Synaptic organization of GABAergic projections from the substantia nigra pars reticulata and the reticular thalamic nucleus to the parafascicular thalamic nucleus in the rat

Brain Research 957 (2002) 231–241 www.elsevier.com / locate / brainres

Research report

Synaptic organization of GABAergic projections from the substantia nigra pars reticulata and the reticular thalamic nucleus to the parafascicular thalamic nucleus in the rat Toshiko Tsumori, Shigefumi Yokota, Katsuhiko Ono, Yukihiko Yasui* Department of Anatomy (2 nd Division), Shimane Medical University, Izumo 693 -8501, Japan Accepted 29 August 2002

Abstract The ventrolateral part of the parafascicular thalamic nucleus (PF), which is considered to take part in the control mechanism of orofacial motor functions, receives projection fibers not only from the dorsolateral part of the substantia nigra pars reticulata (SNr) but also from the ventral part of the reticular thalamic nucleus (RT) [Tsumori et al., Brain Res. 858 (2000) 429]. In order to better understand the influence of these fibers upon the PF projection neurons, the morphology, synaptology and chemical nature of them were examined in the present study. After ipsilateral injections of Phaseolus vulgaris-leucoagglutinin (PHA-L) into the dorsolateral part of the SNr and biotinylated dextran amine (BDA) into the ventral part of the RT, overlapping distributions of PHA-L-labeled SNr fibers and BDA-labeled RT fibers were seen in the ventrolateral part of the PF. At the electron microscopic level, the SNr terminals made synapses predominantly with the medium to small dendrites and far less frequently with the somata and large dendrites, whereas approximately half of the RT terminals made synapses with the somata and large dendrites and the rest did with the medium to small dendrites of PF neurons. Some of single dendritic as well as single somatic profiles received convergent synaptic inputs from both sets of terminals. These terminals were packed with pleomorphic synaptic vesicles and formed symmetrical synapses. After combined injections of PHA-L into the dorsolateral part of the SNr, BDA into the ventral part of the RT and wheat germ agglutinin–horseradish peroxidase (WGA-HRP) into the ventrolateral part of the striatum or into the rostroventral part of the lateral agranular cortex, WGA-HRP-labeled neurons were embedded in the plexus of PHA-L- and BDA-labeled axon terminals within the ventrolateral part of the PF, where the PHA-L- and / or BDA-labeled terminals were in synaptic contact with single somatic and dendritic profiles of the WGA-HRP-labeled neurons. Furthermore, the SNr and RT axon terminals were revealed to be immunoreactive for g-aminobutyric acid (GABA), by using the anterograde BDA tracing technique combined with immunohistochemistry for GABA. The present data suggest that GABAergic SNr and RT fibers may exert different inhibitory influences on the PF neurons for regulating the thalamic outflow from the PF to the cerebral cortex and / or striatum in the control of orofacial movements.  2002 Elsevier Science B.V. All rights reserved. Theme: Motor systems and sensorimotor integration Topic: Thalamus Keywords: Parafascicular thalamic nucleus; Reticular thalamic nucleus; Substantia nigra pars reticulata; GABA; Rat

1. Introduction The parafascicular thalamic nucleus (PF) of the rat, which is one of the intralaminar nuclei, receives afferent

*Corresponding author. Tel.: 181-853-20-2106; fax: 181-853-202105. E-mail address: [email protected] (Y. Yasui).

fibers from the somatosensory and motor cortices [2,7], reticular thalamic nucleus (RT) [7,17,44], zona incerta [7], striatum [7], entopeduncular nucleus [7], substantia nigra pars reticulata (SNr) [44], mesencephalic reticular formation [7], pedunculopontine nucleus [7], vestibular nuclei [21,38] and deep cerebellar nuclei [7], and sends fibers to the striatum [3,9,10,16,21,43,45], RT [8] and subthalamic nucleus [10,41] as well as to the cerebral cortex [2,4,8,16,21,24,43]. On the basis of these hodological data,

0006-8993 / 02 / $ – see front matter  2002 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 02 )03554-0

232

T. Tsumori et al. / Brain Research 957 (2002) 231–241

it has been suggested that the PF of the rat plays a crucial role in motor mechanisms. In a preliminary study [43], we observed that the ventrolateral part of the PF sends projection fibers to the rostroventral part of the lateral agranular cortex (AGl), which corresponds to the orofacial motor cortex [2,28], as well as to the ventrolateral part of the striatum where cortical fibers from the orofacial motor area terminate [50]. Our recent study [44] has also shown that the ventrolateral part of the PF receives projection fibers directly or indirectly via the ventral part of the RT from the dorsolateral part of the SNr which is suggested to be implicated in orofacial behaviors [42,48,49]. On the other hand, the SNr provides g-aminobutyric acid (GABA) inhibitory inputs to the thalamus (see Gerfen and Wilson [11] for review). Neurons in the RT also utilize GABA as a neurotransmitter and exert inhibitory influence upon thalamic neurons [13,29,31]. Judging from the above, it seems likely that dual inhibitory inputs provided by the SNr and RT fibers may affect the PF neurons sending their axons to the cerebral cortex and / or striatum for regulating orofacial movements. However, there have been no studies to examine the synaptic organization between and among these SNr fibers, RT fibers and PF neurons. In the present study, we therefore first demonstrate the overlapping distribution of SNr and RT axon terminals in the PF, and then observe morphology and synaptic organization of the SNr and RT axon terminals, using double anterograde tracing. Secondly, we examine the convergent inputs from the SNr and RT onto single PF neurons projecting to the AGl or striatum, using double anterograde tracing combined with retrograde tracing. Finally, we confirm that the RT and SNr axon terminals in the PF are immunoreactive for GABA, using anterograde tracing combined with the postembedding immunogold method.

2. Materials and methods The experiments were carried out on male Wistar rats ranging in weight from 280 to 330 g. All surgical procedures were performed under general anesthesia with intraperitoneal injection of chloral hydrate (350 mg / kg).

