Selective ablation of dorsal horn NK1 expressing cells reveals a modulation of spinal alpha2-adrenergic inhibition of dorsal horn neurones

Neuropharmacology 54 (2008) 1208–1214

Contents lists available at ScienceDirect

Neuropharmacology journal homepage: www.elsevier.com/locate/neuropharm

Selective ablation of dorsal horn NK1 expressing cells reveals a modulation of spinal alpha2-adrenergic inhibition of dorsal horn neurones Wahida Rahman a, *, Rie Suzuki a, Stephen P. Hunt b, Anthony H. Dickenson a a b

Department of Pharmacology, University College London, Gower Street, London WC1E 6BT, UK Department of Anatomy and Developmental Biology, University College London, Gower Street, London WC1E 6BT, UK

a r t i c l e i n f o

a b s t r a c t

Article history: Received 27 November 2007 Received in revised form 20 March 2008 Accepted 26 March 2008

Activity in descending systems from the brainstem modulates nociceptive transmission through the dorsal horn. Intrathecal injection of the neurotoxin saporin conjugated to SP (SP–SAP) into the lumbar spinal cord results in the selective ablation of NK1 receptor expressing (NK1þve) neurones in the superficial dorsal horn (lamina I/III). Loss of these NK1þve neurones attenuates excitability of deep dorsal horn neurones due to a disruption of both intrinsic spinal circuits and a spino-bulbo-spinal loop, which activates a descending excitatory drive, mediated through spinal 5HT3 receptors. Descending inhibitory pathways also modulate spinal activity and hence control the level of nociceptive transmission relayed to higher centres. To ascertain the spinal origins of the major descending noradrenergic inhibitory pathway we studied the effects of a selective alpha2-adrenoceptor antagonist, atipamezole, on neuronal activity in animals pre-treated with SP–SAP. Intrathecal application of atipamezole dose dependently facilitated the mechanically evoked neuronal responses of deep dorsal horn neurones to low intensity von Frey hairs (5–15 g) and noxious thermal (45–50  C) evoked responses in SAP control animals indicating a physiological alpha2-adrenoceptor control. This facilitatory effect of atipamezole was lost in the SP–SAP treated group. These data suggest that activity within noradrenergic pathways have a dependence on dorsal horn NK1þve cells. Further, noradrenergic descending inhibition may in part be driven by lamina I/III (NK1þve) cells, and mediated via spinal alpha2-adrenoceptor activation. Since the same neuronal population drives descending facilitation and inhibition, the reduced excitability of lamina V/VI WDR neurones seen after loss of these NK1þve neurones indicates a dominant role of descending facilitation. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Electrophysiology Dorsal horn Alpha2-adrenoceptors Descending inhibition NK1 receptor

1. Introduction Neurokinin 1 receptor expressing (NK1þve) neurones, the primary target for Substance P (SP), are integral for the transmission of nociception and in the generation and maintenance of abnormal pain states (Mantyh and Hunt, 2004). Selective ablation of NK1þve neurones in lamina I and/or III of the dorsal horn results in reductions in behavioural hypersensitive responses seen in inflammatory, neuropathic and spinal cord injury models of pain (Mantyh et al., 1997; Mantyh and Hunt, 2004; Nichols et al., 1999; Suzuki et al., 2005; Yezierski, 2005). The majority of these NK1þve cells are located in lamina I of the dorsal horn (Mantyh et al., 1997; Spike et al., 2003) and are the origin of a spino-bulbo-spinal loop, driving a descending excitatory pathway from the brainstem onto spinal neurones. This circuit

* Corresponding author. Tel.: þ44 0207 679 3737; fax: þ44 0207 679 0181. E-mail address: [email protected] (W. Rahman). 0028-3908/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuropharm.2008.03.014

allows full coding of polymodal peripheral inputs under normal conditions, as well being essential for the more persistent chemical evoked responses of deep dorsal horn neurones. Pharmacological block of spinal 5HT3 receptors mimics many of the effects of ablating spinal NK1þve cells (Suzuki et al., 2002), suggesting that these neurones are essential for the facilitatory serotonergic influence from the brainstem (Suzuki et al., 2002). Furthermore, this loop contributes to some of the pathophysiological changes that follow nerve injury (Suzuki et al., 2004a, 2005) but is less important in inflammation (Rahman et al., 2004). Although loss of NK1þve cells reduces diffuse noxious inhibitory controls (Suzuki et al., 2002), it is unknown whether noradrenergic inhibitory modulations change – one possibility would be that an increase in noradrenergic descending inhibition contributes to the observed reduced spinal excitability that follows ablation of superficial NK1þve cells. This could be secondary to the loss of descending facilitations. Descending noradrenergic pathways exert powerful inhibitory influences onto the spinal cord, primarily via spinal alpha2adrenoceptors (Fields and Basbaum, 1994; Millan, 2002), which are

