Distribution of glycine-immunoreactive cell bodies and fibers in the rat brain

~pergamon

Neuroscience Vol. 75, No.3, pp. 737~755, 1996 Copyright © 1996 IBRO. Published by Elsevier Science Ltd Printed in Great Britain

PH: S0306-4522(96)00278-3

0306--4522/96 $15.00+0.00

DISTRIBUTION OF GLYCINE-IMMUNOREACTIVE CELL BODIES AND FIBERS IN THE RAT BRAIN C. RAMPON,* P. H. LUPPI, P. FORT, C. PEYRON and M. JOUVET Laboratoire de Medecine Experimentale, INSERM U52, CNRS, Faculte de Medecine, 8 Avenue Rockefeller, 69373, Lyon, cedex 08 ERS5645, France Abstract-To localize glycinergic cell bodies and fibers in the rat brain, we developed a sensItive immunohistochemical method combining the use of specific glycine antibodies (Campistron G. et al. (1986) Brain Res. 376, 400-405; Wenthold R. J. et al. (1987) Neuroscience 22, 897-912) with the streptavidin-horseradish peroxidase technique and 3,3'-diaminobenzidine'4HCI-nickel intensification. We confirmed the presence of numerous glycine-immunoreactive cell bodies and fibers in the cochlear nuclei, superior olivary complex, nucleus of the trapezoid body, cerebellar cortex, deep cerebellar nuclei and area postrema. For the first time in rats, we described a large to very large number of cell bodies in the medial vestibular ventral part, prepositus hypoglossal, gracile, raphe magnus and sensory trigeminal nuclei. A large number of cells was also observed in the oral and caudal pontine, parvocellular, parvocellular pars alpha, gigantocellular and gigantocellular pars alpha reticular nuclei. In addition, glycine-immunoreactive cells were seen in the ambiguus and subtrigeminal nuclei, the lateral habenula and the subfomical organ. We also provide the first evidence in rats for a very large number of fibers in the trigeminal, facial, ambiguus and hypoglossal motor nuclei, all nuclei of the medullary and pontine reticular formation, and the raphe and trigeminal sensory nuclei. We further revealed the presence of a substantial number of fibers in regions where glycine was not considered as a main inhibitory neurotransmitter, such as the pontine nuclei, the periaqueductal gray, the mesencephalic reticular formation, the anterior pretectal nucleus, the intralaminar thalamic nuclei, the zona incerta, the fields of Forel, the parvocellular parts of the paraventricular nucleus, the posterior hypothalamic areas, the anterior hypothalamic area, and the lateral and medial preoptic areas. These results indicate that, in contrast to previous statements, glycine may be an essential inhibitory neurotransmitter not only in the lower brainstem and spinal cord, but also in the upper brainstem and the forebrain. Copyright © 1996 IBRO. Published by Elsevier Science Ltd. Key words: glycine, immunohistochemistry, hypothalamus, sleep, periaqueductal gray, brainstem.

The first evidence that glycine acts as a strong inhibitory neurotransmitter was provided 30 years ago by the findings that glycine has a regional distribution in the cat spinal cord 5 and an inhibitory action on spinal motoneurons. 76 Subsequent data have clearly confirmed this role in the spinal cord (review in Ref. 4). It has also been shown that glycine is used as an inhibitory neurotransmitter in the lower brainstem?9 Indeed, in rats, cats and guinea-pigs, with specific antisera that recognize glycine in glutaraldehyde-fixed tissue, I 1.61 a large number of glycine-immunoreactive (IR) cell bodies and fibers has been reported in the cochlear nuclei, superior olivary complex and medial nucleus of the trapezoid body,3,26,36,54,55,73 in the cerebellar cortex,21.47 deep cerebellar nuclei l6 ,21 and the area postrema. 20,71 In cats only, a large number of glycine-IR cells and fibers has been further described in the vestibular,

prepositus hypoglossal, sensory trigeminal, and medullary and pontine reticular nuclei.20.21.6o.72 Moreover, a large number of glycine-IR fibers has been also reported in the cranial motor nuclei, locus coeruleus complex, raphe nuclei, pedunculopontine and laterodorsal tegmental nuclei. 20 In rats, cells and fibers have been briefly noted or not detected in these nuclei. I 1,51 Therefore, because the rat is the animal of choice for anatomical and physiological studies, we found it essential to investigate the distribution of glycine-IR cells and fibers in its entire brain. For this purpose, we developed a very sensitive immunohistochemical method combining the use of two highly specific glycine antisera11,73 with streptavidin-horseradish peroxidase (HRP) and 3,3' -diaminobenzidine'4HCl-nickel intensification. EXPERIMENTAL PROCEDURES

Perfusion

Adult male Sprague-Dawley rats (IFFA-Credo, France) weighing 280-320 g were deeply anesthetized with sodium pentobarbital (40 mg/kg, i.p.) and perfused through the ascending aorta, initially with 200 ml of Ringer's lactate solution containing 0.1% heparine, followed by 500 ml of a cold fixative in 0.1 M phosphate buffer (PB; pH 7.4)

*To whom correspondence should be addressed. Abbreviations: HRP, horseradish peroxidase; IR, immunoreactive; PB, phosphate buffer; PBST, phosphatebuffered saline containing 0.3% Triton X-lOO; PBST-Az, phosphate-buffered saline containing 0.3% Triton X-IOO and 0.1% azide. 737

738

C. Rampon et al. Table I. Relative distribution of glycine immunoreactivity in the rat brain

Structure

Medullary level Spinal trigeminal nucleus, interpolar part Spinal trigeminal nucleus, caudal part Spinal trigeminal nucleus, oral part Supraoculomotor nucleus Cuneate nucleus Gracile nucleus K6lliker-Fuse nucleus Locus coeruleus Parabrachial nucleus, lateral part Parabrachial nucleus, medial part Superior colliculus Nucleus of the solitary tract Ambiguus nucleus Prepositus hypoglossal nucleus Area postrema Barrington's nucleus Dorsal tegmental nucleus Ventral tegmental nucleus Laterodorsal tegmental nucleus Deep cerebellar nuclei Granular layer of the cerebellar cortex Molecular layer of the cerebellar cortex Oculomotor nucleus Trochlear nucleus Hypoglossal motor nucleus Facial nucleus Dorsal motor nucleus of the vagus Abducens nucleus Motor trigeminal nucleus Gigantocellular reticular nucleus Gigantocellular reticular nucleus, alpha part Gigantocellular reticular nucleus, ventral part Parvocellular reticular nucleus, alpha part Parvocellular reticular nucleus Intermediate reticular nucleus Lateral paragigantocellular nucleus Dorsal paragigantocellular nucleus Medullary reticular nucleus, dorsal part Medullary reticula.r nucleus, ventral part Raphe magnus nucleus Raphe pallidus nucleus Raphe obscurus nucleus Dorsal raphe nucleus Medial vestibular nucleus, ventral part Superior vestibular nucleus Spinal vestibular nucleus Medial vestibular nucleus Lateral vestibular nucleus Pontine level Medial superior olive Lateral superior olive Ventral preolivary nucleus Nucleus of the trapezoid body Dorsal cochlear nucleus Ventral cochlear nucleus Ventral nucleus of the lateral lemniscus Dorsal nucleus of the lateral lemniscus Inferior colliculus Supratrigeminal nucleus Principal sensory trigeminal nucleus Pontine nucleus Deep mesencephalic nucleus Pontine reticular nucleus, oral part Pontine reticular nucleus, caudal part Periaqueductal gray, dorsolateral part Periaqueductal gray, ventrolateral part Subcoeruleus nucleus, ventral part Subcoeruleus nucleus, dorsal part Edinger-Westphal nucleus Raphe pontis nucleus

SpSI SpSC SpSO Su3 Cu Gr KF LC PbL PbM SC Sol Amb PrH AP Bar DTg VTg LDTg

Cell body

Fiber density

++++ ++ ++

+++ +++ +++ +++ +++ +++ +++ +++ +++ +++ ++ ++ ++++ +++ + ++ ++ ++ ++ ++ ++ + ++ ++ ++++ ++++ ++ ++++ ++++ +++ +++ +++ +++ +++ +++ +++ +++ +++ +++ +++ +++ +++ ++ +++ +++ +++ +++ +++

0

+ ++ + 0 0 0

++ + ++ +++ 0 0 0 0

+++ ++ 3 4 12

7 10

6 MoS Gi GiA GiV PCRtA PCRt IRt LPGi DPGi MdD MdV RMg RPa ROb DR MVeV SuVe SpVe MVe LVe MSO LSO VPO Tz DC VCP VLL DLL IC SuS

