- Email: [email protected]
Characterization of spinal amino acid release and touch-evoked allodynia produced by spinal glycine or GABAA receptor antagonist
Neuroscience Vol. 95, No. 3, pp. 781–786, 2000 781 Copyright q 1999 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306-4522/00 $20.00+0.00
Spinal glycinergic/GABAergic regulation of glutamate release
Pergamon PII: S0306-4522(99)00461-3 www.elsevier.com/locate/neuroscience
CHARACTERIZATION OF SPINAL AMINO ACID RELEASE AND TOUCH-EVOKED ALLODYNIA PRODUCED BY SPINAL GLYCINE OR GABAA RECEPTOR ANTAGONIST T. ISHIKAWA, M. MARSALA, T. SAKABE and T. L. YAKSH* Department of Anesthesiology, University of California at San Diego, 9500 Gilman Drive, La Jolla, CA 92093, U.S.A.
Abstract—Intrathecal strychnine (glycine antagonist) or bicuculline (GABAA antagonist) yields a touch-evoked agitation that is blocked by N-methyl-d-aspartate receptor antagonism. We examined the effects of intrathecal strychnine and bicuculline on touchevoked agitation and the spinal release of amino acids. Fifty-two Sprague–Dawley rats were prepared under halothane anesthesia with a lumbar intrathecal catheter and a loop dialysis catheter. Four days after implantation, rats were randomized to receive an intrathecal injection of N-methyl-d-aspartate (3 mg), strychnine (3 mg) or bicuculline (10 mg), or a combination of N-methyl-daspartate with bicuculline or strychnine. The agitation produced by brief light tactile stroking of the flank (tactile allodynia), and the spontaneous spinal release of glutamate, taurine and serine was measured. Intrathecal N-methyl-d-aspartate, strychnine and bicuculline produced similar touch-evoked allodynia. Intrathecal bicuculline and N-methyl-d-aspartate alone evoked a transient spinal release of glutamate and taurine, but not serine, in the 0–10 min sample, while strychnine did not affect spinal transmitter release at any time. As GABAA but not glycine receptor inhibition at equi-allodynic doses increases glutamate release, while the allodynia of both is blocked by N-methyl-d-aspartate receptor antagonism, we hypothesize that GABAA sites regulate presynaptic glutamate release, while glycine regulates the excitability of neurons postsynaptic to glutamatergic terminals. q 1999 IBRO. Published by Elsevier Science Ltd. Key words: bicuculline, strychnine, tactile stimulation, touch-evoked agitation, modulatory receptor systems, excitatory amino acid release.
The response of spinal neurons to a given afferent stimulus can be modified by segmentally organized systems in the lumbar dorsal horn. 38,39 Electrophysiological studies support an inhibitory role for GABA and glycine (GLY) neurons in the spinal cord, suggesting that these neurons may mediate presynaptic inhibition of primary afferent terminals via an action at axoaxonic synapses, 28 and by a postsynaptic inhibition through action at axodendritic and axosomatic synapses. 37 There is a growing appreciation that these GABAergic and glycinergic systems may serve to regulate the encoding of the sensory message, and that such encoding is responsible for the contextual definition of the non-aversiveness of low-intensity tactile stimuli. Thus, acute antagonism of spinal GABAA receptors with bicuculline (BIC) and GLY receptors with strychnine (STR) will produce a clearly defined, somatotopically limited agitation (touch-evoked agitation, TEA). 22,23,40 Electrophysiologically, this effect of inhibitory amino acids corresponds to an augmented response of dorsal horn wide dynamic range neurons evoked by such light tactile stimuli. 9,25,39 It has been observed in spinal dorsal horn slices that myelinated fiber stimulation evoked short- and long-lasting inhibitory postsynaptic potentials which were respectively blocked by STR and BIC. Interestingly, these potentials appeared to be disynaptic and were blocked by non-Nmethyl-d-aspartate (non-NMDA) receptor antagonists. Such
an organization, although not yet completely defined, suggests that activity in myelinated afferent input may typically evoke a local inhibition. The ability of GABAA and GLY inhibition to induce an allodynic state and an enhanced discharge of dorsal horn neurons thus suggests that the normal encoding of myelinated afferent traffic as “non-aversive” is dependent upon this intrinsic activation of GABAA and/or glycinergic synapses. Characterization of the pharmacology of the tactile allodynia induced by spinal GABAA and GLY antagonists through the use of intrathecally delivered agents has revealed that it is mediated by activation of spinal NMDA receptors. Thus, intrathecally delivered NMDA antagonists will reduce the tactile allodynia produced by i.t. injection of inhibitory amino acid antagonists. 8,19,43 Conversely, the spinal delivery of NMDA will result in a well-defined hyperalgesia and allodynia. 5,14 Previous work has shown that peripheral inflammatory stimuli will evoke a hyperalgesia and this effect is diminished by spinal NMDA receptor antagonism. 43 In these models, when examined, the spinal release of glutamate (GLU) is significantly enhanced. 16,45 These observations jointly support the organizing hypothesis that hyperalgesic and allodynic states may be induced by changes in spinal GLU release. Such release may theoretically be enhanced by increased afferent input or by central changes that might lead to a reduced GABAergic or glycinergic tone. A direct assessment of this hypothesis would be to characterize the spinal release of GLU in the face of spinal GABAA receptor and GLY antagonism, which has been shown to induce tactile allodynia. To accomplish this aim, we examined the release of amino acids from the spinal cord in vivo using a loop microdialysis catheter placed into the lumbar i.t. space. 17
*To whom correspondence should be addressed. Tel.: 1 1-619-543-3597; fax: 1 1-619-543-6070. E-mail address: [email protected] (T. L. Yaksh) Abbreviations: ACSF, artificial cerebrospinal fluid; BIC, bicuculline; GLU, glutamate; GLY, glycine; NMDA, N-methyl-d-aspartate; SER, serine; STR, strychnine; TAU, taurine; TEA, touch-evoked agitation. 781
782
T. Ishikawa et al. EXPERIMENTAL PROCEDURES
This study was performed with approval from the Institutional Animal Care and Use Committee of the University of California at San Diego. Animal model and study paradigm Male Sprague–Dawley rats (52 rats) weighing 262–385 g (Harlan Industries, Indianapolis, IN) were implanted with a PE-10 catheter for drug injection along with a loop dialysis probe under 2.5% halothane anesthesia, according to the method described by Yaksh and Rudy 42 (see below) to permit injection of drugs and dialysis of the lumbar i.t. space, respectively. Rats with neurological dysfunction were killed. Experiments were initiated four days after implantation. To initiate the experiment, the previously prepared rat was lightly anesthetized with halothane (0.5–0.7% in 1:1 air/O2), delivered by a face mask. The dialysis was initiated. After an initial 20-min washout, dialysis samples were taken at 10-min intervals. After two 10-min baseline samples, the animal received an i.t. injection of STR, BIC or NMDA. Dialysis samples were then taken at 10-min intervals for the next 30 min. At this time, the animal was killed with an overdose of barbiturate and the position of the i.t. catheters verified by dissection.
evaluated in rats before and at 10, 20 and 30 min after drug injection. The flank of the rat was briefly stroked once in a rostral and caudal direction with the point of a probe. The response of the rat was graded according to the assessment described previously: 41 0, no response; 1, moderate effort to escape; 2, vigorous effort to escape, episodic bouts of vocalization during stroking; 3, strong and frequent effort to escape, continued and persistent vocalization that continued for a brief interval after termination of application of probe. For each time-point (i.e. 10, 20 and 30 min), the test was repeated twice with a 60-s interval so that the maximum possible score for each time-point was 6.
