Selective inhibition of responses of feline dorsal horn neurones to noxious cutaneous stimuli by tizanidine (DS103-282) and noradrenaline: Involvement ofα2-adrenoceptors

0306-4522/85 $3.00 + 0.00

Neuroscience Vol. 16, No. 3, pp. 673-682, 1985 Printed in Great Britain

Pergamon Press Ltd IBRO

SELECTIVE INHIBITION OF RESPONSES OF FELINE DORSAL HORN NEURONES TO NOXIOUS CUTANEOUS STIMULI BY TIZANIDINE (DS103-282) AND NORADRENALINE: INVOLVEMENT OF a,-ADRENOCEPTORS J. DA~IFS and J. E. QUINLAN M.R.C. Neuropharmacology Group, Department of Pharmacology, The School of Pharmacy, 29/39 Brunswick Square, London WClA 1AX

Abstract-The effects of tizanidine (DSl03-282) were compared with those of noradrenaline and other adrenoceptor agonists on responses of laminae IV and V neurones in the lumbar dorsal horn to noxious and innocuous cutaneous stimuli in the anaesthetized cat. Tiinidine, noradrenaline and the a+ceptor agonist, clonidine, depressed spontaneous activity and responses to noxious, but not those to innocuous, stimuli when administered iontophoretically, either near the recording site in laminae IV-V, or into

laminae II-III, i.e. 300-9OOym dorsal to the recording site. Iontophoretic ejection of dopamine, the B-agonist isoprenaline and the a,-agonists phenylephrine and amidephrine had no effect at either site., or only relatively weak and sometimes non-selective depressant actions on neuronal responses to cutaneous stimuli. The preferential depressant actions of tizanidine, noradrenaline and clonidine were antagonized by the selective a,-antagonist RX781094 administered iontophoretically at the same site as the agonists, and by intravenously administered yohimbine. In contrast, the q-antagonists, prazosin and WB4101, the /3-antagonist, sotalol and opiate antagonist, naloxone did not alter the depressant actions of these agonists on laminae IV and V neurones.

These findings indicate that the selective inhibitory effect of tizanidine and noradrenaline on responses of laminae IV and V neurones to noxious peripheral stimuli are mediated at $-adrenoceptors situated in either laminae IV and V or laminae II-III. The possible physiological relevance of these receptors is

Iontophoretically administered tizanidine (DS 103282) inhibits the excitatory responses of laminae IV and V feline spinal neurones evoked by noxious, but not innocuous, cutaneous stimuli. This observation lead to the proposal that tizanidine may have an analgesic action in addition to its established centrally-mediated muscle relaxant effect.“vz6 Compounds such as morphine, S-hydroxytryptamine and noradrenaline have also been shown to selectively depress responses of spinal laminae IV and V neurones to noxious cutaneous stimuli.l*s*“*‘* Morphine and S-hydroxytryptamine exert selective depressant effects on laminae IV and V neurones only when they are administered dorsal to the cell bodies of these neurones, i.e. into laminae II-III.‘i~l* However, the selective action of tizanidine is obtained at both sites, i.e. when it is ejected into laminae IV-V, or laminae II-III.’ The opiate antagonist naloxone blocks the depressant actions of morphine,” but not that induced by tizanidine,’ thus it seems likely that the site of action of tizanidine is probably different from that of either morphine or Shydroxytryptamine. On the other hand, there are certain similarities between the effects of noradrenaline and tizanidine which suggests these two substances may act at a common site. Firstly, the selective depressant

actions of both agents on responses of laminae IV and V neurones to noxious stimuli is apparent when they are administered either near these neurones or into laminae 11-111.‘~~*‘~ Secondly, noradrenaline and tizanidine, but not morphine or 5-hydroxytrytamine, depress the K-evoked release of Dj3HJaspartate from spinal cord slices,’ and thirdly, tizanidine has structural features in common with the adrenoceptor agonist, clonidine. In view of this the present experiments were designed to examine the possibility that tizanidine and noradrenaline may act at a similar site. To this end, the effects of tizanidine have been compared with those of noradrenaline and a range of other adrenoceptor agonists on responses evoked by noxious and innocuous peripheral stimuli in laminae IV and V neurones and the ability of various adrenoceptor antagonists to modify their actions have been determined. Some of these results have already appeared in a preliminary form.*

673

EXPERIMENTAL PROCEDURES Experiments were performed on cats of either sex anaesthetized with either a-chloralose (50 mg kg-’ IV), following induction with halothane, or halothane (1% in 4). Blood pressure was continuously monitored via an indwelling

674

J.

