Ultrastructural studies of young and old mouse neuromuscular junctions

Summary The ultrastructure of the neuromuscular junction of young and old male CBF-1 mice was analysed both qualitatively and quantitatively. The age-related findings were similar in both the phasic extensor digitorum longus muscle and the tonic soleus muscle but more pronounced in the latter. Presynaptic terminals of old mice compared to young showed decreases in nerve terminal area, mitochondria and synaptic vesicles, but increases in smooth endoplasmic reticulum, coated vesicles, cisternae, microtubules and probably neurofilaments. On the postsynaptic side there were increases in complexity of junctional folds and subsarcolemmal vesicles, and the appearance of lipofuscin deposits. Occasional denervated postsynaptic regions were encountered in old neuromuscular junctions, but the predominant characteristics of aging changes were not those of denervation. Rather, a unique and uniform process involving most of the population of nerve terminals, possibly of physiologically adaptive significance, appears to occur with age in both phasic and tonic limb muscles.     

Neuromuscular transmission in the visceral muscle of locust oviduct

Innervation of the locust oviduct has been investigated with morphological and electrophysiological methods. Using Co2+ and Ni2+ labelling technique, it was found that G7 N2B1 and B2a nerves innervate the oviduct musculature. Ultrastructurally two different terminals could be distinguished: (a) nerve endings containing mainly clear vesicles forming neuromuscular junctions with the muscle fibers; and (b) nerve terminals containing electron-dense granules which showed only "synaptoid" structures, but failed to form junctions with the muscle cells. The neuromuscular junctions proved to be functioning, since it was possible to record intracellularly miniature excitatory postsynaptic potentials and excitatory postsynaptic potentials from the muscle cells. The distribution of the amplitudes of the miniature excitatory postsynaptic potentials suggests a multiterminal innervation. Following electrical stimulation of N2B nerve, excitatory postsynaptic potentials similar to those appearing spontaneously could be evoked. After repetitive stimulation, facilitation or summation of excitatory postsynaptic potentials was observed. The results obtained show that locust oviduct muscle has a double, motor and modulatory innervation.     

Intrinsic differences in sensitivity to 5-HT between high- and low-output terminals innervating the same target

The differential action of neuromodulators on synapses of various efficacy provides additional fine tuning of synaptic regulation beyond frequency induced plasticity. We used the well-characterized high- and low-output motor nerve terminals, of the tonic and phasic neuromuscular junctions (NMJs) in the walking leg extensor muscle of the crayfish, to investigate differential actions of serotonin (5-HT) since both terminals innervate the same target. The excitatory postsynaptic potentials of the tonic NMJ are enhanced to a greater extent than for the phasic NMJs during exposure to 5-HT (100 nM). Macropatch current recordings at identified sites along the motor nerve terminals and quantal analysis indicate that mean quantal content is substantially increased by 5-HT. The overall probability of vesicular release increases to a greater extent at tonic terminals than at phasic terminals when exposed to 100 nM 5-HT. Measures in the area (i.e. charge) of spontaneous quantal currents indicate no difference in postsynaptic receptivity to the glutamatergic synaptic transmission upon exposure to 5-HT. The results provide new details concerning differential modulation of low- and high-output synapses present on the same target tissue.     

Active zones and receptor surfaces of freeze-fractured crayfish phasic and tonic motor synapses

