Excitatory actions of mushroom poison (acromelic acid) on unmyelinated muscular afferents in the rat

Neuroscience Letters 456 (2009) 69–73

Contents lists available at ScienceDirect

Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet

Excitatory actions of mushroom poison (acromelic acid) on unmyelinated muscular afferents in the rat Toru Taguchi, Kimihiko Tomotoshi, Kazue Mizumura ∗ Department of Neuroscience II, Research Institute of Environmental Medicine, Nagoya University, Nagoya, Japan

a r t i c l e

i n f o

Article history: Received 27 January 2009 Received in revised form 27 March 2009 Accepted 29 March 2009 Keywords: C-fiber receptors Skeletal muscle Muscle-nerve preparations in vitro Acromelic acid Mushroom poisoning

a b s t r a c t Ingestion of a poisonous mushroom, Clitocybe acromelalga, results in strong and long-lasting allodynia, burning pain, redness and swelling in the periphery of the body. Acromelic acid (ACRO), a kainate analogue isolated from the mushroom, is assumed to be involved in the poisoning. ACRO has two isomers, ACROA and ACRO-B. The potency of ACRO-A is a million times higher than that of ACRO-B for induction of allodynia when intrathecally administered in mice. The effect of ACRO on the primary afferents of somatic tissues remains largely unknown. The aim of the present study was to examine the effect of ACRO-A on the response behavior of unmyelinated afferents in the skeletal muscle. For this purpose single fiber recordings of C-afferents were made from rat extensor digitorum longus (EDL) muscle-common peroneal nerve preparations in vitro. Intramuscular injections of ACRO-A at three different concentrations (10−12 , 10−10 and 10−8 M, 5 ␮l over 5 s) near the receptive field in the EDL muscle elicited excitation of C-afferents (12%, 50% and 44%, respectively). ACRO-A at the concentration of 10−10 M induced the strongest excitation. The incidence of ACRO-A responsive fibers at the concentration of 10−10 and 10−8 M was significantly higher than that at 10−12 M. The responses to mechanical and heat stimulations did not differ between ACRO-A sensitive and insensitive fibers. These results clearly demonstrated the powerful excitatory action of ACRO-A on mechanosensitive unmyelinated afferents in the rat skeletal muscle. © 2009 Elsevier Ireland Ltd. All rights reserved.

A Japanese poisonous mushroom, Clitocybe acromelalga, causes strong and long-lasting allodynia and burning pain, redness and swelling in the periphery of the body, such as the hands and feet. The symptoms are quite similar to those of erythromelalgia. The symptoms usually appear with some delay (i.e. several days after ingestion), and last for a month or longer. The hand and foot nature of the symptoms is similar to peripheral neuropathies caused by many metabolic disorders (e.g. diabetic neuropathy) and organic solvent intoxication. Recently, mutations of NaV 1.7 channels, which are present at high levels in nociceptive dorsal root ganglion cells and sympathetic ganglion neurons, have been shown to be related to inherited erythromelalgia [19]; however, it is not yet known why this disease affects the hands and feet. Similarities to peripheral neuropathy and inherited erythromelalgia suggest that this mushroom poisoning is peripheral in origin, at least in part. Acromelic acid (ACRO), a kainate analogue isolated from this mushroom, is considered to be responsible for the mushroom poisoning [14]. There are two isomers of ACRO, ACRO-A and ACRO-B

∗ Corresponding author at: Department of Neuroscience II, Division of Stress Recognition and Response, Research Institute of Environmental Medicine, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8601, Japan. Tel.: +81 52 789 3861; fax: +81 52 789 3889. E-mail address: [email protected] (K. Mizumura). 0304-3940/$ – see front matter © 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2009.03.102

