli, including noxious heat, vanilloids, and acidic extracellular pH (1, 810). Inflammatory mediators such as bradykinin, serotonin, histamine, or prostaglandins further stimulate TRPV1 activity either by protein kinase C-dependent signals (1113), by a release from a phosphatidylinositol 4,5-bisphosphate-dependent inhibition (14, 15), by formation of 12-lipoxygenase products (16), or by a protein kinase A-mediated recovery from inactivation (17). Furthermore, extracellular acidification shifts the activation of TRPV1 toward lower temperature or ligand thresholds by protonation of an amino acid located in the vicinity of the pore loop (18). Although the role of the external pHext in modulating TRPV1 activity is well established, a possible impact of TRPV1 activation on the intracellular pH has not been studied.
A Ca2+ influx component through activated voltage- or ligand-gated cation channels has been recognized to lower the intracellular pH (pHi) in neurons or in neuroendocrine cells (19, 20). We therefore asked the question whether activation of the Ca2+-permeable TRPV1 may also mediate intracellular acidification in native rat dorsal root ganglion neurons or in a heterologous expression system. Our results provide evidence for two independent TRPV1-mediated acidification signals, including an as yet unrecognized direct proton conductance through the activated TRPV1 pore.
CFP-YFP and CFP-citrine Tandem ProteinsExpression plasmids encoding intramolecularly linked cyan and yellow fluorescent proteins (CFP-YFP tandem in pcDNA3) were constructed by in-frame ligation of the open reading frames of enhanced YFP or of citrine into the EcoRI and ApaI sites of pECFP-C1 (Invitrogen). Isolation of DRG NeuronsExplanted DRG from lumbar segments of male Wistar rats were sequentially digested with collagenase (type II; 3 mg/ml) for 50 min at 37 °C, and trypsin (type I; 1 mg/ml), followed by mechanical dissociation. Cells were washed twice and resuspended in fresh medium additionally containing 10% horse serum and 50 ng/ml nerve growth factor. Neurons were seeded on poly-L-lysine-coated glass coverslips 2436 h before the experiments.
Fluorescence Spectroscopy of the CFP-YFP TandemCFP-YFP tandem protein was extracted in 20 mM Tris-HCl, pH 7.5, 20 mM Single Cell pHi ImagingLoading of Fura-2 and single cell determination of the intracellular Ca2+ concentration ([Ca2+]i) was carried out essentially as described previously (21). For calibrated single cell pHi determination applying 2',7'-bis-(2-carboxyethyl)-5-(and-6)-carboxyfluorescein (BCECF), cells were loaded with 0.50.8 µM BCECF/AM (Molecular Probes) at 37 °C for 30 min. Excitation spectra of free and protonated BCECF were determined in human embryonic kidney (HEK) 293 cells equilibrated in 128 mM KCl, 10 mM NaCl, 1 mM MgCl2, 1 mM CaCl2, 5.5 mM glucose, 0.2% (w/v) bovine serum albumin, 10 mM HEPES supplemented with nigericin and monensin (10 µM each) and adjusted to pH 9.0 or 4.0, respectively. The relative concentrations of free and protonated BCECF in the probe were calculated by multivariate linear regression analysis (22) to model the spectral properties of experimental data with stored spectra of the free and protonated indicator. Data were then converted to the pHi by using the calculation, pHi = 6.97 log ([BCECF-H]/[BCECF]). Fluorescence Resonance Energy TransferTo determine pKa values and Hill coefficients of the pH sensor proteins, the efficiency E of FRET-coupling between intramolecularly linked green fluorescent protein variants was measured by recording the fluorescence intensities of the donor during acceptor (YFP or citrine) photobleach (23). For assessment of dynamic changes in the FRET efficiency, cells were imaged at three