Deep and superficial flexor muscles in the crayfish abdomen are innervated respectively by small populations of physiologically distinct phasic and tonic motoneurons. Phasic motoneurons typically produce large EPSP's, releasing 100 to 1000 times more transmitter per synapse than their tonic counterparts, and exhibiting more rapid synaptic depression with maintained stimulation. Freeze-fracturing the abdominal flexor muscles yielded images of phasic and tonic synapse-bearing terminals. The two types of synapse are qualitatively similar in ultrastructure, displaying on the presynaptic membrane's P-face synaptic contacts recognized by relatively particle-free oval plaques which are often framed by the muscle fiber's E-face leaflet with its associated receptor particles. Situated within these presynaptic plaques are discrete clusters of large intramembrane particles, forming active zone (AZ) sites specialized for transmitter release. AZs of phasic and tonic synapses are similar: 80% had a range of 15–40 large particles distributed in either paired spherical clusters or in linear form, with a few depressions denoting sites of synaptic vesicle fusion or retrieval around their perimeters. The packing density of particles is similar for phasic and tonic AZs. The E-face of the muscle membrane displays oval-shaped receptor-containing sites made up of tightly packed intramembranous particles. Phasic and tonic receptor particles are packed at similar densities and the measured values resemble those of several other crustacean and insect neuromuscular junctions. Overall, the similarity between phasic and tonic synapses in the packing density of particles at their presynaptic AZs and postsynaptic receptor surfaces suggests similar regulatory mechanisms for channel insertion and spacing. Furthermore, the findings suggest that morphological differences in active zones or receptor surfaces cannot account for large differences in transmitter release per synapse.     

Regeneration of phasic synapses on a crayfish slow muscle following allotransplantation of a mixed phasic-tonic nerve

Separate phasic or tonic nerves allotransplanted to reinnervate a denervated slow superficial flexor muscle (SFM) in the abdomen of adult crayfish regenerate synaptic nerve terminals with phasic or tonic properties. To test competitive interactions between tonic and phasic axons, we allotransplanted the sixth abdominal ganglion with its third nerve root containing a mixture of phasic and tonic axons onto the denervated SFM. The resulting reinnervation of the SFM was compared to the normal innervation on the contralateral intact SFM, which receives innervation only from tonic motoneurons. Variable sizes of excitatory postsynaptic potentials indicated that 2–3 axons innervated each muscle fiber of the SFM in both the allotransplant and normal preparations. Compared to the normal tonic terminals on the intact contralateral side, the allotransplanted synaptic terminals had more phasic-like properties; specificially, they gave rise to larger synaptic potentials, had a lower mitochondrial content and contained a higher density of active zone dense bars per synapse. Moreover, prolific sprouting of the axons in the regenerated nerve, typical of phasic axons, points to more vigorous regeneration of phasic rather than tonic axons to the denervated SFM. In keeping with this prolific axon sprouting, there was both a much higher density of innervation in the allotransplanted SFM compared to the normal SFM, and a higher frequency of extrasynaptic active zones in regenerated terminals of the mixed nerve compared to those of the tonic nerve. Thus, an allotransplanted mixed nerve regenerates mainly phasic axons and synapses on the slow denervated SFM, demonstrating the instructive nature of the neuron in synapse specification, as well as the permissive nature of the target muscle.     

Alterations in neuromuscular junction morphology during fast-to-slow transformation of rabbit skeletal muscles

Summary Chronic low frequency stimulation of motor nerves results in transformation of muscle fibre phenotype from fast- to slow-twitch. We examined the light and electron microscopic structure of neuromuscular junctions in normally fast twitch muscles, tibialis anterior and extensor digitorum longus of rabbit after 3 weeks of stimulation to determine whether synaptic structure is also modified during fibre type transformation. Neuromuscular junctions of stimulated and unstimulated (control) tibialis anterior and extensor digitorum longus muscles and unstimulated slow twitch soleus muscle were visualized with rhodamine-conjugated -bungarotoxin. Video light microscopic images of neuromuscular junctions were digitized to allow quantification of their surface areas, perimeters, lengths and widths. Three weeks of stimulation resulted in a decrease in the maximal velocity of muscle fibre shortening and augmentation of mitochondrial volume in fast muscles, demonstrating the efficacy of the stimulation protocol employed in altering muscle fibre phenotype. Neuromuscular junctions of control tibialis anterior and extensor digitorum longus are thin, compact, and continuous, with complex branching patterns. In contrast, those of slow-twitch soleus are thicker and discontinuous. Neuromuscular junctions in control tibialis anterior and extensor digitorum longus are larger than those in soleus. Three weeks of stimulation causes a marked decrease in the size of neuromuscular junctions in tibialis anterior and extensor digitorum longus, as reflected in the significant reduction in neuromuscular junction surface area, length and width. Electron microscopy of these junctions suggests that secondary postsynaptic folds in stimulated muscles are more closely spaced. Also, axon terminals of stimulated muscles appear to contain more densely packed synaptic vesicles and mitochondria than controls. Decreases in neuromuscular junction dimensions can be partly explained by muscle fibre atrophy. However, the decrease in neuromuscular junction size is proportionately greater than that of muscle fibre diameter in both muscles, indicating that factors other than fibre atrophy may contribute to the reduced neuromuscular junction size in stimulated muscles. Neuromuscular junctions of stimulated tibialis anterior and extensor digitorum longus muscles exhibit some features characteristic of normal soleus neuromuscular junctions, indicating structural adaptations consistent with the altered muscle fibre phenotype. On the other hand, neuromuscular junctions of 3 week stimulated tibialis anterior and extensor digitorum longus and their synaptic branches remain as thin and continuous as those of unstimulated controls, suggesting that the transformation of neuromuscular junctions towards a morphology characteristic of slow muscle, is only partial. These results demonstrate that an altered pattern of impulse activity causes significant synaptic remodelling in adult rabbit skeletal muscles.     

