Postmetamorphic development of neuromuscular junctions and muscle fibers in the frog cutaneous pectoris

Synaptic size, synaptic remodelling, polyneuronal innervation, and synaptic efficacy of neuromuscular junctions were studied as a function of growth in cutaneous pectoris muscles of postmetamorphic Rana pipiens. Recently metamorphosed frogs grew rapidly, and this growth was accompanied by hypertrophy of muscle fibers, myogenesis, and increases in the size and complexity of neuromuscular junctions. There were pronounced gradients in pre- and postsynaptic size across the width of the muscle, with neuromuscular junctions and muscle fibers near the medial edge being smaller than in more lateral regions. The incidence of polyneuronal innervation, measured physiologically and histologically, was also higher near the medial edge. Growth-associated declines in all measures of polyneuronal innervation indicated that synapse elimination occurs throughout life. Electrophysiology also demonstrated regional differences in synaptic efficacy and showed that doubly innervated junctions have lower synaptic efficacy than singly innervated junctions. Repeated, in vivo observations revealed extensive growth and remodelling of motor nerve terminals and confirmed that synapse elimination is a slow process. It was concluded that some processes normally associated with embryonic development persist long into adulthood in frog muscles.     

Elevated levels of polyneuronal innervation persist for as long as two years in reinnervated frog neuromuscular junctions

When the nerve to an adult frog sartorius muscle is crushed, and axons are allowed to regenerate, the level of polyneuronal innervation at reinnervated neuromuscular junctions is higher than normal. With time, much of this polyneuronal innervation is reduced by the process of synapse elimination (Werle and Herrera, 1988). Using intracellular recording, we estimated the level of polyneuronal innervation in adult frog (Rana pipiens) sartorius muscles 2 years (range: 1.7-2.4 years) after crushing the sartorius nerve. We found that 27% (S.E. = 1.4%) of the junctions in muscles 2 years after reinnervation were polyneuronally innervated, whereas only 10% (S.E. = 1.2%) of the junctions in normal frog muscles were polyneuronally innervated. Thus, the synapse elimination that occurs following reinnervation does not restore the normal level of polyneuronal innervation. Histological comparisons of junctional structure between muscles 2 years after reinnervation and normal muscles revealed substantial differences. Reinnervated junctions had a greater length of synaptic gutter apposed by nerve terminal processes, more axonal inputs, more empty synaptic gutter, more instances of single synaptic gutters innervated by more than one axon, and longer lengths of nerve terminal processes that connect synaptic gutters within a junction. On the basis of this physiological and anatomical evidence, we conclude that nerve injury causes persistent changes in the pattern of muscle innervation.     

The relationship of neuromuscular synapse elimination to synaptic degeneration and pathology: Insights from WldS and other mutant mice

Neuromuscular synapse elimination, Wallerian degeneration and peripheral neuropathies are not normally considered as related phenomena. However, recent studies of mutant and transgenic mice, particularly the Wld ^S mutant—in which orthograde degeneration is delayed following axotomy—suggest that re-evaluation of possible links between natural, traumatic and pathogenic regression of synapses may be warranted. During developmental synapse elimination from polyneuronally innervated junctions, some motor nerve terminals progressively and asynchronously vacate motor endplates. A form of asynchronous synapse withdrawal, strongly resembling synapse elimination, also occurs from mononeuronally-innervated motor endplates following axotomy in young adult Wld ^S mutant mice. A similar pattern is observed in skeletal muscles of several neuropathic mutants, including mouse models of dying-back neuropathies, motor neuron disease and—remarkably—models of neurodegenerative diseases such as Huntington's and Alzheimer's diseases. Taken together with recent analysis of synaptic remodelling at neuromuscular junctions in Drosophila, a strong candidate for a common regulatory mechanism in these diverse conditions is one based on protein ubiquitination/deubiquitination. Axotomised neuromuscular junctions in Wld ^S mutant mice offer favourable experimental opportunities for examining developmental mechanisms of synaptic regression, that may also benefit our understanding of how degeneration in the synaptic compartment of a neuron is initiated, and its role in progressive, whole-cell neuronal degeneration.     