W. Rahman et al. / Neuropharmacology 54 (2008) 1208–1214

present in high density within the superficial lamina of the dorsal horn both on postsynaptic dorsal horn cells and primary afferent terminals (Nicholas et al., 1993; Roudet et al., 1994). Activation of noradrenergic receptors with alpha2-adrenoceptor agonists results in antinociception in acute and persistent pain models (Millan, 2002). Furthermore, changes in descending noradrenergic input to the dorsal horn of the spinal cord have been demonstrated (Green et al., 1998; Martin et al., 1999; Ren and Dubner, 1996; Tsuruoka and Willis, 1996; Wei et al., 1999). However, these enhancements in the descending noradrenergic system were only observed in persistent pain models, predominantly chemical inflammation. In this study, we investigated the effects of atipamezole, a selective alpha2-adrenoceptor antagonist (Scheinin et al., 1988; Schwartz and Clark, 1998), on the evoked activity of deep dorsal horn spinal neurones following ablation of NK1þve cells. The purpose was to determine the spinal neuronal origins of the system and to assess whether potential alterations in the alpha2-noradrenergic system contribute to the deficits in mechanical and thermal evoked responses of deep WDR neuronal activity seen after selective ablation of lamina I/III NK1þve cells using SP–SAP (Suzuki et al., 2002, 2004b). 2. Methods Sprague–Dawley rats were employed for this study (Central Biological Services, University College London, UK), and all experimental procedures were approved by the UK Home Office and followed the guidelines under the International Association for the Study of Pain (Zimmermann, 1983). 2.1. Selective ablation of NK1 expressing neurones in the lumbar cord Intrathecal injection of the neurotoxin saporin conjugated to SP (SP–SAP) or saporin (SAP) alone (Advanced Targeting Systems, San Diego, California) was performed as previously described (Suzuki et al., 2002). Rats (130–150 g) were deeply anaesthetised with a combination of diazepam (Phoenix pharmaceuticals, 2.5 mg/kg i.p.) and Hypnorm (Janssen Animal Health, 0.3 mg/kg i.m.). A small incision was made in the atlanto-occipital membrane and a cannula was inserted into the subarachnoid space, terminating in the L4–5 region. The animals received either 10 ml of 1 mM SP–SAP or SAP (control group) injected slowly via a Hamilton syringe attached to the cannula, and flushed with 5 ml saline. The cannula was withdrawn and the wound closed using 3-0 surgical sutures (Sherwood Davis & Geck, USA) and wound clips. Animals were allowed to recover and not used for further study until 28 days after intrathecal injection. 2.2. Immunohistochemistry To confirm selective ablation of superficial NK1 expressing cells in the dorsal horn of SP–SAP treated animals, NK1 immunohistochemistry reactions were carried out as previously described (Suzuki et al., 2002). Briefly, deeply anaesthetised animals were transcardially perfused with 4% paraformaldehyde in PBS (0.15 M, pH 7.4), at 10  C. The spinal cord was removed and cryoprotected overnight in 30% sucrose solution. Transverse sections (40 mm thick) of lumbar spinal cord were cut and serially collected. Antibody reactions were carried out against NK1 (1:10,000, Eurogentec), NeuN (1:1000, Chemicon) and GFAP (1:1000, Dako). Immunoreactive sites were revealed using biotinylated secondary antibodies followed by avidin conjugated fluorescent antibodies Cy3 or FITC. All sections were mounted and coverslipped in Gel Mount (Sigma). 2.3. Electrophysiology In vivo electrophysiology was conducted 4 weeks after intrathecal injection as previously described (Urch and Dickenson, 2003). Animals were anaesthetised with halothane (1.0–1.2%, 66% N2O and 33% O2) and a laminectomy was performed to expose the L4–5 segments of the spinal cord. Extracellular recordings were made from ipsilateral deep dorsal horn (lamina V/VI) wide dynamic range neurones using parylene coated tungsten electrodes (A-M Systems, USA). One neurone was recorded from each animal and all neurones used in the study had defined receptive fields in the toe regions of the hind paw. A train of 16 transcutaneous electrical stimuli (2 ms wide pulses, 0.5 Hz) was applied at three times the threshold current for C-fibres, following which a poststimulus histogram was constructed. Responses evoked by Ab- (0–20 ms), Ad(20–90 ms) and C-fibres (90–350 ms) were separated and quantified on the basis of latency. Responses occurring after the C-fibre latency band were taken to be the post-discharge of the cell (350–800 ms). The peripheral receptive field was also stimulated using a range of natural stimuli for a period of 10 s per stimulus.

