Gaba nerve endings in the rat red nucleus combined detection with serotonin terminals using dual immunocytochemistry

neuroscience Printed

Vol. 23, No.

3, pp. 109~1102,

1987

030&4522/87

S3.00 + 0.00

PergamonJournalsLtd 0 1987IBRO

in Great Britain

GABA NERVE ENDINGS IN THE RAT RED NUCLEUS COMBINED DETECTION WITH SEROTONIN TERMINALS USING DUAL IMMUNOCYTOCHEMISTRY D. ANDR& G. VUIL~N-CACCKJTTO~ and 0. BOSLER* Laboratoire de Neurobiologie, C.N.R.S., B.P. 71, 13402-Marseille tidex 9, France Ab&rac-Immunocytochemical methods were used to examine the ultrastructural features and cellular interrelationships of GABA and serotonin a&rent fibers to the rat red nucleus. GABAergic nerve endings were identified in two ways, either using a pre-embedding immunoperoxidase procedure with an antibody against giutamate dec+rboxylase, the OABA-synthesizing enzyme, or after postembedding immunogold labelling with an anti-GABA antibody. With the latter approach, it was possible to simultaneously visuaiize the GABAergic and serotoninergic innervation of the red nucleus (magnocelhdar part) in electron microscope. preparations. This procedure involved GABA labelling of ultrathin sections obtained from specimens previously immunostained for serotonin using the pre-embedding peroxidase-antiperoxidasc technique. The doubly stained sections showed gold and peroxidase markers to be present in two distinct populations of axonal varicosities. Unlike the GABAergic nerve endings, which were found to be profusely distributed throughout the nucleus, the serotonin nerve endings were relatively scarce. They contacted dendrites of large-sized neurons usually endowed with several GABA-gold labelled terminals. Not uncommonly, direct appositions between serotonin and GABA-positive terminals were also encountered. These data provide morphological evidence that red nucleus outputs may be dually regulated by GABAergic and serotoninergic tierents, while suggesting that presynaptic GABA/serotonin interactions might also play a significant part in red nucleus fucntions.

The combined findings of several biochemical, pharmacological, electrophysiological and morphological studies show that GABA is a key inhibitory transmitter in the.red nucleus (RN). Relatively high levels of endogenous GABA, glutamate decarboxylase (GAD) and/or GABA-transaminase activity have been measured in the RN of rat, cat, monkey and man.‘9*2’*29*33*4 In the cat, a high affinity GABA uptake process by RN synaptosomes49 as well as a release of endogenous GABA from RN slic& have been found to occur, while the inhibitory effects of the transmitter have been documented electrophysiologically both in vivo and in vitro.1*9-32 GABA is thought to mediate the inhibition of rubrospinal neurons triggered by collateral activation of the wrticospinal tract.42*43GABAergic synapses have in fact been shown to exist on identified rubrospinal neurons.20 Moreover, two separate morphological studies on the cat using GAD immunohistochemistry have suggested that the GABAergic innervation of the RN might arise both from intrinsic interneurons and from the adjacent mesencephalic reticular formation.‘6*47 In addition, GAD-positive neurons within the RN have been shown to receive a direct input from the ipsilateral sensorimotor cortex.16 In the present study, the ultrastructural features and cellular relationships of the GABAergic inner*To whom correspondence should be addressed. Abbreviations: BSA, bovine serum albumin; GAD, glutamate decarboxylase; 5-HT, serotonin (5-hydroxytryptamine); IgG, immunoglobulin; PAP, peroxidase-antiperoxidasc; PBS, phosphate.-buffered saline; RN, red nucleus.

vation of the RN were investigated in the rat by complementary immunocytochemical labelling procedures employing antisera directed against GAD or against GABA itself. In addition, the fact that RN large-sized neurons also receive a substantial serotonin (5HT) tierent input, as demonstrated by radioautography in the cat after selective uptake of [3HJ5-HT,8 prompted us to design a dual immunocytochemical procedure for investigating the morphological substrate of possible S-HT/GABA cellular interactions in the RN. Part of these results have been presented at an international satellite symposium of the 10th annual European Neuroscience Association meeting (Hodology and Function of the Red Nucleus, Bandol, France).= EXPERIMENTALPROCEDURES Adult male SpragutDawley

