Cell Tiss. Res. 168, 101-115 (1976)

Cell and Tissue Research 9 by Springer-Verlag 1976

Microtubule--Synaptic Vesicle Associations in Cultured Rat Spinal Cord Neurons Margaret M. Bird Department of Anatomy, University College London, England

Summary. This paper describes new ultrastructural features of neural processes and of synapses in cultured CNS tissue treated with albumin before fixation using a modification of the technique recently introduced by Gray (1975). Nerve fibre bundles in explants of foetal spinal cord grown in vitro for 15-18 days were transected microsurgically. After transection the cultures were exposed to 20% albumin in distilled water and then fixed in unbuffered osmium tetroxide followed by unbuffered glutaraldehyde. In this material, but not in controls (injured but not exposed to albumin; exposed to albumin without injury) microtubules were found within many axonal varicosities, often situated close to presynaptic membrane specializations. These microtubules were closely associated with vesicles resembling synaptic vesicles, which were occasionally aligned in rows along the microtubules. Similar vesicle-microtubule associations were also found in non-terminal axons. Microtubules were also observed very close to some postsynaptic densities. The possibility that the microtubule-vesicle associations are involved in vesicle movements (along axons and/or within axon terminals) is discussed. A more direct involvement of microtubules in terminals in the mechanism of transmitter release is also considered. Key words: Synapses - Microtubules - Albumin - Tissue culture.

Introduction Much excitement has been generated by Gray's recent description of microtubules in vertebrate axon terminals, sometimes clothed in synaptic vesicles, lying Send offprint requests to." Dr. Margaret Bird, Department of Anatomy, London Hospital Medical College, Turner Street, London E1 2AD, England

The author wishes to thank Dr. A.R. Lieberman for his help and advice, Mr. Derek Fraser and Mr. Peter Felton for their technical assistance, Mr. Stuart Waterman for the photographic prints, and Professor D.W. James for laboratory facilities

102

M.M. Bird

close to and apparently anchored to presynaptic densities (Gray, 1975). Although Smith et al. (1970) described a close relationship between microtubules and synaptic vesicles in axons of larval lamprey spinal cord, no convincing subsequent report of comparable observations had been made prior to Gray's dramatic findings. Gray made his observations in tissue from various regions of the frog and mammalian central nervous tissue (CNS) teased in albumin immediately before fixation. This paper reports similar observations made in neurons of embryonic rat spinal cord grown in tissue culture, in which axons were cut and treated with albumin immediately prior to fixation.

Materials and Methods Small pieces (1 mm 3) of embryonic spinal cord from 16-day Wistar rat foetuses were explanted and maintained in a culture medium composed of 90% Eagle's Basal Medium and 10% Horse Serum. The cultures were maintained at 37~ in an atmosphere of 10% carbon dioxide and 90% air and re-fed at intervals of 72 h. After 15-18 days the cultures were treated in one of the following ways: 1. Outgrowing bundles of fibres were cut using a cataract knife attached to a micro-manipulator accurately positioned under an inverted phase contrast microscope. 1-10 rain after the nerve cut a solution of 20% bovine serum albumin in distilled water was added to the preparation. 2. The cultures were placed directly, without cutting, into the 20% albumin solution. 3. Outgrowing fibre bundles were cut but the culture was not exposed to the albumin solution. Five to ten minutes after cutting and/or exposure to albumin all the cultures were fixed according to a modification (Bird and James, 1973) of the technique developed by Kanaseki and Kadota (1969): 4% unbuffered osmium tetroxide followed by 12.5% unbuffered glutaraldehyde. After staining with a saturated solution of aqueous uranyl acetate the specimens were dehydrated through a series of graded ethanols and embedded in Araldite. Thin sections were cut from areas of the explant close and proximal to the cut fibre bundles, areas known from previous work to be rich in synaptic contacts at this stage (Bird, unpublished observations). The sections were grid-stained in lead citrate for 15 min at room temperature.

