Selective motor hyperreinnervation using motor rootlet transfer: an experimental study in rat brachial plexus.

Misdirection of sensory fibers into motor pathways is, in part, responsible for the poor results obtained after peripheral nerve repair. After avulsion of the C-5 root in rats, the authors connected a C-4 ventral rootlet to the musculocutaneous nerve by means of a sural nerve graft. In this way, they were able to increase the number of regenerating motor fibers and avoid growth of sensory fibers into the nerve grafts. Functional recovery was evaluated electrophysiologically and histologically. The origin of the axons that reinnervated the nerve graft was analyzed by means of morphological studies including retrograde labeling procedures. Motor neurons survived and regenerated after the rootlet transfer and there was no functional impairment. Many neurons were retrograde labeled in the ventral horn and widespread biceps muscle reinnervation was demonstrated with recovery of nearly normal electrophysiological properties. Motor hyperreinnervation of the musculocutaneous nerve was observed. This high degree of reinnervation in a long (40-mm) graft was attributed to the good chance that a muscle fiber can be reinnervated by a motor fiber when the number of regenerating motor neurons is increased and when competitive sensory fibers are excluded from reinnervation.

Large dense-core vesicles at the frog neuromuscular junction: characteristics and exocytosis.

The population of large dense-core vesicles (LDCVs) in motor nerve terminals of the frog cutaneous pectoris muscle was analysed after various experimental protocols leading to large acetylcholine release. Three types of LDCVs classified according to their size and the core density were detected. Vesicles, 100-150 nm in diameter, with a large and very dense core (type 1) or with an irregular and diffuse dense core (type 2) were present in similar proportions (45 and 50% respectively) in controls. Smaller vesicles, 50-80 nm in diameter, with a very dense core (type 3) were rare, representing around 5% of the cored vesicles. The relative proportion of type 1 and type 2 LDCVs was not modified after prolonged treatment with 25 mM K+. In contrast, the proportion of type 2 LDCVs significantly increased whereas that of type 1 LDCVs decreased after two or three series of 20 Hz electrical stimuli applied to the nerve at 5 s intervals. These changes suggest that type 2 LDCVs are newly recycled LDCVs in the process of reloading. Images of fusion of LDCVs with the axolemma in regions facing Schwann cell digitations were observed both in K(+)- and in electrically stimulated preparations. They indicate that exocytosis of LDCVs at the frog neuromuscular junction takes place preferentially away from the active zones. The presence of a clathrin-like coat on large pockets still containing a core and of both type 1 and type 2 LDCVs in the vicinity of coated pockets strongly suggests that LDCVs might undergo a combined process of exo-endocytosis at the same site.

[Post-traumatic reconnection of the cervical spinal cord with skeletal striated muscles. Study in adult rats and marmosets]. / Reconnexion post-traumatique de la moelle épinière cervicale avec la musculature striée squelettique. Etude chez le rat et le marmouset adultes.

In an attempt at repairing the injured spinal cord of adult mammals (rat, dog and marmoset) and its damaged muscular connections, we are currently using: 1) peripheral nerve autografts (PNG), containing Schwann cells, to trigger and direct axonal regrowth from host and/or transplanted motoneurons towards denervated muscular targets; 2) foetal spinal cord transplants to replace lost neurons. In adult rats and marmosets, a PNG bridge was used to joint the injured cervical spinal cord to a denervated skeletal muscle (longissimus atlantis [rat] or biceps brachii [rat and marmoset]). The spinal lesion was obtained by the implantation procedure of the PNG. After a post-operative delay ranging from 2 to 22 months, the animals were checked electrophysiologically for functional muscular reconnection and processed for a morphological study including retrograde axonal tracing (HRP, Fast Blue, True Blue), histochemistry (AChE, ATPase), immunocytochemistry (ChAT) and EM. It was thus demonstrated that host motoneurons of the cervical enlargement could extend axons all the way through the PNG bridge as: a) in anaesthetized animals, contraction of the reconnected muscle could be obtained by electrical stimulation of the grafted nerve; b) the retrograde axonal tracing studies indicated that a great number of host cervical neurons extended axons into the PNG bridge up to the muscle; c) many of them were assumed to be motoneurons (double labelling with True Blue and an antibody against ChAT); and even alpha-motoneurons (type C axosomatic synapses in HRP labelled neurons seen in EM in the rat); d) numerous ectopic endplates were seen around the intramuscular tip of the PNG. In larger (cavitation) spinal lesions (rat), foetal motoneurons contained in E14 spinal cord transplants could similarly grow axons through PNG bridges up to the reconnected muscle. Taking all these data into account, it can be concluded that neural transplants are interesting tools for evaluating both the plasticity and the repair capacities of the mammalian spinal cord and of its muscular connections.