spectral settings: 430-nm excitation combined with a 460- to 500-nm emission filter (CFP channel) and 430- or 500-nm excitation combined with a 535- to 580-nm emission filter (FRET channel or YFP channel). The "FRET channel" was corrected for direct emission of CFP (30.6% of the CFP emission detected through the 460- to 500-nm band pass) and for the acceptor cross-excitation at 430 nm (2.41% of the YFP/citrine fluorescence intensity at 500 nm excitation). Correction factors were derived from measurements in cells that expressed either CFP, YFP, or citrine alone. Under the conditions used, the emission via FRET was 1.78-fold brighter than direct CFP emission. This correction was confirmed by the facts that (i) identical FRET efficiencies of about 51% were found in resting cells with either the spectroscopic approach or the acceptor bleach, and (ii) the sum of direct donor emission and corrected FRET signals remained constant during the acceptor bleach (data not shown). Thus, the FRET efficiency E was obtained by using the calculation, E = FFRET/(FFRET + 1.78 FCFP). Confocal Emission SpectroscopyTRPV1- and CFP-YFP tandem-coexpressing cells were scanned with the 458-nm line of an argon laser, and emission was collected with a spectrally resolving array of photo-multipliers (Carl Zeiss, LSM-META). The laser intensity was adjusted to bleach the fluorochromes by less than 2% during the measurement. Emission spectra of the tandem protein were compared with stored spectra of CFP and YFP expressed alone by a regression-based linear unmix procedure to obtain images representing the actual signals emanating from both fluorochromes. These data were converted to pHi by comparing the measured ratio of CFP and YFP intensities with a lookup table representing pH-dependent changes of the fluorochrome emission. ElectrophysiologyWhole cell recordings were performed as described previously (23). Unless otherwise stated, the intracellular solution contained 110 mM cesium methane sulfonate, 25 mM CsCl, 10 mM EGTA, 3.62 mM CaCl2, 30 mM HEPES adjusted to pH 7.2 with CsOH. The standard extracellular solution contained 140 mM NaCl, 2 mM CaCl2, 1 mM MgCl2, 10 mM HEPES, pH 7.4. The NMDG solution consisted of 140 mM NMDG, 0.2 mM EDTA buffered with 10 mM HEPES, MES, or citrate and was titrated to appropriate pH with HCl. Currents in cell-attached patches were recorded using the pipette solution used for whole cell recordings and a bath solution containing 140 mM KCl and 10 mM HEPES (pH 7.4) or 10 mM MES (pH 5.5), titrated with KOH. The bath solution was chosen to depolarize the cell to a potential close to 0 mV and to reduce currents through TRPV1 in the rest of the cell. The pipette potential was set at +60 mV (membrane potential 60 mV) and ramps from 80 to +80 mV applied at 5 s intervals. Currents at the patch potential of 60 mV were measured continuously and at 80 and +80 mV from voltage ramps.
Capsaicin-induced Changes of the pHi in Rat Dorsal Root Ganglion NeuronsIn freshly isolated, fura-2-loaded DRG neurons, application of capsaicin (20 µM) led to a Ca2+ influx in a subpopulation of small to medium diameter neurons (Fig. 1A). The same class of neurons responded to capsaicin treatment with a sustained drop of the pHi from 7.12 ± 0.1 to 6.7 ± 0.08 (n = 5 independent experiments), as tested in BCECF-loaded cells. The acidification kinetic was delayed compared with the capsaicin-induced increases in [Ca2+]i (Fig. 1B). Upon removal of external Ca2+, the intracellular acidification was completely abolished, but was restored by re-addition of Ca2+ (from pHi = 7.04 ± 0.02 in the absence of Ca2+ to 6.48 ± 0.02 60 s after re-addition of 1 mM Ca2+; Fig. 1C) indicating a Ca2+-dependent indirect coupling mechanism.
