Neuroacience Vol. 43, No. 2/3, pp. 601-610, 1991

0306-4522/91 $3.00+ 0.00 Pergamon Press pie ~) 1991IBRO

Printed in Great Britain

RESPONSES OF FUNCTIONALLY IDENTIFIED NEURONES IN THE DORSAL H O R N OF THE CAT SPINAL CORD TO SUBSTANCE P, N E U R O K I N I N A A N D PHYSALAEMIN M. W. SALTER* and J. L. HL~RY? Departments of Physiology, Research in Anaesthesia and Psychiatry, McGill University,Montreal, Quebec, Canada Almtrnet--The mamma!ia~ tachykinins, substance P and neurokinin A, and the non-mammalian tachykinin, physainemin, were teated on functionally identified dorsal horn neurones /n vivo. The experiments were done on cats which were anaesthetized with sodium pentobarbital or were annemically decerebrated. Extracellular ~ngle-unit recordings were made in the lumbar spinal cord and the tachykinins were applied by iontophoresis. Each neurone was classitied functionally as wide dynamic range, non-nociceptive, noeieeptive specific or proprioceptive. The response to tachykinin application was determined for each neurone. Application of each of the tachykininl evoked a characteristic excitatory response which was delayed in omet, slow in developing and prolonged: phy~laemin excited 99/131 neuron~ tested, neurokinin A excited 45/63 neurones and substance P excited 32/49 neurones. With two neurones physaiaemin evoked a depression of the rate of firing, which may have been caused indirectly by excitation of a neighbouring neurone. Such depression was not elicited by either substance P or by neurokinin A. Phyulaemin had a l ~ , , ; - ~ i a i excitatory effect on nociceptive neurones evoking excitation of 76/94 nociceptive neuroms compared with 12/23 non-nociceptive neurones (Z2 ffi 7.9, I d.f., P = 0.005). Substance P aim caused a preferential excitation, with 30/40 nociceptive neurones being excitedwhile all of the non-nociceptive neurones (n •7) were u~affected (~2= 11.5, I d.f., P =0.0007). In contrast, neurokinin A failed to have a preferential effect; 32/46 nociceptive and 9/10 non-nociceptive neurones were excited (Z2 = 1.0, 1 d.f., P = 0.40). Comparing the proportions of nociceptive neurones excited by the different tachykinins indicated that this type of neurone was not differently sensitive to any of the three peptides (X2= 3.2, 2 d.f., P = 0.20). On the other hand, non-nociceptive neurones were preferentially excited by neurokinin A and physaiaemin compared with substance P (X2= 13.4, 2d.f., P =0.001). With regard to the endogenous tachykinins the results of this study may be interpreted in the following ways. The diffe~-,ntlal excitatory effect of substance P on nociceptive neurones supports the proposed role for this peptide in the transmi~fion specifically of nociceptive inputs at the first afferent synapse. On the other hand, as neurokiuin A excited non-nociceptive as well as nociceptive neurones, there may be a functional role for neurokinin A distinct from that of substance P.

The tachykinin, substance P, has been implicated as a chemical mediator of synaptic communication of primary afferent neurones in nociceptive pathways. ~'7.H'24'37 A functional role for neurokinin A, another tachykinin found in primary afferent neurones, remains to be defined. Certain evidence might lead to the expectation that neurokinin A may have functions similar to those of substance P. This evidence includes the fact that both peptides are encoded for in the same gone, preprotachyk~mn-l. • • • ~6.~ Differcntial processing of the R N A transcribed from this gene gives rise to three species of m R N A , a,//and preprotachykinin (PPT); the sequence for ncurokinin A is encoded in ~ and 3,P P T while all three species *Present address for M. W. Salter: Hospital for Sick Children, Division of Neuroscienoe, 555 University Ave., Toronto, Ontario, Canada M5G IX8. ?To whom correspondence should be addressed at: Department of Phytiology, McIntyre Medical Sciences Building, 3655 Drummond Street, Montreal, Quebec, Canada H3G IY6. Abbreviations: PPT, preprotachykimn; PSTH, peristimulus time histogram; WDR, wide dynamic range. NSC43-2/~--K

601

encode substance P. Thus, it would be anticipated that cells which express neurokinin A also produce substance P and, indeed, immunoreactivity for neurokinin A is colocalized with that for substance P in a subpopulation of primary afferent fibres) 2"33 In addition, the two peptides are released together from the spinal cord by capsaicin) 3 On the other hand, from the point of view of effector mechanisms postulating a tight relationship between roles of neurokinin A and substance P may not be justified. F o r example, there appear to be several distinct types of neurokini/3 receptor; substance P is the preferential ligand for the so-called N K - I receptor while neurokinin A has a preferential affinity for the NK-2 receptor, l's'2s Within the spinal cord these two types of receptor are differentially distributed. 3~ It is known, also, that neurokinin A is less potent than substance P in facilitating a nociceptire spinal reflex, the tail-flick reflex. 3 The lack of a clearly defined role for neurokialn A prompted an investigation of its effects on functionally-identified sensory neurones in the spinal dorsal horn and a comparison of these effects with those of

