Enzyme mechanism-based, oxidative DNA-protein cross-links formed with DNA polymerase β in vivo

Abstract

Free radical attack on the C1' position of DNA deoxyribose generates the oxidized abasic (AP) site 2-deoxyribonolactone (dL). Upon encountering dL, AP lyase enzymes such as DNA polymerase β (Polβ) form dead-end, covalent intermediates in vitro during attempted DNA repair. However, the conditions that lead to the in vivo formation of such DNA-protein cross-links (DPC), and their impact on cellular functions, have remained unknown. We adapted an immuno-slot blot approach to detect oxidative Polβ-DPC in vivo. Treatment of mammalian cells with genotoxic oxidants that generate dL in DNA led to the formation of Polβ-DPC in vivo. In a dose-dependent fashion, Polβ-DPC were detected in MDA-MB-231 human cells treated with the antitumor drug tirapazamine (TPZ; much more Polβ-DPC under 1% O2 than under 21% O2) and even more robustly with the "chemical nuclease" 1,10-copper-ortho-phenanthroline, Cu(OP)2. Mouse embryonic fibroblasts challenged with TPZ or Cu(OP)2 also incurred Polβ-DPC. Nonoxidative agents did not generate Polβ-DPC. The cross-linking in vivo was clearly a result of the base excision DNA repair pathway: oxidative Polβ-DPC depended on the Ape1 AP endonuclease, which generates the Polβ lyase substrate, and they required the essential lysine-72 in the Polβ lyase active site. Oxidative Polβ-DPC had an unexpectedly short half-life (∼ 30 min) in both human and mouse cells, and their removal was dependent on the proteasome. Proteasome inhibition under Cu(OP)2 treatment was significantly more cytotoxic to cells expressing wild-type Polβ than to cells with the lyase-defective form. That observation underscores the genotoxic potential of oxidative Polβ-DPC and the biological pressure to repair them.

Keywords: 2-deoxyribonolactone; AP lyase; abasic site; base excision repair; free radical damage.

Fig. 1.

Base excision repair in human cells. Ape1 incises AP sites or dL to generate single-strand DNA breaks with 3′-OH and 5′-dRp termini. Upon attempted excision of dL, Polβ forms stable DPC (Upper Right). For normal AP sites, single-nucleotide (“short-patch”) BER is completed by Polβ inserting 1 nucleotide and excising the 5′-dRp residue. Multinucleotide (“long-patch”) BER can remove the oxidized dL lesion.

Fig. 2.

Sensitivity of BER defective cells to dL-inducing agents. (A) MDA-MB-231 cells were challenged under 1% O₂ with the indicated concentrations of TPZ for 4 h or (B) Cu(OP)₂ for 2 h, or in (C) MEFs with Cu(OP)₂ for 4 h and then shifted to fresh medium before measuring survival by the MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] assay 24 h later. Error bars indicate ±SD for n = 3 independent experiments. *P < 0.05 (two-tailed Student’s t test) between WT and K72A cells (A) or between Polβ-expressing and Polβ-null cells (C).

Fig. 3.

Oxidative trapping of Polβ in DPC in vivo. (A) Cells were treated for 4 h with the indicated concentrations of TPZ under 1% O₂ and assayed for DPC (using 200 ng of DNA). (B and C) Cells were treated with the indicated concentrations of Cu(OP)₂ for 2 h and then immediately assayed for DPC. For each set of experiments, a representative slot blot is shown with quantification for n = 3 independent experiments, normalized to untreated cells. Error bars indicate ±SD. *P < 0.05 (two-tailed Student’s t test) compared with untreated cells.

Fig. 4.

Mechanism-based oxidative trapping of Polβ in DPC in vivo. (A) MDA-MB-231 cells expressing either WT Flag-Polβ or K72A Flag-Polβ were treated with 10 μM Cu(OP)₂ for 2 h and then assayed immediately for DPC (using 200 ng of DNA). (B) CH12F3 cells were treated with Cu(OP)2 for 1 h at the indicated concentrations and then assayed for DPC as described in A. Representative slot blots are shown with quantification for n = 3 independent experiments, normalized to untreated cells. Error bars indicate ±SD. *P < 0.05 (two-tailed Student’s t test) for WT vs. K72A (A) or treated compared with untreated cells (B).

Fig. 5.

Stability of oxidative Polβ-DPC in cells exposed to Cu(OP)₂. (A) MDA-MB-231 cells expressing WT Flag-Polβ were treated with 10 µM Cu(OP)₂ for 2 h and either assayed immediately for Polβ DPC (200 ng) or transferred to fresh medium for the indicated times before assaying for DPC. (B) MDA-MB-231 cells expressing WT Flag-Polβ were treated with 10 μM Cu(OP)₂ for 30 min and then exposed to MG132 for an additional 1 h and either assayed immediately for DPC (0 min) or replaced with fresh media for the indicated time points and then assayed for DPC. For A and B, independent blots were performed with anti-Polβ, anti-Flag, or anti-ubiquitin antibodies. For each set of experiments, a representative slot blot is shown with quantification for n = 3 independent experiments in A and n = 4 in B, normalized to the “0” time point signal, which was taken as 100% DPC signal. *P < 0.05 (two-tailed Student’s t test) for time points compared with time = 0 min.

Fig. 6.

Comparison of cellular sensitivity to Cu(OP)₂ and MG132 cotreatment in MDA-MB-231 cells. MDA-MB-231 cells expressing WT or K72A Polβ were exposed to either 10 µM Cu(OP)₂ for 30 min alone or with 10 µM MG132. For the sequential treatment, cells were first exposed to 10 µM MG132 for 1 h, followed by 10 µM Cu(OP)₂ for 30 min. For the combination treatment, cells were exposed to a mixture of 10 µM MG132 and 10 µM Cu(OP)₂ for 1.5 h. Cell viability was assayed 24 h after incubation in fresh media using the MTT reagent. Data for each panel represent n = 6 independent experiments performed. *P < 0.01 (statistically significant by a two-tailed Student’s t test). Error bars indicate ±SD.