Development of a QSAR model for predicting PPARα activation by PFAS based on human in vitro data of a comprehensive panel of legacy and novel PFAS

Per- and polyfluoroalkyl substances (PFAS) are a large family of persistent environmental contaminants. Some PFAS are known to bioaccumulate, are frequently detected in human serum, and are associated with several adverse effects on the immune system, the endocrine system, and the liver. PFAS-mediated activation of the peroxisome proliferator-activated receptor alpha (PPARα), which plays a key role in lipid and cholesterol homeostasis, is suggested to be an important molecular initiating event triggering PFAS toxicity. The aim of this study was to evaluate the PPARα activation potential of a diverse panel of 34 PFAS congeners, consisting of both legacy and novel compounds, using a PPARα-dependent transactivation assay in transfected HEK293T cells. The resulting concentration-response data were analyzed using benchmark dose (BMD) modelling to quantify PPARα activation potency. A key finding was that PFAS with a sulfonic acid group showed a lower potency compared to those with a carboxylic group. The most potent activators belonged to the perfluoroalkylether carboxylic acid (PFECA) subgroup. Computational descriptors were generated to characterize each congener, and quantitative structure-activity relationship (QSAR) modelling was applied to relate molecular features to in vitro PPARα activation potency, as expressed by BMD estimates. For prioritization purposes in the context of PFAS hazard characterization, the QSAR model was used to screen about 10,000 PFAS congeners. Of these, roughly 10% were within the defined applicability domain of the developed QSAR model. Predicted BMD concentrations for PPARα activation were in the range between 0.3 µM and 420 µM.