Ultrastructural evidence for a separate, small synaptic vesicle (SSV) pathway in ligated bovine splenic nerves, incubated in vitro

BRAIN RESEARCH Brain Research 731 (1996) 101-107

ELSEVIER

Research report

Ultrastructural evidence for a separate, small synaptic vesicle (SSV) pathway in ligated bovine splenic nerves, incubated in vitro J. Quatacker a, *, W. De Potter b a N. Goormaghtigh Institute of Pathology, University Hospital, De Pintelaan 185, B-9000 Ghent, Belgium b Laboratory of Neuropharmacology, University of Antwerp, Antwerp, Belgium

Accepted 9 April 1996

Abstract

In sympathetic nerves the tubules of the axonal reticulum make up the immature elements of the neurosecretory apparatus. The formation of the mature large dense granules occurs via a less dense tubular intermediate, representing the maturing part. At a terminal small synaptophysin-positive vesicles are found intermingled with the dense granules. The biogenesis of these dear, small synaptic vesicles and their relationship with dense granules remains to be determined. In search for the small synaptic vesicles we undertook a careful ultrastructural examination of the axons proximal to a ligation in bovine splenic nerve incubated in vitro for 3 h. The distended axons were crowded with tubules, granulo-tubular elements and dense granules. Occasionally homogeneous clusters of small, uniform vesicles were detected. They were shown to be positive for synaptophysin and were negative for dopamine-[3-hydroxylase, a marker for the granular pathway. The clusters of small vesicles could be found in close spatial relationship with the maturing and mature elements of granular secretion. Our findings argue for the presence of two separate neurosecretory pathways in sympathetic nerves and favour the idea that both small synaptic vesicles and dense granules are a differentiation product of the axonal reticulum. This configuration can explain the biogenesis of small synaptic vesicles and dense granules both in the cell body and at the nerve terminal. Keywords: Axonal reticulum; Axonal transport; Sympathetic nerve; Small synaptic vesicle; Synaptophysin;Large dense-core vesicle

1. Introduction

The axonal reticulum in sympathetic neurons is involved in neurosecretion [24,25]. It is composed of Golgiassociated elements and of a Golgi-derived tubular reticulure descending along the axon [22,23]. This is the immature secretory compartment. In vitro incubation of ligated bovine splenic nerves leads proximally to accumulation of neurosecretory material [10,26]. Immediate fixation optimally preserved the ultrastructure and it was concluded [26] that large dense core vesicles (LDCV) are the mature endproduct of granular secretion [7]. Granule formation seems to pass via less dense vesiculo-tubular complexes [26,28,31,32], forming the maturing neurosecretory compartment.

* Corresponding author. Fax: + 32 (9) 240-4965.

Synaptophysin-positive small synaptic vesicles (SSV) [6,18,19,29], found in conjunction with the dense granules [1,24], are another type of secretory organelle present in the neuronal cell body and axon terminal. Considerable efforts have been made to identify and characterise the proteins of the SSVs [6,18,19,29], but despite these efforts the subcellular site of biogenesis of the SSVs is still not clear. Originally it has been assumed that the small synaptic vesicles at the nerve endings are the organelles by which membranes of neurosecretory granules are recovered following exocytosis [16]. Others have emphasized that SSVs are distinct from the neuronal dense core vesicles [4,6,19,21]. The key question concerning the biogenesis is from which compartment the newly synthesized membrane proteins are first assembled into SSVs [9,18]. The passage of the relevant proteins via the trans Golgi network is widely accepted [11]. But from here the ways may diverge. Rrgnier-Vigouroux et al. [27] postulated a constitutive pathway in which assembly of SSVs occurs at the plasma membrane-endosome level. Exit of the membrane proteins of SSVs from the regulated pathway may be (i) via a separate class of vesicles, (ii) together with the

0006-8993/96/$15.00 Copyright © 1996 Elsevier Science B.V. All rights reserved. Pll S0006- 8993(96)00470-2

