- Email: [email protected]
Origins of GABAB receptor-like immunoreactive terminals in the rat spinal dorsal horn
Brain Research Bulletin, Vol. 58, No. 5, pp. 499–507, 2002 Copyright © 2002 Elsevier Science Inc. All rights reserved. 0361-9230/02/$–see front matter
PII: S0361-9230(02)00824-9
Origins of GABAB receptor-like immunoreactive terminals in the rat spinal dorsal horn Kun Yang, Wen-Ling Ma, Yu-Peng Feng, Yuan-Xiang Dong and Yun-Qing Li∗ Department of Anatomy and K. K. Leung Brain Research Centre, The Fourth Military Medical University, Xi’an, PR China [Received 13 January 2002; Revised 13 May 2002; Accepted 31 May 2002] ABSTRACT: By means of immunohistochemistry for γ -aminobutyric acid receptor B subtype (GABAB R), the origins of GABAB R-like immunoreactive (GABAB R-LI) terminals in the rat spinal dorsal horn were investigated. After dorsal root rhizotomy and/or spinal cord hemisection, the densities of GABAB R-LI terminals were remarkably depleted in the ipsilateral superficial dorsal horn of relevant segments, whereas GABAB R-LI neurons and sparsely distributed GABAB R-LI terminals remained. After injection of Fluoro-Gold (FG) into the left side of superficial lumbar dorsal horn, FG retrograde-labeled neurons were mainly observed in the ipsilateral rostral ventromedial medulla (RVM) and brainstem raphe nuclei. Some of the FG-labeled neurons, especially in the RVM, exhibited GABAB R-like immunoreactivity. Additionally, immunofluorescence histochemical double-staining revealed that the majority of GABAB R-LI neurons in the periaqueductal gray (PAG), RVM and brainstem raphe nuclei showed 5-hydroxytryptamine (5-HT)-like immunoreactivity. The present study morphologically proves that GABAB R-LI terminals in the spinal dorsal horn originate from peripheral afferents, intrinsic neurons and supraspinal structures; GABAB R and 5-HT co-exist in many neurons in the PAG, RVM and brainstem raphe nuclei. Considering that PAG, RVM, brainstem raphe nuclei and spinal dorsal horn are important structures involved in the pain modulation, we suggest that the descending pain modulation system might be mediated, at least in part, by GABAB R. © 2002 Elsevier Science Inc. All rights reserved.
nae (laminae I and II) related to the nociceptive transmission and modulation [13,16,32]. GABAB R-LI neurons are also found in the midbrain periaqueductal gray (PAG), rostral ventromedial medulla (RVM) and brainstem raphe nuclei [16,28], which constitute the central endogenous pain control system and project to the spinal cord (see [7,17,27] for reviews). In the rat, the RVM includes the nucleus raphe magnus (NRM), nucleus reticularis gigantocellularis pars α (Rgcα) and lateral paragigantocellularis nucleus (LPGi), serves as an important relay for descending influences from the PAG to the spinal dorsal horn (see [7,17,27] for reviews). Pharmacological studies have revealed that neurons in the RVM and brainstem raphe nuclei are modulated partly by GABAB R [14,22,23]; activation of GABAB R in these nuclei influences the neurotransmitter release in the brainstem or spinal dorsal horn, by activating GABAB R on the soma of neurons in the RVM or brainstem raphe nuclei [1,2,8,18]. Because receptor is usually expressed on both soma and terminal, these data raise a possibility that some of the GABAB R-LI terminals in the dorsal horn might originate from the above mentioned supraspinal sites. 5-Hydroxytryptamine (5-HT, serotonin) is an important neurotransmitter in the descending pain modulation pathway ([4], see [7] for review). Serotoninergic projections from brainstem to the dorsal horn constitute a gating mechanism to control nociceptive transmission ([4], see [7,17,27] for reviews). Since both 5-HTand GABAB R-LI neurons are extensively encountered in the PAG, RVM and brainstem raphe nuclei [16,21,24,28,32], it is conceivable that GABAB R and 5-HT co-exist in neurons of the PAG, RVM and brainstem raphe nuclei, which send GABAB R-LI terminals to spinal dorsal horn. Although the former studies had raised a possibility that GABAB R-LI terminals in the dorsal horn might originate from primary afferent and intrinsic neurons, but the results were mainly obtained from autoradiographic [20], in situ hybridizational [24] and electrophysiological studies [3,32]. The origins of GABAB R-LI terminals in the dorsal horn are still far from being completely known. In order to clarify the origins of the GABAB R-LI terminals in the spinal dorsal horn, we addressed the above issues by employing immunohistochemistry for GABAB R after dorsal root rhizotomy, spinal cord hemisection and a combination of fluorescence retrograde-tracing with immunofluorescence histochemical staining for GABAB R. In addition, a hypothesis that GABAB R-LI neurons in the PAG, RVM and brainstem raphe nuclei contain
KEY WORDS: GABAB receptor, Spinal cord, Descending pain modulation system, 5-HT, Rat.
INTRODUCTION GABA (γ -aminobutyric acid) is the major inhibitory neurotransmitter in mammalian central nervous system (CNS). GABA receptors, which are the binding sites for GABA, are divided into three subtypes, i.e., ionotropic GABAA receptor, GABAC receptor and metabotropic GABAB receptor (GABAB R) (see [15] for review). GABAB R mediates a variety of inhibitory cellular processes in the CNS by modulating ion channels and adenyl cyclase activity [3,6,9,14,31,32]. Morphological studies have demonstrated that GABAB R is ubiquitous in the vertebrate CNS [5,13,16,20]. In the spinal dorsal horn, GABAB R-like immunoreactive (GABAB R-LI) neurons and terminals are chiefly located in the superficial lami-
∗ Address for correspondence: Dr. Yun-Qing Li, Department of Anatomy, K. K. Leung Brain Research Centre, The Fourth Military Medical University, Xi’an 710032, PR China. Fax: +86-29-3283229; E-mail: [email protected]
499
500
YANG ET AL.
5-HT was further tested by using immunofluorescence histochemical double-staining method. MATERIALS AND METHODS All protocols were approved by the Animal Care and Use Committee for Research and Education in the Fourth Military Medical University (Xi’an, PR China). All surgical procedures were performed under general anesthesia with sodium pentobarbital (40 mg/kg, i.p.). Surgical Procedures For the dorsal root rhizotomy model, six adult Sprague–Dawley (SD) male rats (250–280 g) were used. The surgical field was shaved and a longitudinal incision was made to expose the lumbar-sacral spinal segments. A laminectomy was performed and dorsal root rhizotomy on L3–6 lumbar dorsal roots on the left side was carried out within the vertebral canal. The animals were allowed to survive for 7 days after operation, and then perfused and subjected to the procedures for immunohistochemical staining for GABAB R in the lumbar cord as described below. In the spinal hemisection model, six adult SD male rats were used. A laminectomy and then a hemisection were performed at the T8 level on the left side by using a scalpel blade without damage to the posterior spinal artery or branches. Muscles and fasciae were sutured and closed, animals were allowed to recover. Following surgery, animals were maintained under the same preoperative conditions. The rats began to eat and drink 3 h after surgery, the body weight loss was less than 5% of the total-body mass in the following 7 days. In the dorsal root rhizotomy/spinal hemisection model, six adult SD male rats were subjected to an extensive rhizotomy on the left side dorsal roots (L3–6) and ipsilateral spinal hemisection (T8 level). The animals were allowed to survive for 7 days after operation, and then subjected to the immunohistochemical staining for GABAB R. For control experimental surgery, the rats were subjected to sham surgery or left untreated.
containing 0.3% Triton X-100 (PBS-X) for 2–3 h at room temperature. Subsequently, the sections were placed in 0.05 M Tris–HCl buffer (pH 7.6) containing 0.02% diaminobenzidine-4 HCl (DAB; Dojin, Kumamoto, Japan) and 0.0001% H2 O2 for 30–60 min to visualize peroxidase bound to the ABC complex. The sections in the third dish were treated for control experiment (see the following section). Immunofluorescence Histochemical Staining of GABAB R for FG-Labeled Neurons The procedures for immunofluorescence histochemical staining of GABAB R for retrograde-labeled neurons were essentially the same as described previously [11]. In brief, eight adult SD male rats (7–8 weeks old, 250–280 g) were anesthetized. After exposing the lumbar cord, a 0.05 µl of 4% (w/v) solution of Fluoro-Gold (FG; Fluorochrome Denver, CO) dissolved in distilled water was stereotaxically injected by pressure through a glass micropipette (tip diameter: 40–60 µm) attached to a 1-µl Hamilton microsyringe. Each injection was made slowly for 20 min and the injection needle was kept in place for another 20 min. The animals were
Immunohistochemical Staining for GABAB R in the Spinal Dorsal Horn Seven days after operations, animals of all three models were anesthetized and subjected to the light microscopic observation by the immunohistochemical staining for GABAB R as described elsewhere [11,32]. In brief, the animals were deeply anesthetized with an overdose sodium pentobarbital (80 mg/kg, i.p.), perfused transcardially with 100 ml of chilled saline, followed by 500 ml of 0.1 M phosphate buffer (PB; pH 7.4) containing 4% paraformaldehyde (w/v). After perfusion, the lumbar cord was removed, placed in 0.1 M PB containing 30% (w/v) sucrose overnight at 4◦ C and then transversally cut into serial sections (30 µm thick) on a freezing microtome. The sections were collected into three dishes in 0.01 M phosphate-buffered saline (PBS; pH 7.4). All the sections were washed with PBS. Nissl staining was performed on the sections in the first dish. Identification of spinal laminae was based upon the cytoarchitectonic criteria from Nissl stained sections (see [27] for a review). The sections in the second dish were incubated with affinity-purified guinea-pig IgG anti-GABAB R 1a and 1b isoforms (1 µg/ml; Chemicon, Temecula, CA) diluted with PBS containing 2% normal horse serum and 0.3% Triton X-100 (PBS-NHS) for 48 h at 4◦ C [10,32]. They were then incubated with biotinylated donkey anti-guinea-pig IgG (10 µg/ml; Chemicon) in PBS-NHS for 3–5 h, followed by the Elite ABC (avidin–biotin–peroxidase complex) kit (1:50 dilution; Vector, Burlingame, CA) in PBS
FIG. 1. (A) GABAB R-LI in the 5th segment of the lumbar cord of a rat subjected to unilateral (left) L3–6 dorsal root rhizotomy. (B) and (C) are higher magnification of the rectangle of (b) and (c) in (A), showing GABAB R-LI in the rhizotomy side (B) and control side contralateral to the dorsal rhizotomy (C). I, II and III represent laminae I, II and III, respectively. Bars = 300 µm in (A) and 120 µm in (B) and (C).