[5,6], and the potency of intrathecally administered ACRO-A in inducing allodynia in mice is a million times stronger than that of ACRO-B [12]. ACRO produced selective damage of inhibitory interneurons of the rat lumbar spinal cord when injected systemically or intrathecally [7–9,13,15,18]. Intrathecal administration of ACRO-A and ACRO-B induced long-lasting allodynia 5 min after injection. Even though ACRO-A contains the structure of kainic acid, its allodynia inducing effect was blocked only by N-methyld-aspartate (NMDA) receptor antagonists and Joro spider toxin, a Ca2+ -permeable AMPA receptor antagonist, but not by other AMPA/kainate antagonists [12]. Although the distribution pattern of the symptoms suggests a peripheral nature, there has been no study on the effects of ACRO upon peripheral afferents. We therefore studied whether ACRO-A, the stronger isomer of acromelic acid, ever affected the activities of unmyelinated afferents, using single fiber recording with in vitro rat muscle-nerve preparations that we have recently established [17]. Twenty male Sprague–Dawley rats weighing 355–427 g (11–13 weeks) were used in this study. The animals were purchased from a breeding company (SLC Inc., Shizuoka, Japan) and kept 1–2 per cage under a 12 h light/dark cycle (light between 07.00 and 19.00 h). Room temperature was kept constant (22 ± 1 ◦ C). The animals had free access to food and water. All of the study was conducted according to the Regulations for Animal Experiments in Nagoya

70

T. Taguchi et al. / Neuroscience Letters 456 (2009) 69–73

University, and the Fundamental Guidelines for Proper Conduct of Animal Experiment and Related Activities in Academic Research Institutions in Japan. Rat extensor digitorum longus (EDL) muscle-common peroneal nerve preparation in vitro was used, and the methods of obtaining single fiber activities, stimulation and analysis were basically the same as previously reported [17]. Briefly, the EDL muscle from both hindlimbs was carefully removed with the common peroneal nerve attached under pentobarbital anesthesia (50 mg/kg, i.p.). Animals were killed with an overdose of the same anesthetic after removal of the preparation. The isolated preparation was then placed in an organ bath with the proximal and distal ends of the EDL muscle pinned in the test chamber. The preparation was maintained at 34.0 ± 0.5 ◦ C (pH 7.4) under superfusion with modified Krebs–Henseleit solution (Krebs solution), which contained (in mM) 110.9 NaCl, 4.7 KCl, 2.5 CaCl2 , 1.2 MgSO4 , 1.2 KH2 PO4 , 25.0 NaHCO3 , and 20.0 glucose, and the superfusate was continuously bubbled and equilibrated with a gas mixture of 95% O2 and 5% CO2 . The common peroneal nerve was drawn through a hole to the recording chamber filled with paraffin-oil. Single fiber activities were obtained by dissection method. Action potentials were analyzed on a computer with an analog–digital converter and SPIKE/SPIDI software (Forster C., University of Erlangen-Nuernberg, Germany). The muscle C-fiber sensory receptors used in this experiment fulfilled the same criteria as used in the previous study [17], namely: (1) they were sensitive to mechanical stimulation by probing with a glass rod, (2) had no intensity-dependent increase in the discharge rate while the muscle was stretched by a length of a few millimeters (to exclude muscle spindle afferents), and (3) had conduction velocity slower than 2.0 m/s. The conduction velocity of the fibers was calculated from the distance and conduction latency of a spike induced by electrical stimulation of the receptive field (RF) (intensity just above the threshold to obtain constant action potentials up to 50 V with pulse duration of 100–500 ␮s). Prior to injection the distribution (the size and location) of the RFs of the fibers was investigated with a blunt glass rod, and then with von Frey hairs (VFHs, 0.5 mm in diameter, 1.8–66.8 mN) for a detailed search. The mechanical threshold of a fiber was measured by the method of limits. Each filament was applied twice and the bending force of the filament that induced at least one action potential was considered to be the threshold. The RF was marked with the VFH that was one or two grades stronger than the threshold and that consistently induced a response. Intramuscular injections of ACRO-A into the EDL muscle were made at three different concentrations (10−12 , 10−10 and 10−8 M). The order of the injections was always the same, from the lowest concentration to the highest. With the use of a 3D manipulator the tip of the injection needle (30G) connected to a microsyringe was placed 2 mm from the edge of the RF in the EDL to avoid direct mechanical irritation of the endings. The injections were made at a volume of 5 ␮l over 5 s. The injection volume corresponded to 2.5% of the whole EDL muscle (approximately 200 ␮l [16]). In our pilot study all muscular C-fibers (n = 15) were activated when 5% formalin was injected with this method and none of them were activated with phosphate buffered saline (PBS 0.1 M, pH 7.4, n = 26, data not shown). This observation assured good access of the test substance to the RF of all recorded C-fibers, and mechanical intervention by injection had no influence on discharges. The next injection was made 10 min after the previous one when there was no excitation. If an afferent fiber showed vigorous excitation, then the next injection was made 10 min after cessation of the response from the previous injection. Mechanical and heat stimulations were applied to characterize ACRO-A responsive fibers before a set of injections of ACRO-A. A quantitative mechanical stimulus was provided by a mechanical stimulator that had feedback regulation of the force (manufactured by Aizawa S., Goto College of Medical Arts and Science,

