There is an increasing interest in the use of the urinary levels of 8-oxo dihydrodeoxy guanosine (8-oxodG) and 8-oxo dihydro guanine (8-oxoGua) as a measure of overall oxidative stress. The current interest in these analytes underscores the importance of having accurate and reliable assays. This article reviews the pros and cons of the currently available assay methods; despite its cost, LC-MS/MS remains the gold standard for the assay of 8-oxodG and 8-oxoGua.
by Dr Chiung-Wen Hu and Dr Mu-Rong Chao

Oxidative stress
There is growing scientific evidence suggesting that oxidative stress may play a crucial role in many diseases, including cancer, cardiovascular disease and degenerative disease. Oxidative stress is generally defined as a state in which the antioxidant system is overwhelmed by the amount of reactive oxygen species (ROS) present. Under oxidative stress, the excess levels of ROS may lead to a modification of cellular biomolecules, such as DNA, lipids and proteins, and may have implications for cellular function. 

Urinary 8-oxodG and 8-oxoGua
Oxidative modification of DNA is perhaps the most studied in the literature. Among over 30 nucleobase modifications that have been described, 8-oxo-7,8-dihydro-2’-deoxyguanosine (8-oxodG; also known as  8-OHdG) [Figure 1A] has received particular attention [1,2]. 8-OHdG is generated by hydroxyl radicals, singlet oxygen, and one-electron oxidants in cellular DNA. The detection of this lesion is considered important not only because of its abundance but also because of its mutagenic potential through G-to-T transversion mutations upon replication of DNA [3]. It is generally accepted that oxidatively damaged DNA can be repaired, and that the repair products are released into the bloodstream and consequently appear without further metabolism in the urine [4,5]. The modified base 8-oxo-7,8-dihydroguanine 8-oxoGua [Figure 1B] and modified nucleoside (8-oxodG) that are found in urine represent the major repair products of oxidatively damaged DNA in vivo, presumably the result of the base excision repair (BER) and nucleotide excision repair (NER) pathways, respectively, although the involvement of the NER pathway in 8-oxodG production has not been fully identified [Figure 2]. Urinary 8-oxodG may possibly also originate from enzymatic hydrolysis (e.g. via nudix hydrolases) of oxidized guanine nucleoside 5’-phosphates in the nucleotide pool as well as from the nucleotide incision repair (NIR) or mismatch repair (MMR) pathways [2].

Over the past few decades, urinary 8-oxodG has been widely used as a biomarker of oxidative stress, whereas relatively limited information has been available on urinary 8-oxoGua. Urinary 8-oxoGua was initially rejected as a biomarker of oxidatively damaged DNA, probably because studies on rat urine showed that diet could significantly affect its urinary levels [5]. However, recent studies in humans have shown that urinary levels of 8-oxodG and 8-oxoGua are independent of diet [6,7], opening  up the possibility that the detection of both 8-oxodG and 8-oxoGua in urine would result in a better assessment of the  whole-body burden of oxidative stress.

Methods of urinary 8-oxodG measurement
Various analytical techniques, principally chromatographic and immunoassay methods, have been developed for the quantification of urinary 8-oxodG [8]. Of the chromatographic methods, these include HPLC with electrochemical detection (HPLC-ECD) and gas chromatography-mass spectrometry (GC-MS). Although these methods are in themselves well-established, they can be difficult to carry out in the clinical laboratory and are in any case labour-intensive, require chemical derivatisation, or have inadequate specificity when used for assaying urinary 8-oxodG. In contrast, liquid chromatography-tandem MS (LC-MS/MS) is a relatively new and powerful technology that can overcome the sensitivity and selectivity issues in the analysis of oxidatively damaged DNA. The possibility of using stable isotope-labelled internal standards in mass spectrometry (known as the “isotope dilution” technique) has added greater reliability to the LC-MS/MS methods by accounting for any losses during sample work up and compensating for any variability in MS detection. LC-MS/MS methods have been coupled with online sample extraction using a column-switching device [Figure 3], that allows the automatic preparation of biological samples [9]. After the addition of the stable isotope-labelled internal standard, urine samples can then be directly analysed without any further pretreatment. Briefly, as shown in Figure 3A, when the switching valve is at position A, the urine sample is loaded onto the trap column and only the fraction containing the compound of interest (enriched 8-oxodG) is retained on the column while the remainder of the sample is diverted to waste. When the valve is switched to position B [Figure 3B], the sample is automatically transferred onto the analytical column and MS detector. The advantages of this method include lower ion suppression and relatively short run times, as well as higher sensitivity and selectivity, especially for urine samples, which can  contain a considerable amount of coeluting interferences.