102

J. Quatacker, W.D. Potter~Brain Research 731 (1996) 101-107

granule membrane proteins or (iii) as an integral protein of the mature secretory granule membrane. Release from the Golgi apparatus (i) in a vesicular form and transport to the nerve ending has been considered [12,30]. However, SSVs were not found to accumulate at a block [33]. The dense granules and SSVs (ii) m a y both originate from an immature secretory compartment. Sorting out of proteins to granules and small vesicles may then occur either at the Golgi-associated axonal reticulum [22,23] in the cell body or at a tubular differentiation of the terminal axonal reticulum. However, immunoreactivity for synaptophysin could not be found in the axons [1,12,13,30,34]. The possibility (iii) that the synaptic vesicles originate after exocytosis of mature secretory granules [16] conflicts with the presence in PC 12 cells of two independent pathways [4], leading to small clear vesicles and mature dense granules. In light of our previous results, it seems logical that both dense granules and S S V ' s are derived from the same immature elements. That small vesicles may indeed also accumulate above a ligation was suggested by the appearance of synaptophysin-immunoreactivity above a crush in sciatic nerve [5,15]. Unfortunately, in in vitro incubated nerves [26] the SSVs were not particularly distinct between the compacted tubular and granular material above a ligation. In search for SSVs we undertook a careful examination of the axons proximal to a ligation in bovine splenic nerve incubated in vitro for 3 h. The structures detected were further characterized by immunogold cytochemistry with antibodies against synaptophysin and dopamine-[3-hydroxylase (D[3 I-I). It is the aim of this study to determine the site of origin of the SSVs and to reveal their relationship with the large dense granules [7]. This should improve our insight into the functioning of the SSVs and allow us to find out whether sympathetic nerves have indeed two separate regulated pathways [4].

2. Material and methods 2.1. Nerve incubation Bovine splenic nerves were collected at the slaughterhouse, chilled on ice and transported to the laboratory within 40 rain. The nerves were incubated in vitro [10] basically as described previously [26]. Nerves were dissected and 7 × 4 lengths of approximately 60 m m were fixed with threads to a glass rod. A ligation was placed on the nerve leaving most of it to the proximal side. The glass rod and attached nerve were jet-rinsed with sterile buffer solution and put in a flask containing sterile-filtered Krebs-Ringer bicarbonate buffer equilibrated with 5% C O 2 / 9 5 % 0 2. The flask was placed in a shaking waterbath at 37°C. and the nerves were incubated for 3 h. A 15-min incubation restored the axonal ultrastructure [26]. 2.2. Tissue preparation The nerve, still fixed to the rod, was immediately placed in a tube containing chilled fixative (2% formaldehyde, 0.1% glutaraldehyde in 0.1 M cacodylate buffer, pH 7 . 2 - 7 . 4 ) and kept overnight in fixative. The nerves were rinsed in 0.1 M cacodylate buffer, pH 7.2-7.4. Proximal to the ligation two approximately 1.5 m m segments were cut. They were then both embedded either in glycolmethacrylate ( G M A ) without postfixation [14] or in Epon after OsO 4 (1%) fixation. The small segments were put in gelatine capsules for heat or UV polymerization. 2.3. Immunogold Antibodies to DISH ( 1 / 5 0 0 ) were raised in rabbits, as previously described [24]. Monoclonal anti-synaptophysin (clone SY 38, 1 / 1 0 to 1 / 2 0 ) was from Boehringer, Mannheim. Goat anti-rabbit IgG, labeled with 15 nm gold