Synaptic competition and the elimination of polyneuronal innervation follwoing reinnervation of adult frog sartorius muscles

The elimination of polyneuronal innervation (synapse elimination) that occurs following reinnervation was studied in sartorius muscles of adult Rana pipiens. The percentage of neuromuscular junctions that were polyneuronally innervated declined from 47% at 40–80 days after nerve crush to 22% at greater than 250 days after nerve crush. We measured the size, synaptic strength, and position of competing nerve terminals at identified dually innervated neuromuscular junctions at these two different periods of synapse elimination. Our goal was to determine if any of these parameters play a role in the competition between nerve terminals that ultimately results in the elimination of polyneuronal innervation. Our data support the hypothesis that polyneuronal innervation will persist if competing nerve terminals are of similar synaptic efficacies but will be eliminated if the competing terminals are of different synaptic efficacies. We also tested, but failed to find any evidence, that the spatial proximity of competing nerve terminals at the same synaptic site influences the elimination of polyneuronal innervation.     

The role of synapse elimination in the establishment of neuromuscular compartments

To investigate the specificity of development of initial neuromuscular connections, we examined the compartmental distribution of synapses in neonatal rat lateral gastrocnemius (LG) muscle. Initial neuromuscular connections might be restricted to the compartmental territories present in adults; alternatively, synapse elimination could establish the compartments from a less precise pattern of innervation. We examined 46 pups of ages 0 to 14 postnatal days using a variety of techniques. The principle method was evoked electromyographic (EMG) activity in response to nerve stimulation. The nerve branch to one neuromuscular compartment was cut and the remainder of the nerve was stimulated. The presence of EMG activity was used to identify the areas of muscle contracting in response to nerve stimulation. After cutting a particular branch, EMG activity generally could not be recorded from the denervated compartment. These results indicate that the pattern of innervation at birth is essentially compartment-specific, and that neuromuscular compartments are not shaped from some less precise pattern by postnatal synapse elimination. The factors which operate prenatally to determine this high degree of specificity in neuromuscular connectivity seen at the time of birth, however, remain unknown.     

Lose it to use it

In this issue, Wang et al. (2021. J. Cell Biol. https://doi.org/10.1083/jcb.201911114) describe a phenomenon in which neuromuscular junction synapse elimination triggers myelination of terminal motor axon branches. They propose a mechanism initiated by synaptic pruning that depends on synaptic activity, cytoskeletal maturation, and the associated anterograde transport of trophic factors including Neuregulin 1-III.

