Topographic disorganization of the optic tracts following long-term optic nerve regeneration: a quantitative image analysis study.

Experiments were designed to find the degree to which regenerated optic axons occupied their previous locations in the optic tracts. Following optic nerve crush and regeneration, either the dorsal, ventral, peripheral, temporal, or nasal part of the retina was ablated. The axons of the remaining retinal ganglion cells (RGCs) were labeled with cobalt. Density of the regenerated dorsal and ventral axons in the dorsal vs. ventral optic tracts was determined digitally. In addition, we determined the density of temporal and nasal axons in the temporal vs. nasal compartments of each optic tract and the density of central axons in the central vs. peripheral compartments of both optic tracts. Regenerated axons were not distributed randomly in the optic tracts. Instead, they were slightly but, significantly biased toward growing through the tract or compartment that they had occupied previously. Still, the pathway specificity exhibited by the regenerated axons was closer to random than it was to the pathway specificity seen in normal animals. Dorsal, ventral, and central RGC axons were significantly better localized to their correct tract or compartment than were temporal or nasal RGC axons. Also, over time, dorsal and ventral axons tended to disappear from incorrectly chosen optic tracts. The slight bias toward choosing the appropriate optic tract or optic tract compartment may be enough to account for the topographic specificity of the regenerated retinotectal projection. Near-randomness of the axonal positions in the tracts argues against the presence of any specific guidance cues in the optic tracts of adult animals. Axonal density was highest in the correct compartment and diminished progressively with increasing distance into the incorrect compartment. Such a gradient of axonal density suggests that regenerating axons "drift" away from their previous positions in the optic pathways.