Building a Novel Human Model to Deconvolve the Cancer-Associated Phenotypes of Genetically Dysregulated Pathways in Lung Squamous Cell Carcinoma

Abstract

Advancements in our understanding of the biology of cancer development and possible translational perspectives for clinical management rely on the availability of tractable, patient relevant models which recapitulate key cancer traits. Lung squamous cell carcinoma (LUSC) is a devastating disease which arises from the bronchial epithelium, is often diagnosed late and has very limited therapeutic options. LUSC molecular landscapes show high inter-patient heterogeneity, but some commonalities. These include universal loss of TP53 and CDKN2A tumour suppressors, and common activation of three pivotal pathways: the KEAP1/NRF2 pathway, PI3K signalling and squamous differentiation. These alterations occur early in LUSC development but their requirement for bronchial epithelial cell invasive transformation is unclear. In this thesis, I have generated a novel genetically engineered in vitro model of LUSC using primary human bronchial epithelial cells (hBECs) derived from three donors. In doing so, I have investigated the role of pathway dysregulation in driving phenotypes indicative of LUSC development in the bronchial epithelium. Extensive in vitro analyses confirmed that activation of the squamous differentiation pathway via SOX2 overexpression in hBECs with universal LUSC alterations (TP53 and CDKN2A loss) defines squamous cell fate and initiates early preinvasive LUSC stages. Additionally, I show that SOX2 co-operates with the oxidative stress response (KEAP1-NRF2 pathway) and PI3K signalling to drive more advanced phenotypes, including the expansion of hBECs expressing LUSC molecular markers, epithelial cell invasion, and the formation of tumours resembling LUSC. Importantly, interrogation of gene expression changes associated with each pathway revealed changes in cell-intrinsic processes (e.g., RTK signalling and keratinisation) and immunoregulation (e.g., interferon signalling and antigen presentation). Lastly, I reveal a novel role for SOX2 overexpression in repressing epithelial cell MHC-II expression with potential consequence for tumour/epithelial cell immunogenicity. Overall, I have concluded that this study has succeeded in generating a human model of the classical LUSC molecular subtype. These findings take significant steps towards the building of patient relevant models of LUSC, add to our understanding of the role of pathway dysfunction in driving LUSC development, and the identification of novel translational perspectives which may contribute towards improved patient outcomes.

Date of Award	17 May 2024
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Carlos Lopez Garcia (Supervisor) & Caroline Dive (Supervisor)