Neuromuscular junctions (NMJs) are a favorite model system to study the development, maintenance, and function of neuronal synapses because of their accessibility, size, and simplicity. Although many synaptic mechanisms discovered at the peripheral NMJ have provided important insights into synaptic mechanisms in the central nervous system (CNS), the phenomena of synapse elimination and refinement remain poorly understood in both. In the peripheral nervous system (PNS), synapse elimination is an essential developmental step that removes redundant presynaptic inputs to the muscle fiber. In addition, peripheral motor axon terminals must become myelinated to facilitate rapid and synchronized acetylcholine release to the muscle fiber. However, whether these two essential events during PNS development are coordinately regulated remains unknown.The immature rodent NMJ is first innervated by many axons which are then removed until the synapse reaches a dually innervated state (1). These two axons then further compete for synaptic territory, leaving one “winner” that eventually occupies the motor endplate by the end of the second postnatal week. To determine the relationship between synapse elimination and myelination, Wang et al. (2) used the formation of paranodal junctions between axons and Schwann cells as a surrogate for myelination and then determined whether axons that occupied NMJs in a singly or dually innervated state were more or less likely to be myelinated. They found that when the NMJ is dually innervated, myelination of the terminal axon branch is inhibited; neither synaptic occupancy of the competing axons nor axon diameter influenced myelination. However, once synapse elimination at the NMJ is complete, i.e., a single axon terminal innervates the motor endplate, the winner branch becomes myelinated. Thus, synapse elimination precedes myelination of the terminal axon branch, and competition between dually innervated NMJs restricts myelination.What mechanisms regulate the coordinated maturation of the motor neuron, Schwann cell, and muscle circuit? Since previous studies showed that synapse elimination at the NMJ depends on muscle activity (3), Wang et al. (2) inhibited synapse elimination by blocking acetylcholine receptors with α-bungarotoxin (α-Btx). This inhibition of motor endplate and muscle activity increased not only the number of dually innervated NMJs, but also significantly decreased myelination of terminal axon branches of singly innervated NMJs. Thus, neuromuscular activity must induce retrograde signaling mechanisms that promote not only synapse elimination but also myelination.During synapse elimination, the microtubule cytoskeleton of retracting axons is degraded and reduced (4). In contrast, axons that singly innervate NMJs have a higher microtubule content. α-Btx–dependent block of neuromuscular transmission reduced microtubule content in axons that singly innervate NMJs. Thus, α-Btx treatment simultaneously reduces both microtubule content and myelination.To determine if a mature microtubule-based cytoskeleton is causally related to myelination, Wang et al. used spastin knockout (spastin^KO) mice to artificially stabilize microtubules. Although spastin^KO mice had delayed axon branch removal, stabilization of the microtubule cytoskeleton increased myelination of axons that dually innervated NMJs. Thus, the brake that synaptic competition normally places on terminal branch myelination can be overcome by increasing the mass and maturity of the microtubule cytoskeleton.How does axonal microtubule stability influence terminal axon myelination? Microtubules participate in the anterograde and retrograde transport of diverse cargoes including mitochondria and growth factors. To determine if anterograde axonal transport promotes myelination of axons that singly innervate NMJs, Wang et al. used a dominant-negative mutant of kinesin-1 heavy chain which binds cargo, but lacks the protein’s motor domain, thereby impairing transport. After confirming transport inhibition by tracking impaired movement of the β1 subunit of voltage gated sodium channels, they found that myelination and node of Ranvier formation were significantly delayed in singly innervated NMJs expressing the dominant negative kinesin. Taken together, these results suggest that synapse elimination promotes maturation of the microtubule cytoskeleton which allows more efficient delivery of promyelinating signals to the terminal branch.What could these promyelinating signals be? One obvious candidate is Neuregulin 1 type III (Nrg1-III), which has long been known to promote myelination of peripheral nervous system axons (5). Consistent with this idea, conditional deletion of Nrg1-III dramatically reduced the number of myelinated axon terminals that singly innervate NMJs but did not alter the number of dually innervated NMJs. In contrast, overexpression of Nrg1-III in a transgenic mouse removed the competition-dependent block on myelination resulting in more myelination of both dually and singly innervating axon terminals. In these same transgenic Nrg1-III mice, among those NMJs that were singly innervated, their corresponding axons had higher levels of Nrg1-III. Remarkably, even in these same transgenic overexpressers, inhibition of muscle activity reduced the amount of Nrg1-III found on singly innervated axons, consistent with the observed impairment of the microtubule-based cytoskeleton after α-Btx treatment. ERK1/2 and AKT are downstream effectors of Nrg1-III in Schwann cells and implicated in the myelination pathway. Immunostaining of Schwann cells ensheathing singly innervating axon terminals revealed higher levels of pERK and pAKT.Taken together, the experiments performed by Wang et al. (2) suggest that as multiple axons actively compete for synaptic dominance at the NMJ, the myelination of their terminal branches is delayed. Upon synapse elimination, neuromuscular activity promotes a retrograde signal that increases maturity of the microtubule cytoskeleton. Maturation of the microtubule-based cytoskeleton facilitates the transport of promyelinating signals like Nrg1-III which, when presented to Schwann cells, results in myelination of the “winner” terminal axon branch of a singly innervated NMJ (Fig. 1).Open in a separate windowFigure 1.Synapse elimination promotes myelination of terminal motor axon branches. During early development, NMJs are innervated by multiple axons that compete for endplate territory. During this time, the terminal branches of the axons are not myelinated, and the tubulin cytoskeletal network remains immature. Synaptic activity induces elimination of redundant connections, which leads the winner axon’s microtubule-based cytoskeleton to mature and increase, while the microtubule cytoskeleton is degraded in the retracting axon. The maturity of the cytoskeleton allows for kinesin dependent anterograde transport of Neuregulin 1-III, which then initiates a promyelination signaling cascade via AKT and ERK activation.To the best of our knowledge, this is the first demonstration of plasticity of myelination downstream of activity and synapse refinement in the peripheral motor nervous system. Many studies in the CNS demonstrate that de novo myelination occurs in response to neuronal activity and learning paradigms (6, 7), although the mechanisms responsible remain unknown. Thus, synapse refinement and elimination-dependent myelination may be a paradigm to uncover mechanisms of learning- and activity-dependent myelination in the CNS. Functionally, the addition of myelin to the terminal motor axon branch promotes efficient neurotransmitter release through faster action potential propagation, improved metabolic support of the axon, and more efficient depolarization of the presynaptic terminal by clustered Na⁺ channels at the terminal heminode (8). Whether any or all of these benefits also exist in the CNS remains unknown.This is also the first demonstration of postsynaptic activity driving myelination of a presynaptic axon. Although it is clear that a retrograde signal from the muscle promotes the further maturation and subsequent myelination of the terminal axon, the identity of this cue is unknown. One interesting candidate for a muscle-derived competition and axonal maturation cue is the neurotrophin brain-derived neurotrophic factor (BDNF), which is released during muscle activity (9). Consistent with this idea, BDNF promotes axon maturation by stimulating both actin polymerization and microtubule assembly (10). It will be interesting to test the role of trophic factors in activity-dependent synapse elimination and subsequent myelination in both the CNS and PNS.In conclusion, Wang et al. (2) is an excellent addition to a growing body of research that demonstrates how neuronal activity promotes and modulates myelination. Furthermore, it stands as another example of how using simple model systems, such as the NMJ, may provide insights and have important implications for much more complicated biological systems.     