GABAB R-LI TERMINALS IN SPINAL DORSAL HORN allowed to survive for more than 5 days, and then were deeply anesthetized with an overdose sodium pentobarbital (80 mg/kg, i.p.), perfused transcardially with 100 ml of chilled saline, followed by 500 ml 0.1 M PB containing 4% (w/v) paraformaldehyde. After perfusion, the brainstem and injection site in the spinal cord were removed, placed in 0.1 M PB containing 30% (w/v) sucrose overnight at 4◦ C and then transversally cut into serial sections (30 µm thick) on a freezing microtome. The sections were collected into three dishes in 0.01 M PBS. All of the sections were washed with PBS. The sections in the first dish were mounted onto gelatin-coated glass-slides and used for Nissl staining to assist in identification of the boundaries of the different laminae of the spinal gray matter. Only the sections from the rats (n = 5) whose injection sites were limited to laminae I–III of the dorsal horn were used in the following procedures. The sections in the second dish were incubated with affinity-purified guinea-pig IgG anti-GABAB R (1 µg/ml; Chemicon) diluted with PBS-NHS for 48 h at 4◦ C. The sections were then incubated with biotinylated donkey
FIG. 2. (A) GABAB R-LI in the lumbar cord in a rat subjected to unilateral hemisection at T8 level allowed to survive for 7 days after the operation. On the operation side (B), GABAB R-LI terminals in lamina I and the outer part of lamina II is clearly reduced, while on the side contralateral to the hemisection (C), GABAB R-LI terminals within neuropil of the dorsal horn and neurons in the lateral spinal nucleus (LSN) showed hardly any change. (B) and (C) are higher magnification of the rectangle of (b) and (c) in (A), respectively. Bars = 300 µm in (A) and 120 µm in (B) and (C).
501 anti-guinea-pig IgG (10 µg/ml; Chemicon) in PBS-NHS for 3–5 h. Subsequently, the sections were placed in Avidin-Texas Red (1 µg/ml; Vector) in PBS-X for 2 h to visualize GABAB R-like immunoreactivity. After being stained, the sections were mounted onto clean glass slides, air-dried, coverslipped with 0.01 M PBS containing 50% (v/v) glycerin and 2.5% (w/v) triethylene diamine (anti-fading agent), and observed with a fluorescence microscope (BX-60; Olympus, Tokyo, Japan). An ultraviolet filter approximately providing excitation light at 360 nm and emission light at 395 nm wavelengths was used to illuminate gold-emitting FG-labeled neurons, while a green filter approximately providing excitation light at 550 nm and emission light at 610 nm wavelengths was used to illuminate red-emitting GABAB R-LI neurons. FG-labeled neurons, GABAB R-LI neurons and FG/GABAB R double-labeled neurons were counted on all of the second serial sections. The sections in the third dish were treated for control experiment (see the following section).
FIG. 3. GABAB R-LI in the lumbar cord of a rat subjected to both hemisection at the T8 level and extensive dorsal root (L3–6) rhizotomy on the left side and allowed to survive for 7 days after the operation (A). On the operation side (B), GABAB R-LI terminals in the superficial layers of the dorsal horn are clearly depleted, whereas no definite changes are observed in the lateral spinal nucleus (LSN). Some intrinsic GABAB R-LI neuronal cell bodies (arrows) and sparsely distributed GABAB R-LI terminals still remain. On the side contralateral to the operation (C), there are no changes in GABAB R-LI terminals within laminae I to III. Bars = 300 µm in (A) and 120 µm in (B) and (C).
502
YANG ET AL.
Immunofluorescence Histochemical Double-Staining of GABAB R and 5-HT The other five adult SD male rats were used for immunofluorescence histochemical double-staining of GABAB R and 5-HT as described previously [12]. Briefly, the animals were deeply anesthetized with an overdose sodium pentobarbital (80 mg/kg, i.p.), perfused transcardially with 4% (w/v) paraformaldehyde in 0.1 M PB. The brainstem was removed and cut into serial sections (30 µm thick). The sections were collected into three dishes. The sections in the first dish were mounted onto gelatin-coated glass-slides and used for Nissl staining to assist in identification of nuclei. The sections in the second dish were incubated sequentially with: (1) a mixture of guinea-pig IgG anti-GABAB R (1 µg/ml; Chemicon) and rabbit anti-5-HT (1 µg/ml; Incstar, Still Water, MN) for 48 h in PBS-NHS at 4◦ C; (2)10 µg/ml biotinylated donkey anti-guinea-pig IgG (Jackson, West Groove, PA) for 3 h in PBS-X; and (3) 10 µg/ml FITC-labeled donkey anti-rabbit IgG (Chemicon) and 5 µg/ml Texas Red-labeled avidin D (Vector). The excitation and emission wavelengths were 460 and 515 nm for FITC, respectively. The filter used to observe Texas Red was the same as described above. No significant difference in data was found between animals, so that representative data of 5-HT-LI neurons,
GABAB R-LI neurons and 5-HT/GABAB R co-localized neurons in each nucleus were counted on every third 30 µm-thick sections throughout the brainstem of the second serial sections from one rat. The sections in the third dish were treated for control experiment (see the following section). Control Experiments for Immunohistochemical Staining and Data Analysis Two sets of control experiments were performed to the sections in the third dish. First, normal guinea-pig serum and normal rabbit serum were used to replace the primary antibodies for incubating the sections. Other procedures were the same as that used for the immunohistochemical staining in the second dish. Second, for pre-absorption, the diluted antibodies (1 µg/ml) were added to 10 µg/ml of immunizing conjugate for 4 h at room temperature before addition of the sections. No immunoreactivity was found in both of the control experiments. Image analysis was performed on an IBM computer (IBM, White Plains, NY)-based densitometry system using a solid-state video camera and Leica Q500MC Image Processing and Analysis software (Leica, Wetzlar, Germany). Densities of the control experiment (both antibody omission and pre-absorption) were
FIG. 4. Photomicrographs of the three sections through brainstem showing neurons labeled with FG (A, B, C) injected into the superficial lumbar dorsal horn also exhibiting GABAB R-like immunoreactivity (A , B , C ) in the NRM (A, A ), nucleus reticularis gigantocellularis pars α (Rrgα; B, B ) and nucleus RPa (C, C ). Double arrowheads indicate neuronal cell bodies labeled with FG and also showing GABAB R-like immunoreactivity. FG single-labeled neurons and GABAB R-LI neurons are indicated with arrows (A, B, C) and single arrowhead (A , B , C ), respectively. Bar = 70 µm.
GABAB R-LI TERMINALS IN SPINAL DORSAL HORN treated as 100. The density of immunoreactivity for each rat was from four sections. Measurement for every group was expressed as mean ± SD, statistical significant was determined as p < 0.05 using Student’s t-tests. RESULTS Densities of GABAB R-LI Terminals in the Dorsal Horn Decreased after Unilateral Dorsal Root Rhizotomy and/or Spinal Hemisection On Nissl stained sections of the spinal cord, no detectable cytoarchitectonic change was found after dorsal root rhizotomy and/or spinal hemisection. In the rats that were allowed to survive for 7 days after the unilateral L3–6 dorsal root rhizotomy and/or hemisection at T8 level, GABAB R-LI terminals were densely located on the right superficial laminae (laminae I and II) of the dorsal horn
503 contralateral to the lesions. The control animals subjected to sham operation or left untreated also showed dense GABAB R-LI terminals on both sides of the superficial dorsal horn, with somewhat more intense GABAB R-LI terminals in lamina II than that in lamina I and III. GABAB R-LI neuronal cell bodies were also found in the superficial dorsal horn (Figs. 1–3). However, the densities of GABAB R-LI terminals in the superficial dorsal horn were markedly decreased on the side ipsilateral to the operations (Figs. 1–3). After rhizotomy on the L3–6 dorsal roots, GABAB R-LI terminals were markedly depleted in the corresponding region of the dorsal horn, especially in lamina II that received most of the termination of the fine primary afferent fibers (Fig. 1). The optic density of GABAB R-LI structures on the rhizotomy side was 133.41 ± 5.73, significant lower than that of the control side (162.46 ± 8.92; n = 6, p < 0.05). After unilateral hemisection at T8 level, the density of GABAB R-LI terminals in the dorsal horn was decreased
FIG. 5. Immunofluorescence histochemical double staining showing 5-HT- (A, B, C, D) and GABAB R-LI neurons (A , B , C , D ) in the midbrain periaqueducal gray (PAG; A, A ), dorsal raphe nucleus (DR; B, B ), nucleus raphe magnus (NRM; C, C ) and nucleus reticularis gigantocellularis pars α (Rgcα; D, D ), respectively. Double arrowheads stand for GABAB R/5-HT co-localized neurons. Arrows and arrowheads indicate 5-HT- (A, B, C, D) or GABAB R-LI (A , B , C , D ) neurons, respectively. The stars indicate the same blood vessel. Bar = 35 µm.
