Vitamin C (L-ascorbic acid, ascorbate, VC) is an essential nutrient for the normal maintenance of cellular functions, such as neural pathways, molecule biosynthesis (e.g., collagen, norepinephrine), immune signaling, chromatin remodeling, and cell division
[1][2][3][4]. At physiological concentrations (40–80 µM in human plasma), VC acts as an antioxidant by serving as an electron donor and effectively scavenges reactive oxygen species (ROS). In contrast, at high doses (10–20 mM), VC acts as a pro-oxidant that induces oxidative stress and suppresses tumor growth, without notable damage to normal cells and tissues
[5][6]. Recently, multiple studies have uncovered that VC has multifaceted anti-tumor effects
[7]. For example, it acts as a cofactor for enzymes that regulate gene expression and suppresses oncogenes while reactivating tumor suppressor genes
[8][9][10]. It has also been reported that VC induces various forms of canonical cell death and may be involved in non-canonical mechanisms relating to energy crises resulting from ATP depletion
[11]. Aligned with its pleiotropic effects as a cancer-specific, pro-oxidative cytotoxic agent, anti-cancer epigenetic regulator, and immune modulator, high-dose VC has been proposed as a potent adjuvant treatment for cancer, acting synergistically with numerous standard (chemo-) therapies and alleviating the toxic side effects of chemotherapy (refer to for comprehensive reviews
[12][13] of this). Furthermore, emerging animal model studies suggest that VC enhances anti-tumor effects when used in combination with dietary intervention (intermittent fasting, IF)
[14] or cancer immunotherapy that facilitates anti-tumor immune environments
[15][16], suggesting that VC mediates beneficial anti-cancer effects by targeting both tumor-intrinsic and -extrinsic pathways
[17].
One of the most critical issues in cancer biology, as well as cancer diagnosis and treatment, is cancer cell plasticity, the adaptive and reversible capacity of diverse cancer cell populations to shift between cancer stem cell (CSC) and non-CSC/differentiated cell states in response to the tumor microenvironment
[18]. At the top of the heterogeneous tumor hierarchy, the CSC, a self-renewing and multi-potent cancer cell type, is responsible for tumor recurrence, metastasis, chemoresistance, and mortality
[18]. Phenotypically, CSCs are associated with the epithelial–mesenchymal transition (EMT), which confers cancer cells with increased motility and invasion ability that is characteristic of malignant and drug-resistant cells
[19]. A growing number of studies suggest that CSC heterogeneity and plasticity are influenced not just by genetic factors but also by non-genetic factors, including epigenetic pathways
[20], metabolic processes
[21], and tumor-microimmune environments
[22][23][24]. CSCs develop resistance to traditional chemotherapy due to their adaptable phenotype, which enables them to withstand therapy by overexpressing anti-apoptotic factors, defending against oxidative stress, and effectively repairing DNA damage
[25]. Unfortunately, the vast majority of anti-cancer drugs have been designed to target rapidly dividing non-CSCs, rather than dormant CSCs, frequently leading to tumor relapse and treatment ineffectiveness
[26][27]. Furthermore, many chemotherapy drugs trigger diverse mechanisms of plasticity in cancer cells, including EMT, autophagy, and metabolic reprogramming, contributing to the evolution of therapy-resistant tumors
[28]. While extensive pharmacological efforts have been made to specifically target CSCs, with the aim of eradicating this malignant and drug-resistant cell population
[29][30][31], many synthetic drugs have toxic effects on normal tissues and their use can be accompanied by several detrimental side effects on physiology and behavior
[32]. This underscores the need for safer and more targeted therapeutic approaches that can effectively eliminate CSCs without causing significant harm to normal tissues, thus minimizing adverse effects. In this context, there is a growing recognition that less toxic natural products, possessing anti-cancer stem cell (CSC) activities, such as flavonoids, FDA-approved drugs derived from natural sources, and nutritional herbs commonly employed in traditional Chinese medicine, hold promise as potential alternatives for addressing therapy-resistant cancers
[33][34].
2. Physiological and Anti-Tumor Activities of Vitamin C
Physiologically, VC exists largely in its reduced (ascorbic acid [AA]) or oxidized (dehydroascorbic acid [DHA]) forms, which, depending on its redox state, involves the loss or gain of two electrons
[7] (
Figure 1A).
