0306-4522/92 $5.00 + 0.00 Pergamon Press Ltd ~ 1992 IBRO

Neuroscience Vol. 50, No. 4, pp. 921 933, 1992 Printed in Great Britain

R E G U L A T I O N OF OPIOID B I N D I N G SITES IN THE S U P E R F I C I A L D O R S A L H O R N OF THE R A T S P I N A L C O R D F O L L O W I N G LOOSE L I G A T I O N OF THE SCIATIC NERVE: C O M P A R I S O N WITH SCIATIC N E R V E SECTION AND LUMBAR DORSAL RHIZOTOMY D. BESSE,* M. C. LOMBARD, S. PERROT a n d J. M. BESSON Unit6 de Recherche de Physiopharmacologie du Systrme Nerveux (INSERM, U. 16l) and Laboratoire de Physiopharmacologie de la Douleur, Ecole Pratique des Hautes Etudes, 2 rue d'Alrsia, 75014 Paris, France A~tract--The aim of the present study was to quantify time-related modifications in ~ and 6 opioid binding sites in the superficial layers (laminae I and II) of the L4 lumbar segment in a rat model of mononeuropathy induced by loose ligation of the sciatic nerve. We have shown a 28% (P < 0.01) and 24% (P < 0.01) decrease in ipsi/contralateral side binding ratios for tritiated (Tyr*-D-Ala-Gly-NMe-PheGly-ol) ([3H]DAMGO) and tritiated (Tyr*-D-Thr-Gly-Phe-Leu-Thr) ([3H]DTLET) respectively, at two weeks postlesion which correspond to the delay of maximal hyperalgesia and of maximal alteration of fine diameter primary afferent fibers. In contrast, no change in [3H]U.69593 specific binding could be detected at this postlesion delay. For longer survival delays (four, eight and 15 weeks postlesion), p and 6 binding ratios return towards control values (approximately equal to 1), probably reflecting the occurrence of a long-term neuroplasticity (i.e. a new equilibrium in the metabolism of primary neurons, or collateral sprouting from intact primary afferents) following loose nerve ligation. In addition, a comparison of the results obtained in this model with those measured after sciatic nerve section and lumbar dorsal rhizotomy was performed in order to compare the degree of loss in opioid binding sites in these three types of lesion. The section of the sciatic nerve induced at eight days postlesion an 18% (P < 0.01) and 28% (P < 0.01) decrease in binding ratio for [3H]DAMGO and [3H]DTLET, respectively. At two weeks postlesion the loss was 24% (P < 0.01) for the two ligands, and at longer delays (four and 12 weeks), a progressive recovery in binding ratio was observed. Thus, it appears that both sciatic nerve lesions we have studied result i n / t and 6 binding modifications which have similar intensity and similar time course from two to 12 15 weeks postlesion. In contrast, the unilateral rhizotomy of nine consecutive dorsal roots (T13-$2), which is known to induce a massive degeneration of fine diameter primary afferent fibers, is followed by a dramatic decrease in binding ratios for [3H]DAMGO (53%, P < 0.001) and [3H]DTLET (45%, P < 0.001) at two weeks postlesion. These data suggest that the more deprived the dorsal horn is of fine diameter primary afferent fibers, the more dramatic is the opioid binding loss in the ipsilateral side as compared to the contralateral side. However, when considering specific binding concentration, it must be emphasized that regulation can occur independently of a direct alteration of primary afferent fibers. A significant up-regulation in /~ binding sites was found at every postlesion delay following nerve loose ligation and only at eight days following nerve section. This up-regulation seems to be bilateral although, on the ipsilateral side, this effect could be masked by the fiber degeneration. We propose that this up-regulation in # binding sites could result from the activation of endogenous opioid control systems due to the alteration of nociceptive sensory input. Our results are discussed in view of the recent controversy arising on the efficacy of opiates in neuropathic pains.

