Pre- and post-synaptic roles of action potential activity in synapse elimination revealed by using ectopic neuromuscular junction formation by a foreign nerve.

The formation of the neuromuscular junction (nmj) is based on molecular cascades initiated by neural agrin as well as electrical activity in the neuromuscular structures. This review focuses on the latter factor, emphasizing the multiplicity of its mechanisms in the process of synapse elimination following initial polyneuronal innervation. Pre- and post-synaptic components of activity have in fact been identified through experiments on an adult model of nmj formation: ectopic reinnervation of the rat soleus muscle by the fibular nerve. Two activity-dependent elimination processes are thus compared: competition between distributed nmjs, which depends on evoked muscle impulse activity, and competition between axons converging on single nmjs, which instead depends on differences in the timing of impulses in the converging axons.

Activity-dependent vs. neurotrophic modulation of acetylcholine receptor expression: Evidence from rat soleus and extensor digitorum longus muscles confirms the exclusive role of activity.

Evoked electrical muscle activity suppresses the transcription of mRNAs for acetylcholine receptors in extrajunctional myonuclei. Muscle denervation or disuse releases such inhibition and extrajunctional receptors appear. However, in soleus muscles paralysed with nerve-applied tetrodotoxin, a restricted perijunctional region has been described where myonuclei remain inhibited, a finding attributed to nerve-derived trophic factor(s). Here, we reinvestigate extrajunctional acetylcholine receptor expression in soleus and extensor digitorum longus muscles up to 90 days after denervation or up to 20 days of disuse, to clarify the role of trophic factors, if any. The perijunctional region of soleus muscles strongly expressed acetylcholine receptors during the first 2-3 weeks of denervation. After 2-3 months, this expression had disappeared. No perijunctional expression was seen after paralysis by tetrodotoxin or botulinum toxin A. In contrast, the extensor digitorum longus never displayed suppressed perijunctional acetylcholine receptor expression after any treatment, suggesting that it is an intrinsic property of soleus muscles. Soleus denervation only transiently removed the suppression, and its presence in long-term denervated soleus muscles contradicts any contribution from nerve-derived trophic factor(s). In conclusion, our results confirm that evoked electrical activity is the physiological factor controlling the expression of acetylcholine receptors in the entire extrajunctional membrane of skeletal muscles.

Adult rat motor neurons do not re-establish electrical coupling during axonal regeneration and muscle reinnervation.

Gap junctions (GJs) between neurons are present in both the newborn and the adult nervous system, and although important roles have been suggested or demonstrated in a number of instances, in many other cases a full understanding of their physiological role is still missing. GJs are expressed in the rodent lumbar cord at birth and mediate both dye and electrical coupling between motor neurons. This expression has been proposed to mediate: (i) fast synchronization of motoneuronal spike activity, in turn linked to the process of refinement of neuromuscular connections, and (ii) slow synchronization of locomotor-like oscillatory activity. Soon after birth this coupling disappears. Since in the adult rat regeneration of motor fibers after peripheral nerve injury leads to a recapitulation of synaptic refinement at the target muscles, we tested whether GJs between motor neurons are transiently re-expressed. We found that in conditions of maximal responsiveness of lumbar motor neurons (such as no depression by anesthetics, decerebrate release of activity of subsets of motor neurons, use of temporal and spatial summation by antidromic and orthodromic stimulations, testing of large ensembles of motor neurons) no firing is observed in ventral root axons in response to antidromic spike invasion of nearby counterparts. We conclude that junctional coupling between motor neurons is not required for the refinement of neuromuscular innervation in the adult.

The timing of activity is a regulatory signal during development of neural connections.

In PNS and CNS remarkable rearrangements occur soon after the connections are laid down in the course of embryonic life. These processes clearly follow the period of developmental cell death and mostly take place during the very beginning of postnatal life. They consist in changes of the peripheral fields of neurons, marked by elimination of many inputs, while others undergo further maturation and strengthening. Along the efforts to uncover the signals that regulate development, it turned out that while the initial construction of the circuits is heavily based on chemical cues, the subsequent rearrangement is markedly influence by activity. Here we describe experiments testing the influence on developmental plasticity of a particular aspect of activity, the timing of nerve impulses in the competing inputs. Two recent investigations are reviewed, indicating strikingly similar developmental features in quite different systems, neuromuscular and visual. A sharp contrast between the effects of synchrony and asynchrony emerges, indicating that Hebb-related activity rules are important not only for learning but also for development.

Hebb-based rules of neural plasticity: are they ubiquitously important for the refinement of synaptic connections in development?

Neuronal death and suppression of functional synaptic inputs are well-known regressive events characterizing PNS and CNS development. In the CNS, participation of activity in synapse elimination has been known ever since the pioneering studies of Hubel and Wiesel, but only recently has a Hebb-based mechanism of spike synchrony versus asynchrony received unequivocal experimental support in the visual system. At the neuromuscular junction (NMJ), where synapse elimination was discovered, the specific function of the "timing of activity" was addressed by only one group of studies and did not receive widespread attention. Here we critically review the latest NMJ investigation advocating an "activity-independent" mechanism for synapse elimination and contrast it with an equally recent study demonstrating a key role for spike timing. Finally, we highlight how the striking similarities between the two mentioned studies on spike timing (visual system and NMJ) establish conclusively its role in the development of the nervous system in general.