PrS Pn DpMe PnO PnC CGDL CGVL SubCV SubCD EW RPn

0 0 0

+ 0 0

++++ ++ + +++ +++ +++ ++ ++ ++ ++ +++ 0 0 0

+++ + ++ + + +++ +++ +++ ++++ +++ + +++ 0

+++ ++ 0

++ ++ +++ 0

+++ ++ + 0

+

++++ ++++ ++++ + +++ +++ ++++ ++++ +++ +++ +++ ++ +++ +++ +++ ++ +++ +++ +++ +++ +++

739

Glycine-immunoreactive neurons and fibers in the rat brain Table I, continued Cell body

Fiber density

Dk IMLF APT R

0 0 0 0

++ +++ +++

PF IL PV MD SFO LHb

0 0 0 0 ++ ++

+++ ++

Pa PH PM SuM TM FF ZI Arc

0 0 0 0 0 0 0 0

++ ++

DB VP SI

0 0 0

Structure Pontine level

Nucleus of Darschewitsch Interstitial nucleus of the medial longitudinal fasciculus Anterior pretectal nucleus Red nucleus

+

Thalamus

Parafascicular thalamic nucleus Intralaminar thalamic nucleus Paraventricular thalamic nucleus Mediodorsal thalamic nucleus Subfornical organ Lateral habenular nucleus

++ +

+ ++

Hypothalamus

Paraventricular hypothalamic nucleus Posterior hypothalamic area Premammillary nucleus Supramammillary nucleus Tuberomammillary nucleus Fields of Forel Zona incerta Arcuate nucleus

+ + + +++ +++

+-

Basal forebrain

Diagonal band of Broca Ventral pallidum Substantia innominata

+ + +

Semi-quantitative analysis of glycine-IR cells and fibers in the rat brain. Cell counts were made on frontal sections and reflect the relative number of labeled neurons per structure. We used the following scale: more than 150 labeled cell bodies, ++++; 50-150, +++; 20-50, ++; five to 20, +; less than five, + -; no cell bodies, O. To illustrate the relative density of stained fibers, we used the same nomenclature: ++++ corresponds to a very large number; +++, large; ++, moderate to substantial; +, small; + -, a few in number; 0, no fibers. containing 0.5--4% paraformaldehyde, 0.5--4% glutaraldehyde and 0.2% picric acid. After removal from the skull, the brains were postfixed overnight in a solution of 0.1 M PB containing 2-4% paraformaldehyde and 0.2% picric acid at 4'C. They were rinsed and cryoprotected by immersion in 0.1 M PB containing 30% sucrose for two to three days at 4'C. Subsequently, brains were rapidly frozen with CO 2 gas and cut on a cryostat in coronal IS-11m-thick sections. These free-floating sections were rinsed in 0.1 M phosphatebuffered saline (0.9% NaCI) containing 0.3% Triton X-IOO and 0.1 % azide (PBST-Az). The sections were then submitted to glycine immunohistochemistry, according to the following procedure. Glycine antibodies

The immunohistochemical detection of glycine-IR cell bodies was carried out using two rabbit antisera against glycine. The first one was directed against glycine conjugated to thyroglobulin with glutaraldehyde and reduced by a sodium borohydride solution (Interchim, France). This antiserum has been purified by adsorption on various glutaraldehyde-conjugated protein carriers, and its specificity verified by an enzyme-linked immunosorbent assay. I I The second antiserum (a gift from Dr Wenthold) was obtained with glycine conjugated to bovine serum albumin with glutaraldehyde. 74 Its specificity was increased by affinity chromatography and assessed by either immunoblots of ovalbumin-conjugated amino acids 74 or adsorption tests on free-floating sections?O Immunohistochemistry of glycine

To obtain glycine immunostaining with Interchim's antibody, it was necessary to pretreat the sections by immersion

during 10 min in a solution of natrium borohydride (0.4%) in 0.1 M PB. Such treatment had no effect on the glycine staining obtained with Wenthold's antibody. Immunohistochemical detection of glycine was carried out by sequential incubations of the free-floating sections according to the technique of Hsu et al?8 Following each incubation, the sections were rinsed for 2 x 15 min in phosphate-buffered saline containing 0.3% Triton X-100 (PBST) under gentle stirring at room temperature. The sections were incubated in PBST-Az for four days in the primary rabbit antiserum to glycine (1:5000, Interchim; or 1:20,000, Wenthold's antibody) at 4'C. They were then placed in a biotinylated goat anti-rabbit immunoglobulin solution during 90 min (1:1000 in PBST; Jackson Immunoresearch Labs) at room temperature. Subsequently, the sections were immersed in an avidin-biotin-HRP complex solution (1:1000 in PBST; Vector Elite Kit Vectastain) or streptavidin-HRP (I :40,000; Jackson Immunoresearch Labs) during 90 min. Finally, the glycine immunostaining was visualized by placing the sections for 10-15 min in 0.02% 3,3'-diaminoQenzidineAHCI (Sigma, France) containing 0.003% H 2 0 2 and 0.6% nickel ammonium sulfate in 0.05 M Tris-HCI buffer (pH 7.6) at room temperature. The histochemical reaction was stopped by extensive washes in PBST-Az. The glycine-IR cell bodies and fibers were identified by a reproducible and specific blue-black coloration. Data analysis

The sections were mounted on gelatin-coated glass slides, dried, dehydrated and coverslipped with Depex. They were later observed and drawn with a Leitz orthoplan microscope equipped with an XIV sensitive stage and a video camera connected to a computerized image data analysis system (Biocom, France). Outlines of sections and major

740

C. Rampon et al.

Abbreviations used in the figures

12 12n 3 3V 4 4V 7

7n 10 Amb AP Aq CC CG CLi cp Cu DC DLL DPGi DpMe DR DTg ECIC ECu

g7 Gi GiA GiV Gr ic icp 10 IP IRt KF LC LD LDTg Ifp LP LPGi LRt LRtS5 LSO LVe MdD MdV Me5

hypoglossal nucleus hypoglossal nerve oculomotor nucleus third ventricle trochlear nucleus fourth ventricle facial nucleus facial nerve dorsal motor nucleus of the vagus ambiguus nucleus area postrema aqueduct (Sylvius) central canal periaqueductal gray caudal linear nucleus of the raphe cerebral peduncle, basal part cuneate nucleus dorsal cochlear nucleus dorsal nucleus of the lateral lemniscus dorsal paragigantocellular nucleus deep mesencephalic nucleus dorsal raphe nucleus dorsal tegmental nucleus external cortex of the inferior colliculus external cuneate nucleus genu of the facial nerve gigantocellular reticular nucleus gigantocellular reticular nucleus, alpha part gigantocellular reticular nucleus, ventral part gracile nucleus internal capsule inferior cerebellar peduncle inferior olive interpeduncular nucleus intermediate reticular nucleus K6lliker-Fuse nucleus locus coeruleus laterodorsal thalamic nucleus laterodorsal tegmental nucleus longitudinal fasciculus of the pons lateral posterior thalamic nucleus lateral paragigantocellular nucleus lateral reticular nucleus lateral reticular nucleus, subtrigeminal part lateral superior olive lateral vestibular nucleus medullary reticular nucleus, dorsal part medullary reticular nucleus, ventral part mesencephalic trigeminal nucleus

me5 ml mlf M05 MPB mt MVe MVeV OV PCRt PCRtA pd Pn PnC PnO PnV Pr5 PrH py RMg ROb RPa RPn rs RtTg SC scp Sol sol sp5 Sp5C Sp5I Sp50 SpVe Su3 Su5 SubCA SubCD SubCV SuVe Tz VC VCP VL VLL VPM VPO vtgx xscp

mesencephalic trigeminal tract medial lemniscus medial longitudinal fasciculus motor trigeminal nucleus medial parabrachial nucleus mammi110thalamic tract medial vestibular nucleus medial vestibular nucleus, ventral part vascular organ of the lamina terminalis parvocellular reticular nucleus parvocellular reticular nucleus, alpha part predorsal bundle pontine nuclei pontine reticular nucleus, caudal part pontine reticular nucleus, oral part pontine reticular nucleus, ventral part principal sensory trigeminal nucleus prepositus hypoglossal nucleus pyramidal tract raphe magnus nucleus raphe obscurus nucleus raphe pallidus nucleus raphe pontis nucleus rubrospinal tract reticulotegmental nucleus of the pons superior colliculus superior cerebellar peduncle nucleus of the solitary tract soIitary tract spinal trigeminal tract spinal trigeminal nucleus, caudal part spinal trigeminal nucleus, interpolar part spinal trigeminal nucleus, oral part spinal vestibular nucleus supraoculomotor central gray supratrigeminal nucleus subcoeruleus nucleus, alpha part subcoeruleus nucleus, dorsal part subcoeruleus nucleus, ventral part superior vestibular nucleus nucleus of the trapezoid body ventral cochlear nucleus ventral cochlear nucleus, posterior part ventrolateral thalamic nucleus ventral nucleus of the lateral lemniscus ventral posteromedial thalamic nucleus ventral preolivary nucleus ventral tegmental decussation decussation of the superior cerebellar peduncle

