- Email: [email protected]
Retrograde axonal transport following injection of [3H]-serotonin into the olfactory bulb. II. Radioautographic study
Brain Research, 196 (1980) 417-427 © Elsevier/North-Holland Biomedical Press
417
R E T R O G R A D E A X O N A L T R A N S P O R T F O L L O W I N G I N J E C T I O N OF [aH]S E R O T O N I N INTO T H E O L F A C T O R Y BULB. II. R A D I O A U T O G R A P H I C STUDY*
SILVIA ARANEDA, HALIMA GAMRANI, CHANTAL FONT, ANDRI~ CALAS, JEANFRANCOIS PUJOL and PIERRE BOBILLIER INSERM U 171, Laboratoire de Neurochimie Fonctionnelle, Ddpartement de M~decine Exp&imentale, Universitd Claude Bernard, 8, Avenue Rockefeller, 69008 Lyon and (H.G. and A.C.) CNRS, INP 3 Marseille, 31 Chemin J. Aiguier, 13009 Marseille (France)
(Accepted March 6th, 1980) Key words: serotonin -- retrograde transport - - raphe dorsalis -- olfactory bulb
SUMMARY Radioautography was used to study the intraneuronal distribution of [aH]serotonin (5-HT) and/or its derivatives selectively taken up by the olfactory bulb (OB) serotonergic terminals and subsequently transported to their parent cell bodies in the midbrain raphe nuclei. This was done 24 h after injection of [3H]5-HT into the main OB of rats either pretreated or not with monoamine oxidase (MAO) inhibitor. A prior mechanical obstruction of the rostral ventricular cavities prevented diffusion of the tracer towards cerebrospinal fluid. Heavily labelled nerve cell bodies were found mainly in the ipsilateral raphe dorsalis nucleus (RDN) and to a lesser extent in the raphe centralis nucleus. The radioautographic reaction often extended to dendritic processes while sparing the nucleus. A diffuse reaction was also observed but limited to the raphe area. The supraependymal 5-HT fibers were found to be free of labelling. Neither local destruction of catecholaminergic terminals with 6-OHDA, nor absence of MAO inhibition, impaired this radioautographic pattern, while destruction of serotonergic terminals with 5,6-dihydroxytryptamine in OB resulted in the disappearance of labelled axonal varicosities and neurons in the OB and the R D N respectively. At the electron microscope level, labelled cell bodies in the R D N were mediumsized (12-15 #m). Silver grains were localized mainly on mitochondria and, to a lesser * Presented in part at the Proceedings of the 4th International Catecholamine Symposium, Pacific Grove, Calif., September 17-22, 1978, and at the International Society for Neurochemistry, Satellite Symposium, Athens, Greece, September 11-16, 1979.
418 extent, on lysosomes and endoplasmic reticulum but spared the nucleus and the nucleolus. Silver grains were also found near the nuclear membrane and outside the neuronal membrane. The observation of heavy metal impregnated thick sections confirmed the preferential localization of silver grains on mitochondria with or without inhibition of MAO. These results could account for the subcellular compartments involved in the retrograde axonal transport of [aH]5-HT and its subsequent degradation and/or dendritic release.
INTRODUCTION Previous biochemical results (see part I) have demonstrated a retrograde axonal flow of [3H]serotonin (5-HT) from the main olfactory bulb (OB) to the midbrain raphe nuclei. Pharmacological studies have suggested that this phenomenon was taking place within the 5-HT neurons originating in the raphe nuclei and projecting to the OB. In the present study, radioautography at the light microscope level was used to study the distribution of labelled elements in the OB and the midbrain raphe nuclei following injection of [aH]5-HT in the OB. Further, electron microscope radioautography was performed to investigate the intraneuronal distribution of [SH]5-HT and/or its derivatives retrogradely transported. MATERIALSAND METHODS Radioautographic experiments were performed on 16 rats. A prior mechanical obstruction of the olfactory diverticle of the lateral ventricle was carried out in all animals (for details see part I). A uni- or bilateral injection of [aH]5-HT (Amersham, 8-12 Ci/mmol, 25 or 100 #Ci in 1/~1 of saline) was placed in the OB of rats pretreated with a monoamine oxidase (MAO) inhibitor (Catron, fl-phenylisopropyl hydrazine, 10 mg/kg i.p., 60 min before). All the animals were reanesthetized with Nembutal 24 h after the administration of the tracer and processed for aortic-arch perfusion of the brain. For light microscope radioautography, 8 rats were used. In two animals, 5,6dihydroxytryptamine creatinine sulphate (5,6-DHT, Regis Chemical, 4 #g in 1 #1 of saline) was injected into the right OB. Two other rats were injected into the right OB with 6-hydroxydopamine (6-OHDA, Sigma, 2 #g in 1 #1 of saline with 0. t ~ ascorbic acid added). 5,6-DHT and 6-OHDA pretreatments were respectively made 6 and 10 days before the injection of [aH]5-HT into the OB. Transcardiac perfusion8 was carried out with glutaraldehyde (3.5~) in phosphate buffer (0.12 mM, pH 7.3). The whole brain was dehydrated in ethanol and embedded in paraffin-paraplast (1:1). Sections (7 #m) were then taken at every 100 #m interval, coated by dipping in Ilford K5 emulsion diluted 1:1, exposed for 10-30 days, developed with Phenisol (Ilford) and stained with cresyl violet.