Fig. 1. Epon, Ur & Pb. The distended axons contain densely packed tubular elements and vesicular profiles of the immature tubules of the axonal reticulum. Round to elongated granules are distributed in clusters or aligned (upper right) along the axonal plasma membrane. The diameter of the granules varies and some of the smaller granules have a lower electron density (small arrowheads). A dense tubule can also be seen (large arrowhead). The dense area above contains several tangentially-cut mitochondria. For further details see [26]. × 27,000; bar = 250 nm. Fig. 2. Epon, Ur & Pb. Detail of a granular area. Some vesiculotubular structures have a slightly electron-dense content (small arrowheads). Together with the denser tubular structures (large arrowheads) they form the maturing compartment. The mature large, round to elliptic granules have a high electron density. A small cluster (*) of equally sized microvesicles (SSVs) can also be seen. × 43,000; bar = 250 nm. Fig. 3. Epon, Ur & Pb. The large homogeneous cluster of SSVs is surrounded by an amorphous axoplasm. A few vesicles and granules can be seen in the neighbourhood. M = mitochondrion. × 43,000; bar = 250 nm. Fig. 4. Epon, Ur & Pb. This shows a close spatial relationship between the SSVs (centre) and the granulo-tubular elements. In the latter it is possible to discern large dense granules and granulo-tubular elements of intermediate density (small arrowheads). × 38,000; bar = 250 nm. Fig. 5. Epon, Ur & Pb. When surrounded by amorphous axoplasm, the granular complexes are seen to form separate entities. They are often closely apposed to the axonal plasma membrane. In these complexes dense granules are seen to be integrated with less dense elements. X 13,500; bar -- 250 nm.

J. Quatacker, W.D. Potter/Brain Research 731 (1996) 101-107

103

104

J. Quatacker, W.D. Potter/Brain Research 731 (1996) 101-107

particles and goat a n t i - m o u s e IgG, labeled with 5 n m g o l d particles ( A u r o p r o b e ) w e r e f r o m A m e r s h a m , Poole, U K . In the controls the primary antiserum was omitted or replaced by an appropriately diluted p r e - i m m u n e rabbit serum. The ultrathin G M A sections w e r e labeled by an indirect m e t h o d as p r e v i o u s l y d e s c r i b e d [24].

2.4. Staining Ultrathin and 200 n m E p o n sections w e r e stained with saturated uranyl acetate in 50% alcohol and alkaline l e a d citrate for 3 and 2 rain, respectively. A f t e r i m m u n o l a b e l i n g the G M A sections w e r e stained for 4 m i n on a drop o f 1%

Fig. 6. GMA, PTA staining. The central dense area represents a cluster of SSVs (compare Fig. 3). The small dots result from staining of the content of the SSVs exposed at the surface of the section. The granular and tubular elements around the cluster are strongly contrasted (large arrowheads). Some faintly contrasted tubular elements (small arrowhead) can also be seen. × 38,000; bar = 250 nm. Fig. 7. GMA, Anti-synaptophysin 1/20, PTA staining. Substantial labeling is observed over the cluster. Gold particles are found aggregated over smaller PTA-positive structures. When compared with the dots in Fig. 6, the morphology has somewhat suffered from the immunologic procedure. × 65,000; bar = 250 nm. Fig. 8. GMA, anti-D[3H 1/500, Ur & Pb. Gold labeling is absent over the vesicle cluster. A few structures at the outside are reactive. Labeling is most obvious on granules, but labeling can also be seen on less well defined, presumably membranous, structures. × 38,000; bar = 250 nm.

J. Quatacker, W.D. Potter~Brain Research 731 (1996) 101-107

phosphotungstic acid (PTA) in 1 N HC1 and quickly dried with filterpaper. Sections were examined in a Zeiss EM 900 at 80 kV in high resolution or in high contrast mode.

105

clusters of small vesicles, but labeling could be seen on tubular and granular structures intimately associated with the outside of the clusters (Fig. 8). No labeling was observed when the primary antiserum was replaced by a pre-immune rabbit serum.