Differences in synaptic output between excitatory and inhibitory motoneurons in a crayfish muscle

A pair of antagonistic motoneurons, one excitatory and one inhibitory, innervates the distal accessory flexor muscle in the walking limb of the crayfish Procambarus clarkii. The number and size of synapses formed by these two axons on the muscle fibers (neuromuscular synapses) and on each other (axo-axonal synapses) were estimated using thin-section electron microscopy. Although profiles of nerve terminals of the two axons occur in roughly equal proportions, the frequency of occurrence of neuromuscular synapses differed markedly: 73% were excitatory and 27% were inhibitory. However, inhibitory synapses were 4–5 times larger than excitatory ones, and consequently, the total contact areas devoted to neuromuscular synapses were similar for both axons. Axo-axonal synapses were predominantly from the inhibitory axon to the excitatory axon (86%), and a few were from the excitatory axon to the inhibitory axon (14%). The role of the inhibitory axo-axonal synapse is presynaptic inhibition, but that of the excitatory axo-axonal synapse is not known. The differences in size of neuromuscular synapses between the two axons may reflect intrinsic determinants of the neuron, while the similarity in total synaptic area may reflect retrograde influences from the muscle for regulating synapse number.     

Synaptic competition and the persistence of polyneuronal innervation at frog neuromuscular junctions

Mechanisms governing the elimination of polyneuronal innervation were examined by correlating the morphology and physiology of competing nerve terminals at identified dually innervated neuromuscular junctions in sartorius muscles of adult frogs (Rana pipiens). Synaptic efficacy (endplate potential amplitude per unit nerve terminal length) was presumed to reflect the ability of a terminal to compete for synaptic space. The synaptic efficacies of two terminals at the same synaptic site were found to be surprisingly equal, with a median difference of 33%. Much more variation would be expected if dually innervated junctions were randomly innervated by pairs of terminals having the same range of synaptic efficacy as that found at singly innervated junctions in the same muscle. This finding supports the hypothesis that the weaker input is eliminated from dually innervated junctions when there is a large discrepancy in competitive efficacy, and that both inputs may persist if competitive efficacies are relatively equal. We also tested but failed to find support for the hypothesis that spatial proximity between competing terminals intensifies competition for synaptic space during synapse elimination.     

Pre- and postsynaptic mechanisms in Hebbian activity-dependent synapse modification

We have used a three compartment tissue culture system that involved two separate populations of cholinergic neurons in the side compartments that converged on a common target population of myotubes in the center compartment. Activation of the axons from one population of neurons produced selective down-regulation of the synaptic inputs from the other neuronal population (when the two inputs innervated the same myotubes). The decrease in heterosynaptic inputs was mediated by protein kinase C (PKC). An activity-dependent action of protein kinase A (PKA) was associated with the stimulated input and this served to selectively stabilize this input. These changes associated with PKA and PKC activation were mediated by alterations in the number of acetylcholine receptors at the neuromuscular junction. These results suggest that neuromuscular electrical activity produces postsynaptic activation of both PKA and PKC, with the latter producing generalized synapse weakening and the former a selective synapse stabilization. Treatment of the neuronal cell body and axon to increase PKC activity by putting phorbal ester (PMA) in the side chamber did not affect synaptic transmission (with or without stimulation). By contrast, PKA blockade in the side compartment did produce an activity-dependent decrease in synaptic efficacy, which was due to a decrease in quantal release of neurotransmitter. Thus, when the synapse is activated, it appears that presynaptic PKA action is necessary to maintain transmitter output.     

