Genetic causes of resistance to entinostat in luminal breast cancer model systems.

Abstract

1065 Background: Based on promising phase II data, the histone deacetylase inhibitor Entinostat (ENT) is in phase III trials for patients with metastatic ER-positive breast cancer. Predictors of sensitivity and resistance, however, remain unknown. Methods: Luminal cell lines SKBR3 (ER-/HER2+), BT474 (ER+/HER2+) and MCF7 (ER+/HER2-) were treated with or without ENT at their IC50 doses and their gene expression profiles determined. In addition, a total of 27 MMTV-Neu mouse tumors (luminal) were untreated (N = 8), or treated with ENT at 12 mg/kg for 3 weeks (N = 5), 6 weeks (N = 6), or until progression after complete response (N = 8). We investigated their gene expression profiles by microarray and copy number (CN) by arrayCGH, and utilized the Dawnrank analysis, a network-based bioinformatics tool that integrates DNA and RNA data to identify driver genes, to find predictors of resistance to ENT. Results: Supervised analysis of gene expression data coming from the 3 treated cell lines showed significant upregulation of multiple MYC gene signatures. Therefore, we constitutively overexpressed MYC using lentiviral MYC shRNA in SKBR3 and MCF7, and MYC overexpression made cell lines more resistant to ENT. In MMTV-Neu mice, both MYC gene mRNA and gene signatures were downregulated while cells responded to ENT, and became upregulated when the tumors progressed. aCGH CN analysis revealed that a large portion of mouse Chromosome 4 had DNA CN loss and low gene expression in tumors that progressed while on ENT. Within this region, JUN was computationally identified to be a top driver gene associated with resistance. JUN was next knocked down using lentiviral JUN shRNA in BT474 and T47D, and JUN knock-down repeatedly made cell lines more resistant to ENT. MYC gene-expression was also upregulated in JUN-knockdown BT474 and T47D. Finally, JUN CN loss was found in 22% (132/588) of luminal tumors in The Genome Cancer Atlas breast cancer, and all the MYC signature scores were significantly higher in JUN-deleted TCGA samples. Conclusions: ENT was an effective drug for all of our luminal models, both in vitro and in vivo. Using these models, we selected for resistant variants and identified MYC signature expression, and JUN CN deletion as being associated with resistance.