Bridging Targeted (Zeno MRM-HR) and Untargeted (SWATH) LC-HRMS in a Single Run for Sensitive Exposomics

Vinicius Verri Hernandes,Benedikt Warth
DOI: https://doi.org/10.1021/acs.analchem.4c01630
2024-08-06
Abstract:Traditionally, chemical exposure has been assessed by low-resolution mass spectrometry via targeted approaches due to the typically extremely low concentration of such compounds in biological samples. Nevertheless, untargeted approaches are now becoming a promising tool for a broader investigation of the exposome, covering additional compounds, their biotransformation products, and possible metabolic alterations (metabolomics). However, despite broad compound coverage, untargeted metabolomics still underperforms in ultratrace biomonitoring analysis. To overcome these analytical limitations, we present the development of the first combined targeted/untargeted LC-MS method, merging MRM-HR and SWATH experiments in one analytical run, making use of Zeno technology for improved sensitivity. Multiple reaction monitoring transitions were optimized for 135 highly diverse toxicants including mycotoxins, plasticizers, PFAS, personal care products ingredients, and industrial side products as well as potentially beneficial xenobiotics such as phytohormones. As a proof of concept, standard reference materials of human plasma (SRM 1950) and serum (SRM 1958) were analyzed with both Zeno MRM-HR + SWATH and SWATH-only methodologies. Results demonstrated a significant increase in sensitivity represented by the detection of lower concentration levels in spiked SRM materials (mean value: 2.2 and 3 times lower concentrations for SRMs 1950 and 1958, respectively). Overall, the detection frequency was increased by 68% (19 to 32 positive detections) in the MRM-HR + SWATH mode compared to the SWATH-only. This work presents a promising avenue for addressing the outstanding key challenge in the small-molecule omics field: finding a balance between high sensitivity and broad chemical coverage. It was demonstrated for exposomic applications but might be transferred to lipidomics and metabolomics workflows.
What problem does this paper attempt to address?