- Email: [email protected]
Immunoelectron microscopic examination of orexin-like immunoreactive fibers in the dorsal horn of the rat spinal cord
Brain Research 987 (2003) 86–92 www.elsevier.com / locate / brainres
Research report
Immunoelectron microscopic examination of orexin-like immunoreactive fibers in the dorsal horn of the rat spinal cord Jian-Lian Guan, Qing-Ping Wang, Seiji Shioda* Department of Anatomy, Showa University School of Medicine, 1 -5 -8 Hatanodai, Shinagawa-ku, Tokyo 142 -8555, Japan Accepted 3 July 2003
Abstract The ultrastructure and synaptic relationships of orexin A-like immunoreactive neuronal fibers in the dorsal horn of the rat cervical spinal cord were examined at both the light and electron microscopic levels. At the light microscopic level, many intensely immunostained orexin A-like fibers were found, while at the electron microscopic level, immunoreactivity in these fibers was mostly confined to axon terminals. Most of the axon terminals contained dense-cored vesicles. Immunoreactive and immunonegative dense-cored vesicles were occasionally found within the same orexin A-like immunoreactive axon terminals, which were often found making synapses with immunonegative dendrites. These synapses were both asymmetric and symmetric, with the asymmetric ones predominant. Orexin A-like immunoreactive processes that contained no synaptic vesicles were also found with less frequency. These processes were also observed receiving synaptic inputs from immunonegative axon terminals, but the synapses were mostly asymmetric. Sometimes, such processes were found to receive multiple synaptic inputs for which the presynaptic immunonegative axon terminals could make synapses on other immunonegative dendrites simultaneously. Occasionally, synapses between the orexin A-like immunoreactive axon terminals and orexin A-like immunoreactive processes containing no synaptic vesicles were also found. The present results provide solid morphological evidence that orexin A may be involved in pain-inhibition mechanisms in the spinal cord and suggest that this function may be complex and occur in conjunction with the regulatory effects of other neurotransmitters. 2003 Elsevier B.V. All rights reserved. Theme: Sensory systems Topic: Spinal cord Keywords: Immunocytochemistry; Synapse; Axon terminal; Asymmetric; Ultrastructure
1. Introduction
intrathecal injection of orexin-A revealed directly that OXA could play a role in pain inhibition in the spinal cord [23]. Because the orexin-1 receptor, a G-protein-coupled receptor selective for OXA, has also been reported to be present in the dorsal horn of the spinal cord [11], it is possible that OXA-producing neurons make synapses in the spinal cord. In the present study, an immunocytochemical technique was used to examine the ultrastructural and synaptic relationships of OXA-containing fibers in the dorsal horn of the spinal cord.
Orexin-A (OXA)-producing neurons are present in the lateral hypothalamic and perifornical areas of the brain and have been implicated in the control of arousal [14,22] and the regulation of food intake [10,16]. However, because of an abundance of OXA-like immunoreactive (OXA-LI) neuronal fibers in the dorsal horn of the spinal cord, it has been suggested that OXA could be involved in pain modulation [4,5,18]. A physiological study looking at nociception and hyperalgesia in mouse and rat models also indicated the pain-inhibiting effect of OXA [1]. Furthermore, an experiment designed to use a model of an
2. Materials and methods
*Corresponding author. Tel.: 181-3-3784-8103; fax: 181-3-37846815. E-mail address: [email protected] (S. Shioda).
Four adult male Wistar rats (250–300 g body weight) were used in this study. Animals were placed under deep anesthesia (sodium pentobarbital; 75 mg / kg b.w., i.p.) and
0006-8993 / 03 / $ – see front matter 2003 Elsevier B.V. All rights reserved. doi:10.1016 / S0006-8993(03)03257-8
J.-L. Guan et al. / Brain Research 987 (2003) 86–92
perfused through the left ventricle of the heart with 100 ml of 0.9% sodium chloride, followed by 500 ml of a fixative solution consisting of 4% paraformaldehyde and 0.1% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4). Blocks of cervical spinal cords tissue were dissected from each animal and rinsed with 0.1 M phosphate buffer (pH 7.4). After 2 h of postfixation in 4% paraformaldehyde, the blocks were cut into 30-mm thick frontal sections with a vibratome. The sections were rinsed in 0.1 M phosphate buffer (pH 7.4) overnight to wash out the fixative and were then immunostained. An avidin–biotin–peroxidase complex (ABC) immunoreaction [8] was used to detect the orexin antigen. An affinity-purified anti-orexin-A antiserum raised in goat was purchased from Incstar (Orexin-A (C-19), Santa Cruz Biotechnology, Inc, USA, catalogue [ sc-8070, lot [ E249). The specificity of the antiserum was determined by the supplier (Santa Cruz Biotechnology) and also described in a previous report by this group [7]. This antibody reacts with OXA of mouse, rat and human origin by Western blotting and immunohistochemistry; it is noncross-reactive with orexin-B. In the present study, orexin immunoreactivity was detected by incubation of the sections in the following steps: (1) normal horse serum diluted 1:20 at room temperature for 30 min; (2) primary antiserum diluted 1:20,000 (in 1% normal horse serum) for 16 h at 4 8C; (3) biotinylated anti-goat IgG for 1 h at room temperature; (4) ABC complex for 1 h at room temperature; (5) 3,39-diaminobenzidine (DAB; 0.2 mg / ml) and hydrogen peroxide (0.005%) in 0.1 M phosphate buffer (pH 7.4). The reaction time was monitored by observation of the slices with the aid of a light microscope during the reaction process. With the exception of the first step, the sections were rinsed with 0.1 M phosphate buffer three times after each step, for 5 min each time. Some sections were mounted on glass slides for light microscopic examination. Other reacted sections were postfixed in phosphatebuffered 1% osmium for 1 h at 4 8C. After dehydration in a
87
graded series of ethanol, these sections were embedded in a mixture of Epon–Araldite (TAAB Laboratories Equipment Limited, Berkshire, UK). Selected blocks of tissue were cut with a Reichert Ultracut Microtome (ReichertJung Optische, Vienna, Austria). All ultrathin sections were stained with uranyl acetate and lead citrate. Observations were made using an Hitachi H7000 electron microscope. All the OXA-LI profiles were photographed and counted separately according to their morphology. However, profiles in series sections were only counted as one. As a negative control, some of the spinal cord sections were incubated as described above, but either the first antibody or the second antibody was omitted from the procedure. These control sections were also examined using the electron microscope.
