O.ihl-9230191 53.00
Brain Reserrrch Rrrlkrin, Vol. 26. pp. S-%3-547. 0 Pergamon Press plc, 1991. Printed in the U.S.A.
- .oo
Cholecystokinin-Like Immunoreactivity in the Rat Spinal Cord: Effects of Thoracic Transection D. ZOUAOUI,*
J. J. BENOLIEL,~
F. CESSELIN?
AND M. CONRATH”
*Laborntoire de Cytologic, hstitut des Neurosciences, CNRS URA 1199 Universite’ Pierre et Marie Curie, 9 Quai St. Bernard, 75005 Paris, France ?INSERM U 288, Luboratoire de Neurobiologie Celluiaire et Fonctionnelle FacultP de MPdecine Piti&SalpBtrit?re, 91 Boulevard de l’ffcipital, 75634 Paris Cedex 13, Frunre
Received
1.5 March
1990
%OIJAWI, D., .I. J. BENOLIEL, F. CESSELIN AND M. CONRATH. Cholec:v.srr)kinin-like irnrn~~flreucri~,;~ in the rut spinul cot-cl: E@crs of f~~rucf~ transecrion. BRAIN RES BULL 26(4) 543-547, 1991 .-A study of ~holecystokinin-like immunoreactivity in the lumbar (Ll-LS) spinal cord segments of rats was realised 24-48 hours after complete tboracic transection (TbT8). A comparison was made with corresponding spinal cord segments from control and sham-operated animals. The immuno~yto~hem~~al study with light microscopy showed cholecystokinin-like immunoreactive cell bodies in laminae VII and X at Li-L5. caudal to the transection. In addition, the immunoreactivity was greatly enhanced in bundles of the dorsolateral funiculus compared to sham-op erated animals. Our results suggest that part of cholecystokinin-like cell bodies of laminae VII and X send projections to supraspi-
nal sites. Some of these supraspinal projections would go through the dorsolateral funiculus. In the lumbar dorsal horn of operated animals, the immunoreactivity was greatly enhanced in lamina I, while it was slightly decreased in lamina II, compared to control animals. Using electron microscopy, in lamina I, the immunoreactivity localized in different neurites was generally very intense. Moreover, axon terminals showed swelling: their mean size was 0.8-l .8 pm (OS-l.2 in control animals). This result suggests that some cholecystokinin-like neurons also project to lamina I of rostra1 cervical segments. In lamina II, numerous degenerating axons were observed (24 hours after thoracic spinal transection). This would suggest that part of descending cholecyst~)kInin-like prajections terminate in lamina II.
Cholecystokinin
Rat
Spinal cord
Immunoreactivity
Projection
neurons
Light microscopy
Electron microscopy
METHOD
CHOLECYSTOKININ-LIKE (CCK-LI) immunoreactivity has been described in the rat brain (23). Later immunocytochemical studies have shown that CCK-LI is also present in the spinal cord. CCK-Li fibers in the dorsal horn have different origins: supraspinai cells in the Edinger-Westphal nucleus (9), the raphe magnus nucleus (1 1). or the periaqueductal central grey (21). local circuit neurons f,I ,24f. and peripheral cells of the spinal dorsal ganglia (2,lO). However. the existence of CCK-LI propriospinal projections cannot be dismissed. since it has been shown that many spinal ascending tract neurons send cotlaterals to rostra1 spinal segments (14). Our propose here is to verify this hypothesis and to localize the terminal sites of CCK-LI descending pathways. The thoracic transection (T6-T8) model was chosen since it allows both the accumulation of the peptide in proximal segments and the degenerescence of the distal part of the neurons. Immunocytochemical studies were performed 24-48 hours after thoracic transection, with light and electron microscopy in cervical segments (CZ-CB) rostra1 to the ~ansection and of Iumbar segments (Ll-L5) caudal to the section.
Preparation uf Animals and Tissues Twelve adult male Wistar rats (250-500 g) were studied after thoracic spinal transection and 12 normal rats (6 of them shamoperated) were used as controls. After laminectomy. transection of the spinal cord was performed at T6-T’S, under deep chloral hydrate anesthesia (350 mg/kg). The skin was then sutured and animals were studied 24 to 4X hours later. Animals were perfused via the ascending aorta with a solution containing 100 ml of 0.9% NaCl and 0.1% sodium nitrite to rinse the circulatory system. This was followed by 750 ml of 4% paraformaldehyde in 0.1 M Sijrensen buffer pH 7.4. The perfusion was done under chloral hydrate anesthesia. The spinal cord was then dissected out and postfixed overnight in the same fixative, before being washed in Siirensen buffer 0.15 M pH 7.4. Vibratome sections (50 p,m) were made from cervical (CZC6) and lumbar (Ll-L5) segments,
irnrnuni~c~t~c~ern~st~ Sections were treated using the indirect peroxidase
543
immunocy-
ZOUAOUI, BENOLIEL,CESSELlN AND CONRATH
RIG. 1. CCK-like immunoreactivity in light microscopy of lumbar spinal cord from a normal animal (l), and 48 hours after thoracic (T6) spinal transection (2-7). cc: central canal; dc: dorsal commissure; dh: dorsal horn; ILN: intennediolateral nucleus. l-In a lumbar segment of a normal rat, note the number of immunoteactive fibers in laminae I and II of the dorsal horn. Also note the moderate immunostaining in the dorsolateral funiculus (arrows). X 94. 2-Lumbar dorsal horn (at L4) 24 hours after thoracic transection. Note the great increase of immunostaining in the dorsolateral funiculus (small arrows) compared to the control animal (1). x 94. 3-Higher magnification of the bundle of immunoreactive fibers (arrows) in the dorsolatera1 funicuius. X 376. 4-Lumbar dorsal horn 48 hours after transection. Note the increased density of immunostaining in lamina I, and the decrease of CCK-like immunoreactivity in lamina II, compared to the control animal (1). X 94. 5-CCK-like fibers around cell bodies in the intermediolateral nucleus (24 ltours after transection). X 376. 6-Immunoreactive neuron of lamina VII with a long IR process (small arrows) 48 hours after transection. x 282. 7-CCK-like immunoreactive neurons in lamina X (arrows) 48 hours after transection. x 282.
