Retinal fibers alter tectal positional markers during the expansion of the retinal projection in goldfish.

Although widely accepted, the theory, tha t neurones carry immutable cytochemical markers which specify their synaptic connections, is not consistent with plastic reorganizations. Half retinal fish were therefore tested for changed markers following expansion. Optic nerve crush at the time of the half retinal ablation resulted in regeneration of a normal, restricted projection; but nerve crush following expansion (many months later) resulted in reestablishment of the expanded projection, assessed both by electrophysiological mapping and by radioautography. Since this implied changed markers, the half retina and tectum were tested independently using the ipsilateral tectum and eye as controls. In normal fish, removal of one tectum and deflection of the corresponding optic tract toward the remaining tectum resulted in regeneration of a positionally normal but ipsilateral map. In experimental fish, after the half retina had expanded its projection to the contralateral tectum, i ts optic tract was deflected to the control tectum. After 40 days i t had regenerated a normal, restricted map indicating tha t the retinal markers had not changed. Such restricted projections did not expand in the presence of the normal projection even after a year or more. Similarly, the optic tract from the normal eye was deflected to cause innervation of the tectum containing the expanded half retinal projection. After 40 days, the projection regenerated from the normal eye was similar to the expanded half retinal projection. Areas of the normal retina corresponding to the missing areas of the half retina were not represented. Tectal markers had been altered by the half retinal fibers. In a final group, tecta were denervated and tested at various intervals by innervation from ipsilateral half retinal eyes. After five months of denervation, the regenerating fibers were no longer restricted to the rostra1 tectum but formed an expanded projection initially. Apparently tectal markers are induced by the retinal fibers, changed during expansion, and disappear during long-term denervation. Following crush of the optic nerve in amphibians and teleosts, the retinal ganglion cells regenerate their axons to reestablish orderly connections in the brain (Matthey, ‘25; Sperry, ’44, ’45, ’48). The optic tectum, the major target area, receives a retinotopic projection and i t s regeneration is a highly selective process (At ta rd i and Sperry, ’ 6 3 ; Jacobson and Gaze, ’65). After removing portions of the retina and crushing the optic nerve in goldfish, Attardi and Sperry found histologically tha t the regenerating fibers bypassed empty areas to arborize only in appropriate areas. To explain this selectivity, Sperry ( ’63) postulated that, during developJ . COMP. NEUR., 177: 279-300 ment, the cells of both the retina and tectum had acquired individual “cytochemical tags” and tha t ingrowing axons linked only with those neurones carrying similar tags via “specific chemical affinities.” Later experiments involving ablations to portions of either the retina or tectum of goldfish showed tha t the connections, as assessed by the electrophysiological mapping technique, were capable of rearrangements so tha t whatever portion of retina remained projected topographically to fill the available tectal area (Systems Matching: Gaze and Keating, ’72). Half of the tectum could receive fibers from the entire retina (Gaze and Sharma, ’70;