2.1. Double anterograde labeling Ipsilateral injections of biotinylated dextran amine (BDA, 10 000 MW, Molecular Probes) into the RT and Phaseolus vulgaris-leucoagglutinin (PHA-L, Vector) into the SNr were made stereotaxically by iontophoresis in seven rats. In each rat, a single injection of BDA into the RT was made using a glass micropipette filled with a 10% BDA solution in 0.01 M phosphate buffer (PB, pH 7.3). The driving current (5–6 mA, 200 ms, 2 Hz) was delivered for 15–20 min. After BDA injection, a single injection of PHA-L into the SNr was made through a glass micropipet-

te filled with a 2.5% solution of PHA-L dissolved in 0.01 M PB (pH 7.6). The driving current (6–7 mA, 200 ms, 2 Hz) was delivered for 30–45 min. After 5–7 days of survival, the animals were reanesthetized and perfused transcardially with 150 ml of saline, followed by 500 ml of a solution composed of 4% paraformaldehyde and 0.2% glutaraldehyde in 0.1 M PB (pH 7.3), and then with 150 ml of 10% sucrose in the same buffer. The brains were removed and placed in 20% sucrose in the same buffer at 4 8C overnight. Subsequently, the brains were cut serially into frontal sections of 50 mm thickness on a freezing microtome. The sections were washed in phosphate-buffered saline (PBS), incubated in PBS containing 0.2% Triton X-100 for 3–4 h, and then incubated in PBS containing avidin–biotin–peroxidase complex (ABC, Vector; 1:100) for 1 h. The sections were washed again in PBS and incubated in 25 ml of 0.1 M PB (pH 7.3) containing 10 mg diaminobenzidine (DAB), 5 mg nickel ammonium sulfate and 10 ml of 30% H 2 O 2 . BDA-labeled axons were visualized as black reaction products. The sections were then washed in PBS and incubated overnight in PBS containing 1.5% normal goat serum, 0.2% Triton X-100 and rabbit anti-PHA-L (EY lab; 1:1000). Subsequently, the sections were washed in PBS, incubated in PBS containing biotinylated rabbit anti-goat IgG (Vector; 1:200) for 3 h, washed in PBS, incubated in PBS containing ABC complex for 1 h, washed in PBS, and then incubated in 25 ml of 0.1 M PB (pH 7.3) containing 10 mg DAB and 10 ml of 30% H 2 O 2 . PHA-L-labeled axons were visualized as brown reaction products. After several washing with PBS, the sections were mounted on gelatinized slides, air-dried and cleared in xylene. BDA- and PHA-L-labeled axons were traced by using a camera lucida attached to the microscope (Nikon, OPTIPHOTO-2). Coverslips were then removed and the sections were counterstained with 1% cresyl violet for cytoarchitectural landmarks. In the sections for electron microscopic observation, BDA was first visualized with DAB without nickel ammonium sulfate as mentioned above. After the first DAB reaction to detect BDA, the sections were incubated overnight in PBS containing 1.5% normal goat serum, 0.2% Triton X-100 and rabbit anti-PHA-L (1:1000), washed in PBS and then incubated in PBS containing biotinylated goat anti-rabbit IgG (1:200) for 3 h. After washing with PBS, silver–gold intensification of DAB reaction product of BDA was performed according to the method of Wang and Nakai [47]. Subsequently, the sections were incubated in the ABC solution, and then the second DAB reaction was done for visualization of PHAL. Specimens in which there was a good overlapping distribution of BDA- and PHA-L-labeled axons were cut out from the PF region and collected in 0.1 M PB (pH 7.3). The specimens were postfixed in a solution of 2% osmium tetroxide in the same buffer for 30 min at room temperature. After washing in distilled water, the specimens were stained en bloc with 0.5% uranyl acetate in

T. Tsumori et al. / Brain Research 957 (2002) 231–241

70% ethanol for 1 h, dehydrated in a graded series of ethanol, cleared in propylene oxide and then embedded flat in Epon. Subsequently, serial ultrathin sections were cut on an ultramicrotome, collected on collodion-coated grids and stained with lead nitrate. Finally, the sections were examined under an electron microscope (TOPCON EM002B). To obtain the size distribution of postsynaptic dendrites, the minor diameters of dendrites postsynaptic to BDA- and PHA-L-labeled terminals with synaptic membrane specializations were measured on the photomicrographs using an image analyzing system (Nikon, COSMOZONE-1S). The cross-sectional areas of these terminals were also measured in the same manner. Furthermore, the labeled terminals forming synapse were counted and classified by their origin and postsynaptic target.

2.2. Double anterograde labeling combined with retrograde transport of WGA-HRP After combined injections of BDA into the RT and PHA-L into the SNr, wheat germ agglutinin–horseradish peroxidase (WGA-HRP, Toyobo) injection was made into the AGl or into the striatum on the side ipsilateral to the sites of BDA and PHA-L injections in five rats each. Injections of BDA and PHA-L were performed as described above. Following a survival period of 3–4 days, a 5% solution of WGA-HRP dissolved in 0.1 M Tris buffer (pH 7.9) was injected through a fine glass pipette attached to a 1.0-ml microsyringe. A total volume of 0.04–0.06 ml of WGA-HRP solution was delivered over a period of 15 min. After a further survival period of 3 days, the rats were perfused transcardially with 150 ml of saline, followed by 500 ml of a solution composed of 4% paraformaldehyde and 0.3% glutaraldehyde in 0.1 M PB (pH 7.3). After perfusion, the brains were removed, dissected into 5-mm thick blocks and sectioned 50-mm thick in the frontal plane on a vibrating microtome. The sections through the thalamus were collected in 0.1 M PB and then put in a cryoprotectant solution (0.1 M PB containing 20% sucrose and 10% glycerin) for 30 min. After sinking, they were freeze–thawed in liquid nitrogen and washed in PB. For the histochemical demonstration of WGA-HRP, the sections were treated with tetramethylbenzidine (TMB) according to Mesulam [25]. The sections were then washed in 0.01 M acetate buffer (pH 4.0) and incubated in 25 ml of 0.1 M PB containing 12.5 mg DAB, 5 mg cobalt chloride, and 75 ml of 30% H 2 O 2 for 10 min at room temperature to stabilize TMB reaction product [35]. Subsequently, the sections were processed to reveal the BDA and PHA-L as described above, except that Triton X-100 was not included in the incubation solution. After visualization of BDA and PHA-L, the sections in which there was a good overlap among the distributions of the BDA- and PHA-L-labeled terminals, and WGA-HRP-labeled neurons

233

were selected, and further processed for electron microscopy as described above.

