Abstract 2633: Gene signatures derived from in vitro isolated migratory and invasive breast cancer cells indicate early metastatic relapse.

Abstract

Abstract Background Cancer cell motility and tissue invasion are cornerstones of metastatic dissemination. Adequate molecular profiling of chemotaxis-driven tumor cells is impeded by the transient and stochastic nature of motility occurring in only a small subset of tumor cells. To unveil molecular pathways implicated in breast cancer cell invasion and migration, an in vitro set-up was designed to extract total RNA for microarray hybridisation from isolated migratory and invasive breast cancer cells at an early and a late time point. These target populations were compared to corresponding non-migratory and non-invasive cells for migration- and invasion-related expression profiling to assess potential prognostic relevance. Methods and results A 24-well Transwell set-up was used to perform both migration and invasion experiments on MDA-MB-231 breast cancer cells after 24h serum starvation (400000 cells/insert). A Matrigel coating (20%/15μL) was applied to the bottom-side of insert membranes in both settings. For invasion experiments, a layer of Matrigel (20%/20μL) was added on top of the membranes. Total RNA was isolated at early and late timepoints from both migratory / invasive and reference cells and hybridized onto Illumina HumanHT-12 v4 BeadChips in biological triplicates. Normalization and differential gene expression analysis were performed using the Limma-package in BioConductor. Lists of differentially expressed genes were analyzed using Ingenuity software. Using the nearest shrunken centroid algorithm gene signatures characteristic for early or late invasive or migratory cells were generated. The resulting models were applied onto six distinct gene expression data sets of pre-treatment samples from patients with mostly lymph node negative breast cancer. Survival analysis was performed using distant metastasis-free or disease-free survival as endpoint. Although all signatures were associated with survival in at least one data set, the gene signature associated with late invasion was associated with survival (range HR: 1.403 - 1.929; all P<0.05) in all but two data sets. Of note, gene signatures associated with the late timepoints were prognostically more relevant (significance obtained for 66.7% when using the models for the late timepoint vs. 16.7% and 33.3% when using the models for the early timepoint). Conclusions Gene expression characteristics of invasive and migratory breast cancer cells indicate poor prognosis on several publicly available breast cancer expression data sets. For both migration and invasion, the strongest prognostic relevance was associated with the late time points, possibly by extended differences caused by early events. The role of identified target genes is being functionally validated and additionally, these data can be subjected to multivariate regression analysis to study association with the molecular subtypes. Citation Format: Ridha Limame, Ken Op de Beeck, Steven Van Laere, Guy Van Camp, Marc Peeters, Luc Dirix, Filip Lardon, Olivier De Wever, Patrick Pauwels. Gene signatures derived from in vitro isolated migratory and invasive breast cancer cells indicate early metastatic relapse. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 2633. doi:10.1158/1538-7445.AM2013-2633