INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Capsaicin, the pungent ingredient in chili peppers, is used in various medicinal treatments to alleviate pain, even though its initial application can cause pain and inflammation (Sterner and Szallasi 1999; Szolcsanyi 1977; Waddell and Lawson 1989). The initial painful sensation after capsaicin application arises from the selective activation of vanilloid receptors in capsaicin-sensitive nociceptors that leads to depolarization and the generation of action potentials (APs) (Baumann et al. 1991; Gold et al. 1996a; Heyman and Rang 1985; LaMotte et al. 1991; Williams and Zieglganberger 1982). Our goal is to investigate mechanisms by which capsaicin can initiate hyperalgesia in rat trigeminal ganglion (TG) neurons. Here, we investigate its effects on I_A currents of capsaicin-sensitive (CS) and capsaicin-insensitive (CIS) TG neurons. I_A currents are involved in modulating the action potential shape, threshold, and the regulation of the interspike interval (Yoshimura et al. 1996; Yost 1999).

Primary sensory neurons contain several types of voltage-gated potassium channels (VGPCs), which include those that activate rapidly and inactivate slowly (I_K) and those that activate and inactivate rapidly (I_A) (Gold et al. 1996b; Kostyuk et al. 1991; McCleskey and Gold 1999; Petersen et al. 1987; Rasband et al. 2001; Ricco et al. 1996; Seifert et al. 2001; Stansfeld et al. 1986; Woolf and Costagin 1999; Yoshimura et al. 1996). VGPCs of various types are distributed among different types of neurons (Cardenas et al. 1995; Djouhri and Lawson 1999; Harper and Lawson 1985). In one study, I_A currents were found only in CS neurons (Cardenas et al. 1995), but in another similar study, they were found in both CS and CIS DRG neurons (Gold et al. 1996b).

There have been numerous studies reporting the effects of capsaicin on VGPCs. In an earlier study of the effect of capsaicin on I_A currents, Petersen et al. (1987) found that 30 µM capsaicin reduced the amplitude of I_A and accelerated their inactivation in all chick and guinea pig neurons. Capsaicin blocked I_K currents in Schwann cells with an IC₅₀ (= K_1/2) of 8.7 µM (Baker and Ritchie 1994) and spinal neurons from Xenopus embryos with a K_1/2 = 21 µM (Kuenzi and Dale 1996). Capsaicin inhibited VGPCs in nodose ganglion neurons with a K_1/2 = 20 µM (Bielefeldt 2000), and in a similar study in rat DRG neurons, capsaicin inhibited both I_A and I_K currents with a K_1/2 = 8 µM (Atkins and McCleskey 1993). In recordings from cardiac muscle, capsaicin blocked I_K with a K_1/2 = 11.5 µM (Castle 1992), and I_K currents were blocked with a K_1/2 = 17.4 µM in melanotrophs (Kehl 1994). In a study where various Shaker-type K channels were expressed in cell lines and subsequently exposed to capsaicin, the K_1/2s ranged from 26 µM (for Kv_1.5) to 158 µM (for Kv_3.1) (Grissmer et al. 1994). This wide range of K_1/2s indicates that capsaicin must directly interact with these channels rather than produce some nonspecific membrane effect. Although there is a great range in the K_1/2s, in all these cases, the inhibition occurred at concentrations much larger than that required to activate vanilloid receptors (K_1/2 = 0.68 µM) in CS TG neurons (Liu and Simon 1996). Here, we found that 1 µM capsaicin inhibited I_A currents by 49% in CS neurons and that most of this inhibition is a consequence of the activation of vanilloid receptors.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cell culture

TG neurons from both adult Sprague-Dawley rats and from VR1/ mice, the gift of Dr. David Julius of the University of California at San Francisco, were cultured as described previously (Liu and Simon 1996). Briefly, trigeminal ganglia were dissected aseptically and collected in modified Hank's balanced salt solution (mHBSS). After washing in mHBSS, the ganglia were diced into small pieces and incubated in mHBSS for 30-50 min at 37°C in 0.1% collagenase (Type Xl-S). Individual cells were dissociated by triturating the tissue through a fire-polished glass pipette, followed by a 10-min incubation at 37°C in 10 µg/ml DNase I (Type lV) in F-12 medium (Life Technologies, Gaithersburg, MD). After washing three times with F-12, the cells were cultured in DMEM supplemented with 10% fetal bovine serum. The cells were plated on poly-D-lysine glass coverslips (15 mm diam) and cultured overnight at 37°C in a water saturated atmosphere with 5% CO₂. At the beginning of each experiment, neurons were placed for 10 min in a chamber containing the external solution for measuring I_A. The composition of this solution was (in mM) 80 choline Cl, 80 TEACl, 5 KCl, 2.0 CaCl₂, 1.0; MgCl₂, 10 HEPES, 10 D-glucose, and 1 CdCl₂, adjusted to pH 7.4. CdCl₂ was included to block voltage-gated calcium channels. choline-Cl and TEACl were included to reduce currents from voltage-gated sodium channels, I_K currents, hyperpolarization-activated cation channels, and capsaicin-induced inward currents. Only neurons without or with short processes were used. All experiments were carried out at room temperature (22-25°C)

Care of animals conforms to standards established by the National Institutes of Health. All animal protocols were approved by the Duke University Institutional Animal Care and Use Committee.

Patch-clamp recording

In these experiments, we used glass pipettes (R-6 borosilicate, Drummond Scientific, Broomall, PA) with resistances 2 M. The pipette solution was (in mM) 120 K-aspartate, 20 KCl, 1.0 CaCl₂, 2.0 MgCl₂, 10 EGTA, 10 HEPES, and 5 K₂-ATP, adjusted to pH 7.3. Measurements of I_A were performed in neurons with soma sizes ranging from 18 to 45 µm. The capsaicin-sensitive neurons has soma sizes from 18 to 35 µm (Liu and Simon 1996).