2.3. Anterograde labeling combined with immunohistochemistry for GABA Injections of BDA into the RT or into the SNr were made stereotaxically by iontophoresis in five rats or in four rats, respectively. BDA injection was made as described above. After 5–7 days survival, the animals were reanesthetized and perfused transcardially with 150 ml of saline, followed by 500 ml of a solution composed of 1% paraformaldehyde and 2% glutaraldehyde in 0.1 M PB (pH 7.3). After perfusion, vibratome sections were made by the above-mentioned procedures. Subsequently, the sections were incubated in PBS containing ABC complex for 3–4 h, washed in PBS and then incubated in 25 ml of 0.1 M PB (pH 7.3) containing 10 mg DAB and 10 ml of 30% H 2 O 2 . After visualization of BDA, the specimens were processed for electron microscopic observation as mentioned above, and further processed with the postembedding immunogold technique for revealing GABA immunoreactivity. In detail, the sections were etched in 1% periodic acid for 7 min and 2% sodium metaperiodate for 8 min, pretreated in 1% ovalbumin for 30 min, and incubated overnight at 4 8C in GABA antiserum (1:1500; Sigma) diluted in Tris–HCl buffer (pH 7.2) with 1% normal goat serum. After several washes, the sections were incubated for 1–2 h in goat anti-rabbit IgG conjugated to colloidal gold particles of 10 nm in diameter (1:10; Biocell, Cardiff) diluted in Tris– HCl buffer (pH 8.6) with 1% bovine serum albumin. Subsequently, the sections were stained with uranyl acetate and lead citrate and then examined under an electron microscope. In control sections, which were treated in the same manner except that the primary antiserum was omitted, there was no immunostaining.

3. Results

3.1. Distribution and morphology of SNr and RT terminals 3.1.1. Light microscopic observations After ipsilateral injections of BDA into the ventral part of the RT (Figs. 1A and 2A) and PHA-L into the dorsolateral part of the SNr (Figs. 1B and 2B), PHA-Llabeled SNr fibers stained brown were easily distinguished from BDA-labeled RT fibers stained dark blue to black (Fig. 1C,D). The terminal field of each set of labeled fibers in the thalamus was similar to that shown in our previous study [44], and the good overlapping distribution of PHAL- and BDA-labeled axon terminals was found in the ventrolateral part of the PF (Figs. 2C–F and 3). However, morphological characteristics of these fibers were different: the RT fibers formed a dense plexus and gave rise to many

234

T. Tsumori et al. / Brain Research 957 (2002) 231–241

Fig. 1. Photomicrographs showing the site of BDA injection into the ventral part of the RT (A) and PHA-L injection into the dorsolateral part of the SNr (B), and resulting double anterograde labeling in the ventrolateral part of the PF (C,D). Bouton-like varicosities labeled with PHA-L and -BDA are indicated by large and small arrows, respectively. Bar51 mm in (A), 0.5 mm in (B), 20 mm in (C,D). CG, central gray matter; CPu, caudate putamen; ic, internal capsule; R, red nucleus; RT, reticular thalamic nucleus; SC, superior colliculus; SNr, substantia nigra pars reticulata.

small bouton-like varicosities, whereas the SNr fibers formed a less dense plexus and their terminal boutons were generally large (Fig. 1C,D).

3.1.2. Electron microscopic observations When the ventrolateral part of the PF was examined under the electron microscope, the SNr axon terminals labeled with PHA-L were packed with the electron-dense DAB reaction product filling up the entire space between the vesicles and the mitochondria (Fig. 4A–C). On the other hand, the RT axon terminals labeled with BDA contained highly electron-dense silver–gold grains of various sizes filling up the cytoplasm except for mitochondria (Fig. 4C,D). As shown under the light microscope, the BDA-labeled SNr terminals were larger than the BDA-labeled RT terminals (see Table 1). Furthermore, the SNr terminals formed synapses predominantly with medium to small dendrites and much less frequently with somata and large dendrites, whereas one-half of the RT terminals made synapses with somata and large dendrites and another half were in contact with medium to small dendrites (see Table 1). Other morphological characteristics of the SNr and RT terminals were almost in common: both sets of labeled terminals were rich in mitochondria, were densely packed with pleomorphic synaptic vesicles, and formed symmetrical synapses, although SNr terminals containing round vesicles were occasionally found to form asymmetrical synapses. The relationship between these two sets of labeled terminals and their target structures in the PF is summarized in Table 2. Most of the postsynaptic dendrites received either SNr or RT terminals, although in some cases single dendritic profiles were seen to receive both SNr and RT terminals. The majority of the postsynaptic somata also received either SNr or RT terminals, although approximately one-sixth of the somata observed here was found to receive convergent synaptic inputs from both SNr and RT terminals. In case of the somatic profiles receiving SNr and / or RT terminals, they were frequently in synaptic

contact with more than two labeled terminals of each set (see Table 2).

