- Email: [email protected]
Endogenous adrenergic control of reflexes evoked by mechanical stimulation of the heel in the decerebrated rabbit
Neuroscience Letters 308 (2001) 189±192
www.elsevier.com/locate/neulet
Endogenous adrenergic control of re¯exes evoked by mechanical stimulation of the heel in the decerebrated rabbit R.W. Clarke*, J. Harris, A.K. Houghton 1 Division of Animal Physiology, School of Biosciences, Sutton Bonington Campus, Loughborough LE12 5RD, UK Received 2 May 2001; received in revised form 5 June 2001; accepted 11 June 2001
Abstract In decerebrated rabbits, re¯exes were evoked in medial gastrocnemius motoneurones by mechanical stimulation of the heel, using four pinch strengths from 183 (innocuous) to 4577 (noxious) mN. The a 2-adrenoceptor antagonist idazoxan (1±156 mg intrathecal (i.th.)) signi®cantly increased responses to pinch strengths of 607 mN and above. Subsequent administration of the a 1 adrenoceptor selective antagonist prazosin (200 mg i.th.) decreased re¯exes to 4577 mN pinches but had no other signi®cant effects. The opioid antagonist (2)-quadazocine (25 mg i.th.) caused no further changes in re¯exes. Spinal section at L1 in the presence of this drug combination enhanced gastrocnemius responses to 183 and 607 mN stimuli, had no effect on re¯exes to 1866 mN and signi®cantly decreased responses to 4577 mN pinches. These data con®rm that re¯exes evoked by `natural' stimulation of heel mechanoreceptors are subject to powerful tonic descending inhibition mediated by a 2-adrenoceptors. For the highest strength stimulus, the results of a 2 blockade involved enabling of descending facilitation as well as reduction of descending inhibition. q 2001 Published by Elsevier Science Ireland Ltd. Keywords: Withdrawal re¯ex; Noradrenaline; Spinal cord; Nociception; Descending inhibition; Descending facilitation
Noradrenaline, acting through a 2-adrenoceptors, is thought to be an important mediator of descending inhibitory control of the spinal cord [7], but there are few convincing examples of tonic inhibition that can be relieved by adrenoceptor antagonists. In decerebrated rabbits, re¯exes evoked in the ankle extensor medial gastrocnemius (MG) by electrical stimulation of Abd afferents of the sural nerve are strongly potentiated by intrathecal administration of selective a 2 receptor antagonists [5,9,10]. This effect is mediated by: (i) blockade of descending inhibition; and (ii) enabling or revealing a descending facilitation that is mediated in part by a 1-adrenoceptors [9]. As electrical stimulation is acknowledged be a non-physiological method of afferent activation, we have now investigated the extent to which tonic adrenergic control is exerted over re¯exes evoked by adequate stimulation of cutaneous mechanoreceptors at the heel, which is within the receptive ®eld for the sural nerve. * Corresponding author. Tel.: 144-115-951 6307/6316; fax: 144115-951-6302. E-mail address: [email protected] (R.W. Clarke). 1 Present address: Organon Laboratories, Newhouse, Lanarkshire ML1 5SH, UK.
