Dataset Information

Uncovering Xenobiotics in the Dark Metabolome using Ion Mobility Spectrometry, Mass Defect Analysis and Machine Learning

ABSTRACT: The identification of xenobiotics through nontargeted analysis is a vital step in understanding human exposure, however metabolism, excretion, and co-existence with other endogenous molecules in the metabolome greatly complicate their measurements. Xenobiotics are therefore commonly undetected and exist as part of the dark metabolome. While mass spectrometry (MS)-based platforms are commonly used in metabolomic measurements, deconvoluting endogenous metabolites and xenobiotics is often limited by a lack of xenobiotic parent and metabolite standards as well as the numerous isomers possible for each m/z feature. Here we evaluated the ability of ion mobility spectrometry (IMS) and mass defect filtering techniques to narrow large metabolomic feature lists to xenobiotics of interest. Due to the lack of IMS collision cross section (CCS) values for per- and polyfluoroalkyl substances (PFAS), we initially evaluated 87 PFAS standards with IMS-MS to develop a PFAS CCS library. The detected PFAS CCS and m/z values were then compared to other biomolecule and xenobiotic classes, illustrating clear differentiation between the biomolecules and the halogenated xenobiotics. To address the lack of xenobiotic standards, machine learning was then utilized to predict CCS values. Ultimately, a xenobiotic selection workflow combining experimental and theoretical CCS values and mass defect filtering was employed to evaluate PFAS features in NIST human serum. This workflow reduced the 2,423 LC-IMS-MS features to 98 possible PFAS, and 23 were identified using homologous series information.

INSTRUMENT(S): 6560 Q-TOF LC/MS

ORGANISM(S): Homo Sapiens (ncbitaxon:9606)

SUBMITTER: Erin Baker

PROVIDER: MSV000088215 | MassIVE | Mon Oct 11 13:57:00 BST 2021

REPOSITORIES: MassIVE

ACCESS DATA

Similar Datasets

Project description:The potential effects of poly- and perfluoroalkyl substances (PFAS) are a recent human and environmental health concern. There is a persistent link between PFAS exposure and cancer, but the mechanisms are poorly understood. Although epidemiological evidence supporting PFAS exposure and cancer in general is conflicting, there is a relatively strong evidence linking PFAS and testicular germ cell tumors (TGCTs). However, there have been no mechanistic studies to date cancerning PFAS and TGCTs. In this report, the effects of the legacy PFAS, perfluorooctanesulfonic acid (PFOS) and the newer "clean energy" PFAS, lithium bis(trifluoromethylsulfonyl)imide (LiTFSi and called HQ-115) on TGCT tumorigenicity, survival, metabolism, and gene regulation were investigated. In vitro, proliferation and survival of both chemo-sensitive and -resistant TGCT cells were minimally affected by a wide concentration range of PFOS and HQ-115. However, both chemicals promoted the growth of TGCT cells in mouse xenografts at doses consistenet with human exposure but had minimal acute toxicity, as assessed by total body, kidney, and testis weight. PFOS, but not HQ-115, increased liver weight. Transcriptiomic alterations of PFOS-exposed normal mouse testes were dominated by cancer-related pathways and gene expression altersations associated with the H3K27me3 polycomb pathway and DNA methylation, epigenetic pathways were previously shown to be critical for the survival of TGCT cells after cisplatin-based chemotherapy. Similar pattersn of PFOS-mediated gene expression occured in PFOS-exposed cells in vitro. Metabolomic studies revealed that PFOS also altered metabolites associiated with steroid biosynthesis and fatty acid metabolism in TGCT cells, consistent with the proposed ability of PFAS to mimic fatty acied based ligands controlling lipid metabolism and the proposed role of PFAS as endocrine disrupters. Our data, this first cell and animal based study on PFAS in TGCTs, support a pro-tumorigenic effect of PFAS on TGCT biology and suggests epigenetic, metabolic, and endocrine disruption as potential mechanisms of action that are consistent with the non-mutagenic nature of the PFAS class.

Project description:This Dataset goes alongside the 2023 manuscript: Novel perfluoroalkyl substances (PFAS) discovered in whole blood using automated non-targeted analysis of dried blood spots by Jeremy P. Koelmel, Elizabeth Z. Lin, Emily Parry, Paul Stelben, Emma E.Rennie, Krystal J.Godri Pollitt A small subset of per- and polyfluoroalkyl substances (PFAS) are routinely screened in human blood. These compounds generally explain <50 percent of the total PFAS in human blood. The percentage of known PFAS in human blood has been decreasing as replacement PFAS and more complex PFAS chemistries are introduced to the market. Most of these novel PFAS have not been previous identified. Non-targeted methods are required to characterize this dark matter PFAS. Our objective was to apply non-targeted PFAS analysis to human blood to gain an understanding about the sources, concentrations, and toxicity of these compounds. A high-resolution tandem mass spectrometry (HRMS) and software workflow for PFAS characterization in dried blood spots is reported. Dried blood spots are a less invasive collection technique compared to venous blood draws, allowing collection from vulnerable populations. Biorepositories of archived dried blood spots are available internationally from newborns and present opportunities to study prenatal exposure to PFAS. In this study, dried blood spot cards were analyzed using iterative MS/MS by liquid chromatography HRMS. Data processing was conducted using FluoroMatch Suite including a visualizer tool that presents homologous series, retention time vs m/z plots, MS/MS spectra, feature tables, annotations, and fragments for fragment screening. The researcher performing data-processing and annotation was blinded to the fact that standards were spiked in, and was able to annotated 95 percent of standards spiked on dried blood spot samples, signifying a low false negative rate using FluoroMatch Suite. A total of 28 PFAS (20 standards and 4 exogenous compounds) were detected across five homologous series with Schymanski Level 2 confidence. Of these four, 3 were perfluoroalkyl ether carboxylic acids (PFECA), a chemical class of PFAS which is increasingly being detected in environmental and biological matrices but is not currently screened in most targeted analyses. A further 86 potential PFAS were detected using fragment screening. PFAS are extremely persistent and widespread yet remain largely unregulated. Our findings will contribute to an improved an understanding of exposures. Application of these methods in environmental epidemiology studies have the potential to inform policy with regards to PFAS monitoring, regulation, and individual-level mitigation strategies.

Dataset Information

Uncovering Xenobiotics in the Dark Metabolome using Ion Mobility Spectrometry, Mass Defect Analysis and Machine Learning

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure