Hebbian depression of isolated neuromuscular synapses in vitro

The development of nerve connections in the peripheral nervous system

The structural basis for the functions of the brain is the wiring of nerve cells. The wiring is achieved by 100 billion neurons via some 100,000 billion synaptic connections. Formation of this network is a formidable task which requires intricate, but also reliable mechanisms at work. It is generally recognized that the precise connections between nerve cells are accomplished by two different sorts of mechanisms. First, various molecular guidance cues lead outgrowing axons from specific areas to particular target regions which are rather broadly defined. The second stage includes an exact, point-to-point matching between each axon and the specific target neurons. This is achieved by an intricate interaction between the involved nerve cells. In the following, the attention will primarily regard the second type of events, but some aspects on the initial outgrowth of axons and axon guidance mechanisms will also be given. A large part of the description will focus on the neuromuscular junction, since it has been extensively used as a model for the study of synapse development.

Population coding by cell assemblies--what it really is in the brain

One of the central questions in neuroscience concerns the basic code for information processing in the brain. Much experimental evidence and theoretical consideration have suggested that single-neuron coding is no longer a tenable hypothesis. The present review explains why population neuronal coding is valid and discusses how it is carried out in the brain. The main context is experimental access to real features of the coding in working brains as deduced from experimental research. Several recent studies recording neuronal activities from behaving animals have shown that ensemble activity of neurons represents specific information, indicating the reality of population coding by many neurons. The key concept which can integrate the experimental evidence is the 'cell assembly', i.e., overlapped populations of neurons with flexible functional connections within and among the populations. Correlated activity among the neurons constructs the functional connection. In order to see features of the cell-assembly coding, two main properties of cell assemblies in processing several different kinds of information must be investigated, that is, the overlapping of neurons and the dynamics of synaptic connections. This manner of coding can provide both the experimental and theoretical framework to detect the real dynamic features of information processing by the brain.

Competitive exclusion between axons dependent on a single trophic substance: a mathematical analysis

A mathematical model is presented of competition between axons for a trophic substance, such as is believed to occur particularly during development. The model is biologically realistic. The growth-stimulating activity of the trophic molecules is assumed to result from their binding to high-affinity receptors on neurons and their axons, but the model also incorporates uptake by nonneuronal cells possessing only lower affinity receptors. Plausible and fairly general assumptions are made concerning the kinetics of binding and internalization and the effects on axonal growth. The model takes into account the possibility that trophic factor production may be regulated by the afferent axons or autoregulated. The variables specified are the "axonal vigor" of each axon, representing the ability of each axon to take up trophic molecules, and the concentration of trophic molecules in the extracellular space of the axonal target region. Of the several parameters introduced, the most important turns out to be the "zero vigor-growth parameter," which is defined as the concentration of trophic molecules that gives zero growth of the vigor of a given axon. By means of a Lyapunov function, it is shown that the system will approach asymptotically to a stable equilibrium characterized by the survival of only the axon whose zero-growth parameter is lowest. Or, if several axons share the same lowest zero-growth parameter, these will all survive. The model may be particularly relevant to the elimination of polyneuronal innervation from developing muscle fibers and from autonomic ganglion cells.

Spread of synaptic depression mediated by presynaptic cytoplasmic signaling

Postsynaptic activity may modulate presynaptic functions by transsynaptic retrograde signals. At developing neuromuscular synapses in Xenopus nerve-muscle cultures, a brief increase in the cytosolic calcium ion (Ca2+) concentration in postsynaptic myocytes induced persistent depression of presynaptic transmitter secretion. This depression spread to distant synapses formed by the same neuron. Clearance of extracellular fluid did not prevent the spread of depression, and depression could not be induced by increasing the Ca2+ concentration in a nearby myocyte not in contact with the presynaptic neuron. Thus, the spread of depression is mediated by signaling in the presynaptic cytoplasm, rather than by a retrograde factor in the extracellular space.

Synapse elimination in the central nervous system: functional significance and cellular mechanisms

Recent research into the developmental elimination of supernumerary synapses has increased understanding of this process. In this review we discuss synapse elimination both at the neuromuscular junction and in the central nervous system, considering some possible underlying mechanisms suggested by recent studies. In addition a well-described example of central nervous system synapse elimination, the climbing fiber-Purkinje cell synapse of the cerebellum, is used to explore the functional significance of synaptic regression during brain development.

