Staining of regenerated optic arbors in goldfish tectum: Progressive changes in immature arbors and a comparison of mature regenerated arbors with normal arbors

Abstract
Individual optic arbors, normal and regenerated, were stained via anterograde transport of HRP and viewed in tectal whole mounts. Camera lucida drawings were made of 119 normal optic arbors and of 242 regenerated arbors from fish 2 weeks to 14 months postcrush. These arbors were analyzed for axonal trajectory, spatial extent in the horizontal plane, degree of branching, number of branch endings, average depth, and degree of stratification. Normal optic arbors ranged in size from roughly 100 to 400 μm across in a continuous distribution, had an average of 20 branch endings with average of fifth‐order branching, and were highly stratified into one of three planes within the major optic lamina (SO‐SFGS). Small arbors arising from fine‐caliber axons terminated in the most superficial plane of SO‐SFGS; large arbors from coarse axons terminated in the superficial and middle planes; and medium arbors from medium‐caliber axons terminated in the middle and deep planes of SO‐SFGS, as well as deeper in the central gray and deep white layers. Arbors from central tectum tended to be much more tightly stratified than those in the periphery. No other differences between central and peripheral arbors were noted. Mature regenerated arbors (five months or more postcrush) were normal in their number of branch endings, order of branching, and depth of termination. Their branches covered a wider area of tectum, partially because of their early branching and abnormal trajectories of branches. Axonal trajectories were often abnormal with U‐turns and tortuous paths. Fine‐, medium‐, and coarse‐caliber axons were again present and gave rise to small, medium, and large arbors at roughly the same depths as in the normals. There was frequently a lack of stratification in the medium and large arbors, which spanned much greater depths than normal. Overall, however, regenerates reestablished nearly normal morphology except for axonal trajectory and stratification. Early in regeneration, the arbors went through a series of changes. At 2 weeks postcrush, regenerated axons had grown branches over a wider‐than‐normal extent of tectum, though they were sparsely branched and often tipped with growth cones. At 3 weeks, the branches were more numerous and covered a still wider extent (average of five times normal), many covering more than half the tectal length or width. At 4–5 weeks smaller arbors predominated, although a few enlarged arbors were present for up to 8 weeks. Additional small changes occurred beyond 8 weeks as the arbors became progressively more normal in appearance. The number of branch endings increased from 2 to 4 weeks and dropped slightly between 4 and 6 weeks. This drop, although it did not reach statistical significance, was correlated with the drop in size of the arbors and with the time of sharpening of the retinotopic map. Stratification of the regenerated arbors was poor, improved only very slowly as the arbors matured, and never returned to normal. These findings indicate that the widely branched arbors are a transient stage in regeneration, corresponding to the early diffuse retinotopic map before sharpening. The time of the widely branched arbors and the elimination of extraneous branches corresponds to the period of maximum sensitivity to strobe and to TTX demonstrated in other studies. The role of activity in refining the arbors is discussed.