- Email: [email protected]
Nerve injury induces plasticity that results in spinal inhibitory effects of galanin
Pain 98 (2002) 249–258 www.elsevier.com/locate/pain
Nerve injury induces plasticity that results in spinal inhibitory effects of galanin Sarah J.L. Flatters a,*, Alyson J. Fox b, Anthony H. Dickenson a a
Department of Pharmacology, University College London, Gower Street, London WC1E 6BT, UK Novartis Institute for Medical Sciences, University College London, Gower Place, London WC1E 6BS, UK
b
Accepted 27 November 2001
Abstract Galanin is a 29-amino-acid neuropeptide that has been implicated in the processes of nociception. This study examines the effect of exogenous galanin on dorsal horn neurone activity in vivo in the spinal nerve ligation (SNL) model of neuropathic pain. SNL rats but not naive or sham-operated rats exhibited behaviour indicative of allodynia. In anaesthetized rats, extracellular recordings were made from individual convergent dorsal horn neurones following stimulation of peripheral receptive fields electrically or with natural (innocuous mechanical, noxious mechanical and noxious thermal) stimuli. Spinal administration of galanin (0.5–50 mg) caused a slight facilitation of the neuronal responses to natural and electrical stimuli in naive rats and up to a 65% inhibition of neuronal responses in sham-operated rats following 50 mg galanin. In contrast, there was a marked inhibition of up to 80% of responses to both natural and electrical stimuli in SNL rats following spinal galanin administration. These results suggest that following peripheral nerve injury, there is plasticity in the levels of galanin and/or its receptors at spinal cord level so that the effect of exogenous galanin favours inhibitory function. q 2002 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved. Keywords: Galanin; Nociception; Spinal cord; Neuropathic pain; Rat; Electrophysiology
1. Introduction Galanin is a neuropeptide consisting of 29 amino acids (30 in humans) originally isolated from porcine gut (Tatemoto et al., 1983). It is widely distributed in the central and peripheral nervous systems (Merchenthaler et al., 1993) and has been shown to play a role in several key functions such ¨ gren et al., 1992), memory (Crawley, 1996), as learning (O feeding, and endocrine modulation (for review see Bartfai et al., 1993). In recent years three galanin receptors subtypes (for review see Branchek et al., 2000) have been cloned and termed GalR1 (Habert-Ortholi et al., 1994; Parker et al., 1995), GalR2 (Smith et al., 1997) and GalR3 (Wang et al., 1997; Smith et al., 1998). Activation of GalR1 and GalR3 receptors expressed in cell lines leads to inhibition of cAMP levels via coupling to Gi-protein (Smith et al., 1998; Wang et al., 1998). In contrast, GalR2 has been shown to be excitatory, activating phospholipase C and increasing phosphoinositol and intracellular calcium levels via coupling to Go/Gq-proteins (Wang et al., 1998). The * Corresponding author. Tel.: 144-207-679-3737; fax: 144-207-6793742. E-mail address: [email protected] (S.J.L. Flatters).
expression of the three receptor types has been examined at the mRNA level in rats. Whilst GalR1 is restricted predominantly to the central nervous system. GalR2 is more widespread with expression in the CNS as well as several peripheral tissues. GalR3 appears to be expressed in the periphery with CNS expression being restricted to certain areas (Waters and Krause, 2000). Of relevance to sensory processing, GalR1 and GalR2 have been shown to be present in the adult rat spinal cord and dorsal root ganglion (DRG) (Shi et al., 1997; O’Donnell et al., 1999). GalR1 expression is limited to the superficial dorsal horn and is mainly, but not exclusively, found on large DRG neurones, whilst GalR2 is expressed throughout the dorsal horn and is present on small and medium DRG neurones (O’Donnell et al., 1999). In contrast, GalR3 is expressed at low levels in the spinal cord and in DRG (Waters and Krause, 2000). Much interest has arisen in the role of galanin in nociception, recently reviewed by Xu et al. (2000a). In functional studies, galanin has been shown to have inhibitory effects on ventral root potentials in isolated spinal cord preparations of the newborn rat (Yanagisawa et al., 1986; Nussbaumer et al., 1989) and inhibit tail-flick and hot-plate responses in normal mice following intrathecal administration (Post et al., 1988). Other electrophysiological studies in normal
0304-3959/02/$20.00 q 2002 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved. PII: S 0304-395 9(02)00180-X
250
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
rats report galanin eliciting facilitatory effects, with depolarization of DRG neuronal membranes (Puttick et al., 1994) and facilitation of electrically-evoked responses of spinal neurones following spinal galanin administration (Reeve et al., 2000). In addition, it has been demonstrated that intrathecal galanin produces mechanical hyperalgesia in normal rats (Kuraishi et al., 1991b). Biphasic effects of galanin have also been reported (Wiesenfeld-Hallin et al., 1989a), with low doses causing facilitatory effects and higher doses resulting in inhibition. Nerve injury and inflammation affects the expression of both galanin and its receptors. Following axotomy, partial nerve transection (PSNL) and chronic constriction injury (CCI) in the rat there is a marked up-regulation of galanin levels in the dorsal root ganglion (DRG), superficial dorsal horn and gracile nucleus (Ho¨ kfelt et al., 1987; Ma and Bisby, 1997). Carrageenan-induced inflammation results in an increase in galanin mRNA levels in the superficial dorsal horn, but a reduction in the DRG (Ji et al., 1995). In contrast to the effects on galanin expression, following axotomy GalR1 and GalR2 mRNA levels are decreased in the DRG (Xu et al., 1996; Shi et al., 1997). Inflammation also results in a decrease in GalR1 mRNA levels in adult DRG neurones (Xu et al., 1996), although GalR2 mRNA is up-regulated (Shi et al., 1997). The effect of nerve injury and inflammation on GalR3 expression has not been reported. Previous behavioural studies have also investigated the role of galanin in inflammatory and neuropathic pain models. Anti-galanin antibody has been reported to reverse carrageenan-induced hyperalgesia (Kuraishi et al., 1991a; Satoh et al., 1992) implying pronociceptive effects of galanin whereas intrathecal galanin administration has been shown to alleviate mechanical and cold allodynia in photochemically nerve-injured rats (Hao et al., 1999). Although there is a marked up-regulation of galanin following nerve injury, it remains to be shown whether this up-regulation has functional consequences, particularly in the light of the down-regulation of some galanin receptor populations in the DRG after axotomy. Thus the aim of this study is to investigate the effects of exogenous galanin in vivo on the responses of convergent dorsal horn neurones in normal, sham-operated and neuropathic (spinal nerve ligated; SNL) rats to electrical, mechanical and thermal stimulation. This model of nerve injury uses a partial ligation of the nerve and is sensitive to drugs used in the treatment of neuropathic pain in humans (Chapman et al., 1998a).
2. Materials and methods 2.1. Animal preparation Male Sprague–Dawley rats were housed in groups of five in plastic cages with artificial lighting which had a fixed 12:12-h light/dark cycle. Food and water were available
ad libitum. Guidelines for animal research by the UK Home Office and the International Association for the Study of Pain (Zimmermann, 1983) were adhered to. The spinal nerve ligation (SNL) model of neuropathic pain first described by Kim and Chung (1992) was used. Briefly, rats (130–150 g) were anaesthetized with halothane (Fluothane, Zeneca, UK) in a 50% O2/50% N2O gaseous mixture. The L5 and L6 spinal nerves were exposed and tightly ligated using 6-0 silk and the L4 sciatic nerve branch was left uninjured. The muscle and skin were then closed with sutures and wound clips. The surgery for the sham-operated rats was identical to the SNL rats with the exception of the ligation of L5 and L6 spinal nerves. The animals were allowed to recover and were monitored for 2 weeks post-surgery. Behavioural testing was performed on days 2, 4, 7, 9, 11 and 14 following surgery to assess the development of mechanical and cold allodynia. Rats were placed in individual boxes with a metal mesh floor and allowed to acclimatize. Von Frey hairs with bending forces of 1, 5 and 9 g (Semmes–Weinstein monofilaments) were applied to the plantar surface of both hindpaws in ascending order of force. Each hair was applied ten times with 5-s intervals and the number of withdrawal responses recorded. Cold allodynia was assessed by the response to application of acetone to the plantar surface of both hindpaws. Acetone was applied a total of five times, with 3–4 min left between applications to allow evaporation and so elicit a cooling effect. A withdrawal response was observed as a shaking, flicking or licking of the paw following acetone application. The outcome of these behavioural tests is expressed as a difference score (number of contralateral withdrawals minus number of ipsilateral withdrawals). 2.2. In vivo electrophysiology Experiments were performed 14–18 days after the surgery and on uninjured (normal) animals of a similar weight using methods previously described by this laboratory (Chapman et al., 1998b). Rats were initially anaesthetized with 3.5% halothane in a 33% O2/66% N2O gaseous mixture until areflexia was produced. Cannulation of the trachea was performed to ensure adequate and efficient respiration of the rat. Rats were secured in a stereotaxic frame and a laminectomy was performed to expose segments L4–L5 of the spinal cord. The dura of the exposed spinal cord was then removed and the spine was held rigid by clamps caudal and rostral to the laminectomy. The halothane concentration was lowered (2.5–2.8%) whilst the surgery was performed, and then held at 1.9–2.2% for the duration of the experiment. This level of anaesthesia maintained areflexia. The rats breathed spontaneously at a rate of on average 120 breaths/min and their circulatory system was not compromised. Throughout the experiment the core body temperature of the rat was monitored and maintained at 36.5–37 8C by means of a heating blanket connected to a rectal thermal probe via an automatic feed-