DAVIES

and J. E.

carotid cannula and body temperature was maintained between 37-38°C by means of a thermostatically controlled heating blanket placed beneath the animal. Experiments were terminated if the systolic pressure fell below 95 mmHg. The lumbar spinal cord was exposed by laminectomy, transected at Ll and the L7 and Sl ventral roots were sectioned. Single unit activity in laminae IV-V of the L7-Sl segment of the cord was recorded via either the central barrel (4 M NaCl) of a multibarrelled microelectrode or a single glass microelectrode (4 M NaCl or 2% pontamine sky blue in 1.2M sodium acetate). The location of the recording electrode in laminae IV or V was determined by noting its depth below the surface of the spinal cord and/or subsequently verifying the location of iontophoresed pontamine blue dye marks in histological sections. Neurones were excited by noxious (radiant heat 4550°C with feedback control) and innocuous (electronically controlled air jet) stimuli directed at the ipsilateral foot.’ Amplification and recording of neuronal action potentials were performed as described earlier.4 The total number of action potentials in a stimulus-evoked response was counted using a gated electronic counter. These counts were corrected for background firing rate and used to determine the effects of drugs on evoked responses by expressing the count as a percentage of the control response before drug administration, Changes in the evoked response in excess of 15% following drug administration were regarded as being significant. In some experiments neurones were excited by electrical stimulation of the central ends of the cut tibia], sural or peroneal nerve at intensities sufficient to excite unmyelinated (C) fibres (1 ms, 0.5 Hz, lo-40 V).’ Action potentials evoked by these stimuli were passed through a computer after suitable amplification in order to compile peristimulus time histograms of the synaptic response. Each peripheral barrel of the multibarrelled electrode was filled with one of the following drugs, which were used in various combinations in different electrodes and were administered iontophoretically: tizanidine HCl (Sandoz Ltd) (0.1 M, pH 5.0), clonidine HCl (Boeringer Ingelheim) (0.1 M, pH 5.0), I-noradrenaline bitartrate (Sigma) ladrenaline bitartrate (Sigma), I-phenylephrine HCI (Sigma), isoprenaline HCI (Sigma), amidephrine mesylate (Mead Johnson), RX781094 (Idazoxan, Reckitt & Colman), WB4101 (WB Pharmaceuticals), sotalol HCl (Mead Johndopamine HCl (Sigma), piperoxane HCl son), @hone-Poulenc) (all at 0.2 M, pH 4.0-5.5) phentolamine mesylate (Ciba) (0.2 M, either pH 5.0 or acidified to pH 3.5 with HCl), prazosin HCl (Pfizer) (0.01 M, pH 5.0), (-)-baclofen (Ciba) (0.005 M in 0.165 M NaCl, pH 3.5), GABA (Sigma) (0.5 M, pH 3.5), naloxone HCl (Endo) (0.1 M, pH 5.0), morphine SO, (Boots) (0.07 M, pH 4.5) and yohimbine HCl (Sigma) (0.012 M, pH 4.5). One barrel of each electrode contained pontamine sky blue in Na acetate for current balancing and depositing dye to mark the recordings and drug ejection site at the termination of the experiment. Drugs were either administered from peripheral barrels of the recording microelectrode situated in laminae IV and V or from a separate multibarrelled electrode positioned in laminae II-III using a second micromanipulator.” In the latter case, the vertical displacement of the recording and drug ejecting micropipettes was 300-900 pm. Tizanidine, naloxone and yohimbine were also administered intravenously via the cannulated brachial vein in some experiments, the dose of these substances being expressed in terms of the salt. RESULTS The results described in this paper were obtaiwd from 47 neurones located in laminae IV or V of the dorsal horn. All neurones could be excited by innoc-

QUINLAN

uous as well as noxious peripheral stimuli and many exhibited spontaneous activity. Responses to innocuous stimuli were sometimes variable and the effects of the different drugs studied are only reported on those neurones where such responses were consistent. The data from animals anaesthetized with halothane was qualitatively similar to that from GIchloralose anaesthetized cats and hence the results from these two groups have been combined. Esfects of tizanidine and adrenoceptor

agonists

In keeping with previous observations,’ iontophoretic administration of tizanidine near the record-

ing site in laminae IV-V, or more dorsally into laminae II-III, preferentially and markedly reduced neuronal excitation induced by noxious stimuli, and also often depressed the spontaneous firing rate. The maximum depression of neuronal firing was usually apparent several minutes after terminating the ejection of tizanidine. Recovery occurred 15-40 min, and 30 to sometimes more than 60min, after laminae IV-V and II-III ejections, respectively. In tests on the same neurones, noradrenaline and the selective a,-agonist, clonidine, had qualitatively similar effects to those of tizanidine ejected at the same site (e.g. Figs 1 and 2 and Table 1). None of the three substances depressed responses of laminae IV and V neurones to innocuous stimuli, but all reduced neuronal responses to noxious stimuli and often also inhibited spontaneous activity (Table 1). Higher ejecting currents for all three agents were generally required to produce similar depressions of neuronal responses when ejected into laminae II-III, compared with those needed when ejected near the recorded cells. The duration of the depressant action of noradrenaline was much shorter than that for either tizanidine or clonidine (Table 1). Indeed, the degree and duration of the depressant action of clonidine and tizanidine were remarkably similar (Table 1). Moreover, as seen with tizanidine, responses to noxious stimuli continued to decrease for several minutes after terminating the clonidine ejection (Fig. 1). In contrast, the #l-agonist isoprenaline (20-85 nA, mean 54 nA, five cells) had no significant effect on neuronal responses to cutaneous stimuli and the a,-agonists amidephrine and phenylephrine were ineffective or much less effective in depressing these responses. Relatively high currents (80-100 nA) of amidephrine depressed neuronal responses to noxious stimuli in only one of five neurones when ejected at the recording site and in one of three neurones when ejected into laminae II-III. Phenylephrine (20-90 nA, mean 60 nA) reduced the responses to noxious stimuli of six cells, in four of these responses to innocuous stimuli were also reduced. However, in a further three cells excitations induced by both types of cutaneous stimuli were unaffected by phenylephrine. In tests on five neurones, dopamine (20-80 nA, mean 55 nA) had no effect on responses to cutaneous stimuli, whereas adrenaline (lo-30 nA mean 20 nA) produced selec-