Experiments were performed in accordance with the UK Animals (Scienti®c Procedures) Act of 1986. Eight rabbits (mixed strains and sex, weight range 2.3±3.0 kg) were anaesthetised with methohexitone sodium (Brietal, Eli Lilly, average dose 22 mg kg 21 i.v.) and the trachea cannulated. Anaesthesia was maintained by halothane (2±3.5%) delivered in O2:N2O (30:70). Arterial blood pressure was recorded via a cannula in the left carotid artery. The spinal cord was exposed at L1 and a cannula (outside diameter 0.61 mm) was inserted beneath the dura so that its tip lay near the lumbar enlargement. All animals were then decerebrated by suction to the pre-collicular level, paralyzed with gallamine triethiodide (4 mg kg 21 initially, supplemented as required) and ventilated on room air supplemented with oxygen. The left hind limb was clamped securely and the sciatic nerve and its branches exposed. The nerves to MG muscle were cut, desheathed, and the central end placed over paired platinum recording electrodes. A pinching device driven by compressed air (J. Brown, Edinburgh) [1] was used to apply regulated pinches to the skin over the heel. A choke valve was used to vary the pressure applied to a piston driving the pinching jaws, and the force applied was measured from a strain gauge mounted on the static arm of the pincher. Four pressures
0304-3940/01/$ - see front matter q 2001 Published by Elsevier Science Ireland Ltd. PII: S03 04 - 394 0( 0 1) 02 00 5- 5
190
R.W. Clarke et al. / Neuroscience Letters 308 (2001) 189±192
were used: 0.5, 1, 2 and 4 bar, which gave mean pinch forces of 183 ^ 31, 607 ^ 51, 1866 ^ 98 and 4577 ^ 156 mN over the eight experiments. Pinches were of 1 s duration. Applied to human ®ngers, these stimuli gave rise to sensations ranging from light pressure to pain [3]. The sensation generated by the highest intensity stimulus is suggestive of activation of ®ne myelinated nociceptive afferents, although it is not possible to certain exactly which afferents were recruited by any of the stimuli used. Re¯ex responses were recorded from the MG motor nerve at a gain of 10 k and ®ltered between 500 Hz and 2 kHz to eliminate low frequency movement artefacts. The ampli®ed, ®ltered signal was sent to a window discriminator set to capture all clearly observable action potentials from the motor nerves. The output from this device was delivered to a ratemeter, the smoothed output of which was digitally sampled at 1 kHz, and integrated by computer to give a measure of the size of the re¯ex output in spikes. The four stimuli were each applied once every 4 min. Background activity rate (i.e. that observed in the absence of a stimulus) was also recorded and subtracted from the evoked re¯exes, taking into account the duration of the response. Idazoxan hydrochloride (Research Biochemicals) was dissolved in Ringer's solution and given intrathecally in doses of 1, 5, 25 and 125 mg (see Ref. [9]), at intervals of 24 min, giving a cumulative total dose of 156 mg. Prazosin hydrochloride (P®zer UK) was solubilized in DMF and diluted in 5% d-glucose solution. It was administered as single i.th. dose of 200 mg [9] 24 min after the ®nal dose of idazoxan. The non-selective opioid receptor antagonist (2)-quadazocine (Sterling Winthrop, [3]) was dissolved in 5% d-glucose solution and given in a single i.th. dose of 25 mg, 24 min after prazosin in 7 rabbits. Drugs were injected in volumes not exceeding 200 ml and were ¯ushed in with 60 ml Ringer's solution. The animals were spinalised at L1 24 min after quadazocine, and data collected for a further 30 min. There was little evidence of spinal shock in these experiments, which in our experience is a rather transient phenomenon in rabbit [9,10,14]. Data are expressed as means ^ SEM. The vehicles have been used before and have no effects on re¯exes when administered alone [9]. In the control state, only pinches at 4577 mN induced reliable re¯ex responses in the MG motor nerve, with a mean amplitude of 959 ^ 220 spikes (n 8), a latency of 335 ^ 96 ms from the onset of the pinch and a duration of 2.24 ^ 0.45 s (Fig. 1). Smaller stimuli induced measurable responses in some animals, but overall responses were not signi®cantly different from 0 (Single sample t-tests, P . 0:05). Idazoxan induced dose-related increases in MG re¯exes to 607, 1866 and 4577 mN pinches (repeated measures ANOVA, P , 0:007, Figs. 1 and 2), with the lowest cumulative doses inducing a signi®cant effect being 156, 1 and 6 mg for the three pinch forces, respectively (Tukey±Kramer post-tests, P , 0:01). The adrenoceptor antagonist also decreased latencies (to a mean of 88 ^ 22 ms for 4577
Fig. 1. Re¯exes evoked in the MG muscle nerve by pinching at the heel using forces of 183, 607, 1866 and 4577 mN for 1 s in the control state, after intrathecal idazoxan (156 mg cumulative dose), and after spinalisation at L1. The black bars indicate the application of the stimulus. Each trace is a single sweep.