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
back control unit. At the end of the experiment, rats were killed with an overdose of halothane. A parylene-coated tungsten electrode was lowered into the cord using a SCAT microdrive (Digitimer, UK), which enabled the depth of the neurone from the surface of the dorsal horn of the spinal cord to be measured. In SNL and sham-operated animals, neurones ipsilateral to the surgery were studied. Extracellular recordings were made from single convergent dorsal horn neurones responding to both innocuous (brush) and noxious (pinch) stimuli applied to the receptive fields in the plantar region of the hind paw. Once neurones were identified, 16 electrical impulses (0.5 Hz, 2 ms width) were applied to the centre of the receptive field via needles inserted into the most sensitive area of the receptive field. The threshold current for C-fibre activation was then measured as the current required to elicit a consistent response of the neurone in the C-fibre latency range (see below). Stimulation was then applied to the receptive field at three times the C-fibre threshold current. Using a C.E.D. interface and Spike 2 software (Cambridge Electronic Design, UK), a post-stimulus time histogram (PSTH) was constructed. This enabled Ab-fibre, Ad-fibre and C fibreevoked responses to be separated and quantified by latency following electrical stimulation (0–20, 20–90 and 90–300 ms, respectively). The remaining neuronal response (300– 800 ms post-stimulus), occurring as the neurones exhibited wind-up leading to a hyper-excitability, was denoted as the post-discharge of the neurone. Additionally, the initial Cfibre evoked response and wind-up were recorded and quantified. Initial C-fibre response was taken as the number of action potentials evoked between 90 and 800 ms in response to the first electrical impulse, prior to any wind-up of the neurone. Wind-up was calculated as the difference between the total number of action potentials (90–800 ms) produced by the train of 16 electrical stimuli and the non-potentiated response produced by the train of 16 electrical stimuli (initial C-fibre response multiplied by 16). In addition, the responses of the neurones to natural, mechanical and thermal stimulation were investigated. For mechanical stimulation, two different von Frey hairs (Semmes–Weinstein monofilaments) with bending forces of 8.51 and 28.84 g were applied to the most sensitive zone of the receptive field of the neurone. von Frey hair stimulation of 28.84 g will be termed ‘noxious mechanical stimulus’ as in behavioural studies we find that normal rats withdraw at 23 g von Frey. Each hair (8.51 g hair followed by the 28.84 g hair) was applied for 15 s and the number of action potentials generated was recorded. The spontaneous activity of the neurone was taken into account by recording the number of action potentials over a 15-s period prior to the stimulation by the 8.51 g von Frey hair. This ‘background measurement’ was then subtracted from the response to each of the von Frey hairs. Following mechanical stimulation, a constant jet of 32 8C water was applied to the receptive field for 15 s using a hypodermic needle and large syringe, followed by a constant jet of 45 8C water for
251
15 s. The number of action potentials evoked by the 32 8C stimulus was then subtracted from the number of action potentials evoked by the 45 8C stimulus. This resulted in the response of the neurone to noxious heat alone, having accounted for the mechanical component of the water jet with the 32 8C stimulus. Control tests of the neurone, i.e. electrical stimulation then followed by natural stimulation as described above, were made at 10-min intervals until the responses did not differ by more than 10% which then comprised the pre-drug control response. Porcine galanin (0.5, 5 and 50 mg dissolved in 50 ml saline, Bachem, UK) was then administered spinally, in a cumulative manner, onto the exposed spinal cord (akin to the intrathecal route). Tests were performed every 10 min for a time course of 90 min, following each dose of galanin. This resulted in a 270-min total time course of galanin exposure to the spinal cord. Effects of galanin were expressed as percentages of this pre-drug control value for every evoked response, allowing each neurone to serve as its own control. Neuronal responses to electrical stimulation are expressed as the total number of action potentials evoked by the 16 electrical stimuli. For the natural stimulation, this is the total number of action potentials evoked by the 15-s stimulus application. 2.3. Statistical analysis Analysis of variance (ANOVA) was used to compare the control neuronal responses prior to galanin administration for the three rat groups. For all the neuronal responses, paired t-tests were used to test the differences between the maximum effects of galanin and control neuronal responses prior to galanin administration. Unpaired t-tests were used to test for significant differences between animal groups for each response at each dose. 3. Results Spinal nerve-ligated (SNL) rats displayed marked behavioural sensitivity to innocuous mechanical and cold stimuli compared to sham-operated rats (Fig. 1A–D). This mechanical allodynia was more pronounced as the force of the mechanical stimulus increased. Nine neurones were tested with spinal galanin in both normal and sham-operated (sham) rats. Ten neurones in SNL rats were exposed to spinal galanin. There was no significant difference (ANOVA) in the depth of neurones recorded (Table 1). There was also no significant difference (ANOVA) between the three animal groups in the pre-drug control neuronal responses obtained prior to galanin administration (Table 1). Since the baseline neuronal responses for each stimulus were the same for all rats regardless of surgery undertaken, a valid comparison of the effects of a particular dose of galanin on the neuronal responses between rat groups can be made. Fig. 2A–E illustrates the effects of spinal galanin admin-
252
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
Fig. 1. Behavioural effects of spinal nerve ligation compared to sham operation, n ¼ 11–13. Plots show the mean ^ SEM in the difference score (number of contralateral withdrawals minus number of ipsilateral withdrawals). Von Frey filaments were applied to each paw ten times and acetone five times.
istration on the neuronal responses to C-fibre, Ad-fibre and Ab-fibre stimulation and the initial C-fibre and postdischarge responses. In normal rats, the electrically evoked C-fibre response was slightly facilitated whilst in shamoperated rats this response was significantly inhibited by 50 mg galanin. However, in SNL rats the C-fibre response
was significantly inhibited by all three doses of galanin compared to control readings (P , 0:05) with a maximal 47% inhibition evoked by 50 mg galanin. Statistical comparisons between groups showed significant differences in the neuronal responses of SNL rats and normal rats at all doses (P , 0:05), and between SNL and sham rats at 0.5 and 50
Table 1 Mean depth and pre-drug control responses of neurones a Response
Normal (n ¼ 9)
Sham-operated (n ¼ 9)
Nerve-injured (n ¼ 10)
Mean depth of neurone (mm) C-Fibres (APs) Ad-Fibres (APs) Ab-Fibres (APs) Initial C-fibre (APs) Post discharge (APs) Wind-up (APs) Innocuous mechanical (von Frey 8.51 g) (APs) Noxious mechanical (von Frey 28.84 g) (APs) Heat (APs)
735 ^ 40 281 ^ 46 76 ^ 20 93 ^ 19 19 ^ 3 97 ^ 24 133 ^ 26 130 ^ 20 659 ^ 108 451 ^ 121
693 ^ 42 314 ^ 41 60 ^ 9 99 ^ 8 18 ^ 4 199 ^ 52 292 ^ 84 204 ^ 43 498 ^ 116 415 ^ 68
683 ^ 34 251 ^ 25 56 ^ 6 111 ^ 4 16 ^ 2 183 ^ 46 249 ^ 78 150 ^ 96 306 ^ 94 465 ^ 140
a Values are the mean ^ SEM of the neuronal population. Neuronal responses to electrical stimulation are expressed as the total number of action potentials (APs) evoked by the 16 electrical stimuli. For the natural stimulation, this is the total number of action potentials (APs) evoked by the 15-s stimulus application. No significant differences in any of the pre-drug control responses or depth of the neurones were found between the three animal groups (ANOVA).
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
253
Fig. 2. Effect of spinal galanin on the electrically-evoked neuronal responses, C-fibres (A), Ad-fibres (B), Ab-fibres (C), initial C-fibre response (D) and post discharge (E). Changes in the responses following galanin administration are expressed as a percentage of the control response. (A–E) Mean ^ SEM of the maximum effect which occurred for the neurone population over the 90-min time-course, n ¼ 7–10. *P , 0:05, maximum effect is significantly different from the control response prior to galanin administration for the animal group in question. §P , 0:05, significant difference between the sham neuronal response and the SNL neuronal response.