3.2. Double anterograde labeling combined with retrograde transport of WGA-HRP 3.2.1. Each labeled structure In the rats injected with BDA into the RT and PHA-L into the SNr, WGA-HRP was injected into the rostroventral part of the AGl or into the ventrolateral part of the striatum. The distribution pattern and morphology of each set of labeled terminals were identical with those described above. On the other hand, many labeled neurons were distributed in the ventrolateral part of the PF after the striatal injections, whereas the cortical injections resulted in moderate numbers of labeled neurons in the ventrolateral part of the PF with many labeled neurons in the ventral part of the posterior thalamic nuclear group just lateral to the PF. Within the PF, labeled RT and SNr terminal varicosities were seen in contiguity to the WGA-HRPlabeled neurons in either case. At the electron microscopic level, BDA- and PHA-Llabeled terminals were observed as above. On the other hand, WGA-HRP-labeled neurons were identified as those with TMB reaction product; this product was recognized as electron-dense granules or needle-like crystals, which were randomly dispersed throughout the soma and dendrites (Figs. 5A, and 6A,D). The morphological characteristics of the AGI-projecting neurons are similar to those of the striatum-projecting neurons: they were fusiform, ovoid, or polygonal in shape with a major diameter ranging from 15 to 20 mm and their perikarya contained abundant organelles and a large nucleus with invaginations of its membrane (Figs. 5A and 6A). 3.2.2. Convergent inputs from the RT and SNr onto single PF projection neurons The thalamocortical and thalamostriatal neurons labeled with WGA-HRP were frequently observed to receive synaptic contacts from the RT terminals labeled with BDA

T. Tsumori et al. / Brain Research 957 (2002) 231–241

235

injected into the dorsolateral part of the SNr or into the ventral part of the RT, the sections including the ventrolateral part of the PF were processed to reveal GABA immunoreactivity using the postembedding immunogold method. Structures were determined to be GABA-immunoreactive if the density of immunogold particles on them was at least five times higher than that on adjacent nonimmunoreactive structures. As far as we observed, the BDA-labeled RT terminals were identified to be immunoreactive for GABA (Fig. 7A). Almost all the BDA-labeled SNr terminals were also immunoreactive for GABA (Fig. 7B), although those without immunoreactivity for GABA were occasionally found (Fig. 7C). In the both cases, the BDA-labeled and GABA-positive terminals as well as non-BDA-labeled and GABA-positive ones contained pleomorphic vesicles and formed symmetrical synaptic contact. On the other hand, the BDA-labeled SNr terminals without GABA-like immunoreactivity contained round vesicles and formed asymmetrical synapses. In addition, none of the dendrites and somata observed here were immunoreactive for GABA.

4. Discussion

Fig. 2. Line drawings showing the sites of BDA injection into the ventral part of the RT (shaded area in A) and PHA-L injection into the dorsolateral part of the SNr (shaded area in B), and resulting distributions of BDA- and PHA-L-labeled terminals (black and red dots, respectively) in the ventrolateral part of the PF (C–F, rostral to caudal). Each PF region in the thalamic planes (c–f) is enlarged in (C)–(F), respectively. APTD, dorsal part of the anterior pretectal nucleus; CM, central medial nucleus; f, fornix; fr, fasciculus retroflexus; GP, globus pallidus; LH, lateral habenular nucleus; LP, lateral posterior nucleus; MD, mediodorsal nucleus; MG, medial geniculate nucleus; ml, medial lemniscus; mt, mamilothalamic tract; opt, optic tract; PF, parafascicular nucleus; Po, posterior nuclear group; PrC, precommissural nucleus; PV, paraventricular thalamic nucleus; RI rostral interstitial nucleus; SPF, subparafascicular nucleus; VPMmc, magnocellular division of the ventral posteromedial nucleus; VPMpc, parvicellular division of the ventral posteromedial nucleus. Other abbreviations are as in Fig. 1.

or the SNr terminals labeled with PHA-L. Some of the identified thalamocortical (Fig. 5A–D) and thalamostriatal (Fig. 6A–C) neurons received convergent synaptic inputs from both sets of terminals on their somata and proximal dendrites. In a few examples, single medium to small dendrites of the identified neurons were found to receive convergent synaptic inputs from both sets of terminals (Fig. 6D–F), although many of the dendrites received either RT or SNr terminals.

3.3. GABA immunoreactivity of SNr and RT terminals After visualization of labeled terminals with BDA

In the present study, we confirmed our previous finding that the terminal field of projection fibers from the dorsolateral part of the SNr is overlapped with that from the ventral part of the RT in the ventrolateral part of the PF. We further provided new data as follows: (1) both of the GABA-like immunoreactive SNr and RT axon terminals form symmetrical synaptic contacts with soma and dendrites of PF neurons; (2) the SNr and RT axon terminals make convergent synaptic contacts with single PF neurons, some of which project to the rostroventral part of the AGl or to the ventrolateral part of the striatum. Some of these results were obtained by new technical procedures (see below). In order to investigate the synaptology of two sets of axon terminals of different origin, it is necessary to carry out the experiments using the double anterograde labeling at the electron microscopic level. In the present study, we utilized a new double anterograde labeling technique with BDA and PHA-L for the electron microscopic observation. Alisky and Tolbert [1] used BDA and cholera toxin B subunit as tracers for the double anterograde labeling and employed DAB and DAB enhanced with cobalt (Co-DAB) as chromogens for the double peroxidase reactions. They indicated that the DAB reaction product was easily distinguished from the Co-DAB reaction product, but only at the light microscopic level. In recent double anterograde labeling studies [5,6,39,46], PHA-L and biocytin were used as anterograde tracers, and a new double peroxidase method using two chromogens, DAB and benzidine dihydrochloride (BDHC) was adopted in order to distinguish their reaction products at the electron microscopic

236

T. Tsumori et al. / Brain Research 957 (2002) 231–241

Fig. 3. Line drawing showing a overlapping distribution of BDA- and PHA-L-labeled fibers with bouton-like varicosities (black and red, respectively) is illustrated at a higher magnification. An area enclosed by a small rectangle in the ventrolateral part of the PF indicates a region enlarged. Abbreviations are as in Fig. 2.

