T1711 Dynamic Epigenetic Changes of MT3 Promoter Regulate Its Expression in Esophageal Adenocarcinomas

Abstract

Background: Chronic gastro-esophageal reflux, in which accumulation of reactive oxygen species (ROS) and subsequent oxidative DNA damage occur, is the main risk factor for the development of Barrett's esophagus and progression to esophageal adenocarcinoma (EAC). Metallothionein 3 (MT3) is involved in maintaining intracellular metal homeostasis and was recently reported to be involved in the protection against ROS-induced DNA damage. The MT3 expression and epigenetic alterations have not been reported in EACs. Method and Results: We designed several assays and quantified the changes of the DNA methylation levels of the MT3 promoter CpG island region from -370 to +340 of the transcription start site (TSS) by using the Pyrosequencing technology. The results demonstrated unique dynamic DNA methylation profiles in the MT3 CpG island region. Interestingly, the CpG island region between -370 to -306 from the TSS demonstrated a 48-55% methylation level in all EAC and normal samples. On the other hand, the region from -161 to -130 of TSS showed high levels of DNA methylation in EACs (35.3%) but low methylation levels in NS (mean, 11.2%) (P<.01). Analysis of the CpG island region from -127 to +340 demonstrated less than 10% methylation levels in all normal samples. The methylation level of this region in EAC showed more than 25% methylation. However, the CpG nucleotides adjacent to the TSS (4 to +3) had an unexpected drop in their methylation levels to 10% in EAC samples. Using 10% as the cutoff for DNA hypermethylation, the average methylation level of CpG sites from -127 to -26 was significantly higher in 62% (49/79) of EACs (mean ± SD, 22.6 ± 21.2) than in normals (4.9 ± 2.5). Of note, the methylation level of these sites showed inverse correlations with MT3 expression, confirming that this is a critical region in regulating MT3 transcription. We have also determined the histone code along the CpG island using quantitative chromatin immunoprecipitation (CHIP) assay on two cells lines with different methylation and expression levels of MT3. The results demonstrated more repressive histone markers (H3K9me2, H3K27me2) and a less active histone marker (H3K9ace) on this region of MT3 in OE33 cells, as compared to Flo-1 cells. The results were concordant with the DNA methylation and mRNA expression levels of these cell lines. Treatment of OE33 cells with 5-Aza restored MT3 expression with demethylation of its promoter region and reversal of the repressive histone code to an active code. Conclusions: Our results suggest 1) frequent epigenetic silencing of MT3 in EACs; 2) selecting the CpG sites for functional analysis of DNA methylation is necessary.