mg doses (P , 0:05). The effects of the peptide on electrically evoked Ad-fibre neuronal response for normal, sham and SNL rats followed a similar pattern to the C-fibre evoked responses (Fig. 2B). In SNL rats, the Ad-fibre response was significantly inhibited at all three doses of galanin (P , 0:05). In normal rats, the Ad-fibre response was slightly facilitated by the 0.5-mg dose and in sham rats was inhibited by the 50-mg dose, although these effects
were not significant. The electrically evoked Ab-fibre neuronal response (Fig. 2C) was largely unaltered following exogenous spinal galanin administration in normal, sham and SNL rats. Between-group comparisons also showed no significant difference in the effect of galanin on the evoked Ab-fibre response for the three rat groups. The initial C-fibre neuronal response was facilitated by galanin in normal rats (P , 0:05, 0.5 mg galanin) although
254
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
this was not dose-related (Fig. 2D) and again 50 mg galanin significantly inhibited this response in sham animals (P , 0:05). In SNL rats there was a pronounced doserelated inhibition of the initial C-fibre response by galanin with a maximal 66% inhibition following 50 mg galanin (P , 0:05, 5 and 50 mg galanin). Statistical analysis between animal groups showed significant differences in the effects of galanin on the initial C-fibre responses of SNL rats (P , 0:01) and normal rats at all galanin doses, also between the responses of SNL and sham rats following 50 mg galanin (P , 0:05). Fig. 2E illustrates the effect of exogenous galanin on the electrically evoked post-discharge response, a measure of neuronal hyperexcitability. In normal animals, there was a tendency towards facilitation following galanin administration although this was not significant. In contrast, approximately 50 and 80% inhibition (P , 0:05) of the postdischarge response was seen in both sham and SNL rats following 5 and 50 mg doses of galanin, respectively. There was a significant difference, P , 0:05, in postdischarge responses between normal and SNL rats at 50 mg galanin. Fig. 3 illustrates the effects of exogenous galanin (50 mg) on wind-up, with a comparison of the wind-up graphs of neurones from a normal, sham and SNL rat before and after galanin administration. In SNL animals, galanin inhibited wind-up and in the sham group wind-up was also significantly inhibited following these higher doses of galanin. In normal rats, no significant effects were observed on wind-up responses following galanin administration. Fig. 4A–C depicts the effects of galanin on the neuronal responses to natural stimulation. Spinal galanin resulted in a twofold facilitation of the neuronal response to innocuous mechanical stimulation in normal rats (P , 0:05, Fig. 4A). In marked contrast, this response was significantly inhibited following 5 and 50 mg galanin doses in sham (60 and 76% inhibition, respectively, P , 0:05) and to a slightly greater extent (73 and 91% inhibition, respectively, P , 0:05) in SNL rats. Comparing across the groups, the effect of galanin on the response to low von Frey forces were significantly different (P , 0:01) in SNL rats compared to normal rats at all three doses. Galanin, 0.5 and 50 mg, had a significantly greater effect on this response in SNL rats compared to sham rats at (P , 0:05). Finally, there was a significant difference (P , 0:01) in the effect of 5 and 50 mg doses of galanin between sham and normal rats. A similar profile was seen for the effect of galanin on the noxious mechanical stimulus (Fig. 4B). In normal rats, galanin tended to produce a facilitation although this was not significant. In sham rats, the noxious mechanical response was inhibited by 64% following 50 mg galanin (P , 0:05), whilst in SNL rats 65 and 80% inhibition was observed following 5 and 50 mg galanin, respectively (P , 0:05). Comparing the three animal groups, there is a significant difference in drug effects between SNL rats and normal rats for all three doses (P , 0:01). This is also evident in sham rats compared to normal rats (P , 0:05). In addition, the
Fig. 3. Examples of the effect of exogenous galanin on the wind-up response of a neurone in a (A) normal, (B) sham and (C) SNL rat.
effects of galanin on noxious mechanical simulation in SNL rats are significantly different to these responses in sham rats at 0.5 mg galanin (P , 0:05). Galanin had no significant effect on the neuronal responses to noxious heat in normal and sham rats (Fig. 4C). In contrast, the neuronal responses to noxious heat in SNL rats were significantly inhibited by 80% at all three galanin doses (P , 0:05); these effects are also significantly different to those in normal rats to noxious heat following galanin administration (P , 0:05).
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
255
Fig. 4. Effect of spinal galanin on neuronal responses to innocuous mechanical (A), noxious mechanical (B), and noxious heat (C) stimulation. Changes in the responses following galanin administration are expressed as a percentage of the control response. (A–C) Mean ^ SEM of the maximum effect which occurred for the neurone population over the 90-min time-course, n ¼ 6–9. *P , 0:05, maximum effect is significantly different from the control response prior to galanin administration for the animal group in question. §P , 0:05, significant difference between the sham neuronal response and the SNL neuronal response.
4. Discussion In the present study we have shown that galanin has markedly contrasting effects on spinal nociceptive neurones in normal and neuropathic rats. In normal animals, exogenous galanin produced mild facilitatory effects on dorsal horn neuronal responses to electrical and natural stimulation of peripheral receptive fields to varying extents, with the exception of the unaffected electrically evoked Ab-fibre response. This facilitation was most pronounced for the neuronal responses to innocuous and noxious mechanical stimuli, the electrically evoked post-discharge response and initial C-fibre response. Reeve et al. (2000) have reported similar findings in vivo with exogenous galanin producing facilitation of the initial C-fibre and postdischarge responses. With regard to the neuronal responses evoked by natural stimuli in normal rats, we found a pronounced facilitation of the neuronal responses to innocuous mechanical stimuli, and also to a lesser degree of the neuronal responses to higher intensity, presumed noxious, mechanical stimuli, yet no significant effect on the neuronal
responses to noxious heat. Similar effects have been demonstrated behaviourally (Kuraishi et al., 1991b) where intrathecal galanin administration produced mechanical hyperalgesia but not thermal hyperalgesia in normal rats. More recently, it was reported that chronic intrathecal galanin administration to normal rats caused a significant decrease in mechanical thresholds but had no effect on thermal withdrawal latencies (Kerr et al., 2000). The facilitation seen here could conceivably result from activation of receptors located pre- and post-synaptically. Galanin has been shown to depolarize cultured DRG neurones (Puttick et al., 1994) which could be a pre-synaptic basis for the facilitation of spinal neuronal responses following galanin. Whilst the receptor type responsible is unknown, GalR2 receptors have been shown to have excitatory effects in isolated cell lines resulting in an increase in intracellular calcium (Smith et al., 1997). GalR2 receptors are found both on DRG neurones and in the superficial layers of the dorsal horn and are less abundant than the inhibitory GalR1 receptors in the rat dorsal horn (O’Donnell et al., 1999). Another possibility is the activation of inhibi-
256
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
tory galanin receptors, presumably GalR1, expressed on inhibitory neurones, so that exogenous galanin produces the observed facilitations through disinhibition. Remarkably, in SNL rats exogenous galanin produced a pronounced inhibition of all the neuronal responses with the only exception being the electrically evoked Ab-fibre responses. Electrically evoked C-fibre, Ad-fibre, initial Cfibre and post-discharge responses were inhibited in a doserelated manner, as were the neuronal responses to both innocuous and noxious mechanical stimuli. The wind-up response and the responses of the neurones to noxious heat stimulation were inhibited to the same extent with each dose of galanin. Whilst this is the first study to shown the effect of galanin on evoked dorsal horn neuronal responses in a model of neuropathic pain, they are in agreement with a recent study, where galanin inhibited the spontaneous discharge of dorsal horn neurones in the CCI model of neuropathic pain (Xu et al., 2000b). An earlier study, (Wiesenfeld-Hallin et al., 1989b) found that intrathecal galanin elicited a greater inhibitory effect on the rat flexor reflex in axotomized rats compared to normal rats. Their findings agree with ours, in that the inhibitory effects of galanin seen following nerve injury occur at lower doses. Behavioural studies have shown that intrathecal galanin increases the innocuous mechanical threshold in the photochemical model of neuropathy (Hao et al., 1999), which could relate to the notable inhibition of the neuronal responses to the innocuous mechanical stimuli seen here in the SNL group. The effects of spinal galanin were markedly altered by nerve injury. Following SNL, the neuronal responses were no different from those seen in the sham and normal animals. Since SNL produces an important reduction in afferent inputs into the lumbar segments, the lack of change of neuronal responses, previously reported and discussed (Suzuki et al., 2001), suggests marked compensations in sensory systems following nerve ligation. The mechanisms underlying the marked shift to an inhibitory action of galanin following spinal nerve ligation are unknown. It has been reported that following axotomy both GalR1 mRNA and GalR2 mRNA are down-regulated in DRG neurones (Xu et al., 1996; Shi et al., 1997), although it is yet to be established whether this occurs also at the spinal cord level. These changes could result in a loss of both putative inhibitory and excitatory galanin receptors in the spinal cord (Wang et al., 1998). Such a selective plasticity in terms of GalR1 and GalR2 receptor activation is a possible basis for this inhibitory effect, as are differential changes in the complex circuitry of spinal pathways controlled by galanin receptors. It has been shown that all galanin-containing neurones in the superficial dorsal horn are GABAergic (Simmons et al., 1995). Since nerve injury, in this case sciatic nerve ligation, produces complex changes in the roles of GABA and its receptors after nerve injury (Kontinen et al., 2001), it is quite conceivable that the marked shift to inhibitions results from changes in these inhibitory neuronal populations. Another explanation
for the pronounced inhibition of neuronal responses in SNL rats is a potential up-regulation of inhibitory GalR3 receptors in the spinal cord following peripheral nerve injury. GalR3 receptors are present in low levels in the spinal cord (Smith et al., 1998), but the effect of peripheral nerve injury on this receptor population is currently unknown. Interestingly, neuropathic pain-related behaviour involves genes that map close to the GalR3 receptor (Seltzer et al., 2001). Whatever the case, exogenous galanin administration profoundly inhibits neuronal responses after nerve injury. In general, the inhibitory effects of galanin on evoked neuronal responses in sham-operated rats were only seen with the highest dose. In the case of the post-discharge and wind-up and the responses to innocuous and noxious mechanical stimuli, the inhibition of these responses in sham-operated animals approached those seen in SNL rats. However, the neuronal responses to noxious heat were not significantly inhibited. The sham operation is an essential control experiment for the SNL procedure, and may also represent a state of postsurgical tissue damage with an inflammatory component. It is unlikely that any post-operative inflammation is present 2 weeks following the surgery and the sham-operated animals did not show any behavioural mechanical hyperalgesia (Fig. 1A–C). However, it could be postulated that enduring subthreshold inflammatory processes could be occurring following the surgery, or the initial inflammatory insult could have induced long-lasting changes in galanin receptor expression in the spinal cord. Kuraishi et al. (1991b) found mechanical, but not thermal hyperalgesia induced by carrageenan inflammation was reversed by intrathecal administration of anti-galanin antiserum. This finding suggests that galanin is responsible for the generation of certain hyperalgesias seen after inflammation. We have previously reported that gabapentin, ineffective in normal animals, also inhibits neuronal responses in the sham-operated group (Chapman et al., 1998a). Inflammation has been shown to cause a decrease in GalR1 mRNA (Xu et al., 1996) and an increase in GalR2 mRNA (Shi et al., 1997) in DRG neurones. If these changes have functional importance, a facilitation of the neuronal responses in sham rats could possibly result from GalR2 receptor activation. However, we find other neuronal responses (von Frey 9 g and post discharge) in sham rats to be inhibited to the same extent as in SNL rats. A possible explanation for these modality-selective effects is that the inhibition seen in sham rats arises as a result of up-regulation of the putative inhibitory GalR3, present in low levels in the spinal cord (Smith et al., 1998). There may be regional distributions of the galanin receptors on different neuronal structures that allow these selective effects of galanin.
