Microinjections of calcitonin gene-related peptide within the trigeminal subnucleus caudalis of the cat affects adrenal and autonomic function

Brain Research, 558 (1991) 53-62 © 1991 Elsevier Science Publishers B.V. All rights reserved. 0006-8993/91/$03.50 ADONIS 0006899391169167

53

BRES 16916

Microinjections of calcitonin gene-related peptide within the trigeminal subnucleus caudalis of the cat affects adrenal and autonomic function David A. Bereiter and Albert P. Benetti Section of Neurobiology and Department of Surgery, Brown University~Rhode Island Hospital, Providence, R! 02903 (U.S.A.) (Accepted 9 April 1991)

Key words: Adrenal medulla; Autonomic function; Calcitonin gene-related peptide; Catecholamine; Medullary dorsal horn; Substance P

To assess the role of calcitonin gene-related peptide (CGRP) within the trigeminal subnucleus caudalis (Vc) on adrenal and autonomic function, microinjections were directed at different laminae of Vc in chloralose-anesthetized cats. Microinjections of CGRP (5 pmol, 100 nl) into laminae I-II increased significantly the adrenal secretion of epinephrine, adrenal blood flow, adrenal vascular conductance, mean arterial pressure and heart rate. Injections of CGRP into laminae V-VI decreased significantly the adrenal secretion of epinephrine, however, other measured variables were not affected. To examine if CGRP interacts with substance P within Vc to modify adrenal and autonomic function, subthreshold doses of each peptide were injected alone and simultaneously. Combined subthreshold doses of CGRP and substance P injected into laminae V-VI, but not into laminae I-II or III-IV, evoked increases in arterial pressure and in heart rate that exceeded the responses seen after injection of either peptide alone. The adrenal secretion of catecholamines was not affected by individual or combined subthreshold doses of either peptide, regardless of the laminar site of injection. These data suggest that release of CGRP within laminae I-II of Ve alters adrenal and autonomic function via mechanisms separate from those that mediate substance P-evoked responses. In contrast, CGRP and substance P may act, at least in part, through a common neural substrate within the deeper laminae of Vc to modify arterial pressure and heart rate. Thus, multiple subpopulations of peptide-responsive neurons in the medullary dorsal horn likely contribute to the reflex adrenal and autonomic responses that often accompany nociception.

Noxious sensory stimuli evoke adrenal and autonomic responses as well as causing pain sensation 1'4'38'39. A l -

changes in a u t o n o m i c function were seen after microinjections into laminae I - I I , but not after injections into laminae V - V I 8. H o w e v e r , i m m u n o r e a c t i v e fibers and terminals 14m and receptors for substance p29 are found

though considerable attention has been directed at delineation of the mechanisms underlying the sensorydiscriminative aspects of nociception 9'15'34, the central neural mechanisms that underlie the reflex autonomic responses to nociceptor activation r e m a i n less well defined. To assess the role of putative n e u r o t r a n s m i t t e r release from p r i m a r y afferents within a brainstem region that encodes noxious sensory input, we have previously e x a m i n e d the h o r m o n a l and physiological responses to microinjections of glutamate 6"7 and substance p8 into different laminae of the trigeminal subnucleus caudalis (Vc). G l u t a m a t e activation of neurons within laminae I - I I o r within laminae V - V I of Vc e v o k e d increases in a d r e n a l and a u t o n o m i c function that were consistent with nociceptor activity6'7 and c o r r e s p o n d e d well to the known anatomical distribution of s e c o n d - o r d e r nociceptive neurons within Wc15'34. In contrast, substance P - e v o k e d

within both laminae I - I I and laminae V - V I , and substance P is r e p o r t e d to excite most, if not all, nociceptive neurons within Wc21'36. Collectively, these results have suggested that nociceptive neurons found within the deep magnocellular p o r t i o n of Vc that also r e s p o n d to substance P are m o r e likely to contribute to the sensorydiscriminative aspects of nociception and not to the autonomic aspects of nociception. A l t e r n a t i v e l y , laminae V - V I nociceptive neurons that r e s p o n d to substance P may require additional chemical inputs to affect autonomic function since injections of g l u t a m a t e into the d e e p magnocellular p o r t i o n of Vc alters adrenal and autonomic function 6'7 and further, d e p l e t i o n of substance P does not reduce the p e r c e n t a g e of nociceptive neurons in Vc 35. T h e present experiments assessed the influence of microinjections of calcitonin g e n e - r e l a t e d p e p t i d e

INTRODUCTION

Correspondence: D.A. Bereiter, Brown University/Rhode Island Hospital, Neuroendocrine Laboratory, Division of Surgical Research, Providence, RI 02903, U.S.A.

54 ( C G R P ) within d i f f e r e n t l a m i n a e o f Vc o n a d r e n a l and a u t o n o m i c function. F i b e r s and t e r m i n a l s i m m u n o r e a c tive f o r C G R P are l o c a l i z e d within l a m i n a I - I I , and to a lesser e x t e n t , within l a m i n a V - V I o f Vc 2'24'2s and r e c e p t o r s for C G R P a r e d i s t r i b u t e d similarly 24. W h e r e a s a significant p e r c e n t a g e of trigemina117'2°'27 and dorsal r o o t g a n g l i o n n e u r o n s a°'18'43'44 d e m o n s t r a t e colocalization o f i m m u n o r e a c t i v i t y for C G R P and for s u b s t a n c e P, a n d n o x i o u s s e n s o r y stimuli e v o k e t h e r e l e a s e o f C G R P and s u b s t a n c e P w i t h i n t h e superficial l a m i n a e of t h e spinal c o r d 3°, a s e c o n d aim of this study was to assess the effects of s i m u l t a n e o u s m i c r o i n j e c t i o n s of t h e s e p e p t i d e s

was filled with substance P (0.22 mM + 0.1% BSA) plus 2% Fast green dye. Each pipette was connected to a 1 ~1 syringe (Hamilton) and injections were delivered at the rate of 20 hi/10 s. The patency of the injection pipette was confirmed microscopically prior to placement in the brainstem and again, upon removal from the brain at the end of the experiment. All microinjections were made ipsilateral to the adrenal vein sample. One to 3 sites of injection were explored in each animal within the following region: 1-3 mm caudal to the obex, 2.7-5.2 mm lateral to the midline, and from 0.5 to 2.6 mm below the dorsal surface of the brainstem. At the end of the experiment, the animal was perfused through the heart with normal saline followed by 10% buffered formalin. The sites of injection were identified in coronal sections (40 gin) from the center of the deposit of Fast green dye and mapped onto a brainstem outline adapted from Gobel et al.19.