level. However, it becomes difficult to obtain and use BDHC because of its carcinogenicity. More recently, Lanciego et al. [22] have introduced the double anterograde axonal tracing with BDA and PHA-L in the light microscopic study. Additionally, in our previous study [44] we used silver–gold intensification of DAB reaction product for visualization of transported BDA at the electron microscopic level and obtained good results. In the present study, therefore, BDA-labeled fibers from the RT and PHA-L-labeled fibers from the SNr were visualized with DAB reaction product intensified with silver and gold, and with DAB reaction product, respectively. This technique enabled us to distinguish the BDA-labeled fibers from the PHA-L-labeled fibers and to examine the existence of convergent inputs from the SNr and RT onto single PF neurons. Furthermore, the combined use of this double anterograde labeling and retrograde labeling with WGAHRP allowed us to examine whether single PF neurons projecting to the AGl or to the striatum received convergent inputs from the SNr and RT. Nigrothalamic projections arising from the SNr have been reported in various studies (see review by Parent [33]), and the SNr terminals have been examined electron microscopically in several thalamic nuclei and species [18–20,26,36]. According to Sakai et al. [36], in the rat a large proportion of the SNr terminals form axosomatic synapses in the ventral anterior–ventral lateral complex (VAL), whereas the SNr terminals in the ventromedial nucleus (VM) are rarely in contact with somata and about two-thirds of them make synapses with thin dendrites. In

the monkey, the SNr terminals tend to target the somata and proximal dendrites of thalamocortical projection neurons in the magnocellular part of the ventral anterior nucleus [18]. The SNr terminals of the cat have been reported to make synaptic contacts mainly with dendrites [26] or to be frequently in contact with secondary and tertiary dendrites of VM neurons [20]. In the present study, we examined for the first time the SNr terminals in the PF under the electron microscope, and revealed that these terminals formed synapses predominantly with medium to small dendrites and far less frequently with somata and large dendrites. Thus, it seems likely that the pattern of distribution of the SNr terminals in the PF of the rat is almost identical with that in the VM of the rat and cat. With respect to the basic morphological feature, most of the previous studies indicated that SNr terminals contained pleomorphic synaptic vesicles and formed symmetrical synapses, although Sakai et al. [36] described that the synapses between the SNr terminals and the VAL and VM neurons were of asymmetrical or intermediate type. In the present study, we observed that a large number of SNr terminal boutons contain pleomorphic synaptic vesicles and form symmetrical synapses, although only a small number of SNr terminals containing round vesicles form asymmetrical synapses. Projections from the RT to the thalamic nuclei have been examined and well known to terminate in all thalamic nuclei (see review by Jones [15]). There have been no previous studies to examine the ultrastructural features of the RT terminals in the PF, although those in different

T. Tsumori et al. / Brain Research 957 (2002) 231–241

237

Fig. 4. Electron photomicrographs of BDA- and PHA-L-labeled terminals making symmetrical synaptic contacts with PF neurons after combined injections of BDA into the ventral part of the RT and PHA-L into the dorsolateral part of the SNr. (A) A PHA-L-labeled SNr terminal (t1) makes synaptic contact with a somatic profile (s). (B) A PHA-L-labeled SNr terminal (t2) makes synaptic contact with a dendritic profile (d). (C) A BDA-labeled RT terminal (t3) and a PHA-L-labeled SNr terminal (t4) form synapses with a somatic profile (s). The BDA-labeled terminal (t3) contains highly electron-dense silver–gold grains and is easily distinguished from the PHA-L-labeled terminal (t4). (D) Two BDA-labeled RT terminals (t5, t6) are in synaptic contact with a dendritic profile (d). Arrowheads indicate the synaptic sites. Bar50.5 mm.

Table 1 Cross-sectional area and postsynaptic distribution of the SNr and RT terminals in the PF Source of terminals

Mean area 6SD (mm 2 )

SNr

1.0960.47 (n5101) 0.6060.30 (n5120)

RT

Postsynaptic distribution Soma (%)

Dendrites .2 mm (%)

Dendrites 1–2 mm (%)

Dendrites ,1 mm (%)

12 (11) 58 (41)

3 (3) 9 (6)

65 (58) 50 (35)

32 (29) 26 (18)

Table 2 Interrelationships of the PHA-L-labeled SNr terminals and BDA-labeled RT terminals, and their target structures in the PF Postsynaptic target structure Soma (n5123) Source of terminals

SNr RT SNr / RT

Dendrites (n5285) Number of labeled terminals contacted

Source of terminals

1

2

3%

24 34 –

9 23 13

2 11 7

SNr RT SNr / RT

Number of labeled terminals contacted 1

2

114 152 –

5 9 5

238

T. Tsumori et al. / Brain Research 957 (2002) 231–241

Fig. 5. Electron photomicrographs of a WGA-HRP-labeled PF neuron, which is in contact with BDA- and PHA-L-labeled terminals after combined injections of BDA into the ventral part of the RT, PHA-L into the dorsolateral part of the SNr and WGA-HRP into the rostroventral part of the AGl. (A) The soma (s) and proximal dendrite (d) containing TMB reaction product are surrounded by six labeled terminals, five of which are labeled with BDA (t1, t2, t4, t5, t6) and one with PHA-L (t3). (B–D) Three terminals (t2, t3, t5) in (A) are enlarged at a high magnification in (B–D), respectively (B is rotated clockwise direction through 908). These form symmetrical synaptic contacts (arrowheads). Bar52 mm in (A), 0.5 mm in (B–D).