1209

Mechanical stimulus was applied using hand held von Frey filaments – 2, 6, 8, 15, 26 and 60 g, corresponding to approximately19.6, 58.9, 78.5, 147.2 and 588.6 mN (Touch-test TM, North Coast Medical Inc., San Jose, CA, USA). Heat (35, 40, 45, 48 and 50  C) was applied with a constant water jet onto the centre of the receptive field. Application of each individual von Frey hair was separated by a minimum interval period of 5–10 s, and longer for very responsive neurones at the higher intensity range. Application of each subsequent heat stimulus was separated by a minimum period of 1 min. Data were captured and analysed by a CED 1401 interface coupled to a Pentium computer with Spike 2 software (Cambridge Electronic Design; PSTH and rate functions). The testing procedure was carried out every 20 min and consisted of a train of 16 electrical stimuli followed by natural stimuli as described above. Following three consecutive stable control trials (<10% variation for the C-fibre evoked response, and <20% variation for all other parameters) neuronal responses were averaged to give the pre-drug control values. 2.4. Drug administration Multiple spinal injections of atipamezole (Orion, Finland) were applied directly to the surface of the spinal cord, in a volume of 50 ml, using a Hamilton syringe, in cumulative doses of 1, 10 and 100 mg was given to SAP and SP–SAP treated animals and the effect of each dose or control vehicle application was followed over an hour, with tests carried out at 20, 40 and 60 min before subsequent drug applications were made and effects again followed for an hour. 2.5. Data analysis Data are presented as mean  standard error of mean (S.E.M.). Pre-drug control (electrical, mechanical and thermal) responses between groups were compared using unpaired Student’s t-test. Drug effects are expressed as mean maximal evoked neuronal response. Drug effects are expressed as the mean maximal evoked neuronal response for each dose. Statistical analysis was performed on raw data, with drug effects within a group assessed using one-way analysis of variance with repeated measures (RM-ANOVA). Where a significant effect was seen with increasing dose, Dunnett’s post hoc tests were then used to assess individual dose effects compared with pre-drug baseline controls. Drug effects between groups were analysed using a two-way ANOVA followed by Bonferroni’s post test. All statistical testing was performed using Prism 4.0 software (Graphpad/Prism, San Diego, CA, USA). Level of significance was set at *P < 0.05.

3. Results Following intrathecal injection of SAP or SP–SAP, animals showed normal grooming behaviour and weight gain. 3.1. NK1 immunohistochemistry Histological examination of lumbar cord sections confirmed a marked depletion of NK1 immunoreactivity in lamina I, of SP– SAP treated animals compared with SAP treated animals. A reduction in NK1 receptor immunofluoresence was also observed in lamina III in the SP–SAP group, although this effect was much less marked. Immunoreactivity for neurone-specific nuclear protein (NeuN, a marker for neurones) and glial fibrillary acidic protein (GFAP, a marker for astrocytes) showed no abnormal cell death or glial proliferation or hypertrophy (data not shown). These findings are in keeping with previous studies (Mantyh et al., 1997; Nichols et al., 1999; Rygh et al., 2006; Suzuki et al., 2002). It has also been previously shown that the effects of SAP alone, since it cannot enter the neurone and therefore cannot exert any neurotoxic effects, is no different to intrathecal injection of saline into the lumbar spinal cord (Mantyh et al., 1997; Nichols et al., 1999; Suzuki et al., 2002). 3.2. Neuronal response characteristics in naive, SAP and SP–SAP treated animals The depths of the cells recorded were similar between animal groups (811  68 SAP and 710  63 mm SP–SAP) and correspond to lamina V/VI of the dorsal horn. Similarly, the neuronal thresholds for C-fibre activation (0.63  0.1 SAP and 0.9  0.1 mA SP–SAP) and the electrically evoked responses of the cells in each group (Fig. 1) prior to drug application were not significantly

1210

W. Rahman et al. / Neuropharmacology 54 (2008) 1208–1214

A

Aβ-fibre

B

150

Aδ-fibre 150

SAP SP-SAP

100

50

50

Spikes

100

Control

600

*

*

1

10

100

10

100

0

0

C

* 1

10

100

Control

D

C-fibre

400

600

Postdischarge

Spikes

400

200

*

0

200

0 Control

1

10

100

atipamezole μg

Control

1

atipamezole μg

Fig. 1. The effect of atipamezole on the (A) Ab-fibre evoked response, (B) Ad-fibre evoked response, (C) C-fibre evoked response and (D) post-discharge of spinal neurones in SAP (n ¼ 7) and SP–SAP (n ¼ 7) groups. Atipamezole produced a dose-related inhibition of the Ad- and C-fibre evoked responses of spinal neurones in SP–SAP treated rats but was without significant effect in the SAP group. Data are presented as mean number of action potentials  S.E.M. *P < 0.05 indicates significance of atipamezole effects compared with pre-drug control values (one-way RM-ANOVA with Dunnett’s post test).

different. However, a disruption in the ability of the lamina V WDR cells to code for mechanical and thermal stimuli was seen following ablation of NK1 expressing cells in the superficial lamina of the dorsal horn, similar to those previously reported by Suzuki et al. (2002). We observed marked reductions in the evoked response to mechanical stimuli (von Frey 6 and 15–60 g, P < 0.05 unpaired Student’s t-test) in SP–SAP treated animals compared with SAP controls (Fig. 2). Similarly deficits in neuronal response to heat (45 and 48  C, P < 0.05 unpaired Student’s t-test) (Fig. 3) were also observed in the SP–SAP group compared with SAP controls.