5. Conclusion We find exogenous spinal galanin administration causes a
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
facilitation of the neuronal responses to electrical and natural stimulation in normal rats. In sham rats, galanin at high doses causes an inhibition of their neuronal responses. Exogenous galanin elicits a marked inhibition of all neuronal responses, electrically evoked, mechanical and thermal responses in neuropathic rats. Whatever the mechanisms underlying these changes in the effects of galanin after peripheral nerve ligation, these results suggest that galanin or stable analogues may be a useful target for the treatment of nerve injury pain.
Acknowledgements S.J.L.F. is funded by MRC Industrial Collaborative Studentship.
References ¨ . Galanin – a neuroendocrine peptide. Crit Bartfai T, Ho¨ kfelt T, Langel U Rev Neurobiol 1993;7:229–274. Branchek TA, Smith KE, Gerald C, Walker MW. Galanin receptor subtypes. Trends Pharmacol Sci 2000;21:109–117. Chapman V, Suzuki R, Chamarette HL, Rygh LJ, Dickenson AH. Effects of systemic carbamazepine and gabapentin on spinal neuronal responses in spinal nerve ligated rats. Pain 1998a;75:261–272. Chapman V, Suzuki R, Dickenson AH. Electrophysiological characterization of spinal neuronal response properties in anaesthetized rats after ligation of spinal nerves L5–L6. J Physiol 1998b;507:881–894. Crawley JN. Galanin–acetylcholine interactions: relevance to memory and Alzheimer’s disease. Life Sci 1996;58:2185–2199. Habert-Ortholi E, Amiranoff B, Loquet I, Laburthe M, Mayaux JF. Molecular cloning of a functional human galanin receptor. Proc Natl Acad Sci USA 1994;91:9780–9783. Hao JX, Shi TJ, Xu IS, Kaupilla T, Xu XJ, Hokfelt T, Bartfai T, Wiesenfeld-Hallin Z. Intrathecal galanin alleviates allodynia-like behaviour in rats after partial peripheral nerve injury. Eur J Neurosci 1999;11:427– 432. Ho¨ kfelt T, Wisenfeld-Hallin Z, Villar M, Melander T. Increases of galaninlike immunoreactivity in rat dorsal root ganglion cells after peripheral axotomy. Neurosci Lett 1987;83:217–220. Ji RR, Zhang X, Zhang Q, Dagerlind A, Nilsson S, Wiesenfeld_Hallin Z, Hokfelt T. Central and peripheral expression of galanin in response to inflammation. Neuroscience 1995;68:563–576. Kerr BJ, Cafferty WB, Gupta YK, Bacon A, Wynick D, McMahon SB, Thompson SW. Galanin knockout mice reveal nociceptive deficits following peripheral nerve injury. Eur J Neurosci 2000;12:793–802. Kim SH, Chung JM. An experimental model for peripheral neuropathy produced by segmental spinal nerve ligation in the rat. Pain 1992;50:355–363. Kontinen VK, Stanfa LC, Suzuki R, Dickenson AH. Electrophysiological evidence for increased endogenous GABAergic, but not glycinergic inhibitory tone in the rat spinal nerve ligation model of neuropathy. Anesthesiology 2001;94:329–333. Kuraishi Y, Kawabata S, Matsumoto T, Nakamura A, Fujita H, Satoh M. Involvement of substance P in hyperalgesia induced by intrathecal galanin. Neurosci Res 1991a;11:276–285. Kuraishi Y, Kawamura M, Yamaguchi T, Houtani T, Kawabata S, Futaki S, Fujii N, Satoh M. Intrathecal injections of galanin and its antiserum affect nociceptive response of rat to mechanical, but not thermal, stimuli. Pain 1991b;44:321–324. Ma W, Bisby MA. Differential expression of galanin immunoreactivities in
257
the primary sensory neurons following partial and complete sciatic nerve injuries. Neuroscience 1997;79:1183–1195. Merchenthaler I, Lopez FJ, Negro-Vilar A. Anatomy and physiology of central galanin-containing pathways. Prog Neurobiol 1993;40:1183– 1195. Nussbaumer JC, Yanagisawa M, Otsuka M. Pharmacological properties of a C-fibre response evoked by saphenous nerve stimulation in an isolated spinal cord-nerve preparation of the newborn rat. Br J Pharmacol 1989;98:373–382. O’Donnell D, Ahmad S, Wahlestedt C, Walker P. Expression of the novel galanin receptor subtype GALR2 in the adult rat CNS: distinct distribution from GALR1. J Comp Neurol 1999;409:469–481. ¨ gren SO, Ho¨ kfelt T, Kask K, Langel U ¨ , Bartfai T. Evidence for a role of O galanin on spatial learning. Neuroscience 1992;51:1–5. Parker EM, Izzarelli DG, Nowak HP, Mahle CD, Iben LG, Wang J, Goldstein ME. Cloning and characterization of the rat GALR1 galanin receptor from Rin14B insulinoma cells. Brain Res Mol Brain Res 1995;34:179–189. Post C, Alari L, Hokfelt T. Intrathecal galanin increases the latency in the tail-flick and hot-plate test in mouse. Acta Physiol Scand 1988;132:583–584. Puttick RM, Pinnock RD, Woodruff GN. Galanin-induced membrane depolarization of neonatal rat cultured dorsal root ganglion cells. Eur J Pharmacol 1994;254:303–306. Reeve AJ, Walker K, Urban L, Fox AJ. Excitatory effects of galanin in the spinal cord of intact, anaesthetized rats. Neurosci Lett 2000;295:25–28. Satoh M, Kuraishi Y, Kawamura M. Effects of intrathecal antibodies to substance P, calcitonin gene-related peptide and galanin on repeated cold stress-induced hyperalgesia: comparison with carrageenaninduced hyperalgesia. Pain 1992;49:273–278. Seltzer Z, Wu T, Max BM, Diehl SR. Mapping a gene for neuropathic painrelated behavior following peripheral neurectomy in the mouse. Pain 2001;93:101–106. Shi T-J, Zhang X, Holmberg K, Xu Z-Q, Ho¨ kfelt T. Expression and regulation of galanin-R2 receptors in rat primary sensory neurons: effect of axotomy and inflammation. Neurosci Lett 1997;237:57–60. Simmons DR, Spike RC, Todd AJ. Galanin is contained in GABAergic neurons in the rat spinal dorsal horn. Neurosci Lett 1995;187:119–122. Smith KE, Forray C, Walker MW, Jones KA, Tamm JA, Bard J, Branchek TA, Linemeyer DL, Gerald C. Expression cloning of a rat hypothalamic galanin receptor coupled to phosphoinositide turnover. J Biol Chem 1997;272:24612–24616. Smith KE, Walker MW, Artymyshyn R, Bard J, Borowsky B, Tamm JA, Yao WJ, Vaysse PJ, Branchek TA, Gerald C, Jones KA. Cloned human and rat galanin GALR3 receptors. Pharmacology and activation of Gprotein inwardly rectifying K 1 channels. J Biol Chem 1998;273:23321– 23326. Suzuki R, Matthews EA, Dickenson AH. Comparison of the effects of MK801, ketamine and memantine on responses of spinal dorsal horn neurones in a rat model of mononeuropathy. Pain 2001;91:101–109. ˚ , Jo¨ rnall H, McDonald TJ, Mutt V. Galanin – a Tatemoto K, Ro¨ kaeus A novel biologically active peptide from porcine intestine. FEBS Lett 1983;164:124–128. Wang S, He C, Hashemi T, Bayne M. Cloning and expressional characterization of a novel galanin receptor. Identification of different pharmacophores within galanin for the three galanin receptor subtypes. J Biol Chem 1997;272:31949–31952. Wang S, Hashemi T, Fried S, Clemmons AL, Hawes BE. Differential intracellular signaling of the GalR1 and GalR2 galanin receptor subtypes. Biochemistry 1998;37:6711–6717. Waters SM, Krause JE. Distribution of galanin-1, -2 and -3 receptor messenger RNAs in central and peripheral rat tissues. Neuroscience 2000;95:265–271. Wiesenfeld-Hallin Z, Villar MJ, Hokfelt T. The effects of intrathecal galanin and C-fiber stimulation on the flexor reflex in the rat. Brain Res 1989a;486:205–213. Wiesenfeld-Hallin Z, Xu XJ, Villar MJ, Hokfelt T. The effect of intrathecal
258
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
galanin on the flexor reflex in rat: increased depression after sciatic nerve section. Neurosci Lett 1989b;105:149–154. Xu Z-Q, Shi T-J, Landry M, Ho¨ kfelt T. Evidence for galanin receptors in primary sensory neurones and effect of axotomy and inflammation. Neuroreport 1996;8:237–242. Xu XJ, Hokfelt T, Bartfai T, Wiesenfeld-Hallin Z. Galanin and aspinal nociceptive mechanisms: recent advances and therapeutic implications. Neuropeptides 2000a;34:137–147.
Xu S-L, Zhang Y-P, Lundeberg T, Yu L-C. Effects of galanin on widedynamic range neuron activity in the spinal dorsal horn of rats with sciatic nerve ligation. Regul Pept 2000b;95:19–23. Yanagisawa M, Yagi N, Otsuka M, Yanaihara C, Yanaihara N. Inhibitory effects of galanin on the isolated spinal cord of the newborn rat. Neurosci Lett 1986;70:278–282. Zimmermann M. Ethical guidelines for investigations of experimental pain in conscious animals [editorial]. Pain 1983;16:109–110.