rats were used in this study. They were perfused through the aortic arch, under chloral hydrate anesthesia (400 mg/kg i.p.), with either 4% paraformaldehyde+O.1% dutaraldehyde mixtures or with 3% glutaraldehyde, in 0.12 M phosphate btier (pH 7.4), depending on whether a GAD immunostaining or a combined 5-HT and GABA immunocytochemical labelling was to be carried out. A mesencephalic tissue block containing the RN was obtained from each brain and post-fixed for 1 h by immersion in the fixative. The RN was sampled in 30 brn thick frontal sections, using a Lancer Vibratomiz. T&e were collected in 0.5 M Tris saline, PH 7.6. and Drocessed. floating freely, for immunocytoch&istry &ng ihe doubld bridge peroxidascantiperoxidase (PAP) method,” as previously described.” Glutamate akcarboxylase immunocytochemistry The GAD antiserum, kindly provided by Dr M. Tappaz, was raised in sheep. Its production and characterization

1095

1096

D. ANDI& et al.

have been reported previously.rsJ6 Control immunoblotting experiments showed that this antiserum reveals two bands corresponding to the two GAD subunits previously labelled by an irreversible radioactive suicide substrate for GAD.*-’ GAD immunoreactivity was detected by incubating the sections as follows (room tem~rature): (a) 30 min in normal rabbit serum diluted 1:30; (b) I8 h in primary antiserum or in control preimmune serum diluted 1: 1000; (c) 30 min in rabbit anti-goat immunoglobulin (IgG) serum (Dakopatts) diluted 1:50; (d) 30min in aoat PAP serum (Dakooatts) diluted 1: 100; (ej and (f) 20 &in in rabbit anti-goat Ig& and 20 min in goat PAP diluted I : 100and 1: 200, respectively; (g) 6 min in 3,3’-diamino~nzidine (0. i mg/mI) and hydrogen peroxide (0.01%) in 0.5 M Tris buffer. Incubations with antisera and rinsing were carried out in the presence of 1% normal rabbit serum. Some sections were mounted on gelatin coated slides, dehydrated and coverslipped, for light microscopic examination. The others were post-fixed in 2% phosphate buffered osmium tetroxide (60min) and stained en 610~in 2% aqueous many1 acetate (45 min). After dehydration in a graded series of ethanols, these were eat-ern~dd~ in Epon 812 between two plastic coverslips and re-embedded in Beem capsules for ultrathin sectioning. Tissue blocks from the caudal part of the RN were selected and silver-gold sections were cut from the surface of each of them using a Reichert Ultracut Microtome. The sections were picked up on Formvar-coated grids, counterstained with lead citrate and examined with a Siemens Elmiskop 102 electron microscope.

after rinsing in three changes of distilled water, in 1% freshly prepared sodium metaperiodate for 7mm. Subsequently, they were washed for 5 min in 0.1 M phosphatebuffered saline (PBS) and processed for immunostaining according to the following sequence: (a) pre-incubation for 30min in a blocking buffer (PBS with I-J% BSA) aiso supplemented with 1% normai swine serum (Dakopatts) if the IgG-gold procedure was to be used; (b) 2 h incubation with GABA antiserum diluted 1:1000 in the blockine buffer: (c) 2 h incubation in sheep anti-rabbit IgG or in pktein A adsorbed to 5 or 15nm colloidal gold particles and diluted 1:50 in the corresponding blocking buffer; (d) rinsing in PBS (3 x 5 min) and in distilled water (3 x 5 min). In this step of the procedure, the grids were allowed to dry and tinally counterstained with lead citrate before electron microscopic examination. Staining controls were obtained by incubating some adjacent sections in either IgG gold or protein A-gold alone. RESULTS Glutamate

akcarboxylase ~mrn~~o~e~ctivi~y

Light microscope immunohistochemical preparations showed both somatal and axonal GAD immunostaining. As previously observed in the cat,47 a

