ABSTRACT: Purpose: Tamoxifen is the most commonly used antiestrogen drug for breast cancer treatment, however, acquired resistance is a considerable challenge in clinic. Complexity and redundancy in biological system implies tamoxifen resistance is associated with complex biological processes. The aim of this study is to attempt to reveal the mechanism of tamoxifen resistance using high-throughput sequencing technique. Methods: We applied digital gene expression (DGE), RNA-seq, Small RNA-seq, bisulifte-seq (reduced representation bisuflite sequencing (RRBS)) with Illumina sequencing platforms, and Array CGH to analyze two well-established tamoxifen resistant cell line models: one is the LCC cell line model that was derived from MCF-7 breast cancer cell line and developed tamoxifen resistance following a step-wise strategy. This cell line model includes LCC0 (estrogen-dependent and tamoxifen sensitive), LCC1 (estrogen-independent and tamoxifen sensitive), LCC2 (derived from LCC1 with increased small-dosage tamoxifen treatment, estrogen-independent, and tamoxifen-resistant), and LCC9 (derived from LCC1 with increased small-dosage fluvestrant treatment, estrogen-independent, and cross-resistance to both fluvestrant and tamoxifen). The other is the TAMR cell line model that was derived from MCF-7 breast cancer cell line and developed four tamoxifen-resistant sublines following the strategy of high-dosage of tamoxifen treatment. Each subline was developed from the survived single cell after the tamoxifen treatment. Results: The DGE and RNA-seq results showed characteristic expression features for the both cell line models, for instance, high expression of SOX2 and alterations of other SOX gene family members, decreased expression of PGR, PTEN, TP53, EZH2, SMAD3, TGF-β1, DNMT1, APOBEC3B, FOXO3 and FOXO4, and increased expression of CDH1, APOBEC3H, FOXA1, TET2, TET3, TDG, MBD4, DNMT3a and DNMT3b between the tamoxifen-resistant cell lines with their matched parental cell lines. Small RNA seq showed the decreased expressions of miR-1, miR-196-3p, miR-18a-5p, miR-125b-5p, and the increased expression of miR-23a-3p, miR-26b-5p, miR-320d, miR-141-3p, miR-23a-3p, miR-30a-5p, miR-9-5p, miR-200c, miR-129-5p, miR-125b-5p, miR-1285-3p and miR-30b-5p between the tamoxifen-resistant cell lines and their matched parental cell lines. Additionally, let-7a and let-7c have been found to present high expression levels in the tamoxifen resistant cells in the both cell line models, even it showed significantly decreased in some TAMR cells compared with their parental cell line. The RRBS showed similar DNA methylation levels between the LCC cell lines. Array CGH showed a small number of copy number variants between the LCC cell lines. Conclusion: Our present study suggests the development of tamoxifen resistance is associated with a dynamic process of cell fate changing. Therefore, new therapeutic method that targets whole cell system and cell fate dynamics, will be a promising strategy.
