- Email: [email protected]
Calcitonin gene-related peptide- and substance P-containing primary afferent fibers in the dorsal column of the rat
Brain Research, 495 (It)89) 122-13(; Elsevier
122
BRES 14731)
Calcitonin gene-related peptide- and substance P-containing primary afferent fibers in the dorsal column of the rat Michio Tamatani, Emiko Senba and Masaya Tohyama Department of Anatomy (11), Osaka University Medical School, Osaka (Japan) (Accepted 31 January 1989)
Key words: Calcitonin gene-related peptide; Substance P: Dorsal column; Primary afferent fiber; Rat
The dorsal column and its nuclei exhibit a considerable number of fibers containing neuropeptides, such as calcitonin gene-related peptide (CGRP) and substance P (SP), whose origins and functional roles are as yet unknown. The present study attempts to determine the origin and nature of these fibers by means of immunohistochemistry combined with several experimental manipulations. A similar study was done on scattered substance P (SP) fibers whose presence was confirmed in this study. Transection of the upper cervical cord of rats resulted in an accumulation of CGRP, sometimes with SP also, in the caudal aspect of the lesion, thus indicating the presence of peptide-containing ascending fibers. Hemitransection of the dorsal column at the level of C2_3 caused reduction of CGRP-containing fibers in the dorsal column and its nuclei on the operated side. Electron microscopic observation of the nucleus gracilis revealed that CGRP-Iike immunoreactive terminals made direct axodendritic synaptic contacts. Medium- to large-sized neurons in the dorsal root ganglia were labeled with Fast blue dye which was injected into the dorsal column nuclei. These included medium- to large-sized neurons exhibiting immunoreactivity to CGRP-like substances, and neurons of a medium size which were immunoreactive to SP-like compounds. The incidence of the former was higher at the thoracic level than at the cervical and lumbar levels, while that of the latter was very low. Electron microscopic observation of CGRP-containing fibers in the cervical region of the dorsal column revealed that 88% of these fibers were unmyelinated and the remainder were thinly myelinated. These results indicate that medium- to large-sized CGRP- (and SP)-containing neurons in the dorsal root ganglia give rise to fine fibers into the dorsal column and may terminate in the dorsal column nuclei. INTRODUCTION
for calcitonin m R N A 1'27, and is distributed widely throughout the nervous system 19'27. In the dorsal
It is generally accepted that unmyelinated and thinly myelinated primary afferent fibers originating from small type B cells in the dorsal root ganglia ( D R G ) terminate in the superficial dorsal horn of the spinal cord, whereas a n u m b e r of myelinated fibers derived from type A cells ascend the dorsal column to terminate in the dorsal column nuclei4'35.
root ganglia, medium- to large-sized neurons, as well as the small type B cells, show immunoreactivity to CGRP-Iike substances; some are also immunoreactive to substance P (SP)-like agents 18'21. In addition to dense CGRP-Iike immunoreactive (1R) and substance P (SP)-IR fibers in the superficial dorsal horn, the presence of these fibers in the dorsal column and its nuclei was reported in the rat and h u m a n 13a9'22.
The latter pathway has been largely neglected in chemical neuroanatomical studies of recent years, because of the relative paucity of chemically detectable substances associated with it. In contrast, the type B cell-dorsal horn pathway exhibit a variety of neuropeptides ls.2s. Calcitonin gene-related peptide ( C G R P ) is a neuropeptide incidentally discovered during analyses
Moreover, fine structure of these fibers was precisely described in the recent study 24, in which the authors suggested that these fibers originate from small type B cells in the D R G . However, it is more reasonable to consider that larger C G R P - I R primary afferent neurons project to the dorsal column nuclei via a dorsal column
Correspondence: E. Senba, Department of Anatomy (II), Osaka University Medical School, 4-3-57 Nakanoshima, Kita-ku, Osaka 530, Japan. 0006-8993/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)
123 pathway in view of the generally accepted anatomical knowledge 4'35. This investigation was undertaken to resolve this problem and to elucidate central projections and terminations of the CGRPIR sensory neurons in the dorsal column pathway by means of experimental immunohistochemistry and immunoelectron microscopy. MATERIALS AND METHODS
Animals and tissue preparation Twenty five male Wistar rats each weighing approximately 100 g were used. The animals were divided into 4 groups as follows. (1) One group (n = 6) was prepared for subsequent analysis of the distribution of CGRP-IR and SP-IR structures in the dorsal column and its nuclei in the medulla oblongata. In some animals (n = 2), colchicine (100/~g/100 g b.wt.) was injected into the cisterna magna 2 days prior to their sacrifice. (2) The second group (n = 8) was used to determine the operating direction of CGRP-IR and SP-IR fibers passing through the dorsal column. An incision was made in the dorsal column at the cervical level (C2_3), or at the mid-thoracic level, using sterilized surgical knives with animals under deep anesthesia induced by pentobarbital sodium (50 mg/kg i.p.). These rats were kept alive for 2-3 days. Subsequently, samples from most were subjected to light microscopic or, in some cases, electron microscopic analysis for the presence of CGRP-IR fibers accumulated at the site of the lesion. (3) The third group of animals (n = 3) underwent hemitransection of the dorsal column at the level of the cervical spinal cord (C2_3), and were kept alive for 14 days, in order to allow detection of changes in the CGRP-IR fibers of the dorsal column and its nuclei in the rostral part of the lesion. (4) In the fourth group (n = 8), 0.02/A of a 3% solution (in distilled water) of the fluorescent dye Fast blue (FB), was injected into the dorsal column nuclei of rats under pentobarbital sodium anesthesia (50 mg/kg i.p.). This was achieved using glass micropipettes with tip diameters of 20-50 pm, connected to a microinjector. Animals were immobilized in a stereotaxic instrument and a glass micropipette was inserted into the dorsal column
through a slit opened in the membrane covering the cisterna magna. Three days later, some of these rats were anesthetized again and given an injection of colchicine (100 ~g/100 g b.wt.) through a thin polyethylene tube inserted into the subarachnoid space through a slit in the membrane above the cisterna magna, and lowered to the cervical, thoracic or lumbar levels, respectively. These animals, with or without colchicine treatment, were kept alive for 5 days from the time of FB injection. A separate group of animals (n = 2) were prepared for the electron microscopic analysis of CGRP-IR fibers and terminals in the nucleus gracilis. Rats of all 4 groups were placed under deep pentobarbital sodium anesthesia, and perfused via the ascending aorta with 0.9% saline followed by 400 ml of either Zamboni's fixative3.7, or modified Zamboni's fixative containing 4% paraformaldehyde and 0.2% picric acid in 0.1 M phosphate buffer (pH 7.4). For the specimens to be examined by electron microscopy, 0.05% glutaraldehyde was added to the above solutions. Brains, and spinal cords complete with dorsal roots and dorsal root ganglia were excised and postfixed in the same fixative overnight and then transferred to 0.1 M phosphate buffer containing 30% sucrose, where they were left until they sank. Transverse sections (20/~m thick) of the medulla oblongata and spinal cord were cut in a cryostat and collected in a free-floating state. Sections of D R G were sliced at a thickness of 10-15 pm in a cryostat and mounted on gelatin-coated glass slides. For electron microscopic analysis, tissue blocks were treated with liquid nitrogen and sections (40 pm thick) were cut on a Vibratome.