mN stimulus after 156 mg dose, P , 0:04, paired t-test versus control) and increased duration of the response (mean 3.14 ^ 0.44 s for 4577 mN stimulus after 156 mg dose, P , 0:01 versus control, paired t-test). Background activity in MG was unaffected by idazoxan (mean rate in control state 13 ^ 7 spikes s 21, after idazoxan 11 ^ 9 spikes s 21, Fig. 2). Given after idazoxan, prazosin (200 mg i.th.) tended to reduce MG re¯ex responses to 4577 mN stimuli, so that re¯exes evoked by this pinch force were no longer signi®cantly different from the pre-idazoxan control (paired t-test, P , 0:05, Fig. 3). Responses to other stimuli were unaffected. (2)-Quadazocine (25 mg i.th.) had no signi®cant effects on re¯exes to any pinch force (Fig. 3). Spinalisation at L1 led to signi®cant increases in re¯ex responses to 183 and 607 mN stimuli (paired t-tests, P , 0:05, Figs. 1 and 3),
Fig. 2. Effects of intrathecal idazoxan on medial gastrocnemius background activity and re¯exes evoked in the same muscle nerve by pinching the heel at 183, 607, 1866 and 4577 mN. Each point is the mean of eight experiments ^SEM. *Indicates a point signi®cantly different from the pre-idazoxan control (C, Tukey±Kramer test following repeated measures ANOVA, P , 0:05).
R.W. Clarke et al. / Neuroscience Letters 308 (2001) 189±192
no change in the re¯ex evoked by 1866 mN (paired t-test, P . 0:1) and a decrease in responses to the highest pinch force (Figs. 1 and 3, paired t-test, P , 0:05). None of the post-idazoxan treatments had signi®cant effects on background activity, although it did tend to increase after spinalisation (Fig. 3). Mean arterial blood pressure in the control state was 102 ^ 4 mmHg. It was not signi®cantly affected by idazoxan (101 ^ 4 mmHg after the highest dose), prazosin (101 ^ 4 mmHg) or quadazocine (96 ^ 5 mmHg), but was predictably decreased after spinalisation, to a mean of 67 ^ 5 mmHg. The present data show that idazoxan increases MG re¯exes evoked by `natural' mechanical stimulation of the heel in the same way that it does for those elicited by electrical stimulation of the sural nerve [9]. Although some of this effect may be mediated through imidazoline receptors [4], there is no doubt the blockade of spinal a 2-adrenoceptors was the major action [10]. In the rat, it has been reported that a 2-receptor antagonists increase responses of dorsal horn neurones [8] or spinal re¯exes [12] only when in¯ammation is present, suggesting that the preparative surgery might be responsible for the adrenergic tone observed in the present study. However, we have shown that adrenoceptor antagonists are just as potent in intact, anaesthetised rabbits that have been subject to virtually no surgical interference, as they are in decerebrates [14]. Thus, adrenergic control of the heel-MG re¯ex in rabbit is not an epiphenomenon of the preparation. Given after idazoxan, prazosin decreased only MG re¯exes to the highest strength stimulus. Activation of a 1adrenoceptors is known to increase motoneurone excitability [15], and we have shown that part of the re¯ex-augmenting action of a 2-receptor antagonists is to promote or reveal descending adrenergic facilitation in the ventral horn [9]. The present data suggest that this process is only apparent
191
for re¯exes evoked by more intense stimuli. Either nociceptive afferents preferentially activate motoneurones receiving facilitatory adrenergic tone, or the facilitation is itself recruited by stimulation of high threshold inputs. We favour the latter explanation. The failure of quadazocine to increase mechanicallyevoked re¯exes in this preparation is in marked contrast to the powerful facilitation it causes in spinalised preparations [3], but is not entirely surprising as we have already shown that opioidergic tone is dif®cult to demonstrate in non-spinalized rabbits [2]. We ascribe this difference between preparations to activity in descending inhibitory pathways that act in parallel with segmental opioidergic inhibition and which are capable of holding down re¯exes even when opioid effects are blocked. These pathways must be additional to the adrenergic system, as quadazocine failed to alter re¯exes in the presence of idazoxan. The effects of spinal section indicate that, after blockade of a 2, a 1 and opioid receptors, the net effect of descending pathways was inhibitory to re¯exes evoked by low intensity stimuli and facilitatory to responses evoked by the strongest stimulus used. This set of results can be most easily explained if descending facilitation was recruited by the 4577 mN stimulus. It is likely that idazoxan revealed pre-existing descending facilitation rather than activating it. Comparing control state stimulus-response relationships for mechanically-evoked re¯exes in spinalised and non-spinalised rabbits shows that while the threshold for activating re¯exes is higher in the latter, the re¯exes elicited by the highest stimulus are very similar in magnitude at around 1000 spikes in both preparations [3]. This ®nding may be explained by proposing that noxious stimuli activate descending facilitatory in¯uences that counteract the effects of tonic descending inhibition, thus allowing nocifensive re¯exes to proceed. This idea is consistent with the on-cell concept of Fields and co-workers (see Refs. [6,11,13]). Supported by the BBSRC. We are grateful to P®zer UK for the supply of prazosin and to Sterling Winthrop for (2)quadazocine.
