Deep computational analysis of human cancer and non-cancer tissues details dysregulation of eIF4F components and their interactions in human cancers

Abstract

SUMMARYEukaryotic translation initiation complex (eIF4F) plays roles so diverse in human cancers as to complicate development of an overarching understanding of eIF4F’s functional and regulatory impacts across tumor types. Our analysis of large public data sets yielded several useful findings. EIF4G1 frequently gains gene copies and is overexpressed to achieve characteristic stoichiometries with EIF4E and EIF4A1 in cancers. Varied expressions among EIF4F components distinguish malignant from healthy tissues, regardless of tissue or cancer types. EIF4G1 expression in particular correlates with poor prognosis. Tumors dysregulate biological “house-keeping” pathways typically regulated by cap-dependent initiation in healthy tissues, yet strengthen regulation of cancer-specific pathways in cap-independent contexts. In lung adenocarcinoma, altered interactions among eIF4F subunits are mechanistically linked to eIF4G1 phosphorylation. Tumors may select between cap-dependent and -independent mechanisms, through eIF4G1’s adaptable interactions with eIF4F subunits. Collectively, these results are an important advance towards a general model of translation initiation in cancer.