Tokyo, Japan). The tip size of the probe was 2.28 mm2 . A ramp mechanical stimulus, linearly increasing from 0 to 196 mN in 10 s, was applied to the most sensitive point of the identified RF (Fig. 1A). Heat stimulus was also applied on the surface of the identified RF by superfusing Krebs solution with gradually increasing temperature. The temperature curve was made with a hand-made counter-current tubing system in which the perfusing Krebs solution (at room temperature) ran through the inner tube, which led to the RF, while pre-warmed water ran in the opposite direction in the outer tube. The RF was warmed from 34 ◦ C to approximately 50 ◦ C in 30 s (Fig. 1B). The mechanical and heat thresholds were defined by the same way in our previous study [17]. A fiber was defined to be sensitive to ACRO-A when it fulfilled both of the following two criteria (Fig. 1C2 and C3): (1) the net increase in the discharge rate for 30 s immediately after the onset of the response was more than 0.5 imp/s from the background discharge rate in the control period (60 s, Fig. 1C2), and (2) the instantaneous discharge rate of two consecutive discharges exceeded the mean + 2S.D. of the background discharge rate at least once by 5 min after the injection (Fig. 1C3). The onset of the response was defined as the time point when a fiber induced a discharge exceeding the mean frequency + 2S.D. of the background discharges in the control period (60 s, Fig. 1C3), and when there were two or more consecutive discharges exceeding this level. The latency of the response was defined as the time elapsed from the time the injection was started to the onset of the response. Results were expressed as median with interquartile range (IQR). Data for three groups (concentrations) were compared with the Kruskal–Wallis test. Comparisons between ACRO-A sensitive and insensitive fibers were made using the Mann–Whitney U-test. The incidence of responding fibers was compared for different concentrations of ACRO-A with Fisher’s exact probability test. p < 0.05 was considered significant. A total of 25 muscle C-afferents were identified and recorded from both hindlimbs (13 from the right EDL muscle and 12 from the left). Conduction velocity of the fibers was 0.65 (IQR: 0.54–0.75) m/s. The RFs varied in size and shape: the RFs were round or oval in shape with a diameter ranging from 0.5 to 5 mm. More RFs tended to be found near the musculotendinous junction than in other parts, corresponding to our previous study [17]. The response patterns to ACRO-A (latency, duration and magnitude) were highly variable. Some fibers that responded to ACRO-A stopped firing for a while (30–120 s) following vigorous excitation. Another, rather representative, response of a fiber to three concentrations of ACRO-A is shown in Fig. 2A1–A3. Here the fiber responded relatively weakly to ACRO-A 10−12 M with a response latency of about 40 s after injection (Fig. 2A1). ACRO-A 10−10 and 10−8 M induced irregular discharges with short latencies of a few seconds after injection (Fig. 2A2 and A3). A larger peak discharge rate but shorter response duration was observed in the response to ACRO-A 10−8 M when compared with the response to 10−10 M. The averaged response patterns are shown in Fig. 2B1–B3. The characteristics of the response to ACRO-A (incidence of responsive fibers, background activity before injection, response latency, and magnitude of response) are summarized in Table 1. The incidences at the concentrations of ACRO-A 10−10 and 10−8 M were significantly higher than that at 10−12 M (p < 0.01, 10−12 M vs. 10−10 M, and p < 0.05, 10−12 M vs. 10−8 M, Fisher’s exact probability test). The onset (latency) of the responding fibers was clearly shorter with higher concentrations of ACRO-A, but it was not significantly different with the three concentrations. The magnitude of the response of the ACRO-A sensitive fibers, which was represented by a net increase in discharge rate for 30 s immediately after the onset of the response, was the highest with the concentration of 10−10 M. However, it was not significantly dif-