Statistics Statistical analysis employed a two-way ANOVA followed by Student’s Newman–Keuls test for multiple comparisons. For graphical presentation, release was represented as mean ^ S.E.M. of the percentage change from resting values (basal level). The baseline resting value for each rat was determined as the mean of the two samples collected immediately after the 30-min washout. Comparison of TEA score among drug groups was carried out by the Kruskal–Wallis test followed by the Mann–Whitney rank test.
Spinal drug injection STR sulfate (Sigma), BIC hydrochloride (Sigma) and NMDA (Sigma) were given i.t. with doses prepared to be delivered in a volume of 10 ml. Each i.t. injection was followed immediately by 10 ml of vehicle to flush the catheter. The doses employed (BIC: 10 mg; STR: 3 mg; NMDA: 3 mg) were selected on the basis of previous work indicating that these are the maximum subconvulsive doses and the ability of these doses to evoke a comparable tactile allodynia. 14,20,40 Dialysis and sample collection Construction of the i.t. loop dialysis probe has been described previously. 17 In brief, the catheter assembly consists of a loop of dialysis membrane tubing (10,000 mol. wt exclusion) that is connected to an inflow and outflow arm, each constructed from polyethylene tubing (0.3 mm o.d.). The loop catheters were inserted through an incision in the cisternal membrane and were passed to the rostral edge of the lumbar enlargement (approximately 8 cm). This implantation placed the uncoated (lower 2 cm) section of the dialysis membrane (active segment) of the dialysis catheter at the Th11–L5 spinal segments. The two arms of the dialysis membrane were externalized on the top of the head and the dialysis catheter was flushed with artificial cerebrospinal fluid (ACSF). For the dialysis experiments, the externalized end of one end of the PE-10 tubing was connected to a micro-infusion pump using a PE-50 tubing (inflow). The other end of the PE-10 tubing was attached to approximately 5 cm of PE-10 catheter (outflow). The dialysis system was then perfused with ACSF consisting of (in mM): 151 Na 1, 2.6 K 1, 0.9 Mg 21, 1.3 Ca 21, 123 Cl 2, 21 HCO2 3 , 2.5 HPO4 and 3.5 dextrose. The ACSF was bubbled with 5% O2 in 95% O2 to pH 7.2 and heated to 378C prior to the experiment. The flow rate was 10 ml/min in all experiments. Samples were collected at 10-min intervals in a polypropylene tube on ice and then frozen at 2208C until analysis by high-performance liquid chromatography–UV. For each 10-min sample (100 ml), 25 ml of the perfusate was used for amino acid determination. Chemical analysis GLU, taurine (TAU) and serine (SER) were analysed using a Waters high-performance liquid chromatograph with a reversed phase C18 column and a UV detector, using a phenyl isothiocyanate precolumn derivatization method. Sensitivity was 5–10 pmol per injection. Methionine sulfate was added as an internal standard. External standards containing 40, 400 and 4000 pmol of authentic amino acids were run. The amino acid peak heights were quantified based on a linear relationship between peak height and amounts of corresponding standards. Measurement of touch-evoked agitation The behavioral response to tactile stroking of the skin (TEA) was
RESULTS
Effects of intrathecal strychnine, bicuculline and N-methyl-daspartate on touch-evoked agitation The administration of vehicle (saline) had no effect on general behavior, and there was no TEA. In contrast, intrathecal STR, BIC and NMDA resulted in a prominent and comparable response to tactile stimuli applied to the flank and lower lumbar regions. In all groups, TEA appeared almost immediately and disappeared by 20 min after injection (Fig. 1). These drug-evoked increases in TEA were uniformly reversed or significantly suppressed by the i.t. delivery of dizocilpine maleate (non-competitive NMDA receptor antagonist; Fig. 1). TEA scores in rats receiving a combination of NMDA with either STR or BIC were only modestly enhanced compared with STR/BIC or NMDA alone (Fig. 2). Calculation of the ratio of the effect produced by each drug alone with the effect produced by the drug in combination gave values of 1.5 (NMDA/STR) and 1.97 (NMDA/BIC). These results suggest that co-delivery of the agents produced at most a moderately greater then additive interaction.
Spinal release of amino acids The concentration of GLU, TAU and SER in the lumbar dialysate prior to the delivery of a stimulus was 62 ^ 12, 126 ^ 35 and 184 ^ 29 pg/ml (n 6–12). There were no differences between treatment groups with regard to resting release prior to drug injection. Intrathecal BIC alone evoked a significant spinal release of GLU (Figs 3, 4) in the 0–10 min sample and this gradually declined to baseline level. A similar time-course of release was observed for TAU (Fig. 4). SER was unaltered during the ensuing 30-min perfusion interval in all animals examined (Fig. 4). In contrast to the effects of BIC and NMDA, STR was without effect on any measured spinal amino acid release (see Figs 3 and 4). Co-delivery of NMDA with STR or with BIC resulted in a modest additive effect in evoked glutamate release (Fig. 4).
Spinal glycinergic/GABAergic regulation of glutamate release
783
Fig. 1. Left: time-course for the effect of i.t. STR (3 mg; top), BIC (10 mg; middle) or NMDA (3 mg; bottom) on TEA score. The respective histogram (right) indicates the area under the time–effect curve for TEA expressed as a percentage of the maximum possible score for the interval 10–30 min. Asterisks indicate significant difference compared with control (score 0).
DISCUSSION
Spinal glycine and GABAA regulation of touch-evoked behaviors The present studies with the intrathecal antagonists are in accord with previous work which shows that the response to low-threshold afferent input may be tonically regulated by GABAA or GLY inhibition, and that these behavioral effects are mimicked by the spinal delivery of NMDA. Behavioral and electrophysiological studies have indicated that the i.t.
injection of subconvulsive doses of GABAA or GLY antagonists alter spinal function in such a manner that light, tactile stimuli will induce behavior that has the hallmarks of a pain state, even in the lightly anesthetized animal. These touchevoked changes include: (i) vocalization and organized efforts to escape; 34,41 (ii) increased blood pressure; 22,41 (iii) increased catecholamine release from the locus coeruleus; 18 and (iv) an exaggerated discharge of dorsal wide dynamic range neurons. 9,24,25,31,46 Where examined, these effects appear to be uniformly antagonized by NMDA receptor antagonists. 41
784
T. Ishikawa et al.
Fig. 2. Sum of TEA scores in all experimental groups. Final score in each group was calculated by adding individual TEA scores measured at 10, 20 and 30 min after injection. Note the presence of a modest but significant additive effect after co-administration of STR/NMDA. *P , 0.05.
