The role of long non-coding RNA LINC00839 in oral squamous cell carcinoma based on bioinformatics and experimental research

Abstract

This study explores the role of LINC00839 and its potential interaction with the miR-195-5p/cyclin E1 (CCNE1) axis in oral squamous cell carcinoma (OSCC). Using The Cancer Genome Atlas, we analyzed lncRNA, miRNA, and mRNA sequencing data for OSCC. Different online tools were applied to detect lncRNA-related miRNAs and their target mRNAs, forming a lncRNA/miRNA/mRNA axis. Co-expression analysis determined the correlation between lncRNA and mRNA expression. Afterward, protein-protein interaction network and functional enrichment analyses disclosed the biological activity of target genes. The expression and correlations of LINC00839, miR-195-5p, and CCNE1 were examined in 30 pairs of OSCC and noncancerous tissues. A Chi-square test was used to determine clinicopathological associations, and ROC analysis estimated diagnostic value. A total of 66 differentially expressed lncRNAs, 80 miRNAs, and 1149 mRNAs were identified in OSCC versus non-tumor samples. After filtering lncRNAs based on novelty, and predicting lncRNA-miRNA, and miRNA-mRNA interactions, the LINC00839/miR-195-5p/CCNE1 axis was discovered. RT-qPCR showed upregulation of LINC00839 and CCNE1 was accompanied by the downregulation of miR-195-5p. A significant positive correlation was observed between LINC00839 and CCNE1 mRNA expression, along with a significant negative correlation between LINC00839 and miR-195-5p expression. Moreover, increased LINC00839 was associated with tumor grade and lymph node status, while decreased miR-195-5p was correlated with lymph, depth, and vascular invasion (p < 0.05). The combined ROC curve demonstrated a significant area under the curve of 0.93. This discovery reveals a novel regulatory mechanism underlying OSCC tumorigenesis and may provide effective diagnosis and potential therapeutic targets to cure this devastating cancer.

Keywords: Cyclin E1; Diagnosis; LINC00839; Oral squamous cell carcinoma; Prognosis; miR-195-5p.

Fig. 1

A flowchart demonstrating the bioinformatics methods applied in this research to discover the final regulatory axis for OSCC.

Fig. 2

Volcano plot of differentially expressed (DE) lncRNAs (A), miRNAs (B), and mRNAs (C). The red dots correspond to the upregulated DERNAs, while the blue dots represent the downregulated ones. The black dots signify the DERNAs that were excluded based on two specific criteria: |logFC|≥1.5 and FDR < 0.01.

Fig. 3

Correlation and expression analysis of the LINC00839/miR-195-5p/CCNE1 axis according to TCGA database data. (A) A Pearson correlation analysis was conducted to determine the relationship between the expression levels of LINC00839 and CCNE1. (B) The expression of LINC00839, classified as an oncogenic lncRNA, was compared between OSCC tumor tissues (n = 293) and normal tissues (n = 32). (C) The expression of miR-195-5p, a tumor suppressor miRNA, was analyzed in tumor samples (n = 303) versus normal tissues (n = 32), and (D) the expression of CCNE1, recognized as an oncogene, was assessed in both OSCC tumor and normal tissues. All the results yielded significant p-values and false discovery rates.

Fig. 4

Protein-protein interaction (PPI) network of CCNE1. (A) The CCNE1 PPI network is constituted of 21 nodes and 259 edges. The node size and edge thickness are associated with the degree value and strong data sources, respectively. Additionally, the unique colors assigned to the edges represent different types of interactions. (B) The percentages of different interactions that created the network: physical interactions, co-expression, predicted, co-localization, genetic interactions, pathway, and shared protein domains.

Fig. 5

Dot plot and cnetplot of Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses. Each dot plot indicates significantly enriched terms in GO categories, including (A) biological process (BP), (B) cellular component (CC), and (C) molecular function (MF), accompanied by enriched terms in (D) KEGG. The terms at the top of the dot plot are more significant than those at the bottom. The cnetplot represents the connections between individual proteins of the CCNE1 PPI network and GO categories, including (E) BP, (F) CC, (G) MF, and (H) KEGG. CCNE1 is depicted in bold.

Fig. 6

Analysis of LINC00839/miR-195-5p/cyclin E1 (CCNE1) axis expression and their correlation in OSCC tissues. The relative expression of LINC00839 (A), miR-195-5p (B), and CCNE1 mRNA (C) in 30 OSCC tissues versus adjacent non-tumor tissues was evaluated using RT-qPCR. (D) The protein level of CCNE1 was determined through western blotting (corresponding graph is represented as mean ± SD following normalization against glyceraldehyde 3-phosphate dehydrogenase [GAPDH]). (E) Correlation analysis between LINC00839 and CCNE1 expression levels (Spearman’s nonparametric test). (F) Pearson’s test was conducted to analyze the correlation between LINC00839 and miR-195-5p expression levels. (G) The correlation between miR-195-5p and CCNE1 was determined based on the Pearson correlation coefficient. **p < 0.01 and ****p < 0.0001.

Fig. 7

Immunofluorescence staining of oral squamous cell carcinoma tumor and margin tissues. The nuclei were stained using 4′,6-Diamidino‐2-phenylindole (DAPI), which emits a blue fluorescence, while the cyclin E1 protein was detected using a green fluorescent antibody. Scale bar = 50 μm. *p < 0.05. Original magnification, × 200.

Fig. 8

ROC curve analysis of LINC00839, miR-195-5p, CCNE1, and their combination for OSCC diagnosis in tumor tissues. AUC: area under the curve, CCNE1: cyclin E1.

Fig. 9

Schematic diagram illustrating our bioinformatics analysis and experimental validation of the LINC00839/miR-195-5p/CCNE1 axis in the progression of OSCC.