Fig. 1. Photomicrographs of frontal sections from the lower brainstem after the immunohistochemistry of glycine using Interchim's antibody. (A) Photomicrograph showing the presence of a very large number of glycine-IR fibers in the hypoglossal motor nucleus. Note also the presence of glycine-IR cells in the lateral part of the nucleus and the parvocellular reticular nucleus just lateral to it. (B) Photomicrograph i11ustrating the very large number of glycine-IR cells and fibers localized in the spinal trigeminal nucleus, oral part. (C) Photomicrograph showing the population of glycine-IR neurons localized in the lateral part of the nucleus of the solitary tract. The area postrema appears darkly stained due to the large number of labeled cells. The dorsal motor nucleus of the vagus contained a moderate number of glycine-IR fibers, contrasting with the adjacent hypoglossal motor nucleus covered by numerous terminal-like glycine-IR dots. (D) Photomicrograph showing the large number of glycine-IR cells bodies and fibers in the parvocellular and intermediate reticular nuclei. Note that labeled cells in the intermediate reticular nucleus are medium-sized. The dorsal paragigantocellular and prepositus hypoglossal nuclei displayed a large number of glycine-IR fibers and a moderate number of stained cells. (E) Photomicrograph at the pontine level showing glycine-IR cells localized in the medial part of the caudal pontine reticular nucleus (PnC) between the predorsal bundles. Scale bars = 100/-lm (A-C); 200/-lm (D, E).

Glycine·immunoreactive neurons and fibers in the rat brain

Fig. 1.

741

C. Rampon et al.

742

structures were drawn at low-power magnification, while immunoreactive neurons were plotted at higher magnification (Figs 3, 4). In order to estimate the number of glycine-IR cel1s present in each nucleus, they were counted unilateral1y, in two representative rats, on sections chosen at regular intervals from the caudal medul1a to the caudal mesencephalon. In the text, the fol1owing classification was used: very large (150 cel1s), large (50-ISO), moderate to substantial (20-50), smal1 (five to 20) and few or occasional (less than five). A semi-quantitative analysis of the glycinergic fiber labeling is represented in Table I.

Photomicrographs The photomicrographs were taken with a Leitz microscope connected to a camera (Vario-orthoplan). The negatives were scanned and the digital photomicrographs were stored on a Kodak photo compact disc. To get an optimal reproduction of the glycine staining, we modified the contrast and the luminosity of the crude scans with Photoshop 3.0 (Adobe). The il1ustration plates were then printed with a black and white dye-sublimation printer (Sony UPD 7000 E) (Figs I, 2, 5, 6): RESULTS

Technical considerations

To improve the glycine immunostaining of fibers and cell bodies, we tested different fixatives for each glycine antibody. A strong staining of the glycine-IR fibers and a weak staining of cell bodies was obtained with Wenthold's antibody when using a fixative containing 4% paraformaldehyde and 0.5% glutaraldehyde. With this fixative, the glycine immunostaining obtained with Interchim's antibody was weak. In contrast, a strong labeling of cell bodies and a rather weak fiber staining was obtained for this latter antibody with a fixative containing 0.5% paraformaldehyde and 4% glutaraldehyde. Distribution of glycine-immunoreactive cell bodies

The glycine-IR neurons were distributed from the medulla to the caudal mesencephalon with a decreasing gradient. The indicated size of the cells corresponds to a mean of the larger diameter of five to 10 perikarya. Cell body and fiber distributions are described using the terminology of Paxinos and Watson49 (Figs 1-6).

Medullary level. The caudal (Fig. 3A) and interpolar (Fig. 3B, C) spinal trigeminal nuclei displayed a very large number of round to ovoid medium-sized (20/-lm) glycine-IR cells. The oral spinal trigeminal nucleus (Figs 1B, 3D-F) contained a moderate number of triangular and medium-sized (20/-lm) labeled cells. A substantial number of ovoid and medium-sized or round and small-sized labeled cells was observed in the gracile nucleus (Fig. 3A). The cuneate nucleus (Fig. 3B, C) contained a small number of glycine-IR cells, round or fusiform and medium (15 /-lm) or small (8 /-lm) in size. A large number of round and small-sized (10 /-lm) neurons was clustered in the area postrema (Figs 1C, 3A). A small number of triangular medium-sized glycine-IR neurons was observed in the lateral part of the nucleus of the solitary tract and the caudal part of the ambiguus nucleus (Figs lC, 3A, B). The bipolar and small-sized Golgi cell bodies were immunostained throughout the granular layer of the cerebellum. The deep cerebellar nuclei contained a large number of round and medium-sized (15/-lm) glycine-IR neurons. A small number of ovoid and medium-sized (1520/-lm) neurons was localized in the ventrolateral part of the hypoglossal nucleus (Figs lA, 3A-C). These neurons were intermingled with unstained motoneurons. Occasional round, medium-sized (20/-lm) glycine-IR cells were found in the motor trigeminal (Fig. 4H) and abducens nuclei (Fig. 3F). A few glycine-IR cells were seen on the border of the motor trigeminal nucleus. The trochlear (Fig. 4K), oculomotor (Fig. 4L) and facial (Fig. 3E, F) nuclei were devoid of glycine-IR cells. The nuclei of the medullary reticular formation contained a moderate to large number of glycine-IR cells. The parvocellular (Figs ID, 3B-E) and intermediate reticular nuclei (Fig. 3B-D) contained a large number of labeled cells, round and small-sized (10/-lm) and triangular and medium-sized (20/-lm), respectively. The gigantocellular, alpha part (Figs 2C, 3D-F), and lateral paragigantocellular reticular nuclei (Fig. 3C, D, F) presented a substantial number of multipolar medium-sized (20/-lm) glycine-IR cells.

Fig. 2. Photomicrographs of frontal sections after the immunohistochemistry of glycine with Interchim's antibody. (A, C) High (A) and low (C) power photomicrographs showing glycine-IR neurons and fibers localized in the rostral part of the nucleus raphe magnus and the nucleus gigantocel1ular, alpha part. Note that the nucleus raphe paJlidus is devoid of labeled cel1s. (B) Photomicrograph showing glycine-IR cel1s and fibers localized inside or just lateral to the mesencephalic trigeminal tract, in the supratrigeminal nucleus. (D) Photomicrograph illustrating the numerous glycine-IR cel1s and fibers localized in the prepositus hypoglossal nucleus and medial vestibular nucleus, ventral part. A smal1 number of cell bodies and a large number of fibers are visible in the medial vestibular nucleus. (E) Photomicrograph showing a large number of glycine-IR cel1s in the parvocel1ular reticular nucleus, alpha part, lateral to the descending branch of the facial nerve. (F) Photomicrograph il1ustrating the group of glycine-IR neurons found in the ventrolateral part of the periaqueductal gray. (G) Photomicrograph showing the very large number of glycine-IR cel1s localized in the medial nucleus of the trapezoid body. Note also the large number of fibers localized in the adjacent lateral superior olivary complex. Scale bars = 25 11m (A); 100 11m (B, E-G); 200 11m (C, D).

Glycine-immunoreactive neurons and fibers in the rat brain

743

Fig. 2.

A moderate number of medium-sized (15-20 llm) glycine-IR neurons was observed in the medullary dorsal and ventral (Fig. 3A), gigantocellular (Figs 3B-F) and dorsal paragigantocellular reticular (Fig. lD, 3D, E) nuclei. In the gigantocellular reticular

nucleus, the cells were preferentially localized in the rostral part. A small number of triangular and largesized (2511m) glycine-IR neurons was also observed in the gigantocellular reticular nucleus, ventral part (Fig.3C).

744

C. Rampon et al.

c

Fig. 3.

A substantial to large number of fusiform and large-sized (25 11m) glycine-IR neurons was localized in the rostral part of the nucleus raphe magnus (Figs 2A, C, 4G, H). In contrast, its caudal part contained only a few labeled cells (Fig. 3F). A moderate number of oval and medium-sized glycine-IR cells was observed in an undefined region located dorsally to the rostral part of the raphe magnus nucleus between the predorsal bundles (Figs IE, 4G).

