A new LC/MS method for specific determination of human systemic exposure to bisphenol A, F and S through their metabolites: Application to cord blood samples

Abstract

Due to restriction of the use of BPA, several structural analogues such as BPS and BPF have been proposed for its replacement in many consumer products. This has increased the prevalence of BPS and BPF in urine from tested cohorts. However, these substitutes have similar endocrine disrupting properties to BPA, particularly on reproductive and metabolic functions, which suggests that fetal exposure to these analogues could be of concern for human health. Bisphenols (BPs) are mainly metabolized to glucuronides (BP-Gs), which are considered as inactive but provide a relevant marker of fetal exposure during pregnancy. In most instances, these metabolites are indirectly quantified after hydrolysis and measurement of the corresponding native BPs, which may lead to bias due to spurious BPs contamination during blood collection and/or analyses. We have developed a new method for direct quantification of BP-Gs, which has the advantage of not being affected by errors related to the presence of BPs. First, BP-Gs were extracted from plasma by anion exchange solid phase extraction. They were then labelled with dansyl chloride, using experimentally-optimized incubation conditions, after which the dansyl derivatives were injected into an on-line SPE-UHPLC/MS/MS system. The performance of the method, in terms of sensitivity, precision and accuracy, was evaluated in plasma over a concentration range of 0.05–5 ng/mL. The intra- and inter-day CV% precision were lower than 20% with accuracies ranging from 93% to 115%. The limit of quantification was set at 0.05 ng/mL. The method was then applied to measure BP-Gs in forty-four cord plasma samples. Although no BPF-G was found, BPA-G and BPS-G was determined in almost half of the cord plasma samples with concentration ranges nd-0.089 ng/mL and nd-0.586 ng/mL, respectively.