3. Results At the light microscopic level, many OXA-LI fibers were found distributed in the dorsal horn of the spinal cord, particularly in layer I (Fig. 1). The immunoreactivity was intense, with fibers sometimes observed to have a varicose-like appearance. The OXA-LI fibers were also seen distributed in white matter, especially in the dorsal region of the lateral area. At the electron microscopic level, most OXA-LI structures were axon terminals. Of the total 373 observed OXA-LI structures, 298 (80%) were clearly identified as axon terminals (Figs. 2 and 4). Most of the axon terminals contained dense-cored vesicles (Figs. 2A–C and 4–C; 253 / 298, 85%); some of them could be seen immunoreactive only at the dense-cored vesicle (Figs. 2C and 4A,C; 84 / 253, 33%). A few OXA-LI axon terminals (Fig. 4C, 2%) contained heterogeneous (both immunoreactive and immunonegative) dense-cored vesicles. OXA-LI axon terminals containing dense-cored vesicles were always observed to contain small, clear synaptic vesicles. Al-
Fig. 1. Light micrographs showing orexin A-like immunoreactive (OXA-LI) fibers in the dorsal horn of the rat spinal cord. (A) Full view of the distribution of OXA-LI immunoreactive fibers in the dorsal horn of the spinal cord. (B) An enlarged micrograph from the rectangle in (A) showing immunoreactive fibers in the superior layer of the dorsal horn in greater detail. Bars: (A) 0.2 mm; (B) 0.02 mm.
88
J.-L. Guan et al. / Brain Research 987 (2003) 86–92
Fig. 2. Electron microscopic micrographs showing OXA-LI axon terminals making synapses on immunonegative dendrites. (A) An OXA-LI axon terminal making a symmetric synapse (large arrow) on an immunonegative dendrite. Note the dense-cored vesicle (small arrow) near the presynaptic membrane. (B) An OXA-LI axon terminal making an asymmetric synapse (large arrow) on an immunonegative dendrite. Note that there are several dense-cored vesicles (small arrows) in the axon terminal and that the immunoreactivity of the dense-cored vesicles is more intense than that of the synaptic vesicles. (C) An OXA-LI axon terminal making an asymmetric synapse (large arrow) on an immunonegative dendrite. Note that most of the immunoreactivity is concentrated in the dense-cored vesicles (small arrows). (D) An OXA-LI axon terminal making an asymmetrical synapse (arrow) on a thin immunonegative dendrite. Bars: 0.2 mm.
though examples of dense-cored vesicles near synapse were observed (Fig. 2A), the vesicles were often found dispersed throughout the terminal. A large percentage (176 / 298, 59%) of the OXA-LI axon terminals was found to form synapses. Of these, 87% (153 / 176) formed synapses on immunonegative dendrites (Fig. 2). Most of
these axo-dendritic synapses were asymmetrical (Fig. 2B– D) because the postsynaptic membranes were obviously thicker than the presynaptic membrane, although symmetrical synapses were also found (Fig. 2A). The inner parts of the mitochondria in the OXA-LI axon terminals were immunonegative for OXA. No axo-somatic synapses
J.-L. Guan et al. / Brain Research 987 (2003) 86–92
made by OXA-LI axon terminals were found in the present study. OXA-LI processes with no synaptic vesicles were also observed (Figs. 3 and 4), albeit at a low frequency (75 / 373, 20%). Although the DAB reaction products were
89
diffused over all parts of these processes, the mitochondria were immunonegative for OXA. These processes often received synapses from immunonegative axon terminals (Fig. 3; 24 / 75, 32%). Sometimes, they could be seen receiving two synapses from two different axon terminals
Fig. 3. Electron micrographs showing OXA-LI processes receiving asymmetric synapses from immunonegative axon terminals. (A) An immunonegative axon terminal making a synapse (arrow) on an OXA-LI process. Note the OXA-LI process also makes contact with an OXA-LI axon terminal in the bottom of the panel. (B) An OXA-LI process receives a synapse (arrow) from an immunonegative axon terminal. (C) An OXA-LI process receives two synapses (arrows) from two immunonegative axon terminals. (D) An OXA-LI process receives a synapse (arrow) from an immunonegative axon terminal. Note that the presynaptic axon terminal also makes several synapses on other immunonegative dendrites. Bars: 0.2 mm.
90
J.-L. Guan et al. / Brain Research 987 (2003) 86–92
Fig. 4. Electron micrographs showing synapses between OXA-LI axon terminals and OXA-LI processes that contain no synaptic vesicles. (A, B) An OXA-LI axon terminal making a synapse (large arrow) on an OXA-LI processes that contains no synaptic vesicles. Note that the presynaptic axon terminals contain several dense-cored vesicles (small arrows). (C) An OXA-LI process receives two synapses, one from an OXA-LI axon terminal (large arrow) that contains several immunoreactive dense-cored vesicles (small arrows) and the other from an immunonegative axon terminal (arrowhead). (D) An OXA-LI process receives an asymmetric synapse (large arrow) from an OXA-LI axon terminal that contains no dense-cored vesicles. Bars: 0.2 mm.
(Fig. 3C), while, on occasions, the immunonegative axon terminals that made synapses on OXA-LI processes also made synapses on other immunonegative dendrites (Fig. 3D). Usually, the synapses made on these OXA-LI processes were asymmetric.
In rare cases, the OXA-LI processes that contained no synaptic vesicles were found to receive synapses from OXA-LI axon terminals (Fig. 4). Similar to as seen in immunonegative axon terminals (Fig. 3), these synapses were also asymmetrical (Fig. 4), even with a clear
J.-L. Guan et al. / Brain Research 987 (2003) 86–92
postsynaptic bar beneath the postsynaptic membranes (Fig. 4A).