thalamic nuclei have been examined [14,23,34]. Ilinsky et al. [14] examined projections from the anterior pole of the RT to the motor thalamic nuclei of the monkey. They indicate that most of the RT terminals make contacts on projection and local circuit neurons and target mainly their distal dendrites. In the ventroposterior nucleus of the cat, the RT fibers make synapses predominantly with distal dendrites of projection neurons, and additionally with dendrites of interneurons or with somata [23]. The results of the present study, however, indicate in the rat that one-half of the RT terminals make synapses with somata and proximal dendrites and another half are in contact with mid to distal dendrites of PF neurons. Other ultrastructural features of the RT terminal boutons are identical with those described earlier in different thalamic nuclei and species: they contain pleomorphic synaptic vesicles and form symmetrical synapses. It has been widely considered that the nigrothalamic projections are GABAergic (see review by Gerfen and Wilson [12]). To our knowledge, however, there have been no studies to reveal it at the electron microscopic level. Here we observed electron microscopically that most of the SNr terminals in the PF are immunoreactive for GABA and occasionally found those without GABA-like immunoreactivity, which contained round vesicles and formed asymmetrical synapses. These non-GABAergic SNr terminals might be cholinergic [11,30]. It is also well known that the RT exerts a GABAergic inhibitory action upon thalamic neurons [13,29,31]. At the electron microscopic level, axon terminals from the RT have been reported to display immunoreactivity for GABA in the motor thalamic nuclei of the monkey [14] and in the ventroposterior thalamic nuclei of the cat [23]. The present results also

provided morphological evidence indicating that the RT terminals in the PF of the rat are GABAergic. In the rodent, the dorsal thalamus has been suggested to lack interneurons [37,40], or to contain few GABAergic interneurons [27,32]. In the present study, GABA-immunoreactive somata and dendrites were not found in the PF, suggesting that there are no GABAergic interneurons in the PF. It is, therefore, likely that the PF projection neurons receive GABAergic inhibitory inputs directly from both the SNr and the RT. Furthermore, here we observed convergent inputs from the SNr and RT onto single neurons in the ventrolateral part of the PF, some of which projected to the AGl or to the striatum. Deschenes et al. [8] demonstrated that single neurons in the PF innervate both the motor cortex and the striatum by way of axon collaterals. Therefore, there is a possibility that some of the single PF neurons receiving these two sets of inhibitory inputs send their axon collaterals to the AGl and striatum. As mentioned above, however, the patterns of innervation of PF neurons by the SNr and RT fibers are different: the SNr terminals have a tendency to terminate in more distal region of the PF neurons, whereas the RT terminals are widely distributed along the somata and dendrites. Considering the differential distributions of the postsynaptic targets of these two sets of terminals, it seems likely that the inputs from the SNr and the RT may exert different inhibitory effects on the PF neurons. The RT terminal boutons, approximately half of which make synapses with somata and proximal dendrites, are likely to exert a more powerful inhibitory influence on neurons in the PF than the SNr terminal boutons making synapses mainly with mid to distal dendrites. In addition, our recent study [44] demonstrated that the ventrolateral part of the PF receives not

T. Tsumori et al. / Brain Research 957 (2002) 231–241

239

Fig. 6. Electron photomicrographs of a WGA-HRP-labeled PF neuron, which is in contact with BDA- and PHA-L-labeled terminals after combined injections of BDA into the ventral part of the RT, PHA-L into the dorsolateral part of the SNr and WGA-HRP into the ventrolateral part of the striatum. (A) A somatic profile (s) containing TMB reaction product is contacted by four labeled terminals, one of which is labeled with BDA (t1) and three with PHA-L (t2, t3, t4). (B) A large, elongated t1 terminal is shown at a high magnification. (C) An enlarged view of t3 terminal. (D) A medium-sized dendritic profile (d) containing TMB reaction product is surrounded by both BDA- (t5) and PHA-L- (t6) labeled terminals. (E) An enlarged view of t5 terminal. (F) An enlarged view of t6 terminal. All the synapses indicated above are of symmetrical type (arrowheads). Bar52 mm in (A,D); 0.5 mm in (B,C,E,F).

only direct afferents but also indirect afferents relayed by the ventral part of the RT from the dorsolateral part of the SNr, and suggested that the former and the latter might exert inhibitory and disinhibitory influence on the PF neurons, respectively. In that communication, we also suggested that PF neurons receiving SNr fibers might be different from those receiving axons of SNr-recipient RT neurons because it is unlikely that single PF neurons are under the inhibitory and disinhibitory influences of the SNr at the same time. From this viewpoint, one may say that the convergent inputs on single PF neurons originate from SNr neurons and non-SNr-recipient RT neurons. However, it remains uncertain how these two sets of inhibitory inputs interact for the control of activity of the PF neurons. The projections from the ventrolateral part of the PF to the rostroventral part of the AGl as well as to the ventrolateral part of the striatum are considered to take part

in the control mechanism of orofacial motor functions (for Refs. see Section 1). It is, therefore, suggested that the GABAergic SNr-PF and RT-PF pathways revealed here may exert different inhibitory influences on the PF neurons for regulating the thalamic output stream from the PF to the cerebral cortex and / or striatum in the control of orofacial movements.

Acknowledgements The authors are grateful for photographic help of Mr. Makoto Oshita and Mr. Tsunao Yoneyama. This study was supported in part by Grant-in-Aid for Scientific Research (C) (11680740, 12680732, 13680822) from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

240

T. Tsumori et al. / Brain Research 957 (2002) 231–241

Fig. 7. Electron photomicrographs of BDA-labeled terminals with GABA immunoreactivity in the PF after BDA injection into the ventral part of the RT (A) or into the dorsolateral part of the SNr (B,C). (A) A BDA-labeled RT terminal (t1) showing immunoreactivity for GABA forms a symmetrical synaptic contact (arrowhead) with a somatic profile (s). A non-BDA-labeled terminal with GABA immunoreactivity (t2) is also in contiguity with the soma. (B) A large, elongated BDA-labeled SNr terminal (t3) showing immunoreactivity for GABA forms symmetrical synaptic contacts (arrowheads) with a dendritic profile (d). Note the densely packed pleomorphic vesicles within the terminal. (C) A BDA-labeled SNr terminal (t4) without immunoreactivity for GABA makes an asymmetrical synapse (double arrowhead) with a small dendrite (d). A non-BDA-labeled terminal (t5) showing immunoreactivity for GABA makes a symmetrical synapse (arrowhead) with another dendrite. Bar50.5 mm in (A)–(C).