3. Results

3.1. Ultrastructure: Epon

4. Discussion

After incubation the axons above a constriction were distended (Fig. l and Fig. 3 and 5). Close to the ligation the enlarged axon profiles were often densely packed with cell organelles (Fig. 1 and Fig. 2). The accumulation occurred mainly in the first and to a lesser degree in the second segment proximally. The axons contained tubules, distended tubules and vesicles of varying size. Anastomosing structures of intermediate density were observed mostly in conjunction with dense tubular fragments and granules of high electron density. The majority of axon profiles only revealed these tubular and granular structures, but sporadically clusters of SSVs could be detected (Figs. 2-4). Further away from the ligation vesicular and granular aggregates appeared in an amorphous axoplasm (Fig. 3 and Fig. 5). Clusters of SSVs were most easily observed in these areas (Fig. 3). The uniform, small vesicles (40 nm) formed homogeneous, tightly packed aggregates of varying size (Fig. 3). The clusters of SSVs were usually not found close to the plasma membrane. Phenomena suggestive of exocytosis were not observed. A few tubules and granules could regularly be found associated with the outside of the clusters. A close spatial relationship with granulo-tubular elements was sometimes revealed (Fig. 4). The granulotubular complexes or concentrations of granulo-tubular elements were regularly seen in subplasmalemmal positions (Fig. 5). This suggests a secretory activity.

Although already suggested a long time ago on morphological grounds, [8,20] it is still not generally recognized that the axonal reticulum may be involved in neurosecretory vesicle formation. Since then it was further substantiated that the LDCVs in sympathetic nerves [22-26] may arise mainly through local production at the nerve terminal. This can be deduced from the fact that granulo-tubular complexes, comprising a maturing and a mature part [28,31,32], are found proximal to a ligation in bovine splenic nerves [26]. From a functional point of view, it seems most plausible that the formation of SSVs operates via the structures which are involved in dense granule formation. Ligated adrenergic nerves acquire many properties of sympathetic nerve terminals [2,3] and it may thus be possible to reveal the synaptophysin-bearing membrane component, normally present at a terminal [24], above a ligation. Signs of synaptic vesicle accumulation were not apparent at first [26]; and on a global scale, after homogenisation, synaptophysin appeared even to move away from the site of ligation [1]. However, in individual axons accumulation of synaptophysin-immunoreactivity was demonstrated after a nerve crush [5,15]. In this study a careful search of the axons above a ligation revealed clusters of small vesicles, which were unmistakably recognizable as synaptic vesicles by their extremely uniform size (40-50 nm) and round shape. The demonstration of synaptophysin immunoreactivity over the vesicles clusters indicates that above a ligation the synaptophysin component, in addition to the granulo-tubular component, is a well organized structure. That the synaptophysin is below detectability in the tubular reticulum descending along the axons [1,12,13,30,34] may be due to the fact that SSVs only represent a minor fraction of the neurosecretory organelles formed at the terminals. The colocalization of clusters and granulo-tubular components suggests that both may be formed locally from a common structure. They both can thus be considered to represent sites of differentiation on the axonal reticulum. The homogeneity of the clusters of small vesicles suggests a formation from a differentiation site, which is spatially defined. The clusters were most readily recognized when surrounded by amorphous axoplasm [17], although they could also be found in the more crowded axonal accumulations. This may in the first place represent a problem of identification. In general the SSVs had no peculiar relationship with the axolemma. Accordingly, no signs of exocytosis were seen. The cluster configuration can be considered to

3.2. Ultrastructure: GMA The accumulating tubular and granular elements were revealed by the PTA staining. The larger granulo-tubular profiles showed the highest contrast (Fig. 6). The clusters of SSVs were slightly denser than the background. The individual small vesicles appeared as small faintly contrasted dots (Fig. 6). Granulo-tubular elements could be found in close contact with the outside of the clusters.