The signals were measured using either an Axopatch-1D (early experiments) or an Axopatch 200A (latter experiments) patch clamp amplifier (Axon Instruments, Foster City, CA). The amplifier output was digitized initially with a Digidata 1200 and later with a Digidata1322A converter (Axon Instruments). The sampling rate was 10 kHz. The capacitance and series resistance were compensated, the latter by about 90%. For most neurons, the uncompensated series resistance resulted in voltage-clamp errors within 5 mV and were not corrected. When voltage-clamp errors were greater than 5 mV, the data were discarded. The membrane resistance was measured before, and several times after, capsaicin application (see Fig. 3). The leak current was subtracted from the potassium currents using Clampfit programs.

In experiments where the effects of capsaicin on I_A currents were investigated in CS neurons, our goal was to have calcium enter the cell where it can desensitize vanilloid receptors (Koplas et al. 1997) to an extent that the magnitude of this current is sufficiently small so that its contribution to the leak current can be corrected. To ensure that calcium enters the cell during capsaicin application, TG neurons were held at -60 mV. We only started to measure the inhibitory effect of capsaicin on I_A currents when capsaicin-induced inward currents no longer decreased rapidly and the current desensitized to 0.39 ± 0.21 nA (see Fig. 3). The experiment was terminated if after 3 min of applying capsaicin the inward currents where large or did not decrease to this magnitude. Because intracellular potassium is needed to measure I_A currents, it is impossible to eliminate the current contribution of the capsaicin-induced outward current. However, when the I_A currents are large (see Figs. 2 and 3), the contribution of capsaicin-induced outward currents should be relatively small.

The protocol to measure I_A activation was performed at a holding potential of -80 mV and consisted of 500-ms depolarization pulses from -100 to +50 mV in 10-mV steps with a 2 s interval (see Fig. 3). The amplitude of I_A current was measured at the peak (I_Ap). The maximum of these outward currents is designated I_Apmax.

The steady-state inactivation-voltage protocol consisted of 500-ms preconditioned pulses ranging from -120 to 50 mV followed by 500-ms test pulse depolarizing to 50 mV. The interstimulus interval was 4 s and the holding potential was -80 mV (see Fig. 7).

The protocol used to measure use-dependent blockage consisted of 16 depolarization pulses from a holding potential of -80 to 50 mV with duration 500 ms at frequency of 1 Hz.

Chamber/solution delivery

The recording chamber had a volume of 370 µl and was continuously perfused. Capsaicin, KT5323 and KN-93 were dissolved in dimethylsufoxide (DMSO) to form stock solutions that were further diluted before delivering into the recording chamber using a multibarreled electrode (Adams and List Associated, Westbury, NY) placed about 50 µm from the cell. N-(6-aminohexyl)-5-chloro-1-napthalenesulfonamide HCl (W-7) was dissolved in the solution used for the external buffer. Event markers associated with the opening or closing of valves (Parker Hannifin, Fairfield, NJ) signaled the onset and removal times of the various stimuli.

RT-PCR

Cellular total RNA was prepared from the TGs of six individual adult Sprague-Dawley rats. TG was removed and treated with Trizol Reagent (Gibco BRL No. 15596-026; 100 mg tissue/1 ml Trizol Reagent). The tissues were then quickly homogenized. After treating with ethanol, 20% chloroform, 50% ispropanol, and cold ETOH, the samples were centrifuged and dried under vacuum. The total RNA concentrations were calculated from the absorbance (A) at 260 nm and the purity determined by A260/A280 (the ratio >1.7). To eliminate any residual DNA, all total RNA samples were treated with amplification grade DNase I (Gibco BRL). The cDNA was synthesized by first-strand cDNA synthesis kit for RT-PCR (Boehringer Mannheim). The PCR primers that were used in the present experiment (see Table 1) were synthesized by Gibco BRL. PCR amplification was performed with a PCR core kit (Boehringer Mannheim) with 32 reaction cycles in a Thermal Cycler 3200 (Perkin Elmer, Foster City, CA). Preliminary experiments revealed that under these conditions we were not in the plateau region. The first PCR cycle consisted of a 5-min denaturation time and the last cycle had a 7-min extension time. The other PCR cycles included 1 min of denaturation at 94°C, 1 min of primer annealing at 57°C, and 1 min of extension/synthesis at 72°C. The PCR reaction (10 µl) was separated, and a 2 µl gel loading solution was added to each tube. Agarose gels (2% agarose gel containing 0.6-µg/ml ethidium bromide) were electrophoresed. The gels were digitized using a Foto/analyst TM image analysis system (Fotodyny, Hartland, WI), and the optical density was measured by Image I (Universal Imaging, West Chester, PA.). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control to verify the quality and determine the quantity of RNA. Two controls were used in this experiment. One negative control was performed with all the reactions except primers, and another was run with distilled water to test for possible DNA contamination.

Statistics and curve fitting

Data were analyzed and fitted using pClamp (Axon Instruments) or SigmaPlot (SPSS, Chicago IL) software. Dose-response curves were fitted to a modified Hill equation Y = Yo + ACapⁿ/(K_1/2ⁿ + Capⁿ) characterized by K_1/2 (concentration of stimulus producing a half-maximal inhibition), n, the cooperative index and where the constants A + Yo = 100%. G-V and inactivation parameters were fit to the Boltzmann relation: X = C + {X_max/[1 + exp (V_0.5 - V_m/k)]}, where V_0.5 is the membrane potential at which 50% of activation or inactivation was observed, k is the slope of the function, and C is a constant (= 0 in the G-V relation). The inactivation rate constants were fit to a single exponential (Fig. 2, A and B, R²  0.97). Although fitting some I_A inactivating currents required multiple exponentials (Fig. 2C), we only analyzed rate constants that could be fit to a single exponential in which R²  0.97. The inactivation rate constants () were always taken at the maximum current, which was at the maximum depolarized potential (+50 mV).

Data were analyzed for statistical significance using the paired and unpaired (as indicated in the text), Student's t-test, and presented as the means ± SD. For the dose-response curves, the percent inhibition of the I_A currents, at the same voltage, was calculated by subtracting the inhibited current from the control. From these data, the mean of the differences between these two values gave the mean of the percentage inhibited.