well circumscribed group of positive neuronal perikarya was present rostrally within the rn~n~ph~i~ reticular formation bordering the RN dorsolaterally. Combined GABA and J-hydroxytryptamine immunocyto- Within the RN proper, labelled neuronal cell bodies chemistry were either only occasionally (rostra1 part of the nucleus) or completely absent (caudal part). Punctate Dual immunolabelling was carried out by combining a pre-embedding PAP staining for S-HT with a postreactions typical of positive boutons densely invested embedding ~rn~o~id staining for GABA. We used the whole rostrocaudal extent of the RN, formingrabbit GABA and S-HT antisera (I~unoF~h, Marseille), perisomatic and peridendritic arrangements, esraised using GABA or S-HT coupled to bovine serum pecially around large-sized neurons in the caudat albumin (BSA) with giutaraldehyde as immunogens (for part of the nucleus. Sections initially incubated with preparation and characterization of these. antisera, see Refs 12 and 36). The sections were first immunostained for 5-HT preimmune serum instead of GAD antiserum were using the same protocol as described above for GAD devoid of any specific immunostaining. immunolabeliing (difution of the primary antiserum: When viewed through the electron microscope, l/ZOO@),except that the blocking normal serum was from most GAD-positive profiles in the caudal part of the swine and the PAP serum from rabbit and that the secondary antiserum was a swine anti-rabbit IgG (Dakopatts). RN could be seen to consist of axonal varicosities 0.2 Method specificity was controlked by incubating some secto 1.3 pm in transverse diameter, filled with pleomortions with normal swine serum at the same dilution as that phic synaptic vesicles usually associated with mitoof the 5-HT antiserum. At the end of this PAP staining, chondria, and sometimes with large granular vesicles sections were osmicated, counterstained en bloc with uranyl (Fig. IA). The immunostaining was ~o~only found acetate and flat-embedded in Araldite, as described above. Ultrathin sections from selected blocks were picked up on to be associated with the outer rim of the synaptic Formvar-coated 200 mesh nickel grids for subsequent vesicles, the mitochondrial membrane and the inner GABA post-embedding immunogold staining. leaflet of the axolemma. A few preterminal axon Due to the fact that penetration of the antibodies was segments, identified by the absence of synaptic veslimited in the pre-embedding staining for 5-WT, dual immunolabelling was obtained only on the sections taken from icles, also displayed significant immunoreactivity. the outermost surface of the RN tissue blocks, containing Out of 235 GAD-positive varicosities examined tbe peroxidase reaction. The basic material for analysmg (care being taken to sample only once those visualGABA immunor~~~Fy was derived, however, not only ized on more than one section), not less than I29 from these doubly stained sections but also from sections (55%) were observed to be engaged in diflerentiated cut more deeply from the blocks (more than 2 pm from the surface). These were also used for test experiments designed synaptic junctional complexes, most of which were of to determine the most favourabb procedure for resin etch- the symmetrical variety. These synapses appeared to ing and osmium removal as well as the oPFim& working be preferentially formed on the proximal dendrites of dilution of the GABA immtmoserum and g&F reagents. large-sized neurons. GAD-positive terminals were These cons&Fed of e&her goat anti-IBM T@ aide or protein A (Pbarmaeia) adsorbed to 5 or 15 nm oolloidat Bold also synaptically linked to the cell bodies of these particles as described by De Meyio and Stat and Getme.” neurons as well as to small dendritic profiles possibly Resin etching and deosmieation were carried out as belonging to distal dendritic branches of large sized described by Somogyi und Hodgson,” with r&g&t modi& neurons or to dendrites of small sized neurons. A cations to their procedure. The grids were war&d in dirailled striking arrangement of several GAD-positive variwater and immersed for 5 min in I% periodic acid; then,

GABA and serotonin nerve endings in the red nucleus

cosities, clustered on a single somatal or dendritic target was often observed (Fig. 1A) and axoaxonic appositions between two GAD-immunoreactive presynaptic terminals were not rare. GABA ~~oreuct~ity After GABA immunostaining of electron microscope sections using either the IgG- or the protein A-gold procedure, concentrations of colloidal gold particles overlaid axonal varicosities which were profusely distributed throughout the RN and, occasionally, some profiles which might be termed intervaricose segments. This selective labelling was particularly intense with gold particles 5 nm in size (Fig. 1D and E; Fig. 2D) and could be easily discriminated from the background reaction. This reaction, which was again significantly higher with 5 nm than with 20 nm particles, was the only one to be observed after omission of the GABA antiserum in control experiments. The immunogold reaction overlaid the entire surface of the labelled varicosities. A particularly noteworthy finding when 5 nm rather than 20nm gold particles were used, however, was the fact that GABA immunoreactivity tended to be preferentially localized on mitochondrial profiles (Fig. 1D and E), which is in keeping with earlier findings on other brain Here the labelled varicosities exregions. L1,38,41*45 hibited the same morphological features as the GADpositive varicosities described above (mean diameter: 0.8 pm). They also established axosomatic and axodendritic synaptic contacts (Fig. lC, Fig. 2). In fact, some 63% of the 146 GABA-positive boutons sampled in single sections showed the membrane differentiation typically associated with such contacts. These synapses were essentially symmetrical and were mostly formed on large dendritic trunks

lo!?7

which usually received more than one immunolabelled afFerent (Fig. 1C and D; Fig. 2C and D). In the most favourable planes of section, large longitudinally cut proximal dendrites coufd be traced over a certain distance together with their characteristic arrays of synaptic immunoreactive terminals. Interestingly also, pictures were encountered in which two dendritic processes were simultaneously contacted by the same positive terminal (Fig. 1C). Serotonin i~ore~~t~v~ty. GABA ~~~oreuct~ity

