Brain Research, 519 (1990) 169-182 Elsevier

169

BRES 15563

The involvement of neurokinin receptor subtypes in somatosensory processing in the superficial dorsal horn of the cat S.M. Fleetwood-Walker 1, R. Mitchell 2, P.J. Hope 1, N. E1-Yassir 1, V. Molony I and C.M. Bladon 2 t Department of Preclinical Veterinary Sciences, Royal (Dick) School of Veterinary Studies, Summerhall, Edinburgh (U. K.) and 2MRC Brain Metabolism Unit, Edinburgh (U.K.)

(Accepted 28 November 1989) Key words: Neurokinin; Substance P; Spinal cord; Thermal nociception; Pain; Analgesia; Neurokinin receptor

As well as substance P (SP), neurokinin A (NKA) and neurokinin B (NKB) have recently been found in the superficial dorsal horn of the spinal cord; NKA originating mainly in fine primary afferents. We have investigated the effects of these tachykinins and a range of analogues on somatosensory responses of single identified dorsal horn neurons, when applied ionophoretically to the region of the substantia gelatinosa. Behavioural reflex tests of thermal nociception were carried out in parallel. The role of NK-1, NK-2 and NK-3 receptors was addressed. NK-l-selective agonists attenuated the non-nociceptive responses of identified multireceptive spinocervical tract (SCT) neurons. Of the endogenous tachykinins, both SP and NKB (a weak NK-1 agonist) showed this effect. No role for NK-3 receptors was identified in our experiments. NK-2-selective agonists (including NKA) caused a unique and selective facilitation of thermal nocieeptive responses. NKA also reduced reflex response latency in tail-flick and hot plate tests. NKA as a primary afferent transmitter may thus be involved in mediating or facilitating the expression of thermal nociceptive inputs in the substantia gelatinosa. NKA and SP could be considered as acting in concert in the superficial dorsal horn in an effectively pro-nociceptive modulatory role. Evidence from receptor-selective antagonists supports that obtained with agonists for the roles of particular NK receptors in somatosensory processing. NK-2, but not NK-1 or NK-3 antagonists attenuated endogenous thermal nociceptive responses, supporting the hypothesis that an NK-2 agonist (such as NKA) may normally participate in expression of thermal nociception in the superficial dorsal horn. Behavioural experiments showing increased response latencies with a putative NK-2 selective antagonist further supported the involvement of NK-2 receptors in thermal nociception.

INTRODUCTION The neurotransmitter(s) of cutaneous nociceptive afferent neurons are unknown, but for several years, substance P (SP) has been an important candidate 46. The case for SP being a mediator of nociception, however, remains open. Substance P-like immunoreactivity is contained in some small dorsal root ganglion cells with fine fibres which would require high intensity stimulation for their activation 26. Indeed, SP-like immunoreactive material is released within spinal cord by noxious cutaneous stimuli or by electrical stimulation of afferent fibres at high intensities 1°'35'36'76. Dorsal horn SP-like immunoreactivity is also present in local neurons and descending terminals however 26, so the released material may not necessarily originate from afferents. Along with other peptides in primary afferents, SP can be depleted by capsaicin 38"44. This primary afferent neurotoxin also changes the thresholds in behavioural analgesia tests, but

it now seems likely that this p h e n o m e n o n is dissociated from any depletion of SP 38'44. |ntrathecal administration of SP elicits a behavioural syndrome of hindlimb biting and scratching, similar to that evoked by peripheral irritation 29,66. Recent analysis suggests, however, that the syndrome may not relate specifically to the experience of pain 21. Several groups have reported that intrathecal tachykinins can reduce response thresholds in behavioural analgesia experiments 4'6'45"51'78 and that antagonists can, in some cases, elevate these thresholds 54,56'62'69. Furthermore, nociceptive reflexes in spinal cord can be facilitated by tachykinin receptor agonists and inhibited by antagonists or antisera 2'50'72'74. However, multiple sites of action, inappropriate to the function of endogenous nociception, such as the ventral horn 49, may contribute to these results and the pharmacological specificity of some of the compounds has been questioned 27'65. Several groups have applied SP by microionophoresis

Correspondence: S.M. Fleetwood-Walker, Department of Preclinical Veterinary Sciences, Royal (Dick) School of Veterinary Studies, Summerhall, Edinburgh EH9 1QH, U.K.

170 in the vicinity of dorsal horn neurons and generally observed excitatory effects 23'53'57"64'79. In some cases, complex mixed excitatory and inhibitory effects have been o b s e r v e d 7"41"60'73, suggesting multiple sites or components of action. Although some early work 23'57 suggested that the excitatory effects of SP occurred more frequently on cells with nociceptive rather than solely non-nociceptive response characteristics, the majority of other studies have not supported this. However, in all of these studies, heterogeneous neuronal populations have been examined, which may present considerable variation in the organisation of their cutaneous afferent inputs. Furthermore, the predominant zone of fine afferent termination and also of receptors for the released transmitters, is the substantia gelatinosa52; far away from most of the drug administration sites in the above studies. The possibility that not SP itself, but a related peptide is involved in nociception, has arisen with the discovery of two closely-related tachykinins, neurokinin A (NKA) and neurokinin B (NKB) in spinal cord 33. Neurokinin B does not appear to be associated with cutaneous afferent fibres, whereas the NKA content of the dorsal horn seems to originate almost entirely from small diameter primary afferents 48'67. The NKA presumably arises there from expression of fl- and y-prepropeptides from the preprotachykinin-1 gene 34'45. Neurokinin A (or even the N-terminally extended form, neuropeptide K (NPK) 1'68, present in the fl-prepropeptide) are likely to be coreleased with SP from the relevant afferents and therefore possibly subserve a role in signalling nociception. Moreover, recent evidence suggests that at least 3 distinct types of receptors for tachykinins are present in mare-

A

600 I.tV

malian tissue. The NK-1, NK-2 and NK-3 neurokinin receptors are preferentially activated by the endogenous peptides SP, NKA and NKB, respectively32'51. Although the relative selectivity of the presumed endogenous ligands for these sites is not great, it is possible that quite distinct physiological roles could be exerted by them through these receptors. Binding sites for NK-1 selective ligands are present in dorsal horn, as are sites for ligands which are considered in other locations, to label NK-3 and perhaps NK-2 sites47. Little is known of the functional relevance of each of these sites with respect to sensory processing. We have investigated this using agonist and antagonist analogues, selective for NK-1, NK-2 and NK-3 receptors as well as the endogenous peptides SP, NKA and NKB. Drugs were locally administered by ionophoresis in the zone of fine afferent termination in the dorsal horn (the substantia gelatinosa; lamina II) whilst the integrated sensory responses of single identified ascending neurons were recorded with a separate electrode in deeper laminae. Parallel experiments on behavioural analgesia following intrathecal administration of tachykinins were also carried out in an attempt to corroborate the electrophysiological findings. Preliminary reports of some aspects of this work have been published 12'13'15'16. MATERIALS AND METHODS

Electrophysiological experiments General Recordings were made from single, antidromically-identified neurons of the spinocervical tract (SCT) (which conveys both non-nociceptive and nociceptive information to higher levels of the

]I

I 111! .........

' .......

rrr

1.5 sec

Fig. 1. A. (For legend see Fig. 1. B.)

171 CNS), according to methods described in detail previously14'w. We studied SCT neurons for 3 reasons: (1) because they are relatively suitable for long duration stable recordings - - essential for these studies; (2) because in our studies with other neuropharmacological studies we have found them to be influenced in a manner closely paralleling influences on other ascending tracts14"17; and (3) because although ventrolateral pathways are generally thought to be of prime importance in the rostral transmission of nociceptive information, dorsolateral pathways (such as the SCT) may also be important, particularly in carnivores such as the cat 71. Experiments were carried out on 49 adult cats, anaesthetised with intravenous a-chloralose (60--70 mg/kg) after induction with halothane. Animals were paralysed with gailamine triethiodide (Flaxedil, 15 mg/kg) and artificially respired with room air following bilateral pneumothorax. Additional doses of chloralose (30 mg/kg) were given when required to maintain anaesthesia. Laminectomies were carried out to expose the upper cervical cord (from C 1 to C5) for antidromic stimulation of the axons of SCT cells and the lumbar cord (segments L3-S1) , for recording.

28 24

20 1(8 12 8 4

(:

~m Brush

~ 30 48 ao(*C) Noxious heat

M Noxi pinch

28

Neuron identification

24

Extracellular recordings of multireceptive SCT neurons were made in lumbar segments L6-L7, with glass micropipette electrodes. SCT neurons were identified by their antidromic activation from silver ball electrodes at the upper cervical spinal cord using monophasic square wave pulses, 1 Hz, 0.5 ms pulse width, 600-700 mV amplitude as the search stimulus. SCT neurons were activated from stimulating electrodes on the dorsolateral funiculus at C3, but not C~, nor from electrodes on the dorsal column at C4-C 5. (Restricted surgical section of the dorsal columns was made at C4 to prevent misidentification from crossing fibres.) Identified cells satisfied the standard criteria of collision, fast following frequency, constant latency and distinct threshold for activation, as described in detail previously14.

