Healthcare and Medicine Reference
In-Depth Information
Dark rear
TTX P14 to P40
layer IV
layer IV
3 H-proline
3 H-proline
FIGURE 9.20 Spontaneous retinal activity regulates the formation of stripes. A. Cats were reared in the
dark, and 3 H-proline was injected into one eye to visualize ocular dominance columns. Although the columns
were slightly degraded, they did form. B. Bilateral intraocular TTX injections were performed beginning at
postnatal day 14. When 3 H-proline was injected into one eye to visualize ocular dominance columns, it was
found that segregation of geniculate afferents into stripes failed to occur. Thus, spontaneous retinal activity
is sufficient to influence stripe formation. (Adapted from Stryker and Harris, 1986)
competition hypothesis suggests that synapse modifi-
cation depends on the pattern of activity. Is it possible
that spontaneous activity has a temporal or spatial
pattern? To address this question, the entire retina was
isolated from an embryo and placed in a perfused
recording chamber, where it is possible to record from
many retinal ganglion cells at the same time (Figure
9.21A). Synchronous bursts of action potentials were
recorded from many neurons about every minute or
two, and regions of maximal activity moved slowly across
the retina (Meister et al., 1991) (Figure 9.21B). Record-
ings from both the ferret and the mouse each show that
activity waves sweep across the entire retina during the
first few postnatal weeks (Wong et al., 1993). This pat-
terned activity gradually breaks down at about the time
of eye opening, but visual experience does not seem to
terminate the waves. They are lost at about the same
time even in dark-reared animals (Demas et al., 2003).
Spontaneous action potentials have also been
observed throughout the developing nervous system
and often displays interesting temporal or spatial pat-
terns (Lippe, 1994; O'Donovan et al., 1994; Kotak and
Sanes, 1995). Recordings from the nerve-muscle junc-
tion show that spontaneous activity is very low at birth
and gradually rises during the first two postnatal
weeks (Personius and Balice-Gordon, 2001). Interest-
ingly, motor axon firing is highly correlated during the
first postnatal week (that is, motor synapses are firing
in synchrony), perhaps preserving polyneuronal
innervation at the muscle. The period of synapse elim-
ination occurs when motor activity becomes tempo-
rally decorrelated. In the cortex, large-scale waves of
activity have been found to move slowly, from caudal
to rostral, during the first postnatal week (Figure
9.21C). This activity pattern involves the vast majority
of neurons and seems to end when inhibitory synapses
become hyperpolarizing (Garaschuk et al., 2000).
Are the temporal or spatial patterns of spontaneous
retinal activity required for the formation of specific
connections in the central nervous system? It turns out
that the activity waves within the retina depend on
cholinergic transmission. When cholinergic transmis-
sion is disrupted in mice by knocking out the b2 AChR
subunit, the retinal waves are lost. Retinal ganglion
cells continue to burst, but the rate of bursting differs
across the retina, and neighboring retinal neurons no
longer fire at the same time. Under these conditions,
projections from posterior retinal ganglion cells do not
become restricted to the anterior superior colliculus, as
they would in wild-type animals (McLaughlin et al.,
2003). The projection from retina to LGN is also less
precise as revealed by fine-grained electrophysiologi-
cal mapping (Grubb et al., 2003). The absolute levels of
spontaneous activity from each eye are crucial to estab-
lishing the mature lamination pattern in the LGN.
When the spontaneous activity of one eye is experi-
mentally increased by elevating cAMP levels, its pro-
jection within the LGN is expanded (Stellwagen and
Shatz, 2002). However, the segregation of thalamic
afferents into eye-specific layers may not depend on
the precise pattern of retinal activity in the postnatal
period (Huberman et al., 2003).
Search Pocayo ::

Custom Search