Results Albumin Added Before Fixation; Nerve Fibres Intact The appearance of the tissue was identical to that of tissues fixed in the same

way without prior albumin treatment. Filaments were plentiful in nerve Fig. 1. An axonal varicosity containing broken microtubules (Mr) after nerve section and albumin treatment. A dense flocculent material fills the varicosity (V). Elements of smooth endoplasmic reticulum are associated with some microtubules at the neck of the varicosity Fig. 2. Microtubules surrounded by a fluffy flocculent material enter a large empty varicosity within which all the microtubules are broken Fig. 3a and b. a. Two microtubules pass without interruption through a varicosity. A few vesicles (v) are found at the neck of the expansion. Flocculent material is present throughout the varicosity and is condensed around the microtubules, b. Unevenly distributed light and dark bands are found along some microtubules (arrows). Flocculent material is observed along the length of all three microtubules

Microtubules at Synaptic Terminals in vitro

103

104

M.M. Bird

processes, and microtubules, though present, were less numerous and were not observed in presynaptic zones.

Fixed After Nerve Section; No Albumin Treatment With the exception of local signs of damage directly attributable to nerve sectioning, the tissue was of normal appearance and the distribution of both microtubules and microfilaments was similar to that described above.

Nerves Cut; Albumin Added Before Fixation a) Microtubules in Nerve Processes. Many nerve fibres were packed with microtubules which extended into the numerous varicosities (Figs. 1-5). Microtubules entering varicosities tended to fan out and break (Figs. 1, 2), but unbroken microtubules running through varicosities were also observed (Figs. 3 a, 4, 9). Clouds of moderately electron dense flocculent material were associated with the microtubules, particularly within abnormally large dilations where microtubules were spread apart (Figs. 2, 3a, 3 b). This flocculent material, which was not apparent after other treatments, was also associated with synaptic vesicles (Figs. 4, 5, 6 a) and other membranous elements. Some microtubules, both within axons and varicosities, appeared to possess unequally spaced transverse striations (e.g. Figs. 1-3, 8). The striated appearance may have been due to the overlying flocculent material (Figs. 1-3) but this explanation was not always the obvious one (e.g. Fig. 8). Individual microtubules, or groups of two or three parallel microtubules, were commonly associated with vesicles of the size and general appearance of synaptic vesicles (Figs. 4q5). The vesicles sometimes clothed long stretches of a single microtubule (Fig. 6 b), or a small bundle of microtubules, or formed small clusters at irregular intervals along the course of otherwise naked microtubules (Fig. 4). Although vesicles commonly appeared to be crowded around the associated microtubules with no indication of an orderly arrangement, they sometimes appeared to be lined up in one or more straight rows (Figs. 6a, 8). In most axons and varicosities containing vesicles associated with microtubules, other vesicles of identical appearance, showing no apparent association with microtubules, were observed close by (e.g. Figs. 5, 6b, 7, 14).

Fig. 4. A cluster of vesicles associated with microtubules at the neck of a large varicosity (V). Vesicles (arrowed) are also observed aligned along the microtubules. An intact microtubule (Mt) passing through a smaller varicosity and lying close to a mitochondrion is also seen Figs. 5 and 6. Single a n d small groups of microtubules associated with vesicles (arrows) in several different nerve processes. In Figures 5 and 6a groups of microtubules entering varicosities are clothed in synaptic vesicles. In Figure 6 b a single tubule is clothed in synaptic vesicles. Unattached vesicles are found in all three pictures

Microtubules at Synaptic Terminals in vitro

105

106

M.M. Bird

Microtubules at Synaptic Terminals in vitro

107

b) Presynaptic Microtubule-Vesicle Associations. Microtubule-vesicle associations within varicosities were commonly observed in this material, particularly close to the site of transection, where about 1 in 4 varicosities show this feature. Vesicle-clothed microtubules could be followed in many cases from a pre-terminal or inter-varicose axon into a varicosity (Figs. 4, 5, 6a), and microtubulevesicle associations appeared to be particularly common at the necks of varicosities (Figs. 6 a, 8, 9). Most vesicles associated with microtubules within varicosities were of the clear, spherical variety. Flattened vesicles, though present in this material have not been observed in intimate association with microtubules. Dense cored vesicles were mostly free (Fig. 11) but were occasionally observed in close but equivocal association with microtubules (Fig. 10). Some microtubules within varicosities appeared completely to bypass the region of a synaptic complex (e.g. Fig. 7). Others, however, were situated very