In acidic extracellular media (pHext 6.0), nociceptive neurons responded to capsaicin treatment with a similar drop in pHi. Under these conditions, however, a significant part of the capsaicin-induced intracellular acidification signal persisted in the absence of extracellular Ca2+ (nominally Ca2+-free solution supplemented with 1 mM BAPTA; Fig. 1D). A subsequent readdition of 1 mM Ca2+ to the bath solution further decreased the pHi in the same cells that had already responded to capsaicin in the absence of Ca2+ (Fig. 1D). Determination of [Ca2+]i in fura-2-loaded DRG neurons confirmed that no detectable Ca2+ influx remained in the acidic solution containing 1 mM BAPTA. We conclude that TRPV1-expressing neurons exhibit a second, Ca2+-independent mechanism of intracellular acidification, which operates at acidic pHext. A capsaicin treatment of DRG neurons at the more moderately acidic pHext of 6.5 still evoked a significant intracellular acidification signal in the absence of extracellular Ca2+ (0.092 ± 0.017 pH steps within 60 s at pHext 6.5 as compared with 0.035 ± 0.037 pH steps within 60 s at pHext 7.4; both n = 5, p < 0.02). The weaker response of DRG neurons at pHext 6.5 as compared with pHext 6.0 most likely reflects the roughly 3-fold reduction in the driving force for proton entry. A possible direct proton flux through the TRPV1 pore was characterized in more detail after heterologous expression of TRPV1 in human embryonic kidney (HEK) 293 fibroblasts. Characterization and FRET-based Calibration of Genetically Encoded pH Sensor ProteinsTo combine the transfection marker and the pH-sensing system, we took advantage of the pH sensitivity of yellow fluorescent protein and constructed expression plasmids encoding intramolecularly linked CFP-YFP or CFP-citrine tandem proteins. The heterologously expressed CFP-YFP tandem was localized in the cytosol and nuclei of living HEK293 cells (Fig. 2A). As expected for intramolecular FRET, the extracted CFP-YFP tandem protein showed a dual emission peak with maxima at 476 and 535 nm when excited at 410 nm (Fig. 2B). Protonation of the extracted tandem protein resulted in a drop of YFP fluorescence and an increased emission of CFP. This indicates that the intramolecularly linked fluorochromes fold independently of each other and that FRET occurs between linked fluorochromes. Both FRET acceptors displayed almost no residual fluorescence below pH 5.5. Under the same conditions, CFP had an apparent pKa of 6.15, and its fluorescence dropped by only about 40% at pH 5.0 (data not shown). Thus, the tandem proteins represent genetically encoded pH sensors that can be calibrated on the basis of intramolecular FRET efficiencies.
To more precisely define the FRET efficiencies in living single cells, the recovery of the donor fluorochrome emission was determined during selective bleach of the acceptor (Fig. 2C). In cells that were equilibrated with nigericin and monensin (10 µM each) in a K+-based HBS, pH 8.0, the photobleach of YFP in the CFP-YFP tandem was associated with about a 2-fold increase in the CFP intensity, whereas, under acidic conditions, the donor recovery was markedly reduced (Fig. 2C). The fractional acceptor bleach and the coincident increase in the donor emission (Fig. 2, D and E) fit to a linear dependence, and regression analysis yielded the respective FRET efficiencies. Maximal FRET efficiencies within the CFP-YFP and CFP-citrine tandem proteins at pH 8.0 were 56.6% and 63.1%, respectively. FRET-based titration revealed pKa values and Hill coefficients n (given in parentheses) of 6.2 ± 0.02 (0.83) and 5.8 ± 0.03 (0.59) for the CFP-YFP and CFP-citrine tandem proteins, respectively (Fig. 2, D and E, right panels). Because the FRET efficiency scales with the ability of the acceptor fluorochrome to absorb the donor emission, the pHi can be calibrated by the formula describing the pH-dependent titration of the acceptor, pHi = pKa + n1 log (Emax/E 1), where Emax is the maximal FRET efficiency at pH >> pKa, and E is the actual FRET efficiency. To assess dynamic changes of the actual FRET efficiency, we selected a spectroscopic approach, including a correction for bleeding of the donor and acceptor fluorescences into the FRET channel and a compensation for the higher quantum yield of FRET as compared with the direct emission of CFP (see "Experimental Procedures").