References [1] J.M. Alisky, D.L. Tolbert, Differential labeling of converging afferent pathways using biotinylated dextran amine and cholera toxin subunit B, J. Neurosci. Methods 52 (1994) 143–148. [2] L.D. Aldes, Thalamic connectivity of rat somatic motor cortex, Brain Res. Bull. 20 (1988) 333–348. [3] H.W. Berendse, H.J. Groenewegen, Organization of the thalamostriatal projections in the rat, with special emphasis on the ventral striatum, J. Comp. Neurol. 299 (1990) 187–228. [4] H.W. Berendse, H.J. Groenewegen, Restricted cortical termination fields of the midline and intralaminar thalamic nuclei in the rat, Neuroscience 42 (1991) 73–102. [5] M.D. Bevan, J.P. Bolam, A.R. Crossman, Convergent synaptic input from the neostriatum and the subthalamus onto identified nigrothalamic neurons in the rat, Eur. J. Neurosci. 6 (1994) 320–334. [6] J.P. Bolam, Y. Smith, The striatum and the globus pallidus send convergent synaptic inputs onto single cells in the entopeduncular nucleus of the rat: a double anterograde labelling study combined with postembedding immunocytochemistry for GABA, J. Com. Neurol. 321 (1992) 456–476. [7] J. Cornwall, O.T. Phillipson, Afferent projections to the parafascicular thalamic nucleus of the rat, as shown by the retrograde transport of wheat germ agglutinin, Brain Res. Bull. 20 (1988) 139–150. [8] M. Deschenes, J. Bourassa, V.D. Doan, A. Parent, A single-cell study of the axonal projections arising from the posterior intralaminar thalamic nuclei in the rat, Eur. J. Neurosci. 8 (1996) 329–343. [9] M.E. Erro, J.L. Lanciego, J.M. Gimenez-Amaya, Re-examination of

[10]

[11]

[12] [13]

[14]

[15] [16]

[17]

the thalamostriatal projections in the rat with retrograde tracers, Neurosci. Res. 42 (2002) 45–55. J. Feger, M. Bevan, A.R. Crossman, The projections from the parafascicular thalamic nucleus to the subthalamic nucleus and the striatum arise from separate neuronal populations: a comparison with the corticostriatal and corticosubthalamic efferents in a retrograde fluorescent double-labelling study, Neuroscience 60 (1994) 125–132. C.R. Gerfen, C.J. Wilson, The basal ganglia, in: L.W. Swanson, A. Bjorklund, T. Hokfelt (Eds.), Integrated System of the CNS, Part III, Handbook of Chemical Neuroanatomy, Vol. 12, Elsevier, Amsterdam, 1996, pp. 371–468. E. Gould, L.L. Butcher, Cholinergic neurons in the rat substantia nigra, Neurosci. Lett. 63 (1986) 315–319. C.R. Houser, J.E. Vaughn, R.P. Barber, E. Roberts, GABA neurons are the major cell type of the nucleus reticularis thalami, Brain Res. 200 (1980) 341–354. I.A. Ilinsky, A.V. Ambardekar, K. Kultas-Ilinsky, Organization of projections from the anterior pole of the nucleus reticularis thalami (NRT) to subdivisions of the motor thalamus: light and electron microscopic studies in the rhesus monkey, J. Comp. Neurol. 409 (1999) 369–384. E.G. Jones, Some aspects of the organization of the thalamic reticular complex, J. Comp. Neurol. 162 (1975) 285–308. E.G. Jones, R.Y. Leavitt, Retrograde axonal transport and the demonstration of non-specific projections to the cerebral cortex and striatum from thalamic intralaminar nuclei in the rat, cat and monkey, J. Comp. Neurol. 154 (1974) 349–378. C. Kolmac, J. Mitrofanis, Organization of the reticular thalamic

T. Tsumori et al. / Brain Research 957 (2002) 231–241

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33] [34]

projection to the intralaminar and midline nuclei in rats, J. Comp. Neurol. 377 (1997) 165–178. K. Kultas-Ilinsky, I.A. Ilinsky, Fine structure of the magnocellular subdivision of the ventral anterior thalamic nucleus (VAmc) of Macaca mulatta: II. Organization of nigrothalamic afferents as revealed with EM autoradiography, J. Comp. Neurol. 294 (1990) 479–489. K. Kultas-Ilinsky, I.A. Ilinsky, L.C. Massopust, P.A. Young, K.R. Smith, Nigrothalamic pathway in the cat demonstrated by autoradiography and electron microscopy, Exp. Brain Res. 33 (1978) 481–492. K. Kultas-Ilinsky, I. Ilinsky, S. Warton, K.R. Smith, Fine structure of nigral and pallidal afferents in the thalamus: an EM autoradiography study in the cat, J. Comp. Neurol. 216 (1983) 390–405. H. Lai, T. Tsumori, T. Shiroyama, S. Yokota, K. Nakano, Y. Yasui, Morphological evidence for a vestibulo-thalamo-striatal pathway via the parafascicular nucleus in the rat, Brain Res. 872 (2000) 208– 214. J.L. Lanciego, F.G. Wouterlood, E. Erro, J.M. Gimenez-Amaya, Multiple axonal tracing: simultaneous detection of three tracers in the same section, Histochem. Cell Biol. 110 (1998) 509–515. X.-B. Liu, R. Warren, E. Jones, Synaptic distribution of afferents from reticular nucleus in ventroposterior nucleus of cat thalamus, J. Comp. Neurol. 352 (1995) 187–202. G. Marini, L. Pianca, G. Tredici, Thalamocortical projection from the parafascicular nucleus to layer V pyramidal cells in frontal and cingulate areas of the rat, Neurosci. Lett. 203 (1996) 81–84. M.M. Mesulam, Tetramethyl benzidine for horseradish peroxidase neurohistochemistry: a non-carcinogenic blue reaction product with superior sensitivity for visualizing neural afferents and efferents, J. Histochem. Cytochem. 26 (1978) 106–117. T. Moriizumi, Y. Nakamura, H. Tokuno, M. Kudo, Y. Kitao, Convergence of afferent fibers from the entopeduncular nucleus and the substantia nigra pars reticulata onto single neurons in the ventromedial thalamic nucleus: an electron microscope study in the cat, Neurosci. Lett. 95 (1988) 125–129. E. Mugnaini, W.H. Oertel, An atlas of the distribution of GABAergic neurons and terminals in the rat CNS as revealed by GAD immunohistochemistry, in: A. Bjorklund, T. Hokfelt (Eds.), GABA and Neuropeptides in the CNS, Part I, Handbook of Chemical Neuroanatomy, Vol. 4, Elsevier, Amsterdam, 1985, pp. 436–608. E.J. Neafsey, E.L. Bold, G. Haas, K.M. Hurley-Gius, G. Quirk, C.F. Sievert, R.R. Terreberry, The organization of the rat motor cortex: a microstimulation mapping study, Brain Res. Rev. 11 (1986) 77–96. W.H. Oertel, A.M. Graybiel, E. Mugnaini, R.P. Elde, D.E. Schmechel, I.J. Kopin, Coexistence of glutamic acid decarboxylase and somatostatin-like immunoreactivity in neurons of the feline nucleus reticularis thalami, J. Neurosci. 3 (1983) 1322–1332. J.D. Oh, N.J. Woolf, A. Roghani, R.H. Edwards, L.L. Butcher, Cholinergic neurons in the rat central nervous system demonstrated by in situ hybridization of choline acetyltransferase mRNA, Neuroscience 47 (1992) 807–822. P.T. Ohara, A.R. Liebermann, The thalamic reticular nucleus of the adult rat: Experimental anatomical studies, J. Neurocytol. 14 (1985) 365–411. O.P. Ottersen, J. Storm-Mathisen, GABA-containing neurons in the thalamus and pretectum of the rodent: an immunocytochemical study, Anat. Embryol. 170 (1984) 197–207. A. Parent, in: Comparative Neurobiology of the Basal Ganglia, Wiley, New York, 1986. M. Peschanski, H.J. Ralston, F. Roudier, Reticular thalami afferents to the ventrobasal complex of the rat thalamus: an electron microscope study, Brain Res. 270 (1983) 325–329.