4. Discussion
4.1. Morphological aspects It was reported several years ago [18] that the rat spinal cord is robustly innervated by OXA-LI fibers, a phenomenon which has since been described by other laboratories [4,5]. In the present study, we found, using the immunoelectron microscopic technique, that the OXA-LI fibers were mostly immunoreactive for OXA in their axon terminals. Furthermore, we also found some OXA-LI processes that did not contain synaptic vesicles. From a general point of view, these processes would be thought of as dendrites on the basis of their appearance. However, OXA-LI cell bodies are distributed only within the hypothalamus, and in the lateral area to be more specific [6,16,19], which is a considerable distance from the spinal cord. Although it is feasible from the abundance of OXALI axon terminals in the spinal cord that OXA-containing neurons in the hypothalamus could project long axons directly to the spinal cord, revealed by an experiment combining retrograde transport and immunocytochemistry techniques [18], it is highly unlikely that the OXA-LI neurons located in the hypothalamus would project such long dendrites direct to the spinal cord. Thus, one possible explanation is that these OXA-LI processes containing no synaptic vesicles could be the axon shafts that transfer the OXA peptides from the OXA-LI cell bodies in the hypothalamus to the OXA-LI axon terminals present along the length of the spinal cord [18]. Such a possibility has also been suggested concerning endomorphin 2 in the spinal cord [21]. In this way, the spinal cord has no endomorphin 2-like immunoreactive cell bodies, but it does have some immunoreactive processes containing no synaptic vesicles. It has been reported that OXA acts not only postsynaptically, but also presynaptically [19]. As we did not identify synapses between two OXA-LI axon terminals, the presynaptic action may occur between OXALI axon terminals and OXA-LI axon shafts. There is a discrepancy concerning the asymmetric or symmetric nature of the synapses made by OXA-LI axon terminals [7,12,13]. OXA-LI axon terminals synapse on neuropeptide Y-containing neurons [12] and catecholaminergic neurons [13] in the arcuate nucleus [12] and the locus coeruleus [13], respectively. Both of these reports described the fact that the synapses made by the axon terminals were absolutely asymmetric. However, in one of our previous studies, we found that symmetric synapses were predominant between OXA-LI axon terminals and POMC-producing neurons in the arcuate nucleus [7]. On the other hand, in the present study the synapses made by OXA-LI axon terminals in the dorsal horn of the spinal
91
cord were identified to be mostly asymmetrical. Thus, we suggest that the synapses made by OXA-LI axon terminals may be different types depending on the region in which they are located and the target neurons with which they make synapses. Although no dense-cored vesicles were reported to be present in the OXA-LI axon terminals described in previous studies [12,13], we found here that some of these axon terminals contained a few dense-cored vesicles. This finding is in accordance with other reports [6,7,19,25] and suggests that OXA is transported within the dense-cored vesicles.
4.2. Functional aspects The superficial layers of the dorsal horn of the spinal cord are the area associated with nociceptive transmission. Similar to other hypothalamic peptides such as corticotropin-releasing factor [17], vasopressin [9], neuropeptide FF [15], pituitary adenylate cyclase activating peptide [26], and the recently found endomorphins [24], it has been suggested that OXA is also involved in the modulation of responses to noxious stimuli on account of the existence of the OXA and orexin-1 receptor in this area [4,5,18]. Research into the ultrastructural and synaptic relationships of the OXA-LI fibers in the present study confirmed the existence of OXA in the dorsal horn of the spinal cord and furthermore supports a possible pain-inhibition role of the OXA-LI neurons, which may take place through direct synaptic relationships with other neurotransmitter / neuromodulator-containing neurons in the spinal cord. A recent report [23] indicated that OXA could induce pain inhibition in the spinal cord. These authors found that intrathecal injection of OXA decreased the sum of flinches in the formalin test and increased the hot plate latency; such effects of orexin-A were completely antagonized by pretreatment with SB-334867, a selective orexin-1 receptor antagonist. Furthermore, intrathecal injection of orexin-A suppressed the expression of Fos-like immunoreactivity, induced by paw formalin injection, in laminae I–II of L4-5 of the spinal cord [23]. These data show that OXA could produce analgesic effects in the spinal cord via the orexin1 receptor. That immunopositive and immunonegative dense-cored vesicles coexist within the same OXA-LI axon terminal in some cases suggests that OXA could coexist in axon terminals with other neurotransmitters also transported from the hypothalamus. This coexistence is unlikely to involve melanin-concentrating hormone because although the distributions of both peptides are similar in the lateral hypothalamus, they are from clearly different populations of neurons [2]. One possible neurotransmitter may be dynorphin, and double-label in situ hybridization revealed that nearly all (94%) neurons expressing prepro-orexin mRNA also expressed prodynorphin mRNA in the lateral hypothalamic area [3]. Although the coexistence has been
92
J.-L. Guan et al. / Brain Research 987 (2003) 86–92
expressed in terms of dynorphin playing an important role in the food-intake function of orexin neurons in the lateral hypothalamic area, it could also suggest that dynorphin in the OXA-LI axon terminals in the dorsal horn of the spinal cord plays an important role in pain inhibition because dynorphin in the spinal cord has a strong effect in pain inhibition through the kappa opioid receptor [20]. Coexistence with endorphins is probably impossible because it has been reported that the pain inhibition caused by orexin can be blocked by the orexin-1 receptor antagonist SB-334867, but not naloxone [1]. The fact that some OXA-LI processes receive synapses from immunonegative axon terminals which also make synapses on other immunoreactive dendrites further implies that OXA is involved in a more complex functional mechanism related to pain inhibition.