The present studies indicate that, while these two inhibitory amino acid systems have effects upon spinal processing with similar behavioral consequences, they appear to be distinctly organized. Role of spinal glutamate receptors in tactile allodynia Increased GLU receptor activity can lead to spontaneous pain and to facilitated states of afferent processing in which low-threshold tactile stimuli, presumably activating largeafferent axons, can induce a state indicative of nociception. 1,2,4,5,14 Previous work has indicated that spinal GLU receptor activation evokes a hyperpathia that is mediated in part by a spinal cascade of intermediate events that lead to a state of spinal facilitation. This cascade includes the local release of prostanoids and nitric oxide, 14,15 which have been shown to sensitize the terminals and enhance transmitter release. 29,30,43 In addition, there is ample evidence that NMDA receptors may be located on the terminals of glutamatergic axons 12 and that such a presynaptic action would lead to an exaggerated release of this excitatory amino acid. Systems that serve to regulate spinal GLU release would accordingly be anticipated to predictably alter behavioral sensitivity to light touch. Characterization of glycine and GABA interneurons on allodynia in relation to spinal amino acid release Demonstration that the touch-evoked allodynia evoked by i.t. GABAA and glycine receptor antagonism is reversed by an i.t. NMDA antagonist, along with the role played by spinal NMDA, suggests that loss of GABAA/GLY inhibition may facilitate the release of GLU or enhance the postsynaptic excitability of the system activated by GLU. 35–37 Immunohistochemistry has shown that both GABAergic and glycinergic neurons are present in spinal Rexed laminae II–IV, and these fibers form axoaxonic synapses on Ab terminals and on the soma and dendrites of deeper magnocellular, wide dynamic range neurons. 28,35,36 As summarized schematically in Fig. 5, we hypothesize that, since inhibition of GABAA receptor activity leads to a significant increase in GLU release and a concurrent state of allodynia, GABAA receptors may regulate terminal excitability in a tonically active fashion. Loss of that inhibition results in an enhanced release of GLU. Whether the GLU measured in these dialysates arises from afferent terminals or from interneurons is not known. One unexpected finding in the present studies is the failure of STR to increase GLU release. While higher doses of STR
Fig. 3. Time-course of cerebrospinal fluid GLU release following i.t. NMDA (3 mg), STR (3 mg) and BIC (10 mg) injection. Data represent release expressed as a percentage of baseline. Asterisks indicate significant difference compared with baseline concentration. The cerebrospinal fluid GLU release was evoked by BIC but not by STR.
might have been effective, we emphasize that (i) the STR dose was clearly sufficient to induce an NMDA antagonist-sensitive tactile allodynia and (ii) the allodynia was comparable to that produced by NMDA or by BIC. These observations are thus consistent with the hypothesis that STR acts on a membrane that is postsynaptic to that terminal releasing GLU (see Fig. 5). STR-evoked allodynia would thus remain dependent upon GLU release and NMDA receptor occupancy, but the exaggerated component of the response would reflect an augmentation of the excitability of the cell possessing the postsynaptic NMDA receptor and not the terminal releasing GLU. Although a precise role of TAU in spinal nociceptive processing is not defined, it has been demonstrated that, within the spinal cord, infusion of substance P (1 mM) or kainate (1 mg) elicits release of GLU and aspartate, as well as TAU and GLY, in the dorsal gray matter or in spinal cerebrospinal fluid. 27,45 During development of peripheral inflammation, TAU increases in the extracellular space following release of GLU and substance P. 33 Similarly, in the brain, local delivery of NMDA stimulates a significant release of endogenous TAU in a Ca 21-dependent manner. 10 It has been suggested that TAU in turn inhibits the NMDAinduced Ca 21 influx into the intracellular space, thus providing homeostatic control under conditions of excessive excitation. 11 In accordance, it has been shown that TAU inhibits behaviors evoked by i.t. NMDA or kainate injection. 6,13,21 Measurement of spinal release by microdialysis The present study employed a loop dialysis catheter system, 17 which is implanted into lumbar i.t. space. The placement of the perfusion catheter in the i.t. space has a distinct advantage in that it presents a less invasive procedure than trans-spinal approaches. 26,32 This technique has been usefully employed to define the long-term resting and
Spinal glycinergic/GABAergic regulation of glutamate release
Fig. 4. The effect of i.t. NMDA (3 mg), STR (3 mg) and BIC (10 mg), and the combination of NMDA (3 mg) with either STR (3 mg) or BIC (10 mg) on cerebrospinal fluid GLU, TAU and SER release. Each column indicates the cumulative percentage change in cerebrospinal fluid glutamate for 30 min. The cerebrospinal fluid glutamate release with STR or with BIC was only modestly increased by the combination of NMDA.
stimulus-dependent release of amino acids into the spinal extracellular space following local stimulation (as with i.t. NMDA), as well as in the face of natural stimuli, such as formalin injected into the paw or knee joint inflammation. 16,44,45 Relevance of findings to other allodynic states An important correlate of these studies is that, after nerve injury, a prominent allodynia/hyperalgesia may develop, which is reversible by i.t. NMDA receptor antagonism. 4 It has been demonstrated that, after nerve injury, there is a reduction in GABA-immunoreactive neurons in the dorsal horn. 3,7 In recent studies, we have demonstrated that, after peripheral nerve injury, the observed development of tactile
785
Fig. 5. Schematic drawing of the hypothesized relationship of dorsal horn GABAA and GLY receptors, leading to an NMDA receptor-mediated allodynia with an enhanced release of GLU, being secondary only to the action of GABAA antagonism. Although GABA and GLY may be contained in different neurons, their reported co-containment suggests that the differential effects of the two antagonists on GLU release are accounted for by the different localization of the GABAA and GLY receptors. GABAA receptors are located both pre- and post-terminally to the glutamate-releasing neuron and the GLY receptor only post-terminally. The allodynia is believed to be mediated by large afferents. The excitatory drive into the output neuron may be from a large primary afferent or from an interneuron driven by a large primary afferent. Based on studies showing the role of NMDA in producing spinal facilitation, the GLU-releasing neurons may be driven by small afferents or from small afferent-driven interneurons (see text for further discussion).
allodynia is correlated with a significant increase in spinal GLU release (Marsala, Yang and Yaksh, unpublished observations). These observations jointly offer the possibility that the NMDA antagonist-sensitive allodynia may reflect the loss of such tonic GABAergic inhibition and an increase in spinal GLU release secondary to the loss of that inhibition. Acknowledgements—This work was supported in part by funds from the Ministry of Education of Japan, Grant-in-aid for scientific research, No. 0645444 (T.I.), and by NIH-NS32794 (M.M.) and NIH-NS16541 (T.L.Y.). We would like to thank Shelle Malkmus for skilled technical assistance, and Dr Steve Rossi and Alan Moore for analysis of the amino acids.
REFERENCES
1. Bach F. W., Chaplan S. R. and Yaksh T. L. (1994) Studies on the spinal pharmacology of a new model of allodynia. Ann. Neurol. 36, 288A. 2. Castro-Lopes J. M., Malcangio M., Pan B. H. and Bowery N. G. (1995) Complex changes of GABAA and GABAB receptor binding in the spinal cord dorsal horn following peripheral inflammation or neurectomy. Brain Res. 679, 289–297. 3. Castro-Lopes J. M., Tavares I. and Coimbra A. (1993) GABA decreases in the spinal cord dorsal horn after peripheral neurectomy. Brain Res. 620, 287–291. 4. Chaplan S. R., Malmberg A. B. and Yaksh T. L. (1997) Efficacy of spinal NMDA receptor antagonism in formalin hyperalgesia and nerve injury evoked allodynia in the rat. J. Pharmac. exp. Ther. 280, 829–838. 5. Coderre T. J. and Melzack R. (1991) Central neural mediators of secondary hyperalgesia following heat injury in rats: neuropeptides and excitatory amino acids. Neurosci. Lett. 131, 71–74. 6. Hornfeldt C. S., Smullin D. H., Schamber C. D., Sun X. and Larson A. A. (1992) Antinociceptive effects of intrathecal taurine and calcium in the mouse. Life Sci. 50, 1925–1934. 7. Ibuki T., Hama A. T., Wang X. T., Pappas G. D. and Sagen J. (1997) Loss of GABA-immunoreactivity in the spinal dorsal horn of rats with peripheral nerve injury and promotion of recovery by adrenal medullary grafts. Neuroscience 76, 845–858. 8. Khandwala H., Hodge E. and Loomis C. W. (1997) Comparable dose-dependent inhibition of AP-7 sensitive strychnine-induced allodynia and paw pinch-induced nociception by mexiletine in the rat. Pain 72, 299–308.
786 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46.