3.3. hnmunocytochemistry Labeling for synaptophysin was found over the areas containing the aggregated small vesicles. Discrete clusters of gold particles (Fig. 7) were distributed on faintly PTApositive dots. Substantial labeling of other structures was not detectable. Labeling was abolished in the controls. Labeling for D[3H was intense on tubular and vesicular elements. Dense granules and dense tubular segments were most reactive (Fig. 8). Labeling was absent from the

106

J. Quatacker, W.D. Potter/Brain Research 731 (1996) 101-107

represent the native state at differentiation. The existence of two distinct compartments indicates that in sympathetic neurons the granulo-tubular system with its maturing [28,31,32] and mature [7] elements coexists with the synaptophysin-positive microvesicular system. This implies in the first place that two different functional entities are present. [4,6,24] That we are dealing with two separate compartments can also be deduced from the behaviour of the cytochrome b561 and synaptophysin-positive fractions in differential centrifugation [1]. The granulo-tubular system carries catecholamines and neuropeptides. The precise role of the synaptophysin-positive small vesicles in sympathetic neurons remains unknown. The existence of both systems next to each other does not rule out that after secretion some intermixing of membranes may occur upon endocytosis. This would then necessarily have to be followed by a resorting of D[3H and synaptophysin-positive membrane stretches to reconstitute both types of vesicles. However the existence of different ways for recapturing of the two membrane types has been documented. [21] Our observations may also be discussed in relation to the question of how and where the SSVs are first assembled. A possibility for the biogenesis of small vesicles is a constitutive pathway for plasma membrane proteins [27] and a regulated formation of small vesicles upon endocytosis. SSVs could also be formed from endocytozed LDCVs [16]. However, the organisation in clusters and the absence of signs of exo-endocytosis make those pathways very improbable. The SSVs may directly be formed as a separate class of vesicles [12,30] and transported down the axon. One should however bear in mind that accumulation of SSVs at the node of Ranvier [12] is not necessarily a proof for transport of the small vesicles, but may equally well reflect vesicle formation from axonal reticulum at that site. Such a transition has been described to occur at a terminal [8]. It also remains difficult to conceive how separately transported small vesicles would upon arrival reorganize to homogeneous clusters and co-organize with complex granulo-tubular elements. Our findings are strongly in favour of a common origin of granules and small synaptic vesicles from a differentiation at the Golgiassociated axonal reticulum [22,23] in the cell body and from the tubular complexes at the nerve terminal [26]. That the SSVs at a terminal may be derived from axonal reticulum was already strongly suggested [8]. The proposed configuration can explain the biogenesis of neurosecretory granules and SSVs both in the cell body and at the nerve terminal. In the axons the tubular reticulum should account for the presence of LDCVs. In this scheme the small dense core vesicles of the nerve endings are missing. We consider them to be an artefact, due to vesiculation of the medium-dense tubular elements [26]. When special attention is given to the fixation [26,28,31,32] tubular reticula and complexes predominate. It is proposed that the membrane proteins and content of the neurosecretory organelles [24-26] are transported with the axonal reticulum

to the nerve terminal. A sorting of membrane components may then occur, leading to differentiation in LDCVs [26] and synaptophysin-positive, small vesicles. Our findings thus indicate that sympathetic nerves posses two separate neurosecretory pathways [4,6,29] for local production.

Acknowledgements This work was supported by the IUAP-III fund. This text presents results of the Belgian Programme on International Poles of Attraction initiated by the Belgian State, Prime Minister's Office, Science Policy Programming. The scientific responsibility is assumed by the authors.

