Over the last several years, we have shown that motor neuron pools map onto muscles with a rostrocaudal positional bias. Detailed studies from our lab revealed that this topographic map is detectable in embryonic muscles upon first contact between nerve and muscle, and is partially restored after denervation. We have developed an important model of synaptic competition during reinnervation, where we can predict with 95 percent accuracy the survivor between two competing nerve terminals. We have also developed an in vitro model to identify muscle membrane-bound labels that may be responsible for the positional bias. We have found selective growth of embryonic spinal cord neurites on membranes derived from embryonic rostral or caudal muscles or from transgenic muscle cell lines bearing a heritable memory for rostrocaudal position. We have recently focused our attention on the Eph A/ephrin A subfamily of tyrosine kinase receptors as a class of candidate molecules that regulate neuromuscular topography. We have found that all five members of the ephrin A subfamily are expressed in embryonic muscles, and that membrane expression of ephrin A ligands progressively diminishes during postnatal development. We have further found that overexpression of ephrin A5 or deletion of ephrin A5 and A2 degrades the topographic map. We propose to build on this series of observations in three ways. First, we will study the physiological basis for the altered topographic map by ephrins A using intracellular recording and uptake of activity dependent dyes into living nerve terminals. Second, we will extend our in vitro model for innervation topography using a wide array of neurite growth assays. In particular, we will examine growth on membranes of two particularly selective muscles, the gluteus and serratus anterior where 87 percent to 95 percent of the neurites making a choice grew selectively on membranes of similar axial position. We will also explore selective neurite growth within compartments of a single muscle. Third, we will use this in vitro model to search for molecular guidance cues other than ephrin A ligands that may cooperate in establishing the neuromuscular map. This will include the use of ephrin A5 fusion proteins to block endogenous ephrin A ligands. In addition, we will isolate membranes from mutant mice where ephrin A5 or A2/A5 genes have been deleted. In both cases we will search for residual selective growth by spinal motor neurites as a first step toward isolation of additional guidance molecules. Results of these studies will provide unique insight into how neurites in the peripheral nervous system recognize and synapse with their positionally matched partners. We will also learn whether positional labels in the neuromuscular system are part of a general strategy for encoding position in the nervous system.