N u m e r o u s studies have reported a decrease in opioid binding sites in laminae I a n d II of the spinal dorsal h o r n after deafferentation induced by dorsal rhizotomy 15'24'32'37'45'55'56'75 or peripheral nerve section. 37 T a k e n as a whole, it appears t h a t the intensity of the decrease in binding on the side ipsilateral to the lesion *To whom correspondence should be addressed. Abbreviations: CGRP, calcitonin gene-related peptide; [3H]DAMGO, tritiated (Tyr*-D-Ala-Gly-NMe-Phe-Glyol); [3H]pC1-DPDPE, tritiated (Tyr*-D-Pen-Gly-C1-PheD-Pen); [3H]DTLET, tritiated (Tyr*-D-Thr-Gly-PheLeu-Thr); LSD, Fisher's Least Significant Difference multiple range test; SP, substance P; [3H]U.69593, tritiated (5ct,7e,8/~)-( - )-N-methyl-N-[7-(1-pyrrolidinyl)- 1oxaspiro(4,5)dec-8-yl]benzeneacetamide. 921

is related to the degree of deafferentation. ~7 The loss of opioid binding sites could, in part, explain why some clinical reports claim that the efficacy of opioid substances is weaker against neuropathic, a n d particularly, deafferentation pains t h a n against pains due to an excess of n o c i c e p t i o n ] '43'65 In some cases, however, m o r p h i n e t r e a t m e n t has been reported to be efficient in some types o f n e u r o p a t h i c pains. 58'6° Recently, a new model of m o n o n e u r o p a t h i c pain in rat has been p r o p o s e d by Bennett a n d Xie. ~2 In this model, the n e u r o p a t h y is induced by placing four loose ligatures a r o u n d one sciatic nerve. This procedure produces a chronic b u t reversible nerve constriction. Behavioral studies p e r f o r m e d o n this model have described a pain-related behavior, with mainly

922

D. BEsst! et al.

spontaneous symptoms and both hyperalgesia and allodynia in response to mechanical and thermal stimuli. 7"~2 These behavioral alterations are relevant to pain sensations in humans with peripheral nerve damage.~4 According to the extensive investigation by Attal et al., 7 time courses of pain-related disorders are similar whatever the behavioral test: following an early period of hypoalgesia at five days posttesion hyperalgesia develops, being maximal at two weeks postlesion, with a recovery occurring around two months postlesion. Interestingly, systemic morphine induces dose-dependent naloxone-reversible antinociceptive effects on spontaneous pain and mechanical allodynia at two weeks postlesion. 8"36 As concerns morphological alterations, the sciatic nerve constriction is known to induce an extensive demyelination of primary afferent fibers in the distal portion of the nerve as compared to the proximal one. 12'31 Similarly, nociceptive unmyelinated C fibers appear to be damaged in the distal portion of the nerve. H'2° Proximal alterations of these fibers also seem to occur since, using immunocytochemical techniques, Bennett et al. t3 have observed at 20 days postlesion a reduction in the number of small cells with substance P (SP) and calcitonin gene-related peptide ( C G R P ) immunoreactivities in dorsal root ganglia associated with the sciatic nerve (L4, L5). In parallel, marked depletion of these two peptide immunoreactivities was observed in laminae I and II of the spinal dorsal horn, where fine diameter primary afferent fibers essentially project. These depletions were confirmed by radioimmunoassays showing that, at both 10 and 20 days postlesion, SP and C G R P contents in the dorsal horn are decreased on the side of the nerve damage by comparison to the control side. Since the majority of opioid binding sites in laminae I and II of the dorsal horn have been shown to be associated with primary afferent fibers 28"4s'72 and particularly with fine diameter fibers, 24'3°'54 w e were therefore interested to study the modifications of/~, 6 and x opioid binding sites following a sciatic nerve constriction. Actually, a recent autoradiographic study from Stevens et al. 6~ using selective ligands has shown complex modifications of opioid binding sites in various regions of the spinal cord within the first 10 days following nerve constriction. This period corresponds to the acute phase of the disease. Since this experimental model presents multiphasic behavioral manifestations for several months, our purpose was, firstly, to analyse the time-related plasticity of opioid binding sites at various postlesion delays, including long-term delays. Another aim was to look for opioid binding changes at various postlesion delays following a sciatic nerve section, which would be expected to induce a more marked degeneration of fine diameter primary afferent fibers than does the nerve constriction. Finally, results of these two experimental series were compared to those obtained after a large unilateral lumbar dorsal rhizotomy T13-$2

