Exon-Level Microarray Analyses Identify Alternative Splicing Programs in Breast Cancer

Anna Lapuk(Lawrence Berkeley National Laboratory), Henry Marr(Lawrence Berkeley National Laboratory), Lakshmi R. Jakkula(Lawrence Berkeley National Laboratory), Helder Pedro(Lawrence Berkeley National Laboratory), Sanchita Bhattacharya(Lawrence Berkeley National Laboratory), Elizabeth Purdom(Lawrence Berkeley National Laboratory), Zhi Hu(Lawrence Berkeley National Laboratory), Ken Simpson(Lawrence Berkeley National Laboratory), Lior Pachter(Lawrence Berkeley National Laboratory), Steffen Durinck(Lawrence Berkeley National Laboratory), Nicholas Wang(Lawrence Berkeley National Laboratory), Bahram Parvin(Lawrence Berkeley National Laboratory), Gerald Fontenay(Lawrence Berkeley National Laboratory), Terence P. Speed(Lawrence Berkeley National Laboratory), James C. Garbe(Lawrence Berkeley National Laboratory), Martha R. Stampfer(Lawrence Berkeley National Laboratory), Hovig Bayandorian(Lawrence Berkeley National Laboratory), Shannon Dorton(Lawrence Berkeley National Laboratory), Tyson A. Clark(Lawrence Berkeley National Laboratory), Anthony Schweitzer(Lawrence Berkeley National Laboratory), Andrew J. Wyrobek(Lawrence Berkeley National Laboratory), Heidi S. Feiler(Lawrence Berkeley National Laboratory), Paul T. Spellman(Lawrence Berkeley National Laboratory), John G. Conboy(Lawrence Berkeley National Laboratory), Joe W. Gray(Lawrence Berkeley National Laboratory)
Molecular Cancer Research
July 6, 2010
10.1158/1541-7786.mcr-09-0528
Cited by 116Open Access
Full Text

Abstract

Protein isoforms produced by alternative splicing (AS) of many genes have been implicated in several aspects of cancer genesis and progression. These observations motivated a genome-wide assessment of AS in breast cancer. We accomplished this by measuring exon level expression in 31 breast cancer and nonmalignant immortalized cell lines representing luminal, basal, and claudin-low breast cancer subtypes using Affymetrix Human Junction Arrays. We analyzed these data using a computational pipeline specifically designed to detect AS with a low false-positive rate. This identified 181 splice events representing 156 genes as candidates for AS. Reverse transcription-PCR validation of a subset of predicted AS events confirmed 90%. Approximately half of the AS events were associated with basal, luminal, or claudin-low breast cancer subtypes. Exons involved in claudin-low subtype-specific AS were significantly associated with the presence of evolutionarily conserved binding motifs for the tissue-specific Fox2 splicing factor. Small interfering RNA knockdown of Fox2 confirmed the involvement of this splicing factor in subtype-specific AS. The subtype-specific AS detected in this study likely reflects the splicing pattern in the breast cancer progenitor cells in which the tumor arose and suggests the utility of assays for Fox-mediated AS in cancer subtype definition and early detection. These data also suggest the possibility of reducing the toxicity of protein-targeted breast cancer treatments by targeting protein isoforms that are not present in limiting normal tissues.

Related Papers

No related papers found

Powered by citation graph analysis