419 For electron microscope radioautography, 8 rats were used. In 4 animals Catron pretreatment was omitted. Transcardiac perfusion was carried out with glutaraldehyde (3 ~) in cacodylate buffer (0.2 M, pH 7.4). The raphe was dissected out and immersed in uranyl acetate (5 ~) water solution at 37 °C for 1 h and then treated with lead and copper citrate solution according to the staining technique of Thiery and Rambourg is. The tissue was postfixed overnight in 1 ~ osmium tetroxide in water, dehydrated in ethanol and embedded in Epon. Thick sections (0.5 #m) were placed on celloidincovered slides and dipped in Ilford L4 emulsion (dilution 1:4), exposed for 40 days and developed in Microdol or Phenidon. Then, they were thinned in isoamyl acetate and examined with Elmiskop 102 Siemens electron microscope. Thin sections (70 nm) placed on celloidin film were stained with Reynolds lead citrate before radioautographic treatment as above described. Quantitative study of silver grain distribution was analyzed according to Salpeter et al. 16. RESULTS
Light microscopic radioautography of the injection site An intense radioautographic reaction was observed in the immediate vicinity of the needle track (Fig. 1m) within the internal and external granular layers of the OB. The intensity of the diffuse reaction gradually decreased while extending to the plexiform and the glomerular layers. A small number of dense dusters of silver grains typical of axonal varicosities were observed in the glomerular layer but also in the plexiform and the granular layers (Fig. 1B). Dense accumulations of silver grains were noticed over some nerve cell bodies. No significant reaction was visible in the anterior olfactory nucleus nor in the contralateral OB. Following 5,6-DHT treatment, the diffuse reaction persisted while the reactive axonal sites were abolished (Fig. ID). Following 6-OHDA treatment, the pattern of the radioautographic distribution was similar to that seen in controls (Fig. 1C).
Light microscopic radioautography of the midbrain raphe nuclei In the midbrain raphe area which was reached by the tracer, two types of radioautographic reactions were observed: a diffuse reaction of scattered silver grains surrounding dense accumulations of silver grains localized at the nerve cell bodies. These latter preferential intraneuronal accumulations of silver grains often extended to dendritic processes (Fig. 2B). The nuclei of reactive cells were nearly free of labeling; numerous surrounding perikarya were completely devoid of silver grains. No localized typical axonal reactions could be observed even over the supraependymal surface (Fig. 2D). The majority of labelled cells was localized throughout the anatomical area of midbrain raphe nuclei. Clusters of labelled cells (6-12) were mainly observed in the lateral and medial region of the rostral RDN. Only a small proportion of the neuronal population of the RDN was labelled. Heavily labelled cells were also present, but to a lesser extent, in the raphe centralis nucleus (RCN) (Fig. 2C). A few labelled cells were observed in the periaqueductal substantia grisea, pars centralis (Fig. 2D).
420
Figs. 1. A - D : light microscope radioautographs of the olfactory bulb 24 h after a unilateral intrabulbar injection of [~H]5-HT (25 pCi). Exposure time: 10 days. A: distribution of radioactivity in the injection site, A profuse dispersion of labelling is seen in the glomerular layer while a dense labelling is located in the internal plexiform and granular layers. B : localization of nerve terminals (arrows) in the glomerular, plexiform and granular layers in the untreated rat. C : after 6-OHDA treatment of the olfactory bulb, nerve terminals are still observed in the glomerular layer (arrows). D: 5,6-DHT pretreatment of the olfactory bulb abolished labelled nerve terminals in the glomerular layer. 30 (A) and 560 ~ (B, C, D).
421
7
!
i
Figs. 2. A - D : light microscopic radioautographs 24 h after a unilateral injection of [3H]5-HT (100 pCi) in the olfactory bulb. Exposure time: 1 month. A: ipsilateral labelled cell bodies in the raphe dorsalis nucleus. Medial longitudinal bundle, MLB. B: retrogradely labelled cell bodies in the raphe dorsalis nucleus. Note th.e positive labelling of dendrites. C: retrogradely labelled cell bodies in the raphe centralis nucleus. D : reactive cell body in the periaqueductal substantia grisea. No diffuse reaction is observed over the ependymal layer of the aqueduct (Aq). 80 x (A) and 520 × (B, C, D),
422
Figs. 3. A-C: light microscopic radioautographs 24 h after injection of [3H]5-HT (25/zCi) into the olfactory bulb. Exposure time: 1 month. A: after 6-OHDA treatment of the right olfactory bulb anq subsequent unilateral injection of [aH]5-HT in the treated olfactory bulb, positive-labelled cell bodies are seen in the ipsilateral raphe dorsalis nucleus. Medial longitudinal bundle, MLB. B and C: after 5,6-DHT treatment of the right olfactory bulb and following subsequent bilateral injection of [:~H]5-HT into the olfactory bulb, positive-labelled cell bodies are seen only on the left side of the midline (vertical dotted line) raphe dorsalis nucleus. B : labelled cells in the lateral part of raphe dorsalis nucleus. C: labelled cells in the medial part of raphe dorsalis nucleus. 350 y (A, B, C).
Following a unilateral [aH]5-HT injection in the OB, the distribution o f reactive cells occurred in the ipsilateral R D N (Fig. 2A). Only a few scattered labelled cells could be observed on the contralateral R D N . The radioautographic pattern and the t o p o g r a p h y o f labelled nerve cell bodies were similar either in the absence o f C a t r o n pretreatment or after 6 - O H D A injection in the OB (Fig. 3A). The radioautographic reactions completely disappeared in the ipsilateral R D N following unilateral 5 , 6 - D H T administration in the OB (Fig. 3B and C).
