INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

All living organisms, including worms, exhibit avoidance behavior to prevent injury when exposed to physical or chemical stimuli of potentially damaging intensity (Kumazawa 1996; Sambongi et al. 2000; Wittenburg and Baumeister 1999). Extensive research has examined the transduction mechanisms underlying noxious stimulation in mammalian sensory neurons (see Kress and Reeh 1996; Raja et al. 1999); however, for submammalian species this information is scarce.

In mammals, capsaicin, the pungent agent of red pepper that produces burning pain in humans, induces a nonselective cationic current (I_CAPS) in rat dorsal root ganglion (DRG) neurons (Bevan and Docherty 1993; Wood et al. 2000). I_CAPS is facilitated by acids (Petersen and LaMotte 1993) and undergoes a calcium-dependent tachyphylaxis after repeated capsaicin application (Docherty et al. 1996; Koplas et al. 1997). Noxious heat (43-52°C) induces a similar cationic current (I_HEAT) that can be greatly facilitated by activation of protein kinase C with bradykinin or phorbolesters (Cesare and McNaughton 1996) in the same type of DRG neurons that exhibit sensitivity to capsaicin (Kirschstein et al. 1997).

Identification of the molecular structure of the vanilloid (capsaicin) receptor (VR1) isolated from rat DRG neurons revealed that this receptor is composed of 838 amino acids with six transmembrane domains with a pore forming loop between TM 5 and 6 (Caterina et al. 1997). Apparently four subunits of VR1 are needed to constitute a homomeric tetramer that represents a functioning channel (Jahnel et al. 2001; Kedei et al. 2001). After transfection into oocytes or HEK293 cells, VR1 can be activated by capsaicin, noxious heat (over 43°C) and lowering extracellular pH to 5 (Caterina et al. 1997; Tominaga et al. 1998). This suggested that VR1 is the common transducer for pain-producing stimuli. However, it soon became evident that VR1 is unlikely to be the sole sensor for noxious heat because studies on VR1 knockout mice revealed that the VR1^-/- variant, albeit exhibiting a complete loss of pain behavior when exposed to capsaicin, remained sensitive to noxious heat (Caterina et al. 2000; Davis et al. 2000).

Submammalian species appear to be insensitive to capsaicin (Szolcsanyi 1991), even though they exhibit avoidance to other potentially damaging stimuli, including noxious heat. It has already been demonstrated that noxious heat induces low-threshold membrane currents in DRG neurons isolated from the chick that are insensitive to capsaicin, however, can be blocked by capsaicin antagonists (Marin-Burgin et al. 2000; Nagy and Rang 2000). Recently it has been shown that the predicted protein sequence of the chick heat sensor exhibits 68% identity and 79% similarity to rat VR1 (Jordt and Julius 2002).

The aim of this study was to analyze the cellular mechanisms that underlie the defense reactions induced by noxious heat and acids in DRG neurons isolated from the adult frog, a species that exhibits no sensitivity to capsaicin.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cell cultures

Adult male frogs (Rana pipiens) were killed by decapitation. DRG were dissociated, and the neurons plated as described previously (Philippi et al. 1995). Briefly, DRG were isolated, their connective capsule cut away, and placed in a siliconized glass dish (Sigmacote, Sigma). The DRG were cut into small pieces, treated with collagenase P (3 mg/ml, Boehringer-Mannheim), neutral dispase II (8 mg/ml, Boehringer-Mannheim), and DNase (0.5 mg/ml, Boehringer-Mannheim), in Liebowitz-15 (L-15, Gibco) tissue culture medium (diluted 10 parts L-15 + 3 parts water), containing garamycin (10 mg/l) for 1 h, after which the pieces were triturated to complete dissociation. The neurons were picked up in a siliconized micropipette with a fire polished tip with an opening of ~100 µm, and plated on a poly-[l]-lysine coated glass coverslip (1-h incubation) in L-15 medium. The neurons adhered to the coverslip immediately on contact. The neurons in culture medium were left in ambient air and temperature (23°C) and recordings were made 1-2 days after plating. All experiments were carried out in accordance with the European Community's Council Directive (86/609/EEC) and with the approval of the Institutional Animal Care and Use Committee.