Ultrastructural distribution of synaptophysin and synaptic vesicle recycling at the frog neuromuscular junction.

Synaptic vesicle recycling after intense acetylcholine (ACh) release was studied at the frog neuromuscular junction (NMJ) using the synaptic vesicle transmembrane protein synaptophysin as immunocytochemical marker of the synaptic vesicle membrane during the process of exo-endocytosis. ACh release in cutaneous pectoris nerve-muscle preparations was stimulated by three different means: K+, Cd2+ in Ca(2+)-free medium, and electrical stimulation in the presence of 4-aminopyridine (4-AP). Cd2+ stimulation produced synaptic vesicle depletion and nerve terminal swelling. Electrical stimulation in the presence of 4-AP produced a reduction in the number of synaptic vesicles, deep axolemmal infoldings, coated pits, and coated vesicles. K+ stimulation did not produce any observable ultrastructural changes. Synaptophysin was labeled using silver-intensified immunogold in dissociated muscle fibers. Unstimulated and K(+)-stimulated preparations showed synaptophysin immunolabeling associated only with synaptic vesicles. In contrast, in Cd(2+)-stimulated preparations, synaptophysin appeared along the axolemma, mainly at the active zones, and after electrical stimulation it appeared in both axolemmal infoldings and the remaining synaptic vesicles. The results show that when synaptic vesicle recycling is inhibited by Cd2+ in Ca(2+)-free medium, or when 4-AP is present during electrical stimulation, synaptic vesicle fusion is accompanied by translocation and incorporation of synaptic vesicle membrane proteins into the axolemma. However, during the latter condition, synaptic vesicles are recycled through coated vesicles arising from the axolemmal infoldings. Conversely, during physiological-like stimulation of ACh release by K+ the synaptic vesicles are rapidly recycled at the active zones, by a double and rapid process of exo-endocytosis, without collapse into the axolemma.

Cd(2+)-and K(+)-evoked ACh release induce different synaptophysin and synaptobrevin immunolabelling at the frog neuromuscular junction.

Synaptophysin and synaptobrevin, two integral proteins of synaptic vesicles, have been used as immunocytochemical markers of the synaptic vesicle membrane during Cd(2+)- or K(+)-induced ACH release at the frog neuromuscular junction. ACh release was stimulated in cutaneous pectoris nerve-muscle preparations by: (1) 1 mM Cd2+ in Ca(2+)-free medium for a period of 3 h, (2) 25 or 40 mM K+ in normal Ringer's solution. Synaptophysin and synaptobrevin were immunolabelled in single fibres teased from fixed muscles using rabbit antisera raised against synaptophysin and synaptobrevin revealed with fluorescein-conjugated IgG. The postsynaptic ACh receptors were simultaneously labelled with rhodaminated alpha-bungarotoxin. Unstimulated and K(+)-stimulated preparations showed synaptophysin and synaptobrevin immunolabelling only after membrane permeabilization with 0.1% Triton X-100. In preparations stimulated with Cd2+ in Ca(2+)-free medium, the immunofluorescence was also observed in non Triton X-100 treated muscle fibres. Confocal laser scanning microscopy analysis revealed that in unstimulated and K(+)-stimulated preparations, synaptophysin and synaptobrevin immunofluorescence appears as bands regularly spaced along the permeabilized nerve terminals and that their distribution corresponds to clusters of synaptic vesicles. After Cd2+ stimulation in Ca(2+)-free medium, labelling for both proteins is irregularly distributed, being more intense at the lateral margins of swollen nerve terminals, suggesting a translocation of synaptic vesicle proteins to the axolemma. At the electron microscopic level, Cd2+ stimulation in Ca(2+)-free medium produces nerve terminal swelling and synaptic vesicle depletion. The results show that when ACh release is stimulated under an impairment of synaptic vesicle recycling, which leads to synaptic vesicle depletion, synaptophysin and synaptobrevin translocation occurs. These findings are in favour of a permanent incorporation of synaptic vesicle membrane into the axolemma. In contrast, after K+ stimulation, the immunofluorescence and the normal synaptic vesicle population observed, suggest that a double process of synaptic vesicle exo-endocytosis rapidly occurs, without incorporation of synaptic vesicle components into the axolemma.

