Developing and Characterising Novel Drosophila Models of Invasive Colorectal Cancer

Colorectal cancer is the fourth most common cancer in the UK, with incidence continuing to increase amongst younger adults. It accounts for 10% of cancer mortality and the majority of these deaths follow metastasis.

We use the Drosophila midgut to model human colorectal cancer by inducing stem cells to form tumours by driving mutations in Apc and Ras (AR) commonly seen in human colorectal cancers. While we do not observe metastasis in this model, we found that overexpression of the EMT transcription factor Snail in Apc-Ras-Snail (ARS) flies increases tumour growth and invasion and permits secondary tumour formation. In this project, my aim was to screen other genes expected to increase tumour growth and invasion to look for novel metastatic models and to develop our understanding of our existing models.

As part of this screen, I knocked down ImpL2, the Drosophila homologue of a mammalian gene involved in insulin signalling in the AR and ARS models. Knock down of ImpL2 in AR flies led to increased tumour growth and invasion and further investigation showed that ImpL2 acts as a tumour suppressor in this model. Furthermore, I found that ImpL2 levels are reduced in ARS clones and my results suggested that these reduced ImpL2 levels contribute to the increased tumour growth seen in ARS flies. ImpL2 knockdown was not sufficient for metastasis formation, despite the increased tumour burden and invasion. This suggests that comparison between AR, ARS and AR ImpL2 knockdown flies may help understand how Snail overexpression causes metastasis.

Finally, I characterised ImpL2 expression in wild-type flies and cancer models. This showed activity in three signalling pathways, the WNT, ERK and Pi3K pathways, is required for ImpL2 expression. Genetic perturbations of each pathway suggested that ImpL2 is not blocking growth via Pi3K or insulin receptor signalling.