Immunohistochemistry Light microscopic observation. Free-floating and slide-mounted sections of normal, surgically treated, and FB-injected animals were subjected to the indirect immunofluorescence method of Coons 7 for the detection of CGRP or SP-like immunoreactivities. In some samples, CGRP and SP were detected simultaneously using a double-labeling immunofluorescence method. Sections were first incubated with the respective primary antibodies (polyclonal anti-CGRP antiserum, Amersham; polyclonal anti-SP antiserum
124 raised by us) or in the mixture of the primary antibodies (polyclonal anti-CGRP antiserum, Amersham; monoclonal anti-SP antiserum, Seralab, MAS 035) diluted with 0.02 M phosphatebuffered saline (PBS) (1:1000 dilution for all the antisera) at 4 °C for 2 days. The samples were then washed in cold PBS for 30 min and incubated overnight at 4 °C in PBS containing fluorescein isothiocyanate (F1TC)-conjugated goat anti-rabbit lgG (Miles, 1 : 1000 dilution), or in a mixture of Texas red (TR)-conjugated donkey anti-rabbit IgG (Amersham, 1:250 dilution) and F1TC-conjugated goat anti-rat IgG (Cappel, 1:1000 dilution) for double-labeling. After a final washing with PBS, the sections were mounted, coverslipped with glycerolPBS (1:1), and viewed under a Nikon fluorescence microscope equipped with B2, G and V excitation filters, which permit exclusive visualization of FITC, TR and FB fluorescence, respectively. Control experiments were carried out for some samples by incubating them with primary antisera preabsorbed with l0 ~M of synthetic SP or CGRP (Peptide Inst., Osaka). The specificities of the antisera used in the present study have been described in previous reports 8'15'~1. Electron microscopic observation. Sections for electron microscopic study were processed by the peroxidase-antiperoxidase (PAP) method of Sternberger 33, as reported by Somogyi and Takagi 32. Briefly, they were preincubated in 0.1 M PBS containing 10% normal goat serum (NGS) before being incubated with anti-CGRP antiserum diluted with 0.1 M PBS containing 1% NGS and 1% bovine serum albumin (BSA) at 4 °C for 2 days. Subsequently, the samples were soaked in the dilution mixture for a few hours, incubated with goat anti-rabbit IgG antiserum (Cappel, 1:200 dilution) overnight at 4 °C, and incubated overnight again with rabbit PAP complex (DAKO, 1:500 dilution) overnight at the same temperature. After washing and further rinsing in 0.05 M Tris-HCl buffer (pH 7.6) incubation was carried out for 5-15 min in the same buffer containing 0.02%, 3,3"-diaminobenzidine (DAB) and 0.005% hydrogen peroxide. After washing in Tris-buffer, sections were fixed and stained in 1% osmium tetroxide in 0.1 M phosphate buffer and 1% uranyl acetate in 70% ethyl alcohol. Samples were dehydrated and horizontally embed-
ded in resin on silicon-coated glass slides, and serial ultrathin sections were examined under an electron microscope (Hitachi, H-7000). For the quantitative analysis of immunoreactive fibers, the lateral part of the fasciculus gracilis including the medial part of the fasciculus cuneatus were photographed randomly. Numbers of unmyelinated and myelinated C G R P - I R fibers were counted on the printed photos. RESULTS
Distribution of CGRP-IR fibers in the dorsal column and dorsal column nuclei In the upper cervical region of the dorsal column, CGRP-IR fibers were observed in the fasciculus gracilis and in the dorsomedial border of the fasciculus cuneatus (Fig. ld). The fibers in the latter area seemed to be continuous with small fiber bundles located at the dorsal border of the fasciculus cuneatus (Fig. lc). A prominent aggregation of immunoreactive fiber bundles was observed in the rostral part of the fasciculus cuneatus at the level of area postrema (Fig. lb). Examination of the dorsal column nuclei revealed the presence of C G R P - I R varicose fibers distributed in the dorsal column nuclei with those in the nucleus cuneatus being finer than those in the nucleus gracilis. Most of these fibers seemed to originate from C G R P - I R fiber bundles in the dorsal border of the fasciculus cuneatus (Figs. la,c). Within the nucleus gracilis, CGRP-IR fibers were more numerous in the peripheral regions with the central area of the nucleus usually being devoid of immunoreactive fibers (Fig. la-c). The distribution of SP-IR fibers in the dorsal column and dorsal column nuclei was similar to that of CGRP-IR fibers, although they were fewer compared to C G R P - I R ones. Most of these SP-IR fibers were shown to also contain CGRP. No CGRP-IR or SP-IR neurons were observed in the dorsal column nuclei even in the colchicine-treated animals. No CGRP-IR or SP-IR structures were found in control sections. Transection of the spinal cord Transection of the dorsal column at the C2_ 3 o r mid-thoracic level resulted in an accumulation of CGRP-IR substance in fibers at the caudal aspect of
125
Fig. 1. Fluorescent photomicrographs showing the distribution of CGRP-IR fibers in the dorsal column nuclei at the level of area postrema (a-c) and dorsal column of upper cervical cord. Note a prominent aggregationof CGRP-IR fibers in the fasciculus cuneatus (Cu) (thick arrow) and fine varicose fibers in the nucleus cuneatus (cu) (thin arrows). G, fasciculus gracilis; g, nucleus gracilis; nts, nucleus tractus solitarius. Scale bars; 100/lm (a-c), 50/~m (d).
the lesion. Some of these fibers also displayed immunoreactivity to SP. No substances immunoreactive to CGRP or SP were observed in the rostral part of the lesion. Hemitransection of the dorsal column brought about a decrease in the number of CGRP-IR fibers detected in the dorsal column and its nuclei on the operated side as compared to the control side. Small fiber bundles in the fasciculus gracilis and dorsal border of the fasciculus cuneatus were markedly less evident, while a few immunoreactive fibers persisted in the dorsal column nuclei.
Fluorescent dye tracer and immunohistochemistry Since findings from knife-cut lesion experiments indicated that CGRP-IR fibers in the dorsal column were ascending fibers, they were presumed to originate from the D R G cells or intraspinal neurons.
On this basis, FB was injected into the dorsal column bilaterally or unilaterally at the level of pyramidal decussation. The injection site involved and, in most cases, was restricted to the fasciculus gracilis and cuneatus, and also the most caudal part of the nucleus gracilis. FB-labeled cells were observed in the D R G of any of the spinal levels examined (cervical to lumbar), but none, or very few, were found in the D R G of the contralateral side in the case of unilateral injection. When injection sites involved the medullary dorsal horn or adjacent reticular formation, but excluded the dorsal column and its nuclei, no FB-labeled cells were detected in the DRG. These cases served as controls for vascular diffusion of this dye. FB-labeled cells in the DRGs were exclusively medium (20-40 /~m in diameter) (55%) to large (>40/~m in diameter) (45%) in size (Fig. 2c,d). About 40% of the D R G cells exhibited
126
Fig. 2. Fluorescent photomicrographs showing CGRP-IR (a) and SP-IR (b) neurons in the dorsal root ganglia (L4_5). Cells labeled with Fast blue (FB), which was injected into the dorsal column nuclei, are also shown for the same sections (c,d). Note that FB-labeled cells are exclusively medium- to large-sized cells, some of which show CGRP- or SP-like immunoreactivities (arrows). Scale bar; 100 ,urn.
C G R P - l i k e immunoreactivity, and 3 5 - 5 0 % of these were classified as m e d i u m - to large-sized using the same criteria as for F B - l a b e l e d D R G cells. The r e m a i n d e r were small-sized cells ( < 2 0 ~tm in diameter). A b o u t 2 - 7 % of the medium- to large-sized C G R P - I R cells were simultaneously labeled with FB (Fig. 2a,c), whereas no double-labeling of small C G R P - I R neurons occurred. The incidence of the d o u b l e - l a b e l e d cells in the D R G of the thoracic region was 2 - 3 times higher than those at the cervical or l u m b a r levels (Table I). SP-IR mediumsized neurons were occasionally labeled with FB (Fig. 2b,d). In the colchicine-treated rats, transverse sections of the spinal cord at various levels were processed for immunohistochemical detection of C G R P - l i k e
immunoreactivity. Most of the F B - l a b e l e d cells were located in lamina IV and some of them were scattered in lamina X. This distribution p a t t e r n varied little between the spinal levels, although m o r e labeled cells were observed in the cervical and l u m b a r enlargements. O n the other hand, in addition to the C G R P - I R m o t o n e u r o n s d e t e c t e d in the ventral horn throughout the length of the spinal cord, C G R P q R neurons were discovered in the i n t e r m e d i o m e d i a l (dorsal gray commissure) and intermediolateral cell columns in the t h o r a c o l u m b a r (C8-L2) levels, but only in the i n t e r m e d i o l a t e r a l cell column in the lumbosacral (L6-S1) levels. Identical findings were p r e s e n t e d in a previous study by the current authors 31. No d o u b l e - l a b e l e d cells were observed in the spinal cord.