Combined detection with

Ultrathin sections taken at. the outermost surface of the Vibratome tissue blocks showed peroxidase labelling of 5-HT immunoreactive varicosities in addition to the immunogold reaction (Fig. 2). In no case were any of these varicosities found to ti overlaid by a significant deposit of gold particles, iridicating that S-HT and GABA immunoreactive boutons belonged to separate axonal populations. The internal features of the 5-HT positive varicosities did not differ in any fundamental respects from those of their GABA positive counterparts. They ranged between 0.3 and I pm in transverse diameter and usually showed densely packed synaptic vesicles, which were occasionally associated with one or several immunoreactive dense core vesicles and with mitochondria. They came into direct contact with large dendrites on which GABA immunoreactive terminals were concomitantly detected (Fig. 2) but, unlike the latter, 5-HT positive varicosities were not observed to establish any clearly characterized synaptic junctional complexes. Interestingly also, although they were usually contacted by at least two gold labelled terminals, dendritic elements were never found to receive more than one peroxidase labelled afferent fibre. In fact, 5-HT immunopositive boutons

Fig. 1, GAD and GABA immunocytochemistry. (A) PAP immunostaining for GAD. A cell body fragment from a large-sized neuron receives two positively stained terminals in addition to several unlabelled tierent fibres. Most of the non-i~unor~tive and, presumably, one of the positive terminals are involved in visible synaptic junctional complexes (arrows). (B-C) &G-gold staining for GABA using l5nm gold particles. In micrograph (B), one of the two illustrated positive terminals and a nonimmunoreactive one exhibiting similar internal features are synaptically linked to the soma of a large-sized neuron with symmetrical junctional complexes. In micrograph (C), the positive axon terminals all contact large dendritic shafts (D, , D2) with symmetrical synapses. One of them is juxtaposed to both dendritic profiles and to another gold labelled terminal. (D-E) Protein A-gold labelling using 5 nm gold particles. Other examples of GABA immuno~~ll~ terminals in contact with dendritic elements. The positive profiles show ~~icul~iy heavy gold labelling which appears to be p~fe~nti~ly localized upon mitochondria. In micrograph (D), two of them are contiguous while terminating on one same dendritic trunk. Bars = 0.5 ym. Fig. 2. Combined 5-HT and GABA immunocytochemistry. (A-B) Consecutive sections in which a dendrite can be seen receiving two immunoreactive afferent fibres, the one GABA gold labelled (15 nm particles) and the other 5-HT peroxidase stained. The GABA immunoreactive terminal makes a symmetrical synaptic contact whereas no defined junctional complex is visible, on these planes of section, in the contact area between the 5-HT positive ending and the dendrite. (C-D) Other examples of dendritic elements simultaneously contacted by GABA and 5-HT immunoreactive nerve endings. The dendrite in micrograph C receives two GABA gold labelled tierents, one of which shows a synaptic junction and the other of which is juxtaposed to the 5-HT peroxidase labelled terminal. This type of axoaxonic apposition between the two differentially labelled endings is also visible in micrograph 0). (C) IgG-gold labelling with 15 nm particles; (D) protein A-gold labelling with 5 nm particles. Bars = 0.5 pm.

D. ANDRE:et al.

II00

were relatively scarce in comparison with GABA gold labelled counterparts. They were invariably isolated each from another, whereas frequently established structural relationships GABA positive boutons (Fig. 2C and D).