Heterologously Expressed TRPV1 Mediates an Intracellular AcidificationTRPV1-expressing, fura-2-loaded HEK293 cells bathed in a Ca2+-containing solution, pH 7.4, responded to the addition of 10 µM capsaicin with an immediate and long lasting increase in [Ca2+]i from
When exposed to an acidic extracellular solution (pHext 5.5), the pHi of TRPV1-expressing HEK293 cells dropped only slightly from pHi 7.34 to 7.26 in 60 s, but subsequent addition of capsaicin induced a pronounced and sustained intracellular acidification (pHi 6.48 ± 0.03 60 s after the addition of capsaicin, six experiments; Fig. 4A). This sustained intracellular acidification remained in the absence of extracellular Ca2+ yielding a drop in pHi from 7.22 to 6.29 ± 0.05 60 s after the addition of capsaicin (five experiments; Fig. 4B). Extracellular BAPTA (2 mM in Ca2+-free solution, pH 5.5) was sufficient to prevent capsaicin-evoked Ca2+ entry as tested in parallel in fura-2-loaded cells (data not shown). Unlike recent reports on TRPV1-mediated Ca2+ mobilization from internal stores (24), no Ca2+ release from internal stores was detectable under our conditions. Thus, like in nociceptive neurons, a second mechanism of TRPV1-induced intracellular acidification operates in an acidic environment irrespective of the presence of extracellular Ca2+.
In Ca2+- and Na+-free acidic KCl solutions, protons are the only positive charge carriers that experience a strong driving force to enter the cell upon TRPV1 stimulation. In TRPV1-expressing HEK293 cells kept in isotonic KCl solution containing 2 mM BAPTA and 10 mM MES, pH 5.5, the pHi dropped from 7.08 to 5.88 ± 0.04 within the first minute after capsaicin treatment (five experiments; Fig. 5A). Almost identical values were obtained with CFP-citrine as the pHi indicator (Fig. 5B) indicating that the signals indeed reflect cytosolic acidification rather than increases in the cytosolic Cl concentration.
A permeation of protonated weak acids or bases may shuttle protons from the extracellular space to the cytosol. If protonated MES, the zwitterionic form of the extracellular buffer, permeates through TRPV1, it might mimic a proton conductance. A reduction of the MES concentration from 10 to 1 mM or a further increase to 50 mM (all buffers adjusted to pH 5.5) did not alter the capsaicin-induced acidification in TRPV1-expressing HEK293 cells. In addition, replacement of MES by 2% bovine serum albumin did not prevent the cytosolic proton accumulation. The TRPV1 pore blocker ruthenium red (10 µM), applied shortly after capsaicin, immediately abrogated the intracellular acidification (Fig. 5C). In line with the assumption that the TRPV1-mediated acidification is due to proton entry through the channel pore, expression of a dominant negative TRPV1NML676FAP mutation (25) failed to transmit a capsaicin-induced intracellular acidification (data not shown). We conclude that the TRPV1-mediated acidification at acidic pHext represents a true proton entry pathway rather than a buffer shuttle, and our data provide evidence for a direct H+/H3O+ entry through the activated TRPV1 pore. Because capsaicin is not an endogenous TRPV1 ligand, we tested the ability of N-arachidonoylethanolamine (anandamide) and N-arachidonoyldopamine (NADA) to activate acidification in TRPV1-expressing cells. In line with previous findings that anandamide is a low affinity partial agonist toward rat TRPV1 (10), 10100 µM concentrations of this endocannabinoid were required to induce a significant acidification in TRPV1-expressing HEK293 cells (Fig. 5D). The subsequent addition of capsaicin further accelerated the acidification indicating that anandamide is a partial and/or low affinity agonist. NADA (1030 µM), another endogenous TRPV1 activator (26), induced a strong proton entry signal that was only poorly accelerated by the addition of 10 µM capsaicin (Fig. 5, E and F). Thus, NADA is more potent and effective than anandamide in activating TRPV1-dependent proton entry.