E

0 20 0 =. 16 12 t0

8

O

4

emm Brush

~ 30 48 30 (*C) Noxious heat

Noxi pinch

Recordings and ionophoresis protocols

28 E 24

2G 16 12

8 4 0

~m Brush

~ 30 48 ao (*C) Noxious heat

m Noxi pinch

i

J 20a

Fig. 1. A: raw oscilloscope records showing the initial stages of the excitatory effects of (I) brush, (II) pinch and (III) heat stimuli on the firing rate of an SCT neuron. The cutaneous stimuli were initiated at the step in the analogue marker below each record. The

resting skin temperature of the heat response was 30 °C and the plateau 48 °C. Note the very long latency activation in the noxious heat response. B: ongoing activity record showing typical effects of NKA (applied in the region of the substantia gelatinosa) on sensory responses of an SCT neuron. Ongoing firing frequency was recorded in 400 ms bins and plotted against time. The top sequence shows the excitatory responses to brush, noxious heat and noxious pinch. The middle sequence shows the effect of ionophoresis of NKA (250 nA, beginning 1 min prior to the brush stimulus). Recovery was seen 15 rain later, as shown in the bottom sequence. Ejection of NaCI, at up to 400 nA had no effect.

Recordings were made using either single barrel glass microelectrodes (pulled from 1.2 mm diameter fibred glass and contained 2% Pontamine sky blue in 0.5 M sodium acetate), or the central barrel of a 7-barrelled electrode (which had been constructed from 1.5 mm diameter fibred glass and contained 4 M NaCI). The electrodes had tip sizes of 0.5-1.0/~m and 3.5-5/~m and DC resistances of 10--20 MI2 and 4-8 MD, respectively. The other barrels of the multibarrelied electrodes contained 1 M NaC! (pH 4.0-4.5) for automatic current balancing and independent current controls, Pontamine sky blue (2% in 0.5 M sodium acetate) and selections of the following compounds in distilled water at concentrations of 1.2-10 raM: pH 4.5-5.0 unless otherwise indicated: substance P (SP); [MetOMe11]Sp (SPOMe); [Met-OHll]SP (SPOH); neurokinin A (NKA); kassinin (KASS) (pH 8.0-8.5); [GIp6,D-Prog]SP6_11 (DPro9); [GIp6,L-Pro9]SP6_I] (L-Pro9); neurokinin B (NKB, pH 8.0-8.5); succinyl [Asp6,Me-Phea]SP6_ll (senktide; SENK, pH 8.0-8.5); bombesin; sodium glutamate (100 mM, pH 8.0-8.5); [D-Arg],D-TrpT'9,Leu11]SP (spantide, SPAN); [D-Tyr4,o-Trp 7'9, Nlel]]SP4_1] (D-Tyr4); [D-Pro4,Lys6,o-TrpT"9A°,Phe]llSP4_ll (o-Pro4); and [o-Pro2,o-Trp6'S,Nlel°]NKB (NKB-A, pH 8.0-8.5). Compounds were either obtained from Peninsula, Sigma or Cambridge Research Biochemicals, were gifts from Merck, Sharpe and Dohme ([Glp6, D/L-PrOg]SP6_ll and senktide) or were synthesised in our laboratories (by C.M.B.) using Sheppard's Fmoc-t-butyl polyamide chemistry 11. After purification, the synthesised peptides had the expected amino acid analyses and FAB mass spectra. Peptides were stored at -40 °C over dessicant and solutions were always prepared immediately before use. Compounds were ejected from the electrodes with cathodal currents (KASS, NKB, SENK, NKB-A and glutamate with anodal currents) using a Neurophore BH2 system. Retaining currents of 10-15 nA were used to minimise drug leakage.

Positioning of electrodes In most experiments, a multibarrelled electrode was used to

172 ionophorese compounds into the substantia gelatinosa, and a single barrel electrode to record the responses of SCT neurones in laminae IV/V. To do this, the single barrel electrode was first advanced at a rostral angle of 10° from the perpendicular to locate an SCT neuron. The position of the electrode tip was calculated from depth and angle measurements and then the multibarrelled electrode was slowly advanced at an opposing caudal angle of 25°, from a surface position such that it would reach the substantia gelatinosa at a site directly dorsal to the recording electrode. Positions of both recording and ionophoresis sites were marked by dye spots (50 /zA-min). Data were only included for analysis if histological processing showed dye-spot pairs to be located in the region of appropriate laminae and separated by a rostrocaudal distance of less than 150 ,uM 17. Only a few cells (well-separated rostrocaudally) were recorded in each experiment, so there was no confusion over matching dye-spot pairs.

Responses to cutaneous sensory stimuli Neurons were excited by controlled noxious and innocuous stimuli applied (usually for 10 s periods) to adjacent cutaneous areas within their receptive fields, on the ipsilateral foot. Innocuous

stimuli were provided by a motor-driven, rotating brush and noxious stimuli by a thermocouple-controlled radiant heat lamp (giving a controlled surface temperature ramp between 30 and 48 °C) and by a noxious pinch device which applied controlled displacement, at a calibrated pressure, to a pair of serrated forceps. The noxious or innocuous quality of the stimuli was confirmed in man. A regular cycle of responses to these stimuli was set up. Control trials were always repeated at least twice, as were many of the test responses. Results were rejected if duplicated responses varied by more than 10-15 %. The intensities of the various stimuli were adjusted to give an approximate match for the number of action potentials in each of the (submaximal) responses. Action potentials of the SCT neurons could be clearly discriminated from other field potentials and were continuously monitored for any changes in amplitude or configuration due to non-specific membrane effects of the drugs. Raw oscilloscope records showing the initiation of responses to brush, pinch and thermal stimuli are shown in Fig. 1A. The cycle of stimuli was repeated every 3 min, starting the ejection of most drugs 1 min before the start of a cycle. The possibility of artefactual biases in the sensitivity of different responses to the drugs, due to the order in which the responses were

TABLE I

Summary of the effects of N K receptor agonists (ionophoresed in the region of the substantia gelatinosa) on sensory responses of SCT neurons Experiments were performed and analysed as in Figs. 1-3 and individual plots of percentage of initial control responses against time were examined to find the time point of maximal changes in any of the responses. For the NK-1 and NK-3 agonists this was chosen to be the maximal decrease in the non-nociceptive response and for the NK-2 agonists, NK plus SP, and non-taehykinin controls it was chosen to be the maximal increase in the thermal nociceptive response. In each case at the time of maximum change, the percentage of control values was calculated for each type of response. The mean -+ S.E.M. values for these percentages are shown with n in parentheses. Since it was not always possible to test all types of response in a particular neuron, these n values sometimes vary for the different responses. The mean ± S.E.M. values are also shown for ionophoretic currents and for the time after commencing ionophoresis at which the maximal change (and all the other corresponding values) were recorded. Ionophoresis was usually continued until 6 test cycles had been completed. Statistically significant differences (P < 0.05 by matched-pair t-test on the raw data) are indicated (*) for facilitation of responses to noxious heat and (t) for inhibition of responses to innocuous brush. Effects of [Met-OMe I lISP, but none of the other substances, fell into two apparently distinct categories, indicated as populations a and b.

Substances applied by ionophoresis

Neuronal responses (as percentage of mean pre-drug control) Heat

NK-1 agonists: SP [Met-OMen]SP (a) (b) [Met-OHlqSP [GIp°,L-Prog]sP6_ H

108+ 90 + 90 + 105 + 90 +

NK-2 agonists: NKA Kassinin [GIpO,D-Pro9]SP6_II

Pinch

10 (4) 4 (12) 4 (4) 8 (4) 8 (4)

Brush

80_+12 (5) 81_+9 (6) 87±8 (3)

lonophoretic current (nA )

(4)

11 + 4 9+ 1 12 + 2 7+ 1 10 + 1

400 380+6 400 400 400

5+1 8+2 9+3

290+6 400 400

Spontaneous

63±7 (5) t 4 6 ± 4 (12) t 9 8 ± 1 2 (4)

88_+11 94±3 94_+9 99-+7 96+7

(5) (12) (4)

101±10

(4)

(4)

91+6

(4)

144_+ 3 (18)* 139 ± 11 (5)* 156 ± 15 (5)*

100±1 (11)

95±2

(18)

85±8

(5)

(5)

100_+7

(5)

9 5 + 2 (18) 96-+6 (5) 1 1 0 + 2 (5)

NK-3 agonists: NKB succ[Asp6,Me-Phes]SP6 N

101±8 98_ 4

90±5 (7) 103+11 (6)

52±7 90_+4

(9) t (6)

96+8 100_+5

(9) (6)

10 + 4 8+ 2

400 400

Combinations: SPplusNKA

141± 10 (5)*

7 9 ± 1 8 (5)

60±9

(5) t

103±10 (5)

14 + 3

400 + 400

Non-tachykinins : Bombesin Glutamate

87 + 7 108 ± 8

(4) (4)

97±8 96+11

(4) (4)

97±7 95+8

(4) (8)

Current controls: NaCI pH4.5(150mM) pH8.5

102_+3 93 + 3

(5) (4)

91_+4 98±5

(5) (4)

88+5 94-+5

(5) (4)

(9) (6)

-

Time of maximal change in responses (min)

82_+6

-

104±8

(4)

5+ 1 7-+2

10+3 13+4

400 400

400 400

t

173

[GIp 6, L.Pro 9 ] SP6.11

[ .et-o.d ~ ] sP 150

150

A

B

100

100

E "6

50

I

I I

I

Cl

•

i

C2

T1

I

I

2

3

•

i

4

I

I

5

6

i

'

C " 2

"

i

C1

|

•

i

2

•

3

!