Depression of synaptic efficacy in high- and low-output Drosophila neuromuscular junctions by the molting hormone (20-HE)

The molt-related steroid hormone, 20-hydroxyecdysone (20-HE), was applied to muscles 6 and 7 of third instar larval of Drosophila melanogaster neuromuscular junction preparations to examine if rapid, nongenomic responses could be observed as was shown recently to occur in crustacean neuromuscular junctions. At a dose of 10 microM, the excitatory junction potentials were reduced in amplitude within minutes. To elucidate the site of action of the hormone, focal-macropatch recordings of synaptic currents were obtained over the neuromuscular junctions. The results showed that the high-output (Is) and the low-output (Ib) motor nerve terminals, which innervate muscles 6 and 7, released fewer synaptic vesicles for each stimulation while exposed to 20-HE. Because the size and shape of synaptic currents from spontaneous releases did not change, the effects of the 20-HE are presynaptic. The rapid effects of this hormone may account in part for the quiescent behavior associated with molts among insects and crustaceans.     

Sub-miniature junction potentials in excitatory neuromuscular synapses of the locust

By recording miniature excitatory junction potentials (mejps) intracellularly at two points from a multiterminally innervated muscle fibre it is possible to select mejps whose amplitudes are not substantially affected by electrotonic decay. Many amplitude histograms of such selected mejps from untreated locust jumping muscle show a bimodal distribution with a high proportion of small-amplitude mejps (sub-mejps). Most amplitudes of excitatory junction potentials (ejps) resulting from the release of a single transmitter quantum correspond to the large-mode mejps. Tetanic nerve stimulation, in high [Mg2+]o without Ca2+, greatly reduces the proportion of sub-mejps. It is concluded that there are two modes of spontaneous transmitter release from the motor nerve endings.     

Synaptic vesicle recycling at the neuromuscular junction in the presence of a presynaptic membrane marker

Staining of the presynaptic axonal membrane of the neuromuscular junction with horseradish peroxidase-labeled α-bungarotoxin was utilized as a marker for observing directly the fate of this membrane during the process of synaptic vesicle release and recycling. The neuromuscular junctions of frog sartorius-sciatic nerve preparations were stained with horseradish peroxidase-α-bungarotoxin and stimulated by electrical stimulation of the nerve, high concentration of external potassium ions, and black widow spider venom. Some preparations were stimulated in the presence of exogenous horseradish peroxidase tracer after incubation in the conjugate and were found to contain horseradish peroxidase within many synaptic vesicles, indicating that the conjugate did not affect the process of synaptic vesicle recycling. Stimulation was followed by depletion of synaptic vesicles and appearance of axolemmal infoldings and membranous cisternae. With rest after electrical and potassium stimulation, synaptic vesicles were reconstituted and terminals assumed a more normal appearance. Membrane staining after stimulation occurred in the axolemmal infoldings, some of the intra-axonal cisternae, and in a few coated vesicles. However, all synaptic vesicles were unreactive, in either rested or unrested terminals. Thus, axonal membrane labeled with horseradish peroxidase-α-bungarotoxin did not become incorporated into new synaptic vesicles.These observations support a mechanism of recycling of synaptic vesicles by specific retrieval of vesicle membrane or constituents from the axolemma.     

