Defects in DNMT expression and function can cause global hypomethylation that correlates with genomic instability and tumor progression due to loss of silencing at the loci of oncogenes (Soes et al., 2014; Kushwaha et al., 2016; Sheaffer et al., 2016). One of the most frequently observed upregulated oncogenes in cancer that exhibits a hypomethylated gene locus and is known to cause genomic instability is the transcription factor c-Myc. Large genomic amplifications induced by c-Myc overexpression are attributed to replicative stress caused by unscheduled DNA replication and increased oxidative stress leading to DNA breaks (Dominguez-Sola et al., 2007; Kuzyk and Mai, 2014). While c-Myc DNA hypomethylation is observed in many different cancers, other examples of hypomethylated oncogenic loci are cancer-specific, such as the melanoma-associated antigens (MAGE) gene that is hypomethylated specifically in gastric cancer and colorectal cancer (Honda et al., 2004; Kim et al., 2006) or synuclein γ (SNCG) in breast and ovarian cancers and solid tumors of the liver, gastric and (Gupta et al., 2003; Liu et al., 2005). Overexpression of these oncogenes promote, proliferation, metastasis, disruption of mitotic checkpoints, enhanced transcriptional activity and accelerated rates of chromosomal instability.

Hypomethylated repetitive DNA sequences, including CG repeats in CGIs and in transposable elements (TEs), are commonly found in cancers that could be caused by loss of function in DNA methyltransferase activity. Hypomorphic activity of Dnmt1 in mice induces thymic lymphomas with recurring insertions of a transposable element within the Notch gene leading to its oncogenic activation (Howard et al., 2008). LINE-1 and Alu elements are retrotransposons that constitute over 30% of the human genome and are normally inhibited by DNA methylation yet become frequently hypomethylated in cancer (Daskalos et al., 2009; Pehrsson et al., 2019). Increased LINE-1 and Alu activity causes genomic instability by promoting deletions, chromosome breaks, and translocations (Mandal et al., 2013; Bakshi et al., 2016).

DNA hypomethylation is also observed at microsatellite and pericentromeric heterochromatic regions in chromosomes of DNMT3B-mutant disease. DNMT3B loss of function mutations have been described in ∼60% of patients with an autosomal recessive disorder known as immunodeficiency, centromeric instability and facial anomaly (ICF) syndrome (Hansen et al., 1999; Okano et al., 1999; Xu et al., 1999). ICF is a rare disease showing symptoms in early childhood of recurrent gastrointestinal and pulmonary infections as a result of agammaglobulinemia (Ehrlich et al., 2008; Hagleitner et al., 2008). ICF patients display evident decondensation of heterochromatin caused by DNA hypomethylation at pericentromeric regions in chromosomes 1, 9, and 16, causing upregulated gene expression at these loci, chromosome breaks, and rearrangements in radial structures that are detectable in stimulated lymphocytes and increase the risk of hematological malignancies (Brown et al., 1995; Kaya et al., 2011; Gossling et al., 2017).

Cancer cells on average display up to 30% genome-wide losses in DNA methylation (Ehrlich and Lacey, 2013; Baylin and Jones, 2016) which could significantly impact three-dimensional chromatin architecture and genomic stability by altering oncogene activity and activation of TEs. CTCF, in cooperation with cohesin proteins, regulates long-range looping interactions between enhancers and promoters (Dixon et al., 2012) and preferentially binds at hypomethylated CpGs of cis-regulatory insulator regions of the genome (Nanavaty et al., 2020). DNA hypomethylation could allow for aberrant gains in CTCF binding, creating stronger insulation at oncogenic super-enhancers and an increased frequency of tandem duplications (Gong et al., 2018). Mis-regulated gene expression via stronger TAD boundaries could also cooperate with oncogenic transcription factors to exacerbate genomic instability (Fang et al., 2020). Thus, hypomethylated states may confer an advantage of cancer cells by the increase and maintenance of elevated oncogenic expression during tumor development (Kang et al., 2015).

