Polycyclic aromatic hydrocarbon (PAH) o-quinones produced by the aldo-keto-reductases (AKRs) generate abasic sites, oxidized pyrimidines, and 8-oxo-dGuo via reactive oxygen species

Abstract

Reactive and redox-active polycyclic aromatic hydrocarbon (PAH) o-quinones produced by Aldo-Keto Reductases (AKRs) have the potential to cause depurinating adducts leading to the formation of abasic sites and oxidative base lesions. The aldehyde reactive probe (ARP) was used to detect these lesions in calf thymus DNA treated with three PAH o-quinones (BP-7,8-dione, 7,12-DMBA-3,4-dione, and BA-3,4-dione) in the absence and presence of redox-cycling conditions. In the absence of redox-cycling, a modest amount of abasic sites were detected indicating the formation of a low level of covalent o-quinone depurinating adducts (>3.2 x 10(6) dNs). In the presence of NADPH and CuCl2, the three PAH o-quinones increased the formation of abasic sites due to ROS-derived lesions destabilizing the N-glycosidic bond. The predominant source of AP sites, however, was revealed by coupling the assay with human 8-oxoguanine glycosylase (hOGG1) treatment, showing that 8-oxo-dGuo was the major lesion caused by PAH o-quinones. The levels of 8-oxo-dGuo formation were independently validated by HPLC-ECD analysis. Apyrimidinic sites were also revealed by coupling the assay with Escherichia coli (Endo III) treatment showing that oxidized pyrimidines were formed, but to a lesser extent. Different mechanisms were responsible for the formation of the oxidative lesions depending on whether Cu(II) or Fe(III) was used in the redox-cycling conditions. In the presence of Cu(II)-mediated PAH o-quinone redox-cycling, catalase completely suppressed the formation of the lesions, but mannitol and sodium benzoate were without effect. By contrast, sodium azide, which acts as a *OH and 1O2 scavenger, inhibited the formation of all oxidative lesions, suggesting that the ROS responsible was 1O2. However, in the presence of Fe(III)-mediated PAH o-quinone redox-cycling, the *OH radical scavengers and sodium azide consistently attenuated their formation, indicating that the ROS responsible was *OH.

Figure 1

Detection of AP sites in PAH o-quinone-treated calf thymus DNA produced from heat depurination and depyrimidination for 2 h at 70 °C. (A) BP-7,8-dione, (B) 7,12-DMBA-3,4-dione and (C) BA-3,4-dione. See the Materials and methods for details. a, Basal level of AP sites in methoxyamine-treated calf thymus DNA; b, Heat-treated methoxyamine-treated calf thymus DNA (70 °C). A Wilcoxon two-sample exact test shows a marginally significant effect of temperature, i.e., more adducts were generated at 70 °C than 4 °C (p = 0.05). With the temperature set at 70 °C, Kruskal-Wallis tests show that BP-7,8-dione and 7,12-DMBA-3,4-dione had a significant effect on the quantity of adducts generated (p = 0.0084 and 0.0112, respectively) while BA-3,4-dione had no effect (p = 0.1304). Also, with the temperature set at 70 °C, linear regression analysis showed that the concentration of two quinones (BP-7,8-dione and 7,12-DMBA-3,4-dione) had a highly significant positive effect on the amount of depurinating adducts detected (p <0.0001 for both quinones; a positive effect indicates that the amount of the adducts increases as the concentration of each quinone increases).

Figure 2

Detection of directly formed AP sites in PAH o-quinone-treated calf thymus DNA under redox-cycling conditions. (A) BP-7,8-dione, (B) 7,12-DMBA-3,4-dione and (C) BA-3,4-dione. See the Materials and methods for details. Kruskal-Wallis tests show that all three quinones had a significant effect on the AP sites generated (p = 0.0094, 0.0367 and 0.0093 for BP-7,8-dione, BA-3,4-dione and 7,12-DMBA-3,4-dione, respectively). Linear regression analysis showed that the concentration of all three quinones had a significant effect on the amount of AP sites (p <0.0001 for all three quinones). In each case, sodium azide blocked the formation of AP sites with p value < 0.0001 at all concentrations of PAH o-quinone tested.

Figure 3

Detection of AP sites in calf thymus DNA treated with PAH o-quinones in the presence of NADPH and CuCl₂ revealed by enzymatic base excision by hOGG1. (A) BP-7,8-dione, (B) 7,12-DMBA-3,4-dione and (C) BA-3,4-dione. See the Materials and methods for details. Kruskal-Wallis tests showed that all three quinones had a significant effect on 8-oxo-dGuo generated (p = 0.0304, 0.0094 and 0.0094 for BP-7,8-dione, BA-3,4-dione and 7,12-DMBA-3,4-dione, respectively). Linear regression analysis showed that the concentration of all three quinones had a significant effect on the amount of 8-oxo-dGuo detected (p <0.0001 for all three quinones). In each case, sodium azide blocked the formation of 8-oxo-dGuo with p < 0.0001 at all concentrations of PAH o-quinone tested.

Figure 4

Relationship between ARP and HPLC-ECD methods for the detection of 8-oxo-dGuo. (A) BP-7,8-dione, (B) 7,12-DMBA-3,4-dione and (C) BA-3,4-dione. See the Materials and methods for details.

