No abstract

Axon size in the optic nerves of three adult cats was measured using an accurate planimetric method. A sample of the axon complement was obtained from sets of more than 125 electron micrographs taken at each nerve. In contrast to previous reports, the distribution of axon diameter, as estimated from measurements of axonal cross-sectional area was clearly bimodal. The group of finest axons had a mode at approximately 1 micrometer, and the mode of the group of intermediate size axons was approximately 2 micrometers. Fibers with diameters larger than 3.5 micrometers formed the extensive tail of the distribution. The proportions of large, intermediate, and small axons were estimated to be 5%, 45%, and 50%. These groups probably correspond to the alpha, beta, and gamma classes of retinal ganglion cells. Regional distributions of axon caliber were also bimodal and, in a few instances, even trimodal. A peripheral crescent of each nerve contained a large number of fine fibers. Mean fiber diameter within discrete regions of the optic nerve varied inversely with the density of axon packing. The thickness of the myelin sheath was highly correlated with the inner fiber diameter, hence, myelin sheath width was also distributed in a bimodal manner. These results furnish the first clear evidence that the distribution of fiber caliber within the cat's optic nerve mirrors the principal classes of retinal ganglion cells.

The medial portion of the cat's lateral posterior-pulvinar complex (LPm) receives a prominent ascending projection from the superficial layers of the superior colliculus. This region of the thalamus has been suggested to serve as a relay by which visual information from the midbrain could be conveyed to extrastriate cortex. In order to determine how the functional organization within the LPm compares with that of the superior colliculus, visual response properties of LPm and superior collicular neurons were examined under identical experimental conditions. The majority of neurons in the LPm, as in the superior colliculus, respond vigorously to moving stimuli, and a substantial proportion of these cells also exhibit a preference for movements in a particular direction. Furthermore, most cells in the LPm, in common with those of the tectum, respond only in a phasic manner to flashed stimuli, have homogeneous receptive field organization, and show response suppression to stimuli larger than the activating region of the receptive field. As in the colliculus, the ipsilateral visual field is represented in the LPm. In spite of these similarities, there are also some striking differences between the visual responses of LPm and collicular neurons. First, the average size of receptive fields of neurons within the LPm is at least twice that of units in the superficial gray layers of the tectum. Second, more cells in the colliculus are directionally selective than those in the LPm, and the distribution of preferred directions is different in the two regions. Third, an appreciable proportion (27%) of the cells in the LPm are orientation selective, whereas this response property is only rarely encountered in the cat's tectum. Fourth, many LPm neurons can only be activated by binocular stimulation, whereas most collicular units respond equally well to stimulation of either eye. Collectively, these findings indicate that there is a substantial transformation in the lateral posterior-pulvinar complex of the ascending visual influx provided by the superior colliculus.

The large Spanish wildcat, Felis silvestris tartessia, has retained features of the Pleistocene ancestor of the modern domestic cat, F. catus. To gauge the direction and magnitude of short-term evolutionary change in this lineage, we have compared the retina, the optic nerve, and the dorsal lateral geniculate nucleus (LGN) of Spanish wildcats and their domestic relatives. Retinas of the two species have the same area. However, densities of cone photoreceptors are higher in wildcat--over 100% higher in the area centralis--whereas rod densities are as high, or higher, in the domestic lineage. Densities of retinal ganglion cells are typically 20-100% higher across the wildcat retina, and the total ganglion cell population is nearly 70% higher than in the domestic cat. These differences are confined to the populations of beta and gamma retinal ganglion cells. In contrast, the population of alpha cells is almost precisely the same in both species. The wildcat LGN is much larger than that of the domestic cat and contains approximately 50% more neurons. However, cell size does not differ appreciably in either the retina or LGN of these species. The differences in total numbers of ganglion cells and LGN neurons correspond neatly to the overall decline in brain size in the domestic lineage and to allometric predictions based on average species differences in body size. We suggest that an increase in the severity of naturally occurring cell death is the most plausible mechanism that can account for the rapid evolutionary reduction in cell populations in this feline lineage.

Between the 48th day of gestation (E-48) and maturity, the number of axons in the cat optic nerve is reduced by approximately 50%. On the basis of an electron microscopic assay, the axon population of the E-48 nerve was estimated to be 328,000. In contrast, estimates from two normal adults were 159,000 and 158,000. In utero unilateral enucleation (at E-45 and E-46) attenuated the severity of this loss since the optic nerves of the experimental animals contained 200,000 and 198,000 fibers. These results indicate that prenatal binocular competition is involved in the elimination of ganglion cell axons during the normal development of the cat's visual system. The increased number of axons in the optic nerve of the prenatally enucleated animals could be due to reduced ganglion cell death or a failure to retract supernumerary axon collaterals. It is suggested that the former explanation is more consistent with what is currently known about the development of retinofugal projections.Early in development,, retinal projections to the dorsal lateral geniculate nucleus and the superior colliculus are diffuse (Rakic, 1976(Rakic, , 1977 Cavalcante and Rocha-Miranda, 1978; Frost et al., 1979; Land and Lund, 1979;Linden et al., 1981;Williams and Chalupa, 1982a), whereas the mature connections are characterized by distinct ocular dominance domains and an exquisite topography. The mechanisms which underlie this remarkable transformation are unknown. However, recent studies suggest that cell death may contribute to the development of the mammalian visual system and, in particular, to the refinement of patterns of connection (Cunningham et al., 1979(Cunningham et al., , 1981 Jeffery and Perry, 1981; Finlay, 1981, 1982;Stone et al., 1982). If ganglion cell death underlies the restructuring of retinal projections that we and others have recently demonstrated in the cat (Shatz and DiBerardino, 1980; Kliot and Shatz, 1981; Chalupa, 1980, 1981 1982a), then the optic nerve of the fetus must contain a greater number of fibers than that of the adult. Furthermore, if this attrition is related to axon-target interactions, then prenatal unilateral enucleation should attenuate the severity of in ,utero retinal cell loss since the number of potential postsynaptic target sites available to axons of the spared ey8 could be doubled (cf. Rakic, 1979). The results of th8 present study are consistent with both of these ideas. We have found a significant attrition in the population of retinal ganglion cell axons during normal development. This loss is attenuated by the elimination of prenatal binocular competition. Materials and MethodsFive cats were used in this study. Two were normal adults: a lo-month-old 2.5-kg female tabby and an 18-month-old 3.8-kg black and white male. The two experimental animals had one eye removed more than 2 weeks before their natural birth, one on embryonic day 45 (E-45), the other on day E-46, and were both born on the 63rd day of gestation. These unilateral enucleates were males with black and white coats. Th...