Identification and Characterization of MicroRNAs Associated with Somatic Copy Number Alterations in Cancer

Abstract

MicroRNAs (miRNAs) are key molecules that regulate biological processes such as cell proliferation, differentiation, and apoptosis in cancer. Somatic copy number alterations (SCNAs) are common genetic mutations that play essential roles in cancer development. Here, we investigated the association between miRNAs and SCNAs in cancer. We collected 2538 tumor samples for seven cancer types from The Cancer Genome Atlas. We found that 32-84% of miRNAs are in SCNA regions, with the rate depending on the cancer type. In these regions, we identified 80 SCNA-miRNAs whose expression was mainly associated with SCNAs in at least one cancer type and showed that these SCNA-miRNAs are related to cancer by survival analysis and literature searching. We also identified 58 SCNA-miRNAs common in the seven cancer types (CC-SCNA-miRNAs) and showed that these CC-SCNA-miRNAs are more likely to be related with protein and gene expression than other miRNAs. Furthermore, we experimentally validated the oncogenic role of miR-589. In conclusion, our results suggest that SCNA-miRNAs significantly alter biological processes related to cancer development, confirming the importance of SCNAs in non-coding regions in cancer.

Keywords: Co-regulation; MicroRNA; cancer; copy number alteration.

Figure 1

Characteristics of miRNAs with copy number alterations. (A) Copy number distribution of miRNA gene regions. Red and blue dashed lines show copy number amplification or deletion thresholds, respectively. (B) On the x axis, the fraction of samples containing somatic copy number alterations (SCNAs) is represented. On the y axis, the fraction of miRNAs with SCNAs is presented when an miRNA was declared to be located in the SCNA region if the fraction of samples with SCNAs was larger than the given fraction value on the x axis. (C) The ratio of miRNAs with SCNAs in >10% of samples of each cancer type. (D) The ratio of miRNAs with SCNAs in >15% of samples of each cancer type. In (C,D), red and blue bars represent the percentage of miRNAs with copy number amplifications or deletions, respectively. These ratios were compared to those of SCNAs in randomly selected regions and all genomic regions. “*” indicates that the fraction of miRNAs in the SCNA regions was significantly larger than the fraction of random genomic regions with SCNAs for the KIRC.

Figure 2

Relationship between copy numbers and miRNA expression. Pearson correlation coefficients (PCCs) showing the correlation between copy numbers and miRNA expression levels. Yellow dashed line, PCC threshold.

Figure 3

SCNA-miRNA heatmap obtained for seven cancer types. MiRNAs are listed according to their chromosomal locations. The PCC value, copy number state, and miRNA rank with p-value for each miRNA are shown (*, p-value < 0.05). The brown color intensities represent the degree of PCCs between miRNA copy numbers and expression levels, and red or blue intensities represent the average copy number values of samples amplified or deleted for the given miRNA and cancer type. The last row shows cancer-specific SCNA-miRNA numbers.

Figure 4

Chromosome ideogram obtained by mapping SCNA-miRNAs and pan-cancer SCNA regions. Pan-cancer SCNA regions, amplified (red) or deleted (blue), are shown for each chromosome, and the locations of the SCNA-miRNAs are indicated with cytogenetic bands. The 26 miRNAs in the pan-cancer SCNA regions are represented in bold, and the remaining 54 miRNAs are shown together, which are located within 10.6 mega bases (Mbs) of pan-cancer SCNA regions on average (ranging from 0.3 to 47.4 Mbs).

Figure 5

Survival analysis according to SCNA-miRNAs. Heatmaps showing miRNAs significantly affecting survival. Red and blue, the differences in patient survival when miRNA copy numbers were amplified and deleted, respectively.

Figure 6

Effects of miR-589 on Hep-2 cells. The cells were transfected with miR-589, miR-589 inhibitor, and their negative controls. (A) Expression levels of miR-589 in the cells transfected with miR-589, miR-589 inhibitor, or their negative controls at 48 and 96 h. These are measured by RT-PCR, and 48 and 96 h were examined on different gels. Uncropped images of gels are shown in Figure S4. (B) Cell viability and proliferation rate. The transfected cells were measured by MTT assay at different time periods. The averages of three independent MTT assays performed in triplicate are shown with standard error (SE) bars. The p-values of miR-589 versus its normal control, miR-589 inhibitor versus its normal control, and miR-589 versus miR-589 inhibitor were calculated by using two-tailed t-test.

Figure 7

Boxplots showing the relationship between CC-SCNA-miRNAs and mRNA/protein expression. The mean absolute PCCs of CC-SCNA-miRNAs and genes were compared to the mean absolute PCCs obtained for the correlation between the other 913 miRNAs and gene expression. (A) Expression of all mRNAs, top 1% genes, target genes of miRNAs, top 5% target genes of miRNAs, and (B) all protein expression levels and top 20% proteins were considered.