3.3. Effects of atipamezole on the electrically evoked responses of deep dorsal horn neurones in SAP and SP–SAP treated animals Spinal application of atipamezole (1, 10 and 100 mg) did not produce any significant effects on the electrically evoked Ab-, Ad-, C-fibre responses or post-discharge of spinal neurones in the SAP control group (Fig. 1). These findings are in keeping with previous electrophysiological studies which reported a lack of effect of atipamezole on electrically evoked neuronal activity in normal animals (Green et al., 1998; Stanfa and Dickenson, 1994; Sullivan et al., 1992). Surprisingly, a small yet significant reduction was seen of the C-fibre (P < 0.05 at 100 mg atipamezole; one-way RM-ANOVA with Dunnett’s post test) and the Ad-fibre evoked neuronal responses (P < 0.05 at 1, 10 and 100 mg atipamezole; one-way RM-ANOVA with Dunnett’s post test) in the SP–SAP group, suggesting that loss

of NK1þve cells unmasks the presence of a facilitatory effect of blocking spinal alpha2-adrenoceptors. 3.4. Spinal application of atipamezole enhances the mechanical punctate and thermal evoked responses of deep dorsal horn neurones in SAP control animals only Selectively blocking spinal alpha2-adrenoceptors produced a dose-related facilitation of the evoked neuronal responses to stimulation of the peripheral receptive field with mechanical stimuli in the SAP control group (Fig. 2). These robust enhancements were clearly significant compared with the pre-drug baseline control responses for von Frey 6 g (significant at 10 mg P < 0.01 and 100 mg P < 0.05), von Frey 8 g (significant at 10 and 100 mg P < 0.01) and von Frey 15 g (significant at 10 and 100 mg P < 0.01; one-way RM-ANOVA with Dunnett’s post test). Further, these excitatory effects of atipamezole administration in the SAP group were significantly different compared with the effects of the drug on these neuronal measures in the SP–SAP group (P < 0.05, twoway ANOVA with Bonferroni’s post test). Modest facilitations were also observed with higher intensity noxious mechanical stimuli, however, intrathecal administration of atipamezole, at these doses, did not produce any significant effects on the neuronal responses evoked by these intense suprathreshold mechanical stimuli compared with pre-drug control values (von Frey 26 and 60 g P > 0.05 one-way RM-ANOVA). Dose-related increases to the thermal evoked responses were also observed in the SAP control group; atipamezole produced

W. Rahman et al. / Neuropharmacology 54 (2008) 1208–1214

Spikes

1250

1250

vF 2g

vF 6g SAP

1000

1000

750

750

500

500

250

250

0

0 control

1

10

100

SP-SAP

1250

1000

Spikes

*

1

10

100

*

* *

10

100

10

100

atipamezole µg

vF 8g

vF 15g

1000

* *

750

* *

§ control

atipamezole µg 1250

1211

* *

750

500

500

250

250

0

0

§ control

1

10

100

control

atipamezole µg

Spikes

1250

1

atipamezole µg 1250

vF 26g

1000

1000

750

750

500

500

vF 60g

§ 250

250

§

0

0 control

1

10

100

atipamezole µg

control

1

atipamezole µg

Fig. 2. The facilitatory effect of atipamezole on the mechanically evoked responses of deep dorsal horn neurones is lost in SP–SAP animals compared with the SAP control group. Atipamezole significantly increases the mechanically evoked responses of spinal neurones to von Frey 6, 8 and 15 g in the SAP control group (n ¼ 7). This effect is not seen after SP– SAP treatment (n ¼ 7). Data are presented as mean number of action potentials  S.E.M. xP < 0.05 indicates significant difference in pre-drug baseline responses comparing SAP and SP–SAP groups (unpaired Student’s t-test). *P < 0.05, **P < 0.01 indicate significance of atipamezole effects compared with pre-drug control values (one-way RM-ANOVA with Dunnett’s post test).

significant enhancements of the evoked responses to noxious heat following application of 45  C (significant at 10 and 100 mg P < 0.05) and 48  C (significant at 10 and 100 mg P < 0.01, one-way RM-ANOVA with Dunnett’s post test). A facilitation of evoked responses to 50  C was also seen although only the highest dose proved significant (P < 0.05, one-way RM-ANOVA with Dunnett’s post test). In contrast, the effects of atipamezole on low intensity, innocuous, temperatures (35 and 40  C) were minimal with only a tendency towards an increase in evoked responses observed (Fig. 3). This is unlike the effects of atipamezole on low intensity mechanical stimuli where quite dramatic increases in evoked response was seen. These findings suggest a differential control of mechanical and thermal inputs by descending noradrenergic pathways such that alterations in

alpha2-adrenergic systems could play a role in mechanical allodynia and heat hyperalgesia, respectively. In complete contrast to the facilitatory effects of atipamezole on the natural mechanical and thermal evoked neuronal responses seen in SAP control animals, blocking spinal alpha2-adrenoceptors in SP–SAP treated animals produced little or no effect on these evoked neuronal responses (P > 0.05) (Figs. 2 and 3). Thus after SP–SAP ablation of superficial NK1 expressing neurones atipamezole had no significant effect whatsoever on any of the low and high intensity natural stimuli, both mechanical and thermal. Thus there is a loss of spinal alpha2-adrenoceptor mediated inhibitory controls acting on these deep WDR spinal neurones after SP–SAP treatment.

