A Combination of Microarray-Based Profiling and Biocomputational Analysis Identified miR331-3p and hsa-let-7d-5p as Potential Biomarkers of Ulcerative Colitis Progression to Colorectal Cancer

Abstract

Ulcerative colitis (UC), an inflammatory bowel disease (IBD), may increase the risk of colorectal cancer (CRC) by activating chronic proinflammatory pathways. The goal of this study was to find serum prediction biomarkers in UC to CRC development by combining low-density miRNA microarray and biocomputational approaches. The UC and CRC miRNA expression profiles were compared by low-density miRNA microarray, finding five upregulated miRNAs specific to UC progression to CRC (hsa-let-7d-5p, hsa-miR-16-5p, hsa-miR-145-5p, hsa-miR-223-5p, and hsa-miR-331-3p). The circRNA/miRNA/mRNA competitive endogenous RNA (ceRNA) network analysis showed that the candidate miRNAs were connected to well-known colitis-associated CRC ACVR2A, SOCS1, IGF2BP1, FAM126A, and CCDC85C mRNAs, and circ-SHPRH circRNA. SST and SCARA5 genes regulated by hsa-let-7d-5p, hsa-miR-145-5p, and hsa-miR-331-3p were linked to a poor survival prognosis in a CRC patient dataset from The Cancer Genome Atlas (TCGA). Lastly, our mRNA and miRNA candidates were validated by comparing their expression to differentially expressed mRNAs and miRNAs from colitis-associated CRC tissue databases. A high level of hsa-miR-331-3p and a parallel reduction in SOCS1 mRNA were found in tissue and serum. We propose hsa-miR-331-3p and possibly hsa-let-7d-5p as novel serum biomarkers for predicting UC progression to CRC. More clinical sample analysis is required for further validation.

Keywords: biomarkers inflammatory bowel disease; colorectal cancer; competing endogenous RNA; miRNA; ulcerative colitis.

Figure 1

Microarray low-intensity profile analysis between UC and nonrelated IBD progression to CRC. (A) Global and differential expression levels of miRNA across the screening set as shown by hierarchical clustering heatmap. The analysis summarizes the differential patterns of miRNA expression across the groups, ulcerative colitis (UC) and nonrelated IBD inflammation control miRNA expression profiles compared to colorectal cancer (CRC) miRNA expression pattern. Reddish values indicate upregulation, whereas greenish values show downregulation; (B) logarithm in base 10 FC of our miRNA candidates comparing UC and control to CRC progression.

Figure 2

Upregulated miRNA candidates’ interaction with their inhibitors (circRNAs) and targets (mRNAs). (A) CeRNA circRNA-miRNA-mRNA links. A detailed description of the ceRNA members is provided in the Supplementary Materials; (B) ceRNA subnetwork for upregulated miRNAs. Only the most significant interactions (3 or more links to other molecules) are represented in this subnetwork, in addition to the circRNAs. miRNAs are shown in red; mRNA targeted by different miRNAs (blue) showing 1 (small node size), 2 (medium node size), and more than 3 (bigger node sizes); circRNAs are shown in orange.

Figure 3

PPI network and its top ten hub genes constructed by miRNA targets. (A) PPI network built by using upregulated miRNA-mRNA targets; in total, 453 nodes and 1216 interactions between these nodes were established. In color, the different identified protein clusters are shown in the Supplementary Materials; (B) top ten hub genes analyzed from the upregulated miRNA targets; (C,D) the most significant genes regarding the number of their interactions with other genes; (E) the main clusters in which our top hub genes were identified next to the ACVR2A protein of interest.

Figure 4

GO and KEGG pathway analysis of the upregulated miRNA targets. The GO top 10 terms of the biological processes (A), cellular components (B), and molecular functions (C). The length of each box correlates with the number of genes involved in each process, whereas the color is related to the significance of the value; (D) top 10 most significant KEGG pathways identified. The size of the circles represents the number of genes involved in each one of the pathways, whereas the color represents the significance of the value.

Figure 5

Survival analysis of SST and SCARA5 mRNA expression. Downregulated mRNA expression is represented in red, while upregulated expression is represented in blue. The boxes describe the number of patients analyzed in each period. Each cross shows when a patient was excluded from the analysis. (A) SCARA5 mRNA upregulation was also linked to a worse tumoral prognosis (p-value = 0.034); (B) SST mRNA upregulation was significantly related to a poorer prognosis (p-value = 0.044).