- Email: [email protected]
Differential effects of capsaicin on rat visceral sensory neurons
Neuroscience Vol. 101, No. 3, pp. 727±736, 2000 727 q 2000 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306-4522/00 $20.00+0.00
Capsaicin and nodose neurons
Pergamon
PII: S0306-4522(00)00375-4
www.elsevier.com/locate/neuroscience
DIFFERENTIAL EFFECTS OF CAPSAICIN ON RAT VISCERAL SENSORY NEURONS K. BIELEFELDT* Department of Internal Medicine, Division of Gastroenterology, University of Iowa, 4614 JCP, 200 Hawkins Drive, Iowa City, IA 52242, USA
AbstractÐNodose neurons play an important role in the regulation of visceral function. Recent studies demonstrated that about 80% of these neurons contain messenger RNA for the capsaicin receptor, a heat-sensitive ion channel. Nodose neurons express voltage-sensitive sodium currents that can be differentiated based on their sensitivity to tetrodotoxin. Considering the potential role of tetrodotoxin-resistant sodium currents in somatosensory neurons, sodium channel expression and sodium currents were studied in nodose neurons. The results were correlated with the response to capsaicin. Nodose neurons contain messenger RNA for the tetrodotoxin-resistant sodium channel PN3. Consistent with these ®ndings, about half of the neurons predominantly expressed tetrodotoxin-resistant sodium currents. In 54% (47/87) of the cells, capsaicin triggered an increase in intracellular calcium. Similarly, in 42% (18/43) of the cells, capsaicin elicited an inward current. There was no relationship between cell size (r 0.07) or sodium current properties (r 0.14) and the response to capsaicin. Micromolar concentrations of capsaicin inhibited voltage-dependent sodium, calcium and potassium currents. This effect was use dependent and did not involve the capsaicin receptor. In conclusion, capsaicin changed the excitability of visceral sensory neurons by blocking voltage-dependent ion channels, an effect that may contribute to the analgesic properties of capsaicin. q 2000 IBRO. Published by Elsevier Science Ltd. All rights reserved. Key words: nodose neurons, sodium currents, ion channel block.
The pungent vanilloid capsaicin interacts with an ion channel, the capsaicin or VR-1 receptor, which forms a non-selective cationic channel gated by heat. It has recently been cloned from sensory neurons in dorsal root ganglia. 8 In the dorsal root and trigeminal ganglia, small neurons preferentially respond to capsaicin and its analogs. 13,18,22 Consistent with these results, Michael and Priestley 26 detected mRNA for the VR-1 receptor primarily in small cells that expressed the high-af®nity receptor for nerve growth factor (trkA), a subpopulation of afferent neurons that plays a role in sensation of pain. The importance of the VR-1 receptor in nociception has recently been shown in mice lacking this protein. 7 Prior studies demonstrated that vagal afferents also respond to capsaicin. 24 About 80% of the neurons in the nodose ganglion, the primary sensory ganglion of the vagus, contain mRNA for the VR-1 receptor independent of size or expression of neurotrophin receptors. 25 Although the majority of neurons contain mRNA for the receptor, the functional importance of the VR-1 receptor in vagal neurons remains unknown. Prior studies revealed that a subgroup of nodose neurons responded to capsaicin, indicating that the cells indeed expressed the ion channel. 24,39 However, the percentage and characterisitics of these capsaicin-responsive neurons are not known. While vagal afferents provide the main route of sensory input necessary for the regulation of visceral function, they are generally believed not to contribute to nociception. 25 Recent studies indicate that vagal ®bers may be involved in chemo-sensation 34 and perception of noxious gastric distension. 37
Nodose neurons express voltage-dependent sodium currents that can be distinguished based on their electrophysiological and pharmacological properties. 21,33 The current sensitive to the sodium channel blocker tetrodotoxin (TTX) activates and inactivates at more negative voltages than the current resistant to the toxin. In the somatosensory afferents within the dorsal root ganglia, the TTX-resistant sodium current is preferentially found in small-diameter neurons involved in nociception. 5,29,38 Using mRNA from rat dorsal root ganglia, the sequence encoding this TTXresistant sodium channel was determined and named PN3 or SNS. 1,31 In addition to PN3, another sodium channel with low sensitivity to TTX has recently been cloned from dorsal root ganglia. 14 The high-threshold for activation of the native and heterologously expressed channel correlates well with the high threshold of polymodal nociceptive afferents observed in vivo, suggesting that this channel may indeed play a key role in determining the response characteristics of nociceptors. To examine whether nodose neurons functionally express the VR-1 receptor, the effects of capsaicin were examined in cultured neurons isolated from the nodose ganglion. In addition, the relationship between sodium currents and capsaicin effects was addressed to assess whether nodose neurons preferentially express the VR-1 receptor in cells with TTXresistant sodium currents.
Experimental animals
*Tel.: 11-319-384-9841; fax: 11-319-353-6399. E-mail address: [email protected] (K. Bielefeldt). Abbreviations: EGTA, ethyleneglycolbis(aminoethyl ether)tetra-acetate; HEPES, N-2-hydroxyethylpiperazine-N 0 -2-ethanesulfonic acid; PCR, polymerase chain reaction; TTX, tetrodotoxin; VR, vanilloid receptor subtype.
Male Sprague±Dawley rats (aged two to three months) were used for the experiments. The animals were housed under a 12-h light/dark cycle with free access to water and food. Animal handling followed the guidelines of the American Physiological Society. The experimental protocol had previously been approved by the Animal Care Committee of the University of Iowa. All efforts were made to minimize the suffering and reduce the number of animals used.
EXPERIMENTAL PROCEDURES
727
728
K. Bielefeldt Table 1. Primers used for the identi®cation of sodium channel RNA present in nodose neurons (the GenBank access code is given in parentheses)
Channel PN1 PN3 NaG b-SCH-1 b-SCH-2 b-SCH-3 SNP-6
Sense primer
Anti-sense primer
Reference
CACTGTTGACAACCCTCTGC 3381±3400* CAGTGATTCTGGAGAACTTC 5181±5200 TGGGGATTGACACAGCCTC 1589±1608 AGAGACCATGTGGGACTGCA 3110±3129 AGAGACCATGTGGGACTGCA 3042±3061 AGAGACCATGTGGGACTGCA 3101±3120 ACCCTCTGAGTGAGGATGAC 5321±5340
GTCCAGCCAACACCAGGCAT 3780±3761* TCCAGACAGTGGATCTTATC 5430±5411 GAACACACTGCCATGACTA 1782±1790 CTGCCGGTCCCTATGCCACT 3520±3501 TTATTCAGATCTTCCAGCGG 3350±3331 TTTTCGGAAGCACTCCCGTA 3350±3331 GATGTCCAACTCCCCACTGT 5580±5561
Toledo-Aral 36 (U79568) Sangameswaran 31 (U53833) Chen 9 (U60590) Noda 27 (M22253) Noda 28 (X03639) Kayano 17 (Y00766) Schaller 32 (L39018)
*Numbers indicate the corresponding region in the published cDNA sequence.
Cell dissociation and culture Rats were anesthetized and decapitated, and the nodose ganglia were quickly removed under a dissection microscope. The tissue was minced with a surgical blade and incubated for 50 min in modi®ed Leibowitz L-15 medium containing collagenase (type 1A, 1 mg/ml), trypsin (type III, 1 mg/ml), and DNase (type IV, 0.1 mg/ml) at 378C. The enzymatic digestion was terminated by adding soybean trypsin inhibitor (2 mg/ml), 3 mM CaCl2 and bovine serum albumin (1 mg/ml). After gentle trituration, the tissue fragments were centrifuged at 800 r.p.m. for 5 min and then resuspended in modi®ed L-15 medium with 5% rat serum and 2% chick embryo extract (Life Technologies, NY). The cells were plated on poly-l-lysine-coated glass coverslips and incubated overnight at 378C prior to the electrophysiological studies. At that time, the nerve cell bodies were round and did not have processes, reducing the potential for space-clamp problems during electrophysiological experiments. Electrophysiological recordings The cells attached to the coverslips were transferred into a 0.5-ml recording chamber on the stage of an inverted microscope. Sodium currents were recorded using the whole-cell patch-clamp technique with an Axopatch 200A ampli®er (Axon Instruments, California) interfaced with a personal computer. The patch pipettes were pulled from thin-walled borosilicate glass (TW150-4, World Precision Instruments, Florida) with tip resistances of 1±3 MV after ®re polishing. Current recordings were ®ltered at 2 kHz with a four-pole Bessel ®lter and digitized at 10 kHz using a Digidata 1200 interface (Axon Instruments). The series resistance and whole-cell capacitance were compensated by more than 80%. The passive membrane properties were checked repeatedly during the course of the experiments. Cells were used for data analysis only if these properties remained stable. The software package pCLAMP6.0 (Axon Instruments) was used for data acquisition and analysis. The leak current and residual capacitative transients were digitally subtracted using the p/n protocol for leak subtraction with n 4. In current-clamp experiments, the cells were initially kept at their resting membrane potential of 258.7 ^ 1.8 mV. The threshold for action potential generation was determined with depolarizing current injection of increasing amplitude. The cells were then stimulated with a current injection of 1.5 times the threshold value lasting 1.8 s. All experiments were performed at room temperature (218C). Determination of the intracellular calcium concentration Cultured nodose neurons were plated on 2.5-cm poly-l-lysinecoated coverslips. The cells were washed with normal extracellular solution, and incubated for 45±60 min at 378C with the membranepermeable ester of the calcium-sensitive ¯uorescent indicator fura-2, fura-2-AM, at a concentration of 3 mM and 0.02% pluronic. 3 After removal of the loading solution, the cells were kept in normal extracellular medium for at least 30 min to allow de-esteri®cation of the acetoxymethyl ester. The cells were then mounted in a recording chamber on an inverted ¯uorescence microscope. A 75-W xenon lamp served as light source. The light passed through interference
®lters mounted on a computer-controlled shutter allowing subsequent imaging at excitation wavelengths of 340 and 380 nm. The emitted light passed through a ®lter set to a wavelength of 510 nm and was measured with a CCD camera (Photon Technology International). All images were automatically corrected for background ¯uorescence. The camera data were fed to an image analysing system, digitized and stored on-line on a personal computer (Photon Technology International). Regions of interest were selected visually for the data analysis. Typically, the loading was homogeneous and did not allow identifying subcellular compartments. In some cells, the indicator appeared to be compartmentalized in areas with higher calcium concentrations, as indicated by a ¯uorescence ratio that exceeded the remainder of the cytoplasm by a factor of two or more. These regions were excluded from further analysis. The images obtained at 340 and 380 nm were divided pixel by pixel to generate the ratio values for the selected regions of interest. As the spatial resolution of the CCD camera is poor, the images represent averages of the free calcium concentration in the cytosol and compartmentalized calcium in small organelles. Therefore, the data were not converted into calcium concentration but expressed as ¯uorescence ratio. Reverse transcription±polymerase chain reaction mRNA was extracted from rat brain and nodose ganglia following a protocol described previously. 10 After reverse transcription with Moloney-monkey leukemia virus, reverse transcriptase ®rst-strand cDNA templates were ampli®ed in 30 cycles using 2 U of Taq DNA polymerase (Gibco BRL) in 50 ml of 100 mM Tris buffer containing 50 mM KCl and 1.4±1.9 mM MgCl2. Each cycle consisted of denaturing at 948C for 1 min, followed by annealing at 56±648C for 1 min and extension at 728C for 1.5 min. The polymerase chain reaction (PCR) products were fractionated on 1% agarose containing ethidium bromide. Isoform-speci®c primers were used to further de®ne the sodium channel subunits expressed (Table 1). The PCR products were puri®ed and sequenced using the Applied Biosystems 373 A Stretch Sequencer at the DNA Core Facility of the University of Iowa. The obtained sequences were then analysed for similarity with known sequences by searching the NCBI data bank at the National Institutes of Health. Data analysis All data are expressed as mean ^ S.E.M. Exponential ®ts were obtained using the curve-®tting routine of the Origin software package (Microcal, California). The t-test was used to discern signi®cant difference after intervention. Statistical signi®cance was determined at P , 0.05. Solutions and chemicals To measure sodium or calcium currents, the pipette solution contained (in mM) 100 cesium aspartate, 20 CsCl, 2.3 CaCl2, 4.8 MgCl2, 10 EGTA, 10 HEPES, 4 Mg-ATP and 0.5 Na-GTP, buffered to pH 7.2 with CsOH. For potassium currents and current-clamp experiments, pipettes were ®lled with a solution containing (in mM)
729
Capsaicin and nodose neurons
Fig. 1. Voltage-dependent sodium currents and action potentials in nodose neurons. Depolarizations from 270 mV to various test potentials between 260 and 40 mV trigger rapidly activating and inactivating inward currents that differ in their voltage dependence and kinetics. Superimposed sample traces are shown for two representative neurons with fast (A) and slow (B) kinetics (calibration bars: 2 nA and 5 ms, respectively). Current-clamp experiments demonstrate two distinct response characteristics to long depolarizating current injections. Representative voltage traces obtained after stimulation of neurons with a depolarizing current injection at 50% above threshold from 260 mV show cells responding either with a single spike (C) or a burst (D) of action potentials (calibration bars: 40 mV and 500 ms, respectively). The steady-state inactivation was tested with 200-ms prepulses to potentials between 2100 and 20 mV, followed by a step to 20 mV (E). The results were ®tted with a Boltzmann equation (dotted lines) for rapidly activating neurons (circle; n 17) and slowly activating neurons (squares; n 17). The V1/2 was 240.8 mV with a slope factor of 9.3 for group 1 neurons and 212.5 mV with a slope factor of 5.1 for group 2 neurons. The histogram (F) demonstrates the existence of two overlapping populations of neurons based on their voltage of half-inactivation (n 198). A histogram describing the effects of the sodium channel blocker TTX in 105 neurons supports the existence of two groups of neurons (G).
