Toxic effects of perfluoroalkyl and polyfluoroalkyl substances (PFAS) on the gut microenvironment and their potential...

Per- and polyfluoroalkyl substances (PFAS) are persistent surfactants with ingestion as a major exposure route, positioning the intestine as a primary site of contact. This narrative review integrates mechanistic toxicology, multi-omics microbiology, and human observational studies to evaluate whether PFAS-associated disruption of intestinal homeostasis could contribute to colorectal cancer (CRC). In vitro epithelial systems and animal models indicate that selected PFAS can impair barrier function through altered membrane properties and reduced tight-junction expression, increasing paracellular permeability and luminal antigen translocation. PFAS may also perturb goblet-cell secretion and mucus organization, in part through endoplasmic reticulum (ER) stress and altered autophagy, thereby facilitating mucosa-associated bacterial adherence. Stress signaling can converge on mitochondrial dysfunction and reactive oxygen species (ROS) generation that primes inflammasome activity and cytokine-mediated inflammation. In colon cell models, PFOA and PFOS have been associated with modulation of Wnt/β-catenin signaling and downstream transcriptional programs linked to proliferative and invasive phenotypes. At the community level, exposure-associated dysbiosis includes loss of butyrate-producing taxa and disruption of bile acid pools, consistent with reduced short-chain fatty acids (SCFAs) and altered farnesoid X receptor signaling. However, epidemiologic findings remain inconsistent, including null and inverse associations that may reflect reverse causation from occult bleeding, exposure misclassification, residual dietary confounding, non-monotonic dose responses, congener heterogeneity, and species-specific toxicokinetics. We propose priorities for future work including long-lag prospective sampling, physiologically based pharmacokinetic (PBPK)-informed exposure reconstruction, and adverse outcome pathway (AOP)-anchored multi-omics endpoints for causal inference and regulation.