Figure 1. Physiological and anti-cancer mechanisms of vitamin C activity. (
A) Physiological vitamin C (VC) exists largely in its reduced (ascorbic acid (AA)) or oxidized (dehydroascorbic acid (DHA)) forms, determined by either the gain or loss of two electrons and two hydrogens (reduction: +2e
− +2H
+; oxidation: −2e
− −2H
+). (
B) Pharmacological VC can induce cancer cell death through two complementary mechanisms that elevate oxidative stress. Following VC treatment, hydrogen peroxide (H
2O
2) is produced in the extracellular environment by AA oxidation via Fenton chemistry that is facilitated by the presence of labile ferric iron (Fe
3+) that enters cancer cells from the tumor microenvironment through either aquaporins or passive diffusion. VC enters cells through sodium-dependent vitamin C transporters (mainly SVCT2) when it is in its reduced form (AA), or via glucose transporters (mainly GLUT1) when it is in its oxidized form (DHA). Once inside the cell, dehydroascorbic acid (DHA) is rapidly converted to ascorbic acid (AA) through the action of the reducing agent glutathione (GSH). This process depletes the intracellular glutathione, resulting in elevated levels of intracellular H
2O
2 and several detrimental effects, including DNA damage, lipid peroxidation, and protein oxidation. In particular, DNA damage triggers the activation of the DNA repair enzyme poly (ADP-ribose) polymerase (PARP), which depletes cellular NAD
+ levels. This depletion, in turn, inhibits the activity of glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and glycolysis in cancer cells, resulting in decreased ATP production and cell death. (
C) VC plays a pivotal role in numerous biological processes by serving as a cofactor for Fe
2+ and alpha-ketoglutarate-dependent dioxygenases (Fe
2+/α-KGDDs). These enzymes encompass a range of proteins, including collagen prolyl hydroxylases (CP4H), JmjC histone demethylases (JHDMs), ten–eleven translocation (TET) DNA hydroxylases, and hypoxia-inducible factor (HIF) hydroxylases (such as proline hydroxylase domain proteins (PHDs), and asparagine hydroxylase (factor-inhibiting HIF [FIH])). These enzymes have diverse functions, such as regulating collagen synthesis to maintain skin tissue and extracellular matrix (ECM) integrity as well as to facilitate efficient wound healing. They can also promote histone and DNA demethylation, thereby enhancing induced pluripotent stem cell (iPSC) reprogramming and suppressing leukemia progression. Furthermore, they can modulate various responses under low-oxygen conditions (hypoxia). This figure was created using BioRender, with modifications inspired by
[42][43][44].
Notably, VC exhibits several drug-like properties that make it a promising therapeutic agent. VC has a low molecular weight (<500 Da, 176.12 g/mol), is water-soluble, and has a high oral bioavailability. The bioavailability of VC in foods is generally considered equivalent to the purified form within the recommended nutritional range of 15–200 mg
[45]. Yet, this bioavailability diminishes by over 50% with higher amounts, such as doses exceeding 1000 mg. Since VC was first chemically synthesized in 1933, the bioavailability of synthetic and natural VC has been a subject of extensive research
[46]. Animal studies indicate varying bioavailability between synthetic and natural VC, depending on the study design and animal model. In contrast, human studies have consistently shown no significant differences in their bioavailability
[47]. While synthetic and natural VC share the same molecular and physicochemical properties, it has been reported that fruits and vegetables offer a wealth of micronutrients, dietary fiber, and phytochemicals that can modulate the absorption and utilization of VC
[47][48]. Furthermore, VC is highly susceptible to oxidation and degradation, particularly in biological fluids such as plasma and blood
[47][48]. The stability of VC in these fluids is influenced by several factors that can occur during processing (e.g., heat and light, pH, metal ions) and storage (e.g., temperature, oxygen exposure). Particularly, in vivo VC levels are determined by a balance between uptake, metabolism, and excretion (refer to
[49] for a comprehensive review of these processes). Notably, the conversion of ascorbic acid to DHA in foods or the gastrointestinal tract can diminish the bioactivity of VC
[49]. To enhance the chemical stability and bioavailability of VC, various chemically synthesized ascorbic analogs, such as ascorbate 2-sulfate, ascorbate 2-monophosphate, and ascorbate 2-triphosphate, have been developed
[45]. Additionally, encapsulating VC in specific nanoparticles has been shown to improve stability during storage and delivery
[45].
