Untargeted Metabolomics Reveals the Effects of Pre-Analytic Storage Conditions on Serum Metabolite Profiles in Cats

Published: 7 January 2022| Version 1 | DOI: 10.17632/bkw33tpg36.1
Patrick Barko


Untargeted metabolomics analysis of serum samples has been used to characterize metabolic disturbances associated with various clinical disorders in cats. However, the pre-analytic stability of serum metabolites exposed to extended storage in clinical freezers at -20C compared with preservation at temperatures below -70C is unknown. Our objective was to compare serum metabolite profiles in samples collected from healthy cats that were stored at -20C for up to 12 months to samples stored at -80C. We collected serum from 8 healthy cats and stored aliquots at -20C for 6 months, -20C for 12 months, and -80 for 12 months. We used untargeted liquid chromatography-mass spectrometry to generate serum metabolites profiles. We compared the relative concentrations of serum metabolites and assessed the repeatability of metabolite measurements among the different storage conditions. 733 named metabolites were detected in the sera of 8 healthy cats. Principal component analysis and hierarchical clustering of samples based on the Euclidian distance revealed clustering of samples by individual cat rather than storage condition. Univariate statistics identified 23 serum metabolites that differed significantly among storage conditions. Intraclass correlation analysis revealed that highly stable metabolites were enriched in the benzoate, creatine, bile acid, and various amino acid sub-pathways, whereas unstable metabolites were enriched in the monoacylglycerol, fatty acid, and gamma-glutamyl amino acid sub-pathways. In random forest analysis, several amino acids and gamma glutamyl acids held discriminatory power in differentiating among storage conditions. Our findings revealed that the majority of serum metabolites were stable during storage at -20C and were consistent with increasing oxidation of samples with extended storage time at -20C.


Steps to reproduce

Two files have been uploaded to this archive: 1) "feline_stab_meta.csv" contains the sample metadata for the 8 cats included in this study 2) "feline_stab_scaledimp.csv" contains the serum metabolite measurements (median scaled with missing values imputed) and the metabolite pathway details for each biochemical R code used for the analysis is located in the GitHub repository associated with this project: https://github.com/pcbarko/Barko-Feline-Serum-Metabolome-Stability


University of Illinois at Urbana-Champaign


Metabolome Analysis