o n a d r e n a l and a u t o n o m i c function. M i c r o i n j e c t i o n s of

Experimental design

p e p t i d e s w e r e d i r e c t e d at l a m i n a e of Vc that process

The effects of CGRP alone were assessed after microinjections of 2 pmol (40 nl) and 5 pmol (100 nl), delivered in random order at 11 of 19 sites. The remaining 8 of the 19 rites received injections of the larger (5 pmol) dose only. These data were collected from 11 cats. An additional 11 cats were used to assess the effects of simultaneous injections of CGRP and substance P. Subthreshold doses of CGRP (3 pmol, 60 nl) and substance P (13 pmol, 60 nl), were injected alone, in random order, followed by the simultaneous injection of both peptides (120 nl) at each of 18 sites. The doses used for the dual-injection experiments were determined previouslys to evoke little or no change in the measured variables. Although no separate group of animals received cerebrospinal fluid (CSF) injections alone, we have shown repeatedly that 160 nl injections of CSF into different laminae of Vc had no effect on adrenal or autonomic function6-s. Each subsequent sampling period was separated by 30 min. Peripheral arterial (1.5 ml) and adrenal venous blood samples (0.4-1.0 ml) were collected over 20-60 s at -5 and 0 min (prestimulus controls) and at 1, 3, 6 and 10 min after the onset of injection.

nociceptive

input

(I-II

and

V-VI)

and

directed

at

l a m i n a e t h a t process m a i n l y n o n - n o x i o u s sensory input ( I I I - I V ) in a n e s t h e t i z e d cats. A p r e l i m i n a r y r e p o r t of t h e s e results has b e e n p r e s e n t e d 3.

MATERIALS AND METHODS

General methods Adult cats of either sex (2.4-5.9 kg b. wt.) were deprived of food overnight but were given free access to water. After initial induction with ketamine-HC1 (25 mg/kg, i.m.), anesthesia was maintained by a-chloralose (75 mg/kg, i.v.) given as a full dose prior to surgery and supplemental doses of chloralose (7.5 mg/kg, i.v.) were given every 60-90 rain throughout the day. Depth of anesthesia was determined by loss of flexion withdrawal reflex to pinch of the hindpaw. Arterial pressure was monitored from a catheter placed in the descending aorta (via the femoral artery) and heart rate was assessed from a standard 3-lead electrocardiogram. Peripheral blood samples were collected from a catheter placed in the femoral artery. The left lumboadrenal vein was catheterized distal to the lateral edge of the adrenal gland as described previously5. A catheter was placed in the cephalic vein for counterbalanced infusions of normal saline during blood sampling, for reinfusion of red cells after sampling, and for the administration of all drugs. After tracheostomy, animals were respired artificially (15 strokes/min) with O2-supplemented room air. Expiratory CO 2 was monitored continuously with an infrared detector (Anarad) and stroke volume was adjusted to maintain a normal range (3-4.5%). Animals were paralyzed with gallamine triethiodide (5 mg/kg, i.v.) after completion of all surgical procedures, and supplemental doses (3 mg/kg, i.v.) were given hourly. Body temperature was monitored rectally and kept at 38 °C with a heating blanket. All surgical incisions were infiltrated with 2% lidocaine jelly.

Brainstem exposure, microinjection technique and histology After placement of all catheters, the animal was mounted in a stereotaxic frame (Kopf) using blunt ear bars. The muscle overlying the occipital bone was reflected to allow visualization of the dorsal surface of the brainstem and the obex as a reference point for placement of microinjection pipettes. Microinjections of peptides were delivered from glass micropipettes (tip diameter = 50-75/~m) positioned stereotaxically at the targeted brainstem site 10-15 min prior to injection. Care was taken to avoid rupture of blood vessels on the dorsal surface of the brainstem. Single micropipettes were used to assess the effects of CGRP alone (0.05 raM, rat a-CGRP, Bachem) plus 0.1% bovine serum albumin (BSA) and 2% Fast green dye. To assess possible interactive effects between the release of CGRP with that of substance P, twin-barreled pipettes were made and one barrel was filled with CGRP as above and the other barrel

Biochemical determinations Peripheral arterial and adrenal venous blood samples were collected on ice in tubes treated with EDTA (0.16 M). After centrifugation, the plasma was stored at -60 °C for subsequent analyses. The red blood cells were resuspended in saline and returned to the animal prior to each subsequent sampling period. The concentration of adrenal venous catecholamines was determined from 100 gl of plasma after alumina extraction by highperformance liquid chromatography (HPLC) with electrochemical detection as described previously4. The intra-assay and interassay coefficients of variation for a cat plasma pool containing 2.0 ng/extraction volume of norepinephrine were 2.9 and 6.8%, respectively; containing 1.8 ng/extraction volume of epinephrine were 3.3 and 11.6%, respectively; and containing 2.4 rig/extraction volume of dopamine were 4.6 and 11.8%, respectively. Adrenal secretory rates were calculated as the product of (adrenal plasma concentration (ng/ml)-peripheral plasma concentration (ng/ml))xadrenal plasma flow (ml/min) = ng/min for each catecholamine species. Plasma adrenocorticotrophic hormone (ACTH) was determined by direct radioimmunoassay 31 using an antibody generated against a conjugate of ACTH-(1-24) that reacted completely with ACTH(1-39) and ACTH-(1-24), but not with either the NH2-terminal end (ACTH-(1-13) or the COOH-terminal end (ACTH-(18-39)) fragments of the ACTH molecule. Charcoal-stripped cat plasma was added to the ACTH-(1-39) standards (Bachem) to control for matrix effects. The intra-assay and interassay coefficients of variation for a plasma pool with an ACTH concentration of 34 pg/ml were 12 and 13%, respectively.

