High-Coverage Quantitative Metabolomics of Human Urine: Effects of Freeze–Thaw Cycles on the Urine Metabolome and Biomarker Discovery

• Published in: Analytical chemistry (Washington) Vol. 94; no. 27; pp. 9880 - 9887
• Main Authors: Chen, Deying, Chan, Wan, Zhao, Shuang, Li, Lanjuan, Li, Liang
• Format: Journal Article
• Language: English
• Published: Washington American Chemical Society 12.07.2022
• Subjects: analytical chemistry; Biomarkers; Canada; Chemistry; China; Dansylation; Freeze thaw cycles; Freeze-thawing; Human wastes; humans; isotopes; Low temperature; Metabolites; metabolome; Metabolomics; phenol; Phenols; Sample preparation; Temperature requirements; Urine; Workflow; Canada; China
• ISSN: 0003-2700; 1520-6882; 1520-6882
• DOI: 10.1021/acs.analchem.2c01816

Urine sample storage after collection at ultra-low-temperature (e.g., −80 °C) is normally required for comparative metabolome analysis of many samples, and therefore, freeze–thaw cycles (FTCs) are unavoidable. However, the reported effects of FTCs on the urine metabolome are controversial. Moreover, there is no report on the study of how urine FTCs affect biomarker discovery. Herein, we present our study of the FTC effects on the urine metabolome and biomarker discovery using a high-coverage quantitative metabolomics platform. Our study involved two centers located in Hangzhou, China, and Edmonton, Canada, to perform metabolome analysis of two separate cohorts of urine samples. The same workflow of sample preparation and dansylation isotope labeling LC–MS was used for in-depth analysis of the amine/phenol submetabolome. The analysis of 320 samples from the Hangzhou cohort consisting of 80 healthy subjects with each urine being subjected to four FTCs resulted in relative quantification of 3682 metabolites with 3307 identified or mass-matched. The analysis of 176 samples from the Edmonton cohort of 44 subjects with four FTCs quantified 3516 metabolites with 3166 identified or mass-matched. Multivariate and univariate analyses indicated that significant variations (fold change ≥ 1.5 with q-value ≤ 0.05) from FTCs were only observed in a very small fraction of the metabolites (<0.3%). Moreover, various metabolites did not show a consistent pattern of concentration changes from one to four FTCs, allowing the use of two separate cohorts of samples to remove these randomly changed metabolites. Three metabolite biomarkers for separating males and females were discovered, and FTC did not influence their discovery.