504
YANG ET AL.
GABAB R-LI TERMINALS IN SPINAL DORSAL HORN
505
TABLE 2 GABAB R AND 5-HT CO-LOCALIZED NEURONS IN THE PAG, NRM, Rgcα, ROb AND RPa
Nuclei PAG NRM Rgcα ROb RPa
GABAB R and 5-HT Co-existed Neurons/Total GABAB R-LI Neurons (%) 287/336 766/925 453/492 427/465 525/554
GABAB R and 5-HT Co-existed Neurons/Total 5-HT-LI Neurons (%)
(85.4) (82.8) (92.1) (91.8) (94.8)
apparently, especially in lamina I and the outer part of lamina II (Fig. 2). The optic density of GABAB R-LI structures for the hemisection side was 146.01 ± 6.93, significant lower than that of the control side (163.74±6.19; n = 6, p < 0.05). Extensive dorsal root (L3–6) rhizotomy and hemisection on the ipsilateral side resulted in marked depletion of GABAB R-LI terminals in the superficial laminae of the dorsal horn, whereas some GABAB R-LI intrinsic neurons in laminae I and II and sparse GABAB R-LI terminals remained (Fig. 3). The combined injury model (hemisection/dorsal root cut) also showed decreased optic density of GABAB R-LI structures on the lesion side (113.12 ± 6.11), which is much lower than that on the right control side (166.37 ± 6.25; n = 6, p < 0.01). FG Retrograde-Labeled Neurons in Supraspinal Nuclei Showed GABAB R-Like Immunoreactivity After FG injection into the lumbar cord, the injection sites were mainly confined within laminae I–III of the dorsal horn. FG retrograde-labeled neuronal cell bodies were found, bilaterally with an ipsilateral predominance in the brainstem. FG retrograde-labeled neuronal cell bodies were frequently observed in the NRM, Rgcα, nucleus raphe obscurus (ROb), nucleus raphe pallidus (RPa), lateral reticular nucleus, medullary reticular field, nucleus reticularis gigantocellularis (NGC), LPGi, and randomly distributed in ventral PAG and dorsal raphe nucleus (DR) (Fig. 4A–C). In contrast to the FG-labeled neurons predominantly distributed on the side ipsilateral to the FG injection, examination of brainstem and midbrain sections labeled with GABAB R antiserum revealed that GABAB R-LI neuronal cell bodies bilaterally occupied the NRM, Rgcα, ROb, RPa and DR (Fig. 4A –C ), as well as the ventral PAG. Some of the FG-labeled neuronal cell bodies also showed GABAB R immunoreactivity (Fig. 4). Most of the FG/GABAB R double-labeled neuronal cell bodies located unilaterally on the side ipsilateral to the FG injection site were observed in the RVM, including NRM, Rgcα, NGC, ROb, RPa (Fig. 4). Only few of the FG-labeled neurons showing GABAB R immunoreactivity were found in the PAG and DR. These double-labeled neurons accounted for approximately 30–70% of the total FG-labeled neuronal population and 10–15% of the total GABAB R-LI neurons in the above mentioned structures (Table 1). Most of the GABAB R-LI Neurons in RVM and PAG Contained 5-HT The data of immunofluorescence histochemical double-staining for GABAB R and 5-HT revealed that neuronal cell bodies showing GABAB R- or 5-HT-like immunoreactivity and both GABAB Rand 5-HT-like immunoreactivities were encountered throughout the midline structures of the brainstem (Fig. 5). Co-localization of GABAB R and 5-HT was observed in most neurons showing GABAB R-like immunoreactivity in the NRM, Rgcα, NGC, ROb, RPa, DR, and the ventral PAG (Fig. 5). The rates of GABAB R/5-HT co-existing neurons to the total number of
287/544 766/987 453/527 427/502 525/615
(52.8) (77.6) (86.0) (85.1) (85.4)
GABAB R- and 5-HT-LI neurons were 82.8–94.8 and 52.8–86.0%, respectively (Table 2). DISCUSSION Origins of GABAB R-LI Terminals in the Spinal Dorsal Horn Our previous electrophysiological study has shown that in lamina II of the superficial dorsal horn, there are three possible origins of functional glutamatergic terminals: peripheral, intrinsic and supraspinal structures [30]. Theoretically, there are also three possible origins for GABAB R-LI terminals in the spinal dorsal horn: primary afferent central terminals, intrinsic dorsal horn neurons and descending pain modulation system (see [27] for review). Although the possibility of peripheral and intrinsic origins has been raised by the former works [3,5,13,24,32], our present study provides the direct morphological evidence for that. Moreover, since the spinal hemisection decreases the density of the GABAB R-LI terminals in the ipsilateral dorsal horn, the results anatomically indicate that GABAB R-LI terminals also originate from supraspinal structures. It is interesting to find GABAB R-LI neurons in the brainstem projecting to the superficial dorsal horn. The dense distribution and complex origins of GABAB R-LI structures in the dorsal horn suggest that GABAB R might play a complex role in the information transmission and/or modulation in the spinal cord. Previous studies have reported that baclofen, a selective GABAB R agonist, depresses the presynaptic excitatory neurotransmitter release in spinal dorsal horn from either primary afferent central terminals [3,32] or terminals originating from intrinsic and central structures [9,29,32]. Moreover, activation of GABAB R decreases inhibitory neurotransmitter release in the dorsal horn [9,31]. Although neurons in the brainstem sending projection fibers to the dorsal horn principally contain inhibitory neurotransmitters, our present data indicate that activation of GABAB R-LI terminals in the spinal cord might also affect the descending inhibitory neurotransmitter release through GABA by binding to these GABAB R on the descending terminals. It should be noted that, the lesion of the descending tract might affect the normal GABAB R expression on the central terminals of the primary afferents, or conversely, that rhizotomy of the dorsal roots could also affect normal GABAB R expression on the terminals of descending projections. On the other hand, both injury models might also influent the GABAB R expression of the intrinsic neurons in the dorsal horn. To interpret the present results correctly, these possibilities should not be neglected. However, these possibilities are limited because we have also observed that no detectable cytoarchitectonic changes are found after dorsal root rhizotomy and/or spinal hemisection. Significance of GABAB R-LI Neurons in the Central Endogenous Pain Control System In the present study, many GABAB R-LI neuronal cell bodies were observed in the PAG, RVM and their adjacent structures. The
506 distribution pattern of GABAB R-LI neurons in these structures is in agreement with that reported by Margeta-Mitrovic et al. [16]. Our present study further proves that some FG retrograde-labeled neurons in the PAG, RVM and brainstem raphe nuclei show GABAB R-like immunoreactivity, these results support the above mentioned hypothesis that there are some GABAB R-LI terminals descending from supraspinal structures to the dorsal horn. Many FG retrograde-labeled neurons have been found to exhibit GABAB R-like immunoreactivity in the central endogenous pain control system, which indicates that these neurons in the brainstem projecting to spinal dorsal horn are probably modulated by GABA via binding to the GABAB R in the supraspinal sites [1,2,8,18]. In the present study, although many GABAB R-LI cell bodies were found in ventral PAG, not many FG-labeled neurons were encountered in the PAG. The scarcity of FG-labeled neurons might be due to the fact that few PAG neurons project directly to the spinal dorsal horn, whereas most of them terminate in the RVM (see [7] for review). Co-localization of 5-HT and GABAB R-Like Immunoreactivities in the RVM The descending pain modulation effect is mainly mediated by 5-HT release in the dorsal horn (see [7,17] for reviews). Co-localization of GABAB R with monoamine in midbrain monoaminergic neurons has been reported [28]. It is also presumed that GABAB R-LI neurons in the RVM, probably containing 5-HT, project directly to dorsal horn [1]. Activation of GABAB R in cell bodies and dendrites elicits slow inhibitory postsynaptic potentials, whereas activation of GABAB R in presynaptic nerve terminals inhibits neurotransmitter release [3,9,29,31,32]. Since GABAB R could exert its inhibition by either presynaptic or postsynaptic mechanisms, likewise, the activation of GABAB R in RVM could either inhibit presynaptic neurotransmitter release or hyperpolarize postsynaptic neurons that project to the dorsal horn [14]. The net effect of the GABAB R activation in RVM depends on the different rate of pre- and postsynaptic GABAB R being activated. Based upon this hypothesis, Abellán et al. [2] and Tao et al. [22] have discovered that different doses of baclofen administrated to the RVM affect different ratios of presynaptic or postsynaptic GABAB R, and thus regulate the 5-HT release in the dorsal horn. The present results suggest morphologically that those 5-HT containing neurons, which project to dorsal horn, might be modulated by GABAB R. To address this issue in detail, further study will be required. Colocalization of 5-HT- and GABAB R-Like Immunoreactivities in the PAG In the present study, many GABAB R-LI neurons in the PAG were found containing 5-HT. It has been characterized that PAG neurons provides a major serotoninergic input to the RVM and terminate onto serotoninergic neurons, ON-cells, and OFF-cells (see [7] for review). Activation of opioid µ or κ receptors in the PAG disinhibits NRM-projection neurons and results in increasing 5-HT release within the dorsal horn and thus influences spinal nociception transmission [19,25]. Osborne et al. [19] demonstrated that both GABAB R agonist baclofen and opioid µ receptor agonist methionine–enkephalin produce an outward current (hyperpolarization) in PAG neurons with a similar characterization. Contrary to both presynaptic and postsynaptic actions of GABAB R activation in RVM, in PAG, opioid receptor activation predominantly inhibits synaptic transmission by reducing the probability of presynaptic neurotransmitter release, controlled by a presynaptic voltage-dependent potassium conductance [26]. Both opioid µ receptor and GABAB R are G protein-coupled inhibitory receptor in the CNS. The activation of µ receptor or GABAB R is known to inhibit calcium channel and activate potassium channel. It is
