Ascorbate acts as a highly potent inducer of chromate mutagenesis and clastogenesis: linkage to DNA breaks in G2 phase by mismatch repair

Abstract

Here we examined the role of cellular vitamin C in genotoxicity of carcinogenic chromium(VI) that requires reduction to induce DNA damage. In the presence of ascorbate (Asc), low 0.2-2 microM doses of Cr(VI) caused 10-15 times more chromosomal breakage in primary human bronchial epithelial cells or lung fibroblasts. DNA double-strand breaks (DSB) were preferentially generated in G2 phase as detected by colocalization of H2AX and 53BP1 foci in cyclin B1-expressing cells. Asc dramatically increased the formation of centromere-negative micronuclei, demonstrating that induced DSB were inefficiently repaired. DSB in G2 cells were caused by aberrant mismatch repair of Cr damage in replicated DNA, as DNA polymerase inhibitor aphidicolin and silencing of MSH2 or MLH1 by shRNA suppressed induction of H2AX and micronuclei. Cr(VI) was also up to 10 times more mutagenic in cells containing Asc. Increasing Asc concentrations generated progressively more mutations and DSB, revealing the genotoxic potential of otherwise nontoxic Cr(VI) doses. Asc amplified genotoxicity of Cr(VI) by altering the spectrum of DNA damage, as total Cr-DNA binding was unchanged and post-Cr loading of Asc exhibited no effects. Collectively, these studies demonstrated that Asc-dependent metabolism is the main source of genotoxic and mutagenic damage in Cr(VI)-exposed cells.

Figure 1

Increased genotoxicity of Cr(VI) in IMR90 cells containing Asc. (A) Representative HPLC profiles for Asc determinations in protein-normalized extracts. (B) Confocal images of IMR90 cells immunostained for γH2AX, cyclin B1 and counterstained with DAPI. Control or 1 mM Asc-loaded cells were treated with 2 μM Cr(VI) for 3 h and fixed 6 h later. (C) Preloading with 1 mM Asc increased the number of cells with γH2AX foci at different Cr doses and (D) times after exposure. Dose-dependence was measured at 6 h post-exposure and time-dependence was determined with 3 μM Cr(VI). Cells with four or more foci were defined as positive. Results are means ± SD for at least four slides with >100 cells scored per slide. Data in panel D are after subtraction of background values in Cr-untreated samples (1.9 ± 0.7 and 1.6 ± 0.6 for Asc− and Asc+ samples, respectively). (E) Elevated frequency of micronuclei in cells loaded with 1 mM Asc prior to Cr(VI) exposure. Micronuclei were scored 48 h after Cr exposure. Closed symbols—total micronuclei, open symbols—CREST-negative micronuclei. Data are means ± SD for four slides with >500 cells counted per slide. (F) Cells preloaded with 1 mM Asc had lower uptake of Cr(VI) and (G) lower Cr-DNA binding. Results are means ± SD for 3–5 independent samples.

Figure 2

Asc caused concentration-dependent increases in DSB foci and micronuclei in primary HBE cells. (A) Accumulation of Asc by HBE cells. Data are means ± SD for six independent samples. (B) Cr(VI) uptake by HBE cells containing different Asc concentrations. Data are means ± SD for four independent samples. (C) Confocal images of cells immunostained for γH2AX, 53BP1 and counterstained with DAPI. Control and preloaded with 1.7 mM Asc-cells were treated with 2 μM Cr(VI) for 3 h and fixed 6 h later. (D) Preloading with Asc increased the number of cells containing foci of γH2AX and (E) 53BP1. Slides were fixed for immunostaining at 6 h after Cr treatment. Cells containing ≥4 foci were scored as positive. Data are means ± SD for four to eight slides with >100 cells counted per slide. (F) Asc-promoted γH2AX foci were preferentially formed in cells expressing a G₂ phase-specific marker cyclin B1. Cells lacking or containing 1.7 mM Asc were treated with 2 μM (left panel) or 5 μM Cr(VI) (right panel). H2AX—percentage of γH2AX+ cells, dual—percentage of cells that are positive for both γH2AX and cyclin B1. Data are means ± SD for four slides with >100 cells/slide. (G) Physiological but not low levels of Asc caused delayed induction of γH2AX after exposure to subtoxic 0.5 μM Cr(VI). Foci were scored at 9 and 18 h after Cr treatment of cells containing 0, 0.6 or 1.7 mM Asc. Results are means ± SD for four slides with >450 cells counted. Data are after subtraction of background values in Cr-untreated samples (9 h: 0.7 ± 0.5, 1.6 ± 1.2 and 1.4 ± 1.2%; 18 h: 0.4 ± 0.8, 0.2 ± 0.4 and 0.5 ± 0.5% for 0, 0.6 and 1.7 mM Asc samples, respectively). (H) Percentage of G₂ cells detected by positive immunostaining for cyclin B1. Cells with (1.7 mM) and without Asc were treated with 2 μM Cr(VI) and fixed for immunofluorescence at 3 and 12 h post-exposure. Data are means ± SD for three slides with >100 cells/slide counted. (I) Preloading with Asc (1.7 mM) increased the induction of micronuclei by Cr(VI). Results are means ± SD for eight slides with >1000 cells/slide counted. Closed symbols—total micronuclei, open symbols—CREST-negative micronuclei.