T. Ishikawa et al. Khayyat G. F., Yu U. J. and King R. B. (1975) Response patterns to noxious and non-noxious stimuli in rostral trigeminal relay nuclei. Brain Res. 97, 47–60. Lehmann A., Hagberg H. and Hamberger A. (1984) A role for taurine in the maintenance of homeostasis in the central nervous system during hyperexcitation? Neurosci. Lett. 52, 341–346. Lehmann A., Hagberg H., Lazarewicz J. W., Jacobson I. and Hamberger A. (1986) Alterations in extracellular amino acids and Ca 21 following excitotoxin administration and during status epilepticus. Adv. exp. Med. Biol. 203, 363–373. Liu H., Mantyh P. W. and Basbaum A. I. (1997) NMDA-receptor regulation of substance P release from primary afferent nociceptors. Nature 386, 721–724. Lombardini J. (1976) Regional and subcellular studies on taurine in the rat central nervous system. In Taurine (eds Huxtable R. and Barbeau A.), pp. 311–326. Raven, New York. Malmberg A. B. and Yaksh T. L. (1992) Hyperalgesia mediated by spinal glutamate or substance P receptor blocked by spinal cyclooxygenase inhibition. Science 257, 1276–1279. Malmberg A. B. and Yaksh T. L. (1993) Spinal nitric oxide synthesis inhibition blocks NMDA-induced thermal hyperalgesia and produces antinociception in the formalin test in rats. Pain 54, 291–300. Malmberg A. B. and Yaksh T. L. (1995) Cyclooxygenase inhibition and the spinal release of prostaglandin E2 and amino acids evoked by paw formalin injection: a microdialysis study in unanesthetized rats. J. Neurosci. 15, 2768–2776. Marsala M., Malmberg A. B. and Yaksh T. L. (1995) The spinal loop dialysis catheter: characterization of use in the unanesthetized rat. J. Neurosci. Meth. 62, 43–53. Milne B., Duggan S., Jhamandas K. and Loomis C. (1996) Innocuous hair deflection evokes a nociceptive-like activation of catechol oxidation in the rat locus coeruleus following intrathecal strychnine: a biochemical index of allodynia using in vivo voltammetry. Brain Res. 718, 198–202. Onaka M., Minami T., Nishihara I. and Ito S. (1996) Involvement of glutamate receptors in strychnine- and bicuculline-induced allodynia in conscious mice. Anesthesiology 84, 1215–1222. Partridge B. J., Chaplan S. R., Sakamoto E. and Yaksh T. L. (1998) Characterization of the effects of gabapentin and 3-isobutyl-gamma-aminobutyric acid on substance P-induced thermal hyperalgesia. Anesthesiology 88, 196–205. Serrano J. S., Serrano M. I., Guerrero M. R., Ruiz R. and Polo J. (1990) Antinociceptive effect of taurine and its inhibition by naxolone. Gen. Pharmac. 21, 333–336. Sherman S. E. and Loomis C. W. (1994) Morphine insensitive allodynia is produced by intrathecal strychnine in the lightly anesthetized rat. Pain 56, 17–29. Sherman S. E. and Loomis C. W. (1995) Strychnine-dependent allodynia in the urethane-anesthetized rat is segmentally distributed and prevented by intrathecal glycine and betaine. Can. J. Physiol. Pharmac. 73, 1698–1705. Sherman S. E. and Loomis C. W. (1996) Strychnine-sensitive modulation is selective for non-noxious somatosensory input in the spinal cord of the rat. Pain 66, 321–330. Sivilotti L. and Woolf C. J. (1994) The contribution of GABAA and glycine receptors to central sensitization: disinhibition and touch-evoked allodynia in the spinal cord. J. Neurophysiol. 72, 169–179. Skilling S. R., Smullin D. H., Beitz A. J. and Larson A. A. (1988) Extracellular amino acid concentrations in the dorsal spinal cord of freely moving rats following veratridine and nociceptive stimulation. J. Neurochem. 51, 127–132. Smullin D. H., Skilling S. R. and Larson A. A. (1990) Interactions between substance P, calcitonin gene-related peptide, taurine and excitatory amino acids in the spinal cord. Pain 42, 93–101. Solodkin M., Jime´nez I. and Rudomin P. (1984) Identification of common interneurons mediating pre- and postsynaptic inhibition in the cat spinal cord. Science 224, 1453–1456. Sorkin L. S. (1993) NMDA evokes an l-NAME sensitive spinal release of glutamate and citrulline. NeuroReport 4, 479–482. Sorkin L. S. and McAdoo D. J. (1993) Amino acids and serotonin are released into the lumbar spinal cord of the anesthetized cat following intradermal capsaicin injections. Brain Res. 607, 89–98. Sorkin L. S. and Puig S. (1996) Neuronal model of tactile allodynia produced by spinal strychnine: effects of excitatory amino acid receptor antagonists and a mu-opiate receptor agonist. Pain 68, 283–292. Sorkin L. S., Steinman J. L., Hughes M. G., Willis W. D. and McAdoo D. J. (1988) Microdialysis recovery of serotonin released in spinal cord dorsal horn. J. Neurosci. Meth. 23, 131–138. Sorkin L. S., Westlund K. N., Sluka K. A., Dougherty P. M. and Willis W. D. (1992) Neural changes in acute arthritis in monkeys. IV. Time-course of amino acid release into the lumbar dorsal horn. Brain Res. Rev. 17, 39–50. Sosnowski M. and Yaksh T. L. (1989) Role of spinal adenosine receptors in modulating the hyperesthesia produced by spinal glycine receptor antagonism. Anesth. Analg. 69, 587–592. Todd A. J. (1996) GABA and glycine in synaptic glomeruli of the rat spinal dorsal horn. Eur. J. Neurosci. 8, 2492–2498. Todd A. J. and Sullivan A. C. (1990) Light microscope study of the coexistence of GABA-like and glycine-like immunoreactivities in the spinal cord of the rat. J. comp. Neurol. 296, 496–505. Todd A. J., Watt C., Spike R. C. and Sieghart W. (1996) Colocalization of GABA, glycine, and their receptors at synapses in the rat spinal cord. J. Neurosci. 16, 974–982. Wall P. D. (1979) Substantia gelatinosa and the control of somato-sensory transmission. Int. Rehab. Med. 1, 106–110. Wall P. D. and Yaksh T. L. (1978) Effect of Lissauer tract stimulation on activity in dorsal roots and in ventral roots. Expl Neurol. 60, 570–583. Yaksh T. L. (1989) Behavioral and autonomic correlates of the tactile evoked allodynia produced by spinal glycine inhibition: effects of modulatory receptor systems and excitatory amino acid antagonists. Pain 37, 111–123. Yaksh T. L. (1997) Preclinical models of nociception. In Anesthesiology: Biologic Foundations (eds Yaksh T. L., Lynch C. III, Zapol W. M., Maize M., Biebuyck J. F. and Saidman L. J.), pp. 685–718. Lippincott-Raven, Philadelphia. Yaksh T. L. and Rudy T. A. (1976) Chronic catheterization of the spinal subarachnoid space. Physiol. Behav. 17, 1031–1036. Yamamoto T. and Yaksh T. L. (1992) Studies on the spinal interaction of morphine and the NMDA antagonist MK-801 on the hyperesthesia observed in a rat model of sciatic mononeuropathy. Neurosci. Lett. 135, 67–70. Yang L. C., Marsala M. and Yaksh T. L. (1996) Characterization of time course of spinal amino acids, citrulline and PGE2 release after carrageenan/ kaolin-induced knee joint inflammation: a chronic microdialysis study. Pain 67, 345–354. Yang L. C., Marsala M. and Yaksh T. L. (1996) Effect of spinal kainic acid receptor activation on spinal amino acid and prostaglandin E2 release in rat. Neuroscience 75, 453–461. Yokota T., Nishikawa N. and Nishikawa Y. (1979) Effects of strychnine upon different classes of trigeminal subnucleus caudalis neurons. Brain Res. 168, 430–434. (Accepted 7 September 1999)
Spinal glycinergic/GABAergic regulation of glutamate release
Pergamon PII: S0306-4522(99)00461-3 www.elsevier.com/locate/neuroscience
CHARACTERIZATION OF SPINAL AMINO ACID RELEASE AND TOUCH-EVOKED ALLODYNIA PRODUCED BY SPINAL GLYCINE OR GABAA RECEPTOR ANTAGONIST T. ISHIKAWA, M. MARSALA, T. SAKABE and T. L. YAKSH* Department of Anesthesiology, University of California at San Diego, 9500 Gilman Drive, La Jolla, CA 92093, U.S.A.
