Non-targeted analysis of lipidic extracts by high-resolution mass spectrometry to characterise the chemical exposome: Comparison of four clean-up strategies applied to egg

Antonin Padioleau,Ronan Cariou,Ingrid Guiffard,Bruno Le Bizec,Beate I. Escher,Jean-Philippe Antignac,Gaud Dervilly

DOI: https://doi.org/10.1016/j.jchromb.2023.123963

IF: 3.318

2024-01-01

Journal of Chromatography B

Abstract:Biota samples are used to monitor chemical stressors and their impact on the ecosystem and to describe dietary chemical exposure. These complex matrices require an extraction step followed by clean-up to avoid damaging sensitive analytical instruments based on chromatography coupled to mass spectrometry. While interest for non-targeted analysis (NTA) is increasing, there is no versatile or generic sample preparation for a wide range of contaminants suitable for a diversity of biotic matrices. Among the contaminants' variety, persistent contaminants are mostly hydrophobic (mid- to non-polar) and bio-magnify through the lipidic fraction. During their extraction, lipids are generally co-extracted, which may cause matrix effect during the analysis such as hindering the acquired signal. The aim of this study was to evaluate the efficacy of four clean-up methods to selectively remove lipids from extracts prior to NTA. We evaluated (i) gel permeation chromatography (GPC), (ii) Captiva EMR-lipid cartridge (EMR), (iii) sulphuric acid degradation (H₂SO₄) and (iv) polydimethyl siloxane (PDMS) for their efficiency to remove lipids from hen egg extracts. Gas and liquid chromatography coupled with high-resolution mass spectrometry fitted with either electron ionisation or electrospray ionisation sources operating in positive and negative modes were used to determine the performances of the clean-up methods. A set of 102 chemicals with a wide range of physico-chemical properties that covers the chemical space of mid- to non-polar contaminants, was used to assess and compare recoveries and matrix effects. Matrix effects, that could hinder the mass spectrometer signal, were lower for extracts cleaned-up with H₂SO₄ than for the ones cleaned-up with PDMS, EMR and GPC. The recoveries were satisfactory for both GPC and EMR while those determined for PDMS and H₂SO₄ were low due to poor partitioning and degradation/dissociation of the compounds, respectively. The choice of the clean-up methods, among those assessed, should be a compromise that takes into account the matrix under consideration, the levels and the physico-chemical properties of the contaminants.

biochemical research methods,chemistry, analytical