Differential squamous cell fates elicited by NRF2 gain-of-function versus KEAP1 loss-of-function

Clinical evidence has revealed that high-level activation of NRF2 caused by somatic mutations in NRF2 is frequently detected in esophageal squamous cell carcinoma (ESCC), whereas that by somatic mutations in KEAP1, a negative regulator of NRF2, is not. Here, we challenged to generate a mouse model of NRF2-activated ESCC using the cancer-derived NRF2L30F mutation and cancer-driver mutant Trp53R172H. Concomitant expression of NRF2L30F and Trp53R172H induced proliferation of squamous cell epithelia and resulted in NRF2-activated ESCC-like lesions. In contrast, while squamous cell-specific deletion of KEAP1 induced similar NRF2 hyper-activation, the loss-of-KEAP1 combined with Trp53R172H did not elicit the proliferation and formation of ESCC-like lesions. Instead, KEAP1-deleted cells disappeared from the esophageal epithelium over time by cell competition. These findings provide insights into the observation that somatic mutations are more frequently observed in NRF2 than in KEAP1, and the mouse model developed here will be instrumental in elucidating the mechanistic basis leading to NRF2-activated ESCC. Overall design: To examine how mutations in NRF2 gene affect esophageal squamous cell carcinogenesis, we generated a knock-in mouse line with human cancer-derived NRF2L30F, in which leucine (L) at codon 30 in DLGex motif was replaced with phenylalanine (F).To iduce NRF2L30F-expression by Cre-mediated recombination in squamous cells, we exploited K5Cre mice, and K5CreERT2 mice which harbor tamoxifen-inducible squamous cell specific CreERT2 recombinase system. We generated NRF2L30/FL30F::K5Cre, NRF2L30F/L30F::K5CreERT2, and NRF2L30F/+::K5CreERT2 mice.To compare the phenotype of squamous cell specific NRF2L30F-expressing mice with that of conditionally KEAP1-deficiency mice, we also generated KEAP1flox/flox::K5Cre and KEAP1flox/flox::K5CreERT2 mice. We performed RNA-sequence analysis using RNA samples from the esophagus. In the first experiment, we utilized RNA samples from esophagus of Wild-type, NRF2L30F/L30F::K5Cre, and KEAP1flox/flox::K5Cre mice, in age of 7-10 day-old. Second, we utilized RNA samples of esophagus derived from wild-type, NRF2L30F/+::K5CreERT2, NRF2L30F/L30F::K5CreERT2, and KEAP1flox/flox::K5CreERT2 mice at 1 week after tamoxifen treatment (100 µg/g body weight, tyree constitutive days), in age of 6-9 week-old.