The role of cancer-associated fibroblasts in altering signal transduction and drug resistance in KRAS mutated cancers

Abstract

KRAS is commonly mutated in solid tumours but effective treatment is limited due to resistance. Cancer-associated fibroblasts (CAFs) are a key tumour microenvironment component but how it interacts with KRAS mutant-cancer cells from different tissue types is not well explored, particularly in the context of drug resistance. My project investigates whether CAF secreted proteins alter signal transduction and drug response in KRAS mutant-cancers from colorectal, lung and pancreatic tissues. Proteins from lysates and media harvested from cell culture, termed conditional media (CM), were quantified using mass spectrometry. The differential proteome and secretome profiles of cancer cells and CAFs were characterised where novel markers in the proteome, including heat shock protein b6 in CAFs, were validated in tumour specimens. CAF-enriched secreted proteins were also identified, including WNT5B. A bespoke panel of 97 anti-cancer drugs revealed variations in the differential drug response across the cancer cell lines, which corresponds to the varying effect of CAF CM on the (phospho)proteome at the basal state level and highlights the complexity in CAF interactions beyond KRAS mutational status. CAF-mediated erdafitnib (fibroblast growth factor receptor (FGFR) inhibitor) sensitivity and resistance to 6-mercaptopurine and methotrexate were consistently observed in lung cancer H1792 where the response was further assessed using (phospho)proteomics. Greater downregulation of cell cycle and FGFR downstream pathways were associated with erdafitinib sensitivity, which could be due to upregulated focal adhesion kinase activity by CAF secreted proteins at the basal state level. Changes in mitogenic and signal transducer and activator of transcription 3 (STAT3) signalling are associated with 6-mercaptopurine and methotrexate resistance, respectively, which could be exploited as a therapeutic strategy. Overall, this data-driven project using a wide drug screen panel and proteomic platforms provided insights into differences between the two cell types and how CAFs could influence drug response, which are of potential clinical relevance.