Transient sensory neuromuscular junctions in developing rats.

Sensory nerve terminals form transient neuromuscular contacts at myotendinous junctions of rat muscles in regions of developing Golgi tendon organs. The terminals contact myotubular tips beginning with the 19th day of gestation and become detached from muscle fibres by the 4–5th postnatal day. Although the terminals are sensory—which has been confirmed by de-efferentation experiments—they mainly contain clear and dense cored vesicles, resembling in this respect motor endings. The width of the cleft varies from 20 to 60 nm; the basal lamina is either missing or interposed between the axolemma and the plasma membrane.Sensory terminals forming temporary contacts apparently induce gradual withdrawal of innervated myotubular tips from the aponeurosis, which triggers the differentiation of the tendon organ body. The mechanism underlying neural induction is not known; it is possible that the terminals release an inductive substance which affects the innervated myotubes.     

The effects of chronic stimulation on the morphology of the frog neuromuscular junction

Summary A quantitative study was made of the effects of 24 h continuous stimulation on the morphology of the frog neuromuscular junction. The synaptic vesicle concentration in the nerve endings of frog sartorius muscles stimulatedin vitro for 24 h at 2 Hz was the same as that in controls stimulated for only 0.3 h at 2 Hz. The control preparations were either freshly dissected or maintained at restin vitro for 23 h prior to stimulation. Chronically stimulated terminals differed from their controls only in having more cisternae and fewer dense cored vesicles. Varying the lengths of the nerves to both chronically stimulated andin vitro control muscles had little effect on the morphology of the nerve endings.Continuous recording of muscle twitch tension demonstrated that neurotransmission was effective throughout the 24 h period of stimulation. Additional evidence that nerve failure or degeneration was not a factor in the results came from a second set of control and chronically stimulated preparations that were tetanized at 30 Hz for 0.3 h before fixation. Changes attributable to rapid stimulation were evident in 87 to 100% of their nerve terminals.Although the distribution of membrane among various membrane organelles differed from one treatment group to another, the total amount of measurable membrane in the nerve terminals was the same in all of the treatment groups; that is, the total amount of membrane was not altered by maintenancein vitro, chronic stimulation at 2 Hz, rapid stimulation at 30 Hz, reduced nerve length, or any tested combination of these treatments. This conservation of total membrane suggests that membrane exchange between axon and nerve terminal occurs at a relatively slow rate which is unaffected by synaptic activity, and that the local mechanism for recycling synaptic vesicle membrane in frog neuromuscular junctions is more autonomous and durable than has been suspected.     

Reorganization of synaptic ultrastructure at facilitated lobster neuromuscular terminals

Summary Prolonged stimulation of the single excitor axon to the lobster distal accessory flexor muscle in the presence of ouabain caused long-term facilitation at its neuromuscular synapses. Hence the extracellularly recorded synaptic potentials failed less frequently and increased their mean amplitude, compared to the non-facilitated (control) potentials from homologous sites in the contralateral muscle. The fine structure of synaptic terminals between matched pairs of facilitated and control preparations was compared with the aid of serial section electron microscopy. Differences between facilitated and control preparations were similar both when the latter were bathed in normal saline or ouabain-containing saline, suggesting that the changes were related to the electrical stimulation rather than to the presence of ouabain. First, the facilitated terminals were smaller in surface area than the control. Second, the number and size of synaptic contacts in the facilitated terminals resembled those in the control. Third, presynaptic dense bodies or active sites increased in number although their sizes remained unaltered in the facilitated terminal. This increase is attributed to the addition of dense bodies at existing synaptic contacts since synaptic contacts remained constant in number between facilitated and control preparations. Fourth, the number and size of synaptic vesicles were unaffected by prolonged stimulation although there was a redistribution of vesicles such that they appeared to be channelled in distinct streams to synaptic contacts. Fifth, mitochondria increased in number and were situated closer to the dense bodies at facilitated nerve terminals than at control terminals. Overall, these changes denote considerable reorganization of the synaptic terminals associated with elevated transmitter release.     

