Sequence-dependent variation in the reactivity of 8-Oxo-7,8-dihydro-2'-deoxyguanosine toward oxidation

Abstract

The goal of this study was to define the effect of DNA sequence on the reactivity of 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-oxodG) toward oxidation. To this end, we developed a quadrupole/time-of-flight (QTOF) mass spectrometric method to quantify the reactivity of site specifically modified oligodeoxyribonucleotides with two model oxidants: nitrosoperoxycarbonate (ONOOCO(2)(-)), a chemical mediator of inflammation, and photoactivated riboflavin, a classical one-electron oxidant widely studied in mutagenesis and charge transport in DNA. In contrast to previous observations with guanine [ Margolin , Y. , ( 2006 ) Nat. Chem. Biol. 2 , 365 ], sequence context did not affect the reactivity of ONOOCO(2)(-) with 8-oxodG, but photosensitized riboflavin showed a strong sequence preference in its reactivity with the following order (8-oxodG = O): COA ≈ AOG > GOG ≥ COT > TOC > AOC. That the COA context was the most reactive was unexpected and suggests a new sequence context where mutation hotspots might occur. These results point to both sequence- and agent-specific effects on 8-oxodG oxidation.

Figure 1

LC-QTOF analysis of 8-oxodG-containing duplex oligodeoxyribonucleotides. (a) Extracted ion chromatograms of the GOG oligodeoxyribonucleotide duplex before (upper) and after (lower) treatment with 2 mM ONOOCO₂⁻. GATCTCGATC internal standard was added after treatment. (b) TOF mass spectrum of the triply-charged GOG oligodeoxyribonucleotide showing a monoisotopic m/z value of 1024.50920 and a well resolved isotopic envelope. (c, d) Scatter plots showing results from (c) QTOF and (d) QQQ analysis of the relative reactivity of ONOOCO₂⁻ with oligodeoxyribonucleotide duplexes containing either GGG or GOG sequence contexts. (e) Illustration of CID localization of 8-oxodG in the GOG oligodeoxyribonucleotide, with expected (in italics, upper row) and measured (lower row) m/z values.

Figure 2

Reactivity of 8-oxodG-containing oligodeoxyribonucleotides with ONOOCO₂⁻. (a,b) LCQTOF analysis of duplex oligodeoxyribonucleotides were treated with (a) 0, 0.1, 0.2, 0.5, or 2 mM ONOOCO₂⁻, or (b) 0, 0.0025, 0.05, 0.1 or 0.2 mM ONOOCO₂⁻. Each data point represents mean ± SD of three independent experiments, with Student's T-test at P<0.05 revealing no significant differences in reactivity. (c) Direct analysis of 8-oxodG in AOC and AOG oligodeoxyribonucleotides treated with varying concentrations of ONOOCO₂⁻. The oligodeoxyribonucleotides were digested to 2'-deoxyribonucleosides before LC-QQQ quantification of 8-oxodG. Each data point represents an average of two separate samples.

Figure 3

Comparison of sequence-dependent reactivity of 8-oxodG with ONOOCO₂⁻ and photoactivated riboflavin. Loss of oligodeoxyribonucleotides containing 8-oxodG in different sequence contexts following oxidation with either (a) ONOOCO₂⁻ or (b) photoactivated riboflavin was quantified as the area under the curve (AUC) values for the dose-response graphs in Figures 2A and 4A. AUC values were calculated by the trapezoidal method for each replicate data set, with mean and standard deviation calculated for 3 independent experiments. The lower the AUC value the higher the reactivity. Correlation analysis of ONOOCO₂⁻ and riboflavin reactivity dose-response curves revealed significant differences (P <0.05 by Student's t-test) for pair-wise comparisons of AUC values for 8-oxodG sequence contexts (abbreviations in Table 1). For ONOOCO₂⁻, only GOG and COT were significantly different; for riboflavin-mediated photooxidation, all comparisons were significantly different except for COA vs. AOG, GOG vs. COT, AOC vs. TOC and COT vs. TOC.

Figure 4

Reactivity of 8-oxodG-containing oligodeoxyribonucleotides with photosensitized riboflavin. 8-oxodG-containing duplex oligodeoxyribonucleotides were treated with varying concentrations of riboflavin (0, 3, 6, 10, 30 and 90 μM) and a constant level of UVA irradiation at 4 °C for 20 min, as described in Materials and Methods. The oligodeoxyribonucleotides were then analyzed by (a) LCQTOF or (b) LC-QQQ. Each data point represents the mean ± S.D. of 3 samples; asterisks denote significant differences by ANOVA, P < 0.01. (c) Targeted quantification of 8-oxodG in AOC and AOG oligodeoxyribonucleotides treated with varying concentrations of photosensitized riboflavin (0–90 μM) and then digested to nucleosides for LC-QQQ analysis. Data represent mean ± error about the mean for analysis of two samples.

Scheme 1

Mechanisms and products of 8-oxodG oxidation by photoactivated riboflavin and nitrosoperoxycarbonate.