Postsynaptic elevation of calcium induces persistent depression of developing neuromuscular synapses

Synaptic activity is known to modulate neuronal connectivity in the nervous system. At developing Xenopus neuromuscular synapses in culture, repetitive postsynaptic application of ACh near the synapse leads to immediate and persistent synaptic depression, which was shown to be caused by reduction of presynaptic evoked transmitter release. However, little depression was found when ACh was applied to the muscle 20 microns or further from the synapse. Fluorescence imaging of cytosolic Ca2+ ([Ca2+]i) showed that each ACh pulse induced a transient elevation of myocyte [Ca2+]i that spread approximately 20 microns. Local photoactivated release of Ca2+ from the caged Ca2+ chelators nitr-5 or nitrophen in the postsynaptic cell was sufficient to induce persistent synaptic depression. These results support a model in which localized Ca2+ influx into the postsynaptic myocyte initiates transsynaptic retrograde modulation of presynaptic secretion mechanisms.

Expression of constitutively active CaMKII in target tissue modifies presynaptic axon arbor growth

Calcium/calmodulin-dependent protein kinase II (CaMKII) can be regulated by synaptic activity and could therefore be involved in activity-dependent control of neuronal growth. We tested whether increased CaMKII activity in postsynaptic optic tectal neurons can modify the development of retinotectal axons in Xenopus. The elaboration of individual presynaptic retinal axons was observed in vivo before and up to 3 days after infecting the tectal cells with vaccinia virus carrying the gene for constitutively active truncated CaMKII (tCaMKII). Elevated postsynaptic CaMKII activity prevented the axons from developing the complexity of normal arbors by increasing the normal rate of branch retractions. Some effects of tCaMKII on arbor morphology were seen 1 day after infection, but they became more dramatic by the third day. The results suggest that postsynaptic CaMKII plays a role in the development of presynaptic arbor structure.

Tetanic stimulation and cyclic adenosine monophosphate regulate segregation of presynaptic inputs on a common postsynaptic target neuron in vitro

Previous studies indicated that Aplysia sensory neurons (SNs) compete when reestablishing synapses with a motor cell target (L7) in vitro. The competition is characterized by a cell number-dependent decrease in the efficacy of each connection, an increase in the elimination of SN varicosities, a reduction in the formation of new SN varicosities, and the segregation of varicosities of each SN to restricted portions of the target axons. The changes do not require spike activity, since both the SNs and L7 do not fire spontaneously. Here, we examined whether adding activity to SNs during the early stages of synapse formation with stimuli known to evoke facilitatory responses in stable SN-L7 connections--tetanic stimulation or increase in intracellular cyclic adenosine monophosphate (cAMP)--would modulate the intrinsic segregatory process. Tetanic stimulation to one SN increased synapse efficacy and the number of varicosities of the stimulated SNs while reducing the functional changes by the nonstimulated SNs in the same cultures. An increase in the stability of preexisting varicosities contributed to the overall increase in varicosities evoked by tetanus. The functional changes evoked by tetanus were not expressed when the same tetanic stimulation was also given to the other SN, or when L7 was hyperpolarized during the tetanus to the SN. Raising cAMP levels in one SN increased synapse efficacy and the rate of new varicosity formation by the injected SNs without affecting the development of the connections formed by the noninjected SNs. These results suggest that different forms of presynaptic and postsynaptic activities in neurons can regulate specific aspects of the competitive process associated with the fine-tuning of connections formed by converging presynaptic inputs.

Does Hebbian synaptic plasticity explain learning-induced sensory plasticity in adult mammals?

Over the last decade, a large number of studies have demonstrated that sensory systems undergo functional reorganizations in adult mammals. In the auditory system, highly specific reorganizations were observed during learning situations in which a particular tone frequency predicts the occurrence of an aversive event. After a brief overview of the specific receptive field changes observed after associative learning in cortical and thalamic neurons, I will raise the question concerning whether or not Hebbian synaptic plasticity adequately accounts for these data. The required conditions for Hebbian synaptic plasticity to act do not seem to be met in situations in which learning-induced receptive field plasticity occurs. This analysis points out the weakness of the traditional Hebbian scheme to provide realistic bases for learning-induced neuronal plasticity and stresses the need to look for other potential mechanisms involving neuromodulators.