Abstract—Intrathecal strychnine (glycine antagonist) or bicuculline (GABAA antagonist) yields a touch-evoked agitation that is blocked by N-methyl-d-aspartate receptor antagonism. We examined the effects of intrathecal strychnine and bicuculline on touchevoked agitation and the spinal release of amino acids. Fifty-two Sprague–Dawley rats were prepared under halothane anesthesia with a lumbar intrathecal catheter and a loop dialysis catheter. Four days after implantation, rats were randomized to receive an intrathecal injection of N-methyl-d-aspartate (3 mg), strychnine (3 mg) or bicuculline (10 mg), or a combination of N-methyl-daspartate with bicuculline or strychnine. The agitation produced by brief light tactile stroking of the flank (tactile allodynia), and the spontaneous spinal release of glutamate, taurine and serine was measured. Intrathecal N-methyl-d-aspartate, strychnine and bicuculline produced similar touch-evoked allodynia. Intrathecal bicuculline and N-methyl-d-aspartate alone evoked a transient spinal release of glutamate and taurine, but not serine, in the 0–10 min sample, while strychnine did not affect spinal transmitter release at any time. As GABAA but not glycine receptor inhibition at equi-allodynic doses increases glutamate release, while the allodynia of both is blocked by N-methyl-d-aspartate receptor antagonism, we hypothesize that GABAA sites regulate presynaptic glutamate release, while glycine regulates the excitability of neurons postsynaptic to glutamatergic terminals. q 1999 IBRO. Published by Elsevier Science Ltd. Key words: bicuculline, strychnine, tactile stimulation, touch-evoked agitation, modulatory receptor systems, excitatory amino acid release.
The response of spinal neurons to a given afferent stimulus can be modified by segmentally organized systems in the lumbar dorsal horn. 38,39 Electrophysiological studies support an inhibitory role for GABA and glycine (GLY) neurons in the spinal cord, suggesting that these neurons may mediate presynaptic inhibition of primary afferent terminals via an action at axoaxonic synapses, 28 and by a postsynaptic inhibition through action at axodendritic and axosomatic synapses. 37 There is a growing appreciation that these GABAergic and glycinergic systems may serve to regulate the encoding of the sensory message, and that such encoding is responsible for the contextual definition of the non-aversiveness of low-intensity tactile stimuli. Thus, acute antagonism of spinal GABAA receptors with bicuculline (BIC) and GLY receptors with strychnine (STR) will produce a clearly defined, somatotopically limited agitation (touch-evoked agitation, TEA). 22,23,40 Electrophysiologically, this effect of inhibitory amino acids corresponds to an augmented response of dorsal horn wide dynamic range neurons evoked by such light tactile stimuli. 9,25,39 It has been observed in spinal dorsal horn slices that myelinated fiber stimulation evoked short- and long-lasting inhibitory postsynaptic potentials which were respectively blocked by STR and BIC. Interestingly, these potentials appeared to be disynaptic and were blocked by non-Nmethyl-d-aspartate (non-NMDA) receptor antagonists. Such
an organization, although not yet completely defined, suggests that activity in myelinated afferent input may typically evoke a local inhibition. The ability of GABAA and GLY inhibition to induce an allodynic state and an enhanced discharge of dorsal horn neurons thus suggests that the normal encoding of myelinated afferent traffic as “non-aversive” is dependent upon this intrinsic activation of GABAA and/or glycinergic synapses. Characterization of the pharmacology of the tactile allodynia induced by spinal GABAA and GLY antagonists through the use of intrathecally delivered agents has revealed that it is mediated by activation of spinal NMDA receptors. Thus, intrathecally delivered NMDA antagonists will reduce the tactile allodynia produced by i.t. injection of inhibitory amino acid antagonists. 8,19,43 Conversely, the spinal delivery of NMDA will result in a well-defined hyperalgesia and allodynia. 5,14 Previous work has shown that peripheral inflammatory stimuli will evoke a hyperalgesia and this effect is diminished by spinal NMDA receptor antagonism. 43 In these models, when examined, the spinal release of glutamate (GLU) is significantly enhanced. 16,45 These observations jointly support the organizing hypothesis that hyperalgesic and allodynic states may be induced by changes in spinal GLU release. Such release may theoretically be enhanced by increased afferent input or by central changes that might lead to a reduced GABAergic or glycinergic tone. A direct assessment of this hypothesis would be to characterize the spinal release of GLU in the face of spinal GABAA receptor and GLY antagonism, which has been shown to induce tactile allodynia. To accomplish this aim, we examined the release of amino acids from the spinal cord in vivo using a loop microdialysis catheter placed into the lumbar i.t. space. 17
*To whom correspondence should be addressed. Tel.: 1 1-619-543-3597; fax: 1 1-619-543-6070. E-mail address: [email protected] (T. L. Yaksh) Abbreviations: ACSF, artificial cerebrospinal fluid; BIC, bicuculline; GLU, glutamate; GLY, glycine; NMDA, N-methyl-d-aspartate; SER, serine; STR, strychnine; TAU, taurine; TEA, touch-evoked agitation. 781
782
T. Ishikawa et al. EXPERIMENTAL PROCEDURES
This study was performed with approval from the Institutional Animal Care and Use Committee of the University of California at San Diego. Animal model and study paradigm Male Sprague–Dawley rats (52 rats) weighing 262–385 g (Harlan Industries, Indianapolis, IN) were implanted with a PE-10 catheter for drug injection along with a loop dialysis probe under 2.5% halothane anesthesia, according to the method described by Yaksh and Rudy 42 (see below) to permit injection of drugs and dialysis of the lumbar i.t. space, respectively. Rats with neurological dysfunction were killed. Experiments were initiated four days after implantation. To initiate the experiment, the previously prepared rat was lightly anesthetized with halothane (0.5–0.7% in 1:1 air/O2), delivered by a face mask. The dialysis was initiated. After an initial 20-min washout, dialysis samples were taken at 10-min intervals. After two 10-min baseline samples, the animal received an i.t. injection of STR, BIC or NMDA. Dialysis samples were then taken at 10-min intervals for the next 30 min. At this time, the animal was killed with an overdose of barbiturate and the position of the i.t. catheters verified by dissection.
evaluated in rats before and at 10, 20 and 30 min after drug injection. The flank of the rat was briefly stroked once in a rostral and caudal direction with the point of a probe. The response of the rat was graded according to the assessment described previously: 41 0, no response; 1, moderate effort to escape; 2, vigorous effort to escape, episodic bouts of vocalization during stroking; 3, strong and frequent effort to escape, continued and persistent vocalization that continued for a brief interval after termination of application of probe. For each time-point (i.e. 10, 20 and 30 min), the test was repeated twice with a 60-s interval so that the maximum possible score for each time-point was 6.
Statistics Statistical analysis employed a two-way ANOVA followed by Student’s Newman–Keuls test for multiple comparisons. For graphical presentation, release was represented as mean ^ S.E.M. of the percentage change from resting values (basal level). The baseline resting value for each rat was determined as the mean of the two samples collected immediately after the 30-min washout. Comparison of TEA score among drug groups was carried out by the Kruskal–Wallis test followed by the Mann–Whitney rank test.