145 potassium aspartate, 4 NaCl, 10 EGTA, 1 MgCl2, 2 ATP, 0.5 GTP, 10 HEPES and 2.3 CaCl2, adjusted with KOH to a pH of 7.2. The extracellular solution for the measurement of sodium currents was composed of 20 mM NaCl, 120 nM choline chloride, 10 mM tetraethylammonium chloride, 3 mM MgCl2, 10 mM HEPES and 5.5 mM glucose buffered at pH 7.3 with tetraethylammonium hydroxide. To record calcium currents, the extracellular solution was composed of (in mM) 5 CaCl2, 140 choline chloride, 10 HEPES and 5.5 glucose, buffered at pH 7.3 with tetraethylammonium hydroxide. For currentclamp recordings and calcium imaging, the extracellular solution contained (in mM) 145 NaCl, 4.5 KCl, 1 MgCl2, 2 CaCl2, 5 glucose, buffered with HEPES and NaOH to pH 7.4. For the recording of potassium currents, the extracellular solution contained (in mM) 140 choline chloride, 4.5 KCl, 3 MgCl2, 5 glucose, buffered with HEPES and KOH to pH 7.4. Capsaicin was dissolved in ethanol at a concentration of 1 mM and was diluted immediately before use. The ®nal concentration of ethanol in the bathing solution never exceeded 0.01%. In control experiments, this concentration did not lead to any changes in membrane properties or intracellular calcium concentration. In some experiments, capsaicin was applied by a pressure application. A glass pipette with a tip resistance of approximately 1 MV was placed at approximately 100 mm from the cell using a micromanipulator. The drug was applied with a 100-ms pressure pulse of 10 psi using a pneumatic PicoPump (World Precision Instruments, Sarasota, FL). Fura-2-AM and pluronic were purchased from Molecular Probes
(Eugene, OR). All other chemicals were analysis grade and were obtained from Sigma Chemical (St Louis, MO). RESULTS
Sodium currents in nodose neurons Consistent with previous observations, 21,33 nodose neurons expressed voltage-dependent sodium currents that could be differentiated based on their physiological and pharmacological properties. Brief depolarizations from 270 mV to test potentials positive to 250 mV triggered rapidly activating and inactivating inward currents (Fig. 1A, B). Similarly, two distinct patterns were observed when cells were examined under current-clamp conditions. After determining the threshold for action potential generation, the neurons were stimulated using 1.5 times the current amplitude required to generate a single action potential. At a resting membrane potential of 260 mV, such current injections triggered action potentials at a frequency of 2.8 ^ 0.8 Hz (n 35). Interestingly, again two groups emerged. Nineteen neurons
730
K. Bielefeldt
Fig. 2. Molecular identi®cation of sodium channels expressed in nodose ganglia. Primers for sodium channels expressed primarily in central (A) or peripheral (B) neurons were used to amplify cDNA obtained from rat brain (I) and rat nodose ganglia (II). The ®rst lane shows molecular markers with a 123-bp ladder.
responded with only a single action potential at the beginning of the stimulus (Fig. 1C). In contrast, 16 neurons generated several action potentials at a frequency between two and 12.5 Hz (Fig. 1D). The mean amplitude of the action potential differed slightly between these two groups: 119.6 ^ 2.7 mV in cells with repeated action potentials vs 112.1 ^ 2.3 mV in the remaining neurons (P , 0.04). The duration of the action potential measured at the half-maximal voltage was slightly but not signi®cantly prolonged in the latter group of cells (4.9 ^ 0.6 vs 8.5 ^ 2.3 ms; P 0.1). The sodium current density was signi®cantly higher in cells with persistent electrical activity during prolonged depolarization compared to quiescent cells (242.8 ^ 25.9 vs 161.2 ^ 18.5 pA/pF; P , 0.01). We examined the effect of 2 mM TTX on the sodium current in 25 cells that were also studied under current-clamp conditions. The sodium channel blocker inhibited 65.9 ^ 10.7% of the peak inward current in cells that generated only a single action potential (n 13) compared to 28.2 ^ 10.2% in neurons that responded with two or more action potentials during prolonged depolarization (P , 0.03; n 12). When examined under voltage-clamp conditions, cells differed signi®cantly in the steady-state inactivation (Fig. 1E). A histogram was constructed based on the half-inactivation (V1/2) obtained by ®ts with the Boltzmann equation, which demonstrated two overlapping populations of neurons (Fig. 1F). Currents with rapid activation and inactivation kinetics had a V1/2 of around 245 mV, while currents with slower activation and inactivation had a V1/2 that was about 20 mV more positive. Prior studies have demonstrated differences in the sensitivity of nodose neurons to the neurotoxin TTX. Consistent with these ®ndings, 2 mM TTX inhibited sodium currents to less than 20% of control in about 50% of the cells. However, in the remaining cells, this concentration of TTX led only to a partial or no inhibition of the sodium inward current (Fig. 1G). The pharmacological characteristics correlated with the electrophysiological characteristics as suggested by the correlation coef®cient of r 20.87 (P , 0.01) between the voltage of half-inactivation and the fraction of residual current after administration of TTX. Previously published studies in neurons from the dorsal
root ganglion indicated that the TTX-resistant sodium channel is preferentially expressed in small neurons with unmyelinated ®bers. 4,5,29 To test whether such a relationship exists in nodose neurons the cell capacitance, a measure of the membrane surface area and thus the size of a cell, was correlated with the inhibitory effect of 2 mM TTX. The correlation of coef®cient of 20.37 (P , 0.01; n 67) shows the opposite result in nodose neurons with large cells predominantly expressing TTX-resistant sodium currents. Sodium channel isoforms present in nodose neurons The electrophysiological and pharmacological data suggest that at least two different sodium channels are expressed in nodose neurons. To determine which of the known sodium channels are present in the cells, a PCR-based approach was chosen. Speci®c primers were designed for neuronal sodium channels (Table 1). As expected, mRNA for these channels can be found in the brain, which served as a positive control. In contrast, only primers for b-SCH-I and for PN3 yielded a positive reaction with cDNA from nodose neurons (Fig. 2). Interestingly, primers for NaG, a TTX-resistant sodium channel that was detected in neurons from the canine nodose ganglion, 9 did not yield a positive reaction product even when tested under low stringency conditions (data not shown). The PCR products were puri®ed and sequenced. A computerbased search of the NCBI database at the National Institutes of Health demonstrated a complete identity between the sequenced PCR products and the published primary structure for the ampli®ed region of b-SCH-I and PN3 (data not shown). Effects of capsaicin on the intracellular calcium concentration To determine whether nodose neurons respond to capsaicin, the cells were loaded with fura-2 to measure changes in the intracellular calcium concentration. After obtaining a stable baseline, capsaicin or vehicle solution were added to the extracellular solution. Vehicle did not cause signi®cant changes in the intracellular calcium concentration (decrease
Capsaicin and nodose neurons
731
Capsaicin response and sodium current expression When cells were studied under voltage-clamp conditions, capsaicin triggered an inward current in 18 of 43 cells (42%) examined (Fig. 3B). Repeated administration within 10 s demonstrated a signi®cant decrease in the peak amplitude, suggesting a desensitization of the ion channel (53 ^ 12% of control; P , 0.01; n 8). A similar, though slower and persisting increase was seen in 12 of 23 cells when capsaicin was applied through exchange of the bathing solution (Fig. 3C). At a concentration of 10 mM, the capsazepine inhibited the response to 1 mM capsaicin to 20 ^ 11% of control (P , 0.05; n 3). Prior studies on dorsal root and trigeminal ganglia revealed that small-diameter neurons primarily respond to capsaicin. 12,13,18,22,23 In contrast, the cell size measured by the hole cell capacitance did not correlate with the capsaicin response in nodose neurons (r 0.07; P 0.65; n 45). To determine whether the capsaicin response differs between cells expressing TTX-sensitive or TTX-resistant sodium current, cells were depolarized from 260 to 0 mV before and after the administration of 2 mM TTX. The sodium current inhibition by TTX did not correlate with the peak inward current triggered by capsaicin (r 0.14; P 0.54; n 22). Thus, neither cell size nor the expression of sodium currents allowed prediction of the response to capsaicin. Capsaicin and voltage-dependent ion currents
Fig. 3. Effect of capsaicin on nodose neurons. (A) Sample traces obtained after administration of 500 nM capsaicin or vehicle to extracellular solution (arrow). (B) A representative current trace of a nodose neuron held at 260 mV. The horizontal bar indicates the pressure application of 1 mM capsaicin (100 ms; 10 psi) using a second pipette positioned at a 100-mm distance from the cell (calibration bars: 200 pA; 0.5 s). (C) Capsaicin was applied by rapid exchange of the perfusion solution (calibration bars: 200 pA; 2 s).
by 2 ^ 1%; n 12). However, 47 of the 87 cells examined (54%) responded with a rise in intracellular calcium (Fig. 3A). At a concentration of 1 mM, capsaicin caused an increase in the intracellular calcium level by 54.0 ^ 16.0% (n 5). To assess whether unresponsive cells were viable, potassium depolarizations were performed. Interestingly, prior exposure to high concentration of capsaicin blunted the response to subsequent depolarization with high potassium solution in these cells, suggesting a change in excitability. While the calcium level rose by 34.8 ^ 5.1% in control conditions, it only increased by 18.4 ^ 6.2% in cells pretreated with 10 mM capsaicin (P , 0.05; n 4).
As described above, high concentrations of capsaicin blunted the rise in intracellular calcium triggered by potassium depolarization. To determine the potential mechanism of this effect and its relationship to the capsaicin receptor, the effects of capsaicin on sodium, calcium and potassium currents were studied. Capsaicin administration to the extracellular solution concentration dependently and reversibly decreased the peak sodium current (Fig. 4). At 10 mM, the peak sodium current decreased by 22.7 ^ 4.3% (P , 0.01; n 31). This effect was completely reversible upon washout (peak current 97 ^ 3% of control; not signi®cant). The kinetics of activation and inactivation were not signi®cantly altered (data not shown). In four cells with more than 90% TTX-resistant sodium current, capsaicin inhibited the peak sodium current by 31.9 ^ 20%; similarly, in six neurons with more than 90% TTX-sensitive current, capsaicin blocked 31.2 ^ 13% of the peak sodium current. The correlation coef®cient of 0.06 between the sodium current inhibition by capsaicin and the effect of 2 mM TTX further supports that TTX-sensitive and TTX-resistant current were equally affected (P 0.74; n 31). Capsaicin blocked sodium currents even in cells that did not respond to capsaicin with an inward current (Fig. 5). Similarly, the capsazepine (25 mM) signi®cantly decreased voltage-dependent sodium currents (22.9 ^ 1.9%; n 3). Considering the previously described lack of a correlation between the capsaicin-evoked current and the inhibitory effect of capsaicin on sodium current, it is unlikely that capsaicin alters sodium currents through the capsaicin receptor. This is further supported by the results obtained with a high concentration of capsazepine, which also inhibits sodium current, thereby arguing against a speci®c effect of the vanilloid that requires activation of the VR-1 receptor. To examine whether the inhibitory effect of capsaicin was
732
K. Bielefeldt
Fig. 5. Capsaicin response and inhibitory effect of vanilloids. The peak sodium current in the presence of 10 mM capsaicin was plotted as a function of the peak inward current triggered by capsaicin, demonstrating no relationship between the response of the capsaicin receptor and the inhibitory effect of capsaicin on voltage-dependent sodium currents.