1212

W. Rahman et al. / Neuropharmacology 54 (2008) 1208–1214

Spikes

1400

1400

35ºC

1200

1200

1000

1000

800

800

600

600

400

400

200

200

40ºC

SP-SAP

0

0 control

1

10

control

100

atipamezole µg 1400

1400

Spikes

*

*

10

100

48ºC

* *

* *

10

100

1200

1000

1000

800

800

600

600

400

400

§

200

§

0

0 control

1

10

100

control

1

atipamezole µg

atipamezole µg 1400

1

atipamezole µg

45ºC

1200

200

SAP

*

50ºC

1200

Spikes

1000 800 600 400 200 0 control

1

10

100

atipamezole µg Fig. 3. Effects of atipamezole on the heat evoked response of deep dorsal horn neurones in SAP (n ¼ 7) and SP–SAP (n ¼ 7) treated animals. In contrast to the minimal effects of atipamezole on the heat evoked responses on deep dorsal horn neurones in SP–SAP treated group, atipamezole produced a dose-related increase in the evoked responses to noxious heat stimuli (45 and 48  C only) in SAP control animals. Data are presented as mean number of action potentials  S.E.M. xP < 0.05 indicates significant difference in pre-drug baseline responses comparing SAP and SP–SAP groups (unpaired Student’s t-test). *P < 0.05, **P < 0.01 indicate significance of atipamezole effects compared with pre-drug control values (one-way RM-ANOVA with Dunnett’s post test).

4. Discussion Noradrenaline, acting at spinal alpha2-adrenoceptors, is one of the main sources for descending inhibitory modulation of spinal nociceptive processing, with a large body of evidence showing a reduction in pain behaviours and noxious evoked neuronal activity with alpha2 agonists in models of acute and persistent pain states (Millan, 2002). Indeed, clonidine has been used as a clinical analgesic (Eisenach et al., 1996). However, whereas agonist studies report on the consequences of activation of a receptor by an exogenous drug, antagonism of physiological responses reveals the role of released transmitter at the receptor, which will depend on activity in related neuronal pathways.

4.1. Descending noradrenergic systems have a pronounced inhibitory effect, mediated by spinal alpha2-adrenoceptor activation, on mechanical and thermal evoked deep dorsal horn neuronal activity Earlier electrophysiological studies suggested little tonic activity within descending noradrenergic pathways in non-pathological conditions. These studies, however, only investigated suprathreshold electrically evoked responses of dorsal horn neurones (Green et al., 1998; Millan, 2002; Stanfa and Dickenson, 1994; Sullivan et al., 1992). In contrast we have shown here that atipamezole significantly facilitated evoked neuronal responses to von Frey 6, 8 and 15 g in the SAP group (Fig. 3). Since these are low

W. Rahman et al. / Neuropharmacology 54 (2008) 1208–1214

intensity innocuous (vF 6 and 8 g) and mildly noxious (vF 15 g) mechanical stimuli, and the effects of the antagonist on stronger stimuli were much less pronounced, it is possible that potential changes in alpha2-adrenoceptor mediated controls could relate primarily to mechanical punctate allodynia rather than mechanical hyperalgesia. Dose-related facilitation of the evoked responses to 45, 48 and 50  C heat was seen in the SAP group, with little or no effect of atipamezole seen on the innocuous temperatures tested in these animals. This suggests that a strong descending noradrenergic inhibitory control of noxious thermal evoked responses exists in naı¨ve animals. Interestingly, 45 and 48  C heat evoked responses are subject to 5HT3 receptor mediated facilitation, since acute spinal application of ondansetron (selective 5HT3 antagonist) produced significant reductions in the evoked responses to these temperatures (Suzuki et al., 2002). Therefore these findings suggest that under normal conditions noxious thermal stimuli are subject to both descending inhibition and facilitation. In contrast, there appears to be a differential modulation of mechanical evoked neuronal responses by descending controls, with lower intensity mechanical punctate stimuli subject to greater noradrenergic inhibition as demonstrated in this study and mechanical punctate stimuli in the noxious range under a greater facilitatory modulation mediated by 5HT3 receptor activation (Suzuki et al., 2002). 4.2. Loss of NK1þve cells results in a loss of descending inhibition of mechanical and thermal evoked responses of lamina V/VI WDR neurones In this study we show a tonic and/or evoked descending noradrenergic inhibitory control of low intensity mechanical punctate and noxious thermal evoked responses of deep dorsal horn neurones in SAP rats. In comparison, atipamezole produced little or no effect on the naturally evoked responses in SP–SAP treated animals, demonstrating a link between spinal NK1þve cells and activity in descending noradrenergic pathways, which may reside in intrinsic spinal synaptic contacts or possibly through activation of descending controls. Noradrenergic innervation of the dorsal horn originates exclusively from the locus coeruleus and other brainstem sites (Kwiat and Basbaum, 1992; Millan, 2002), thus it is possible that ablation of NK1þve neurones removes the cells of origin of ascending pathways that then drive descending noradrenergic inhibition of the spinal cord. Spinal NK1þve neurones are critical for excitatory transmission through lamina V/VI WDR neurones, since loss of NK1þve cells results in significant reduction in the full polymodal coding response characteristics of these cells (Suzuki et al., 2002). Pharmacological evidence identified 5HT acting on spinal excitatory 5HT3 receptors as an important physiological substrate for this circuit (Suzuki et al., 2002). Thus these cells are at the origin of a spinobulbo-spinal loop that drives descending 5HT3 mediated facilitation. However, central sensitization of deep dorsal horn neurones as revealed by wind-up and long term facilitation also depend on these NK1þve cells, but these events are independent of the integrity of this 5HT3 receptor mediated system, and so are most likely intrinsic spinal events (Rygh et al., 2006; Suzuki et al., 2002). In addition to this example of excitability regulated within intrinsic spinal circuits, and through descending facilitation, nociceptive processing can also be modulated by changes in inhibitory systems, which, like the excitatory circuits, can be spinal or descending from the brain. In this study we provide electrophysiological data to show that in addition to the descending facilitatory 5HT3 spino-bulbo-spinal loop described by Suzuki et al. (2002), the same NK1þve cells likely participate in a spino-ponto-spinal loop, since the facilitatory effect of blocking spinal alpha2-adrenoceptors on the evoked activity of deep dorsal horn WDR neurones in SAP