YANG ET AL. conceivable that GABAB R plays a similar role as that of opioid µ receptor in PAG with disinhibitory effect on PAG neurons, and thus affects the 5-HT release in RVM. Some studies have been performed to verify this hypothesis [1,2]. CONCLUSION Our present study provides direct morphological evidence for the hypothesis that in addition to peripheral and intrinsic origins, some GABAB R-LI terminals in the superficial dorsal horn originate from supraspinal structures; moreover, some GABAB R-LI neurons in the PAG, RVM and brainstem raphe nuclei also contain 5-HT. The results further confirm the modulatory role of GABAB R in nociceptive transmission. ACKNOWLEDGEMENTS
The authors are grateful for photographic help of Ms. Yue-Ping Yuan. This work was supported by Grants-in-Aid from the National Natural Science Foundation of China (30000052, 39970239, 30070389), Foundation for University Key Teacher by the Ministry of Education of China and the National Program of Basic Research of China (G1999054004) and Grant for Doctoral Dissertation from the Fourth Military Medical University. REFERENCES 1. Abellán, M. T.; Adell, A.; Honrubia, M. A.; Mengod, G.; Artigas, F. GABAB -R1 receptors in serotonergic neurons: Effects of baclofen on 5-HT output in rat brain. NeuroReport 11:941–945; 2000. 2. Abellán, M. T.; Jolas, T.; Aghajanian, G. K.; Artigas, F. Dual control of dorsal raphe serotonergic neurons by GABAB receptors. Electrophysiological and microdialysis studies. Synapse 36:21–34; 2000. 3. Ataka, T.; Kumamoto, E.; Shimoji, K.; Yoshimura, M. Baclofen inhibits more effectively C-afferent than Aδ-afferent glutamatergic transmission in substantia gelatinosa neurons of adult rat spinal cord slices. Pain 86:273–282; 2000. 4. Bowker, R. M.; Reddy, V. K.; Fung, S. J.; Chan, J. Y. H.; Barnes, C. D. Serotonergic and non-serotonergic raphe neurons projecting to the feline lumbar and cervical spinal cord: A quantitative horseradish peroxidase-immunocytochemical study. Neurosci. Lett. 75:31–37; 1987. 5. Charles, K. J.; Evans, M. L.; Robbins, M. J.; Calver, A. R.; Leslie, R. A.; Pangalos, M. N. Comparative immunohistochemical localization of GABAB 1a, GABAB 1b and GABAB 2 subunits in rat brain, spinal cord and dorsal root ganglion. Neuroscience 106:447–467; 2001. 6. Dirig, D. M.; Yaksh, T. L. Intrathecal baclofen and muscimol, but not midazolam, are antinociceptive using the rat-formalin model. J. Pharmacol. Exp. Ther. 275:219–227; 1995. 7. Fields, H. L.; Heinricher, M. M.; Mason, P. Neurotransmitters in nociceptive modulatory circuits. Annu. Rev. Neurosci. 14:219–245; 1991. 8. Innis, R. B.; Nestler, E. J.; Aghajanian, G. K. Evidence for G protein mediation of serotonin- and GABAB -induced hyperpolarization of rat dorsal raphe neurons. Brain Res. 459:27–36; 1988. 9. Iyadomi, M.; Iyadomi, I.; Kumamoto, E.; Tomokuni, K.; Yoshimura, M. Presynaptic inhibition by baclofen of miniature EPSCs and IPSCs in substantia gelatinosa neurons of the adult rat spinal dorsal horn. Pain 85:385–393; 2000. 10. Jones, K. A.; Borowsky, B.; Tamm, J. A.; Craig, D. A.; Durkin, M. M.; Dai, M.; Yao, W.-J.; Johnson, M.; Gunwaldsen, C.; Huang, L.-Y.; Tang, C.; Shen, Q.; Salon, J. A.; Morse, K.; Laz, T.; Smith, K. E.; Nagarathnam, D.; Noble, S. A.; Branchek, T. A.; Gerald, C. GABAB receptors function as a heteromeric assembly of the subunits GABAB R1 and GABAB R2. Nature 396:674–679; 1998. 11. Li, Y.-Q.; Wang, Z.-M.; Zheng, H.-X.; Shi, J.-W. Central origins of substance P-like immunoreactive fibers and terminals in the spinal trigeminal caudal subnucleus in the rat. Brain Res. 719:219–224; 1996.
GABAB R-LI TERMINALS IN SPINAL DORSAL HORN 12. Li, Y.-Q.; Wu, S.-X.; Li, J.-L.; Li, J.-S.; Kaneko, T.; Mizuno, N. Co-existence of calcium-binding proteins in neurons of the medullary dorsal horn of the rat. Neurosci. Lett. 286:103–106; 2000. 13. Liang, F.; Hatanaka, Y.; Saito, H.; Yamamori, T.; Hashikawa, T. Differential expression of γ -aminobutyric acid type B receptor-1a and -1b mRNA variants in GABA and non-GABAergic neurons of the rat brain. J. Comp. Neurol. 416:475–495; 2000. 14. Lin, H. H.; Dun, N. J. Post- and presynaptic GABAB receptor activation in neonatal rat rostral ventrolateral medulla neurons in vitro. Neuroscience 86:211–220; 1998. 15. Malcangio, M.; Bowery, N. G. GABA and its receptors in the spinal cord. Trends Pharmacol. Sci. 17:457–462; 1996. 16. Margeta-Mitrovic, M.; Mitrovic, I.; Riley, R. C.; Jan, L. Y.; Basbaum, A. I. Immunohistochemical localization of GABAB receptors in the rat central nervous system. J. Comp. Neurol. 405:299–321; 1999. 17. Mason, P. Central mechanisms of pain modulation. Curr. Opin. Neurobiol. 9:436–441; 1999. 18. McGowan, M. K.; Hammond, D. L. Intrathecal GABAB antagonists attenuate the antinociception produced by microinjection of l-glutamate into the ventromedial medulla of the rat. Brain Res. 607:39–46; 1993. 19. Osborne, P. B.; Vaughan, C. W.; Wilson, H. I.; Christie, M. J. Opioid inhibition of rat periaqueductal gray neurones with identified projections to rostral ventromedial medulla in vitro. J. Physiol. (London) 490:383–389; 1996. 20. Price, G. W.; Wilkin, G. P.; Tunbull, M. J.; Bowery, N. G. Are baclofen-sensitive GABAB receptors present on primary afferent terminals of the spinal cord? Nature 307:71–74; 1984. 21. Steinbusch, H. W. M. Distribution of serotonin-immunoreactivity in the central nervous system of the rat-cell bodies and terminals. Neuroscience 6:557–618; 1981. 22. Tao, R.; Ma, Z.; Auerbach, S. B. Differential regulation of 5-hydroxytryptamine release by GABAA and GABAB receptors in midbrain raphe nuclei and forebrain of rats. Br. J. Pharmacol. 119:1375–1384; 1996.
507 23. Thomas, D. A.; McGowan, M. K.; Hammond, D. L. Microinjection of baclofen in the ventromedial medulla of rats: Antinociception at low doses and hyperalgesia at high doses. J. Pharmacol. Exp. Ther. 275:274–284; 1995. 24. Towers, S.; Princivalle, A.; Billinton, A.; Edmunds, M.; Bettler, B.; Urban, L.; Castro-Lopes, J.; Bowery, N. G. GABAB receptor protein and mRNA distribution in rat spinal cord and dorsal root ganglia. Eur. J. Neurosci. 12:3201–3210; 2000. 25. Vaughan, C. W.; Connor, M.; Jennings, E. A.; Marinelli, S.; Allen, R. G.; Christie, M. J. Actions of nociceptin/orphanin FQ and other prepronociceptin products on rat rostral ventromedial medulla neurons in vitro. J. Physiol. (London) 534:849–859; 2001. 26. Vaughan, C. W.; Ingram, S. L.; Connor, M. A.; Christie, M. J. How opioids inhibit GABA-mediated neurotransmission. Nature 390:611– 614; 1997. 27. Willis, W. D.; Coggeshall, R. E. Sensory mechanisms of the spinal cord, 2nd ed. New York: Plenum Press, 1991. 28. Wirtshafter, D.; Sheppard, A. C. Localization of GABAB receptors in midbrain monoamine containing neurons in the rat. Brain Res. Bull. 56:1–5; 2001. 29. Yang, K.; Kumamoto, E.; Furue, H.; Li, Y.-Q.; Yoshimura, M. Capsaicin induces a slow inward current which is not mediated by substance P in substantia gelatinosa neurons of the rat spinal cord. Neuropharmacology 39:2185–2194; 2000. 30. Yang, K.; Li, Y.-Q. Origins of spontaneous and noxious stimuli-evoked miniature EPSCs in substantia gelatinosa. NeuroReport 12:39–42; 2001. 31. Yang, K.; Feng, Y.-P.; Li, Y.-Q. Baclofen inhibition of dorsal root-evoked inhibitory postsynaptic currents in substantia gelatinosa neurons of rat spinal cord slice. Brain Res. 900:320–323; 2001. 32. Yang, K.; Wang, D.; Li, Y.-Q. Distribution and depression of the GABAB receptor in the spinal dorsal horn of adult rat. Brain Res. Bull. 55:279–285; 2001.