A substantial number of round and medium-sized (15 11m) neurons was stained in the medial vestibular nucleus, ventral part (Figs 2D, 3E), and the spinal vestibular nucleus (Fig. 3D). At the same level, the medial (Fig. 3D, E), lateral (Fig. 3E, F) and superior (Figs 2D, 3F) vestibular nuclei contained a small number of glycine-IR cells, small to medium-sized (10-20 J..lm) (Figs 2D, 3D-F). The prepositus hypoglossal nucleus exhibited a small number of

745

Glycine-immunoreactive neurons and fibers in the rat brain

G

I

J ~CG

.. ~.

.

......

PnO ......

Ifp

K

Fig. 4. Figs 3 and 4. Drawing of IS-11m frontal sections (from caudal to rostral levels) illustrating the localization of glycine-IR neurons. Each dot represents one glycine-IR neuron.

large-sized (20-30 11m) fusiform cells in its rostral part (Figs 2D, 3E) and a substantial number of small-sized (10-15 11m) round neurons in its caudal part (Fig. 3D).

Pontine level. The ventral nucleus of the lateral lemniscus (Fig. 41) and the lateral superior olive (Figs 2G, 4G, H) contained a large number of round, medium-sized (15 11m) and a small number of small

746

C. Rampon et al.

(10 )lm) glycine-IR neurons. In the latter structure, the majority of the glycine-IR neurons was found dorsally. A large number of round or triangular glycine-IR cells was also localized in the ventral preolivary nucleus (Fig. 4G, H). A very high concentration of strongly stained glycine-IR neurons was observed in the trapezoid body nucleus (Figs 2G, 4G, H). The dorsal cochlear nucleus contained a large number of ovoid or round labeled cells (Figs 3E, F, 4G). Small cells (10 )lm) were mainly found in the deep layer of this nucleus, whereas medium-sized cells (15 )lm) were rather observed in its superficial layer. Only a small number of round and small-sized (10 )lm) labeled cells was seen in the ventral cochlear nucleus (Figs 3E, F, 4G). Round and small-sized (12 )lm) labeled cells were present in moderate numbers in the principal sensory trigeminal (Fig. 4G, H) and the supratrigeminal nuclei (Figs 2B, 4G, H). A small number of glycine-IR cells was observed in the lateral part of the raphe pontis nucleus and just lateral to it (Fig. 4H, I). Neurons showing glycine immunoreactivity were not seen in the other pontine raphe nuclei (Figs 2e, 3D). At the pontomedullary junction, the parvocellular reticular nucleus, alpha part, lateral to the descending branch of the facial nerve and medial to the principal sensory trigeminal nucleus, contained a large number of triangular, large-sized (25 )lm) stained neurons (Figs 2E, 3F, 4G). The caudal pontine reticular nucleus exhibited a large number of small-sized glycine-IR neurons (10 )lm), rather localized in its ventral part, and intermingled with giant unstained cells (Figs IE, 4G, H). Ventral to the motor trigeminal nucleus, the ventral subcoeruleus nucleus contained a large number of oval and medium-sized (20 )lm) glycine-IR cells, whereas the dorsal subcoeruleus nucleus presented a small number of fusiform and large-sized (25 )lm) labeled cells (Fig. 4H). The K611iker-Fuse nucleus contained a small number of glycine-IR cells, round or fusiform and medium (15 )lm) or small (8 )lm) in size (Fig. 41). At the level of the inferior colliculus, the oral pontine reticular nucleus exhibited a large number of labeled cells (Fig. 41), particularly in its ventral part. The cells were round and medium-sized (15-20 )lm) ventrally, and fusiform and large-sized (25-30 )lm) dorsally.

Mesencephalic level. A large number of round to ovoid and small- to medium-sized (15-20 )lm) glycine-IR cells was observed in the ventral and ventrolateral parts of the periaqueductal gray (Figs 2F, 4J-L). This group of cells extended rostrocaudally from the caudal part of the oculomotor nucleus to the level of the caudal extension of the dorsal raphe nucleus. At the same level, a moderate number of oval and medium-sized (20 )lm) glycine-IR cells was localized in the deep mesencephalic nucleus, lateral to the periaqueductal gray, and in the oral pontine reticular nucleus, dorsolateral and dorsal to the medial lemniscus (Fig. 4J-L). Diencephalic and telencephalic levels. No glycine-IR cells were found in the forebrain, except in the subfomical organ (Fig. 6A) and the lateral habenular nucleus, which presented a moderate number of round, small-sized neurons (10 )lm). Distribution of glycine-immunoreactive fibers

The density of the glycine-IR fibers decreased from the medulla to the basal forebrain, and only occasional fibers were seen in the cortex, excepting the cingulate cortex. In the regions containing glycine-IR varicose fibers, basket-like structures were often encountered around both glycine-positive and -negative cell bodies. Medullary level. The highest density of glycine-IR fibers was observed in the facial, ambiguus and hypoglossal motor nuclei. A large number of glycine-IR fibers was also seen in the vestibular (Fig. ID), cochlear, cuneate, gracile and prepositus hypoglossal (Figs ID, 2D) nuclei, and in the magnus, obscurus and pallidus raphe nuclei (Fig. 2C). The caudal, intermediate and oral parts of the spinal trigeminal nuclei (Fig. IB), the parvocellular (Fig. ID), dorsal and lateral paragigantocellular, gigantocellular and alpha (Fig. 2C) and ventral gigantocellular reticular nuclei also contained a large number of stained fibers. In contrast, a moderate number of fibers was seen in the nucleus of the solitary tract, the dorsal motor nucleus of the vagus and the area postrema (Fig. lC). A few fibers were distributed in the inferior olive, except its most caudal part, at the level of the central canal, which contained a substantial number of glycine-IR fibers.

Fig. 5. Photomicrographs of frontal sections illustrating glycine-IR fibers after immunohistochemical staining of glycine with Wenthold's antibody. (A) Photomicrograph illustrating the high density of glycine-IR fibers localized in the periaqueductal gray. Note that the ventrolateral, lateral and dorsomedial parts of the periaqueductal gray contained more fibers than its dorsolateral part and the neighboring dorsal raphe nucleus. (B) Photomicrograph showing the large number of labeled fibers in the supraoculomotor nucleus, contrasting with the small number of fibers in the adjacent oculomotor nucleus. (C) Photomicrograph illustrating a bundle of non-varicose glycine-IR fibers vertically oriented in the posterior hypothalamic area. (D) High-power photomicrograph illustrating the glycine-IR varicose fibers localized in the dorsal raphe nucleus. (E) Photomicrograph showing the glycine-IR varicose fibers localized in the pontine nucleus. Scale bars = 200 /lm (A); 100 /lm (B, D, E); 50/lm (C).

Glycine-immunoreactive neurons and fibers in the rat brain

~

Fig. 5.

747

748

C. Rampon et al.

Pontine level. In the caudal part of the pons, a very large number of varicose fibers was distributed in the medial and lateral superior olive (Fig. 2G), the abducens and trigeminal motor nuclei. At the same level, the principal sensory trigeminal nucleus and the parvocellular reticular nucleus, alpha part, also exhibited a large number of glycine-IR fibers. A substantial number of varicose fibers was localized in the deep cerebellar nuclei and the granular layer of the cerebellar cortex. The molecular layer of the cerebellar cortex contained only a moderate number of labeled fibers. More rostrally, the ventral, intermediate and dorsal nuclei of the lateral lemniscus contained a very large number of glycine-IR fibers. A large number of stained fibers was distributed in the locus coeruleus, the K6lliker-Fuse, the lateral and medial parabrachial nuclei, the inferior colliculus, the subcoeruleus nuclei, and also in the caudal and oral pontine reticular nuclei. The laterodorsal tegmental, pedunculopontine, dorsal raphe (Fig. 5D) and Barrington's nuclei contained a substantial number of varicose glycine-IR fibers. A moderate number of fibers was observed in the dorsal and ventral tegmental nuclei and in the cuneiform nucleus. Mesencephalic level. A large number of glycine-IR varicose fibers was seen in the supraoculomotor region just dorsal to the oculomotor nucleus (Fig. 5B). A large number of glycine-IR fibers was found in the ventral, ventrolateral, lateral and dorsomedial parts of the pontine and mesencephalic periaqueductal gray (Fig. 5A), the interstitial nucleus of the medial longitudinal fasciculus, the Edinger-Westphal and the anterior pretectal nuclei. The oculomotor and trochlear motor nuclei, the nucleus of Darschewitsch and the dorsolateral part of the periaqueductal gray (Fig. 5A) contained only a moderate number of stained fibers. The entire mesencephalic reticular formation exhibited a large number of varicose fibers and bundles of non-varicose fibers cut transversally. A small number of varicose fibers was localized in the red nucleus. A substantial number of varicose glycine-IR fibers was observed in the deep and intermediate gray layers of the superior colliculus. At the same level,