241

[35] D.B. Rye, C.B. Saper, B.H. Wainer, Stabilization of the tetramethylbenzidine (TMB) reaction product, J. Histochem. Cytochem. 32 (1984) 1145–1153. [36] S.T. Sakai, I. Grofova, K. Bruce, Nigrothalamic projections and nigrothalamocortical pathway to the medial agranular cortex in the rat: single- and double-labeling light and electron microscopic studies, J. Comp. Neurol. 391 (1998) 506–525. [37] S.F. Sawyer, M.E. Martone, P.M. Groves, A GABA immunocytochemical study of rat motor thalamus: light and electron microscopic observation, Neuroscience 42 (1991) 103–124. [38] T. Shiroyama, T. Kayahara, Y. Yasui, J. Nomura, K. Nakano, Projections of the vestibular nuclei to the thalamus in the rat: a Phaseolus vulgaris leucoagglutinin study, J. Comp. Neurol. 407 (1999) 318–332. [39] Y. Smith, J.P. Bolam, Convergence of synaptic inputs from the striatum and the globus pallidus onto identified nigrocollicular cells in the rat: a double anterograde labelling study, Neuroscience 44 (1991) 45–73. [40] R. Spreafico, C. Frassoni, M.C. Regondi, P. Arcelli, S. DeBiasi, Interneurons in the mammalian thalamus: a marker of species, in: D. Minciacchi, M. Morinari, G. macchi, E. Jones (Eds.), Thalamic Networks for Relay and Modulation, Pergamon, Oxford, 1993, pp. 17–28. [41] T. Sugimoto, T. Hattori, N. Mizuno, K. Itoh, M. Sato, Direct projections from the centre median-parafascicular complex to the subthalamic nucleus in the cat and rat, J. Comp. Neurol. 214 (1983) 209–216. [42] T. Tsumori, Y. Yasui, Organization of the nigro-tecto-bulbar pathway to the parvicellular reticular formation: a light- and electron-microscopic study in the rat, Exp. Brain Res. 116 (1997) 341–350. [43] T. Tsumori, K. Ono, S. Yokota, T. Kishi, Y. Yasui, Projections from the substantia nigra pars reticulata to the cortex and the striatum relayed by the parafascicular thalamic nucleus in the rat, Neurosci. Res. Suppl. 22 (1998) S172. [44] T. Tsumori, S. Yokota, H. Lai, Y. Yasui, Monosynaptic and disynaptic projections from the substantia nigra pars reticulata to the parafascicular thalamic nucleus in the rat, Brain Res. 858 (2000) 429–435. [45] J.G. Veening, F.M. Cornelissen, P.A.J.M. Lieven, The topical organization of the afferents to the caudatoputamen of the rat. A horseradish peroxidase study, Neuroscience 5 (1980) 1253–1268. [46] M. von Krosigk, Y. Smith, J.P. Bolam, A.D. Smith, Synaptic organization of GABAergic inputs from the striatum and the globus pallidus onto neurons in the substantia nigra and retrorubral field which project to the medullary reticular formation, Neuroscience 50 (1992) 531–549. [47] Q.-P. Wang, Y. Nakai, Double preembedding immunostaining for the study by electron microscopy of synaptic relationships between immunocytochemically identified neurons, Neurosci. Prot. 96-05005 (1996) 1–13. [48] Y. Yasui, K. Nakano, Y. Nakagawa, T. Kayahara, T. Shiroyama, N. Mizuno, Non-dopaminergic neurons in the substantia nigra project to the reticular formation around the trigeminal motor nucleus in the rat, Brain Res. 585 (1992) 361–366. [49] Y. Yasui, T. Tsumori, K. Ono, T. Kishi, Nigral axon terminals are in contact with parvicellular reticular neurons which project to the motor trigeminal nucleus in the rat, Brain Res. 775 (1997) 219–224. [50] G. Zhang, K. Sasamoto, Projections of two separate cortical areas for rhythmical jaw movements in the rat, Brain Res. Bull. 24 (1990) 221–230.