545
CCK-Ll IN THE RAT SPINAL CORD
tochemical technique. After preincubation in 0. I M Sorensen buffer containing 1% normal sheep serum, they were incubated for 3 hours with the anti-CCK8 (1500). Anti-CCK8 antiserum (kindly provided by G. Tramu, INSERM) was prepared in rabbit by repeated injections of CCK8 coupled to thyroglobulin by glutaraldehyde. Sections were then rinsed in Sorensen buffer and incubated in the second antibody (anti-rabbit IgG coupled to peroxidase) (1500). After washing, the peroxidase was revealed by 3-3’diaminobenzidine. For electron microscope study, the dorsal horn was dissected out, postfixed for 2 hours in 2% 0~0, and, after dehydration, embedded in araldite. Ultrathin sections were made, stained with lead citrate and observed with a Philips EM 300 microscope. Immunocytochemical
Controls
The specificity of the reaction was controlled by preabsorption of the anti-CCK8 antiserum with the following peptides: synthetic CCK8S (l-50 )*.g/ml), pentagastrin (l-100 pg/ml), thyroglobulin (l-100 kg/ml) or calcitonin gene-related peptide (CGRP: 50 kg/ ml). Preabsorptions were performed at room temperature for 2-3 hours before addition of the sections. RESULTS In control animals, in light microscopy, numerous CCK-Ll fibers and varicosities were observed in lumbar segments in laminae I and II of the dorsal horn. Some of the CCK-Ll fibers extended to lamina Ill (Fig. l-l). Preincubation of the antiserum with thyroglobulin or CGRP did not modify the immunolabeling. In contrast, no labeling was observed after preincubation of the antiserum with synthetic CCK8, or pentagastrin. In lumbar segments from thoracic transected animals, the intensity of the immunoreaction was increased in lamina I (Fig. 1-4). and in the dorsolateral funiculus (Fig. l-2). The labeled axons (compared with control level Fig. l-l) formed a bundle and exhibited heavy accumulations of reactivity (Fig. 1-3). CCK-Ll fibers were also observed in the intermediolateral nucleus (ILN) (Fig. l-5). In contrast. in lamina II a slight decrease in the number of CCK-Ll immunoreactive fibers was observed 48 hours after thoracic transection (Fig. l-2). In addition. some CCK-Ll immunoreactive neurons were observed in lamina VII (Fig. l-6) and lamina X (Fig. l-7). They were relatively large neurons (30-40 pm), with long immunoreactive processes. The study of cervical segments (C2ZC6) rostra1 to the thoracic spinal transection showed the same immunolabeling as that in control animals. Ultrastructural study of lamina I of the lumbar dorsal horn in control animals showed immunoreactive axons and axon terminals. These neurites measured 0.5-l .2 IJ-m and often made synaptic contacts with unlabeled dendrites (Fig. 2-l). In experimental animals. the number of immunoreactive axons and axon terminals was greatly enhanced in the same region. They were strongly labeled, often swollen and contained synaptic vesicles and large granular vesicles (Fig. 2-2-4). Moreover, these profiles are larger than in control animals (0.8-1.8 km). Some of the axon terminals made synaptic contacts with unlabeled dendrites (Fig. 2-2, 3)
or with small immunoreactive axons (Fig. 2-4). In addition, labeled axons were observed in apposition with unlabeled degenerating axons (Fig. 2-2). In lamina II of lumbar segments, degenerating axons were observed 24 hours after thoracic transection (Fig. 2-5). These small degenerating axons, (0.54.8 pm) contained numerous neurofilaments (Fig. 2-5). Some of them were in apposition with immunoreactive dendrites or small axons (Fig. 2-5). In addition, primary afferent terminals or glomeruli were also observed. they were unlabeled and made synaptic contacts with strongly labeled dendrites (Fig. 2-6). DISCUSSION
Our study of the lumbar spinal cord 2448 hours after thoracic transection showed an accumulation of the reactivity in lamina I, in fibers of the dorsolateral funiculus and in numerous cell bodies of laminae X and VII, along with a decrease of the number of CCK-Ll fibers in laminae Il. The changes observed here probably do not result from a local reorganisation of spinal neurons. since the transection of the thoracic spinal cord was recent (24 to 48 hours) (3). The interpretation of this result would be that the axonal transport of CCK-Ll in ascending fibers was interrupted by the transection. This resulted in an accumulation of the peptide proximal to the transection. In fact, spinal transection, when studied soon after the operation, was comparable to the colchitine treatment. known to interrupt axonal transport (17). The exact nature of the peptide detected here is still uncertain. Contradictory results were observed between immunocytochemistry and radioimmunoassay, using animals treated at birth with capsaicin which induces a partial sensory deafferentation (12. 15, 19. 