Recording and perfusion techniques

The single-electrode patch-clamp technique was used to record whole cell membrane currents using an Axopatch 1D preamplifier, and pCLAMP 8 programs (Axon Instruments) with a laboratory PC for storing and evaluating the data. Electrodes were pulled from borosilicate glass and were not fire polished. After filling they had a resistance of 4-5 M. The series resistance was <10 M and was compensated to ~80%.

For drug application, a system for fast superfusion of the neurons was used. It consisted of a manifold of seven fused silica capillaries (0.36 mm ID; Composite Metal Services) connected to a common outlet made from a glass capillary coiled with insulated copper wire (20 µm OD) that was used to pass DC current for heating the solutions superfusing the neuron (Dittert et al. 1998). Each of the seven tubes was connected to a reservoir containing a solution separated from the tube by a valve that was closed under resting conditions. A microprocessor controlled the solenoid valves (General Valve) and heating of the coil. The orifice of the outlet capillary was placed <100 µm in front of the soma of the neuron to be studied. Before and after application of any of the test solutions, neurons were superfused with control extracellular solution (ECS) that contained (in mM) 120 NaCl, 1.9 KCl, 0.75 CaCl₂, 1.5 MgCl₂, 7.5 HEPES, and 7.5 glucose; pH was adjusted to 7.3 with NaOH. In acidic ECS, 2-(N-morpholino)ethanesulphonic acid (MES) was used instead of HEPES and the pH was adjusted as indicated. The osmolarity was adjusted to 250 mosM. The intracellular solution (ICS) contained (in mM) 105 KCl, 0.4 CaCl₂, 1.5 MgCl₂, 4 EGTA, 7.5 HEPES, and 2 Mg-ATP, and pH was adjusted to 7.3 with KOH. The osmolarity was 220 mosM. In experiments to determine the reversal of the membrane currents induced by heat, the ICS was changed to: 94 gluconic acid delta-lactone, 11 CsCl, 4 EGTA, 7.5 HEPES, and 0.4 CaCl₂; pH adjusted to 7.3 with CsOH. Capsaicin (CAPS) was dissolved in 100 µl 96% ethanol, diluted with distilled water to 1 mM and then used to prepare the test solutions by adding ECS. All drugs were purchased from Sigma Chemical. Data are given as means ± SE. Heat ramps had a 3-s duration.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Behavioral experiments

We confirmed the results of earlier studies showing that frogs are completely insensitive to capsaicin (Szolcsanyi 1991). Frogs exhibited no signs of pain behavior when capsaicin was dropped into the eyes at a concentration of 100 µM (n = 5).

To determine the temperature threshold for inducing the leg withdrawal reflex, pithed frogs were hung with their feet submerged in water while it was gradually heated. The threshold for the withdrawal reflex was 38 ± 0.5 °C (n = 6). The pH threshold for inducing wiping reflexes at room temperature (23°C) was between pH 2 and 1.5, with pH 1 inducing a violent wiping. The wiping reflex was abolished immediately by washing the feet with a solution at neutral pH (n = 5).

Responses to noxious heat in DRG neurons

DRG neurons with the size between 18 and 25 µm in diameter were used for recording. They exhibited resting membrane potentials (r.m.p.) between 50 and 86 mV when measured in current-clamp mode with electrodes filled with a solution containing K⁺ as the predominant cation. Figure 1A demonstrates an example of the membrane current in a frog DRG neuron induced by a 3-s ramp of rising temperature. I_HEAT exhibited no detectable resting membrane current at 70 mV and reached the amplitude of 1.6 nA at 47°C. The temperature coefficient (Q₁₀), calculated from the Arrhenius plot in the range between 41 and 47°C, was 3.6 in this cell (Fig. 1B).