Memory, sleep, and dynamic stabilization of neural circuitry: evolutionary perspectives

Some aspects of the evolution of mechanisms for enhancement and maintenance of synaptic efficacy are treated. After the origin of use-dependent synaptic plasticity, frequent synaptic activation (dynamic stabilization, DS) probably prolonged transient efficacy enhancements induced by single activations. In many "primitive" invertebrates inhabiting essentially unvarying aqueous environments, DS of synapses occurs primarily in the course of frequent functional use. In advanced locomoting ectotherms encountering highly varied environments, DS is thought to occur both through frequent functional use and by spontaneous "non-utilitarian" activations that occur primarily during rest. Non-utilitarian activations are induced by endogenous oscillatory neuronal activity, the need for which might have been one of the sources of selective pressure for the evolution of neurons with oscillatory firing capacities. As non-sleeping animals evolved increasingly complex brains, ever greater amounts of circuitry encoding inherited and experiential information (memories) required maintenance. The selective pressure for the evolution of sleep may have been the need to depress perception and processing of sensory inputs to minimize interference with DS of this circuitry. As the higher body temperatures and metabolic rates of endothermy evolved, mere skeletal muscle hypotonia evidently did not suffice to prevent sleep-disrupting skeletal muscle contractions during DS of motor circuitry. Selection against sleep disruption may have led to the evolution of further decreases in muscle tone, paralleling the increase in metabolic rate, and culminating in the postural atonia of REM (rapid eye movement) sleep. Phasic variations in heart and respiratory rates during REM sleep may result from superposition of activations accomplishing non-utilitarian DS of redundant and modulatory motor circuitry on the rhythmic autonomic control mechanisms. Accompanying non-utilitarian DS of circuitry during sleep, authentic and variously modified information encoded in the circuitry achieves the level of unconscious awareness as dreams and other sleep mentation.

Geometry, kinetics and plasticity of release and clearance of ATP and noradrenaline as sympathetic cotransmitters: roles for the neurogenic contraction

The paper compares the microphysiology of sympathetic neuromuscular transmission in three model preparations: the guinea-pig and mouse vas deferens and rat tail artery. The first section describes the quantal release of ATP and noradrenaline from individual sites. The data are proposed to support a string model in which: (i) most sites (> or = 99%) ignore the nerve impulse and a few (< or = 1%) release a single quantum of ATP and noradrenaline; (ii) the probability of monoquantal release is extremely non-uniform; (iii) high probability varicosities form 'active' strings; and (iv) an impulse train causes repeated quantal release from these sites. Analogy with molecular mechanisms regulating transmitter exocytosis in other systems is proposed to imply that coincidence of at least two factors at the active zone, Ca2+ and specific cytosolic protein(s), may be required to remove a 'fusion clamp', form a 'fusion complex' and trigger exocytosis of a sympathetic transmitter quantum, and that the availability of these proteins may regulate the release probability. The second section shows that clearance of noradrenaline in rat tail artery is basically > or = 30-fold slower than of co-released ATP, and that saturation of local reuptake and binding to local buffering sites maintain the noradrenaline concentration at the receptors, in spite of a profound decline in per pulse release during high frequency trains. The third section describes differences in the strategies by which mouse vas deferens and rat tail artery use ATP and noradrenaline to trigger and maintain the neurogenic contraction.

Nitric oxide mediates activity-dependent synaptic suppression at developing neuromuscular synapses

Temporal correlation between pre- and postsynaptic activities is an important mechanism that regulates synaptic connectivity during development and synaptic plasticity in the adult. In developing neuromuscular junctions, postsynaptic activity is critical in functional suppression and, ultimately, elimination of the synapses. Although repetitive postsynaptic firing asynchronous to the presynaptic activity results in a persistent synaptic suppression, the underlying molecular mechanism remains unknown. Here we provide evidence that nitric oxide (NO), a free radical implicated in several forms of synaptic plasticity, may serve as a retrograde signal for activity-dependent suppression in the neuromuscular synapse. NO donors and activators of the cyclic GMP pathway suppressed spontaneous and evoked synaptic currents. Moreover, the synaptic suppression induced by repetitive postsynaptic depolarization was prevented by the NO-binding protein haemoglobin and by inhibitors of NO synthase. Thus, synaptic suppression may be triggered by NO released from a postsynaptic myocyte that fires asynchronously to the presynaptic terminal.