To investigate whether proton entry is a common property of TRP channels, we tested TRPV1-related non-selective channels with known direct activators for direct proton permeability. Neither stimulation of TRPV4 by 4 The pH-dependent spectral shifts of the CFP-YFP tandem can be assessed by detecting either excitation spectra, emission spectra, or both. Confocal emission spectroscopy (Carl Zeiss, LSM510-META) revealed a marked drop of the acceptor peak in the emission spectrum and a coincident increase in the donor emission (Fig. 6, AD, upper panels). A regression-based spectral dissection of CFP- and YFP-derived signals and a subsequent computer-assisted comparison of the experimentally determined CFP/YFP ratios with a calibration curve of the CFP-YFP tandem protein yielded a pixel-by-pixel calibration of the pHi. Both the Ca2+-dependent acidification at neutral pH ext and the Ca2+-independent intracellular proton accumulation in acidic isotonic KCl not only lowered the cytosolic pH but also reached the nucleus without measurable delay (Fig. 6, AD, lower panels). Thus, if TRPV1 is activated in the vicinity of the cell body, intracellular acidification signals may also reach the nucleoplasm. Because locally or systemically administered capsaicin or resiniferatoxin selectively eliminate TRPV1-expressing nociceptive neurons (21, 27), further studies will have to address the question whether intracellular acidification contributes to the cytotoxicity of TRPV1-activating agents.
Electrophysiological Recording of Proton Currents through TRPV1A crude estimate of the proton entry that would be necessary to lower the cytosolic pH by about 1.5 unit revealed that a detectable amount of inward current should be carried by protons in acidic extracellular media. In whole cell patch-clamp experiments (clamped at 60 mV), capsaicin-stimulated heterologously expressed TRPV1 carried large inward currents with typical amplitudes of 515 nA, which would largely obscure an additional proton influx component. Moreover, the detection of a proton current is complicated by the proton-induced TRPV1 potentiation. Thus, a direct demonstration of proton currents requires the removal of other readily permeating cations. In neutral extracellular solutions with N-methyl-d-glucamine (NMDG+) as the only positive charge carrier besides H+/H3O+, capsaicin treatment still activated currents that reversed at 76.2 ± 1.9 mV (mean ± S.E. of 11 cells; Fig. 7, A and B). The relative permeability of NMDG+ compared with Cs+ ( ) indicates an incomplete exclusion of the large cation from the TRPV1 pore. Likewise, other organic cations such as tetraethylammonium (TEA+) carried significant inward currents at least in the absence of extracellular mono- or divalent metal ions (Fig. 7A).
Because NMDG+ had a low fractional permeability and is relatively insensitive to changes of the pH in the range of interest, we tested the effect of extracellular pH steps on capsaicin-induced currents in bath solutions with NMDG+ as the sole charge carrier besides H+/H3O+. A stepwise reduction of the pHext from 7.4 to 5.5 consistently induced reversible right-ward-shifts of the reversal potential by about 11 mV. When exposed to pHext 4.5, the reversal potentials shifted by 19 mV, and lowering the pHext from 7.4 to 3.5 induced a 41-mV positive shift of the reversal potential (Fig. 7, B and C). Positive shifts of the reversal potential of similar amplitude were observed on switching from pH 7.4 to 4.5 or 3.5 after replacing the Cs+ in the pipette solution by NMDG+ indicating that the rightward shift is caused by H+/H3O+ conductance rather than by a pH-dependent change of the . Assuming independent permeation of different cations and a of 0.05, the reversal potentials gave permeability ratios of about 1127, 245, and 87 at pHext 5.5, 4.5, and 3.5, respectively. The high relative proton permeability of TRPV1 largely contrasts with the recent observation that TRPV1 does not distinguish between monovalent metal ions and, thus, points to a distinct mechanism of permeation. Evidence for