Mobility and cycling of synaptic protein-containing vesicles in axonal growth cone filopodia

The spatial distribution and coordination of vesicular dynamics within growth cones are poorly understood. It has long been thought that membranous organelles are concentrated in the central regions of growth cones and excluded from filopodia; this view has dramatically shaped conceptual models of the cellular mechanisms of axonal growth and presynaptic terminal formation. To begin to test these models, we studied membrane dynamics within axonal growth cones of living rat cortical neurons. We demonstrate that growth cone filopodia contain vesicles that transport synaptic vesicle proteins bidirectionally along filopodia and fuse with the filopodial surface in response to focal stimulation, allowing for both local secretion of vesicular contents and rapid changes in the plasma membrane composition of individual filopodia. Our results suggest a new model in which growth cone filopodia are actively involved in both emitting and responding to local signals related to axon growth and early synapse formation.     

Ultrastructural immunocytochemical localization of B-50/GAP43, a protein kinase C substrate, in isolated presynaptic nerve terminals and neuronal growth cones

Summary Accumulating evidence indicates that the neuron-specific B-50/GAP43, a substrate for protein kinase C, plays a role in neuronal differentiation and neuritogenesis during nervous tissue development and axonal regeneration. An ultrastructural immunocytochemical study on the localization of B-50 in presynaptic terminals (synaptosomes) and neuronal growth cones was carried out by means of cryoultramicrotomy with affinity-purified B-50 antibodies. Detection was accomplished with colloidal gold, conjugated either to protein-A or goat anti-rabbit immunoglobulins. In synaptosomes, isolated from the frontal cortex of 6-week-old rats, and in neuronal growth cones, isolated from forebrains of 5-day-old rats, the majority of B-50 is detected at the surrounding neuronal plasma membrane. In both neuronal growth cones and synaptosomes, a relatively small fraction of B-50 in the cytoplasm was not evidently associated with internal membranes. Our results indicate that B-50 is mainly located at the cytoplasmic face of the synaptosomal and neuronal growth cone plasma membrane. The similar B-50 localization in neuronal growth cones and synaptosomes suggests that, both in extending axons and mature synaptic terminals, B-50 may exert identical functions as a protein kinase C substrate at the plasma membrane.     

AMPA receptors regulate dynamic equilibrium of presynaptic terminals in mature hippocampal networks

The formation and disassembly of synapses in mature neuropil could provide a substrate to encode experience in the brain. Although there is evidence for postsynaptic spine dynamics in mature systems, contributions to circuit rearrangements by presynaptic terminals have remained unclear. We used hippocampal slice cultures from mice expressing spectral variants of green fluorescent protein (GFP) that are targeted to the membrane and/or synaptic vesicles in neuronal subsets to image identified presynaptic terminals. In mature tissues with no net change in synapse numbers, subpopulations of presynaptic terminals appeared and disappeared within 1-3 days. The three terminal types established by mossy fibers had distinct properties. High-frequency stimulation increased the fraction of dynamic terminals for 1-2 days, a process mediated by activation of AMPA receptors, protein kinase A (PKA) and protein synthesis. Thus, synaptic activity can make stable presynaptic terminals become dynamic, providing a candidate mechanism to convert experience into changes in network connectivity.     

Distribution of neurotransmitter secretion in growing axons

Neurotransmitter secretion from the nerve terminal is mediated by the fusion of synaptic vesicles with the plasma membrane. It is generally believed that neurotransmitter release in mature synapses is localized to the presynaptic nerve terminals. To probe the topology of neurotransmitter secretion along developing axons in culture, we recorded membrane currents from myocytes manipulated into contact with axons. At the early stages of growth, exocytotic events were detected along the axon as well as at the growth cone. At the later stages of growth, neurotransmitter secretion adopted the form of a smooth proximodistal gradient, with the highest level of activity at the growth cone region. Our results reveal the existence of a previously undetected early stage of axonal growth and suggest developmental regulation in the pattern of neurotransmitter secretion along the growing nerve processes.     

Presynaptic dendrites and serial synapses in the intermediate and deep layers of the cat superior coliculus.

Presynaptic dendrites (PSDs) which participate in the serial synapses have frequently been found in the intermediate and deep layers of the cat superior colliculus. The PSDs are presynaptic to small dendritic shafts or spines with symmetrical membrane thikening, and postsynaptic to axon terminals with asymmetrical synaptic contact. Two types of the axon terminals are observed, both of which contain pleomorphic vesicles.     

