Identifying an lncRNA-Related ceRNA Network to Reveal Novel Targets for a Cutaneous Squamous Cell Carcinoma

Abstract

A cutaneous squamous cell carcinoma (cSCC) derived from keratinocytes is the second most common cause of non-melanoma skin cancer. The accumulation of the mutational burden of genes and cellular DNA damage caused by the risk factors (e.g., exposure to ultraviolet radiation) contribute to the aberrant proliferation of keratinocytes and the formation of a cSCC. A cSCC encompasses a spectrum of diseases that range from recursor actinic keratosis (AK) and squamous cell carcinoma (SCC) in situ (SCCIS) to invasive cSCCs and further metastatic SCCs. Emerging evidence has revealed that lncRNAs are involved in the biological process of a cSCC. According to the ceRNA regulatory theory, lncRNAs act as natural miRNA sponges and interact with miRNA response elements, thereby regulating the mRNA expression of their down-stream targets. This study was designed to search for the potential lncRNAs that may become potential therapeutic targets or biomarkers of a cSCC. Considering the spirit of the study to be adequately justified, we collected microarray-based datasets of 19 cSCC tissues and 12 normal skin samples from the GEO database (GSE42677 and GSE45164). After screening the differentially expressed genes via a limma package, we identified 24 differentially expressed lncRNAs (DElncRNAs) and 3221 differentially expressed mRNAs (DEmRNAs). The miRcode, miRTarBase, miRDB and TargetScan databases were used to predict miRNAs that could interact with DElncRNAs and DEmRNAs. A total of 137 miRNA-lncRNA and 221 miRNA-mRNA pairs were retained in the ceRNA network, consisting of 31 miRNAs, 11 DElncRNAs and 155 DEmRNAs. For the functional analysis, the top enriched biological process was enhancer sequence-specific DNA binding in Gene Ontology (GO) terms. The FoxO signaling pathway, autophagy and cellular senescence were the top enrichment terms based on a Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis. The combination of a STRING tool and Cytoscape software (plug-in MCODE) identified five core mRNAs and built a core mRNA-associated ceRNA network. The expression for five identified core mRNAs and their related nine lncRNAs was validated using the external dataset GSE7553. Finally, one lncRNA HLA-F-AS1 and three mRNAs named AGO4, E2F1 and CCND1 were validated with the same expression patterns. We speculate that lncRNA HLA-F-AS1 may sponge miR-17-5p or miR-20b-5p to regulate the expression of CCND1 and E2F1 in the cSCC. The present study may provide potential diagnostic and therapeutic targets for cSCC patients.

Detection and analysis of long noncoding RNA expression profiles related to epithelial–mesenchymal transition in keloids

“…As an emerging class of regulatory RNAs, lncRNAs may participate in fundamental processes including cell cycle regulation, proliferation, apoptosis, self-renewal, metastasis and the DNA damage response through transcriptional or posttranscriptional regulation and thus give rise to diseases. Abnormally expressed lncRNAs play a key role in fibrosis [ 10 ] and different types of cancer, including breast cancer [ 11 – 13 ], lung cancer [ 14 , 15 ], colorectal cancer [ 16 , 17 ], esophageal squamous cell carcinoma [ 18 ], and cutaneous squamous cell carcinoma [ 19 , 20 ].…”

The long non-coding RNA PVT1 promotes tumorigenesis of cutaneous squamous cell carcinoma via interaction with 4EBP1

“…LncRNAs have been reported to be involved in various biological processes in tumors through multiple coordinative mechanisms, such as direct binding with DNA, RNA, and specific proteins [36]. The conventional view holds that lncRNAs primarily act as competitive endogenous RNAs (ceRNAs) to facilitate cSCC tumor progression by regulating miRNA-lncRNA interactions [37][38][39]. In contrast, accumulating evidence has indicated that lncRNAs, which are predominantly located in the cytoplasm, can exert their function by binding with targeted proteins and modulating the expression of corresponding proteins [40][41][42][43][44].…”

Integrated analysis of the molecular mechanisms in idiopathic pulmonary fibrosis

“…We used perl script to convert the probe IDs in the probe expression matrices to the corresponding gene symbols in the platform files 22 . Then, we used Limma package to analyze each obtained matrix file of gene symbols, calculated p values and |logFC|, and screen differential genes according to same criteria 23 . Our criteria for screening DEGs were P < 0.05 and an absolute value of logFC > 1.…”