their also they with

DISCUSSION

Methodological considerations

Several authors have confirmed the validity of our GAD, GAISA, and 5-HT i~unola~l~ng procedures for s~fi~lly identifying GAEAergic or serotoninergic structures (see for example Refs 6, 12, 24, 35 and 40). Depending on whether GAD or GABA was to be detected, we used two different approaches for localizing GABAergic elements, namely a pm-embedding peroxidase procedure and a ~st~m~ding i~unogold procedure, respectively. The GABA immunolabelling procedure holds several advantages in addition to those which can be gained by using the classical post-embedding methods (see Ref. 30) and colloidal gold markers? these related to the characteristics of the GABA antibody. Due to the app~~bility of GABA immunostaining to specimens fixed with relatively high concentrations of glutaraldehyde, electron microscope preparations immunostained for GABA were more satisfactory than those immunostained for GAD as far as tissue preservation was concerned, while allowing a more accurate id~ti~~tion of the cellular relationships established by GABAergic terminals. Another advantage of post-embedding immunogold labelling was the fact that it could be successfully combined with pre-embedding peroxidase staining, so that it was possible to simultaneously detect GABA and S-HT immunoreacti~ties in a single electron microscope section. This was due to the fact that the fixation prerequisites were the same for the immunodetection of both antigens. The combination of pre- and post-embedding methods for dual immunolocalization of distinct tissular antigens constitutes a useful means of investigating chemical connectivity in the brain. This has already been illustrated in the hypothalamus, where post-embedding GABA immunostaining has been recently used with a double labelling approach similar to that used in the present study, to identify GABAergic synaptic inputs onto oxytocin and vasopressin neurons of the supraoptic nucleus.4’@ In the present study, while demonstrating that RN largesized neurons receive a dual GABAergic and serotoninergic innervation, we noted the fact that colloidal gold and peroxidase product immunodeposits invariably labelled separate axonal populations. The question therefore arises as to whether or not the pre-/~st-ern~~ng dual approach is valid for investigating potential cellular co-localization of the two substances to he immunodetected. With this procedure, the recognition of one antigenic determinant by the first specific antiserum might be ex-

pected to impede immunostaining for a second determinant if both are present in the same cellular element. In fact, if any GABA/S-HT axonal colocalization occurred in the RN, the proportion of GABAergic terminais with a presumed serotoninergic potentiality would certainly be very low since the GABAergic input to the nucleus is known to mainly originate from local intemeuronsi6.” whereas the 5-HT afferent fibres may originate from the midbrain raphe nuclei. Nevertheless, since GABA/ 5-HT co-localization has been d~ument~ in other brain areas,3,4*13it would be of special interest to formally investigate this issue in the RN, This should be feasible after immunostaining adjacent electron microscope sections for GABA and 5-HT (see Ref. 28) and/or by using combined immunocytochemistry and ~dioauto~~hy to simultaneously localize endogenous GABA or GAD and [‘H&FIT uptake sites (see Ref. 7). GABA and serotonin innervations of the red nucleus

The present data show GABAergic fibres to constitute a major afferent input to RN large-sized neurons. They thus confirm and extend previous morphological data obtained on the cat using GAD immunocytochemistry. N),47They are also in keeping with initial immunohistochemical observations after application to 5-HT antibodies, on the basis of which the RN was identified as a structure with a low to moderate serotoniner~c innervation in the rat,39 However, judging from previous radioautog~phic data obtained after selective uptake of [‘HIS-HT in the cat’ as well as in the rat (Descarries, personal communication), which demonstrated the existence of a dense plexus of serotoninergic 8bres in the RN, it is probable that a non-ne~~ble number of S-HT-cont~ning varicosities has escaped immunodetection, presumably for technical reasans inherent to the pre-embedding approach. Despite these limitations, the fact that RN neuronal elements were never found to be contacted by more than one serotoninergic terminal is of significant interest in that this feature contrasts with the considerable neuronal investment by GARAergic afferent fibres observed not only after GABA post-embedding but also after GAD pre-embedding immunolabelling. The relatively high frequency with which GABAergic immunoreactive profiles were found to establish weI!-dewed synaptic junctions suggests that all the GABAergic afferent fibres terminating in the RN may actually be synaptic. As for the 5-MT immunoreactive varicosities, the fact that they were not found to be engaged in differentiated synaptic junctional complexes does not necessarily mean that they never establish true synaptic contacts. There is still a possibility, indeed, that serotoninergic synaptic terminals, which have been found using [‘HIS-HT uptake radioautograph? to exist in the cat,’ have heen missed in the present study due to the heavy covering of 5-HT immunostained profiles by the peroxidase reaction