View larger version (13K): [in this window] [in a new window]   Fig. 1. Whole cell membrane current induced by noxious heat in a cultured frog dorsal root ganglion (DRG) neuron. A: whole cell membrane current induced by a 3-s ramp of temperature increase to 47°C in a DRG neuron clamped at 70 mV. Top: the temperature measured by a miniature thermocouple in the orifice of the outlet capillary of the superfusing system. B: Arrhenius plot in which the current normalized to that at 24°C (ordinate, log scale) was plotted against the reciprocal of the absolute temperature (abscissa). The slope exhibits Q10 = 3.6. C: superimposed current responses from representative 28 DRG neurons normalized to 45°C are plotted against temperature.

We detected I_HEAT in 78 of the 82 neurons; however, a large variability in the magnitude size was observed (490 ± 70 pA at 70 mV, at 47°C, n = 78). In contrast to the rat DRG neurons, in which I_HEAT rises steeply when the temperature is elevated above ~43°C (Cesare and McNaughton 1996; Kirschstein et al. 1997; Vyklický et al. 1999), we found it more difficult to estimate the threshold temperature for inducing I_HEAT in frog DRG neurons. To demonstrate the variability and the limited accuracy of measurements of the temperature threshold of I_HEAT in frog DRG neurons, superimposed responses from 28 representative neurons normalized to 45°C are plotted against temperature in Fig. 1C. Nevertheless, we attempted to estimate the temperature threshold from the Arrhenius plot as the point at which the heat-induced current began to deviate from linearity (Fig. 1B) or from the current-temperature plot, as the point at which there was a detectable deviance from the zero resting membrane current. Allowing for the limited accuracy of the measurements, the mean threshold estimated from the Arrhenius plot (n = 15) and from the current temperature relationship (n = 71) was close to 38°C. The mean Q₁₀ of the I_HEAT was 8.7 ± 2.4 between 41 and 47°C (n = 15).

We examined whether the membrane currents induced by noxious heat in the frog DRG neurons can be influenced by agents that have been shown to block or facilitate membrane currents induced by noxious heat in the rat DRG neurons or in VR1-transfected HEK 293 cells.

Although capsaicin (5 µM) facilitates membrane currents induced by noxious heat in rat DRG neurons (Kirschstein et al. 1997; Vlachova et al. 2001) and in VR1-transfected HEK 293 cells (Tominaga et al. 1998), it does not increase I_HEAT for frog DRG neurons (Fig. 2A). Both capsazepine (10 µM), the competitive antagonist at the capsaicin receptor (Bevan et al. 1992), and ruthenium red (10 µM), the noncompetitive antagonist at the capsaicin receptor (Dray et al. 1990), produced no significant change of I_HEAT (Fig. 2, B and C). The control I_HEAT in Fig. 2B was recorded immediately before and after testing capsazepine. The average ratio of the heat-induced current in the presence of capsaicin, capsazepine, and ruthenium red as compared with the control heat response was 1.1 ± 0.1 (n = 9), 1.1 ± 0.1 (n = 9), and 1.3 ± 0.4 (n = 6) at 47°C, respectively. These values were not significantly different from 1 (paired t-test, P > 0.05).

View larger version (9K): [in this window] [in a new window]   Fig. 2. The effects of capsaicin, capsazepine, and ruthenium red on whole cell membrane current induced by noxious heat. A: IHEAT in the control extracellular solution (ECS) and in the presence of capsaicin (CAPS, 5 µM). B: IHEAT in the control ECS immediately before application (left) of capsazepine (CZP, 10 µM; middle). The time of capsazepine application is indicated by the bar above the record. IHEAT in the control ECS after application of capsazepine (right). C: ruthenium red (RR, 10 µM) and wash in control ECS. The cell was clamped to 70 mV. The dashed lines indicate 0 membrane current. The trace above the 1st record in A shows the ramp of temperature increase to 49°C used in all records. The bars indicate the time of drug application.