The rat brachial plexus and its terminal branches: an experimental model for the study of peripheral nerve regeneration.

Despite the introduction of microsurgical techniques into clinical practice, the results of surgical procedures involving the brachial plexus and peripheral nerves are still far from spectacular. We therefore studied the rat brachial plexus and its terminal branches in 203 rats. Detailed anatomic and morphologic analyses of the biceps brachii and musculocutaneous nerve, finger flexors, flexor carpi radialis, and the median nerve were performed. Various sources of conventional and vascularized nerve grafts were explored. After musculocutaneous nerve section or median nerve section, there were no articular contractures or automutilations, which constitutes an advantage for these experimental models over the sciatic nerve model. The brachial plexus and its terminal branches provide a good experimental model which can be used to assess the development and normal control of muscle function, examine the mechanisms underlying functional recovery, and test the effects of treatments to enhance recovery.

Electrophysiological evaluation of the effects of naftidrofuryl on skeletal muscle reinnervation in the rat.

The effects of naftidrofuryl on the reinnervation of the rat gastrocnemius muscle after its denervation by localized freezing of the sciatic nerve were tested with electrophysiological techniques. Daily intraperitoneal injections of 30 mg/kg of naftidrofuryl do not increase the rate of axonal regeneration since early signs of reinnervation appeared as in controls around the 10th day after surgery. However, axonal sprouting is markedly increased since the percentage of muscle fibers with polyneuronal innervations was almost twice as high as in controls at the 15 and 21 day postoperative stages. The promoting effects of naftidrofuryl on polyneuronal innervation which gives rise to a redundant innervation during the first period of reinnervation constitutes an improvement of motor function which might be efficient for treatment of nerve injury and neuropathies.

Persistence of functional neuromuscular junctions formed in a denervated skeletal muscle of the adult rat by axons that have regrown from the injured spinal cord through a peripheral nerve autograft.

In previous 'short-term' (2 to 7 months) experiments, we had demonstrated, in the adult rat, that motoneurons of the injured cervical spinal cord could extend lengthy axons into an autologous peripheral nerve segment which was connected to a nearby denervated skeletal muscle. In addition, we had shown that new functional motor endplates were formed by these axons both at the original sites of innervation and at ectopic locations of the denervated muscle. This substitution motor system, although quite functional, was anatomically very different from the original model of innervation in the intact animal, relating to its motoneuronal pool, the course of its motor axons and the sites of terminal innervation. The present 'long-term' (11 to 21 months) experiments demonstrate the anatomical and functional permanency of the new motor circuitry, despite a lack of strict specificity in the new neuromuscular connections. However, some minor modifications or adjustments were observed with time: (i) the maintenance of functional ectopic endplates could be consistently demonstrated, while functional reinnervated endplates at the initial sites of innervation were rare or even lacking; (ii) there was a definitive withdrawal of all non target-specific regenerated axons from the vicinity of the muscle. It is now necessary to address the question to what extent this substitution motor system is actually controlled by central and/or peripheral inputs.

Membrane events related to transmitter release in mouse motor nerve terminals captured by ultrarapid cryofixation.