Hypermethylation of the genome is associated with aberrant silencing of many tumor suppressors that could drive or accelerate carcinogenesis. DNA hypermethylation can be caused by a decrease in the activity of DNA demethylases such as the TET proteins leading to the silencing of DNA repair enzymes and tumor suppressors. Silencing due to aberrant DNA hypermethylation at imprinted loci (Rainier et al., 1993; Moulton et al., 1994; Steenman et al., 1994; Plass and Soloway, 2002), senescence genes (Kamb et al., 1994; Zhao et al., 2016; Xie et al., 2019) and lineage specific transcription factors, that normally slow proliferation and drive differentiation, reviewed in (Suelves et al., 2016; Pfeifer, 2018), may cause a buildup in unchecked DNA damage and DNA replication errors.

Methylated cytosines are approximately fivefold more likely to undergo mutagenesis than unmethylated cytosines (Poulos et al., 2017). This could be due to a greater tendency for 5mC to spontaneously deaminate compared to unmethylated cytosines (Supek et al., 2014; Blokzijl et al., 2016). Furthermore, C > T and G > A mutations disproportionally affect CpG dinucleotides. In one study, approximately 18% of C > T and G > A missense and nonsense mutations in genes of inherited human diseases occurred within CpGs, a frequency 10-fold higher would have been expected by chance alone (Cooper et al., 2010). Cytosine deaminases (AID and APOBEC1-3) also exhibit greater activity toward 5mC than 5hmC (Nabel et al., 2012; Rangam et al., 2012), and 5mC deamination generates T:G mismatches that can potentially recruit error-prone mismatch repair (MMR) complexes (Cortellino et al., 2011). Importantly, C > T mutation rates are 50% lower for C or 5hmC compared to 5mC (Tomkova et al., 2016). Given that C > T transition mutations are the most frequent, age-associated mutation signature in cancer (Alexandrov et al., 2013; Kandoth et al., 2013) loss of TET activity could mimic a premature aging phenotype with regard to the frequency at which C > T mutations accumulate. Distinct mutational signatures are also caused by aberrant deaminase activity, with AID/APOBEC-driven signatures dominated by high levels of C > T transitions in specific cancers (Alexandrov et al., 2020). Upregulated oncogenic activity of AID/APOBEC deaminases, in combination with elevated levels of 5mC in the genome, could synergistically accelerate malignant transformation.

Studies in ESCs have shown that loss of TET function causes decreased 5hmC and DNA hypermethylation in telomeric regions, that leads to telomere shortening, reduced telomere recombination and chromosome segregation defects (Lu et al., 2014; Kafer et al., 2016; Yang et al., 2016). Sub-telomeres are also hypermethylated in TET depleted ESCs, which may further impede telomere elongation by recombination (Yang et al., 2016). DNA hypermethylation caused by loss of TET activity increases nucleosome occupancy, and subsequently, loss of CTCF binding and a block in the downstream recruitment of the cohesin complex, reducing the formation of CTCF/cohesion-mediated chromatin loops thereby downregulating the expression of neighboring genes (Wiehle et al., 2019; Nanavaty et al., 2020). Loss of CTCF itself can also induce DNA hypermethylation at CTCF binding sites and the fusion of TAD boundaries that establish oncogenic expression patterns and increased cancer progression (Akdemir et al., 2020; Damaschke et al., 2020). TETs also regulate chromosomal architecture by protecting large undermethylated genomic regions known as DNA methylation “canyons” from becoming hypermethylated (Wiehle et al., 2016; Zhang X. et al., 2016). Loss of Dnmt3a or Tet2 causes canyon edges to collapse and become hypermethylated, suggesting that de novo methyltransferase activity and Tet2-mediated hydroxylation of 5mC work together to maintain 5hmC and hypomethylation at these loci (Jeong et al., 2014; Zhang X. et al., 2016).

