Molecular Characterization of Endocrine Resistance.

Abstract

Abstract Background. Tamoxifen (TAM), anastrozole (ANA), letrozole (LET), and exemestane (EXE) are major drugs to treat estrogen (E)-dependent breast cancer. Although they are effective, disease recurrence often occurs as a result of acquired resistance to them. To unravel the mechanisms involved, MCF-7aro cell lines resistant to each of the 4 drugs or undergone long-term E deprivation (i.e., LTEDaro) were generated. LET-R, ANA-R, and LTEDaro cells contain a constitutively active ERa. TAM-R cells are thought to have hypersensitive ERa. Furthermore, EXE has weak E-like properties, leading to an ER-dependent crosstalk with EGFR signaling. Thus, our results indicate that resistance mechanisms of the 4 drugs are not identical, activation of ERa is critical for endocrine resistance, and our cell lines are valuable for studying the resistance mechanisms to these drugs. Results. Genome-wide gene expression microarray analysis of these lines was carried out. Hierarchical clustering identified many E-responsive genes with varied expression in our cell lines. At least 3 types of E-responsive genes were observed: up-regulated in all cell lines, up-regulated in all lines except LTEDaro, and up-regulated in all lines but LTEDaro and TAM-R. To better understand the roles of ERa, we initiated ChIP-sequencing experiments for a genome-wide analysis of ERa-binding sites in MCF-7aro and LTEDaro, and plan to correlate the ERa-binding data and the gene expression data. There was a significant overlap of ERa-binding sites identified in the two lines. Furthermore, a substantial number of ERa-binding sites were identified in LTEDaro without E. To study the post-transcriptional regulatory mechanisms, miRNA microarray profiling analysis of resistant lines was also initiated. Global miRNA expression was correlated with a similarity matrix analysis which revealed 4 miRNA clusters, with 2 having E-driven miRNAs. We identified expression of 9 miRNAs as down-regulated in AI resistant cells but up-regulated in TAM-R and LTEDaro, expression of 16 miRNAs as up-regulated in LET-R and ANA-R, and expression of 24 miRNAs as up-regulated in EXE-R. Functional analysis also revealed that miRNA-128a down-regulated the TGFb response in LET-R. Finally, to determine the ER-Signal Transduction cross-talk mechanisms, we initiated an unbiased screen of kinases that participate in the development of resistance to the 4 drugs. Our cell lines were screened against a focused dicer substrate siRNA library that targets 544 kinases. The kinomic approach revealed that hits from MCF-7aro and LTEDaro were different. Twenty hits were identified using LTEDaro, as indicated by more than 40% suppression of cell proliferation with 3 or 4 siRNAs/target. These hits included several kinases known for their roles in endocrine resistance as well as several novel kinases. Discussion. We anticipate that these bioinformatics studies will produce valuable molecular information regarding the mechanisms of endocrine resistance, and the information will help design approaches to reduce disease recurrence and improve the efficacy of endocrine treatments of breast cancer. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 407.