In-depth study of breast cancer tumor promoting cell transcriptome using RNA-seq and microarrays

Background - Several evidences support the presence of a subpopulation of tumorpromoting cells (TPCs) in many tumors, including breast cancer (BC), which might be responsible for treatment failure due to their resistance to treatment and the inability to specifically target the TPCs subpopulation. RNA-seq allows an in-depth transcriptome analysis under the quantitative and qualitative (isoforms, ncRNAs) aspects. Materials and Methods - Expression profiles of highly tumorigenic mammospheres (MCFS) and of parental MCF7 were obtained using Illumina microarrays and SOLiD RNA-seq after linear isothermal DNA amplification. Different approaches for RNA-seq analysis were used and results compared. Deregulated pathways were identified by gene set enrichment analysis and alternative splicing events were predicted using AltAnalyze. Results - RNA-seq showed higher dynamic range than microarrays. Partial agreement between different software for the analysis of RNA-seq data was obtained. More than 150 gene sets were enriched at FDR<1% using either microarray or RNA-seq data. Despite the partial overlap in the enriched gene sets, microarrays and RNA-seq data led to the identification of the same deregulated pathways indicating that RNA-seq from 5000 cells provides reliable information despite RNA amplification. MCFS showed an upregulation of NK-kB related pathways and cytokines, like IL8 and MCP-1, a downregulation of estrogen receptor-related genes, a lower response to estradiol and lower sensitivity to hormone therapy. Moreover, several gene sets associated with epigenetic control, cholesterol metabolism and progenitor phenotype were identified. RNAseq data allowed the identification of more than 100 differentially expressed ncRNAs and more than 2000 splicing events. The most interesting findings were validated by independent techniques. Conclusions - A marked transcriptional reprogramming occurs in breast cancer TPCs involving the expression of both coding and non coding genes and alternative splicing. These findings could contribute to the understanding of the mechanism of resistance to hormonal therapy and suggest possible TPCs-specific therapeutic targets.