thick non-varicose fibers were crossing the midline through the superior colliculus commissures. A substantial number of fibers was distributed in the pontine nuclei (Fig. 5E) and the lateral subnucleus of the interpeduncular nucleus. The superior central raphe nucleus contained a small number of horizontal varicose fibers. A few immunostained fibers were localized in the ventral tegmental area and the substantia nigra. Diencephalic and telencephalic levels. The parafascicular and central lateral thalamic nuclei contained a substantial number of glycine-IR fibers. A moderate number of varicose fibers was observed in the paraventricular, central medial and paracentral thalamic nuclei and the lateral habenula. The mediodorsal, lateral dorsal and lateral posterior thalamic nuclei contained a small number of glycine-IR fibers. No or occasional fiber immunostaining was seen in the medial and lateral geniculate nuclei, the reticular, posterior and ventrobasal thalamic nuclei. In the hypothalamus, at the rostral level of the paraventricular nucleus, plexuses of glycine-IR fibers were localized in the lateral hypothalamic area around the fornix and in the anterior part of the anterior hypothalamic nucleus, dorsal to the optic tract. The lateral and medial preoptic areas (Fig. 6B), the parvocellular parts of the paraventricular hypothalamic nucleus, and the lateral, dorsal and posterior areas of the posterior hypothalamus (Fig. 6C, E) contained a substantial number of varicose fibers. The magnocellular part of the paraventricular nucleus, the premammillary, supramammillary and tuberomammillary nuclei contained a small number of glycine-IR fibers. Only a few fibers were found in the arcuate hypothalamic nucleus, the mammillary nuclei, the supraoptic nucleus and the retrochiasmatic area. Three bundles of non-varicose glycine-IR fibers were also visible: the first one was seen in the medial forebrain bundle and the superior cerebellar peduncle at the level of the posterior hypothalamus (Fig. 6C), the second was dorsal and mediodorsal to the optic tract (Fig. 6B). This bundle of fibers crossed the midline through the supraoptic commissure of Meynert. The third bundle of fibers was composed of long non-varicose fibers oriented vertically in the posterior hypothalamic area at the mammillary level. At the same level, a large number of varicose

Fig. 6. Photomicrographs of frontal sections illustrating glycine-IR fibers in the forebrain after immunohistochemical detection of glycine with Wenthold's antibody. (A) Photomicrograph showing the presence of glycine-IR cells and fibers in the subfomical organ. (B) High-power photomicrograph illustrating long glycine-IR varicose fibers and terminal-like dots in the medial preoptic area lateral to the vascular organ of the lamina terminalis. (C) Photomicrograph showing the large number of glycine-IR fibers present in the lateral area of the posterior hypothalamus. Bundles of fibers are visible in the superior cerebellar peduncle and ventral to it in the dorsal part of the lateral hypothalamic area. (D) High-power photomicrograph showing glycine-IR fibers forming a basket-like structure around an unstained cell body of the substantia innominata. (E) Photomicrograph showing numerous glycine-IR varicose fibers in the dorsal hypothalamic area. (F) Photomicrograph illustrating the presence of numerous glycine-IR varicose fibers in the intralaminar nuclei of the thalamus. Scale bars = 100 11m (A, C, E, F); 50 11m (B, D).

and fibers in the rat brain Glycine-immunoreactive neurons

;,'.;, "';""~ ~ '! ~

..

Fig. 6.

749

C.

750

Rampon et al.

glycine-IR fibers was observed in the fields of Forel and the zona incerta. The adjacent entopeduncular and subthalamic nuclei contained no or only occasional fibers. A moderate number of glycine-IR fibers was seen in the diagonal band of Broca, the ventral pallidum and the substantia innominata (Fig. 6D). The medial and lateral septal nuclei contained a small number of immunoreactive fibers. A few glycine-IR varicose fibers were observed in the globus pallidus, the claustrum, the anterior cortical and central amygdaloid nuclei, and the magnocellular preoptic nucleus. Occasional fibers were localized in the caudateputamen. A small number of varicose fibers was observed in the cingulate cortex. The other cortical areas contained only occasional immunostained fibers. In the hippocampus, a few varicose fibers were immunostained. DISCUSSION

Combining the use of highly specific antibodies to glycine and a sensitive immunohistochemical method, we demonstrated in rats the presence of glycine-IR neurons and fibers in a large number of lower brainstem structures. Furthermore, we revealed the presence of glycine-IR fibers in upper brainstem areas such as the periaqueductal gray, intralaminar thalamic nuclei and hypothalamus. These results indicate that, in contrast to previous statements, glycine might be an essential inhibitory neurotransmitter not only in the lower brainstem and spinal cord, but also in more rostral regions. Technical considerations

A number of arguments indicates that the widespread distribution of neurons and fibers we report here is due to an increase of the sensibility of our immunohistochemical method and not to a problem of specificity. First, the specificity of the two glycine antisera used has been extensively studied and assessed. II ,73 Second, both gave rise to the labeling of the same populations of cell bodies and fibers. Third, the distribution of glycine-IR cell bodies was heterogeneous and well-known non-glycinergic cells such as the brainstem motoneurons or the cortical pyramidal cells remained unstained. Fourth, we observed glycine-IR neurons and fibers with a similar distribution as in previous studies in the trigeminal sensory nuclei,20.50 gracile and cuneate nuclei,21,51 nucleus of the solitary tract, area postrema,20,72 cerebellum,II,21,48,51,70 deep cerebellar nuclei,16,21,50 nucleus of the trapezoid body, the cochlear nuclei, the superior olivary complex, the nuclei of the lateral lemniscus and the inferior colliculus. 3,20,21,36,50,51,73,77 These and our results are further supported by the finding of glycine receptors in these nuclei by autoradiography,56,79 immunohistochemistry 6 and in situ

hybridization. 57 ,68 Finally, a strong strychninesensitive inhibitory effect of glycine has been described when applied on neurons of these nuclei.?,13,18,19.31 In addition to these structures, we described glycine-IR cell bodies and fibers in numerous other areas. We discuss these data in detail below. Medullary level

We showed, for the first time in rats, an extremely dense concentration of glycine-IR fibers in the trigeminal, facial and hypoglossal motor nuclei. Fort et al.,20,21 in cats, reported similar results. Our data are strongly supported by the finding of a very high density of glycinergic receptors in these nuclei. 9,52,57,79 Moreover, it has been shown that glycine was responsible for the tonic hyperpolarization of cat trigeminal motoneurons during paradoxical sl eep l4,59 and for their inhibition evoked by stimulation of the trigeminal sensory nerves. 33 The inhibition of hypoglossal motoneurons evoked by stimulation of the lingual nerve is also mediated by glycine. 38 In agreement with Wentzel et al. 75 in rabbits, Spencer et al. 60 and Fort et al. 20 in cats, the abducens nucleus contained a large number of fibers, whereas the trochlear and oculomotor nuclei presented only a moderate density of labeled fibers. In support of our results, Fort et al.,20 in cats, observed a substantial to large number of glycine-IR cells and fibers in the parvocellular, intermediate, gigantocellular alpha part and lateral paragigantocellular nuclei, and a moderate number of cells and a large number of fibers in the dorsal and ventral, gigantocellular and dorsal paragigantocellular reticular nuclei. We further saw glycine-IR cells in the nucleus reticular parvocellular alpha and the supratrigeminal nucleus. Previous studies in rats 11,51 noted the presence of glycine-IR cell bodies in the medullary reticular formation, but did not precisely describe their location. Our results and those of Fort et al.,20 in cats, indicate that a large proportion of the cells in the medullary reticular formation are glycinergic and that nearly all cells in these nuclei receive a glycinergic innervation. This assumption is supported by the high density of glycine receptors found in these nuclei. 6 ,57,79 Moreover, a strong strychnine-sensitive inhibitory effect of glycine has been described when applied on neurons from the medullary reticular formation.25.62-64 Anatomical 53 and electrophysi010gica1 12,69 studies in rats showed that glycinergic neurons in the parvocellular, parvocellular alpha part and supratrigeminal nuclei project to the trigeminal or hypoglossal motor nuclei. Holstege and Bongers27 have also shown, in rats, that glycinergic neurons in the nucleus gigantocellular alpha project to thoracic spinal motoneurons. These and our results corroborate the hypothesis that these nuclei might contain the glycinergic premotor neurons responsible for the