Synaptic organization of the nucleus raphe dorsalis of the cat

Summary This electron microscopic study describes the different types of synaptic terminals found in the nucleus raphe dorsalis of the adult cat. Serial section analysis was used extensively to confirm the nature of the synaptic contact established by the various classes of terminals.Five different classes of terminals are identified according to the shape and packing density of the synaptic vesicles and type of contact they establish. The most common class (RDI-type) contains densely packed, round, agranular synaptic vesicles and establishes asymmetrical synaptic contacts. A second class (RDII-type) also contains spherical synaptic vesicles, but establishes symmetrical synaptic contacts with dendrites of all sizes. Most of the terminals in these two classes contain a few dense-cored synaptic vesicles, but a small sub-group contains many dense-cored vesicles. A third, less frequent, class (RSI-type) contains sparsely packed spherical synaptic vesicles and the majority of these terminals have asymmetrical contacts. A fourth terminal class contains pleomorphic synaptic vesicles (P-type), contacts dendrites of all sizes, and usually establishes symmetrical synaptic contacts. Finally, boutons thought to be the vesicle-filled excrescences of dendrites (postsynaptic dendrites) are found and in some cases the dendritic origin of these profiles was confirmed by serial sectioning. Such boutons containing pleomorphic vesicles are presynaptic to other such dendrites as well as conventional dendrites, and are postsynapticto the other terminal types described.Somata within the nucleus exhibit somatic spines but receive few synaptic contacts. Most axo-somatic terminals contain either round or pleomorphic vesicles and have postsynaptic thickenings intermediate to the symmetric and asymmetric types.     

Electron microscopic study of GABA-immunoreactive neuronal processes in the superficial gray layer of the rat superior colliculus: their relationships with degenerating retinal nerve endings

Summary GABA-immunoreactive neuronal elements were detected in the stratum griseum superficiale or superficial gray layer of the rat superior colliculus in an electron microscopic study, using postembedding immunocytochemistry with protein A-gold as a marker. In addition to neuronal somata, two types of GABA-immunoreactive neuronal processes were observed. Numerous profiles of axon terminals (1 m in diameter) with clear round or pleomorphic synaptic vesicles and mitochondria were found to establish mostly symmetrical synaptic contacts with GABA-immunonegative dendrites of various diameters. Some axosomatic synapses could also be observed. The gold particle density in this axon terminal compartment was between seven and 13 times the background level. The stratum griseum superficiale also included GABA-immunoreactive dendrites, some of which contained clear synaptic vesicles. These dendritic profiles always formed the presynaptic component of dendrodendritic synaptic contacts. The density of the gold particles in the dendritic compartment, taken as a whole, was between three and 13 times the background level. Furthermore, the relationship between the GABA-immunoreactive neuronal elements and degenerating retinal nerve endings identified in the left stratum griseum superficiale following enucleation of the right eye was investigated after a 7-day survival period. The profiles of degenerating retinal nerve endings (0.7 m in diameter) were found to be devoid of any specific labelling. Most of the retinal boutons established axodendritic synapses of the asymmetrical type with an immunonegative dendrite, which was also contacted in some cases by a GABA-immunopositive axon terminal. Other retinal endings were presynaptic to GABA-immunopositive dendritic profiles with synaptic vesicles, some of which were found to contact in turn an unlabelled dendrite, thereby completing serial synaptic relationships. More rarely, retinal endings formed the presynaptic component of possible axoaxonic synapses with GABA-positive terminals presumed to be axonic in nature. It can be concluded that the retinal input to the superficial gray layer often converges with a GABAergic axonal input on a dendritic target, the neurotransmitter specificity of which is unknown. In other cases, retinal terminals synaptically contact GABA-immunolabelled conventional and presynaptic dendrites and probably also some axon terminals; this might provide an anatomical substrate for the control of GABA release from these GABAergic processes. These results indicate that transmitter GABA plays an important role in retinocollicular transmission.     