Spinal drug injection STR sulfate (Sigma), BIC hydrochloride (Sigma) and NMDA (Sigma) were given i.t. with doses prepared to be delivered in a volume of 10 ml. Each i.t. injection was followed immediately by 10 ml of vehicle to flush the catheter. The doses employed (BIC: 10 mg; STR: 3 mg; NMDA: 3 mg) were selected on the basis of previous work indicating that these are the maximum subconvulsive doses and the ability of these doses to evoke a comparable tactile allodynia. 14,20,40 Dialysis and sample collection Construction of the i.t. loop dialysis probe has been described previously. 17 In brief, the catheter assembly consists of a loop of dialysis membrane tubing (10,000 mol. wt exclusion) that is connected to an inflow and outflow arm, each constructed from polyethylene tubing (0.3 mm o.d.). The loop catheters were inserted through an incision in the cisternal membrane and were passed to the rostral edge of the lumbar enlargement (approximately 8 cm). This implantation placed the uncoated (lower 2 cm) section of the dialysis membrane (active segment) of the dialysis catheter at the Th11–L5 spinal segments. The two arms of the dialysis membrane were externalized on the top of the head and the dialysis catheter was flushed with artificial cerebrospinal fluid (ACSF). For the dialysis experiments, the externalized end of one end of the PE-10 tubing was connected to a micro-infusion pump using a PE-50 tubing (inflow). The other end of the PE-10 tubing was attached to approximately 5 cm of PE-10 catheter (outflow). The dialysis system was then perfused with ACSF consisting of (in mM): 151 Na 1, 2.6 K 1, 0.9 Mg 21, 1.3 Ca 21, 123 Cl 2, 21 HCO2 3 , 2.5 HPO4 and 3.5 dextrose. The ACSF was bubbled with 5% O2 in 95% O2 to pH 7.2 and heated to 378C prior to the experiment. The flow rate was 10 ml/min in all experiments. Samples were collected at 10-min intervals in a polypropylene tube on ice and then frozen at 2208C until analysis by high-performance liquid chromatography–UV. For each 10-min sample (100 ml), 25 ml of the perfusate was used for amino acid determination. Chemical analysis GLU, taurine (TAU) and serine (SER) were analysed using a Waters high-performance liquid chromatograph with a reversed phase C18 column and a UV detector, using a phenyl isothiocyanate precolumn derivatization method. Sensitivity was 5–10 pmol per injection. Methionine sulfate was added as an internal standard. External standards containing 40, 400 and 4000 pmol of authentic amino acids were run. The amino acid peak heights were quantified based on a linear relationship between peak height and amounts of corresponding standards. Measurement of touch-evoked agitation The behavioral response to tactile stroking of the skin (TEA) was
RESULTS
Effects of intrathecal strychnine, bicuculline and N-methyl-daspartate on touch-evoked agitation The administration of vehicle (saline) had no effect on general behavior, and there was no TEA. In contrast, intrathecal STR, BIC and NMDA resulted in a prominent and comparable response to tactile stimuli applied to the flank and lower lumbar regions. In all groups, TEA appeared almost immediately and disappeared by 20 min after injection (Fig. 1). These drug-evoked increases in TEA were uniformly reversed or significantly suppressed by the i.t. delivery of dizocilpine maleate (non-competitive NMDA receptor antagonist; Fig. 1). TEA scores in rats receiving a combination of NMDA with either STR or BIC were only modestly enhanced compared with STR/BIC or NMDA alone (Fig. 2). Calculation of the ratio of the effect produced by each drug alone with the effect produced by the drug in combination gave values of 1.5 (NMDA/STR) and 1.97 (NMDA/BIC). These results suggest that co-delivery of the agents produced at most a moderately greater then additive interaction.
Spinal release of amino acids The concentration of GLU, TAU and SER in the lumbar dialysate prior to the delivery of a stimulus was 62 ^ 12, 126 ^ 35 and 184 ^ 29 pg/ml (n 6–12). There were no differences between treatment groups with regard to resting release prior to drug injection. Intrathecal BIC alone evoked a significant spinal release of GLU (Figs 3, 4) in the 0–10 min sample and this gradually declined to baseline level. A similar time-course of release was observed for TAU (Fig. 4). SER was unaltered during the ensuing 30-min perfusion interval in all animals examined (Fig. 4). In contrast to the effects of BIC and NMDA, STR was without effect on any measured spinal amino acid release (see Figs 3 and 4). Co-delivery of NMDA with STR or with BIC resulted in a modest additive effect in evoked glutamate release (Fig. 4).
Spinal glycinergic/GABAergic regulation of glutamate release
783
Fig. 1. Left: time-course for the effect of i.t. STR (3 mg; top), BIC (10 mg; middle) or NMDA (3 mg; bottom) on TEA score. The respective histogram (right) indicates the area under the time–effect curve for TEA expressed as a percentage of the maximum possible score for the interval 10–30 min. Asterisks indicate significant difference compared with control (score 0).
DISCUSSION
Spinal glycine and GABAA regulation of touch-evoked behaviors The present studies with the intrathecal antagonists are in accord with previous work which shows that the response to low-threshold afferent input may be tonically regulated by GABAA or GLY inhibition, and that these behavioral effects are mimicked by the spinal delivery of NMDA. Behavioral and electrophysiological studies have indicated that the i.t.
injection of subconvulsive doses of GABAA or GLY antagonists alter spinal function in such a manner that light, tactile stimuli will induce behavior that has the hallmarks of a pain state, even in the lightly anesthetized animal. These touchevoked changes include: (i) vocalization and organized efforts to escape; 34,41 (ii) increased blood pressure; 22,41 (iii) increased catecholamine release from the locus coeruleus; 18 and (iv) an exaggerated discharge of dorsal wide dynamic range neurons. 9,24,25,31,46 Where examined, these effects appear to be uniformly antagonized by NMDA receptor antagonists. 41
784
T. Ishikawa et al.
Fig. 2. Sum of TEA scores in all experimental groups. Final score in each group was calculated by adding individual TEA scores measured at 10, 20 and 30 min after injection. Note the presence of a modest but significant additive effect after co-administration of STR/NMDA. *P , 0.05.
The present studies indicate that, while these two inhibitory amino acid systems have effects upon spinal processing with similar behavioral consequences, they appear to be distinctly organized. Role of spinal glutamate receptors in tactile allodynia Increased GLU receptor activity can lead to spontaneous pain and to facilitated states of afferent processing in which low-threshold tactile stimuli, presumably activating largeafferent axons, can induce a state indicative of nociception. 1,2,4,5,14 Previous work has indicated that spinal GLU receptor activation evokes a hyperpathia that is mediated in part by a spinal cascade of intermediate events that lead to a state of spinal facilitation. This cascade includes the local release of prostanoids and nitric oxide, 14,15 which have been shown to sensitize the terminals and enhance transmitter release. 29,30,43 In addition, there is ample evidence that NMDA receptors may be located on the terminals of glutamatergic axons 12 and that such a presynaptic action would lead to an exaggerated release of this excitatory amino acid. Systems that serve to regulate spinal GLU release would accordingly be anticipated to predictably alter behavioral sensitivity to light touch. Characterization of glycine and GABA interneurons on allodynia in relation to spinal amino acid release Demonstration that the touch-evoked allodynia evoked by i.t. GABAA and glycine receptor antagonism is reversed by an i.t. NMDA antagonist, along with the role played by spinal NMDA, suggests that loss of GABAA/GLY inhibition may facilitate the release of GLU or enhance the postsynaptic excitability of the system activated by GLU. 35–37 Immunohistochemistry has shown that both GABAergic and glycinergic neurons are present in spinal Rexed laminae II–IV, and these fibers form axoaxonic synapses on Ab terminals and on the soma and dendrites of deeper magnocellular, wide dynamic range neurons. 28,35,36 As summarized schematically in Fig. 5, we hypothesize that, since inhibition of GABAA receptor activity leads to a significant increase in GLU release and a concurrent state of allodynia, GABAA receptors may regulate terminal excitability in a tonically active fashion. Loss of that inhibition results in an enhanced release of GLU. Whether the GLU measured in these dialysates arises from afferent terminals or from interneurons is not known. One unexpected finding in the present studies is the failure of STR to increase GLU release. While higher doses of STR
Fig. 3. Time-course of cerebrospinal fluid GLU release following i.t. NMDA (3 mg), STR (3 mg) and BIC (10 mg) injection. Data represent release expressed as a percentage of baseline. Asterisks indicate significant difference compared with baseline concentration. The cerebrospinal fluid GLU release was evoked by BIC but not by STR.