T. Taguchi et al. / Neuroscience Letters 456 (2009) 69–73

71

Fig. 1. Responses of an afferent fiber to mechanical, heat and ACRO-A stimulation. (A) Mechanical response. Original registration (A1), corresponding peri-stimulus time histogram of the response (bin width: 1 s, A2), instantaneous frequency of discharges (Hz, A3, each action potential is shown as one dot) and recording of force applied by a servo-controlled mechanical stimulator to the receptive field of the fiber (A4), respectively. In A2 mean discharge rate during the control (60 s) and the stimulus period (10 s, horizontal black bar) is shown above the histogram. In A3 values of mean (thin line) and mean + 2 standard deviation (dashed line) are given during the control period to determine the mechanical threshold (see text for detail). Mechanical threshold of this fiber was 48.8 mN when a ramp mechanical stimulus, linearly increasing from 0 to 196 mN in 10 s, was applied (A4). Time scale on abscissa in A4 is identical also with A1–A3. (B) No response to heat: peri-stimulus time histogram (bin width: 1 s, top) and temperature curve (bottom). (C) Response to ACRO-A (10−10 M): original registration (C1), corresponding peri-stimulus time histogram of the response (bin width: 1 s, C2), and instantaneous frequency of discharges (Hz, C3, each action potential is shown as one dot), respectively. In C2 mean discharge rate during the control (60 s) and 30 s immediately after the onset of the response is shown above the histogram. In C3 horizontal lines of mean (thin) and mean + 2 standard deviation (thick dashed line) are given during the control period to determine onset of the response (see text for detail).

ferent between the three concentrations, possibly because of the small number of fibers responsive to ACRO-A 10−12 M. It should be noted that the background activities before the injections of ACRO-A were not different with the three concentrations (Table 1), which means that the response induced by ACRO-A diminished before the next application at a higher concentration. These results together show that ACRO-A 10−10 M had the most powerful and long-lasting excitatory action on unmyelinated muscular afferents. Table 1 Characteristics of the response. Concentration of ACRO-A 10−12 M No. of responding/tested fibers (%) Background activity before ACRO-A (imp/s) Response latency (s) Net response magnitude (imp/s)

10−10 M

10−8 M **

3/25 (12)

12/24 (50)

10/23 (44)*

0.10 (0–0.45)

0.12 (0–0.42)

0.10 (0–0.36)

0.3, 82.1, 124.0 1.1, 1.7, 2.1

7.7 (3.6–20.5) 5.0 (2.1–6.0)

6.5 (2.3–15.4) 2.9 (2.2–3.6)

* p < 0.05 and ** p < 0.01 significant difference when compared with data at the lowest concentration of ACRO-A (10−12 M), Fisher’s exact probability test.