Chemicals

W-7, KN-93, and KT5823 were purchased from Calbiochem (La Jolla, Ca). Unless stated otherwise, all other chemicals came from Sigma Chemical (St. Louis, MO). Cell culture materials were purchased from GIBCO (Life Technologies, Rockville, MD).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

and  subunits of VGPCs in rat trigeminal ganglion neurons

RT-PCR analysis of adult TG neurons showed the presence of number of  and  subunits of VGPCs. In the gel seen in Fig. 1A, with the exception of the K_v3 subunit, all the  and  subunits tested were present. Figure 1B shows the concentrations of the different subunits relative to the "housekeeping" gene [(G) glyceraldehyde-3-phosphate dehydrogenase]. The relative concentration of the  and  subunits were: K_v1.2 ~ K_v1.5 > K_v1.6, ~Kv 1.3 ~ K_v1.4, and K_V1 ~ K_v2 > K_v3.

View larger version (24K): [in this window] [in a new window]   Fig. 1. Potassium channel subunits in rat trigeminal ganglion (TG) neurons. Trigeminal ganglion neurons were analyzed by RT-PCR for the presence of mRNA for -subunits (Kv 1.2-1.6) and -subunits (Kv 1-3) of the named voltage-gated potassium channels. A: the gel shows the molecular weight maker on the left side and indications of the presence of the various subunits of voltage-gated potassium channels. G, glyceraldehyde-3-phosphate dehydrogenase. B: a histogram showing the mean ± SD (n = 6) of the expression levels of the different subunits with respect to that of glyceraldehyde-3-phosphate dehydrogenase (G).

I_K and I_A currents in TG neurons

TG neurons exhibit several functional types of VGPCs that exhibit distinctive kinetic responses (Fig. 2, A-C). The total potassium current was separated into I_A and I_K currents by using TEA to inhibit I_K currents. In Fig. 2A, the I_K current did not markedly inactivate, whereas the I_A current inactivated in an exponential manner. At the maximum peak current, I_Apmax, the inactivation rate constant () was 94 ms. In another type of response shown in Fig. 2B, the total outward current did not markedly inactivate because of the compensation of the slower inactivation times of I_A and the slower activation times of I_K. In a third type of response, the total outward current also did not inactivate, the I_K current activated slowly and the I_A current exhibited both a rapid and a slow inactivating component (Fig. 2C).

View larger version (33K): [in this window] [in a new window]   Fig. 2. Pharmacological separation of IA and IK currents. Neurons were held at -80 mV and stimulated from -100 to 50 mV in 10-mV intervals. Shown are the currents evoked form 3 different TG neurons. Left: total outward current; middle: the current remaining after TEA was included in the extracellular buffer (IA). The difference between the total and IA currents is the IK current (right). Three examples of the morphologies of these currents are presented. A: the total current has a large inactivating component, the IA current decays exponentially ( = 94 ms), whereas the IK current does not inactivate. B: the total current does not inactivate, the IA current is large and inactivates in an exponential manner, and the noninactivating IK current has relatively slow onset kinetics. C: the total current does not inactivate, the IA current inactivates with a rapid and slow component, and the IK component activates slowly.

I_A currents in CS and CIS neurons

Figure 3 shows a recording from a CS neuron. After obtaining an I_A current, 1 µM capsaicin evoked a slowly activating and desensitizing inward current (V_h = 80 mV). Because the external solution had the impermeant cations choline and TEA, the maximal inward current was small 1.2 ± 0.6 nA (n = 13), relative to the current obtained in a NaCl-containing buffer, -4.5 nA (Liu and Simon 1996). After applying capsaicin continuously for 1-3 min, the current desensitized to a near steady state value of 0.39 ± 0.21 nA (n = 12). At this time, compared with control values, the membrane resistance was lower, and the I_A currents were reduced 52.4 + 18.0% [from 10.4 ± 5.3 nA (control) to 5.6 ± 6.7 nA (n = 12)]. Because some of this inhibition could arise as a consequence of the capsaicin-induced leak current, we performed leak subtraction (see Fig. 3, box). This correction was calculated from the membrane resistance at -80 mV, and, with this correction, I_A was inhibited by 55.3 ± 18.2% (n = 12). However, because the I-V relation for capsaicin-induced currents is outwardly rectifying, the leak correction may not have been fully compensated at more positive potentials. To obtain a better estimate of the "true" inhibition of I_A by capsaicin, the cell was washed until the membrane resistance returned to its control value. That is before capsaicin application, the membrane resistance was 770.4 ± 129 M (n = 12) and after about a 30-s wash, it was 763 ± 137 M (n = 12). Using this criterion, the I_A currents after the washout were inhibited by 49.2 ± 17.5% (n = 12). This latter procedure was used to calculate the inhibition of I_A by capsaicin (see Fig. 6). Under these conditions, the percentage blockage is not dependent on the voltage (see I_Ap-V plots) even though the I-V relation of capsaicin-activated currents is strongly voltage dependent (Liu and Simon 2000), suggesting that the capsaicin-induced currents do not markedly contribute to the blockage. Finally, after washing these cells for another 3-6 min, the I_A currents recovered (78.7 ± 21.2% n = 12) to their precapsaicin values.