Electron microscopic radioautography of RDN Observation o f thin sections o f the R D N confirm dethe complete lack o f
423 labelling o f the supraependymal fibers (Fig. 4A). The diffuse reaction was very low t h r o u g h o u t the whole R D N . Against this absence o f background, some neurons appeared conspicuously labelled (Fig. 4B and C). These were medium-sized neurons (12-15 # m ) with indented nuclei. Silver grains were sparsely distributed on the perikarya, and in general spared the nuclei, but frequently overlapped the mitoch-
Figs. 4. A-C: high resolution radioautographs of thin sections of rat raphe dorsalis 24 h after injection of [aH]5-I-1T(100/~Ci) into the olfactory bulb. Exposure time: 40 days. A: ependyma of the 4th ventricle: supraependymal fibers (arrows) are completely devoid of labelling. Microdol development. B and C: labelled neurons: silver grains are mainly localized on mitochondrial profiles (M) and on lysosomes (Ly). The nucleus is not labelled but some silver grains overlap the nuclear membrane in C (arrow) and its immediate vicinity. Phenidon development. 6000 × (A), 7500 x (B) and 10,000 × (C).
Figs. 5. A - C : high resolution radioautographs of thick sections of Ur-Cu-Pb-stained raphe dorsalis, 24 h after injection of [3H]5-HT (100/iCi) into the olfactory bulb of Catron-pretreated (A-C) or untreated (B) rats. (Phenidon developer). A: labelled neuron at low magnification : silver grains clearly spare the tangential nucleus profile (N) and seem to be selectively localized on mitochondria (M). Endoplasmic reticulum: ER. An adjoining neuron is devoid of labelling (*). B and C: Silver grains seem to be preferentially localized on the mitochondria. 4500 ~ (A, B) and 6000 (C).
425 ondria and lysosomes. Some grains were observed near the neuronal nuclear membrane and also outside the cell, in its vicinity. Observation of heavy metal impregnated thick sections of the RDN showed cell bodies which displayed more numerous silver grains with respect to the low surrounding diffuse reaction (Fig. 5). These mediumsized neurons displayed intensely stained mitochondria which appeared elongated and sometimes branched, and the endoplasmic reticulum (ER) extended as a continuous structure across the whole perikaryon. In the radioautographic study of these thick sections, silver grain distribution with respect to mitochondria was analyzed according to Salpeter et al. 16. In Catronpretreated animals (Fig. 5A and C), 74 ~ of the silver grains occurred within 12 nm, measured from the edge of the grain to the membrane of the nearest mitochondria, and 93 ~ within 50 nm. In untreated animals (Fig. 5B), 53 ~o of the silver grains occurred within 12 nm of the nearest mitochondria and 88 ~ within 50 nm. Under both experimental conditions, grains not localized on mitochondria occurred on the ER. There were few, if any, grains overlying the nuclei. DISCUSSION In this study, the lack of labelling of periaqueductal 5-HT fibers, the restricted area of diffusion, and the small proportion of labelled cell bodies in the RDN strongly contrast with the pattern of labelling observed following intraventricular administration of [aH]5-HT4,10. In this latter case, a gradient of general diffuse reaction is observed in the periaqueductal region which displays intensely labelled supraependymal fibers or nerve terminals, combined with a third of the RDN labelled neurons 7. In our model, the radioautographic results confirm the biochemical finding(s) (see part I) and exclude the possibility that the tracer could reach the RDN by ventricular diffusion. Previous studies have demonstrated the ipsilateral projections of the RDN 6 and their bilateral organization13. In fact, after a single injection of [ZH]5-HT into the OB, the RDN was ipsilaterally labelled. This strongly suggests that [3H]5-HT is retrogradely transported by raphe fibers and/or terminal afferents to the OB. In addition, the labelling of the RDN was suppressed when the terminals in the OB were locally destroyed by 5,6-DHT, while local administration of 6-OHDA failed to impair the RDN labelling. These observations and previous biochemical results (see part I) show that the retrogradely labelled cells observed in the RDN are serotonergic. Moreover, 5-HT cell bodies have been previously well defined in these areas (B7 and Bs cell groups) using histofluorescence5, light and high resolution radioautographyT,10 or immunocytochemistry15. The topography and morphology of the [3H]5-HT retrogradely labelled neurons correspond to these serotonergic groups. The subcellular distribution of retrogradely transported [aH]5-HT or derivatives was also found to be different from that observed when the 5-HT cell bodies are labelled following either ventricular or intratissular injection of [aH]5-HT. In this case, the nuclei and nucleoli were strongly labelled10 but the animals were sacrificed after 3 h of [3H]5-HT administration. In our experiments, however, the maximum accumula-
