Establishing a transcription factors protein network for lymphangiogenesis: identification of binding domains, impact of pathological mutations and perturbations by small molecules

Abstract

As a major route of cancer metastasis is via the lymphatic vasculature, therapeutic manipulation of neo‐lymphangiogenesis is an important adjunct treatment for cancer. So far only anti‐VEGF signalling molecules have been developed to counteract tumour‐induced de novo vessel formation, yet transcription factors could provide a wealth of potential molecular targets.Classic reverse genetic approaches in mice have validated individual roles for Sox18, Prox1, Nr2f2 (COUP‐TFII) and Gata2 in lymphangiogenesis. Expression of SOX18, COUP‐TFII and GATA2 coincide only in discrete locations within the vasculature and result in the induction of lymphatic endothelial progenitor cell specification. However, no physical interaction network between those key players has been identified yet.Here we map the protein‐protein interactions between transcription factors important for lymphatic vascular development to identify key transcription factor interactions as novel potential therapeutic targets. The direct interactions between SOX18, COUP‐TFII, GATA2, MEF2C and components of the Notch pathway (RBPJ, HEY1) have been assessed in vitro using a proximity assay (AlphaScreen) coupled to cell‐free expression of the proteins. These data allowed to reconstitute the network of interactions involved in the control of lymphatic endothelial cell specification.Using truncations, we established a comprehensive molecular mapping of the dimerization and protein binding domains of SOX18. Further, we defined how those interactions are affected by mutations associated with lymphatic diseases, such as Hypotrichosis‐Lymphedema‐Telangiectasia.We then used this network as a tool for drug discovery. Perturbations created by small molecules that impact lymphangiogenesis in vivo were measured on the whole network. Inhibition of direct interactions was measured and long‐scale modifications of the network could be inferred. This molecular dissection then provides new information to understand how the physiological effects of drugs are mediated.