Fig. 4. Effect of capsaicin on sodium currents. Sodium currents were triggered by depolarization from 260 to 0 mV. (A) The time-course of current inhibition and recovery is shown for a single cell. A test pulse was applied every 15 s (points). The horizontal bar indicates the addition of 10 mM capsaicin. The insert shows superimposed sample traces obtained before and after the addition of 10 mM capsaicin (calibration bars: 2 nA; 2 ms). (B) The inhibitory effect of capsaicin plotted as a function of the concentration and ®tted with the Hill equation, revealing an IC50 of 40 mM with a Hill coef®cient of 1.0.
restricted to voltage-dependent sodium currents, calcium currents were recorded before and after the administration of capsaicin to the extracellular solution. At a concentration of 10 mM capsaicin, the peak calcium current decreased by 31.9 ^ 6.6% (P , 0.01; n 6). This effect was dose dependent and reversible, as the peak current recovered to 95 ^ 1% of control after washout (Fig. 6). To isolate low-threshold calcium currents, the response to a depolarization from 240 to 10 mV was digitally subtracted from that triggered by a step from 290 to 10 mV. The inhibitory effect of capsaicin on low-threshold calcium current did not differ from that on high-threshold currents (30.0 ^ 8.8%; n 4). An analysis of the current decay revealed that capsaicin signi®cantly accelerated the current inactivation (time constants of 82.1 ^ 20.7 vs 40.1 ^ 9 s for control and capsaicin, respectively; P , 0.05; n 9). While pretreatment with 25 mM capsazepine only decreased calcium currents by 4 ^ 2% (not signi®cant), the antagonist did not block the inhibitory effect of 10 mM capsaicin given subsequently (40.7 ^ 15.8% decrease in peak current; n 4). The effects of capsaicin on sodium and calcium currents indicate that the vanilloid may non-speci®cally block voltagedependent ion channels. Therefore, the effect of capsaicin was
studied on potassium currents. Capsaicin (10 mM) decreased potassium currents by 42.1 ^ 7.5% (P , 0.01; n 8). This inhibition was fully reversible upon washout, as the current amplitude recovered to 100 ^ 3% compared to control. The effect was dose dependent, as shown in Fig. 7. To examine whether capsaicin preferentially inhibits sustained or transient outward currents, the A current was isolated by digital subtraction of the current response to a depolarization from 240 to 40 mV from that triggered by a step from 290 to 40 mV. Only four of the 13 cells tested expressed a signi®cant fraction of transient outward current. Capsaicin inhibited both the transient and sustained outward currents in these cells by 29 ^ 8% and 41 ^ 9%, respectively (n 4). Similar to the effects on calcium currents, capsaicin signi®cantly accelerated the current decay. Under control conditions, the currents did not appreciably inactivate within 100 ms after a depolarization from 260 to 40 mV. After the addition of capsaicin, the current inactivated with a time constant of 8.1 ^ 1.8 ms (P , 0.01; n 5). Capsazepine (25 mM) decreased the peak current by 48.1 ^ 14.7% (n 3). The accelerated decay of potassium and calcium currents seen in the presence of capsaicin is suggestive of a usedependent block. Moreover, the time-course of activation for calcium currents slowed in the presence of capsaicin (10.9 ^ 1.2 vs 16.7 ^ 3.1 ms for control and capsaicin), as demonstrated in Fig. 8A. To further examine whether capsaicin indeed preferentially inhibits open channels, the peak current was measured in cells that were repeatedly depolarized. When the voltage was stepped from 260 to 0 mV for 50 ms at a frequency of 0.5 Hz, cumulative inhibition caused a decrease in the peak calcium current by 16 ^ 6% within 1 min (Fig. 8B). In the presence of 20 mM capsaicin, the peak current declined by 35 ^ 8% (P , 0.05; n 6). Using a similar paradigm for potassium currents, no signi®cant inhibition was observed under control conditions (Fig. 8C). In contrast, capsaicin caused a progressive inhibition with a decrease in the peak potassium current by 14 ^ 4% within 1 min (P , 0.05; n 9). To examine the effect of capsaicin on the cumulative inhibition of sodium currents, cells were depolarized from 270 to 0 mV for 5 ms at a frequency of
Capsaicin and nodose neurons
Fig. 6. Effect of capsaicin on calcium currents. (A) The time-course of current inhibition and recovery is shown for a single cell. A test pulse was applied every 15 s (points). Calcium currents were triggered by depolarization from 260 to 10 mV. The horizontal bar indicates the addition of 10 mM capsaicin. The insert shows superimposed sample traces obtained before and after the addition of 10 mM capsaicin (calibration bars: 2 nA; 20 ms). (B) The inhibitory effect of capsaicin plotted as a function of the concentration and ®tted with the Hill equation, revealing an IC50 of 8.6 mM with a Hill coef®cient of 0.7.
10 Hz. The higher stimulation frequency was chosen as the sodium current essentially completely recovered from inactivation within 100 ms. In control cells, there was a cumulative inhibition with a peak current decrease of 15 ^ 2% within 10 s (Fig. 8D). This inhibition was more pronounced in the presence of capsaicin (P , 0.05; n 8). DISCUSSION
The key ®ndings of this study are: (i) two populations of nodose neurons can be distinguished based on the pharmacological and biophysical characteristics of their sodium currents; (ii) about half of the nodose neurons respond to capsaicin; (iii) the response to capsaicin does not correlate with cell size or the expression of a TTX-resistant sodium current; (iv) high concentrations of capsaicin and its competitive antagonist capsazepine inhibit voltage-sensitive ion channels. The data presented con®rm previous observations that nodose neurons express voltage-dependent sodium currents that can be differentiated based on their sensitivity to the neurotoxin TTX. The current-clamp experiments support the potential physiological importance of this ®nding,
733
Fig. 7. Effect of capsaicin on potassium currents. (A) The time-course of current inhibition and recovery is shown for a single cell. A test pulse was applied every 15 s (points). Potassium currents were triggered by depolarization from 260 to 40 mV. The horizontal bar indicates the addition of 10 mM capsaicin. The insert shows superimposed sample traces obtained before and after the addition of 10 mM capsaicin (calibration bars: 1 nA; 20 ms). (B) The inhibitory effect of capsaicin plotted as a function of the concentration and ®tted with the Hill equation, revealing an IC50 of 20 mM with a Hill coef®cient of 0.83.
demonstrating that cells with a high current density and mostly TTX-resistant sodium current were able to generate action potentials during prolonged stimulations. Similar TTX-sensitive and TTX-resistant sodium currents have been described in somatosensory and colon sensory neurons. 4,5,11,15,35 However, a more detailed evaluation of these results reveals several differences if one compares somatosensory neurons from the dorsal root with the visceral sensory neurons from the nodose ganglion. Both TTX-sensitive and TTX-resistant sodium currents in nodose neurons inactivate at potentials that are shifted by about 20 mV to the right compared to results obtained from somatosensory neurons. 11,15 Furthermore, large cells were more likely to express TTX-resistant current than small cells, a ®nding that is contrary to results described in neurons from dorsal root ganglia. 5,29 Finally, mRNA for only two sodium channel a-subunits was found in nodose neurons, while at least seven sodium channel isoforms were detected in dorsal root ganglia when a similar approach was used. 4 The recently cloned PN3 channel subunit encodes a TTXresistant sodium channel that is preferentially expressed in nociceptive neurons. 27 Similarly, the capsaicin response or expression of the VR-1 receptor has also been used as a marker of nociceptive neurons. 7,8 Interestingly, only about 50% of the nodose neurons responded to capsaicin, while
734
K. Bielefeldt
Fig. 8. Use-dependent inhibition of voltage-dependent currents by capsaicin. (A) Changes in the time-course of activation. Calcium current traces were normalized and superimposed for control conditions (solid line) and after the addition of 20 mM capsaicin (broken line). (B) The effect of capsaicin on peak calcium currents triggered by repeated depolarization from 260 to 0 mV for 50 ms at 0.5 Hz (®lled symbols: control; open symbols: capsaicin). (C) The effect of capsaicin on peak potassium currents triggered by repeated depolarization from 260 to 40 mV for 50 ms at 0.5 Hz (®lled symbols: control; open symbols: capsaicin). (D) The effect of capsaicin on peak sodium currents triggered by repeated depolarization from 270 to 10 mV for 5 ms at 10 Hz (®lled symbols: control; open symbols: capsaicin).
