Analytical strategies to profile the internal chemical exposome and the metabolome of human placenta.

ABSTRACT

As a non-invasive biological matrix, the placenta offers great and novel opportunities to monitor fetal exposure to exogenous chemicals and their biotransformation products (the internal chemical exposome), as well as the biological responses associated, in large-scale epidemiological studies. However, it is first crucial to ensure that analytical methods based on high-resolution mass spectrometry (HRMS) can detect the low abundant components of the internal chemical exposome present in these complex biological matrices. In this study, we aimed to develop a robust analytical method (extraction and sample preparation) sensitive enough to profile the internal chemical exposome and the metabolome of placenta using high-resolution mass spectrometry (HRMS) for future application in mother child cohorts. Several extraction solvents (methanol, methanol/H2O (50/50 v/v), methyl tert-butyl ether/methanol/H2O) were tested and their ability to extract components of the internal exposome and metabolome were compared. Then, sample preparation methods commonly used for metabolomics application (methanolic protein precipitation) were compared with solid phase extraction (SPE), protein and phospholipid removal plates (PPRP) and combination of SPE and PPRP. The methods were compared and validated using qualitative (i.e., numbers of features and chemical classes ID), quantitative parameters adopted from targeted multi-residue analysis (recovery experiments, repeatability and matrix effect) as well as the ability of these methods to be implemented for high-throughput applications. The analytical repeatability of the two most effective methods (methanolic extraction followed by either protein precipitation or PPRP) were tested at the batch level to determine the best concentration factors to be used for improving detection of components of the internal chemical exposome and metabolome without impacting on the analytical response. Finally, these two methods based on protein precipitation and PPRP were tested on 40 placenta samples from the French PELAGIE birth cohort, and annotation was performed on the related datasets to compare the respective impacts of PPT and PPRP. A wide range of exogenous (e.g., biocides, pharmaceuticals, personal care products) and endogenous chemicals (steroids, prostanoids, lipids, carnitins) could be detected and annotated (some of them for the first time in placenta). We show that both methods are complementary but that PPRP allows the injection of more concentrated extracts without impacting the LC repeatability and therefore improve the detection (presence and signal area fold change) of many exogenous and endogenous chemicals.