PII: S0361-9230(02)00824-9
Origins of GABAB receptor-like immunoreactive terminals in the rat spinal dorsal horn Kun Yang, Wen-Ling Ma, Yu-Peng Feng, Yuan-Xiang Dong and Yun-Qing Li∗ Department of Anatomy and K. K. Leung Brain Research Centre, The Fourth Military Medical University, Xi’an, PR China [Received 13 January 2002; Revised 13 May 2002; Accepted 31 May 2002] ABSTRACT: By means of immunohistochemistry for γ -aminobutyric acid receptor B subtype (GABAB R), the origins of GABAB R-like immunoreactive (GABAB R-LI) terminals in the rat spinal dorsal horn were investigated. After dorsal root rhizotomy and/or spinal cord hemisection, the densities of GABAB R-LI terminals were remarkably depleted in the ipsilateral superficial dorsal horn of relevant segments, whereas GABAB R-LI neurons and sparsely distributed GABAB R-LI terminals remained. After injection of Fluoro-Gold (FG) into the left side of superficial lumbar dorsal horn, FG retrograde-labeled neurons were mainly observed in the ipsilateral rostral ventromedial medulla (RVM) and brainstem raphe nuclei. Some of the FG-labeled neurons, especially in the RVM, exhibited GABAB R-like immunoreactivity. Additionally, immunofluorescence histochemical double-staining revealed that the majority of GABAB R-LI neurons in the periaqueductal gray (PAG), RVM and brainstem raphe nuclei showed 5-hydroxytryptamine (5-HT)-like immunoreactivity. The present study morphologically proves that GABAB R-LI terminals in the spinal dorsal horn originate from peripheral afferents, intrinsic neurons and supraspinal structures; GABAB R and 5-HT co-exist in many neurons in the PAG, RVM and brainstem raphe nuclei. Considering that PAG, RVM, brainstem raphe nuclei and spinal dorsal horn are important structures involved in the pain modulation, we suggest that the descending pain modulation system might be mediated, at least in part, by GABAB R. © 2002 Elsevier Science Inc. All rights reserved.
nae (laminae I and II) related to the nociceptive transmission and modulation [13,16,32]. GABAB R-LI neurons are also found in the midbrain periaqueductal gray (PAG), rostral ventromedial medulla (RVM) and brainstem raphe nuclei [16,28], which constitute the central endogenous pain control system and project to the spinal cord (see [7,17,27] for reviews). In the rat, the RVM includes the nucleus raphe magnus (NRM), nucleus reticularis gigantocellularis pars α (Rgcα) and lateral paragigantocellularis nucleus (LPGi), serves as an important relay for descending influences from the PAG to the spinal dorsal horn (see [7,17,27] for reviews). Pharmacological studies have revealed that neurons in the RVM and brainstem raphe nuclei are modulated partly by GABAB R [14,22,23]; activation of GABAB R in these nuclei influences the neurotransmitter release in the brainstem or spinal dorsal horn, by activating GABAB R on the soma of neurons in the RVM or brainstem raphe nuclei [1,2,8,18]. Because receptor is usually expressed on both soma and terminal, these data raise a possibility that some of the GABAB R-LI terminals in the dorsal horn might originate from the above mentioned supraspinal sites. 5-Hydroxytryptamine (5-HT, serotonin) is an important neurotransmitter in the descending pain modulation pathway ([4], see [7] for review). Serotoninergic projections from brainstem to the dorsal horn constitute a gating mechanism to control nociceptive transmission ([4], see [7,17,27] for reviews). Since both 5-HTand GABAB R-LI neurons are extensively encountered in the PAG, RVM and brainstem raphe nuclei [16,21,24,28,32], it is conceivable that GABAB R and 5-HT co-exist in neurons of the PAG, RVM and brainstem raphe nuclei, which send GABAB R-LI terminals to spinal dorsal horn. Although the former studies had raised a possibility that GABAB R-LI terminals in the dorsal horn might originate from primary afferent and intrinsic neurons, but the results were mainly obtained from autoradiographic [20], in situ hybridizational [24] and electrophysiological studies [3,32]. The origins of GABAB R-LI terminals in the dorsal horn are still far from being completely known. In order to clarify the origins of the GABAB R-LI terminals in the spinal dorsal horn, we addressed the above issues by employing immunohistochemistry for GABAB R after dorsal root rhizotomy, spinal cord hemisection and a combination of fluorescence retrograde-tracing with immunofluorescence histochemical staining for GABAB R. In addition, a hypothesis that GABAB R-LI neurons in the PAG, RVM and brainstem raphe nuclei contain
KEY WORDS: GABAB receptor, Spinal cord, Descending pain modulation system, 5-HT, Rat.
INTRODUCTION GABA (γ -aminobutyric acid) is the major inhibitory neurotransmitter in mammalian central nervous system (CNS). GABA receptors, which are the binding sites for GABA, are divided into three subtypes, i.e., ionotropic GABAA receptor, GABAC receptor and metabotropic GABAB receptor (GABAB R) (see [15] for review). GABAB R mediates a variety of inhibitory cellular processes in the CNS by modulating ion channels and adenyl cyclase activity [3,6,9,14,31,32]. Morphological studies have demonstrated that GABAB R is ubiquitous in the vertebrate CNS [5,13,16,20]. In the spinal dorsal horn, GABAB R-like immunoreactive (GABAB R-LI) neurons and terminals are chiefly located in the superficial lami-
∗ Address for correspondence: Dr. Yun-Qing Li, Department of Anatomy, K. K. Leung Brain Research Centre, The Fourth Military Medical University, Xi’an 710032, PR China. Fax: +86-29-3283229; E-mail: [email protected]
499
500
YANG ET AL.
5-HT was further tested by using immunofluorescence histochemical double-staining method. MATERIALS AND METHODS All protocols were approved by the Animal Care and Use Committee for Research and Education in the Fourth Military Medical University (Xi’an, PR China). All surgical procedures were performed under general anesthesia with sodium pentobarbital (40 mg/kg, i.p.). Surgical Procedures For the dorsal root rhizotomy model, six adult Sprague–Dawley (SD) male rats (250–280 g) were used. The surgical field was shaved and a longitudinal incision was made to expose the lumbar-sacral spinal segments. A laminectomy was performed and dorsal root rhizotomy on L3–6 lumbar dorsal roots on the left side was carried out within the vertebral canal. The animals were allowed to survive for 7 days after operation, and then perfused and subjected to the procedures for immunohistochemical staining for GABAB R in the lumbar cord as described below. In the spinal hemisection model, six adult SD male rats were used. A laminectomy and then a hemisection were performed at the T8 level on the left side by using a scalpel blade without damage to the posterior spinal artery or branches. Muscles and fasciae were sutured and closed, animals were allowed to recover. Following surgery, animals were maintained under the same preoperative conditions. The rats began to eat and drink 3 h after surgery, the body weight loss was less than 5% of the total-body mass in the following 7 days. In the dorsal root rhizotomy/spinal hemisection model, six adult SD male rats were subjected to an extensive rhizotomy on the left side dorsal roots (L3–6) and ipsilateral spinal hemisection (T8 level). The animals were allowed to survive for 7 days after operation, and then subjected to the immunohistochemical staining for GABAB R. For control experimental surgery, the rats were subjected to sham surgery or left untreated.
containing 0.3% Triton X-100 (PBS-X) for 2–3 h at room temperature. Subsequently, the sections were placed in 0.05 M Tris–HCl buffer (pH 7.6) containing 0.02% diaminobenzidine-4 HCl (DAB; Dojin, Kumamoto, Japan) and 0.0001% H2 O2 for 30–60 min to visualize peroxidase bound to the ABC complex. The sections in the third dish were treated for control experiment (see the following section). Immunofluorescence Histochemical Staining of GABAB R for FG-Labeled Neurons The procedures for immunofluorescence histochemical staining of GABAB R for retrograde-labeled neurons were essentially the same as described previously [11]. In brief, eight adult SD male rats (7–8 weeks old, 250–280 g) were anesthetized. After exposing the lumbar cord, a 0.05 µl of 4% (w/v) solution of Fluoro-Gold (FG; Fluorochrome Denver, CO) dissolved in distilled water was stereotaxically injected by pressure through a glass micropipette (tip diameter: 40–60 µm) attached to a 1-µl Hamilton microsyringe. Each injection was made slowly for 20 min and the injection needle was kept in place for another 20 min. The animals were
Immunohistochemical Staining for GABAB R in the Spinal Dorsal Horn Seven days after operations, animals of all three models were anesthetized and subjected to the light microscopic observation by the immunohistochemical staining for GABAB R as described elsewhere [11,32]. In brief, the animals were deeply anesthetized with an overdose sodium pentobarbital (80 mg/kg, i.p.), perfused transcardially with 100 ml of chilled saline, followed by 500 ml of 0.1 M phosphate buffer (PB; pH 7.4) containing 4% paraformaldehyde (w/v). After perfusion, the lumbar cord was removed, placed in 0.1 M PB containing 30% (w/v) sucrose overnight at 4◦ C and then transversally cut into serial sections (30 µm thick) on a freezing microtome. The sections were collected into three dishes in 0.01 M phosphate-buffered saline (PBS; pH 7.4). All the sections were washed with PBS. Nissl staining was performed on the sections in the first dish. Identification of spinal laminae was based upon the cytoarchitectonic criteria from Nissl stained sections (see [27] for a review). The sections in the second dish were incubated with affinity-purified guinea-pig IgG anti-GABAB R 1a and 1b isoforms (1 µg/ml; Chemicon, Temecula, CA) diluted with PBS containing 2% normal horse serum and 0.3% Triton X-100 (PBS-NHS) for 48 h at 4◦ C [10,32]. They were then incubated with biotinylated donkey anti-guinea-pig IgG (10 µg/ml; Chemicon) in PBS-NHS for 3–5 h, followed by the Elite ABC (avidin–biotin–peroxidase complex) kit (1:50 dilution; Vector, Burlingame, CA) in PBS
FIG. 1. (A) GABAB R-LI in the 5th segment of the lumbar cord of a rat subjected to unilateral (left) L3–6 dorsal root rhizotomy. (B) and (C) are higher magnification of the rectangle of (b) and (c) in (A), showing GABAB R-LI in the rhizotomy side (B) and control side contralateral to the dorsal rhizotomy (C). I, II and III represent laminae I, II and III, respectively. Bars = 300 µm in (A) and 120 µm in (B) and (C).