20). Indeed, it has been suggested that some anti-CCKS antisera cross-react with CGRP (4.5). However, the preabsorption of our antiserum with lop4 M CGRP did not modify the immunolabeling. In addition, biochemical controls showed that our anti-CCK was unable to recognize CGRP even when added 2.5 kg/ tube (18). Recently, we have shown in a study comparing immunocytochemistry and high pressure liquid chromatography that CCK-Ll in ascending and probably descending neurons might contain genuine CCK8 (26). The presence of CCK-Ll neurons around the central canal and lamina VII. caudal to the thoracic transection. suggested that these neurons send projections to higher spinal levels or to supraspinal sites. CCK-Ll neurons were previously observed in this region (6). It was further demonstrated that some of them project to supraspinal levels (7). The axons of the dorsolateral fasciculi cauda1 to the transection exhibit an accumulation of the reactivity. CCK-Ll fibers in this bundle may also represent ascending pathways, spinothalamic and spinoreticular projections (8). In transected animals, intense immunoreactivity of CCK-Ll fibers was observed in lamina I with light microscopy. The study of the same level with electron microscopy showed numerous strongly labeled and swollen axons and axon terminals measuring between 0.8-l .8 I.r.rn (compared with control animals 0.5-l .? pm). These data suggest that CCK-like neurons of laminae X and VII also send projections to the dorsal horn terminating or going
FOLLOWING PAGE FIG. 2. Ultrastructural study of laminae I and II at lumbar level from a normal animal (1) and from animals 24 hours after thoracic (T6) transection. (26). A: axon terminal; D: dendrite; dg: degenerating axon. 1-Immunoreactive axon terminal in lamina I in apposition with a dendrite. x 26,320. 2-Large immunoreactive axon terminal and immunoreactive axon in apposition with an unlabeled degenerating axon in lamina I. x 28.200. 3-IR axon terminals in synaptic contact (arrow) or in apposition with two unlabeled dendrites. X 26,320. 4-Immunoreactive axon terminals in lamina I. x 28,200. 5-In lamina II. note the degenerating process, probably an axon, in apposition with a labeled dendrite. Note the accumulation of nemofilaments. X 26,320. h-Nonimmunoreactive glomerular terminal (C) observed in lamina II surrounded by labeled or unlabeled dendrites and by unlabeled axon terminals. X 28.200.
546
ZOUAOUI,
BENOLIEL.
CESSELIN AIL’D CONRA’IH
547
CCK-LI IN THE RAT SPINAL CORD
through lamina 1. Indeed, MenCtrey et al. (13) have demonstrated, using retrograde axonal transport of peroxidase, that neurons of laminae X and VII send axons to other segments. We have shown here that some of them might contain CCK-LI immunoreactivity. At lumbar levels, following thoracic transection, we observed CCK-LI neurons around the central canal and CCK-LI fibers in the intermediolateral nucleus (ILN), areas known to possess sympathetic preganglionic neurons (16). This was in agreement with previous results showing that some CCK-LI fibers in lumbar level are associated with sympathetic preganglionic neurons. The decrease of CCK-LI immunoreactivity in lamina II of the lumbar dorsal horn observed in light microscopy caudal to the section, indicates that some descending fibers could project to lamina II. With electron microscopy. in lamina II, we observed numerous degenerating structures and a diminished number of small immunoreactive axon terminals (0.54.8 km) 24 hours af-
ter thoracic transection. Some of them were totally degenerated, others were at the beginning of the degenerating process. They could be distinguished by their osmophilia and dissolution of their synaptic vesicles (22). These descending fibers might come from rostra1 spinal levels or supraspinal nuclei. However, rostra1 to the thoracic transection, we were never able to visualize CCK-LI neurons. Numerous studies have demonstrated that supraspinal descending fibers come from the Edinger-Westphal nucleus (9). raphe magnus nucleus (1 I) and periaqueductal central grey (21). We have shown here that some of them might project to lamina II. ACKNOWLEDGEMENTS
We would like to thank Paulette Cloup for the photography. This research was supported by a grant from the Foundation Singer-Polignac.
REFERENCES I. Conrath-Verrier. M.: Dietl, M.: Tramu. G. Cholecystokinin-like immunoreactivity in the dorsal horn of the spinal cord of the rat: a light and electron microscopic study. Neuroscience 13:871-885; 1983. 2. Dalsgaard, C. J.; Vincent, S. R.; Hdkfelt, T.; Lundberg. J. M.; Dahlstrom, A.; Schultzberg, M.; Dockray, G.; Cuello, A. C. Coexistence of CCK and SP-like peptides in neurons of the dorsal ganglia in the rat. Neurosci. Lett. 33:159-165; 1982. 3. Gamier, J. P.; Gauthier. S.; Sourkes, T. L. Differential effects of transection of the spinal cord and splanchnic nerve on adrenal tyrosine hydroxylase and catecholamines. Neuroscience 14:907-920; 1985. 4. Hiikfelt. T.; Herrera-Marschitz. M.; Seroogy, K.; Ju, G.; Stainea, W. A.; Fischer, V.; Dockray. G.; Hamaoka, T.; Walsh, J. H.; Goldstein, M. lmmunohistochemical studies on cholecystokinin (CCK)immunoreactive neurons in the rat using sequence specific antisera and with special reference to the caudate nucleus and primary sensory neurons. J. Chem. Neuroanat. I:1 l-52; 1988. 