Abstract
Abstract 3618
DNA methylation is the most stable epigenetic modification and has a major role in cancer initiation and progression. The two main aims for this research were, firstly, to use the genome wide analysis of DNA methylation to better understand the development of acute myeloid leukemia (AML). The second aim was to detect differentially methylated genes/regions between certain subtypes of AML and normal bone marrow (NBM). We used the methylated DNA immunoprecipitation technique followed by high-throughput sequencing by Illumina Genome Analyser II (MeDIP -seq) for 9 AML samples for which ethical approval has been obtained. The selected leukemias included three with the t(8; 21), three with the t(15; 17) translocations and three with normal karyotypes (NK). The control samples were 3 normal bone marrows (NBMs) from healthy donors. The number of reads generated from Illumina ranged between 18– 20 million paired-end reads/lane with a good base quality from both ends (base quality > 30 represented 75%-85% of reads). The reads were aligned using 2 algorithms (Maq and Bowtie) and the methylation analysis was performed by Batman software (Bayesian Tool for Methylation Analysis).
The creation of this genome-wide methylation map for AML permits the examination of the patterns for key genetic elements. Investigation of the 35,072 promoter regions identified 80 genes, which showed a significant differential methylation levels in leukemic cases in comparison to NBM; consistently high methylation levels in leukaemia were detected in the promoters of 70 genes e.g. DPP6, ID4, DCC, whereas high methylation levels in NBM, lost in leukaemia was observed in 10 genes e.g. ATF4. For each AML subtype, we also identified significant differentially methylated promoter regions e.g. PAX1 for t(8; 21), GRM7 for t(15; 17), NPM2 for NK.
An analysis of gene body methylation identified 49 genes with significantly higher methylation in AML in comparison to NBM e.g. MYOD1 and 31 genes with a higher methylation in NBMs than AML e.g. GNG8. A similar analysis of 23,600 CpG islands identified 400 CpG islands with significant differential methylation levels between leukaemia and NBMs (212 CpG islands were found to have significantly increased methylation in leukaemia and 188 CpG islands had significantly higher methylation in NBMs). The pattern of methylation in CpG island “shores” (2 KB from either side of each CpG island) has been investigated and 312 CpG island shores showed a higher methylation in leukaemia and 88 CpG shores had a significant increase methylation levels in NBMs. This genome wide methylation map has been validated by using direct bisulfite sequencing of the regions identified above (Spearman r= 0.8, P <0.0001) and also by using Illumina Infinium assay (Spearman r= 0.7 P <0.0001) which interrogates regions at single representative CpGs. Comparison of previous array based gene expression data with this methylation map revealed a significant negative correlation between promoter methylation and gene expression (Pearson r= -0.9, P< 0.0001) while, gene body methylation showed a small negative correlation with gene expression, that was found in genes of CpG density >3% (Pearson r= -0.3, P< 0.0001).
Conclusion:
we have established a high-resolution (100bp) map of DNA methylation in AML and thus identified a novel list of genes, which have significantly differential methylation levels in AML.
Disclosures:
No relevant conflicts of interest to declare.
Genome-wide investigation of the dynamic changes of epigenome modifications after global DNA methylation editing