which causes a total degeneration of thc central process of primary sensory neurons. We used quantitative autoradiography and the highly selective ligands tritiated (Tyr*-D-Ala-Gly-NMe-Phe-Gly-ol) ([3H]DAMGO), tritiated (Tyr*-o-Thr-Gly-Phe-LeuThr) ([3H]DTLET), or tritiated (Tyr*-o-Pen-Gly-C1Phe-D-Pen) ([3H]pCI-DPDPE), and [3H]]U(695~3t0 study #, 6 and x opioid binding sites respectively in laminae I and II of the spinal dorsal horn. A preliminary report of this study has been presented, a~

EXPERIMENTAL PROCEDURES Animal and tissue preparation

Experiments were performed on adult male Sprague-Dawley albino rats weighing 250-300 g (Charles Rivers, France), according to the guidelines of the Committee for Research and Ethical Issues of the IASP. 23 Rats were housed one to six per cage with sawdust bedding, and given food and water ad libitum. Loose ligation of the sciatic nerve. The procedure was comparable to the original description] 2 Rats were anesthetized with sodium pentobarbital (50 mg/kg i.p.). Four loose ligations (5.0 chromic gut) were tied around the common sciatic nerve at the level of the mid-thigh, so that circulation through the epineural vasculature was not totally interrupted. In each rat, the same dissection was performed on the opposite side, but the sciatic nerve was not ligated. All animals were operated on at the same age. During the first five postoperative days, rats were housed in individual cages, and thereafter six to a cage. Animals with a sciatic nerve constriction were tested using a modified version of Randall-Selitto's method adapted by Kayser and Guilbaud. ~° All the rats were tested before surgery and 1 h before being killed. In addition, in order to be sure that the animals used for long postlesion delays exhibited hyperalgesia, they were tested repeatedly during the survival delay. Only rats having demonstrated behavioral reactions corresponding to those reported by Attal et aL 7 in this experimental model were used in our autoradiographic study. Briefly, they presented hypoalgesia to mechanical stimulations up until five days. Then, hyperalgesia peaked at two weeks postlesion and progressively declined, with a complete recovery at eight weeks postlesion. Three experimental groups were investigated: intact rats (n = 6); sham rats (shamA) submitted to an exposure of the sciatic nerve on the left side, the right side being left intact (n = 4); mononeuropathic rats (lig.), submitted to a sciatic loose ligation on the right side and to a sham operation on the left side (n = 18). These last two groups were studied at the following postlesion delays: five days, two, four, eight and 15 weeks (n = 3-45 per delay). Sham A rats were studied only at two weeks after surgery. Section o f the sciatic nerve. Animals were anesthetized with ketamine (100 mg/kg i.p.). The right sciatic nerve was exposed through an incision of the skin in the mid-thigh and a separation of the muscle fibers. Two ligatures (3 mm spaced) of 4-0 black silk were tied tightly around the nerve. It was then cut between the two ligatures and a 2-ram-long portion was excised. After repositioning of the nerve in the thigh, muscle and skin were closed by suture. The other side was not operated on. Several survival delays were studied: eight days, two, four and 12 weeks (n = 3 rats per group). Sham rats were the same as in the case of the sciatic nerve ligation. Lumbar dorsal rhizotomy. The surgical procedure was performed using ketamine (100 mg/kg i.p.) as general and xylocaine (2%) as local anesthetics. A dorsal hemilaminectomy of four vertebrae (T13-L3) was made on the right side and nine dorsal roots (T13-$2) were sectioned intradurally