1213

controls was lost after SP–SAP treatment. An alpha2A-adrenoceptor mediated feedback inhibition of capsaicin-induced hyperalgesia was recently described (Mansikka et al., 2004). Since activation of NK1þve cells by capsaicin is a likely mechanism by which this hyperalgesia occurs, the findings of Mansikka et al. (2004) support the suggestion that NK1þve cells drive a descending noradrenergic inhibition of spinal neurones via activation of spinal alpha2-adrenoceptors. Further, the anatomical termination pattern of NK1þve cells implies that they are well placed to activate descending pathways to the dorsal horn. The large majority of lamina I/III NK1þve cells project to brainstem and thalamic sites, important in sensory, affective and autonomic aspects of pain, which in turn modulate descending monoaminergic pathways from the brainstem (Gauriau and Bernard, 2002; Naim et al., 1997; Todd et al., 2000). Additionally, a small proportion of NK1þve cells project to the periaqueductal grey (Spike et al., 2003), which targets noradrenergic neurones localised in the pontine A5, A6 (locus coeruleus) and the pontine A7 (sub coeruleus) regions which are key supraspinal sources of descending noradrenergic pathways projecting to the spinal cord (Cameron et al., 1995; Kwiat and Basbaum, 1992). The increased neuronal response to low intensity mechanical stimuli after atipamezole administration in the SAP group, which is lost after SP–SAP treatment, is difficult to explain. However, a recent study demonstrated the presence of A-fibre input to lamina I/ III NK1þve neurones, which is normally under strong GABAergic inhibitory control. The authors showed that blocking spinal inhibition with bicuculline enabled lamina I and III NK1þve neurones to respond A-fibre stimulation (Torsney and MacDermott, 2006). A small proportion of alpha2-adrenoceptors are located on inhibitory GABAergic interneurones (Olave and Maxwell, 2002) thus it is a possible that atipamezole could, via a mechanism of disinhibition, effectively ‘‘turn’’ high threshold NK1þve neurones into ‘‘WDR’’ cells and potentially activate the excitatory loop. It is also possible that synaptic reorganization and/or loss of alpha2-adrenoceptor number/function, following SP–SAP treatment, could underlie the loss of effect of atipamezole in the SP–SAP group. Further studies are needed to investigate whether the atipamezole mediated increase in neuronal responses observed in the SAP group is due to activation of descending facilitation or via activity in intrinsic connections of NK1þve cells in the dorsal horn, or both. 4.3. Conclusion This study demonstrates a link between lamina I/III dorsal horn NK1þve cells and activity within descending noradrenergic pathways since the excitatory effects of atipamezole on deep dorsal horn neuronal responses in control animals are lost after SP–SAP treatment. Interestingly, superficial NK1þve projection cells receive synaptic contacts from glutamatergic interneurones, which possess alpha2C-adrenoceptors, and this could well be a mechanism by which alpha2 agonists exert their antinociceptive effects and provide an immunohistochemical basis for noradrenergic modulation of NK1þve neurones (Olave and Maxwell, 2002). Nonetheless, the crucial finding from this study is that the pronounced reduction in deep dorsal horn neuronal excitability seen after ablation of NK1þve cells is not due to an enhancement of descending inhibitions mediated via spinal alpha2-adrenoceptors. Furthermore we have previously shown that ablation of lamina I/III NK1þve neurones also results in a loss of activity in intrinsic GABAergic transmission in the spinal cord compared with SAP controls (Rahman et al., 2007). Taken together these data further support the findings of Suzuki et al. (2002) who showed that deficits in spinal cord excitability resulting from ablation of superficial NK1þve cells were largely due to a loss of a descending facilitatory drive mediated by 5HT3 receptors. This suggests that under normal conditions, acute nociceptive stimuli result in the activation of lamina I/III