Glycine-immunoreactive neurons and fibers in the rat brain

hyperpolarization of cranial and spinal motoneurons occurring during paradoxical sleep.15,20 For the first time, we observed a large number of glycine-IR cell bodies and fibers in the rostral part of the nucleus raphe magnus. In agreement with Fort et al.2° in cats, we also observed a substantial to large number of glycine-IR fibers in the dorsal, pontis, obscurus and pallidus raphe nuclei. Further supporting these results, glycine receptors have been localized in the raphe nuclei 6,79 and an inhibitory effect of glycine has been shown On neurons of the dorsal raphe nucleus,23 The functional role of these glycinergic inhibitory inputs remains to be identified. We suggested that it could be responsible for the cessation of activity of serotoninergic neurons during paradoxical sleep.20 We saw a substantial to large number of mediumsized (15 I-lm) neurons and fibers in the medial vestibular, ventral part, spinal vestibular and prepositus hypoglossal nuclei. In contrast, Pourcho et al.,51 in rats, reported the presence of large multipolar cells (35--40 I-lm) in both medial and lateral vestibular nuclei. However, supporting our results, Walberg et al. 72 in guinea-pigs, Fort et al. 21 and Yingcharoen et al. 78 in cats, observed small or medium-sized glycine-IR neurons and a large number of glycine-IR fibers in the medial, descending, lateral vestibular and prepositus hypoglossal nuclei. These results are supported by the presence in these structures of glycinergic receptors 6.57 ,79 and the inhibitory effect of glycine application On vestibular neurons. 39 Pontine level

The ventral parts of the oral and caudal pontine reticular nuclei and the ventral subcoeruleus nucleus exhibited a large number of small-sized glycine-IR neurons intermingled with giant unstained cells. Pourcho et al.,51 in rats, previously reported the presence of cells in these regions, but did not detail their distribution, while, in cats, Fort et al.2° found only few glycine-IR neurons. The presence of glycinergic neurons is further supported by in vitro electrophysiological recordings in rat slices, showing that electrical stimulation of the ventral part of these nuclei induces a strong strychnine-sensitive inhibition of contralateral neurons. 24 We confirmed the presence of a large number of glycine-IR fibers in the locus coeruleus. 41 In agreement with Fort et al. 20 in cats, we also observed a large number of glycine-IR fibers in the K6I1iker-Fuse, the lateral and medial parabrachial nuclei, the dorsal and ventral parts of the subcoeruleus nuclei, and the oral and caudal pontine reticular nuclei. Supporting our results, numerous glycinergic receptors have been seen in these structures. 6,57,79 Moreover, it has been shown that local application of glycine induces inhibition of neurons located in the locus coeruleus 41 and the pontine medial reticular [ormation,24 in which a very high glycine content has also been found. 10

751

A substantial number of fibers was distributed in the pontine nuclei. Aas and Brodal, I in cats and rats, observed only a few glycine-IR fibers and Zarbin et al. 79 in rats and Probst et al. 52 in humans reported no strychnine binding in these nuclei. However, supporting our results, Frostholm and Rotter,22 in mice, found an intense strychnine labeling and Araki et al.,6 in rats, observed glycine receptor immunoreactivity in the pontine nuclei. Nevertheless, additional electrophysiological experiments are necessary to confirm that glycine is used as an inhibitory neurotransmitter in these structures. Mesencephalic level

We precisely mapped the distribution of a population of glycine-IR neurons in the ventrolateral part of the periaqueductal gray and the deep mesencephalic nucleus. In agreement with our data, glycine-IR neurons were seen previously in these structures in rats 11 ,51 and cats. 21 In addition, we observed a high density of varicose immunoreactive fibers in the ventral, ventrolateral, lateral and dorsomedial parts of the periaqueductal gray, the supraoculomotor region and the deep mesencephalic nucleus. Supporting our results, glycine receptors have been localized in these regions in rats. 6,79 These and our data suggest that glycine might play a crucial inhibitory role in the periaqueductal gray. We also demonstrated for the first time a large number of glycine-IR varicose fibers in the interstitial nucleus of the medial longitudinal fasciculus, the Edinger-Westphal and the anterior pretectal nuclei. Frostholm and Rotter,22 in mice, showed a moderate grain density of strychnine binding in the pretectal thalamic nucleus. However, no or only a small number of glycine receptors have been localized in these nuclei in rats. 6,57,79 Additional studies are therefore necessary to confirm the role of glycine in these structures. We observed only a few immunostained fibers in the ventral tegmental area and the substantia nigra. In agreement with our data, a low density of glycine receptors has been reported in the ventral tegmental area and the pars reticulata of the substantia nigra in humans 52 or rats. 6,57,66,79 However, in vitro and in vivo electrophysiological recordings showed that dopaminergic neurons are sensitive to glycine through activation of postsynaptic strychninesensitive receptorsy,43 Additional studies are therefore necessary to resolve these discrepancies. Diencephalic and telencephalic levels

We revealed the presence, in rats, of a substantial number of glycine-IR fibers in the anterior hypothalamic area, the lateral and medial preoptic areas, the parvocellular parts of the paraventricular hypothalamic nucleus, the lateral, dorsal and posterior

752

C. Rampon et al.

areas of the hypothalamus, the fields of Forel and the zona incerta. Van den Pol and Gorcs,70 in rats, reported only the presence of fibers in the hypothalamus, with no further description. Autoradiographic studies demonstrated detectable levels of [3H]strychnine binding sites in rats and mice only in the zona incerta. 22 ,79 Araki et a/. 6 and Van den Pol and Gorcs,7° using immunohistochemistry, found a labeling of glycine receptors only in the mammillary region and the magnocellular supraoptic and paraventricular nuclei. In contrast, neurons containing glycine receptor mRNA were localized in the rat's horizontal and vertical limbs of the diagonal band and anterior hypothalamic nuclei, the lateral hypothalamus and the zona incerta. 57 Moreover, in vivo microiontophoretic application of glycine decreased the spontaneous or evoked activity of neurons localized in the preoptic-anterior hypothalamic region, the medial hypothalamus and the lateral hypothalamic area, and strychnine blocked this inhibition. 8,34.35,42 Furthermore, when applied on intracellularly recorded dissociated hypothalamic neurons, glycine induced current reversal at the chloride equilibrium potentia1. 2,45,67 Together, these and our results suggest that glycine is used as an inhibitory neurotransmitter in several structures of the hypothalamus. One of the most striking results of the present study is the finding of a substantial number of glycine-IR fibers in the parafascicular and central lateral thalamic nuclei, and a moderate number in the paraventricular and central medial thalamic nuclei. Supporting our data, a strong density of glycine receptors has been found in the rat and mouse parafascicular nucleus,22.57.79 and a small to moderate density in the central medial and laterodorsal thalamic nuclei. 56 In addition, electrophysiological studies in cats have demonstrated that the inhibition of intralaminar thalamic neurons evoked by electrical stimulation is sensitive to strychnine. 22 ,29,37 The subfornical organ and the lateral habenula nucleus presented a moderate number of glycine-IR cells and fibers. This is the first report of the presence of glycine in these nuclei. Further studies are therefore needed to confirm our findings. In agreement with Van den Pol and Gorcs,7° in rats, a few glycine-IR fibers were seen in the hippocampus. We further reported the presence of a more

substantial plexus of fibers in the frontal and cingulate cortices. A negligible density of glycinergic receptors has been detected in the cortex and hippocampus using either immunohistochemistry,6,66 in situ hybridization 57 or [3H]strychnine binding techniques 9,22,79 in rats and mice. However, it has been shown in vivo and in vitro that glycine application induces an inhibition of cortical neuron activity which is reversed by strychnine. 3o,32,40,58 Our results suggest that glycine may be used as an inhibitory neurotransmitter in the frontal and cingulate, but not in other cortices. CONCLUSIONS AND PHYSIOLOGICAL IMPLICATIONS

In this study, we observed a large number of glycine-IR fibers in nearly all brainstem structures. Moreover, these fibers were localized in the vicinity of neurons and often formed basket-like structures around them. We also visualized a large number of glycine-IR cell bodies in numerous medullary and pontine structures. These findings indicate that glycine may be a major inhibitory neurotransmitter in the brainstem. A large number of GABAergic neurons and fibers has also been observed in the brainstem, indicating that GABA is another major inhibitory neurotransmitter at this level. 44 ,47 Additional double-labeling studies for each brainstem nucleus are therefore necessary to determine the importance and degree of co-localization of these two amino acid inhibitory neurotransmitters. Indeed, they have been found to co-localize in certain, but not all, populations of brainstem and spinal cord neurons. We also revealed the presence of glycine-IR fibers in a substantial number of mesencephalic and forebrain structures in which glycine has only marginally been considered as a neurotransmitter candidate. Although in these regions the density of glycine-IR fibers is far inferior to that of GABA, our results indicate that, in the future, glycine should also be considered as a candidate for inhibitory neurotransmission in these areas. 46 ,65 Acknowledgements-This work was supported by INSERM (U52) and CNRS (CRS 5645). The authors wish to thank

Dr R. J. Wenthold for kindly supplying the primary antibody and Dr J. S. Lin for helpful comments.