M. W. S ~ J . ~ and J. L. HENRY

602

substance P. A principal objective was to determine whether neurokinin A shows the same differential excitatory effect o n nociceptive vs non-nociceptive neurones as does substance P. In addition, the nonmammalian tachykinin, p h y ~ l ~ m i n , wag studied for comparative purposes, as • second ligand w h k h shows preferential amnlty for the N K - I receptor, l~° Phy~lae~min has been found to have facifitatory effects o n the talI-fl/ck reflex which are indistinguishable from those of substance P ? Some of the work in the present study has been reported previously. ~'~ EXPgMIMI~TAL PIIOCgDUBI[S

Animal preparation Experiments were performed on 46 adult cats: 31 were with sodium peatobarbitul (40mg/kg, i.p.; supplemental dcm= 5 mg/kg i.v. every 3 h), 12 were a n a ~ with ~,-chloralme (60n~Ag i.v. after induction with lmlothg~/oxygen) and ~ were _~_,~-l~zted under induced by lmlotha~/oxyJma, which wag mbseqnently ~ t i n n e d . Thronshom the ~ t s the carotid arterial blood presmn~ewas continuously digplsyed on a Grim P5 polygraph. The mean arterial pressure was maintained above 80ram Hg with i.v. infusion of 10% dextran (Macrodex, Watrmada) or n o ~ bitartrate (Levophed, Winthrop; 0.002% in normal saline), if neommry. Spinal segments L5 to L7 were surgically expmed for l'e¢ol~din~ and were Govered with ~ ~ Off to prevent d ~ and cooling. In all experiments the spinal cord was tmmected st thc first lumbar level to l~nove d e ~ - n d i ~ influeuo, and to dtml.ate the possibility that the effects on lumbar donal hem neurone* might be mediated via m p r l u ~ m l g r u c t u ~ . Prior to the t r m m ~ o n 0.1 ml of 1.0% lidocaine hydroc~oride was injected into the Li segment to mJnlmim spinal sheck. After bilateral pnenmothorax, the anirr~ls were l~t~ysed with pancuroninm bromide (Pavulon, Orpnon; I mg/kg i.v., repeated when required) and ventilated artificially. End-tidal CO2 concentration was maintained between 3.5 and 5.0%. Rectal temperature was nutintained at 38°C with a servo-controlled infrared bulb. Periodically, spinal cord circulation was monitored visually u~ng a dissecting stereomicroscope.

Record~ and data acquisiaon Single unit spikes were recorded cxtracellularly with multibarrelled #ass m i c r o l ~ (overall tip diameter 4-10/~m). A solution of 2.7M NtCI was placed i the central recording barrel (impedance 4-10 Ml'l t o u r e d at 100 Hz). The raw data were amplified, d/splayed on o,cillos:ope8 and via an audio monitor, and recordad on rna.mefic cauette tape and on film as detailed in Salter and Henry.~ The rate of discharge was continuously d/splayed on a Gram P5 polygraph. Intervals between spikes and between stimuli were computed to +0.5 ms using an IBM personal Mad hardware and software developed in our laboratory. ~ These intervals were stored on fixed disk.

lontophore~is One of the outer barrels of the micropipette contained a solution of sodium L-glutamate (! M, pH 7.4, Sigma). Only units excited by iontophoretic application of glutamate were studied, to eliminate recordhlge from fibres. L e a l u ~ of glDtamate between Rlpqpl/¢ationswas minimized by a retaining current (10 hA). Other barrels of the micropipette contained one of the following: substance P (0.graM in 165raM NaC1, pH 5.5, Peninsula or Institute Armand-Frappier), neurokinin A

(1 mM in 165 mM NaCI, pH 5A, Institute Armand-Frappier), phymlnemm (l mM in 165 mM NaCl, pH 5.5, Protein Research Foundation or Bachem), control solution (NaCI, 165raM, pH 5.5) or Pontamine.Sky Blue 6BX (0.5% in 0.5 M sodium acetate, Ourr). With the exception of glutamate, iontophoretic application was with outward current. Current passed through the barrel containing NaCI was mud to detect artefacts due to local c_-btnges in current density at the tip of the electrode. Controls for possible currant artefacts were done either by ejecting outward cunent continuously between peptide applications (see Fig. IB) or by making control ejections of the same nu~nitude and duration as the peptide applications (see Fig. 2C).

Functional classification of units Details of the claszifa~tion scheme are given ehtewhere.2* In brief, the following natural peripheral stimuli were used: movement of single hairs, fight touch of the skin with a fiune paper, firm manual pressure 0m:l~d to be nonno~ons when applied to the experimenter), noxious pinch of the Iddn with a serrated forceps, noxious hes~ag of the skin (temperature >47°C), and pamve movement of the limb. According to its responses to these stimuli each neurone was claamified as non-nociceptive, wide dynamic ran a~ (WDR), nociceptive specific or proprioceptive. For each neurone the excitatory receptive fields for hair movemint, light touch and noxious pinch were represented on a ~mmatic dia~am of the hindlimb. In all cues, the neurone wM ~ functionally before glutamate or any of the tachykinins were tested.

Data analysis Quantitative analysis of data was done off-line. To determine an effect, the number of spikes during the period of each application of a given pept~le for each neurone was measured and compared with the number of spikes during a control period oft.he same duration immediatelypreceding each application. The percentage change in the number of spikes was calculated for each application. The average percentage change was determined u m g up to five consecutive applications and statistical si~nif~ance was calculated using the paired t-test or the sign test, as appropriate. Effects were considered statistically significant for P