Potentiation of transmitter release by ciliary neurotrophic factor requires somatic signaling

Neurotrophic factors participate in the development and maintenance of the nervous system. Application of ciliary neurotrophic factor (CNTF), a protein that promotes survival of motor neurons, resulted in an immediate potentiation of spontaneous and impulse-evoked transmitter release at developing neuromuscular synapses in Xenopus cell cultures. When CNTF was applied at the synapse, the onset of the potentiation was slower than that produced by application at the cell body of the presynaptic neuron. The potentiation effect was abolished when the neurite shaft was severed from the cell body. Thus, transmitter secretion from the nerve terminals is under immediate somatic control and can be regulated by CNTF.

Synapse elimination, the size principle, and Hebbian synapses

Synapse elimination at the vertebrate neuromuscular junction reduces a polyinnervated population of muscle fibers to a monoinnervated state. The function of this developmental phenomenon (if any) is unproven. A theoretical analysis of Hebbian (correlation) rules connecting presynaptic and postsynaptic activity and synaptic strength at the neuromuscular junction is presented. The following points are demonstrated: (1) Correlational competition leads to the reduction of polyinnervation to a stable monoinnervated state; (2) the competition gives rise to the size principle over a wide range of the plausible parameter space; (3) over a significant subrange, the competition selectively eliminates topographically incorrect synapses; and (4) in cases in which topographic projection errors overwhelm the system, both error correction and the development of the size principle are disrupted. Correlational competition may explain contradictory experimental results concerning the effects of stimulating or silencing subpopulations of motor neurons. It may also explain an otherwise puzzling instance of a breakdown in the size principle seen in humans undergoing neural regeneration. Taken together, these findings suggest a novel hypothesis for the function of synapse elimination at the neuromuscular junction: the establishment of the size principle.

Plasticity of developing neuromuscular synapses

Developing neuromuscular synapses are susceptible to modulation by the presence of synaptic activity and a number of chemical factors originated from either pre- or postsynaptic cells. In vitro studies of functional modulation of synaptic strength by electrical stimulation of pre- and postsynaptic cells suggested an essential role of retrograde interactions at developing synapses.

Long-term synapse loss induced by focal blockade of postsynaptic receptors

Focal application in vivo of alpha-bungarotoxin to block neurotransmission in a small region of a neuromuscular junction causes long-lasting synapse elimination at that site. In contrast, blockade of neurotransmission throughout a junction does not cause synapse elimination. These and related experiments indicate that active synaptic sites can destabilize inactive synapses in their vicinity.

Sleep and dynamic stabilization of neural circuitry: a review and synthesis

A common mechanism is advanced for the lengthy stabilization of neural circuitry encoding information of both hereditary and experimental origin. Stabilization is proposed to occur through the following means and interrelationships. Synaptic function is intrinsically plastic because of greatly restricted entry of essential, relatively short-lived molecules into synaptic terminals. Alterations that accompany synaptic transmission transiently facilitate this entry ("facilitated entry"). Synaptic efficacy is enhanced as the concentration of these molecules increases following a transmission event but subsequently declines if depletion of the molecules occurs without commensurate replacement. Accordingly, if lengthy persistence of information encoded by enhancements of synaptic efficacy is to be achieved, the enhancements must be reinforced repeatedly by synaptic transmission ("dynamic stabilization"). Synapses of circuits not in frequent functional use are thought to be dynamically stabilized by spontaneous, internally generated, "non-utilitarian" excitations occurring primarily during rest or sleep. In species with complex, highly developed brains, requirements for dynamic stabilization of infrequently used circuits apparently cannot be met during rest, a restriction that may underlie the origin of sleep. Dynamic stabilization of infrequently used motor circuits of endotherms appears to occur predominantly during REM sleep.