Nerve injury induces plasticity that results in spinal inhibitory effects of galanin Sarah J.L. Flatters a,*, Alyson J. Fox b, Anthony H. Dickenson a a
Department of Pharmacology, University College London, Gower Street, London WC1E 6BT, UK Novartis Institute for Medical Sciences, University College London, Gower Place, London WC1E 6BS, UK
b
Accepted 27 November 2001
Abstract Galanin is a 29-amino-acid neuropeptide that has been implicated in the processes of nociception. This study examines the effect of exogenous galanin on dorsal horn neurone activity in vivo in the spinal nerve ligation (SNL) model of neuropathic pain. SNL rats but not naive or sham-operated rats exhibited behaviour indicative of allodynia. In anaesthetized rats, extracellular recordings were made from individual convergent dorsal horn neurones following stimulation of peripheral receptive fields electrically or with natural (innocuous mechanical, noxious mechanical and noxious thermal) stimuli. Spinal administration of galanin (0.5–50 mg) caused a slight facilitation of the neuronal responses to natural and electrical stimuli in naive rats and up to a 65% inhibition of neuronal responses in sham-operated rats following 50 mg galanin. In contrast, there was a marked inhibition of up to 80% of responses to both natural and electrical stimuli in SNL rats following spinal galanin administration. These results suggest that following peripheral nerve injury, there is plasticity in the levels of galanin and/or its receptors at spinal cord level so that the effect of exogenous galanin favours inhibitory function. q 2002 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved. Keywords: Galanin; Nociception; Spinal cord; Neuropathic pain; Rat; Electrophysiology
1. Introduction Galanin is a neuropeptide consisting of 29 amino acids (30 in humans) originally isolated from porcine gut (Tatemoto et al., 1983). It is widely distributed in the central and peripheral nervous systems (Merchenthaler et al., 1993) and has been shown to play a role in several key functions such ¨ gren et al., 1992), memory (Crawley, 1996), as learning (O feeding, and endocrine modulation (for review see Bartfai et al., 1993). In recent years three galanin receptors subtypes (for review see Branchek et al., 2000) have been cloned and termed GalR1 (Habert-Ortholi et al., 1994; Parker et al., 1995), GalR2 (Smith et al., 1997) and GalR3 (Wang et al., 1997; Smith et al., 1998). Activation of GalR1 and GalR3 receptors expressed in cell lines leads to inhibition of cAMP levels via coupling to Gi-protein (Smith et al., 1998; Wang et al., 1998). In contrast, GalR2 has been shown to be excitatory, activating phospholipase C and increasing phosphoinositol and intracellular calcium levels via coupling to Go/Gq-proteins (Wang et al., 1998). The * Corresponding author. Tel.: 144-207-679-3737; fax: 144-207-6793742. E-mail address: [email protected] (S.J.L. Flatters).
expression of the three receptor types has been examined at the mRNA level in rats. Whilst GalR1 is restricted predominantly to the central nervous system. GalR2 is more widespread with expression in the CNS as well as several peripheral tissues. GalR3 appears to be expressed in the periphery with CNS expression being restricted to certain areas (Waters and Krause, 2000). Of relevance to sensory processing, GalR1 and GalR2 have been shown to be present in the adult rat spinal cord and dorsal root ganglion (DRG) (Shi et al., 1997; O’Donnell et al., 1999). GalR1 expression is limited to the superficial dorsal horn and is mainly, but not exclusively, found on large DRG neurones, whilst GalR2 is expressed throughout the dorsal horn and is present on small and medium DRG neurones (O’Donnell et al., 1999). In contrast, GalR3 is expressed at low levels in the spinal cord and in DRG (Waters and Krause, 2000). Much interest has arisen in the role of galanin in nociception, recently reviewed by Xu et al. (2000a). In functional studies, galanin has been shown to have inhibitory effects on ventral root potentials in isolated spinal cord preparations of the newborn rat (Yanagisawa et al., 1986; Nussbaumer et al., 1989) and inhibit tail-flick and hot-plate responses in normal mice following intrathecal administration (Post et al., 1988). Other electrophysiological studies in normal
0304-3959/02/$20.00 q 2002 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved. PII: S 0304-395 9(02)00180-X
250
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
rats report galanin eliciting facilitatory effects, with depolarization of DRG neuronal membranes (Puttick et al., 1994) and facilitation of electrically-evoked responses of spinal neurones following spinal galanin administration (Reeve et al., 2000). In addition, it has been demonstrated that intrathecal galanin produces mechanical hyperalgesia in normal rats (Kuraishi et al., 1991b). Biphasic effects of galanin have also been reported (Wiesenfeld-Hallin et al., 1989a), with low doses causing facilitatory effects and higher doses resulting in inhibition. Nerve injury and inflammation affects the expression of both galanin and its receptors. Following axotomy, partial nerve transection (PSNL) and chronic constriction injury (CCI) in the rat there is a marked up-regulation of galanin levels in the dorsal root ganglion (DRG), superficial dorsal horn and gracile nucleus (Ho¨ kfelt et al., 1987; Ma and Bisby, 1997). Carrageenan-induced inflammation results in an increase in galanin mRNA levels in the superficial dorsal horn, but a reduction in the DRG (Ji et al., 1995). In contrast to the effects on galanin expression, following axotomy GalR1 and GalR2 mRNA levels are decreased in the DRG (Xu et al., 1996; Shi et al., 1997). Inflammation also results in a decrease in GalR1 mRNA levels in adult DRG neurones (Xu et al., 1996), although GalR2 mRNA is up-regulated (Shi et al., 1997). The effect of nerve injury and inflammation on GalR3 expression has not been reported. Previous behavioural studies have also investigated the role of galanin in inflammatory and neuropathic pain models. Anti-galanin antibody has been reported to reverse carrageenan-induced hyperalgesia (Kuraishi et al., 1991a; Satoh et al., 1992) implying pronociceptive effects of galanin whereas intrathecal galanin administration has been shown to alleviate mechanical and cold allodynia in photochemically nerve-injured rats (Hao et al., 1999). Although there is a marked up-regulation of galanin following nerve injury, it remains to be shown whether this up-regulation has functional consequences, particularly in the light of the down-regulation of some galanin receptor populations in the DRG after axotomy. Thus the aim of this study is to investigate the effects of exogenous galanin in vivo on the responses of convergent dorsal horn neurones in normal, sham-operated and neuropathic (spinal nerve ligated; SNL) rats to electrical, mechanical and thermal stimulation. This model of nerve injury uses a partial ligation of the nerve and is sensitive to drugs used in the treatment of neuropathic pain in humans (Chapman et al., 1998a).
2. Materials and methods 2.1. Animal preparation Male Sprague–Dawley rats were housed in groups of five in plastic cages with artificial lighting which had a fixed 12:12-h light/dark cycle. Food and water were available
ad libitum. Guidelines for animal research by the UK Home Office and the International Association for the Study of Pain (Zimmermann, 1983) were adhered to. The spinal nerve ligation (SNL) model of neuropathic pain first described by Kim and Chung (1992) was used. Briefly, rats (130–150 g) were anaesthetized with halothane (Fluothane, Zeneca, UK) in a 50% O2/50% N2O gaseous mixture. The L5 and L6 spinal nerves were exposed and tightly ligated using 6-0 silk and the L4 sciatic nerve branch was left uninjured. The muscle and skin were then closed with sutures and wound clips. The surgery for the sham-operated rats was identical to the SNL rats with the exception of the ligation of L5 and L6 spinal nerves. The animals were allowed to recover and were monitored for 2 weeks post-surgery. Behavioural testing was performed on days 2, 4, 7, 9, 11 and 14 following surgery to assess the development of mechanical and cold allodynia. Rats were placed in individual boxes with a metal mesh floor and allowed to acclimatize. Von Frey hairs with bending forces of 1, 5 and 9 g (Semmes–Weinstein monofilaments) were applied to the plantar surface of both hindpaws in ascending order of force. Each hair was applied ten times with 5-s intervals and the number of withdrawal responses recorded. Cold allodynia was assessed by the response to application of acetone to the plantar surface of both hindpaws. Acetone was applied a total of five times, with 3–4 min left between applications to allow evaporation and so elicit a cooling effect. A withdrawal response was observed as a shaking, flicking or licking of the paw following acetone application. The outcome of these behavioural tests is expressed as a difference score (number of contralateral withdrawals minus number of ipsilateral withdrawals). 2.2. In vivo electrophysiology Experiments were performed 14–18 days after the surgery and on uninjured (normal) animals of a similar weight using methods previously described by this laboratory (Chapman et al., 1998b). Rats were initially anaesthetized with 3.5% halothane in a 33% O2/66% N2O gaseous mixture until areflexia was produced. Cannulation of the trachea was performed to ensure adequate and efficient respiration of the rat. Rats were secured in a stereotaxic frame and a laminectomy was performed to expose segments L4–L5 of the spinal cord. The dura of the exposed spinal cord was then removed and the spine was held rigid by clamps caudal and rostral to the laminectomy. The halothane concentration was lowered (2.5–2.8%) whilst the surgery was performed, and then held at 1.9–2.2% for the duration of the experiment. This level of anaesthesia maintained areflexia. The rats breathed spontaneously at a rate of on average 120 breaths/min and their circulatory system was not compromised. Throughout the experiment the core body temperature of the rat was monitored and maintained at 36.5–37 8C by means of a heating blanket connected to a rectal thermal probe via an automatic feed-
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