Fig. 7. A single microtubule (Mt) enters a synaptic varicosity. The microtubule does not pass close to the presynaptic specialization Figs. 8 and 9. Microtubules with vesicles aligned along them (arrows) in nerve processes leading to presynaptic terminals. In Figure 8 note the regular light and dark band-like striations along the tubules (small arrows) in the preterminal axon Figs. 10 and 11. Dark-cored vesicles (dcv) and clear synaptic vesicles (cv) close to microtubules. In Figure l0 some dcv's are closely associated with microtubules (arrow) whereas in Figure 11 no close associations are apparent Fig. 12. Three microtubules bypass a presynaptic vesicle cluster. A fourth tubule (arrows) bends towards the presynaptic specialization and may be anchored to it. Unattached clear and dark-cored vesicles and an element of smooth endoplasmic reticulum (SER) with bulbous dilations, are also present Fig. 13. Several smooth endoplasmic reticulum tubules (SER) with dilations of varying size and large SER vesicles are present within a presynaptic terminal. A single microtubule is directed towards the presynaptic region (arrow) Fig. 14. Vesicles crowded around microtubules are directed towards presynaptic densities Fig. 15. Microtubules are arranged parallel to a presynaptic site. One microtubule associated with clear vesicles (cv) curves towards the synapse Fig. 16. A single microtubule by-passing an "immature" presynaptic site is adjacent to two microtubules which appear to terminate close to the cluster of presynaptic vesicles Fig. 17. A single microtubule runs towards a presynaptic membrane: synaptic vesicles are associated only with the presynaptic portion of the microtubule (arrows) Fig. 18. No microtubules are visible, but the linear arrangement of the vesicles indicates that they may be aligned along a microtubule Fig. 19. A microtubule (Mt) situated close to a presynaptic specialization. Few vesicles are associated with this microtubule except in the region of the synapse. Microtubules (arrow) are also found close to the postsynaptic density

108

M.M. Bird

Microtubules at Synaplic Terminals in vitro

109

110

M.M. Bird

Microtubules at Synaptic Terminals in vitro

111

close to the presynaptic complex (Figs. 12-22). Such microtubules either ran approximately parallel to (or curved towards) the presynaptic membrane and continued beyond it (Figs. 15, 16, 19, 20, 23, 24), or ran perpendicular and more or less directly to the presynaptic membrane (Figs. 13, 17, 18, 21, 22). It sometimes appeared as though the microtubules might be attached to the presynaptic dense projections but unequivocal micrographs of this relationship were not obtained. Some microtubules running close to presynaptic sites were associated with few vesicles (Figs. 13, 15, 16, 19); others displayed a dense accumulation of vesicles in the region immediately adjacent to the presynaptic membrane (Figs. 14, 17, 18).

c) Postsynaptic Microtubules. A few examples of dendritic microtubules in close relation to postsynaptic membrane specializations were also observed (Figs. 2325), either in the form of microtubules running parallel with, and immediately below, the postsynaptic density (Figs. 23, 24), or, more dramatically, runnning directly perpendicular to it and apparently terminating in the postsynaptic density (Fig. 25). d) Smooth Endoplasmic Reticulum. The smooth endoplasmic reticulum (SER) is prominent in this material, both in axons and in varicosities (Figs. 1, 4, 12-14). Much of the SER is in the form of narrow tubules the diameters of which are similar to the diameter of the microtubules. A clear distinction between microtubules and tubular SER was almost always possible, however, because the latter had a less regular profile with intermittent varicosities (e.g. Figs. 5, 12, 14). Microtubule-SER complexes of the type described in frog CNS tissue by Lieberman (1971) and in chick spinal cord explants by Grainger and James (1969), were not observed. Discussion