The responsiveness of ACRO-A sensitive fibers to other stimulus modalities was analyzed with the responses to ACRO-A 10−10 M (n = 24 tested fibers), because that concentration was the most effective in eliciting excitation of the afferents. There were no significant differences between the ACRO-A sensitive and insensitive fibers in the mechanical thresholds determined by VFHs and the mechanical stimulator, the magnitude of the mechanical response (i.e. net evoked discharges during the quantitative mechanical stimulus), the incidence of heat sensitive fibers, or the response threshold temperature to heat. The magnitude of the heat response tended to be higher in the ACRO-A insensitive fibers than in the sensitive ones (p = 0.053, Mann–Whitney U-test). Conversely, the response magnitude and the latency of ACRO-A-induced excitation were not significantly different between heat sensitive and insensitive fibers. This is the first study to demonstrate the powerful excitatory action of acromelic acid-A on unmyelinated muscular afferents. Half of the mechanically sensitive C-fibers in the skeletal muscle showed sensitivity to ACRO-A, and the most effective concentration was 10−10 M. The mechanical and heat responses of muscular C-fibers were not significantly different between ACRO-A sensitive and insensitive fibers, suggesting that ACRO-A sensitivity is not determined by mechanical or heat sensitivity. Thus receptor

72

T. Taguchi et al. / Neuroscience Letters 456 (2009) 69–73

Fig. 2. Responses to ACRO-A. (A) Sample recording from a fiber. Original registrations (upper graphs) and corresponding peri-stimulus time histograms (lower graphs) of a fiber responding to ACRO-A 10−12 M (A1), 10−10 M (A2), and 10−8 M (A3). Left inset in C represents the location of a receptive field (black area) at the muscle-tendon junction (tendon: hatched area) and the point of injection (tip of the needle). Right inset represents the spike shape of this C-fiber. Conduction velocity was 0.43 m/s. Mechanical and heat responses of this fiber are shown in Fig. 1A and B. Sample recordings of A2 and Fig. 1C are identical. (B) Averaged responses of 23–25 fibers. n = 25, 24 and 23 for ACRO-A 10−12 M (B1), 10−10 M (B2) and 10−8 M (B3), respectively. The averaged responses are shown by calculating mean (black) discharge rate + S.E.M. (grey) of all the fibers recorded (bin width: 1 s). Abscissa: time in s. Ordinate: discharge rate (imp/s). The time point of injection is indicated by a broken line with a solid arrowhead. Note (1) a small response to ACRO-A 10−12 M, (2) a strong and long-lasting response to ACRO-A 10−10 M, and (3) a strong, but short-lasting response to ACRO-A 10−8 M.

molecules for ACRO-A could be present independent of the presence or absence of mechano-/heat-transducers/ion channels. The mechanisms of ACRO-related mushroom poisoning, which is characterized by redness, swelling, burning pain and tactile allodynia in the periphery of the body, are largely unknown. Ingestion of a few toadstools seems to be enough to develop the symptoms. The concentration of ACRO-A has been estimated at 283 ␮g/g in C. acromelalga [2]. Although there are no detailed reports as to the amount, a 10 g toadstool is likely to be enough to induce the poisoning judging from some individual cases. Supposing that a 60 kg man ingested this amount of mushroom, the dose of ACRO-A would be estimated at 47.2 × 10−6 g/kg body weight [2]. In a study by Minami et al. [12], the mechanical allodynia inducing effect was maximum with a dose of 50 × 10−15 g/kg when intrathecally administered to mice. In this study we injected ACRO-A at the concentration of 10−12 , 10−10 and 10−8 M (5 ␮l) into the receptive field of the EDL muscle (c.a. 200 mg). This corresponds to 7.75 × 10−12 , 10−10 , 10−8 g/kg muscle, respectively. Judging from these calculations, it seems that oral ingestion of the mushroom needs a greater amount of ACRO-A for poisoning than for excitation of unmyelinated muscular afferents, and induction of allodynia by intrathecal administration needs a much lower amount of ACRO-A than for excitation of afferents. There is also some discrepancy in the concentrations of ACRO-A needed to induce neural excitation. In previous studies the concentration of ACRO-A to depolarize C-fibers in the dorsal root of 1–7-day-old Wister rats [4] and to induce calcium influx in the spinal cord slice of 2-week-old mice [12] was around 10−6 M but not under 10−8 M, but it was higher than that to elicit excitation of muscular C-fibers of adult Sprague–Dawley rats in the present study. This discrepancy might have come from differences in species, strain or the developmental stage of the used animals.

