NRF2 is a redox-responsive transcription factor that regulates expression of cytoprotective genes via its interaction with DNA sequences known as antioxidant response elements (AREs). NRF2 activity is induced by oxidative stress, but oxidative stress is not the only context in which NRF2 can be activated. Mutations that disrupt the interaction between NRF2 and KEAP1, an inhibitor of NRF2, lead to NRF2 hyperactivation and promote oncogenesis. The mechanisms underlying NRF2's oncogenic properties remain unclear, but likely involve aberrant expression of select NRF2 target genes. We tested this possibility using an integrative genomics approach to get a precise view of the direct NRF2 target genes dysregulated in tumors with NRF2 hyperactivating mutations. This approach revealed a core set of 32 direct NRF2 targets that are consistently upregulated in NRF2 hyperactivated tumors. This set of NRF2 “cancer target genes” includes canonical redox-related NRF2 targets, as well as target genes that have not been previously linked to NRF2 activation. Importantly, NRF2-driven upregulation of this gene set is largely independent of the organ system where the tumor developed. One key distinguishing feature of these NRF2 cancer target genes is that they are regulated by high affinity AREs that fall within genomic regions possessing a ubiquitously permissive chromatin signature. This implies that these NRF2 cancer target genes are responsive to oncogenic NRF2 in most tissues because they lack the regulatory constraints that restrict expression of most other NRF2 target genes. This NRF2 cancer target gene set also serves as a reliable proxy for NRF2 activity, and high NRF2 activity is associated with significant decreases in survival in multiple cancer types. Overall, the pervasive upregulation of these NRF2 cancer targets across multiple cancers, and their association with negative outcomes, suggests that these will be central to dissecting the functional implications of NRF2 hyperactivation in several cancer contexts.
Bibliographical noteFunding Information:
The authors wish to acknowledge Sarah Lacher for helpful discussions. We are also indebted to the researchers who worked to generate and process the data cited. This work was supported by funding from the National Institute of General Medical Sciences, National Institutes of Health, USA ( R35-GM-119553 to M.S.) and the Intramural Research Program of the National Institute of Environmental Health Sciences, National Institutes of Health , USA [ Z01-ES-100475 and Z01-ES-46008 (D.A.B)].