To determine the reversal potential of I_HEAT in frog DRG neurons, voltage ramps were used according to the protocol demonstrated in the Fig. 3A. The holding potential of 70 mV was switched to +80 mV, and during a 1-s ramp, the cell was polarized to 80 mV at room temperature or during a 5-s step of increased temperature to 45°C. To minimize interference from voltage-gated currents, 0.5 mM CdCl₂ and 2 µM TTX was added to the ECS, and the electrodes were filled with a solution containing gluconic acid delta-lactone and cesium as the prevalent ion. In 4 of 20 neurons, we were able to obtain smooth records without a hump due to activation of voltage-gated channels that enabled us to construct current-voltage relationships shown in Fig. 3. The responses at room temperature (Fig. 3A) were subtracted from the responses induced during the 5-s steps of increased temperature to (Fig. 3B) to plot the current-voltage relationship (Fig. 3C). I_HEAT exhibited reversal close to 10 mV and a pronounced outward rectification, similar to that in the rat DRG neurons, suggesting that I_HEAT is carried through nonselective cation channels (Cesare and McNaughton 1996).

View larger version (13K): [in this window] [in a new window]   Fig. 3. Reversal of the membrane current induced by noxious heat. A: response of a frog DRG neuron clamped at 70 mV to a voltage ramp at room temperature (25°C). The protocol of the voltage ramp is shown above the record. The cell was depolarized from 70 to +80 mV and repolarized to 80 mV during a ramp of 1-s duration. The electrode was filled with cesium-delta-lactone. B: the response to the same voltage ramp when the neuron was exposed to a 5-s step of heat (45°C). C: current-voltage relationship of the heat induced membrane current (the current induced by the voltage ramp at room temperature was subtracted).

Effects of acids on frog DRG neurons

Acids are well-recognized algogens. In small and medium rat DRG neurons, reducing extracellular pH to 6 produces a sustained membrane current carried nonselectively by cations that is likely to generate spike activity inducing pain (Bevan and Yeats 1991). In frog DRG neurons, reducing the extracellular pH to 5.0 produced a huge depolarizing slowly inactivating membrane current (I_ACID) in 42 of 45 neurons examined with an average peak amplitude of 7.7 ± 0.7 nA.

Figure 4A shows a neuron with a r.m.p. of 86 mV in which the extracellular solution at pH 5.0 produced a sustained depolarization with an overshoot to +20 mV. A 3-s ramp of elevating temperature 38°C induced a small increase in this depolarization. However, when the temperature was elevated >40°C (arrow), there was a marked inhibition of this proton-induced depolarization. Washing the neuron with ECS at pH 7.3 rapidly brought the r.m.p. to 80 mV. To illustrate the effects of the same heat ramp at pH 7.3, control V_HEAT is superimposed in this record (control). Figure 4B shows the membrane current induced by extracellular pH 5.0 in the same cell clamped at 70 mV. The inward current (7.5 nA) exhibited a fast raising phase (time constant, ~200 ms) and a very slow time constant of inactivation. A ramp of elevating temperature 38°C resulted in a small increase of I_ACID, while its profound inhibition was observed when the temperature exceeded 40°C (arrow). Responses produced by the same heat ramp at pH 7.3 are superimposed in the records in Fig. 4 (control). The average inhibition of I_ACID produced by increased temperature to 46°C was 82 ± 2% (n = 42).

View larger version (16K): [in this window] [in a new window]   Fig. 4. The effects of noxious heat on depolarization and membrane current induced by pH 5 in a DRG neuron of the frog. A: depolarization of the DRG neuron evoked by acidic pH 5. The neuron was superfused for the time indicated by the bar and the heat ramp to 46°C is shown above the record. Observe the spike on the initial phase of the depolarization produced by acidic pH. B: membrane current induced by pH 5 in the same neuron clamped at 70 mV. , the threshold temperature for the inhibition of the responses. The responses to the same heat ramps at pH 7.3 are superimposed to the records at the identical scale (control).