426 tion of transported 5-HT was found in the RDN, 24 h after [aH]5-HT injection into the OB (see part I) and the nuclei and nucleoli were practically free of silver grains. Such a difference in the pattern of labelling could correspond either to a time-dependent evolution of the cellular distribution of 5-HT and/or to a different intracellular compartmentation of the transported 5-HT. As shown by our quantitative approach, the labelling in both Catron-treated and untreated animals seems to be principally on the mitochondria 1~. The same preferential mitochondrial labelling was observed following direct uptake of the [3H]5-HT in the absence of MAO inhibition and within 3 h of tracer administration 10. This similar localization could suggest that the major part of the retrogradely transported 5-HT could have access to mitochondrial MAO in the cell bodies. Further support for this view is given by the fact that only about 30 0/o of the total radioactivity was detected as unmetabolized [~H]5-HT (see part I) in this structure. The biochemical results (see part I) have suggested that [3H]5-HT, injected into the OB, could be transported in several compartments. A kinetic quantitative analysis of radioactivity localization within serotonergic axons would be necessary to identify these subcellular pools. In addition to the mitochondria, labelling of the ER and lysosomes by retrogradely transported 5-HT or metabolites has been observed. These latter two compartments have been postulated to be concerned with, respectively, the mechanism of retrograde transport and cellular degradation of retrogradely transported macromolecular material2,9,11,17. However, the involvement of the agranular reticulum in retrograde transport of HRP remains controversial3,1L This does not, however, presuppose the identity of the intracellular pathways involved in transporting either H R P or 5-HT and/or its derivatives pinocytosed into the cell. The continuity between perikaryal and axonal labelled organelles has to be established. Finally, the labelling of dendrites, somatic membranes and their surrounding areas suggests that a fraction of the transported 5-HT could subsequently be released into the extracellular space. This accords with the hypothesis of a 'dendritic release' already suggested for dopamine in dopaminergic dendrites in the substantia nigra 1,14,L% ACKNOWLEDGEMENTS The authors wish to thank Dr. Lucienne Leger and Dr. Laurent Descarries for helpful discussions. This work was supported by INSERM U 171 (CRL no. 77 5 110 6) and INSERM U 52, D G R S T (Grant 78 7 2782) and CNRS INP 08 Marseille. S.A. was a recipient of a fellowship from the government of France. REFERENCES 1 Bjrrklund, A. and Lindvall, O., Dopamine in dendrites of substantia nigra neurons: suggestions for a role in dendritic terminals, Brain Research, 83 (1975) 531-537. 2 Broadwell, R. D. and Brightman, M. W., Cytochemistry of undamaged neurons transporting exogenous protein in vivo, J. comp. NeuroL, 185 (1979) 31-74. 3 Broadwell et al., Neuronal transport of acid hydrolase and peroxidase within the lysosomal system of organelles: possible involvement of agranular reticulum like cisterns, J. comp. Neurol., in press.
427 4 Chan-Palay, V., Serotonin axons in the supra and subependymal plexuses and in the leptomeninges; their roles in local alterations of cerebrospinal fluid and vasomotor activity, Brain Research, 102 (1976) 103-130. 5 Dahlstr6m, A. and Fuxe, K., Evidence for the existence of monoamine-containing neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of brain stem neurons, Acta physiol, scand., Suppl. 232 (1964) 1-55. 6 De Pimps, J., Hardy, H. and Heimer, L., The afferent connections of the main and the accessory olfactory bulb formations in the rat: an experimental HRP-study, J. comp. Neurol., 181 (1978) 213-244. 7 Descarries, L., Beaudet, A., Watkins, K. and Garcia, S., The serotonin neurons in nucleus raphe dorsalis of adult rat, Anat. Rec., 193 (1979) 520. 8 Descarries, L. et Schroder, J. M., Fixation du tissu nerveux par perfusion a grand dSbit, J. Microsc., 7 (1968) 281-286. 9 Droz, B., Rambourg, A. and Koenig, H. L., The smooth endoplasmic reticulum: structure and role in the renewal of axonal membrane and synaptic vesicles by fast axonal transport, Brain Research, 93 (1975) 1-13. 10 Gamrani, H. and Calas, A., Cytochemical, stereological and radioautographic studies of rat raphe neurons, J. Microsc., in press. 11 LaVail, J. H. and LaVail, M. M., The retrograde intraaxonal transport of horseradish peroxidase in the chick visual system: a light and electron microscopic study, J. comp. Neurol., 157 (1974) 303-358. 12 LaVail et al., Evidence against the smooth endoplasmic reticulum as a continuous channel for the retrograde axonal transport of HRP, Brain Research, in press. 13 Levitt, P. and Moore, R. Y., Developmental organization of raphe serotonin neuron groups in the rat, Anat. Embryol., 154 (1978) 241-251. 14 Nieoullon, A., Cheramy, A. and Glowinski, J., Release of DA in vivo from cat substantia nigra, Nature (Lond.), 266 (1977) 375-377. 15 Pickel, V., Joh, T. H. and Reis, D., Monoamine-synthesizing enzymes in central dopaminergic, noradrenergic and serotonergic neurons. Immunocytochemical localization by light and electron microscopy, J. Histochem. Cytochem., 24 (1976) 792-806. 16 Salpeter, M. M., Bachmann, L. and Salpeter, E. E., Resolution in electron microscope radioautography, J. Cell Biol., 41 (1969) 1-20. 17 Sotelo, C. and Riche, D., The smooth endoplasmic reticulum and the retrograde and fast orthograde transport of horseradish peroxidase in the nigro-striatonigral loop, Anat. Embryol., 146 (1974) 209-218. 18 Thierry, G. and Rambourg, A., A new staining technique for studying thick sections in the electron microscope, J. Microsc. Biol. Cell., 26 (1976) 103-106. 19 Wilson, C. J., Groves, P. M. and Fifkova, E., Monoaminergic synapses, including dendro-dendritic synapses in the rat substantia nigra, Exp. Brain Res., 30 (1977) 161-174.
