Axon trajectories and pattern of terminal arborization during the prenatal development of the cat's retinogeniculate pathway.

In this study we have examined the trajectories taken by populations of ganglion cell axons and the spatial gradients of terminal arbor maturity within the lateral geniculate nucleus (LGN) during the prenatal development of the cat's visual system. To do so, an in vitro method of labeling optic tract axons from fetal brains between embryonic day 37 (E37) and postnatal day 2 (P2) with horseradish peroxidase (HRP) was used. At the earliest ages studied (E37-E53), optic axons leave the optic tract to run across the LGN toward their sites of termination in straight trajectories parallel to each other. At later ages (E57-P2), however, axons with abrupt changes in their course across the nucleus can be clearly identified. When the detailed terminal arbor morphology of the set of retinogeniculate axons filled with HRP at a given age was examined, two different spatial gradients of maturation could be detected. The terminal arbors of axons within LGN layer A are always more mature than those ending in layer A1, an observation consistent with previous findings that axons from the contralateral eye arrive within the LGN several days before those from the ipsilateral eye. Moreover, the terminal arbors of axons projecting to the medial portions of each layer are always more mature than their more lateral counterparts. These gradients are likely to be a direct reflection of the central-first, peripheral-last gradient associated with the neurogenesis of the retinal ganglion cells themselves. In the oldest animals studied (E58-P2), a remarkable periodic pattern of terminal arbor labeling was seen following a localized HRP injection into the optic tract. Within the labeled portions of the LGN, densely filled axon terminal arbors are separated by unlabeled gaps of similar width. This pattern of labeling could reflect local topographic disorder within the optic tract or could arise if axons of different classes of retinal ganglion cells run in separate portions of the optic tract. Taken together, all of these observations suggest that there may be a fair degree of topographic order in the retinogeniculate projection within the cat's LGN early on in development. However, when topographic errors are present, some can be corrected by minor readjustments in axonal trajectories.