Splicing factor ratio as an index of epithelial-mesenchymal transition and tumor aggressiveness in breast cancer

• Published in: Oncotarget 2017-01, Vol.8 (2), p.2423-2436
• Main Authors: Fici, Pietro, Gallerani, Giulia, Morel, Anne-Pierre, Mercatali, Laura, Ibrahim, Toni, Scarpi, Emanuela, Amadori, Dino, Puisieux, Alain, Rigaud, Michel, Fabbri, Francesco
• Format: Article
• Language: English
• Subjects: Adult; Aged; Aged, 80 and over; Alternative Splicing; Biomarkers, Tumor - genetics; Biomarkers, Tumor - metabolism; Breast Neoplasms - genetics; Breast Neoplasms - metabolism; Breast Neoplasms - pathology; Cell Line, Tumor; Epithelial-Mesenchymal Transition; Female; Gene Expression Regulation, Neoplastic; Humans; Middle Aged; Neoplasm Grading; Prognosis; Repressor Proteins - genetics; Repressor Proteins - metabolism; Research Paper; RNA Splicing Factors - genetics; RNA Splicing Factors - metabolism; RNA-Binding Proteins - genetics; RNA-Binding Proteins - metabolism
• ISSN: 1949-2553; 1949-2553
• DOI: 10.18632/oncotarget.13682

Epithelial-to-mesenchymal transition (EMT) has been shown to be associated with tumor progression and metastasis. During this process in breast cancer, a crucial role is played by alternative splicing systems. To identify a new early prognostic marker of metastasis, we evaluated EMT-related gene expression in breast cell lines, and in primary tumor tissue from 31 patients with early breast cancer, focusing our attention on EMT-related splicing factors ESRP1, ESRP2 and RBFOX2. Results showed that the expression patterns of these genes were indicative of the onset of EMT in in-vitro models, but not in tissue samples. However, the ratio between ESRP1 or ESRP2 and RBFOX2 significantly decreased during EMT and positively correlated with the EMT-specific phenotype in cell models, representing a promising prognostic markers. Low ESRP1/RBFOX2 ratio value was associated with a higher risk of metastasis (p < 0.005) in early breast cancer patients, regardless other clinical features. A cut-off of ratio of 1.067 was determined by ROC curve analysis (AUC 0.8375; 95% CI 0.6963-0.9787). Our study show evidence that a decrease in this ratio correlates with cancer progression. The results provide a rationale for using ESRP1/RBFOX2 ratio as a new prognostic biomarker for the early prediction of metastatic potential in breast cancer.