5. Ju, G.; Hiikfelt, T.: Fischer, J. A.; Frey, P.; Rehfeld, J. F.; Dockray, G. Doer cholecystokinin-like immunoreactivity in rat primary sensory neurons represent calcitonin gene-related peptide? Neurosci. Lett. 68:305-310; 1986. 6. La Motte. C. C. Lamina X of primate spinal cord: Distribution of five neuropeptides and serotonin. Neuroscience 25:639-658; 1988. 7. Leah. J.; Mcnetrey. D.; De Pommery, J. Neuropeptides in long ascending spinal tract cells in the rat: Evidence for parallel processing of ascending information. Neuroscience 24: 197-207; 1988. 8. MC Mahon. S. B.; Wall. P. D. A system of rat spinal cord lamna I cells projecting through the contralateral dorsolateral funiculus. J. Comp. Neural. 214:217-223; 1983. 9. Maciewicz, M.: Philipps, 8. S.: Grenier, Y.; Poletti. C. E. EdingerWestphal nucleus: Cholecystokinin immunocytochemistry and projections to spinal cord and trigeminal nucleus of the cat. Brain Rec. 299:139-145; 1984, IO. Maderdrut. .l. L.; Yaksh. T. L.: Petrusz. P.; Go, V. L. W. Origin and distribution of cholecystokinin-containing nerve terminals in the lumbar dorsal horn and nucleus caudalis of the cat. Brain Res. 243: 363-368: 19X2. 1 I Mantyh. P. W. The spinothalamic tract in the primate a reexamination using uheatgerm agglutinin conjugated to horseradish peroxidase. Neuroscience 9:847-862; 1983. 12. Marley, P. D.; Nagy, J. I.: Emson, P. C.; Rehfeld, J. F. Cholecystokinin in the rat spinal cord: distribution and lack of effect of neonatal capsaicin treatment and rhizotomy. Brain Res. 238:494198; 1982. 13. Menetrey. D.; Chaouch, A.; Blinder. D.: Besson, J. M. Neurons at the origin of the spinomesencephalic tract in the rat: an anatomical study using the retrograde transport of horseradish peroxidase. J. Comp. Neural. 206: 197-207: 1982. 14. Menetrey. D.; De Pommery. J.; Roudier, J. F. Propriospinal fibers reaching the lumbar enlargement in the rat. Neurosci. Lett. 58:257261; 1985.
15. Nagy, J. I.: Hunt. S. P.: Iversen, L. L.: Emson, P. C. Biochemical and anatomical observations of the degeneration of peptide-containing primary afferent neurons after neonatal capsaicin. Neuroscience 6:1923-1934: 1981. 16. Oldfield, B. J.; Sheppard, A.; Nilaver. G. A study of substance P innervation of the intermediate zone of the thoracolumbar spinal cord. J. Comp. Neurol. 236:127-140; 1985. 17. Pickel, V. M.; Miller, M.; Chan. J.: Sumal. K. K. Substance P and enkephalin in transected axons of medulla and spinal cord. Regul. Pept. 6:121-137: 1983. 18. Pohl. M.; Benoliel. J. J.; Bourgoin. S.: Lombard, M C.: Mauborgne, A.; Taquet. H.; Carayon. A.; Besson, J. M.; Cesselin, F.: Hamon, M. Regional distribution of calcitonin gene-related peptide-, substance P-, cholecystokinin-. met-enkephalin-. and dynorphin A ( l-8)-like materials in the spinal cord and dorsal root ganglia of adult rats: effects of dorsal rhizotomy and neonatal capsaicin. J. Neurochem. 55:1122-l 131; 1990. 19. Priestley. J. V.; Bramwell. S.; Butcher. L. L.: Cuello, A. C. Effect of capsaicin on neuropeptides in areas of termination of primary sensory neurons. Neurochemistry 4:57-65; 1982. 20. Schultzberg, M.; Dokrfy,, G.; William\, R. G. Capsaicin depletes CCK-like immunoreactlvlty detected by immunocytochemistrq, but not that measured by radioimmunoassay m rat dorsal spinal cord. Brain Res. 235:198-204; 1982. 21. Skirboll, L.: HGkfelt. T.; Dokray, 0.; Rehfeld. J. F.; Brownstein. M.; Cuello, A. C. Evidence of periaqueductal cholecystokinin-substance P neurons projecting to the spinal cord. Neuroscience 3: I I5 l1158; 1983. 22. Sumal. K. K.; Pickel. V. M.: Miller, R. J.: Reis. D. J. Enkephalincontaining neurons in substantia gelatinosa of spinal trigeminal complex: ultrastructure and synaptic interaction with primary sensory afferents. Brain Res. 248:223-236; 1982. 23. Vanderhaeghen, J. J.; Deschepper. C.; Lostra, T.: Schonen, J. Immunocytochemical evidence for cholecystokinin-like peptides in neuronal cell bodies of the rat spinal cord. Cell Tissue Res. 223:463467: 1982. 24. Vanderhaeghen. J. J.; Lotstra. F.; Vierendeels. G.; Deschepper, C.: Verhaa. M.; Verbanck, P.: Gilles. C. Cholecystokinin in the central nervous system: relationship with cerebral cortex. dopaminergic and limbic systems, spinal cord and hypothalamo-hypophyseal pathways. Int. Cong. Series 568:298-31 I; 1981. 25. Williams. R. G.: Dimaline, R.: Varro. A.; I\etta. A. M.; Trizio, D.: Dockray, G. Cholecystokinin octapeptide in rat central nervous system: immunocytochemical studies using a monoclonal antibody that does not react with CGRP. Neurochem. Int. 11:433442: 1987. 26. Zouaoui. D.; Benoliel. J. J.; Conrath. M.: Cesaelin. F. Cholecystokinin-like immunoreactivity in the dorsal horn of the rat spinal cord: An attempt to analyse contradictory results between immunocytochemistry and radioimmunoassay. Neuropeptides 17: 177-185: 1990.
