Background
Yaksh et al. [
1] first reported that an intrathecal injection of capsaicin (CAP) depletes substance P from primary sensory neurons and causes a prolonged increase in the thermal and chemical, but not mechanical, pain thresholds. Since then, a number of studies have shown that CAP produces analgesic action at the spinal level [
2‐
4]. However, as reported by Nagy et al. [
5], the depletion of spinal cord substance P by CAP may not be sufficient to explain the changes induced by CAP.
Recent studies in peripheral axons revealed that CAP acts on vanilloid receptors 1 (TRPV1 channels) that are present on fine afferent fibers [
6]. The receptor channels are activated by painful heat stimuli (>43°C), as well as noxious chemical stimuli [
6,
7]. The flow of Ca
++ and Na
+ through the channels induces depolarizing receptor potentials at the peripheral terminals, which generate action potentials that transmit pain information [
6]. Following such primary excitatory actions, treatment with CAP often produces secondary inhibitory actions. The application of TRPV1 channel agonists at afferent fiber bundles blocks conduction of the action potential [
8‐
11]. Adenosine triphosphate (ATP) also affects TRPV1 channels indirectly through the activation of G-protein, and sensitizes them so as to be activated at a lower temperature [
7]. In the superficial layers (laminae I-II) of the dorsal horn, TRPV1 channels are abundant, especially at presynaptic elements of unmyelinated C-fiber origins [
12]. However, their roles in afferent signal transmission are unclear.
We have been recording neuronal excitation in central slices by using an optical method with a voltage-sensitive dye [
13,
14], and investigated the neuronal circuitry and plasticity in spinal cord slices [
15,
16]. Recently, we succeeded in staining primary afferent fibers anterogradely from the dorsal root, and optically recorded the presynaptic excitation exclusively [
17]. Utilizing these two types of optical imaging methods, the visualization of presynaptic excitation and net excitation of pre- and postsynaptic elements, we investigated the mechanism of CAP action in the spinal cord. In addition, we examined the temperature dependence of CAP activity, because, unlike other brain regions such as the olfactory bulb and the hippocampus, we noticed that the magnitude of optical signal, especially in the superficial dorsal horn, changes depending on the perfusate temperature. Further, we speculated that TRPV1 channels contribute to this temperature dependence. In this study, we report that CAP acts on TRPV1 channels, most likely at presynaptic elements, that are consistently sensitized via P2Y receptors, and, therefore, that CAP can inhibit the afferent signal transmission even at physiological temperature.
Discussion
In the present study, we showed that CAP, anandamide and an increase in perfusate temperature inhibited the net excitation of pre and postsynaptic elements and the presynaptic excitation evoked by dorsal root stimulation in the superficial dorsal horn. These inhibitions were Ca++-dependent, antagonized by a TRPV1 channel antagonist and a P2X and P2Y antagonist, and facilitated by a P2Y agonist, but not by a P2X agonist. In addition, the inhibitions were unaffected by antagonists of GABAA and glycine receptors.
Several studies have shown that an intrathecal injection of CAP induces analgesic action at the spinal level [
1‐
5]. In addition, electrophysiological studies have revealed that CAP suppresses C-fiber-evoked synaptic transmission in the spinal dorsal horn [
19,
20]. A bath application of CAP for 1 min produces a large increase in the spontaneous EPSCs frequency using patch clamp recording from spinal dorsal horn neurons [
21]. However, the mechanism of these effects has not been clarified. In this study, we showed that CAP inhibits neuronal excitation, especially in superficial layers of the spinal dorsal horn. This distribution of inhibited area by CAP agrees with the distribution of TRPV1 channels reported by immunohistochemical studies [
12]. It is well known that the activation of TRPV1 channels induces an influx of Ca
++ and Na
+ ions via this receptor [
6]. In this study, we showed that inhibition of the neuronal excitation by CAP was not observed in the presence of CPZ. In addition, the CAP effect was not dependent on the activation of inhibitory amino acid-containing neurons, since this inhibition was unaffected by BMI and STRY. Therefore, it is reasonable to conclude that CAP inhibited the neuronal excitation via TRPV1 channels in superficial layer of the spinal dorsal horn, although a possibility remains that CAP-sensitive interneurons containing other inhibitory transmitter(s) and/or modulator(s) contributed to the CAP effect.