Synaptic strength and horseradish peroxidase uptake in crayfish nerve terminals

Summary Uptake of horseradish peroxidase was studied by examining percentages of labelled synaptic vesicles in nerve endings of the excitatory axon innervating the opener muscle of the walking leg in the crayfish (Procambarus clarkii). Terminals on fibres with large excitatory postsynaptic potentials (EPSP) had higher percentage of labelled vesicles than terminals of fibres with small EJPs. The extent of labelling in the synaptic vesicle pool was greater for terminals with higher transmitter output. Evidence for three possible routes of synaptic vesicle formation was found. Movement of vesicles within the terminal as a whole appeared to be constrained, but rapid movement of vesicles within local populations probably occurs.     

Fine structural study of the abdominal muscle receptor organs of the crayfish (Procambarus clarkii). Sensory endings and synaptic structures

Summary The sensory endings, neuromuscular junctions and interneuronal synapses in the crayfish muscle receptor organ have been studied by electron microscopy. The dendrites of the receptor neuron terminate as endings which are either free in the connective tissue matrix of the central region of the receptor strands, or abut on the muscle membrane forming a specialized junction with a narrow cleft of about 18 nm. Efferent nerve endings are classified into three types on the basis of their fine structural features. Type 1 endings contain mainly spherical vesicles with a diameter of about 55 nm and a few large granular vesicles with a diameter of about 100 nm, and synapse exclusively on muscle fibres. Type 2 endings have a high proportion of elongated vesicles measuring about 30 × 80 nm and a few large granular vesicles, and synapse on both sensory neurons and muscle. Type 3 endings are characterized by the high electron density of the axoplasm and numerous large granular vesicles with a diameter of about 100 nm; they synapse only on the sensory neuron of the slow receptor unit.It is suggested that Type 1 endings are excitatory, and Type 2 and 3 endings are inhibitory. Several differences in postsynaptic structure were observed between the putative excitatory and inhibitory neuromuscular junctions. Axo-axonal synapses between endings of Type 1 and Type 2, the latter being presynaptic to the former, are also found. Functional implications and possible roles of these structures are discussed.     

The lateral vestibular nucleus of the toadfish Opsanus tau: Ultrastructural and electrophysiological observations with special reference to electrotonic transmission

