Arsenite-induced ROS/RNS generation causes zinc loss and inhibits the activity of poly(ADP-ribose) polymerase-1

Abstract

Arsenic enhances the genotoxicity of other carcinogenic agents such as ultraviolet radiation and benzo[a]pyrene. Recent reports suggest that inhibition of DNA repair is an important aspect of arsenic cocarcinogenesis, and DNA repair proteins such as poly(ADP ribose) polymerase (PARP)-1 are direct molecular targets of arsenic. Although arsenic has been shown to generate reactive oxygen/nitrogen species (ROS/RNS), little is known about the role of arsenic-induced ROS/RNS in the mechanism underlying arsenic inhibition of DNA repair. We report herein that arsenite-generated ROS/RNS inhibits PARP-1 activity in cells. Cellular exposure to arsenite, as well as hydrogen peroxide and NONOate (nitric oxide donor), decreased PARP-1 zinc content, enzymatic activity, and PARP-1 DNA binding. Furthermore, the effects of arsenite on PARP-1 activity, DNA binding, and zinc content were partially reversed by the antioxidant ascorbic acid, catalase, and the NOS inhibitor, aminoguanidine. Most importantly, arsenite incubation with purified PARP-1 protein in vitro did not alter PARP-1 activity or DNA-binding ability, whereas hydrogen peroxide or NONOate retained PARP-1 inhibitory activity. These results strongly suggest that cellular generation of ROS/RNS plays an important role in arsenite inhibition of PARP-1 activity, leading to the loss of PARP-1 DNA-binding ability and enzymatic activity.

Keywords: Arsenic; PARP-1; RNS; ROS.

Fig. 1

Effect of arsenite (As) on PARP-1 protein level in HaCat cells. A) Western blotting analysis of PARP-1 protein level in cell lysates. HaCat cells were treated with indicated concentrations of sodium arsenite for 48 h. B) Quantification of western blot result with control group (As, 0 μM) normalized to 1.0. Data are presented as mean ± S.D., n=6.

Fig. 2

Effects of arsenite, hydrogen peroxide, and NONOate on PARP-1 activity in HaCat (A, B, C) and HEKn (D, E, F) cells. Cells were treated with sodium arsenite (A, D), hydrogen peroxide (B, E), or NONOate (C, F) separately for 48hours, and then PARP-1 protein was isolated by immunoprecipitation. PARP-1 activity was determined using a HT Colorimetric PARP/Apoptosis Assay kit. Data are presented as means ± S.D., * P<0.05 vs. 0 group, n=6 for HaCat (A, B, C), n=3 for HEKn (D, E, F).

Fig. 3

Effects of vitamin C, AG and catalase on the activity of PARP-1 in arsenite treated cells. HaCat cells were treated with arsenite as described in “Methods” except the indicated concentration of either vitamin C or AG was added for 30 min before arsenite addition. In catalase experiment, HEKn cells were treated with indicated concentrations of catalase for 30 min before arsenite addition. After incubation for 48 hours, PARP-1 was isolated by immunoprecipitation and the activity of PARP-1 was measured with a HT Colorimetric PARP/Apoptosis Assay kit. A) Effects of vitamin C on the activity of PARP-1. B) Effects of AG on the activity of PARP-1. C) Effects of catalase on the activity of PARP-1. Data are presented as means ± S.D. * P<0.05 vs. As alone group, n=6.

Fig. 4

Effects of arsenite, hydrogen peroxide, or NONOate on PARP-1 DNA binding. Cells were treated with indicated concentrations of sodium arsenite, hydrogen peroxide, or NONOate for 48 hrs. PARP-1 was isolated by immunoprecipitation and PARP-1 binding to DNA was assayed by EMSA. A) Images of EMSA results after electrophoresis, visualization of nucleic acids with SYBR green. B) Effects of sodium arsenite on the DNA-binding of PARP-1. C) Effects of hydrogen peroxide on the DNA-binding of PARP-1. D) Effects of NONOate on the DNA-binding of PARP-1. E) Effects of vitamin C (400 μM) and AG (400 μM) on the DNA-binding of PARP-1 with arsenite (2 μM). Data are presented as means± S.D. * P<0.05 vs. concentration at 0 group, n=4.

Fig. 5

Effects of arsenite, hydrogen peroxide, NONOate on the zinc content of PARP-1 isolated from treated HaCat cells. Cells were treated with sodium arsenite, hydrogen peroxide, or NONOate for 48 hrs with or without pretreatment with vitamin C (400 μM) or AG (400 μM). PARP-1 was isolated from the treated cells by immunoprecipitation. The content of zinc in the isolated PARP-1 was detected by spectrophotometric method. Data were presented as means ± S.D., * P<0.05 vs. 0 group, n=6. A) Effects of arsenite; B) Effects of hydrogen peroxide; C) Effects of NONOate; D) Effects of arsenite (2 μM) with or without pretreatment with vitamin C (400 μM) or AG (400 μM).

Fig. 6

Effects of arsenite, hydrogen peroxide, or NONOate on the activity of PARP-1 isolated from untreated HaCat cells. HaCat cell lysates were prepared from untreated, subconfluent cells then PARP-1 was isolated by immunoprecipitation. The isolated PARP-1 was treated with the indicated concentrations of sodium arsenite, hydrogen peroxide, or NONOate at 4°C for 48 hrs before the activity of PARP-1 was determined using an HT Colorimetric PARP/Apoptosis Assay kit. Data were presented as means ± S.D., * P<0.05 vs. 0 group, n=6. A) Effects of arsenite. B) Effects of hydrogen peroxide. C) Effects of NONOate.

Fig. 7

Effects of arsenite, hydrogen peroxide, and NONOate on DNA binding of purified PARP- 1. PARP-1 was isolated by immunoprecipitation from untreated HaCat cells, then the purified PARP-1 protein was treated with indicated concentrations of sodium arsenite, hydrogen peroxide, and NONOate for 48 hours at 4°C, and the DNA-binding ability was tested with EMSA as described in the legend to Fig. 4. Data are presented as means± S.D. * P<0.05 vs. concentration of 0 group, n=4. A) The images of EMSA. B) Effects of sodium arsenite. C) Effects of hydrogen peroxide. D) Effects of NONOate.

Fig 8

Effects of arsenite, hydrogen peroxide, or NONOate on the zinc content of purified PARP-1. PARP-1 protein was isolated from untreated HaCat cells by immunoprecipitation, then the isolated PARP-1 was treated with sodium arsenite (50 μM), hydrogen peroxide (50 μM), and NONOate (20 μM) for 48 hrs at 4°C The zinc content in the isolated PARP-1 was detected spectrophotometrically (DU 800, Beckman Coulter). Data were presented as means ± S.D., * P<0.05 vs. control group; n=6.