127 TABLE I Number of CGRP-like immunoreactive (IR) cells and double-labeled cells in the dorsal root ganglia at various spinal levels
The last column shows the percentage of medium- to large-sized CGRP-IR cells that are double-labeled at each level. Spinal level
CGRP-IR cells (small)
CGRP-IR cells~ (medium-large)
FB-labeled CGRP-IR cells (small)
FB-labeled CGRP-1R cells b (medium-large)
b/a (%)
Cervical Upper thoracic (TI_6) Lower thoracic (T7_13) Lumbar
267 (63.1%) 564 (65.7%) 410 (51.2%) 271 (53.0%)
156 (36.9%) 294 (34.3%) 390 (48.8%) 240 (47.0%)
0 0 0 0
4 19 16 5
2.6 6.5 4.1 2.1
Electron m i c r o s c o p i c observation
C G R P - I R structures in the transverse sections of the dorsal column at the caudal aspect of the lesion (C2_3) were examined by electron microscopy. Immunoreactive fibers were distributed in the fasciculus gracilis and dorsomedial part of the fasciculus cuneatus. Twenty electron microscopic photographs were taken randomly. Of 420 C G R P - I R axons identified on these photos, 368 were unmyelinated (87.6%) and 52 were thinly myelinated (12.4%) (Fig. 3). C G R P - I R axon terminals in the nucleus gracilis were also examined. Observation was focused in the dorsomedial part of this nucleus at the level of area postrema. Most of these terminals were small in
diameter (0.6-1.2 ~tm; mean, 0.86 # m ) and formed direct synaptic contacts of asymmetric type with dendrites whose diameters were not larger than 1.2 /~m (Fig. 4a,b). Some large C G R P - I R axon terminals or swellings (2.0-2.6 ~tm) were also observed (Fig. 4c). DISCUSSION The dorsal column of the rat is far more heterogenous than previously supposed. It is comprised of long ascending and descending fibers which are listed as follows: (1) primary afferent fibers; (2) second-order postsynaptic dorsal column fibers 3°, (3) descending fibers from the dorsal column nucleiS;
Fig. 3. An electron microscopic photograph showing CGRP-IR fibers in the dorsal column. Note that most of these fibers are unmyelinated and some of them are thinly myelinated (arrows). Scale bar: 1/am.
128 and (4) corticospinal fibers. The former two types are ascending fibers and the latter two are descending. In addition to these long forms, about 25% of the dorsal column fibers are estimated to be propriospinal ascending and descending fibers 6.
Fig. 4. Electron microscopicphotographs showing CGRP-IR terminals in the nucleus gracilis. Most of these terminals are small and make direct synaptic contacts (arrowheads) with small dendrites (D) (a,b). CGRP-IR large boutons are occasionally observed (c). Scale bars: 0.5 ~um (a,b), 1 #m (c).
The present immunohistochemical study clearly shows that all CGRP-IR fibers in the dorsal column are of the ascending type, and some at least are primary afferent fibers originating from medium- to large-sized CGRP-IR neurons in the DRG. It was further demonstrated that a rather small number of medium- to large-sized CGRP-containing D R G cells (less than 10%) reach the dorsal column nuclei. This seems reasonable in view of the report by Glees and Soler ~° which suggests that relatively few fibers in the dorsal column (about 25%) reach the dorsal column nuclei, the remainder seeming to enter the spinal gray during their course of ascent. However, the possibility that a proportion of the ascending CGRP-IR fibers originate intraspinally cannot be discounted, because it has been demonstrated that at least 15% of the ascending axons in the dorsal column are non-primary afferent forms zg. These fibers have been shown to arise from neurons in laminae IIl and IV in various species 3'23'3°'34, although no comparative anatomical data concerning this is available for the rat. Unfortunately, CGRPIR dorsal column postsynaptic (DCPS) neurons could not be detected by the experimental procedures employed in the present study. Recently, McNeill et al. 24 reported that about 95% of CGRPIR fibers in the dorsal column disappear after dorsal rhizotomy. These findings suggest that DCPS neurons contribute very little to CGRP-IR dorsal column pathway. CGRP-IR fibers were found to be distributed throughout the entire length of the dorsal column nuclei. This distribution pattern corresponds closely with that of ascending fibers or primary afferent fibers in the dorsal column of the rat 2'9 and the cat 12 as determined by degeneration methods. The authors of these reports suggest that primary afferent fibers terminate with even distribution between the rostral and caudal aspects of the nuclei. The present study further showed that CGRP-IR fibers in the dorsal column and its nuclei decreased in number following transection of the dorsal column. These findings, taken together, indicate that most CGRPIR fibers are ascending axons and terminate in the dorsal column nuclei, although additional innervation from some other central region cannot be excluded at this juncture. It is generally accepted that primary afferent
129 fibers in the dorsal column convey information from several different types of receptors in the skin, muscles and joints 36. However, little is known about the functional role of the dorsal column pathway involving CGRP. It may be that these fibers are associated with the transmission of certain types of sensory modalities, such as proprioception, tactile sensation or visceral sensation. Differences in the routes by which proprioceptive information is transmitted from muscles and joints are reflected by variation in fiber composition in the fasciculus cuneatus and fasciculus gracilis, i.e. the fasciculus gracilis above the upper lumbar cord is devoid of proprioceptive fibers 4. Therefore, since C G R P - I R fibers were observed in both fasciculi at all spinal levels examined, it is less likely that these convey proprioceptive signals. In recent investigations, the current authors determined that large myelinated fibers innervating Meissner's tactile corpuscles and Merkel's discs are devoid of CGRP-Iike immunoreactivity 16'~7. Thus, it is unlikely that CGRP-IR primary afferent fibers in the dorsal column transmit impulses from low threshold mechanoreceptors in the skin. The higher incidence of CGRP-IR sensory cells projecting to the dorsal column nuclei in the thoracic region shown in the present study may indicate that some of these CGRP-IR fibers in the dorsal column convey visceral information to the central nervous system. In support of this speculation, it has been shown that some visceral afferent fibers pass through dorsal column and reach the dorsal column nuclei 26 and also that most of the visceral sensory neurons are CGRP-IR 11,25. REFERENCES
Kuo et al. have shown that most of the visceral afferent fibers (90%) in the cat are unmyelinated2°; the classical view has been that the dorsal column contains no unmyelinated fibers 35. However, this has also been challenged by Chung et al. 6 in an anatomical study which revealed that a considerable proportion (23%) of primary afferent fibers in the dorsal column are unmyelinated. In line with this, McNeil124 et al. reported the presence of CGRP-IR fine fibers in the dorsal column of the rat. They showed that 60-70% of these fibers are unmyelinated, which is lower than the incidence we have shown in the present study. This discrepancy may be explained by the difference of levels and regions of the dorsal column examined or technical differences of immunocytochemistry. Anyway, it is obvious from these studies that most of the CGRP-IR fibers in the dorsal column are unmyelinated or thinly myelinated. McNeill et al. 24 considered that these fine fibers are derived from small type B cells in the D R G which convey nociceptive information. However, we clearly demonstrated in the present study that these fibers originate from medium- to largesized D R G cells. This implies that a subpopulation of medium- to large-sized D R G neurons gives rise to unmyelinated fibers, and coincides well with the recent electrophysiological and morphological study 14 which showed unmyelinated afferent fibers do not originate exclusively from small D R G cells. ACKNOWLEDGEMENT This study was supported in part by a grant from the Ministry of Education of Japan. cord from medullary somatosensoryrelay nuclei, J. Comp. Neurol., 173 (1977) 773-792.