417
R E T R O G R A D E A X O N A L T R A N S P O R T F O L L O W I N G I N J E C T I O N OF [aH]S E R O T O N I N INTO T H E O L F A C T O R Y BULB. II. R A D I O A U T O G R A P H I C STUDY*
SILVIA ARANEDA, HALIMA GAMRANI, CHANTAL FONT, ANDRI~ CALAS, JEANFRANCOIS PUJOL and PIERRE BOBILLIER INSERM U 171, Laboratoire de Neurochimie Fonctionnelle, Ddpartement de M~decine Exp&imentale, Universitd Claude Bernard, 8, Avenue Rockefeller, 69008 Lyon and (H.G. and A.C.) CNRS, INP 3 Marseille, 31 Chemin J. Aiguier, 13009 Marseille (France)
(Accepted March 6th, 1980) Key words: serotonin -- retrograde transport - - raphe dorsalis -- olfactory bulb
SUMMARY Radioautography was used to study the intraneuronal distribution of [aH]serotonin (5-HT) and/or its derivatives selectively taken up by the olfactory bulb (OB) serotonergic terminals and subsequently transported to their parent cell bodies in the midbrain raphe nuclei. This was done 24 h after injection of [3H]5-HT into the main OB of rats either pretreated or not with monoamine oxidase (MAO) inhibitor. A prior mechanical obstruction of the rostral ventricular cavities prevented diffusion of the tracer towards cerebrospinal fluid. Heavily labelled nerve cell bodies were found mainly in the ipsilateral raphe dorsalis nucleus (RDN) and to a lesser extent in the raphe centralis nucleus. The radioautographic reaction often extended to dendritic processes while sparing the nucleus. A diffuse reaction was also observed but limited to the raphe area. The supraependymal 5-HT fibers were found to be free of labelling. Neither local destruction of catecholaminergic terminals with 6-OHDA, nor absence of MAO inhibition, impaired this radioautographic pattern, while destruction of serotonergic terminals with 5,6-dihydroxytryptamine in OB resulted in the disappearance of labelled axonal varicosities and neurons in the OB and the R D N respectively. At the electron microscope level, labelled cell bodies in the R D N were mediumsized (12-15 #m). Silver grains were localized mainly on mitochondria and, to a lesser * Presented in part at the Proceedings of the 4th International Catecholamine Symposium, Pacific Grove, Calif., September 17-22, 1978, and at the International Society for Neurochemistry, Satellite Symposium, Athens, Greece, September 11-16, 1979.
418 extent, on lysosomes and endoplasmic reticulum but spared the nucleus and the nucleolus. Silver grains were also found near the nuclear membrane and outside the neuronal membrane. The observation of heavy metal impregnated thick sections confirmed the preferential localization of silver grains on mitochondria with or without inhibition of MAO. These results could account for the subcellular compartments involved in the retrograde axonal transport of [aH]5-HT and its subsequent degradation and/or dendritic release.
INTRODUCTION Previous biochemical results (see part I) have demonstrated a retrograde axonal flow of [3H]serotonin (5-HT) from the main olfactory bulb (OB) to the midbrain raphe nuclei. Pharmacological studies have suggested that this phenomenon was taking place within the 5-HT neurons originating in the raphe nuclei and projecting to the OB. In the present study, radioautography at the light microscope level was used to study the distribution of labelled elements in the OB and the midbrain raphe nuclei following injection of [aH]5-HT in the OB. Further, electron microscope radioautography was performed to investigate the intraneuronal distribution of [SH]5-HT and/or its derivatives retrogradely transported. MATERIALSAND METHODS Radioautographic experiments were performed on 16 rats. A prior mechanical obstruction of the olfactory diverticle of the lateral ventricle was carried out in all animals (for details see part I). A uni- or bilateral injection of [aH]5-HT (Amersham, 8-12 Ci/mmol, 25 or 100 #Ci in 1/~1 of saline) was placed in the OB of rats pretreated with a monoamine oxidase (MAO) inhibitor (Catron, fl-phenylisopropyl hydrazine, 10 mg/kg i.p., 60 min before). All the animals were reanesthetized with Nembutal 24 h after the administration of the tracer and processed for aortic-arch perfusion of the brain. For light microscope radioautography, 8 rats were used. In two animals, 5,6dihydroxytryptamine creatinine sulphate (5,6-DHT, Regis Chemical, 4 #g in 1 #1 of saline) was injected into the right OB. Two other rats were injected into the right OB with 6-hydroxydopamine (6-OHDA, Sigma, 2 #g in 1 #1 of saline with 0. t ~ ascorbic acid added). 5,6-DHT and 6-OHDA pretreatments were respectively made 6 and 10 days before the injection of [aH]5-HT into the OB. Transcardiac perfusion8 was carried out with glutaraldehyde (3.5~) in phosphate buffer (0.12 mM, pH 7.3). The whole brain was dehydrated in ethanol and embedded in paraffin-paraplast (1:1). Sections (7 #m) were then taken at every 100 #m interval, coated by dipping in Ilford K5 emulsion diluted 1:1, exposed for 10-30 days, developed with Phenisol (Ilford) and stained with cresyl violet.