References [1] S. Bingham, P.T. Davey, A.J. Babbs, E.A. Irving, M.J. Sammons, M. Wyles, P. Jeffrey, L. Cutler, I. Riba, A. Johns, R.A. Porter, N. Upton, A.J. Hunter, A.A. Parsons, Orexin-A, an hypothalamic peptide with analgesic properties, Pain 92 (2001) 81–90. ¨ [2] C. Broberger, L. De Lecea, J.G. Sutcliffe, T. Hokfelt, Hypocretin / orexin- and melanin-concentrating hormone-expressing cells form distinct populations in the rodent lateral hypothalamus: relationship to the neuropeptide Y and agouti gene-related protein systems, J. Comp. Neurol. 402 (1998) 460–474. [3] T.C. Chou, C.E. Lee, J. Lu, J.K. Elmquist, J. Hara, J.T. Willie, C.T. Beuckmann, R.M. Chemelli, T. Sakurai, M. Yanagisawa, C.B. Saper, T.E. Scammell, Orexin (hypocretin) neurons contain dynorphin, J. Neurosci. 21 (2001) RC168. [4] D.J. Cutler, R. Morris, V. Sheridhar, T.A. Wattam, S. Holmes, S. Patel, J.R. Arch, S. Wilson, R.E. Buckingham, M.L. Evans, R.A. Leslie, G. Williams, Differential distribution of orexin-A and orexinB immunoreactivity in the rat brain and spinal cord, Peptides 20 (1999) 1455–1470. [5] Y. Date, M.S. Mondal, S. Matsukura, M. Nakazato, Distribution of orexin-A and orexin-B (hypocretins) in the rat spinal cord, Neurosci. Lett. 288 (2000) 87–90. [6] L. de Lecea, T. Kilduff, C. Peyron, X.-B. Gao, P.E. Foye, P.E. Danuelson, C. Fukuhara, E.L.F. Battenberg, V.T. Gautvik, F.S. Bartlett II, W.N. Frankel, A.N. van den Pol, F.G. Bloom, K.M. Gautvik, J.G. Sutcliffe, The hypocretins: hypothalamus-specific peptides with neuroexcitatory activity, Proc. Natl. Acad. Sci. USA 95 (1998) 322–327. [7] J.-L. Guan, T. Saotome, Q.-P. Wang, H. Funahashi, T. Hori, S. Tanaka, S. Shioda, Orexinergic innervation of POMC-containing neurons in the rat arcuate nucleus, NeuroReport 12 (2001) 547–551. [8] J.-L. Guan, Q.-P. Wang, Y. Nakai, Electron microscopic observation of mu-opioid receptor in the rat area postrema, Peptides 20 (1999) 873–880. [9] M. Hallbeck, A. Blomqvist, Spinal cord-projection vasopressinergic neurons in the rat paraventricular hypothalamus, J. Comp. Neurol. 411 (1999) 201–211.
[10] A.C. Haynes, B. Jackson, H. Chapman, M. Tadayyon, A. Johns, R.A. Porter, J.R. Arch, A selective orexin-1 receptor antagonist reduces food consumption in male and female rats, Regul. Pept. 96 (2000) 45–51. [11] G.J. Hervieu, J.E. Cluderay, D.C. Harrison, J.C. Roberts, R.A. Leslie, Gene expression and protein distribution of the orexin-1 receptor in the rat brain and spinal cord, Neuroscience 103 (2001) 777–797. [12] T.L. Horvath, S. Diano, A.N. van den Pol, Synaptic interaction between hypocretin (orexin) and neuropeptide Y cells in the rodent and primate hypothalamus: a novel circuit implicated in metabolic and endocrine regulations, J. Neurosci. 19 (1999) 1072–1087. [13] T.L. Horvath, C. Peyron, S. Diano, A. Ivanov, G. Aston Jones, T.S. Kilduff, A.N. van den Pol, Hypocretin (orexin) activation and synaptic innervation of the locus coeruleus noradrenergic system, J. Comp. Neurol. 415 (1999) 145–159. [14] M.M. Methippara, M.N. Alam, R. Szymusiak, D. McGinty, Effects of lateral preoptic area application of orexin-A on sleep–wakefulness, NeuroReport 11 (2000) 3423–3426. [15] P. Penula, A.A. Aarnisalo, K. Wasowics, Neuropeptide FF, a mammalian neuropeptide with multiple function, Prog. Neurobiol. 48 (1996) 461–487. [16] T. Sakurai, A. Amemiya, M. Ishii, I. Matsuzaki, R.M. Chemelli, H. Tanaka, S.C. Williams, J.A. Richardson, G.P. Kozlowski, S. Wilson, J.R.S. Arch, R.E. Buckingham, A.C. Haynes, S.A. Carr, R.S. Annan, D.E. McNulty, W.-S. Liu, J.A. Terrett, N.A. Elshourbagy, D.J. Bergsma, M. Yanagisawa, Orexins and orexin receptors: a family of hypothalamic neuropeptides and G protein-coupled receptors that regulate feeding behavior, Cell 92 (1998) 573–585. [17] M. Schafer, S.A. Mousa, C. Stein, Corticotropin-releasing factor in antinociception and inflammation, Eur. J. Pharmacol. 323 (1997) 1–10. [18] A.N. van den Pol, Hypothalamic hypocretin (orexin): robust innervation of the spinal cord, J. Neurosci. 19 (1999) 3171–3182. [19] A.N. van den Pol, X.B. Gao, K. Obrietan, T.S. Kilduff, A.B. Belousov, Presynaptic and postsynaptic actions and modulation of neuroendocrine neurons by a new hypothalamic peptide, hypocretin / orexin, J. Neurosci. 18 (1998) 7962–7971. [20] T.W. Vanderah, M.H. Ossipov, J. Lai, T.P.J. Malan, F. Porreca, Mechanisms of opioid-induced pain and antinociceptive tolerance: descending facilitation and spinal dynorphin, Pain 92 (2001) 5–9. [21] Q.-P. Wang, J.E. Zadina, J.-L. Guan, A.J. Kastin, Endomorphin-2 immunoreactivity in the cervical dorsal horn of the rat spinal cord at the electron microscopic level, Neuroscience 113 (2002) 593–605. [22] M.-C. Xi, F.R. Morales, M.H. Chase, Effects on sleep and wakefulness of the injection of hypocretin-1 (orexin-A) into the laterodorsal tegmental nucleus of the cat, Brain Res. 901 (2001) 259–264. [23] T. Yamamoto, N. Nozaki-Taguchi, T. Chiba, Analgesic effect of intrathecally administered orexin-A in the rat formalin test and in the rat hot plate test, Br. J. Pharmacol. 137 (2002) 170–176. [24] J.E. Zadina, L. Hackler, L.-J. Ge, A.J. Kastin, A potent and selective endogenous agonist for the m-opiate receptor, Nature 386 (1997) 499–502. [25] J.H. Zhang, S. Sampogna, F.R. Morales, M.H. Chase, Orexin (hypocretin)-like immunoreactivity in the cat hypothalamus: a light and electron microscopic study, Sleep 24 (2001) 67–76. [26] Y.Z. Zhang, A.B. Malmberg, B. Sjolund, T.L. Yaksh, The effect of pituitary adenylate cyclase activating peptide (PACAP) on the nociceptive formalin test, Neurosci. Lett. 207 (1996) 187–190.
