Development of eyespecific lamination in the lateral geniculate nucleus

The adult pattern of segregated inputs from the two eyes to distinct layers in the LGN is not present early in development. Instead, axons from the two eyes are initially completely overlapping throughout a large portion of the LGN in both carnivores and primates (Linden et al., 1981; Rakic,

Figure 8.1. Ocular dominance columns from a normal monkey (left) and a monkey subjected to monocular deprivation during the critical period (right). In the normal monkey the ipsilateral eye was injected with [3H]proline, which was transported transneu-ronally through the LGN and can be seen as the white stripes in the figure. In the normal monkey, the ocular dominance columns from the injected eye (white stripes) and the uninjected eye (dark stripes) are equal in width. In the monocularly deprived monkey, [3H]proline was injected into the deprived eye. The white stripes, representing the ocular dominance columns from this deprived eye, are much narrower than those seen in the normal animal or than those from the nondeprived eye (dark stripes) in this animal. (From LeVay et al., 1980.)

Figure 8.1. Ocular dominance columns from a normal monkey (left) and a monkey subjected to monocular deprivation during the critical period (right). In the normal monkey the ipsilateral eye was injected with [3H]proline, which was transported transneu-ronally through the LGN and can be seen as the white stripes in the figure. In the normal monkey, the ocular dominance columns from the injected eye (white stripes) and the uninjected eye (dark stripes) are equal in width. In the monocularly deprived monkey, [3H]proline was injected into the deprived eye. The white stripes, representing the ocular dominance columns from this deprived eye, are much narrower than those seen in the normal animal or than those from the nondeprived eye (dark stripes) in this animal. (From LeVay et al., 1980.)

1976; Shatz, 1983). During normal development in the carnivore, adult-like eye-specific segregation of afferents gradually appears as axonal branches in inappropriate locations are pruned and branches in appropriate locations grow and elaborate (Sretavan and Shatz, 1986). In the primate, few inappropriate axonal branches are seen during development, and eye-specific segregation of axons appears to occur through the loss of entire retinal fibers innervating inappropriate regions of the LGN (Lachia and Casagrande, 1988; Snider et al., 1998). Evidence that the segregation of retinal axons into eye-specific layers might be occurring through a competitive process in both carnivores and primates came originally from experiments in which one eye was removed from an animal early in development. When the axonal projection from the remaining eye was labeled later in development or in adulthood, axons were found to occupy nearly the entire LGN (Chalupa and Williams, 1984; Rakic, 1981). This indicates that interactions between the afferents from the two eyes, not just interactions between afferents and LGN cells, are necessary for normal eye-specific segregation in the LGN. The normal segregation of retinal afferents into layers in the LGN occurs in utero in most species and before eye opening in others; therefore, this process clearly does not depend on visual experience. However, the process of segregation has been found to depend on spontaneous activity of retinal ganglion cells. Retinal ganglion cells have been shown to be spontaneously active in utero (Galli and Maffei, 1988). Pharmacological blockade of this spontaneous activity prevents the segregation of retino-geniculate afferents (Penn et al., 1998; Shatz and Stryker, 1988), resulting in an LGN where afferents from the two eyes remain overlapping throughout most of the nucleus. The activity blockade, however, does not prevent axonal growth: axons in treated animals grow extensively (Sretavan et al., 1988). Nor does the activity blockade completely abolish the axons' ability to interact with postsynaptic targets: the axons in treated animals are not confined to the normal layers of the LGN, but they are confined to the LGN itself and do not grow into adjacent nuclei (Sretavan et al., 1988). These results suggest that during normal development a competitive, activity-dependent process is responsible for driving eye-specific segregation in the LGN.

Further evidence for competition between axons from the two eyes during development of eye-specific layers in the LGN comes from experiments in which the amount of neuronal activity in one eye is pharmacologically altered. If spontaneous activity in one eye is completely blocked, then the LGN territory occupied by that silenced eye's ganglion cell axons is reduced, while the territory occupied by axons from the normal eye is expanded (Penn et al., 1998). Conversely, if the amount of spontaneous activity in one eye is pharmacologically increased, that eye's axons gain territory in the LGN (Stellwagen and Shatz, 2002). These experiments show that the more active eye always "wins," regardless of whether it has normal or enhanced levels of activity, while the less active eye always "loses," regardless of whether it has normal or reduced levels of activity. The LGN layers, however, develop normally when both eyes are pharmacologically induced to have more spontaneous activity than normal (Stellwagen and Shatz, 2002), again indicating the importance of relative levels of activity between the two retinae.

The activity-dependent, competitive processes involved in the development of eye-specific segregation in the LGN

also appear to be important (at least for some period of time) in maintaining segregation. If spontaneous activity is blocked during a developmental time immediately after the retinogeniculate afferents have segregated, the axons actually desegregate and the projections from the two retinas become completely overlapping in the LGN (Chapman, 2000).

0 0

Post a comment