1214

W. Rahman et al. / Neuropharmacology 54 (2008) 1208–1214

NK1þve cells and under these conditions the balance between facilitation and inhibition shifts to a dominant role of descending facilitation. These animal findings may relate to a study where patients with fibromyalgia, a diffuse pain condition, had normal inhibitory controls, meaning that loss/disruption of inhibitory controls is unlikely to play a role in this pain state and perhaps suggests for a role for diffuse monoamine facilitations in this condition (Staud et al., 2004). Interestingly there is clinical data showing that a block of 5HT3 receptors can be effective in treating fibromyalgia (Haus et al., 2004; Stratz et al., 2004). These findings taken together lend support for a key role for descending facilitation in mediating different pain states, as opposed to a disruption/ alteration in descending noradrenergic inhibition. Acknowledgements This work was supported by the Wellcome Trust. References Cameron, A.A., Khan, I.A., Westlund, K.N., Willis, W.D., 1995. The efferent projections of the periaqueductal gray in the rat: a Phaseolus vulgaris-leucoagglutinin study. II. Descending projections. J. Comp. Neurol. 351, 585–601. Eisenach, J., De Kock, M., Klimscha, W., 1996. a2-Adrenergic agonists for regional anesthesia – a clinical review of clonidine (1984–1995. Anesthesiology 85, 655–674. Fields, H., Basbaum, A., 1994. Central nervous system mechanisms of pain modulation. In: Wall, P., Melzack, R. (Eds.), Textbook of Pain. Churchill Livingston, Edinburgh, pp. 243–257. Gauriau, C., Bernard, J.F., 2002. Pain pathways and parabrachial circuits in the rat. Exp. Physiol. 87, 251–258. Green, G.M., Lyons, L., Dickenson, A.H., 1998. Alpha2-adrenoceptor antagonists enhance responses of dorsal horn neurones to formalin induced inflammation. Eur. J. Pharmacol. 347, 201–204. Haus, U., Spath, M., Farber, L., 2004. Spectrum of use and tolerability of 5-HT3 receptor antagonists. Scand. J. Rheumatol. Suppl., 12–18. Kwiat, G.C., Basbaum, A.I., 1992. The origin of brainstem noradrenergic and serotonergic projections to the spinal cord dorsal horn in the rat. Somatosens. Mot. Res. 9, 157–173. Mansikka, H., Lahdesmaki, J., Scheinin, M., Pertovaara, A., 2004. Alpha(2A) adrenoceptors contribute to feedback inhibition of capsaicin-induced hyperalgesia. Anesthesiology 101, 185–190. Mantyh, P., Rogers, S., Honore, P., Allen, B., Ghilardi, J., Li, J., Daughters, R., Lappi, D., Wiley, R., Simone, D., 1997. Inhibition of hyperalgesia by ablation of lamina I spinal neurons expressing the substance P receptor. Science 278, 275–279. Mantyh, P.W., Hunt, S.P., 2004. Setting the tone: superficial dorsal horn projection neurons regulate pain sensitivity. Trends Neurosci. 27, 582–584. Martin, W., Gupta, N., Loo, C., Rohde, D., Basbaum, A., 1999. Differential effects of neurotoxic destruction of descending noradrenergic pathways on acute and persistent nociceptive processing. Pain 80, 57–65. Millan, M.J., 2002. Descending control of pain. Prog. Neurobiol. 66, 355–474. Naim, M., Spike, R.C., Watt, C., Shehab, S.A., Todd, A.J., 1997. Cells in laminae III and IV of the rat spinal cord that possess the neurokinin-1 receptor and have dorsally directed dendrites receive a major synaptic input from tachykinincontaining primary afferents. J. Neurosci. 17, 5536–5548. Nicholas, A., Pieribone, V., Hokfelt, T., 1993. Distributions for mRNA’s for alpha2adrenergic receptor subtypes in rat brain: an in situ hybridization study. J. Comp. Neurol. 328, 575–594. Nichols, M., Allen, B., Rogers, S., Ghilardi, J., Honore, P., Luger, N., Finke, M., Li, J., Lappi, D., Simone, D., Mantyh, P., 1999. Transmission of chronic nociception by spinal neurons expressing the substance P receptor. Science 286, 1558–1561.