419 For electron microscope radioautography, 8 rats were used. In 4 animals Catron pretreatment was omitted. Transcardiac perfusion was carried out with glutaraldehyde (3 ~) in cacodylate buffer (0.2 M, pH 7.4). The raphe was dissected out and immersed in uranyl acetate (5 ~) water solution at 37 °C for 1 h and then treated with lead and copper citrate solution according to the staining technique of Thiery and Rambourg is. The tissue was postfixed overnight in 1 ~ osmium tetroxide in water, dehydrated in ethanol and embedded in Epon. Thick sections (0.5 #m) were placed on celloidincovered slides and dipped in Ilford L4 emulsion (dilution 1:4), exposed for 40 days and developed in Microdol or Phenidon. Then, they were thinned in isoamyl acetate and examined with Elmiskop 102 Siemens electron microscope. Thin sections (70 nm) placed on celloidin film were stained with Reynolds lead citrate before radioautographic treatment as above described. Quantitative study of silver grain distribution was analyzed according to Salpeter et al. 16. RESULTS
Light microscopic radioautography of the injection site An intense radioautographic reaction was observed in the immediate vicinity of the needle track (Fig. 1m) within the internal and external granular layers of the OB. The intensity of the diffuse reaction gradually decreased while extending to the plexiform and the glomerular layers. A small number of dense dusters of silver grains typical of axonal varicosities were observed in the glomerular layer but also in the plexiform and the granular layers (Fig. 1B). Dense accumulations of silver grains were noticed over some nerve cell bodies. No significant reaction was visible in the anterior olfactory nucleus nor in the contralateral OB. Following 5,6-DHT treatment, the diffuse reaction persisted while the reactive axonal sites were abolished (Fig. ID). Following 6-OHDA treatment, the pattern of the radioautographic distribution was similar to that seen in controls (Fig. 1C).
Light microscopic radioautography of the midbrain raphe nuclei In the midbrain raphe area which was reached by the tracer, two types of radioautographic reactions were observed: a diffuse reaction of scattered silver grains surrounding dense accumulations of silver grains localized at the nerve cell bodies. These latter preferential intraneuronal accumulations of silver grains often extended to dendritic processes (Fig. 2B). The nuclei of reactive cells were nearly free of labeling; numerous surrounding perikarya were completely devoid of silver grains. No localized typical axonal reactions could be observed even over the supraependymal surface (Fig. 2D). The majority of labelled cells was localized throughout the anatomical area of midbrain raphe nuclei. Clusters of labelled cells (6-12) were mainly observed in the lateral and medial region of the rostral RDN. Only a small proportion of the neuronal population of the RDN was labelled. Heavily labelled cells were also present, but to a lesser extent, in the raphe centralis nucleus (RCN) (Fig. 2C). A few labelled cells were observed in the periaqueductal substantia grisea, pars centralis (Fig. 2D).
420
Figs. 1. A - D : light microscope radioautographs of the olfactory bulb 24 h after a unilateral intrabulbar injection of [~H]5-HT (25 pCi). Exposure time: 10 days. A: distribution of radioactivity in the injection site, A profuse dispersion of labelling is seen in the glomerular layer while a dense labelling is located in the internal plexiform and granular layers. B : localization of nerve terminals (arrows) in the glomerular, plexiform and granular layers in the untreated rat. C : after 6-OHDA treatment of the olfactory bulb, nerve terminals are still observed in the glomerular layer (arrows). D: 5,6-DHT pretreatment of the olfactory bulb abolished labelled nerve terminals in the glomerular layer. 30 (A) and 560 ~ (B, C, D).
421
7
!
i
Figs. 2. A - D : light microscopic radioautographs 24 h after a unilateral injection of [3H]5-HT (100 pCi) in the olfactory bulb. Exposure time: 1 month. A: ipsilateral labelled cell bodies in the raphe dorsalis nucleus. Medial longitudinal bundle, MLB. B: retrogradely labelled cell bodies in the raphe dorsalis nucleus. Note th.e positive labelling of dendrites. C: retrogradely labelled cell bodies in the raphe centralis nucleus. D : reactive cell body in the periaqueductal substantia grisea. No diffuse reaction is observed over the ependymal layer of the aqueduct (Aq). 80 x (A) and 520 × (B, C, D),
422
Figs. 3. A-C: light microscopic radioautographs 24 h after injection of [3H]5-HT (25/zCi) into the olfactory bulb. Exposure time: 1 month. A: after 6-OHDA treatment of the right olfactory bulb anq subsequent unilateral injection of [aH]5-HT in the treated olfactory bulb, positive-labelled cell bodies are seen in the ipsilateral raphe dorsalis nucleus. Medial longitudinal bundle, MLB. B and C: after 5,6-DHT treatment of the right olfactory bulb and following subsequent bilateral injection of [:~H]5-HT into the olfactory bulb, positive-labelled cell bodies are seen only on the left side of the midline (vertical dotted line) raphe dorsalis nucleus. B : labelled cells in the lateral part of raphe dorsalis nucleus. C: labelled cells in the medial part of raphe dorsalis nucleus. 350 y (A, B, C).
Following a unilateral [aH]5-HT injection in the OB, the distribution o f reactive cells occurred in the ipsilateral R D N (Fig. 2A). Only a few scattered labelled cells could be observed on the contralateral R D N . The radioautographic pattern and the t o p o g r a p h y o f labelled nerve cell bodies were similar either in the absence o f C a t r o n pretreatment or after 6 - O H D A injection in the OB (Fig. 3A). The radioautographic reactions completely disappeared in the ipsilateral R D N following unilateral 5 , 6 - D H T administration in the OB (Fig. 3B and C).