Fig. 3. Effects of successive intrathecal administration of idazoxan (IDAZ, 156 mg); prazosin (PRAZ, 200 mg); (2)-quadazocine (QUAD, 25 mg) and spinalisation (SPINAL) on medial gastrocnemius background activity and re¯exes evoked in the same muscle nerve by pinching the heel at 183, 607, 1866 and 4577 mN. Control, IDAZ and PRAZ columns are the averages 1 SEM from eight replicates, for the others n 7. *Signi®cantly greater than pre-idazoxan control (paired t-tests, P , 0:05); ²signi®cantly less than post-idazoxan level (paired t-test, P , 0:05).
[1] Brown, A.C., Headley, P.M. and West, D.C., A device for producing reproducible electronically-timed pinch to provide noxious mechanical input, J. Physiol., 349 (1984) 6 p. [2] Clarke, R.W., Ford, T.W. and Taylor, J.S., Adrenergic and opioidergic modulation of a spinal re¯ex in the rabbit, J. Physiol., 404 (1988) 407±417. [3] Clarke, R.W., Harris, J. and Houghton, A.K., Opioidergic modulation of spinal re¯exes activated by mechanical stimulation of the heel in the decerebrated, spinalized rabbit, Exp. Brain Res., 120 (1998) 18±24. [4] Clarke, R.W., Harris, J. and Ogilvie, J., Imidazoline I2-receptors and spinal re¯exes in the decerebrated rabbit, Neuropharmacology, 39 (2000) 1904±1912. [5] Clarke, R.W., Parry-Baggott, C., Houghton, A.K. and Ogilvie, J., The involvement of bulbo-spinal pathways in fentanyl-
192
[6] [7] [8]
[9] [10]
R.W. Clarke et al. / Neuroscience Letters 308 (2001) 189±192 induced inhibition of spinal withdrawal re¯exes in the decerebrated rabbit, Pain, 78 (1998) 197±207. Fields, H.L., Heinricher, M.M. and Mason, P., Neurotransmitters in nociceptive modulatory circuits, Annu. Rev. Neurosci., 14 (1991) 219±245. Furst, S., Transmitters involved in antinociception in the spinal cord, Brain Res. Bull., 48 (1999) 129±141. Green, G.M., Lyons, L. and Dickenson, A.H., a2-adrenoceptor antagonists enhance responses of dorsal horn neurones to formalin induced in¯ammation, Eur. J. Pharmacol., 347 (1998) 201±204. Harris, J. and Clarke, R.W., An analysis of adrenergic in¯uences on the sural gastrocnemius re¯ex of the decerebrated rabbit, Exp. Brain Res., 92 (1992) 310±317. Harris, J. and Clarke, R.W., Motor and cardiovascular effects of selective a2-adrenoceptor antagonists in the decerebrated rabbit, Eur. J. Pharmacol., 237 (1993) 323± 328.
[11] Heinricher, M.M., Organizational characteristics of supraspinally mediated responses to nociceptive input, In T.L. Yaksh, C. Lynch, W.M. Zapol, M. Maze, J.F. Biebuyck and L.J. Saidman (Eds.), Anesthesia: Biologic Foundations, Lippincott-Raven, New York, 1997, pp. 643±661. [12] Herrero, J.F. and Solano, R.E., The antinociceptive effect of the m-opioid fentanyl is reduced in the presence of the a2adrenergic antagonist idazoxan in in¯ammation, Brain Res., 840 (1999) 106±114. [13] Mason, P., Central mechanisms of pain modulation, Curr. Op. Neurobiol., 9 (1999) 436±441. [14] Ogilvie, J., Simpson, D.A.A. and Clarke, R.W., Tonic adrenergic and serotonergic inhibition of a withdrawal re¯ex in rabbits subjected to different levels of surgical preparation, Neuroscience, 89 (1999) 1247±1258. [15] Rekling, J.C., Funk, G.D., Bayliss, D.A., Dong, X.W. and Feldman, J.L., Synaptic control of motoneuronal excitability, Physiol. Rev., 80 (2000) 767±852.