Olave, M.J., Maxwell, D.J., 2002. An investigation of neurones that possess the alpha 2C-adrenergic receptor in the rat dorsal horn. Neuroscience 115, 31–40. Rahman, W., Sikander, S., Suzuki, R., Hunt, S.P., Dickenson, A.H., 2007. Superficial NK1 expressing spinal dorsal horn neurones modulate inhibitory neurotransmission mediated by spinal GABA(A) receptors. Neurosci. Lett. 419, 278–283. Rahman, W., Suzuki, R., Rygh, L.J., Dickenson, A.H., 2004. Descending serotonergic facilitation mediated through rat spinal 5HT3 receptors is unaltered following carrageenan inflammation. Neurosci. Lett. 361, 229–231. Ren, K., Dubner, R., 1996. Enhanced descending modulation of nociception in rats with persistent hindpaw inflammation. J. Neurophysiol. 76, 3025–3037. Roudet, C., Mouchet, P., Feuerstein, C., Savasta, M., 1994. Normal distribution of alpha 2-adrenoceptors in the rat spinal cord and its modification after noradrenergic denervation: a quantitative autoradiographic study. J. Neurosci. Res. 39, 319–329. Rygh, L.J., Suzuki, R., Rahman, W., Wong, Y., Vonsy, J.L., Sandhu, H., Webber, M., Hunt, S., Dickenson, A.H., 2006. Local and descending circuits regulate long-term potentiation and zif268 expression in spinal neurons. Eur. J. Neurosci. 24, 761–772. Scheinin, H., MacDonald, E., Scheinin, M., 1988. Behavioural and neurochemical effects of antipamezole, a novel alpha 2-adrenoceptor antagonist. Eur. J. Pharmacol. 151, 35–42. Schwartz, D.D., Clark, T.P., 1998. Selectivity of atipamezole, yohimbine and tolazoline for alpha-2 adrenergic receptor subtypes: implications for clinical reversal of alpha-2 adrenergic receptor mediated sedation in sheep. J. Vet. Pharmacol. Ther. 21, 342–347. Spike, R.C., Puskar, Z., Andrew, D., Todd, A.J., 2003. A quantitative and morphological study of projection neurons in lamina I of the rat lumbar spinal cord. Eur. J. Neurosci. 18, 2433–2448. Stanfa, L.C., Dickenson, A.H., 1994. Enhanced alpha-2 adrenergic controls and spinal morphine potency in inflammation. Neuroreport 5, 469–472. Staud, R., Vierck, C.J., Robinson, M.E., Price, D.D., 2004. Spatial summation of heat pain within and across dermatomes in fibromyalgia patients and pain-free subjects. Pain 111, 342–350. Stratz, T., Fiebich, B., Haus, U., Muller, W., 2004. Influence of tropisetron on the serum substance P levels in fibromyalgia patients. Scand. J. Rheumatol. Suppl., 41–43. Sullivan, A.F., Kalso, E.A., McQuay, H.J., Dickenson, A.H., 1992. The antinociceptive actions of dexmedetomidine on dorsal horn neuronal responses in the anaesthetized rat. Eur. J. Pharmacol. 215, 127–133. Suzuki, R., Morcuende, S., Webber, M., Hunt, S.P., Dickenson, A.H., 2002. Superficial NK1-expressing neurons control spinal excitability through activation of descending pathways. Nat. Neurosci. 5, 1319–1326. Suzuki, R., Rahman, W., Hunt, S.P., Dickenson, A.H., 2004a. Descending facilitatory control of mechanically evoked responses is enhanced in deep dorsal horn neurones following peripheral nerve injury. Brain Res. 1019, 68–76. Suzuki, R., Rahman, W., Rygh, L.J., Webber, M., Hunt, S.P., Dickenson, A.H., 2005. Spinal–supraspinal serotonergic circuits regulating neuropathic pain and its treatment with gabapentin. Pain 117, 292–303. Suzuki, R., Rygh, L.J., Dickenson, A.H., 2004b. Bad news from the brain: descending 5-HT pathways that control spinal pain processing. Trends Pharmacol. Sci. 25, 613–617. Todd, A.J., McGill, M.M., Shehab, S.A., 2000. Neurokinin 1 receptor expression by neurons in laminae I, III and IV of the rat spinal dorsal horn that project to the brainstem. Eur. J. Neurosci. 12, 689–700. Torsney, C., MacDermott, A.B., 2006. Disinhibition opens the gate to pathological pain signaling in superficial neurokinin 1 receptor-expressing neurons in rat spinal cord. J. Neurosci. 26, 1833–1843. Tsuruoka, M., Willis, W., 1996. Descending modulation from the region of the locus coeruleus on nociceptive sensitivity in a rat model of inflammatory hyperalgesia. Brain Res. 743, 86–92. Urch, C.E., Dickenson, A.H., 2003. In vivo single unit extracellular recordings from spinal cord neurones of rats. Brain Res. Protoc. 12, 26–34. Wei, F., Dubner, R., Ren, K., 1999. Laminar-selective noradrenergic and serotoninergic modulation includes spinoparabrachial cells after inflammation. Neuroreport 10, 1757–1761. Yezierski, R.P., 2005. Spinal cord injury: a model of central neuropathic pain. Neurosignals 14, 182–193. Zimmermann, M., 1983. Ethical guidelines for investigations of experimental pain in conscious animals. Pain 16, 109–110.