Integrative network and computational toxicology reveal the molecular mechanisms in PFOA-induced spermatogenic disorder

Abstract

Perfluorooctanoic acid (PFOA), a widely used industrial chemical, poses significant environmental and biological toxicity, particularly affecting reproductive health. This study aimed to integrate network toxicology, machine learning, and molecular dynamics simulations (MDS) to uncover the molecular mechanisms of PFOA-induced spermatogenic toxicity. Toxicity profiling using admetSAR revealed that PFOA exhibited pronounced reproductive toxicity and a strong binding affinity to nuclear receptors, including estrogen, androgen, and PPAR gamma. By integrating PFOA targets derived from toxicology databases with differentially expressed genes associated with non-obstructive azoospermia, we pinpointed 256 differentially expressed spermatogenic toxicity targets from an initial pool of 4311 potential PFOA targets. Gene ontology (GO) and KEGG pathway enrichment analyses highlighted biological processes, such as spermatogenesis and cell cycle regulation, along with pathways related to cell division and intercellular communication. Protein-protein interaction networks and machine learning algorithms (LASSO, SVM-RFE, RF) pinpointed five core genes-RAD51, KIF15, PTTG1, BIRC5, and CDC25C-that serve as potential diagnostic biomarkers. Molecular docking revealed strong binding affinities between PFOA and these proteins, with RAD51 showing the highest binding stability (-8.467 kcal/mol). Furthermore, MDS confirmed stable interactions, with low RMSD, RMSF, and Rg values, indicating structural stability. In vivo studies showed that PFOA exposure (1 and 5 mg/kg) caused testicular damage in mice in a dose-dependent manner, with significant downregulation of core target proteins; in vitro experiments demonstrated a concentration-dependent reduction in GC1 cell viability and substantial alterations in its gene expression. This study highlights the critical roles of these mechanisms through which PFOA disrupts spermatogenesis, emphasizing core biomarkers that may serve as therapeutic targets. Our findings contribute insights into the reproductive toxicity of PFOA and similar environmental pollutants, offering a basis for developing strategies to protect male fertility.

Keywords: Machine learning; Molecular docking; Molecular dynamics simulation; Network toxicology; PFOA; Reproductive toxicity; Spermatogenesis.