Effects of tizanidine and noradrenaline

615

at spinal a,-receptors

Norodranoline 20nA

m

l&t

D Air

0

-

I

Clonidine

I

I

I

0

0

0

min

0

(I

I

cl

C

0

30nA

Tiranidine

30nA I

loo

Spikes set-’

I

I 0

I 0

m

I

a

0

m

I 0

0

m

m a

0 1

Min. Laminoe 11-111

Fig. 1. Ratemeter record showing the effects of noradrenaline 20 nA, clonidine 30 nA and tizanidine 30 nA on the responses of the same lamina V neurone induced by noxious (heat 46.5”C) and innocuous (air-jet) stimuli. All drugs were ejected into laminae II-III. All three drugs depressed responses of the neurone to noxious stimuli and spontaneous activity but not responses to innocuous stimuli. Note the slower onset of action of tizanidine and clonidine compared to that of noradrenaline and the more prolonged duration of action of the former two agents.

tive depressant effects comparable to those observed with noradrenaline. None of the agents tested had any observable excitatory action on laminae IV and V neurones

when they were administered into laminae II-III. However, phenylephrine and noradrenaline increased the background firing rate of one and three neurones respectively when ejected near these cells. Responses

Table 1. Comparison between the effect of iontophoretically administered tizanidine, noradrenaline and clonidine on resnonses to noxious cutaneous stimuli and snontaneous activity of the same laminae IV-V neurones Effect on: Spontaneous activity

Responses to noxious heat Ejection site

Laminae IV-V

Laminae II-III

No. dep/ Agonist [mean f SD current (nA) No. and duration (min)] tested Tizanidine (13.3 f 5.0nA for 3.9 f 1.9 min) Noradrenaline (12.7 f 7.2 nA for 3.5 f 0.9 min) Clonidine (16.5 f 7.2 nA for 4.8 f 2.2 min)

% Depression*

Recovery time (min)*

No. dep/ No. tested

10/10

50-94 (78)

20-45 (31)

617

50-100 (80)

2cMo (3 1)

lO/lO

&loo

(75)

2-7 (5)

717

25-100 (76)

2-7 (4)

IO/10

@loo

(70)

16-70 (38)

617

40-100 (70)

16-55 (33)

40-100 (68)

25-80 (50)

10112

5CklOO(67)

25-80 (67)

3(r93 (63)

3-31(13)

10112

2Ck80 (50)

3-15 (10)

53-100 (73)

25-100 (58)

lo/12

5&100 (62)

35-70 (51)

Tizanidine (33.0 * 1.9 nA 13/13 for 6.7 f 0.8 min) Noradrenaline (33.1 f 13/13 12.9 nA for 7.2 + 0.8 min) Clonidine (33.3 f 11.5 nA 13/13 for 6.9 f 0.8 min)

% Depression*

Recovery time (min)’

*The values in these columns refer to, the range and mean percentage depression (cells not depressed being excluded) and the range and mean recovery time (min). N.B. None of the agonists depressed responses of neurones to innocuous (air-jet) stimuli.

J. DAVIES and J. E. QUINLAN

676

RX 781094 30nA 70 -

Noradrenaline

Noradrenaline

2OnA -c.

60 -

20nA

1

I

50 40 r 30 20 IO O-

I

1

1

1

0

10

20

30

40

RX78 W@4

7Or A....4 Spontaneous

60. 50 40 30 20 10 0,

mr 10 -

60

I

50 40 30 20 10 QL

I

0

1

I

20

40

TIM

E (MIN)

I

60

I

SO

Effects of tizanidine and noradrenaline at spinal a,-receptors

671

Table 2. Effects of adrenoceptor antagonists on the depressant actions of tizanidine, noradrenaline and clonidine on responses of laminae IV and V neurones to cutaneous stimuli and spontaneous activity when iontophoresed into laminae IV-V or II-III Effect on depressant action of:

Antagonistt Sotalol (2CMOO nA, 10-25 min) WB4101 (20-60 nA, 5-20 min) (4&50 nA, 1l-20 mitt) Prazosin (6tMO nA, 4-12 min) Phentolamine pH 5.5 (l&20 nA, 4-12 mitt) Phentolamine pH 3.5 30 nA, 9-24 min) (20-80 nA, 1141 mitt) Yohimbine (20 nA, 5-7 min) $(40-120 nA, 11-24 mitt) Piperozane (15-30 nA, 7-16 in) $(2&100 nA, 9-26 min) Rx781094 (2CMOnA, 5-21 min) (20-80 nA, 10-43 mitt)

