Molecular mechanisms of PFAS toxicity in in vitro human liver cell models, with a focus on lipid metabolism

Perfluoroalkyl substances (PFAS) are industrial chemicals with many applications, including food packaging. These chemicals have been detected in our environment and human tissues, raising concerns regarding their potential adverse health effects on humans. Exposure to PFAS is associated with hepatotoxicity, lipid metabolism disruption, and decreased immune response. This study aimed to investigate the toxicity mechanisms of several PFAS using 2D and 3D human liver cell-based in vitro models with increasing biological complexity. In Chapter 2, 2D cultures of in vitro human hepatocyte-like cells (HepaRG) were used to assess the impact of legacy PFAS (PFOS, PFOA, PFBA, PFDA, PFHxA, PFHxS, and PFNA) and alternative PFAS for these legacy PFAS (PMPP (Adona), HFPO-DA (GenX), and PFBS) on lipid composition using a lipidomics approach. The study showed that PFOS and PFOA increased ceramide, diacylglyceride, N-acylethanolamine, phosphatidylcholine, and triacylglyceride levels, whereas PMPP, HFPO-DA, PFBA, PFBS, PFHxA, and PFHxS decreased these lipid levels. In Chapter 3, six understudied PFAS (PFOSA, PFPeA, PFPrA, 6:2 FTOH, 6:2 FTSA, and 8:2 FTSA) were screened in vitro for their capacity to activate nuclear receptors in HEK293T reporter gene assays and to affect the expression of genes involved in lipid metabolism in HepaRG 2D cultures. The findings indicate PFOSA as the most potent among the six tested PFAS, capable of interfering with lipid metabolic pathways. In Chapter 4, we examined the impact of PFOSA, 6:2 FTSA, and 8:2 FTSA on HepG2 (2D) cultures, focusing on their influence on lipid metabolism. Subsequently, 3D spheroids of HepG2 cells were exposed to PFOSA, the most disruptive PFAS on gene expression and lipid profiles, to understand its effects on hepatic lipid metabolism, with PFOS used as reference compound. PFOSA suppressed the expression of cholesterol synthesis genes in both models. In PFOSA-exposed HepG2 monolayers, glycerophospholipid metabolism genes were suppressed and increased glycerophospholipid levels. In HepG2 3D spheroids, PFOSA inhibited genes involved in fatty acid metabolism with most CAR and NAE species levels decreased. In Chapter 5, 3D spheroid models consisting of HepaRG or HepG2 cells cocultured with primary human hepatic stellate cells (HSCs) were used to evaluate the impact of PFOSA and PFOS on lipid composition and gene expression. The cocultured spheroids (HepaRG+HSC and HepG2+HSC) were exposed to PFOSA or PFOS for 24 or 72 h. After PFAS exposure, coculture spheroids exhibited altered lipid profiles and gene expression in a time- and cell type-dependent manner. In HepG2+HSC spheroids, PFOSA stimulated lipogenesis at 24 h and promoted lipid degradation at 72 h, whereas PFOS enhanced lipogenesis at both 24 and 72 h. In HepaRG+HSC spheroids, PFOSA promoted lipogenesis at 24 and 72 h. Conversely, PFOS stimulated lipid degradation at 24 h and lipogenesis at 72 h. PFOSA was more potent than PFOS in suppressing genes involved in lipid metabolism in HepaRG+HSC and HepG2+HSC spheroids. In Chapter 6, the overall findings obtained in the previous Chapters 2 to 5 indicates that several PFAS examined in this dissertation, including PFOSA and PFOS, disrupted lipid metabolism by altering gene expression and lipid composition in 2D and 3D in vitro hepatocyte models. Therefore, the existence of approximately 15,000 PFAS structures, many of which lack toxicity data, is concerning. Although toxicity data exist for PFOA, PFOS, PFNA, and PFHxS, rodent studies suggest that other understudied PFAS may also be toxic to humans. We propose a tiered testing strategy for identifying PFAS that disrupt hepatic lipid metabolism. Testing begins with a 2D monolayer hepatocyte culture screening in Tier I, followed by testing of selected PFAS in a 3D spheroid hepatocyte monoculture in Tier II, and a 3D spheroid hepatocyte-stellate cell coculture, possessing cell-cell interactions in Tier III to confirm the mechanism.