Simultaneous measurement of kynurenine metabolites and explorative metabolomics using liquid chromatography-mass spectrometry: A novel accurate method applied to serum and plasma samples from a large healthy cohort

Kynurenine metabolites are emerging as promising clinical biomarkers in several diseases, especially within psychiatry. Unfortunately, they are difficult to detect, particularly the challenging neurotoxic metabolite qui-nolinic acid (QUIN). The aim of this study was twofold: First, to develop a liquid chromatography-mass spec-trometry method (LC-MS) for simultaneous targeted quantification of key kynurenine metabolites together with untargeted metabolomics, and second, to demonstrate the feasibility of the method by exploring serum/plasma and gender differences in 120 healthy young adults between 18 and 30 years of age. A range of analytical columns (C18 and biphenyl columns) and mobile phases (acidic and alkaline) were systematically evaluated. The optimized LC-MS method was based on a biphenyl column, a water-methanol gradient with 0.2% formic acid, and authentic isotope-labeled standards for each kynurenine metabolite. Precision and accuracy of targeted quantification of the key kynurenine metabolites tryptophan (TRP), kynurenine (KYN), kynurenic acid (KYNA), 3-hydroxykynurenine (3-HK), and QUIN were excellent, far exceeding the acceptance criteria specified by in-ternational guidelines. Median inter-and intra-day precision were < 6% in serum and plasma; the median ac-curacy was 2.4% in serum and 8% in plasma. Serum concentrations were < 10% different from the corresponding concentrations in plasma for all kynurenine metabolites in healthy young adults. Men had higher levels (8-18%) of TRP, KYN, and KYNA than women (p < 0.009), while no differences were observed for 3-HK and QUIN (p > 0.70). Incurred sample reanalysis of 10% of the samples yielded a median difference < 5% from the initial measurement, demonstrating the robustness of the method. Besides the targeted quantification of key kynurenine metabolites, our method was found to be suitable for simultaneous untargeted metabolomics ana-lyses of hundreds of metabolites. A range of compound classes could be detected including amino acids, nucleic acids, dipeptides, antioxidants, and acylcarnitines, making explorative studies highly feasible. For example, we identified an additional kynurenine metabolite, 2-Quinolinecarboxylic acid, which was 47% higher in males than females (adjusted p-value = 0.001). In conclusion, in this study, we present a reliable and robust LC-MS method for simultaneous targeted and untargeted metabolomics ready for both research and clinical use. We show that both serum and plasma can be used for kynurenine studies, and the reported gender differences are in accordance with the literature. Future studies should consider using biphenyl-based LC-MS columns to successfully detect QUIN.

Automated summary

In this study, a reliable and robust LC-MS method was developed for simultaneous targeted and untargeted metabolomics. This method was shown to be suitable for the quantification of key kynurenine metabolites and the analysis of other metabolites. The study also revealed gender differences in kynurenine metabolite levels.