For immunoassay, competitive enzyme-linked immunosorbent assay (ELISA) has also received widespread use in many laboratories, particularly with the introduction of commercially available kits (e.g. the JaICA kit). The benefits of ELISA include its ease of use, its high-throughput and the fact that no specialist equipment is needed [2].

There is however, a significant discrepancy reported in the literature between the basal urinary levels of 8-oxodG as determined by the chromatographic method and those produced by the ELISA method, which can be 4-fold or higher [8,10]. In 2006, the European Standards Committee of Urinary (DNA) Lesions Analysis (ESCULA) was formed; it represents an international group of more than 25 laboratories dedicated to the identification of the sources of errors and variability in urinary 8-oxodG assays. Progress has been made in this respect, including the demonstration, in ELISAs carried out at 37°C, that 8-oxo-7,8-dihydro-2’-deoxyguanosine 5’-monophosphate (8-oxodGMP) and 8-oxodG oligomers bind several times more of the mouse N45.1 monoclonal antibody used in the assay than 8-oxodG monomer [11]. This may result in a mean level of 8-oxodG higher than that obtained by chromatographic measurement. To evaluate in more detail the accuracy of intra/intertechnique and interlaboratory measurements, samples of phosphate buffered saline and urine spiked with different concentrations of 8-oxodG, together with a series of urine samples from healthy subjects were distributed to ESCULA members, covering 20 methods (broadly classified as MS, HPLC-ECD and ELISA) [12]. The results showed a good agreement between the MS and ECD methods but there was a concentration-dependent deviation for ELISA (i.e., less agreement between the chromatographic and ELISA techniques at higher levels of urinary 8-oxodG. ELISA also showed more within-method variation than did the chromatographic techniques [12].  

Methods of urinary 8-oxoGua measurement
Unlike in the case of urinary 8-oxodG measurement, there is a relatively limited number of  chromatographic methods for 8-oxoGua detection in human urine. To the best of our knowledge, there is also no commercially available ELISA kit specific for 8-oxoGua detection. Previous theories suggested that the diet could significantly contribute to urinary 8-oxoGua thus complicating the assay; in fact, the quantitative analysis of 8-oxoGua is difficult because of its poor solubility and stability. 8-oxoGua is completely soluble only in 0.01–0.1 N NaOH solution and the resulting stock solution is only stable for < 2 day at room temperature or -20°C. The stability of such stock solutions can be greatly improved by decreasing the pH and/or the storage temperature (although avoiding complete freezing). From the practical viewpoint, it has been suggested that once the 8-oxoGua powder is completely dissolved in NaOH, it could be better to dilute it in an aqueous solution of methanol and keep it in the freezer for long-term storage (e.g., a stock solution 8-oxoGua in 5% methanol {pH ~11} is stable for ~ 112 days at -20°C [9]). Meanwhile, it should be mentioned that unlike the 8-oxoGua stock solution that can  be tricky to deal with, 8-oxoGua in urine is actually quite stable (> 60 days at -20°C).

