Healthcare and Medicine Reference
In-Depth Information
larva, the growth cones of about 40 motor neurons
touch about 30 different muscles before they select
those one or two onto which they will synapse (Nose
et al., 1992b; Broadie et al., 1993). The differences
between the muscles are subtle since they are, by and
large, similar to one another. Each muscle has a variety
of molecules on its surface, many of which are the
same as the molecules expressed on the membranes of
all its neighbors. The difference is that an individual
muscle cell also expresses cell surface molecules that
are shared with only some of its neighbors, and it
usually expresses different concentrations of the same
molecules as its neighbors (Winberg et al., 1998).
Some motor neurons normally innervate muscles
that express high levels of Netrin. In embryos in which
Netrin is not expressed, the axons that would normally
innervate some of these muscles go to inappropriate
muscles (Mitchell et al., 1996). However, two other
muscle cells that normally express Netrin remain
innervated in these mutants, suggesting that addi-
tional recognition molecules must be involved in these
muscles. Connectin is a second molecule that plays a
role in nerve-muscle specificity in this system. It is a
homophilic cell adhesion molecule that is expressed,
under the direct control of a homeotic gene on the
surface of a subset of motor neurons and the muscle
cells that they innervate (Gould and White, 1992; Nose
et al., 1992a; Meadows et al., 1994; Nose et al., 1994,
1997; 1997; Raghavan and White, 1997). There are
few neuromuscular innervation defects in connectin
mutants, however; when connectin is expressed
ectopically on all muscles in transgenic flies, motor
axons frequently make targeting errors and invade
nontarget muscles adjacent to their normal targets. The
defects seen with connectin overexpression may be
attributed to increased adhesion between different
muscles that do not normally adhere to each other,
making it difficult for the axon to take its usual
pathway through the muscle field.
A third factor that plays a role in this system is the
homophilic adhesion molecule, FasII, which is also
expressed differentially on subsets of muscle fibers
(Schuster et al., 1996a, b). Since FasII is expressed on
many muscle cells at different levels, a more subtle
experiment was done, which was to use various cell-
type specific promoters to switch the relative levels of
FasII expressed on specific muscles. The result is that
extra synapses form on muscles that express higher
levels of FasII at the expense of synapses formed on
neighboring muscles that do no not have increased
FasII. This is true when the level of FasII on the less
innervated muscles is high or low, so it is the relative
and not the absolute level of FasII that is important.
FasIII, another homophilic adhesion molecule, and
SemaII, a secreted growth cone repulsive factor, are
also expressed on overlapping specific muscle subsets
in Drosophila . As with connectin mutants, loss of func-
tion mutants in these molecules displays no serious
effects on neuromuscular targeting (Winberg et al.,
1998). But as for the other molecules described, misex-
pression of FasIII or SemaII in inappropriate muscles
leads to dramatic targeting effects. The change in prob-
ability of particular motor neurons targeting particu-
lar muscles caused by experimentally changing the
levels of the single-cell adhesion molecule is consistent
with the idea that growth cones are able to distinguish
targets by relative changes in the concentrations of a
number of such molecules. Furthermore, targeting
errors caused by the increase in an attractive or adhe-
sive factor can be compensated by a simultaneous
increase in a repulsive factor, showing that indeed the
combination of amounts of various such factors is
what counts. In summary, the results with FasII, FasIII,
connectin, and SemaII, and the netrin suggest that
cellular targeting in the Drosophila neuromuscular
is based on a combinatorial code involving all these
molecules and perhaps others (Figure 6.24).
In nematodes, a very intriguing case of cellular tar-
geting is provided by the hermaphroditic specific
motor neurons that innervate the vulva. In syg-1 and
syg-2 mutants, vulval muscles remain uninnervated,
and the neurons make ectopic synapses on inappropri-
ate targets (Shen and Bargmann, 2003; Shen et al., 2004).
The Syg-1 and Syg-2 proteins are adhesion molecules
of the IgG superfamily. Syg-1 is expressed in the neuron
and its binding partner Syg-2 is normally expressed
transiently not on the postsynaptic targets but on a
vulval epithelial guidepost cell. This interaction is crit-
ical for the maturation of the axonal terminal in prepa-
ration for synapse formation on the adjacent region of
the neuron. Interestingly, if Syg-2 is expressed under
the control of a promoter that causes it to be localized
on other epithelial cells, the hermaphrodite-specific
motor neurons begin to make presynaptic specializa-
tions at these ectopic sites. Thus, heterophilic binding
between Syg-1 and Syg-2 is involved in setting up the
formation of appropriate synapses.
SynCAMS and cadherins (especially protocad-
herins) are large families of homophilic adhesion
molecules that may add another level of specificity by
helping presynaptic terminals make contacts at the
correct subcellular locations (Abbas, 2003; Yamagata et
al., 2003). There are more than 50 different cadherins,
and many of them are expressed at subsets of synapses.
As many as 20 different genes of the cadherin super-
family genes are expressed in restricted patterns in the
developing tectum (Miskevich et al., 1998), and recent
speculation is that these types of molecules, like the
Search Pocayo ::

Custom Search