might have been effective, we emphasize that (i) the STR dose was clearly sufficient to induce an NMDA antagonist-sensitive tactile allodynia and (ii) the allodynia was comparable to that produced by NMDA or by BIC. These observations are thus consistent with the hypothesis that STR acts on a membrane that is postsynaptic to that terminal releasing GLU (see Fig. 5). STR-evoked allodynia would thus remain dependent upon GLU release and NMDA receptor occupancy, but the exaggerated component of the response would reflect an augmentation of the excitability of the cell possessing the postsynaptic NMDA receptor and not the terminal releasing GLU. Although a precise role of TAU in spinal nociceptive processing is not defined, it has been demonstrated that, within the spinal cord, infusion of substance P (1 mM) or kainate (1 mg) elicits release of GLU and aspartate, as well as TAU and GLY, in the dorsal gray matter or in spinal cerebrospinal fluid. 27,45 During development of peripheral inflammation, TAU increases in the extracellular space following release of GLU and substance P. 33 Similarly, in the brain, local delivery of NMDA stimulates a significant release of endogenous TAU in a Ca 21-dependent manner. 10 It has been suggested that TAU in turn inhibits the NMDAinduced Ca 21 influx into the intracellular space, thus providing homeostatic control under conditions of excessive excitation. 11 In accordance, it has been shown that TAU inhibits behaviors evoked by i.t. NMDA or kainate injection. 6,13,21 Measurement of spinal release by microdialysis The present study employed a loop dialysis catheter system, 17 which is implanted into lumbar i.t. space. The placement of the perfusion catheter in the i.t. space has a distinct advantage in that it presents a less invasive procedure than trans-spinal approaches. 26,32 This technique has been usefully employed to define the long-term resting and
Spinal glycinergic/GABAergic regulation of glutamate release
Fig. 4. The effect of i.t. NMDA (3 mg), STR (3 mg) and BIC (10 mg), and the combination of NMDA (3 mg) with either STR (3 mg) or BIC (10 mg) on cerebrospinal fluid GLU, TAU and SER release. Each column indicates the cumulative percentage change in cerebrospinal fluid glutamate for 30 min. The cerebrospinal fluid glutamate release with STR or with BIC was only modestly increased by the combination of NMDA.
stimulus-dependent release of amino acids into the spinal extracellular space following local stimulation (as with i.t. NMDA), as well as in the face of natural stimuli, such as formalin injected into the paw or knee joint inflammation. 16,44,45 Relevance of findings to other allodynic states An important correlate of these studies is that, after nerve injury, a prominent allodynia/hyperalgesia may develop, which is reversible by i.t. NMDA receptor antagonism. 4 It has been demonstrated that, after nerve injury, there is a reduction in GABA-immunoreactive neurons in the dorsal horn. 3,7 In recent studies, we have demonstrated that, after peripheral nerve injury, the observed development of tactile
785
Fig. 5. Schematic drawing of the hypothesized relationship of dorsal horn GABAA and GLY receptors, leading to an NMDA receptor-mediated allodynia with an enhanced release of GLU, being secondary only to the action of GABAA antagonism. Although GABA and GLY may be contained in different neurons, their reported co-containment suggests that the differential effects of the two antagonists on GLU release are accounted for by the different localization of the GABAA and GLY receptors. GABAA receptors are located both pre- and post-terminally to the glutamate-releasing neuron and the GLY receptor only post-terminally. The allodynia is believed to be mediated by large afferents. The excitatory drive into the output neuron may be from a large primary afferent or from an interneuron driven by a large primary afferent. Based on studies showing the role of NMDA in producing spinal facilitation, the GLU-releasing neurons may be driven by small afferents or from small afferent-driven interneurons (see text for further discussion).
allodynia is correlated with a significant increase in spinal GLU release (Marsala, Yang and Yaksh, unpublished observations). These observations jointly offer the possibility that the NMDA antagonist-sensitive allodynia may reflect the loss of such tonic GABAergic inhibition and an increase in spinal GLU release secondary to the loss of that inhibition. Acknowledgements—This work was supported in part by funds from the Ministry of Education of Japan, Grant-in-aid for scientific research, No. 0645444 (T.I.), and by NIH-NS32794 (M.M.) and NIH-NS16541 (T.L.Y.). We would like to thank Shelle Malkmus for skilled technical assistance, and Dr Steve Rossi and Alan Moore for analysis of the amino acids.
REFERENCES
1. Bach F. W., Chaplan S. R. and Yaksh T. L. (1994) Studies on the spinal pharmacology of a new model of allodynia. Ann. Neurol. 36, 288A. 2. Castro-Lopes J. M., Malcangio M., Pan B. H. and Bowery N. G. (1995) Complex changes of GABAA and GABAB receptor binding in the spinal cord dorsal horn following peripheral inflammation or neurectomy. Brain Res. 679, 289–297. 3. Castro-Lopes J. M., Tavares I. and Coimbra A. (1993) GABA decreases in the spinal cord dorsal horn after peripheral neurectomy. Brain Res. 620, 287–291. 4. Chaplan S. R., Malmberg A. B. and Yaksh T. L. (1997) Efficacy of spinal NMDA receptor antagonism in formalin hyperalgesia and nerve injury evoked allodynia in the rat. J. Pharmac. exp. Ther. 280, 829–838. 5. Coderre T. J. and Melzack R. (1991) Central neural mediators of secondary hyperalgesia following heat injury in rats: neuropeptides and excitatory amino acids. Neurosci. Lett. 131, 71–74. 6. Hornfeldt C. S., Smullin D. H., Schamber C. D., Sun X. and Larson A. A. (1992) Antinociceptive effects of intrathecal taurine and calcium in the mouse. Life Sci. 50, 1925–1934. 7. Ibuki T., Hama A. T., Wang X. T., Pappas G. D. and Sagen J. (1997) Loss of GABA-immunoreactivity in the spinal dorsal horn of rats with peripheral nerve injury and promotion of recovery by adrenal medullary grafts. Neuroscience 76, 845–858. 8. Khandwala H., Hodge E. and Loomis C. W. (1997) Comparable dose-dependent inhibition of AP-7 sensitive strychnine-induced allodynia and paw pinch-induced nociception by mexiletine in the rat. Pain 72, 299–308.
786 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46.