back control unit. At the end of the experiment, rats were killed with an overdose of halothane. A parylene-coated tungsten electrode was lowered into the cord using a SCAT microdrive (Digitimer, UK), which enabled the depth of the neurone from the surface of the dorsal horn of the spinal cord to be measured. In SNL and sham-operated animals, neurones ipsilateral to the surgery were studied. Extracellular recordings were made from single convergent dorsal horn neurones responding to both innocuous (brush) and noxious (pinch) stimuli applied to the receptive fields in the plantar region of the hind paw. Once neurones were identified, 16 electrical impulses (0.5 Hz, 2 ms width) were applied to the centre of the receptive field via needles inserted into the most sensitive area of the receptive field. The threshold current for C-fibre activation was then measured as the current required to elicit a consistent response of the neurone in the C-fibre latency range (see below). Stimulation was then applied to the receptive field at three times the C-fibre threshold current. Using a C.E.D. interface and Spike 2 software (Cambridge Electronic Design, UK), a post-stimulus time histogram (PSTH) was constructed. This enabled Ab-fibre, Ad-fibre and C fibreevoked responses to be separated and quantified by latency following electrical stimulation (0–20, 20–90 and 90–300 ms, respectively). The remaining neuronal response (300– 800 ms post-stimulus), occurring as the neurones exhibited wind-up leading to a hyper-excitability, was denoted as the post-discharge of the neurone. Additionally, the initial Cfibre evoked response and wind-up were recorded and quantified. Initial C-fibre response was taken as the number of action potentials evoked between 90 and 800 ms in response to the first electrical impulse, prior to any wind-up of the neurone. Wind-up was calculated as the difference between the total number of action potentials (90–800 ms) produced by the train of 16 electrical stimuli and the non-potentiated response produced by the train of 16 electrical stimuli (initial C-fibre response multiplied by 16). In addition, the responses of the neurones to natural, mechanical and thermal stimulation were investigated. For mechanical stimulation, two different von Frey hairs (Semmes–Weinstein monofilaments) with bending forces of 8.51 and 28.84 g were applied to the most sensitive zone of the receptive field of the neurone. von Frey hair stimulation of 28.84 g will be termed ‘noxious mechanical stimulus’ as in behavioural studies we find that normal rats withdraw at 23 g von Frey. Each hair (8.51 g hair followed by the 28.84 g hair) was applied for 15 s and the number of action potentials generated was recorded. The spontaneous activity of the neurone was taken into account by recording the number of action potentials over a 15-s period prior to the stimulation by the 8.51 g von Frey hair. This ‘background measurement’ was then subtracted from the response to each of the von Frey hairs. Following mechanical stimulation, a constant jet of 32 8C water was applied to the receptive field for 15 s using a hypodermic needle and large syringe, followed by a constant jet of 45 8C water for
251
15 s. The number of action potentials evoked by the 32 8C stimulus was then subtracted from the number of action potentials evoked by the 45 8C stimulus. This resulted in the response of the neurone to noxious heat alone, having accounted for the mechanical component of the water jet with the 32 8C stimulus. Control tests of the neurone, i.e. electrical stimulation then followed by natural stimulation as described above, were made at 10-min intervals until the responses did not differ by more than 10% which then comprised the pre-drug control response. Porcine galanin (0.5, 5 and 50 mg dissolved in 50 ml saline, Bachem, UK) was then administered spinally, in a cumulative manner, onto the exposed spinal cord (akin to the intrathecal route). Tests were performed every 10 min for a time course of 90 min, following each dose of galanin. This resulted in a 270-min total time course of galanin exposure to the spinal cord. Effects of galanin were expressed as percentages of this pre-drug control value for every evoked response, allowing each neurone to serve as its own control. Neuronal responses to electrical stimulation are expressed as the total number of action potentials evoked by the 16 electrical stimuli. For the natural stimulation, this is the total number of action potentials evoked by the 15-s stimulus application. 2.3. Statistical analysis Analysis of variance (ANOVA) was used to compare the control neuronal responses prior to galanin administration for the three rat groups. For all the neuronal responses, paired t-tests were used to test the differences between the maximum effects of galanin and control neuronal responses prior to galanin administration. Unpaired t-tests were used to test for significant differences between animal groups for each response at each dose. 3. Results Spinal nerve-ligated (SNL) rats displayed marked behavioural sensitivity to innocuous mechanical and cold stimuli compared to sham-operated rats (Fig. 1A–D). This mechanical allodynia was more pronounced as the force of the mechanical stimulus increased. Nine neurones were tested with spinal galanin in both normal and sham-operated (sham) rats. Ten neurones in SNL rats were exposed to spinal galanin. There was no significant difference (ANOVA) in the depth of neurones recorded (Table 1). There was also no significant difference (ANOVA) between the three animal groups in the pre-drug control neuronal responses obtained prior to galanin administration (Table 1). Since the baseline neuronal responses for each stimulus were the same for all rats regardless of surgery undertaken, a valid comparison of the effects of a particular dose of galanin on the neuronal responses between rat groups can be made. Fig. 2A–E illustrates the effects of spinal galanin admin-
252
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
Fig. 1. Behavioural effects of spinal nerve ligation compared to sham operation, n ¼ 11–13. Plots show the mean ^ SEM in the difference score (number of contralateral withdrawals minus number of ipsilateral withdrawals). Von Frey filaments were applied to each paw ten times and acetone five times.
istration on the neuronal responses to C-fibre, Ad-fibre and Ab-fibre stimulation and the initial C-fibre and postdischarge responses. In normal rats, the electrically evoked C-fibre response was slightly facilitated whilst in shamoperated rats this response was significantly inhibited by 50 mg galanin. However, in SNL rats the C-fibre response
was significantly inhibited by all three doses of galanin compared to control readings (P , 0:05) with a maximal 47% inhibition evoked by 50 mg galanin. Statistical comparisons between groups showed significant differences in the neuronal responses of SNL rats and normal rats at all doses (P , 0:05), and between SNL and sham rats at 0.5 and 50
Table 1 Mean depth and pre-drug control responses of neurones a Response
Normal (n ¼ 9)
Sham-operated (n ¼ 9)
Nerve-injured (n ¼ 10)
Mean depth of neurone (mm) C-Fibres (APs) Ad-Fibres (APs) Ab-Fibres (APs) Initial C-fibre (APs) Post discharge (APs) Wind-up (APs) Innocuous mechanical (von Frey 8.51 g) (APs) Noxious mechanical (von Frey 28.84 g) (APs) Heat (APs)
735 ^ 40 281 ^ 46 76 ^ 20 93 ^ 19 19 ^ 3 97 ^ 24 133 ^ 26 130 ^ 20 659 ^ 108 451 ^ 121
693 ^ 42 314 ^ 41 60 ^ 9 99 ^ 8 18 ^ 4 199 ^ 52 292 ^ 84 204 ^ 43 498 ^ 116 415 ^ 68
683 ^ 34 251 ^ 25 56 ^ 6 111 ^ 4 16 ^ 2 183 ^ 46 249 ^ 78 150 ^ 96 306 ^ 94 465 ^ 140
a Values are the mean ^ SEM of the neuronal population. Neuronal responses to electrical stimulation are expressed as the total number of action potentials (APs) evoked by the 16 electrical stimuli. For the natural stimulation, this is the total number of action potentials (APs) evoked by the 15-s stimulus application. No significant differences in any of the pre-drug control responses or depth of the neurones were found between the three animal groups (ANOVA).
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
253
Fig. 2. Effect of spinal galanin on the electrically-evoked neuronal responses, C-fibres (A), Ad-fibres (B), Ab-fibres (C), initial C-fibre response (D) and post discharge (E). Changes in the responses following galanin administration are expressed as a percentage of the control response. (A–E) Mean ^ SEM of the maximum effect which occurred for the neurone population over the 90-min time-course, n ¼ 7–10. *P , 0:05, maximum effect is significantly different from the control response prior to galanin administration for the animal group in question. §P , 0:05, significant difference between the sham neuronal response and the SNL neuronal response.
mg doses (P , 0:05). The effects of the peptide on electrically evoked Ad-fibre neuronal response for normal, sham and SNL rats followed a similar pattern to the C-fibre evoked responses (Fig. 2B). In SNL rats, the Ad-fibre response was significantly inhibited at all three doses of galanin (P , 0:05). In normal rats, the Ad-fibre response was slightly facilitated by the 0.5-mg dose and in sham rats was inhibited by the 50-mg dose, although these effects
were not significant. The electrically evoked Ab-fibre neuronal response (Fig. 2C) was largely unaltered following exogenous spinal galanin administration in normal, sham and SNL rats. Between-group comparisons also showed no significant difference in the effect of galanin on the evoked Ab-fibre response for the three rat groups. The initial C-fibre neuronal response was facilitated by galanin in normal rats (P , 0:05, 0.5 mg galanin) although
254
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
this was not dose-related (Fig. 2D) and again 50 mg galanin significantly inhibited this response in sham animals (P , 0:05). In SNL rats there was a pronounced doserelated inhibition of the initial C-fibre response by galanin with a maximal 66% inhibition following 50 mg galanin (P , 0:05, 5 and 50 mg galanin). Statistical analysis between animal groups showed significant differences in the effects of galanin on the initial C-fibre responses of SNL rats (P , 0:01) and normal rats at all galanin doses, also between the responses of SNL and sham rats following 50 mg galanin (P , 0:05). Fig. 2E illustrates the effect of exogenous galanin on the electrically evoked post-discharge response, a measure of neuronal hyperexcitability. In normal animals, there was a tendency towards facilitation following galanin administration although this was not significant. In contrast, approximately 50 and 80% inhibition (P , 0:05) of the postdischarge response was seen in both sham and SNL rats following 5 and 50 mg doses of galanin, respectively. There was a significant difference, P , 0:05, in postdischarge responses between normal and SNL rats at 50 mg galanin. Fig. 3 illustrates the effects of exogenous galanin (50 mg) on wind-up, with a comparison of the wind-up graphs of neurones from a normal, sham and SNL rat before and after galanin administration. In SNL animals, galanin inhibited wind-up and in the sham group wind-up was also significantly inhibited following these higher doses of galanin. In normal rats, no significant effects were observed on wind-up responses following galanin administration. Fig. 4A–C depicts the effects of galanin on the neuronal responses to natural stimulation. Spinal galanin resulted in a twofold facilitation of the neuronal response to innocuous mechanical stimulation in normal rats (P , 0:05, Fig. 4A). In marked contrast, this response was significantly inhibited following 5 and 50 mg galanin doses in sham (60 and 76% inhibition, respectively, P , 0:05) and to a slightly greater extent (73 and 91% inhibition, respectively, P , 0:05) in SNL rats. Comparing across the groups, the effect of galanin on the response to low von Frey forces were significantly different (P , 0:01) in SNL rats compared to normal rats at all three doses. Galanin, 0.5 and 50 mg, had a significantly greater effect on this response in SNL rats compared to sham rats at (P , 0:05). Finally, there was a significant difference (P , 0:01) in the effect of 5 and 50 mg doses of galanin between sham and normal rats. A similar profile was seen for the effect of galanin on the noxious mechanical stimulus (Fig. 4B). In normal rats, galanin tended to produce a facilitation although this was not significant. In sham rats, the noxious mechanical response was inhibited by 64% following 50 mg galanin (P , 0:05), whilst in SNL rats 65 and 80% inhibition was observed following 5 and 50 mg galanin, respectively (P , 0:05). Comparing the three animal groups, there is a significant difference in drug effects between SNL rats and normal rats for all three doses (P , 0:01). This is also evident in sham rats compared to normal rats (P , 0:05). In addition, the
Fig. 3. Examples of the effect of exogenous galanin on the wind-up response of a neurone in a (A) normal, (B) sham and (C) SNL rat.