GABA and serotonin nerve endings in the red nucleus

1101

existence of structural relationships between homoendproduct, which made it difficult to establish logous GABAergic terminals as well as between whether the electron density associated with synaptic GABA-ergic and serotoninergic terminals, providing membrane differentiation was present. However, cytological support for the existence of axoaxonic even assuming that serotoninergic terminals involved interactions between both atTerent systems to the in true synaptic junctions do exist in the rat RN, their incidence would certainly be much lower than that of RN. The possible involvement of these presumed GARA/GABA and GABA/S-HT presynaptic intersynaptic GABAergic terminals, since many of these were unequivocally identified even after GAD im- actions in the regulation of motor functions has still to be investigated. munoperoxidase labelling. In any event, even nonjunctional S-IT varicosities are currently thought to be capable of releasing their endogenous t~s~tter ~cknowfedge~~ts-us study was supported by a grant (for review, see Ref. 2). Hence the present study from the Mini&&e de la Recherche et de la Technologie provides morphological evidence that RN large-sized (ATP 85.C. 1196). The authors are areatlv indebted to Dr neurons of a putative cholinergic identity’5*17*27*3’M. Tappaz who’generously supplid them with anti-GAD (but also see Refs 14, 34, 48), which are known to immunserum, to Dr A. Nieoullon for helpful comments and to Dr J. Blanc for her detailed revision of the English. The mainly give off the RN output system to the spinal technical assistance of Mr A. Yvinec with photographic cord, might undergo a duaf GABAergic and serowork and that of Mrs B. Derd&ian in preparing the manuscript is also gratefully acknowledged. toninergic control. Our observations aiso point to the

REFERENCES 1. Altmann H., Bruggencate G., Pickehnann P. and Steinberg R. (1976) Effects of GABA, glycine, picrotoxin and bicuculline me&chloride on rubrospinal neurons in cats. Brain Res. 111, 337-345. identifi~tion of serotonin neurons. In ~ono~~rgic Neurons: 2. Beaudet A. and Demarries L. (1987) ~t~t~~ur~ Light Microscopy and ~Itr~t~ture, IBRO Handbook Series: Met/m& in the N~ros~~n~es (ed. Steinbusch H. W. hi.), Vol. 10, pp. 265-313. Wiley, Chichester. 3. Belin M. F., Gamrani H., Aguera M., Calas A. and Pujol J. F. (1980) Selective uptake of [3Hlgamma-aminobutyrate by rat supra- and subependymal nerve fibers. Histological and high resolution radioautogaphic studies. Neuroscience $241-254. 4. Belin M. F., Nanopoulos D., Didier M., Aguera M., Steinbusch H., Verhofstad A., Maitre M. and Pujol J. F. (1983) I~uno~st~h~~ evidence for the presence of gab-~nobuty~c acid and serotonin in one nerve cell. A study on the raphe nuclei of the rat using antibodies to glutamate decarboxylase and serotonin. Bruin Res. 275, 329-339. 5. Bendayan M. (1984) Protein A-gold electron microscopic immunocytochemistry: Methods, applications and limitations. J. Electron Microsc. Techn. 1, 243-270. 6. B&rod A., Chat M., Paut I_..and Tappax M. (1984) Immunocytochemical visualization of tyrosine-hydroxylase (TH), phenylethanolamine N-methyl transferase (PNMT) and glutamic acid decarboxylase (GAD) on serial semi-thin plastic embedded sections in the rat substantia nigra, locus coeruleus and hypothalamic median eminence. J. Hisfochem. Cyrochem. 32, 1131-l 138. of chemically detined neurons and their cellular 7. Bosler O., Beaudet A. and Pickel V. M. (1986) Chara~t~~tion ~lationships by combined immun~yt~hemist~ and radioauto~ap~~ localization of transmitter uptake sites. J. EIectron Microsc. Techn. 4, 21-39. 8. Bosler O., Nieoullon A., Onteniente B. and Dusticier N. (1983) In vitro radioautographic study of the monoaminergic innervation of cat red nucleus. Identification of serotoninergic terminals. Bruin Res. 259, 288-292. 9. Davis R. and Vaughan P. C. (1969) Pharmacological properties of feline red nucleus. Znt. J. Neuropharmac. 8,475488. 10. De. Mey J. R. (1983) The preparation of immunoglobulin gold conjugates (IgG reagents) and their use as markers for light and electron microscopic immun~yt~hemis~y. In Immunohistochemistry, IBRO Handbook Series: Methods in the Neurosciences (ed. Cuello A. C.), Vol. 3, pp. 341-372. Wiley, Chichester. neurons in the rat 11. Gabbott P. L. A., Somogyi J., Stewart M. G. and Hamori J. (1986) GABA-i~~o~~ive cerebellum: A light and electron microscope study. J. camp. Neural. 251, 474-490. 12. Geffard M., Heinrich-Rock A. M., Dulluc J. and S&&la P. (1985) Antisera against small neurotransmitter-like molecules. Neurochem. Int. 7, 403-413. 13. Harandi M., Didier M., Aguera M., Calas A. and Belin M. F. (1986) GABA and serotonin (5-HT) pattern in the suuraenendymal fibers of the rat euithalamus: combined radioautographic and immunocytochemical studies. Effect of 5-HT &n&t on PHJGABA accumulation. Brain Res. 370, 241-249: 14. House C. R., Crawford G. D., Barber R. P., Salvaterra P. M. and Vaughn J. E. (1983) Organization and rno~holo~~ characteristics of choline@ neurons: an immunocytochemical study with a monoclonal antibody to choline acetyltransferase. Brain Res. 266, 97-l 19. 15. Jones B. E., Pare M. and Beaudet A. (1986) Retrograde labeling of neurons in the brain stem following injections of r HIcholine into the rat spinal cord. Neuroscience 18, 901-916. 16. Katsumaru HI., Murakami F., Wu J. Y. and Tsukahara N. (1984) GABAergic intrinsic intemeurons in the red nucleus of the cat demonstrated with combined immunocytochemistry and anterograde degeneration methods. Neurosci. Res. 1,35-44. 17. Kimura, H., McGee1 P. L., Peng J. H. and McGee1 E. G. (1981) The central cholinergic system studied by choline acetyltransferase immunohistochemistry in the cat. J. camp. Neurof. MO, 151-201. 18. Kubota M., Sakaguchi H. and Tsukahara N. (1983) Release of endogenous GABA from the cat red nucleus slices. Brain Res. 270, 190-192. 19. Muller P. B. and Langemann H. (1962) Distribution of glutamic acid decarboxylase activity in human brain. J. Nemo&em. 2, 399-401.