Statistical analyses The adrenal and autonomic responses were analyzed by first grouping the injection sites according to the laminar location within Vc prior to knowledge of the hormonal response. Sites within the

55 superficial layers of Vc included laminae I and II; sites within the magnocellular portion of Vc included laminae III and IV; sites within the deep magnocellular portion of Vc included laminae V and VI. The delineation of laminae was according to the description by Gobel et 81.19. All data were collected from 22 animals where the mean arterial pressure remained above 75 mm Hg and the patency of the injection cannula was verified at the end of the experiment. The responses of adrenal secretion of catecholamines, adrenal blood flow, adrenal vascular conductance (adrenal blood flow/mean arterial pressure), plasma concentrations of ACTH, mean arterial pressure and heart rate to microinjections of peptides, across laminar sites of injection, were assessed by two-way analysis of variance (injection volume and time) corrected for repeated measures45. Analyses were performed on the absolute change from the prestimulus value (mean of -5 and 0 min values) and individual comparisons used the Newman-Keuls test after analysis of variance45. The effects of simultaneous injections of subthreshold doses of CGRP and substance P were assessed by comparing the responses after individual injections, after simultaneous injections, and to the calculated sum of the responses to individual injections. Two-way analyses of variance used the absolute change from the prestimulus values. Only those sites of injection that were located within the laminar boundary of Vc were included in these analyses. To encode each injection site on the brainstem outline for an evoked change in adrenal secretion of epinephrine, the 95% confidence limit for the absolute difference between the prestimulus values (-5 and 0 min) was determined for all sampling periods across all animals. The 95% confidence interval was determined both as an absolute difference (2.09 ng/min) and as a percentage difference (37.40%) from all sites of injection (n = 37). The symbols on the brainstem outline of Fig. 2 reflect the peak change in adrenal secretion seen during the 10 min postinjection sampling period with respect to both of these confidence intervals. All data in the text and in the figures represent the mean + S.E.M. Possible relationships between response variables after microinjections of CGRP alone were assessed by Spearman rank-order correlation analyses. These analyses considered the peak change in adrenal secretion of catecholamines in relation to the change in adrenal blood flow, in adrenal vascular conductance or in mean arterial pressure and included the responses to both doses of CGRP (2 and 5 pmol).

I - I I , whereas injections into laminae V - V I caused a significant (P < 0.01) decrease in secretion. However, the magnitude of these changes were smaller than those of adrenal epinephrine. The adrenal secretion of dopamine was not affected by microinjections of CGRP, regardless of the laminar site of injection within Vc (data not shown). The location of the sites of microinjection of C G R P are seen in Fig. 2, where each site has b e e n encoded for the peak change in the adrenal secretion of epinephrine evoked by injection of 5 pmol of CGRP. As seen in Fig. 2, the adrenal epinephrine secretory responses to microinjections of C G R P were consistent within a given lamina. Six of the 7 sites located within lamina I - I I caused a prompt (by 1 or 3 min) increase in the adrenal secretion of epinephrine, whereas 6 of the 7 sites located within laminae I I I - 1 V had no effect, and 4 of the 5 sites located within laminae V - V I evoked a significant decrease in secretion. Peripheral arterial concentrations of n o r e p i n e p h r i n e were 0.40 + 0.05 ng/ml (n = 30) prior to stimulation and increased signficantly (+0.15 + 0.05 ng/ml, P < 0.01) by 10 min after injections of C G R P within laminae I - I I ; however, peripheral concentrations of epinephrine or of d o p a m i n e were not changed.

a .

8.=_

==E t-

a

A.

b

64-

,i-

20-

-2-

RESULTS

"4" -6

Microinjections of CGRP The prestimulus adrenal secretion of epinephrine, of n o r e p i n e p h r i n e and of d o p a m i n e was 6.27 _+ 0.77, 6.27 + 1.41 and 0.06 _+ 0.01 ng/min (n = 30), respectively, prior to injection of either dose of C G R P (2 pmol, n = 11; 5 pmol, n -- 19) into Vc. Microinjections of C G R P (5 pmol) into the superficial laminae ( I - I I ) caused a prompt increase in the adrenal secretion of epinephrine, whereas injections into the deep magnocellular layers (lamina V - V I ) of Vc evoked a significant decrease in the adrenal secretion of epinephrine (Fig. 1A). C G R P injections located within laminae I I I - I V had no significant effect on the adrenal secretion of catecholamines. The increase in the adrenal secretion of epinephrine evoked by injections of C G R P within laminae I - I I was dose-related and n o n e of the injections of smaller doses (2 pmol) of C G R P into laminae I - I I caused a significant change in adrenal secretion (Fig. 1B). The adrenal secretion of norepinephfine was not affected by injections of C G R P into lamina

8-

l

l

l

l

l

a*

l

l

l

l

l

l

B

a

•

~" 6" •: - ~ ..C

4("

Q-~ •

2"

o.

U.l ~ -2" ~ ~ -4" -6

,~ .

.

.

.

.

.

.

.

.

.

0 1 2 3 4 5 6 7 8 9 1 0 Time (min)

Fig. 1. A: adrenal epinephrine secretory responses to microinjections of CGRP (5 pmol) into different laminae of the trigeminal subnucleus caudalis. &, laminae I-II sites (n = 7); O, laminae Ill-IV sites (n = 7); O, laminae V-VI sites (n = 5). B: adrenal epinephrine secretory responses to microinjections of 2 pmol (A) and 5 pmol ( • ) into laminae I-II of trigeminal subnudeus caudalis. • P < 0.05, **P < 0.01 versus prestimulus value; a = P < 0.05, b = P < 0.01 versus other groups. Solid bar at time 0 represents microinjection period.

56 a

12

~ -r-

*

A.

8

E

E

v

4'

D.

~

O' I,

-4

,

,

,

i

J

l

,

,

,

,

**

S l

16: "E 12-

Fig. 2. Location of sites of microinjection of CGRP (5 pmol) encoded for the peak change in the adrenal secretion of epinephrine. &, increase in secretion; O, decrease; O, no change; a, nucleus ambiguus; TLC, trigeminal extension of the lateral cervical nucleus; I-II, III-IV, and V-VI, laminae of trigeminal subnucleus caudalis; Vtr, spinal trigeminal tract; -3, distance caudal to obex (mm). Calibration = 500/~m.

Injections of C G R P into the d e e p e r laminae ( I l l - I V or V - V I ) of Vc h a d no effect on p e r i p h e r a l concentrations of catecholamines (data not shown). T h e prestimulus p l a s m a concentration of A C T H was 90 + 11 pg/ml (n = 30) prior to injection of C G R P and no significant differences between t r e a t m e n t groups

~

8-

~ •I-

4-

"~

O-

J

i

i

|

i

i

i

|

i

i

,

0 1 2 3 4 5 6 7 8 9 1 0 Time (rain)

Fig. 4. Mean arterial pressure (A) and heart rate (B) responses to microinjections of CGRP (5 pmol) into different laminae of the trigeminal subnucleus eandalis. • laminae I-II sites (n = 7); O, laminae Ill-IV sites (n = 7); O, laminae V-VI sites (n = 5); MAP, mean arterial pressure; HR, heart rate; *P < 0.05, **P < 0.01 versus prestimulus value; a = P < 0.05 versus other groups. Solid bar at time 0 represents microinjection period.