Ejection site

Tizanidine* + -

Noradrenaline* + -

Clonidine* + -

IV-V

0

3

0

5

0

2

IV-V II-III

0 0

3 2

0 0

7 2

0 0

2 1

II-III

0

2

0

3

IV-V

0

2

0

6

IV-V II-III

2 1

0 0

3 2

0

2 1

0 0

IV-V II-III

0 0

1 1

0 0

2 3

0 0

2 1

IV-V II-III

0 0

3 2

0 0

6 6

0 0

3 2

IV-V II-III

8 5

0 0

11 8

2 2

6 5

0 0

I

*The numbers in each column refer to the number of neurones where the depressant actions of the agonists were antagonized (+) of unaffected (-) by the antagonists. tThe range of ejecting currents (nA) and duration of ejection (min) for each antagonist is shown in parentheses. #High ejecting currents of yohimbine and piperoxane were achieved by ejecting them from two or more adjacent barrels of the microelectrode simultaneously.

of these particular neurones to noxious stimuli were unaffected by phenylephrine, but were depressed by noradrenaline. Eficts of adrenoceptor antagonists on the inhibitory actions of tizanidine, noradrenaline and clonidine

The effects of adrenoceptor antagonists on the depressant action of tizanidine, noradrenaline and clonidine on responses of laminae IV and V neurones to noxious stimuli are summarized in Table 2. The antagonists were ejected for some minutes before and also usually during the ejection of agonist; both agonists and antagonists being ejected at the same site. The antagonists were initially tested against the actions of noradrenaline, since the depressant effect of this substance was more rapidly reversible than that of tizanidine or clonidine. However, the effects of an antagonist were not always evaluated against all

three agonists on each cell studied because this usually entailed maintaining stable recordings from that particular neurone for at least 4 h in order to establish adequate controls with each agonist prior to testing the antagonist. The fl-antagonist, sotalol (only tested in laminae IV and V), and al- antagonists, WB4101 and prazosin, had no effect on the depressant actions of tizanidine, clonidine or noradrenaline (Table 2). However, WB4101 (25 nA) antagonized the noradrenaline- and phenylephrine-induced increase in spontaneous firing rate observed on one lamina IV neurone. The a,-antagonist, piperoxane, ejected at either site, and the non-selective a-antagonist, phentolamine, ejected near laminae IV and V neurones from solutions at pH 5.5, also failed to antagonize the actions of these agonists. However, some difficulty was experienced with assessing the effects of piper-

Fig. 2. Depression of responses of a lamina V neurone to noxious heat stimuli (47°C l----0) but not responses to innocuous air-jet stimuli (O---O) by noradrenaline 20 nA, tizanidine 15 nA and clonidine 15 nA and the subsequent antagonism of these effects by the a,-adrenoceptor antagonist RX78 1094 30 nA ejected at the same site (cell body) for several minutes before and during the ejection of agonists. The graph shows the firing rate (spikes/s), corrected for spontaneous firing rate, in response to each cutaneous stimuli. At least 35 min intervals elapsed between recording the effects of each agonist after RX781094 administration to allow for the reversal of the action of this antagonist. Spontaneous tiring rates are also plotted on these graphs (A.. .A).

3. DAVIESand .J. E. QUIPU’LAN

oxane and phentolamine as both substances often reduced spike amplitude and depressed neuronal responses to all cutaneous stimuli when ejected with relatively low currents f < 20 nA). Furthermore, the passage of current through microp~~tte barrefs containing these antagonists frequent& generated eiectrical noise with subsequent failure to maintain a set ejecting current. Few neurones were tested with the %-antagonist yohimbine because current could rarely be passed through soiutions of if, presumably due to its low aqueous soiub~Jjty; in no cell was dear antagonism observed with this compound. Jrr contrast to the Jack of effect of the aforementioned antagonists, the selective cr,-antagonist, RX78J094,‘0 reversibly and reprudnc~bly antagonized the depressant actions of noradrenaline, tizanidine and clonidine on both neuronal responses to noxious stimuli, and spontaneous activity (Table 2). More detailed summaries of the antagonist action of RX78 LO94are included in Table 3. When ejected either near the Jaminae IV and V neurones or into laminae II,--JJI, RX781094 markedly reduced or abolished the effects of noradrenaline, tizanidine and clonidine on ali neurones tested. The antagonism by ~X781~4 was manifested as a reduction in the degree and the duration of the depressant actions of the agonists (Table 3). An example of the antagonism by RX781094 of the effects of all three agonists is illustrated in Fig. 2. RX781094 was also capable of reversing an established agonist-induce depression of neuronaf responses to noxious stimuli (Fig. 3j. Spike amplitude and neuronal responses to cutaneous stimuli where unaffected by RX781094 when it was ejected into laminae IL-JII, but both were sometimes slightly reduced (by 1%30%) when RX781094 was ejected close to laminae IV and V cells. These effects of RX781094 were immediately reversible on terminating the ejection. In such cases, the RX781094 ejection was terminated just before testing the agonist. The duration of the antagonist action of RX781094 against noradrenaline was determined on 10 neurones. The depressant action of noradrenaJine fully recovered 10-40min (23 -fr IO, mean t-SD, n = 7) and 16-45 min (30 & 14, n = 3) after administering RX781094 into laminae IV-V and II-III, respectively. RX781094 had no effect on: (a) the non-selective depressant actions of f - )baclofen (two cells) or y -aminobutyrate (three cells) on responses of laminae JV and V cells to cutaneous stimuli when ejected near these neurones, or (b) the selective antin~iceptive action of morphine jthree cells) ejected into Jaminae II-III (Fig. 3). Jn each case the depressant action of noradrenaline was antagonized on all cells tested. During the course of this study it was observed that phentolamine could tK more readiiy iontophoresed from solutions acidtied to pW 3.5 (with WI) compared with unadjusted aqueous soluajons of pH 5.5. Ejection of phentolamine from t&se acidified solu-

Effects of tizanidine and noradrenaline at spinal a,-receptors

679

RX 781091 30nA ,-.-_-.

