at a level of detail and rapidity that far surpasses current mammalian
models.
Two of the most common initiating events in human CRC is
activation of WNT and RAS signaling pathways [3]. This can be
modeled in intestinal clones in adult Drosophila by combining
mutations in the negative regulators of the Wnt pathway, Apc and
Apc2, with overexpression of the oncogenic form of Ras, UAS-Ras
[V12] using the midgut stem cell driver, Escargot (Esg)-Gal4—
generating the so-called “ApcRas” model [4]. This genetic manipulation leads to the formation of tumor-like overgrowths with
many similarities to human CRC tumors, including increased proliferation, a block in cell differentiation and cell polarity, and disrupted organ architecture [5]. However, there are no secondary
tumors found in these flies, with ApcRas clones confined to the gut
and surrounded by a thick layer of basement membrane. Therefore,
these tumors can be considered benign CRC tumors.
Key steps of metastasis include invasion/dissemination of cells
from primary tumors, migration through the body, colonization of
distant sites, and growth of secondary metastases. Research on the
metastatic process has been hindered by a lack of genetically tractable experimental models reproducing these key steps, which are also
amenable to high-throughput analyses. Recently, when investigating the requirement of an epithelial-to-mesenchymal transition
(EMT) for metastatic dissemination, we found that over-expression
of the EMT master regulator Snail (Sna) in ApcRas intestinal
tumors induces tumor metastasis in adult Drosophila [6]. In the
“ApcRasSna” model, Sna drives a partial-EMT in intestinal tumor
cells, which, although they retain some epithelial markers, subsequently break through the basal lamina of the midgut, undergo a
collective migration, and seed polyclonal metastases. In this model,
secondary growths can be found in multiple distant locations outside the gut, including the abdomen, thorax, and head [6]. Tumor
cells generated in this metastatic CRC model are labeled with both
GFP and luciferase reporters. Here we report a detailed protocol
for firstly generating the ApcRasSna model, and then for assaying
for each step of the metastatic process, including notes on how it
can be made high-throughput. We expect that this can be exploited
to examine the effects of many different risk factors such as diet,
alcohol, toxins, and social behavior on tumor metastasis, as well as
to perform rapid large-scale genetic and drug screens.
2 Materials
2.1 Flies
Lines for making controls (clones with just the reporters: GFP and
luciferase)
l
yw hsp-flipase; esg Gal4, UAS-GFP/CyO; UAS-Luciferase
FRT82B Gal80/ TM6b.
162
Jamie Adams et al.
Précédent

- 169/425

Suivant