View larger version (35K): [in this window] [in a new window]   Fig. 3. Capsaicin effects on IA activation in a capsaicin-sensitive neuron. A: IA currents were generated in a 26 µm neuron whose membrane resistance was 780 M. The application of 1 µM capsaicin induced a maximum inward current of 1.7 nA. In the continued presence of capsaicin, the inward current desensitized to a value of 0.56 nA after about 60 s. At this time, the membrane resistance was 142 M and IApmax was reduced from 19.1 to 12.8 nA. After leak subtraction (in box), IApmax was 12.3 nA. On washing the cell for 20 s, the capsaicin-induced inward current returned to baseline and the membrane resistance increased to 784 M. At this time, IApmax recovered to 13.1 nA. After washing for 4 min, the IA currents completely recovered. The peak current (IAp)  voltage (V) relation is shown in absence and presence of capsaicin (with and without leak subtraction) as well as the 20-s and 4-min washes. The G-V curves for all conditions were essentially unchanged. The 20-s wash data was fit to the Boltzmann equation and yielded (control: V0.5 = 9.8 mV and k = 15.1, 1 µM capsaicin: V0.5 = 9.9 mV and k =17.1. Bottom right: 3 superimposed currents that illustrate the changes in the time course between the IApmax's (at V = 50 mV) in the absence and presence of 1 µM capsaicin (with leak subtraction). The IA current in the presence of capsaicin was normalized to the peak of the control peak IApmax. No change in activation time is produced, but the inactivation time course decreased. Bars indicate duration of capsaicin-application.

Figure 3 also shows that capsaicin did not markedly affect the time to peak (at I_Apmax) but significantly increased  (in bottom corner, compare the "control" and "normalized" traces). In the presence of 1 µM capsaicin,  the inactivation rate constant at I_Apmax, increased from 141 ± 58 to 231 ± 47 ms (n = 12, P < 0.01, paired t-test). The percentage inactivation of I_A current was estimated by the equation (I_Apmax - I_Apmax500ms)/I_Apmax, where I_Amax500ms is I_Apmax at 500 ms. In the presence of 1 µM capsaicin the percentage inactivation was reduced from 61 ± 8 to 39 ± 16% (n = 12, P < 0.01, paired t-test).

Capsaicin-insensitive neurons (CIS)

RAT TG NEURONS. In CIS neurons, the inhibitory effect of capsaicin, even at 30 µM, was small relative to CS neurons at 1 µM capsaicin. For the I_A recordings shown in Fig. 4, I_Apmax was reduced by 17.2 and 32.7% by 1 and 30 µM capsaicin, respectively. Figure 4 also shows the I-V relations at these two different concentrations. In the presence of 30 µM capsaicin, the G-V relation did not differ from the control. On average, the G-V curves were unaffected by either 1 or 30 µM capsaicin (control: V_0.5 = 3.1 ± 9.1 mV, k =18.3 ± 4.9; 1 µM capsaicin: V_0.5 = 3.1 ± 8.1 mV, k =18.5 ± 4.6; n = 6); (control: V_0.5 = 3.2 ± 4.3 mV, k =15.8 ± 1.6; 30 µM capsaicin: V_0.5 = 3.3 ± 3.7 mV, k =15.7 ± 1.4; n = 11). The maximal I_A current was reduced by 9.3 ± 6.1% (n = 5) in the presence of 1 µM capsaicin. At I_Apmax,  was not significantly altered being 89.2 ± 66.1 ms for the control and 91.4 ± 59.5 ms in the presence of 1 µM capsaicin (P = 0.52, n = 7 paired t-test).

View larger version (30K): [in this window] [in a new window]   Fig. 4. Effects of capsaicin on IA currents in capsaicin-insensitive neurons. IApmax was reduced from 12.2 nA (control) to 10.1 and 8.2 nA, on exposure to 1 and 30 µM capsaicin, respectively (see IAp-V relationships). After a 3-min wash, the current slightly recovered from exposure to 30 µM capsaicin. The G-V curve was not altered in the presence of 30 µM capsaicin (solid line, fits to the Boltzmann equation: control: V0.5 = 2.7, k =16.7; 30 µM capsaicin: V0.5 = -0.2 mV and k = 17.2). Neither the activation nor inactivation rate constants were markedly altered by 1 µM capsaicin. Bars indicate duration of capsaicin-application. Holding potential = 80 mV.

Mouse TG neurons from VR1/ mice

To eliminate the possibility that the subset of rat CIS neurons that we tested did not resemble nociceptors, we chose a population of mouse VR1/ TG neurons that have soma diameters between 15 and 22 µm. Neurons in this size range have long-duration action potentials, TTX-resistant sodium currents (unpublished observation) and other receptors that indicate that they are nociceptors (Caterina et al. 1999). The inhibitory effects of 1 and 10 µM capsaicin on I_A currents, obtained from two different neurons, are seen in Fig. 5. On average, 1 and 10 µM capsaicin inhibited I_A currents by 3.0 ± 6.9 (n = 6) and 29.7 ± 11.9 (n = 13), respectively. The G-V relationship significantly was not altered at either of these concentrations (data not shown).

View larger version (36K): [in this window] [in a new window]   Fig. 5. Effect of capsaicin on TG neurons from VR1/ mice. A: IA currents obtained from a TG neuron from a VR1/ mouse with a soma diameter of 16 µm. In this trace, IApmax was inhibited 9% by 1 µM capsaicin. B: IA currents obtained from a TG neuron from a VR1/ mouse with a soma diameter of 18 µm where 10 µM capsaicin inhibited IApmax by 11%. The inhibition was reversible after a 3-min wash.

Concentration dependence of capsaicin inhibition of I_A

Figure 6 shows the concentration dependence of capsaicin's inhibition of I_Apmax in rat CS and CIS neurons. In CS neurons, the total I_A inhibition was concentration dependent, ranging from 5.0 ± 4.6% (n = 4) at 0.1 µM to 72.0 ± 17.7% (n = 7) at 3 µM capsaicin (with 10 µM, the leak resistance was too high to compensate). In CIS neurons, the inhibition of I_A currents increased linearly from 6.4 ± 5.4% (n = 7) at 1 µM capsaicin to 24.6 ± 9.7% (n = 11) at 30 µM capsaicin. To obtain an estimate of the inhibition that can be attributed to the activation of vanilloid receptors (- - -), the CIS inhibition was subtracted from the total inhibition, and the data were fit to the Hill equation and yielded the following parameters: K_1/2 =0.67 µM, n = 1.3 and a maximal inhibition equal to 67.8%. These data show a clear distinction between the responses of these two types of neurons to capsaicin and suggest that the activation of vanilloid receptors can result in a marked inhibition of I_A currents in CS neurons.