(laminae) were noted. Plasma A C T H was not affected significantly by C G R P regardless o f the l a m i n a r site of ~:

0.3"

b

A I

b

o

"

0.2

"o

o.1

mE 0.0 •'o ,,¢

-0.1

<

-0.2

•

0.002"

~'~

0.001-

8

o.ooo,

O t'--

J

i

i

i

i

i

i

b

i

i

i

i

B.

b

-~'~ < <3

-0.002

~

l

l

l

l

l

l

l

l

l

l

0 1 2 3 4 5 6 7 8 9 1 0 Time (min)

F i g . 3. A d r e n a l b l o o d f l o w ( A ) a n d a d r e n a l v a s c u l a r c o n d u c t a n c e

(B) responses to microinjection of CGRP (5 pmol) into different laminae of the trigeminal subnucleus caudalis. • , laminae I-II sites (n = 7); O, laminae Ill-IV sites (n = 7); O, laminae V-VI sites (n = 5). *P < 0.05, **P < 0.01 versus prestimulus value; b = P < 0.01 versus laminae V-VI injection group. Solid bar at time 0 represents microinjection period.

microinjection. The m e a n prestimulus total a d r e n a l b l o o d flow was 0.73 + 0.06 ml/min (n = 30). Microinjections of C G R P (5 p m o l ) into laminae I - I I or into laminae I l l - I V e v o k e d a significant ( P < 0.025) increase in a d r e n a l b l o o d flow, whereas injections into l a m i n a e V - V I h a d no effect (Fig. 3A). T h e adrenal b l o o d flow responses to injections of C G R P were very consistent as 6 of the 7 sites located within laminae I - I I as well as 6 o f the 7 sites l o c a t e d within laminae I l l - I V e v o k e d increases in adrenal b l o o d flow. Injections of the smaller dose of C G R P (2 pmol) had no effect regardless of the site of injection within Vc. To examine if the change in a d r e n a l b l o o d flow to 5 p m o l of C G R P was the result of active vasodilatation o r simply due to an increase in arterial pressure, a d r e n a l vascular conductance values were d e t e r m i n e d ( a d r e n a l b l o o d flow (ml/min)/mean arterial pressure ( m m H g ) = ml/min/mm Hg). Prestimulus a d r e n a l vascular conductance a v e r a g e d 0.0069 + 0.0005 ml/min/mm H g (n = 30) and microinjections of C G R P within either l a m i n a e I - I I or laminae I l l - I V caused a significant ( P < 0.025) increase in adrenal vascular conductance that r e t u r n e d towards prestimulus values (Fig. 3B). M e a n arterial pressure and h e a r t rate were 106 + 3 m m H g and 208 + 5 beats/min

57

A A

•

=

1512-

'~

2

~.

0~

96

LLI

"4"

"5,E, ~'~

-6

F= ~

.

.

.

.

..

,

,

,

,

,

, m

8ffl -~

B.

¢~

2"

6

-4-"

32 i

m

i

i

0

1

2

3

15

i

i

i

i

i

4 5 6 7 Time (min)

i

8

i

i

i

i

i

;

i

~

B.

b a

i

-3

.

A.

9 6 3 0 -3

t

12

a

i

i

i

t

i

i

i

i

T i

B.

9

32 0 -3"

-6

i

t

01

i

J

i

i

J

i

J

910

12

t,-

J

12

-2

Fig. 5. Adrenal epinephrine secretory responses to microinjections of subthreshold doses of CGRP alone, of substance P alone or coincidently into laminae I-II (A) or into laminae V-VI (B) of the trigeminal subnucleus caudalis. A: laminae I-II sites (n = 5); B; laminae V-VI sites (n = 6). O, CGRP, 3 pmol; C), substance P, 13 pmol; A, coincident injections of CGRP and substance P. Solid bar at time 0 represents microinjection period.

'~E

i

.I-

-6

.=

i

15 4"

"O

<~

0 -3

e-

~

3

J

i

J

i

2 3 4 5 6 7 8 9 1 0 Time (min)

Fig. 6. Mean arterial pressure responses to microinjections of subthreshold doses of CGRP alone, of substance P alone or coincidently into laminae I-II (A) or into laminae V-VI (B) of the trigeminal subnucleus caudalis. A: laminae I-II sites (n = 5); B: laminae V-VI sites (n = 6). 0 , CGRP, 3 pmol; C), substance P, 13 pmoi; &, coincident injections of CGRP and substance P. *P < 0.05, **P < 0.01 versus prestimulus value; a = P < 0.05 versus other groups. Solid bar at time 0 represents microinjection period.

0

J

i

i

i

i

j

j

!

1

2

3

4

5

6

7

8

i

i

9 10

Time (rain) Fig. 7. Heart rate responses to microinjections of subthreshold doses of CGRP alone, of substance P alone or coincidently into laminae I-II (A) or into laminae V-VI (B) of trigeminal subnucleus caudalis. A: laminae I-II sites (n = 5); B: laminae V-VI sites (n = 6). 0 , CGRP, 3 pmol; @, substance P, 13 pmol; A, coincident injections of CGRP and substance R *P < 0.05, **P < 0.01 versus prestimulus value; a = P < 0.05, b = P < 0.01 versus other groups. Solid bar at time 0 represents microinjection period.

