Vitamin C increases viral mimicry induced by 5-aza-2'-deoxycytidine

Abstract

Vitamin C deficiency is found in patients with cancer and might complicate various therapy paradigms. Here we show how this deficiency may influence the use of DNA methyltransferase inhibitors (DNMTis) for treatment of hematological neoplasias. In vitro, when vitamin C is added at physiological levels to low doses of the DNMTi 5-aza-2'-deoxycytidine (5-aza-CdR), there is a synergistic inhibition of cancer-cell proliferation and increased apoptosis. These effects are associated with enhanced immune signals including increased expression of bidirectionally transcribed endogenous retrovirus (ERV) transcripts, increased cytosolic dsRNA, and activation of an IFN-inducing cellular response. This synergistic effect is likely the result of both passive DNA demethylation by DNMTi and active conversion of 5-methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC) by ten-eleven translocation (TET) enzymes at LTR regions of ERVs, because vitamin C acts as a cofactor for TET proteins. In addition, TET2 knockout reduces the synergy between the two compounds. Furthermore, we show that many patients with hematological neoplasia are markedly vitamin C deficient. Thus, our data suggest that correction of vitamin C deficiency in patients with hematological and other cancers may improve responses to epigenetic therapy with DNMTis.

Keywords: 5-hydroxymethylcytosine; DNA methyltransferase inhibitor; endogenous retrovirus; epigenetic therapy; vitamin C.

Fig. 1.

The combination of 5-aza-CdR and physiological levels of vitamin C inhibits cell growth and enhances apoptosis. (A) Population-doubling time after treatment with PBS, vitamin C, 5-aza-CdR, or the combination. HCT116 and HL60 cells were treated with one dose of 300 nM 5-aza-CdR for 24 h; MCF7, SNU398, and HepG2 cells were treated with three consecutive daily doses of 5-aza-CdR at 100 nM, with drug withdrawal at 72 h. All cells were treated with daily doses of 57 μM vitamin C. Values are mean ± SD of three independent experiments. *P < 0.05 by paired Student’s t test comparing combination vs. 5-aza-CdR treatment. (B) Apoptosis analysis showing an increased percentage of apoptotic cells after combination treatment relative to either 5-aza-CdR or vitamin C alone. Cells were stained with annexin V-FITC and PI and were analyzed by flow cytometry.

Fig. 2.

Combination treatment of vitamin C and 5-aza-CdR up-regulates viral-defense genes. (A) Comparison of the expression of IFN-responsive genes upon combination treatment vs. 5-aza-CdR treatment in HCT116, HL60, and SNU398 cells. Treatment was as described in Fig. 1A. Transcripts were analyzed by microarray at 5 d after treatment. Orange indicates genes up-regulated in at least two cell lines (genes are listed at the right); gray denotes genes up-regulated in only one line. (B) A log2 plot of the expression of dsRNA defense genes after treatment with vitamin C, 5-aza-CdR, or their combination vs. untreated HCT116 cells. Values are calculated from fragments per kilobase per million fragments mapped and are averaged from two independent RNA-seq datasets.

Fig. 3.

Combination treatment up-regulates endogenous retroviruses in both ssRNA and dsRNA forms in HCT116 cells. (A) Log2 fold change (FC) of transcripts uniquely mapped to individual ERV loci, comparing vitamin C, 5-aza-CdR, and combination treatment vs. untreated cells. Diagonal shading shows the percentage of bidirectional transcripts. (B) Percentage of double-stranded ERV (dsERV) counts relative to total reads in HCT116 cells after PBS, 5-aza-CdR, or combination treatment. *P < 0.05 by χ² test for equality of proportions without continuity correction, compared with untreated cells. (C) Log2 fold change of dsERV levels comparing 5-aza-CdR and combination treatment (purple) relative to untreated cells. (D) The Venn diagram shows that ERVs that are bidirectionally transcribed largely overlap with ERVs that form dsRNA. The 50 most abundant bidirectionally transcribed ERVs and dsERVs in HCT116 cells after the combination treatment were determined by directional RNA-seq and dsRNA-seq, respectively.

Fig. 4.

Combination treatment promotes 5hmC production at ERV LTRs. (A) Dot blot analysis of global 5hmC levels in cancer cells on day 5 after treatment with an antibody against 5hmC. (B) Genomewide analysis of 5hmC at up-regulated ERV loci in HCT116 cells on days 1 and 5 after vitamin C and combination treatment. The top 100 most up-regulated ERVs after combination treatment compared with untreated cells were identified by RNA-seq, and 5hmC enrichment and were analyzed using the Hydroxymethyl Collector Kit (Active Motif) followed by next-generation sequencing. (Upper) The heat map shows the expression levels of individual ERVs in the unit of natural log read counts. (Lower) The average base counts of 5hmC reads at all ERV loci and their ± 3-kb flanking region. (C, Lower) Quantification of 5hmC in HCT116 cells at LTRs of ERVs on days 1, 3, and 5 after treatment by qPCR. (Upper) Primer design. The promoter region of β-actin (ACTB) was used as a negative control. Values are presented as mean ± SD of five independent experiments. *P < 0.05 by Student’s t test. (D) Expression of HERV-Fc1 and LTR12C transcripts in HCT116 cells at days 1, 3, and 5 after treatment relative to the TATA-binding protein (TBP) levels by quantitative RT-PCR. Values are mean ± SD of three independent experiments.

Fig. 5.

DHA elicits effects similar to those of vitamin C, but other antioxidants do not. (A) Dot blot analysis of global 5hmC levels with an antibody against 5hmC in HCT116 cells untreated or treated with 300 nM 5-aza-CdR for 24 h, daily doses of vitamin C or DHA at 57 μM, daily doses of GSH at 4.9 mM or DTT at 500 μM, and the combination of 5-aza-CdR with vitamin C, DHA, GSH, or DTT. (B and C) Expression levels of HERV-Fc1 (B) and LTR12C (C) by quantitative RT-PCR in HCT116 cells on day 5 after treatment. Values are presented as mean ± SD of three independent experiments.

Fig. 6.

TET2 is involved in ERV up-regulation upon combination treatment. (A) Dot blot analysis of global 5hmC levels in wild-type A2780 and TET2-KO cells after treatment. (B) Expression levels of HERV-Fc1 and LTR12C by quantitative RT-PCR in wild-type A2780 and TET2-KO cells on day 5 after treatment. Values are presented as mean ± SD of four independent experiments. *P < 0.05 by Student’s t test.

Fig. 7.

Plasma vitamin C concentrations in patients with hematopoietic malignancies. (A) The bar graph shows the distribution of plasma vitamin C levels in 24 patients with hematopoietic malignancies who were not taking vitamin C supplements. (B) The pie chart shows the percentage of patients with vitamin C deficiency (plasma level <11.4 μM), patients with levels below normal (11.4–26 μM), and patients with levels within the normal range (26–84 μM).