1 Amara, S.G., Jonas, V., Rosenfeld, M.G., Ong, E.S. and Evans, R.M., Alternative RNA processing in calcitonin gene expression generate mRNAs encoding different polypeptide products, Nature (Lond.), 298 (1982) 240-244. 2 Basbaum, A.I. and Hand, P.J., Projections of cervicothoracic dorsal roots to the cuneate nucleus of the rat, with observations on cellular 'Bricks', J. Cornp. Neurol., 148 (1973) 347-360. 3 Brown, A.G. and Fyffe, R.E.W., Form and function of dorsal horn neurones with axons ascending the dorsal columns in cat, J. Physiol. (Lond.), 321 (1981) 31-47. 4 Brown, A.G. and Gordon, G., Subcortical mechanisms concerned in somatic sensation, Br. Med. Bull., 33 (1977) 122-128. 5 Burton, H. and Loewy, A.D., Projections to the spinal
6 Chung, K., Langford, L.A. and Coggeshall, R.E., Primary afferent and propriospinal fibers in the rat dorsal and dorsolateral funiculi, J. Comp. Neurol., 263 (1987) 68-75. 7 Coons, A.H., Fluorescent antibody methods. In J.F. Danielli (Ed.), General Cytochemical Method, Academic, New York, 1958, pp. 399-422. 8 Cueilo, A.C., Gaifre, G. and Milstein, C., Detection of substance P in the central nervous system by a monoclonal antibody, Proc. Natl. Acad. Sci. U.S.A., 76 (1979) 35323536. 9 Ganchrow, D. and Bernstein, J.J., Projections of caudal fasciculus gracilis to nucleus gracilis and other medullary structures, and Clarke's nucleus in the rat, Brain Research, 205 (1981) 383-390. 10 Glees, P. and Soler, J., Fibre content of the posterior
130
11
12
13
14
15
16
17
18
19
20
21
22
column and synaptic connections of nucleus gracilis, Z. Zellforsch., 36 (1951) 381-400. Green, T. and Dockray, G.J., Calcitonin gene-related peptide and substance P in afferents to the upper gastrointestinal tract in the rat, Neurosci. Lett., 76 (1987) 151-156. Hand, P.J., Lumbosacral dorsal root terminations in the nucleus gracilis of the cat, J. Comp. Neurol., 126 (1966) 137-156. Harmann, P.A., Chung, K., Briner, R.P., Westlund, K.N. and Carlton, S.M., Calcitonin gene-related peptide (CGRP) in the human spinal cord: a light and electron microscopic analysis, J. Comp. Neurol., 269 (1988) 371380. Hoheisel, U and Mense, S., Non-myelinated afferent fibers do not originate exclusively from the smallest dorsal root ganglion cells in the cat, Neurosci. Lett.. 72 (1986) 153-157. Inagaki, S., Sakanaka, M., Shiosaka, S., Senba, E., Takatsuki, K., Takagi, H., Kawai, Y., Minagawa, H. and Tohyama, M., Ontogeny of substance P-containing neuron system of the rat: immunohistochemical analysis. I. Forebrain and upper brainstem, Neuroscience, 7 (1982) 251277. Ishida-Yamamoto, A., Senba, E. and Tohyama, M., Calcitonin gene-related peptide and substance P immunoreactive nerve fibers in Meissner's corpuscles of rats: an immunohistochemical analysis, Brain Research, 453 (1988) 362-366. Ishida-Yamamoto, A., Senba, E. and Tohyama, M., Distribution and fine structure of calcitonin gene-related peptide-like immunoreactive nerve fibers in the rat skin, Brain Research, in press. .lu. G., H6kfelt, T., Brodin, E., Fahrenkrug, J., Fisher, J.A., Frey, P., Elde, R.P. and Brown, J.C., Primary sensory neurons of the rat showing calcitonin gene-related peptide immunoreactivity and their relation to substance P-, somatostatin-, galanin-, vasoactive intestinal polypeptide- and cholecystokinin-immunoreactive ganglion cells, Cell Tissue Res., 247 (1987) 417-431. Kawai, Y., Takami, K., Shiosaka, S., Emson, P.C., Hillyard, C.J., Girgis, S., Maclntyre, I. and Tohyama, M., Topographic localization of calcitonin gene-related peptide in the rat brain: an immunohistochemical analysis, Neuroscience, 15 (1985) 747-763. Kuo, D.C., Yang, G.C.H., Yamasaki, D.S. and Krauthamer, G.H., A wide field electron microscopic analysis of the fiber constituents of the major splanchnic nerve in cat, J. Comp. Neurol., 210 (1982) 49-58. Lee, Y., Takami, K., Kawai, Y., Girgis, S., Hillyard, C.J., Maclntyre, I., Emson, P.C. and Tohyama, M., Distribution of calcitonin gene-related peptide in the rat peripheral nervous system with reference to its coexistence with substance P, Neuroscience, 15 (1985) 1227-1237. Ljungdahl, A., H6kfelt, T. and Nilsson, G., Distribution of substance P-like immunoreactivity in the central nervous system of the rat. I. Cell bodies and nerve terminals,
Neuroscience, 3 (1978) 861-943. 23 Lu, G.-W., Bennett, G.J., Nishikawa, N., Hoffert, M.J. and Dubner, R., Extra- and intracellular recordings from dorsal column postsynaptic spinomedullary neurons in the cat, Exp, NeuroL, 82 (1983) 456-477. 24 McNeill, D.L., Chung, K., Carlton, S.M. and Coggeshall, R.E., Calcitonin gene-related peptide immunostained axons provide evidence for fine primary afferent fibers in the dorsal and dorsolateral funiculi of the rat spinal cord, J. Comp, Neurol., 272 (1988) 303-308. 25 Molander, C., Ygge, J. and Dalsgaard, C.J., Substance P-, somatostatin- and calcitonin gene-related peptide-like immunoreactivity and fluoride resistant acid phosphataseactivity in relation to retrogradely labeled cutaneous, muscular and visceral primary sensory neurons in the rat, Neurosci. Lett., 74 (1987) 37-42. 26 Neuhuber, W., The central projections of visceral primary afferent neurons of the inferior mesenteric plexus and hypogastric nerve and the location of the related sensory and preganglionic sympathetic cell bodies in the rat, Anat. Embryol., 164 (1982) 413-425. 27 Rosenfeld, M.G., Mermod, J.J., Amara, S.G., Swanson, L.W., Sawchenko, P.E., Rivier, J., Vale, W.W. and Evans, R.M., Production of novel neuropeptide encoded by the calcitonin gene via tissue specific RNA processing, Nature (Lond.), 304 (1983) 129-135. 28 Ruda, M.A., Bennett, G.J. and Dubner, R., Neurochemistry and neural circuitry in the dorsal horn, Progr. Brain Res., 66 (1986) 219-268. 29 Rustioni, A., Non-primary afferents to the nucleus gracilis from the lumbar cord of the cat, Brain Research, 51 (1973) 81-95. 30 Rustioni, A., Spinal neurons project to the dorsal column nuclei of rhesus monkeys, Science, 196 (1977) 656-658. 31 Senba, E. and Tohyama, M., Calcitonin gene-related peptide containing autonomic afferent pathways to the pelvic ganglia of the rat, Brain Research, 449 (1988) 386-390. 32 Somogyi, P. and Takagi, H., A note on the use of picric acid-paraformaldehyde-glutaraldehyde fixative for correlated high and electron microscopic immunocytochemistry, Neuroscience, 7 (1982) 1779-1783. 33 Sternberger, L.A., lmmunocytochemistry, 2nd edn,, Wiley, New York, 1979. 34 Uddenberg, N., Functional organization of long, secondorder afferents in the dorsal funiculus, Exp. Brain Res., 4 (1968) 377-382. 35 Webster, K.E., Somaesthetic pathways, Br. Med. Bull., 33 (1977) 113-120. 36 Willis, W.D., and Coggeshall, R.E., Sensory pathways in the dorsal columns. In W.D. Willis and R.E. Coggeshall (Eds.), Sensory Mechanisms of the Spinal Cord, Plenum, New York, 1978, pp. 197-259. 37 Zamboni, L. and De Martino, C., Buffered picric acid formaldehyde: a new rapid fixative for electron microscopy, J. Cell. Biol., 35 (1967) 148A.
