Phospho-proteomics and functional genomics for the characterisation of resistance mechanisms to agents targeting the EGFR, PI3K and MAPK pathways and their application in guiding drug combinations in colorectal cancer

Abstract

KRAS mutant and RAS/BRAF wild type (RAS WT) colorectal cancer (CRC) cells respond differently to targeted agents, as exemplified by cetuximab, an anti-EGFR antibody, which is only effective in a proportion of patients with metastatic RAS WT CRC. I hypothesised that this may be due to differences in the early adaptive signalling changes following exposure to targeted anticancer therapies. Therefore, I compared the effects on signalling transduction, using phospho(p)-proteomics, between RAS WT and KRAS mutant CRC cells, following 1h exposure to agents targeting the EGFR, PI3K and MAPK pathways in 25 cell lines and cells derived from 13 CRC patients' serous effusions. In KRAS mutant cells increase in the phosphorylation levels of intracellular effectors of the MAPK and PI3K pathways was evident, whereas in RAS WT CRC cells increase in the p-receptor tyrosine kinases (RTK) predominated, following exposure to gefitinib and PI3K inhibitor (GDC0941). As cetuximab is only effective in a proportion of patients with metastatic RAS WT CRC patients, I compared the p-proteomic changes in RAS WT cetuximab sensitive and resistant CRC cells. In addition, I carried out a whole genome CRISPR screen, in order to identify novel mechanisms of resistance to anti-EGFR therapies in RAS WT CRC. Collectively, the p-proteomic and functional genomic data converged on findings that involve the PI3K and MAPK pathways. Subsequent concurrent exposure to cetuximab and GDC0941 in resistant RAS WT cells and patient derived organoids led to a significant increase in antiproliferative activity, when compared to each single agent. The CRISPR screen and subsequent validation work suggested loss of NF1, a negative regulator of RAS proteins, as a cause of anti-EGFR resistance in RAS WT CRC. These results suggest that the combination of functional genomic and p-proteomic approaches can provide powerful insights in drug resistance mechanisms to anti-cancer agents and can guide future therapeutic strategies.