Targeting Vasculogenic Mimicry in Cancer
Like any tissue, tumours depend on a constant blood supply to deliver oxygen and nutrients. Tumour neovascularization is achieved most frequently through the process of angiogenesis, in which tumour cells secrete extracellular factors to recruit host endothelium to grow into the neoplastic tissue. A tumour’s dependency on a blood supply made angiogenesis an attractive target for cancer therapy, but disappointingly, anti-angiogenic drugs often times show limited efficacy in the context of cancer, with benefits only being transient followed by relapse and resistance. In retrospect, this is not surprising because it is now well-appreciated that some aggressive cancers are able to supply themselves with blood via an alternate mechanism to angiogenesis known as vascular mimicry (VM). In VM capable tumours, pseudo blood vessels are formed by tumour cells that have acquired endothelial-like properties independently of host endothelium. VM is a marker of poor prognosis and metastasis in patients, however it is still poorly understood at the molecular level. Recent advances have empirically demonstrated that VM drives resistance to anti-angiogenic therapy (AAT), which informs a motivation to identify VM-related targets that can be exploited therapeutically in combination with existing AATs as a rational approach to target both routes of tumour neovascularization to effectively starve tumours of the blood supply needed for survival. To this end, this thesis aims to identify drivers of VM across a range of cell types, with the goal of uncovering common mechanisms that VM- capable cancer cells use to maintain their pseudo endothelial phenotype, to identify nodes that can be exploited therapeutically to improve patient response to AAT. To achieve this, first we explored what triggers VM formation, through the manipulation of oxygen levels and a previously identified master regulator of VM, FOXC2. Next, we established a high-throughput method to sort cells for VM propensity using acetylated LDL uptake and leveraged this to conduct a genome-wide CRISPR screen using a bespoke dual gRNA-tRNA expression system. Through these approaches, we identified a shortlist of candidate genes to functionally investigate for VM capabilities to validate already known and uncover previously unknown VM-related targets, which can be leveraged as a promising therapeutic route towards sensitizing tumours to AATs.