The sequence of structural changes occurring in the presynaptic membrane during transmitter release was studied at the mouse neuromuscular junction using the combined quick-freezing and cryosubstitution techniques. The mouse levator auris longus (LAL) muscle was stimulated by two means: either, chemically, by soaking 5 min before freezing in a physiological solution containing 25 mM potassium chloride or, electrically, by applying, 10 ms before freezing, a single supramaximal stimulus to the nerve-muscle preparation treated with 50 microM 3,4-diaminopyridine (3,4-DAP) and 100 microM (+)tubocurarine. In both cases, the preparations were maintained at approximately 5 degrees C, 5 min prior to freezing, in order to prolong nerve membrane changes. In most experiments, tannic acid (0.1%) was added to the substitution medium for better preservation of membranes. The different steps of warming in the substitution medium were strictly controlled from -90 degrees C to 4 degrees C. When fixed under chemical stimulation, the presynaptic membrane appeared very sinuous and synaptic vesicles were seen apposed to specialized sites facing subjunctional folds. When submitted to a single electrical stimulus, after treatment with 3,4-diaminopyridine, features of synaptic vesicle fusion were observed at these specialized sites which appear similar by their morphology, their macromolecular organization (already described) and their functional changes to active zones of the frog neuromuscular junction. Other images suggested that with 3,4-diaminopyridine which causes a pronounced and long-lasting release of transmitter, some vesicles collapse after exocytosis instead of being locally reformed by endocytosis.

Synaptophysin (p38) immunolabelling at the mouse neuromuscular junction.

The synaptophysin (p38), a transmembrane glycoprotein of synaptic vesicles, has been used as a marker in order to study the membrane events that take place during transmitter release at the mouse neuromuscular junction (NMJ). p38 has been labelled by immunofluorescence using a monoclonal anti-p38 antibody and fluorescein-conjugated IgG on dissociated muscle fibres (biceps brachialis m.). Its localization has been compared to that of the acetylcholine (ACh) receptors labelled with rhodaminated alpha-bungarotoxin. A weak labelling was obtained in nerve-muscle preparations at rest only when the muscle fibres were permeabilized with Triton X-100. By contrast, an intense immunofluorescence of the NMJ was observed after an exhaustive ACh release induced by Cd2+ in Ca(2+)-free medium, which leads to a synaptic vesicle depletion and an increase in the membranous structures in nerve terminals. Treatment with Cd2+ in Ca(2+)-free solution leads to both synaptic vesicle depletion and p38 immunolabelling, which is in favour of synaptic vesicle fusion and incorporation into the axolemma.

Changes in large dense-core vesicles during maturation of the rat neuromuscular junction.

The proportion of large dense-core vesicles (LDCVs) in motor nerve terminals of the rat biceps brachialis muscle was evaluated from embryonic day 20 to 4 weeks postpartum as well as in the adult. A progressive decrease was observed up to 3 weeks postpartum when maturation of the endplates is achieved. Differences compared with the adult were no longer significant at 4 weeks postpartum. Three types of LDVCs, classified according to their size and the core density, were detected. Their relative proportion did not vary significantly during the period of life examined. The high proportion of LDCVs during early development and their persistence at a low level in the adult suggest that they might play a role in the maturation and maintenance of the endplates.

Ultrastructure of botulinum type-A poisoned frog motor nerve terminals after enhanced quantal transmitter release caused by carbonyl cyanide m-chlorophenylhydrazone.

The effects of carbonyl cyanide m-chlorophenylhydrazone (CCCP) on spontaneous quantal transmitter release and nerve terminal ultrastructure were studied on isolated cutaneous pectoris nerve-muscle preparations from frogs that were completely paralysed by a single sublethal dose of Clostridium botulinum type A toxin (BoTx). CCCP enhanced miniature endplate potential frequency at poisoned junctions and caused a reduction in the density of clear synaptic vesicles and of large dense core vesicles in motor nerve terminals. However, the intensity of these effects was much less important than that previously reported at unpoisoned junctions. The moderate depletion of synaptic vesicles can be related to the low levels of transmitter release detected with CCCP at BoTx-poisoned terminals.

Nerve sprouting induced by a piece of peripheral nerve placed over a normally innervated frog muscle.