The lateral vestibular nucleus of the toadfish Opsanus tau was localized by means of axonal iontophoresis of Procion Yellow. The ultrastructure of the lateral vestibular nucleus neurons was then correlated with their electrophysiological properties. The lateral vestibular nucleus consists of neurons of various sizes which are distributed in small clusters over a heavily myelinated neuropil. The perikarya and main dendrites of the large and the small neurons are surrounded by a synaptic bed, which is separated from the neighboring neuropil by a layer of thin astrocytic processes. The synaptic bed contains three main classes of axon terminals, club endings, large and small terminals, the first being quite infrequent. All the large terminals as well as the occasionally observed club endings contain a pure population of rounded synaptic vesicles. In some of the small axon terminals there are also rounded vesicles; however, the majority contain flattened vesicles or a pleomorphic population. These data indicate that the small terminals originate from different afferent sources. The synaptic interfaces of the large boutons and of the club endings bear three types of junctional complexes: attachment plates, gap junctions and active zones. Those showing both gap junctions and active zones were designated as morphologically ‘mixed synapses’. Gap junctions, although in large number, have only been observed at the synaptic interfaces between terminals with rounded vesicles and the perikarya or the dendrite of the lateral vestibular nucleus neurons. Therefore electrotonic coupling would only be possible by way of presynaptic fibers. Some axons observed in the neuropil were found to establish gap junctional complexes with two different dendritec profiles and this observation is in favour of electrotonic coupling by way of presynaptic terminals.Field and intracellular potentials were recorded in the lateral vestibular nucleus. The field potential evoked by stimulation of the vestibular nerve consisted of an early positive-negative wave followed by a slow negativity, and that evoked by spinal cord stimulation was composed of an antidromic potential followed by a slow negative wave. Vestibulo-spinal neurons were identified by their antidromic spikes. In these cells, stimulation of the ipsilateral vestibular nerve evoked an excitatory postsynaptic potential with two components. The short delay of the first component of this excitatory postsynaptic potential and its ability to follow paired stimulation at close intervals without reduction of the second response suggest that it is transmitted electrotonically from primary vestibular afferent fibers. By contrast the latency of the second peak of the vestibular evoked excitatory postsynaptic potential and its sensitivity to high stimulus frequencies are compatible with monosynaptic chemically mediated transmission from primary vestibular afferents. Spinal stimulation evoked graded antidromic depolarizations in vestibulo-spinal neurons. The latency of these potentials was too short to allow for chemical transmission through afferents or recurrent collaterals and suggests electrotonic spread of antidromic activity from neighboring neurons. An important finding is that the graded antidromic depolarizations can initiate spikes; thus coupling between neurons in the lateral vestibular nucleus is sufficiently close that a cell can be excited by activity spread from neighboring cells. Similar graded depolarizations were recorded in identified primary vestibular afferents; their latencies and time course indicate that they were brought about by electrotonic spread of postsynaptic potentials and spikes to the impaled presynaptic fibers; this confirms the morphological evidence that coupling between lateral vestibular nucleus neurons occurs, at least in part, by way of presynaptic vestibular axons. As the spinal stimulus strength was increased, these graded depolarizations became large enough to initiate spikes which presumably propagate to the vestibular receptors. Thus antidromic invasion of the presynaptic terminals may provide negative feedback by preventing their re-excitation at short intervals after a synchronous discharge of an adequate number of postsynaptic cells. Excitatory inputs to the neurons of the lateral vestibular nucleus were identified from the spinal cord and from the contralateral vestibular nerve. Long latency excitatory postsynaptic potentials large enough to excite the cells were recorded following spinal stimulation; the threshold intensity for evoking them was consistently higher than that adequate to generate the graded antidromic depolarizations. Field potentials recorded after stimulation of the contra lateral vestibular nerve consisted of an initial positive negative wave followed by a slow negative wave. the stimulus intensity for evoking these potentials was the same or slightly above the threshold for those evoked in the lateral vestibular nucleus on the stimulated side. Also lateral vestibular nucleus neurons exhibited excitatory postsynaptic potentials large enough to excite the cells following stimulation of the contralateral vestibular nerve. but no inhibitory postsynaptic potentials were detected. This lack of commissural inhibition indicates a qualitative difference between the central organization of these cells in the toadfish and in mammals.The presence of neurons in the lateral vestibular nucleus which send their axons to the labyrinth was confirmed by their heavy staining with Procion Yellow following axonal iontophoresis. In a number of vestibular neurons. abruptly rising spikes were evoked at short latencies after adequate stimulation of the ipsilateral vestibular nerve. Graded stimuli applied to the vestibular nerve evoked graded short latency depolarizations as well as long latency excitatory postsynaptic potentials in these presumed efferent neurons to the labyrinth; the former could indicate electrotonic coupling of the efferent cells or electrotonic transmission from primary afferents, resulting in a short latency feedback loop.From these studies, the synaptic organization of the lateral vestibular nucleus neurons is compared with that of the Mauthner cells of teleosts, and the possibility of a dual mode of transmission, electrical and chemical, by primary vestibular afferents is discussed.     

Visual identification of two kinds of nerve cells and their synaptic contacts in a living autonomic ganglion of the mudpuppy (Necturus maculosus).