(n = 30), respectively, prior to injection of C G R P and no significant differences between t r e a t m e n t groups were noted. Microinjections of C G R P (5 pmol) into laminae I - I I evoked prompt (by 1 min) increases in m e a n arterial pressure (Fig. 4A) and in heart rate (Fig. 4B) that returned towards prestimulus values by 3 min. C G R P injections into the deeper laminae of Vc had no consistent effect on arterial pressure or on heart rate; however, a delayed increase (by +10 min) in heart rate after injections into laminae I I I - I V was noted. The arterial hematocrit averaged 0.42 + 0.07 (n = 30) prior to microinjection and r e m a i n e d constant during the experiment. Possible relationships between changes in response variables after microinjections of C G R P (2 and 5 pmol doses) into Vc were assessed by Spearman rank-order correlation analysis on the peak change observed during the 10 min sampling period. The peak change in adrenal secretion of epinephrine was well correlated with that of n o r e p i n e p h r i n e (rs = 0.563, P < 0.01). However, the peak change in adrenal secretion of catecholamines was not correlated with that of adrenal blood flow, of adrenal vascular conductance or with that of arterial pressure (P > 0.05) suggesting that the sites that evoked the greatest change in the adrenal secretion of catecholamines did not

58 TABLE I Summary comparison of the autonomic effects of microinjections of substance P, of CGRP, and simultaneous injections of both peptides into those laminae of trigeminal subnucleus caudalis that process nociceptive sensory input

ABF, adrenal blood flow; ACTH, plasma adrenocorticotropin; AdCon, adrenal vascular conductance; Epi, adrenal secretion of epinephrine; HR, heart rate, MAP, mean arterial pressure; SP, substance P. + or - = P < 0.05, and + + = P < 0.01 versus prestimulus value. Laminae

Injection

Dose

I-II

SP SP CGRP CGRP SP/CGRP

low high* low high low

SP SP CRGP CGRP SP/CGRP

low high* low high low

V-VI

Epi

ACTH

++

++

++

ABF

+

AdCon

+

MAP

HR

++

++

+ +

++ +

++

++

-

* from Brain Research, 490 (1989) 307-319.

also evoke the greatest change in cardiovascular function. Coincident microinjections o f C G R P and substance P Subthreshold doses of CGRP (3 pmol) and substance P (13 pmol) were injected individually, in random order, and then simultaneously at each brainstem site in an additional group of 11 cats to examine the possibility of peptide interactive effects within different laminae of Vc on adrenal and autonomic function. Prior to the intitial microinjection at each site, the mean adrenal secretory rates for epinephrine, for norepinephrine and for dopamine were 6.29 _+ 1.31, 3.53 + 0.76 and 0.04 + 0.01 ng/min (n = 18), respectively, and were not affected significantly (P > 0.05) by subsequent microinjections and the corresponding sampling periods. As seen in Fig. 5, injections of individual subthreshold doses of CGRP, of substance P, or simultaneous injections within laminae I - I I (Fig. 5A), or within laminae V - V I (Fig. 5B) had no significant effect on the adrenal secretion of epinephrine. Similar results were obtained for the 7 sites located within laminae III-IV. Individual or coincident microinjections of subthreshold doses of CGRP and substance P also did not influence significantly the peripheral plasma concentrations of catecholamines, or of ACTH and did not affect adrenal blood flow regardless of the laminar site of injection (data not shown). In contrast, coincident subthreshold doses of CGRP and substance P injected into laminae I - I I or into laminae V - V I evoked significant increases in mean arterial pressure (MAP) and in heart rate, whereas individual injections alone had no effect. Prior to the initial microinjection period, MAP and heart rate averaged 102.2 + 4.1 mm Hg and 183.2 _+ 6.2 beats/min (n = 18), respectively, and remained similar for subsequent experimental periods. Simultaneous injections of CGRP

plus substance P into laminae I - I I caused an increase in MAP (P < 0.05), however, this response was not different from that due to individual injections alone (Fig. 6A). Coincident injections of CGRP and substance P into laminae V - V I (Fig. 6B) evoked an increase in MAP that was significantly greater (P < 0.05) than the response to injection of either peptide alone and was greater (P < 0.05) than that predicted from the summation of the responses to individual injections. Simultaneous injections of CGRP and substance P within laminae I - I I caused an increase in heart rate (P < 0.05), however, this response was not different from that due to individual injections (Fig. 7A). However, as was seen for MAP, simultaneous injections of CGRP and substance P within laminae V - V I evoked an increase in heart rate (Fig. 7B) that was significantly greater (P < 0.01) than that seen after injection of either peptide alone and was greater (P < 0.01) than that predicted from summation of the responses to individual injections. The apparent interaction between CGRP and substance P within laminae V - V I of Vc to modify MAP and heart rate was a consistent finding since 5 of these 6 sites responded to simultaneous subthreshold doses by evoking significant increases in MAP and in heart rate. Individual or coincident microinjections of subthreshold doses of CGRP and of substance P into laminae I l l - I V (n = 7) had no effect on MAP or on heart rate. DISCUSSION The identity of the neurotransmitters released from the central terminals of primary sensory neurons in response to nociceptor activation is not well defined; however, based on anatomical and physiological evidence, several peptides as well as excitatory amino acid

59 transmitters remain as possible candidates (see refs. 9, 15). The physiological evidence in support of putative transmitter involvement in experimental nociception has often been based on their effects on behavioral responses to noxious stimuli and on electrophysiologically characterized neural activity9'15'34 as well as on direct estimates of transmitter release to noxious stimuli 3°'4°. Alternatively, identification of the neurotransmitters that mediate the reflex adrenal and autonomic responses 1'4'38'39 to noxious stimuli may provide additional criteria for the definition of the neurochemical basis for nociception. To this end, we have examined previously the influence of microinjections of glutamate 6,7 and of substance p8 within different laminae of Vc on autonomic responses that have often been interpreted as correlates of pain sensation. Those results indicated that microinjections of glutamate or of substance P into laminae I-II of Vc increased the adrenal secretion of catecholamines, plasma concentrations of ACTH, and cardiovascular responses consistent with the location of second-order nociceptive neurons in the superficial laminae of the medullary dorsal ho1T115'34. In contrast, although microinjections of glutamate 6'7 into laminae V - V I of Vc also evoked adrenal and autonomic responses similar to those seen after injections into laminae I-II, injections of substance p8 into laminae V - V I failed to alter adrenal or autonomic function. These findings suggested that nociceptive neurons located within the deep magnocellular portion of Vc, that also respond to substance P, do not contribute to the autonomic correlates of nociception. The present study extends these results to include the effects of microinjections of CGRP into different laminae of Vc and to consider the possible interaction of CGRP with substance P in control of adrenal and autonomic function. The results indicated that microinjections of CGRP into the superficial laminae of Vc evoked significant increases in the adrenal secretion of catecholamines and in autonomic function (adrenal blood flow, adrenal vascular conductance, arterial pressure, heart rate) consistent with nociceptor activation. In contrast, injections of CGRP into the deep magnocellular laminae (V-VI) of Vc had no effect on cardiovascular function, although a decrease in the adrenal secretion of catecholamines was seen. CGRP immunoreactivity is distributed widely throughout the nervous system22"24, suggesting that CGRP is involved in various aspects of neural function. However, it is also likely that CGRP-containing neurons contribute to the encoding of noxious sensory input. Within the medullary and spinal dorsal horns, CGRPcontaining fibers and terminals are localized mainly within laminae I-II and laminae V - V | 2'11'18'24'25"42, laminae that contain the greatest percentage of nocicep-