Redistribution of GAP-43 during growth cone developmentin vitro; immunocytochemical studies

Summary The growth-associated protein GAP-43 (B-50, F1, pp46), has been found in elongating axons during development and regeneration, and has also been associated with synaptic plasticity in mature neurons. We have examined the loss of GAP-43 labelling from cerebellar granule cells with immunocytochemical localization of a polyclonal antibody to GAP-43. One day after plating, the plasma membrane of cell bodies, neurites and growth cones were all labelled with anti-GAP-43. By 10 days, most of the cell body labelling was lost, and by 20 days the neuritic and growth cone labelling was greatly reduced. Beginning at six days, anti-GAP-43 labelling of growth cones, which was initially uniform, became clustered. When growth cones were double-labelled with antibodies to GAP-43 and the synaptic vesicle protein, p65, inverse changes in the distribution of label was observed. While growth cone labelling with anti-p65 increased from three to 20 days in culture, GAP-43 label began to be lost from some growth cones by six days and showed continuing decline through 20 days. For individual growth cones, the loss of GAP-43 appeared to parallel the accumulation of p65, and first growth cones to lose GAP-43 appeared to be the first to accumulate p65 label. When cultures were grown on a substrate of basement membrane material, the time frames of neuritic outgrowth, loss of GAP-43 labelling, and increase in p65 labelling were all accelerated. At five days, labelling for GAP-43 was weak and labelling for p65 was strong, in a pattern comparable to that seen in older cultures on a polylysine substrate. These results suggest several conclusions concerning the expression and loss of GAP-43 in cultured cerebellar granule neurons. First, GAP-43 label is initially distributed in all parts of these cells. With increasing time in culture the label is first lost from cell bodies and later from neurites and growth cones. Second, the loss of GAP-43 label from growth cones is correlated with the appearance of the synaptic vesicle protein p65. Finally,in vitro developmental changes in the loss of GAP-43 can be altered by changing the growth substrate.     

GABAergic neurons in the mouse superficial dorsal horn with special emphasis on their relation to primary afferent central terminals

Immunocytochemical studies were carried out on the morphological relation between primary afferent central terminals (C-terminals) and GABAergic neurons in the mouse superficial dorsal horn. The superficial dorsal horn is composed of many synaptic glomeruli comprising two types: Type I with centrally located CI-terminals surrounded by several dendrites and few axonal endings, and Type II with centrally located CII-terminals surrounded by several dendrites and a few axonal endings. The CI-terminals are sinuous or scalloped with densely packed agranular synaptic vesicles, a few granular synaptic vesicles and mitochondria, and show an electron dense axoplasm, whereas the CII-terminals are large and round or rectangular with evenly distributed agranular synaptic vesicles, a number of granular synaptic vesicles and mitochondria, and show an electron opaque axoplasm. The immunoreaction of GABA was remarkable in the superficial laminae of the dorsal horn. Many interneuronal somata in the substantia gelatinosa showed GABAergic immunoreactivity. The immunoreaction was seen in the entire GABAergic neuroplasm, but not in the nucleus and its envelope. Most GABAergic features appeared as dendrites making postsynaptic contact with CI- or CII-terminals; i.e., numerous C-terminals made presynaptic contact with GABAergic dendrites. GABA immunoreactivity was seen over round synaptic vesicles and mitochondrial membranes. A few CII-terminals made presynaptic contact with GABAergic interneuronal somata. Previous physiological and anatomical studies have suggested that not only the cutaneous nociceptive primary afferent C-terminals but also mechanoreceptive primary afferent C-terminals make presynaptic contact with the GABAergic dendrites, boutons and soma. The presynaptic relation of these primary afferents with GABAergic neurons seems to provide morphological support for the essential feature of the gate control theory: primary afferent fibers may play a part in the modulation of nociceptive information via GABAergic neurons in the superficial dorsal horn. Small GABAergic terminals were found to make contact with blood capillaries suggesting the release of GABA into circulation.     

Development of synaptic structure and function in organotypic cultures of the rat hippocampus

The sequence of development of synapses, as well as the ultrastructure of axonal growth cones, has been investigated electron microscopically in tissue cultures of the newborn rat hippocampus. During differentiation of the tissue cultures, the formation of synapses is preceded by identifiable growth cones. A characteristic feature of axonal growth cones is the presence of numerous large clear vesicles which vary in diameter from ~100 to 150 nm. The first immature synapses were formed on the 5th, 6th or 7th day in vitro on the growth cones of differentiating neuronal processes. Axonal growth cones are occasionally found to be presynaptic to a dendrite. At first axo-dendritic synapses, most of them being en passant, arise, whereas axo-somatic and axo-spinous-dendritic synapses of different complex structures appear later.It is suggested that the earliest signs of synaptogenesis are vesicular structures (‘growth’ vesicles and few synaptic vesicles), which occur in growth cones, axons and presynaptic boutons of immature synaptic contacts even before formation of the specialized pre- and postsynaptic membranes.     