www.elsevier.com/locate/neulet
Endogenous adrenergic control of re¯exes evoked by mechanical stimulation of the heel in the decerebrated rabbit R.W. Clarke*, J. Harris, A.K. Houghton 1 Division of Animal Physiology, School of Biosciences, Sutton Bonington Campus, Loughborough LE12 5RD, UK Received 2 May 2001; received in revised form 5 June 2001; accepted 11 June 2001
Abstract In decerebrated rabbits, re¯exes were evoked in medial gastrocnemius motoneurones by mechanical stimulation of the heel, using four pinch strengths from 183 (innocuous) to 4577 (noxious) mN. The a 2-adrenoceptor antagonist idazoxan (1±156 mg intrathecal (i.th.)) signi®cantly increased responses to pinch strengths of 607 mN and above. Subsequent administration of the a 1 adrenoceptor selective antagonist prazosin (200 mg i.th.) decreased re¯exes to 4577 mN pinches but had no other signi®cant effects. The opioid antagonist (2)-quadazocine (25 mg i.th.) caused no further changes in re¯exes. Spinal section at L1 in the presence of this drug combination enhanced gastrocnemius responses to 183 and 607 mN stimuli, had no effect on re¯exes to 1866 mN and signi®cantly decreased responses to 4577 mN pinches. These data con®rm that re¯exes evoked by `natural' stimulation of heel mechanoreceptors are subject to powerful tonic descending inhibition mediated by a 2-adrenoceptors. For the highest strength stimulus, the results of a 2 blockade involved enabling of descending facilitation as well as reduction of descending inhibition. q 2001 Published by Elsevier Science Ireland Ltd. Keywords: Withdrawal re¯ex; Noradrenaline; Spinal cord; Nociception; Descending inhibition; Descending facilitation
Noradrenaline, acting through a 2-adrenoceptors, is thought to be an important mediator of descending inhibitory control of the spinal cord [7], but there are few convincing examples of tonic inhibition that can be relieved by adrenoceptor antagonists. In decerebrated rabbits, re¯exes evoked in the ankle extensor medial gastrocnemius (MG) by electrical stimulation of Abd afferents of the sural nerve are strongly potentiated by intrathecal administration of selective a 2 receptor antagonists [5,9,10]. This effect is mediated by: (i) blockade of descending inhibition; and (ii) enabling or revealing a descending facilitation that is mediated in part by a 1-adrenoceptors [9]. As electrical stimulation is acknowledged be a non-physiological method of afferent activation, we have now investigated the extent to which tonic adrenergic control is exerted over re¯exes evoked by adequate stimulation of cutaneous mechanoreceptors at the heel, which is within the receptive ®eld for the sural nerve. * Corresponding author. Tel.: 144-115-951 6307/6316; fax: 144115-951-6302. E-mail address: [email protected] (R.W. Clarke). 1 Present address: Organon Laboratories, Newhouse, Lanarkshire ML1 5SH, UK.
Experiments were performed in accordance with the UK Animals (Scienti®c Procedures) Act of 1986. Eight rabbits (mixed strains and sex, weight range 2.3±3.0 kg) were anaesthetised with methohexitone sodium (Brietal, Eli Lilly, average dose 22 mg kg 21 i.v.) and the trachea cannulated. Anaesthesia was maintained by halothane (2±3.5%) delivered in O2:N2O (30:70). Arterial blood pressure was recorded via a cannula in the left carotid artery. The spinal cord was exposed at L1 and a cannula (outside diameter 0.61 mm) was inserted beneath the dura so that its tip lay near the lumbar enlargement. All animals were then decerebrated by suction to the pre-collicular level, paralyzed with gallamine triethiodide (4 mg kg 21 initially, supplemented as required) and ventilated on room air supplemented with oxygen. The left hind limb was clamped securely and the sciatic nerve and its branches exposed. The nerves to MG muscle were cut, desheathed, and the central end placed over paired platinum recording electrodes. A pinching device driven by compressed air (J. Brown, Edinburgh) [1] was used to apply regulated pinches to the skin over the heel. A choke valve was used to vary the pressure applied to a piston driving the pinching jaws, and the force applied was measured from a strain gauge mounted on the static arm of the pincher. Four pressures