The upregulation of acetylcholine release at endplates of alpha-bungarotoxin-treated rats: its dependency on calcium

1. The presynaptic component of an adaptive feedback mechanism leading to increased acetylcholine (ACh) release was studied in endplates of diaphragms from rats treated chronically with alpha-bungarotoxin (alpha BTX). 2. Quantal contents were calculated 'directly' from the amplitude of miniature endplate potentials (MEPPs) and endplate potentials (EPPs) which were recorded after mu-conotoxin treatment to prevent muscle action potentials. 3. In vitro application of the Ca2+ channel blockers nifedipine (10 microM) or omega-conotoxin (40 nM) had no significant effect on the increased quantal content of endplates from alpha BTX-treated rats. 4. At control endplates, in vitro block of presynaptic K+ channels by 5 microM 3,4-diaminopyridine did increase the quantal content to a level which was similar to that found in endplates of alpha BTX-treated rats but also induced a broadening of EPPs, which was not found at endplates after alpha BTX treatment. 5. The difference between quantal contents of alpha BTX-treated and control rats was highly dependent on the [Ca2+]o/[Mg2+]o ratio when [Mg2+]o was fixed at 1 mM. At low [Ca2+]o, the quantal content of endplates from alpha BTX-treated rats was lower than that of controls while at [Ca2+]o in the normal and high range this was reversed. However, changing the [Ca2+]o/[Mg2+]o ratio by means of [Mg2+]o, at a fixed [Ca2+]o of 2 mM, did not influence the relative increase of quantal contents at endplates from alpha BTX-treated rats. Double logarithmic plots of the 'toxin-induced' myasthenia gravis (TIMG) and control quantal content versus [Ca2+]o had an approximately linear part between 0.2 and 1.5 mM [Ca2+]o. The slopes of the TIMG and control lines were 1.81 and 0.96, indicating that the ACh release in TIMG muscles was more sensitive to changes of [Ca2+]o than controls. 6. At normal [Ca2+]o and [Mg2+]o, the depression of EPP amplitude during stimulation of the phrenic nerve at 30-50 Hz was somewhat larger at endplates from alpha BTX-treated rats than at control endplates. At low [Ca2+]o, the potentiation of EPP amplitudes during a stimulus train was much larger at endplates from alpha BTX-treated rats than from controls. 7. The results do not support the idea that the increased release of ACh is caused via regulatory effects on the presynaptic Ca2+ or K+ channels. Instead, the anomalous dependency of ACh release on Ca2+ in muscles of alpha BTX-treated rats suggests that a cytoplasmic, Ca(2+)-dependent, component is involved in the adaptive change of transmitter release.

Postsynaptic regulation of the development and long-term plasticity of Aplysia sensorimotor synapses in cell culture

The monosynaptic component of the neuronal circuit that mediates the withdrawal reflex of Aplysia californica can be reconstituted in dissociated cell culture. Study of these in vitro monosynaptic connections has yielded insights into the basic cellular mechanisms of synaptogenesis and long-term synaptic plasticity. One such insight has been that the development of the presynaptic sensory neurons is strongly regulated by the postsynaptic motor neuron. Sensory neurons which have been cocultured with a target motor neuron have more elaborate structures--characterized by neurites with more branches and varicosities--than do sensory neurons grown alone in culture or sensory neurons that have been cocultured with an inappropriate target cell. Another way in which the motor neuron regulates the development of sensory neurons is apparent when sensorimotor cocultures with two presynaptic cells are examined. In such cocultures the outgrowth from the different presynaptic cells is obviously segregated on the processes of the postsynaptic cell. By contrast, when two sensory neurons are placed into cell culture without a motor neuron, their processes readily grow together. In addition to regulating the in vitro development of sensory neurons, the motor neuron also regulates learning-related changes in the structure of sensory neurons. Application of the endogenous facilitatory transmitter serotonin (5-HT) causes long-term facilitation of in vitro sensorimotor synapses due in part to growth of new presynaptic varicosities. But 5-HT applied to sensory neurons alone in culture does not produce structural changes in these cells. More recently it has been found that sensorimotor synapses in cell culture can exhibit long-term potentiation (LTP). Like LTP of some hippocampal synapses, LTP of in vitro Aplysia synapses is regulated by the voltage of the postsynaptic cell. Pairing high-frequency stimulation of sensory neurons with strong hyperpolarization of the motor neuron blocks the induction of LTP. Moreover, LTP of sensorimotor synapses can be induced in Hebbian fashion by pairing weak presynaptic stimulation with strong postsynaptic depolarization. These findings implicate a Hebbian mechanism in classical conditioning in Aplysia. They also indicate that Hebbian LTP is a phylogenetically ancient form of synaptic plasticity.