1. Many of the nerve cells comprising the cardiac parasympathetic ganglion of the mudpuppy are spread out in a thin, transparent sheet of tissue, enabling one to see cellular details in living preparations with differential interference contrast optics. The aim of this study was twofold: to establish the morphology of the nerve cells and their synaptic connections by light and electron microscopy, and to determine which aspects of the ganglion's structure could be reliably identified in the living tissue. 2. There are two types of neurones in the ganglion: (a) principal cells that send post-ganglionic axons to cardiac muscle fibres, and (b) interneurones whose processes are confined to the ganglion. 3. Interneurones are distinguished from principal cells by the presence of numerous granular vesicles seen with the electron microscope, and by intense formaldehyde-induced fluorescence. The interneurones are thus similar to catecholamine-containing interneurones in autonomic ganglia of other vertebrates. 4. Principal cells are innervated by processes that terminate mainly on the cell body, forming up to forty-five synaptic boutons and covering, on the average, 5% of the perikaryal surface. The synaptic terminals are derived from three sources: (a) axons from the vagus nerves, (b) interneurones and (c) other principal cells. Vagal terminals contacting principal cells contain agranular vesicles typical of preganglionic cholinergic endings. At regions of contact between processes of interneurones and principal cells, the interneurones have granular vesicles focused at membrane specializations; in addition there are small areas of close plasma membrane apposition, probably gap junctions. Some of the contacts between principal cells are characterized by gap junctions; others are structurally similar to vagal endings but persist after vagal degeneration. 6. Interneurones are innervated by axons that make contact mainly with their processes. The axon terminals on processes of interneurones contain agranular vesicles similar to vagal terminals on principal cells. 7. In live preparations principal cells are distinguished from interneurones by their size and the appearance of their organelles. Synaptic contacts on principal cells could often be identified and, in some cases, large contacts from interneurones or those from other nearby principal cells could be traced back to their cell bodies of origin. The validity of these identifications was confirmed by subsequent electron microscopic examination of the same cells.     

The fine structure of innervated myotendinous cylinders in extraocular muscles of rhesus monkeys

Summary Many of the myelinated nerve fibres of the distal myotendinous region of rectus muscles terminate on muscle fibre tips. The terminal expansions contain aggregated, small clear vesicles and mitochondria. Neuromuscular clefts at the contacts measure 20–40 nm and are uninterrupted by a basal lamina; the sarcoplasm opposite the contacts is unmodified. Some terminals invaginate the muscle fibre tips and others contact the sides of processes formed by splitting of the tips. The muscle fibre termination, its tendon and the nerve fibre branches are encapsulated to form an end-organ averaging 125 m in length and described as a myotendinous cylinder.Approximately 350 innervated myotendinous cylinders were estimated to be present in the horizontal recti with rather fewer in the vertical rectus muscles. Many of them occur shortly before the main myotendinous junction. All muscle fibres contributing to myotendinous cylinders were identified as the compact, felderstruktur, multi-innervated variety with directly apposed myofibrils that are known to be non-twitch fibres. All felderstruktur fibre terminations examined were encapsulated but 19% of them were not innervated.The nerve terminals of myotendinous cylinders are similar to those described by Dogiel (1906) as palisade endings and it is argued that they meet the morphological criteria of sensory neuromuscular endings. Their disposition suggests a capacity to monitor felderstruktur muscle fibre contraction.     

Action of brown widow spider venom and botulinum toxin on the frog neuromuscular junction examined with the freeze-fracture technique.