Spike timing plays a key role in synapse elimination at the neuromuscular junction.

Nerve impulse activity produces both developmental and adult plastic changes in neural networks. For development, however, its precise role and the mechanisms involved remain elusive. Using the classic model of synapse competition and elimination at newly formed neuromuscular junctions, we asked whether spike timing is the instructive signal at inputs competing for synaptic space. Using a rat strain whose soleus muscle is innervated by two nerves, we chronically evoked different temporal spike patterns in the two nerves during synapse formation in the adult. We found that asynchronous activity imposed upon the two nerves promotes synapse elimination, provided that their relative spikes are separated by 25 ms or more; remarkably, this elimination occurs even though an equal number of spikes were evoked in the competing axons. On the other hand, when spikes are separated by 20 ms or less, activity is perceived as synchronous, and elimination is prevented. Thus, in development, as in adult plasticity, precise spike timing plays an instructive role in synaptic modification.

Expression of small-conductance calcium-activated potassium channels (SK3) in skeletal muscle: regulation by muscle activity.

The type 3 small conductance calcium-activated potassium channel (SK3) is expressed in embryonic and adult denervated skeletal muscles where it contributes to hyperexcitability. This study aimed at determining the role of muscle activity in regulating SK3 channels. Soleus muscles of adult rats were denervated by cutting the sciatic nerve. In reinnervation studies, the soleus nerve was crushed: in one group, muscles were reinnervated with electrically silent axons, by chronic sciatic nerve perfusion with tetrodotoxin. Several groups of denervated muscles were subjected to chronic direct electrical stimulation, using either fast (100 Hz) or slower patterns (20 or 30 Hz). The SK3 mRNA and protein levels in soleus muscle were determined by reverse transcriptional-PCR, Western blot and immunofluorescence. Both denervated and reinnervated-paralysed soleus muscles displayed similar up-regulation of SK3 mRNA and protein. Reinnervation with electrically active axons instead inhibited SK3 up-regulation. Chronic muscle direct stimulation in vivo, irrespective of the pattern used, reversed the denervation-induced up-regulation of SK3 expression or prevented it when initiated at the time of denervation. Chronic electrical stimulation of denervated muscles also completely prevented the development of the after-hyperpolarization (AHP) following the action potential, normally induced in the muscle fibres by denervation. We conclude that action potential activity evoked by motor neurones in muscle fibres is both necessary and sufficient to account for the physiological down-regulation of SK3 channels in the non-junctional membrane of skeletal muscle.

Synapse formation and elimination: role of activity studied in different models of adult muscle reinnervation.

Synapse competition and elimination are a general developmental process both in central and in peripheral nervous systems that is strongly activity dependent. Some common features regulate synapse competition, and one of these is an application to development of the Hebb's postulate of learning: repeated coincident spike activity in competing presynaptic inputs on the same target cell inhibits competition, whereas noncoincident activity promotes weakening of some of the inputs and ultimately their elimination. Here we report experiments that indicate that the development of muscle innervation (initial polyneuronal innervation and subsequent synapse elimination) follows the Hebb's paradigm. We utilized two different models of muscle reinnervation in the adult rat: 1) we crushed nerves going to soleus or extensor digitorum longus muscles, to activate regeneration of the presynaptic component of the neuromuscular junctions (NMJ), or 2) we injected the soleus muscle with Marcaine (a myotoxic agent) to activate regeneration of the postsynaptic component, the muscle fiber. A condition of transient polyneuronal innervation occurs during NMJ regeneration in both cases, although the two models differ insofar as the relative strength of the competing inputs is concerned. During the period of competition (a few days or weeks, in Marcaine or crush experiments, respectively), we imposed a synchronous firing pattern on the competing inputs by stimulating motor axons distal to a chronic conduction block and demonstrated that this procedure strongly inhibits synapse elimination, with respect to control muscles in which regeneration occurs under natural impulse activity of motoneurons.

Activity-dependent synaptic competition at mammalian neuromuscular junctions.

Synapse elimination is a widespread developmental process in the peripheral and central nervous system that brings about refinement of neural connections through epigenetic mechanisms. Here we describe recent advances concerning the role of the pattern of motoneuronal firing, synchronous or asynchronous, in neuromuscular synapse elimination.

Perinatal switch from synchronous to asynchronous activity of motoneurons: link with synapse elimination.

Synaptic competition is a basic feature of developing neural connections. To shed light on its dependence on the activity pattern of competing inputs, we investigated in vivo rat motoneuronal firing during late embryonic and early neonatal life, when synapse elimination occurs in muscle. Electromyographic recordings with floating microelectrodes from tibialis anterior and soleus muscles revealed that action potentials of motoneurons belonging to the same pool have high temporal correlation. The very tight linkage, a few tens of milliseconds, corresponds to the narrow time windows of published paradigms of activity-dependent synaptic plasticity. A striking change occurs, however, soon after birth when motoneuronal firing switches to the adult uncorrelated type. The switch precedes the onset of synapse elimination, whose time course was determined with confocal microscopy. Interestingly, the soleus muscle, whose motoneurons switch to desynchronized activity later than those of the tibialis anterior muscle, also exhibits delayed synapse elimination. Our findings support a developmental model in which synchronous activity first favors polyneuronal innervation, whereas an asynchronous one subsequently promotes synapse elimination.