References [1] Annaert, W.G., Quatacker, J., Llona, I. and De Potter, W.P., Difference in the distribution of cytochrome b561 and synaptophysin in dog splenic nerve: a biochemical and immunocytochemicalstudy, J. Neurochem., 62 (1994) 265-274. [2] Calvo, S., Gonzalez-Garcia, C. and Cefia, V., Axoplasmic transport of [3H] ouabain binding sites and catecholamine secretion from an adrenergic nerve trunk, Mol. Pharmacol., 42 (1992) 141-146. [3] Cefia, V., Garcia, A.G., Pi, F. and Sanchez-Garcia,P., Presence of a noradrenaline uptake system on a ligated cat sympathetic nerve, J. Physiol., 372 (1986) 351-362. [4] Cutler, D.F. and Cramer, L.P., Sorting during transport to the surface of PC 12 cells: divergence of synaptic vesicles and the secretory granule protein, J. Cell Biol., 110 (1990) 721-30. [5] Dahlstrtim, A.B., Czernik, A.J. and Li, J-Y., Organelles in fast axonal transport. What molecules do they carry in anterograde versus retrograde directions, as observed in mammalian systems?, Mol. Neurobiol., 6 (1992) 157-177. [6] De Camilli, P. and Jahn, R., Pathways to regulated exocytosis in neurons, Annu. Rev. Physiol., 52 (1990) 625-645. [7] De Potter, W.P., Kurzawa, B., Miserez, B. and Coen, E.P., Evidence against differential release of noradrenaline, neuropeptide Y and dopamine-13-hydroxylasefrom adrenergic nerves in the isolated perfused sheep spleen, Synapse, 19 (1995) 67-76. [8] Droz, B., Rambourg, A. and Koenig, H.L., The smooth endoplasmatic reticulum: structure and role in the renewal of axonal membrane and synaptic vesicles by fast axonal transport, Brain Res., 93 (1975) 1-13. [9] Feany, M.B., Yee, A.G., Delvy, M.L. and Buckley, K.M., The synaptic vesicle protein SV2, synaptotagminand synaptophysinare sorted to separate cellular compartments in CHO fibroblasts, J. Cell Biol., 123 (1993) 575-584. [10] Fried G., Terenius L., Brodin E., Efendic S., Dockray G., Fahrenkrug J., Goldstein M. and H~SkfeltT., Neuropeptide Y, enkephalin and noradrenaline coexist in sympathetic neurons innervating the bovine spleen, Cell Tiss. Res., 243 (1986) 495-508. [11] Hammerschlag, R., Stone, G.C., Bolen, F.A., Lindsey, J.D. and EUisman, M.H., Evidence that all newly synthesized proteins destined for fast axonal transport pass through the Golgi apparatus, J. Cell Biol., 93 (1982) 568-575. [12] Janetzko, A., Zimmermann, H. and Volknand, W., Intraneuronal distribution of a synaptic vesicle-membraneprotein: antibody binding sites at axonal membrane compartments and trans-Golgi network and accumulation at the nodes of Ranvier, Neuroscience, 32 (1989) 65-77. [13] Kadota, T., Fujita, M. and Kadota, K., Immunocytochemicallocal-