Microtubules are not normally seen within synaptic endings in cultured or intact, developing or mature nervous tissue prepared for ultrastructural observations by the currently most widely used techniques. These "conventional" techniques generally involve initial fixation with buffered glutaraldehyde and/or paraformaldehyde, followed by postfixation in osmium tetroxide, and are thought to result in optimal preservation of cytoplasmic microtubules. In the present study, however,microtubules were observed within axonal varicosities and endings of cultured CNS tissue, often in close association with vesicles

Figs. 2 0 - 2 2 . Microtubules are located close to presynaptic specializations (arrows). In Figures 21 and 22 the microtubules (Mt) are directed vertically towards the synaptic cleft Figs. 23 and 24. Microtubules (arrows) in close association with both pre- and postsynaptic membranes Fig. 25. A microtubule (arrow) is directed vertically towards a postsynaptic membrane. Light and dark irregular striations occur along its length

112

M.M. Bird

and other components of the synaptic apparatus. These observations were made following selective mechanical damage to nerve fibres and exposure of the unfixed tissue to albumin. Similar observations were made by Gray (1975) on adult CNS tissue (minced in albumin before fixation), and by Westrum and Gray (1976) on similarly treated neonatal CNS tissue. These findings suggest that conventional fixation may not always reveal a complete picture of cytoplasmic microtubule distribution and indicate that the albumin technique should probably be used to re-examine microtubules not only in neurons but also in other cells, particularly those with a secretory function. It would be of considerable interest to establish whether microtubules and microtubule-vesicle associations are revealed with this method, in cells such as chromaffin cells and pancreatic islet cells, close to plasma membranes at which exocytosis of secretory products occurs. The basis for the preservation of microtubules and of microtubule-vesicle associations following albumin treatment, and the reason for the failure of conventional techniques to reveal them, is not clear. The fact that the tissue must be damaged before microtubules are revealed close to the presynaptic membrane is important. Equally significant, however, is the fact, established in this study, that extensive disruption of the tissue is not necessary. In the present study, nerve fibre bundles were carefully transected using a microsurgical technique before albumin was added to the cultures. Direct damage to the fibres and endings in the regions studied by electron microscopy was minimal. It is extremely unlikely, therefore, that the presence of microtubules close to synaptic specializations is due to displacement or distortion of microtubules in nonterminal segments of the nerve fibres. The significance of the damage is presumably that it allows albumin to enter and permeate the damaged fibres. It may be, as Gray (1975, 1976) has speculated, that microtubules in synaptic regions are more labile than microtubules elsewhere in the nerve cell, and are only preserved and revealed in these sites as a result of some particular effect of albumin, preventing or slowing down their disintegration. The flocculent material observed within neural processes and condensed around microtubules and other organelles probably represents fixed albumin that had entered the axon, since it was seen in these sites only after albumin had been added to cultures in which nerve fibres had been cut. On the other hand the material condensed around some microtubules giving them a striated appearance may be caused by a phenomen on akin to that described by Behnke (1975), and Jacobs et al. (1975), in negatively stained material. If penetration of albumin into the axon is an essential prerequisite for demonstrating microtubules and microtubule-vesicle associations at synapses, a ready explanation for the fact that only a small proportion of axons and terminals displays these features, can be offered. Presumably, profiles of conventional appearance were parts of undamaged or only slightly damaged fibres, or of fibres into which albumin was unable to penetrate for some other reason. It is, however, possible that neural processes showing these features are different, or contain some component (? a special class of vesicle-see below) that is different from processes not affected in a similar manner. Although not incompatible with the proposal that albumin stabilizes labile synaptic microtubules,