mRNA for the VR-1 receptor can be detected in 80% of the neurons. 26 Marsh et al. 24 focused on C-type nodose neurons and demonstrated that about 60±70% of these cells exhibited electrical responses to capsaicin. Even more neurons demonstrated ultrastructural changes after prolonged exposure to capsaicin at concentrations between 1 and 10 mM. 24 Considering the non-speci®c effects of capsaicin described in this paper, it remains unclear whether these effects were solely mediated by activation of the capsaicin receptor. Similarly, Winter 39 observed responses of nodose neurons to 10 mM capsaicin. However, she did not provide information about the percentage or characteristics of neurons that were affected by capsaicin. Consistent with these results, Benson et al. 2 demonstrated that about 60% of the cardiac sensory neurons in the nodose ganglion respond to capsaicin. As neurons from mice lacking the VR-1 receptor do not respond to capsaicin, 6 it appears reasonable to use the physiological results obtained with capsaicin as a surrogate for the presence of the VR-1 receptor. Therefore, the data presented in this article suggest that only slightly more than half of the neurons functionally express the VR-1 receptor. This contrasts with the dorsal root ganglia, where about 50% of the neurons contain mRNA for the VR-1 receptor, 26 which corresponds very well with physiological data demonstrating a response to capsaicin in about 50±70% of the cells examined. 12,13,18,23 Neither cell size nor the sensitivity to TTX allowed
prediction of the response to the vanilloid capsaicin in nodose neurons. Similarly, Marsh et al. 24 did not see a correlation between the capsaicin effect and the electrophysiological properties of nodose cells. These results differ from ®ndings in somatosensory neurons. Using in situ hybridization, mRNA for the VR-1 receptor was found primarily in small and medium-sized cells. 26 Moreover, small cells were also more likely to respond to capsaicin when compared with large neurons from the dorsal root ganglia. 12,13,18,23 Su et al. 35 recently reported no relationship between cell size and response to capsaicin in visceral sensory neurons from dorsal root ganglia. In contrast to the ®ndings in nodose neurons, these investigators could not elicit capsaicin responses in neurons that only expressed TTX-sensitive sodium current. These ®ndings demonstrate that the properties of visceral sensory neurons differ from those of somatosensory neurons. However, additional differences may exist between the visceral sensory neurons in the placode-derived nodose ganglion 30 compared to the colonic sensory neurons in the dorsal root that originate in the neural crest. 6 These data raise the question whether, at least in the context of visceral neurons, cell size, sodium channel expression or capsaicin sensitivity can reliably be used as markers of nociception. Previously published studies have demonstrated that the native and heterologously expressed VR-1 receptor has a Kd for capsaicin between 0.5 and 1 mM. 8,12,13,18,23,35 Consistent
735
Capsaicin and nodose neurons
with these ®ndings, 1 mM capsaicin triggered calcium in¯ux and inward currents in about 50% of nodose neurons. Interestingly, higher concentrations decreased the cellular excitability in cells otherwise not affected by the vanilloid, as judged by a blunting of the calcium increase caused by potassium depolarization. As capsaicin activates a non-speci®c cation channel, the resulting prolonged depolarization in the ongoing presence of capsaicin could cause the inactivation of sodium and calcium channels. To circumvent this possibility, the electrophysiological experiments described were performed under voltage-clamp conditions, thus keeping the membrane potential stable and eliminating voltagedependent inactivation processes. Under those conditions, capsaicin directly inhibited sodium, calcium and potassium currents. Several ®ndings suggest that this effect is independent of the VR-1 receptor. First, the IC50 for the inhibitory effect of capsaicin was one order of magnitude higher than the known Kd for the VR-1 receptor. Second, while the response to capsaicin desensitizes in the presence of extracellular calcium, 8 the inhibition of voltage-dependent ion current remained stable for several minutes, even with an extracellular calcium concentration of 5 mM. Conversely, capsaicin blocked sodium currents in the absence of extracellular calcium, thereby arguing against an indirect effect of calcium in¯ux through the VR-1 receptor, leading to an activation of intracellular signaling pathways, such as the activation of calcium-dependent protein kinases. Third, capsazepine also inhibited voltage-dependent ion channels and did not block the effect of capsaicin. Finally, there was no correlation between the capsaicin-evoked inward current and the inhibitory effect of capsaicin in the cells examined. Su et al. 35 described a similar inhibitory effect of capsaicin on colonic sensory neurons. However, the inhibition of sodium currents did not reverse after removal of capsaicin and was blunted by the competitive antagonist capsazepine. Moreover, the sodium current in cells that did not respond to capsaicin was not affected by the vanilloid. Based on these results, the authors hypothesized that capsaicin-induced changes in the intracellular calcium concentration may secondarily alter the channel properties. As summarized above, such a mechanism is unlikely for the capsaicininduced inhibition of voltage-dependent currents in nodose neurons. As mentioned previously, calcium was omitted
from the extracellular solution when sodium currents were measured. In the presence of capsaicin, calcium and potassium currents inactivated much more rapidly than under control conditions. Such a result is typical for an open channel or use-dependent block, where the ion channel has to open before the interaction between the inhibitor and the channel pore can take place. This is further supported by experiments demonstrating a progressive decrease in peak calcium, sodium and potassium current with repeated stimulations. Previous studies in vivo have demonstrated that capsaicin augments the anesthetic effect of locally administered TTX. 19 Capsaicin also reduced the extracellularly recorded compound action potential in human sural nerve preparations. In contrast to the observation in nodose neurons, capsaicin primarily reduced the TTX-resistant component of the compound action potential, suggesting a more selective effect on the TTX-resistant sodium current. 16 Kuenzi and Dale 20 observed that vanilloids dose-dependently blocked potassium and calcium currents in Xenopus spinal neurons. In Xenopus cells, micromolar concentrations of capsaicin inhibited voltage-dependent currents in all cells tested, while only about one-®fth of the neurons showed the typical inward current triggered by capsaicin. The inhibitory effect of capsaicin exhibited the characteristics of an open channel block, consistent with the ®ndings in nodose neurons. CONCLUSION
Despite the presence of mRNA for the VR-1 receptor in about 80% of the nodose neurons, only about half of the cells functionally express the VR-1 receptor as de®ned by a response to the pungent vanilloid capsaicin. While low concentrations activate an inward current in these neurons, higher concentrations non-speci®cally inhibit voltage-gated ion currents through an open channel block. This effect is independent of the known vanilloid receptor. The inhibition of sodium and calcium currents will decrease neuronal excitability and may contribute to the analgesic properties of capsaicin. AcknowledgementsÐThis study was supported by grants from the National Institutes of Health (DK01548-01).
REFERENCES
1. Akopian A. N., Silvilotti L. and Wood J. N. (1996) A tetrodotoxin-resistant voltage-gated sodium channel expressed by sensory neurons. Nature 379, 257±262. 2. Benson C. J., Eckert S. P. and McCleskey W. E. (1999) Acid-evoked currents in cardiac sensory neurons: a possible mediator of myocardial ischemic sensation. Circulation Res. 84, 921±928. 3. Bielefeldt K., Martin M., Whiteis C., Yedidag E. and Abboud F. M. (1997) Modulation of calcium release channels in intestinal smooth muscle cells. Cell Calcium 22, 507±514. 4. Black J. A., Dib-Hajj S., McNabola K., Jeste S., Rizzo M. A., Kocsis J. D. and Waxman S. G. (1996) Spinal sensory neurons express multiple sodium channel a-subunits. Molec. Brain Res. 43, 117±131. 5. Caffrey J. M., Eng D. L., Black S. G., Waxman S. G. and Kocsis J. D. (1992) Three types of sodium channels in adult rat dorsal root ganglion neurons. Brain Res. 592, 283±297. 6. Carey M. B. and Matsumoto S. G. (1999) Neurons differentiating from murine neural crest in culture exhibit sensory or sympathetic-like calcium currents. J. Neurobiol. 39, 501±514. 7. Caterina M. J., Lef¯er A., Malmberg A. B., Martin W. J., Trafton J., Peterson-Zeitz K. R., Klotzenburg M., Basbaum A. I. and Julius D. (2000) Impaired nociception and pain in mice lacking the capsaicin receptor. Science 288, 306±313. 8. Caterina M. J., Schumacher M. A., Tominaga M., Rosen T. A., Levine J. D. and Julius D. (1997) The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature 389, 816±824. 9. Chen J., Ikeda S. R., Lang W., Isales C. M. and Wei X. (1997) Molecular cloning of a putative tetrodotoxin-resistant sodium channel from dog nodose ganglion neurons. Gene 202, 7±14. 10. Chomczynski P. and Sacchi N. (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate±phenol±chloroform extraction. Analyt. Biochem. 162, 156±159.
736
K. Bielefeldt
11. Cummins T. R. and Waxman S. G. (1997) Downregulation of tetrodotoxin-resistant sodium currents and upregulation of a rapidly repriming tetrodotoxin-sensitive sodium current in small spinal sensory neurons after nerve injury. J. Neurosci. 17, 3503±3514. 12. Dedov V. N. and Roufogalis B. D. (1998) Rat dorsal root ganglion neurones express different capsaicin-evoked Ca 21 transients and permeabilities to Mn 21. Neurosci. Lett. 248, 151±154. 13. Del Mar L. P. and Cardenas C. G. (1996) Capsaicin preferentially affects small-diameter acutely isolated rat dorsal root ganglion cell bodies. Expl Brain Res. 111, 30±34. 14. Dib-Hajj S. D., Tyrell L., Black J. A. and Waxman S. G. (1998) NaN, a novel voltage-gated Na channel, is expressed preferentially in peripheral sensory neurons and down-regulated after axotomy. Proc. natn. Acad. Sci. USA 95, 8963±8968. 15. Elliott A. A. and Elliott J. R. (1993) Characterization of TTX-sensitive and TTX-resistant sodium currents in small cells from adult dorsal root ganglia. J. Physiol. 463, 39±56. 16. Grosskreutz J., Quasthoff S., KuÈhn M. and Grafe P. (1996) Capsaicin blocks tetrodotoxin-resistant sodium potentials and calcium potentials in unmyelinated C ®bres of biopsied human sural nerve in vitro. Neurosci. Lett. 208, 49±52. 17. Kayano T., Noda M., Flockerzi V., Takahashi H. and Numa S. (1988) Primary structure of rat brain sodium channel III deduced from the cDNA sequence. Fedn Eur. biochem. Socs Lett. 228, 187±194. 18. Kirschstein T., BuÈsselberg D. and Treede R. D. (1997) Coexpression of heat-evoked and capsaicin-evoked inward currents in acutely dissociated rat dorsal root ganglion neurons. Neurosci. Lett. 231, 33±36. 19. Kohane D. S., Kuang Y., Lu N. T., Langer R., Strichartz G. R. and Berde C. B. (1999) Vanilloid receptor agonists potentiate the in vivo local anesthetic activity of percutaneously injected site 1 sodium channel blockers. Anesthesiology 90, 524±534. 20. Kuenzi F. M. and Dale N. (1996) Effect of capsaicin and analogues on potassium and calcium currents and vanilloid receptors in Xenopus embryo spinal neurones. Br. J. Pharmac. 119, 81±90. 21. Li Z., Chapleau M. W., Bates J. N., Bielefeldt K., Lee H. C. and Abboud F. M. (1998) Nitric oxide as an autocrine regulator of sodium currents in baroreceptor neurons. Neuron 20, 1039±1049. 22. Liu L. and Simon S. A. (1996) Capsaicin-induced currents with distinct desensitization and Ca 21 dependence in rat trigeminal ganglion cells. J. Neurophysiol. 75, 1503±1514. 23. Liu L., Wang Y. and Simon S. A. (1996) Capsaicin activated currents in rat dorsal root ganglion cells. Pain 65, 191±195. 24. Marsh S. J., Stansfeld C. E., Brown D. A., Davey R. and McCarthy D. (1987) The mechanism of action of capsaicin on sensory C-type neurons and their axons in vitro. Neuroscience 23, 275±289. 25. Mayer E. A. and Gebhart G. F. (1994) Basic and clinical aspects of visceral hyperalgesia. Gastroenterology 107, 271±293. 26. Michael G. J. and Priestley J. V. (1999) Differential expression of the mRNA for the vanilloid receptor subtype 1 in cells of the adult rat dorsal root and nodose ganglia and its downregulation by axotomy. J. Neurosci. 19, 1844±1854. 27. Noda M., Ikeda T., Kayano T., Suzuki H., Takeshima H., Kurasaki M., Takahashi H. and Numa S. (1996) Existence of distinct sodium channel messenger RNAs in rat brain. Nature 320, 188±192. 28. Noda M. and Numa S. (1987) Structure and function of sodium channel. J. Recept. Res. 7, 467±497. 29. Novakovic S. D., Tzoumaka E., McGivern J. G., Haraguchi M., Sangameswaran L., Gogas K. R., Eglen R. M. and Hunter J. C. (1998) Distribution of the tetrodotoxin-resistant sodium channel PN3 in rat sensory neurons in normal and neuropathic conditions. J. Neurosci. 18, 2174±2187. 30. Okabe H., Okubo T., Adachi H., Ishikawa T. and Ochi Y. (1997) Immunohistochemical demonstration of cytokeratin in human embryonic neurons arising from placodes. Brain Dev. 19, 347±352. 31. Sangameswaran L. B., Delgado S. G., Fish L. M., Koch B. D., Jakeman L. B., Stewart G. R., Sze P., Hunter J. C., Eglen R. M. and Herman R. C. (1997) Structure and function of a novel voltage-gated, tetrodotoxin-resistant sodium channel speci®c to sensory neurons. J. biol. Chem. 271, 5953±5956. 32. Schaller K. L., Krzemien D. M, Yarowsky P. J., Krueger B. K. and Caldwell J. H. (1995) A novel, abundant sodium channel expressed in neurons and glia. J. Neurosci. 15, 3231±3242. 33. Schild J. H. and Kunze D. L. (1997) Experimental and modeling study of Na 1 current heterogeneity in rat nodose neurons and its impact on neuronal discharge. J. Neurophysiol. 78, 3198±3209. 34. Schuligoi R., Jocic M., Heinemann A., SchoÈninkle E., Pabst M. A. and Holzer P. (1998) Gastric acid-evoked c-fos messenger RNA expression in rat brainstem is signaled by capsaicin-resistant vagal afferents. Gastroenterology 115, 649±660. 35. Su X., Wachtel R. E. and Gebhart G. F. (1999) Capsaicin sensitivity and voltage-gated sodium currents in colon sensory neurons from rat dorsal root ganglia. J. Neurophysiol. 277, G1180±G1188. 36. Toledo-Aral J. J., Moss B. L., He Z. J., Koszowski A. G., Whisenend T., Levinson S. R., Wolf J. J., Silos-Santiago I., Halegoua S. and Mandel G. (1997) Identi®cation of PN1, a predominant voltage-dependent sodium channel expressed principally in peripheral neurons. Proc. natn. Acad. Sci. USA 94, 1527±1532. 37. Traub R. J., Sengupta J. N. and Gebhart G. F. (1996) Differential c-fos expression in the nucleus of the solitary tract and spinal cord following noxious gastric distension in the rat. Neuroscience 74, 873±884. 38. Waxman S. G., Dib-Hajj S., Cummins T. R. and Black J. A. (1999) Sodium channels and pain. Proc. natn. Acad. Sci. USA 96, 7635±7639. 39. Winter J. (1998) Brain derived neurotrophic factor, but not nerve growth factor, regulates capsaicin sensitivity of rat vagal ganglion neurons. Neurosci. Lett. 241, 21±24. (Accepted 16 August 2000)
Capsaicin and nodose neurons
Pergamon
PII: S0306-4522(00)00375-4
www.elsevier.com/locate/neuroscience
DIFFERENTIAL EFFECTS OF CAPSAICIN ON RAT VISCERAL SENSORY NEURONS K. BIELEFELDT* Department of Internal Medicine, Division of Gastroenterology, University of Iowa, 4614 JCP, 200 Hawkins Drive, Iowa City, IA 52242, USA
AbstractÐNodose neurons play an important role in the regulation of visceral function. Recent studies demonstrated that about 80% of these neurons contain messenger RNA for the capsaicin receptor, a heat-sensitive ion channel. Nodose neurons express voltage-sensitive sodium currents that can be differentiated based on their sensitivity to tetrodotoxin. Considering the potential role of tetrodotoxin-resistant sodium currents in somatosensory neurons, sodium channel expression and sodium currents were studied in nodose neurons. The results were correlated with the response to capsaicin. Nodose neurons contain messenger RNA for the tetrodotoxin-resistant sodium channel PN3. Consistent with these ®ndings, about half of the neurons predominantly expressed tetrodotoxin-resistant sodium currents. In 54% (47/87) of the cells, capsaicin triggered an increase in intracellular calcium. Similarly, in 42% (18/43) of the cells, capsaicin elicited an inward current. There was no relationship between cell size (r 0.07) or sodium current properties (r 0.14) and the response to capsaicin. Micromolar concentrations of capsaicin inhibited voltage-dependent sodium, calcium and potassium currents. This effect was use dependent and did not involve the capsaicin receptor. In conclusion, capsaicin changed the excitability of visceral sensory neurons by blocking voltage-dependent ion channels, an effect that may contribute to the analgesic properties of capsaicin. q 2000 IBRO. Published by Elsevier Science Ltd. All rights reserved. Key words: nodose neurons, sodium currents, ion channel block.