Research report
Immunoelectron microscopic examination of orexin-like immunoreactive fibers in the dorsal horn of the rat spinal cord Jian-Lian Guan, Qing-Ping Wang, Seiji Shioda* Department of Anatomy, Showa University School of Medicine, 1 -5 -8 Hatanodai, Shinagawa-ku, Tokyo 142 -8555, Japan Accepted 3 July 2003
Abstract The ultrastructure and synaptic relationships of orexin A-like immunoreactive neuronal fibers in the dorsal horn of the rat cervical spinal cord were examined at both the light and electron microscopic levels. At the light microscopic level, many intensely immunostained orexin A-like fibers were found, while at the electron microscopic level, immunoreactivity in these fibers was mostly confined to axon terminals. Most of the axon terminals contained dense-cored vesicles. Immunoreactive and immunonegative dense-cored vesicles were occasionally found within the same orexin A-like immunoreactive axon terminals, which were often found making synapses with immunonegative dendrites. These synapses were both asymmetric and symmetric, with the asymmetric ones predominant. Orexin A-like immunoreactive processes that contained no synaptic vesicles were also found with less frequency. These processes were also observed receiving synaptic inputs from immunonegative axon terminals, but the synapses were mostly asymmetric. Sometimes, such processes were found to receive multiple synaptic inputs for which the presynaptic immunonegative axon terminals could make synapses on other immunonegative dendrites simultaneously. Occasionally, synapses between the orexin A-like immunoreactive axon terminals and orexin A-like immunoreactive processes containing no synaptic vesicles were also found. The present results provide solid morphological evidence that orexin A may be involved in pain-inhibition mechanisms in the spinal cord and suggest that this function may be complex and occur in conjunction with the regulatory effects of other neurotransmitters. 2003 Elsevier B.V. All rights reserved. Theme: Sensory systems Topic: Spinal cord Keywords: Immunocytochemistry; Synapse; Axon terminal; Asymmetric; Ultrastructure
1. Introduction
intrathecal injection of orexin-A revealed directly that OXA could play a role in pain inhibition in the spinal cord [23]. Because the orexin-1 receptor, a G-protein-coupled receptor selective for OXA, has also been reported to be present in the dorsal horn of the spinal cord [11], it is possible that OXA-producing neurons make synapses in the spinal cord. In the present study, an immunocytochemical technique was used to examine the ultrastructural and synaptic relationships of OXA-containing fibers in the dorsal horn of the spinal cord.
Orexin-A (OXA)-producing neurons are present in the lateral hypothalamic and perifornical areas of the brain and have been implicated in the control of arousal [14,22] and the regulation of food intake [10,16]. However, because of an abundance of OXA-like immunoreactive (OXA-LI) neuronal fibers in the dorsal horn of the spinal cord, it has been suggested that OXA could be involved in pain modulation [4,5,18]. A physiological study looking at nociception and hyperalgesia in mouse and rat models also indicated the pain-inhibiting effect of OXA [1]. Furthermore, an experiment designed to use a model of an
2. Materials and methods
*Corresponding author. Tel.: 181-3-3784-8103; fax: 181-3-37846815. E-mail address: [email protected] (S. Shioda).
Four adult male Wistar rats (250–300 g body weight) were used in this study. Animals were placed under deep anesthesia (sodium pentobarbital; 75 mg / kg b.w., i.p.) and
0006-8993 / 03 / $ – see front matter 2003 Elsevier B.V. All rights reserved. doi:10.1016 / S0006-8993(03)03257-8
J.-L. Guan et al. / Brain Research 987 (2003) 86–92
perfused through the left ventricle of the heart with 100 ml of 0.9% sodium chloride, followed by 500 ml of a fixative solution consisting of 4% paraformaldehyde and 0.1% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4). Blocks of cervical spinal cords tissue were dissected from each animal and rinsed with 0.1 M phosphate buffer (pH 7.4). After 2 h of postfixation in 4% paraformaldehyde, the blocks were cut into 30-mm thick frontal sections with a vibratome. The sections were rinsed in 0.1 M phosphate buffer (pH 7.4) overnight to wash out the fixative and were then immunostained. An avidin–biotin–peroxidase complex (ABC) immunoreaction [8] was used to detect the orexin antigen. An affinity-purified anti-orexin-A antiserum raised in goat was purchased from Incstar (Orexin-A (C-19), Santa Cruz Biotechnology, Inc, USA, catalogue [ sc-8070, lot [ E249). The specificity of the antiserum was determined by the supplier (Santa Cruz Biotechnology) and also described in a previous report by this group [7]. This antibody reacts with OXA of mouse, rat and human origin by Western blotting and immunohistochemistry; it is noncross-reactive with orexin-B. In the present study, orexin immunoreactivity was detected by incubation of the sections in the following steps: (1) normal horse serum diluted 1:20 at room temperature for 30 min; (2) primary antiserum diluted 1:20,000 (in 1% normal horse serum) for 16 h at 4 8C; (3) biotinylated anti-goat IgG for 1 h at room temperature; (4) ABC complex for 1 h at room temperature; (5) 3,39-diaminobenzidine (DAB; 0.2 mg / ml) and hydrogen peroxide (0.005%) in 0.1 M phosphate buffer (pH 7.4). The reaction time was monitored by observation of the slices with the aid of a light microscope during the reaction process. With the exception of the first step, the sections were rinsed with 0.1 M phosphate buffer three times after each step, for 5 min each time. Some sections were mounted on glass slides for light microscopic examination. Other reacted sections were postfixed in phosphatebuffered 1% osmium for 1 h at 4 8C. After dehydration in a
87
graded series of ethanol, these sections were embedded in a mixture of Epon–Araldite (TAAB Laboratories Equipment Limited, Berkshire, UK). Selected blocks of tissue were cut with a Reichert Ultracut Microtome (ReichertJung Optische, Vienna, Austria). All ultrathin sections were stained with uranyl acetate and lead citrate. Observations were made using an Hitachi H7000 electron microscope. All the OXA-LI profiles were photographed and counted separately according to their morphology. However, profiles in series sections were only counted as one. As a negative control, some of the spinal cord sections were incubated as described above, but either the first antibody or the second antibody was omitted from the procedure. These control sections were also examined using the electron microscope.
