Evaluation of enzyme-linked immunosorbent assay and liquid chromatography–tandem mass spectrometry methodology for the analysis of 8-oxo-7,8-dihydro-2′-deoxyguanosine in saliva and urine

Abstract

While ELISA is a frequently used means of assessing 8-oxo-7,8-dihydro-2-deoxyguanosine (8-oxodG) in biological fluids, differences in baseline urinary 8-oxodG levels, compared to chromatographic techniques, have raised questions regarding the specificity of immunoassays. Recently, ELISA of salivary 8-oxodG has been used to report on periodontal disease. We compared salivary 8-oxodG levels, determined by two commercial ELISA kits, to liquid chromatography-tandem mass spectrometry (LC-MS/MS) with prior purification using solid-phase extraction. While values were obtained with both ELISA kits, salivary 8-oxodG values were below or around the limit of detection of our LC-MS/MS assay. As the limit of detection for the LC-MS/MS procedure is much lower than ELISA, we concluded that the assessment of salivary 8-oxodG by ELISA is not accurate. In contrast to previous studies, ELISA levels of urinary 8-oxodG (1.67 ± 0.53 pmol/μmol creatinine) were within the range reported previously only for chromatographic assays, although still significantly different than LC-MS/MS (0.41 ± 0.39 pmol/μmol creatinine; p = 0.002). Furthermore, no correlation with LC-MS/MS was seen. These results question the ability of ELISA approaches, at present, to specifically determine absolute levels of 8-oxodG in saliva and urine. Ongoing investigation in our laboratories aims to identify the basis of the discrepancy between ELISA and LC-MS/MS.

Introduction

Oxidative stress and, in particular, oxidative damage to DNA, are reported to be involved in numerous pathological conditions [1], having both genetic [2] and epigenetic consequences [3]. The ability to assess oxidative stress in vivo is essential to our understanding of the importance of this phenomenon in both normal and pathological processes [4]. Furthermore, methods for the minimally, or noninvasive, assessment of oxidative stress are highly desirable, as these will be most applicable to large-scale human studies and clinical applications [5]. In this context, urine is principally the matrix of choice [6], and the most frequently measured biomarker of DNA oxidation is 8-oxo-7,8-dihydro-2-deoxyguanosine (8-oxodG) [7]. The principal methods applied to the analysis of this lesion in urine are liquid chromatography prepurification followed by gas chromatography (LC-GC/MS) [8], liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) [9], immunoassay [10], and (column switching) liquid chromatography with electrochemical detection (HPLC-ECD) [11], [12]. While immunoassay has, by far, demonstrated the greatest versatility, in terms of the matrices to which it can be applied (urine, serum, plasma,¹ cell culture medium, DNA hydrolysates) [13], [14], [15], and simplicity of use, comparison with urinary 8-oxodG levels by chromatographic techniques has revealed a discrepancy [16], [17], [18]. Recent typical mean values for urinary 8-oxodG, in healthy volunteers, by a commercial ELISA kit can range from 8.16 to 11.13 μg/g creatinine [17], [18], compared to 0.79–2.13 μg/g creatinine, for example, by chromatographic techniques (discussed in [19]). Nevertheless, the results for ELISA and chromatographic techniques have been shown to correlate: r = 0.46, p < 0.001 using HPLC-ECD [18] and r = 0.73, p = 0.002 using LC-GC/MS [20]. The monoclonal antibody used in a commercial ELISA manufactured by JICA, denoted N45.1, has been well characterised as recognising 8-oxodG, with the closest competitor, 8-oxoguanosine (8-oxoGuo), requiring to be present in concentrations two orders of magnitude greater than 8-oxodG to compete effectively [21]. It seems unlikely that this accounts for the discrepancy, as 8-oxoGuo appears to be present in urine at concentrations only twice that of 8-oxodG (8-oxoGuo, 48, vs 8-oxodG, 28 nmol/24 h [9]). We have proposed that this discrepancy might arise from, among others, contributions from 8-oxodGMP, via the activity of enzymes such as the human Nudix hydrolases [22], which may have the potential to leave the cell [23], along with 8-oxodG-containing oligomers, via the activity of nucleotide excision repair [24], although the presence of the latter has been disputed [9]. The precise reason for the higher ELISA values remains unknown.

Recently, there have been reports that 8-oxodG can be measured by ELISA in the saliva of clinically healthy individuals, with salivary 8-oxodG levels in individuals exhibiting periodontal disease, being significantly elevated [25], and related to the presence of periodontopathic bacteria [26]. With the exception of Bogdanov et al. [27], there are no reports of chromatographic techniques attempting to measure 8-oxodG in saliva. Furthermore, given the presence of 8-oxodG in many body fluids, it is not clear whether salivary 8-oxodG is specific solely for oxidative stress in the buccal cavity, or may be regarded as another matrix for the noninvasive assessment of whole body oxidative stress. Herein we investigate the applicability of two commercially available ELISAs to the measurement of salivary and urinary 8-oxodG, along with comparison to an LC-MS/MS assay.