TRPC5 is a regulator of hippocampal neurite length and growth cone morphology

Growth cone motility is regulated by both fast voltage-dependent Ca2+ channels and by unknown receptor-operated Ca2+ entry mechanisms. Transient receptor potential (TRP) homomeric TRPC5 ion channels are receptor-operated, Ca2+-permeable channels predominantly expressed in the brain. Here we show that TRPC5 is expressed in growth cones of young rat hippocampal neurons. Our results indicate that TRPC5 channel subunits interact with the growth cone-enriched protein stathmin 2, are packaged into vesicles and are carried to newly forming growth cones and synapses. Once in the growth cone, TRPC5 channels regulate neurite extension and growth-cone morphology. Dominant-negative TRPC5 expression allowed significantly longer neurites and filopodia to form. We conclude that TRPC5 channels are important components of the mechanism controlling neurite extension and growth cone morphology.     

Modulation of growth cone calcium current is mediated by a PTX-sensitive G protein.

Growth cones of isolated neurons B5 of Helisoma were voltage clamped in the whole-cell configuration. Depolarization of growth cones to -20 mV or greater activated a high-voltage-activated (HVA) calcium current. Addition of the neuropeptide FMRFamide (1 microM), which causes a presynaptic inhibition of synaptic transmission, reversibly reduced the calcium current magnitude. This inhibitory effect is mediated by a pertussis toxin (PTX)-sensitive G protein. Dialysis with the non-hydrolyzable GTP analogs GTP gamma S and Gpp(NH)p caused FMRFamide's effect to become irreversible. Dialysis with GDP beta S or preincubation with PTX prevented FMRFamide from reducing the calcium current. Thus, one role of growth cone G proteins is to modulate ion channels in growth cone membrane which in turn may control growth cone motility.     

Growth cones in differentiated neuroblastoma: A time-lapse cinematographic and electron microscopic study

Summary Growth cones of differentiating neuroblastoma cells in monolayer culture were studied by time-lapse cinematography and electron microscopy. Morphological differentiation, and thus growth cone formation, was induced by the glucocorticoid dexamethasone. Growth cones lengthened gradually at an average rate of 30 m/h, advancing in stages that involved alternating extensions and retractions of the filopodia and lamellar sheets. During neurite growth the cell body usually remained stationary. The ultrastructure of growth cones was typified by several filopodia, each containing a bundle of microfilaments, agranular endoplasmic reticulum, aggregates of large agranular vesicles lying adjacent to filopodia (previously termed vesicle-filled mounds), many dense-cored vesicles, 100–140 nm in diameter, microtubules, bizarre and distorted mitochondria, and scattered free ribosomes. Comparing the findings with previous ultrastructural accounts of growth cones of cultured ganglion cells, similarities outnumbered differences. The organization of the microfilament bundles and the abundance of free ribosomes were remarkable in the neuroblastoma cell as was the profusion of dense-cored vesicles which were most numerous in the proximal portion of the growth cone.     

Functional GABA(B) receptors are expressed at the cone photoreceptor terminals in bullfrog retina

GABA(B) receptors at the cone terminals in bullfrog retina were characterized by immunocytochemical and whole-cell patch clamp techniques in retinal slice preparations. Somata, axons and synaptic terminals (pedicles) of cones were both GABA(B) receptor (GABA(B)R) 1 and GABA(B)R2 immunoreactive. Physiologically, barium/calcium currents of cones to voltage steps were significantly reduced in size when GABA was puffed to cone terminals in the presence of picrotoxin that is supposed to block both GABA(A) and GABA(C) receptors. Similar reduction in barium currents was obtained with puff application of baclofen to cone terminals. These results suggest the presence of functional GABA(B) receptors at the bullfrog cone terminals. Suppression of barium currents of cones by baclofen was dose-dependent. Moreover, barium currents of cones were potentiated by background illumination, as compared with those recorded in the dark. 6,7-Dinitroquinoxaline-2,3-dione, an antagonist of non-NMDA receptors that hyperpolarizes horizontal cells and reduces GABA release from these cells, and saclofen, a GABA(B) receptor antagonist, both potentiated barium currents of cones in the dark, thereby mimicking the effects of background illumination. It is suggested that changes in calcium influx into the cone synaptic terminals due to activation of GABA(B) receptors may provide a negative feedback mechanism for regulating signal transmission between cones and second-order neurons in the retina by modifying the amount of glutamate released from the cones.     

