P12.06.A Relationship between ascorbate and DNA methylation markers in clinical glioma tumours

Abstract

Abstract Background Members of the 2-oxoglutarate-dependent dioxygenase (OGDD) enzyme family play an important role in gliomas as they regulate epigenetic modifications and response to hypoxia. The OGDDs require 2-OG and O2 as substrates, and ferrous iron and ascorbate as cofactors. Both hypoxia and aberrant DNA methylation are prognostic indicators for gliomas. The ten-eleven translocase (TET) DNA demethylases are OGDDs that convert 5-methyl cytosine (5mC) to 5-hydroxymethylcytosine (5hmC), with 5hmC levels related to better prognosis. Despite this, there is limited data on the OGDD enzymes and their substrates/cofactors in glioma tissues. Our previous study showed an association between ascorbate content and markers of the hypoxic response in glioblastoma tissue. Here we determine whether there is an association between ascorbate and DNA methylation in glioma. In addition, we assess whether methylation of the methylguanine-DNA methyltransferase (DNA repair enzyme MGMT) promoter is associated with ascorbate content. Materials and methods Frozen clinical glioma samples from 37 patients (n=11 WHO grade I-III, n=26 glioblastoma) were obtained from the Cancer Society Tissue Bank (Ethics approval H19/163). Isocitrate dehydrogenase 1 (IDH1) mutation status was determined by sequencing. Samples were processed on dry ice in liquid nitrogen and analysed for ascorbate (high-performance liquid chromatography), global DNA methylation (mass spectrometry) and MGMT promoter analyses (methylation specific PCR). Results Many grade I-III tumours were IDH1 R132H mutant (6/11), and most glioblastomas were not (2/26). Glioblastoma had significantly lower ascorbate content than grade I-III tumours (p=0.026). Glioblastoma also had lower global 5hmC levels (p=0.0013). IDH1 R132H tumours tended to have a lower ascorbate content (p=0.09). Ascorbate and 5hmC levels were directly correlated (Spearman r= 0.466, p=0.004). However, cytosine and 5mC showed no association with grade or ascorbate. MGMT promoter methylation status was not associated with global methylation or ascorbate content (p=0.97, p=0.96, respectively). Conclusion Our data suggests that ascorbate supports TET activity in clinical glioma. It also appears that site-specific (promoter) methylation was not affected by ascorbate availability. These findings may have clinical implications, as higher 5hmC levels are associated with improved outcome, whilst continued MGMT suppression suggests chemotherapy responsiveness. However, evidence that raising tumour ascorbate leads to increased 5hmC levels, or an associated improvement in survival, requires intervention trials.