Spinal cord injury treatment by induction of a shift from cholinergic to glutamatergic innervation of muscle fibers

A series of experimental and clinical reports published over the last few years have suggested a possible new strategy for treatment of spinal cord injury (SCI). The purpose of this editorial is to bring this new approach to the attention of the readers of this journal. The basic concept is that direct connection of a paralyzed muscle to the spinal cord rostral to the lesion by means of a nerve graft can lead to: (1) reinnervation of that muscle by supraspinal glutamatergic descending axons, (2) a shift from cholinergic to glutamatergic innervation of muscle fibers and, eventually, (3) the partial recovery of voluntary control of that muscle.16 The idea to treat the consequences of SCI by directly reinnervating paralyzed muscles with axons coming from rostral to the spinal lesion level emerged from experimental evidence that found that interrupted central nervous system (CNS) axons (the regeneration of which is strongly limited by the “nonpermissive” milieu of the CNS) are in fact able to elongate inside a more “permissive” peripheral nervous system (PNS) milieu represented by a nerve graft.17 These observations opened the possibility of regaining lost motor function by connecting the spinal cord rostral to the lesion site directly to the periphery with the interposition of a nerve graft. The distal stump of the nerve graft can then be sutured either to a nerve, through which axons will eventually reach one or more paralyzed muscles, or directly to muscles. Two different strategies based on the direct CNS–PNS connection have been attempted so far. The first approach employs a nerve graft to connect the periphery with the motoneurons of the ventral horn rostral to the injury. In a series of experimental studies using various animal models including nonhuman primates, it has been shown that these motoneurons can extend axons to the paralyzed skeletal muscles to form functional endplates1, 10, 11, 13, 15 and the first application of this surgical approach to a human patient led to encouraging results.19 The focus of this editorial, however, is on the CNS–PNS bypass surgical approach proposed by Brunelli et al.6 in 1983. Those authors sought to provide a regeneration pathway toward the periphery, not for motoneuron axons but for descending supraspinal axons that were interrupted at the lesion site. In their first experimental studies in rats they showed that supraspinal axons can: (1) grow along a peripheral nerve, bypassing the motoneuron,6 (2) extend axons to skeletal muscles, and (3) reinnervate them providing a partial recovery of function.7 The potential implications of these results for the treatment of SCI were obvious. The application of Brunelli's surgical technique in a nonhuman primate model of SCI was reported throughout the second half of the 1990s.3, 8 In Macaca fascicularis, the authors anastomosed the lateral bundle of the severed spinal cord directly with the sciatic nerve by means of a peroneal nerve graft and showed that, as seen in rodents, upper CNS axons can grow along a nerve graft connected to a distal nerve, elongate along the distal nerve itself, reach the target muscle, and partially restore lost motor function. The results of these experimental studies in nonhuman primates led to the application of this surgical technique on a paraplegic patient in July 2000.3 In a preliminary report of the postoperative follow-up, the patient was reported to show some voluntary and selective contractions 12 months after surgery3 and after 2 years the patient was able to stand, walk with a light walker for short distances, and use a cyclette.4 Further follow-up showed that improvement continued until the patient was able to stand for long times and walk with tripod sticks.5 The effectiveness of the technique, which awaits confirmation by others, raises two key questions: which are the central axons that regenerate toward the muscles, and what are the neurotransmitters of the new neuromuscular endplates formed by descending supraspinal axons? A recent experimental study9 provided answers to both questions: axons regenerating along the nerve graft originate from glutamatergic supraspinal neurons belonging to the rubrospinal, reticulospinal, and vestibulospinal tracts; and glutamatergic axons form functionally active endplates with skeletal muscle fibers. The newly formed neuromuscular junctions express AMPA glutamate receptors and are selectively blocked by glutamate receptor antagonist and not by cholinergic receptor blockers, to which the reinnervated muscle became resistant. Although the existence of glutamatergic neuromuscular junctions has been demonstrated in invertebrates and fish,12, 20 these recent experiments showed for the first time that, in mammals, neuromuscular junctions that are normally cholinergic can shift to glutamatergic in particular conditions and function effectively to control skeletal muscle. The demonstration of glutamate transporter expression in mammalian skeletal muscles,2, 18 where glutamate is thought to play a role as a cotransmitter at the neuromuscular junction,2 suggests that some elements needed for glutamate signaling are normally present in the mammalian neuromuscular junction, thus providing an explantation of how it is possible that skeletal muscles are able to shift from cholinergic to glutamatergic innervation. Besides the potential revolutionary value of this paradigm in the treatment of SCI, these results could also have a significant conceptual impact on research on the basic biology of skeletal muscle, as they challenge a widely held paradigm of neuromuscular function.14 The claim of new strategies in the treatment of clinical conditions with high social impact, such as SCI, should always be assessed with caution to avoid the risk of generating unjustified hopes in patients and their families. Several important steps are needed before the effective relevance of Brunelli's strategy can be assessed as a means of treatment of SCI patients. However, the perspectives of this technique are potentially very important and they are definitely worth further exploration both by basic and clinical researchers. Still to be elucidated are the involved biological mechanisms, the range of clinical applications, and ways to improve its effectiveness. For instance, this technique does not seem applicable to persons with cervical SCI because the process of inserting grafts in the rostral cervical spinal cord would produce secondary damage impairing existing function, whereas impairing the function of thoracic spinal levels would not significantly alter an individual's functional capacity. In the following years it is to be hoped that involvement of researchers in the study of this peculiar type of plasticity of the neuromuscular system will occur, with potential benefit at a fundamental and clinical level.