Primary afferent terminals in the nucleus of the solitary tract of the frog: An electron microscopic study

In the frog solitarius nucleus, primary afferent terminals of the facial and glossopharyngeal-vagal nerves were identified with cobalt labelling and electron microscopy. The labelled terminals were grouped in two main categories, one with small (1-2 micron) and pale terminals, and another with large (3-5 micron) and dark terminals. The small terminals greatly outnumbered the large ones. In addition many terminals intermediate in size and staining reactions were found. All kinds of labelled boutons contained medium-size clear synaptic vesicles, among which dense-core vesicles of the smaller type frequently occurred. The labelled primary afferent terminals established axo-dendritic contacts of the asymmetric type. Close to these contact sites they were themselves very frequently contacted by a profile interpreted as presynaptic in relation to them. Such profiles contained spherical, pleomorphic (including dense-core) or flattened vesicles; a fourth kind was interpreted as presynaptic dendrites. It is concluded that viscerosensory fibres, as opposed to somatosensory fibres, predominantly generate small and lightly stained terminals. It is likely that the effect of synaptic transmission at the solitarius tract terminals is modulated in a very versatile manner by the various presynaptic profiles converging on these terminals.     

The cytoskeletons of isolated, neuronal growth cones

We have examined by electron microscopy the cytoskeletons of growth cones isolated from neonatal rat forebrain by the method of Gordon-Weeks and Lockerbie [Gordon-Weeks and Lockerbie (1984) Neuroscience 13, 119-136]. When fixed in suspension with conventional fixatives, isolated growth cones contain a central region filled with a branching system of smooth endoplasmic reticulum and a cortical region immediately beneath the plasma membrane that is relatively free of organelles and is composed of an amorphous granular cytoplasm. The filopodia of isolated growth cones are also devoid of organelles and contain a cytoplasm that is similar in appearance to that in the cortical region. No microtubules or neurofilaments have been found in these growth cones. When isolated growth cones were prepared for electron microscopy by a method which preserves actin filaments [Boyles, Anderson and Hutcherson (1985) J. Histochem. Cytochem. 33, 1116-1128], microfilaments were found throughout the cortical cytoplasm. In the filopodia, the microfilaments were bundled together and oriented longitudinally. Filopodial microfilament bundles often extended into the body of the growth cone and could traverse it completely. Inclusion of Triton X-100 (1% v/v) in the fixative solubilized the membranes and soluble cytoplasmic proteins of growth cones, allowing an unobscured view of the microfilament cytoskeleton including the core bundle of microfilaments in filopodia. Suspended within the cytoskeleton were the coats of coated vesicles. These were particularly numerous at the broad bases of filopodia. Microfilaments bound heavy meromyosin and were cytochalasin B (2.0 X 10(-7) M) sensitive. Individual microfilaments branched and within filopodia they were extensively cross-linked by thin (7 nm) filaments. Microtubules and neurofilaments were not seen in these cytoskeletons despite the fact that the fixative contained a Ca2+ chelator. When growth cones were preincubated in taxol (14 microM) their cytoskeletons were found to contain microtubules. These were located mainly in the centre of the growth cone, were absent from the filopodia and were contiguous with microfilaments. We conclude that the cytoskeletons of isolated neuronal growth cones from neurones of the central nervous system are mainly composed of actin microfilaments. Although microtubules are not normally present, there is a pool of soluble tubulin which will form microtubules in the presence of taxol. This may imply that those microtubule-associated proteins that promote tubulin polymerization are absent in the growth cone or are below the concentration threshold for polymerization.     

The distribution and movement of organelles in maturing growth cones: correlated video-enhanced and electron microscopic studies

Summary The morphology of growth cones from identified neurons ofAplysia californica was analysed both with video-enhanced contrast differential-interference contrast (VEC-DIC) microscopy, and through serial electron microscopic reconstructions of the same growth cones. The largest structures seen in the living growth cones, the large irregular refractile bodies (LIRBs), were shown in electron micrographs to be unique structures, composed predominantly of dense-core vesicles but including mitochondria and smooth membrane profiles. The LIRBs were stratified in the growth cones, occurring predominantly in sections distant from the substrate and relatively devoid of microtubules. VEC-DIC observations showed that LIRBs formed in the peripheral regions of the organelle-rich central growth cone, and grew in size through fusion with other LIRBs, accumulating into a large central mass in more proximal regions. The distribution of microtubules and LIRBs and the movements of LIRB suggest that there is an overall circulatory pattern in the growth cones, with the delivery of new vesicles occurring at distal areas close to the substrate, and the accumulation and retrograde processing of organelles occurring in more proximal areas away from adhesive contacts.     