0304-3940/01/$ - see front matter q 2001 Published by Elsevier Science Ireland Ltd. PII: S03 04 - 394 0( 0 1) 02 00 5- 5
190
R.W. Clarke et al. / Neuroscience Letters 308 (2001) 189±192
were used: 0.5, 1, 2 and 4 bar, which gave mean pinch forces of 183 ^ 31, 607 ^ 51, 1866 ^ 98 and 4577 ^ 156 mN over the eight experiments. Pinches were of 1 s duration. Applied to human ®ngers, these stimuli gave rise to sensations ranging from light pressure to pain [3]. The sensation generated by the highest intensity stimulus is suggestive of activation of ®ne myelinated nociceptive afferents, although it is not possible to certain exactly which afferents were recruited by any of the stimuli used. Re¯ex responses were recorded from the MG motor nerve at a gain of 10 k and ®ltered between 500 Hz and 2 kHz to eliminate low frequency movement artefacts. The ampli®ed, ®ltered signal was sent to a window discriminator set to capture all clearly observable action potentials from the motor nerves. The output from this device was delivered to a ratemeter, the smoothed output of which was digitally sampled at 1 kHz, and integrated by computer to give a measure of the size of the re¯ex output in spikes. The four stimuli were each applied once every 4 min. Background activity rate (i.e. that observed in the absence of a stimulus) was also recorded and subtracted from the evoked re¯exes, taking into account the duration of the response. Idazoxan hydrochloride (Research Biochemicals) was dissolved in Ringer's solution and given intrathecally in doses of 1, 5, 25 and 125 mg (see Ref. [9]), at intervals of 24 min, giving a cumulative total dose of 156 mg. Prazosin hydrochloride (P®zer UK) was solubilized in DMF and diluted in 5% d-glucose solution. It was administered as single i.th. dose of 200 mg [9] 24 min after the ®nal dose of idazoxan. The non-selective opioid receptor antagonist (2)-quadazocine (Sterling Winthrop, [3]) was dissolved in 5% d-glucose solution and given in a single i.th. dose of 25 mg, 24 min after prazosin in 7 rabbits. Drugs were injected in volumes not exceeding 200 ml and were ¯ushed in with 60 ml Ringer's solution. The animals were spinalised at L1 24 min after quadazocine, and data collected for a further 30 min. There was little evidence of spinal shock in these experiments, which in our experience is a rather transient phenomenon in rabbit [9,10,14]. Data are expressed as means ^ SEM. The vehicles have been used before and have no effects on re¯exes when administered alone [9]. In the control state, only pinches at 4577 mN induced reliable re¯ex responses in the MG motor nerve, with a mean amplitude of 959 ^ 220 spikes (n 8), a latency of 335 ^ 96 ms from the onset of the pinch and a duration of 2.24 ^ 0.45 s (Fig. 1). Smaller stimuli induced measurable responses in some animals, but overall responses were not signi®cantly different from 0 (Single sample t-tests, P . 0:05). Idazoxan induced dose-related increases in MG re¯exes to 607, 1866 and 4577 mN pinches (repeated measures ANOVA, P , 0:007, Figs. 1 and 2), with the lowest cumulative doses inducing a signi®cant effect being 156, 1 and 6 mg for the three pinch forces, respectively (Tukey±Kramer post-tests, P , 0:01). The adrenoceptor antagonist also decreased latencies (to a mean of 88 ^ 22 ms for 4577
Fig. 1. Re¯exes evoked in the MG muscle nerve by pinching at the heel using forces of 183, 607, 1866 and 4577 mN for 1 s in the control state, after intrathecal idazoxan (156 mg cumulative dose), and after spinalisation at L1. The black bars indicate the application of the stimulus. Each trace is a single sweep.
