Individual and combined effects of DNA methylation and copy number alterations on miRNA expression in breast tumors

Abstract

Background: The global effect of copy number and epigenetic alterations on miRNA expression in cancer is poorly understood. In the present study, we integrate genome-wide DNA methylation, copy number and miRNA expression and identify genetic mechanisms underlying miRNA dysregulation in breast cancer.

Results: We identify 70 miRNAs whose expression was associated with alterations in copy number or methylation, or both. Among these, five miRNA families are represented. Interestingly, the members of these families are encoded on different chromosomes and are complementarily altered by gain or hypomethylation across the patients. In an independent breast cancer cohort of 123 patients, 41 of the 70 miRNAs were confirmed with respect to aberration pattern and association to expression. In vitro functional experiments were performed in breast cancer cell lines with miRNA mimics to evaluate the phenotype of the replicated miRNAs. let-7e-3p, which in tumors is found associated with hypermethylation, is shown to induce apoptosis and reduce cell viability, and low let-7e-3p expression is associated with poorer prognosis. The overexpression of three other miRNAs associated with copy number gain, miR-21-3p, miR-148b-3p and miR-151a-5p, increases proliferation of breast cancer cell lines. In addition, miR-151a-5p enhances the levels of phosphorylated AKT protein.

Conclusions: Our data provide novel evidence of the mechanisms behind miRNA dysregulation in breast cancer. The study contributes to the understanding of how methylation and copy number alterations influence miRNA expression, emphasizing miRNA functionality through redundant encoding, and suggests novel miRNAs important in breast cancer.

Figure 1

Genomic localization of the 70 in-cis miRNAs. The 70 in-cis miRNAs were identified using Wilcoxon rank-sum tests to identify differential miRNA expression between patients in different aberration groups. miRNAs are color coded according to main aberration type of the miRNA locus (see legend). The number in parentheses under aberration type represents the number of in-cis miRNAs in the given aberration category. Inter-chromosomal and intra-chromosomal lines (the latter is seen as independent lines on chromosomes 8 and 13) link in-cis miRNAs to other members of the same miRNA family (see legend). miRNA family annotation is taken from TargetScan (release 6.2) [2,33]. The plot was made using the Circos software package [34].

Figure 2

Examples of in-cis miRNAs. Six different miRNAs are depicted whose expression was associated with a certain aberration type in the breast cancer discovery cohort. The boxplots show the miRNA expression among aberration groups and the number in parentheses indicate the number of patients in a given aberration group. Hypomet, hypomethylation; hypermet, hypermethylation.

Figure 3

In-cis miRNAs show functional effects when overexpressed. Breast cancer cell lines were transfected with miRNA mimics (20 nM) and assayed for (a) cell proliferation (Ki67), (b) phosphorylated AKT (p-AKT) levels, (c) cell viability and (d) apoptosis (cleaved PARP (cPARP)), 72 hours after transfection. The dashed lines indicate cut-off points that were considered significant (see Materials and methods). Asterisks denote significant effects. The data for cell viability are from two replicate experiments with error bars showing standard deviations.

Figure 4

miRNA expression in luminal and basal-like patients. miRNA expression when patients were divided into luminal and basal-like subgroups in the discovery and replication cohort. (a) miR-148b-3p expression, (b) let-7e-3p expression. The P-values are from Wilcoxon rank-sum tests.

Figure 5

Kaplan-Meier curves showing relapse-free survival when dividing samples into high and low let-7e-3p expression groups. The ‘low’ group denotes samples with expression below the median (n = 43) and the ‘high’ group contains samples with expression above the median (n = 43). The P-value is from a log-rank test. The plot shows survival data of the discovery cohort.