GABAB R-LI TERMINALS IN SPINAL DORSAL HORN allowed to survive for more than 5 days, and then were deeply anesthetized with an overdose sodium pentobarbital (80 mg/kg, i.p.), perfused transcardially with 100 ml of chilled saline, followed by 500 ml 0.1 M PB containing 4% (w/v) paraformaldehyde. After perfusion, the brainstem and injection site in the spinal cord were removed, placed in 0.1 M PB containing 30% (w/v) sucrose overnight at 4◦ C and then transversally cut into serial sections (30 µm thick) on a freezing microtome. The sections were collected into three dishes in 0.01 M PBS. All of the sections were washed with PBS. The sections in the first dish were mounted onto gelatin-coated glass-slides and used for Nissl staining to assist in identification of the boundaries of the different laminae of the spinal gray matter. Only the sections from the rats (n = 5) whose injection sites were limited to laminae I–III of the dorsal horn were used in the following procedures. The sections in the second dish were incubated with affinity-purified guinea-pig IgG anti-GABAB R (1 µg/ml; Chemicon) diluted with PBS-NHS for 48 h at 4◦ C. The sections were then incubated with biotinylated donkey
FIG. 2. (A) GABAB R-LI in the lumbar cord in a rat subjected to unilateral hemisection at T8 level allowed to survive for 7 days after the operation. On the operation side (B), GABAB R-LI terminals in lamina I and the outer part of lamina II is clearly reduced, while on the side contralateral to the hemisection (C), GABAB R-LI terminals within neuropil of the dorsal horn and neurons in the lateral spinal nucleus (LSN) showed hardly any change. (B) and (C) are higher magnification of the rectangle of (b) and (c) in (A), respectively. Bars = 300 µm in (A) and 120 µm in (B) and (C).
501 anti-guinea-pig IgG (10 µg/ml; Chemicon) in PBS-NHS for 3–5 h. Subsequently, the sections were placed in Avidin-Texas Red (1 µg/ml; Vector) in PBS-X for 2 h to visualize GABAB R-like immunoreactivity. After being stained, the sections were mounted onto clean glass slides, air-dried, coverslipped with 0.01 M PBS containing 50% (v/v) glycerin and 2.5% (w/v) triethylene diamine (anti-fading agent), and observed with a fluorescence microscope (BX-60; Olympus, Tokyo, Japan). An ultraviolet filter approximately providing excitation light at 360 nm and emission light at 395 nm wavelengths was used to illuminate gold-emitting FG-labeled neurons, while a green filter approximately providing excitation light at 550 nm and emission light at 610 nm wavelengths was used to illuminate red-emitting GABAB R-LI neurons. FG-labeled neurons, GABAB R-LI neurons and FG/GABAB R double-labeled neurons were counted on all of the second serial sections. The sections in the third dish were treated for control experiment (see the following section).
FIG. 3. GABAB R-LI in the lumbar cord of a rat subjected to both hemisection at the T8 level and extensive dorsal root (L3–6) rhizotomy on the left side and allowed to survive for 7 days after the operation (A). On the operation side (B), GABAB R-LI terminals in the superficial layers of the dorsal horn are clearly depleted, whereas no definite changes are observed in the lateral spinal nucleus (LSN). Some intrinsic GABAB R-LI neuronal cell bodies (arrows) and sparsely distributed GABAB R-LI terminals still remain. On the side contralateral to the operation (C), there are no changes in GABAB R-LI terminals within laminae I to III. Bars = 300 µm in (A) and 120 µm in (B) and (C).
502
YANG ET AL.
Immunofluorescence Histochemical Double-Staining of GABAB R and 5-HT The other five adult SD male rats were used for immunofluorescence histochemical double-staining of GABAB R and 5-HT as described previously [12]. Briefly, the animals were deeply anesthetized with an overdose sodium pentobarbital (80 mg/kg, i.p.), perfused transcardially with 4% (w/v) paraformaldehyde in 0.1 M PB. The brainstem was removed and cut into serial sections (30 µm thick). The sections were collected into three dishes. The sections in the first dish were mounted onto gelatin-coated glass-slides and used for Nissl staining to assist in identification of nuclei. The sections in the second dish were incubated sequentially with: (1) a mixture of guinea-pig IgG anti-GABAB R (1 µg/ml; Chemicon) and rabbit anti-5-HT (1 µg/ml; Incstar, Still Water, MN) for 48 h in PBS-NHS at 4◦ C; (2)10 µg/ml biotinylated donkey anti-guinea-pig IgG (Jackson, West Groove, PA) for 3 h in PBS-X; and (3) 10 µg/ml FITC-labeled donkey anti-rabbit IgG (Chemicon) and 5 µg/ml Texas Red-labeled avidin D (Vector). The excitation and emission wavelengths were 460 and 515 nm for FITC, respectively. The filter used to observe Texas Red was the same as described above. No significant difference in data was found between animals, so that representative data of 5-HT-LI neurons,
GABAB R-LI neurons and 5-HT/GABAB R co-localized neurons in each nucleus were counted on every third 30 µm-thick sections throughout the brainstem of the second serial sections from one rat. The sections in the third dish were treated for control experiment (see the following section). Control Experiments for Immunohistochemical Staining and Data Analysis Two sets of control experiments were performed to the sections in the third dish. First, normal guinea-pig serum and normal rabbit serum were used to replace the primary antibodies for incubating the sections. Other procedures were the same as that used for the immunohistochemical staining in the second dish. Second, for pre-absorption, the diluted antibodies (1 µg/ml) were added to 10 µg/ml of immunizing conjugate for 4 h at room temperature before addition of the sections. No immunoreactivity was found in both of the control experiments. Image analysis was performed on an IBM computer (IBM, White Plains, NY)-based densitometry system using a solid-state video camera and Leica Q500MC Image Processing and Analysis software (Leica, Wetzlar, Germany). Densities of the control experiment (both antibody omission and pre-absorption) were
FIG. 4. Photomicrographs of the three sections through brainstem showing neurons labeled with FG (A, B, C) injected into the superficial lumbar dorsal horn also exhibiting GABAB R-like immunoreactivity (A , B , C ) in the NRM (A, A ), nucleus reticularis gigantocellularis pars α (Rrgα; B, B ) and nucleus RPa (C, C ). Double arrowheads indicate neuronal cell bodies labeled with FG and also showing GABAB R-like immunoreactivity. FG single-labeled neurons and GABAB R-LI neurons are indicated with arrows (A, B, C) and single arrowhead (A , B , C ), respectively. Bar = 70 µm.
GABAB R-LI TERMINALS IN SPINAL DORSAL HORN treated as 100. The density of immunoreactivity for each rat was from four sections. Measurement for every group was expressed as mean ± SD, statistical significant was determined as p < 0.05 using Student’s t-tests. RESULTS Densities of GABAB R-LI Terminals in the Dorsal Horn Decreased after Unilateral Dorsal Root Rhizotomy and/or Spinal Hemisection On Nissl stained sections of the spinal cord, no detectable cytoarchitectonic change was found after dorsal root rhizotomy and/or spinal hemisection. In the rats that were allowed to survive for 7 days after the unilateral L3–6 dorsal root rhizotomy and/or hemisection at T8 level, GABAB R-LI terminals were densely located on the right superficial laminae (laminae I and II) of the dorsal horn
503 contralateral to the lesions. The control animals subjected to sham operation or left untreated also showed dense GABAB R-LI terminals on both sides of the superficial dorsal horn, with somewhat more intense GABAB R-LI terminals in lamina II than that in lamina I and III. GABAB R-LI neuronal cell bodies were also found in the superficial dorsal horn (Figs. 1–3). However, the densities of GABAB R-LI terminals in the superficial dorsal horn were markedly decreased on the side ipsilateral to the operations (Figs. 1–3). After rhizotomy on the L3–6 dorsal roots, GABAB R-LI terminals were markedly depleted in the corresponding region of the dorsal horn, especially in lamina II that received most of the termination of the fine primary afferent fibers (Fig. 1). The optic density of GABAB R-LI structures on the rhizotomy side was 133.41 ± 5.73, significant lower than that of the control side (162.46 ± 8.92; n = 6, p < 0.05). After unilateral hemisection at T8 level, the density of GABAB R-LI terminals in the dorsal horn was decreased
FIG. 5. Immunofluorescence histochemical double staining showing 5-HT- (A, B, C, D) and GABAB R-LI neurons (A , B , C , D ) in the midbrain periaqueducal gray (PAG; A, A ), dorsal raphe nucleus (DR; B, B ), nucleus raphe magnus (NRM; C, C ) and nucleus reticularis gigantocellularis pars α (Rgcα; D, D ), respectively. Double arrowheads stand for GABAB R/5-HT co-localized neurons. Arrows and arrowheads indicate 5-HT- (A, B, C, D) or GABAB R-LI (A , B , C , D ) neurons, respectively. The stars indicate the same blood vessel. Bar = 35 µm.
