Biology Reference
In-Depth Information
manifestation. At the core of the mechanism of convergence and extension by
cell intercalation is the idea that cells can forcibly intercalate between one
another along one axis, thus reducing that dimension of the tissue and
simultaneously forcing an increase in another, transverse dimension. The
reduction in one dimension and increase in another is coupled biomechani-
cally by conservation of volume. That is, as cells intercalate between one other
along one axis, they force an increase in the dimension of a perpendicular axis.
Ideally, halving the width of the tissue could produce a doubling of its length
for maximum eciency of translating convergence into extension (Figure
18.4A, left). But in fact, convergence generally results in increase in both
length and thickness (Figure 18.4A, right). For example, convergence in
Xenopus mesoderm and neural tissue produces both extension and thickening,
and each tissue appears to have a specific conversion ratio of convergence-in-
to-extension and convergence-in-to-thickening (Keller et al., 1989; Wilson et
al., 1989; Shih and Keller, 1992a,b). Complete translation of convergence to
extension seems never to occur, but instead, part of the reduction in width is
absorbed by thickening. According to our model, as cells actively intercalate
mediolaterally, they wedge between one another, and push one another
apart along the anterior-posterior axis, thus producing the extension (Figure
18.3B-D). However, the compression forces that cells come under in the
anterior-posterior axis during this wedging would also tend to force the cells
to move either above or below the plane of convergence and extension; that is,
it would tend to form a second layer and thicken the tissue (Figure 18.4B).
What is the process that resists thickening, or restores cells to their original
layer? The fact that tissues vary in the e
ciency of translation of convergence
into extension versus thickening suggests that whatever the process, the tissues
vary in its effectiveness. The obvious process is radial intercalation, the process
that thinned the tissue in the first place (Figure 18.2A). The initial event in the
overall process of convergence and extension is thinning and extension, driven
by radial intercalation (Keller and Danilchik, 1988). If the thinning forces
generated by radial intercalation continue during mediolateral intercalation, it
would tend to reduce the tendency of the latter to produce thickening by
reversing the process of radial intercalation (Figure 18.4C).
The converse question is why does radial intercalation give rise to extension
rather than uniform spreading in all directions during the initial phases of
convergence and extension (Figure 18.2A) (Wilson et al., 1989; Wilson and
Keller, 1991)? Radial intercalation seems to be an isotropic process in the
plane of the tissue (Wilson and Keller, 1991), and therefore it should give rise
to spreading in all directions. It does so during epiboly of the animal cap
(Keller, 1978, 1980), but in the initial phases of convergence and extension of
the marginal zone, it produces extension. Using the same argument applied to
mediolateral intercalation above, radial intercalation, even if it were biased to
occur only between anterior-posterior neighbours, would tend to also produce