Figure 5

Detection of AP sites in calf thymus DNA treated with PAH o-quinones in the presence of NADPH and CuCl₂ revealed by enzymatic base excision by Endo III. (A) BP-7,8-dione, (B) 7,12-DMBA-3,4-dione and (C) BA-3,4-dione. See the Materials and methods for details. Kruskal-Wallis tests showed that all three quinones had a significant effect on the amount of oxidized pyrimidines formed (p = 0.0299, 0.0094 and 0.0094 for BP-7,8-dione, BA-3,4-dione and 7,12-DMBA-3,4-dione, respectively). In each case, sodium azide blocked the formation of oxidized pyrimidines with p < 0.0001 at all concentrations of PAH o-quinone tested.

Figure 6

PAH o-quinones can produce 100 times higher 8-oxo-dGuo than the heat-labile unstable DNA depurinating adducts in vitro. The experiments were conducted with 10 μM concentration of three PAH o-quinones in the presence or absence of NADPH and Cu(II). (A) BP-7,8-dione, (B) 7,12-DMBA-3,4-dione, (C) BA-3,4-dione. AP sites formed during redox-cycling were significantly different from the basal level of AP sites seen for all three quinones (p = 0.0101, 0.0297 and 0.0101 for BP-7,8-dione, BA-3,4-dione and 7,12-DMBA-3,4-dione, respectively). Oxidized pyrimidines generated during redox-cycling were significantly different from the basal level of AP sites observed with all three quinones (p = 0.0101 for all three quinones). 8-oxo-dGuo lesions formed during redox-cycling was significantly different from the basal level of AP sites seen for all three quinones (p = 0.0101 for all three quinones).

Figure 7

Detection of AP sites, oxidized pyrimidines and 8-oxo-dGuo in calf thymus DNA treated with BP-7,8-dione under Cu(II)- and Fe(III)-mediated redox-cycling conditions. (A) Directly formed AP sites, (B) Oxidized pyrimidines (by Endo III-coupled assay) and (C) 8-oxo-dGuo (by hOGG1-coupled assay). See the Materials and methods for details. For both systems, the difference in the amount of three lesions was concentration dependent. For Cu(II) plus NADPH system, the p-value for the interaction between lesions and the linear component of quinone concentration was <0.0001. For Fe(III) plus NADPH system, the p-value for the interaction between lesions and the linear component of quinone concentration was <0.0001.

Figure 8

Detection of AP sites, oxidized pyrimidines and 8-oxo-dGuo in calf thymus DNA treated with hydrogen peroxide under Cu(II)- and Fe(III)-mediated redox-cycling conditions. (A) Directly formed AP sites, (B) Oxidized pyrimidines (by Endo III-coupled assay) and (C) 8-oxo-dGuo (by hOGG1-coupled assay). See the Materials and methods for details. For both redox-cycling conditions, the difference in the amount of all three lesions was significant; 8-oxo-dGuo gave the most lesions, while oxidized pyrimidines gave the fewest lesions. However, the differences were dependent upon hydrogen peroxide concentration. The p-value for the linear interaction between lesions and hydrogen peroxide concentration was <0.0001 for both Cu(II) and Fe(III) systems.

Figure 9

The effects of ROS scavenging agents on the formation of direct AP sites, oxidized pyrimidines and 8-oxo-dGuo in calf thymus DNA treated with 1 μM BP-7,8-dione under Cu(II)-and Fe(III)-mediated redox-cycling conditions. (A) Directly formed AP sites, (B) Oxidized pyrimidines (by Endo III-coupled assay) and (C) 8-oxo-dGuo (by hOGG1-coupled assay). The concentration of each scavenger was as follows. Catalase was added as 800 U/mL. Mannitol, sodium benzoate and sodium azide were added to achieve a final concentration of 0.1 M. In the data presented, the effects of each scavenger on the levels of directly formed AP site, oxidized pyrimidines (by Endo III-coupled assay) and 8-oxo-dGuo (by hOGG1-coupled assay) were compared with the levels of each control group (complete system), respectively. Kruskal-Wallis tests were conducted to determine which scavengers altered the amount of each lesions in the Cu(II) plus NADPH or Fe(III) plus NADPH redox-cycling conditions. With the Cu(II) plus NADPH system, catalase reduced the amount of all three adducts significantly (p = 0.0495 for each lesion type); mannitol produced significantly more AP sites (p = 0.0495) but had no significant effect on oxidized pyrimidines (p = 0.1266) or 8-oxo-dGuo ( p = 0.8273); sodium benzoate produced significantly more AP sites (p = 0.0495) but had no significant effect on oxidized pyrimidines or 8-oxo-dGuo (p = 0.2754 and 0.1266, respectively); sodium azide reduced the amounts of all three adducts significantly (p = 0.0495 for each lesion type). Inspection of the data shows that the effects of mannitol and sodium benzoate were statistically significant but their effects were marginal when compared with sodium azide. With the Fe(III) plus NADPH system, catalase reduced the amount of all three adducts significantly (p = 0.0495 for each adduct); mannitol reduced the amount of 8-oxo-dGuo and AP sites significantly (p = 0.0495) but not oxidized pyrimidines (p = 0.1266); Both sodium benzoate and sodium azide reduced the amount of all three adducts significantly (p = 0.0495 for each lesion type).

Scheme 1

Spectrum of DNA modifications that PAH o-quinones can cause in DNA