A comparative survey of functional footprints of EGFR pathway mutations in human cancers

Abstract

Genes functioning in epidermal growth factor receptor (EGFR) signaling pathways are among the most frequently activated oncogenes in human cancers. We have conducted a comparative analysis of functional footprints (that is, effect on signaling and transcriptional landscapes in cells) associated with oncogenic and tumor suppressor mutations in EGFR pathway genes in human cancers. We have found that mutations in the EGFR pathway differentially have an impact on signaling and metabolic pathways in cancer cells in a mutation- and tissue-selective manner. For example, although signaling and metabolic profiles of breast tumors with PIK3CA or AKT1 mutations are, as expected, highly similar, they display markedly different, sometimes even opposite, profiles to those with ERBB2 or EGFR amplifications. On the other hand, although low-grade gliomas and glioblastomas, both brain cancers, driven by EGFR amplifications are highly functionally similar, their functional footprints are significantly different from lung and breast tumors driven by EGFR or ERBB2. Overall, these observations argue that, contrary to expectations, the mechanisms of tumorigenicity associated with mutations in different genes along the same pathway, or in the same gene across different tissues, may be highly different. We present evidence that oncogenic functional footprints in cancer cell lines have significantly diverged from those in tumor tissues, which potentially explains the discrepancy of our findings with the current knowledge. Nevertheless, our analyses reveal a common inflammatory response signature in EGFR-driven human cancers of different tissue origins. Our results may have implications in the design of therapeutic strategies in cancers driven by these oncogenes.