Figure 3

Cellular Asc strongly increases mutagenicity of Cr(VI). (A) Cr uptake and (B) Cr-DNA binding in control (15 μM Asc) and Asc-preloaded (1.4 mM) CHO cells. Means ± SD for 3–6 independent samples. (C) Yield of Cr-DNA adducts in CHO cells as a function of intracellular Cr dose. (D) Frequency of Hprt mutants in control (15 μM Asc) and Asc-preloaded (1.4 mM) CHO cells. Cells were treated with 0–40 μM Cr(VI). Mutation frequencies are means±SD for four independent populations. (E) Clonogenic survival of control and 1.4 mM Asc-containing CHO cells. Data are means ± SD from two clonogenic assays with triplicate dishes. (F) Cr uptake (means ± SD, n = 4) and (G) Cr-induced Hprt mutagenesis in V79 cells containing different concentrations of Asc. Cells were exposed to 10 μM Cr(VI). Data are means ± SD for 4–8 independently treated populations. Background frequencies were 3.6 × 10⁻⁵ for control (0.8 μM Asc), 3.4 × 10⁻⁵ for 1.3 mM, 3.9 × 10⁻⁵ for 2.2 mM and 2.8 × 10⁻⁵ for 3.4 mM Asc-containing cells.

Figure 4

Post-exposure loading of Asc had no effect on genotoxicity and mutagenicity of Cr(VI). Asc was loaded into cells 1 h after Cr(VI) exposure. (A) Formation of micronuclei in IMR90 cells. (B) Frequency of γH2AX foci-containing HBE cells and (C) Hprt mutagenesis in CHO cells. CHO cells were loaded with 1.4 mM Asc either before (Asc→Cr) or 1 h after (Cr→Asc) exposure to 10 μM Cr(VI). Results are means ± SD.

Figure 5

Suppression of γH2AX and micronuclei formation by stable downregulation of MLH1 and MSH2 proteins. (A) Western blots for MSH2 and MLH1 in IMR90 cells expressing targeting (MLH1, MSH2) and non-specific (Luc) shRNA. Frequency of micronuclei in IMR90 cells (B) preloaded with 1 mM Asc (C) in Asc-deficient cells. Data are means ± SD for four slides with >1000 cells counted per slide. Induction of γH2AX by Cr(VI) in cells (D) containing 1 mM Asc and (E) lacking Asc. Slides were fixed for immunofluorescence at 18 h post-Cr. Data are means ± SD for four slides with >200 cells/slide counted. Frequency of γH2AX/cyclin B1 double positive cells in (F) Asc-loaded and (G) Asc-deficient samples. Experimental conditions were as in panels (D) and (E).

Figure 6

DNA polymerase inhibitor aphidicolin abolished induction of DSB in G₂ cells. All experiments were conducted with 3 μM Cr(VI) exposures for 3 h. Aphidicolin (1 μM) was added 15 min before Cr and was present during and after Cr exposures. Cells were fixed for immunostaining at 1, 3 and 6 h post-Cr treatments. Data are means ± SD for four slides with >100 cells/slide counted. (A) Effect of aphidicolin on frequency of total γH2AX+ (H2AX) and γH2AX/cyclin B1 dual positive (dual) HBE cells lacking Asc, or (B) preloaded with 1.7 mM Asc. Panels A and B have the same legend. (C) Effect of aphidicolin on frequency of total γH2AX+ (H2AX) and γH2AX/cyclin B1 dual positive (dual) IMR90 cells lacking Asc, or (D) preloaded with 1 mM Asc. Panels C and D have the same legend. In untreated controls, frequency of γH2AX+ cells varied from 0.3 to 0.9% for IMR90 cells, and from 0.3 to 0.6% for HBE cells. (E) Percentage of cyclin B1-positive HBE and (F) IMR90 cells in the presence or absence of aphidicolin. Closed symbols—no aphidicolin (−Aph), open symbols—1 μM aphidicolin (+Aph); squares +Asc, diamonds −Asc. (G) Assessment of S to G₂ progression in Cr-treated or (H) untreated IMR90 cells containing or lacking Asc. Cells were labeled with 10 μM BrdU for 15 min prior to 3 μM Cr and fixed for immunostaining at 1, 3 and 6 h post-Cr. Data are means±SD for three slides with >100 cells counted per slide.

Figure 7

A model of Asc-promoted genotoxicity and mutagenicity of carcinogenic chromium(VI).