J. Quatacker, W.D. Potter/Brain Research 731 (1996) 101-107

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

ization of synaptophysin on the smooth-surfaced tubular membranes present in nerve terminal and preterminal areas in the rat cerebellar cortex, Arch. Histol. Cytol., 54 (1991) 519-525. Leduc, E.H. and Bernhard, W., Recent modification of the glycolmethacrylate embedding procedure, J. Ultrastructure Res., 19 (1967) 196-199. Li, J.-Y, Jahn, R. and Dahlstr~Sm, A., Synaptotagmin I is present mainly in autonomic and sensory neurons of the rat peripheral nervous system, Neuroscience, 63 (1994) 837-850, Lowe, A.W., Madeddu, L. and Kelly, R.B., Endocrine secretory granules and neuronal synaptic vesicles have three integral membrane proteins in common, J. Cell Biol., 106 (1988) 51-59. Matthews, M.R., An ultrastructural study of axonal changes following constriction of postganglionic branches of the superior cervical ganglion of the rat, Phil. Trans. R. Soc. Ser. B, 264 (1973) 479-508. Mundigl, O., Matteoli, M., Daniell, L., Thomas-Reetz, A., Metcalf, A., Jahn, R. and De Camilli P., Synaptic vesicle proteins and early endosomes in cultured hippocampal neurons: Differential effect of Brefeldin A in axons and dendrites, J. Cell Biol., 122 (1993) 1207-1221. Navone, F., Di Gioia, G., Jahn, R., Browning, M., Greengard, P. and De Camilli, P., Microvesicles of the neurohypophysis are biochemically related to small synaptic vesicles of presynaptic nerve terminals, J. Cell Biol., 109 (1989)3425-3433. Pellegrino de Iraldi, A. and De Robertis, E., The neurotubular system of the axon and the origin of granulated and non-granulated vesicles in regenerating nerves, Z Zellforsch., 87 (1968) 330-344. Pow, D.V. and Morris, J.F., Membrane routing during exocytosis and endocytosis in neuroendocrine neuroues and endocrine ceils: use of colloidal gold particles and immunocytochemical discrimination of membrane compartments, Cell Tissue Res., 264 (1991) 299-316. Quatacker, J., The axonal reticulum in the neurons of the superior cervical ganglion of the rat as a direct extension of the Golgi apparatus, Histochem. J., 13 (1981) 109-124. Quatacker, J.R., Annaert, W.G. and De Potter, W.P., Detection of sialic acid residues in the axonal reticutum of rat superior cervical ganglion cells by lectin-gold cytochemistry, J. Histochem. Cytochem., 38 (1990) 1445-1449.

107

[24] Quatacker, J.R., Annaert, W.G., Miserez, B.J. and De Potter, W.P., Imrnunocytochemical demonstration of dopamine-13-hydroxylase and cytochrome B561 on the axonal reticulum in bovine sympathetic neurons, J. Histochem. Cytochem., 40 (1992) 1599-1604. [25] Quatacker, J., Miserez, B., Annaert, W. and De Potter, W., Ultrastructural localization of neuropeptide Y-immunoreactivity in the axonal reticulum elements, accumulating anterogradely in transected rat sciatic nerve, Brain Res., 627 (1993) 249-253. [26] Quatacker, J., Annaert, W. and De Potter, W., The organisation of the axonal reticulum at a ligation, in in vitro incubated bovine splenic nerves, Brain Res., 680 (1995) 36-42. [27] R6gnier-Vigouroux, A., Tooze, S.A. and Huttner, W.B., Newly synthesized synaptophysin is transported to synaptic microvesicles via constitutive secretory vesicles and the plasma membrane, EMBO J., 10 (1991) 3589-3601. [28] Richards, J.G. and Tranzer, J.P., Localization of amine storage sites in adrenergic cell bodies. A study of the superior cervical ganglion of the rat by fine structural cytochemistry, J. Ultrastruct. Res., 53 (1975) 204-216. [29] Siidhof, T.C. and Jabn, R., Protein of synaptic vesicles involved in exocytosis and membrane recycling, Neuron, 6 (1991) 665-677. [30] Tixier-Vidal, A., Faivre-Bauman, A., Picart, R. and Wiedenmann, B., Immunoelectron microscopic localization of synaptophysin in a Golgi compartment of developing hypothalamic neurons, Neuroscience, 26 (1988) 847-861. [31] Tranzer, J.P., A new amine storing compartment in adrenergic axons, Nature New Biol., 237 (1972) 57-58. [32] Tranzer, J.P. and Richards, J.G., Ultrastructural cytochemistry of biogenic amines in nervous tissue: methodological improvements, J. Histochem. Cytochem., 24 (1976) 1178-1193. [33] Tsukita, S. and Ishikawa, H., The movement of membranous organelles in axons: electron microscopic identification of anterogradely and retogradely transported organelles, J. Cell Biol., 84 (1980) 513-550. [34] Zimmermann, H., Volknandt, W., Henkel, A., Bonzelius, F., Janeztko, A. and Kanaseki, T., The synaptic vesicle membrane: origin, protein component, exocytosis and recycling, Cell Biol. Int. Report, 13 (1989) 993-1006.