Opioid receptor regulation in various neuropathic states proximal to the spinal cord, through an opening of the dura mater, under a dissection microscope. The wound was washed with saline. Before closing the wound, a piece of Parafilm was placed on the surface of the spinal cord and the skin was sutured. In sham-operated rats (shamB), a similar procedure was performed, but dorsal roots were not sectioned. The survival delay following dorsal rhizotomy or sham operation was two weeks (n = 3 rats per group). At the end of the survival time, rats were killed by decapitation. The lumbar enlargements were removed rapidly, frozen in isopentane at --40°C and preserved at -80'('.

Autoradiographic procedure Tritiated ligands used in this study were [3H]DAMGO (1.85 TBq/mmol; CEA Saclay, France), [3H]DTLET (2.22 TBq/mmol; CEA Saclay, France), [3HlpCI-DPDPE (1.58 TBq/mmol; NEN Research Products) and [3H]U.69593 (1.61 TBq/mmol; NEN Research Products). Unlabeled levorphanol (tartrate) was purchased from Sigma. Spinal cord lumbar enlargements were cut from L3 to L5 (frontal sections 20/~m thick) on a cryostat ( - 2 O C ) , thaw-mounted onto gelatinized slides and kept at -80~C up to incubation. Three sets of slides, each bearing adjacent sections, were prepared: one for [3H]DAMGO, one for [3H]DTLET (or [3HlpCI-DPDPE) and the last one for [3H]U.69593. This latter ligand was only used in rats with a sciatic nerve constriction, at two weeks postlesion. Spinal cord sections were brought to room temperature. For experiments at constant concentration, sections were incubated in 350 ml of 50 mM Tris HCI buffer (pH 7.4) at 2 5 C for 60 rain, containing either 3 nM [3H]DAMGO to label V sites, 3 nM [3H]DTLET (or [3HlpCI-DPDPE) to label 6 sites, or 5nM [3H]U.69593 to label • sites. The nonspecific binding was looked for on some spinal cord sections under the same conditions of incubation, in the presence of 1 VM levorphanol. Saturation experiments were performed with [3H]DAMGO and [3H]pCI-DPDPE, in similar conditions of incubation, but seven concentrations ranging from 0.5 to 20nM were used for total and non-specific binding. At the end of the incubation, spinal cord sections were washed twice in 350 ml of ice-cold 50 mM Tris HC1 buffer (pH 7.4) for 10rain each, rinsed in ice-cold bidistillated water for 2 s and air-dried. Labeled sections were then juxtaposed against tritium-sensitive films (Hyperfilm-3H, Amersham) and exposed at 4'C during 15 weeks for [3H]DAMGO and [3H]DTLET (or [3H]pC1-DPDPE), and 20 weeks for [3H]U.69593. Autoradiograms were developed in Kodak D19 for 2 min at 20'~C, fixed for 5 min, washed for 20 min under running water and dried. Quantitative analysis Autoradiograms were analysed by densitometry using a Biocom 200 analyser system (Biocom, France). We measured binding in the area corresponding to the whole mediolateral extent of Rexed's laminae I and lI as defined for the rat by Molander et al.5~ We purposely chose to measure binding in the whole area in order to make more reproducible measurements and to be able to compare the effects of the three types of lesion. Actually, the morphology of the spinal cord fluctuates as a function of the inter-animal variability, the age of animals and the severity of the lesion. Consequently, the boundaries of the area of measurement corresponding to laminae I and II were fixed according to Nissl poststained sections. Moreover, in order to accurately provide a representative binding value for each rat in each spinal segment for a given ligand, 10-15 different spinal cord sections were successively analysed every 200 ~m on both sides of the spinal cord in each segment from L3 to L5 and averaged. These segments have been chosen according to the fact that this is the main site of projection of the sciatic nerve 25,64

