Analysis of Human Papillomavirus-Associated Cervical Cancer Differentially Expressed Genes and Identification of Prognostic Factors using Integrated Bioinformatics Approaches

Abstract

Background: Human papillomavirus (HPV)-induced cervical cancer progresses through a series of steps. Despite our limited understanding of the mechanisms driving this progression, identifying the key genes involved could significantly improve early detection and treatment.

Materials and methods: Two gene expression profiles of GSE9750 and GSE6791, which included cervical cancer HPV-positive and -negative samples, were evaluated using the R limma package with established cut-off criteria of P value < 0.05 and | fold change| ≥ 1. KEGG pathway enrichment was performed to identify potential pathways. Weighted gene co-expression network analysis (WGCNA) was used to discover co-expressed gene modules and trait-module connections.

Results: Considering the defined criteria, 115 differentially expressed genes (DEGs) were identified. The DEG's KEGG pathway enrichment analysis revealed enrichment in highly relevant pathways to the HPV infection, including cell cycle, viral carcinogenesis, autophagy-animal, Epstein-Barr virus infection, human T-cell leukemia virus 1 infection, and microRNAs in cancer. WGCNA results in 13 co-expression modules, and the magenta module is identified with significant relations to HPV, cervical cancer stage, and metastasis traits. The survival analysis identified BEX1 and CDC45 as potential prognostic factors in HPV-associated cervical cancer.

Conclusion: The innovation of our work lies in identifying essential genes associated with the multi-step process of cervical carcinogenesis. In fact, the current study has the potential to give a distinct viewpoint on the molecular pathways linked to cervical cancer. Considering the potential importance of the hub genes, we recommend conducting in-depth wet lab research to determine their impact on the biological mechanisms of cervical cancer.

Keywords: BEX1; CDC45; HPV; cell cycle; cervical cancer; gene expression; gene modules; gene networks.

Figure 1

The boxplot of datasets and DEGs’ volcano plan. Screening for DEGs was done using a P value < 0.05 and | FC| ≥ 1

Figure 2

(a) The bubble chart of the KEGG pathway analysis of DEGs. (b) The clustering between the KEGG pathway enrichments based on the shared genes

Figure 3

Viral carcinogenesis. The cell cycle involvement in the HPV carcinogenesis is demonstrated in the viral carcinogenesis pathways of different viruses

Figure 4

WGCNA analysis. (a) Sample clustering dendrogram and identification of outliers (the outliers were removed after identification). (b) The soft threshold selection. Scale-free topology fitting index R2 analysis (left) and mean connectivity for various soft threshold powers (right). The left panel’s red line indicates that R2 = 4. (c) A clustering diagram of gene modules denoted by distinct colors

Figure 5

(a) Clinical trait heatmap and sample dendrogram. The sample dendrogram’s threshold value is set at 100 to remove samples with considerable variability. (b) Module–trait relationship heatmap. Hierarchical clustering of module eigengenes that represent the clustering analysis’s modules. The module is represented by the row, while the column represents the trait. The P values in the box show the correlation and P value. (c) The gene significance across the modules. The magenta module showed significant module–trait relationship among the other modules (P value = 4.5e-37). (d). BEX1, CDC45, and FAM107A identified the magenta module gene–gene interaction as the module hub genes

Figure 6

The relationship between the magenta module hub-gene expression and the clinical outcome in cervical cancer patients. The lower expression of BEX1 (Logrank P = 0.01) and CDC45 (Logrank P = 0.00049) exhibited significantly worse overall survival. The GAPDH gene was utilized to normalize data from targeted genes