122
BRES 14731)
Calcitonin gene-related peptide- and substance P-containing primary afferent fibers in the dorsal column of the rat Michio Tamatani, Emiko Senba and Masaya Tohyama Department of Anatomy (11), Osaka University Medical School, Osaka (Japan) (Accepted 31 January 1989)
Key words: Calcitonin gene-related peptide; Substance P: Dorsal column; Primary afferent fiber; Rat
The dorsal column and its nuclei exhibit a considerable number of fibers containing neuropeptides, such as calcitonin gene-related peptide (CGRP) and substance P (SP), whose origins and functional roles are as yet unknown. The present study attempts to determine the origin and nature of these fibers by means of immunohistochemistry combined with several experimental manipulations. A similar study was done on scattered substance P (SP) fibers whose presence was confirmed in this study. Transection of the upper cervical cord of rats resulted in an accumulation of CGRP, sometimes with SP also, in the caudal aspect of the lesion, thus indicating the presence of peptide-containing ascending fibers. Hemitransection of the dorsal column at the level of C2_3 caused reduction of CGRP-containing fibers in the dorsal column and its nuclei on the operated side. Electron microscopic observation of the nucleus gracilis revealed that CGRP-Iike immunoreactive terminals made direct axodendritic synaptic contacts. Medium- to large-sized neurons in the dorsal root ganglia were labeled with Fast blue dye which was injected into the dorsal column nuclei. These included medium- to large-sized neurons exhibiting immunoreactivity to CGRP-like substances, and neurons of a medium size which were immunoreactive to SP-like compounds. The incidence of the former was higher at the thoracic level than at the cervical and lumbar levels, while that of the latter was very low. Electron microscopic observation of CGRP-containing fibers in the cervical region of the dorsal column revealed that 88% of these fibers were unmyelinated and the remainder were thinly myelinated. These results indicate that medium- to large-sized CGRP- (and SP)-containing neurons in the dorsal root ganglia give rise to fine fibers into the dorsal column and may terminate in the dorsal column nuclei. INTRODUCTION
for calcitonin m R N A 1'27, and is distributed widely throughout the nervous system 19'27. In the dorsal
It is generally accepted that unmyelinated and thinly myelinated primary afferent fibers originating from small type B cells in the dorsal root ganglia ( D R G ) terminate in the superficial dorsal horn of the spinal cord, whereas a n u m b e r of myelinated fibers derived from type A cells ascend the dorsal column to terminate in the dorsal column nuclei4'35.
root ganglia, medium- to large-sized neurons, as well as the small type B cells, show immunoreactivity to CGRP-Iike substances; some are also immunoreactive to substance P (SP)-like agents 18'21. In addition to dense CGRP-Iike immunoreactive (1R) and substance P (SP)-IR fibers in the superficial dorsal horn, the presence of these fibers in the dorsal column and its nuclei was reported in the rat and h u m a n 13a9'22.
The latter pathway has been largely neglected in chemical neuroanatomical studies of recent years, because of the relative paucity of chemically detectable substances associated with it. In contrast, the type B cell-dorsal horn pathway exhibit a variety of neuropeptides ls.2s. Calcitonin gene-related peptide ( C G R P ) is a neuropeptide incidentally discovered during analyses
Moreover, fine structure of these fibers was precisely described in the recent study 24, in which the authors suggested that these fibers originate from small type B cells in the D R G . However, it is more reasonable to consider that larger C G R P - I R primary afferent neurons project to the dorsal column nuclei via a dorsal column
Correspondence: E. Senba, Department of Anatomy (II), Osaka University Medical School, 4-3-57 Nakanoshima, Kita-ku, Osaka 530, Japan. 0006-8993/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)
123 pathway in view of the generally accepted anatomical knowledge 4'35. This investigation was undertaken to resolve this problem and to elucidate central projections and terminations of the CGRPIR sensory neurons in the dorsal column pathway by means of experimental immunohistochemistry and immunoelectron microscopy. MATERIALS AND METHODS
Animals and tissue preparation Twenty five male Wistar rats each weighing approximately 100 g were used. The animals were divided into 4 groups as follows. (1) One group (n = 6) was prepared for subsequent analysis of the distribution of CGRP-IR and SP-IR structures in the dorsal column and its nuclei in the medulla oblongata. In some animals (n = 2), colchicine (100/~g/100 g b.wt.) was injected into the cisterna magna 2 days prior to their sacrifice. (2) The second group (n = 8) was used to determine the operating direction of CGRP-IR and SP-IR fibers passing through the dorsal column. An incision was made in the dorsal column at the cervical level (C2_3), or at the mid-thoracic level, using sterilized surgical knives with animals under deep anesthesia induced by pentobarbital sodium (50 mg/kg i.p.). These rats were kept alive for 2-3 days. Subsequently, samples from most were subjected to light microscopic or, in some cases, electron microscopic analysis for the presence of CGRP-IR fibers accumulated at the site of the lesion. (3) The third group of animals (n = 3) underwent hemitransection of the dorsal column at the level of the cervical spinal cord (C2_3), and were kept alive for 14 days, in order to allow detection of changes in the CGRP-IR fibers of the dorsal column and its nuclei in the rostral part of the lesion. (4) In the fourth group (n = 8), 0.02/A of a 3% solution (in distilled water) of the fluorescent dye Fast blue (FB), was injected into the dorsal column nuclei of rats under pentobarbital sodium anesthesia (50 mg/kg i.p.). This was achieved using glass micropipettes with tip diameters of 20-50 pm, connected to a microinjector. Animals were immobilized in a stereotaxic instrument and a glass micropipette was inserted into the dorsal column
through a slit opened in the membrane covering the cisterna magna. Three days later, some of these rats were anesthetized again and given an injection of colchicine (100 ~g/100 g b.wt.) through a thin polyethylene tube inserted into the subarachnoid space through a slit in the membrane above the cisterna magna, and lowered to the cervical, thoracic or lumbar levels, respectively. These animals, with or without colchicine treatment, were kept alive for 5 days from the time of FB injection. A separate group of animals (n = 2) were prepared for the electron microscopic analysis of CGRP-IR fibers and terminals in the nucleus gracilis. Rats of all 4 groups were placed under deep pentobarbital sodium anesthesia, and perfused via the ascending aorta with 0.9% saline followed by 400 ml of either Zamboni's fixative3.7, or modified Zamboni's fixative containing 4% paraformaldehyde and 0.2% picric acid in 0.1 M phosphate buffer (pH 7.4). For the specimens to be examined by electron microscopy, 0.05% glutaraldehyde was added to the above solutions. Brains, and spinal cords complete with dorsal roots and dorsal root ganglia were excised and postfixed in the same fixative overnight and then transferred to 0.1 M phosphate buffer containing 30% sucrose, where they were left until they sank. Transverse sections (20/~m thick) of the medulla oblongata and spinal cord were cut in a cryostat and collected in a free-floating state. Sections of D R G were sliced at a thickness of 10-15 pm in a cryostat and mounted on gelatin-coated glass slides. For electron microscopic analysis, tissue blocks were treated with liquid nitrogen and sections (40 pm thick) were cut on a Vibratome.