Brain Reserrrch Rrrlkrin, Vol. 26. pp. S-%3-547. 0 Pergamon Press plc, 1991. Printed in the U.S.A.
- .oo
Cholecystokinin-Like Immunoreactivity in the Rat Spinal Cord: Effects of Thoracic Transection D. ZOUAOUI,*
J. J. BENOLIEL,~
F. CESSELIN?
AND M. CONRATH”
*Laborntoire de Cytologic, hstitut des Neurosciences, CNRS URA 1199 Universite’ Pierre et Marie Curie, 9 Quai St. Bernard, 75005 Paris, France ?INSERM U 288, Luboratoire de Neurobiologie Celluiaire et Fonctionnelle FacultP de MPdecine Piti&SalpBtrit?re, 91 Boulevard de l’ffcipital, 75634 Paris Cedex 13, Frunre
Received
1.5 March
1990
%OIJAWI, D., .I. J. BENOLIEL, F. CESSELIN AND M. CONRATH. Cholec:v.srr)kinin-like irnrn~~flreucri~,;~ in the rut spinul cot-cl: E@crs of f~~rucf~ transecrion. BRAIN RES BULL 26(4) 543-547, 1991 .-A study of ~holecystokinin-like immunoreactivity in the lumbar (Ll-LS) spinal cord segments of rats was realised 24-48 hours after complete tboracic transection (TbT8). A comparison was made with corresponding spinal cord segments from control and sham-operated animals. The immuno~yto~hem~~al study with light microscopy showed cholecystokinin-like immunoreactive cell bodies in laminae VII and X at Li-L5. caudal to the transection. In addition, the immunoreactivity was greatly enhanced in bundles of the dorsolateral funiculus compared to sham-op erated animals. Our results suggest that part of cholecystokinin-like cell bodies of laminae VII and X send projections to supraspi-
nal sites. Some of these supraspinal projections would go through the dorsolateral funiculus. In the lumbar dorsal horn of operated animals, the immunoreactivity was greatly enhanced in lamina I, while it was slightly decreased in lamina II, compared to control animals. Using electron microscopy, in lamina I, the immunoreactivity localized in different neurites was generally very intense. Moreover, axon terminals showed swelling: their mean size was 0.8-l .8 pm (OS-l.2 in control animals). This result suggests that some cholecystokinin-like neurons also project to lamina I of rostra1 cervical segments. In lamina II, numerous degenerating axons were observed (24 hours after thoracic spinal transection). This would suggest that part of descending cholecyst~)kInin-like prajections terminate in lamina II.
Cholecystokinin
Rat
Spinal cord
Immunoreactivity
Projection
neurons
Light microscopy
Electron microscopy
METHOD
CHOLECYSTOKININ-LIKE (CCK-LI) immunoreactivity has been described in the rat brain (23). Later immunocytochemical studies have shown that CCK-LI is also present in the spinal cord. CCK-Li fibers in the dorsal horn have different origins: supraspinai cells in the Edinger-Westphal nucleus (9), the raphe magnus nucleus (1 1). or the periaqueductal central grey (21). local circuit neurons f,I ,24f. and peripheral cells of the spinal dorsal ganglia (2,lO). However. the existence of CCK-LI propriospinal projections cannot be dismissed. since it has been shown that many spinal ascending tract neurons send cotlaterals to rostra1 spinal segments (14). Our propose here is to verify this hypothesis and to localize the terminal sites of CCK-LI descending pathways. The thoracic transection (T6-T8) model was chosen since it allows both the accumulation of the peptide in proximal segments and the degenerescence of the distal part of the neurons. Immunocytochemical studies were performed 24-48 hours after thoracic transection, with light and electron microscopy in cervical segments (CZ-CB) rostra1 to the ~ansection and of Iumbar segments (Ll-L5) caudal to the section.
Preparation uf Animals and Tissues Twelve adult male Wistar rats (250-500 g) were studied after thoracic spinal transection and 12 normal rats (6 of them shamoperated) were used as controls. After laminectomy. transection of the spinal cord was performed at T6-T’S, under deep chloral hydrate anesthesia (350 mg/kg). The skin was then sutured and animals were studied 24 to 4X hours later. Animals were perfused via the ascending aorta with a solution containing 100 ml of 0.9% NaCl and 0.1% sodium nitrite to rinse the circulatory system. This was followed by 750 ml of 4% paraformaldehyde in 0.1 M Sijrensen buffer pH 7.4. The perfusion was done under chloral hydrate anesthesia. The spinal cord was then dissected out and postfixed overnight in the same fixative, before being washed in Siirensen buffer 0.15 M pH 7.4. Vibratome sections (50 p,m) were made from cervical (CZC6) and lumbar (Ll-L5) segments,
irnrnuni~c~t~c~ern~st~ Sections were treated using the indirect peroxidase
543
immunocy-
ZOUAOUI, BENOLIEL,CESSELlN AND CONRATH
RIG. 1. CCK-like immunoreactivity in light microscopy of lumbar spinal cord from a normal animal (l), and 48 hours after thoracic (T6) spinal transection (2-7). cc: central canal; dc: dorsal commissure; dh: dorsal horn; ILN: intennediolateral nucleus. l-In a lumbar segment of a normal rat, note the number of immunoteactive fibers in laminae I and II of the dorsal horn. Also note the moderate immunostaining in the dorsolateral funiculus (arrows). X 94. 2-Lumbar dorsal horn (at L4) 24 hours after thoracic transection. Note the great increase of immunostaining in the dorsolateral funiculus (small arrows) compared to the control animal (1). x 94. 3-Higher magnification of the bundle of immunoreactive fibers (arrows) in the dorsolatera1 funicuius. X 376. 4-Lumbar dorsal horn 48 hours after transection. Note the increased density of immunostaining in lamina I, and the decrease of CCK-like immunoreactivity in lamina II, compared to the control animal (1). X 94. 5-CCK-like fibers around cell bodies in the intermediolateral nucleus (24 ltours after transection). X 376. 6-Immunoreactive neuron of lamina VII with a long IR process (small arrows) 48 hours after transection. x 282. 7-CCK-like immunoreactive neurons in lamina X (arrows) 48 hours after transection. x 282.