View larger version (17K): [in this window] [in a new window]   Fig. 6. Concentration dependence of IA blockage by capsaicin. The dose-response data are plotted as the percentage inhibition by capsaicin on the maximal IA currents (IApmax) in capsaicin-sensitive () and capsaicin-insensitive () rat TG neurons. In capsaicin-sensitive neurons, the inhibition is dose dependent, whereas in capsaicin-insensitive (CIS) neurons, the inhibition is linear. To obtain the effect of the activation of vanilloid receptors in CS neurons, the mean values of these 2 plots were subtracted and fit to a Hill equation ( - ). In fitting this curve, we assumed that at 0.1 and 0.3 µM capsaicin the inhibition is 0 in CIS neurons. The fitting parameters were K1/2 =0.67 µM, n = 1.3, and the maximal inhibition was 68.7%.

Effects of capsaicin on I_A on inactivation-voltage curves

To obtain a better understanding of how capsaicin may inhibit I_A currents, we explored its effects in CS (Fig. 7A) and CIS (Fig. 7B) neurons on the inactivation-voltage parameters. Figure 7A shows the response of a CS neuron, in which 1 µM capsaicin reduced the amplitude of I_Amax by 41% and shifted the midpoint of the inactivation curve toward hyperpolarizing voltages by 17 mV. Note that for depolarizing voltages greater than 40 mV, the current decreased about 90%. In contrast, in rat CIS TG neurons, 1 µM capsaicin did not produce a change in the inactivation-voltage relation (Fig. 7B). Interestingly, in many of these neurons the current could only be maximally inactivated by about 50%. Gold et al. (1996b) saw a similar behavior in studies using DRG neurons.

View larger version (28K): [in this window] [in a new window]   Fig. 7. Effect of capsaicin on IA inactivation-voltage plots. A: capsaicin-sensitive neuron. The holding potential was -80 mV, and the inactivation-voltage protocol is shown below the control. It is seen that in the presence of 1 µM capsaicin IA amplitude was reduced the the inactivation-voltage curve was shifted by about 17 mV in the hyperpolarizing direction. The inactivation curves were fitted to a modified Boltzmann equation having the parameters: (control: V0.5 = 54.8 mV, k = 6.5, C = 0.11; 1 µM capsaicin: V0.5 = 72.1 mV, k = 9.9, C = 0.09). B: capsaicin-insensitive neuron. In this neuron, the presence of 1 µM capsaicin did not markedly affect the inactivation-voltage relationship: (control: V0.5 = 39.5 mV, k = 7.8, C = 0.56 and 1 µM capsaicin: V0.5 = 37.4 mV, k = 8.1, C =0.53). Note that a persistent current remained in both types of neurons even at very depolarizing voltages.

In CS neurons, the mean conductance-voltage (G/G_max) and inactivation-voltage relations for the I_A currents in the presence and absence of 1 µM capsaicin are shown in Fig. 8. The G-V curve was not significantly changed [P = 0.43; control: V_0.5 = 13.3 ± 14.3 mV, k =16.6 ± 2.7 mV (n = 10); 1 µM capsaicin: V_0.5 = 11.1 ± 15.4 mV, k =19.1 ± 3.5; n = 10], but the inactivation-voltage relation shifted to hyperpolarizing voltages by 14.7 mV (control: V_0.5 = 66.1 ± 4.9 mV, k = 6.0 ± 1.1, C = 0.14 ± 0.04; 1 µM capsaicin, V_0.5 = 80.8 ± 8.1 mV, k = 7.4 ± 0.4, C = 0.08 ± 0.04; n = 10). Although no significant differences were found for the parameter, k, the shift in V_0.5 and the values for C were significantly different (P < 0.05; paired t-test). For CIS neurons, the inactivation-voltage relation was not significantly different from control in the presence of capsaicin (control: V_0.5 = 51.6 ± 8.5 mV, k = 6.2 ± 1.03, C = 0.35 ± 0.15; 1 µM capsaicin: 53.9 ± 10.7 mV, k = 6.7 ± 1.04, C = 0.32 ± 0.13 (n = 5); P = 0.13, 0.36 and 0.23 for V_0.5, k, and C, respectively).

View larger version (22K): [in this window] [in a new window]   Fig. 8. IA conductance- and inactivation-voltage curves: effects of capsaicin. The solid lines shown for the mean G-V curves were fit using the following parameters: control: V0.5 = 11.8 mV, k =18.0; 1 µM capsaicin: V0.5 = 12.1 mV, k = 21.1. The solid lines for the mean inactivation-voltage curves were fit with the following parameters: control: V0.5 = 63.9 mV, k = 8.2, C = 0.11; 1 µM capsaicin, V0.5 = 79.7 mV, k = 9.1, C = 0.09). The data are given as means ± SD. The mean G-V relationship for IA currents in the presence and absence of 1 µM capsaicin were constructed by eliminating 2 neurons, in which IA was almost completely blocked because the G-V curves were hard to analyze. Consequently, 10 neurons were analyzed in which the capsaicin (1 µM) blockage was between 20 and 60%. See text for additional details.

Use-dependent block

It is well known that voltage-gated ion channels can be inhibited by a process named "use-dependent block" (Hille 1993). To test whether capsaicin blocks I_A in a use-dependent manner, the cells were held at - 80 mV and in the presence and absence of 1 µM capsaicin, a sequence of depolarizing pulses to 50 mV were delivered every second. For nine CS neurons, the ratio between the first and last of 16 pulses was 0.77 ± 0.04 to 0.74 ± 0.77 (n = 9, mean ± SD) in the absence and presence of 1 µM capsaicin (data not shown). For five CIS neurons, the ratio between the first and last of 16 pulses was 0.92 ± 0.08 to 0.90 ± 0.09 in the absence and presence of 1 µM capsaicin (data not shown). These differences are not significantly different.