tive neurons 15'34. Ultrastructural analyses reveal that CGRP-containing terminals make direct contact with spinothalamic tract neurons in laminae I-II and in laminae V-V112. Dorsal rhizotomy decreases markedly the density of CGRP immunostaining of terminals in the spinal cord 1s'42 suggesting that most, if not all, CGRP input to the dorsal horn derives from sensory ganglion cells. Noxious mechanical or thermal cutaneous stimuli evokes the release of CGRP from the superficial laminae of the cat spinal cord 3°. Behavioral studies indicate that intrathecally administered CGRP lowers the nociceptive threshold to mechanical stimuli 32 and increases the flexion reflex to C-fiber activation 47, whereas intrathecal pretreatment with antibodies to CGRP increases the nociceptive threshold to paw pressure in rats 26. Considerable anatomical and physiological evidence has suggested a functional relationship between CGRP and substance P, especially with respect to nociception. A significant percentage of individual trigemina117'2°'27 and dorsal root ganglion neurons 1°'18'43'44 demonstrate colocalization of immunoreactivity for CGRP and substance P. Within the medullary and spinal dorsal horns, the distribution of fibers and terminals immunoreactive for CGRP is similar to that for substance P. The greatest density of immunoreactivity for both peptides is found within laminae I-II; however, significant densities are also seen in laminae V - V I 2'14'24'25'36'42. Batten et al. 2 have noted that many fibers in the superficial and deep magnocellular layers of Vc in the cat stain positively for both CGRP and substance P. Colocalization of CGRP and substance P has also been confirmed in individual terminal varicosities in the rat superficial spinal dorsal horn 33'43. Similarly, the distribution of receptor binding for both CGRP 24 and for substance p29 is localized in those laminae of Vc that contain a significant percentage of nociceptive neurons. Noxious mechanical or thermal stimuli evoke the release of CGRP and of substance P within the substantia gelatinosa of the cat lumbar spinal cord 3°. In the rat dorsal spinal cord slice preparation, both CGRP and substance P are released after exposure to capsaicin 13'37, and both CGRP and substance P facilitate the release of excitatory amino acids evoked by electrical stimulation of dorsal rootlets 23. These findings suggest that CGRP and substance P are released coincidently from primary afferent terminals in response to noxious stimuli. To examine qualitatively the adrenal and autonomic responses evoked by microinjections of CGRP into Vc versus those responses evoked by substance 1,8, a summary is presented (Table I). After injections into laminae I-II, CGRP and substance P share the ability to increase the adrenal secretion of catecholamines, arterial pressure and heart rate. Similarly, after injections into laminae V-VI, neither CGRP

60 nor substance P evoked an increase in any of the measured variables. However, significant differences in the responses to microinjections of C G R P and substance P were noted. Although injection of either peptide alone into laminae I - I I evoked an increase in the adrenal secretion of catecholamines, C G R P also caused a significant increase in total adrenal blood flow and adrenal conductance, whereas substance P had no effect on either of these variables. The similarities in response profiles and in effective laminar locations for microinjections suggests that some of the effects of C G R P and substance P on adrenal and autonomic function may be mediated via common neural elements within Vc. Although only a single subthreshold dose of each peptide was tested, simultaneous injections of C G R P and substance P into laminae I - I I revealed no evidence for an interaction between C G R P and substance P beyond that predicted from injection of either peptide alone. In contrast, simultaneous injections of subthreshold doses of C G R P and substance P into laminae V - V I evoked increases in arterial pressure and in heart rate that exceeded the response to either peptide alone. Summary Table I suggests the existence of at least two subpopulations of peptide-responsive neurons within laminae I - I I with respect to control of adrenal and autonomic function: (1). CGRP-responsive neurons increased the adrenal secretion of catecholamines, and increased adrenal blood flow with a corresponding increase in adrenal vascular conductance, however, plasma concentrations of A C T H were not affected; and (2) substance P-responsive neurons increased the adrenal secretion of catecholamines without affecting adrenal blood flow or adrenal vascular conductance, whereas plasma concentrations of A C T H were elevated. The failure to observe a significant interaction between C G R P and substance P on adrenal secretion, on adrenal blood flow or on plasma A C T H was consistent with separate neuronal subpopulations within laminae I - I I . However, it cannot be excluded that microinjections of doses greater than those used here may have revealed significant interactive effects. The present results and those reported previously 8 indicated that injections of threshold doses of C G R P (5 pmol) or of substance P (35 pmol) alone into laminae V - V I did not evoke significant increases in adrenal or autonomic function that would be consistent with nociceptor activation, although C G R P caused a decrease in the adrenal secretion of catecholamines. Thus, it is less certain, even though simultaneous injections of subthreshold doses of C G R P (3 pmol) and substance P (13 pmol) into laminae V - V I caused increases in arterial pressure and in heart rate that exceeded those predicted from the responses to either peptide alone, that such responses may be interpreted as autonomic correlates of nociception. Elec-

