Investigating how the GATA transcription factor Serpent drives epithelial-mesenchymal transitions in development and disease

The posterior endoderm in Drosophila embryos undergoes an epithelial-mesenchymal transition (EMT), driving the collective migration of endoderm cells. This EMT shows many similarities with the so-called ‘partial-EMTs’ that have been described in facilitating cancer metastasis. A single GATA-binding transcription factor, serpent (srp), is responsible for driving the entire process of EMT in the posterior endoderm. srp was previously shown to directly repress transcription of the key apical-basal polarity regulator crumbs (Crb). However, many changes other than loss of apical-basal polarity also occur during EMT, therefore in my project I aimed to understand how srp activates these cell changes.

To explore which other genes srp may be regulating to drive EMT, I used Targeted DNA adenine methyltransferase identification (TaDa). This technique yielded a list of 1,177 genes directly bound by srp, of which 200 were related to mediating the response to ecdysone (20E), the major steroid hormone in Drosophila. To investigate this, I perturbed ecdysone signalling using two separate methods, which both prevented EMT in the posterior endoderm. This lack of EMT was attributed to the improper endocytic recycling of apical-basal polarity regulating proteins and adherens junction components. To test this theory further, I exposed embryos to exogenous sources of 20E, which recapitulated phenotypes observed during EMT at far earlier stages of embryogenesis than is normal. Together, these experiments strongly support a novel role for the co-regulation of srp and 20E in driving the precisely timed changes required for EMT.

The characterisation of developmental EMTs provides an understanding of the underlying molecular mechanisms of how EMTs normally occur during tissue morphogenesis. With the finding that activation of EMT in primary tumours can lead to cancer dissemination and metastasis, these findings may aid in the generation of improved therapeutics for metastatic cancer. To test for conserved mechanisms of developmental EMTs in disease using Drosophila, fly models first needed to be generated that accurately recapitulate aspects of human metastasis. I aided in the generation of a Drosophila model for colorectal cancer (CRC) metastasis, that recapitulates many hallmarks of human metastatic CRC. A pilot screen using this model identified 3 compounds with anti-EMT properties, capable of blocking the dissemination of circulating tumour cells. In the future, this model has the potential to be used to investigate and compare mechanisms of developmental EMTs with those involved in cancer metastasis, in particular if steroid hormones play a role in co-activating the process during disease progression.