Effects of cAMP, cGMP, and PDBu on I_A currents

CTP-CAMP. Because capsaicin appears to modulate I_A currents, in part through the activation of vanilloid receptors, we have explored whether intracellular pathways that are activated by capsaicin such as cAMP (Liu et al. 2001), cGMP (Wood et al. 1989), and PKC (Harvey et al. 1995) can modulate I_A currents.

View larger version (36K): [in this window] [in a new window]   Fig. 9. Effects of cAMP and phorbol-12, 13-dibuterate (PDBu) on IA currents. A: a 3-min incubation of 1 mM CTP-cAMP caused an increase in IApmax from 11.6 to 12.2 nA. B: a 3-min incubation of 500 nM PDBu produced a increase in IApmax from 9.9 to 10.3 nA. Holding potential = 80 mV.

PHORBOL-12,13-DIBUTERATE (PDBU). We next investigated the effects of 500 nM PDBu, a PKC agonist, on I_A currents. In neither CS nor CIS neurons did 500 nM PDBu alter I_A currents. As seen in Fig. 9B (in a CS neuron), a 3-min application of PDBu did not alter the amplitude of the I_A current. The G-V curve (data not shown) and activation and inactivation time courses were also not altered. On average, 500 nM PDBu reduced the amplitude of I_A current by 0.3 ± 6.4% (n = 5). PDBu (500 nM) did not alter either the G-V curves (control: V_0.5 = 4.5 ± 15.2, k =13.4 ± 2.6; 1 µM capsaicin: V₀.₅ 5.3 ± 13.0 mV, k = 14.2 ± 2.7; n = 5), or the inactivation rate constants (control: 195.5 ± 91.0; 1 µM capsaicin: 193.7 ± 83.8).

CTP-C GMP. The application of 1 mM CPT-cGMP to TG neurons reduced I_A in both CS (Fig. 10) and CIS neurons with no major differences between them. In the CS neuron shown in Fig. 10, a 3-min exposure of 1 mM CPT- cGMP reduced I_Apmax by 33% (also see I_Ap-V curve) and after 3-min wash, the current partially recovered. The application of CTP-cGMP (1 mM) did not shift the G-V curve, but the inactivation rate constant decreased from 226 to 193 ms.

View larger version (32K): [in this window] [in a new window]   Fig. 10. Effects of CTP-cGMP on IA currents. An IA current was recorded from a capsaicin-sensitive neuron (not shown) at holding potential -80 mV. After a 3-min exposure to 1 mM CPT-cAMP, IApmax was reduced from 4.5 to 3.0 nA (as indicated in the IAp-V plot). The IA current recovered after a 3-min wash. The G-V curve was unaltered by 1 mM CTP-cAMP (control: V0.5 = 0.3 mV, k = 14.5; 1mM CPT-cGMP: V0.5 = 1.3 mV, k = 15.3). CTP-cGMP did not alter the activation time course but slightly increased the inactivation rate constant from 197 to 228 ms. Bottom: 1 mM CTP-cGMP did not alter the inactivation-voltage curve (control: V0.5 = 60.4 mV, k = 7.6, C = 0.22; 1 µM capsaicin: V0.5 = 61.5 mV, k = 8.55, C = 0.22).

KT5823, a PKG inhibitor, partially reverses the inhibitory effect of capsaicin

We have shown that, by itself, capsaicin can inhibit I_A currents (Fig. 3) and have also shown that I_A currents are inhibited by cGMP (Fig. 10). Given this information, we determined whether capsaicin would be a less effective inhibitor if the cGMP-PKG pathway was inhibited. We therefore tested whether the presence of a PKG inhibitor, KT5823, would reduce the extent that capsaicin inhibits I_A currents.

We used 0.5 µM KT5823 to test capsaicin effect because it virtually abolished the effects of 8-Br-cGMP on N-type calcium channels (D'Ascenzo 2002). A representative tracing, seen in Fig. 11A, shows that a 3-min incubation of 5µM KT5823 caused a small 8.9 ± 11.7% (n = 8) increase in the I_A current. In the presence of 0.5 µM KT5823, 1 µM capsaicin activated an inward current that desensitized in a few minutes to a value <0.3 nA. At this time, I_A currents were recorded, and after wash, when the membrane resistance recovered, the I_A current was inhibited by 19%. After a 3-min wash, the I_A current almost completely recovered (Fig. 11B). Whereas, 1 µM capsaicin inhibits the total I_A by 49% (Fig. 6), in the presence of 0.5 µM KT5823, it inhibits by 24.8 ± 13.6% (n = 8, P < 0.01).

View larger version (29K): [in this window] [in a new window]   Fig. 11. Inhibitory effect of capsaicin on IA current is partially reversed by KT 5823. A: IA current was recorded in capsaicin-sensitive neuron (data not shown). After a 3-min preincubation of 0.5µM KT5823, IApmax was reduced to 10.2 nA from 10.5 nA. On the addition of 1 µM capsaicin to 0.5µM KT5823, only 18% of the IA current was blocked. The IA current recovered after 3-min wash. B: a histogram showing the percentage inhibition of IA current by 1 µM capsaicin (49.2 ± 17.5%, n = 12) and 5 µM KT5823 + 1 µM capsaicin (24.8 ± 13.6%, n = 8, mean ± SD).

Calmodulin-dependent pathways

RESTING CONDITIONS. W-7: a Ca²+-calmodulin inhibitor. Because the activation of vanilloid receptor-mediated second-messenger pathways permits the entry of calcium and because calcium associates with calmodulin (CaM), we investigated the effects of W-7 (25 µM), a calcium-calmodulin-inhibitor (Fig. 12). This concentration was chosen because it has previously been shown to be an inhibitor of VGPCs (Peretz et al. 2002). Figure 11 shows that 25 µM W-7 inhibited I_Apmax by about 30% and that the inhibition was partially reversible after a 3-min wash. On average,  was significantly decreased (control: 100 ± 44.7 ms; 25 µM W-7: 63.2 ± 36.9 ms; n = 8, P = 0.009), but the G-V curve was not altered significantly (control: V_0.5 = 6.86 ± 9.07 mV, k = 17.9 ± 4.23; 25 µM W-7: V_0.5 = 7.49 ± 9.9 mV, k = 18.32 ± 4.77; P  0.53). Finally, W-7 produced a 5-mV hyperpolarizing shift in the voltage dependence of inactivation (control: V_0.5 = 66. 31 ± 10.2 mV, k = 8.2 ± 1.3, C = 0.26 ± 0.22 25 µM W-7: V_0.5 = 71.86 ± 13.2 mV, k = 10.6 ± 2.9, C = 0.20 ± 0.14, n = 7; P = 0.0065).

