Studying fibroblast growth factor (FGF) mediated cell migration in "Drosophila" larval air sacs

Invertebrates and vertebrates use FGF signaling in many developmental processes. Mesoderm
formation, limb outgrowth but also the development of the vascular system and the lung rely
on FGF ligands. We have chosen to study the Drosophila FGF signaling pathway that has
been shown to be required for mesodermal- as well as tracheal cell migration. We aimed at a
better understanding of FGF signaling to elucidate how the extracellular information,
provided by the FGF/Bnl ligand is interpreted in tracheal cells. Using Downstream of FGFR
(Dof), an adaptor protein of the FGF signaling pathway, as an entry point, we have previously
identified interacting proteins and focused on one prime candidate as a potential linker of
FGFR to the cytoskeleton. This candidate protein Receptor of protein kinase C (Rack1) is
conserved throughout evolution. rack1 is expressed in the early embryonic tracheal system
and has been proposed to play important roles in cell migration as well as in the regulation of
the actin cytoskeleton. We have identified and characterized rack1 mutants; these mutants are
zygotic lethal but neither show a detectable embryonic- nor any other larval phenotype, due to
a very high maternal contribution. Removing the maternal store by generating germline
clones results in eggs that fail to develop. This developmental arrest is due to an incomplete
transfer of maternal product into the oocyte (nurse cell dumping).
In order to characterize the function of rack1 in the context of FGF signaling, we started to
characterize the development of third instar larval air sacs.
It has been reported that this structure develops via cell migration as well as cell division in
response to FGF/Bnl signaling. First we confirm the occurrence of cell division and found
that in early air sacs, division is ubiquitous and becomes restricted later to the central part of
the air sac. We also documented cell behavior during cell migration using live imaging.
To initiate a genetic analysis of rack1 and other candidate target genes in tracheal cell
migration, strains and methods were established, allowing the generation of mosaic air sacs
consisting of marked wild-type or mutant cells in an otherwise heterozygous background
based on the MARCM system. This system was also applied to characterize cellular shape
and dynamics of individual or small groups of air sac tracheoblasts in different parts of the air
sac. We found that air sac tip cells extend long and dynamic actin based protrusions and
further demonstrated that cells not directly located at the tip do form similar protrusions.
Finally, we took advantage of the our knowledge of air sac architecture and development to
study the cell-autonomous requirement of candidate genes in genetic mosaics. We showed
that marked wild-type clones have a preference to be positioned at the tip. Mutants lacking btl
or dof, two genes required for embryonic tracheal cell migration, never populate regions at the
migratory front. We inferred that air sac tracheoblast cells lacking btl or dof are deficient in
migration and take this as a readout for measuring cell migration.
Having established criteria for measuring cell migration in air sacs, we tested rack1 mutants
for their involvement in air sac tracheoblast migration and find that this gene is not required
for this process. We also analyzed other candidate genes as well as components of the FGF
signaling pathway and found evidence that Ras plays a dual role during third instar air sac
formation. It appears to integrate signaling input from the EGFR pathway to trigger cell
division as well as input from the FGF pathway to activate a cell migratory response. In
contrast to border cells, mutants affecting the transcription factor Slow border cells (Slbo), the
VEGFR (PVR) or DE-Cadherin (Shg) do not impede air sac tracheoblast migration.
Components shown to regulate the actin cytoskeleton in response to PVR signaling such as
Myoblast city (Mbc) the Drosophila Dock180 homologue or the small Rho family GTPases
Rac1, Rac2 and Mig-2-like (Mtl) as well as the effector Chickadee, the Drosophila
homologue of Profilin, are essential for air sac tracheoblast migration. Thus, recruitment of
these actin cytoskeleton regulators and effectors is mediated via different ligands/receptors in
trachea and border cells.
Our studies demonstrate that the development of the air sac during late larval stages is a good
system to study guided cell migration and allows the genetic dissection of the FGF signaling
pathway.
The tools we developed allow to assay any candidate gene for which a mutant is available and
also laid the foundation for the isolation and characterization of genes in a genome wide EMS
screen.