Results

Effects of Ascorbic Acid in Pharmacologic Concentrations on Survival of Tumor and Normal Cells. We first investigated whether ascorbate in pharmacologic concentrations selectively affected the survival of cancer cells by studying nine cancer cell lines, four normal cell types, and clinically relevant conditions. Clinical pharmacokinetics analyses show that pharmacologic concentrations of plasma ascorbate, from 0.3 to 15 mM, are achievable only from i.v. administration (7). These concentrations are cleared within hours by renal filtration and excretion. In contrast, plasma ascorbate concentrations from maximum possible oral doses cannot exceed 0.22 mM because of limited intestinal absorption, which is bypassed with i.v. administration (7–9). To mimic potential clinical i.v. use, tested cells were incubated for 1 h with either pharmacologic ascorbate concentrations (0.3-20 mM) or a high physiologic concentration (0.1 mM) as control. Once ascorbate was removed, cell survival was determined by nuclear staining or MTT after 24 h (Fig. 1A ). For five of the nine cancer cell lines, ascorbate concentrations causing a 50% decrease in cell survival (EC50 values) were less than 5 mM, a concentration easily achievable from i.v. infusion (7). All tested normal cells were insensitive to 20 mM ascorbate.

Colony formation assays were used as an additional means to determine cell survival (21). Four cancer cell lines were incubated with 5 mM ascorbate or untreated media for 1 h. Cells were diluted and plated and growth assessed after 14 days (Fig. 1B ). All four untreated cell lines grew in soft agar, whereas three of four exposed to ascorbate displayed at least 99% growth inhibition.

Effects of Ascorbic Acid on Death of Human Lymphoma Cells. Human lymphoma cells (JLP-119) were studied in detail to determine the effects of ascorbate on cell death. Lymphoma cells were selected because of their sensitivity to ascorbate (Fig. 1A ), the suitability of these cells for nuclear staining to characterize the mode of cell death (16, 19, 28), and the report of a positive clinical response of lymphoma to i.v. ascorbate (14) (unpublished work). Cells were incubated for 1 h with 0.1-5 mM ascorbate and washed, and Hoechst/PI nuclear staining was performed 18 h later to determine the amount and type of cell death (Fig. 2A ). Ascorbate induced concentration-dependent cell death, which was nearly 100% at 2 mM. As ascorbate concentration increased, the pattern of death changed from apoptosis to pyknosis/necrosis, a pattern suggestive of H2O2-mediated cell death (19). We determined the time necessary for cell death after exposure to 2 mM ascorbate for 1 h (Fig. 2B ). Apoptosis occurred by 6 h after exposure, and cell death by pyknosis was ≈90% at 14 h after exposure. In contrast to lymphoma cells, there was little or no killing of normal lymphocytes and monocytes by ascorbate (Fig. 2C ).

Discussion

Our data show that ascorbic acid selectively killed cancer but not normal cells, using concentrations that could only be achieved by i.v. administration and conditions that reflect potential clinical use. The effect was due only to extracellular and not intracellular ascorbate, consistent with clinical i.v. dosing. Ascorbate-mediated cell death was due to protein-dependent extracellular H2O2 generation, via ascorbate radical formation from ascorbate as the electron donor. Like glucose, when ascorbate is infused i.v., the resulting pharmacologic concentrations should distribute rapidly in the extracellular water space (42). We showed that such pharmacologic ascorbate concentrations in media, as a surrogate for extracellular fluid, generated ascorbate radical and H2O2. In contrast, the same pharmacologic ascorbate concentrations in whole blood generated little detectable ascorbate radical and no detectable H2O2. These findings can be accounted for by efficient and redundant H2O2 catabolic pathways in whole blood (e.g., catalase and glutathione peroxidase) relative to those in media or extracellular fluid (38–41). The totality of the data are consistent with the interpretation that ascorbic acid administered i.v. in pharmacologic concentrations may serve as a pro-drug for H2O2 delivery to the extracellular milieu, but without H2O2 accumulation in blood.

Although it is possible that H2O2 might accumulate in blood, this would occur only under specific conditions that reflect on the general safety of i.v. ascorbate. Ascorbate administered i.v. is likely to be safe in most patients, with virtually no toxicity compared to most currently available cancer chemotherapeutic agents. The occurrence of one predicted complication, oxalate kidney stones, is controversial (13). In patients with glucose-6-phosphate dehydrogenase deficiency, i.v. ascorbate is contraindicated because it causes intravascular hemolysis (13). The mechanism of this previously unexplained observation is now straightforward, based on the results here. H2O2 generated in blood is normally removed by catalase and glutathione peroxidase within red blood cells, with internal glutathione providing reducing equivalents. The electron source for glutathione is NADPH from the pentose shunt, via glucose-6-phosphate dehydrogenase. If activity of this enzyme is diminished, the predicted outcome is impaired H2O2 removal causing intravascular hemolysis, the observed clinical finding.