545
CCK-Ll IN THE RAT SPINAL CORD
tochemical technique. After preincubation in 0. I M Sorensen buffer containing 1% normal sheep serum, they were incubated for 3 hours with the anti-CCK8 (1500). Anti-CCK8 antiserum (kindly provided by G. Tramu, INSERM) was prepared in rabbit by repeated injections of CCK8 coupled to thyroglobulin by glutaraldehyde. Sections were then rinsed in Sorensen buffer and incubated in the second antibody (anti-rabbit IgG coupled to peroxidase) (1500). After washing, the peroxidase was revealed by 3-3’diaminobenzidine. For electron microscope study, the dorsal horn was dissected out, postfixed for 2 hours in 2% 0~0, and, after dehydration, embedded in araldite. Ultrathin sections were made, stained with lead citrate and observed with a Philips EM 300 microscope. Immunocytochemical
Controls
The specificity of the reaction was controlled by preabsorption of the anti-CCK8 antiserum with the following peptides: synthetic CCK8S (l-50 )*.g/ml), pentagastrin (l-100 pg/ml), thyroglobulin (l-100 kg/ml) or calcitonin gene-related peptide (CGRP: 50 kg/ ml). Preabsorptions were performed at room temperature for 2-3 hours before addition of the sections. RESULTS In control animals, in light microscopy, numerous CCK-Ll fibers and varicosities were observed in lumbar segments in laminae I and II of the dorsal horn. Some of the CCK-Ll fibers extended to lamina Ill (Fig. l-l). Preincubation of the antiserum with thyroglobulin or CGRP did not modify the immunolabeling. In contrast, no labeling was observed after preincubation of the antiserum with synthetic CCK8, or pentagastrin. In lumbar segments from thoracic transected animals, the intensity of the immunoreaction was increased in lamina I (Fig. 1-4). and in the dorsolateral funiculus (Fig. l-2). The labeled axons (compared with control level Fig. l-l) formed a bundle and exhibited heavy accumulations of reactivity (Fig. 1-3). CCK-Ll fibers were also observed in the intermediolateral nucleus (ILN) (Fig. l-5). In contrast. in lamina II a slight decrease in the number of CCK-Ll immunoreactive fibers was observed 48 hours after thoracic transection (Fig. l-2). In addition. some CCK-Ll immunoreactive neurons were observed in lamina VII (Fig. l-6) and lamina X (Fig. l-7). They were relatively large neurons (30-40 pm), with long immunoreactive processes. The study of cervical segments (C2ZC6) rostra1 to the thoracic spinal transection showed the same immunolabeling as that in control animals. Ultrastructural study of lamina I of the lumbar dorsal horn in control animals showed immunoreactive axons and axon terminals. These neurites measured 0.5-l .2 IJ-m and often made synaptic contacts with unlabeled dendrites (Fig. 2-l). In experimental animals. the number of immunoreactive axons and axon terminals was greatly enhanced in the same region. They were strongly labeled, often swollen and contained synaptic vesicles and large granular vesicles (Fig. 2-2-4). Moreover, these profiles are larger than in control animals (0.8-1.8 km). Some of the axon terminals made synaptic contacts with unlabeled dendrites (Fig. 2-2, 3)
or with small immunoreactive axons (Fig. 2-4). In addition, labeled axons were observed in apposition with unlabeled degenerating axons (Fig. 2-2). In lamina II of lumbar segments, degenerating axons were observed 24 hours after thoracic transection (Fig. 2-5). These small degenerating axons, (0.54.8 pm) contained numerous neurofilaments (Fig. 2-5). Some of them were in apposition with immunoreactive dendrites or small axons (Fig. 2-5). In addition, primary afferent terminals or glomeruli were also observed. they were unlabeled and made synaptic contacts with strongly labeled dendrites (Fig. 2-6). DISCUSSION
Our study of the lumbar spinal cord 2448 hours after thoracic transection showed an accumulation of the reactivity in lamina I, in fibers of the dorsolateral funiculus and in numerous cell bodies of laminae X and VII, along with a decrease of the number of CCK-Ll fibers in laminae Il. The changes observed here probably do not result from a local reorganisation of spinal neurons. since the transection of the thoracic spinal cord was recent (24 to 48 hours) (3). The interpretation of this result would be that the axonal transport of CCK-Ll in ascending fibers was interrupted by the transection. This resulted in an accumulation of the peptide proximal to the transection. In fact, spinal transection, when studied soon after the operation, was comparable to the colchitine treatment. known to interrupt axonal transport (17). The exact nature of the peptide detected here is still uncertain. Contradictory results were observed between immunocytochemistry and radioimmunoassay, using animals treated at birth with capsaicin which induces a partial sensory deafferentation (12. 15, 19. 