Studies using ESCs with a triple knockout of all three TET enzymes revealed an increase of 5mC signal across all chromosomes by whole-genome bisulfite sequencing (Lu et al., 2014) and an increased frequency in telomere loss and chromosomal fusion (Yang et al., 2016). Genomic instability has also been observed in Tet2/3 double knockout hematopoietic stem and progenitor cells (HSPCs) and immature myeloid cells, which show impairment of DNA repair responses, lower homologous recombination (HR) or non-homologous end-joining (NHEJ) gene expression and spontaneously accumulate DNA double strand breaks (DSBs) marked by increased phosphorylated histone H2A.X (γH2AX) in comparison to WT cells (An et al., 2015). These studies, in combination with numerous reports of TET deficiency in other solid tumors (Haffner et al., 2011; Kudo et al., 2012; Yang et al., 2013; Munari et al., 2016) demonstrate the loss of 5hmC by defective TET enzymatic activity as a hallmark of cancer that may drive genomic instability.

The effect of vitamin C treatment on cell viability and genomic stability under steady-state conditions or in response to DNA damage can vary depending on developmental stage or normal versus malignant cell context. Vitamin C treatment can promote survival and maintain proliferation while protecting stem cells from DNA damage and senescence, whereas in cancer cells, the reactivation of tumor suppressors can reduce cell viability and alter sensitivity to therapeutic agents. Given that TET enzymes play a primary role as tumor suppressors, the ability to enhance 5hmC formation and DNA hypomethylation by vitamin C is of great interest in the maintenance of genomic stability, and cancer prevention and treatment (Figure 2B).

CpG islands in the promoters of DNA damage response genes, such as the tumor suppressor p16^INK4a and p21, are frequently hypermethylated in cancer (Roman-Gomez et al., 2002; Zhao et al., 2016; Ocker et al., 2019). The p16 and p21 proteins belong to a family of cyclin–dependent kinase (CDK) inhibitors, also known as CDKN2A and CDKN1A, respectively, that bind and inactivate CDKs to block cell cycle progression. Reactivation of their expression by vitamin C has been shown to induce senescence, apoptosis and halt proliferation of cancer cells. Vitamin C treatment of human skin and colon cancer cells increases 5hmC formation and reduces 5mC in the promoter CGIs of p16^INK4a and p21 that correlates with upregulated expression (Lin et al., 2014; Gerecke et al., 2018). Vitamin C also upregulates p53 and p21 protein expression in other cancer cells, such as oral squamous cell (Zhou et al., 2020) and hepatocellular carcinoma (Lv et al., 2018) causing growth arrest and apoptosis, however, in these studies the effect on 5hmC/5mC levels at these gene loci was not measured.

The studies in cancer cells on the effect of vitamin C in regulating p16^INK4a and p21 expression seem at odds with the observation in ESCs, iPSCs and tissue specific stem cell cultures in which vitamin C treatment silences the p16^INK4a/p19^ARF locus and is associated with reduced expression of p53 and p21 (He et al., 2008; Li et al., 2009; Esteban et al., 2010; Li Y. et al., 2016; Zhang P. et al., 2016). Furthermore, TET activity at promoter CGIs in stem cells may be counteracted by vitamin C-mediated co-activation of the H3K36 demethylase JHDM1b (KDM2B), which removes gene body H3K36me2/3 marks leading to p16^INK4a/p19^ARF gene silencing (He et al., 2008; Tzatsos et al., 2009). These different responses highlight the context specific effect of vitamin C in the regulation of senescence in normal stem cells compared to cancer cells. Overexpression of the reprogramming factors such as OCT4, SOX2, KLF4, and MYC increase ROS production by up to threefold compared to controls which could induce oxidative damage and trigger the premature senescence of non-cancerous stem cells (Krishnamurthy et al., 2004; Li et al., 2009). Vitamin C antioxidant effects can mitigate the high levels of ROS induced during reprogramming to prevent oxidative damage, thereby blocking ROS-induced senescence mechanisms (Krishnamurthy et al., 2004; Li et al., 2009).

Vitamin C treatment also induces the expression of a TET2-dependent gene expression signature in human leukemia cell lines and primary murine HSPCs involved in BER such as GADD45, PARP, and DNA glycosylases (Cimmino et al., 2017). The ability of vitamin C to drive increased BER activity could be a direct consequence of the increased need to actively remove and replace TET catalyzed oxidation products of mC in the DNA.