View larger version (32K): [in this window] [in a new window]   Fig. 12. Modulation of IA currents by N-(6-aminohexyl)-5-chloro-1-napthalenesulfonamide HCl (W-7). An IA current was recorded from a capsaicin-sensitive neuron (not shown) that was being held at -80 mV. On a 3-min exposure to 25 µM W-7, the amplitude of IApmax was reduced from 6.7 to 5.05 nA, as indicated in the IAp-V plot. In the presence of 25 µM W-7, the G-V curve slightly shifted in a depolarizing direction (control: V0.5 = 7.9 mV, k = 16.6; 25 µM W-7: V0.5 = 3.9 mV, k = 17.9), whereas the inactivation-voltage curves were shifted 6 mV in the hyperpolarizing direction (control: V0.5 = 62.4 mV, k = 7.89, C = 0.11; 25 µM W-7: V0.5 = 68.5 mV, k = 7.85, C = 0.17). Comparison of the activation and inactivation time courses of IA in the control and with the raw and normalized induced traces, showed that W-7 decreased the inactivation rate constant from 165 to 123ms.

View larger version (30K): [in this window] [in a new window]   Fig. 13. Modulation of IA currents by KN-93. An IA current was recorded from a capsaicin-sensitive neuron (not shown) at holding potential -80 mV. After a 3-min exposure to 5 µM KN-93, the amplitude of IApmax was reduced from 8.8 (control) to 7.4 nA, as seen in the current-voltage plot. After a 3-min wash, the amplitude of IA recovered. but its inactivation rate constant did not. The presence of 5 µM KN-93 did not cause a shift in the G-V curve (control: V0.5 = 0.95 mV, k = 18.9; 5 µM KN-93: V0.5 = 0.96 mV, k = 18.1). KN-93 shifted the inactivation-voltage curves in the hyperpolarizing direction (control: V0.5 = 63.9 mV, k = 6.29, C = 0.069; KN-93: V0.5 = 70.53 mV, k = 8.95, C = 0.043). The activation time course did not change, while inactivation rate constant decreased from 183.9 to 40.1 ms in the presence of 5 µM KN-93.

STIMULATED CONDITIONS. Because the activation of vanilloid receptors increases intracellular calcium, which will then associate with calmodulin, it is expected that the CaMKII concentration would increase which could result in a decrease in I_A currents. We therefore investigated whether inhibiting CaMKII (with KN-93) would decrease capsaicin's ability to inhibit I_A currents.

View larger version (25K): [in this window] [in a new window]   Fig. 14. Inhibitory effect of capsaicin on IA current is partially reversed by the CaMKII inhibitor, KN-93. A: an IA current was recorded in a capsaicin-sensitive neuron (data not shown) that was preincubated for 3 min in 5 µM KN-93. On the addition of 1 µM capsaicin (in the presence of KN-93), IApmax decreased from 7.3 to 6.4 nA. Washing capsaicin, but not KN-93, from the cell reversed its effects on the rate constant but not the amplitude. B: a histogram showing the percentage inhibition of IA currents by 1 µM capsaicin (49.2 ± 17.5%, n = 12) and 5 µM KN-93 + 1 µM capsaicin (23.0 ± 13.5%, n = 7). **, P  0.01. HP = 80 mV.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

It is well known that nociceptors that are activated by capsaicin can become sensitized (Baumann et al. 1991; Green 1989). To determine whether I_A currents may contribute to this behavior, we have investigated how capsaicin modulates I_A currents in TG neurons. We studied I_A currents because inhibiting them, through nerve injury, or pharmacologically, will lead to hyperexcitability and hyperalgesia (Nashmi and Fehlings 2001; Pearce and Duchen 1994; Rasband et al. 2001). In CS neurons, we found that the activation of vanilloid receptors evokes intracellular cascades that will inhibit I_A currents. Although pathways involving PKC and cAMP do not appear to be involved, the activation of cGMP-PKG and CaMKII pathways contributes to capsaicin-mediated inhibition of I_A currents.

I_A currents in TG neurons

The RT-PCR experiments revealed that rat TG neurons, like DRG neurons, contain many types of  (K_v 1.2-1.6) and  (K_v1-3) VGPC subunits (Fig. 1) (Ishikawa et al. 2001; Seifert et al. 2001). In addition, TG neurons have been found to contain other VGPC subunits (K_v 1.1, 2.1, 2.2) (Seifert et al. 2001). Some  subunits alone, or when associated with  subunits give rise to characteristic I_K- or I_A-type currents (Dolly and Parcej 1996, Grissmer et al. 1994; Rasband et al. 2001; Sheih et al. 2000). The electrophysiological data reveal that TG or DRG neurons contain multiple types of functional VGPCs, including more than one type of I_A current (see Fig. 2 and Gold et al. 1996b) that can arise from presence of receptors constituted by different subunits (Rasband et al. 2001) as well as their phosphorylation state (An et al. 2000). I_A currents are present in CS neurons, suggesting that they are present in nociceptors (Fig. 3). In unmyelinated axons, I_A currents consist of K_v1.1 and/or K_v1.2 subunits that are co-assembled with K_v 1.4 subunits (Cooper et al. 1998). Also, CS neurons with TTX-r Na⁺ channels (nociceptors) contain I_A currents that have been attributed to the presence of K_v1.4 subunits (Rasband et al. 2001). In this study, we did not distinguish between the various types of I_A currents because our goal was to determine how I_A currents are modulated by capsaicin in CS and CIS neurons.