REFERENCES

Aas J. E. and Brodal P. (1990) GABA and glycine as putative transmitters in subcortical pathways to the pontine nuclei. A combined immunocytochemical and retrograde tracing study in the cat with some observations in the rat. Neuroscience 34, 149-162. 2. Akaike N. and Kaneda M. (1989) Glycine-gated chloride current in acutely isolated rat hypothalamic neurons. J. Neurophysiol. 62, 1400-1409. 3. Aoki E., Semba R., Keino H., Kato K. and Kashiwamata S. (1988) Glycine-like immunoreactivity in the rat auditory pathway. Brain Res. 442,63-71. 4. Aprison M. H. (1990) The discovery of the neurotransmitter role of glycine In Glycine Neurotransmission (eds Ottersen O. P. and Storm-Mathisen J.), pp. 1-23. John Wiley, Chichester. 5. Aprison M. H. and Werman R. (1965) The distribution of glycine in cat spinal cord and roots. Life Sci. 4,2075-2083. 1.

Glycine-immunoreactive neurons and fibers in the rat brain 6.

753

Araki T., Yamano M., Murakami T., Wanaka A., Betz H. and Tohyama M. (1988) Localization of glycine receptors in the rat central nervous system: an immunocytochemical analysis using monoclonal antibody. Neuroscience 25, 613-624. 7. Bennett J. A., McWilliam P. N. and Shepheard S. L. (1987) A gamma-aminobutyric-acid-mediated inhibition of neurones in the nucleus tractus solitarius of the cat. J. Physiol., Lond. 392,417-430. 8. Blume H. W., Pittman Q. J. and Renaud L. P. (1981) Sensitivity of identified medial hypothalamic neurons to GABA, glycine and related amino acids; influence of bicuculline, picrotoxine and strychnine on synaptic inhibition. Brain Res. 209, 145-158. 9. Bristow D. R., Bowery N. G. and Woodruff G. N. (1986) Light microscopic autoradiographic localization of [3H]glycine and [3H]strychnine binding sites in rat brain. Eur. J. Pharmac. 126, 303-307. 10. Camacho-Arroyo 1. and Tapia R. (1992) Levels, uptake and release of glycine and glutamate in the rat pontine reticular formation. Neurochem. Res. 17,463-467. 11. Campistron G., Buijs R. M. and Geffard M. (1986) Glycine neurons in the brain and spinal cord. Antibody production and immunocytochemical localization. Brain Res. 376,400-405. 12. Castillo P., Pedroarena c., Chase M. H. and Morales F. R. (1991) Strychnine blockade of the non-reciprocal inhibition of trigeminal motoneurons induced by stimulation of the parvocellular reticular formation. Brain Res. 567, 346-349. 13. Champagnat J., Denavit-Saubie M., Grant K. and Shen K. F. (1986) Organization of synaptic transmission in the mammalian solitary complex, studied in vitro. 1. Physio/., Lond. 381, 551-573. 14. Chase M. H., Chandler S. H. and Nakamura Y. (1980) Intracellular determination of the membrane potential of trigeminal motoneurons during sleep and wakefulness. J. Neurophysiol. 44, 349-358. 15. Chase M. H., Soja P. J. and Morales F. R. (1989) Evidence that glycine mediates the postsynaptic potentials that inhibit lumbar motoneurons during the atonia of active sleep. J. Neurosci. 9, 743-751. 16. Chen S. and Hillman D. E. (1993) Colocalization of neurotransmitters in the deep cerebellar nuclei. J. Neurocytol. 22, 81-90. 17. Collingridge G. L. and Davies J. (1981) The influence of striatal stimulation and putative neurotransmitters on identified neurones in the rat substantia nigra. Brain Res. 212, 345-359. 18. Curtis D. R. and Crawford J. M. (1968) Central synaptic transmission-Microelectrophoretic studies. A. Rev. Pharmac. 9, 209-240. 19. Feldman P. D. and Felder R. B. (1991) Effects of gamma-aminobutyric acid and glycine on synaptic excitability of neurones in the solitary tract nucleus. Neuropharmacology 30, 225-236. 20. Fort P., Luppi P. H. and Jouvet M. (1993) Glycine-immunoreactive neurones in the cat brain stem reticular formation. NeuroReport 4, 1123-1126. 21. Fort P., Luppi P. H., Wenthold R. and Jouvet M. (1990) Glycine immunoreactive neurons in the medulla oblongata in cats. C. R. Acad. Sci., Paris 311, 205-212. 22. Frostholm A. and Rotter A. (1985) Glycine receptor distribution in mouse CNS: autoradiographic localization of [3H]strychnine binding sites. Brain Res. Bull. 15, 473-486. 23. Gallager D. W. and Aghajanian G. K. (1976) Effect of antipsychotic drugs on the firing of dorsal raphe nucleus cells. 1. Reversal by picrotoxin. Eur. J. Pharmac. 39, 357-364. 24. Greene R. W. and Carpenter D. O. (1985) Actions of neurotransmitters on pontine medical reticular formation neurons of the cat. J. Neurophysio/. 54, 520-531. 25. Haji A., Remmers J. E., Connelly C. and Takeda R. (1990) Effects of glycine and GABA on bulbar respiratory neurons of cat. J. Neurophysiol. 63, 955-965. 26. Helfert R. H., Bonneau J. M., Wenthold R. J. and Altschuler R. A. (1989) GABA and glycine immunoreactivity in the guinea pig superior olivary. Brain Res. 501, 269-286. 27. Holstege J. C. and Bongers C. M. (1991) A glycinergic projection from the ventromedial lower brainstem to spinal motoneurons. An ultrastructural double labeling study in rat. Brain Res. 566, 308-315. 28. Hsu S., Raine L. and Fanger H. (1981) A comparative study of the peroxidase-antiperoxidase method and an avidin-biotin complex method for studying polypeptide hormones with radioimmunoassay antibodies. Am. 1. din. Path. 75, 734-738. 29. Ishida Y. (1977) Intralaminar thalamic responses to somatic and fastigial stimulation: cortical inhibition and its blockade by strychnine in cats. Neuropharmacology 16, 163-170. 30. Ito S. and Cherubini E. (1991) Strychnine-sensitive glycine responses of neonatal rat hippocampal neurones. J. Physiol., Lond. 440, 67-83. 31. Jordan D., Mifflin S. W. and Spyer H. M. (1988) Hypothalamic inhibition ofneurones in the nucleus tractus solitarius of the cat is GABA mediated. 1. Physiol., Lond. 399, 389-404. 32. Kelly J. S. and Krnjevic K. (1968) Effects of gamma-aminobutyric acid and glycine on cortical neurones. Nature 219, 1380-1381. 33. Kidokoro Y., Kubota K., Shuto S. and Sumino R. (1968) Possible interneurons responsible for reflex inhibition of motoneurons of jaw-closing muscles from the inferior dental nerve. J. Neurophysiol. 31,709-716. 34. Kita H. and Oomura Y. (1981) Reciprocal connections between the lateral hypothalamus and the frontal cortex in the rat: electrophysiological and anatomical observations. Brain Res. 213, 1-16. 35. Kita H. and Oomura Y. (1982) Evidence for a glycinergic cortico-Iateral hypothalamic inhibitory pathway in the rat. Brain Res. 235, 131-136. 36. Kolston J., Osen K. K., Hackney C. M., Ottersen O. P. and Storm-Mathisen J. (1992) An atlas of glycine- and GABA-like immunoreactivity and colocalization in the cochlear nuclear complex of the guinea pig. Anat. Embryol. 186, 443-465. 37. Krauthamer G. and Yamaguchi Y. (1964) Suppression par la strychnine de I'action inhibitrice des corpus stries. J. Physio/., Paris 56, 586. 38. Kubin L., Kimura H., Tojima H., Davies R. O. and Pack A. I. (1993) Suppression of hypoglossal motoneurons during the carbachol-induced atonia of REM sleep is not caused by fast synaptic inhibition. Brain Res. 611,300-312. 39. Lapeyre P. N. M. and de Waele C. (1995) Glycinergic inhibition of spontaneously active guinea-pig medial vestibular nucleus neurons in vitro. Neurosci. Lett. 188, 155-158.

754 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57.

58. 59.

60.

61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73.