The overshoot of the depolarization shown in Fig. 4A suggested that I_ACID is carried by Na⁺. To confirm this, we attempted to reverse I_ACID using the same protocol as for I_HEAT; however, we were unable to reverse it even at high positive membrane potentials (+80 mV). Therefore we substituted cesium for sodium in the extracellular solution to determine whether cesium can pass through the channels involved in the generation of I_ACID. Figure 5A shows I_HEAT induced by a ramp of increasing temperature to 47°C in control ECS at pH 7.3 that reached 1.5 nA. Substitution of extracellular Cs⁺ for Na⁺ resulted in a small resting inward current (~200 pA), and the heat ramp produced I_HEAT similar to that in control ECS (Fig. 5B). In the presence of extracellular Na⁺, superfusion of the neuron with ECS at pH 6.1 resulted in a fast inactivating and a slowly inactivating current with I_HEAT superimposed (Fig. 5C). When Cs⁺ was substituted for Na⁺ (Fig. 5D), the acid-induced currents were completely abolished, while I_HEAT was not significantly influenced. These results indicate that cesium does not pass through the channels involved in generating I_ACID, while it passes freely through the channels activated by noxious heat.

View larger version (18K): [in this window] [in a new window]   Fig. 5. The effects of substitution of extracellular sodium by cesium on the membrane currents induced by low pH and noxious heat. A: membrane current induced by a ramp of increasing temperature to 47°C in control ECS at pH 7.3. B: membrane current induced by the ramp of increasing temperature when equimolar CsCl was substituted for NaCl in superfusing solution. C: membrane current induced by pH 6.1 and a ramp of increasing temperature to 47°C in ECS containing Na+. D: membrane current induced by pH 6.1 and a ramp of increasing temperature to 47°C in ECS in which CsCl was substituted for NaCl. , the threshold for membrane currents induced by noxious heat.

The magnitude of I_ACID in frog DRG neurons depends on the proton concentration. Figure 6A shows a continuous record of the membrane current from a cell clamped at 70 mV when pH of the superfusing solution was decreased at 6-s intervals from 7.3, to 6.8, 5.74, 5.46, and 5. Acid (pH) dose-response curves (Fig. 6B) were constructed from the responses normalized to the current induced by pH 5.0 (n = 6). The estimated mean pH₅₀ was 5.65 ± 0.04 and the Hill coefficient 2.3 ± 0.6. 

View larger version (12K): [in this window] [in a new window]   Fig. 6. pH dependence of the sustained proton-induced membrane current in frog DRG neurons. A: membrane currents induced in a frog DRG neuron superfused by ECS of decreasing pH from 7.3 to 6.8, 5.7, 5.46, and 5 in steps of 6-s duration. B: dose-response curve constructed from 6 experiments similar to that shown in A. Responses were normalized to pH 5 and fitted with a logistic equation yielding a mean pH50 5.65 ± 0.04 and the Hill coefficient of 2.3 ± 0.6.

The large slowly inactivating proton-induced currents were profoundly inhibited by a high concentration of amiloride (1 mM; Fig. 7). Figure 7A shows I_ACID induced by extracellular application of pH 5 for 7 s indicated by an open bar above the record. Amiloride (1 mM) at pH 5 applied for 2 s at room temperature inhibited I_ACID to ~20%. Comparing the records of I_ACID in the control (B) and in the presence of 1 mM amiloride (C), it can be seen that amiloride profoundly inhibited I_ACID both at room temperature and during its increase in elevating temperature. The effect of amiloride was fast reversible (D). The average inhibition of pH 5.0 induced peak responses by 1 mM amiloride was 78 ± 4% (n = 14). I_ACID was completely insensitive to TTX (2 µM), Cd²⁺ (0.5 mM), capsaicin (5 µM), capsazepine (10 µM), and ruthenium red (10 µM; not demonstrated).

View larger version (9K): [in this window] [in a new window]   Fig. 7. The effects of amiloride on the sustained proton-induced membrane current. A: pH 5 and amiloride 1 mM were applied to a DRG neuron as indicated by the bars above the record. B: control membrane current induced by pH 5 exposed to a ramp of temperature increase to 47°C (shown below the records). C: membrane current induced by pH 5 exposed to a temperature increase to 47°C in the presence of 1 mM amiloride. D: recovery. The bars above the records show the time of drug application. Membrane potential 70 mV and vertical scale 2 nA apply for A. Membrane potential 40 mV and 5 nA apply for B-D. A and B and C, 2 different cells.