The transport of the reduced form of VC (AA) occurs through specialized transporters known as sodium-dependent vitamin C transporters (SVCT) 1 and 2, which are conserved across mammalian species, including humans
[50][51][52][53] (
Figure 1B). Notably, it has been reported that SVCT2 is a key protein for VC uptake in both normal
[54][55] and cancer cells
[37][56]. Moreover, VC has been shown to inhibit breast cancer cell growth
[56] or preferentially kill CSC populations in live cancer in an SVCT2-dependent manner
[37]. In contrast, transport of the oxidized form of VC (DHA) into cells is primarily facilitated by glucose transporters known as GLUTs (GLUTs 1–4 and 8)
[57][58][59][60]. According to the prevailing model, DHA is the most active anti-cancer form of VC in tumors, as it generates cytotoxic reactive oxygen species (ROS) upon its intracellular conversion to AA following its entry into cells
[61][62]. However, recent studies involving direct treatment of various cancer cell lines, including human breast cancer and neuroblastoma cells, with DHA have consistently shown that DHA has minimal or no significant impact on cell death
[42][43][63]. This suggests the possibility that the cytotoxic responses to DHA can vary depending on the specific cancer cell type or experimental conditions. Taken together, these results underscore the need for further studies in a wide range of cancer cell types to explore how the distinct redox forms of VC contribute to VC-induced cell death.
Numerous preclinical investigations of various human cancer models have indicated that the extracellular generation of hydrogen peroxide (H
2O
2) is a pivotal factor in the anti-cancer efficacy of high-dose VC
[13][63]. AA readily oxidizes to DHA through a two-electron oxidation process in the presence of catalytic metals, like copper (Cu
+/Cu
2+) and iron (Fe
3+/Fe
2+), leading to elevated H
2O
2 concentrations in the extracellular space of tumors
[7][64][65]. The H
2O
2 can subsequently permeate cells by utilizing peroxiporins within the plasma membrane to exert its influence on redox-dependent signaling and metabolic pathways pertinent to the viability of cancer cells, including the pathways regulating processes such as cell-cycle arrest, DNA damage, and apoptosis
[66] (
Figure 1B). However, the role of iron in the anti-cancer action of VC has recently been debated, with varying findings in different in vitro studies. Some investigations have observed that reducing or depleting intracellular iron levels enhances the growth inhibition and apoptosis induced by VC in neuroblastoma and K562 leukemic cells
[67][68], while others have reported that extracellular iron diminishes the anti-cancer effects of VC in PC-3 and LNCaP prostate cancer cell lines
[69]. More recently, it has been documented that exogenous iron impairs the anti-cancer effects of VC in specific cancer cell lines, both in vitro and in vivo
[70]. These findings suggest that the impact of iron on VC-induced cytotoxicity may vary depending on the cell type or experimental method employed, such as inhibiting intracellular iron using iron chelators or exogenous iron treatment, necessitating further exploration in other cancer types.
Notably, the mechanisms of cell death underlying the anti-cancer effects of VC have undergone extensive investigation. Previous research has suggested that pharmacological VC can trigger various forms of cell death, including apoptosis, necroptosis, and autophagy, with the specific outcome contingent upon the concentration and cell type employed in the experiment
[43]. Earlier studies have also indicated that VC-induced cytotoxicity is primarily mediated through caspase-dependent apoptosis or necrosis, based on assessments of changes in the protein levels of key cell death effectors, such as caspases, BAX, BID, and receptor-interacting protein kinase (RIPK1), in response to VC exposure
[71][72][73]. However, emerging evidence indicates that classical inhibitors of apoptosis or necrosis, such as the pan-caspase inhibitor Z-VAD-FMK and the RIP1-targeted necroptosis inhibitor Nec-1, do not prevent the cell death induced by pharmacological VC, suggesting that there are non-canonical cell death mechanisms at play
[74][75]. Recently, non-apoptotic forms of cell death, such as ferroptosis, parthanatos, and pyroptosis, have garnered attention as promising targets for cancer therapy with natural or synthetic compounds that induce ROS
[76]. These findings suggest that the cytotoxic effects induced by high-dose VC may entail multiple cell death pathways operating synergistically, rather than a single pathway.
