Colorectal carcinoma (CRC) is the third most commonly diagnosed and the fourth most deathly cancer, leading to 8% of all cancer deaths globally [
Recently, miRNA, a group of noncoding, small RNA molecules, was found to be a powerful biomarker with bright prospect in cancer applications. MicroRNA participates in several pathophysiological processes and pathways, and its level is related to cancer burden and survival of cancer patients [
The Cancer Genome Atlas project is aimed at profiling genomic changes in more than 33 different cancer types including CRC [
The sequencing based mRNA and miRNA expression data of 255 CRC patients were obtained from TCGA project [
The pairwise Pearson correlation coefficient between mRNA and miRNA expression was calculated based on 17,300 mRNAs and 612 miRNAs, yielding a correlation coefficient data matrix with 17,300 mRNAs in row and 612 miRNAs in columns. A Pearson correlation coefficient less than −0.5 was used as a cutoff to obtain the most probable biologically relevant miRNA-mRNA regulations. Data preprocessing and Pearson correlation coefficient calculation were performed in the R (
Gene coexpression networks were built according to the normalized gene expression values. We constructed the network adjacency between two genes,
Coexpressed genes identified by correlation analysis were used to query the KEGG pathway database [
The TCGA colon cancer dataset included both mRNA and miRNA gene expression profiles of 255 tumor samples. Given the regulatory relationship between miRNA and mRNA, we assumed that the correlation between miRNA and the expression of its target genes was negative. By using a cutoff of negative Pearson correlation coefficient less than −0.5, 13 mRNA and miRNA association pairs were identified in colon cancer (Table
miRNA and mRNA correlation analysis.
miRNA | Gene Symbol∣Entrez ID | Correlation coefficient | |
---|---|---|---|
1 | hsa-mir-200c | DCN∣1634 | −0.54 |
2 | hsa-mir-15a | IGF2∣3481 | −0.51 |
3 | hsa-mir-15a | INS-IGF2∣723961 | −0.51 |
4 | hsa-mir-200c | IGFBP7∣3490 | −0.51 |
5 | hsa-mir-200c | PDLIM3∣27295 | −0.51 |
6 | hsa-mir-200a | HSPB7∣27129 | −0.51 |
7 | hsa-mir-200c | SERPINF1∣5176 | −0.50 |
8 | hsa-mir-200a | SFRP4∣6424 | −0.50 |
9 | hsa-mir-200a | MAB21L2∣10586 | −0.50 |
10 | hsa-mir-552 | TRIB2∣28951 | −0.50 |
11 | hsa-mir-625 | C6orf15∣29113 | −0.50 |
12 | hsa-mir-200a | THBS4∣7060 | −0.50 |
13 | hsa-mir-16-1 | IGF2∣3481 | −0.50 |
Scatter plot for miRNA and mRNA gene expression profiles. Cor indicates the Pearson correlation coefficient.
We obtained the clinical information from TCGA colon cancer dataset. The relevance of miR-200c expression level and patients’ clinic pathological features was evaluated to reveal the possible influence of miR-200c on patients’ clinical outcomes. The association of miR-200c with age, overall survival, tumor metastasis, colon polyps history, lymphatic invasion, tumor stages, and other clinical features was evaluated.
Clinical relevance of miR-200c.
Number | miRNA expression |
|
||
---|---|---|---|---|
High | Low | |||
|
|
|||
>60 | 198 | 108 | 24 | |
≤60 | 31 | 90 | 27 | |
|
|
|||
0 | 204 | 107 | 97 | |
1 | 41 | 22 | 19 | |
|
| |||
Yes | 105 | 47 | 58 | |
No | 133 | 78 | 55 | |
|
|
|||
Yes | 127 | 69 | 58 | |
No | 101 | 43 | 58 | |
|
|
|||
I | 50 | 24 | 26 | |
II | 89 | 47 | 42 | |
III | 65 | 36 | 29 | |
IV | 41 | 22 | 19 | |
|
|
|||
With tumor | 214 | 110 | 104 | |
Tumor free | 32 | 13 | 19 | |
|
|
|||
|
83 | 37 | 46 | |
<2 | 160 | 83 | 77 | |
|
|
|||
Yes | 58 | 29 | 29 | |
No | 151 | 63 | 88 | |
|
|
|||
|
156 | 66 | 90 | |
>31 days (median) | 92 | 58 | 34 |
In order to reveal the gene-gene interactions underlying in colon cancer pathogenesis, we constructed three coexpression networks according to the clustering of Pearson correlation of gene expression. Genes with similar function in a biological process were hypothesized to have similar expression patterns. Coexpression analysis identified gene-gene interaction network through the correlation of gene expression profile and clustering of thousands of transcript into a functional module. As shown in Figure
Coexpression network identified with mRNA expression of colon cancer. Node with interaction larger than 20 was colored in yellow.