On the role of the 200-kDa neurofilament protein at the developing neuromuscular junction

To examine whether the 200-kDa neurofilament protein (200K NFP) is involved in mechanically stabilizing axons, we studied the developmental appearance of immunoreactivity to nonphosphorylated and phosphorylated 200K NFP at the neuromuscular junction. Polyinnervated rat muscle fibers become singly innervated during the first 3 weeks of postnatal life through the process of synapse elimination. If production or post-translational modification of the 200K NFP is actively involved in imparting mechanical stability on neuromuscular synapses, then the selective presence of this protein in only one of several axons at each developing end plate region might make that one axon selectively resistant to elimination. The remaining axons would then be eliminated. Immunoreactivity to the 200K NFP is present on Gestational Day 14 and can be seen in more than one preterminal axon in the end plate region of a muscle fiber during the period of synapse elimination. These results suggest that the 200K NFP is present and phosphorylated early in development and, although the 200K NFP may increase the mechanical stability of axons, this increased stability does not determine the final outcome of synapse elimination.     

Synapse elimination occurs late in the hormone-sensitive levator ani muscle of the rat

Using tetranitroblue tetrazolium (TNBT) to stain neuromuscular synapses, we compared the development of the adult pattern of innervation in two fast-twitch muscles in the rat: the androgen-sensitive levator ani (LA) and the extensor digitorum longus (EDL), which is not thought to be androgen sensitive. We found that about 18% of adult LA muscle fibers, but only about 2% of adult EDL fibers, are multiply innervated. Moreover, synapse elimination occurs substantially later in the LA compared with the EDL. At 2 weeks after birth, the EDL is already predominantly singly innervated, whereas the LA is still predominantly multiply innervated. The apparent delay in the normal time course of synapse elimination in the LA corresponds to a similar delay in other aspects of neuromuscular development (the time course of appearance of axonal retraction bulbs, the growth of fibers, and the development of adult motor terminal morphology). Finally, motor terminals change during synapse elimination from morphologies resembling growth cones to the adult form of neuromuscular synapses. Because the period of synapse elimination is significantly different for muscles that differ in their androgen sensitivity, hormonal sensitivity may represent an important property of motoneurons or muscle fibers influencing the normal time course of neuromuscular synapse elimination in rats. Thus, androgen might regulate the normal ontogenetic process of synapse elimination.     

An anatomical and electrophysiological study of synapse elimination at the developing frog neuromuscular junction

Synapse elimination was examined in the developing frog cutaneous pectoris muscle using histological and electrophysiological techniques. Morphological synapse elimination occurred in two phases. The first phase, which began at the time of metamorphosis and continued until the second to third postmetamorphic week, was characterized by a rapid decline in the number of endplates receiving greater than or equal to 3 synaptic inputs. However, 50% of the muscle fibers still remained dually innervated. This dual innervation decreased with a much slower time course; approximately 20% of the muscle fibers were dually innervated in 1- to 2-year-old frogs. During the first phase of synapse elimination no difference was noted between the distribution of acetylcholine receptors or acetylcholinesterase activity associated with the terminal arborizations formed by separate axons at one synaptic site. However, terminal arborizations formed by small diameter axons and consisting of varicosities separated by thin interconnectives became apparent during this period. Such varicose arborizations responded to nerve stimulation and released acetylcholine in proportion to their terminal length as did the nonvaricose arborizations. In addition, the number of morphological and physiological inputs at one endplate site was well correlated throughout the first phase of synapse elimination.     

Pre‐ and postsynaptic mechanisms in Hebbian activity‐dependent synapse modification

We have used a three compartment tissue culture system that involved two separate populations of cholinergic neurons in the side compartments that converged on a common target population of myotubes in the center compartment. Activation of the axons from one population of neurons produced selective down‐regulation of the synaptic inputs from the other neuronal population (when the two inputs innervated the same myotubes). The decrease in heterosynaptic inputs was mediated by protein kinase C (PKC). An activity‐dependent action of protein kinase A (PKA) was associated with the stimulated input and this served to selectively stabilize this input. These changes associated with PKA and PKC activation were mediated by alterations in the number of acetylcholine receptors at the neuromuscular junction. These results suggest that neuromuscular electrical activity produces postsynaptic activation of both PKA and PKC, with the latter producing generalized synapse weakening and the former a selective synapse stabilization. Treatment of the neuronal cell body and axon to increase PKC activity by putting phorbal ester (PMA) in the side chamber did not affect synaptic transmission (with or without stimulation). By contrast, PKA blockade in the side compartment did produce an activity‐dependent decrease in synaptic efficacy, which was due to a decrease in quantal release of neurotransmitter. Thus, when the synapse is activated, it appears that presynaptic PKA action is necessary to maintain transmitter output. Published 2002 Wiley Periodicals, Inc. J Neurobiol 52: 241–250, 2002     