i

4

5

3 min cycles

9

succ [~,MePhe 8 ]SP 6-~

[GIp ,D-Pro ] SP6.11

150

150

P

•

T1

3 rain cycles

6

I

D

100

8

8 50

50

E "6

"6

I

I ,

Cl

•

,

C2

•

,

•

T1

,

2

•

,

3

-

,

4

-

,

•

5

3 rain cycles

,

Cl

-

,

C2

-

I

,

-

T1

,

2

-

,

3

•

,

4

-

,

-

5

3 min cycles

Fig. 2. Effects of some selective NK receptor agonists (applied in the region of the substantia gelatinosa) on sensory responses of SCT neurons. Individual experiments were carried out as described in Fig. 1. Stimulus-evoked activity (and a 10 s epoch of spontaneous activity prior to the brush stimulus) was integrated and expressed as a percentage of the mean of control responses C 1 and C2. Consecutive 3 min cycles of stimuli were then continued (T1-Tn) during ionophoresis of the drug (in these cases at 400 nA) as shown by the bar. Responses to noxious heat (O), noxious pinch (D), innocuous brush (A) and spontaneous activity (O) are shown as mean + S.E.M. values from 5-10 experiments with each agonist. Statistically significant differences from controls are indicated (*), P < 0.05 by matched-pair t-test on the raw data.

tested, was overcome by varying the order of the stimuli. Since maximal effects were generally slow to develop (in the order of 5-10 min), drugs were usually applied at a set ionophoretic current over several test cycles (see for example, Fig. 2). In some cases, ionophoretic currents were increased stepwise, allowing the demonstration of current-dependent drug effects on the same neuron (see Fig. 4). After any test, the responses were allowed to recover fully (up to 30 min) before further study. Continuous records of firing rate (over 400-2000 ms bin widths) were plotted, together with the analogue signals from the stimulators and Neurophore, and then stored on computer disc. Data were analysed by integrating the numbers of stimulus-induced action potentials in selected epochs. Responses to the mechanical brush were integrated from the 10 s epoch over which the stimuli were presented. Responses to pinch were integrated over a total of 20 s to include the prolonged excitatory response which outlasted the stimulus. Responses to the noxious thermal stimuli were integrated over 25-30 s from the start of the 48 °C plateau to include the long-latency activity evoked by this stimulus. The integrated responses were normalised and expressed graphically, to allow direct comparison of the effects of a particular drug on the different types of activity.

Behavioural experiments Behavioural analgesia testing with intrathecal administration of

drugs was carried out according to Yaksh and Rudy 77. In brief, male Wistar rats (140-220 g) were anaesthetised with O2/halothane and using sterile technique, an indwelling polyethylene cannula was inserted intrathecally from the base of the skull to terminate around spinal segments L2-L 4 (verified at postmortem). The tubing was of outer diameter 1.8 mm, internal diameter 0.4 mm, but drawn out to approximately 0.5 mm outer diameter for the caudal 2-3 cm. The cannula passed out of the skin through a secure injection port. Antibiotics were given both at the site of the wound and systemically. Within 15 min animals were conscious with normal sensorimotor function. After 5-8 days, animals were tested following injection of 5-10 /~1 drug followed by 5-10/A saline flush. The tail flick test was carried out using a radiant heat source designed to raise the surface temperature to 55-57 °C over a rise time of 5.5-6.0 s. The region between 5 and 7 cm from the tail tip was always tested. The cut-off time was set at 15 s. The hot plate test was conducted at 56-57 °C. All normal/sham operated or vehicle alone control rats gave a marked locomotor reaction within 1.8-2.1 s. The cut-off time was set at 6 s. Tail temperature was measured regularly and only showed clear changes (fall of up to 4 °C) in cases of flaccid paralysis and failure at the rota-rod test (6 rpm for 1 min). Histological examination of spinal cords after testing revealed no gross neurotoxic changes due to any of the drugs.

174 RESULTS

I, Fig. 2). T h e r e d u c t i o n was o n l y partial, with a m a x i m u m in the o r d e r of 5 0 % ; similar to o u r observa-

Electrophysiological experiments

tions with o t h e r n e u r o p e p t i d e s applied in superficial dorsal h o r n ~7. A small n u m b e r of n e u r o n s were u n affected by S P O M e , b u t it is possible that technical

General T h e locations, receptive fields a n d electrophysiological p r o p e r t i e s of the S C T n e u r o n s were similar to those r e p o r t e d previously. O n l y m u l t i r e c e p t i v e n e u r o n s excited by b o t h i n n o c u o u s a n d n o x i o u s c u t a n e o u s stimuli were e x a m i n e d in this study. A l l c u t a n e o u s receptive fields were o n the ipsilateral h i n d l i m b a n d r e m a i n e d t o p o g r a p h ically c o n s t a n t t h r o u g h o u t the r e c o r d i n g period. Most cells r e s p o n d e d to n o x i o u s pinch as well as to i n n o c u o u s b r u s h i n g a n d n o x i o u s t h e r m a l stimuli. None

of the e x p e r i m e n t a l l y - o b s e r v e d drug effects

could be r e p r o d u c e d by the e j e c t i o n of s o d i u m chloride at the m a x i m u m c a t h o d a l or a n o d a l c u r r e n t s used (see Table I).

causes could have b e e n responsible. A similar i n h i b i t i o n of n o n - n o c i c e p t i v e r e s p o n s e s was s h o w n by N K B ; b u t this p e p t i d e shows o n l y m a r g i n a l selectivity for NK-3 over NK-1 receptors 3t'32,5~. T h e highly-selective NK-3 agonist, S E N K 51'75 was w i t h o u t effect (Table I, Fig. 2), suggesting that an NK-1 a n d n o t an NK-3 site was involved. I n contrast, several N K - 2 agonists ( N K A , K A S S a n d D-Pro 9, b u t n o t its s t e r e o i s o m e r , L-Pro 9 (refs. 31, 32, 51)) p r o d u c e d a m a r k e d facilitation of responses to n o x i o u s t h e r m a l stimuli (Figs. 1B a n d 2, Table I). O t h e r n e u r o n a l activity, i n c l u d i n g r e s p o n s e s e v o k e d by

T h e m a i n p h a r m a c o l o g i c a l results are s u m m a r i s e d in

150

Table IV. Text a b b r e v i a t i o n s for c o m p o u n d s are o u t l i n e d in M a t e r i a l s a n d M e t h o d s u n d e r r e c o r d i n g a n d i o n o p h o resis protocols. to

leo

Effects of neurokinin receptor agonists A g o n i s t s with selectivity for NK-1 receptors, SP a n d S P O M e , b u t n o t the inactive free acid S P O H 18'31'32'51, surprisingly caused a selective r e d u c t i o n of responses to i n n o c u o u s b r u s h , b u t were w i t h o u t effect o n responses to n o x i o u s h e a t or pinch o r o n s p o n t a n e o u s activity (Table

Q.

5o "5

100

Span

c

0 c o o r-

• 50

E

•

0

t

I

i

i

i

I

I

i

I

Cl

C2

T1

2

3

4

5

6

7

//// recovery

3min cycles

Fig. 3. Effects of an NK-2 receptor antagonist [o-Pro4,Lys 6, D-TrpT'9'1°,PheH]SP4_11 (applied in the region of the substantia gelatinosa) on sensory responses of SCT neurons. Experiments and data analysis were carried out as described in Fig. 2. The peptide was ionophoresed at 400 nA for all of the 8 neurons shown here. Recovery was clearly apparent for all 5 of the neurons in which it was sought, 14 + 2 min after stopping the ionophoresis, showing that the peptide effects were reversible and not due to non-specific neuronal damage. Responses to noxious heat (O), noxious pinch (r-q), innocuous brush (A) and spontaneous activity (®) are shown as mean + S.E.M. values. Statistically significant differences from controls are indicated (*), P < 0.05 by matched-pair t-test on raw data.

D-Tyr

D-Pro

D-Pro

D-Pro

350nA

250nA

350nA

400hA

NKB-A

Fig. 4. Direct effects of neurokinin receptor antagonists on somatosensory responses of SCT neurons. Ionophoresis of drugs at various currents (400 nA, unless otherwise indicated) was usually carried out for 6 test cycles. Stimulus-evoked activity (or spontaneous activity) was integrated over the relevant epochs (as in Figs. 2 and 3) and expressed as a percentage of the mean of pre-drug control responses. In each experiment, values for each type of response were taken at the time when the greatest change (in any response) occurred (see Table I). For antagonists in the first and last panel (NK-1 and NK-3 antagonists) the point of greatest change in brush responses was recorded. For all other panels (NK-2 antagonists) the time of greatest change in thermal nociceptive responses was taken. Responses to noxious heat ( i ) , noxious pinch (tW), innocuous brush (ll~) and spontaneous activity (D) are shown as mean + S.E.M. values. The different antagonists, ejection currents, time of maximal change in a particular response and numbers of experiments were as follows. Span: [D-Argl,D-TrpT'9,Leull]SP (spantide) at 400 nA for 7 + 1 min (n = 12). D-Tyr: [D-Tyra,D-Trp7'9,NIe11]SP4_11at 350 nA for 12 + 3 min (n = 6). D-Pro: [D-Pro4,LyS6,D-Trp7'9' 1°,Phell]SP4_11 at 250 nA for 12 + 3 min (n = 11), at 350 nA for 8 + 2 min (n = 5) and at 400 nA for 10 _ 3 min (n = 8), respectively. NKB-A: [D-Proz,D-Trp6'8,Nlel°]NKB at 400 nA for 8 + 2 min (n = 4). Statistically significant attenuation of thermal nociceptive resonses is indicated (*) (P < 0.05, matched-pair t-test on raw data).