Noradrenaline

30 nA

loo Spikes set-’ 0

RX 701094 30nA

lOOnA

Morphi ne_fO nA

,

1°~[~~~~~~~~ m

m

m

YOlilMSlNE

___ RX 781094 1OOnA

2msI 1

i.v.

I

-

w

NALOXONE

O.lmg i.v

Fig. 3. Ratemeter record illustrating the selectivity of the antagonist action of RX781094. Recordings were obtained of responses of a lamina IV neurone induced by noxious heat (48C) stimuli. All drugs were administers into laminae II-III. ~o~~enaline 30 nA depressed responses to thermal stimuli. This effect was reversed by the additional ejection of RX781904 30 nA. One hour later an ejection of morphine SOnA depressed responses of the neurone to noxious stimuli but this was not antagonized by the subsequent and prolonged ejection of RX781094 at 30nA and then lOOnA, nor by the intravenous injection of yobimbine (2 mg/kg). However, the depressant action of morphine was reversed following an intravenous injection of naloxone (0.1 mg/kg).

tions antagonized the depressant actions of noradrenaline, clonidine and tizanidine on responses to noxious stimuli and on spontaneous activity (Table 2). This antagonism appeared to be long-lasting since recovery of the depressant effect of noradrenaline was only observed on one neurone within 2.5 h. It is feasible that agonists ejected into laminae II-III may diffuse more ventrally and act at the same sites as those activated by ejection of agonists near the laminae IV and V neurones. To test this hypothesis the ability of RX781094 ejected near laminae IV and V neurones to antagonize noradrenaline ejected into laminae II-III was determined. In five such tests RX781094 failed to alter the depressant action of noradrenaline ejected into laminae II-III, but did antagonize its depressant action when ejected near the recording site (i.e. the same site as the antagonist). Efects

ofintravenous administration of yohimbine

Since attempts to eject yohimbine from micropipettes were mostly unsuccessful, the effects of intra-

venous injection of this agent were examined on the antinociceptive action of tizanidine, noradrenaline Yohimbine 0.3-2 mg/kg was administered intravenously to 10 cats after establishing control responses to noradrenaline, clonidine and/or tizanidine on 12 laminae IV and V neurones. The agonists were iontophoresed near the recorded neurones (five cells) or into laminae II-III (seven cells). Doses of up to 0.5 mg/kg had no significant effect on responses of spinal neurones to cutaneous stimuli or on arterial blood pressure. However, doses of 0.75-2 mg/kg yohimbine produced an immediate dose dependent decrease (25-60%) of mean arterial blood pressure and an increase in pulse pressure. During this hypotensive episode responses to noxious, but not innocuous, stimuli were enhanced by 30 to more than 100% in five cells and were unaffected in three cells. Spontaneous firing rates were atso increased in three of seven neurones. Within approximately 5 min of injecting 0.75-l mg/kg yohimbine arterial blood pressure

6.0

.PA

PA 8.0 c

2cm

2cm

Pig. I. Diagrammatic representation of the regions richly innervated with catecholamine-containing fibers. The diagrams of coronal (left column) and parasagital (right column) sections were modified from the atlas by Karten and Hodos.16 The density of innervation is indicated by the density of stippling. Position of some regions shown in Fig. 2 are indicated with arrows which bear the corresponding letters (encircled). The numbers indicate the anteroposterior and mediolateral position of the respective sections. Abbreviations: AD, archistriatum, pars dorsalis; APH, area parahippocampalis, corticoid; HV, hyperstriatum ventrale; LFS, lamina frontalis superior; LPQ lobus parolfactorius; NC, neostriatum caudale; PA, paleostriatum augmentatum; SL, septum laterale; TPO, area temporo-parieto-occipitalis; V, ventricle. For identification of other structures consult the corresponding diagrams in Ref. 16. 679 This is a replacement

art page for NSC

15 :3, pp. 679480.

J.

