Healthcare and Medicine Reference
In-Depth Information
from three axons per muscle fiber to only one during
the second postnatal week.
The precise time course over which synapse elimi-
nation occurs, and the number of afferents lost, vary
greatly between areas of the nervous system, even
within a single species. In the rat cerebellum, the
elimination of climbing fiber synapses onto Purkinje
cells occur during the second postnatal week (Mariani
and Changeux, 1981). In contrast, the elimination of
preganglionic synapses onto neurons of the rat sub-
mandibular ganglion occurs over at least five postna-
tal weeks (Lichtman, 1977), far longer than is required
for elimination at the neuromuscular junction (Figure
9.3). The number of cochlear nerve synapses on
neurons of the chick cochlear nucleus declines rapidly,
from about four to two afferents (Figure 9.3), and
reaches a mature state even before hatching (Jackson
and Parks, 1982).
As the number of afferents increases, it becomes dif-
ficult to resolve small differences in PSP size. However,
functional estimates of synaptic convergence suggest
that elimination occurs even in systems that remain
multiply innervated as adults (Lichtman and Purves,
1980; Sanes, 1993). This physiological method is not
sensitive to certain forms of synapse elimination. For
example, climbing fiber axons innervate the soma and
dendrite of cerebellar Purkinje cells in neonates, but
the somatic synapses are eliminated during develop-
ment (Altman, 1972). A novel method for studying the
functional elimination of synapses makes use of a
technique in which glutamate is focally elevated in the
projecting population while an intracellular recording
is made from a postsynaptic neuron. This technique
was used to demonstrate a fourfold decrease in the
number of MNTB neurons innervating a single LSO
neuron during the first postnatal week in rats (Kim
and Kandler, 2003). Therefore, synapse elimination
appears to be a widespread phenomenon, although
there are no general rules about the percentage of affer-
ents that are lost or the duration of time required.
How is it possible to know whether axon or synapse
elimination occurs at central neurons that receive con-
tacts from hundreds or thousands of afferents? The
size of individual PSPs is too small and their ampli-
tude is too variable, leaving one with a cloud of synap-
tic potentials that do not increase in crisp steps. One
approach is to count all the synaptic contacts on a
neuron at several postnatal ages. For example, meas-
ures of synapse number and length taken from motor
neurons provide an estimate that 50% of synaptic con-
tacts are lost during development. Similar estimates
have been made for human cortex (Conradi and
Ronnevi, 1975; Huttenlocher and de Courten, 1987).
Impressive as these numbers are, there are two poten-
tial problems. First, they do not tell us whether the actual
number of axons making synapses onto postsynaptic
neurons change during development. For example, a
single afferent could initially make 100 weak synapses
that gradually transform into 50 strong ones. Second,
total synapse number may increase in some systems
during the time when a small fraction of synapses are
being eliminated, and this would go unnoticed.
If we assume for a moment that axons make
synapses wherever they arborize, then it should be
possible to study synapse elimination indirectly by
looking at the amount of territory occupied by the
axonal arborization. The most obvious case occurs
when a projection is eliminated entirely due to the
death of nerve cell bodies. In chicks, there is a projec-
tion from a brainstem nucleus, the isthmo-optic
nucleus, to the retina in which nearly 60% of the pro-
jecting cells die, particularly those projecting to the
wrong place in the retina (Casticas et al., 1987). The
elimination of errant projections from the olfactory
epithelium to the olfactory bulb may also result from
the selective loss of sensory neurons (Zou et al., 2004).
2 mV
5 mV
2 ms
2 ms
30 mV
20 ms
2 mV
4 mV
2 ms
2 ms
Rat Muscle
FIGURE 9.3 Three examples of decreased convergence as meas-
ured electrophysiologically (see Figure 9.2). On the right are shown
the increases in afferent-evoked PSP recorded in immature neurons
of the chick cochlear nucleus, rat autonomic ganglion, and rat NMJ.
There are 3-5 quantal increases in PSP amplitude. On the left are
shown the increases in afferent-evoked PSP amplitude in mature
neurons. There are 1-2 quantal increases in PSP amplitude, indicat-
ing the functional elimination of inputs. (Adapted from Jackson and
Parks, 1982; Lichtman, 1977; O'Brien et al., 1978)
Search Pocayo ::

Custom Search