Figures 4 and 7 show that elevating temperature in the innocuous range 40°C increases I_ACID, while even a short-lasting exposure to the noxious range >40°C results in its profound inhibition. We analyzed this finding in more detail. Figure 8A shows I_HEAT induced by a ramp of temperature increasing to 48°C at pH 7.3. The records in Fig. 8, B-D, show I_HEAT superimposed on I_ACID induced by pH 6.5, 6.1, and 5, respectively. Between D and E pH 5 was continuously applied, and the record in E shows that I_ACID remained profoundly inhibited after the 30-s interval of continuous exposure to pH 5. Figure 8F demonstrates that after completing this procedure, the cell exhibited a r.m.p. 67 mV and that the effects of pH 5 and heat were completely reversible. This finding suggests that either the noxious temperature itself converts the proton-activated channels from the open to the closed state or that some of the second-messenger systems control activity of the ion channels from the intracellular side.

View larger version (19K): [in this window] [in a new window]   Fig. 8. The effect of noxious heat on IACID in a DRG neuron of the frog. A: IHEAT induced by a 3-s heat ramp to 48°C in a DRG neuron of the frog at pH 7.3 clamped at 70 mV. The temperature measured with the thermocouple in the orifice of the common outlet capillary placed in front of the neuron is shown above the records. B: IHEAT at pH 6.5. The time of low pH applications are indicated by the bars above the records. C: IHEAT at pH 6.1. D: IHEAT at pH 5. Observe the change of the vertical scale. Acidic pH was applied continuously for 30 s between D and E, after which the 2nd heat ramp was applied. F: depolarization of the same neuron produced by pH 5 to which a heat ramp was superimposed (recorded in current-clamp mode 30 s later). , the temperature 40°C.

To explore whether intracellular messengers control I_ACID in frog DRG neurons, we tested the effects of phorbol ester myristate acetate (PMA), the membrane-permeable activator of protein kinase C that greatly facilitates I_HEAT in rat DRG neurons (Cesare et al. 1999) or forskolin, the activator of protein kinase A (Kress and Zeilhofer 1999; Kress et al. 1996). Figure 9 demonstrates that PMA (1 µM) markedly inhibits I_ACID produced by extracellular pH 5 over the temperature range 23-47°C. Figure 9A shows control I_HEAT induced by a 3-s ramp of increasing temperature to 47°C. The effects of PMA on I_ACID to which I_HEAT is superimposed are shown in D. The control records of I_ACID with I_HEAT superimposed and after wash are shown in C and D. Figure 9E demonstrates the effects of PMA on the membrane current induced by pH 5 at room temperature. Forskolin (1 µM) also inhibited the proton induced current, although to a lesser extent (not demonstrated). PMA inhibited the proton induced current to 37 ± 5% (n = 5) and forskolin to 78 ± 8% (n = 4). Neither PMA nor forskolin significantly influenced the magnitude of I_HEAT or the depolarization produced by noxious heat (not demonstrated). The effects of increasing temperature on I_ACID were not influenced by staurosporine (250 nM), the nonspecific inhibitor of protein kinases (not demonstrated).

View larger version (10K): [in this window] [in a new window]   Fig. 9. The effects of Phorbol ester myristate acetate (PMA) on the responses induced by pH 5 in a DRG neuron from the frog. A: control membrane current induced at pH 7.3 by a 3-s ramp of increasing temperature to 45°C indicated above the record. B: membrane current induced by pH 5 at room temperature and during a 3-s ramp of increasing temperature to 47°C. , the temperature 40°C. C: membrane current induced by pH 5 at room temperature and during increasing temperature in the presence of 1 µM PMA. D: membrane current induced by pH 5 at room temperature and during increasing temperature, 30 s after testing the effects of PMA. E: profound inhibition of proton induced membrane current induced by pH 5 in the presence of PMA (1 µM) at room temperature.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our results demonstrate that noxious heat applied to frog DRG neurons produces membrane current that exhibits similar properties to that found in mammalian primary sensory neurons. However, contrary to mammalian primary sensory neurons, this current is not influenced by capsaicin and its antagonists. Acids induce a large slowly inactivating membrane current in frog DRG neurons that is selectively carried by Na⁺ and that is profoundly inhibited by noxious heat.