3. Results At the light microscopic level, many OXA-LI fibers were found distributed in the dorsal horn of the spinal cord, particularly in layer I (Fig. 1). The immunoreactivity was intense, with fibers sometimes observed to have a varicose-like appearance. The OXA-LI fibers were also seen distributed in white matter, especially in the dorsal region of the lateral area. At the electron microscopic level, most OXA-LI structures were axon terminals. Of the total 373 observed OXA-LI structures, 298 (80%) were clearly identified as axon terminals (Figs. 2 and 4). Most of the axon terminals contained dense-cored vesicles (Figs. 2A–C and 4–C; 253 / 298, 85%); some of them could be seen immunoreactive only at the dense-cored vesicle (Figs. 2C and 4A,C; 84 / 253, 33%). A few OXA-LI axon terminals (Fig. 4C, 2%) contained heterogeneous (both immunoreactive and immunonegative) dense-cored vesicles. OXA-LI axon terminals containing dense-cored vesicles were always observed to contain small, clear synaptic vesicles. Al-
Fig. 1. Light micrographs showing orexin A-like immunoreactive (OXA-LI) fibers in the dorsal horn of the rat spinal cord. (A) Full view of the distribution of OXA-LI immunoreactive fibers in the dorsal horn of the spinal cord. (B) An enlarged micrograph from the rectangle in (A) showing immunoreactive fibers in the superior layer of the dorsal horn in greater detail. Bars: (A) 0.2 mm; (B) 0.02 mm.
88
J.-L. Guan et al. / Brain Research 987 (2003) 86–92
Fig. 2. Electron microscopic micrographs showing OXA-LI axon terminals making synapses on immunonegative dendrites. (A) An OXA-LI axon terminal making a symmetric synapse (large arrow) on an immunonegative dendrite. Note the dense-cored vesicle (small arrow) near the presynaptic membrane. (B) An OXA-LI axon terminal making an asymmetric synapse (large arrow) on an immunonegative dendrite. Note that there are several dense-cored vesicles (small arrows) in the axon terminal and that the immunoreactivity of the dense-cored vesicles is more intense than that of the synaptic vesicles. (C) An OXA-LI axon terminal making an asymmetric synapse (large arrow) on an immunonegative dendrite. Note that most of the immunoreactivity is concentrated in the dense-cored vesicles (small arrows). (D) An OXA-LI axon terminal making an asymmetrical synapse (arrow) on a thin immunonegative dendrite. Bars: 0.2 mm.
though examples of dense-cored vesicles near synapse were observed (Fig. 2A), the vesicles were often found dispersed throughout the terminal. A large percentage (176 / 298, 59%) of the OXA-LI axon terminals was found to form synapses. Of these, 87% (153 / 176) formed synapses on immunonegative dendrites (Fig. 2). Most of
these axo-dendritic synapses were asymmetrical (Fig. 2B– D) because the postsynaptic membranes were obviously thicker than the presynaptic membrane, although symmetrical synapses were also found (Fig. 2A). The inner parts of the mitochondria in the OXA-LI axon terminals were immunonegative for OXA. No axo-somatic synapses
J.-L. Guan et al. / Brain Research 987 (2003) 86–92
made by OXA-LI axon terminals were found in the present study. OXA-LI processes with no synaptic vesicles were also observed (Figs. 3 and 4), albeit at a low frequency (75 / 373, 20%). Although the DAB reaction products were
89
diffused over all parts of these processes, the mitochondria were immunonegative for OXA. These processes often received synapses from immunonegative axon terminals (Fig. 3; 24 / 75, 32%). Sometimes, they could be seen receiving two synapses from two different axon terminals
Fig. 3. Electron micrographs showing OXA-LI processes receiving asymmetric synapses from immunonegative axon terminals. (A) An immunonegative axon terminal making a synapse (arrow) on an OXA-LI process. Note the OXA-LI process also makes contact with an OXA-LI axon terminal in the bottom of the panel. (B) An OXA-LI process receives a synapse (arrow) from an immunonegative axon terminal. (C) An OXA-LI process receives two synapses (arrows) from two immunonegative axon terminals. (D) An OXA-LI process receives a synapse (arrow) from an immunonegative axon terminal. Note that the presynaptic axon terminal also makes several synapses on other immunonegative dendrites. Bars: 0.2 mm.
90
J.-L. Guan et al. / Brain Research 987 (2003) 86–92
Fig. 4. Electron micrographs showing synapses between OXA-LI axon terminals and OXA-LI processes that contain no synaptic vesicles. (A, B) An OXA-LI axon terminal making a synapse (large arrow) on an OXA-LI processes that contains no synaptic vesicles. Note that the presynaptic axon terminals contain several dense-cored vesicles (small arrows). (C) An OXA-LI process receives two synapses, one from an OXA-LI axon terminal (large arrow) that contains several immunoreactive dense-cored vesicles (small arrows) and the other from an immunonegative axon terminal (arrowhead). (D) An OXA-LI process receives an asymmetric synapse (large arrow) from an OXA-LI axon terminal that contains no dense-cored vesicles. Bars: 0.2 mm.
(Fig. 3C), while, on occasions, the immunonegative axon terminals that made synapses on OXA-LI processes also made synapses on other immunonegative dendrites (Fig. 3D). Usually, the synapses made on these OXA-LI processes were asymmetric.
In rare cases, the OXA-LI processes that contained no synaptic vesicles were found to receive synapses from OXA-LI axon terminals (Fig. 4). Similar to as seen in immunonegative axon terminals (Fig. 3), these synapses were also asymmetrical (Fig. 4), even with a clear
J.-L. Guan et al. / Brain Research 987 (2003) 86–92
postsynaptic bar beneath the postsynaptic membranes (Fig. 4A).
