Integrative analysis of gene expression and DNA methylation to identify biomarkers of non-genital warts induced by low-risk human papillomaviruses infection

Abstract

Background: Human papillomaviruses have been shown to dysregulate the gene expression and DNA methylation profiles of their host cells over the course of infection. However, there is a lack of information on the impact of low-risk HPV infection and wart formation on host cell's expression and methylation patterns. Therefore, the objective of this study is to analyse the genome and methylome of common warts using an integrative approach.

Methods: In the present study, gene expression (GSE136347) and methylation (GSE213888) datasets of common warts were obtained from the GEO database. Identification of the differentially expressed and differentially methylated genes was carried out using the RnBeads R package and the edgeR Bioconductor package. Next, functional annotation of the identified genes was obtained using the Database for Annotation, Visualization, and Integrated Discovery (DAVID). Network construction and analyses of the gene-gene, protein-protein, and signaling interactions of the differentially expressed and differentially methylated genes was performed using the GeneMANIA web interface, the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database, and the Signaling Network Open Resource 2.0 (SIGNOR 2.0), respectively. Lastly, significant hub genes were identified using the Cytoscape application CytoHubba.

Results: A total of 276 genes were identified as differentially expressed and differentially methylated in common warts, with 52% being upregulated and hypermethylated. Functional enrichment analysis identified extracellular components as the most enriched annotations, while network analyses identified ELN, ITGB1, TIMP1, MMP2, LGALS3, COL1A1 and ANPEP as significant hub genes.

Conclusions: To the best knowledge of the authors, this is the first integrative study to be carried out on non-genital warts induced by low-risk HPV types. Future studies are required to re-validate the findings in larger populations using alternative approaches.

Keywords: DNA methylation; Gene expression; HPV; Human papillomavirus; Warts.

Fig. 1

Expression and methylation status (a) and chromosomal location of DEDM genes (b) in common warts.

Fig. 2

Analysis of the top five most significant GO terms obtained for the DEDM genes in terms of expression (A) and methylation (B).

Fig. 3

Gene-gene networks of the downregulated and hypermethylated (A) as well as the downregulated and hypomethylated (B) DEDM genes in warts. The size of a circle is proportional to the number of correlations with other genes in the gene-gene network. Striped circles represent the DEDM genes, while unstriped circles indicate associated genes.

Fig. 4

Gene-gene network of the upregulated and hypermethylated (A) as well as the upregulated and hypomethylated (B) DEDM genes in warts. The size of a circle is proportional to the number of correlations with other genes in the gene-gene network. Striped circles represent the DEDM genes, while unstriped circles indicate associated genes.

Fig. 5

Protein-protein network of DEDM genes in warts. The interaction score was set at the highest confidence (0.900), and disconnected nodes are not shown.

Fig. 6

Signaling network of DEDM genes in warts. The interaction score was set at the highest confidence (0.9).