Immunohistochemistry Light microscopic observation. Free-floating and slide-mounted sections of normal, surgically treated, and FB-injected animals were subjected to the indirect immunofluorescence method of Coons 7 for the detection of CGRP or SP-like immunoreactivities. In some samples, CGRP and SP were detected simultaneously using a double-labeling immunofluorescence method. Sections were first incubated with the respective primary antibodies (polyclonal anti-CGRP antiserum, Amersham; polyclonal anti-SP antiserum
124 raised by us) or in the mixture of the primary antibodies (polyclonal anti-CGRP antiserum, Amersham; monoclonal anti-SP antiserum, Seralab, MAS 035) diluted with 0.02 M phosphatebuffered saline (PBS) (1:1000 dilution for all the antisera) at 4 °C for 2 days. The samples were then washed in cold PBS for 30 min and incubated overnight at 4 °C in PBS containing fluorescein isothiocyanate (F1TC)-conjugated goat anti-rabbit lgG (Miles, 1 : 1000 dilution), or in a mixture of Texas red (TR)-conjugated donkey anti-rabbit IgG (Amersham, 1:250 dilution) and F1TC-conjugated goat anti-rat IgG (Cappel, 1:1000 dilution) for double-labeling. After a final washing with PBS, the sections were mounted, coverslipped with glycerolPBS (1:1), and viewed under a Nikon fluorescence microscope equipped with B2, G and V excitation filters, which permit exclusive visualization of FITC, TR and FB fluorescence, respectively. Control experiments were carried out for some samples by incubating them with primary antisera preabsorbed with l0 ~M of synthetic SP or CGRP (Peptide Inst., Osaka). The specificities of the antisera used in the present study have been described in previous reports 8'15'~1. Electron microscopic observation. Sections for electron microscopic study were processed by the peroxidase-antiperoxidase (PAP) method of Sternberger 33, as reported by Somogyi and Takagi 32. Briefly, they were preincubated in 0.1 M PBS containing 10% normal goat serum (NGS) before being incubated with anti-CGRP antiserum diluted with 0.1 M PBS containing 1% NGS and 1% bovine serum albumin (BSA) at 4 °C for 2 days. Subsequently, the samples were soaked in the dilution mixture for a few hours, incubated with goat anti-rabbit IgG antiserum (Cappel, 1:200 dilution) overnight at 4 °C, and incubated overnight again with rabbit PAP complex (DAKO, 1:500 dilution) overnight at the same temperature. After washing and further rinsing in 0.05 M Tris-HCl buffer (pH 7.6) incubation was carried out for 5-15 min in the same buffer containing 0.02%, 3,3"-diaminobenzidine (DAB) and 0.005% hydrogen peroxide. After washing in Tris-buffer, sections were fixed and stained in 1% osmium tetroxide in 0.1 M phosphate buffer and 1% uranyl acetate in 70% ethyl alcohol. Samples were dehydrated and horizontally embed-
ded in resin on silicon-coated glass slides, and serial ultrathin sections were examined under an electron microscope (Hitachi, H-7000). For the quantitative analysis of immunoreactive fibers, the lateral part of the fasciculus gracilis including the medial part of the fasciculus cuneatus were photographed randomly. Numbers of unmyelinated and myelinated C G R P - I R fibers were counted on the printed photos. RESULTS
Distribution of CGRP-IR fibers in the dorsal column and dorsal column nuclei In the upper cervical region of the dorsal column, CGRP-IR fibers were observed in the fasciculus gracilis and in the dorsomedial border of the fasciculus cuneatus (Fig. ld). The fibers in the latter area seemed to be continuous with small fiber bundles located at the dorsal border of the fasciculus cuneatus (Fig. lc). A prominent aggregation of immunoreactive fiber bundles was observed in the rostral part of the fasciculus cuneatus at the level of area postrema (Fig. lb). Examination of the dorsal column nuclei revealed the presence of C G R P - I R varicose fibers distributed in the dorsal column nuclei with those in the nucleus cuneatus being finer than those in the nucleus gracilis. Most of these fibers seemed to originate from C G R P - I R fiber bundles in the dorsal border of the fasciculus cuneatus (Figs. la,c). Within the nucleus gracilis, CGRP-IR fibers were more numerous in the peripheral regions with the central area of the nucleus usually being devoid of immunoreactive fibers (Fig. la-c). The distribution of SP-IR fibers in the dorsal column and dorsal column nuclei was similar to that of CGRP-IR fibers, although they were fewer compared to C G R P - I R ones. Most of these SP-IR fibers were shown to also contain CGRP. No CGRP-IR or SP-IR neurons were observed in the dorsal column nuclei even in the colchicine-treated animals. No CGRP-IR or SP-IR structures were found in control sections. Transection of the spinal cord Transection of the dorsal column at the C2_ 3 o r mid-thoracic level resulted in an accumulation of CGRP-IR substance in fibers at the caudal aspect of
125
Fig. 1. Fluorescent photomicrographs showing the distribution of CGRP-IR fibers in the dorsal column nuclei at the level of area postrema (a-c) and dorsal column of upper cervical cord. Note a prominent aggregationof CGRP-IR fibers in the fasciculus cuneatus (Cu) (thick arrow) and fine varicose fibers in the nucleus cuneatus (cu) (thin arrows). G, fasciculus gracilis; g, nucleus gracilis; nts, nucleus tractus solitarius. Scale bars; 100/lm (a-c), 50/~m (d).
the lesion. Some of these fibers also displayed immunoreactivity to SP. No substances immunoreactive to CGRP or SP were observed in the rostral part of the lesion. Hemitransection of the dorsal column brought about a decrease in the number of CGRP-IR fibers detected in the dorsal column and its nuclei on the operated side as compared to the control side. Small fiber bundles in the fasciculus gracilis and dorsal border of the fasciculus cuneatus were markedly less evident, while a few immunoreactive fibers persisted in the dorsal column nuclei.
Fluorescent dye tracer and immunohistochemistry Since findings from knife-cut lesion experiments indicated that CGRP-IR fibers in the dorsal column were ascending fibers, they were presumed to originate from the D R G cells or intraspinal neurons.
On this basis, FB was injected into the dorsal column bilaterally or unilaterally at the level of pyramidal decussation. The injection site involved and, in most cases, was restricted to the fasciculus gracilis and cuneatus, and also the most caudal part of the nucleus gracilis. FB-labeled cells were observed in the D R G of any of the spinal levels examined (cervical to lumbar), but none, or very few, were found in the D R G of the contralateral side in the case of unilateral injection. When injection sites involved the medullary dorsal horn or adjacent reticular formation, but excluded the dorsal column and its nuclei, no FB-labeled cells were detected in the DRG. These cases served as controls for vascular diffusion of this dye. FB-labeled cells in the DRGs were exclusively medium (20-40 /~m in diameter) (55%) to large (>40/~m in diameter) (45%) in size (Fig. 2c,d). About 40% of the D R G cells exhibited
126
Fig. 2. Fluorescent photomicrographs showing CGRP-IR (a) and SP-IR (b) neurons in the dorsal root ganglia (L4_5). Cells labeled with Fast blue (FB), which was injected into the dorsal column nuclei, are also shown for the same sections (c,d). Note that FB-labeled cells are exclusively medium- to large-sized cells, some of which show CGRP- or SP-like immunoreactivities (arrows). Scale bar; 100 ,urn.
C G R P - l i k e immunoreactivity, and 3 5 - 5 0 % of these were classified as m e d i u m - to large-sized using the same criteria as for F B - l a b e l e d D R G cells. The r e m a i n d e r were small-sized cells ( < 2 0 ~tm in diameter). A b o u t 2 - 7 % of the medium- to large-sized C G R P - I R cells were simultaneously labeled with FB (Fig. 2a,c), whereas no double-labeling of small C G R P - I R neurons occurred. The incidence of the d o u b l e - l a b e l e d cells in the D R G of the thoracic region was 2 - 3 times higher than those at the cervical or l u m b a r levels (Table I). SP-IR mediumsized neurons were occasionally labeled with FB (Fig. 2b,d). In the colchicine-treated rats, transverse sections of the spinal cord at various levels were processed for immunohistochemical detection of C G R P - l i k e
immunoreactivity. Most of the F B - l a b e l e d cells were located in lamina IV and some of them were scattered in lamina X. This distribution p a t t e r n varied little between the spinal levels, although m o r e labeled cells were observed in the cervical and l u m b a r enlargements. O n the other hand, in addition to the C G R P - I R m o t o n e u r o n s d e t e c t e d in the ventral horn throughout the length of the spinal cord, C G R P q R neurons were discovered in the i n t e r m e d i o m e d i a l (dorsal gray commissure) and intermediolateral cell columns in the t h o r a c o l u m b a r (C8-L2) levels, but only in the i n t e r m e d i o l a t e r a l cell column in the lumbosacral (L6-S1) levels. Identical findings were p r e s e n t e d in a previous study by the current authors 31. No d o u b l e - l a b e l e d cells were observed in the spinal cord.