The pungent vanilloid capsaicin interacts with an ion channel, the capsaicin or VR-1 receptor, which forms a non-selective cationic channel gated by heat. It has recently been cloned from sensory neurons in dorsal root ganglia. 8 In the dorsal root and trigeminal ganglia, small neurons preferentially respond to capsaicin and its analogs. 13,18,22 Consistent with these results, Michael and Priestley 26 detected mRNA for the VR-1 receptor primarily in small cells that expressed the high-af®nity receptor for nerve growth factor (trkA), a subpopulation of afferent neurons that plays a role in sensation of pain. The importance of the VR-1 receptor in nociception has recently been shown in mice lacking this protein. 7 Prior studies demonstrated that vagal afferents also respond to capsaicin. 24 About 80% of the neurons in the nodose ganglion, the primary sensory ganglion of the vagus, contain mRNA for the VR-1 receptor independent of size or expression of neurotrophin receptors. 25 Although the majority of neurons contain mRNA for the receptor, the functional importance of the VR-1 receptor in vagal neurons remains unknown. Prior studies revealed that a subgroup of nodose neurons responded to capsaicin, indicating that the cells indeed expressed the ion channel. 24,39 However, the percentage and characterisitics of these capsaicin-responsive neurons are not known. While vagal afferents provide the main route of sensory input necessary for the regulation of visceral function, they are generally believed not to contribute to nociception. 25 Recent studies indicate that vagal ®bers may be involved in chemo-sensation 34 and perception of noxious gastric distension. 37
Nodose neurons express voltage-dependent sodium currents that can be distinguished based on their electrophysiological and pharmacological properties. 21,33 The current sensitive to the sodium channel blocker tetrodotoxin (TTX) activates and inactivates at more negative voltages than the current resistant to the toxin. In the somatosensory afferents within the dorsal root ganglia, the TTX-resistant sodium current is preferentially found in small-diameter neurons involved in nociception. 5,29,38 Using mRNA from rat dorsal root ganglia, the sequence encoding this TTXresistant sodium channel was determined and named PN3 or SNS. 1,31 In addition to PN3, another sodium channel with low sensitivity to TTX has recently been cloned from dorsal root ganglia. 14 The high-threshold for activation of the native and heterologously expressed channel correlates well with the high threshold of polymodal nociceptive afferents observed in vivo, suggesting that this channel may indeed play a key role in determining the response characteristics of nociceptors. To examine whether nodose neurons functionally express the VR-1 receptor, the effects of capsaicin were examined in cultured neurons isolated from the nodose ganglion. In addition, the relationship between sodium currents and capsaicin effects was addressed to assess whether nodose neurons preferentially express the VR-1 receptor in cells with TTXresistant sodium currents.
Experimental animals
*Tel.: 11-319-384-9841; fax: 11-319-353-6399. E-mail address: [email protected] (K. Bielefeldt). Abbreviations: EGTA, ethyleneglycolbis(aminoethyl ether)tetra-acetate; HEPES, N-2-hydroxyethylpiperazine-N 0 -2-ethanesulfonic acid; PCR, polymerase chain reaction; TTX, tetrodotoxin; VR, vanilloid receptor subtype.
Male Sprague±Dawley rats (aged two to three months) were used for the experiments. The animals were housed under a 12-h light/dark cycle with free access to water and food. Animal handling followed the guidelines of the American Physiological Society. The experimental protocol had previously been approved by the Animal Care Committee of the University of Iowa. All efforts were made to minimize the suffering and reduce the number of animals used.
EXPERIMENTAL PROCEDURES
727
728
K. Bielefeldt Table 1. Primers used for the identi®cation of sodium channel RNA present in nodose neurons (the GenBank access code is given in parentheses)
Channel PN1 PN3 NaG b-SCH-1 b-SCH-2 b-SCH-3 SNP-6
Sense primer
Anti-sense primer
Reference
CACTGTTGACAACCCTCTGC 3381±3400* CAGTGATTCTGGAGAACTTC 5181±5200 TGGGGATTGACACAGCCTC 1589±1608 AGAGACCATGTGGGACTGCA 3110±3129 AGAGACCATGTGGGACTGCA 3042±3061 AGAGACCATGTGGGACTGCA 3101±3120 ACCCTCTGAGTGAGGATGAC 5321±5340
GTCCAGCCAACACCAGGCAT 3780±3761* TCCAGACAGTGGATCTTATC 5430±5411 GAACACACTGCCATGACTA 1782±1790 CTGCCGGTCCCTATGCCACT 3520±3501 TTATTCAGATCTTCCAGCGG 3350±3331 TTTTCGGAAGCACTCCCGTA 3350±3331 GATGTCCAACTCCCCACTGT 5580±5561
Toledo-Aral 36 (U79568) Sangameswaran 31 (U53833) Chen 9 (U60590) Noda 27 (M22253) Noda 28 (X03639) Kayano 17 (Y00766) Schaller 32 (L39018)
*Numbers indicate the corresponding region in the published cDNA sequence.
Cell dissociation and culture Rats were anesthetized and decapitated, and the nodose ganglia were quickly removed under a dissection microscope. The tissue was minced with a surgical blade and incubated for 50 min in modi®ed Leibowitz L-15 medium containing collagenase (type 1A, 1 mg/ml), trypsin (type III, 1 mg/ml), and DNase (type IV, 0.1 mg/ml) at 378C. The enzymatic digestion was terminated by adding soybean trypsin inhibitor (2 mg/ml), 3 mM CaCl2 and bovine serum albumin (1 mg/ml). After gentle trituration, the tissue fragments were centrifuged at 800 r.p.m. for 5 min and then resuspended in modi®ed L-15 medium with 5% rat serum and 2% chick embryo extract (Life Technologies, NY). The cells were plated on poly-l-lysine-coated glass coverslips and incubated overnight at 378C prior to the electrophysiological studies. At that time, the nerve cell bodies were round and did not have processes, reducing the potential for space-clamp problems during electrophysiological experiments. Electrophysiological recordings The cells attached to the coverslips were transferred into a 0.5-ml recording chamber on the stage of an inverted microscope. Sodium currents were recorded using the whole-cell patch-clamp technique with an Axopatch 200A ampli®er (Axon Instruments, California) interfaced with a personal computer. The patch pipettes were pulled from thin-walled borosilicate glass (TW150-4, World Precision Instruments, Florida) with tip resistances of 1±3 MV after ®re polishing. Current recordings were ®ltered at 2 kHz with a four-pole Bessel ®lter and digitized at 10 kHz using a Digidata 1200 interface (Axon Instruments). The series resistance and whole-cell capacitance were compensated by more than 80%. The passive membrane properties were checked repeatedly during the course of the experiments. Cells were used for data analysis only if these properties remained stable. The software package pCLAMP6.0 (Axon Instruments) was used for data acquisition and analysis. The leak current and residual capacitative transients were digitally subtracted using the p/n protocol for leak subtraction with n 4. In current-clamp experiments, the cells were initially kept at their resting membrane potential of 258.7 ^ 1.8 mV. The threshold for action potential generation was determined with depolarizing current injection of increasing amplitude. The cells were then stimulated with a current injection of 1.5 times the threshold value lasting 1.8 s. All experiments were performed at room temperature (218C). Determination of the intracellular calcium concentration Cultured nodose neurons were plated on 2.5-cm poly-l-lysinecoated coverslips. The cells were washed with normal extracellular solution, and incubated for 45±60 min at 378C with the membranepermeable ester of the calcium-sensitive ¯uorescent indicator fura-2, fura-2-AM, at a concentration of 3 mM and 0.02% pluronic. 3 After removal of the loading solution, the cells were kept in normal extracellular medium for at least 30 min to allow de-esteri®cation of the acetoxymethyl ester. The cells were then mounted in a recording chamber on an inverted ¯uorescence microscope. A 75-W xenon lamp served as light source. The light passed through interference
®lters mounted on a computer-controlled shutter allowing subsequent imaging at excitation wavelengths of 340 and 380 nm. The emitted light passed through a ®lter set to a wavelength of 510 nm and was measured with a CCD camera (Photon Technology International). All images were automatically corrected for background ¯uorescence. The camera data were fed to an image analysing system, digitized and stored on-line on a personal computer (Photon Technology International). Regions of interest were selected visually for the data analysis. Typically, the loading was homogeneous and did not allow identifying subcellular compartments. In some cells, the indicator appeared to be compartmentalized in areas with higher calcium concentrations, as indicated by a ¯uorescence ratio that exceeded the remainder of the cytoplasm by a factor of two or more. These regions were excluded from further analysis. The images obtained at 340 and 380 nm were divided pixel by pixel to generate the ratio values for the selected regions of interest. As the spatial resolution of the CCD camera is poor, the images represent averages of the free calcium concentration in the cytosol and compartmentalized calcium in small organelles. Therefore, the data were not converted into calcium concentration but expressed as ¯uorescence ratio. Reverse transcription±polymerase chain reaction mRNA was extracted from rat brain and nodose ganglia following a protocol described previously. 10 After reverse transcription with Moloney-monkey leukemia virus, reverse transcriptase ®rst-strand cDNA templates were ampli®ed in 30 cycles using 2 U of Taq DNA polymerase (Gibco BRL) in 50 ml of 100 mM Tris buffer containing 50 mM KCl and 1.4±1.9 mM MgCl2. Each cycle consisted of denaturing at 948C for 1 min, followed by annealing at 56±648C for 1 min and extension at 728C for 1.5 min. The polymerase chain reaction (PCR) products were fractionated on 1% agarose containing ethidium bromide. Isoform-speci®c primers were used to further de®ne the sodium channel subunits expressed (Table 1). The PCR products were puri®ed and sequenced using the Applied Biosystems 373 A Stretch Sequencer at the DNA Core Facility of the University of Iowa. The obtained sequences were then analysed for similarity with known sequences by searching the NCBI data bank at the National Institutes of Health. Data analysis All data are expressed as mean ^ S.E.M. Exponential ®ts were obtained using the curve-®tting routine of the Origin software package (Microcal, California). The t-test was used to discern signi®cant difference after intervention. Statistical signi®cance was determined at P , 0.05. Solutions and chemicals To measure sodium or calcium currents, the pipette solution contained (in mM) 100 cesium aspartate, 20 CsCl, 2.3 CaCl2, 4.8 MgCl2, 10 EGTA, 10 HEPES, 4 Mg-ATP and 0.5 Na-GTP, buffered to pH 7.2 with CsOH. For potassium currents and current-clamp experiments, pipettes were ®lled with a solution containing (in mM)
729
Capsaicin and nodose neurons
Fig. 1. Voltage-dependent sodium currents and action potentials in nodose neurons. Depolarizations from 270 mV to various test potentials between 260 and 40 mV trigger rapidly activating and inactivating inward currents that differ in their voltage dependence and kinetics. Superimposed sample traces are shown for two representative neurons with fast (A) and slow (B) kinetics (calibration bars: 2 nA and 5 ms, respectively). Current-clamp experiments demonstrate two distinct response characteristics to long depolarizating current injections. Representative voltage traces obtained after stimulation of neurons with a depolarizing current injection at 50% above threshold from 260 mV show cells responding either with a single spike (C) or a burst (D) of action potentials (calibration bars: 40 mV and 500 ms, respectively). The steady-state inactivation was tested with 200-ms prepulses to potentials between 2100 and 20 mV, followed by a step to 20 mV (E). The results were ®tted with a Boltzmann equation (dotted lines) for rapidly activating neurons (circle; n 17) and slowly activating neurons (squares; n 17). The V1/2 was 240.8 mV with a slope factor of 9.3 for group 1 neurons and 212.5 mV with a slope factor of 5.1 for group 2 neurons. The histogram (F) demonstrates the existence of two overlapping populations of neurons based on their voltage of half-inactivation (n 198). A histogram describing the effects of the sodium channel blocker TTX in 105 neurons supports the existence of two groups of neurons (G).
