Ascorbate content of clinical glioma tissues is related to tumour grade and to global levels of 5-hydroxymethyl cytosine

Abstract

Gliomas are incurable brain cancers with poor prognosis, with epigenetic dysregulation being a distinctive feature. 5-hydroxymethylcytosine (5-hmC), an intermediate generated in the demethylation of 5-methylcytosine, is present at reduced levels in glioma tissue compared with normal brain, and that higher levels of 5-hmC are associated with improved patient survival. DNA demethylation is enzymatically driven by the ten-eleven translocation (TET) dioxygenases that require ascorbate as an essential cofactor. There is limited data on ascorbate in gliomas and the relationship between ascorbate and 5-hmC in gliomas has never been reported. Clinical glioma samples (11 low-grade, 26 high-grade) were analysed for ascorbate, global DNA methylation and hydroxymethylation, and methylation status of the O-6-methylguanine-DNA methyltransferase (MGMT) promoter. Low-grade gliomas contained significantly higher levels of ascorbate than high-grade gliomas (p = 0.026). Levels of 5-hmC were significantly higher in low-grade than high-grade glioma (p = 0.0013). There was a strong association between higher ascorbate and higher 5-hmC (p = 0.004). Gliomas with unmethylated and methylated MGMT promoters had similar ascorbate levels (p = 0.96). One mechanism by which epigenetic modifications could occur is through ascorbate-mediated optimisation of TET activity in gliomas. These findings open the door to clinical intervention trials in patients with glioma to provide both mechanistic information and potential avenues for adjuvant ascorbate therapy.

Figure 1

Ascorbate content of clinical glioma tissue from 37 patients. (a) Correlation between ascorbate content normalised to gDNA or normalised to tissue weight. Pearson correlation, significance indicated. (b) Ascorbate levels were higher in grade I-III (LGG, n = 11) than grade IV (HGG, n = 26) gliomas; unpaired two-tailed t-test. (c) Ascorbate levels varied across gliomas with differing subtypes, including ganglioglioma (n = 1), ependymoma (n = 2), astrocytoma (n = 4), oligodendroglioma (n = 4), and glioblastoma (n = 26). (d) Tumour ascorbate levels were similar in female (n = 11) and male (n = 28) patients, and (e) ascorbate levels did not differ by age (below (n = 19) or above (n = 18) 60 years), unpaired two-tailed t-tests. Filled circle WHO grade I–III, Open circle WHO grade IV; median is indicated by a horizontal line; ns, no significance.

Figure 2

Relative levels of global cytosine species in clinical glioma samples according to tumour grade and ascorbate. Cytosine species include methylcytosine (5-mC) and hydroxymethylcytosine (5-hmC). Association between WHO grade I–III (n = 11) and grade IV (n = 25) tumours and (a) 5-mC and (b) 5-hmC levels. Association between below and above median ascorbate content (n = 18 each) and (c) 5-mC and (d) 5-hmC levels. (e) There was no association between 5-mC and ascorbate. (f) There was an association between 5-hmC and ascorbate. Individual samples (n = 36) are shown with the median as a horizontal line; Filled circle WHO grade I–III, Open circle WHO grade IV.

Figure 3

Infiltrating macrophages in glioma tumours. (a) Relative CD163 was assessed using Western blotting; a representative blot is shown. (b) High grade gliomas (HGG) had higher levels of CD163 than low grade gliomas (LGG), indicating a higher macrophage infiltrate. (c) Ascorbate content was similar between CD163 negative and positive tumours. (d) Hydroxymethylcytosine (5-hmC) was significantly lower in CD163 positive tumours, Mann Whitney test. Individual samples (n = 36) are shown with the median as a horizontal line; Filled circle WHO grade I–III, Open circle WHO grade IV; T, tumour; ns not significant.

Figure 4

MGMT promoter hypermethylation status of clinical glioma samples. (a) Gel electrophoresis images of unmethylated (93 bp) and methylated (81 bp) MSP products for each glioma sample. The assay included an unmethylated (U), a methylated (M) positive control and a negative (N) control without DNA. Bisulfite conversion of all samples was performed simultaneously and the positive control reaction was visualised (BC). N = 39, including the two repeat samples (#24 and #36). (b) Association of ascorbate levels (nmol/µg/DNA) between gliomas with either unmethylated or methylated MGMT promoters. (c) Association of global 5-hmC (% of total cytosine) with unmethylated/methylated MGMT promoter. Filled circle WHO grade I-III, Open circle WHO grade IV, Square samples with apparent low methylation; median is shown as horizontal line, n = 38.