mN stimulus after 156 mg dose, P , 0:04, paired t-test versus control) and increased duration of the response (mean 3.14 ^ 0.44 s for 4577 mN stimulus after 156 mg dose, P , 0:01 versus control, paired t-test). Background activity in MG was unaffected by idazoxan (mean rate in control state 13 ^ 7 spikes s 21, after idazoxan 11 ^ 9 spikes s 21, Fig. 2). Given after idazoxan, prazosin (200 mg i.th.) tended to reduce MG re¯ex responses to 4577 mN stimuli, so that re¯exes evoked by this pinch force were no longer signi®cantly different from the pre-idazoxan control (paired t-test, P , 0:05, Fig. 3). Responses to other stimuli were unaffected. (2)-Quadazocine (25 mg i.th.) had no signi®cant effects on re¯exes to any pinch force (Fig. 3). Spinalisation at L1 led to signi®cant increases in re¯ex responses to 183 and 607 mN stimuli (paired t-tests, P , 0:05, Figs. 1 and 3),
Fig. 2. Effects of intrathecal idazoxan on medial gastrocnemius background activity and re¯exes evoked in the same muscle nerve by pinching the heel at 183, 607, 1866 and 4577 mN. Each point is the mean of eight experiments ^SEM. *Indicates a point signi®cantly different from the pre-idazoxan control (C, Tukey±Kramer test following repeated measures ANOVA, P , 0:05).
R.W. Clarke et al. / Neuroscience Letters 308 (2001) 189±192
no change in the re¯ex evoked by 1866 mN (paired t-test, P . 0:1) and a decrease in responses to the highest pinch force (Figs. 1 and 3, paired t-test, P , 0:05). None of the post-idazoxan treatments had signi®cant effects on background activity, although it did tend to increase after spinalisation (Fig. 3). Mean arterial blood pressure in the control state was 102 ^ 4 mmHg. It was not signi®cantly affected by idazoxan (101 ^ 4 mmHg after the highest dose), prazosin (101 ^ 4 mmHg) or quadazocine (96 ^ 5 mmHg), but was predictably decreased after spinalisation, to a mean of 67 ^ 5 mmHg. The present data show that idazoxan increases MG re¯exes evoked by `natural' mechanical stimulation of the heel in the same way that it does for those elicited by electrical stimulation of the sural nerve [9]. Although some of this effect may be mediated through imidazoline receptors [4], there is no doubt the blockade of spinal a 2-adrenoceptors was the major action [10]. In the rat, it has been reported that a 2-receptor antagonists increase responses of dorsal horn neurones [8] or spinal re¯exes [12] only when in¯ammation is present, suggesting that the preparative surgery might be responsible for the adrenergic tone observed in the present study. However, we have shown that adrenoceptor antagonists are just as potent in intact, anaesthetised rabbits that have been subject to virtually no surgical interference, as they are in decerebrates [14]. Thus, adrenergic control of the heel-MG re¯ex in rabbit is not an epiphenomenon of the preparation. Given after idazoxan, prazosin decreased only MG re¯exes to the highest strength stimulus. Activation of a 1adrenoceptors is known to increase motoneurone excitability [15], and we have shown that part of the re¯ex-augmenting action of a 2-receptor antagonists is to promote or reveal descending adrenergic facilitation in the ventral horn [9]. The present data suggest that this process is only apparent
191
for re¯exes evoked by more intense stimuli. Either nociceptive afferents preferentially activate motoneurones receiving facilitatory adrenergic tone, or the facilitation is itself recruited by stimulation of high threshold inputs. We favour the latter explanation. The failure of quadazocine to increase mechanicallyevoked re¯exes in this preparation is in marked contrast to the powerful facilitation it causes in spinalised preparations [3], but is not entirely surprising as we have already shown that opioidergic tone is dif®cult to demonstrate in non-spinalized rabbits [2]. We ascribe this difference between preparations to activity in descending inhibitory pathways that act in parallel with segmental opioidergic inhibition and which are capable of holding down re¯exes even when opioid effects are blocked. These pathways must be additional to the adrenergic system, as quadazocine failed to alter re¯exes in the presence of idazoxan. The effects of spinal section indicate that, after blockade of a 2, a 1 and opioid receptors, the net effect of descending pathways was inhibitory to re¯exes evoked by low intensity stimuli and facilitatory to responses evoked by the strongest stimulus used. This set of results can be most easily explained if descending facilitation was recruited by the 4577 mN stimulus. It is likely that idazoxan revealed pre-existing descending facilitation rather than activating it. Comparing control state stimulus-response relationships for mechanically-evoked re¯exes in spinalised and non-spinalised rabbits shows that while the threshold for activating re¯exes is higher in the latter, the re¯exes elicited by the highest stimulus are very similar in magnitude at around 1000 spikes in both preparations [3]. This ®nding may be explained by proposing that noxious stimuli activate descending facilitatory in¯uences that counteract the effects of tonic descending inhibition, thus allowing nocifensive re¯exes to proceed. This idea is consistent with the on-cell concept of Fields and co-workers (see Refs. [6,11,13]). Supported by the BBSRC. We are grateful to P®zer UK for the supply of prazosin and to Sterling Winthrop for (2)quadazocine.
