Biomedical Engineering Reference
In-Depth Information
tions between eggshells with zero, one, and two dorsal appendages. Finally, sta-
ble patterns with three or four peaks emerge in the model when the two-peaked
pattern is destabilized by the same mechanisms that generate the two-peaked
pattern itself. Since the number of peaks in the pattern corresponds to the num-
ber of dorsal appendages, this finding provides the mechanistic basis for ex-
plaining complex morphologies in mutants of
Drosophila melanogaster
(50,51).
In addition, this versatility in patterning may account for more complex eggshell
morphologies in related fly species (52,53).
This phenomenological model did not explicitly account for the details of
EGFR interaction with its ligands. Our main goal was to examine the pattern
formation capability of the localized input (i.e., Gurken) modulated by a net-
work of spatially distributed feedback loops (i.e., Argos and Spitz). As with any
phenomenological modeling, our model tested the sufficiency of the proposed
mechanism, but did not prove it at the genetic or biochemical level. A mechanis-
tic approach to modeling of EGFR-mediated signaling in oogenesis is now pos-
sible, based on the biochemical analysis of EGFR/Argos/Spitz interactions (9).
4.
CONCLUSIONS AND OUTLOOK
At this time, only a few dozen out of ~30,000 EGFR-related PubMed en-
tries are dedicated to modeling and computational analysis of EGFR signaling.
Most of the existing models are formulated at the molecular and cellular levels
(5). However, to understand how this system operates in vivo we need modeling
at the level of tissues. Even the simplest models of EGFR signaling in multicel-
lular systems must simultaneously account for ligand release, transport, binding,
intracellular signaling, and gene expression. Given this complexity, the inte-
grated models are nontrivial to test experimentally. We believe that a combined
modeling-experimental approach is possible in
Drosophila
, where a number of
genetic tools are available for implementing the model-directed manipulations in
vivo.
We have described two systems from
Drosophila
development where mod-
eling seems both feasible and necessary. In both cases, a large amount of data
was summarized in the form of a complex patterning mechanism. The feasibility
of these mechanisms depends on the quantitative parameters, such as the spatial
ranges of secreted signals and the strengths of the feedback loops. Modeling can
be used to elucidate the quantitative constraints on the proposed patterning
mechanisms and to dissect the relative contributions of multiple cellular proc-
esses. We are just beginning to develop and test mechanistic models of EGFR
signaling in tissues. Currently, we rely on the large amount of biochemical and
cellular experiments in mammalian systems. In the future, direct biophysical
characterization of the
Drosophila
EGFR network will be required in order to
develop truly mechanistic models of EGFR signaling in fruit fly development.