1. The effects of a segment of peripheral nerve on the innervation of a skeletal muscle were investigated in the cutaneous pectoris muscle of the frog Rana esculenta. An explant of the sciatic nerve was placed in an aneural region of the muscle at a distance of 1-3 mm from the zone of neuromuscular junctions. The muscles were examined morphologically and electrophysiologically at different post-operative times. 2. After the first month nerve sprouts were seen arising from both nerve terminals and intramuscular nodes of Ranvier. Both the number of sites of sprouting and the relative distance to the explant tip increased with time (up to 5 months), suggesting spread of a nerve sprouting-promoting stimulus. 3. Most neurites resulting from sprouting were seen growing towards the nerve explant, in the vicinity of which active neurite proliferation occurred. Some of them entered the explant as observed in semi-thin and thin sections which revealed the presence of both myelinated and unmyelinated nerve fibres. 4. Electrical stimulation of the nerve explant in preparations autografted for more than 2 months resulted in the contraction of bundles of muscle fibres. Endplate potentials of similar amplitudes were recorded intracellularly in some muscle fibres of such preparations when stimulating either the nerve explant or the cutaneous pectoris nerve. When two stimuli were paired by gradually reducing the interval of time between them, the second response was gradually facilitated. This confirmed that nerve fibres stimulated through the explant corresponded to new neuritic processes resulting from motor nerve sprouting. 5. Pieces of perineural tissue and segments of peripheral nerve killed by alcohol treatment or by repeated freezing and thawing were used as controls. They did not induce any nerve sprouting. 6. The results indicate that cells surviving in a segment of peripheral nerve trunk actively induce intact axons to sprout both from their terminals and intramuscular nodes of Ranvier; moreover these cells promote attraction and proliferation of growing neurites. The possibility of release of a diffusible factor by glial cells is discussed.

Presynaptic actions of botulinal neurotoxins at vertebrate neuromuscular junctions.

1. In the present paper we review some presynaptic aspects of the mode of action of botulinal toxins (BoTxs) at vertebrate neuromuscular junctions with emphasis on studies carried out in our laboratories using electrophysiological and morphological techniques. 2. Spontaneous quantal transmitter release recorded as miniature end-plate potentials is drastically affected by BoTxs. The low probability of release at poisoned terminals can be enhanced by carbonyl cyanide m-chlorophenylhydrazone (CCCP), Cd2+ and La3+. However, CCCP and La3+ which drastically deplete clear synaptic vesicles from unpoisoned terminals failed to markedly affect the density of synaptic vesicles at poisoned terminals. It is concluded that poisoned terminals have a reduced sensitivity to the release-promoting action of Ca2+, Cd2+ and La3+. 3. When comparing the effect of the various BoTxs on nerve-impulse evoked transmitter release it appears that increasing phasic Ca2+ entry into the terminals enhances evoked synchronized quantal release only from terminals poisoned with serotypes A and E. In contrast, enhanced Ca2+ entry into terminals poisoned with serotypes B, D and F induced a period of high frequency asynchronous release suggesting that these BoTxs may affect a presynaptic step beyond the influx of Ca2+, that may be involved in the synchronization of transmitter quanta. These data suggest that the actions of BoTxs involve several steps of the acetylcholine release process. 4. The analysis of presynaptic currents which depend on both Ca2+ entry and intraterminal background Ca2+ levels strongly suggests that neither Ca2+ entry nor intraterminal Ca2+ levels are altered by BoTxs. Furthermore, poisoned terminals are no more efficient than unpoisoned ones in dealing with Ca2+ overloads. 5. Finally, the morphological examination of junctions paralysed by BoTx-A indicates that the toxin triggers a particularly important overgrowth of the nerve terminals and suggests that the in vivo functional recovery may occur from an extension of the original nerve terminal arborization and the concomitant remodelling of postsynaptic structures.

Terminal nerve sprouting at the frog neuromuscular junction induced by prolonged tetrodotoxin blockade of nerve conduction.