680

DAMES

and J. E.

recovered to 85-95% of the pre-injection level and thereafter remained stable. In parallel with the blood pressure changes the neuronal responses to noxious stimuli decreased either to pre-injection ievels or to levels 2O-50% greater than control values. However, the marked hypotension and enhancement of neuronal responses to noxious stimuli seen following doses of 2mg/kg yohimbine often persisted for 20 min before stabilizing at levels nearer the control values. Agonists were not tested until arterial blood pressure and neuronal responses to cutaneous stimuli had stabilized. Under these conditions, doses of OS-2 mg/kg yohimbine appeared to completely antagonize the depressant actions of noradrenaline @/IO cells), clonidine (4/4 cells) and tizanidine (6/6 cells) on both neuronal responses to noxious stimuli and spontaneous activity (when evident). The antagonism produced by yoh~mbine was reversible 30-45 min after its administration. It is possible that the enhancement of neuronal responses to noxious stimuli following the injection of yohimbine may have masked the depressant actions of the agonists. However, in studies on two neurones the intensity of the noxious stimulus was adjusted after the yohimbine (2 mg/kg) injection to evoke neuronal responses comparable to the control responses. On both neurones clear antagonism of the depressant effects of noradrenaline, clonidine and tizanidine were still observed. Yohimbine (2 mg/kg) did not antagonize the selective antin~i~tive action of morphine (SO-100 nA) ejected into laminae II-III on two cells although this action of morphine was reversed by the subsequent intravenous injection of 0.1 mg/kg naloxone (Fig. 3). As a further check on the validity of the antagonism produced by systemic yohimbine, its ability to influence the actions of the agonists on excitation of laminae IV and V neurones evoked by electrical stimulation of cutaneous nerves at intensities sufficient to activate unmyelinated fibres (C fibres), were determined in experiments on three cats. In

Q~~LAN

these experiments 1.0-l .5 mg/kg yohimbine caused similar hypotensive effects to those described above, but this was not accompanied by changes in the electrically evoked synaptic excitation. All three agonists virtually abohshed the long latency excitation evoked by such stimuli in three spinal neurones, and in each cell yohimbine antagonized these effects. In studies on two of the above neurones tizanidine was also administered intravenously. A dose of 0.1 mg/kg tizanidine completely abolished the long latency electrically evoked excitation in each cell in keeping with a previous report.’ The sub~quent intravenous administration of yohimbine 1 mgjkg reversed this action of tizanidine (Fig. 4).

DISCUSSION

We have previously reported that both tizanidine and noradrenaline depress responses of laminae IV and V neurones to noxious, but not innocuous, cutaneous stimuli, when ejected near these neurones or more dorsally into laminae H-11.’ Others have also demonstrated a similar selective antinociceptive action for noradrenaline on dorsal horn netnones.‘.” The present results indicate that the selective antinociceptive action of noradrenaline is probably mediated by an interaction with a+adrenoceptors; since it is mimicked by the a,-agonist clonidine. Moreover, dopamine, the &agonist isoprenaline and the selective a,-agonists, amidephrme and phenylephrine, either had no effect or only relatively weak and sometimes non-selective depressant e&r% on neuronal resjKnntes to Cutaneous stimuli. Preliminary data reported by ~leetwo~-Walker and c&agnes’* also suggest a,-adrenoceptors may be involved in the selective depressant action of noradrenaline on responses of dorsal horn neurones to noxious stimuli. The observation that the effects of tizanidine ejected into laminae II-III or in the vicinity of laminae

N"ci l5 Spikes

0

synaptiudly evoked response.

Fig. 2. Catecholamioe fluorescence of different regions in the pigeon brain. (A) posterodorsolateral neostriatum, typical view. (B) Posterodorsolateral neostriatum, a region with pericellular nests. (C) Pe~llular fluorescent nest in ~~er~o~~ateml n~t~atum of a bird pretreated with reserpine, the peripheral inhibitor of DOPA dezarboxylase and L-WPA. (D) The prepiriform cortex fsee the coronal sections at (A) 14.00-14.50 in Ref. 161. (E) The archistriatum. (F) The septal region and lobus parolfactorius on each side of the ventricle in a parasagital section at about L = 1.0. Each bar is 50 pm long. The approximate position of these photographs is indicated in Fig. 1 by means of arrows bearing the corresponding letters (encircled). c, corticoid; s, septum; v, ventricle.

680

Effects of tizanidine and noradrenaline at spinal qreceptors IV and V neurones were remarkably similar in all respects to those of clonidine suggests that tizanidine may also act at a,-adrenoceptors. Certainly opiate receptors are not implicated in the inhibitory actions of these agents, since their actions were not affected by the opiate antagonist, naloxone. The notion that a,-receptors may be involved is more firmly established by the finding that: (a) the selective a,-adrenoceptor antagonist, RX781094,” reversibly antagonized the preferential depressant actions of tizanidine as well as those of noradrenaline and clonidine, (b) the actions of these three agonists were unaffected by the selective a,-antagonists, WB4101 and prazosin, and the j-antagonist, sotalol. The failure of the iontophoretic administration of the a,-agonists piperoxane and yohimbine and the a-antagonist phentolamine @H 5.0) to modify the actions of noradrenaline, clonidine and tizanidine may seem at variance with the suggestion that the latter agents act at a,-receptors. However, it was difficult to assess the effects of these antagonists in view of the direct effects they had on neuronal excitability and problems encountered in administering them iontophoretically. Others have experienced similar difficulties with adrenoceptor antagonists in the spinal cord.‘,‘3 On the other hand, the inhibitory effects of noradrenaline and clonidine on spinal sympathetic preganglionic neurones, and on noradrenergic neurones in the nucleus commissuralis, are apparently antagonized by iontophoretic piperoxane and phentolamine. ‘s*23The latter investigators used microelectrodes with tip sizes of lo-20 pm compared with those of 4-6 pm in the present experiments, which may account for these conflicting results. It is unlikely that the antagonistic action of RX781094 was the result of some non-specific membrane action, since the non-selective inhibitory effects of y-aminobutyrate and (-)baclofen, and the selective antinociceptive effect of morphine, were not modified by RX781094. The ability of intravenously administered yohimbine to antagonize the selective depressant effects of tizanidine, noradrenaline and clonidine, but not those of morphine, is also consistent with an action of the former three agents on a,-adrenoceptors. The occasional enhancement of neuronal responses to noxious stimuli following the injection of yohimbine was probably secondary to peripheral circulatory changes altering the rate of heat transfer to nociceptors.6+12 This is supported by the observation that systemic administration of yohimbine also antagonized the depressant actions of tizanidine, clonidine and noradrenaline on responses of laminae IV and V neurones evoked by electrical stimulation of high-threshold cutaneous nerves (C fibres), whilst itself having no obvious effect on these synaptically evoked responses. In addition to inhibitory effects, excitatory actions have been reported for noradrenaline on central neurones which seem to be mediated at