Next, we investigated the Ca
++ dependency of the CAP and temperature effects. Calcium ions are the major cation carried through the TRPV1 channel [
6]. Both the CAPs- and temperature-dependent inhibitions were blocked in Ca
++-free condition. Although the TRPV1 channel has a relatively high permeability to Ca
++, both CAP and heat are known to activate the channels in the absence of external Ca
++ [
6,
22], and it seems difficult to explain the complete block of CAP and heat effects in the Ca
++ free condition. It is possible that TRPV1 channels were activated by CAP and heat in the Ca
++ free condition, but that the influx of cations other than Ca
++ via TRPV1 channels, and the following presynaptic depolarization, might not be sufficient to inhibit the neuronal excitation.
In the periphery, the acute administration of CAP induces burning pain by activating C fibers via TRPV1 channels. It depolarizes the terminals generating action potentials. They are conducted to the spinal cord and induce nociceptive sensation. Larger doses and/or prolonged administration of CAP have analgesic effect due to the desensitization of TRPV1 channels [
23]. We recorded the excitation centrally. The depolarization of afferent terminals and/or fibers branching points would cause reduction of action potential size and even failure of action potential generation [
26‐
29]. The central inhibitory action of CAP observed in this study accords well with this presynaptic inhibition mechanism. When CAP was applied for 10 min, the inhibition of neuronal excitation was reversible. However, when CAP was applied for 20 min, it became irreversible. This prolonged irreversible effect of CAP might be due to other mechanisms, such as CB1 activation.
In this study, the neuronal excitation was also inhibited by anandamide to a larger extent (-23.3%) than by CAP (-11.8%), and the effect of anandamide was partially antagonized by CPZ. It is known that the cannabinoid CB1 receptors are also activated by anandamide [
30]. Therefore, it is possible that the inhibition of the neuronal excitation by anandamide is the combination of TRPV1 and CB1 channel activation.
TRPV1 channels in the periphery are activated by heat, as well as by CAP [
6,
7]. However, the temperature dependence of TRPV1 channels in the spinal dorsal horn has not been reported. In this study, we showed that increasing the perfusate temperature inhibited the neuronal excitation in the superficial laminae of the spinal dorsal horn. This inhibition was Ca
++-dependent, antagonized by CPZ, and was unaffected by BMI and STRY, as well as, by the inhibition by CAP. From these results, the inhibition of neuronal excitation by increasing the perfusate temperature may be induced by the same mechanism as the inhibition induced by CAP. That is, the natural ligand of TRPV1 channels is present in the normal condition, and is continuously activating TRPV1 channels. In fact, in the natural condition, anandamide has been found in the dorsal horn [
18], and we showed in this study that anandamide depressed the neuronal excitation in a similar manner to CAP.
Normally, the thermal threshold for activation of TRPV1 channels is more than 43°C [
6,
7]. However, in this study, the perfusate temperature increase from 27 to 33°C was sufficient to inhibit the neuronal excitation mediated by TRPV1 channels. Tominaga et al. [
7] reported that the temperature threshold for TRPV1 channels activation was reduced from 42 to 35°C in the presence of ATP by measuring the current in HEK293 cells and that this effect was mediated by the P
2Y receptor. In this study, the effect of CAP was blocked by suramin and was facilitated by UTP, and the effect of increasing the perfusate temperature was blocked by suramin, but not by TNP-ATP. These results indicate that the threshold for activation of TRPV1 channels in the spinal dorsal horn is decreased by the activation of the P
2Y receptor. ATP has been reported to be a fast synaptic transmitter in the spinal dorsal horn [
24,
25]. ATP release from primary afferent terminals may contribute to the decrease of threshold for TRPV1 channels activation. The presence of ATP in the unstimulated condition is also highly likely. Therefore, TRPV1 channels in the spinal dorsal horn may be activated even at physiological temperatures.
In this study, the optical signal was recorded by anterograde staining of only primary afferents. Since this optical response was not decreased in a Ca
++-free medium (Fig.
6B) or in D-AP5 and CNQX [
17], it reflects exclusively the voltage change of presynaptic elements. The inhibition of the presynaptic response by CAP and by the increase in perfusate temperature suggests a decrease in the amplitude or number of presynaptic action potentials in presynaptic elements. TRPV1 channels exist, not only on primary afferent terminals, but also on postsynaptic neurons in the superficial dorsal horn [
12] though not conclusive at present. Therefore, the decrease of presynaptic excitation may be due to a direct effect on TRPV1 channels in primary afferent terminals, or indirect effects mediated by any factor(s) released from TRPV1 channel-activated postsynaptic neurons. However, the latter possibility is less likely since the effect of CAP was observed in the presence of D-AP5 and CNQX, when most of the fast afferent signal transmission to the postsynaptic sites is blocked, although a possibility remains that the afferent stimulation we used activated CAP-sensitive glutamate-containing interneurons via transmitter(s) other than glutamate. Therefore, the inhibition of the presynaptic response by CAP and the increase in perfusate temperature is likely due to a decrease in the amplitude or number of presynaptic action potentials via the activation of TRPV1 channels.
