LC-MS-based untargeted metabolomics reveals benzoic acid as a predictive biomarker for embryo implantation potential

Evaluating the quality of embryos and implantation potential is a critical determinant of <i>in vitro</i> fertilization and embryo transfer, and it is also one of the main challenges of assisted reproductive technology. A reliable non-invasive method to choose the best candidate with real implantation potential for transfer from two day-3 embryos with equal morphological quality is still lacking clinically. In this article, a sensitive LC-MS method was developed and metabolomic profiling analysis in a 3-day embryo culture medium was performed. Differential metabolites were analysed in two kinds of commercial culture media, and a total of 66 common metabolites were obtained from 106 independent samples in 5 batches. The relationship between changes in key metabolite, benzoic acid, concentration and the embryo implantation result was discovered. This work improved coverage through conditional optimization, enhanced the reliability of omics data through multi-batch validation, and provided a potential biomarker for evaluating the implantation potential of day-3 embryos.