a Large Diameter of the Activated TRPV1 PoreIf proton currents through the open TRPV1 channel pore were carried by a column of free water molecules connecting the intracellular and extracellular space, we would expect a large pore diameter for the activated TRPV1. Indeed, our electrophysiological data demonstrate that TRPV1-mediated cation currents can be carried by the large organic cations TEA+ or NMDG+ (see Fig. 7, A and B). On the other hand, permeability for large cations may result from an anomalous mole fraction behavior, which allows permeation of large cations only in the absence of small metal cations. Taking advantage of the fluorescent cation 6-methoxy-N-ethylquinolinium (MEQ), a cell-impermeant Cl indicator dye, we observed a capsaicin-induced selective loading of this heterocyclic dye into TRPV1-expressing HEK293 cells, but not into untransfected control cells (Fig. 7D). MEQ is easily detectable upon entry into the cytosol, because it escapes the quench by extracellular Cl (140 mM). Of note, this permeation occurs in the presence of extracellular Na+. Consistent with a model in which MEQ directly permeates through the TRPV1 pore in a bidirectional manner, the intracellularly loaded dye rapidly leaked back into the bath solution upon removal of the extracellular dye (Fig. 7D). If the pore-blocking agent ruthenium red (10 µM) was applied prior to the removal of the extracellular dye, the loading process was stopped, and subsequent dye leakage was prevented (Fig. 7E), thus confirming that both entry and exit pathways rely on an open pore conformation of TRPV1. N-(Ethoxycarbonylmethyl)-6-methoxyquinolinium (MQAE), a more bulky derivative of MEQ, also entered into TRPV1-expressing HEK293 cells upon capsaicin treatment (Fig. 7F). The non-cationic internal salt of MQAE (ethoxycarbonyl group replaced by a negatively charged sulfoethyl group), 6-methoxy-N-(3-sulfopropyl)quinolinium (SPQ), however, was unable to enter the cells upon TRPV1 activation (Fig. 7G) indicating a maintained selectivity for cation permeation. These data indicate that large cations permeate through the TRPV1 pore even in the presence of physiological extracellular cations. Thus, our data are compatible with a model in which the activated TRPV1 pore is large enough to hold a column of water molecules connecting the intracellular and extracellular space. Moreover, although TRPV1 accepts bulky charge carriers, permeation is still restricted to cations. Effects of Changes in Intracellular pH on Currents through TRPV1To test whether changes in intracellular pH resulting from proton entry through TRPV1 can influence the activity of ion channels in the cell, we studied the effect of intracellular acidification on the activity of TRPV1 itself in HEK293 cells. For these experiments, we measured capsaicin-activated currents in cell-attached patches to protect TRPV1 channels located in the membrane patch from potentiation by acidic bath solutions and to avoid activation of ASICs. The bath solution contained 140 mM KCl to depolarize the cell. Under the conditions used, we observed spontaneous currents in the patches with levels of activity ranging from single channels to macroscopic currents of several hundred picoamperes. These currents displayed the outwardly rectifying IV relation typical of TRPV1 and had reversal potentials close to 0 mV. Currents through TRPV1 in the patch were further increased by the addition of capsaicin (1 µM) to the bath solution (Fig. 8). Subsequently, the pH of the bath solution was changed from 7.4 to 5.5 to promote intracellular acidification in the continuous presence of capsaicin. The intracellular acidification caused a reduction of the outward current component (p = 0.0018 at +80 mV) in all cells and, in some cells, a reproducible, but statistically insignificant (p = 0.135 at 80 mV), reduction in the inward current. More prominently, the slope conductance at 0 mV consistently decreased by about 2540% upon intracellular acidification. Thus, intracellular acidification causes a voltage-dependent block of TRPV1.