Activity-dependent switch from synapse formation to synapse elimination during development of neuromuscular junctions

The embryonic development of neuromuscular junctions consists of two successive epochs, an early period marked by exuberant synapse formation and a later period marked by synapse elimination. In the frog muscles we have studied, myogenesis is protracted and overlaps the periods of synapse formation and elimination. Thus, the formative and regressive events of synaptic development do not occur in synchrony across different fibers in the muscle. We propose that local activity orchestrates a shift from synaptogenesis to synapse elimination at the level of single muscle fibers. We also present evidence that perisynaptic Schwann cells and the expression of ion channels in the sarcolemma play important roles in the development of neuromuscular junctions. Questions for future study are outlined.     

Synaptic competition and the elimination of polyneuronal innervation following reinnervation of adult frog sartorius muscles

The elimination of polyneuronal innervation (synapse elimination) that occurs following reinnervation was studied in sartorius muscles of adult Rana pipiens. The percentage of neuromuscular junctions that were polyneuronally innervated declined from 47% at 40-80 days after nerve crush to 22% at greater than 250 days after nerve crush. We measured the size, synaptic strength, and position of competing nerve terminals at identified dually innervated neuromuscular junctions at these two different periods of synapse elimination. Our goal was to determine if any of these parameters play a role in the competition between nerve terminals that ultimately results in the elimination of polyneuronal innervation. Our data support the hypothesis that polyneuronal innervation will persist if competing nerve terminals are of similar synaptic efficacies but will be eliminated if the competing terminals are of different synaptic efficacies. We also tested, but failed to find any evidence, that the spatial proximity of competing nerve terminals at the same synaptic site influences the elimination of polyneuronal innervation.     

Extracellular ATP in activity-dependent remodeling of the neuromuscular junction

Electrical activity during early development affects the development and maintenance of synapses (Spitzer [2006]: Nature 4447:707-712), but the intercellular signals regulating maintenance of synapses are not well identified. At the neuromuscular junction, adenosine 5-triphosphate (ATP) is coreleased with acetylcholine at activated nerve terminals to modulate synaptic function. Here we use cocultured mouse motor neurons and muscle cells in a three-compartment cell culture chamber to test whether endogenously released ATP plays a role in activity-dependent maintenance of neuromuscular synapses. The results suggest that ATP release at the synapse counters the negative effect of electrical activity, thus stabilizing activated synapses. Confirming our previous work (Li et al. [2001]: Nat Neurosci 4:871-872), we found that in doubly innervated muscles, electrical stimulation induced heterosynaptic downregulation of the nonstimulated convergent input to the muscle fiber with no or little change of the stimulated inputs. However, in preparations that were stimulated in the presence of apyrase, an enzyme that degrades extracellular ATP, synapse downregulation of stimulated inputs was substantial and significant, and end plate potentials were reduced. Apyrase treatment for 20 h in the absence of stimulation did result in moderate diminution, but this was prevented by blocking spontaneous neural activity with tetrodotoxin. The P2 receptor blocker, suramin, also induced activity-dependent synapse diminution. The decrease in synaptic efficacy produced by prolonged stimulation in the presence of apyrase persisted for greater than 20 h, consistent with a developmental time-course and distinct from the rapid neuromodulatory actions of ATP that have been demonstrated by others. We conclude that extracellular ATP promotes stabilization of the neuromuscular junction and may play a role in activity-dependent synaptic modification during development.     