145 potassium aspartate, 4 NaCl, 10 EGTA, 1 MgCl2, 2 ATP, 0.5 GTP, 10 HEPES and 2.3 CaCl2, adjusted with KOH to a pH of 7.2. The extracellular solution for the measurement of sodium currents was composed of 20 mM NaCl, 120 nM choline chloride, 10 mM tetraethylammonium chloride, 3 mM MgCl2, 10 mM HEPES and 5.5 mM glucose buffered at pH 7.3 with tetraethylammonium hydroxide. To record calcium currents, the extracellular solution was composed of (in mM) 5 CaCl2, 140 choline chloride, 10 HEPES and 5.5 glucose, buffered at pH 7.3 with tetraethylammonium hydroxide. For currentclamp recordings and calcium imaging, the extracellular solution contained (in mM) 145 NaCl, 4.5 KCl, 1 MgCl2, 2 CaCl2, 5 glucose, buffered with HEPES and NaOH to pH 7.4. For the recording of potassium currents, the extracellular solution contained (in mM) 140 choline chloride, 4.5 KCl, 3 MgCl2, 5 glucose, buffered with HEPES and KOH to pH 7.4. Capsaicin was dissolved in ethanol at a concentration of 1 mM and was diluted immediately before use. The ®nal concentration of ethanol in the bathing solution never exceeded 0.01%. In control experiments, this concentration did not lead to any changes in membrane properties or intracellular calcium concentration. In some experiments, capsaicin was applied by a pressure application. A glass pipette with a tip resistance of approximately 1 MV was placed at approximately 100 mm from the cell using a micromanipulator. The drug was applied with a 100-ms pressure pulse of 10 psi using a pneumatic PicoPump (World Precision Instruments, Sarasota, FL). Fura-2-AM and pluronic were purchased from Molecular Probes
(Eugene, OR). All other chemicals were analysis grade and were obtained from Sigma Chemical (St Louis, MO). RESULTS
Sodium currents in nodose neurons Consistent with previous observations, 21,33 nodose neurons expressed voltage-dependent sodium currents that could be differentiated based on their physiological and pharmacological properties. Brief depolarizations from 270 mV to test potentials positive to 250 mV triggered rapidly activating and inactivating inward currents (Fig. 1A, B). Similarly, two distinct patterns were observed when cells were examined under current-clamp conditions. After determining the threshold for action potential generation, the neurons were stimulated using 1.5 times the current amplitude required to generate a single action potential. At a resting membrane potential of 260 mV, such current injections triggered action potentials at a frequency of 2.8 ^ 0.8 Hz (n 35). Interestingly, again two groups emerged. Nineteen neurons
730
K. Bielefeldt
Fig. 2. Molecular identi®cation of sodium channels expressed in nodose ganglia. Primers for sodium channels expressed primarily in central (A) or peripheral (B) neurons were used to amplify cDNA obtained from rat brain (I) and rat nodose ganglia (II). The ®rst lane shows molecular markers with a 123-bp ladder.
responded with only a single action potential at the beginning of the stimulus (Fig. 1C). In contrast, 16 neurons generated several action potentials at a frequency between two and 12.5 Hz (Fig. 1D). The mean amplitude of the action potential differed slightly between these two groups: 119.6 ^ 2.7 mV in cells with repeated action potentials vs 112.1 ^ 2.3 mV in the remaining neurons (P , 0.04). The duration of the action potential measured at the half-maximal voltage was slightly but not signi®cantly prolonged in the latter group of cells (4.9 ^ 0.6 vs 8.5 ^ 2.3 ms; P 0.1). The sodium current density was signi®cantly higher in cells with persistent electrical activity during prolonged depolarization compared to quiescent cells (242.8 ^ 25.9 vs 161.2 ^ 18.5 pA/pF; P , 0.01). We examined the effect of 2 mM TTX on the sodium current in 25 cells that were also studied under current-clamp conditions. The sodium channel blocker inhibited 65.9 ^ 10.7% of the peak inward current in cells that generated only a single action potential (n 13) compared to 28.2 ^ 10.2% in neurons that responded with two or more action potentials during prolonged depolarization (P , 0.03; n 12). When examined under voltage-clamp conditions, cells differed signi®cantly in the steady-state inactivation (Fig. 1E). A histogram was constructed based on the half-inactivation (V1/2) obtained by ®ts with the Boltzmann equation, which demonstrated two overlapping populations of neurons (Fig. 1F). Currents with rapid activation and inactivation kinetics had a V1/2 of around 245 mV, while currents with slower activation and inactivation had a V1/2 that was about 20 mV more positive. Prior studies have demonstrated differences in the sensitivity of nodose neurons to the neurotoxin TTX. Consistent with these ®ndings, 2 mM TTX inhibited sodium currents to less than 20% of control in about 50% of the cells. However, in the remaining cells, this concentration of TTX led only to a partial or no inhibition of the sodium inward current (Fig. 1G). The pharmacological characteristics correlated with the electrophysiological characteristics as suggested by the correlation coef®cient of r 20.87 (P , 0.01) between the voltage of half-inactivation and the fraction of residual current after administration of TTX. Previously published studies in neurons from the dorsal
root ganglion indicated that the TTX-resistant sodium channel is preferentially expressed in small neurons with unmyelinated ®bers. 4,5,29 To test whether such a relationship exists in nodose neurons the cell capacitance, a measure of the membrane surface area and thus the size of a cell, was correlated with the inhibitory effect of 2 mM TTX. The correlation of coef®cient of 20.37 (P , 0.01; n 67) shows the opposite result in nodose neurons with large cells predominantly expressing TTX-resistant sodium currents. Sodium channel isoforms present in nodose neurons The electrophysiological and pharmacological data suggest that at least two different sodium channels are expressed in nodose neurons. To determine which of the known sodium channels are present in the cells, a PCR-based approach was chosen. Speci®c primers were designed for neuronal sodium channels (Table 1). As expected, mRNA for these channels can be found in the brain, which served as a positive control. In contrast, only primers for b-SCH-I and for PN3 yielded a positive reaction with cDNA from nodose neurons (Fig. 2). Interestingly, primers for NaG, a TTX-resistant sodium channel that was detected in neurons from the canine nodose ganglion, 9 did not yield a positive reaction product even when tested under low stringency conditions (data not shown). The PCR products were puri®ed and sequenced. A computerbased search of the NCBI database at the National Institutes of Health demonstrated a complete identity between the sequenced PCR products and the published primary structure for the ampli®ed region of b-SCH-I and PN3 (data not shown). Effects of capsaicin on the intracellular calcium concentration To determine whether nodose neurons respond to capsaicin, the cells were loaded with fura-2 to measure changes in the intracellular calcium concentration. After obtaining a stable baseline, capsaicin or vehicle solution were added to the extracellular solution. Vehicle did not cause signi®cant changes in the intracellular calcium concentration (decrease
Capsaicin and nodose neurons
731
Capsaicin response and sodium current expression When cells were studied under voltage-clamp conditions, capsaicin triggered an inward current in 18 of 43 cells (42%) examined (Fig. 3B). Repeated administration within 10 s demonstrated a signi®cant decrease in the peak amplitude, suggesting a desensitization of the ion channel (53 ^ 12% of control; P , 0.01; n 8). A similar, though slower and persisting increase was seen in 12 of 23 cells when capsaicin was applied through exchange of the bathing solution (Fig. 3C). At a concentration of 10 mM, the capsazepine inhibited the response to 1 mM capsaicin to 20 ^ 11% of control (P , 0.05; n 3). Prior studies on dorsal root and trigeminal ganglia revealed that small-diameter neurons primarily respond to capsaicin. 12,13,18,22,23 In contrast, the cell size measured by the hole cell capacitance did not correlate with the capsaicin response in nodose neurons (r 0.07; P 0.65; n 45). To determine whether the capsaicin response differs between cells expressing TTX-sensitive or TTX-resistant sodium current, cells were depolarized from 260 to 0 mV before and after the administration of 2 mM TTX. The sodium current inhibition by TTX did not correlate with the peak inward current triggered by capsaicin (r 0.14; P 0.54; n 22). Thus, neither cell size nor the expression of sodium currents allowed prediction of the response to capsaicin. Capsaicin and voltage-dependent ion currents
Fig. 3. Effect of capsaicin on nodose neurons. (A) Sample traces obtained after administration of 500 nM capsaicin or vehicle to extracellular solution (arrow). (B) A representative current trace of a nodose neuron held at 260 mV. The horizontal bar indicates the pressure application of 1 mM capsaicin (100 ms; 10 psi) using a second pipette positioned at a 100-mm distance from the cell (calibration bars: 200 pA; 0.5 s). (C) Capsaicin was applied by rapid exchange of the perfusion solution (calibration bars: 200 pA; 2 s).
by 2 ^ 1%; n 12). However, 47 of the 87 cells examined (54%) responded with a rise in intracellular calcium (Fig. 3A). At a concentration of 1 mM, capsaicin caused an increase in the intracellular calcium level by 54.0 ^ 16.0% (n 5). To assess whether unresponsive cells were viable, potassium depolarizations were performed. Interestingly, prior exposure to high concentration of capsaicin blunted the response to subsequent depolarization with high potassium solution in these cells, suggesting a change in excitability. While the calcium level rose by 34.8 ^ 5.1% in control conditions, it only increased by 18.4 ^ 6.2% in cells pretreated with 10 mM capsaicin (P , 0.05; n 4).
As described above, high concentrations of capsaicin blunted the rise in intracellular calcium triggered by potassium depolarization. To determine the potential mechanism of this effect and its relationship to the capsaicin receptor, the effects of capsaicin on sodium, calcium and potassium currents were studied. Capsaicin administration to the extracellular solution concentration dependently and reversibly decreased the peak sodium current (Fig. 4). At 10 mM, the peak sodium current decreased by 22.7 ^ 4.3% (P , 0.01; n 31). This effect was completely reversible upon washout (peak current 97 ^ 3% of control; not signi®cant). The kinetics of activation and inactivation were not signi®cantly altered (data not shown). In four cells with more than 90% TTX-resistant sodium current, capsaicin inhibited the peak sodium current by 31.9 ^ 20%; similarly, in six neurons with more than 90% TTX-sensitive current, capsaicin blocked 31.2 ^ 13% of the peak sodium current. The correlation coef®cient of 0.06 between the sodium current inhibition by capsaicin and the effect of 2 mM TTX further supports that TTX-sensitive and TTX-resistant current were equally affected (P 0.74; n 31). Capsaicin blocked sodium currents even in cells that did not respond to capsaicin with an inward current (Fig. 5). Similarly, the capsazepine (25 mM) signi®cantly decreased voltage-dependent sodium currents (22.9 ^ 1.9%; n 3). Considering the previously described lack of a correlation between the capsaicin-evoked current and the inhibitory effect of capsaicin on sodium current, it is unlikely that capsaicin alters sodium currents through the capsaicin receptor. This is further supported by the results obtained with a high concentration of capsazepine, which also inhibits sodium current, thereby arguing against a speci®c effect of the vanilloid that requires activation of the VR-1 receptor. To examine whether the inhibitory effect of capsaicin was
732
K. Bielefeldt
Fig. 5. Capsaicin response and inhibitory effect of vanilloids. The peak sodium current in the presence of 10 mM capsaicin was plotted as a function of the peak inward current triggered by capsaicin, demonstrating no relationship between the response of the capsaicin receptor and the inhibitory effect of capsaicin on voltage-dependent sodium currents.
