The calf muscles of the rat hindlimb are composed of smaller entities, called neuromuscular compartments, which are the territories of muscle innervated by a single, naturally occurring primary (first-order) muscle nerve branch. While it is quite clear that a precise connectivity exists very early in development between motoneuron pools and individual muscles, the mechanisms responsible for producing the adult pattern of compartmental innervation are unknown. This study uses intracellular recording techniques to demonstrate that neuromuscular compartments are essentially established at birth and that postnatal synapse elimination has little role in establishing neuromuscular compartments. Our results demonstrate the existence of a small number of cross-compartmental connections in neonates which are not present in adults. Examining the removal of these cross-compartmental connections in both normal muscles and in muscles that have had synapse elimination delayed by tenotomy reveals that the synapses responsible for this innervation are eliminated in a selective manner.

The selective reinnervation of muscles suggests that muscles have intrinsic recognition cues that promote selective synaptogenesis. For example, the anterior and posterior heads of the axolotl iliotibialis (ILT) muscle are preferentially reinnervated by their original motoneurons even after surgically exchanging them. The nature and location of cues that promote such selectivity are unknown, although previous work suggests that the muscle fibers themselves might harbor the relevant molecules. To address this question, we removed anterior and posterior ILT muscles, destroyed their myofibers by surgically damaging them and treating them with bupivacaine, and reimplanted them in either a normal or a reversed anterior/posterior orientation. After the regenerated myofibers became innervated, we stimulated different spinal nerves and recorded the synaptic potentials evoked in muscle fibers. Our results showed that if the muscles were removed, damaged, and reimplanted in their original positions, the segmental origin of inputs to the regenerated myofibers was similar to that seen in normal muscles and in muscles reimplanted with their myofibers intact. However, muscles that were removed and damaged but regenerated in new positions were innervated differently from normal muscles and from muscles whose myofibers survived transplantation. Thus, the site at which a muscle regenerates has an influence on the source of the muscle's reinnervation. Nevertheless, the innervation of muscles that regenerated after transplantation to a foreign site was not strictly appropriate for the new position, but was biased towards the muscle's original innervation pattern. Therefore, some, but not all, of the cues that reflect the original identity of the transplanted muscles survive the replacement of its myofibers.