Per- and polyfluoroalkyl substances (PFAS) are widely used persistent synthetic chemicals that have been linked to adverse health effects. While the behavior of PFAS has been evaluated in the environment, our understanding of reaction products in mammalian systems is limited. This study identified biological PFAS transformation products and generated mass spectral libraries to facilitate an automated search and identification. The biological transformation products of 27 PFAS, spanning 5 chemical subclasses (alcohols, sulfonamides, carboxylic acids, ethers, and esters), were evaluated following enzymatic reaction with mouse liver S9 fractions. Four major pathways were identified by liquid chromatography-high-resolution mass spectrometry: glucuronidation, sulfation, dealkylation, and oxidation. Class-based fragmentation rules and associated PFAS transformation product libraries were generated and integrated into an automated nontargeted PFAS data analysis software (FluoroMatch). Fragmentation was additionally predicted for the potential transformation products of more than 2,500 PFAS in the EPA CompTox Chemicals Dashboard PFASSTRUCTv4. Generated mass spectral libraries were validated by applying FluoroMatch to a data set of urine from aqueous film-forming foam (AFFF)-dosed mice. Toxicity predictions showed identified PFAS transformation products to be potential developmental and mutagenic toxicants. This research enables more comprehensive PFAS characterization in biological systems, which will improve the assessment of exposures and evaluation of the associated health impacts.
Keywords: FluoroMatch; LC-MS/MS; biological transformation; glucuronide conjugate; mass spectrum library; per- and polyfluoroalkyl substances (PFAS).