Capsaicin inhibits I_A currents

NONSPECIFIC INHIBITION. In CIS neurons, we found that the magnitude of the I_A current decreased linearly with increasing capsaicin concentration such that at 30 µM it reached 24.6% (Fig. 6). In TGs from VR1/ mice that were obtained from neurons having characteristics of nociceptors, 10 µM capsaicin inhibited I_A currents 29% (Fig. 5). Both these values are much less than the 49% inhibition by 1 µM capsaicin in CS neurons (Fig. 6). Capsaicin has previously been shown to inhibit I_A currents at K_1/2's ranging from 6 to 158 µM in a manner that depends on the cell and/or the particular subtype of receptor (Castle 1992, Grissmer et al. 1994). In all cases, except for what we found in CS neurons (Fig. 6), these values are much higher than the concentration needed to activate vanilloid receptors (0.7 µM) (Caterina et al. 1997). In many of these studies, especially those where the K_1/2 were comparatively low (6-10 µM), it is likely that some of the cells tested contained vanilloid receptors (e.g., Atkins and McCleskey 1993).

CS neurons: inhibition of I_A through the activation of vanilloid receptors

We have shown that in CS neurons that capsaicin inhibits I_A currents largely through the activation of vanilloid receptors (Figs. 3-6). For 1 µM capsaicin, 43% of the blockage of I_A currents can be attributed to the activation of vanilloid receptors (Figs. 6 and 15). That the inhibition is a consequence of the activation of vanilloid receptors is supported by the following evidence: the blockage of I_A currents in CS neurons occurs at concentrations comparable to the activation of vanilloid receptors (Fig. 6), the inhibition persists even after the decrease in resistance produced by capsaicin has recovered (Fig. 3), and at 1 µM capsaicin in small diameter VR1/ TG neurons, the inhibition is small (Fig. 5).

View larger version (13K): [in this window] [in a new window]   Fig. 15. A: modulation of IApmax by intracellular messengers. Blockages of IApmax by 1 µM capsaicin (49.2 ± 17.5%, n = 12), 1 mM CTP-cAMP (4.1 ± 5.3%; n = 6), 1 mM CTP-cGMP (32.0 ± 13.8%; n = 8), 500 nM PDBu (0.3 ± 6.4%; n = 5), 25 µM W-7 (21.6 ± 13.4%, n = 8), and 5 µM KN-93 (15.6 ± 10.6%: n = 8). - - -, the percent inhibition (43%) that can be attributed to the activation of vanilloid receptors. Data represents means ± SD.

If the channel I_A conductance is unaffected by capsaicin, then reduction in the current may be due to a decrease in the open state probability, perhaps as a consequence of the hyperpolarizing shift in the inactivation-voltage relation (Dubois 1982).

I_A currents are modulated through capsaicin-activated intracellular pathways

It is well known that potassium channels, like other voltage-gated ion channels, are modulated by intracellular compounds that regulate their degree of phosphorylation (Ishikawa et al. 2001; Nicol et al. 1997; Wickman and Clapham 1995, Zhang et al. 2001). Because the activation of vanilloid receptors activates a variety of pathways, it is to be expected that some of them would also modulate I_A currents.

A summary of the effects of all the compounds tested on I_A currents is presented in Fig. 15. The data clearly show that the ability to modulate I_A currents is very dependent on which pathways are activated. They also indicate that by themselves, none of these pathways inhibit the current to as great an extent as 1 µM capsaicin (49 or 43% attributed to the activation of vanilloid receptors). In DRGs and/or TG neurons, it has been found that the activation of vanilloid receptors induces increases in cAMP, cGMP, and PKC (Harvey et al. 1995; Liu et al. 2001, Wood et al. 1989). Consequently, we tested whether the activation of these three pathways can modulate I_A currents. Neither 1 mM CTP-cAMP nor 500 nM PDBu (a PKC activator) had a marked effect on I_A currents (Figs. 9, A and B, and 15). We therefore can eliminate these two pathways as those that contribute to capsaicin's inhibition of I_A.

The data that we have obtained provides good evidence that vanilloid receptor activation of the cGMP-PKG pathway, perhaps through the production of NO (Fazekas et al. 1994), is involved in the capsaicin-induced inhibition of I_A currents. The evidence is as follows. In CS neurons, the application of capsaicin increases intracellular Ca²⁺ (Cholewinski et al. 1993) and consequently the calcium-dependent cGMP concentration (Wood et al. 1989). We showed that the application of 1 mM CTP-cGMP to TG neurons reduced the I_A currents by 30% (Fig. 11). Another argument for a role of the cGMP-PKG pathway is that the ability of capsaicin to inhibit I_A currents was significantly reduced in the presence of the PKG antagonist KT5823 (Fig. 11). In addition, I_A currents were increased by KT5823. Finally, in the absence of external stimuli, W-7, a CaM inhibitor, also decreased I_A currents. The inhibitory effect of W-7 may, at least partially be attributed to its ability to increase cGMP concentrations (Parfenova et al. 1993).

However, it is evident that by itself, the cGMP-PKG pathway cannot account for the magnitude of capsaicin-mediated inhibition or the 16 mV shift in the inactivation-voltage curve (Fig. 8). This suggests that other mechanisms (pathways) must be contributing to the inhibitory effect of capsaicin on the I_A currents.

The other obvious pathway that we explored is the CaMCaMKII pathway. Because the activation of vanilloid receptors permits the entry of calcium (Cholewinski et al. 1993) and because calcium associates with calmodulin (CaM), the activation of VR1 receptors should increase the CaMKII concentrations. If a direct connection between the activation of VR1 receptors and increases in CaMKII was present, then decreasing the CaMKII concentration with KN-93 would result in an increase in I_A currents. This is exactly what was found (Fig. 15). That is, by itself 1 µM capsaicin inhibits I_A 49% and in the presence of 5 µM KN-93, capsaicin inhibited I_A currents by 23%. These data sugg