DNA methyltransferase inhibitors (DNMTis), such as 5-azacytidine and decitabine, are cytidine analogs used for the treatment of hematological malignancies by inducing global DNA hypomethylation (Hackanson et al., 2005; Shadduck et al., 2007; Santos et al., 2010). Restoring TET function by vitamin C administration, in combination with DNMTi therapy, may help erase DNA hypermethylation at tumor suppressor loci to promote differentiation and cell death. In addition, studies in a variety of cancers cells have shown that DNA hypomethylation by DNMTis causes the increased expression of endogenous retroviruses (ERVs) that mimic a viral infection and trigger an innate immune response leading to apoptosis (Chiappinelli et al., 2015; Roulois et al., 2015; Liu et al., 2016). Vitamin C was shown to synergize with DNMTi treatment in a TET2-dependent manner to increase 5hmC, and drive DNA hypomethylation to further increase ERV expression and enhance apoptosis of leukemia and solid tumor cell lines (Liu et al., 2016). Interestingly, ERVs in ESCs treated with vitamin C have been shown to acquire and retain high levels of 5hmC that do not become demethylated for several days, in contrast to the rapid and dynamic hydroxymethylation and demethylation induced at CGIs in promoters and TSS sites of hypermethylated pluripotency and blastocyst-like genes (Blaschke et al., 2013). This distinction may be a key factor in how vitamin C treatment, by promoting TET activity, can induce the expression of ERVs while preventing genomic instability that would be caused by a lack of 5mC or 5hmC at these retrotransposable elements (REs). REs are one of the two classes of TEs, which unlike the cut and paste mechanisms of DNA transposons, insert themselves by duplicating elements into a new genomic location via an RNA intermediate (Friedli and Trono, 2015; Chuong et al., 2016). In ESCs, TET2 has been shown to be recruited to ERV loci by an ERV RNA-binding protein Paraspeckle Component -1 (PSPC1) to regulate the expression of adjacent genes during embryonic development (Guallar et al., 2018). Maintenance of 5hmC levels at ERVs by vitamin C may therefore reduce the risk of insertional mutagenesis while exploiting these cis-regulatory sequences for transcriptional control of neighboring genes.

Human telomeres shorten by 20–200 bp per cell division (Harley et al., 1990) and telomere length in circulating blood cells has been used as a biomarker of human aging (von Zglinicki and Martin-Ruiz, 2005). Higher intake of antioxidants by multivitamin supplementation or high vegetable intake is associated with increased telomere length in human studies (Xu et al., 2009; Marcon et al., 2012; Tsoukalas et al., 2019), however, direct evidence that vitamin C supplementation alone can suppress telomere attrition in vivo requires further study. Using in vitro cellular models derived from human iPSCs, vitamin C treatment has been shown to increase telomerase activity and the expression of genes encoding telomerase-related RNA and protein components that protect telomere stability (Wei et al., 2012; Kim et al., 2013). Patients with Werner Syndrome (WS) harbor a mutation in the WRN gene that leads to loss of telomere maintenance, premature aging and increased cancer rates (Burtner and Kennedy, 2010; Zhang et al., 2015). Vitamin C treatment of a human WS mesenchymal stem cell model was shown to slow down telomere loss and downregulate the senescence-inducing p16^Ink4a protein (Zhang et al., 2013, 2015). In ESCs, the protection and maintenance of telomere length by vitamin C has not yet been correlated directly to its role as a TET coactivator; however, given that mouse ESCs with TET deficiency exhibit shorter telomeres, chromosomal instability, sub-telomere DNA hypermethylation and reduced telomere recombination (Lu et al., 2014; Kafer et al., 2016; Lazzerini-Denchi and Sfeir, 2016; Yang et al., 2016), vitamin C most likely plays a direct role in TET-mediated telomere maintenance.