The three networks identified with gene expression of tumor tissues may contribute to the initiation and development of colon cancer. In order to characterize the molecular functions of the networks in colon cancer, the pathway enrichment analysis was performed. All the genes involved in the networks were used to query the KEGG database to identify enriched pathway. Significantly enriched KEGG pathways with Fisher exact
Pathway enrichment analysis of network 1.
Term | Count | % |
|
Genes |
---|---|---|---|---|
hsa04110: Cell cycle | 8 | 1.91 |
|
CCNE2, CDK1, E2F5, DBF4, TTK, ANAPC10, RB1, CDC27 |
hsa04114: Oocyte meiosis | 6 | 1.43 |
|
CCNE2, CDK1, SLK, FBXO5, ANAPC10, CDC27 |
hsa00230: Purine metabolism | 7 | 1.67 |
|
POLR3G, POLE2, POLR2K, NT5C3, PDE4D, RRM2B, PPAT |
hsa05222: Small cell lung cancer | 5 | 1.19 |
|
CCNE2, PIK3CB, ITGAV, RB1, BIRC2 |
hsa00240: Pyrimidine metabolism | 5 | 1.19 |
|
POLR3G, POLE2, POLR2K, NT5C3, RRM2B |
hsa04120: Ubiquitin mediated proteolysis | 6 | 1.43 |
|
TRIM37, UBE2W, UBA6, ANAPC10, BIRC2, CDC27 |
hsa00512: O-Glycan biosynthesis | 3 | 0.72 |
|
GALNT3, GALNT7, C1GALT1 |
hsa05200: Pathways in cancer | 10 | 2.39 |
|
CCNE2, NRAS, HIF1A, PIK3CB, ITGAV, BRCA2, KITLG, RB1, BIRC2, FZD6 |
Pathway enrichment analysis of network 2 and network 3
Term | Count | % |
|
Genes |
---|---|---|---|---|
hsa04144: Endocytosis | 9 | 2.85 |
|
ARFGAP1, GIT1, HRAS, AP2A1, GRK6, RAB11B, HGS, EPN1, SH3GL1 |
hsa04150: mTOR signaling pathway | 5 | 1.58 |
|
ULK1, STK11, TSC2, MLST8, RPTOR |
hsa04330: NOTCH signaling pathway | 4 | 1.27 |
|
NOTCH3, RFNG, NCOR2, DVL1 |
Recent studies have shown the importance of miRNA and its regulatory network in colorectal carcinoma and the potential role of miRNA in colon cancer initiation and progression has been partly explored [
In this study, the gene expression of miRNA-200c was associated with DCN expression. The potential role of miR-200c has been well investigated in many cancer types, including colon cancer, ovarian cancer [
In addition, to date, miR-200 is divided into two groups according to a single nucleotide change in the seed sequence. MiR-141 and miR-200a are classified in group A, while miR-200b, miR-200c, and miR-429 are classified in group B. Notably, gene expression of miR-200a was also shown to be associated with 6 genes in this dataset. MiR-200a has been reported to be involved in the regulation of E-cadherin and involved in the pathogenesis of several types of cancer [
Lastly, three coexpression networks have been constructed in this study. Hundreds of genes are involved in the networks, which may be related to the development and progression of colon cancer. Genes involved in cell cycle and oocyte meiosis pathways were significantly affected, which indicated that coexpression networks were related to tumor proliferation and growth. In addition, tumor pathways were also significantly enriched, including small cell lung cancer, NOTCH, and mTOR signaling pathways which were often altered in cancer tissues [
Overall, we identified the miRNA and mRNA association in CRC, and the DCN might be a potential target of miR-200c, which indicated the important role of miR-200c and DCN in CRC. Clinical manifestations also implied the significant relation between colon polyps history, overall survival, and the expression level of miR-200c. This study provided an insight into cancer related mRNA coexpression network of CRC.
Colorectal carcinoma
Decorin
Carcinoembryonic antigen
Prostate-specific antigen
Carcinoma antigens
Circulating tumor DNA
Circulating tumor cells
Chronic lymphocytic leukemia
Patient-derived orthotopic xenograft.
The authors declare that they have no conflicts of interest.