Fig. 3. Effects of successive intrathecal administration of idazoxan (IDAZ, 156 mg); prazosin (PRAZ, 200 mg); (2)-quadazocine (QUAD, 25 mg) and spinalisation (SPINAL) on medial gastrocnemius background activity and re¯exes evoked in the same muscle nerve by pinching the heel at 183, 607, 1866 and 4577 mN. Control, IDAZ and PRAZ columns are the averages 1 SEM from eight replicates, for the others n 7. *Signi®cantly greater than pre-idazoxan control (paired t-tests, P , 0:05); ²signi®cantly less than post-idazoxan level (paired t-test, P , 0:05).
[1] Brown, A.C., Headley, P.M. and West, D.C., A device for producing reproducible electronically-timed pinch to provide noxious mechanical input, J. Physiol., 349 (1984) 6 p. [2] Clarke, R.W., Ford, T.W. and Taylor, J.S., Adrenergic and opioidergic modulation of a spinal re¯ex in the rabbit, J. Physiol., 404 (1988) 407±417. [3] Clarke, R.W., Harris, J. and Houghton, A.K., Opioidergic modulation of spinal re¯exes activated by mechanical stimulation of the heel in the decerebrated, spinalized rabbit, Exp. Brain Res., 120 (1998) 18±24. [4] Clarke, R.W., Harris, J. and Ogilvie, J., Imidazoline I2-receptors and spinal re¯exes in the decerebrated rabbit, Neuropharmacology, 39 (2000) 1904±1912. [5] Clarke, R.W., Parry-Baggott, C., Houghton, A.K. and Ogilvie, J., The involvement of bulbo-spinal pathways in fentanyl-
192
[6] [7] [8]
[9] [10]
R.W. Clarke et al. / Neuroscience Letters 308 (2001) 189±192 induced inhibition of spinal withdrawal re¯exes in the decerebrated rabbit, Pain, 78 (1998) 197±207. Fields, H.L., Heinricher, M.M. and Mason, P., Neurotransmitters in nociceptive modulatory circuits, Annu. Rev. Neurosci., 14 (1991) 219±245. Furst, S., Transmitters involved in antinociception in the spinal cord, Brain Res. Bull., 48 (1999) 129±141. Green, G.M., Lyons, L. and Dickenson, A.H., a2-adrenoceptor antagonists enhance responses of dorsal horn neurones to formalin induced in¯ammation, Eur. J. Pharmacol., 347 (1998) 201±204. Harris, J. and Clarke, R.W., An analysis of adrenergic in¯uences on the sural gastrocnemius re¯ex of the decerebrated rabbit, Exp. Brain Res., 92 (1992) 310±317. Harris, J. and Clarke, R.W., Motor and cardiovascular effects of selective a2-adrenoceptor antagonists in the decerebrated rabbit, Eur. J. Pharmacol., 237 (1993) 323± 328.
[11] Heinricher, M.M., Organizational characteristics of supraspinally mediated responses to nociceptive input, In T.L. Yaksh, C. Lynch, W.M. Zapol, M. Maze, J.F. Biebuyck and L.J. Saidman (Eds.), Anesthesia: Biologic Foundations, Lippincott-Raven, New York, 1997, pp. 643±661. [12] Herrero, J.F. and Solano, R.E., The antinociceptive effect of the m-opioid fentanyl is reduced in the presence of the a2adrenergic antagonist idazoxan in in¯ammation, Brain Res., 840 (1999) 106±114. [13] Mason, P., Central mechanisms of pain modulation, Curr. Op. Neurobiol., 9 (1999) 436±441. [14] Ogilvie, J., Simpson, D.A.A. and Clarke, R.W., Tonic adrenergic and serotonergic inhibition of a withdrawal re¯ex in rabbits subjected to different levels of surgical preparation, Neuroscience, 89 (1999) 1247±1258. [15] Rekling, J.C., Funk, G.D., Bayliss, D.A., Dong, X.W. and Feldman, J.L., Synaptic control of motoneuronal excitability, Physiol. Rev., 80 (2000) 767±852.