Electron microscopic radioautography of RDN Observation o f thin sections o f the R D N confirm dethe complete lack o f
423 labelling o f the supraependymal fibers (Fig. 4A). The diffuse reaction was very low t h r o u g h o u t the whole R D N . Against this absence o f background, some neurons appeared conspicuously labelled (Fig. 4B and C). These were medium-sized neurons (12-15 # m ) with indented nuclei. Silver grains were sparsely distributed on the perikarya, and in general spared the nuclei, but frequently overlapped the mitoch-
Figs. 4. A-C: high resolution radioautographs of thin sections of rat raphe dorsalis 24 h after injection of [aH]5-I-1T(100/~Ci) into the olfactory bulb. Exposure time: 40 days. A: ependyma of the 4th ventricle: supraependymal fibers (arrows) are completely devoid of labelling. Microdol development. B and C: labelled neurons: silver grains are mainly localized on mitochondrial profiles (M) and on lysosomes (Ly). The nucleus is not labelled but some silver grains overlap the nuclear membrane in C (arrow) and its immediate vicinity. Phenidon development. 6000 × (A), 7500 x (B) and 10,000 × (C).
Figs. 5. A - C : high resolution radioautographs of thick sections of Ur-Cu-Pb-stained raphe dorsalis, 24 h after injection of [3H]5-HT (100/iCi) into the olfactory bulb of Catron-pretreated (A-C) or untreated (B) rats. (Phenidon developer). A: labelled neuron at low magnification : silver grains clearly spare the tangential nucleus profile (N) and seem to be selectively localized on mitochondria (M). Endoplasmic reticulum: ER. An adjoining neuron is devoid of labelling (*). B and C: Silver grains seem to be preferentially localized on the mitochondria. 4500 ~ (A, B) and 6000 (C).
425 ondria and lysosomes. Some grains were observed near the neuronal nuclear membrane and also outside the cell, in its vicinity. Observation of heavy metal impregnated thick sections of the RDN showed cell bodies which displayed more numerous silver grains with respect to the low surrounding diffuse reaction (Fig. 5). These mediumsized neurons displayed intensely stained mitochondria which appeared elongated and sometimes branched, and the endoplasmic reticulum (ER) extended as a continuous structure across the whole perikaryon. In the radioautographic study of these thick sections, silver grain distribution with respect to mitochondria was analyzed according to Salpeter et al. 16. In Catronpretreated animals (Fig. 5A and C), 74 ~ of the silver grains occurred within 12 nm, measured from the edge of the grain to the membrane of the nearest mitochondria, and 93 ~ within 50 nm. In untreated animals (Fig. 5B), 53 ~o of the silver grains occurred within 12 nm of the nearest mitochondria and 88 ~ within 50 nm. Under both experimental conditions, grains not localized on mitochondria occurred on the ER. There were few, if any, grains overlying the nuclei. DISCUSSION In this study, the lack of labelling of periaqueductal 5-HT fibers, the restricted area of diffusion, and the small proportion of labelled cell bodies in the RDN strongly contrast with the pattern of labelling observed following intraventricular administration of [aH]5-HT4,10. In this latter case, a gradient of general diffuse reaction is observed in the periaqueductal region which displays intensely labelled supraependymal fibers or nerve terminals, combined with a third of the RDN labelled neurons 7. In our model, the radioautographic results confirm the biochemical finding(s) (see part I) and exclude the possibility that the tracer could reach the RDN by ventricular diffusion. Previous studies have demonstrated the ipsilateral projections of the RDN 6 and their bilateral organization13. In fact, after a single injection of [ZH]5-HT into the OB, the RDN was ipsilaterally labelled. This strongly suggests that [3H]5-HT is retrogradely transported by raphe fibers and/or terminal afferents to the OB. In addition, the labelling of the RDN was suppressed when the terminals in the OB were locally destroyed by 5,6-DHT, while local administration of 6-OHDA failed to impair the RDN labelling. These observations and previous biochemical results (see part I) show that the retrogradely labelled cells observed in the RDN are serotonergic. Moreover, 5-HT cell bodies have been previously well defined in these areas (B7 and Bs cell groups) using histofluorescence5, light and high resolution radioautographyT,10 or immunocytochemistry15. The topography and morphology of the [3H]5-HT retrogradely labelled neurons correspond to these serotonergic groups. The subcellular distribution of retrogradely transported [aH]5-HT or derivatives was also found to be different from that observed when the 5-HT cell bodies are labelled following either ventricular or intratissular injection of [aH]5-HT. In this case, the nuclei and nucleoli were strongly labelled10 but the animals were sacrificed after 3 h of [3H]5-HT administration. In our experiments, however, the maximum accumula-