1102

D. ANDR& el ai.

20. Murakami F., Katsumaru H., Wu J. Y., Matsuda T. and Tsukahara N. (1983) Immunocytochemical demonstration of GABAerigc synapses on identified rubrospinal neurons. Brain iies. 267, 357-360. 21. Nieoullon AT and Dusticier N. (1981) Glutamate decarboxylase distribution in discrete motor nuclei in the cat brain. J. Neurochem. 37, 202-209. 22. Nieoullon A., Vuillon-Cacciuttolo G., Dusticier N., Kerkirian L., Andrt D. and Bosler 0. (In press) Putative neurotransmitters in the red nucleus and their involvement in post-lesion adaptive mechanisms. Behav. Brain Res. 23. Oertel W. H., Schmechel D. E., Daly J. W., Tappaz M. L. and Kopin I. .I. (1980) Localization of glutamate decarboxylase on tine-immunoet~trophor~sis and two-dimensional ekzctrophoresis by use of radioactive suicide substrate 2[3Hjgamma-acetylenic GABA. Life Sci. 27, 2 133-2 141. 24. Oertel W. H., Schmechel D. E., Mugnaini E., Tappaz M. L. and Kopin I. J. (1981) Immunocytochemical localization of glutamate decarboxylase in rat cerebellum with a new antiserum. Neuroscience 6, 2715-2735. 25. Oertel W. H., Schmechel D. E., Tappaz M. L. and Kopin I. J. (1981) Production of a specific antiserum to rat brain glutamic acid decarboxylase (GAD) by injection of an antigen-antibody complex. Neuroscience 6, 2689-2700. 26. Oertel W. H., Tappaz M. L., Kopin 1. J., Ransom D. H. and Schmechel D. E. (1980) Production of an antiserum to rat brain glutamate (GAD)/cysteine sulfinate (CSD) decarboxylase. Brain Res. Bull. 5, suppl. 2, 713-719. 27. Onteniente B. and Conde F. (1986) Improved choline-acetyltransferase immunohistochemistry in the red nucleus by blocking the axonal transport. Abstr. In Hodo!ogy and Function qfthe Red Nueteus, p. 43. INSERM Int. Symp., 1I--14 Sept. 1986, Bandol, France. 28. Pelletier G., Steinbusch H. W. M. and Verhofstad A. A. J. (1981) Immunoreactive substance P and serotonin present in the same dense-core vesicles. Nature 293, 71-72. 29. Perry T. L., Berry K., Hansen S., Diamond S. and Mok C. (1971) Regional distribution of aminoacids in human brain obtained at autopsy. J. Neurochem. 18, 513-519. 30. Polak J. M. and Varndell I. M. (1984) I~~o~~el~~ng~r Electron microscopy. Elsevier, Amsterdam. 31. Ross M. E., Park D. H., Teitelman G., Pickel V. M., Reis D. J. and Job T. H. (1983) Immunohistochemical lacalization of choline acetyltransferase using a monoclonal antibody: A radioautographic method. Newroscience IO, W-922. 32. Sakaguchi H., Kubota M., Nakamura M. and Tsukahara N. (1984) Effects of amino acids on cat red nucieus neurons in vitro. Expr Brain Res. 54, i Se-1 56. 33. Salvador R. A. and Atbers R. W. (1959) The dist~bution of glut~ic-g~a-~inobuty~~ transaminasc in the nervous system of the rhesus monkey. J. biol. Chem. 234, 922-925. 34. Satoh K., Armstrong D. M. and Fibiger H. C. (1983) A comparison of the distribution of central choline@ neurons as demonstrated by acetylcholinesterase pharmaco-histochemistry and choline acctyltransferase immuno&sto&emi$~ry. Brain Res. Bull. 11, 693-720. 