Background
Mounting evidence suggests that noncoding RNAs (lncRNAs) were involved in various human cancers. However, the role of these lncRNAs in HPV-driven cervical cancer (CC) has not been extensively studied. Considering that HR-HPV infections contribute to cervical carcinogenesis by regulating the expression of lncRNAs, miRNAs and mRNAs, we
Conclusions
Together, these data identify co-expression modules that provide valuable information to understand the pathogenesis of HPV-mediated tumorigenesis, which highlights the pivotal function of the LINC00511-PGK1 co-expression network in cervical carcinogenesis. Furthermore, our CES model has a reliable predicting ability that could stratify CC patients into low- and high-risk groups of poor survival. This study provides a bioinformatics method to screen prognostic biomarkers which leads to lncRNA-mRNA co-expression network identification and construction for patients' survival prediction and potential drug applications in other cancers.
Methods
LncRNA/mRNA microarray technology was utilized to identify the differentially expressed lncRNAs (DElncRNAs) and mRNAs (DEmRNAs) in HPV-16 and HPV-18 cervical carcinogenesis compared to normal cervical tissues. Venn diagram and weighted gene co-expression network analysis (WGCNA) were used to identify the hub DElncRNAs/DEmRNAs that were both significantly correlated with HPV-16 and HPV-18 CC patients. LncRNA-mRNA correlation analysis and functional enrichment pathway analysis were performed on these key DElncRNAs/DEmRNAs in HPV-16 and HPV-18 CC patients to explore their mutual mechanism in HPV-driven CC. A lncRNA-mRNA co-expression score (CES) model was established and validated by using the Cox regression method. Afterward, the clinicopathological characteristics were analyzed between CES-high and CES-low groups. In vitro, functional experiments were performed to evaluate the role of LINC00511 and PGK1 in cell proliferation, migration and invasion in CC cells. To understand whether LINC00511 play as an oncogenic role partially via modulating the expression of PGK1, rescue assays were used.
Results
We identified 81 lncRNAs and 211 mRNAs that were commonly differentially expressed in HPV-16 and HPV-18 CC tissues compared to normal tissues. The results of lncRNA-mRNA correlation analysis and functional enrichment pathway analysis showed that the LINC00511-PGK1 co-expression network may make an important contribution to HPV-mediated tumorigenesis and be closely associated with metabolism-related mechanisms. Combined with clinical survival data, the prognostic lncRNA-mRNA co-expression score (CES) model based on LINC00511 and PGK1 could precisely predict patients' overall survival (OS). CES-high patients had a worse prognosis than CES-low patients and the enriched pathways and potential targets of applicable drugs were explored in CES-high patients. In vitro experiments confirmed the oncogenic functions of LINC00511 and PGK1 in the progression of CC, and revealed that LINC00511 functions in an oncogenic role in CC cells partially via modulating the expression of PGK1. Conclusions: Together, these data identify co-expression modules that provide valuable information to understand the pathogenesis of HPV-mediated tumorigenesis, which highlights the pivotal function of the LINC00511-PGK1 co-expression network in cervical carcinogenesis. Furthermore, our CES model has a reliable predicting ability that could stratify CC patients into low- and high-risk groups of poor survival. This study provides a bioinformatics method to screen prognostic biomarkers which leads to lncRNA-mRNA co-expression network identification and construction for patients' survival prediction and potential drug applications in other cancers.