A chemical derivatization-based pseudotargeted liquid chromatography-tandem mass spectrometry method for sensitive and high coverage determination of bile acids in human serum

Global profiling of bile acids (BAs) is imperative for understand their function and disease pathogenesis. But it is still a challenging task, as the collision-induced dissociation (CID) fragment ions of unconjugated BAs showed low ion intensities to insufficient analysis. Herein, we developed a highly sensitive method for pseudotargeted profiling of BAs by chemical derivatization. In the developed method, a labeling reagent, 2-dimethylaminoethylamine (DMED), was adopted to label the carboxyl group of BAs. The results demonstrated that the detection sensitivities of unconjugated BAs were increased by 4-200 folds after DMED-labeling. Moreover, to profile other potential BAs not included in the 91 known targets, diverse survey experiments were performed on Qtrap-MS to search BAs for both precursor and fragment ion species, and retention index (RI) strategy was adopted to facilitate the identification of isomers. Finally, MRM-based LC-MS/MS method was validated for the pseudotargeted profiling of the BAs submetabolome with good linearity (r2 ≥ 0.990 for 89 known BAs) and high sensitivity (0.05-0.5 ng/mL for unconjugated BAs), covering unconjugated, glycine, taurine, sulfuric acid, glucuronic acid, and as well as those doubly-conjugated with above types. With this method, a total of 107 BAs, covering 54 BAs identified by authentic standards and 53 BAs candidates, were successfully determined in human serum of women with intrahepatic cholestasis of pregnancy (ICP). Multivariate analysis revealed deferentially expressed BAs. ICP disease altered the BAs profile with a reduced proportion of unconjugated, sulfate- and doubly-conjugated BAs and an increased proportion of glycine and taurine conjugates. Altered proportion and profile of BAs in ICP groups were gradually recovered during the ursodeoxycholic acid (UDCA) therapy. Overall, the strategy of DMED-labeling technique combined with diverse survey experiments is sufficiently sensitive and robust to comprehensively analysis of metabolic profiling of BAs in human serum.