Calcium-dependent postsynaptic exocytosis: a possible mechanism for activity-dependent synaptic modulation

Elevation of cytosolic Ca2+ level in the postsynaptic cell is critical for the induction of many forms of activity-dependent synaptic modulation. Based on our recent evidence that in muscle cells and fibroblasts constitutive exocytosis is increased by elevating cytosolic Ca2+ levels, we hypothesize that Ca(2+)-dependent exocytosis at the postsynaptic site may provide a mechanism for a localized, activity-dependent synaptic modulation.

Proteolytic activity, synapse elimination, and the Hebb synapse

The Hebb synapse has been postulated to serve as a mechanism subserving both regulation of synaptic strength in the adult nervous system (long-term potentiation and depression) and developmental activity-dependent plasticity. According to this model, pre- and postsynaptic temporal concordance of activity results in strengthening of connections, while discordant activity results in synapse weakening. Evidence is presented that proteases and protease inhibitors may be involved in modification of synaptic strength. This leads to a modification of the Hebb assumptions, namely that postsynaptic activity results in protease elaboration with a consequent general reduction of synaptic connections to the active postsynaptic element. Further, presynaptic activity, if strong enough, induces local release of a protease inhibitor, such as protease nexin I, which neutralizes proteolytic activity and produces a relative preservation of the active input. This formulation produces many of the effects of the classical Hebbian construction, but the protease/inhibitor model suggests additional specific mechanistic features for activity-dependent plasticity.

Retrograde interactions during formation and elimination of neuromuscular synapses

Maturation of neuromuscular synapses depends on dynamic interactions between presynaptic motor neurons and postsynaptic muscle cells. Recent studies have addressed the cellular mechanisms underlying these interactions in cell cultures and in developing animals. Retrograde signals from the postsynaptic muscle cells appear to play critical roles in all stages of synapse development, from the initial synaptogenesis to the stabilization or elimination of the synapse.

Pre- and postsynaptic changes mediated by two second messengers contribute to expression of Aplysia long-term heterosynaptic inhibition

FMRFamide evokes long-term inhibition of the sensorimotor connection of Aplysia that includes structural alterations in the presynaptic sensory cell. FMRFamide also evokes a down-regulation of the adhesion molecule apCAM from the surface of the postsynaptic motor cell L7. We examined the second messenger pathways mediating the long-term actions of FMRFamide on both the pre- and postsynaptic cells and determined whether the activation of each pathway is required for the expression of long-term functional and structural plasticity. Inhibition of the lipoxygenase pathway of arachidonic acid metabolism, but not the cyclooxygenase pathway, blocks the long-term changes in the presynaptic sensory cell evoked by FMRFamide. The down-regulation of apCAM in L7 appears to be mediated by cAMP-dependent activation of protein kinase A. Blocking the cAMP-dependent changes also blocks FMRFamide-induced long-term functional and structural changes. These results suggest that the expression of long-term heterosynaptic inhibition in Aplysia may require concomitant presynaptic and postsynaptic changes, each transduced by specific second messenger systems.

Activity-dependent modulation of developing neuromuscular synapses

Spontaneous and impulse-evoked synaptic currents were observed immediately following nerve-muscle contact in Xenopus cell cultures. The functional significance of this early synaptic activity was examined. Stimulation of pre- and/or post-synaptic cells was found to exert immediate and persistent effects on the efficacy of synaptic transmission. Exogenous application of calcitonin gene-related peptide (CGRP) and neurotrophins, factors that may be coreleased with ACh in activity-dependent manner at the developing neuromuscular junctions, also modulate either the postsynaptic ACh response or presynaptic ACh release. These results underscore the plasticity of developing neuromuscular synapses and suggest a complex interplay between electrical activity and chemical factors during the formation and maturation of neuronal connections.

Synapse elimination from the mouse neuromuscular junction in vitro: a non-Hebbian activity-dependent process