20). Indeed, it has been suggested that some anti-CCKS antisera cross-react with CGRP (4.5). However, the preabsorption of our antiserum with lop4 M CGRP did not modify the immunolabeling. In addition, biochemical controls showed that our anti-CCK was unable to recognize CGRP even when added 2.5 kg/ tube (18). Recently, we have shown in a study comparing immunocytochemistry and high pressure liquid chromatography that CCK-Ll in ascending and probably descending neurons might contain genuine CCK8 (26). The presence of CCK-Ll neurons around the central canal and lamina VII. caudal to the thoracic transection. suggested that these neurons send projections to higher spinal levels or to supraspinal sites. CCK-Ll neurons were previously observed in this region (6). It was further demonstrated that some of them project to supraspinal levels (7). The axons of the dorsolateral fasciculi cauda1 to the transection exhibit an accumulation of the reactivity. CCK-Ll fibers in this bundle may also represent ascending pathways, spinothalamic and spinoreticular projections (8). In transected animals, intense immunoreactivity of CCK-Ll fibers was observed in lamina I with light microscopy. The study of the same level with electron microscopy showed numerous strongly labeled and swollen axons and axon terminals measuring between 0.8-l .8 I.r.rn (compared with control animals 0.5-l .? pm). These data suggest that CCK-like neurons of laminae X and VII also send projections to the dorsal horn terminating or going
FOLLOWING PAGE FIG. 2. Ultrastructural study of laminae I and II at lumbar level from a normal animal (1) and from animals 24 hours after thoracic (T6) transection. (26). A: axon terminal; D: dendrite; dg: degenerating axon. 1-Immunoreactive axon terminal in lamina I in apposition with a dendrite. x 26,320. 2-Large immunoreactive axon terminal and immunoreactive axon in apposition with an unlabeled degenerating axon in lamina I. x 28.200. 3-IR axon terminals in synaptic contact (arrow) or in apposition with two unlabeled dendrites. X 26,320. 4-Immunoreactive axon terminals in lamina I. x 28,200. 5-In lamina II. note the degenerating process, probably an axon, in apposition with a labeled dendrite. Note the accumulation of nemofilaments. X 26,320. h-Nonimmunoreactive glomerular terminal (C) observed in lamina II surrounded by labeled or unlabeled dendrites and by unlabeled axon terminals. X 28.200.
546
ZOUAOUI,
BENOLIEL.
CESSELIN AIL’D CONRA’IH
547
CCK-LI IN THE RAT SPINAL CORD
through lamina 1. Indeed, MenCtrey et al. (13) have demonstrated, using retrograde axonal transport of peroxidase, that neurons of laminae X and VII send axons to other segments. We have shown here that some of them might contain CCK-LI immunoreactivity. At lumbar levels, following thoracic transection, we observed CCK-LI neurons around the central canal and CCK-LI fibers in the intermediolateral nucleus (ILN), areas known to possess sympathetic preganglionic neurons (16). This was in agreement with previous results showing that some CCK-LI fibers in lumbar level are associated with sympathetic preganglionic neurons. The decrease of CCK-LI immunoreactivity in lamina II of the lumbar dorsal horn observed in light microscopy caudal to the section, indicates that some descending fibers could project to lamina II. With electron microscopy. in lamina II, we observed numerous degenerating structures and a diminished number of small immunoreactive axon terminals (0.54.8 km) 24 hours af-
ter thoracic transection. Some of them were totally degenerated, others were at the beginning of the degenerating process. They could be distinguished by their osmophilia and dissolution of their synaptic vesicles (22). These descending fibers might come from rostra1 spinal levels or supraspinal nuclei. However, rostra1 to the thoracic transection, we were never able to visualize CCK-LI neurons. Numerous studies have demonstrated that supraspinal descending fibers come from the Edinger-Westphal nucleus (9). raphe magnus nucleus (1 I) and periaqueductal central grey (21). We have shown here that some of them might project to lamina II. ACKNOWLEDGEMENTS
We would like to thank Paulette Cloup for the photography. This research was supported by a grant from the Foundation Singer-Polignac.
REFERENCES I. Conrath-Verrier. M.: Dietl, M.: Tramu. G. Cholecystokinin-like immunoreactivity in the dorsal horn of the spinal cord of the rat: a light and electron microscopic study. Neuroscience 13:871-885; 1983. 2. Dalsgaard, C. J.; Vincent, S. R.; Hdkfelt, T.; Lundberg. J. M.; Dahlstrom, A.; Schultzberg, M.; Dockray, G.; Cuello, A. C. Coexistence of CCK and SP-like peptides in neurons of the dorsal ganglia in the rat. Neurosci. Lett. 33:159-165; 1982. 3. Gamier, J. P.; Gauthier. S.; Sourkes, T. L. Differential effects of transection of the spinal cord and splanchnic nerve on adrenal tyrosine hydroxylase and catecholamines. Neuroscience 14:907-920; 1985. 4. Hiikfelt. T.; Herrera-Marschitz. M.; Seroogy, K.; Ju, G.; Stainea, W. A.; Fischer, V.; Dockray. G.; Hamaoka, T.; Walsh, J. H.; Goldstein, M. lmmunohistochemical studies on cholecystokinin (CCK)immunoreactive neurons in the rat using sequence specific antisera and with special reference to the caudate nucleus and primary sensory neurons. J. Chem. Neuroanat. I:1 l-52; 1988. 