Prenatal development of cat retinogeniculate axon arbors in the absence of binocular interactions.

During prenatal development of the cat's retinogeniculate projection, inputs from the ganglion cell axons of the two eyes are initially intermixed with each other within the lateral geniculate nucleus (LGN). As development proceeds, the inputs sort out to give rise to the eye-specific layers characteristic of the adult. During this sorting out process, individual axons undergo a stereotyped sequence of morphological changes that ultimately produce the characteristic pattern of arborization in which axon arbors are restricted in extent only to those layers of the LGN appropriate for the eye of origin (Sretavan and Shatz, 1984, 1986). Here, we examine whether binocular interactions between retinal ganglion cell axons from the two eyes are required for the formation of this restricted pattern of terminal arborization. To examine this question, one eye was removed at embryonic day 23 (E23), when ganglion cell axons have not yet reached the optic chiasm, and the ganglion cell axons from the remaining eye were allowed to develop in the complete absence of binocular interactions. At E59, when segregation into eye-specific layers is normally almost complete, the retinogeniculate projection from the remaining eye was then examined both by anterograde transport following intraocular injections of 3H-leucine and by the in vitro filling of individual ganglion cell axons with HRP. Results from the intraocular injections showed that in the absence of one eye, the remaining eye is still capable of forming both ipsilateral and contralateral optic tracts; however, the projection was distributed diffusely throughout each LGN, rather than being confined to normal eye-specific territories. When individual HRP-filled axons were reconstructed and examined, it was remarkable to find that the pattern of terminal arborization was virtually indistinguishable from normal axons. As usual, arbors were restricted to the distal portion of each axon trunk, and measurements showed that the total linear length of axon contributing to each arbor was within the normal range (enucleated, 2310 +/- 920 microns; normal, 2520 +/- 810 microns). Furthermore, the terminal arborizations of axons appeared to be organized into a series of tiers within the LGN in a pattern surprisingly similar to the pattern of eye-specific layers normally present by E59. Also unchanged was the normally occurring loss of axons from the remaining optic nerve: Counts at E59 showed that about 2.3 X 10(5) axons were present in enucleated animals as compared to 2.5 X 10(5) axons in controls.(ABSTRACT TRUNCATED AT 400 WORDS)