This study revealed that half of mechano-sensitive unmyelinated muscular afferents were sensitive to ACRO-A. A similar result could be expected also in the skin. In the cutaneous tissue, strong and long-lasting excitation of C-fibers may cause neurogenic inflammation by the release of powerful vasodilatatory and pro-inflammatory neuropeptides such as substance P (SP) and calcitonin gene-related peptide (CGRP) from nerve terminals [11]. It is possible that strong and long-lasting excitation of C-fibers induced by ACRO results in neurogenic inflammation, and this would explain the erythromelalgia-like symptoms in the periphery of the body. Acromelic acid is a kainoid having similar chemical structure to kainic acid, and kainate receptors exist in peripheral afferents [10]. However, the neuroexcitatory and neurotoxic effects of ACRO and kainic acid are quite different. For example (1) ingestion of ACRO causes poisoning similar to the symptoms of erythromelalgia, whereas kainic acid, which is an extract from a kind of seaweed (Digenea simplex), has been traditionally used as a vermifuge, (2) intrathecal administration of ACRO induces strong and long-lasting allodynia while that of kainic acid does not, with ACRO-A being effective at much lower dosage than ACRO-B [12], and (3) systemic or intrathecal administration of ACRO leads to histological damage restricted to the lower spinal neurons, tonic extension of the hindlimbs, subsequent spastic paraplegia, and so on [7–9,15], whereas the administration of kainic acid results in selective neuronal damage in the limbic system (e.g. thalamus, hypothalamus and hippocampal CA1) and limbic seizure [1]. No neuronal damage can be seen in the limbic area from ACRO, or in the lower spinal neurons from kainic acid. In addition, ACRO-A induced allodynia was blocked by NMDA receptor antagonists and Joro spider toxin, a Ca2+ -permeable AMPA receptor antagonist, but the ACROB induced allodynia was not [12]. There is evidence that NMDA receptors exist on muscle afferents [3]. Therefore, ACRO-induced

T. Taguchi et al. / Neuroscience Letters 456 (2009) 69–73

excitation might have occurred through NMDA or AMPA receptors in the periphery. However, the allodynia induced by ACRO was not affected by other AMPA/kainate antagonists. These results suggest that ACRO acts through as yet unidentified receptors, and that ACRO is stereospecific for the induction of allodynia. In summary, half of the mechanically sensitive unmyelinated afferents in the skeletal muscle were sensitive to acromelic acidA. Further trials to find ACRO-specific receptor molecules and to elucidate the mechanism leading to the mushroom poisoning after ingestion are needed. Acknowledgements The authors wish to thank to Dr. S. Ito, the Department of Medical Chemistry, Kansai Medical University and Dr. T. Minami, the Department of Anesthesiology, Osaka Medical College, for giving us an opportunity to perform this study, and to Dr. K. Furuta, the United Graduate School of Drug Discovery and Medical Information Sciences, Gifu University, for kindly providing us acromelic acid. This work was supported by The Uehara Memorial Foundation. References [1] Y. Ben-Ari, Limbic seizure and brain damage produced by kainic acid: mechanisms and relevance to human temporal lobe epilepsy, Neuroscience 14 (1985) 375–403. [2] J. Bessard, P. Saviuc, Y. Chane-Yene, S. Monnet, G. Bessard, Mass spectrometric determination of acromelic acid A from a new poisonous mushroom: Clitocybe amoenolens, J. Chromatogr. A 1055 (2004) 99–107. [3] B.E. Cairns, P. Svensson, K. Wang, S. Hupfeld, T. Graven-Nielsen, B.J. Sessle, C.B. Berde, L. Arendt-Nielsen, Activation of peripheral NMDA receptors contributes to human pain and rat afferent discharges evoked by injection of glutamate into the masseter muscle, J. Neurophysiol. 90 (2003) 2098–2105. [4] M. Ishida, H. Shinozaki, Novel kainate derivatives: potent depolarizing actions on spinal motoneurones and dorsal root fibres in newborn rats, Br. J. Pharmacol. 104 (1991) 873–878.