127 TABLE I Number of CGRP-like immunoreactive (IR) cells and double-labeled cells in the dorsal root ganglia at various spinal levels
The last column shows the percentage of medium- to large-sized CGRP-IR cells that are double-labeled at each level. Spinal level
CGRP-IR cells (small)
CGRP-IR cells~ (medium-large)
FB-labeled CGRP-IR cells (small)
FB-labeled CGRP-1R cells b (medium-large)
b/a (%)
Cervical Upper thoracic (TI_6) Lower thoracic (T7_13) Lumbar
267 (63.1%) 564 (65.7%) 410 (51.2%) 271 (53.0%)
156 (36.9%) 294 (34.3%) 390 (48.8%) 240 (47.0%)
0 0 0 0
4 19 16 5
2.6 6.5 4.1 2.1
Electron m i c r o s c o p i c observation
C G R P - I R structures in the transverse sections of the dorsal column at the caudal aspect of the lesion (C2_3) were examined by electron microscopy. Immunoreactive fibers were distributed in the fasciculus gracilis and dorsomedial part of the fasciculus cuneatus. Twenty electron microscopic photographs were taken randomly. Of 420 C G R P - I R axons identified on these photos, 368 were unmyelinated (87.6%) and 52 were thinly myelinated (12.4%) (Fig. 3). C G R P - I R axon terminals in the nucleus gracilis were also examined. Observation was focused in the dorsomedial part of this nucleus at the level of area postrema. Most of these terminals were small in
diameter (0.6-1.2 ~tm; mean, 0.86 # m ) and formed direct synaptic contacts of asymmetric type with dendrites whose diameters were not larger than 1.2 /~m (Fig. 4a,b). Some large C G R P - I R axon terminals or swellings (2.0-2.6 ~tm) were also observed (Fig. 4c). DISCUSSION The dorsal column of the rat is far more heterogenous than previously supposed. It is comprised of long ascending and descending fibers which are listed as follows: (1) primary afferent fibers; (2) second-order postsynaptic dorsal column fibers 3°, (3) descending fibers from the dorsal column nucleiS;
Fig. 3. An electron microscopic photograph showing CGRP-IR fibers in the dorsal column. Note that most of these fibers are unmyelinated and some of them are thinly myelinated (arrows). Scale bar: 1/am.
128 and (4) corticospinal fibers. The former two types are ascending fibers and the latter two are descending. In addition to these long forms, about 25% of the dorsal column fibers are estimated to be propriospinal ascending and descending fibers 6.
Fig. 4. Electron microscopicphotographs showing CGRP-IR terminals in the nucleus gracilis. Most of these terminals are small and make direct synaptic contacts (arrowheads) with small dendrites (D) (a,b). CGRP-IR large boutons are occasionally observed (c). Scale bars: 0.5 ~um (a,b), 1 #m (c).
The present immunohistochemical study clearly shows that all CGRP-IR fibers in the dorsal column are of the ascending type, and some at least are primary afferent fibers originating from medium- to large-sized CGRP-IR neurons in the DRG. It was further demonstrated that a rather small number of medium- to large-sized CGRP-containing D R G cells (less than 10%) reach the dorsal column nuclei. This seems reasonable in view of the report by Glees and Soler ~° which suggests that relatively few fibers in the dorsal column (about 25%) reach the dorsal column nuclei, the remainder seeming to enter the spinal gray during their course of ascent. However, the possibility that a proportion of the ascending CGRP-IR fibers originate intraspinally cannot be discounted, because it has been demonstrated that at least 15% of the ascending axons in the dorsal column are non-primary afferent forms zg. These fibers have been shown to arise from neurons in laminae IIl and IV in various species 3'23'3°'34, although no comparative anatomical data concerning this is available for the rat. Unfortunately, CGRPIR dorsal column postsynaptic (DCPS) neurons could not be detected by the experimental procedures employed in the present study. Recently, McNeill et al. 24 reported that about 95% of CGRPIR fibers in the dorsal column disappear after dorsal rhizotomy. These findings suggest that DCPS neurons contribute very little to CGRP-IR dorsal column pathway. CGRP-IR fibers were found to be distributed throughout the entire length of the dorsal column nuclei. This distribution pattern corresponds closely with that of ascending fibers or primary afferent fibers in the dorsal column of the rat 2'9 and the cat 12 as determined by degeneration methods. The authors of these reports suggest that primary afferent fibers terminate with even distribution between the rostral and caudal aspects of the nuclei. The present study further showed that CGRP-IR fibers in the dorsal column and its nuclei decreased in number following transection of the dorsal column. These findings, taken together, indicate that most CGRPIR fibers are ascending axons and terminate in the dorsal column nuclei, although additional innervation from some other central region cannot be excluded at this juncture. It is generally accepted that primary afferent
129 fibers in the dorsal column convey information from several different types of receptors in the skin, muscles and joints 36. However, little is known about the functional role of the dorsal column pathway involving CGRP. It may be that these fibers are associated with the transmission of certain types of sensory modalities, such as proprioception, tactile sensation or visceral sensation. Differences in the routes by which proprioceptive information is transmitted from muscles and joints are reflected by variation in fiber composition in the fasciculus cuneatus and fasciculus gracilis, i.e. the fasciculus gracilis above the upper lumbar cord is devoid of proprioceptive fibers 4. Therefore, since C G R P - I R fibers were observed in both fasciculi at all spinal levels examined, it is less likely that these convey proprioceptive signals. In recent investigations, the current authors determined that large myelinated fibers innervating Meissner's tactile corpuscles and Merkel's discs are devoid of CGRP-Iike immunoreactivity 16'~7. Thus, it is unlikely that CGRP-IR primary afferent fibers in the dorsal column transmit impulses from low threshold mechanoreceptors in the skin. The higher incidence of CGRP-IR sensory cells projecting to the dorsal column nuclei in the thoracic region shown in the present study may indicate that some of these CGRP-IR fibers in the dorsal column convey visceral information to the central nervous system. In support of this speculation, it has been shown that some visceral afferent fibers pass through dorsal column and reach the dorsal column nuclei 26 and also that most of the visceral sensory neurons are CGRP-IR 11,25. REFERENCES
Kuo et al. have shown that most of the visceral afferent fibers (90%) in the cat are unmyelinated2°; the classical view has been that the dorsal column contains no unmyelinated fibers 35. However, this has also been challenged by Chung et al. 6 in an anatomical study which revealed that a considerable proportion (23%) of primary afferent fibers in the dorsal column are unmyelinated. In line with this, McNeil124 et al. reported the presence of CGRP-IR fine fibers in the dorsal column of the rat. They showed that 60-70% of these fibers are unmyelinated, which is lower than the incidence we have shown in the present study. This discrepancy may be explained by the difference of levels and regions of the dorsal column examined or technical differences of immunocytochemistry. Anyway, it is obvious from these studies that most of the CGRP-IR fibers in the dorsal column are unmyelinated or thinly myelinated. McNeill et al. 24 considered that these fine fibers are derived from small type B cells in the D R G which convey nociceptive information. However, we clearly demonstrated in the present study that these fibers originate from medium- to largesized D R G cells. This implies that a subpopulation of medium- to large-sized D R G neurons gives rise to unmyelinated fibers, and coincides well with the recent electrophysiological and morphological study 14 which showed unmyelinated afferent fibers do not originate exclusively from small D R G cells. ACKNOWLEDGEMENT This study was supported in part by a grant from the Ministry of Education of Japan. cord from medullary somatosensoryrelay nuclei, J. Comp. Neurol., 173 (1977) 773-792.