681

a,-adrenoceptors.22* The observation that WB4101 antagonized the excitations induced by noradrenaline and the a,-agonist, phenylephrine, in a few laminae IV and V cells is consistent with the above mentioned findings. These a,-mediated excitatory events in the dorsal horn appear to be extremely sparse judging by the few cells so affected in this study, and in that reported by North and Yoshimura.24 The marked selectivity of the depressant action of tizanidine, clonidine and noradrenaline suggests that the receptor sites involved in this effect may be closely associated with synaptic connections between nociceptive afferents and laminae IV and V neurones. Indeed, high densities of a,-adrenoceptor binding sites have been demonstrated in the superficial laminae of the dorsal hom.27*2*Such sites could be localized pre-or postsynaptically. A recent study has shown that noradrenaline has a direct hyperpolarizing action on rat substantia gelatinosa neurones in vitro, which is blocked by a,-adrenoceptor antagonists. 24 It has been suggested that this postsynaptic hyperpolarization may account for the anti-nociceptive action of noradrenaline on dorsal horn neurones. A similar mechanism could account for the selective antinociceptive action of tizanidine and clonidine ejected into laminae II-III. However, large diameter afferents innervating hair follicles terminate in laminae III-VI,2*3 and it is therefore difficult to reconcile such a postsynaptic inhibitory action, with the lack of effect of these agents on responses to non-noxious stimuli, when they are administered near the cell bodies of laminae IV and V neurones. One possible explanation could be that these agents act at the same site irrespective of whether they are administered into laminae II-III or IV-V. This seems unlikely, however, as the a,-antagonist RX781094 ejected into laminae IV-V failed to influence the action of noradrenaline ejected into laminae II-III, but blocked its affect when ejected into laminae IV-V. Thus, it would appear that the selective inhibitory action of noradrenaline, and probably also that of tizanidine and clonidine, on laminae IV and V neurones can occur as a result of an interaction with a,-adrenoceptors situated in either laminae II-III or IV-V. Whether one site is more important than the other, in relation to the selective inhibitory action of tizanidine on laminae IV and V neurones, when it is administered systemically,’ or to the behavioural analgesia observed following intrathecal administration of noradrenaline or clonidine,‘6s20 remains to be determined. Finally, it should be mentioned that electrical or chemical stimulation of certain brain stem areas containing noradrenergic neurones with projections to the spinal cord can induce spinal analgesia and depress the responses of dorsal horn neurones to noxious stimuli.‘6*‘g*21 It is possible, therefore, that the a,-adrenoceptors mediating the effects of tizanidine and noradrenaline in the present experiments may be of physiological relevance.

682

J. DAVIESand J. E.

Acknowledgements-The authors are grateful to the M.R.C. and Sandoz Ltd, Basle, Switzerland for financial support and the following companies for gifts of drugs: Endo

QIJINLAN

Laboratories (naloxone), Ciba (baclofen), Reckitt and Colmann (RX 781094), Boehringer lngelheime (clonidine) and Rhone-Poulenc (piperoxane).

REFERENCES

9. 10. 11. 12. 13.

