Biomedical Engineering Reference
In-Depth Information
reduced into two L-shaped stripes positioned symmetrically around the midline
(Figure 4B). The cells in which the EGFR activity is initially very high and then
abruptly extinguished by Argos adopt the dorsal midline fate; the cells in which
the EGFR activity is maintained high adopt the dorsolateral fate; the cells in
which the EGFR activity is moderate and then quenched by Argos become the
dorsal appendage anlagen. In short, the initial gradient of EGFR activity induced
by a localized input (Gurken) is first amplified by a positive feedback (Spitz)
and afterward refined by a negative feedback (Argos), thereby creating three
distinct fates.
In both patterning systems, EGFR signaling mediates the patterning of gene
expression in the epithelial cells. Just as analyzing the similarities between the
two systems enables us to derive common regulatory principles in EGFR signal-
ing, examining the differences between the two systems may reveal further
mechanistic insights. In the ovary, Argos inhibits EGFR signaling in the Argos-
producing cells and their neighbors (11). In the embryo, on the other hand, the
Argos-producing cells appear to be refractory to its inhibitory action (10,14).
Understanding this observed difference may lead to a better grasp of the way
Argos functions. Another difference between the two systems is the absence of
positive feedback in the ventral ectoderm.
2
The ventral ectodermal patterning
has been shown to be highly robust. Halving or doubling the level of input from
the midline does not alter the pattern (10). In contrast, varying the level of input
from the oocyte generates diverse eggshell morphology phenotypes (20). In par-
ticular, defects in the positive feedback lead to eggshells with fused dorsal ap-
pendages (11). Hence, it appears that the positive feedback is necessary to
transform a pattern mediated by a simple inductive signal (Gurken) into a more
complex spatial pattern.
Long-range inhibition by secreted Argos plays a central role in both mecha-
nisms. Most likely, these mechanisms will have to be revised to account for the
recent discovery of the fact that Argos directly interacts with the EGFR ligand
Spitz (9), and not with the receptor itself.
3.
MODELING AND COMPUTATIONAL ANALYSIS
OF AUTOCRINE AND PARACRINE NETWORKS
Here we describe our initial steps toward building mechanistic models of
autocrine and paracrine EGFR signaling in epithelial layers. In each case we
briefly describe the model and illustrate the representative questions that the
model can help to address. The details of the derivations and computational
analysis can be found elsewhere (25-30). All the parameters used from this sec-
tion onward are listed in Table 1.