175 noxious mechanical stimuli, was unaffected. Effects appeared with a relatively rapid onset, using quite moderate ionophoretic currents in the case of NKA particularly, and recovery could be observed 15-30 min after ceasing drug application (Fig. 1B). In approximately 50% of neurons tested with the NK-2 agonists, the effect declined from a maximum after 8-14 min of continued drug administration, consistent with receptor desensitisation. A general excitant amino acid, glutamate, did not mimic the facilitation of thermal nociceptive responses (Table I). Simultaneous administration of SP and NKA, intended to simulate the co-release of these peptides from fine afferent terminals, produced the expected pattern of both NK-1 and NK-2 receptor activation with amplified thermal nociceptive responses and suppressed non-nociceptive responses (Table I). --

150

*

*

Effects of neurokinin receptor antagonists The physiological role of different neurokinin receptors in synaptic transmission of sensory inputs was assessed using selective antagonists. Spantide (SPAN) is a potent antagonist of tachykinin action in tissues with a predominance of NK-1 receptors 3"19 whereas it is virtually inactive at NK-2 receptors in, for example, hamster urinary bladder 3. Spantide itself had no apparent effect on any neuronal responses or on spontaneous activity of SCT neurons (Fig. 4). The putative NK-3 antagonist NKB-A 42'7° was synthesised and tested but was also without effect (Fig. 4). A small number of compounds have been described 58 as effective tachykinin antagonists

/

150 I

/

~

o

Span

SPOMo Span +SPOMe

NKA

Span +NKA

Span +NKB

Fig. 5. Effects of the putative NK-1 receptor-selective antagonist [D-Argl,D-Trp7'9,Leull]SP (spantide) on the modulation of somatosensory responses by agonists. Integrated responses to different sensory stimuli were expressed (as previously) as a percentage of the mean of pre-drug control responses. Responses to noxious heat (11), noxious pinch (~1), innocuous brush ( a ) and spontaneous activity (D) are shown as mean _+ S.E.M. values. In each experiment, values for each type of response were recorded at the time of maximal change in any response (innocuous brush for experiments with spantide alone or in combination with [Met-OMeI1]SP or NKB, and noxious heat for experiments involving NKA). Ionophoresis of agonists or antagonist alone was usually carried out for 6 test cycles. For the combination experiments, antagonist was always applied for at least two 3 min cycles prior to (and then continued during) ionophoresis of the agonist. The different analogues, ejection currents, time of maximal change in responses since starting ionophoresis (of agonist in combination experiments) and number of experiments, were as follows. Span: [D-Argl,D-Trp7'9,LeuN]SP (spantide) at 400 nA for 7 + 1 min (n = 12) (in all cases where applied, spantide was ionophoresed at 400 nA). SPOMe: [MetO M e l l ] s p at 400 nA for 8 + 3 rain (n = 6). Span + SPOMe: [Met-OMeH]S at 400 nA for 6 + 2 min (n = 6). N K A : NKA at 320 + 15 nA for 8 + 2 min (n = 8). Span + N K A : NKA at 350 _+ 20 nA for 10 + 2 min (n = 6). NKB: NKB at 400 nA for 8 + 2 min (n = 7). Span + NKB: NKB at 400 nA for 5 + 2 min (n = 5). Statistically significant alterations from control sensory responses are indicated (*) (P < 0.05, matched-pair t-test on raw data). Significant reversals of agonist effects are indicated ('1") (P < 0.05, Mann-Whitney U-test).

0 D-Pro

SPOMe

D-Pro +SPOMe

NKA

D-Pro +NKA

Fig. 6. Effects of the putative NK-2 receptor-selective antagonist [D-Pro4,Lys6,D-Trp7'9'l°,Phell]sP~_ll on the modulation of somatosensory responses by agonists. Integrated responses to different sensory stimuli were expressed (as previously) as a percentage of the mean of pre-drug control responses. Responses to noxious heat (11), noxious pinch (~1), innocuous brush ( a ) and spontaneous activity ([~) are shown as mean _+ S.E.M. values. In each experiment, values for each type of response were recorded at the time of maximal change in any response (innocuous brush for experiments with [Met-OMelt]SP, and noxious heat for experiments with [o-Pro4,Lys6,D-Trp7'9'l°,Phell]SP4_ll alone or in combination with NKA). Ionophoresis of agonists or antagonist alone was usually carried out for 6 test cycles. For the combination experiments, antagonist was always applied for at least two 3 min cycles prior to (and then continued during) ionophoresis of the agonist. The different analogues, ejection currents, time of maximal change in responses since starting ionophoresis (of agonist in combination experiments) and number of experiments, were as follows. D-Pro: [D-Pro4,Lys6,D-Trp7'9'l°,Phell]Sp4_ll at 250 nA for 10 + 3 min (n = 9) (in all cases where applied, [D-Pro4,Lys6,-D-TrpT'9'l°,Phell]SP4_ll was ionophoresed at 250 nA). SPOMe: [Met-OMell]SP at 400 nA for 8 + 3 min (n = 6). o-Pro + SPOMe: [Met-OMeu]SP at 400 nA for 10 + 2 min (n = 6). N K A : NKA at 320 + 20 nA for 6 + 2 min (n = 6). o-Pro + N K A : NKA at 320 + 20 nA for 8 + 1 min (n = 6). Statistically significant alterations from control sensory responses are indicated (*) (P < 0.05, matched-pair t-test on raw data). Significant reversals of agonist effects are indicated (t) (P < 0.05, Mann-Whitney U-test).

176 TABLE II Effects of intrathecally-administered NKA in (a) tail-flick and (b) hot plate tests Behavioural testing was carried out, as described in Materials and Methods, on male Wistar rats, which had been previously implanted with intrathecal cannulae. Control response latencies in both tail flick and hot plate tests were indistinguishable in unoperated, sham-operated and saline-injected rats and showed no significant changes through the course of experimental testing. All such values were within the range of pre-drug latencies shown in the table. All values are the mean ± S.E.M. from 5-7 experiments. Statistically significant changes from pre-drug response latencies are indicated (*), P < 0.05 by Mann-Whitney U-test. Drug dose (nmol)

Response latency (s) Pre-drug

Post-drug (time in min) 5

10

20

(a) Tail flick test

1 3 5

6.7±0.4 6.7 + 0.4 7.1 ±0.2

7.0±0.6 6.1 + 0.7 9.8± 1,6

7.1 ±0.5 3.5 ± 0.2* >15

6.8+0.7 6.9 ± 0.2 >15

(b) Hot plate test

1 3 5

2.0±0.2 2.1±0.1 2.0±0.3

2.1±0.2 0.8±0.1" 2.0±0.2

2.0±0.1 1.5±0,3 3.7±1.0

2.1±0.2 2.0±0.2 >6

at NK-2, but not NK-1 receptors. Two of these, D-Tyr4 and D-Pro 4 were synthesised and tested, resulting in a marked inhibition of thermal nociceptive responses of SCT neurons. The effect was highly selective since other activity was quite unaffected, its magnitude was dependent on ionophoretic current, and recovery from it could be demonstrated (Figs. 3 and 4). o-Pro 4 is also an effective antagonist of bombesin 58, we tested the latter peptide to see whether it could reproduce the facilitation of thermal nociceptive responses shown by NK-2 agonists. Bombesin had no apparent effect (Table I) suggesting that the effects of D-Tyr4 and D-Pro 4 (the inverse of NK-2 agonist action), were indeed due to NK-2 receptor

Interactions o f neurokinin receptor agonists and antagonists Experiments were carried out to confirm the agonist evidence that the effects of SPOMe and N K A were mediated by NK-1 and NK-2 receptors respectively. The NK-1 antagonist SPAN, readily reversed the inhibitory effect of SPOMe and indeed that of NKB, on nonnociceptive n e u r o n a l responses (Fig. 5). In view of the inactivity of the potent and highly-selective NK-3 agonist SENK, it thus seems likely that the observed action of NKB, as well as that of [Met-OMeH]SP, are mediated by NK-1 receptors. Spantide did not affect the amplification by N K A of noxious thermal responses, indicating that N K A action was through a SPAN-resistant site such as the NK-2 receptor (Fig. 5). The putative NK-2-selective

blockade.

TABLE III Effects of intrathecally-administered [o-PrJ, L ys6, D-Trp7'9.io, Phel l] S P 4 ii in (a) tail-flick and (b) hot plate tests Behavioural testing was carried out, as described in Materials and Methods, on male Wistar rats, which had been previously implanted with intrathecal cannulae. Control response latencies in both tail flick and hot plate tests were indistinguishable in unoperated, sham-operated and saline-injected rats and showed no significant changes through the course of experimental testing. All such values were within the range of pre-drug latencies shown in the table. All values are the mean + S.E.M. from 5-7 experiments. Statistically significant changes from pre-drug response latencies are indicated (*), P < 0.05 by Mann-Whitney U-test. Drug dose (nmol)

Response latency (s) Pre-drug

(a) Tail flick test

2.5 5.0 7.5

6.4±0.3 6.3±0.3 7.4±0.6

(b) Hot plate test

2.5 5.0 7.5

2.1±0.2 2.0±0.1 2.2±0.2

Post-drug (time in min) 10

30

60

6.7 ± 0.6 10.6 ± 1.4* 11.8 ± 0.35*

6.5 ± 0.3 14.0 ± 0.7* >15

6.2 + 0.5 7.0 ± 0.2 >15

2.0±0.2 4.5±0.2* >6

2.1±0.2 2.1±0.2 >6

2.1±0.2 2.4±0.2 4.7±0.6

177 TABLE IV Summary of the main influences of tachykinins observed on the somatosensory responses of SCTneurons

The data summarised is displayed in detail in Figs. 4-6 and Table I. See text for compound abbreviations. The indicated activity of the compounds at various NK receptors is not intended to imply that this is their exclusive action, merely that which is probably involved here (see text for details). Observation

