Clinical relevance of guanine-derived urinary biomarkers of oxidative stress, determined by LC-MS/MS

Abstract

A reliable and fast liquid chromatography-tandem mass spectrometry method has been developed for the simultaneous determination of three oxidized nucleic acid damage products in urine, 8-oxoguanine (8-oxoGua), 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-oxodGuo) and 8-oxo-7,8-dihydroguanosine (8-oxoGuo). We applied this method to assess the effect of various urine workup procedures on the urinary concentrations of the oxidized nucleic acid products. Our results showed that frozen urine samples must be warmed (i.e., to 37 °C) to re-dissolve any precipitates prior to analysis. We showed that common workup procedures, such as thawing at room temperature or dilution with deionized water, are not capable of releasing fully the oxidized nucleic acid products from the precipitates, and result in significant underestimation (up to ~ 100% for 8-oxoGua, ~ 86% for both 8-oxodGuo and 8-oxoGuo). With this method, we further assessed and compared the ability of the three oxidized nucleic acid products, as well as malondialdehyde (MDA, a product of lipid peroxidation), to biomonitor oxidative stress in vivo. We measured a total of 315 urine samples from subjects with burdens of oxidative stress from low to high, including healthy subjects, patients with chronic obstructive pulmonary disease (COPD), and patients on mechanical ventilation (MV). The results showed that both the MV and COPD patients had significantly higher urinary levels of 8-oxoGua, 8-oxodGuo, and 8-oxoGuo (P < 0.001), but lower MDA levels, compared to healthy controls. Receiver operating characteristic curve analysis revealed that urinary 8-oxoGuo is the most sensitive biomarker for oxidative stress with area under the curve (AUC) of 0.91, followed by 8-oxodGuo (AUC: 0.80) and 8-oxoGua (AUC: 0.76). Interestingly, MDA with AUC of 0.34 failed to discriminate the patients from healthy controls. Emerging evidence suggests a potential clinical utility for the measurement of urinary 8-oxoGuo, and to a lesser extent 8-oxodGuo, which is strongly supported by our findings.

Keywords: COPD; LC-MS/MS; Lipid peroxidation; Mechanical ventilation; Nucleic acid oxidation; Urine precipitates.

Fig. 1

Representative chromatograms of a COPD patient urine spiked with stable isotope labeled internal standards, as measured by LC-MS/MS coupled with online SPE. Multiple reaction monitoring transitions of (A) 8-oxoGua at 9.4 min, (B) 8-oxoGuo at 12.4 min and (C) 8-oxodGuo at 13.4 min.

Fig. 2

Comparisons of urinary concentrations of (A) 8-oxoGua, (B) 8-oxodGuo, and (C) 8-oxoGuo in urine samples (n = 20 from healthy subjects) following thawing either at RT or 37 °C. The concentration comparison between RT and 37 °C was performed by Wilcoxon Signed Rank Test.

Fig. 3

Distribution of (A) 8-oxoGua, (B) 8-oxodGuo, and (C) 8-oxoGuo concentrations in healthy controls (n = 60 non-smokers and 68 smokers), COPD patients (n = 43 non-smokers and 44 smokers), and MV patients (n = 100). Each point represents an individual subject and the horizontal lines represent median values. The comparison was performed by Mann-Whitney U test.

Fig. 4

Correlations between the three urinary nucleic acid-derived biomarker concentrations: (A) 8-oxodGuo vs. 8-oxoGua, (B) 8-oxodGuo vs. 8-oxoGuo, and (C) 8-oxoGuo vs. 8-oxoGua. The correlation was estimated by Spearman correlation coefficient.

Fig. 5

ROC curve analysis of four urinary biomarkers of oxidative stress (measured by LC-MS/MS) for distinguishing patients and healthy subjects. The analysis yielded AUC of 0.91 for 8-oxoGuo, 0.80 for 8-oxodGuo, 0.76 for 8-oxoGua and 0.34 for MDA.