1 Amara, S.G., Jonas, V., Rosenfeld, M.G., Ong, E.S. and Evans, R.M., Alternative RNA processing in calcitonin gene expression generate mRNAs encoding different polypeptide products, Nature (Lond.), 298 (1982) 240-244. 2 Basbaum, A.I. and Hand, P.J., Projections of cervicothoracic dorsal roots to the cuneate nucleus of the rat, with observations on cellular 'Bricks', J. Cornp. Neurol., 148 (1973) 347-360. 3 Brown, A.G. and Fyffe, R.E.W., Form and function of dorsal horn neurones with axons ascending the dorsal columns in cat, J. Physiol. (Lond.), 321 (1981) 31-47. 4 Brown, A.G. and Gordon, G., Subcortical mechanisms concerned in somatic sensation, Br. Med. Bull., 33 (1977) 122-128. 5 Burton, H. and Loewy, A.D., Projections to the spinal
6 Chung, K., Langford, L.A. and Coggeshall, R.E., Primary afferent and propriospinal fibers in the rat dorsal and dorsolateral funiculi, J. Comp. Neurol., 263 (1987) 68-75. 7 Coons, A.H., Fluorescent antibody methods. In J.F. Danielli (Ed.), General Cytochemical Method, Academic, New York, 1958, pp. 399-422. 8 Cueilo, A.C., Gaifre, G. and Milstein, C., Detection of substance P in the central nervous system by a monoclonal antibody, Proc. Natl. Acad. Sci. U.S.A., 76 (1979) 35323536. 9 Ganchrow, D. and Bernstein, J.J., Projections of caudal fasciculus gracilis to nucleus gracilis and other medullary structures, and Clarke's nucleus in the rat, Brain Research, 205 (1981) 383-390. 10 Glees, P. and Soler, J., Fibre content of the posterior
130
11
12
13
14
15
16
17
18
19
20
21
22
column and synaptic connections of nucleus gracilis, Z. Zellforsch., 36 (1951) 381-400. Green, T. and Dockray, G.J., Calcitonin gene-related peptide and substance P in afferents to the upper gastrointestinal tract in the rat, Neurosci. Lett., 76 (1987) 151-156. Hand, P.J., Lumbosacral dorsal root terminations in the nucleus gracilis of the cat, J. Comp. Neurol., 126 (1966) 137-156. Harmann, P.A., Chung, K., Briner, R.P., Westlund, K.N. and Carlton, S.M., Calcitonin gene-related peptide (CGRP) in the human spinal cord: a light and electron microscopic analysis, J. Comp. Neurol., 269 (1988) 371380. Hoheisel, U and Mense, S., Non-myelinated afferent fibers do not originate exclusively from the smallest dorsal root ganglion cells in the cat, Neurosci. Lett.. 72 (1986) 153-157. Inagaki, S., Sakanaka, M., Shiosaka, S., Senba, E., Takatsuki, K., Takagi, H., Kawai, Y., Minagawa, H. and Tohyama, M., Ontogeny of substance P-containing neuron system of the rat: immunohistochemical analysis. I. Forebrain and upper brainstem, Neuroscience, 7 (1982) 251277. Ishida-Yamamoto, A., Senba, E. and Tohyama, M., Calcitonin gene-related peptide and substance P immunoreactive nerve fibers in Meissner's corpuscles of rats: an immunohistochemical analysis, Brain Research, 453 (1988) 362-366. Ishida-Yamamoto, A., Senba, E. and Tohyama, M., Distribution and fine structure of calcitonin gene-related peptide-like immunoreactive nerve fibers in the rat skin, Brain Research, in press. .lu. G., H6kfelt, T., Brodin, E., Fahrenkrug, J., Fisher, J.A., Frey, P., Elde, R.P. and Brown, J.C., Primary sensory neurons of the rat showing calcitonin gene-related peptide immunoreactivity and their relation to substance P-, somatostatin-, galanin-, vasoactive intestinal polypeptide- and cholecystokinin-immunoreactive ganglion cells, Cell Tissue Res., 247 (1987) 417-431. Kawai, Y., Takami, K., Shiosaka, S., Emson, P.C., Hillyard, C.J., Girgis, S., Maclntyre, I. and Tohyama, M., Topographic localization of calcitonin gene-related peptide in the rat brain: an immunohistochemical analysis, Neuroscience, 15 (1985) 747-763. Kuo, D.C., Yang, G.C.H., Yamasaki, D.S. and Krauthamer, G.H., A wide field electron microscopic analysis of the fiber constituents of the major splanchnic nerve in cat, J. Comp. Neurol., 210 (1982) 49-58. Lee, Y., Takami, K., Kawai, Y., Girgis, S., Hillyard, C.J., Maclntyre, I., Emson, P.C. and Tohyama, M., Distribution of calcitonin gene-related peptide in the rat peripheral nervous system with reference to its coexistence with substance P, Neuroscience, 15 (1985) 1227-1237. Ljungdahl, A., H6kfelt, T. and Nilsson, G., Distribution of substance P-like immunoreactivity in the central nervous system of the rat. I. Cell bodies and nerve terminals,
Neuroscience, 3 (1978) 861-943. 23 Lu, G.-W., Bennett, G.J., Nishikawa, N., Hoffert, M.J. and Dubner, R., Extra- and intracellular recordings from dorsal column postsynaptic spinomedullary neurons in the cat, Exp, NeuroL, 82 (1983) 456-477. 24 McNeill, D.L., Chung, K., Carlton, S.M. and Coggeshall, R.E., Calcitonin gene-related peptide immunostained axons provide evidence for fine primary afferent fibers in the dorsal and dorsolateral funiculi of the rat spinal cord, J. Comp, Neurol., 272 (1988) 303-308. 25 Molander, C., Ygge, J. and Dalsgaard, C.J., Substance P-, somatostatin- and calcitonin gene-related peptide-like immunoreactivity and fluoride resistant acid phosphataseactivity in relation to retrogradely labeled cutaneous, muscular and visceral primary sensory neurons in the rat, Neurosci. Lett., 74 (1987) 37-42. 26 Neuhuber, W., The central projections of visceral primary afferent neurons of the inferior mesenteric plexus and hypogastric nerve and the location of the related sensory and preganglionic sympathetic cell bodies in the rat, Anat. Embryol., 164 (1982) 413-425. 27 Rosenfeld, M.G., Mermod, J.J., Amara, S.G., Swanson, L.W., Sawchenko, P.E., Rivier, J., Vale, W.W. and Evans, R.M., Production of novel neuropeptide encoded by the calcitonin gene via tissue specific RNA processing, Nature (Lond.), 304 (1983) 129-135. 28 Ruda, M.A., Bennett, G.J. and Dubner, R., Neurochemistry and neural circuitry in the dorsal horn, Progr. Brain Res., 66 (1986) 219-268. 29 Rustioni, A., Non-primary afferents to the nucleus gracilis from the lumbar cord of the cat, Brain Research, 51 (1973) 81-95. 30 Rustioni, A., Spinal neurons project to the dorsal column nuclei of rhesus monkeys, Science, 196 (1977) 656-658. 31 Senba, E. and Tohyama, M., Calcitonin gene-related peptide containing autonomic afferent pathways to the pelvic ganglia of the rat, Brain Research, 449 (1988) 386-390. 32 Somogyi, P. and Takagi, H., A note on the use of picric acid-paraformaldehyde-glutaraldehyde fixative for correlated high and electron microscopic immunocytochemistry, Neuroscience, 7 (1982) 1779-1783. 33 Sternberger, L.A., lmmunocytochemistry, 2nd edn,, Wiley, New York, 1979. 34 Uddenberg, N., Functional organization of long, secondorder afferents in the dorsal funiculus, Exp. Brain Res., 4 (1968) 377-382. 35 Webster, K.E., Somaesthetic pathways, Br. Med. Bull., 33 (1977) 113-120. 36 Willis, W.D., and Coggeshall, R.E., Sensory pathways in the dorsal columns. In W.D. Willis and R.E. Coggeshall (Eds.), Sensory Mechanisms of the Spinal Cord, Plenum, New York, 1978, pp. 197-259. 37 Zamboni, L. and De Martino, C., Buffered picric acid formaldehyde: a new rapid fixative for electron microscopy, J. Cell. Biol., 35 (1967) 148A.