Effective compounds

Predominant activity at NK receptors

(a) Inhibition of responses to innocuous brush:

SP SPOMe NKB SPAN

NK-1 agonist

NKA KASS D-Pro 9 D-Pro 4

NK-2 agonist NK-2 agonist NK-2 agonist NK-2 antagonist

(c) Inhibition of responses to noxious heat:

D-Pro4 D-Tyr4

NK-2 antagonist NK-2 antagonist

(d) No effect:

SPOH

little weak NK-1 agonist NK-3 agonist NK-1 antagonist NK-3 antagonist

- reversal by: (b) Facilitation of responses to noxious heat: - reversal by:

L-Pro 9

SENK SPAN NKB-A

NK-1 agonist NK-3 (NK-1) agonist NK-1 antagonist

antagonist D-Pro 4 (when applied at currents selected to give no overt effect alone), significantly attenuated the effect of N K A without any alteration in the effect of SPOMe (Fig. 6). Therefore both agonist and antagonist actions were consistent with their receptor selectivity reported in the literature, reinforcing the conclusion that NK-2 and NK-1 receptors are distinctly involved in the two phenomena observed here. Behavioural experiments

In the two tests of thermal nociception here, intrathecally-administered N K A caused transient decreases in response latency (Table II), consistent with the facilitation of thermal nociception that it produced in the electrophysiological experiments. Other workers have described a similar effect of N K A in the tail flick test 6'22. Effects of 3 nmol N K A in both tail flick and hot plate tests were transient and full recovery occurred within 20 min (Table II). Doses of 5 nmol and above invariably caused flaccid paralysis. No irritation-like behaviour or vocalisation was apparent at the doses tested. The inverse effect was seen with the NK-2 antagonist t~-Pro4, that is an increased response latency in tail flick and hot plate tests (Table III). This matches the reduced responsiveness of single neurons to noxious thermal stimuli seen during ionophoresis of the same antagonist.

Effects in the behavioural tests were transient with full recovery, but by a dose of 7.5 nmol, flaccid paralysis was evident. Some experiments were also carried out with the NK-1 antagonist SPAN which has been described to cause a mild behavioural analgesia in the tail flick test 56. In our experiments, doses of SPAN as low as 0.5 nmol resulted in all 4 animals tested in signs of profound distress with vocalisation and hind limb irritation, so testing was discontinued, despite no changes in reflex latencies being observed. DISCUSSION

The present results describe effects of tachykinins on the processing of somatosensory inputs to SCT neurons. Whilst there is need for caution in extrapolating to other populations of nociceptive dorsal horn neurons, our observation of similar effects (of NK-2 agonists and antagonists at least) on unidentified lamina IV/V neurons of cat and rat lead us to suspect that (as seen previously with other neuropharmacological agents 14A7) the observed phenomena are of general importance. The behavioural experiments clearly support this contention. The results demonstrate two quite distinct effects of tachykinins in the region of the superficial dorsal horn on the processing of somatosensory inputs. Firstly, responses of SCT neurons to innocuous cutaneous stimuli were selectively attenuated by SP (Table I). The effect was replicated in the majority of cases by the selective NK-1 agonist SPOMe 1s'32 suggesting that it is mediated by an NK-1 receptor. A similar effect of NKB, but not the highly-selective NK-3 agonist SENK is consistent with action at an NK-1 site 6'7'2°. The effect of SPOMe was reversed by the NK-1 antagonist SPAN (Fig. 5), which is virtually inactive at NK-2 sites 3A9, whereas an NK-2 antagonist D-Pro 4 (refs. 29, 30) was ineffective (Fig. 6). Spantide alone had no effect (Fig. 4), suggesting that there is little tonic activation of the NK-1 receptor that responds to exogenous agonists. Interestingly, none of the sensory responses were attenuated by SPAN, indicating that an endogenous tachykinin whose action is susceptible to an NK-1 antagonist (for example, SP), is not mediating any of the sensory inputs examined in the present experimental conditions. It is possible of course, that relevant sites were not accessed even though exogenous agonist actions could be reversed. Indeed there is evidence that beyond the NK-1 influence on non-nociceptive inputs reported here, there is a further NK-1 site of relevance in deeper dorsal horn. Substance P ionophoresed close to deeper dorsal horn neurons generally causes e x c i t a t i o n 23'53'57'64'79, but sometimes more complex mixed effects 7'41"6°'73. Neurokinin A and

178 NKB also cause excitation when administered close to deeper dorsal horn neurons, both in vivo 24'55 and in vitro 61, but this action may be via an NK-1 site since these neurokinins both display considerable affinity at NK-1 receptors as well as at NK-2 and NK-3 sites, respectively3~'32'51. Excitation is also seen with the partially selective NK-1 agonist physalaemin 32'51, both in vivo 63 and in vitro 6~. We have recently recorded from identified SCT neurons using one barrel of a multibarrel pipette (ionophoresing drugs in close proximity to the neurons) and found consistent general excitation in 5 out of 5 cells with the selective NK-1 agonist SPOMe. The weight of evidence therefore suggests that the predominant tachykinin receptor, on or near deeper dorsal horn neurons which mediates excitation, is of the NK-1 type. Interestingly, fine afferent fibres (certainly of the A6 type, but perhaps also C) send a small number of collaterals to this area 52, so it remains possible that this NK-1 receptor may mediate responses to SP from primary afferents. The present experiments, however, underline a modulatory role for NK-1 receptors in superficial dorsal horn, presumably responding to SP which may be released not only from afferents but potentially even from local or descending neurons. The location of receptors mediating selective influences on sensory inputs cannot be discerned from these experiments. It seems most likely that interneurons intercalated between primary afferent terminals and the SCT neurons may be the relevant sites. It is possible that the receptors could be restricted to particular zones of dorsally-extending SCT dendrites, but we have no evidence for this and it is clear that tachykinins applied close to SCT somata have quite different non-selective excitatory effects. A diminution in the non-nociceptive responsiveness of cells as seen here, has overtly very little connection with pain or analgesia. Nevertheless, it would bias the responsiveness of multireceptive cells towards them becoming predominantly nociceptive-selective cells. If supraspinal sites are able to decode this, perhaps by interpretation of parallel processing or frequency codes, the overall influence of SP (and indeed also, as described below, NKA) in the superficial dorsal horn would be pro-nociceptive. The second major observation of this study (which would also have an overall pro-nociceptive influence on the responsiveness of multireceptive cells) was the selective facilitation by N K A of responses to noxious thermal stimuli (Fig. 1, Table I). Other activity and sensory responses (even to noxious mechanical stimuli) were unaffected, indicating an action at some site restricted to the processing of inputs from thermal nociceptive afferents. A similar effect to that of NKA was seen with NK-2 agonists KASS 51 and D-Pro9 but not L-Pro9 (Fig. 2,

Table I). The potency ratio of D-PrO9:L-Pro 9 is 0.01--0.2 at NK-1 sites, 200-300 at NK-2 sites and 3-25 at NK-3 sites 18'32. The marked activity of the o-Pro 9 here, with inactivity of L-Pro9, is inconsistent with an NK-1 receptor being involved, but consistent with an NK-2 (or perhaps NK-3 site). Since, however, N K A (which also facilitated thermal nociceptive responses) is several-fold more potent than D-Pro 9 at NK-2 sites whereas NKB and SENK (which had no effect on thermal nociceptive responses) are at least 500-fold more potent than D-Pro 9 at NK-3 sites 18'32'51, it seems probable that an NK-2 receptor mediates the effect. Although L-Pro9 is an NK-1 agonist, it is of lower potency than SP, SPOMe or NKB 18'32, consistent with its failure to reproduce the attenuation of non-nociceptive responses seen here with those agonists. Similarly, although NKA does have significant potency at NK-1 sites 18'32, it is less than that of any of the agonists that were effective attenuators of non-nociceptive responses. Studies with antagonists support the idea that an NK-2 receptor is responsible for the facilitation of thermal nociceptive responses caused by NKA. Spantide, an antagonist at NK-1 but not NK-2 sites 3'19 caused no change in the effect of NKA (Fig. 5). Using relatively low ejection currents, an NK-2-selective antagonist o-Pro 4 (refs. 58, 59), however, successfully attenuated the effect of N K A (Fig. 6). This antagonist shows more than 14-fold greater affinity at NK-2 than NK-3 sites 59. Taken in conjunction with the data on selectivity of different agonists in facilitating thermal nociceptive responses, the effect seems most likely to be mediated by an NK-2 receptor. Interestingly, when ionophoresed at rather higher currents, both o-Pro 4 and another NK-2 antagonist D-Tyr4 (refs. 58, 59), were alone able to attenuate thermal nociceptive responses (Figs. 3 and 4). Effects were reversible and dependent in magnitude on ionophoretic current, indicating a degree of specificity. Like D-Pro4, D-Tyr4 is a more effective antagonist at NK-2 than at NK-3 sites, but it can variably act as an agonist or antagonist at NK-1 sites in different tissues 58'59. The inhibition of thermal nociceptive responses may well therefore be due to blockade of NK-2 receptors occupied during the response by release of an endogenous NK-2 agonist such as NKA. Such a contention is supported circumstantially by several pieces of evidence: (i) within this tissue, D-Pro 4 effectively prevented the action of NK-2, but not NK-1 agonists; (ii) the two NK-2 antagonists acted similarly on thermal nociceptive responses whereas they have divergent effects in some NK-1 receptor systems; (iii) the effect of the NK-2 antagonists is precisely (and very selectively) the inverse of the effect of NK-2 agonists. Furthermore since SPAN failed to