Activity-dependent liberation of synaptic neuropeptide vesicles

Despite the importance of neuropeptide release, which is evoked by long bouts of action potential activity and which regulates behavior, peptidergic vesicle movement has not been examined in living nerve terminals. Previous in vitro studies have found that secretory vesicle motion at many sites of release is constitutive: Ca(2+) does not affect the movement of small synaptic vesicles in nerve terminals or the movement of large dense core vesicles in growth cones and endocrine cells. However, in vivo imaging of a neuropeptide, atrial natriuretic factor, tagged with green fluorescent protein in larval Drosophila melanogaster neuromuscular junctions shows that peptidergic vesicle behavior in nerve terminals is sensitive to activity-induced Ca(2+) influx. Specifically, peptidergic vesicles are immobile in resting synaptic boutons but become mobile after seconds of stimulation. Vesicle movement is undirected, occurs without the use of axonal transport motors or F-actin, and aids in the depletion of undocked neuropeptide vesicles. Peptidergic vesicle mobilization and post-tetanic potentiation of neuropeptide release are sustained for minutes.     

GAP-43 distribution is correlated with development of growth cones and presynaptic terminals

Summary GAP-43 (F1, B-50, pp46) has been associated with neuronal development and regeneration, but precise localization within neurons is not known. Pre-embedding electron microscopic immunocytochemistry using silver-enhanced 1 nm gold particles was used to localize GAP-43 label in cell cultures of cerebellar neurons. In the plasma membranes of early cultures, high levels of GAP-43 were seen in all parts of the neuron. In older cultures, consistent with previous reports, the first loss of GAP-43 label was seen in the soma and then the axon. Growth cones had high levels of GAP-43 label on the plasma membrane, with increased distribution over unattached relative to attached filopodia. The amount of GAP-43 seen over the plasma membrane of forming presynaptic terminals is lower than over growth cones, indicating a possible correlation between the presence of GAP-43 and the stage of presynaptic terminal development. Intracellular GAP-43 in axons and growth cones was highest in membranes of smooth cisternae. The levels of GAP-43 in smooth cisternae in axons fell by seven days in culture while the levels of GAP-43 in smooth cisternae of growth cones fell at 14 days. When mini-explant cerebellar cultures were examined with light microscopic immunocytochemistry, GAP-43 label of plasma membrane was highest at the periphery of the radial axonal outgrowth, suggesting that addition of GAP-43 to the plasma membrane can occur in the distal axon or at the growth cone.     

Filopodia, lamellipodia and retractions at mouse neuromuscular junctions

Summary In order to determine if mature motor nerve terminals retain structures associated with development such as filopodia and lamellipodia, we studied whole mounts of mature mouse neuromuscular junctions stained with both fluorescent-labelled tetanus toxin C-fragment and alpha-bungarotoxin, and employed electron microscopy in parallel. The rapid fluorescent stain may be of general usefulness. Both filopodia and lamellipodia were found, extending beyond the border of the established postsynaptic receptors. Filopodia often appeared in clusters, were devoid of a synaptic vesicle antigen, and many withdrew in response to cytochalasin D. Control experiments demonstrated that filopodia were not induced by the toxin treatment. The mean number of filopodia per endplate varied from about one in phasic muscle to three in tonic muscle, and was twice as great in immature mouse muscle. Postsynaptic receptor-rich regions without overlying terminals were less numerous than filopodial and lamellipodial projections without underlying receptors. Electron microscopy showed that lamellipodia contained actin-like filaments and immunoreactivity to actin, but no neurofilaments, microtubules, mitochondria or vesicles. Therefore, these structures would not be visualized byin vivo mitochondrial stains. The lamellipodia protruded into the gap between muscle and a closely overlying Schwann cell process. Lamellipodia occupied about 5% of the linear extent of the terminal arbor in whole mounts, but appeared in 16% of random electron micrographic fields. Thus, the lamellipodia and filopodia typical of developing terminals are present in adulthood and represent a distinctive specialization of the nerve terminal, which may interact with the adjacent Schwann and muscle cell. The frequency of filopodia is a function of age and of muscle or motoneuron type. We suggest that some of the factors known to regulate growth of filopodia and lamellipodiain vitro or in development may continue to act at adult presynaptic nerve terminals.