Activity and synapse elimination at the neuromuscular junction

The neuromuscular junction undergoes a loss of synaptic connections during early development. This loss converts the innervation of each muscle fiber from polyneuronal to single. During this change the number of motor neurons remains constant but the number of muscle fibers innervated by each motor neuron is reduced. Evidence indicates that a local competition among the inputs on each muscle fiber determines which inputs are eliminated. The role of synapse elimination in the development of neuromuscular circuits, other than ensuring a single innervation of each fiber, is unclear. Most evidence suggests that the elimination plays little or no role in correcting for errant connections. Rather, it seems that connections are initially highly specific, in terms of both which motor neurons connect to which muscles and which neurons connect to which particular fibers within these muscles. A number of attempts have been made to determine the importance of neuromuscular activity during early development for this rearrangement of synaptic connections. Experiments reducing neuromuscular activity by muscle tenotomy, deafferentation and spinal cord section, block of nerve impulse conduction with tetrodotoxin, and the use of postsynaptic and presynaptic blocking agents have all shown that normal activity is required for normal synapse elimination. Most experiments in which complete muscle paralysis has been achieved show that activity may be essential for the occurrence of synapse elimination. Furthermore, experiments in which neuromuscular activity has been augmented by external stimulation show that synapse elimination is accelerated. A plausible hypothesis to explain the activity dependence of neuromuscular synapse elimination is that a neuromuscular trophic agent is produced by the muscle fibers and that this production is controlled by muscle-fiber activity. The terminals on each fiber compete for the substance produced by that fiber. Inactive fibers produce large quantities of this substance; on the other hand, muscle activity suppresses the level of synthesis of this agent to the point where only a single synaptic terminal can be maintained. Inactive muscle fibers would be expected to be able to maintain more nerve terminals. The attractiveness of this scheme is that it provides a simple feedback mechanism to ensure that each fiber retains a single effective input.     

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.     

Assembly,plasticity and selective vulnerability to disease of mouse neuromuscular junctions

Although physiological differences among neuromuscular junctions (NMJs) have long been known, NMJs have usually been considered as one type of synapse, restricting their potential value as model systems to investigate mechanisms controlling synapse assembly and plasticity. Here we discuss recent evidence that skeletal muscles in the mouse can be subdivided into two previously unrecognized subtypes, designated FaSyn and DeSyn muscles. These muscles differ in the pattern of neuromuscular synaptogenesis during embryonic development. Differences between classes are intrinsic to the muscles, and manifest in the absence of innervation or agrin. The distinct rates of synaptogenesis in the periphery may influence processes of circuit maturation through retrograde signals. While NMJs on FaSyn and DeSyn muscles exhibit a comparable anatomical organization in postnatal mice, treatments that challenge synaptic stability result in nerve sprouting, NMJ remodeling, and ectopic synaptogenesis selectively on DeSyn muscles. This anatomical plasticity of NMJs diminishes greatly between 2 and 6 months postnatally. NMJs lacking this plasticity are lost selectively and very early on in mouse models of motoneuron disease, suggesting that disease-associated motoneuron dysfunction may fail to initiate maintenance processes at “non-plastic” NMJs. Transgenic mice overexpressing growth-promoting proteins in motoneurons exhibit greatly enhanced stimulus-induced sprouting restricted to DeSyn muscles, supporting the notion that anatomical plasticity at the NMJ is primarily controlled by processes in the postsynaptic muscle. The discovery that entire muscles in the mouse differ substantially in the anatomical plasticity of their synapses establishes NMJs as a uniquely advantageous experimental system to investigate mechanisms controlling synaptic rearrangements at defined synapses in vivo.     

Activity-dependent elimination of neuromuscular synapses

At developing neuromuscular synapses in vertebrates, different motor axon inputs to muscle fibers compete for maintenance of their synapses. Competition results in progressive changes in synaptic structure and strength that lead to the weakening and loss of some inputs, a process that has been called synapse elimination. At the same time, a single input is strengthened and maintained throughout adult life, consistently recruiting muscle fibers to contract even at rapid firing rates. Work over the last decade has led to an understanding of some of the cell biological mechanisms that underlie competition and how these culminate in synapse elimination. We discuss current ideas about how activity modulates neuromuscular synaptic competition, how competition leads to synapse loss, and how these processes are modulated by cell-cell signaling. A common feature of competition at neuromuscular as well as CNS synapses is that temporally correlated activity seems to slow or prevent competition, while uncorrelated activity seems to trigger or enhance competition. Important questions that remain to be addressed include how patterns of motor neuron activity affect synaptic strength, what is the temporal relationship between changes in synaptic strength and structure, and what cellular signals mediate synapse loss. Answers to these questions will expand our understanding of the mechanisms by which activity edits synaptic structure and function, writing permanent changes in neural circuitry.