trophysiological studies have shown that substance P can alter the firing rate of non-nociceptive neurons 36 and depletion of neuropeptides by capsaicin does not reduce the percentage of nociceptive neurons in the Vc of the rat 35. The mechanisms that underlie the interaction between C G R P and substance P within laminae V - V I of Vc are not known. In vitro studies have indicated that C G R P increased the capsaicin-evoked release of substance P from the dorsal horn slices32 and inhibited the degradation of substance P in human CSF 2a. The addition of C G R P or of substance P evoked the release of excitatory amino acids from dorsal spinal cord slices and potentiated the release of amino acids after electrical stimulation of dorsal root segments 23. However, in the intact dorsal spinal cord, the addition of substance P, but not of C G R P to the microdialysate fluid, increased the release of excitatory amino acids 4°. Interestingly, simultaneous addition of C G R P and substance P to the dialysate fluid potentiated the release of taurine, a putative inhibitory neurotransmitter, but did not affect the release of glutamate or aspartate compared to that seen in response to substance P alone 4°. Other studies have indicated that C G R P and substance P interact to modify nocifensive behavior 44, motor reflexes 47, and peripheral vascular control 16'17, responses that would not have been expected from the application of either peptide alone. Recent electrophysiological studies have indicated that substance P and glutamate interact to potentiate the response of lamina V spinothalamic tract neurons to mechanical stimuli 46. It is not yet known if C G R P or substance P interact with glutamate to modify the adrenal or autonomic responses that would be seen after glutamate alone. The present results argue for multiple subpopulations of peptide-responsive neurons at the level of the medullary dorsal horn in differential control of adrenal and autonomic function. An unresolved issue, not addressed by this study, remains the degree of overlap between the neural substrate that underlies the sensory-discriminative aspects of nociception and that which underlies the autonomic or affective aspects of nociception. The diversity of adrenal and autonomic response profiles seen after microinjections of individual putative neurotransmitters, is consistent with the existence of subpopulations of dorsal horn cells concerned with specific affective responses to nociception as well as additional subpopulations of cells concerned with sensation.

Acknowledgements. The authors thank Carol Cornell for assistance with the biochemical assays. This study was supported in part by NIH Grant NS-26137.

61 REFERENCES 1 Araki, T., Ito, K., Kurosawa, M. and Sato, A., Responses of adrenal sympathetic nerve activity and catecholamine secretion to cutaneous stimulation in anesthetized rats, Neuroscience, 12 (1984) 289-299. 2 Batten, T.EC., Lo, V.K.E, Maqbool, A. and McWiUiam, P.N., Distribution of calcitonin gene-related peptide-like immunoreactivity in the medulla oblongata of the cat, in relation to choline acetyltransferase-immunoreactive motoneurones and substance P-immunoreactive fibres, J. Chem. Neuroanat., 2 (1989) 163176. 3 Bereiter, D.A. and Benetti, A.P., Microinjections of CGRP within the medullary dorsal horn: effects on adrenal and autonomic function and interaction with substance P, Soc. Neurosci. Abstr., 16 (1990) 434. 4 Bereiter, D.A., Benetti, A.E and Thrivikraman, K.V., Thermal nociception potentiates the release of ACTH and norepinephrine by blood loss, Am. J. Physiol., 259 (1990) R1236-R1242. 5 Bereiter, D.A., Engeland, W.C. and Gann, D.S., Peripheral venous catecholamines versus adrenal secretory rates after brain stem stimulation in cats, Am. J. Physiol., 251 (1986) E14-E20. 6 Bereiter, D.A. and Gann, D.S., Adrenal secretion of catecholamines evoked by chemical stimulation of trigeminal nucleus caudalis in the cat, Neuroscience, 25 (1988) 697-704. 7 Bereiter, D.A. and Gann, D.S., Glutamate activation of neurons within trigeminal nucleus caudalis increases adrenocorticotropin in the cat, Pain, 33 (1988) 341-348. 8 Bereiter, D. and Gann, D.S., Substance P and GABAergic effects on adrenal and autonomic function evoked by microinjections into trigeminal subnucleus caudalis in the cat, Brain Research, 490 (1989) 307-319. 9 Besson, J.-M. and Chaouch, A., Peripheral and spinal mechanisms of nociception, Physiol. Rev., 67 (1987) 67-186. 10 Cameron, A.A., Leah, J.D. and Snow, EJ., The coexistence of neuropeptides in feline sensory neurons, Neuroscience, 27 (1988) 969-979. 11 Carlton, S.M., McNeill, D.L., Chung, K. and CoggeshaU, R.E., Organization of calcitonin gene-related peptide-immunoreactive terminals in the primate dorsal horn, J. Comp. Neurol., 276 (1988) 527-536. 12 Carlton, S.M., Westlund, K.N., Zhang, D., Sorkin, L.S. and Willis, W.D., Calcitonin gene-related peptide containing primary afferent fibers synapse on primate spinothalamic tract cells, Neurosci. Left., 109 (1990) 76-81. 13 Donnerer, J. and Amann, R., Capsaicin-evoked neuropeptide release is not dependent on membrane potential changes, Neurosci. Lett., 117 (1990) 331-334. 14 Drew, J.E, Westrum, L.E. and Ho, R.H., Mapping of the normal distribution of substance P-like immunoreactivity in the spinal trigeminal nucleus of the cat, Exp. Neurol., 93 (1986) 168-179. 15 Dubner, R. and Bennett, G.J., Spinal and trigeminal mechanisms of nociception, Annu. Rev. Neurosci., 6 (1983) 381-418. 16 Gamse, R. and Saria, A., Potentiation of tachykinin-induced plasma protein extravasation by calcitonin gene-related peptide, Eur. J. Pharmacol., 114 (1985) 61-66. 17 Gazelius, B., Edwall, B., Olgart, L., Lundberg, J.M., H6kfelt, T. and Fischer, J.A., Vasodilatory effects and coexistence of calcitonin gene-related peptide (CGRP) and substance P in sensory nerves of cat dental pulp, Acta Physiol. Scand., 130 (1987) 33-40. 18 Gibson, S.J., Polak, J.M., Bloom, S.R., Sabate, I.M., Mulderry, P.M., Ghatei, M.A., McGregor, G.P., Morrison, J.EB., Kelly, J.S., Evans, R.M. and Rosenfeld, M.G., Calcitonin gene-related peptide immunoreactivity in the spinal cord of man and of eight other species, J. Neurosci., 4 (1984) 3101-3111. 19 Gobel, S., Falls, W.M. and Hockfield, S., The division of the dorsal and ventral horns of the mammalian caudal medulla into eight layers using anatomical criteria. In D.J. Anderson and B.