504
YANG ET AL.
GABAB R-LI TERMINALS IN SPINAL DORSAL HORN
505
TABLE 2 GABAB R AND 5-HT CO-LOCALIZED NEURONS IN THE PAG, NRM, Rgcα, ROb AND RPa
Nuclei PAG NRM Rgcα ROb RPa
GABAB R and 5-HT Co-existed Neurons/Total GABAB R-LI Neurons (%) 287/336 766/925 453/492 427/465 525/554
GABAB R and 5-HT Co-existed Neurons/Total 5-HT-LI Neurons (%)
(85.4) (82.8) (92.1) (91.8) (94.8)
apparently, especially in lamina I and the outer part of lamina II (Fig. 2). The optic density of GABAB R-LI structures for the hemisection side was 146.01 ± 6.93, significant lower than that of the control side (163.74±6.19; n = 6, p < 0.05). Extensive dorsal root (L3–6) rhizotomy and hemisection on the ipsilateral side resulted in marked depletion of GABAB R-LI terminals in the superficial laminae of the dorsal horn, whereas some GABAB R-LI intrinsic neurons in laminae I and II and sparse GABAB R-LI terminals remained (Fig. 3). The combined injury model (hemisection/dorsal root cut) also showed decreased optic density of GABAB R-LI structures on the lesion side (113.12 ± 6.11), which is much lower than that on the right control side (166.37 ± 6.25; n = 6, p < 0.01). FG Retrograde-Labeled Neurons in Supraspinal Nuclei Showed GABAB R-Like Immunoreactivity After FG injection into the lumbar cord, the injection sites were mainly confined within laminae I–III of the dorsal horn. FG retrograde-labeled neuronal cell bodies were found, bilaterally with an ipsilateral predominance in the brainstem. FG retrograde-labeled neuronal cell bodies were frequently observed in the NRM, Rgcα, nucleus raphe obscurus (ROb), nucleus raphe pallidus (RPa), lateral reticular nucleus, medullary reticular field, nucleus reticularis gigantocellularis (NGC), LPGi, and randomly distributed in ventral PAG and dorsal raphe nucleus (DR) (Fig. 4A–C). In contrast to the FG-labeled neurons predominantly distributed on the side ipsilateral to the FG injection, examination of brainstem and midbrain sections labeled with GABAB R antiserum revealed that GABAB R-LI neuronal cell bodies bilaterally occupied the NRM, Rgcα, ROb, RPa and DR (Fig. 4A –C ), as well as the ventral PAG. Some of the FG-labeled neuronal cell bodies also showed GABAB R immunoreactivity (Fig. 4). Most of the FG/GABAB R double-labeled neuronal cell bodies located unilaterally on the side ipsilateral to the FG injection site were observed in the RVM, including NRM, Rgcα, NGC, ROb, RPa (Fig. 4). Only few of the FG-labeled neurons showing GABAB R immunoreactivity were found in the PAG and DR. These double-labeled neurons accounted for approximately 30–70% of the total FG-labeled neuronal population and 10–15% of the total GABAB R-LI neurons in the above mentioned structures (Table 1). Most of the GABAB R-LI Neurons in RVM and PAG Contained 5-HT The data of immunofluorescence histochemical double-staining for GABAB R and 5-HT revealed that neuronal cell bodies showing GABAB R- or 5-HT-like immunoreactivity and both GABAB Rand 5-HT-like immunoreactivities were encountered throughout the midline structures of the brainstem (Fig. 5). Co-localization of GABAB R and 5-HT was observed in most neurons showing GABAB R-like immunoreactivity in the NRM, Rgcα, NGC, ROb, RPa, DR, and the ventral PAG (Fig. 5). The rates of GABAB R/5-HT co-existing neurons to the total number of
287/544 766/987 453/527 427/502 525/615
(52.8) (77.6) (86.0) (85.1) (85.4)
GABAB R- and 5-HT-LI neurons were 82.8–94.8 and 52.8–86.0%, respectively (Table 2). DISCUSSION Origins of GABAB R-LI Terminals in the Spinal Dorsal Horn Our previous electrophysiological study has shown that in lamina II of the superficial dorsal horn, there are three possible origins of functional glutamatergic terminals: peripheral, intrinsic and supraspinal structures [30]. Theoretically, there are also three possible origins for GABAB R-LI terminals in the spinal dorsal horn: primary afferent central terminals, intrinsic dorsal horn neurons and descending pain modulation system (see [27] for review). Although the possibility of peripheral and intrinsic origins has been raised by the former works [3,5,13,24,32], our present study provides the direct morphological evidence for that. Moreover, since the spinal hemisection decreases the density of the GABAB R-LI terminals in the ipsilateral dorsal horn, the results anatomically indicate that GABAB R-LI terminals also originate from supraspinal structures. It is interesting to find GABAB R-LI neurons in the brainstem projecting to the superficial dorsal horn. The dense distribution and complex origins of GABAB R-LI structures in the dorsal horn suggest that GABAB R might play a complex role in the information transmission and/or modulation in the spinal cord. Previous studies have reported that baclofen, a selective GABAB R agonist, depresses the presynaptic excitatory neurotransmitter release in spinal dorsal horn from either primary afferent central terminals [3,32] or terminals originating from intrinsic and central structures [9,29,32]. Moreover, activation of GABAB R decreases inhibitory neurotransmitter release in the dorsal horn [9,31]. Although neurons in the brainstem sending projection fibers to the dorsal horn principally contain inhibitory neurotransmitters, our present data indicate that activation of GABAB R-LI terminals in the spinal cord might also affect the descending inhibitory neurotransmitter release through GABA by binding to these GABAB R on the descending terminals. It should be noted that, the lesion of the descending tract might affect the normal GABAB R expression on the central terminals of the primary afferents, or conversely, that rhizotomy of the dorsal roots could also affect normal GABAB R expression on the terminals of descending projections. On the other hand, both injury models might also influent the GABAB R expression of the intrinsic neurons in the dorsal horn. To interpret the present results correctly, these possibilities should not be neglected. However, these possibilities are limited because we have also observed that no detectable cytoarchitectonic changes are found after dorsal root rhizotomy and/or spinal hemisection. Significance of GABAB R-LI Neurons in the Central Endogenous Pain Control System In the present study, many GABAB R-LI neuronal cell bodies were observed in the PAG, RVM and their adjacent structures. The
506 distribution pattern of GABAB R-LI neurons in these structures is in agreement with that reported by Margeta-Mitrovic et al. [16]. Our present study further proves that some FG retrograde-labeled neurons in the PAG, RVM and brainstem raphe nuclei show GABAB R-like immunoreactivity, these results support the above mentioned hypothesis that there are some GABAB R-LI terminals descending from supraspinal structures to the dorsal horn. Many FG retrograde-labeled neurons have been found to exhibit GABAB R-like immunoreactivity in the central endogenous pain control system, which indicates that these neurons in the brainstem projecting to spinal dorsal horn are probably modulated by GABA via binding to the GABAB R in the supraspinal sites [1,2,8,18]. In the present study, although many GABAB R-LI cell bodies were found in ventral PAG, not many FG-labeled neurons were encountered in the PAG. The scarcity of FG-labeled neurons might be due to the fact that few PAG neurons project directly to the spinal dorsal horn, whereas most of them terminate in the RVM (see [7] for review). Co-localization of 5-HT and GABAB R-Like Immunoreactivities in the RVM The descending pain modulation effect is mainly mediated by 5-HT release in the dorsal horn (see [7,17] for reviews). Co-localization of GABAB R with monoamine in midbrain monoaminergic neurons has been reported [28]. It is also presumed that GABAB R-LI neurons in the RVM, probably containing 5-HT, project directly to dorsal horn [1]. Activation of GABAB R in cell bodies and dendrites elicits slow inhibitory postsynaptic potentials, whereas activation of GABAB R in presynaptic nerve terminals inhibits neurotransmitter release [3,9,29,31,32]. Since GABAB R could exert its inhibition by either presynaptic or postsynaptic mechanisms, likewise, the activation of GABAB R in RVM could either inhibit presynaptic neurotransmitter release or hyperpolarize postsynaptic neurons that project to the dorsal horn [14]. The net effect of the GABAB R activation in RVM depends on the different rate of pre- and postsynaptic GABAB R being activated. Based upon this hypothesis, Abellán et al. [2] and Tao et al. [22] have discovered that different doses of baclofen administrated to the RVM affect different ratios of presynaptic or postsynaptic GABAB R, and thus regulate the 5-HT release in the dorsal horn. The present results suggest morphologically that those 5-HT containing neurons, which project to dorsal horn, might be modulated by GABAB R. To address this issue in detail, further study will be required. Colocalization of 5-HT- and GABAB R-Like Immunoreactivities in the PAG In the present study, many GABAB R-LI neurons in the PAG were found containing 5-HT. It has been characterized that PAG neurons provides a major serotoninergic input to the RVM and terminate onto serotoninergic neurons, ON-cells, and OFF-cells (see [7] for review). Activation of opioid µ or κ receptors in the PAG disinhibits NRM-projection neurons and results in increasing 5-HT release within the dorsal horn and thus influences spinal nociception transmission [19,25]. Osborne et al. [19] demonstrated that both GABAB R agonist baclofen and opioid µ receptor agonist methionine–enkephalin produce an outward current (hyperpolarization) in PAG neurons with a similar characterization. Contrary to both presynaptic and postsynaptic actions of GABAB R activation in RVM, in PAG, opioid receptor activation predominantly inhibits synaptic transmission by reducing the probability of presynaptic neurotransmitter release, controlled by a presynaptic voltage-dependent potassium conductance [26]. Both opioid µ receptor and GABAB R are G protein-coupled inhibitory receptor in the CNS. The activation of µ receptor or GABAB R is known to inhibit calcium channel and activate potassium channel. It is