4. Discussion
4.1. Morphological aspects It was reported several years ago [18] that the rat spinal cord is robustly innervated by OXA-LI fibers, a phenomenon which has since been described by other laboratories [4,5]. In the present study, we found, using the immunoelectron microscopic technique, that the OXA-LI fibers were mostly immunoreactive for OXA in their axon terminals. Furthermore, we also found some OXA-LI processes that did not contain synaptic vesicles. From a general point of view, these processes would be thought of as dendrites on the basis of their appearance. However, OXA-LI cell bodies are distributed only within the hypothalamus, and in the lateral area to be more specific [6,16,19], which is a considerable distance from the spinal cord. Although it is feasible from the abundance of OXALI axon terminals in the spinal cord that OXA-containing neurons in the hypothalamus could project long axons directly to the spinal cord, revealed by an experiment combining retrograde transport and immunocytochemistry techniques [18], it is highly unlikely that the OXA-LI neurons located in the hypothalamus would project such long dendrites direct to the spinal cord. Thus, one possible explanation is that these OXA-LI processes containing no synaptic vesicles could be the axon shafts that transfer the OXA peptides from the OXA-LI cell bodies in the hypothalamus to the OXA-LI axon terminals present along the length of the spinal cord [18]. Such a possibility has also been suggested concerning endomorphin 2 in the spinal cord [21]. In this way, the spinal cord has no endomorphin 2-like immunoreactive cell bodies, but it does have some immunoreactive processes containing no synaptic vesicles. It has been reported that OXA acts not only postsynaptically, but also presynaptically [19]. As we did not identify synapses between two OXA-LI axon terminals, the presynaptic action may occur between OXALI axon terminals and OXA-LI axon shafts. There is a discrepancy concerning the asymmetric or symmetric nature of the synapses made by OXA-LI axon terminals [7,12,13]. OXA-LI axon terminals synapse on neuropeptide Y-containing neurons [12] and catecholaminergic neurons [13] in the arcuate nucleus [12] and the locus coeruleus [13], respectively. Both of these reports described the fact that the synapses made by the axon terminals were absolutely asymmetric. However, in one of our previous studies, we found that symmetric synapses were predominant between OXA-LI axon terminals and POMC-producing neurons in the arcuate nucleus [7]. On the other hand, in the present study the synapses made by OXA-LI axon terminals in the dorsal horn of the spinal
91
cord were identified to be mostly asymmetrical. Thus, we suggest that the synapses made by OXA-LI axon terminals may be different types depending on the region in which they are located and the target neurons with which they make synapses. Although no dense-cored vesicles were reported to be present in the OXA-LI axon terminals described in previous studies [12,13], we found here that some of these axon terminals contained a few dense-cored vesicles. This finding is in accordance with other reports [6,7,19,25] and suggests that OXA is transported within the dense-cored vesicles.
4.2. Functional aspects The superficial layers of the dorsal horn of the spinal cord are the area associated with nociceptive transmission. Similar to other hypothalamic peptides such as corticotropin-releasing factor [17], vasopressin [9], neuropeptide FF [15], pituitary adenylate cyclase activating peptide [26], and the recently found endomorphins [24], it has been suggested that OXA is also involved in the modulation of responses to noxious stimuli on account of the existence of the OXA and orexin-1 receptor in this area [4,5,18]. Research into the ultrastructural and synaptic relationships of the OXA-LI fibers in the present study confirmed the existence of OXA in the dorsal horn of the spinal cord and furthermore supports a possible pain-inhibition role of the OXA-LI neurons, which may take place through direct synaptic relationships with other neurotransmitter / neuromodulator-containing neurons in the spinal cord. A recent report [23] indicated that OXA could induce pain inhibition in the spinal cord. These authors found that intrathecal injection of OXA decreased the sum of flinches in the formalin test and increased the hot plate latency; such effects of orexin-A were completely antagonized by pretreatment with SB-334867, a selective orexin-1 receptor antagonist. Furthermore, intrathecal injection of orexin-A suppressed the expression of Fos-like immunoreactivity, induced by paw formalin injection, in laminae I–II of L4-5 of the spinal cord [23]. These data show that OXA could produce analgesic effects in the spinal cord via the orexin1 receptor. That immunopositive and immunonegative dense-cored vesicles coexist within the same OXA-LI axon terminal in some cases suggests that OXA could coexist in axon terminals with other neurotransmitters also transported from the hypothalamus. This coexistence is unlikely to involve melanin-concentrating hormone because although the distributions of both peptides are similar in the lateral hypothalamus, they are from clearly different populations of neurons [2]. One possible neurotransmitter may be dynorphin, and double-label in situ hybridization revealed that nearly all (94%) neurons expressing prepro-orexin mRNA also expressed prodynorphin mRNA in the lateral hypothalamic area [3]. Although the coexistence has been
92
J.-L. Guan et al. / Brain Research 987 (2003) 86–92
expressed in terms of dynorphin playing an important role in the food-intake function of orexin neurons in the lateral hypothalamic area, it could also suggest that dynorphin in the OXA-LI axon terminals in the dorsal horn of the spinal cord plays an important role in pain inhibition because dynorphin in the spinal cord has a strong effect in pain inhibition through the kappa opioid receptor [20]. Coexistence with endorphins is probably impossible because it has been reported that the pain inhibition caused by orexin can be blocked by the orexin-1 receptor antagonist SB-334867, but not naloxone [1]. The fact that some OXA-LI processes receive synapses from immunonegative axon terminals which also make synapses on other immunoreactive dendrites further implies that OXA is involved in a more complex functional mechanism related to pain inhibition.