Mathews (Eds.), Pain in the Trigeminal Region, Elsevier, Amsterdam, 1977, pp. 443-453. 20 Hardebo, J.E., Suzuki, N. and Owman, C., Origins of substance P- and calcitonin gene related peptide-containing nerves in the internal carotid artery of rat, Neurosci. Lett., 101 (1989) 39-45. 21 Henry, J.L., Sessle, B.J., Lucier, G.E. and Hu, J.W., Effects of substance P on nociceptive and non-nociceptive trigeminal brainstem neurons, Pain, 8 (1980) 33-45. 22 Ishida-Yamamoto, A. and Tohyama, M., Calcitonin generelated peptide in the nervous tissue, Prog. Neurobiol., 33 (1989) 335-386. 23 Kangrga, I. and Randic, M., Tachykinins and calcitonin generelated peptide enhance release of endogenous glutamate and aspartate from the rat spinal dorsal horn slice, J. Neurosci., 10 (1990) 2026-2038. 24 Kruger, L., Mantyh, P.W., Sternini, C., Brecha, N.C. and Mantyh, C.R., Calcitonin gene-related peptide (CGRP) in the rat central nervous system: patterns of immunoreactivity and receptor binding, Brain Research, 463 (1988) 223-244. 25 Kruger, L., Sternini, C., Brecha, N.C. and Mantyh, P.W., Distribution of calcitonin gene-related peptide immunoreactivity in relation to the rat central somatosensory projection, J. Comp. Neurol., 273 (1988) 149-162. 26 Kuraishi, Y., Nanayama, T., Ohno, H., Minami, M. and Satoh, M., Antinociception induced in rats by intrathecal administration of antiserum against calcitonin gene-related peptide, Neurosci. Lett., 92 (1988) 325-329. 27 Lee, Y., Kawai, Y., Shiosaka, S., Takami, K., Kiyama, H., Hillyard, C.J., Girgis, S., MacIntyre, I., Emson, P.C. and Tohyama, M., Coexistence of calcitonin gene-related peptide and substance P-like peptide in single cells of the trigeminal ganglion of the rat: immunohistochemical analysis, Brain Research, 330 (1985) 194-196. 28 Le Greves, P., Nyberg, E, Terenius, L. and H6kfelt, T., Calcitonin gene-related peptide is a potent inhibitor of substance P degradation, Eur. J. Pharmacol., 115 (1985) 309-311. 29 Mantyh, P.W. and Hunt, S.P., The autoradiographic localization of substance P receptors in rat and bovine spinal cord and the rat and cat spinal trigeminal pars caudalis and the effects of neonatal capsaicin, Brain Research, 332 (1985) 315-324. 30 Morton, C.R. and Hutchison, W.D., Release of sensory neuropeptides in the spinal cord: studies with calcitonin gene-related peptide and galanin, Neuroscience, 31 (1989) 807-815. 31 Nicholson, W.E., Davis, D.R., Sherell, B.J. and Orth, D.N., Rapid radioimmunoassay for corticotropin in unextracted human plasma, Clin. Chem., 30 (1984) 259-265. 32 Oku, R., Satoh, M., Fujii, N., Otaka, A., Yajima, H. and Takagi, H., Calcitonin gene-related peptide promotes mechanical nociception by potentiating release of substance P from the spinal dorsal horn in rats, Brain Research, 403 (1987) 350-354. 33 Plenderleith, M.B., Hailer, C.J. and Snow, P.J., Peptide coexistence in axon terminals within the superficial dorsal horn of the rat spinal cord, Synapse, 6 (1990) 344-350. 34 Price, D.D. and Dubner, R., Neurones that subserve the sensori-discriminative aspects of pain, Pain, 3 (1977) 307-338. 35 Salt, T.E., Crozier, C.S. and Hill, R.G., The effects of capsaicin pre-treatment on the response of single neurones to sensory stimuli in the trigeminal nucleus caudalis of the rat: evidence against a role for substance P as the neurotransmitter serving thermal nociception, Neuroscience, 7 (1982) 1141-1148. 36 Salt, T.E., Morris, R. and Hill, R.G., Distribution of substance P-responsive and nociceptive neurones in relation to substance P-immunoreactivity within the caudal trigeminal nucleus of the rat, Brain Research, 273 (1983) 217-228. 37 Saria, A., Gamse, R,, Peternaann, J., Fischer, J.A., Theodorsson-Norheim, E. and Lundberg, J.M., Simultaneous release of several tachykinins and calcitonin gene-related peptide from rat spinal cord slices, Neurosci. Lett., 63 (1986) 310-314. 38 Sato, A., Sato, Y. and Schmidt, R.F., Changes in blood pressure and heart rate induced by movements of normal and inflamed

62 knee joints, Neurosci. Lett., 52 (1984) 55-60. 39 Sato, A., Sato, Y. and Schmidt, R.E, Catecholamine secretion and adrenal nerve activity in response to movements of normal and inflamed knee joints in cats, J. Physiol., 375 (1986) 611-624. 40 Smullin, D.H., Skilling, S.R. and Larson, A.A., Interactions between substance P, calcitonin gene-related peptide, taurine and excitatory amino acids in the spinal cord, Pain, 42 (1990) 93-101. 41 Tashiro, T., Ruda, M.A., Satoda, T., Matsushima, R. and Mizuno, N., Convergence of serotonin-, enkephalin- and substance P-like immunoreactive afferent fibers onto cat medullary dorsal horn neurons: a triple immunocytochemical staining technique combined with the retrograde HRP-tracing method, Brain Research, 491 (1989) 360-365. 42 Traub, R.J., Solodkin, A. and Ruda, M.A., Calcitonin generelated peptide immunoreactivity in the cat lumbosacral spinal cord and the effects of multiple dorsal rhizotomies, J. Comp. Neurot., 287 (1989) 225-237. 43 Tuchscherer, M.M. and Seybold, V.S., A quantitative study of the coexistence of peptides in varicosities within the superficial

44

45 46

47

laminae of the dorsal horn of the rat spinal cord, J. Neurosci., 9 (1989) 195-205. Wiesenfeld-Hallin, Z., H6kfelt, T., Lundberg, J.M., Forssmann, W.G., Reinecke, M., Tschopp, F.A. and Fiseber, J.A., Immunoreactive calcitonin gene-related peptide and substance P coexist in sensory neurons to the spinal cord and interact in spinal behavioral responses of the rat, Neurosci. Lett., 52 (1984) 199-204. Winer, B.J., Statistical Principles in Experimental Design., 2nd edn., McGraw-Hill, New York, 1971, pp. 514-603. Willis, W.D. and Dougherty, P.M., Long-lasting enhancement of responses of spinothalamic tract (SST) cells to mechanical stimulation of the skin during potentiation of amino acid (EAA's) responses by substance P (SP), Soc. Neurosci. Abstr., 16 (1990) 161. Woolf, C. and Wiesenfeld-HaUin, Z., Substance P and calcitonin gene-related peptide synergistically modulate the gain of the nociceptive flexor withdrawal reflex in the rat, Neurosci. Left., 66 (1986) 226--230.