The effects of a prolonged blockade of nerve conduction by tetrodotoxin on frog motor innervation were studied in the cutaneous pectoris muscle of Rana esculenta. Prolonged nerve blockade (up to 22 days) was obtained by repeated subperineural injections of tetrodotoxin. Changes in morphological parameters of neuromuscular junctions were investigated in muscles after staining with a combined cholinesterase-silver method. In addition, changes in the incidence of polyneuronal innervation were investigated conjointly by electrophysiology and morphology. Morphometric analysis of singly innervated muscle fibres of 60 microns diameter revealed insignificant changes during the first week of tetrodotoxin-nerve blockade. After 15 days of paralysis, the mean length of synaptic contacts and the mean length of terminal arborization per synapse were significantly increased as compared to controls (contralateral muscles and citrate buffer-injected controls). After 20-22 days, differences in synaptic and aborization mean lengths were accentuated and reached 44 and 43%, respectively. At that time, the mean number of terminal branching points and of continuous synaptic contacts were also significantly increased (around 20 and 50%, respectively). No changes in the length of abandoned gutters were observed. The incidence of focal polyaxonal innervation (detected morphologically) and of polyneuronal innervation (determined electrophysiologically) was unchanged. The results show that prolonged tetrodotoxin blockade induces sprouting of the terminal arborization which results in an extension of pre-existing nerve terminals and an increase in the complexity of terminal arborization by addition of new branches. Nodal (collateral) sprouting was not changed.

Effects of cadmium on quantal transmitter release and ultrastructure of frog motor nerve endings.

Exposure of frog cutaneous pectoris nerve-muscle preparations to cadmium (0.1-1 mM) results in an increase in miniature end-plate potential (m.e.p.p.) frequency. The increase is dependent on the concentration, the time of exposure and the co-presence of other divalent cations in the extracellular fluid. The stimulatory effect of cadmium is most marked in a calcium-free medium. Increased levels of calcium (4-10 mM) or of magnesium (10 mM) reduce the stimulatory effect suggesting that those cations interfere with the entry of cadmium into nerve endings. Once the effect of cadmium on m.e.p.p. frequency is attained, washing with a cadmium-free solution fails to abolish its effect. The action of cadmium on m.e.p.p. frequency slowly declines towards zero after about 3 hrs. An ultrastructural study of nerve terminals exposed for one hr to 1 mM cadmium reveals that neither in calcium-containing nor in a nominally calcium-free medium are there any significant changes in the number of synaptic vesicles as compared to controls. However, after 3 hrs of cadmium action in a calcium-free medium there is about 65% depletion of synaptic vesicles, while in calcium-containing media there is only about 25% depletion. The results suggest that cadmium by itself can support transmitter release but not synaptic vesicle recycling which instead might depend upon calcium.

Formation of functional endplates by spinal axons regenerating through a peripheral nerve graft. A study in the adult rat.

Peripheral nerve (PN) autografts were used in the adult rat to join the midcervical spinal cord to a nearby denervated skeletal muscle. Retrograde tracing, morphological and electrophysiological studies indicated the following: 1) a great number of neurons, located bilaterally, between C3 and C7 in most laminae of the grey matter, extended axons into the PN grafts, 2) a lesser number of neurons regenerated up to the reconnected muscle, but most of them were typical motoneurons, 3) neuromuscular junctions were formed in ectopic locations, around the tip of the grafted nerve, and at the sites of original endplates, 4) these junctions were functional and formed by axons that had regenerated into the PN bridges, as muscle contraction was obtained by electrical stimulation of the grafted nerves, 5) they were proved to be cholinergic since endplate potentials, evoked by stimulating the PN graft, were suppressed by curare. These results strongly suggest that spinal neurons, and especially motoneurons, are involved in the formation, through PN bridges, of new functional cholinergic connections with denervated skeletal muscles.

Sprouting of frog motor nerve terminals after long-term paralysis by botulinum type A toxin.

A single sublethal injection of botulinum type A toxin (BoTx-A) to winter frogs induced a general and complete paralysis of skeletal muscles, which lasted several months. Quantitative analysis of 483 end-plates from 8 BoTx-A poisoned muscles and 495 endplates from 8 control muscles revealed a higher and significant incidence of terminal and ultraterminal sprouts in poisoned junctions when taking into account the normal remodelling of motor innervation. We conclude that prolonged neuromuscular blockade by BoTx-A results in the extension of the nerve terminal arborization.