35. S&g&la P., Gamrani H., Geffard, M., Calas A. and Le Moat M. (1985) UI~~truct~al immun~yt~hemistry of y-aminobutyrate in the cerebral and cerebellar cortex of the rat. Neuroscience 16, 865-874. 36. S&g&la P., Geffard M., Buijs R. M. and Le Moal M. (1984) Antibodies against y-aminobutyric acid: specificity studies and immunocytochemical results. Proc. nom. Acad. Sci. U.S.A. 81, 3888-3892. 37. Slot J. W. and Geuze H. J. (1983) The use of protein A-colloidal goId (PAG) complexes as immunolabels in ultra-thin frozen sections. In Zmm~oh~fochem~fr.y, IBRO Handbook Series: Me&o& in the N~ros~i~ces {ed. Cu&o A. C.), Vol. 3, pp. 323-346. Wiley, Chichester. 38. Somogyi P. and Hodgson A. J. (1985) Antisera to y-aminobutyric acid. III. Demonstration of GA3A in Go&iimpregnated neurons and in conventional electron microscopic sections of cat striate-cortex. J. H&&em. Cyta&w. 33, 249-257. 39. Steinb~ch H. W. M. (198 I) ~is~ibution of ~rotonin-~mmunoreacti~ty in the central nervous system of the rat. C&l bodies and terminals. Neuroscience 6, 557-618. 40. Tappaz M. L., Wassef M., Oertel W. H., Paut L. and Pujol J. F. (1983) Light and electron microscopic immunocytochemistry of glutamic acid decarboxylase (GAD) in the basal hypothalamus: morph&gical &dance for neuroendocrine gamma-aminobutyrate (GABA). Neuroscience 9, 271-287. analysis of the GABA~F& inflation of 41. Theodosis D. T., Paut L. and Tappaz M. L. (1986) ~~un~~~~c~ oxytocin- and vasopressin-secreting neurons in the rat snpraoptic nucleus. NrurosCence 19, 267--222. 42. Tsukahara N., Fuller D. R. G. and Brooks V. B. (1968) Collateral pyramidal in&ten= of the co&o-rubrospinal system. J. Neurophysiol. 31, 467-484. 43. Tsukahara and Kosaka K. (1968) The mode of cerebral excitation of red nucleus neurons. Ex@ Brain Res. 5,102-d 17. (ed§ 44. Vacca L. L. (1982) “Doubte bridge” techniques of ~un~~~~~~. fn Techie in I ~~~~c~~r~ Bullock G. R. and Petrusz P.), pp. 156-182. Academic Press, London. 45. Van den Pol A. N. (1985) Dual ultrastructural localization of two neurotranamitter-related antigens: colloi&tl gold-labelled neurophysin-immunoreactive supraoptic neurons receive peroxidase-tabelled ghttamate dsoarbox~~se- or gold-labelled GABA-immunoreactive synapses. 3. Netuosci. 11, 2940-2954. 46. Van der Heyden J. A. M., De Kloet E. R., Korf J. and Versteeg D. H. G. (1979) GABA content of disc@e brain nuclei and spinal cord of the rat. J. Newochem. 33, 857-861. 47. Vuillon-Cacciuttolo G., Bosler 0. and Nieouflon A. (1984) GABA neurons in the cat red nucleus: a biochemical and immunohistochemical demonstration. Neurosci. Left. 52, 129-134. 48. Wainer B. H., Levey A. I., Mufson E. J. and Mesulam M. M. (1984) Cholinergic systems in mammalian brain id&fled with antibodies against choline acetyltransferase. Neurocliem. Inr. 6, 163-182. &&mate uptake in s&cortical areas &&-ted 49. Young A. B., Rromberg M. B. and Penney J, B. (I981) m by sensorimotor cortical ablation in the cat. J. Neztrosci. 1, 241-249. (Accepted I1 Muy 1987)