73

[5] K. Konno, K. Hashimoto, Y. Ohfune, H. Shirahama, T. Matsumoto, Acromelic acids A and B. Potent neuroexcitatory amino acids isolated from Clitocybe acromelalga, J. Am. Chem. Soc. 110 (1988) 4807–4815. [6] K. Konno, H. Shirahama, T. Matsumoto, Isolation and structure of acromelic acid A and B. New kainoids of Clitocybe acromelalga, Tetrahedron Lett. 24 (1983) 939–942. [7] S. Kwak, H. Aizawa, M. Ishida, H. Shinozaki, Systemic administration of acromelic acid induces selective neuron damage in the rat spinal cord, Life Sci. 49 (1991) PL91–96. [8] S. Kwak, H. Aizawa, M. Ishida, H. Shinozaki, Acromelic acid, a novel kainate analogue, induces long-lasting paraparesis with selective degeneration of interneurons in the rat spinal cord, Exp. Neurol. 116 (1992) 145–155. [9] S. Kwak, R. Nakamura, Selective degeneration of inhibitory interneurons in the rat spinal cord induced by intrathecal infusion of acromelic acid, Brain Res. 702 (1995) 61–71. [10] C.J. Lee, H. Kong, M.C. Manzini, C. Albuquerque, M.V. Chao, A.B. MacDermott, Kainate receptors expressed by a subpopulation of developing nociceptors rapidly switch from high to low Ca2+ permeability, J. Neurosci. 21 (2001) 4572–4581. [11] B. Lynn, Neurogenic inflammation caused by cutaneous polymodal receptors, Prog. Brain Res. 113 (1996) 361–368. [12] T. Minami, S. Matsumura, M. Nishizawa, Y. Sasaguri, N. Hamanaka, S. Ito, Acute and late effects on induction of allodynia by acromelic acid, a mushroom poison related structurally to kainic acid, Br. J. Pharmacol. 142 (2004) 679–688. [13] T. Ogata, Y. Nakamura, K. Tsuji, T. Shibata, K. Kataoka, M. Ishida, H. Shinozaki, A marked increase in intracellular Ca2+ concentration induced by acromelic acid in cultured rat spinal neurons, Neuropharmacology 33 (1994) 1079–1085. [14] P. Saviuc, V. Danel, New syndromes in mushroom poisoning, Toxicol. Rev. 25 (2006) 199–209. [15] H. Shinozaki, M. Ishida, Y. Gotoh, S. Kwak, Specific lesions of rat spinal interneurons induced by systemic administration of acromelic acid, a new potent kainate analogue, Brain Res. 503 (1989) 330–333. [16] T. Taguchi, Y. Kozaki, K. Katanosaka, K. Mizumura, Compression-induced ATP release from rat skeletal muscle with and without lengthening contraction, Neurosci. Lett. 434 (2008) 277–281. [17] T. Taguchi, J. Sato, K. Mizumura, Augmented mechanical response of muscle thin-fiber sensory receptors recorded from rat muscle-nerve preparations in vitro after eccentric contraction, J. Neurophysiol. 94 (2005) 2822–2831. [18] K. Tsuji, Y. Nakamura, T. Ogata, A. Mitani, K. Kataoka, T. Shibata, M. Ishida, H. Shinozaki, Neurotoxicity of acromelic acid in cultured neurons from rat spinal cord, Neuroscience 68 (1995) 585–591. [19] S.G. Waxman, Nav1.7, its mutations, and the syndromes that they cause, Neurology 69 (2007) 505–507.