426 tion of transported 5-HT was found in the RDN, 24 h after [aH]5-HT injection into the OB (see part I) and the nuclei and nucleoli were practically free of silver grains. Such a difference in the pattern of labelling could correspond either to a time-dependent evolution of the cellular distribution of 5-HT and/or to a different intracellular compartmentation of the transported 5-HT. As shown by our quantitative approach, the labelling in both Catron-treated and untreated animals seems to be principally on the mitochondria 1~. The same preferential mitochondrial labelling was observed following direct uptake of the [3H]5-HT in the absence of MAO inhibition and within 3 h of tracer administration 10. This similar localization could suggest that the major part of the retrogradely transported 5-HT could have access to mitochondrial MAO in the cell bodies. Further support for this view is given by the fact that only about 30 0/o of the total radioactivity was detected as unmetabolized [~H]5-HT (see part I) in this structure. The biochemical results (see part I) have suggested that [3H]5-HT, injected into the OB, could be transported in several compartments. A kinetic quantitative analysis of radioactivity localization within serotonergic axons would be necessary to identify these subcellular pools. In addition to the mitochondria, labelling of the ER and lysosomes by retrogradely transported 5-HT or metabolites has been observed. These latter two compartments have been postulated to be concerned with, respectively, the mechanism of retrograde transport and cellular degradation of retrogradely transported macromolecular material2,9,11,17. However, the involvement of the agranular reticulum in retrograde transport of HRP remains controversial3,1L This does not, however, presuppose the identity of the intracellular pathways involved in transporting either H R P or 5-HT and/or its derivatives pinocytosed into the cell. The continuity between perikaryal and axonal labelled organelles has to be established. Finally, the labelling of dendrites, somatic membranes and their surrounding areas suggests that a fraction of the transported 5-HT could subsequently be released into the extracellular space. This accords with the hypothesis of a 'dendritic release' already suggested for dopamine in dopaminergic dendrites in the substantia nigra 1,14,L% ACKNOWLEDGEMENTS The authors wish to thank Dr. Lucienne Leger and Dr. Laurent Descarries for helpful discussions. This work was supported by INSERM U 171 (CRL no. 77 5 110 6) and INSERM U 52, D G R S T (Grant 78 7 2782) and CNRS INP 08 Marseille. S.A. was a recipient of a fellowship from the government of France. REFERENCES 1 Bjrrklund, A. and Lindvall, O., Dopamine in dendrites of substantia nigra neurons: suggestions for a role in dendritic terminals, Brain Research, 83 (1975) 531-537. 2 Broadwell, R. D. and Brightman, M. W., Cytochemistry of undamaged neurons transporting exogenous protein in vivo, J. comp. NeuroL, 185 (1979) 31-74. 3 Broadwell et al., Neuronal transport of acid hydrolase and peroxidase within the lysosomal system of organelles: possible involvement of agranular reticulum like cisterns, J. comp. Neurol., in press.
427 4 Chan-Palay, V., Serotonin axons in the supra and subependymal plexuses and in the leptomeninges; their roles in local alterations of cerebrospinal fluid and vasomotor activity, Brain Research, 102 (1976) 103-130. 5 Dahlstr6m, A. and Fuxe, K., Evidence for the existence of monoamine-containing neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of brain stem neurons, Acta physiol, scand., Suppl. 232 (1964) 1-55. 6 De Pimps, J., Hardy, H. and Heimer, L., The afferent connections of the main and the accessory olfactory bulb formations in the rat: an experimental HRP-study, J. comp. Neurol., 181 (1978) 213-244. 7 Descarries, L., Beaudet, A., Watkins, K. and Garcia, S., The serotonin neurons in nucleus raphe dorsalis of adult rat, Anat. Rec., 193 (1979) 520. 8 Descarries, L. et Schroder, J. M., Fixation du tissu nerveux par perfusion a grand dSbit, J. Microsc., 7 (1968) 281-286. 9 Droz, B., Rambourg, A. and Koenig, H. L., The smooth endoplasmic reticulum: structure and role in the renewal of axonal membrane and synaptic vesicles by fast axonal transport, Brain Research, 93 (1975) 1-13. 10 Gamrani, H. and Calas, A., Cytochemical, stereological and radioautographic studies of rat raphe neurons, J. Microsc., in press. 11 LaVail, J. H. and LaVail, M. M., The retrograde intraaxonal transport of horseradish peroxidase in the chick visual system: a light and electron microscopic study, J. comp. Neurol., 157 (1974) 303-358. 12 LaVail et al., Evidence against the smooth endoplasmic reticulum as a continuous channel for the retrograde axonal transport of HRP, Brain Research, in press. 13 Levitt, P. and Moore, R. Y., Developmental organization of raphe serotonin neuron groups in the rat, Anat. Embryol., 154 (1978) 241-251. 14 Nieoullon, A., Cheramy, A. and Glowinski, J., Release of DA in vivo from cat substantia nigra, Nature (Lond.), 266 (1977) 375-377. 15 Pickel, V., Joh, T. H. and Reis, D., Monoamine-synthesizing enzymes in central dopaminergic, noradrenergic and serotonergic neurons. Immunocytochemical localization by light and electron microscopy, J. Histochem. Cytochem., 24 (1976) 792-806. 16 Salpeter, M. M., Bachmann, L. and Salpeter, E. E., Resolution in electron microscope radioautography, J. Cell Biol., 41 (1969) 1-20. 17 Sotelo, C. and Riche, D., The smooth endoplasmic reticulum and the retrograde and fast orthograde transport of horseradish peroxidase in the nigro-striatonigral loop, Anat. Embryol., 146 (1974) 209-218. 18 Thierry, G. and Rambourg, A., A new staining technique for studying thick sections in the electron microscope, J. Microsc. Biol. Cell., 26 (1976) 103-106. 19 Wilson, C. J., Groves, P. M. and Fifkova, E., Monoaminergic synapses, including dendro-dendritic synapses in the rat substantia nigra, Exp. Brain Res., 30 (1977) 161-174.