Ascorbate as a potential cancer therapeutic agent has a controversial and emotionally charged past (1, 3–6). Clinical observational studies reported possible benefit in selected patients, but double-blind placebo-controlled studies reported no benefit, and ascorbate was discarded as a potential therapy by conventional practitioners. Only recently has it been understood that the discordant clinical findings can be explained by previously unrecognized fundamental pharmacokinetics properties of ascorbate (7). In vitro effects of ascorbate on death and survival of cell lines have been reported, but there are multiple experimental concerns. For example, reports compared an experimental condition to that with no ascorbate at all (43, 44), but such a condition has had unclear physiologic relevance, because ascorbate outside and inside cells is always present unless there is severe scurvy. It was unclear whether observed effects were due to extracellular or intracellular ascorbate, or both (12, 43–46). Some experiments have used widely varying incubation times and ascorbate concentrations that have had no corresponding clinical context, making interpretation difficult. H2O2 generation by ascorbate oxidation in culture media was variously interpreted as artifact (47, 48), even though chelators had no effect (49), or reported to mediate damage internally due to diminished intracellular ascorbate, but using an H2O2 assay in which ascorbate could interfere (43, 44).

The experiments presented here provide a clear clinical context for ascorbate action. Conditions were selected to reflect peak ranges of i.v. ascorbate concentrations, which clinically might last a few hours at most, depending on the infusion rate (7). Intracellular transport of ascorbate is tightly controlled in relation to extracellular concentration (8, 9, 29). Intravenous ascorbate infusion is expected to drastically change extracellular but not intracellular concentrations (8, 9). For i.v. ascorbate to be clinically useful in killing cancer cells, pharmacologic but not physiologic extracellular concentrations should be effective, independent of intracellular ascorbate concentrations. This was what was observed here. The experiments here provide a cohesive explanation for ascorbate action in generating H2O2 outside cells, without H2O2 accumulation in blood, leading to the conclusion that ascorbate at pharmacologic concentrations in blood is a pro-drug for H2O2 delivery to tissues.

We observed that H2O2 generation was independent of metal chelators and dependent on at least 0.5% extracellular protein. The responsible proteins were between 10 and 30 kDa (data not shown). It is reasonable that extracellular milieu contains these proteins, given that extracellular milieu protein is as much as 20% of serum protein, and favors lower-molecular-weight proteins (50). Although identities of the proteins responsible are unknown, we postulate that they may have redox-active metal centers. While chelators may marginally affect these metals, they could participate in the oxidation of ascorbate when it is at pharmacologic concentrations, with subsequent formation of superoxide and H2O2 (34). It is also possible that in vivo, cell membranes and their associated proteins could harbor metals accessible to extracellular fluid and could react similarly. In either case, ascorbate, an electron-donor in such reactions, ironically initiates pro-oxidant chemistry and H2O2 formation (34, 51).

It is unknown why ascorbate, via H2O2, killed some cancer cells but not normal cells. There was no correlation with ascorbate-induced cell death and glutathione, catalase activity, or glutathione peroxidase activity. The data here showed that ascorbate initiated H2O2 formation extracellularly, but H2O2 targets could be either intracellular or extracellular, because H2O2 is membrane permeant (38, 52). For example, extracellular H2O2might target membrane lipids, forming hydroperoxides or reactive intermediates that are quenched or repaired in normal cells but not in sensitive cancer cells. In sensitive but not resistant cancer cells, intracellular H2O2could target DNA, DNA repair proteins, or mitochondria because of diminished superoxide dismutase activity (53). New insights may follow from future studies of a very broad range of tumor cells or from microarray analysis of resistant and sensitive cells derived from the same genetic lineage.

H2O2, as the product of pharmacologic ascorbate concentrations, has potential therapeutic uses in addition to cancer treatment, especially in infections. H2O2 is a potent mammalian antimicrobial defense mechanism (54). Neutrophils generate H2O2 from superoxide, in turn formed by NADPH oxidase-catalyzed reduction of molecular oxygen. There may be particular therapeutic application in patients with chronic granulomatous disease who have diminished superoxide production (55). Old observational animal experiments, although uncontrolled, suggest that i.v. ascorbate is effective in some viral infections (56, 57). This finding is also consistent with in vitro experiments, in which H2O2 is toxic to hepatitis C (58). Use of ascorbate as an H2O2-delivery system against sensitive pathogens, viral or bacterial, has substantial clinical implications that deserve rapid exploration.

To proceed clinically in potential treatment of infectious diseases and cancer, clear safety documentation of i.v. ascorbate administration is necessary. More than 100 patients have been described, presumably without glucose-6-phosphate dehydrogenase deficiency, who received 10 g or more of i.v. ascorbate with no reported adverse effects other than tumor lysis (3, 4, 15, 59). However, these descriptions lack formal safety documentation. Complementary and alternative medicine practitioners worldwide currently use ascorbate i.v. in doses as high as 70 g over several hours (14, 15, 59). Because i.v. ascorbate is easily available to people who seek it, a phase I safety trial in patients with advanced cancer is justified and underway.

see  this link for complete article with references:   http://www.pnas.org/content/102/38/13604.long