923

Gray levels measured on autoradiograms were converted into binding site concentration values by reference to tritium standards (Amersham). Specific binding was calculated by substracting the amount of non-specific binding from total binding. For experiments using constant concentrations of ligand, results were expressed as specific binding concentrations (fmol/mg of tissue) on each side and as ipsi/contralateral side binding ratios in order to diminish inter-animal variations. For saturation experiments, binding parameters at equilibrium such as specific binding capacity (Bmax) and dissociation constant (Kd) were determined from the linear regression analysis of Scatchard. For each animal, the binding value corresponding to a given concentration was the mean of binding values obtained from five different spinal cord sections. Statistical analysis was performed using either the tpaired test, or an ANOVA, followed by the Least Significant Difference (LSD) multiple range test. ~9~ RESULTS AS expected, in intact rats, [ 3 H ] D A M G O specific binding was densest in the superficial layers of the dorsal h o r n (laminae I and II). The distribution of [3H]DTLET a n d [3HlpC1-DPDPE is more h o m o geneous in the gray substance, but is still predomin a n t in laminae I a n d II. The c o m p a r i s o n of the binding of these two ligands shows that 3 n M [3H]pCI-DPDPE binds to a b o u t 35% fewer sites t h a n does [3H]DTLET at the same concentration. For [3H]U.69593, a l t h o u g h very weak, the specific binding seems mainly related to the two superficial laminae. S h a m operations, consisting of separation of muscle fibers a n d exposure of the nerve in the case of the sciatic nerve lesions (shamA), or h e m i l a m i n e c t o m y a n d incision of the d u r a m a t e r in the case of the dorsal rhizotomy (shamB), did not produce any effect on opioid binding as c o m p a r e d to intact rats. Consequently, intact rats have been chosen as reference for statistical analysis of results o b t a i n e d in each experimental group. Sciatic nerve a n d dorsal root lesions induced modifications o f / t a n d 3 binding sites in the superficial layers of the dorsal horn. A l t h o u g h m e a s u r e m e n t s were p e r f o r m e d from L3 to L5, the maximal effect invariably being found in L4 segment, only results o b t a i n e d in this segment are presented in this study. Individual examples of [3H]DAMGO and [3H]DTLET binding are shown in Fig. 1 for sham A rats a n d rats with (i) a right sciatic nerve loose ligation, (ii) a right sciatic nerve section a n d (iii) a right T 1 3 - $ 2 dorsal rhizotomy. With regard to [3H]U.69593 binding, individual examples are s h o w n in Fig. 2 only for shamA rats a n d rats with a sciatic nerve loose ligation.

Effects induced by a loose ligation o f the sciatic nerve Autoradiography of [3H]DAMGO (3 nM), [3H]DTLET (3 riM), [3H]pC1-DPDPE (3 riM) a n d [3H]U.69593 (5 n M ) specific binding shows complex modifications of p, 3 a n d ~c binding sites at five days, two, four, eight a n d 15 weeks, in laminae l a n d II of the L4 segment.

924

D. BESSEet al.

p Sites. Modifications in [3H]DAMGO specific binding are observed on both sides, ipsilateral (ANOVA, Fdd~= ~,15= 4.5, P < 0.01) and contralateral (ANOVA, Fdj~=5,15=6.3, P ~ , . %..,..,., d . ,

..,.,, -.-'

Regulation of opioid binding sites in the superficial dorsal horn of the rat spinal cord following loose ligation of the sciatic nerve: comparison with sciatic nerve section and lumbar dorsal rhizotomy.

The aim of the present study was to quantify time-related modifications in mu and delta opioid binding sites in the superficial layers (laminae I and ...
4MB Sizes 0 Downloads 0 Views