The Drosophila model system is increasingly being used to uncover new areas of investigation for cancer researchers, and a recent report in The EMBO Journal proves no exception. Anthony Brumby and Helena Richardson show that the loss of tumour suppressor scribble, which is involved in cell polarity, induces an overgrowth phenotype, and suggest that loss of such proteins could also contribute to tumour formation in humans.

Loss of scrib in the larval eye imaginal disc — using FLP/FRT-mediated recombination — resulted in cells losing their columnar shape and monolayered morphology, to become more rounded and multilayered. They also lost differentiation markers, and expressed increased levels of cyclin E, which induces proliferation. This correlated with increased labelling with the S-phase marker bromodeoxyuridine, to confirm the increased proliferation.

However, as larval development progresses, the scrib tissue is lost through apoptosis. This seems to be mediated through the c-Jun amino-terminal kinase (Jnk) stress response, as inhibiting Jnk activity using a dominant-negative mutant prevented this apoptosis and increased the size of the scrib clones. Interestingly, this response was not a cell-intrinsic response — removing the surrounding wild-type tissue resulted in massive overproliferation, presumably because apoptosis was prevented.

The authors next investigated whether other oncogenic mutations could overcome the suppressive effect of Jnk pathway activation on the scrib phenotype. Whereas activated forms of Ras and Notch cooperated with scrib to induce overgrowth — the mutant tissue proliferated in three dimensions and did not differentiate — activation of the Wingless, Hedgehog and Decapentaplegic pathways had somewhat different effects. Although they showed some differentiation defects and the scrib clones were generally increased in size, they did not have the same overgrowth phenotype as that observed with activated Ras and Notch.

So, which Ras effector was important for the overgrowth response? Neither overexpression of PI3K, nor its effectors, could mimic the effect of activated Ras, but an activated Raf allele could induce overproliferation of scrib clones. The mitogen-activated protein kinase (MAPK) pathway therefore mediates this cooperation, but it, in turn, can influence processes such as survival, growth, proliferation and differentiation. Surprisingly, recapitulating proliferation by expressing cyclin E, and apoptosis inhibition by expressing the caspase inhibitor p35, could still not mimic the overgrowth phenotype that is caused by activated Ras. Other effectors of Ras could therefore also be important.

Although scrib homologues have not yet been shown to be tumour suppressors in humans, this research does indicate that proteins that dictate cell polarity could be important in tumorigenesis, and should be investigated further.