The majority of recently developed chromatographic methods (including HPLC-ECD, GC/MS following prior HPLC pre-purification, and LC-MS/MS) are capable of simultaneously measuring 8-oxodG and 8-oxoGua in urine [13-18]. The literature values of urinary 8-oxoGua concentrations in healthy subjects were at least twice as high as 8-oxodG, as shown in Table 1. LC-MS/MS methods using a direct injection of microlitre volumes of urine without tedious manual sample pre-purification are the simplest to carry out in practice [9,17,18].

Future perspectives
There are a large number of pathological conditions in which oxidative stress may have a role. Urinary biomarkers of oxidatively damaged DNA (i.e. 8-oxodG and 8-oxoGua) offer a means by which oxidative  stress could be monitored in a noninvasive way. The analytes may also have the potential to act as markers of disease development risk or assess efficacy of therapy [19]. Nevertheless, before the measurement of oxidatively damaged DNA can be used clinically, assay methods need to be fully validated. This is exactly what ESCULA is aiming to achieve through continued international cooperation. When these urinary biomarkers can be measured accurately and sensitively, we will then be able to have a better understanding of the role of oxidative damage in disease as well as being able to establish the range of urine values that are clinically normal or abnormal.

Conclusion
For sensitivity, selectivity and reliability, LC-MS/MS is the preferred technique and, when coupled with a column-switching device, is also suitable for high-throughput routine clinical use. Despite its cost and specialised nature, LC-MS/MS may remain the gold standard for measuring oxidatively damaged DNA in urine. 

References
1. Cadet J, Douki T, Ravanat JL. Acc Chem Res. 2008; 41: 1075-83.
2. Cooke MS, Henderson PT, Evans MD. J Clin Biochem Nutr 2009; 45: 255-70.
3. Cheng KC et al. J Biol Chem 1992; 267:166-72.
4. Shigenaga MK, Gimeno CJ, Ames BN. Proc Natl Acad Sci USA 1989; 86: 9697-701.
5. Fraga CG et al. Proc Natl Acad Sci USA 1990; 87:4533-7.
6. Gackowski D et al. Free Radic Res. 2001; 35: 825-32.
7. Cooke MS et al. Mutat Res. 2005; 574: 58-66.
8. Cooke MS, Olinski R, Loft S, European Standards Committee on Urinary (DNA) Lesion Analysis. Cancer Epidemiol Biomarkers Prev 2008; 17: 3-14.
9. Hu CW, Chao MR, Sie CH. Free Radic Biol Med 2010; 48: 89-97.
10. Shimoi K et al. Cancer Epidemiol Biomarkers Prev. 2002; 11: 767-70 .
11. Evans MD et al. Free Radic Res 2008; 42: 831-40.
12. Evans MD et al. European Standards Committee on Urinary (DNA) Lesion Analysis, FASEB J. 2010; 24:1249-60.
13. Helbock HJ et al. Proc Natl Acad Sci USA 1998; 95: 288-93.
14. Svoboda P et al. Antioxid Redox Signaling. 2006; 8: 985-92.
15. Rozalski R et al. Cancer Epidemiol Biomarkers Prev 2002; 11: 1072-5.
16. Roszkowski K et al. Int J Cancer. 2008; 123: 1964-7.
17. Weimann A, Belling D, Poulsen HE. Nucleic Acids Res 2002; 30:E7.
18. Machowetz A et al. FASEB J 2007; 21: 45-52.
19. Cooke MS, Olinski R, Evans MD. Clin Chim Acta 2006; 365: 30-49.

The authors
Chiung-Wen Hua and
Mu-Rong Chaob
a Department of Public Health
Chung Shan Medical University
No.110, Sec.1,
Chien-Kuo N Road
Taichung 402
Taiwan
e-mail: windyhu@csmu.edu.tw

b Department of Occupational Safety and
Health Chung Shan Medical University
No.110, Sec.1,
Chien-Kuo N Road
Taichung 402
Taiwan
e-mail: mrchao@csmu.edu.tw