Antiproliferative effect of ascorbic acid is associated with the inhibition of genes necessary to cell cycle progression

Abstract

Background: Ascorbic acid (AA), or Vitamin C, is most well known as a nutritional supplement with antioxidant properties. Recently, we demonstrated that high concentrations of AA act on PMP22 gene expression and partially correct the Charcot-Marie-Tooth disease phenotype in a mouse model. This is due to the capacity of AA, but not other antioxidants, to down-modulate cAMP intracellular concentration by a competitive inhibition of the adenylate cyclase enzymatic activity. Because of the critical role of cAMP in intracellular signalling, we decided to explore the possibility that ascorbic acid could modulate the expression of other genes.

Methods and findings: Using human pangenomic microarrays, we found that AA inhibited the expression of two categories of genes necessary for cell cycle progression, tRNA synthetases and translation initiation factor subunits. In in vitro assays, we demonstrated that AA induced the S-phase arrest of proliferative normal and tumor cells. Highest concentrations of AA leaded to necrotic cell death. However, quiescent cells were not susceptible to AA toxicity, suggesting the blockage of protein synthesis was mainly detrimental in metabolically-active cells. Using animal models, we found that high concentrations of AA inhibited tumor progression in nude mice grafted with HT29 cells (derived from human colon carcinoma). Consistently, expression of tRNA synthetases and ieF2 appeared to be specifically decreased in tumors upon AA treatment.

Conclusions: AA has an antiproliferative activity, at elevated concentration that could be obtained using IV injection. This activity has been observed in vitro as well in vivo and likely results from the inhibition of expression of genes involved in protein synthesis. Implications for a clinical use in anticancer therapies will be discussed.

Figure 1. Human pangenomic microarrays and qPCR analysis.

(a) Primary cultures of human skin fibroblasts were incubated in medium containing 0, 0.3, 0.6, and 0.8 mM AA. After 24 h, RNA was extracted; reverse transcribed, and hybridized on AGILENT human pangenomic microarrays. Dye swap experiments were also conducted. Data were analyzed using the Rosetta Luminator software package. And only those genes that were up- or downregulated in AA-treated cells were analyzed further. Among the downregulated genes, 40% belong to two classes (tRNA synthetases and translation initiation factor subunits) involved in protein synthesis. (b) Validation of microarray results using qPCR and Roche UPL-specific probes.

Figure 2. Impact of AA treatment on cell growth.

Growth curves of healthy human fibroblasts (a) or Raji cells (b) incubated with various AA concentrations. Cell density estimation of the reversal of AA treatment on Raji cells incubated with 0.6 mM (c), 2 mM (d), or 3 mM (e) AA.

Figure 3. Cell cycle analysis.

Cell cycles in control and treated human primary fibroblasts (a) or Raji cells (b), were analyzed using propidium iodide. DNA content was analyzed using FACS to determine the cell cycle distribution. All cell cycle results were analyzed using the ModFit software provided with the FACSCalibur flow cytometer. The FACS profile is shown in (c).

Figure 4. Cell viability analysis.

(a) Cell viability of Raji cells treated with varying concentrations of AA was evaluated using trypan blue staining. (b) The same cells were analyzed using a propidium iodide (PI) /Annexin-V double label (see Methods). ELITE Flow Cytometry (Beckman Coulter) and Cell Quest software was used to analyze the different cell populations. Morphology of untreated (c) and treated (3mM AA) (d) colon adenocarcinoma (HT29) cells. Cells were observed with a light microscope using the phase contrast setting.

Figure 5. Tumor growth and weight evaluation.

Kaplan Meier survival curves. Gene expression in tumor. HT29 cells were injected into nude male mice. Mice presenting with a tumor after 10 days were treated every day with either a placebo (physiological serum) or one of three AA concentrations. (a) Tumor growth in placebo-treated mice and mice treated with a daily intraperitoneal injection of AA (1000 mg/kg/d). (b) Mice were sacrificed after 1 month of a daily AA treatment and tumors were weighted to determine the effect of AA treatment. An asterisk indicates a P value<0,05. (c) Kaplan-Meier survival curves of mice treated with placebo or with 15 mg/kg/d, 100 mg/kg/d, or 1000 mg/kg/d AA. Once visible tumors were established (10 days after injection) mice were randomized and treated either with ascorbic acid or with placebo. Comparison of placebo treated curve and 1000 mg/kg/d treated animals, using a long rank statistical test, indicates a high statistical significance (P value = 0.0022). (d) Relative quantification of mRNA coding for tRNA synthetase and a translation initiation factor subunit using qPCR technology and Universal Probe Library (Roche). An asterisk indicates a P value<0,05.