Noxious heat

We found I_HEAT in a majority of small- to medium-sized frog DRG neurons similar to what is found for rat DRG neurons that also exhibit sensitivity to vanilloids (Cesare and McNaughton 1996; Kirschstein et al. 1997; Vyklický et al. 1999). For both frog and rat neurons, I_HEAT exhibits a strong outward rectification with reversal at a slightly positive membrane potential, suggesting the activation of nonselective cation channels (Cesare and McNaughton 1996).

In contrast to rat DRG neurons in which the temperature threshold is uniformly 43-44°C (Cesare and McNaughton 1996; Kirschstein et al. 1997; Vyklický et al. 1999), the temperature threshold for inducing I_HEAT in frog DRG neurons exhibits a large scatter with the mean of ~38°C. The most striking difference between the I_HEAT in rat and frog DRG neurons is the sensitivity to drugs that block or facilitate activity of the VR1 receptor. I_HEAT in rat DRG neurons and in VR1-transfected HEK cells can be blocked by capsazepine and ruthenium red, which act on the capsaicin receptor as competitive and noncompetitive antagonist, respectively (Bevan et al. 1992; Caterina et al. 1997; Dray et al. 1990; Tominaga et al. 1998). For frog DRG neurons, these drugs are completely ineffective (Fig. 3). Capsaicin that dramatically facilitates I_HEAT in the rat (Caterina et al. 1997; Tominaga et al. 1998; Vlachova et al. 2001) does not produce any membrane current in frog DRG neurons at any temperature. This suggests that the molecular structure of the heat sensor in the frog is lacking the domain for binding vanilloids.

The vanilloid receptor, VR1, identified in rat primary nociceptors and analyzed in heterologously expressing systems, plays the major role in generating I_HEAT (Caterina et al. 1997). It is generally agreed that VR1 serves the role of the polymodal nociceptor because it is sensitive to capsaicin, acids, and noxious heat (Caterina and Julius 2001; Tominaga et al. 1998). The sensitivity of VR1 can be influenced by protein kinase C from the intracellular side (Caterina 2001; Vellani et al. 2001), while the reducing agent dithiothreitol increases its sensitivity from the extracellular side (Vyklický et al. 2002).

In contrast to mammals, the molecular structure of the heat sensor in the frog is not known. It seems unlikely that VRL1, a heat-sensitive VR1 variant, which is not sensitive to capsaicin (Caterina et al. 1999), could be the heat sensor in the frog because its temperature threshold for activation is much higher (>50°C). It also seems unlikely that the heat receptor in chick DRG neurons (Jordt and Julius 2002; Marin-Burgin et al. 2000; Nagy and Rang 2000) could play the same role in the frog. Chick belongs to a species that is not sensitive to capsaicin (Szolcsanyi 1991), although it exhibits an escape response to heat of a potentially damaging intensity (Hughes and Sufka 1990). Chick DRG neurons exposed to noxious heat exhibit low-threshold I_HEAT that, unlike what is seen in the frog, can be blocked by capsazepine and ruthenium red (Marin-Burgin et al. 2000). The molecular structure of the heat sensor in birds has been recently analyzed and shown to be a vanilloid-insensitive homologue of rat VR1 exhibiting 68% identity and 79% similarity (Jordt and Julius 2002).

The present results support the idea that the frog heat sensor represents an ortholog of VR1; however, identification of the primary structure is needed to judge the extent of the homology. It cannot also be excluded that this heat receptor is closely related to that in DRG neurons in VR1 knockout mice that lack sensitivity to capsaicin yet still exhibit an escape reaction to noxious heat similar to that observed in intact animals (Caterina et al. 2000; Davis et al. 2000).