179 affect sensory responses, despite readily reversing SPAN action, it seems most unlikely that NK-1 receptors in superficial dorsal horn are involved in endogenous transmission of nociceptive inputs. Any involvement of NK-3 receptors in the action of the NK-2 antagonists cannot be excluded with certainty, but the low affinity of the antagonists for NK-3 sites, the lack of mimicry an NK-3 antagonist NKB-A 42'7° (see Fig. 4) and the failure of NK-3 agonists to produce any semblance of an inverse effect (Table I, Fig. 3), all suggest that NK-3 sites are not relevant here. It is of course possible that the NK-2 antagonist action on thermal nociceptive responses is not actually due to neurokinin receptor blockade, but to other actions. It is known that D-Pro4 is an effective antagonist of bombesin 58, but when ionophoresed into superficial dorsal horn here (Table I) bombesin produced no effect such as excitation or facilitation of responses (the antagonism of which would be consistent with the reduction in thermal nociceptive responses seen with D-Pro4). A number of tachykinin antagonists release histamine 58'65, although the activity of octapeptide analogues (such as the two NK-2 antagonists used here) tends to be rather low 2°'58. Trial experimen'cs showed, however, that ionophoretic administration of histamine was without effect here (3 out of 3 cells). An unequivocal confirmation that an NK-2 receptor is involved in mediating endogenous thermal nociception here, will only come with the use of a range of highly selective antagonists; for example, L659,877 and L659,874 which show greater than 100-fold selectivity for NK-2 over other neurokinin receptors 43. Behavioural experiments supported the electrophysiological data by showing that intrathecal NKA reduced response latency in two reflex tests of thermal nociception (Table II). This effect of NKA accords with similar effects on tail-flick latency described by two other groups 6'22. The receptor type involved, clearly cannot be interpreted from our data, but the described agonist potency ranking 4'6'4°'51 seems most consistent with an NK-1 site, although antagonist experiments 5 suggest that the site may not present the classical NK-1 profile. Ionophoretic experiments on rat trigeminal neurons also show failure to reverse SP actions by peripherallyeffective NK-1 antagonists 25. The receptor type involved in tachykinin-induced hindlimb scratching and irritation behaviour seems most likely to be of the NK-1 type 4' 9,22,29.51,54,66, but the relation of this phenomenon to endogenous nociception is unclear and has recently been questioned 21. Interestingly, intrathecally-injected NK-3 agonists are reported to cause increases in tail-flick latency, followed closely by flaccid paralysis 4'51. This supports our conclusion that the reduction in thermal nociception caused by D-Pro4 in both electrophysiological

and behavioural experiments, is more probably due to NK-2 rather than NK-3 receptor blockade. With multiple sites of action possible in these behavioural experiments (for example, motoneurons and premotoneurons), it is of course always possible that heterogenous profiles may arise from the contribution of several receptor types. Our experiments also demonstrated increased response latency due to D-Pro4 in the tail flick and hot plate tests (Table III). This supports the idea that an NK-2 receptor may be involved in permitting endogenous thermal nociceptive responsiveness. In our hands, the NK-1 antagonist SPAN could not be adequately tested and since no further analogues were investigated, no analysis of the receptor type involved can be made from our behavioural experiments. A variety of putative antagonists have been tested by other groups, some of which prolong response latency in these tests 54'56'62'69. In general, no clear picture has emerged, with some 56'62, but not all56 NK-1 antagonists being effective. Varying side-effects of the different analogues may be responsible. In just one report 7°, a putative NK-3 antagonist increased hot-plate-, but not tail-flick-response latency, but its effects were delayed and long-lasting and its receptor specificity has not been fully characterized. Nevertheless, a recent report by Dourish et al. 9 indicates that the highly selective NK-2 antagonist L659,837 and another NK-2 selective compound U-67,202 (ref. 54) increased response latency in hot plate and tail flick tests, whereas highly selective NK-1 antagonists were ineffective. The results here with NK-2 agonists and antagonists, are consistent with the idea that NKA released from primary afferents may serve a physiological role in either mediating or promoting thermal nociception. NK-2 agonists, however, did not themselves increase spontaneous activity of the cells, suggesting either that other substances and inputs, in addition, may be necessary to evoke a full thermal nociceptive response, or that not all of the appropriate sites are being accessed here by sufficient concentrations of agonist. It may well be that other substances, such as neuropeptides and amino acids, perhaps even released from the same afferents and interacting with the influence of tachykinins, may have a role in nociception. The responses to the noxious mechanical stimuli used here will always include some activity evoked in low threshold non-nociceptive fibres, so any influence on nociceptive inputs may be less marked on mechanical than thermal responses. Responses to noxious pinch were slightly, but not significantly affected by NK-1 agonists, which had clearly reduced non-nociceptive responses (Fig. 2), suggesting that there was not a very large non-nociceptive component in the response to pinch.

180 The selective involvement of dorsal horn NK-2 receptors with thermal but not mechanical nociception is u n e x p e c t e d since the p r e d o m i n a n t population of C fibre nociceptors in the cat r e s p o n d regularly to thermal as well as to mechanical stimuli and the majority of A 6 nociceptors either fail to respond or require sensitisation by r e p e a t e d noxious thermal stimuli to do so 52. Nevertheless, mechano-specific C nociceptors have been described 3°, and experiments on afferent peptide release t°' 35, spinal reflexes 72 and effects of the primary afferent neurotoxin, capsaicin, on behavioural reflex responses 44 suggest that thermal and mechanical nociceptive inputs m a y differ, at least partly, at the biochemical level. It seems possible that the large group of polymodal nociceptive afferents m a y be heterogeneous, containing sub-groups with connectivity and perhaps biochemical differences that can be at least relatively selectively activated by mechanical/thermal stimuli. A n o t h e r consideration is that the mechanical but not thermal noxious stimuli (Fig. 1A) may recruit A 6 high threshold mechanoreceptors to a significant extent. A subpopulation of dorsal root ganglion cells with A d conduction velocities do contain SP-like immunoreactivity 37 and the observation of SP release in spinal cord by mechanical but not thermal nociceptive stimuli 35 could be taken to suggest that the released SP originates from A d mechanoreceptive rather than p o l y m o d a l C nociceptive afferents. T h e r e may of course be o t h e r explanations. In contrast to SP, N K A is readily released by both mechanical and thermal noxious stimuli 28 (see below) and may therefore be more closely associated with transmission from polymodal nociceptors. A n u m b e r of experiments have shown the release of tachykinins into the spinal cord induced by stimulation of afferents at high intensity, or by noxious cutaneous stimuli t°'28'35"36'76. Using p u s h - p u l l cannulae, Kuraishi et

al. described the release of SP-like immunoreactivity by noxious mechanical, but not thermal stimuli 35. H o w e v e r , the use of a n t i b o d y - c o a t e d microelectrodes and thermal stimuli m o r e certainly into the noxious range, e n a b l e d Duggan et al. to d e m o n s t r a t e that SP-like immunoreactivity was released during noxious heating of the skin 1°. Since SP in the dorsal horn arises only partly from afferents 48, it is difficult to fully discern its role as a primary afferent transmitter from these experiments. R e c e n t studies in our laboratories 28 have investigated the release of neurokinin-like immunoreactivity using probes coated with antiserum directed against N K A . Noxious thermal (46-47 °C) and noxious mechanical stimuli, e v o k e d the release of N K A - l i k e immunoreactivity into dorsal horn. Significant basal levels of N K A - l i k e immunoreactivity and persistence of released material were also observed. In view of the present results, it appears that N K A (or N P K ) m a y be one of the substances released from p o l y m o d a l nociceptors that is necessary for the transmission of thermal nociception in superficial dorsal horn. It is not clear w h e t h e r or not they alone are sufficient to elicit a response to activity of p o l y m o d a l C-afferents in s e c o n d - o r d e r neurons. N K A and N P K are present at much lower levels than Spl'aS"but a p p e a r to be m o r e directly relevant to nociception. Nevertheless, both N K A / N P K and SP may well act in concert in superficial dorsal horn, to facilitate nociceptive responsiveness and to attenuate non-nociceptive responsiveness, respectively.

Acknowledgements. This work was supported by the Wellcome Trust, the AFRC, ICI Pharmaceuticals plc and the MRC. Thanks to Leslie Iversen, Arthur Duggan, Ainsley Iggo and Norrie Russell for helpful comments and to Merck, Sharp and Dohme and ICI for gifts of peptides. Also thanks to Ms. A. Leask for technical assistance, the Wellcome Animal Unit for facilities, and to Celia Leitch for typing the manuscript.