YANG ET AL. conceivable that GABAB R plays a similar role as that of opioid µ receptor in PAG with disinhibitory effect on PAG neurons, and thus affects the 5-HT release in RVM. Some studies have been performed to verify this hypothesis [1,2]. CONCLUSION Our present study provides direct morphological evidence for the hypothesis that in addition to peripheral and intrinsic origins, some GABAB R-LI terminals in the superficial dorsal horn originate from supraspinal structures; moreover, some GABAB R-LI neurons in the PAG, RVM and brainstem raphe nuclei also contain 5-HT. The results further confirm the modulatory role of GABAB R in nociceptive transmission. ACKNOWLEDGEMENTS
The authors are grateful for photographic help of Ms. Yue-Ping Yuan. This work was supported by Grants-in-Aid from the National Natural Science Foundation of China (30000052, 39970239, 30070389), Foundation for University Key Teacher by the Ministry of Education of China and the National Program of Basic Research of China (G1999054004) and Grant for Doctoral Dissertation from the Fourth Military Medical University. REFERENCES 1. Abellán, M. T.; Adell, A.; Honrubia, M. A.; Mengod, G.; Artigas, F. GABAB -R1 receptors in serotonergic neurons: Effects of baclofen on 5-HT output in rat brain. NeuroReport 11:941–945; 2000. 2. Abellán, M. T.; Jolas, T.; Aghajanian, G. K.; Artigas, F. Dual control of dorsal raphe serotonergic neurons by GABAB receptors. Electrophysiological and microdialysis studies. Synapse 36:21–34; 2000. 3. Ataka, T.; Kumamoto, E.; Shimoji, K.; Yoshimura, M. Baclofen inhibits more effectively C-afferent than Aδ-afferent glutamatergic transmission in substantia gelatinosa neurons of adult rat spinal cord slices. Pain 86:273–282; 2000. 4. Bowker, R. M.; Reddy, V. K.; Fung, S. J.; Chan, J. Y. H.; Barnes, C. D. Serotonergic and non-serotonergic raphe neurons projecting to the feline lumbar and cervical spinal cord: A quantitative horseradish peroxidase-immunocytochemical study. Neurosci. Lett. 75:31–37; 1987. 5. Charles, K. J.; Evans, M. L.; Robbins, M. J.; Calver, A. R.; Leslie, R. A.; Pangalos, M. N. Comparative immunohistochemical localization of GABAB 1a, GABAB 1b and GABAB 2 subunits in rat brain, spinal cord and dorsal root ganglion. Neuroscience 106:447–467; 2001. 6. Dirig, D. M.; Yaksh, T. L. Intrathecal baclofen and muscimol, but not midazolam, are antinociceptive using the rat-formalin model. J. Pharmacol. Exp. Ther. 275:219–227; 1995. 7. Fields, H. L.; Heinricher, M. M.; Mason, P. Neurotransmitters in nociceptive modulatory circuits. Annu. Rev. Neurosci. 14:219–245; 1991. 8. Innis, R. B.; Nestler, E. J.; Aghajanian, G. K. Evidence for G protein mediation of serotonin- and GABAB -induced hyperpolarization of rat dorsal raphe neurons. Brain Res. 459:27–36; 1988. 9. Iyadomi, M.; Iyadomi, I.; Kumamoto, E.; Tomokuni, K.; Yoshimura, M. Presynaptic inhibition by baclofen of miniature EPSCs and IPSCs in substantia gelatinosa neurons of the adult rat spinal dorsal horn. Pain 85:385–393; 2000. 10. Jones, K. A.; Borowsky, B.; Tamm, J. A.; Craig, D. A.; Durkin, M. M.; Dai, M.; Yao, W.-J.; Johnson, M.; Gunwaldsen, C.; Huang, L.-Y.; Tang, C.; Shen, Q.; Salon, J. A.; Morse, K.; Laz, T.; Smith, K. E.; Nagarathnam, D.; Noble, S. A.; Branchek, T. A.; Gerald, C. GABAB receptors function as a heteromeric assembly of the subunits GABAB R1 and GABAB R2. Nature 396:674–679; 1998. 11. Li, Y.-Q.; Wang, Z.-M.; Zheng, H.-X.; Shi, J.-W. Central origins of substance P-like immunoreactive fibers and terminals in the spinal trigeminal caudal subnucleus in the rat. Brain Res. 719:219–224; 1996.
GABAB R-LI TERMINALS IN SPINAL DORSAL HORN 12. Li, Y.-Q.; Wu, S.-X.; Li, J.-L.; Li, J.-S.; Kaneko, T.; Mizuno, N. Co-existence of calcium-binding proteins in neurons of the medullary dorsal horn of the rat. Neurosci. Lett. 286:103–106; 2000. 13. Liang, F.; Hatanaka, Y.; Saito, H.; Yamamori, T.; Hashikawa, T. Differential expression of γ -aminobutyric acid type B receptor-1a and -1b mRNA variants in GABA and non-GABAergic neurons of the rat brain. J. Comp. Neurol. 416:475–495; 2000. 14. Lin, H. H.; Dun, N. J. Post- and presynaptic GABAB receptor activation in neonatal rat rostral ventrolateral medulla neurons in vitro. Neuroscience 86:211–220; 1998. 15. Malcangio, M.; Bowery, N. G. GABA and its receptors in the spinal cord. Trends Pharmacol. Sci. 17:457–462; 1996. 16. Margeta-Mitrovic, M.; Mitrovic, I.; Riley, R. C.; Jan, L. Y.; Basbaum, A. I. Immunohistochemical localization of GABAB receptors in the rat central nervous system. J. Comp. Neurol. 405:299–321; 1999. 17. Mason, P. Central mechanisms of pain modulation. Curr. Opin. Neurobiol. 9:436–441; 1999. 18. McGowan, M. K.; Hammond, D. L. Intrathecal GABAB antagonists attenuate the antinociception produced by microinjection of l-glutamate into the ventromedial medulla of the rat. Brain Res. 607:39–46; 1993. 19. Osborne, P. B.; Vaughan, C. W.; Wilson, H. I.; Christie, M. J. Opioid inhibition of rat periaqueductal gray neurones with identified projections to rostral ventromedial medulla in vitro. J. Physiol. (London) 490:383–389; 1996. 20. Price, G. W.; Wilkin, G. P.; Tunbull, M. J.; Bowery, N. G. Are baclofen-sensitive GABAB receptors present on primary afferent terminals of the spinal cord? Nature 307:71–74; 1984. 21. Steinbusch, H. W. M. Distribution of serotonin-immunoreactivity in the central nervous system of the rat-cell bodies and terminals. Neuroscience 6:557–618; 1981. 22. Tao, R.; Ma, Z.; Auerbach, S. B. Differential regulation of 5-hydroxytryptamine release by GABAA and GABAB receptors in midbrain raphe nuclei and forebrain of rats. Br. J. Pharmacol. 119:1375–1384; 1996.
507 23. Thomas, D. A.; McGowan, M. K.; Hammond, D. L. Microinjection of baclofen in the ventromedial medulla of rats: Antinociception at low doses and hyperalgesia at high doses. J. Pharmacol. Exp. Ther. 275:274–284; 1995. 24. Towers, S.; Princivalle, A.; Billinton, A.; Edmunds, M.; Bettler, B.; Urban, L.; Castro-Lopes, J.; Bowery, N. G. GABAB receptor protein and mRNA distribution in rat spinal cord and dorsal root ganglia. Eur. J. Neurosci. 12:3201–3210; 2000. 25. Vaughan, C. W.; Connor, M.; Jennings, E. A.; Marinelli, S.; Allen, R. G.; Christie, M. J. Actions of nociceptin/orphanin FQ and other prepronociceptin products on rat rostral ventromedial medulla neurons in vitro. J. Physiol. (London) 534:849–859; 2001. 26. Vaughan, C. W.; Ingram, S. L.; Connor, M. A.; Christie, M. J. How opioids inhibit GABA-mediated neurotransmission. Nature 390:611– 614; 1997. 27. Willis, W. D.; Coggeshall, R. E. Sensory mechanisms of the spinal cord, 2nd ed. New York: Plenum Press, 1991. 28. Wirtshafter, D.; Sheppard, A. C. Localization of GABAB receptors in midbrain monoamine containing neurons in the rat. Brain Res. Bull. 56:1–5; 2001. 29. Yang, K.; Kumamoto, E.; Furue, H.; Li, Y.-Q.; Yoshimura, M. Capsaicin induces a slow inward current which is not mediated by substance P in substantia gelatinosa neurons of the rat spinal cord. Neuropharmacology 39:2185–2194; 2000. 30. Yang, K.; Li, Y.-Q. Origins of spontaneous and noxious stimuli-evoked miniature EPSCs in substantia gelatinosa. NeuroReport 12:39–42; 2001. 31. Yang, K.; Feng, Y.-P.; Li, Y.-Q. Baclofen inhibition of dorsal root-evoked inhibitory postsynaptic currents in substantia gelatinosa neurons of rat spinal cord slice. Brain Res. 900:320–323; 2001. 32. Yang, K.; Wang, D.; Li, Y.-Q. Distribution and depression of the GABAB receptor in the spinal dorsal horn of adult rat. Brain Res. Bull. 55:279–285; 2001.