Acids

Our results suggest that the large proton-induced membrane current observed in the majority of the frog DRG neurons underlies generation of the afferent activity that induces wiping reflexes in vivo. This I_ACID is a slowly inactivating inward current that exhibits pH₅₀ ~5.7 and at extracellular pH 5 frequently exceeds 10 nA at 70 mV. This pH might seem too high for evoking wiping because pH 2-1 was needed to be applied on the skin in vivo. This discrepancy can be explained by rich blood circulation in frog skin that may greatly dilute the proton build-up, thus preventing it from reaching a concentration that would activate peripheral endings of the afferent fibers.

It has been shown that low pH ~5 induces appreciable membrane currents in small DRG neurons in the rat (Bevan and Yeats 1991) and in rat VR1- or human VR1-transfected into oocytes or HEK293 cells by opening nonselective cation channels (Caterina et al. 1997; Hayes et al. 2000; Tominaga et al. 1998). However, our results present evidence that in the frog, I_ACID is selectively carried by Na⁺. First, with the intracellular solution used for filling the electrode containing nominally zero Na⁺, we were unable to reverse the low-pH-induced membrane current at high positive membrane potential. Second, substitution of cesium for sodium in ECS completely blocked I_ACID without affecting I_HEAT (Fig. 5). Third, depolarization produced by pH 5 frequently exhibited an overshoot to a positive membrane potential (Fig. 4A). We demonstrate that the channels involved in generation of I_ACID in frog DRG neurons can be blocked to ~20% with amiloride (1 mM); however, they are completely insensitive to TTX (2 µM). The channels may belong to H⁺-gated cation channels of the NaC/DEG family (Waldmann and Lazdunski 1998); however, more evidence about the molecular structure is needed to classify them exactly.

We found I_ACID coexpressed with I_HEAT; however, no correlation was detected in the respect of their magnitudes. We also frequently observed that the large, slowly inactivating I_ACID was coexpressed with the fast inactivating currents induced at less acidic pH (see Fig. 5C) (Davies et al. 1988; Krishtal and Pidoplichko 1981). However, the fast-inactivating proton-induced currents are unlikely to be responsible for wiping that lasts as long as the acidic solution is applied to the skin.

The most striking observation is the effects of elevated temperature on I_ACID that were invariably observed. In the innocuous range, the magnitude of I_ACID increased in parallel with elevating temperature up to ~38°C. This can be understood because in general ion channels increase their open frequency if the temperature is elevated (Hille 1992). However, the finding that even a short-lasting application of noxious heat (>40°C) profoundly inhibits I_ACID (Fig. 8) is striking and to our knowledge not yet observed. The inhibition of the proton-induced current outlasts the duration of noxious temperature for 30 s; however, it is fully reversible and repeatable in the same neurons. It can be speculated that activation of intracellular messenger systems can convert the proton-activated channels from an active to an inactive state, e.g., by phosphorylation or dephosphorylation. This idea is congruent with our finding that PMA profoundly, and forskolin to a smaller extent, inhibit the sustained proton-induced current (Fig. 9). These effects are opposite to that observed in rat DRG neurons in which PMA dramatically increases I_HEAT it and this effect can be blocked by staurosporine, the nonspecific inhibitor of protein kinases (Cesare and McNaughton 1996; Vellani et al. 2001). However, in our experiments on frog DRG neurons, the inhibition of I_ACID produced by noxious heat was not blocked by 250 nM staurosporine. This suggests that the mechanisms that underlie inhibitory effects of noxious heat on I_ACID are more complex and warrant further study.

We conclude that the cellular mechanisms underlying defense reactions induced by noxious heat in frogs are distinct from those in mammals. While a single protein structure VR1 plays the role of polymodal detector for noxious stimuli produced by heat and acids in mammals, in frog these functions are subserved by distinct ion channels.