REFERENCES 1 Arai, H. and Emson, P.C., Regional distribution of neuropeptide K and other tachykinins (neurokinin A, neurokinin B and substance P) in rat central nervous system, Brain Research, 399 (1986) 240-249. 2 Brown, J.R., Calthrop, J.C., Hancock, A.B. and Jordan, C.C., Studies with tachykinin antagonists on neuronal preparations in vitro. In R. Hakanson and E Sundler (Eds.), Tachykinin Antagonists, Elsevier, Amsterdam, 1985, pp. 355-365. 3 Brown, J.R., Jordan, C.C., Ward, P. and Whittington, A.R., Selectivity of tachykinin antagonists on smooth muscle preparations. In R. Hakanson and E Sundler (Eds.), Tachykinin Antagonists,'Elsevier, Amsterdam, 1985, pp. 305-312. 4 Couture, R., Launeville, O. and Dorais, J., Characterization of neurokinin receptors in a spinal nociceptive reflex of the rat by the use of selective agonists, Regul. Peptides, 22 (1988) 52. 5 Couture, R., Yashpal, K., Henry, J.L., Escher, E. and Regoli, D., Characterization of spinal actions of four substance P analogues, Eur. J. Pharmacol., 110 (1985) 63-69. 6 Cridland, R.A. and Henry, J.L., Comparison of the effects of

7 8 9

10

11

substance P, neurokinin A, physalaemin and eledoisin in facilitating a nociceptive reflex in the rat, Brain Research, 381 (1986) 93-99. Davies, J. and Dray, A., Depression and facilitation of synaptic responses in cat dorsal horn by substance P administered into substantia gelatinosa, Life Sci., 27 (1980) 2037-2042. Doi, T. and Jurna, J., Intrathecal substance P depresses the tail flick response - - antagonism by naloxone, Naunyn-Schmiedeberg's Arch. Pharmacol., 317 (1981) 135-139. Dourish, C.T., Clark, M.L., Hawley, D., Williams, B.J. and lversen, S.D., Antinociceptive effects of novel, selective tachykinin receptor antagonists in thermal and chemical analgesia tests, Regul. Peptides, 22 (1988) 58. Duggan, A.W., Morton, C.R., Zhao, Z.Q. and Hendry, I.A., Noxious heating of the skin releases immunoreactive substance P in the substantia gelatinosa of the cat: a study with antibody microprobes, Brain Research, 403 (1987) 345-349. Eberle, A.N., Atherton, E., Dryland, A. and Sheppard, R.C., Peptide synthesis. Part 9: solid phase synthesis of melaninconcentrating hormone using a continuous flow polyamide method, J. Chem. Soc. (Perkin Trans.), 1 (1986) 361.

181 12 Fleetwood-Walker, S.M., Hope, P.J., Mitchell, R. and EIYassir, N., Effects of agonists selective for NK-1, NK-2 and NK-3 receptors on somatosensory transmission in superficial dorsal horn, Regul. Peptides, 22 (1988) 67. 13 Fleetwood-Walker, S.M., Hope, P.J., Mitchell, R., Molony, V. and EI-Yassir, N., Effects of tachykinins on sensory processing in the superficial dorsal horn of the cat, J. Physiol. (Lond.), 382 (1987) 67P. 14 Fleetwood-Walker, S.M., Mitchell, R., Hope, P.J., Molony, V. and Iggo, A., An a 2 receptor mediates the selective inhibition by noradrenaline of nociceptive responses of identified dorsal horn neurones, Brain Research, 334 (1985) 243-254. 15 Fleetwood-Walker, S.M., Mitchell, R., Hope, P.J., El-Yassir, N. and Molony, V., The roles of tachykinin and opioid receptor types in nociceptive and non-nociceptive processing in superficial dorsal horn. In R.E Schmidt, H.-G. Schaible and C. Vahle-Hinz (Eds.), Fine Afferent Nerve Fibres and Pain, VCH, Weinheim, F.R.G., 1987, pp. 239-247. 16 Fleetwood-Walker, S.M., Mitchell, R., Hope, P.J., EI-Yassir, N. and Molony, V., Neurokinin A as a transmitter of primary afferents involved in thermal nociception. In J.L. Henry, R. Couture, A.C. Cuello, G. Pelletier, R. Quirion and D. Regoli (Eds.), Substance P and Neurokinins, Springer, New York, 1987, pp. 291-293. 17 Fleetwood-Walker, S.M., Hope, P.J., Mitchell, R., EI-Yassir, N. and Molony, V., The influence of opioid receptor subtypes on the processing of nociceptive inputs in the spinal dorsal horn of the cat, Brain Research, 451 (1988) 213-226. 18 Fletcher, A.E., McKnight, A.T. and Maguire, J.J., The use of [Glp6,L-Pro9]SP6_ll and [GIp6,D-Pro9]SP6_ll to distinguish receptors for tachykinins, Br. J. Pharmacol., 90 (1987) 266P. 19 Folkers, K., Rosell, S., Hakanson, R., Chu, J.-Y., Lu, L.-A., Leander, S., Tang, P.-EL. and Ljungquvist, A., Chemical design of antagonists of substance P. In R. Hakanson and E Sundler (Eds.), Tachykinin Antagonists, Elsevier, Amsterdam, 1985, pp. 259-266. 20 Foreman, J.C., Jordan, C.C., Oehme, P. and Renner, H., Structure-activity relationships for some substance P-related peptides that cause wheal and flare reactions in human skin, J. Physiol. (Lond.), 335 (1983) 449-465. 21 Frenk, H., Bossut, D., Urca, G. and Mayer, D.J., Is substance P a primary afferent transmitter for nociceptive input? I. Analysis of pain-related behaviours resulting from intrathecal administration of substance P and six excitatory compounds, Brain Research, 455 (1988) 223-231. 22 Gamse, R. and Saria, A., Nociceptive behaviour after intrathecal injections of substance P, neurokinin A and calcitonin gene-related peptide in mice, Neurosci. Lea., 70 (1986) 143-147. 23 Henry, J.L., Effects of substance P on functionally-identified units in cat spinal cord, Brain Research, 114 (1976) 439-451. 24 Henry, J.L. and Salter, M.W., Neurokinin A excites nociceptive and non-nociceptive dorsal horn neurones in the cat spinal cord. In J.L. Henry, R. Couture, A.C. Cuello, G. Pelletier, R. Quirion and D. Regoli (Eds.), Substance P and Neurokinins, Springer, New York, 1987, pp. 279-281. 25 Hill, R.G., Salt, T.E. and Leighton, G.E., Actions of putative substance P antagonists on neurones in the CNS. In R. Hakanson and E Sundler (Eds.), Tachykinin Antagonists, Elsevier, Amsterdam, 1985, pp. 377-382. 26 H6kfelt, T., Johansson, O., Ljundahl, A. and Schultzberg, M., Peptidergic neurones, Nature (Lond.), 284 (1980) 515-521. 27 HOkfelt, T., Vincent, S., Hellsten, L., Rosell, S., Folkers, K., Markey, K., Goldstein, M. and Cuello, C., Immunohistochemical evidence for a neurotoxic action of [D-Pro2,o-Trp 7'9] substance P, an analogue with substance P antagonistic activity, Acta Physiol. Scand., 113 (1981) 571-573. 28 Hope, P.J., Duggan, A.W., Jarrott, B., Schaible, H.-G. and Fleetwood-Walker, S.M., Release, spread and persistence of immunoreactive neurokinin A in the dorsal horn of the cat following noxious cutaneous stimulation: studies with antibody

microprobes, J. Neurosci., in preparation. 29 Hylden, J.L.K. and Wilcox, G.L., Intrathecal substance P elicits a caudally-directed biting and scratching behaviour, Brain Research, 217 (1981) 212-215. 30 Iggo, A., Activation of cutaneous nociceptors and their actions on dorsal horn neurons, Adv. Neurol., 4 (1974) 1-9. 31 Iversen, L.L., Multiple receptors for tachykinins. In R. Hakanson and F. Sundler (Eds.), Tachykinin Antagonists, Elsevier, Amsterdam, 1985, pp. 291-304. 32 Iversen, L.L., Foster, A.C., Watling, K.J., McKnight, A.T., Williams, B.J. and Lee, C.M., Multiple receptors and binding sites for tachykinins. In J.L. Henry, R. Couture, A.C. Cuelio, G. Pelletier, R. Quirion and D. Regoli (Eds.), Substance P and Neurokinins, Springer, New York, 1987, pp. 40-43. 33 Kimura, A., Okada, M., Sugita, Y., Kanazawa, I. and Munekata, E., Novel neuropeptides, neurokinin ct and r, isolated from porcine spinal cord, Proc. Jap. Acad., 59(B) (1983) 101-104. 34 Krause, J.E., Carter, M.S., Xu, Z.S. and MacDonald, M.R., Structure, expression and regulation of the rat preprotachykinin I gene. In J.L. Henry, R. Couture, A.C. Cuello, G. Pelletier, R. Quirion and D. Regoli (Eds.), Substance P and Neurokinins, Springer, New York, 1987, pp. 5-7. 35 Kuraishi, Y., Hirota, N., Sato, Y., Hino, Y., Satoh, M. and Takagi, H., Evidence that substance P and somatostatin transmit separate information related to pain in the spinal dorsal horn, Brain Research, 325 (1985) 294-298. 36 Kuraishi, Y., Hirota, N., Sugimoto, M., Satoh, M. and Takagi, H., Effects of morphine on noxious-stimuli-induced release of substance P from rabbit dorsal horn in vivo, Life Sci., 33, Suppl. 1 (1983) 693-696. 37 Lawson, S.N., Waddell, P.J. and McCarthy, P.W., A comparison of the electrophysiological and immunocytochemical properties of rat dorsal root ganglion fibers with A and C fibers. In R.F. Schmidt, H.-G. Schaible and C. Vahle-Hinz (Eds.), Fine Afferent Nerve Fibers and Pain, VCH, Weinheim, ER.G., 1987, pp. 193-203. 38 Lembeck, E and Donnerer, J., Time course of capsaicin-induced functional impairments in comparison with changes in neuronal substance P content, Naunyn-Schmiedeberg's Arch. Pharmacol., 316 (1981) 240-243. 39 Lynn, B. and Baranowski, R., A comparison of the relative numbers and properties of cutaneous nociceptive afferents in different mammalian species. In R.E Schmidt, H.-G. Schaible and C. Vahle-Hinz (Eds.), Fine Afferent Nerve Fibers and Pain, VCH, Weinheim, ER.G., 1987, pp. 85-94. 40 Moochhala, S.M. and Sawynok, J., Hyperalgesia produced by intrathecal substance P and related peptides: desensitization and cross-desensitization, Br. J. Pharmacol., 82 (1984) 381-388. 41 Murase, K. and Randic, M., Actions of substance P on rat spinal dorsal horn neurones, J. Physiol. (Lond.), 346 (1984) 203-217. 42 Murray, C.W., Cowan, A., Wright, D., Vaught, J.L. and Jacoby, H., Evidence for three distinct peripheral neurokinin receptors using a putative neurokinin B antagonist. In J.L. Henry, R. Couture, A.C. Cuello, G. Pelletier, R. Quirion and D. Regoli (Eds.), Substance P and Neurokinins, Springer, New York, 1987, pp. 149-151. 43 McKnight, A.T., Maguire, J.J., Williams, B.J., Foster, A.C., Tridgett, R. and Iversen, L.L., Pharmacological specificity of synthetic peptides as antagonists at tachykinin receptors, Regul. Peptides, 22 (1988) 127. 44 Nagy, J.I. and Van Der Kooy, D., Effects of neonatal capsaicin treatment on nociceptive thresholds in the rat, J. Neurosci., 3 (1983) 1145-1150. 45 Nawa, H., Kotani, H. and Nakanishi, S., Tissue-specific generation of two preprotachykinin mRNAs from one gene by alternative RNA splicing, Nature (Lond.), 312 (1984) 729-734. 46 Nicoll, R.A., Schenker, C. and Leeman, S.E., Substance P as a transmitter candidate, Annu. Rev. Neurosci., 3 (1980) 227268.