Fig. 4. Effect of capsaicin on sodium currents. Sodium currents were triggered by depolarization from 260 to 0 mV. (A) The time-course of current inhibition and recovery is shown for a single cell. A test pulse was applied every 15 s (points). The horizontal bar indicates the addition of 10 mM capsaicin. The insert shows superimposed sample traces obtained before and after the addition of 10 mM capsaicin (calibration bars: 2 nA; 2 ms). (B) The inhibitory effect of capsaicin plotted as a function of the concentration and ®tted with the Hill equation, revealing an IC50 of 40 mM with a Hill coef®cient of 1.0.
restricted to voltage-dependent sodium currents, calcium currents were recorded before and after the administration of capsaicin to the extracellular solution. At a concentration of 10 mM capsaicin, the peak calcium current decreased by 31.9 ^ 6.6% (P , 0.01; n 6). This effect was dose dependent and reversible, as the peak current recovered to 95 ^ 1% of control after washout (Fig. 6). To isolate low-threshold calcium currents, the response to a depolarization from 240 to 10 mV was digitally subtracted from that triggered by a step from 290 to 10 mV. The inhibitory effect of capsaicin on low-threshold calcium current did not differ from that on high-threshold currents (30.0 ^ 8.8%; n 4). An analysis of the current decay revealed that capsaicin signi®cantly accelerated the current inactivation (time constants of 82.1 ^ 20.7 vs 40.1 ^ 9 s for control and capsaicin, respectively; P , 0.05; n 9). While pretreatment with 25 mM capsazepine only decreased calcium currents by 4 ^ 2% (not signi®cant), the antagonist did not block the inhibitory effect of 10 mM capsaicin given subsequently (40.7 ^ 15.8% decrease in peak current; n 4). The effects of capsaicin on sodium and calcium currents indicate that the vanilloid may non-speci®cally block voltagedependent ion channels. Therefore, the effect of capsaicin was
studied on potassium currents. Capsaicin (10 mM) decreased potassium currents by 42.1 ^ 7.5% (P , 0.01; n 8). This inhibition was fully reversible upon washout, as the current amplitude recovered to 100 ^ 3% compared to control. The effect was dose dependent, as shown in Fig. 7. To examine whether capsaicin preferentially inhibits sustained or transient outward currents, the A current was isolated by digital subtraction of the current response to a depolarization from 240 to 40 mV from that triggered by a step from 290 to 40 mV. Only four of the 13 cells tested expressed a signi®cant fraction of transient outward current. Capsaicin inhibited both the transient and sustained outward currents in these cells by 29 ^ 8% and 41 ^ 9%, respectively (n 4). Similar to the effects on calcium currents, capsaicin signi®cantly accelerated the current decay. Under control conditions, the currents did not appreciably inactivate within 100 ms after a depolarization from 260 to 40 mV. After the addition of capsaicin, the current inactivated with a time constant of 8.1 ^ 1.8 ms (P , 0.01; n 5). Capsazepine (25 mM) decreased the peak current by 48.1 ^ 14.7% (n 3). The accelerated decay of potassium and calcium currents seen in the presence of capsaicin is suggestive of a usedependent block. Moreover, the time-course of activation for calcium currents slowed in the presence of capsaicin (10.9 ^ 1.2 vs 16.7 ^ 3.1 ms for control and capsaicin), as demonstrated in Fig. 8A. To further examine whether capsaicin indeed preferentially inhibits open channels, the peak current was measured in cells that were repeatedly depolarized. When the voltage was stepped from 260 to 0 mV for 50 ms at a frequency of 0.5 Hz, cumulative inhibition caused a decrease in the peak calcium current by 16 ^ 6% within 1 min (Fig. 8B). In the presence of 20 mM capsaicin, the peak current declined by 35 ^ 8% (P , 0.05; n 6). Using a similar paradigm for potassium currents, no signi®cant inhibition was observed under control conditions (Fig. 8C). In contrast, capsaicin caused a progressive inhibition with a decrease in the peak potassium current by 14 ^ 4% within 1 min (P , 0.05; n 9). To examine the effect of capsaicin on the cumulative inhibition of sodium currents, cells were depolarized from 270 to 0 mV for 5 ms at a frequency of
Capsaicin and nodose neurons
Fig. 6. Effect of capsaicin on calcium currents. (A) The time-course of current inhibition and recovery is shown for a single cell. A test pulse was applied every 15 s (points). Calcium currents were triggered by depolarization from 260 to 10 mV. The horizontal bar indicates the addition of 10 mM capsaicin. The insert shows superimposed sample traces obtained before and after the addition of 10 mM capsaicin (calibration bars: 2 nA; 20 ms). (B) The inhibitory effect of capsaicin plotted as a function of the concentration and ®tted with the Hill equation, revealing an IC50 of 8.6 mM with a Hill coef®cient of 0.7.
10 Hz. The higher stimulation frequency was chosen as the sodium current essentially completely recovered from inactivation within 100 ms. In control cells, there was a cumulative inhibition with a peak current decrease of 15 ^ 2% within 10 s (Fig. 8D). This inhibition was more pronounced in the presence of capsaicin (P , 0.05; n 8). DISCUSSION
The key ®ndings of this study are: (i) two populations of nodose neurons can be distinguished based on the pharmacological and biophysical characteristics of their sodium currents; (ii) about half of the nodose neurons respond to capsaicin; (iii) the response to capsaicin does not correlate with cell size or the expression of a TTX-resistant sodium current; (iv) high concentrations of capsaicin and its competitive antagonist capsazepine inhibit voltage-sensitive ion channels. The data presented con®rm previous observations that nodose neurons express voltage-dependent sodium currents that can be differentiated based on their sensitivity to the neurotoxin TTX. The current-clamp experiments support the potential physiological importance of this ®nding,
733
Fig. 7. Effect of capsaicin on potassium currents. (A) The time-course of current inhibition and recovery is shown for a single cell. A test pulse was applied every 15 s (points). Potassium currents were triggered by depolarization from 260 to 40 mV. The horizontal bar indicates the addition of 10 mM capsaicin. The insert shows superimposed sample traces obtained before and after the addition of 10 mM capsaicin (calibration bars: 1 nA; 20 ms). (B) The inhibitory effect of capsaicin plotted as a function of the concentration and ®tted with the Hill equation, revealing an IC50 of 20 mM with a Hill coef®cient of 0.83.
demonstrating that cells with a high current density and mostly TTX-resistant sodium current were able to generate action potentials during prolonged stimulations. Similar TTX-sensitive and TTX-resistant sodium currents have been described in somatosensory and colon sensory neurons. 4,5,11,15,35 However, a more detailed evaluation of these results reveals several differences if one compares somatosensory neurons from the dorsal root with the visceral sensory neurons from the nodose ganglion. Both TTX-sensitive and TTX-resistant sodium currents in nodose neurons inactivate at potentials that are shifted by about 20 mV to the right compared to results obtained from somatosensory neurons. 11,15 Furthermore, large cells were more likely to express TTX-resistant current than small cells, a ®nding that is contrary to results described in neurons from dorsal root ganglia. 5,29 Finally, mRNA for only two sodium channel a-subunits was found in nodose neurons, while at least seven sodium channel isoforms were detected in dorsal root ganglia when a similar approach was used. 4 The recently cloned PN3 channel subunit encodes a TTXresistant sodium channel that is preferentially expressed in nociceptive neurons. 27 Similarly, the capsaicin response or expression of the VR-1 receptor has also been used as a marker of nociceptive neurons. 7,8 Interestingly, only about 50% of the nodose neurons responded to capsaicin, while
734
K. Bielefeldt
Fig. 8. Use-dependent inhibition of voltage-dependent currents by capsaicin. (A) Changes in the time-course of activation. Calcium current traces were normalized and superimposed for control conditions (solid line) and after the addition of 20 mM capsaicin (broken line). (B) The effect of capsaicin on peak calcium currents triggered by repeated depolarization from 260 to 0 mV for 50 ms at 0.5 Hz (®lled symbols: control; open symbols: capsaicin). (C) The effect of capsaicin on peak potassium currents triggered by repeated depolarization from 260 to 40 mV for 50 ms at 0.5 Hz (®lled symbols: control; open symbols: capsaicin). (D) The effect of capsaicin on peak sodium currents triggered by repeated depolarization from 270 to 10 mV for 5 ms at 10 Hz (®lled symbols: control; open symbols: capsaicin).
mRNA for the VR-1 receptor can be detected in 80% of the neurons. 26 Marsh et al. 24 focused on C-type nodose neurons and demonstrated that about 60±70% of these cells exhibited electrical responses to capsaicin. Even more neurons demonstrated ultrastructural changes after prolonged exposure to capsaicin at concentrations between 1 and 10 mM. 24 Considering the non-speci®c effects of capsaicin described in this paper, it remains unclear whether these effects were solely mediated by activation of the capsaicin receptor. Similarly, Winter 39 observed responses of nodose neurons to 10 mM capsaicin. However, she did not provide information about the percentage or characteristics of neurons that were affected by capsaicin. Consistent with these results, Benson et al. 2 demonstrated that about 60% of the cardiac sensory neurons in the nodose ganglion respond to capsaicin. As neurons from mice lacking the VR-1 receptor do not respond to capsaicin, 6 it appears reasonable to use the physiological results obtained with capsaicin as a surrogate for the presence of the VR-1 receptor. Therefore, the data presented in this article suggest that only slightly more than half of the neurons functionally express the VR-1 receptor. This contrasts with the dorsal root ganglia, where about 50% of the neurons contain mRNA for the VR-1 receptor, 26 which corresponds very well with physiological data demonstrating a response to capsaicin in about 50±70% of the cells examined. 12,13,18,23 Neither cell size nor the sensitivity to TTX allowed
prediction of the response to the vanilloid capsaicin in nodose neurons. Similarly, Marsh et al. 24 did not see a correlation between the capsaicin effect and the electrophysiological properties of nodose cells. These results differ from ®ndings in somatosensory neurons. Using in situ hybridization, mRNA for the VR-1 receptor was found primarily in small and medium-sized cells. 26 Moreover, small cells were also more likely to respond to capsaicin when compared with large neurons from the dorsal root ganglia. 12,13,18,23 Su et al. 35 recently reported no relationship between cell size and response to capsaicin in visceral sensory neurons from dorsal root ganglia. In contrast to the ®ndings in nodose neurons, these investigators could not elicit capsaicin responses in neurons that only expressed TTX-sensitive sodium current. These ®ndings demonstrate that the properties of visceral sensory neurons differ from those of somatosensory neurons. However, additional differences may exist between the visceral sensory neurons in the placode-derived nodose ganglion 30 compared to the colonic sensory neurons in the dorsal root that originate in the neural crest. 6 These data raise the question whether, at least in the context of visceral neurons, cell size, sodium channel expression or capsaicin sensitivity can reliably be used as markers of nociception. Previously published studies have demonstrated that the native and heterologously expressed VR-1 receptor has a Kd for capsaicin between 0.5 and 1 mM. 8,12,13,18,23,35 Consistent
735
Capsaicin and nodose neurons
with these ®ndings, 1 mM capsaicin triggered calcium in¯ux and inward currents in about 50% of nodose neurons. Interestingly, higher concentrations decreased the cellular excitability in cells otherwise not affected by the vanilloid, as judged by a blunting of the calcium increase caused by potassium depolarization. As capsaicin activates a non-speci®c cation channel, the resulting prolonged depolarization in the ongoing presence of capsaicin could cause the inactivation of sodium and calcium channels. To circumvent this possibility, the electrophysiological experiments described were performed under voltage-clamp conditions, thus keeping the membrane potential stable and eliminating voltagedependent inactivation processes. Under those conditions, capsaicin directly inhibited sodium, calcium and potassium currents. Several ®ndings suggest that this effect is independent of the VR-1 receptor. First, the IC50 for the inhibitory effect of capsaicin was one order of magnitude higher than the known Kd for the VR-1 receptor. Second, while the response to capsaicin desensitizes in the presence of extracellular calcium, 8 the inhibition of voltage-dependent ion current remained stable for several minutes, even with an extracellular calcium concentration of 5 mM. Conversely, capsaicin blocked sodium currents in the absence of extracellular calcium, thereby arguing against an indirect effect of calcium in¯ux through the VR-1 receptor, leading to an activation of intracellular signaling pathways, such as the activation of calcium-dependent protein kinases. Third, capsazepine also inhibited voltage-dependent ion channels and did not block the effect of capsaicin. Finally, there was no correlation between the capsaicin-evoked inward current and the inhibitory effect of capsaicin in the cells examined. Su et al. 35 described a similar inhibitory effect of capsaicin on colonic sensory neurons. However, the inhibition of sodium currents did not reverse after removal of capsaicin and was blunted by the competitive antagonist capsazepine. Moreover, the sodium current in cells that did not respond to capsaicin was not affected by the vanilloid. Based on these results, the authors hypothesized that capsaicin-induced changes in the intracellular calcium concentration may secondarily alter the channel properties. As summarized above, such a mechanism is unlikely for the capsaicininduced inhibition of voltage-dependent currents in nodose neurons. As mentioned previously, calcium was omitted
from the extracellular solution when sodium currents were measured. In the presence of capsaicin, calcium and potassium currents inactivated much more rapidly than under control conditions. Such a result is typical for an open channel or use-dependent block, where the ion channel has to open before the interaction between the inhibitor and the channel pore can take place. This is further supported by experiments demonstrating a progressive decrease in peak calcium, sodium and potassium current with repeated stimulations. Previous studies in vivo have demonstrated that capsaicin augments the anesthetic effect of locally administered TTX. 19 Capsaicin also reduced the extracellularly recorded compound action potential in human sural nerve preparations. In contrast to the observation in nodose neurons, capsaicin primarily reduced the TTX-resistant component of the compound action potential, suggesting a more selective effect on the TTX-resistant sodium current. 16 Kuenzi and Dale 20 observed that vanilloids dose-dependently blocked potassium and calcium currents in Xenopus spinal neurons. In Xenopus cells, micromolar concentrations of capsaicin inhibited voltage-dependent currents in all cells tested, while only about one-®fth of the neurons showed the typical inward current triggered by capsaicin. The inhibitory effect of capsaicin exhibited the characteristics of an open channel block, consistent with the ®ndings in nodose neurons. CONCLUSION
Despite the presence of mRNA for the VR-1 receptor in about 80% of the nodose neurons, only about half of the cells functionally express the VR-1 receptor as de®ned by a response to the pungent vanilloid capsaicin. While low concentrations activate an inward current in these neurons, higher concentrations non-speci®cally inhibit voltage-gated ion currents through an open channel block. This effect is independent of the known vanilloid receptor. The inhibition of sodium and calcium currents will decrease neuronal excitability and may contribute to the analgesic properties of capsaicin. AcknowledgementsÐThis study was supported by grants from the National Institutes of Health (DK01548-01).
