Abstract
Purpose :
Retinoblastoma (RB) is a cancer occurring in retinal precursors of young children. RB1 inactivation is a rate limiting step in RB initiation. Additional genetic/epigenetic events are necessary for malignant transformation of retinal cell. We performed epigenomic analysis to identify the potential prognostic/therapeutic targets in RB.
Methods :
We analysed global methylation and gene expression pattern in 7 RB tumors with 2 neural retina controls from donor tissues using Infinium MethylationEPIC Array and RNAseq. The methylation data analysis was carried out using shinyEPICo tool. After Noob-quantile normalisation, differentially methylated positions (DMPs) were identified from the delta beta values and differentially methylated regions (DMRs) were derived from DMPs. Transcription factors were selected from the DMRs and further validated. Integration of DMRs with differentially expressed genes (DEGs) was done to identify the epigenetically regulated genes.
Results :
Global methylation analysis identified 64807 DMPs and the enrichment of hyper methylated genes showed phototransduction or visual cycle as primary hit indicating the loss of photo receptorness in the retinoblastoma. Active proliferation of RB tumor cells was evident from hypo methylated genes that were enriched for glucose metabolism. DMR analysis revealed 1289 differentially methylated promoters in the key genes including APC2, VHL, MLH1, SYK, UHRF1, EZH2 GATA5, PAX5 and TFAP2A. Further analysis of 21 transcription factors confirmed the hypermethylation of PAX5 and its downregulation at transcript level. Integration of the 1289 promoters and 80 DEGs identified TMEM123 as a new candidate in retinoblastoma tumorigenesis.
Conclusions :
Our multi-omic approach has identified distinct epigenetic alterations impacting the transcription leading to RB tumorigenesis. The identified alterations can be used as prognostic or therapeutic indicators and suitable drugs that can revert these alterations may provide better treatment for retinoblastoma.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.