Belcher G., Ryall R. W. and Schaffner R., (1978) The differential effects of 5-hydroxytryptamine, noradrenaline and raphe stimulation on nociceptive and non-nociceptive dorsal horn neurones in the cat. Brain Res. 151, 307-321. Brown A. Cl., (1977) Cutaneous axons and sensory neurones in the spinal cord. Br. Med. Bull. 32, 109-t 12. Brown A. G., Rose P. K. and Snow P. J., (1977) The morphology of hair follicle afferent fibre collaterals in the spinal cord of the cat. J. Physiol. 272, 779-797. Davies J. (1982) Selective depression of synaptic transmission of spinal neurones in the cat by a new centrally acting muscle relaxant, 5-chloro-4-(2-imidazolin-2-yl-amino)-2,1,3-benzothiodazole (DS 103-282). Br. J. Pharmac. 76, 473-48 1. Davies J. and Dray A. (1978) Pharmacological and electrbphysiological studies of morphine and enkephalin on rat supraspinal and cat spinal neurones. Br. J. Pharrnac. 63, 87-96. Davies J. and Dray A. (1980) Vascular effects of substance P changes synaptic responsiveness of rat dorsal horn neurones. Life Sci. 26, 1851-1856. Davies J. and Johnston S. J. (1984) Selective antinociceptive effects of tizanidine (DS 103-282) a centrally acting muscle relaxant, on dorsal horn neurones in the feline spinal cord. Br. J. Pharmac. 82, 409421. Davies J., Johnston S. E., Hill D. R. and Quinlan J. E. (1984) Tizanidine (DS103-282), a centrally acting muscle relaxant, selectively depresses excitation of feline dorsal horn neurones to noxious peripheral stimuli by an action at a,-adrenoceptors. Neurosci. Lett. 48, 1977202. Davies J., Johnston S. E. and Lovering R. (1983) Inhibition by DS 103-282 of D[‘H)asparate release from spinal cord slices. Br. J. Pharmacol. 78, 2P. Doxey J. C., Roach A. G. and Smith C. F. C. (1983) Studies on RX781094: selective, potent and specific antagonist of a,-adrenoceptors. Br. J. Pharmacol. 78, 489-505. Duggan A. W., Hall J. G. and Headley P. M. (1977) Suppression of transmission of nociceptive impulses by morphine: selective effects of morphine administered in the region of the substantia gelatinosa. Br. J. Pharmacol. 61, 65-76. Duggan A. W., Griersmith B. T., Headley P. M. and Maher J. B. (1978) The need to control skin temperature when using radiant heat in test of analgesia. Expi Neurol. 61, 471478. Engberg I. and Ryall R. W. (1966) The inhibitory action of noradrenaline and other monoamines on spinal neurones. J. Physiol. 185, 298-322.

14. Fleetwood-Walker S. M., Hope P. J., Iggo A., Mitchell R. and Maloney V. (1983) Characterization of receptors mediating the selective effect of noradrenaline on the cutaneous response of identified dorsal horn neurones in the cat. J. Physiol. 343, 67-68P.

15. Guyenet P. G. and Cabot J. B. (1981) Inhibition of sympathetic preganglionic neurones by catecholamines and clonidine: mediation by an a-adrenergic receptor. J. Neurosci. 1, 908-917. antinociception by intrathecal 16. Hammond D. L. and Yaksh T. L. (1984) Antagonism of stimulation-produced administration of methysergide or phentolamine. Brain Res. 298, 329-337. 17. Hassan N. and McLellan D. L. (1980) Double blind comparison of single doses of DSl03-282, baclofen and placebo for suppression of spasticity. J. Neurol. Neurosurg. Psychiar. 43, 1132-l 136. 18. Headley P. M., Duggan A. W. and Griersmith B. T. (1978) Selective reduction by noradrenaline and 5-hydroxytryptamine of nociceptive responses of cat dorsal horn neurones. Brain Res. 145, 185-189. 19. Hodge C. J., Apkarian A. V., Stevens R., Vogelsang G. and Wisnicki H. J. (1981) Locus coeruleus modulation of dorsal horn unit responses to cutaneous stimulation. Brain Res. 204, 415420. 20. Howe J. R., Wang J.-Y. and Yaksh T. L. (1983) Selective antagonism of the antinociceptive effect of intrathecally applied alpha adrenergic agonists by intrathecal prazosin and intrathecal yohimbine. J. Pharmacol. exp. Ther. 224, 552-558.

21. Jenson T. S. and Yaksh T. L. (1984) Spinal monoamine and opiate systems partly mediate the antinociceptive effects produced by glutamate at brainstem sites. Brain Res. 321, 287-297. 22. Menkes D. B., Baroban J. M. and Aghajanian G. K. (1981) Prazosin selectively antagonizes neuronal responses mediated by a,-adrenoceptors in brain. Naunyn-Schmieakbergs Arch. Pharmac. 317, 273-275. 23. Moore S. D. and Guyenet P. G. (1983) Alpha-receptor mediated inhibition of A2 noradrenergic neurones. Brain Res. 276, 188-191.

24. North R. A. and Yoshimura M. (1984) The actions of noradrenaline on neurones of the rat substantia gelatinosa in vitro. J. Physiol. 349, 43-55.

25. Reddy S. V. R., Maderdrut J. L. and Yaksh T. L. (1980) Spinal cord pharmacology of adrenergic agonist-mediated antinociception. J. Pharmacol. exp. Ther. 213, 525-533. 26. Sayers A. C. Burki, H. R. and Eichenberger E. (1980) The pharmacology of 5-chloro-4-(2-imidazolin-2-yl-amino)2,1,3benzothiodazole (DS 103-282), a novel myotonolytic agent. Drug Res. 30, 793-803. 27. Unnerstall J. R., Kopajtic T. A. and Kuhar M. J. (1984) Distribution of a,-agonist binding sites in the rat and human central nervous system: analysis of some functional, anatomic correlates of the pharmacologic effects of clonidine and related adrenergic agents. Brain Res. Rev. 7, 69-101. 28. Young W. S. and Kuhar M. J. (1980) Noradrenergic a,- and a,-receptors: light microscopic autoradiographic localization. Proc. Nain. Acad. Sci. 77, 1690-1700. (Accepted 6 August 1985)