Regionally specific distribution of corticospinal synapses because of activity-dependent synapse elimination in vitro.

We have shown previously that the corticospinal tract (CST) with functional connections can be reconstructed in vitro in slice cocultures. Using that system, we stimulated the deep cortical layer and recorded field EPSPs (fEPSPs) along a 100 microm-interval lattice in the spinal gray matter. The specific, spatial synapse distribution on the dorsal side at 14 d in vitro (DIV) basically corresponded to the in vivo area in which CST axons terminate. Anterograde labeling of corticospinal axons with biocytin showed a similar terminal distribution on that side. In vitro development of synapse spatial distribution was investigated. fEPSPs were recorded all across the gray matter at 7 DIV, but amplitudes began to decrease on the ventral side at 9 DIV, dorsal-dominant distribution being nearly complete at 14 DIV. Anterograde labeling showed that the decrease in fEPSP amplitudes was associated with a decrease in the number of axon terminals on the ventral area. Decreases in the synaptic responses and terminals were blocked by applications of D-2-amino-5-phosphonovaleric acid and tetrodotoxin, whereas 6-cyano-7-nitroquinoxaline-2,3-dione had a partial effect. These findings suggest that this regressive event, which occurs during development, is activity and NMDA dependent. Retrograde labeling with two colors of beads and an electrophysiological study that investigated the axon reflex showed that at 7 DIV most corticospinal neurons project to both the ventral and dorsal spinal cord, indicating that synapse decrease on the ventral side is attributable primarily to axon branch elimination rather than to death of cortical cells that send axons solely to that side.