Discovery of lipid biomarkers between bovine colostrum and milk using UHPLC-Q-TOF-MS lipidomics

Abstract

Sixty-three significantly different lipids from bovine colostrum (BC) and bovine milk (BM) were quantified using ultra-high-performance liquid chromatography-Q exactive orbitrap-mass spectrometry lipidomics and selected based criteria that included area under curve (AUC) value of the receiving operator characteristic (ROC) curve analysis >0.9. Four potential lipid biomarkers, including two up-regulated lipids, namely PE (P-16:0/22:5), and PE (P-16:0/20:4), and two down-regulated lipids, Hex2Cer (d14:0/24:0), and Hex2Cer (d15:0/24:1), were selected as a potential lipid biomarkers for further analysis. The orthogonal projections to latent structures-discriminate analysis (OPLS-DA), permutation tests of OPLS-DA, and receiving operator characteristic (ROC) curve analysis were further implemented to validate the reliability and stability of the potential lipid biomarkers. The four lipid biomarkers had an AUC value of 1 in the ROC curve analysis, showing a good classification performance. These results could contribute to purity analysis of BC products and a theoretical basis for the detection of lipids over lactation.