In the spinal dorsal horn, the regulation of presynaptic spiking is known in the context of presynaptic inhibition. The depolarization of primary afferent terminals decreases the amplitude of action potentials and, thus, produces the presynaptic inhibition [
27,
28]. In addition, the silent presynaptic afferent terminals may become activated after their liberation from the shunting of the action potentials at primary afferent terminals [
27,
29,
31‐
33]. In the superficial laminae of the dorsal horn, TRPV1 channels are abundant, especially at presynaptic elements of unmyelinated C-fiber origin. Therefore, CAP and the increase in perfusate temperature may inhibit neuronal excitation in the spinal dorsal horn via TRPV1 channels at primary afferent terminals.
TRPV1 channels exist not only at sensory neuron somata and primary afferent terminals, but also on axonal membranes [
34]. Furthermore, it is reported that the application of TRPV1 channel agonists at afferent fiber bundles blocks action potential conduction [
8‐
11]. Therefore, another possibility is that the inhibition of neuronal excitation in the spinal dorsal horn by CAP and an increase in perfusate temperature may be caused by
a conduction block of the action potential in primary afferent fibers. It has been reported that removal of external Ca
++ does not affect the C fiber conduction block caused by CAP [
9]. Therefore, the decrease of neuronal excitation by CAP and temperature increase is less likely due to a conduction block. It is difficult to distinguish, however, whether the effect of CAP and an increase in perfusate temperature is caused by a change of action potential at the level of primary afferent fibers or their terminals in the dorsal horn, because the presynaptic response that we measured in this study was the sum of excitations in both afferent fibers and terminals along the slice depth. To elucidate this point, it is necessary to record the action potentials in individual afferent fibers with higher-resolution camera in the future.
Methods
The preparation, apparatus, and data processing for the optical imaging were identical to those used in our previous studies [[
13‐
16,
35]; for detailed descriptions see [
16]]. A brief summary follows.
Preparation
All animal studies were undertaken with protocols approved by the university animal ethics committee. Transverse slices (350–450 μm thick) with dorsal roots attached (5–10 mm in length) were prepared from lumbosacral enlargements of 12- to 25-day-old Sprague-Dawley rat spinal cords, as described elsewhere [
36]. A slice, stained with a voltage-sensitive absorption dye, RH-482 (0.1 mg/ml, 20 min), was set in a submersion-type chamber (0.2 ml) on an inverted microscope (IMT, Olympus, Tokyo) equipped with a 150-W halogen lamp. The slice was perfused with Ringer solution at 27°C containing 124 mM NaCl, 5 mM KCl, 1.2 mM KH
2PO
4, 1.3 mM MgSO
4, 2.4 mM CaCl
2, 26 mM NaHCO
3, 0.2 mM thiourea, 0.2 mM ascorbic acid, and 10 mM glucose (oxygenated with 95% O
2 and 5% CO
2).
The RH-482 (NK-3630) dye was obtained from Nippon Kanko Shikiso (Okayama, Japan); the DL-2-amino-5-phosphonovaleric acid (D-AP5), bicuculline methiodide, strychnine hemisulfate, and nifedipine were from Sigma (St. Louis, MO); and the other chemicals were from Nacalai Tesque (Kyoto, Japan). Each of these chemicals was dissolved in distilled water at high concentration, divided into aliquots, and kept frozen at -40°C until use. The aliquots were dissolved in the bathing solution at known concentrations during the experiments. The perfusion solution of the slice was switched to the drug-containing solution for a fixed period.
Neonatal capsaicin treatment
Capsaicin was dissolved in 100% ethanol, and then diluted with Tween80/PBS solution until ethanol concentration became 10%. Rats of postnatal day 2–3 were injected the capsaicin solution (50 mg/kg) subcutaneously at the dorsal cervix. Three weeks after the injection, rats were tested on a hot plate (65°C). While normal untreated rats raised and licked the feet within 10 seconds, successfully treated rats did not react over one minute.