effects of galanin on noxious mechanical simulation in SNL rats are significantly different to these responses in sham rats at 0.5 mg galanin (P , 0:05). Galanin had no significant effect on the neuronal responses to noxious heat in normal and sham rats (Fig. 4C). In contrast, the neuronal responses to noxious heat in SNL rats were significantly inhibited by 80% at all three galanin doses (P , 0:05); these effects are also significantly different to those in normal rats to noxious heat following galanin administration (P , 0:05).
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
255
Fig. 4. Effect of spinal galanin on neuronal responses to innocuous mechanical (A), noxious mechanical (B), and noxious heat (C) stimulation. Changes in the responses following galanin administration are expressed as a percentage of the control response. (A–C) Mean ^ SEM of the maximum effect which occurred for the neurone population over the 90-min time-course, n ¼ 6–9. *P , 0:05, maximum effect is significantly different from the control response prior to galanin administration for the animal group in question. §P , 0:05, significant difference between the sham neuronal response and the SNL neuronal response.
4. Discussion In the present study we have shown that galanin has markedly contrasting effects on spinal nociceptive neurones in normal and neuropathic rats. In normal animals, exogenous galanin produced mild facilitatory effects on dorsal horn neuronal responses to electrical and natural stimulation of peripheral receptive fields to varying extents, with the exception of the unaffected electrically evoked Ab-fibre response. This facilitation was most pronounced for the neuronal responses to innocuous and noxious mechanical stimuli, the electrically evoked post-discharge response and initial C-fibre response. Reeve et al. (2000) have reported similar findings in vivo with exogenous galanin producing facilitation of the initial C-fibre and postdischarge responses. With regard to the neuronal responses evoked by natural stimuli in normal rats, we found a pronounced facilitation of the neuronal responses to innocuous mechanical stimuli, and also to a lesser degree of the neuronal responses to higher intensity, presumed noxious, mechanical stimuli, yet no significant effect on the neuronal
responses to noxious heat. Similar effects have been demonstrated behaviourally (Kuraishi et al., 1991b) where intrathecal galanin administration produced mechanical hyperalgesia but not thermal hyperalgesia in normal rats. More recently, it was reported that chronic intrathecal galanin administration to normal rats caused a significant decrease in mechanical thresholds but had no effect on thermal withdrawal latencies (Kerr et al., 2000). The facilitation seen here could conceivably result from activation of receptors located pre- and post-synaptically. Galanin has been shown to depolarize cultured DRG neurones (Puttick et al., 1994) which could be a pre-synaptic basis for the facilitation of spinal neuronal responses following galanin. Whilst the receptor type responsible is unknown, GalR2 receptors have been shown to have excitatory effects in isolated cell lines resulting in an increase in intracellular calcium (Smith et al., 1997). GalR2 receptors are found both on DRG neurones and in the superficial layers of the dorsal horn and are less abundant than the inhibitory GalR1 receptors in the rat dorsal horn (O’Donnell et al., 1999). Another possibility is the activation of inhibi-
256
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
tory galanin receptors, presumably GalR1, expressed on inhibitory neurones, so that exogenous galanin produces the observed facilitations through disinhibition. Remarkably, in SNL rats exogenous galanin produced a pronounced inhibition of all the neuronal responses with the only exception being the electrically evoked Ab-fibre responses. Electrically evoked C-fibre, Ad-fibre, initial Cfibre and post-discharge responses were inhibited in a doserelated manner, as were the neuronal responses to both innocuous and noxious mechanical stimuli. The wind-up response and the responses of the neurones to noxious heat stimulation were inhibited to the same extent with each dose of galanin. Whilst this is the first study to shown the effect of galanin on evoked dorsal horn neuronal responses in a model of neuropathic pain, they are in agreement with a recent study, where galanin inhibited the spontaneous discharge of dorsal horn neurones in the CCI model of neuropathic pain (Xu et al., 2000b). An earlier study, (Wiesenfeld-Hallin et al., 1989b) found that intrathecal galanin elicited a greater inhibitory effect on the rat flexor reflex in axotomized rats compared to normal rats. Their findings agree with ours, in that the inhibitory effects of galanin seen following nerve injury occur at lower doses. Behavioural studies have shown that intrathecal galanin increases the innocuous mechanical threshold in the photochemical model of neuropathy (Hao et al., 1999), which could relate to the notable inhibition of the neuronal responses to the innocuous mechanical stimuli seen here in the SNL group. The effects of spinal galanin were markedly altered by nerve injury. Following SNL, the neuronal responses were no different from those seen in the sham and normal animals. Since SNL produces an important reduction in afferent inputs into the lumbar segments, the lack of change of neuronal responses, previously reported and discussed (Suzuki et al., 2001), suggests marked compensations in sensory systems following nerve ligation. The mechanisms underlying the marked shift to an inhibitory action of galanin following spinal nerve ligation are unknown. It has been reported that following axotomy both GalR1 mRNA and GalR2 mRNA are down-regulated in DRG neurones (Xu et al., 1996; Shi et al., 1997), although it is yet to be established whether this occurs also at the spinal cord level. These changes could result in a loss of both putative inhibitory and excitatory galanin receptors in the spinal cord (Wang et al., 1998). Such a selective plasticity in terms of GalR1 and GalR2 receptor activation is a possible basis for this inhibitory effect, as are differential changes in the complex circuitry of spinal pathways controlled by galanin receptors. It has been shown that all galanin-containing neurones in the superficial dorsal horn are GABAergic (Simmons et al., 1995). Since nerve injury, in this case sciatic nerve ligation, produces complex changes in the roles of GABA and its receptors after nerve injury (Kontinen et al., 2001), it is quite conceivable that the marked shift to inhibitions results from changes in these inhibitory neuronal populations. Another explanation
for the pronounced inhibition of neuronal responses in SNL rats is a potential up-regulation of inhibitory GalR3 receptors in the spinal cord following peripheral nerve injury. GalR3 receptors are present in low levels in the spinal cord (Smith et al., 1998), but the effect of peripheral nerve injury on this receptor population is currently unknown. Interestingly, neuropathic pain-related behaviour involves genes that map close to the GalR3 receptor (Seltzer et al., 2001). Whatever the case, exogenous galanin administration profoundly inhibits neuronal responses after nerve injury. In general, the inhibitory effects of galanin on evoked neuronal responses in sham-operated rats were only seen with the highest dose. In the case of the post-discharge and wind-up and the responses to innocuous and noxious mechanical stimuli, the inhibition of these responses in sham-operated animals approached those seen in SNL rats. However, the neuronal responses to noxious heat were not significantly inhibited. The sham operation is an essential control experiment for the SNL procedure, and may also represent a state of postsurgical tissue damage with an inflammatory component. It is unlikely that any post-operative inflammation is present 2 weeks following the surgery and the sham-operated animals did not show any behavioural mechanical hyperalgesia (Fig. 1A–C). However, it could be postulated that enduring subthreshold inflammatory processes could be occurring following the surgery, or the initial inflammatory insult could have induced long-lasting changes in galanin receptor expression in the spinal cord. Kuraishi et al. (1991b) found mechanical, but not thermal hyperalgesia induced by carrageenan inflammation was reversed by intrathecal administration of anti-galanin antiserum. This finding suggests that galanin is responsible for the generation of certain hyperalgesias seen after inflammation. We have previously reported that gabapentin, ineffective in normal animals, also inhibits neuronal responses in the sham-operated group (Chapman et al., 1998a). Inflammation has been shown to cause a decrease in GalR1 mRNA (Xu et al., 1996) and an increase in GalR2 mRNA (Shi et al., 1997) in DRG neurones. If these changes have functional importance, a facilitation of the neuronal responses in sham rats could possibly result from GalR2 receptor activation. However, we find other neuronal responses (von Frey 9 g and post discharge) in sham rats to be inhibited to the same extent as in SNL rats. A possible explanation for these modality-selective effects is that the inhibition seen in sham rats arises as a result of up-regulation of the putative inhibitory GalR3, present in low levels in the spinal cord (Smith et al., 1998). There may be regional distributions of the galanin receptors on different neuronal structures that allow these selective effects of galanin.
