English

Abstract

While the alternative transcription and splicing mechanisms have long been known for some genes like oncogenes, their prevalence in almost all multi-exon genes has been recently realized with the increasing application of high-throughput experimental methods, named Next Generation RNA Sequencing (NGS). Henceforth, understanding the regulation of these processes in comparisons between cell types and cancer requests, sensitive and specific bioinformatics as well as bio-statistic approaches, that is, Cufflinks/Cuffdiff, DEXseq and RESEM, detecting gene transcript/isoforms and exons abundance is necessary. Isoforms and exons expression analysis by NGS is complicated by several sources of measurement variability causing numerous statistical defies. Here, with the purpose of minimizing this statistical challenge, we integrated both Cufflinks/Cuffdiff and DEXseq bioinformatics approaches assessing whole alternative splicing (AS) events, focusing on alternative transcripts regulation and their exons modulation respectively, by processing our previous prepared Estrogen Receptor β (Erβ+ and Erβ-) breast cancer (BC) cells, stimulated by estradiol (E2). Results showed that Estradiol (E2) induced Erβ+ BC (Erβ+E2), exhibited dissimilar reply as opposed to the other’s analyzed BC cell lines in terms of intragenic, exons and junction reads count ratio. Relationship analysis between expressed genes and transcript isoforms, suggested a substantial role of alternative promoters in AS event occurrence in Erβ+ BC as opposed to Erβ- BC. Indeed, merging Cufflinks/Cuffdiff and DEXseq approaches, 79 multi-exon genes were detected as statistically differentially modulated (spliced) in Erβ+ hormone induced BC cell line, and around 38% of these spliced genes claimed to be induced by alternative promoters. The present survey discriminated between several cancer specific alternative splicing genes like LIFR a BC metastasis suppressor, PBX1 a pioneer factor defining aggressive Erβ- BC and PHLPP2 a tumor suppressor, as exhibiting significant exon modulation in early AS occurrence in hormone responsive Erβ+ BC exclusively. Although, our findings supported dissimilar reply comparing both Cuflinks/Cuffdiff and DEXseq approaches called AS events, it is noteworthy to underline their relative agreement, evaluating spliced genes functional annotation as well as their complementarity performing whole AS survey. This study therefore proposed the integration between Cufflinks/Cuffdiff and DEXseq tools as a reasonable complementary methodology assessing full AS pattern in hormone responsive Erβ BC cells. Key words: Cufflinks/Cuffdiff, DEXseq, RNA-Seq, alternative splicing (AS), exons, transcript isoforms, estrogen receptor β (ERβ), breast cancer (BC) cells.