Vitamin C is required for the optimal activity and demethylation capacity of multiple KDMs that hydroxylate and remove mono-, di-, or trimethyl-lysines in histones (Klose et al., 2006; Tsukada et al., 2006). During somatic cell reprogramming, vitamin C-mediated removal of histone modifications such as H3K4me3 and H3K36me3, potentially by KDM5 and KDM2, respectively, can turn off the expression of genes, while demethylation of H3K9me2/me3 by members of the KDM3/4/7 family can remodel heterochromatin regions that can facilitate DNA hypomethylation at these otherwise silenced loci (Li et al., 2009; Tzatsos et al., 2009; Esteban et al., 2010; Huang et al., 2019). Histone modifications and KDMs have a wide variety of roles in regulating genomic stability that could potentially be influenced by vitamin C with different biological outcomes in either stem cells or cancer cells. Suppression of H3K9me2/me3 demethylase activity such as KDM3A (JMJD1A), KDM4C (JMJD2C), and KDM7C (PHF2) in tissue specific stem cell cultures and cancer cells can lead to premature senescence, increased DNA damage and genomic instability (Huang et al., 2019; Pappa et al., 2019; Fan et al., 2020). Vitamin C can coactivate KDM5B/C to demethylate H3K4me3, and KDM5B (JARID1B) has been shown to increase DNA DSB repair by recruiting factors Ku70 and BRCA1 in osteosarcoma (U2OS) cells (Li et al., 2014). A deficiency in KDM5B was shown to disengage the DNA repair process, promote spontaneous DNA damage, activate p53 signaling, and sensitize cells to genotoxic insults (Li et al., 2014). Demethylation of H3K4me3 by KDM5C (JARID1C) leads to heterochromatin formation, and renal cell carcinoma patients with JARID1C mutations exhibit genome-wide DNA hypomethylation, increased genomic rearrangements, and an overall worse prognosis (Rondinelli et al., 2015). In these settings, vitamin C could enhance KDM activity to prevent genomic instability. However, other KDM activity, such as overexpression of KDM2B (JHDM1B) and KDM4A (JHDM3A) that target H3K36 for demethylation could promote increased DNA damage through decreased HR repair (Pfister et al., 2014; Staberg et al., 2018).

HIF-1 and HIF-2 are transcription factors induced in low oxygen (hypoxic) environments typical of tumor niches, and are repressed in normoxic conditions by prolyl hydroxylases (PHDs) (Rankin and Giaccia, 2016). Vitamin C has been shown to function as a direct cofactor of HIF PHDs (Osipyants et al., 2018) that catalyze the hydroxylation of HIF-1α (Appelhoff et al., 2004). HIF-1α is involved in microsatellite instability and mismatch repair deficiency in a colon cancer model (Koshiji et al., 2005) and induces transcriptional changes leading to the downregulation of several DNA damage repair genes in an oral squamous cell carcinoma model (Nakamura et al., 2018). Treatment with vitamin C could promote the degradation of HIF-1α by stimulating HIF prolyl hydroxylases that would restore DNA repair activity and remove HIF-1 α mediated protection from genomic instability (Knowles et al., 2003; Zhao et al., 2015). HIF stabilization is also associated with therapeutic resistance to DNA damage. Hypoxic mouse stromal cells are shown to be more resistant to irradiation than the same cells under normoxic conditions in a HIF-1 dependent manner (Calvo-Asensio et al., 2018), and a lung cancer model expressing a HIF-1-stabilizing micro RNA exhibits a hypoxic phenotype and increased radioresistance (Grosso et al., 2013). Inhibition of leukemia cell growth by vitamin C treatment also correlates with the downregulation of HIF-1α mRNA expression (Kawada et al., 2013). Vitamin C could therefore target the PHD family of α-KGDDs to maintain genomic stability and counteract hypoxic tumor conditions to influence disease progression or sensitivity to DNA damage.