5. Conclusion We find exogenous spinal galanin administration causes a
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
facilitation of the neuronal responses to electrical and natural stimulation in normal rats. In sham rats, galanin at high doses causes an inhibition of their neuronal responses. Exogenous galanin elicits a marked inhibition of all neuronal responses, electrically evoked, mechanical and thermal responses in neuropathic rats. Whatever the mechanisms underlying these changes in the effects of galanin after peripheral nerve ligation, these results suggest that galanin or stable analogues may be a useful target for the treatment of nerve injury pain.
Acknowledgements S.J.L.F. is funded by MRC Industrial Collaborative Studentship.
References ¨ . Galanin – a neuroendocrine peptide. Crit Bartfai T, Ho¨ kfelt T, Langel U Rev Neurobiol 1993;7:229–274. Branchek TA, Smith KE, Gerald C, Walker MW. Galanin receptor subtypes. Trends Pharmacol Sci 2000;21:109–117. Chapman V, Suzuki R, Chamarette HL, Rygh LJ, Dickenson AH. Effects of systemic carbamazepine and gabapentin on spinal neuronal responses in spinal nerve ligated rats. Pain 1998a;75:261–272. Chapman V, Suzuki R, Dickenson AH. Electrophysiological characterization of spinal neuronal response properties in anaesthetized rats after ligation of spinal nerves L5–L6. J Physiol 1998b;507:881–894. Crawley JN. Galanin–acetylcholine interactions: relevance to memory and Alzheimer’s disease. Life Sci 1996;58:2185–2199. Habert-Ortholi E, Amiranoff B, Loquet I, Laburthe M, Mayaux JF. Molecular cloning of a functional human galanin receptor. Proc Natl Acad Sci USA 1994;91:9780–9783. Hao JX, Shi TJ, Xu IS, Kaupilla T, Xu XJ, Hokfelt T, Bartfai T, Wiesenfeld-Hallin Z. Intrathecal galanin alleviates allodynia-like behaviour in rats after partial peripheral nerve injury. Eur J Neurosci 1999;11:427– 432. Ho¨ kfelt T, Wisenfeld-Hallin Z, Villar M, Melander T. Increases of galaninlike immunoreactivity in rat dorsal root ganglion cells after peripheral axotomy. Neurosci Lett 1987;83:217–220. Ji RR, Zhang X, Zhang Q, Dagerlind A, Nilsson S, Wiesenfeld_Hallin Z, Hokfelt T. Central and peripheral expression of galanin in response to inflammation. Neuroscience 1995;68:563–576. Kerr BJ, Cafferty WB, Gupta YK, Bacon A, Wynick D, McMahon SB, Thompson SW. Galanin knockout mice reveal nociceptive deficits following peripheral nerve injury. Eur J Neurosci 2000;12:793–802. Kim SH, Chung JM. An experimental model for peripheral neuropathy produced by segmental spinal nerve ligation in the rat. Pain 1992;50:355–363. Kontinen VK, Stanfa LC, Suzuki R, Dickenson AH. Electrophysiological evidence for increased endogenous GABAergic, but not glycinergic inhibitory tone in the rat spinal nerve ligation model of neuropathy. Anesthesiology 2001;94:329–333. Kuraishi Y, Kawabata S, Matsumoto T, Nakamura A, Fujita H, Satoh M. Involvement of substance P in hyperalgesia induced by intrathecal galanin. Neurosci Res 1991a;11:276–285. Kuraishi Y, Kawamura M, Yamaguchi T, Houtani T, Kawabata S, Futaki S, Fujii N, Satoh M. Intrathecal injections of galanin and its antiserum affect nociceptive response of rat to mechanical, but not thermal, stimuli. Pain 1991b;44:321–324. Ma W, Bisby MA. Differential expression of galanin immunoreactivities in
257
the primary sensory neurons following partial and complete sciatic nerve injuries. Neuroscience 1997;79:1183–1195. Merchenthaler I, Lopez FJ, Negro-Vilar A. Anatomy and physiology of central galanin-containing pathways. Prog Neurobiol 1993;40:1183– 1195. Nussbaumer JC, Yanagisawa M, Otsuka M. Pharmacological properties of a C-fibre response evoked by saphenous nerve stimulation in an isolated spinal cord-nerve preparation of the newborn rat. Br J Pharmacol 1989;98:373–382. O’Donnell D, Ahmad S, Wahlestedt C, Walker P. Expression of the novel galanin receptor subtype GALR2 in the adult rat CNS: distinct distribution from GALR1. J Comp Neurol 1999;409:469–481. ¨ gren SO, Ho¨ kfelt T, Kask K, Langel U ¨ , Bartfai T. Evidence for a role of O galanin on spatial learning. Neuroscience 1992;51:1–5. Parker EM, Izzarelli DG, Nowak HP, Mahle CD, Iben LG, Wang J, Goldstein ME. Cloning and characterization of the rat GALR1 galanin receptor from Rin14B insulinoma cells. Brain Res Mol Brain Res 1995;34:179–189. Post C, Alari L, Hokfelt T. Intrathecal galanin increases the latency in the tail-flick and hot-plate test in mouse. Acta Physiol Scand 1988;132:583–584. Puttick RM, Pinnock RD, Woodruff GN. Galanin-induced membrane depolarization of neonatal rat cultured dorsal root ganglion cells. Eur J Pharmacol 1994;254:303–306. Reeve AJ, Walker K, Urban L, Fox AJ. Excitatory effects of galanin in the spinal cord of intact, anaesthetized rats. Neurosci Lett 2000;295:25–28. Satoh M, Kuraishi Y, Kawamura M. Effects of intrathecal antibodies to substance P, calcitonin gene-related peptide and galanin on repeated cold stress-induced hyperalgesia: comparison with carrageenaninduced hyperalgesia. Pain 1992;49:273–278. Seltzer Z, Wu T, Max BM, Diehl SR. Mapping a gene for neuropathic painrelated behavior following peripheral neurectomy in the mouse. Pain 2001;93:101–106. Shi T-J, Zhang X, Holmberg K, Xu Z-Q, Ho¨ kfelt T. Expression and regulation of galanin-R2 receptors in rat primary sensory neurons: effect of axotomy and inflammation. Neurosci Lett 1997;237:57–60. Simmons DR, Spike RC, Todd AJ. Galanin is contained in GABAergic neurons in the rat spinal dorsal horn. Neurosci Lett 1995;187:119–122. Smith KE, Forray C, Walker MW, Jones KA, Tamm JA, Bard J, Branchek TA, Linemeyer DL, Gerald C. Expression cloning of a rat hypothalamic galanin receptor coupled to phosphoinositide turnover. J Biol Chem 1997;272:24612–24616. Smith KE, Walker MW, Artymyshyn R, Bard J, Borowsky B, Tamm JA, Yao WJ, Vaysse PJ, Branchek TA, Gerald C, Jones KA. Cloned human and rat galanin GALR3 receptors. Pharmacology and activation of Gprotein inwardly rectifying K 1 channels. J Biol Chem 1998;273:23321– 23326. Suzuki R, Matthews EA, Dickenson AH. Comparison of the effects of MK801, ketamine and memantine on responses of spinal dorsal horn neurones in a rat model of mononeuropathy. Pain 2001;91:101–109. ˚ , Jo¨ rnall H, McDonald TJ, Mutt V. Galanin – a Tatemoto K, Ro¨ kaeus A novel biologically active peptide from porcine intestine. FEBS Lett 1983;164:124–128. Wang S, He C, Hashemi T, Bayne M. Cloning and expressional characterization of a novel galanin receptor. Identification of different pharmacophores within galanin for the three galanin receptor subtypes. J Biol Chem 1997;272:31949–31952. Wang S, Hashemi T, Fried S, Clemmons AL, Hawes BE. Differential intracellular signaling of the GalR1 and GalR2 galanin receptor subtypes. Biochemistry 1998;37:6711–6717. Waters SM, Krause JE. Distribution of galanin-1, -2 and -3 receptor messenger RNAs in central and peripheral rat tissues. Neuroscience 2000;95:265–271. Wiesenfeld-Hallin Z, Villar MJ, Hokfelt T. The effects of intrathecal galanin and C-fiber stimulation on the flexor reflex in the rat. Brain Res 1989a;486:205–213. Wiesenfeld-Hallin Z, Xu XJ, Villar MJ, Hokfelt T. The effect of intrathecal
258
S.J.L. Flatters et al. / Pain 98 (2002) 249–258
galanin on the flexor reflex in rat: increased depression after sciatic nerve section. Neurosci Lett 1989b;105:149–154. Xu Z-Q, Shi T-J, Landry M, Ho¨ kfelt T. Evidence for galanin receptors in primary sensory neurones and effect of axotomy and inflammation. Neuroreport 1996;8:237–242. Xu XJ, Hokfelt T, Bartfai T, Wiesenfeld-Hallin Z. Galanin and aspinal nociceptive mechanisms: recent advances and therapeutic implications. Neuropeptides 2000a;34:137–147.
Xu S-L, Zhang Y-P, Lundeberg T, Yu L-C. Effects of galanin on widedynamic range neuron activity in the spinal dorsal horn of rats with sciatic nerve ligation. Regul Pept 2000b;95:19–23. Yanagisawa M, Yagi N, Otsuka M, Yanaihara C, Yanaihara N. Inhibitory effects of galanin on the isolated spinal cord of the newborn rat. Neurosci Lett 1986;70:278–282. Zimmermann M. Ethical guidelines for investigations of experimental pain in conscious animals [editorial]. Pain 1983;16:109–110.