REFERENCES
1. Akopian A. N., Silvilotti L. and Wood J. N. (1996) A tetrodotoxin-resistant voltage-gated sodium channel expressed by sensory neurons. Nature 379, 257±262. 2. Benson C. J., Eckert S. P. and McCleskey W. E. (1999) Acid-evoked currents in cardiac sensory neurons: a possible mediator of myocardial ischemic sensation. Circulation Res. 84, 921±928. 3. Bielefeldt K., Martin M., Whiteis C., Yedidag E. and Abboud F. M. (1997) Modulation of calcium release channels in intestinal smooth muscle cells. Cell Calcium 22, 507±514. 4. Black J. A., Dib-Hajj S., McNabola K., Jeste S., Rizzo M. A., Kocsis J. D. and Waxman S. G. (1996) Spinal sensory neurons express multiple sodium channel a-subunits. Molec. Brain Res. 43, 117±131. 5. Caffrey J. M., Eng D. L., Black S. G., Waxman S. G. and Kocsis J. D. (1992) Three types of sodium channels in adult rat dorsal root ganglion neurons. Brain Res. 592, 283±297. 6. Carey M. B. and Matsumoto S. G. (1999) Neurons differentiating from murine neural crest in culture exhibit sensory or sympathetic-like calcium currents. J. Neurobiol. 39, 501±514. 7. Caterina M. J., Lef¯er A., Malmberg A. B., Martin W. J., Trafton J., Peterson-Zeitz K. R., Klotzenburg M., Basbaum A. I. and Julius D. (2000) Impaired nociception and pain in mice lacking the capsaicin receptor. Science 288, 306±313. 8. Caterina M. J., Schumacher M. A., Tominaga M., Rosen T. A., Levine J. D. and Julius D. (1997) The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature 389, 816±824. 9. Chen J., Ikeda S. R., Lang W., Isales C. M. and Wei X. (1997) Molecular cloning of a putative tetrodotoxin-resistant sodium channel from dog nodose ganglion neurons. Gene 202, 7±14. 10. Chomczynski P. and Sacchi N. (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate±phenol±chloroform extraction. Analyt. Biochem. 162, 156±159.
736
K. Bielefeldt
11. Cummins T. R. and Waxman S. G. (1997) Downregulation of tetrodotoxin-resistant sodium currents and upregulation of a rapidly repriming tetrodotoxin-sensitive sodium current in small spinal sensory neurons after nerve injury. J. Neurosci. 17, 3503±3514. 12. Dedov V. N. and Roufogalis B. D. (1998) Rat dorsal root ganglion neurones express different capsaicin-evoked Ca 21 transients and permeabilities to Mn 21. Neurosci. Lett. 248, 151±154. 13. Del Mar L. P. and Cardenas C. G. (1996) Capsaicin preferentially affects small-diameter acutely isolated rat dorsal root ganglion cell bodies. Expl Brain Res. 111, 30±34. 14. Dib-Hajj S. D., Tyrell L., Black J. A. and Waxman S. G. (1998) NaN, a novel voltage-gated Na channel, is expressed preferentially in peripheral sensory neurons and down-regulated after axotomy. Proc. natn. Acad. Sci. USA 95, 8963±8968. 15. Elliott A. A. and Elliott J. R. (1993) Characterization of TTX-sensitive and TTX-resistant sodium currents in small cells from adult dorsal root ganglia. J. Physiol. 463, 39±56. 16. Grosskreutz J., Quasthoff S., KuÈhn M. and Grafe P. (1996) Capsaicin blocks tetrodotoxin-resistant sodium potentials and calcium potentials in unmyelinated C ®bres of biopsied human sural nerve in vitro. Neurosci. Lett. 208, 49±52. 17. Kayano T., Noda M., Flockerzi V., Takahashi H. and Numa S. (1988) Primary structure of rat brain sodium channel III deduced from the cDNA sequence. Fedn Eur. biochem. Socs Lett. 228, 187±194. 18. Kirschstein T., BuÈsselberg D. and Treede R. D. (1997) Coexpression of heat-evoked and capsaicin-evoked inward currents in acutely dissociated rat dorsal root ganglion neurons. Neurosci. Lett. 231, 33±36. 19. Kohane D. S., Kuang Y., Lu N. T., Langer R., Strichartz G. R. and Berde C. B. (1999) Vanilloid receptor agonists potentiate the in vivo local anesthetic activity of percutaneously injected site 1 sodium channel blockers. Anesthesiology 90, 524±534. 20. Kuenzi F. M. and Dale N. (1996) Effect of capsaicin and analogues on potassium and calcium currents and vanilloid receptors in Xenopus embryo spinal neurones. Br. J. Pharmac. 119, 81±90. 21. Li Z., Chapleau M. W., Bates J. N., Bielefeldt K., Lee H. C. and Abboud F. M. (1998) Nitric oxide as an autocrine regulator of sodium currents in baroreceptor neurons. Neuron 20, 1039±1049. 22. Liu L. and Simon S. A. (1996) Capsaicin-induced currents with distinct desensitization and Ca 21 dependence in rat trigeminal ganglion cells. J. Neurophysiol. 75, 1503±1514. 23. Liu L., Wang Y. and Simon S. A. (1996) Capsaicin activated currents in rat dorsal root ganglion cells. Pain 65, 191±195. 24. Marsh S. J., Stansfeld C. E., Brown D. A., Davey R. and McCarthy D. (1987) The mechanism of action of capsaicin on sensory C-type neurons and their axons in vitro. Neuroscience 23, 275±289. 25. Mayer E. A. and Gebhart G. F. (1994) Basic and clinical aspects of visceral hyperalgesia. Gastroenterology 107, 271±293. 26. Michael G. J. and Priestley J. V. (1999) Differential expression of the mRNA for the vanilloid receptor subtype 1 in cells of the adult rat dorsal root and nodose ganglia and its downregulation by axotomy. J. Neurosci. 19, 1844±1854. 27. Noda M., Ikeda T., Kayano T., Suzuki H., Takeshima H., Kurasaki M., Takahashi H. and Numa S. (1996) Existence of distinct sodium channel messenger RNAs in rat brain. Nature 320, 188±192. 28. Noda M. and Numa S. (1987) Structure and function of sodium channel. J. Recept. Res. 7, 467±497. 29. Novakovic S. D., Tzoumaka E., McGivern J. G., Haraguchi M., Sangameswaran L., Gogas K. R., Eglen R. M. and Hunter J. C. (1998) Distribution of the tetrodotoxin-resistant sodium channel PN3 in rat sensory neurons in normal and neuropathic conditions. J. Neurosci. 18, 2174±2187. 30. Okabe H., Okubo T., Adachi H., Ishikawa T. and Ochi Y. (1997) Immunohistochemical demonstration of cytokeratin in human embryonic neurons arising from placodes. Brain Dev. 19, 347±352. 31. Sangameswaran L. B., Delgado S. G., Fish L. M., Koch B. D., Jakeman L. B., Stewart G. R., Sze P., Hunter J. C., Eglen R. M. and Herman R. C. (1997) Structure and function of a novel voltage-gated, tetrodotoxin-resistant sodium channel speci®c to sensory neurons. J. biol. Chem. 271, 5953±5956. 32. Schaller K. L., Krzemien D. M, Yarowsky P. J., Krueger B. K. and Caldwell J. H. (1995) A novel, abundant sodium channel expressed in neurons and glia. J. Neurosci. 15, 3231±3242. 33. Schild J. H. and Kunze D. L. (1997) Experimental and modeling study of Na 1 current heterogeneity in rat nodose neurons and its impact on neuronal discharge. J. Neurophysiol. 78, 3198±3209. 34. Schuligoi R., Jocic M., Heinemann A., SchoÈninkle E., Pabst M. A. and Holzer P. (1998) Gastric acid-evoked c-fos messenger RNA expression in rat brainstem is signaled by capsaicin-resistant vagal afferents. Gastroenterology 115, 649±660. 35. Su X., Wachtel R. E. and Gebhart G. F. (1999) Capsaicin sensitivity and voltage-gated sodium currents in colon sensory neurons from rat dorsal root ganglia. J. Neurophysiol. 277, G1180±G1188. 36. Toledo-Aral J. J., Moss B. L., He Z. J., Koszowski A. G., Whisenend T., Levinson S. R., Wolf J. J., Silos-Santiago I., Halegoua S. and Mandel G. (1997) Identi®cation of PN1, a predominant voltage-dependent sodium channel expressed principally in peripheral neurons. Proc. natn. Acad. Sci. USA 94, 1527±1532. 37. Traub R. J., Sengupta J. N. and Gebhart G. F. (1996) Differential c-fos expression in the nucleus of the solitary tract and spinal cord following noxious gastric distension in the rat. Neuroscience 74, 873±884. 38. Waxman S. G., Dib-Hajj S., Cummins T. R. and Black J. A. (1999) Sodium channels and pain. Proc. natn. Acad. Sci. USA 96, 7635±7639. 39. Winter J. (1998) Brain derived neurotrophic factor, but not nerve growth factor, regulates capsaicin sensitivity of rat vagal ganglion neurons. Neurosci. Lett. 241, 21±24. (Accepted 16 August 2000)