182 47 O'Donohue, T.L., Helke, C.J., Burcher, E., Shults, C.W. and Buck, S.H., Autoradiographic distribution of binding sites for iodinated tachykinins in the rat central nervous system. In J.L. Henry, R. Couture, A.C. Cuello, G. Pelletier, R. Quirion and D. Regoli (Eds.), Substance P and Neurokinins, Springer, New York, 1987, pp. 93-95. 48 Ogawa, T., Kanazawa, I. and Kimura, S., Regional distribution of substance P, neurokinin A and neurokinin B in rat spinal cord, nerve roots and dorsal root ganglia, and the effects of dorsal root or spinal transection, Brain Research, 359 (1985) 152-157. 49 Otsuka, M. and Yanagisawa, M., The effects of substance P and baclofen on motoneurones of isolated spinal cord of the newborn rat, J. Exp. Biol., 89 (1980) 201-214. 50 Otsuka, M. and Yanagisawa, M., Effect of tachykinin antagonist on a nociceptive reflex in the isolated spinal cord-tail preparation of the newborn rat, J. Physiol. (Lond.), 395 (1988) 255-270. 51 Papir-Kricheli, D., Frey, J., Laufer, R., Gilon, C., Chorev, M., Selinger, Z. and Devor, M., Behavioural effects of receptorspecific substance P agonists, Pain, 31 (1987) 263-276. 52 Perl, E.R., Characterisation of nociceptors and their activation of neurons in the superficial dorsal horn: first steps for the sensation of pain, Adv. Pain Res. Ther., 6 (1984) 23-51. 53 Piercey, M.E, Einspahr, EJ., Dobry, P.J.K., Schroeder, L.A. and Hollister, R.P., Morphine does not antagonize the substance P mediated excitation of dorsal horn neurones, Brain Research, 186 (1980) 421-434. 54 Piercey, M.E, Moon, M.W., Blinn, J.R. and Dobry-Schreur, P.J.K., Analgesic activities of spinal cord substance P antagonists implicate substance P as a neurotransmitter of pain sensation, Brain Research, 385 (1986) 74-85. 55 Pini, A. and Ryall, R.W., The action of neurokinins on dorsal horn neurones in anaesthetised rats and cats, J. Physiol. (Lond.), 384 (1987) 68P. 56 Post, J. and Folkers, K., Behavioural and antinociceptive effects of intrathecally injected substance P analogues in mice, Eur. J. Pharmacol., 113 (1985) 335-342. 57 Randic, M. and Miletic, V., Effect of substance P in cat dorsal horn neurones activated by noxious stimuli, Brain Research, 128 (1977) 164-169. 58 Regoli, D., D'Orleans-Juste, P., Drapeau, G., Dion, S. and Escher, E., Pharmacological characterisation of substance P antagonists. In R. Hakanson and E Sundler (Eds.), Tachykinin Antagonists, Elsevier, Amsterdam, 1985, pp. 277-287. 59 Regoli, D. Drapeau, G., Dion, S. and D'Orleans-Juste, P., Receptors for neurokinins in peripheral organs. In J.L. Henry, R. Couture, A.C. Cuello, G. Pelletier, R. Quirion and D. Regoli (Eds.), Substance P and Neurokinins, Springer, New York, 1987, pp. 99-107. 60 Ryall, R.W. and Pini, A., A depressant action of substance P on spinal nociceptive neurons in the rat and cat. In J.L. Henry, R. Couture, A.C. Cuello, G. Pelletier, R. Quirion and D. Regoli (Eds.), Substance P and Neurokinins, Springer, New York, 1987, pp. 239-240. 61 Ryu, P.D., Jeftinija, S. and Randic, M., Action of tachykinins on rat spinal dorsal horn neurones, Soc. Neurosci. Abstr., 12 (1986) 153. 62 Sakurada, T., Kuwahara, H., Sakurada, S., Kirara, K., Ohba, M. and Munekata, E., Behavioural assessment of substance P antagonists in mice, Neuropeptides, 9 (1987) 197-206. 63 Salter, M.W. and Henry, J.L., A depressant action of substance P on spinal nociceptive neurons in the rat and cat, Soc. Neurosci.

Abstr., 11 (1985) 968. 64 Sastry, B.R., Substance P effects on spinal nociceptive neurones, Life Sci., 24 (1979) 2169-2178. 65 Schultheiss, H. and Horig, J., Histamine release from rat peritoneal cells by substance P antagonists and fragments: a study on structure-activity relationship. In R. Hakanson and E Sundler (Eds.), Tachykinin Antagonists, Elsevier, Amsterdam, 1985, pp. 413-420. 66 Seybold, V.S., Hylden, J.L.K. and Wilcox, G.L., Intrathecal substance P and somatostatin in rats: behaviours indicative of nociception, Peptides, 3 (1982) 49-54. 67 Sundler, F., Brodin, E., Ekblad, E., Hakanson, R. and Uddman, R., Sensory nerve fibres: distribution of substance P, neurokinin A and caicitonin gene-related peptide. In R. Hakanson and E Sundler (Eds.), Tachykinin Antagonists, Elsevier, Amsterdam, 1985, pp. 3-14. 68 Tatemoto, K., Lundberg, J.M., Jonvall, H. and Mutt, V., Neuropeptide K: isolation, structure and biological activity of a novel brain tachykinin, Biochem. Biophys. Res. Commun., 128 (1985) 947-953. 69 Vaught, .1.L., Scott, R. and Jacoby, M., Substance P antagonists and analgesia: do they or don't they? In J.L. Henry, R. Couture, A.C. Cuello, G. PeUetier, R. Quirion and D. Regoli (Eds.), Substance P and Neurokinins, Springer, New York, 1987, pp. 176-178. 70 Vaught, J.L., Scott, R., Wright, D. and Jacoby, H., [DPro2,o-Trp6'8,Nlel°]-Neurokinin B: pharmacological profile of a novel neurokinin antagonist. In J.L. Henry, R. Couture, A.C. Cuello, G. Pelletier, R. Quirion and D. Regoli (Eds.), Substance P and Neurokinins, Springer, New York, 1987, pp. 122-124. 71 Vierck, C.J. Jr., Greenspan, J.D., Ritz, L.A. and Yeomans, D.C., The spinal pathways contributing to the ascending conduction and the descending modulation of pain sensations and reactions. In T.L. Yaksh (Ed.), Spinal Afferent Processing, Plenum, New York, 1986, pp. 275-329. 72 Wiesenfeld-Hallin, Z., Substance P and somatostatin modulate spinal cord excitability via physiologically different sensory pathways, Brain Research, 372 (1985) 172-175. 73 Willcockson, W.S., Chung, J.M., Hori, V., Lee, K.H. and Willis, W.D., Effects of iontophoretically-released peptides on primate spinothalamic cells, J. Neurosci., 4 (1984) 741-750. 74 Woolf, C. and Wiesenfeld-Hallin, Z., Substance P and calcitonin gene-related peptide synergistically modulate the gain of the nociceptive flexor withdrawal reflex in the rat, Neurosci. Len., 66 (1986) 226-230. 75 Wormser, U., Laufer, R., Hart, Y., Chorev, M., Gilon, C. and Selinger, Z., Highly selective agonists for substance P receptor subtypes, EMBO J., 5 (1986) 2805-2808. 76 Yaksh, T.L., Jessell, T.M., Gamse, R., Mudge, A.W. and Leeman, S.E., Intrathecal morphine inhibits substance P release from mammalian spinal cord in vivo, Nature (Lond.), 286 (1980) 155-157. 77 Yaksh, T. and Rudy, R., Chronic catheterization of the spinal subarachnoid space, Physiol. Behav., 17 (1976) 1031-1036. 78 Yasphal, K., Wright, D. and Henry, J.L., Substance P reduces tail-flick latency: implications for chronic pain syndromes, Pain, 14 (1982) 155-167. 79 Zieglgansberger, W. and Tulloch, I.E, Effects of substance P on neurones in the dorsal horn of the spinal cord of the cat, Brain Research, 166 (1979) 273-282.