Abstract

Age-related macular degeneration (AMD) is a common vision-threatening disease. The current study sought to integrate DNA methylation with transcriptome profile to explore key features in AMD. Gene expression data were obtained from the Gene Expression Omnibus (GEO, accession ID: GSE135092) and DNA methylation data were obtained from the ArrayExpress repository (E-MTAB-7183). A total of 456 differentially expressed genes (DEGs) and 4827 intragenic differentially methylated CpGs (DMCs) were identified between AMD and controls. DEGs and DMCs were intersected and 19 epigenetically induced (EI) genes and 15 epigenetically suppressed (ES) genes were identified. Immune cell infiltration analysis was performed to estimate the abundance of different types of immune cell in each sample. Enrichment scores of inflammatory response and tumor necrosis factor-alpha (TNFα) signaling via nuclear factor kappa B (NF-κb) were positively correlated with abundance of activated memory CD4 T cells and M1 macrophages. Subsequently, two significant random forest classifiers were constructed based on DNA methylation and transcriptome data. SMAD2 and NGFR were selected as key genes through functional epigenetic modules (FEM) analysis. Expression level of SMAD2, NGFR and their integrating proteins was validated in hydrogen peroxide (H2O2) and TNFα co-treated retinal pigment epithelium (RPE) in vitro. The findings of the current study showed that local inflammation and systemic inflammatory host response play key roles in pathogenesis of AMD. SMAD2 and NGFR provide new insight in understanding the molecular mechanism and are potential therapeutic targets for development of AMD therapy.