Network toxicology deciphers micro- and nanoplastics-mediated mixture hazard, predictive risk assessment, and regulatory translation.

Micro- and nanoplastics (MNPs) are pervasive environmental contaminants and efficient carriers of coexisting pollutants, including heavy metals, organic chemicals, and antibiotics. Their capacity to adsorb, transport, and release contaminants has raised growing concern over mixture toxicity under realistic exposure scenarios. This review systematically examines the mechanistic basis of MNPs-mediated combined toxicity and evaluates the emerging role of network toxicology as a systems-based tool for hazard assessment. MNPs alter the bioavailability, environmental fate, tissue distribution, and intracellular delivery of associated pollutants through hydrophobic, electrostatic, and other intermolecular interactions. Carrier-mediated uptake, particularly the "Trojan horse" effect, appears to be a major driver of non-additive toxicity in co-exposure systems. We then outline the core workflow of network toxicology, encompassing target identification, network construction, pathway enrichment, and experimental validation, and discuss its application in decoding mixture toxicity. Oxidative stress, inflammatory signaling, metabolic disturbance, barrier dysfunction, and programmed cell death emerge as conserved and interconnected pathways underlying synergistic multi-organ injury. These findings indicate that toxicity in complex exposure systems is governed not only by the intrinsic properties of particles or chemicals, but also by their dynamic physicochemical and biological interactions. We further assess current advances and limitations in network toxicology and propose a next-generation risk assessment (NGRA)-oriented framework to support mechanism-based risk assessment and regulatory decision-making. Although current evidence is dominated by binary, high-dose laboratory studies, network toxicology offers strong potential as a new approach methodology (NAM) for predictive evaluation of environmentally relevant mixtures. Future priorities include standardized multi-omics integration, dose-time-response modeling, human-relevant validation, and regulatory translation.