In mammalian cells, homologues of the E. coli DNA dealkylation enzyme AlkB act as RNA demethylases, catalyzing the oxidative demethylation of N⁶-methyladenosine (m⁶A) (Fedeles et al., 2015). Nine mammalian homologues have been identified, ALKBH1-8 and the fat mass and obesity associated protein (FTO), with each displaying unique roles in genomic stability (Fedeles et al., 2015). ALKBH5 and FTO are the most well characterized of the AlkB homologs and their silencing or deletion reduces the expression of a number of HR and other repair genes in glioblastoma stem cells (Kowalski-Chauvel et al., 2020) and osteoblasts (Zhang et al., 2019), increasing their susceptibility to genotoxic damage. Interestingly, ALKBH2 and ALKBH3 have been shown biochemically to act as DNA repair enzymes that oxidize 5mC to generate 5hmC, 5fC, and 5caC, similarly to TET enzymes (Bian et al., 2019). However, in certain cancer cell models, ALKBH activity may promote survival or resistance to chemotherapy. ALKBH2 knockdown sensitize lung cancer cells to cisplatin (Wu S.S. et al., 2011) and glioblastoma cells to temozolomide (Johannessen et al., 2013) and ALKBH8-knockout in mouse embryonic fibroblasts elevates ROS levels that could promote DNA damage (Endres et al., 2015). While vitamin C was shown to stimulate oxoglutarate turnover in E. coli AlkB (Welford et al., 2003), studies on the role of vitamin C in the regulation of mammalian ALKBH proteins are lacking. Given the recent confirmation that vitamin C binds directly to FTO (Wang et al., 2020) and the diverse roles of m⁶A in the regulation of gene expression in cancer (He et al., 2019) the role of vitamin C in m⁶A demethylation and the implication on genomic stability and cancer therapy requires further investigation.

Maximal plasma concentrations from oral vitamin C doses above 500 mg/day do not exceed 150 μM, due to homeostatic down-regulation of vitamin C transporters in enterocytes and kidney cells, leading to reduced absorption and increased urinary excretion (Young et al., 2015). High-dose vitamin C (up to 100 g) administered by intravenous (IV) infusion can bypass the limited oral bioavailability without reported toxicity, and raise plasma concentrations to high millimolar levels that remain above 100 μM for up to 6 h (Padayatty et al., 2004). An intriguing effect of high-dose vitamin C treatment in cancer cells is its ability to create ROS and act as a pro-oxidant.

Vitamin C in the form of ascorbate, under normal growth conditions, behaves as an anti-oxidant that donates electrons to quench damage-inducing free radicals, becoming a relatively stable ascorbate radical while to protecting cells from lipid, protein, and DNA oxidation (Padayatty et al., 2003). In contrast to the effect vitamin C has in reprogramming and stem cell cultures, which is to reduce ROS, increase proliferation and protect cells from apoptosis, vitamin C treatment of cancer cells correlates with an increase in ROS-mediated oxidative stress and reduced viability (Chen et al., 2008; Noguera et al., 2017; Ghanem et al., 2020; Wu et al., 2020; Zhou et al., 2020). Moreover, it has been demonstrated that while diverse cancer cell types are sensitive to low- and sub-millimolar doses of vitamin C, even with high millimolar doses, limited cytotoxicity is observed in normal tissue (Chen et al., 2008; Noguera et al., 2017). This points to a unique synthetic lethality of cancer cells to high-dose vitamin C treatment that can be exploited for cancer therapy. In a Burkitt’s lymphoma model, toxicity to ascorbate was associated with ROS-induced peroxide formation that could be abrogated by catalase treatment (Chen et al., 2005). Follow up work by the same group showed elevated ascorbate radical and peroxide levels in vivo in the microenvironment of three separate models where vitamin C reduced tumor burden in mice (Chen et al., 2008). Another study showed mitochondrial hyperpolarization associated with high-dose ascorbate and ROS production (Ghanem et al., 2020).

The pro-oxidant effects of vitamin C in these studies has been attributed to excess uptake and formation of redox active iron. In a series of reactions, ascorbate can reduce ferric (Fe³⁺) iron to ferrous (Fe²⁺), which can then be returned to the ferric state after reducing O₂ to superoxide. Superoxide reacts with itself and protons to generate hydrogen peroxide and O₂, and the hydrogen peroxide can be reduced by Fe²⁺ in a Fenton reaction to form hydroxide radicals that can create the pro-oxidant killing effects observed in cancer cells (Du et al., 2012). Another possibility is that high-dose vitamin C can exhaust the primary cellular antioxidant GSH when taken up in higher amounts by more metabolically active cells, thus rendering cancer cells more vulnerable to oxidative stress. In this capacity, high-dose vitamin C was shown to be selectively toxic to KRAS or BRAF mutant colorectal cancer cells due to increased cellular uptake of oxidized vitamin C (DHA) through upregulated GLUT transporters, which led to GSH depletion and lethal levels of ROS (Yun et al., 2015).