5. Ju, G.; Hiikfelt, T.: Fischer, J. A.; Frey, P.; Rehfeld, J. F.; Dockray, G. Doer cholecystokinin-like immunoreactivity in rat primary sensory neurons represent calcitonin gene-related peptide? Neurosci. Lett. 68:305-310; 1986. 6. La Motte. C. C. Lamina X of primate spinal cord: Distribution of five neuropeptides and serotonin. Neuroscience 25:639-658; 1988. 7. Leah. J.; Mcnetrey. D.; De Pommery, J. Neuropeptides in long ascending spinal tract cells in the rat: Evidence for parallel processing of ascending information. Neuroscience 24: 197-207; 1988. 8. MC Mahon. S. B.; Wall. P. D. A system of rat spinal cord lamna I cells projecting through the contralateral dorsolateral funiculus. J. Comp. Neural. 214:217-223; 1983. 9. Maciewicz, M.: Philipps, 8. S.: Grenier, Y.; Poletti. C. E. EdingerWestphal nucleus: Cholecystokinin immunocytochemistry and projections to spinal cord and trigeminal nucleus of the cat. Brain Rec. 299:139-145; 1984, IO. Maderdrut. .l. L.; Yaksh. T. L.: Petrusz. P.; Go, V. L. W. Origin and distribution of cholecystokinin-containing nerve terminals in the lumbar dorsal horn and nucleus caudalis of the cat. Brain Res. 243: 363-368: 19X2. 1 I Mantyh. P. W. The spinothalamic tract in the primate a reexamination using uheatgerm agglutinin conjugated to horseradish peroxidase. Neuroscience 9:847-862; 1983. 12. Marley, P. D.; Nagy, J. I.: Emson, P. C.; Rehfeld, J. F. Cholecystokinin in the rat spinal cord: distribution and lack of effect of neonatal capsaicin treatment and rhizotomy. Brain Res. 238:494198; 1982. 13. Menetrey. D.; Chaouch, A.; Blinder. D.: Besson, J. M. Neurons at the origin of the spinomesencephalic tract in the rat: an anatomical study using the retrograde transport of horseradish peroxidase. J. Comp. Neural. 206: 197-207: 1982. 14. Menetrey. D.; De Pommery. J.; Roudier, J. F. Propriospinal fibers reaching the lumbar enlargement in the rat. Neurosci. Lett. 58:257261; 1985.
15. Nagy, J. I.: Hunt. S. P.: Iversen, L. L.: Emson, P. C. Biochemical and anatomical observations of the degeneration of peptide-containing primary afferent neurons after neonatal capsaicin. Neuroscience 6:1923-1934: 1981. 16. Oldfield, B. J.; Sheppard, A.; Nilaver. G. A study of substance P innervation of the intermediate zone of the thoracolumbar spinal cord. J. Comp. Neurol. 236:127-140; 1985. 17. Pickel, V. M.; Miller, M.; Chan. J.: Sumal. K. K. Substance P and enkephalin in transected axons of medulla and spinal cord. Regul. Pept. 6:121-137: 1983. 18. Pohl. M.; Benoliel. J. J.; Bourgoin. S.: Lombard, M C.: Mauborgne, A.; Taquet. H.; Carayon. A.; Besson, J. M.; Cesselin, F.: Hamon, M. Regional distribution of calcitonin gene-related peptide-, substance P-, cholecystokinin-. met-enkephalin-. and dynorphin A ( l-8)-like materials in the spinal cord and dorsal root ganglia of adult rats: effects of dorsal rhizotomy and neonatal capsaicin. J. Neurochem. 55:1122-l 131; 1990. 19. Priestley. J. V.; Bramwell. S.; Butcher. L. L.: Cuello, A. C. Effect of capsaicin on neuropeptides in areas of termination of primary sensory neurons. Neurochemistry 4:57-65; 1982. 20. Schultzberg, M.; Dokrfy,, G.; William\, R. G. Capsaicin depletes CCK-like immunoreactlvlty detected by immunocytochemistrq, but not that measured by radioimmunoassay m rat dorsal spinal cord. Brain Res. 235:198-204; 1982. 21. Skirboll, L.: HGkfelt. T.; Dokray, 0.; Rehfeld. J. F.; Brownstein. M.; Cuello, A. C. Evidence of periaqueductal cholecystokinin-substance P neurons projecting to the spinal cord. Neuroscience 3: I I5 l1158; 1983. 22. Sumal. K. K.; Pickel. V. M.: Miller, R. J.: Reis. D. J. Enkephalincontaining neurons in substantia gelatinosa of spinal trigeminal complex: ultrastructure and synaptic interaction with primary sensory afferents. Brain Res. 248:223-236; 1982. 23. Vanderhaeghen, J. J.; Deschepper. C.; Lostra, T.: Schonen, J. Immunocytochemical evidence for cholecystokinin-like peptides in neuronal cell bodies of the rat spinal cord. Cell Tissue Res. 223:463467: 1982. 24. Vanderhaeghen. J. J.; Lotstra. F.; Vierendeels. G.; Deschepper, C.: Verhaa. M.; Verbanck, P.: Gilles. C. Cholecystokinin in the central nervous system: relationship with cerebral cortex. dopaminergic and limbic systems, spinal cord and hypothalamo-hypophyseal pathways. Int. Cong. Series 568:298-31 I; 1981. 25. Williams. R. G.: Dimaline, R.: Varro. A.; I\etta. A. M.; Trizio, D.: Dockray, G. Cholecystokinin octapeptide in rat central nervous system: immunocytochemical studies using a monoclonal antibody that does not react with CGRP. Neurochem. Int. 11:433442: 1987. 26. Zouaoui. D.; Benoliel. J. J.; Conrath. M.: Cesaelin. F. Cholecystokinin-like immunoreactivity in the dorsal horn of the rat spinal cord: An attempt to analyse contradictory results between immunocytochemistry and radioimmunoassay. Neuropeptides 17: 177-185: 1990.