This study demonstrates that TRPV1 is not only modulated by the extracellular proton concentration, but it also transmits intracellular acidification in a heterologous expression system as well as in TRPV1-expressing nociceptive DRG neurons. At neutral pHext, the capsaicin-induced intracellular acidification was strictly coupled to a Ca2+ influx component. Similar Ca2+-dependent acidification responses have been observed in other neuronal cell models for Ca2+ permeation through activated NMDA receptors (20, 28) or voltage-gated Ca2+ channels (19, 29, 30). The extent of the Ca2+-dependent intracellular acidification that is conferred by the activated TRPV1, however, was more dramatic than in most recent studies on other Ca2+-permeable cation channels. This difference may reflect the significant fractional Ca2+ permeability and the strong expression of endogenous TRPV1 in DRG neurons or of heterologously expressed TRPV1 in HEK293 cells. Moreover, in its natural context, TRPV1 is activated in free nerve endings, which are characterized by a higher surface-to-volume than the cell body of the DRG neuron. Thus, the acidification in the subcellular context of nociceptive nerve endings may be even stronger than that observed in the soma of TRPV1-expressing DRG neurons. The proposed mechanisms by which increases in [Ca2+]i transmit intracellular acidification include: (i) the mitochondrial Ca2+-scavenging pathway leading to a collapse of the negative potential across the inner mitochondrial membrane and a subsequent leak of protons (31), (ii) the displacement of protons from Ca2+-buffering proteins (32, 33), or (iii) a Ca2+-H+ antiport by plasmalemmal Ca2+-ATPases (29, 34). A mitochondrial Ca2+ uptake has already been shown to underly a Ca2+ influx-dependent acidification of DRG neurons in response to membrane depolarization (35). Moreover, a capsaicin-induced and Ca2+-dependent dissipation of the mitochondrial membrane potential has been directly evidenced in DRG neurons (36), suggesting that mitochondrial H+ release may indeed account for the Ca2+-dependent component of TRPV1-induced acidification signals. In the absence of extracellular Ca2+, activated TRPV1 is no longer inactivated by high intracellular Ca2+ and exhibits an even stronger and long lasting Na+ influx (1). Under these conditions, we could not observe a significant intracellular acidification in capsaicin-stimulated TRPV1-expressing HEK293 cells or in primary DRG neurons. Thus, the remaining Na+ conductance of TRPV1 that may reduce the driving force for proton extrusion via Na+/H+ exchanger activity (37) is not responsible for the acidification. Comparing different monovalent cations, TRPV1 appears to be rather non-selective (1). This property prompted us to investigate a possible direct proton or hydronium cation conductance of TRPV1. In acidic KCl solutions, protons are the only positive charge carriers that experience a driving force to enter the cell through TRPV1. Under these conditions, capsaicin-activated TRPV1 still conferred a dramatic and sustained intracellular acidification, which could be disrupted by the pore blocker ruthenium red. An almost complete equilibration of extracellular and intracellular proton concentrations could be observed within 4 min in most TRPV1-expressing cells. We conclude that a proton current through the poorly selective TRPV1 pore is the most likely explanation for the observed effects.
Direct evidence for proton permeation through a channel is gathered by demonstration of proton currents. Voltage-gated proton-selective conductances have been described in molluscan neurons (38) as well as in various non-excitable mammalian cell types in which they help maintain the function of NADPH oxidase (39). Although these currents exhibit classic voltage-, time-, and pH-dependent gating properties, indirect evidence points to a permeation mechanism other than that expected for a classic, water-filled ion channel pore (40). Experimental determination of proton currents through a classic cation channel pore is complicated by the low micromolar availability of the charge carrier. Moreover, a residual permeability of TRPV1 for the large cations NMDG+ or TEA+, both of which are commonly used in replacement experiments, still resulted in capsaicin-inducible inward currents through the activated TRPV1 pore of several hundreds of pA at a potential of 100 mV. Nonetheless, an additional proton current component should shift the reversal potential in a pHext-dependent manner. Indeed, stepping the external pH from 7.4 to 5.5 or lower consistently and reversibly shifted the reversal potentials toward more positive potentials. Because the concentrations of other cations remained constant, the rightward shift must reflect the additional current component that is carried by protons and/or hydrogenium ions. Because similar pHext-dependent shifts of the reversal potentials were observed in symmetrical NMDG+ solutions, we exclude the possibility that extracellular protons may shift the reversal potential by further decreasing the fractional permeability of NMDG+ ( As recently described for Na+ channels, one may argue that proton conductance may result from an anomalous mole fraction behavior