C. Rampon et al. Levi G., Bernardi G., Cherubini E., Gallo V., Marciani M. G. and Stanzione P. (1982) Evidence in favor of a neurotransmitter role of glycine in the rat cerebral cortex. Brain Res. 236, 121-131. Luppi P. H., Charlety P. J., Fort P., Akaoka H., Chouvet G. and Jouvet M. (1991) Anatomical and electrophysiological evidence for a glycinergic inhibitory innervation of the rat locus coeruleus. Neurosci. Lett. 128, 33-36. Mayer M. L. (1981) Electrophysiological analysis of inhibitory synaptic mechanisms in the preoptic area of the rat. J. Physio!., Lond. 316, 327-346. Mercuri N. B., Calabresi P. and Bernardi G. (1990) Effects of glycine on neurons in the rat substantia nigra zona compacta: in vitro electrophysiological study. Synapse 5, 190-200. Mugnaini E. and Oertel W. H. (1985) An atlas of the distribution ofGABAergic neurons and terminals in the rat CNS as revealed by GAD immunohistochemistry. In Handbook of Chemical Neuroanatomy (eds Bjorklund A. and Hokfelt T.), pp. 436-608. Elsevier, Amsterdam. Nose I., Higashi H., Inokuchi H. and Nishi S. (1991) Synaptic responses of guinea pig and rat central amygdala neurons in vitro. J Neurophysio!. 65, 1227-1241. Ottersen O. P., Davanger S. and Storm-Mathisen J. (1987) Glycine-like immunoreactivity in the cerebellum of rat and Senegalese baboon, Papio papio: a comparison with the distribution of GABA-like immunoreactivity and with [3H]glycine and [3H]GABA uptake. Expl Brain Res. 66, 211-221. Ottersen O. P. and Storm-Mathisen J. (1984) Glutamate and GABA containing neurons in the mouse and rat brain, as demonstrated with a new immunocytochemical technique. J compo Neurol. 229, 374-392. Ottersen O. P., Storm-Mathisen J. and Somogyi P. (1988) Colocalization of glycine-like and GABA-like immunoreactivities in Golgi cell terminals in the rat cerebellum: a postembedding light and electron microscopic study. Brain Res. 450, 342-353. Paxinos G. and Watson C. (1986) The Rat Brain in Stereotaxic Coordinates, 2nd edn. Academic Press, San Diego. Peyret D., Campistron G., Geffard M. and Aran J. M. (1987) Glycine immunoreactivity in the brainstem auditory and vestibular nuclei of the guinea pig. Acta otolar., Stockh. 104,71-76. Pourcho R. G., Goebel D. J., Jojich L. and Hazlett J. C. (1992) Immunocytochemical evidence for the involvement of glycine in sensory centers of the rat brain. Neuroscience 46, 643-656. Probst A., Cortes R. and Palacios J. M. (1986) The distribution of glycine receptors in the human brain. A light microscopic autoradiographic study using eH]strychnine. Neuroscience 17, 11-35. Rampon c., Luppi P. H., Peyron c., Fort P., Petit J. M. and Jouvet M. (1995) Origin of the glycinergic afferent to the trigeminal motor nucleus of the rat. Eur. J. Neurosci. 8 Suppl., 33-35. Saint Marie R. L., Benson C. G., OstapoffE. M. and Morest D. K. (1991) Glycine immunoreactive projections from the dorsal to the anteroventral cochlear nucleus. Hear. Res. 51, 11-28. Saint Marie R. L., OstapoffE. M., Morest D. K. and Wenthold R. 1. (1989) Glycine-immunoreactive projection of the cat lateral superior olive: possible role in midbrain ear dominance. J compo Neurol. 279, 382-396. Sanes D. H., Geary W. A., Wooten G. F. and Rubel E. W. (1987) Quantitative distribution of the glycine receptor in the auditory brain stem of the gerbil. J. Neurosci. 7, 3793-3802. Sato K., Zhang J. H., Saika T., Sato M., Tada K. and Tohyama M. (1991) Localization of glycine receptor alpha I subunit mRNA-containing neurons in the rat brain: an analysis using in situ hybridization histochemistry. Neuroscience 43, 381-395. Shirasaki T., Klee M. R., Nakaye T. and Akaike N. (1991) Differential blockade of bicuculline and strychnine on GABA- and glycine-induced responses in dissociated rat hippocampal pyramidal cells. Brain Res. 561, 77-83. Soja P. J., Morales F. R., Baranyi A. and Chase M. H. (1987) Effect of inhibitory amino acid antagonists on IPSPs induced in lumbar motoneurons upon stimulation of the nucleus reticularis gigantocellularis during active sleep. Brain Res. 423, 353-358. Spencer R. F., Wenthold R. J. and Baker R. (1989) Evidence for glycine as an inhibitory neurotransmitter of vestibular, reticular, and prepositus hypoglossi neurons that project to the cat abducens nucleus. J Neurosci. 9, 2718-2736. Storm-Mathisen J., Leknes A. K., Bore A. T., Vaaland J. L., Edminson P., Haug F. M. S. and Ottersen O. P. (1983) First visualization of glutamate and GABA in neurons by immunocytochemistry. Nature 301, 517-519. Tebecis A. K. (1973) Transmitters and reticulospinal neurons. Expl Neurol. 40,297-308. Tebecis A. K. and Di Maria A. (1972) Strychnine-sensitive inhibition in the medullary reticular formation: evidence for glycine as an inhibitory transmitter. Brain Res. 40, 373-383. Tebecis A. K. and Ishikawa T. (1973) Glycine and GABA as inhibitory transmitters in the medullary reticular formation. Pflugers Arch. 338, 273-278. Todd A. J. and Sullivan A. C. (1990) Light microscope study of the coexistence of GABA-like and glycine-like immunoreactivities in the spinal cord of the rat. J. compo Neuro!. 296, 496-505. Tohyama M., Wanaka A., Araki T., Betz H. and Malbon C. C. (1989) Localization of glycine and beta-adrenergic receptors in the rat brain. Archs Histo!' Cytol. 52, 39-48. Tokutomi N., Agopyan N. and Akaike N. (1992) Penicillin-induced potentiation of glycine receptor-operated chloride current in rat ventro-medial hypothalamic neurones. Br. J Pharmac. 106, 73-78. Triller A., Cluzeaud F. and Korn H. (1987) Gamma-aminobutyric acid-containing terminals can be apposed to glycine receptors at central synapses. J. Cell Bioi. 104,947-956. Umemiya M. and Berger A. J. (1995) Presynaptic inhibition by serotonin of glycinergic inhibitory synaptic currents in the rat brain stem. J. Neurophysiol. 73, 1192-1200. Van den Pol A. N. and Gorcs T. (1988) Glycine and glycine receptor immunoreactivity in brain and spinal cord. J. Neurosci. 8, 472-492. Walberg F. and Ottersen O. P. (1992) Neuroactive amino acids in the area postrema. An immunocytochemical investigation in rat with some observations in cat and monkey (Macaca fascicularis). Anat. Embryo!. 185, 529-545. Walberg F., Ottersen O. P. and Rinvik E. (1990) GABA, glycine, aspartate, glutamate and taurine in the vestibular nuclei: an immunocytochemical investigation in the cat. Expl Brain Res. 79, 547-563. Wenthold R. J. (1987) Evidence for a glycinergic pathway connecting the two cochlear nuclei: an immunocytochemical and retrograde transport study. Brain Res. 415, 183-187.

Glycine-immunoreactive neurons and fibers in the rat brain 74. 75. 76.

77. 78.

79.

755

Wenthold R. J., Huie D., Altschuler R. A. and Reeks K. A. (1987) Glycine immunoreactivity localized in the cochlear nucleus and superior olivary complex. Neuroscience 22, 897-912. Wentzel P. R., de Zeeuw C. 1., Holstege J. C. and Gerrits N. M. (1993) Colocalization of GABA and glycine in the rabbit oculomotor nucleus. Neurosci. Lett. 164, 25-29. Werman R., Davidoff R. A. and Aprison M. H. (1967) Inhibition of motoneurons by iontophoresis of glycine. Nature 214,681-683. Winer J. A., Larue D. T. and Pollak G. D. (1995) GABA and glycine in the central auditory system of the mustache bat: structural substrates for inhibitory neuronal organization. 1. compo Neural. 355,317-353. Yingcharoen K., Rinvik E., Storm-Mathisen J. and Ottersen O. P. (1989) GABA, glycine, glutamate, aspartate and taurine in the perihypoglossal nuclei: an immunocytochemical investigation in the cat with particular reference to the issue of amino acid colocalization. Expl Brain Res. 78, 345-357. Zarbin M. A., Wamsley J. K. and Kuhar M. J. (1981) Glycine receptors: light microscopic autoradiographic localization with eH]strychnine. 1. Neurosci. 5,532-547. (Accepted 1 May 1996)