T. Ishikawa et al. Khayyat G. F., Yu U. J. and King R. B. (1975) Response patterns to noxious and non-noxious stimuli in rostral trigeminal relay nuclei. Brain Res. 97, 47–60. Lehmann A., Hagberg H. and Hamberger A. (1984) A role for taurine in the maintenance of homeostasis in the central nervous system during hyperexcitation? Neurosci. Lett. 52, 341–346. Lehmann A., Hagberg H., Lazarewicz J. W., Jacobson I. and Hamberger A. (1986) Alterations in extracellular amino acids and Ca 21 following excitotoxin administration and during status epilepticus. Adv. exp. Med. Biol. 203, 363–373. Liu H., Mantyh P. W. and Basbaum A. I. (1997) NMDA-receptor regulation of substance P release from primary afferent nociceptors. Nature 386, 721–724. Lombardini J. (1976) Regional and subcellular studies on taurine in the rat central nervous system. In Taurine (eds Huxtable R. and Barbeau A.), pp. 311–326. Raven, New York. Malmberg A. B. and Yaksh T. L. (1992) Hyperalgesia mediated by spinal glutamate or substance P receptor blocked by spinal cyclooxygenase inhibition. Science 257, 1276–1279. Malmberg A. B. and Yaksh T. L. (1993) Spinal nitric oxide synthesis inhibition blocks NMDA-induced thermal hyperalgesia and produces antinociception in the formalin test in rats. Pain 54, 291–300. Malmberg A. B. and Yaksh T. L. (1995) Cyclooxygenase inhibition and the spinal release of prostaglandin E2 and amino acids evoked by paw formalin injection: a microdialysis study in unanesthetized rats. J. Neurosci. 15, 2768–2776. Marsala M., Malmberg A. B. and Yaksh T. L. (1995) The spinal loop dialysis catheter: characterization of use in the unanesthetized rat. J. Neurosci. Meth. 62, 43–53. Milne B., Duggan S., Jhamandas K. and Loomis C. (1996) Innocuous hair deflection evokes a nociceptive-like activation of catechol oxidation in the rat locus coeruleus following intrathecal strychnine: a biochemical index of allodynia using in vivo voltammetry. Brain Res. 718, 198–202. Onaka M., Minami T., Nishihara I. and Ito S. (1996) Involvement of glutamate receptors in strychnine- and bicuculline-induced allodynia in conscious mice. Anesthesiology 84, 1215–1222. Partridge B. J., Chaplan S. R., Sakamoto E. and Yaksh T. L. (1998) Characterization of the effects of gabapentin and 3-isobutyl-gamma-aminobutyric acid on substance P-induced thermal hyperalgesia. Anesthesiology 88, 196–205. Serrano J. S., Serrano M. I., Guerrero M. R., Ruiz R. and Polo J. (1990) Antinociceptive effect of taurine and its inhibition by naxolone. Gen. Pharmac. 21, 333–336. Sherman S. E. and Loomis C. W. (1994) Morphine insensitive allodynia is produced by intrathecal strychnine in the lightly anesthetized rat. Pain 56, 17–29. Sherman S. E. and Loomis C. W. (1995) Strychnine-dependent allodynia in the urethane-anesthetized rat is segmentally distributed and prevented by intrathecal glycine and betaine. Can. J. Physiol. Pharmac. 73, 1698–1705. Sherman S. E. and Loomis C. W. (1996) Strychnine-sensitive modulation is selective for non-noxious somatosensory input in the spinal cord of the rat. Pain 66, 321–330. Sivilotti L. and Woolf C. J. (1994) The contribution of GABAA and glycine receptors to central sensitization: disinhibition and touch-evoked allodynia in the spinal cord. J. Neurophysiol. 72, 169–179. Skilling S. R., Smullin D. H., Beitz A. J. and Larson A. A. (1988) Extracellular amino acid concentrations in the dorsal spinal cord of freely moving rats following veratridine and nociceptive stimulation. J. Neurochem. 51, 127–132. Smullin D. H., Skilling S. R. and Larson A. A. (1990) Interactions between substance P, calcitonin gene-related peptide, taurine and excitatory amino acids in the spinal cord. Pain 42, 93–101. Solodkin M., Jime´nez I. and Rudomin P. (1984) Identification of common interneurons mediating pre- and postsynaptic inhibition in the cat spinal cord. Science 224, 1453–1456. Sorkin L. S. (1993) NMDA evokes an l-NAME sensitive spinal release of glutamate and citrulline. NeuroReport 4, 479–482. Sorkin L. S. and McAdoo D. J. (1993) Amino acids and serotonin are released into the lumbar spinal cord of the anesthetized cat following intradermal capsaicin injections. Brain Res. 607, 89–98. Sorkin L. S. and Puig S. (1996) Neuronal model of tactile allodynia produced by spinal strychnine: effects of excitatory amino acid receptor antagonists and a mu-opiate receptor agonist. Pain 68, 283–292. Sorkin L. S., Steinman J. L., Hughes M. G., Willis W. D. and McAdoo D. J. (1988) Microdialysis recovery of serotonin released in spinal cord dorsal horn. J. Neurosci. Meth. 23, 131–138. Sorkin L. S., Westlund K. N., Sluka K. A., Dougherty P. M. and Willis W. D. (1992) Neural changes in acute arthritis in monkeys. IV. Time-course of amino acid release into the lumbar dorsal horn. Brain Res. Rev. 17, 39–50. Sosnowski M. and Yaksh T. L. (1989) Role of spinal adenosine receptors in modulating the hyperesthesia produced by spinal glycine receptor antagonism. Anesth. Analg. 69, 587–592. Todd A. J. (1996) GABA and glycine in synaptic glomeruli of the rat spinal dorsal horn. Eur. J. Neurosci. 8, 2492–2498. Todd A. J. and Sullivan A. C. (1990) Light microscope study of the coexistence of GABA-like and glycine-like immunoreactivities in the spinal cord of the rat. J. comp. Neurol. 296, 496–505. Todd A. J., Watt C., Spike R. C. and Sieghart W. (1996) Colocalization of GABA, glycine, and their receptors at synapses in the rat spinal cord. J. Neurosci. 16, 974–982. Wall P. D. (1979) Substantia gelatinosa and the control of somato-sensory transmission. Int. Rehab. Med. 1, 106–110. Wall P. D. and Yaksh T. L. (1978) Effect of Lissauer tract stimulation on activity in dorsal roots and in ventral roots. Expl Neurol. 60, 570–583. Yaksh T. L. (1989) Behavioral and autonomic correlates of the tactile evoked allodynia produced by spinal glycine inhibition: effects of modulatory receptor systems and excitatory amino acid antagonists. Pain 37, 111–123. Yaksh T. L. (1997) Preclinical models of nociception. In Anesthesiology: Biologic Foundations (eds Yaksh T. L., Lynch C. III, Zapol W. M., Maize M., Biebuyck J. F. and Saidman L. J.), pp. 685–718. Lippincott-Raven, Philadelphia. Yaksh T. L. and Rudy T. A. (1976) Chronic catheterization of the spinal subarachnoid space. Physiol. Behav. 17, 1031–1036. Yamamoto T. and Yaksh T. L. (1992) Studies on the spinal interaction of morphine and the NMDA antagonist MK-801 on the hyperesthesia observed in a rat model of sciatic mononeuropathy. Neurosci. Lett. 135, 67–70. Yang L. C., Marsala M. and Yaksh T. L. (1996) Characterization of time course of spinal amino acids, citrulline and PGE2 release after carrageenan/ kaolin-induced knee joint inflammation: a chronic microdialysis study. Pain 67, 345–354. Yang L. C., Marsala M. and Yaksh T. L. (1996) Effect of spinal kainic acid receptor activation on spinal amino acid and prostaglandin E2 release in rat. Neuroscience 75, 453–461. Yokota T., Nishikawa N. and Nishikawa Y. (1979) Effects of strychnine upon different classes of trigeminal subnucleus caudalis neurons. Brain Res. 168, 430–434. (Accepted 7 September 1999)