Microtubules at Synaptic Terminals in vitro

113

the observation by Gray (1976) of at least some microtubule-vesicle associations in tissue teased in distilled water and other hypotonic solutions before fixation, suggests that other factors may be involved. It has often been suggested that synaptic vesicles are manufactured in neuronal perikarya, probably from the Golgi apparatus, and are transported along the axon to the axon terminals (see for example Van Breemen et al., 1958; F.O. Schmitt, 1968; Smith et al., 1975). Gray, (1975) has further proposed that microtubules directed towards presynaptic specializations are responsible for guiding such vesicles to appropriate sites on the presynaptic membrane. The present observations, while not inconsistent with these suggestions, provide no additional evidence to support them, and in recent years it has become apparent that at peripheral cholinergic and at least some central synapses, synaptic vesicle membrane lost to the axon terminal membrane during transmitter exocytosis is recovered within the terminal by an endocytotic mechanism (Ceccarelli et al., 1973; Heuser and Reese, 1973; Pysh and Wiley, 1974; Model et al., 1975). Thus, notwithstanding the general dependence of axon terminals on their parent cell bodies, the proposal that preformed synaptic vesicles roll down axons along microtubules (Schmitt, 1968) is less attractive as a general concept now, than it perhaps was a few years ago. In any case, the vesicle-microtubule associations observed in this and other studies, even if indicative of vesicle translocation, do not necessarily indicate unidirectional, somatofugal transport or movement. The vesicles associated with microtubules are as likely to be moving in the somatopetal as the somatofugal direction. However, even if synaptic vesicles do recycle within the terminal and the associated vesicles turn out not to be synaptic vesicles en route from perikaryon to nerve ending, it is known that vesicle membrane is recovered from axon terminal membrane some distance from the site of exocytosis (Heuser and Reese, 1973), and microtubule "guide lines" within terminals might play a part in directing the recovered vesicles back to the active zones. The suggestion that microtubules guide or actively transport vesicles to and from specific sites within the nerve cell, stems naturally from an extensive body of evidence implicating microtubules in such functions in various cell types (reviewed by Smith et al., 1975). There is, however, an alternative explanation for the significance of microtubules and of microtubule-vesicle associations in the vicinity of the presynaptic membrane. There is good evidence that microtubules are directly involved in the mechanism of excitation-secretion coupling in the release of insulin from the pancreatic islet cell (Malaisse et al., 1975) and ofcatecholamines from the cells of the adrenal medulla (Poisner and Cooke, 1975). Furthermore, there is persuasive evidence for a similar involvement of microtubules in transmitter release from at least some types of nerve ending (Sorimachi et al., 1973; Wooten et al., 1975). It is tempting to speculate that the close microtubule associations with the presynaptic apparatus may be the morphological correlate of the biochemical, pharmacological and physiological evidence suggesting the involvement of microtubules in the mechanism of transmitter release. Finally, it is possible that vesicles seen in close association with microtubules following albumin treatment are not true synaptic vesicles. Associated vesicles

114

M.M. Bird

were usually in the minority in varicosities displaying these special features, and many other vesicles were dispersed within the axoplasm. Thus, although the associated and other vesicles appear similar and may be identical, there is a possibility that the associated vesicles represent a distinct class of vesicle. In view of the suggestion in the preceding paragraph, and because of the known importance of calcium both in the mechanism of transmitter release and in the control of microtubule stability, the possibility that the associated vesicles contain calcium should be considered and investigated. A close association between microtubules and postsynaptic densities, as occasionally observed in this study, was not reported by Gray (1975) but has also been seen by Westrum and Gray (1976) in neonatal CNS tissue prepared by the albumin method. It remains to be established whether or not these postsynaptic microtubules are related to the presence of tubulin (the microtubule subunit protein) in the postsynaptic membrane (Banker et al., 1974; Matus et al., 1975).