References [1] S. Bingham, P.T. Davey, A.J. Babbs, E.A. Irving, M.J. Sammons, M. Wyles, P. Jeffrey, L. Cutler, I. Riba, A. Johns, R.A. Porter, N. Upton, A.J. Hunter, A.A. Parsons, Orexin-A, an hypothalamic peptide with analgesic properties, Pain 92 (2001) 81–90. ¨ [2] C. Broberger, L. De Lecea, J.G. Sutcliffe, T. Hokfelt, Hypocretin / orexin- and melanin-concentrating hormone-expressing cells form distinct populations in the rodent lateral hypothalamus: relationship to the neuropeptide Y and agouti gene-related protein systems, J. Comp. Neurol. 402 (1998) 460–474. [3] T.C. Chou, C.E. Lee, J. Lu, J.K. Elmquist, J. Hara, J.T. Willie, C.T. Beuckmann, R.M. Chemelli, T. Sakurai, M. Yanagisawa, C.B. Saper, T.E. Scammell, Orexin (hypocretin) neurons contain dynorphin, J. Neurosci. 21 (2001) RC168. [4] D.J. Cutler, R. Morris, V. Sheridhar, T.A. Wattam, S. Holmes, S. Patel, J.R. Arch, S. Wilson, R.E. Buckingham, M.L. Evans, R.A. Leslie, G. Williams, Differential distribution of orexin-A and orexinB immunoreactivity in the rat brain and spinal cord, Peptides 20 (1999) 1455–1470. [5] Y. Date, M.S. Mondal, S. Matsukura, M. Nakazato, Distribution of orexin-A and orexin-B (hypocretins) in the rat spinal cord, Neurosci. Lett. 288 (2000) 87–90. [6] L. de Lecea, T. Kilduff, C. Peyron, X.-B. Gao, P.E. Foye, P.E. Danuelson, C. Fukuhara, E.L.F. Battenberg, V.T. Gautvik, F.S. Bartlett II, W.N. Frankel, A.N. van den Pol, F.G. Bloom, K.M. Gautvik, J.G. Sutcliffe, The hypocretins: hypothalamus-specific peptides with neuroexcitatory activity, Proc. Natl. Acad. Sci. USA 95 (1998) 322–327. [7] J.-L. Guan, T. Saotome, Q.-P. Wang, H. Funahashi, T. Hori, S. Tanaka, S. Shioda, Orexinergic innervation of POMC-containing neurons in the rat arcuate nucleus, NeuroReport 12 (2001) 547–551. [8] J.-L. Guan, Q.-P. Wang, Y. Nakai, Electron microscopic observation of mu-opioid receptor in the rat area postrema, Peptides 20 (1999) 873–880. [9] M. Hallbeck, A. Blomqvist, Spinal cord-projection vasopressinergic neurons in the rat paraventricular hypothalamus, J. Comp. Neurol. 411 (1999) 201–211.
[10] A.C. Haynes, B. Jackson, H. Chapman, M. Tadayyon, A. Johns, R.A. Porter, J.R. Arch, A selective orexin-1 receptor antagonist reduces food consumption in male and female rats, Regul. Pept. 96 (2000) 45–51. [11] G.J. Hervieu, J.E. Cluderay, D.C. Harrison, J.C. Roberts, R.A. Leslie, Gene expression and protein distribution of the orexin-1 receptor in the rat brain and spinal cord, Neuroscience 103 (2001) 777–797. [12] T.L. Horvath, S. Diano, A.N. van den Pol, Synaptic interaction between hypocretin (orexin) and neuropeptide Y cells in the rodent and primate hypothalamus: a novel circuit implicated in metabolic and endocrine regulations, J. Neurosci. 19 (1999) 1072–1087. [13] T.L. Horvath, C. Peyron, S. Diano, A. Ivanov, G. Aston Jones, T.S. Kilduff, A.N. van den Pol, Hypocretin (orexin) activation and synaptic innervation of the locus coeruleus noradrenergic system, J. Comp. Neurol. 415 (1999) 145–159. [14] M.M. Methippara, M.N. Alam, R. Szymusiak, D. McGinty, Effects of lateral preoptic area application of orexin-A on sleep–wakefulness, NeuroReport 11 (2000) 3423–3426. [15] P. Penula, A.A. Aarnisalo, K. Wasowics, Neuropeptide FF, a mammalian neuropeptide with multiple function, Prog. Neurobiol. 48 (1996) 461–487. [16] T. Sakurai, A. Amemiya, M. Ishii, I. Matsuzaki, R.M. Chemelli, H. Tanaka, S.C. Williams, J.A. Richardson, G.P. Kozlowski, S. Wilson, J.R.S. Arch, R.E. Buckingham, A.C. Haynes, S.A. Carr, R.S. Annan, D.E. McNulty, W.-S. Liu, J.A. Terrett, N.A. Elshourbagy, D.J. Bergsma, M. Yanagisawa, Orexins and orexin receptors: a family of hypothalamic neuropeptides and G protein-coupled receptors that regulate feeding behavior, Cell 92 (1998) 573–585. [17] M. Schafer, S.A. Mousa, C. Stein, Corticotropin-releasing factor in antinociception and inflammation, Eur. J. Pharmacol. 323 (1997) 1–10. [18] A.N. van den Pol, Hypothalamic hypocretin (orexin): robust innervation of the spinal cord, J. Neurosci. 19 (1999) 3171–3182. [19] A.N. van den Pol, X.B. Gao, K. Obrietan, T.S. Kilduff, A.B. Belousov, Presynaptic and postsynaptic actions and modulation of neuroendocrine neurons by a new hypothalamic peptide, hypocretin / orexin, J. Neurosci. 18 (1998) 7962–7971. [20] T.W. Vanderah, M.H. Ossipov, J. Lai, T.P.J. Malan, F. Porreca, Mechanisms of opioid-induced pain and antinociceptive tolerance: descending facilitation and spinal dynorphin, Pain 92 (2001) 5–9. [21] Q.-P. Wang, J.E. Zadina, J.-L. Guan, A.J. Kastin, Endomorphin-2 immunoreactivity in the cervical dorsal horn of the rat spinal cord at the electron microscopic level, Neuroscience 113 (2002) 593–605. [22] M.-C. Xi, F.R. Morales, M.H. Chase, Effects on sleep and wakefulness of the injection of hypocretin-1 (orexin-A) into the laterodorsal tegmental nucleus of the cat, Brain Res. 901 (2001) 259–264. [23] T. Yamamoto, N. Nozaki-Taguchi, T. Chiba, Analgesic effect of intrathecally administered orexin-A in the rat formalin test and in the rat hot plate test, Br. J. Pharmacol. 137 (2002) 170–176. [24] J.E. Zadina, L. Hackler, L.-J. Ge, A.J. Kastin, A potent and selective endogenous agonist for the m-opiate receptor, Nature 386 (1997) 499–502. [25] J.H. Zhang, S. Sampogna, F.R. Morales, M.H. Chase, Orexin (hypocretin)-like immunoreactivity in the cat hypothalamus: a light and electron microscopic study, Sleep 24 (2001) 67–76. [26] Y.Z. Zhang, A.B. Malmberg, B. Sjolund, T.L. Yaksh, The effect of pituitary adenylate cyclase activating peptide (PACAP) on the nociceptive formalin test, Neurosci. Lett. 207 (1996) 187–190.