of TRPV1 pore taking effect only in the absence of extracellular Na+ and divalent cations. A similar TRPV1-mediated acidification signal was, however, observed in isotonic KCl or in the presence of physiological concentrations of extracellular Na+, Mg2+, and Ca2+ as long as a proton gradient across the plasma membrane was applied. Because permeation of large heterocyclic cations through the activated TRPV1 remained in the presence of physiological ion concentrations, a proton permeation mechanism along a water wire may still operate under these conditions. Because a major difference between protons and monovalent metal cations in water is their higher degree of lateral mobility, the mechanistic basis of rapid mobility of protons across water-filled channel pores is given by proton hopping between adjacent water molecules (Grotthuss mechanism) and reorientation of the hydrogen-bonded network (41). Although the general applicability of this model to proton migration along water-filled channel pores has been demonstrated for artificial gramicidin channels, structural or functional evidence for the presence of a continuous wire of water molecules within a classic ion channel pore is scarce. In line with reports on poorly selective P2X cation channels (42, 43), we here present evidence for a wide pore diameter of the activated TRPV1 by measuring inward currents carried by the large cations NMDG+ and TEA+. In addition to the electrophysiological evidence, the bidirectional and ruthenium red-sensitive permeation of the heterocyclic cationic dyes MEQ and MQAE cannot be explained by endocytosis (see Fig. 7). These findings are in agreement with the previously reported direct permeability of the cationic dye FM143 through native or heterologously expressed TRPV1 (44). Thus, our data support the model that the open TRPV1 pore holds enough water molecules to form a continuous water wire allowing proton hopping along adjacent free water molecules. If protons were to diffuse through the pore or to interact with the amino acid side chains or backbones within the pore region, their mobility would be strongly reduced resulting in a suppression of proton currents. Unlike most other effectors, intracellular acidification does not require specialized receptors to cause a plethora of cellular effects. Besides effects on cellular survival (45) or modulation of virtually all enzymatic activities, more specific effects may be exerted in the TRPV1-expressing free nerve endings, including a modulation of the conversion of the generator potential to action potentials or their further propagation along the axon. In line with this assumption, lowering the intracellular pH has been demonstrated to suppress the activity of high threshold voltage-gated Ca2+ channels in sensory DRG neurons (46, 47) as well as in other neuronal systems (48, 49). Moreover, we could demonstrate that intracellular acidification modulates TRPV1 activity by inducing a voltage-dependent block taking effect mainly at potentials higher than 30 mV. In conclusion, we present evidence for physiologically significant proton permeation through TRPV1. The high proton permeability as well as the fact that nociceptive neurons frequently become activated in the context of inflammation or ischemia, situations in which the extracellular pH may drop to values at which proton influx through TRPV1 can be experimentally demonstrated, constitute a novel role for proton permeation through TRPV1 in mediating intracellular acidosis. Future work will test specific downstream effects of the TRPV1-mediated dual acidification mechanism in nociceptive neurons.
Note Added in ProofDuring preparation of the manuscript, a proton conductance of human TRPV1 has been shown by Vulcu et al. (50) upon heterologous expression in Xenopus laevis oocytes.
* This work was supported by the Fonds der Chemischen Industrie and by the Deutsche Forschungsgemeinschaft (Grants Scha 941/1 and SFB 366). 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. ¶ To whom correspondence should be addressed. Tel.: 49-30-8445-1863; Fax: 49-30-8445-1818; E-mail: schae{at}zedat.fu-berlin.de.
1 The abbreviations used are: pHext, extracellular pH; pHi, intracellular pH; TRPV1, transient receptor potential vanilloid type 1; DRG, dorsal root ganglia; CFP and YFP, cyan and yellow fluorescent proteins; BCECF/AM,2',7'-bis-(2-carboxyethyl)-5-(and-6)-carboxyfluoresceinacetoxymethyl ester; HEK293, human embryonic kidney 293; FRET, fluorescence resonance energy transfer; [Ca2+]i, intracellular Ca2+ concentration; AEA or anandamide, N-arachidonoylethanolamine; NADA, N-arachidonoyldopamine; NMDG+, N-methyl-d-glucamine; TEA+, tetraethylammonium; MEQ, N-ethyl-6-methoxyquinolinium; MQAE, N-(ethoxycarbonylmethyl)-6-methoxyquinolinium; SPQ, 6-methoxy-N-(3-sulfopropyl)quinolinium; MES, 4-morpholineethanesulfonic acid.
We are grateful to Roger Y. Tsien, San Diego, CA who kindly supplied us with the YC-3.3-pcDNA3 plasmid containing the open reading frame of citrine. We also thank Nadine Albrecht for expert technical assistance.
This article has been cited by other articles:
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Advertisement | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||