straight and unbranched, as were axons in unlesioned cultures. Collaterals of the axons growing toward the lesion, in contrast, were highly branched, had clear growth cones, and extended in circuitous, meandering trajectories. In addition, the number of presynaptic boutons per unit length along the meandering axons near the lesion was increased twofold compared with axons extending away from the lesion or axons in control cultures.

We demonstrated that these collaterals were not merely preexisting axons that had become rearranged because of the proximity to the lesion but were in fact sprouted de novo by using an antibody directed against the growth-associated protein GAP-43. This protein is highly expressed early in development in vivo, when axons elongate, but is down-regulated to very low levels after synaptogenesis is complete (Skene, 1989). Consistent with these observations, GAP-43 immunoreactive fibers are numerous in hippocampal slice cultures after 3 days in vitro but are never seen by 14 days in vitro. In cultures in which the Schaffer collaterals were transected after 14 days in vitro, in contrast, there were numerous immunoreactive fibers with readily identifiable growth cones, beginning 3 days post lesion. By 14 days post lesion, large numbers of immunoreactive fibers crossed the lesion cavity itself and entered area CA1 (Figure 4). We showed that these regenerated axons form functional synapses with CA1 pyramidal cells and restore synaptic transmission across the lesion (McKinney et al., 1999a). By 21 days post lesion, there were no more GAP-43 immuno-reactive fibers, suggesting that sprouting had ceased. However, transmission across the lesion persisted, indicating that the newly generated axons remained and were functional. These data thus provided direct evidence that hippocampal pyramidal cells can sprout new axon collaterals after injury (see also Perez et al., 1996). The slow time course of this growth is interesting in the context of delayed development of posttraumatic epilepsy.

The functional consequences of this injury-induced axonal sprouting were examined electrophysiologically (McKinney et al., 1997). High frequencies of spontaneous EPSPs were observed in most intracellular recordings made from CA3 pyramidal cells 14 days post lesion. In addition, local stimulation within area CA3 elicited unusual poly-synaptic EPSPs that were often suprathreshold for action potential initiation. Similar pathological synaptic responses have been observed in ex vivo slices taken from the vicinity of neocortical lesions (Prince and Tseng, 1993). There was a strong positive correlation between the levels of GAP-43 immunoreactivity and the frequency of spontaneous EPSPs in individual lesioned cultures. Lesioned and unlesioned cultures were indistinguishable in the amount of GABA immunoreactivity and in the amplitude of pharmacologically isolated GABAergic inhibitory postsynaptic potentials (IPSPs). Finally, most lesioned cultures responded to a challenge with a subconvulsive concentration of the GABAa

A Five days after transection CA3

Was this article helpful?

0 0

Post a comment