A metabolomics approach for monitoring thermal processing temperature of bovine milk using ultra-performance liquid chromatography tandem Q-Exactive mass spectrometry and multivariate data analysis

Due to the lack of effective assessment method, overheated milk commodities are often marketed as pasteurized milk on the market, which was sold in high price by fraud. Thus, this article aims to establish an approach based on metabolomics to monitor thermal processing temperature of bovine milk. Metabolomics data of bovine milk samples heated at temperatures ranging from 60 °C to 150 °C were achieved by ultra-performance liquid chromatography-high-resolution mass spectrometry (UPLC-HRMS) platform, followed by multivariate data analysis. A regular variation pattern of chemical composition as temperature rises was pictured, furthermore, Nε-carboxymethyl lysine (CML), Nε-carboxyethyl lysine (CEL), pentosidine, pyrraline and lysinoalanine (LAL) were identified as 5 of the most contributed compounds to discriminate pasteurized and ultra-high-temperature (UHT) milk. By the comprehensive study on their content changes, we concluded that the optimal temperature range was 90−100 °C for the generation of CML and CEL in this experiment, moreover, 110−120 °C for LAL, 80−100 °C for pentosidine and 130−140 °C for pyrraline. Finally, a predicted rule to discriminate pasteurized and UHT milk was preliminarily established based on the ratios of CML/CEL, CEL/pentosidine and CML/pentosidine, which could be applied in food labelling authentication of commercial bovine milk after further validation.