Boundaries to Epithelial Movement - Mechanisms of Morphogenesis - page 256

CHAPTER
21

Boundari es to Epithelial M ovement

One theme that has run through the chapters in this section of the topic is that the spatial
relationships between epithelial cells of animals are flexible and fluid, quite unlike those seen
in plants. Driven by internal forces mediated by cytoskeletal elements or by cell-cell adhe-
sions, cells can move within the plane of the epithelium, swapping neighbours with relative
ease. The same is true of endothelial cells. This flexibility is a tremendous asset to a devel-
oping embryo and it underpins the complex morphogenetic processes that make our own
bodies what they are. It does, however, create a potential problem: if cells can move so easily
within the plane of an epithelium, how can they be prevented from mixing freely when the
embryo needs to divide an epithelium up into distinct zones of differentiation?

AV OIDING THE PROBLEM OF BOUNDARY CONTR OL

One way of solving the problem of a cell's tendency to move and disrupt an orderly
boundary is to allow movement of cells but to force any that have crossed a boundary to
change state appropriately. 1 This approach relies on a 'continuously active' patterning mech-
anism. In practice, 'continuously active' does not mean throughout life, but just for the period
of development during which movement is still possible and not prevented by further
anatomical changes or by cell differentiation. An example of this type of system seems,
according to current evidence, to be provided by the boundary between the midbrain and
hindbrain of chick embryos ( Figure 21.1 a). Cells in the midbrain neurepithelium are in
a slightly different state to those in the hindbrain neurepithelium, confirming that these are
distinct zones of the neural tube. This difference is demonstrated by the expression of the tran-
scription factor Otx2 by midbrain cells and Gbx2 by hindbrain cells. Unlike cells from adjacent
zones (rhombomeres) within the hindbrain itself, which refuse to mix when placed together in
cell culture, cells from the midbrain and the adjacent hindbrain will mix freely. 2 What is more,
dye labelling of small numbers of midbrain or hindbrain cells in chick embryos shows them to
cross the midbrain-hindbrain border even in vivo ( Figure 21.1 b) Nevertheless, despite evident
cell movement, the boundary between the zones as visualized by Otx2 and Gbx2 in situ
hybridization remains clean and smooth. A solution to this apparent paradox is suggested
by cell transplantation experiments, in which Otx2-positive cells from the midbrain were
transplanted into Gbx2-expressing hindbrain regions; the cells switched to expressing
Gbx2 ref3 ( Figure 21.1 c). This strongly suggests that the cells are responding to their