The effect of action potentials on elimination of mouse neuromuscular junctions (NMJ) was studied in a three-compartment cell culture preparation. Axons from superior cervical ganglion or ventral spinal cord neurons in two lateral compartments formed multiple neuromuscular junctions with muscle cells in a central compartment. The loss of synapses over a 2-7-day period was determined by serial electrophysiological recording and a functional assay. Electrical stimulation of axons from one side compartment during this period, using 30-Hz bursts of 2-s duration, repeated at 10-s intervals, caused a significant increase in synapse elimination compared to unstimulated cultures (p < 0.001). The extent of homosynaptic and heterosynaptic elimination was comparable, i.e., of the 226 functional synapses of each type studied, 111 (49%) of the synapses that had been stimulated were eliminated, and 87 (39%) of unstimulated synapses on the same muscle cells were eliminated. Also, simultaneous bilateral stimulation caused significantly greater elimination of synapses than unilateral stimulation (p < 0.005). These observations are contrary to the Hebbian hypothesis of synaptic plasticity. A spatial effect of stimulus-induced synapse elimination was also evident following simultaneous bilateral stimulation. Prior to stimulation, most muscle cells were innervated by axons from both side compartments, but after bilateral stimulation, muscle cells were predominantly unilaterally innervated by axons from the closer compartment. These experiments suggest that synapse elimination at the NMJ is an activity-dependent process, but it does not follow Hebbian or anti-Hebbian rules of synaptic plasticity. Rather, elimination is a consequence of postsynaptic activation and a function of location of the muscle cell relative to the neuron. An interaction between spatial and activity-dependent effects on synapse elimination could help produce optimal refinement of synaptic connections during postnatal development.

Synapse formation and elimination during growth of the pectoral muscle in Xenopus laevis

1. Synapse formation and synapse elimination were studied in the pectoral muscle of Xenopus laevis. 2. Histology showed that fibres were not added during postmetamorphic growth. Most fibres were innervated at two widely separated junctions and this number did not change as frogs grew. 3. Intracellular recording revealed that fibres with two junctions could be mononeuronally innervated, or innervated in one of three different polyneuronal patterns. A growth-related shift in innervation pattern was observed, with the polyneuronal patterns replaced by mononeuronal innervation. 4. Endplate potentials (EPPs) evoked by low-frequency nerve stimulation were simultaneously measured at both junctions on individual fibres. For each fibre, the ratio of EPP amplitudes (smaller/larger) was calculated. When the two junctions were innervated by different motoneurones (A-B), the median EPP ratio was smaller than when the two junctions were innervated by the same motoneurone (A-A), although the difference was not significant. 5. The difference in the ratio of EPP amplitudes became significant, however, if junctions were conditioned by a train of fifty stimuli at 10 Hz. Immediately after such a train, EPP ratios for A-B fibres were significantly smaller than ratios for A-A fibres. This difference was due to greater synaptic depression at one of the junctions on A-B fibres. 6. We concluded that enhanced depression of the EPP upon repetitive stimulation is a physiological correlate of the competition that underlies synapse elimination.

The role of spatio-temporal firing patterns in neuronal development of sensory systems

The emergence of precise and orderly sets of neuronal connections often depends upon coordinated electrical activity during the early stages of development. In recent years, an increasing number of reports have shown that neurons of immature sensory systems can spontaneously generate electrical activity that occurs synchronously amongst adjacent cells. These patterns of correlated activity seem to be well suited to the role of providing the cues that are necessary for the activity-dependent refinement of the neural connections in the developing visual, auditory and somatosensory pathways.

Development of the neuromuscular synapse

Major advances have occurred in our understanding of the signaling events involved in neuromuscular synapse formation. In particular, it has recently been shown that agrin is necessary for synapse formation, that acetylcholine receptor genes are specifically transcribed by synaptic nuclei in response to signals from the synaptic basal lamina, and that synaptic competition between motor neurons can occur by a Hebbian mechanism in cell culture.

Retrograde modulation at developing neuromuscular synapses: involvement of G protein and arachidonic acid cascade

Intracellular loading of nonhydrolyzable GTP analogs into innervated muscle cells in Xenopus cultures led to a marked increase in the frequency of spontaneous synaptic currents (SSCs), while extracellular application of the drugs at the same concentration was without effect. The increase in SSC frequency appeared to be unrelated to changes in the muscle membrane sensitivity toward acetylcholine (ACh), but resulted from an elevated spontaneous ACh secretion from the presynaptic nerve terminal. Postsynaptic loading of arachidonic acid (AA) produced a similar effect as the GTP analogs, and the potentiation effect of both GTP analogs and AA was reversed by an inhibitor of AA metabolism, AA861. Further studies indicate that a lipoxygenase metabolite, 5-HPETE, appears to be a likely candidate for the retrograde factor involved in modulating ACh secretion. These results suggest that G protein activation of the AA cascade in the postsynaptic cell could produce a retrograde signal to modulate transmitter secretion from the presynaptic nerve terminal at developing synapses.