1. Structural changes which normally accompany transmitter release at frog neuromuscular junctions are visualized with the freeze-fracture technique. The effects of brown widow spider venom and botulinum toxin were evaluated in terms of their ability to block or produce these structural changes. Changes produced by these neuropoisons were correlated with their known effects on neurotransmitter release. 3. Fusion of synaptic vesicles with the presynaptic plasmalemma, normally evoked by electrical stimulation, was abolished at neuromuscular junctions from frogs treated with botulinum toxin. 3. The concentration of large intramembranous particles in the presynaptic plasmalemma, an indication of the excess of synaptic vesicle fusion over recovery of synaptic vesicle membrane, was increased by treatment with brown widow spider venom, even in the presence of botulinum toxin. 4. When external calcium was present, sites of vesicle fusion induced by brown widow spider venom, as well as by electrical stimulation, were located mainly in the active zone. In the absence of external calcium, many plasmalemmal deformations, also though to be sites of vesicle fusion, were more evenly dispersed over the presynaptic surface of nerve terminals. 5. Botulinum toxin decreased the number of vesicle fusion sites in the active zone induced by spider venom in the presence of external calcium but had little effect on the number of fusion sites induced by spider venom in the absence of external calcium. 6. Nerve terminals soaked in a sodium-free Ringer solution were partially depleted of vesicles. Addition of spider venom to this Ringer did not cause additional depletion of vesicles. 7. Formation of cation-permeable channels in the presynaptic membrane could account for these effects of spider venom on the frog neuromuscular junction. Botulinum toxin blocks vesicle fusion by some means which is not yet understood.     

Scanning electron microscope study of neuromuscular junctions in different muscle fiber types in the zebra finch and rat

Examination by scanning electron microscopy revealed differences between neuromuscular junctions in the muscle fibers of the zebra finch (bird) and rat. The neuromuscular junctions between the anterior and posterior latissimus dorsi muscles of the zebra finch were compared. The junctions of the former, exclusively slow tonic fibers, were small and numerous along the long axis of a single muscle fiber. The synaptic depressions per junction were few. The junctions of the latter, exclusively fast twitch fibers, were large and consisted of more synaptic depressions than the former. Junctional folds were occasionally found in some depressions. The neuromuscular junctions between the extensor digitorum longus and soleus muscles of the rat were also compared. The former consisted almost entirely of fast twitch muscle fibers, whereas the latter consisted of both slow twitch fibers (75%) and fast twitch fibers (25%). The junctions in the extensor digitorum longus muscle were almost all labyrinthine gutters containing exclusively slit-like junctional folds. In the soleus muscle, two types of junctions were observed. One type was similar to that of the extensor digitorum longus muscle; the other was characterized by labyrinthine gutters containing sparse, narrow slit-like and pit-like junctional folds. We suggest from these structural differences of the subneural apparatuses that the junction of the fast twitch muscle is characterized by the subneural apparatus containing numerous slit-like junctional folds, and that of the slow twitch muscle fiber characterized by the apparatus containing sparse, narrow slit-like and pit-like junctional folds.     

Stimulation-associated changes in frog neuromuscular junctions. A quantitative ultrastructural comparison of rapid-frozen and chemically fixed nerve terminals

The ultrastructural effects of stimulation and subsequent rest were measured in frog neuromuscular junctions preserved by rapid-freezing and freeze-substitution, a method that minimizes fixation-associated membrane rearrangements. The effects were compared to those measured in junctions preserved by aldehyde fixation in order to identify artifacts attributable to the method of preservation. Effects of stimulation previously observed in tissue preserved by aldehyde fixation were evident in both the rapid-frozen and aldehyde-fixed neuromuscular junctions in the present study. Synaptic vesicles were reduced in number and cisternal profiles were increased. However, the sizes and shapes of the cisternae differed with the method of preservation. In addition, it was found that mitochondria underwent a change in shape with stimulation. This was accompanied by swelling in the fixed preparations, but not in the rapid-frozen ones. Fixation after stimulation also produced swelling of the nerve terminals, a stimulation-associated change not evident in preparations that were preserved by rapid-freezing. After stimulation and 60 min of rest, nerve terminals showed recovery towards control morphology, evidence that the effects of the stimulation parameters used in the study were reversible. This study, utilizing rapid-frozen material, confirms previous reports based on chemically fixed tissue that stimulation reduces the number of synaptic vesicles and increases the number of cisternae. The findings are in accord with the hypotheses of exocytotic neurotransmitter release and local recycling of synaptic membrane. In addition, the study emphasizes that accurate quantitative assessments of membrane redistribution in active secretory systems cannot depend on chemically fixed tissues. It also shows that mitochondria are susceptible to radical distortion by aldehyde fixatives, and that the degree of susceptibility differs with the physiological state of the tissue.