Optical recording
The light absorption change in a 0.83-mm-square area in the dorsal horn at a wavelength of 700 ± 32 nm was recorded by an imaging system (SD1001, Fuji Film Microdevice, Tokyo) with 128 × 128-pixel photo sensors at a frame rate of 0.6 ms [
37]. Thirty-two single pulses were given to the dorsal root at a constant interval of 12–15 s. Starting at 10 ms before each stimulus, 128 consecutive frames of the light-absorption images were taken by the image sensor with a sampling interval of 0.6 ms. The reference frame, which was taken immediately before each series of 128 frames, was subtracted from each of the subsequent 128 frames. Thirty-two series of such difference images were averaged and stored in the system memory. We determined the initial frame by averaging the first 15 frames of the difference image and then subtracting this from each of the 128 frames of the image data on a pixel-by-pixel basis in order to eliminate the effects of noise contained in the reference frame. The ratio image was then calculated by dividing the image data by the reference frame. In most cases, the ratio image was filtered by a 3-point moving average over time [see [
14] for detail].
Dorsal root stimulation
The dorsal root was stimulated by a glass suction electrode. The types of primary afferent fibers activated by the electrical stimuli were initially identified by the field potentials recorded by a glass microelectrode positioned either in the superficial dorsal horn or at the entry zone of the root, as described previously [
14].
The single current-pulse stimulation of the dorsal root elicits the following optical responses in the dorsal horn [
14]: (1) a brief (< 3 ms) and small, almost undetectable, response is evoked at the entry zone of the dorsal root and, occasionally, in the deep dorsal horn by a 0.05-mA current pulse of 0.05 ms duration; (2) by increasing the stimulus intensity to 0.1 mA, an optical response of longer duration (< 100 ms) appears in lamina I extending to the outer part of lamina II, and in lamina III and deeper laminae; (3) additional increases in intensity (> 0.3 mA) and/or duration (> 0.5 ms) lead to the generation of an intense, prolonged (> 200 ms) response in the superficial laminae I-III, most prominently in lamina II. The long-lasting response in lamina II is delayed, with the latency corresponding to the slow conduction velocity of C fibers (ca. 1 m/s). The onset of optical responses in lamina I and lamina III or deeper laminae elicited by any of these conditions takes place within one image frame (i.e., the latency is less than 0.6 ms). The conduction velocity of fibers responsible for the induction of the immediate response should be faster than 6 m/s (dorsal root length of 4 mm/0.6 ms), which corresponds to the conduction velocity of A fibers. These spatial and temporal profiles of optical responses agree fairly well with the cytoarchitectonic organization of the dorsal horn [
38‐
41], giving additional indications of the fiber types activated by the stimuli.
Therefore,
t he activation conditions consisted of a 0.05-mA current pulse of 0.05 ms duration for Aα/β fibers (A fib
e r other than Aδ fibers), a 0.1-mA current pulse of 0.05 ms duration (low-intensity stimulation) for Aα/β/δ fibers (all types of A fibers), and a 1.5-mA current pulse of 0.5 ms duration (high-intensity stimulation) for A and C fibers. These parameters were similar to those used in other studies [
42‐
46].
Temperature control
The temperature of perfusate in the recording chamber was maintained with a heating coil along the polyethylene tubing (50 cm in length and 1 mm in diameter) between a peristalic pump and the inlet of chamber. Solution in a reservoir that was maintained at a room temperature of 25°C was heated to 27°C while passing through the tubing. The solution temperature in the chamber was continuously monitored with a small sensor positioned near the specimen. Since the room temperature was maintained constant and the solution in the reservoir was equilibrated to the room, the solution temperature in the chamber was usually stable, within ± 0.2°C, during the duration of each experiment (1–3 hours). Since the capacity of the chamber (0.5 ml) and the volume in the tubing (0.1 ml) were small in respect to the flow (3 ml/min), the temperature of the solution in the chamber could be controlled by the amount of electric current. In control experiments, the temperature measured at the surface of the slice was increased from 27°C to 33°C within 180 seconds by a current increase, and, following the reduction after 20 min at 33°C, the temperature at the surface of the slice was returned to 27°C within 300 seconds.
Statistical analysis
Results were expressed as the mean ± SE. Non-parametric ANOVA (Tukey-Kramer test) was used for statistical differences. Significant differences were calculated using the t-test, and differences were considered significant when p < 0.05 (*) or p < 0.01 (**) and # (p > 0.5).