In addition to cell death mechanisms, metabolic crises are a recurring phenomenon in cancer cell death triggered by pharmacological VC treatment
[18]. Consistent with the Warburg hypothesis, which indicates cancer cells’ preference for glycolysis over oxidative phosphorylation for energy production, ATP depletion and cell demise in response to VC are primarily attributed to the hindered glycolysis caused by the VC-induced H
2O
2-mediated inhibition of glyceraldehyde 3-phosphate dehydrogenase (GAPDH) activity
[43][62]. In this context, recent studies have proposed a model suggesting that VC-induced H
2O
2 inflicts DNA damage, consequently promoting poly (ADP-ribose) polymerase (PARP) activation, which, in turn, consumes NAD and depletes ATP through the reduction of GAPDH activity and glycolysis
[43][61][77][78]. However, subsequent investigations have reported that treatment with the PARP inhibitor Olaparib maintains NAD
+ and ATP levels but results in increased DNA double-strand breaks and does not prevent ascorbate-induced cell death
[43]. This suggests that the PARP-associated DNA damage response may not be the exclusive cause of this cytotoxicity, implying that supplementary mechanisms contribute to the NAD
+- and ATP depletion-dependent cytotoxicity of VC treatment
[43] (
Figure 1B).
In addition to its pro-oxidant properties, several studies have uncovered additional VC-mediated anti-tumor mechanisms involving epigenetic and post-translational pathways (
Figure 1C). VC has been found to serve as a key cofactor that catalyzes the activity of various iron-containing dioxygenase enzymes, such as ferrous iron Fe
2+ and α-ketoglutarate (αKG)-dependent dioxygenases (Fe
2+/α-KGDDs), that play diverse roles in many biological processes, including the regulation of metabolic adaptations to hypoxia, the epigenetic regulation of gene transcription, and the reprogramming of cellular metabolism
[7]. These enzymes include hypoxia-inducible factor (HIF) hydroxylases (e.g., prolyl-hydroxylase domain-containing proteins (PHDs 1–3) and factor inhibiting HIF (FIH)), as well as DNA demethylases (e.g., TET1–3)
[8][9][10]. The activities of these enzymes contribute to the suppression of oncogenes and the re-expression of tumor suppressor genes, resulting in post-translational and/or epigenetic anti-tumor effects in both hematological and solid tumors
[36][79][80]. Indeed, a growing number of studies report the involvement of these enzymes in VC-induced tumor suppression across multiple cancer types, including leukemia, melanoma, and renal cell carcinoma
[4][79][80][81][82].
Harnessing its multifaceted effects in cancer, high-dose VC has emerged as a promising therapeutic strategy, either as a standalone treatment or in combination with various standard (chemo-) therapies, potentially alleviating the toxic side effects associated with chemotherapy
[12][13][83]. Notably, even intravenous administration of very high doses of VC, ranging from 1 to 200 g and administered repeatedly, was reported to be well tolerated in the majority of patients
[84]. However, caution has been noted regarding the administration of high doses, as they may lead to overt side effects in certain susceptible patients, such as the formation of oxalate renal stones
[85]. Moreover, it is important to note that some patients may experience side effects, including diarrhea, nausea, abdominal cramps, and other gastrointestinal issues
[85]. In light of these considerations, national clinical trials (NCT) have been actively investigating the effects of vitamin C as a standalone treatment or in combined therapies across various cancers, including EGFR mutant non-small cell lung cancer (NSCLC) (NCT04033107), recurrent high-grade glioma (NCT01891747), metastatic colorectal cancer (NCT04516681, NCT02969681), KRAS and BRAF mutant colon cancer (NCT04035096), hepatocellular carcinoma, pancreatic cancer, gastric cancer, colorectal cancer (NCT04033107), and acute myeloid leukemia (AML) (NCT02877277).