Abstract
MicroRNAs (miRNAs) are aberrantly expressed in virtually all cancer types, including digestive cancers. Herein, we aggregated and systematically analyzed miRNA expression profiles of 1765 tumor samples, including esophageal, gastric, liver, pancreatic, colon and rectal cancers, obtained through small RNA sequencing by The Cancer Genome Atlas. We found that digestive cancers of different tissue origins could be differentiated according to their miRNA expression profiles. In particular, esophageal squamous cell carcinoma and esophageal adenocarcinoma exhibited distinct miRNA expression patterns. Thirteen (e.g. miR-135b, miR-182) and sixteen (e.g. miR-139, miR-133a-1, miR-490) miRNAs were commonly upregulated and downregulated in more than four cancer types, respectively. Pertinent to pathological features, low miR-181d expression was associated with microsatellite instability in colon and gastric cancers whereas low miR-106a expression was associated with hepatitis B virus infection in hepatocellular carcinoma. Progression in colon cancer could also be predicted by low let-7f-2 and high miR-106a expression. Molecular subtypes with distinct prognostic outcomes independent of tumor-node-metastasis staging were identified in hepatocellular carcinoma and colon cancer. In total, 4 novel and 6 reported associations between specific miRNAs and patients’ survival were identified. Collectively, novel miRNA markers were identified to stratify digestive cancers with different pathological features and survival outcomes.
Highlights
MicroRNAs are a group of small non-coding RNAs that are ~22 nucleotides in length. miRNAs are aberrantly expressed in virtually all types of human cancers, including digestive cancers[9,10,11,12], in which they could alter cellular phenotypes, such as proliferation, apoptosis and invasiveness, through their interactions www.nature.com/scientificreports/
Principal component analysis (PCA) using the differential miRNA expression data of 1765 tumor samples from six major digestive cancers, namely esophageal cancer, hepatocellular carcinoma (HCC), gastric adenocarcinoma, pancreatic adenocarcinoma, colon adenocarcinoma and rectal roughly divided samples into 5 subgroups consistent with their tumor origins but esophageal cancer samples were mixed with gastric adenocarcinoma (Fig. 1A,B)
Further clustering using PCA divided esophageal cancer and gastric adenocarcinoma samples into two subgroups, in which one subgroup was dominated by esophageal squamous cell carcinoma (ESCC) while the other consisted of esophageal adenocarcinoma (EAC) and gastric adenocarcinoma (Fig. 1C), thereby yielding a total of seven distinct clusters in PCA (Fig. 1D)
Summary
MicroRNAs (miRNAs) are a group of small non-coding RNAs that are ~22 nucleotides in length. miRNAs are aberrantly expressed in virtually all types of human cancers, including digestive cancers[9,10,11,12], in which they could alter cellular phenotypes, such as proliferation, apoptosis and invasiveness, through their interactions www.nature.com/scientificreports/. MiRNAs are aberrantly expressed in virtually all types of human cancers, including digestive cancers[9,10,11,12], in which they could alter cellular phenotypes, such as proliferation, apoptosis and invasiveness, through their interactions www.nature.com/scientificreports/. Some miRNAs have been shown to correlate with cancer progression and may be used as prognostic markers[13]. The use of the generated datasets for discovery of novel miRNA markers for clinical utilization, prognostication, has not yet been achieved. We report an integrative analysis of digestive cancers by their miRNA expression profiles obtained from The Cancer Genome Atlas (TCGA). We demonstrated that miRNA expression profiles could be used to differentiate digestive cancers of different tissue origins. We identified molecular subtypes and specific miRNAs that were associated with clinicopathological features, including patients’ survival
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have