Multiple cellular and animal models and recent clinical trials have shown that high-dose vitamin C treatment can reduce cancer cell viability and improve treatment outcome in combination therapies. In patients with ovarian cancer, high-dose intravenous vitamin C (IVC) administered in combination with carboplatin and paclitaxel enhanced chemosensitivity and reduced adverse side effects of chemotherapy (Ma et al., 2014). Glioblastoma patients receiving IVC in combination with radiation therapy and temozolomide have also shown improved survival compared to previous studies (Schoenfeld et al., 2017). Another notable adjuvant use of vitamin C was with PARP inhibitor treatment for a small group of patients with HR-deficient advanced metastatic cancers (primarily BRCA mutations) where partial or complete responses were observed in all eight patients (Demiray, 2020).

High-dose parenteral vitamin C treatment can be modeled in mice via IP administration, where 4 g/Kg delivered IP can induce mM concentrations in plasma similar to pharmacokinetic studies in patients treated with high-dose IVC (Padayatty et al., 2004; Yun et al., 2015). Vitamin C treatment of Tet2-deficient mouse HSPCs and patient-derived AML leads to increased 5hmC formation, DNA hypomethylation and a block in aberrant self-renewal in vitro with suppression of disease progression in vivo (Cimmino et al., 2017). High-dose vitamin C also enhances immune checkpoint inhibitor therapy, that in animals models was shown to be curative when used to treat tumors with high mutational burdens such as mismatch repair–deficient (MMRd) or microsatellite instable (MSI) breast and colorectal mouse tumors that otherwise would be unresponsive to single immune checkpoint inhibitors (Magri et al., 2020). MSI cancers exhibit microsatellite DNA hypermethylation and accumulate 5mC mutations in CpG in the absence MMR that has been attributed to methylation-associated repair deficiencies (Ling et al., 2010; Poulos et al., 2017). While a role for TET mediated hypomethylation in microsatellite stability was not assessed, this may be one mechanism by which vitamin C treatment could play a role in MMR-deficient cancer treatment. Importantly the therapeutic efficacy of vitamin C in combination with checkpoint inhibitors was dependent on a functional immune system, suggesting that tumor cell-intrinsic and immune microenvironmental effects of vitamin C can work together to slow tumor growth.

Multiple studies continue to identify new potential combination cancer treatment strategies with vitamin C. In cellular and preclinical studies, high-dose vitamin C enhances the sensitivity of hematological cell malignancies to arsenic trioxide (Huijskens et al., 2016; Noguera et al., 2017) and increases chemosensitivity and radiosensitivity of various solid tumor cells including ovarian (Ma et al., 2014), pancreatic (Du et al., 2015) glioblastoma and non-small cell lung carcinoma cells (Schoenfeld et al., 2017). Vitamin C has also been shown to sensitize melanoma cells to bromodomain inhibitors (Mustafi et al., 2018). These studies argued in favor of combination therapies based primarily on the mechanism of oxidative stress generated by high-dose ascorbate, that could increase cancer cell susceptibility to standard therapies by exacerbating DNA damage. The effect of vitamin C as a co-factor of α-KGDDs and specifically as a regulator of DNA demethylation via increased TET function were not investigated. However, it has been shown that TET-mediated DNA oxidation induced by vitamin C can create a synthetic lethality, where AML cells being forced to undergo active DNA demethylation renders them hypersensitive to PARP inhibition (Cimmino et al., 2017). In this study, vitamin C treatment increases the expression of genes involved in BER that may be a response to the increased oxidation of 5mC and in combination with the PARP inhibitor, Olaparib, enhances the killing of human AML cells greater than either agent alone (Cimmino et al., 2017). Future studies on the use of vitamin C as a therapeutic agent would benefit from the inclusion of additional correlative studies such as DNA methylation changes, oxidized mC formation or even histone and RNA methylation levels in order to fully appreciate the cell-intrinsic or microenvironmental epigenetic biomarkers of its anti-cancer efficacy.