References Banker, G., Churchill, L., Cotman, C.W.: Proteins of the postsynaptic density. J. Cell Biol. 63, 456~,65 (1974) Bird, M.M., James, D.W.: The development of synapses in vitro between previously dissociated chick spinal cord neurons. Z. Zellforsch. 140, 203-216 (1973) Behnke, O. : An outer component of microtubules. Nature (Lond.) 257, 709-710 (1975) Breemen, V.L.van, Anderson, E., Reger, J.F.: An attempt to determine the origin of synaptic vesicles. Exp. Cell Res., Suppl. 5, 153-167 (1958) Ceccarelli, B., Hurlbut, W.P., Mauro, A. : Depletion of vesicles from frog neuromuscular junctions by prolonged tetanic stimulation. J. Cell Biol. 54, 30-38 (1972) Grainger, A.F., James, D.W. : Mitochondrial extensions associated with microtubules in outgrowing processes from chick spinal cord in vitro. J. Cell Sci. 4, 729-737 (1969) Gray, E.G.: Pre-synaptic microtubules and their associations with synaptic vesicles. Proc. roy. Soc. B 190, 369-372 (1975) Gray, E.G. : Microtubules in synapses of the retina. J. Neurocytol. 5 (in press) (1976) Heuser, J.E., Reese, T.S.: Evidence for recycling of synaptic vesicle membrane during transmitter release at the frog neuromuscular junction. J. Cell Biol. 57, 315-344 (1973) Jacobs, M., Bennett, P.M., Dickens, M.J. : Duplex microtubule is a new form of Tubulin assembly induced by polycations. Nature (Lond.) 257, 709-710 (1975) Kanaseki, T., Kadota, E. : The vesicle in a basket. J. Cell Biol. 42, 202-220 (1969) Lieberman, A.R.: Microtubule-associated smooth endoplasmic reticulum in the frog's brain. Z. Zellforsch. 116, 564--574 (1971) Malaisse, W.J., Malaisse-Lagae, F., Van Obberghen, E., Somers, G., Devis, G., Ravazzola, M., Orci, L.: R61e of microtubules in the phasic pattern of insulin release. Ann. N.Y. Acad. Sci. 253, 630~552 (1975) Matus, A.I., Walters, B.B., Mughal, S. : Immunohistochemical demonstration of Tubulin associated with microtubules and synaptic junctions in mammalian brain. J. Neurocytol. 4, 733 744 (1975) Model, P.G., Highstein, A.M., Bennett, M.V.L.: Depletion of vesicles and fatigue of transmission at a vertebrate synapse. Brain Res. 98, 209-228 (1975) Poisner, A.M., Cooke, P.: Microtubules and the adrenal medulla. Ann. N.Y. Acad. Sci. 253, 653~669 (1975) Pysh, J.J., Wiley, R.G. : Synaptic depletion and recovery in cat sympathetic ganglia electrically stimulated in vivo. J. Cell Biol. 60, 365-374 (1974) Schmitt, F.O. : The molecular biology of neuronal fibrous protein. Neurosci. Res. Prog. Bull. 6, 119-144 (1968) Smith, D.S., JS.rlfors, U., B6ranek, R.: The organisation of synaptic axoplasm in the lamprey (Petromyzon marinus) central nervous tissue. J. Cell Biol. 46, 199-219 (1970)

Microtubules at Synaptic Terminals in vitro

115

Smith, D.S., J/irlfors, U., Cameron, B.F. : Morphological evidence for the participation of microtubules in axonal transport. Ann. N.Y. Acad. Sci. 253, 472-506 (1975) Sorimachi, M.F., Oesch, F., Thoenen, H.: Effect of colchicine and cytochalasin-B on the release of 3H-norephinephrine from guinea pig atria evoked by high potassium, nicotine, and tyramine. Naunyn-Schmiedebergs Arch. exp. Path. Pharmak. 276, 1-12 (1975) Westrum, L.E., Gray, E.G.: Microtubules and membrane specializations. Brain Res., in press (1976) Wooton, G.F., Kopin, I.J., Axelrod, J.: Effects of colchicine and vinblastine on axonal transport and transmitter release in sympathetic nerves. Ann. N.Y. Acad. Sci. 253, 528-534 (1975)

Received January 19, 1976

Microtubule--synaptic vesicle associations in cultured rat spinal cord neurons.

This paper describes new ultrastructural features of neural processes and of synapses in cultured CNS tissue treated with albumin before fixation usin...
5MB Sizes 0 Downloads 0 Views