Integrative network toxicology and machine learning identify AKR1C3 as a candidate functional target of cypermethrin in colorectal cancer.

The potential contribution of foodborne pesticide residues to colorectal cancer (CRC) remains insufficiently understood. In this study, we integrated computational toxicology, network analysis, machine learning, molecular docking, and in vitro validation to investigate possible molecular links between representative foodborne pesticides and CRC. The toxicological properties of ten pesticides were evaluated using ADMETlab 3.0 and ProTox-III. Pesticide-associated target genes were predicted using TargetNet, SuperPred, and the Similarity Ensemble Approach and were intersected with CRC-related genes obtained from GeneCards and OMIM. Functional enrichment analysis showed that the overlapping genes were enriched in xenobiotic response, oxidative stress, and cancer-related pathways. Using random survival forest and least absolute shrinkage and selection operator regression, six hub genes, including PDGFRA, AKR1C3, PDGFRB, CDC42, PIK3CA, and CYP2C9, were identified and used to construct a prognostic model with favorable predictive performance. Among these candidates, AKR1C3 exhibited strong predicted binding affinity with several pesticides, particularly cypermethrin and deltamethrin. Cellular thermal shift assay further supported a direct interaction between cypermethrin and AKR1C3. Cypermethrin reduced cell viability, DNA synthesis, and AKR1C3 protein expression in NCM460 cells. To improve CRC relevance and functional validation, HCT116 colorectal cancer cells were further included. Cypermethrin decreased HCT116 cell viability and AKR1C3 expression, whereas AKR1C3 overexpression partially restored cypermethrin-induced viability reduction and attenuated intracellular ROS accumulation. Together, these findings identify AKR1C3 as a candidate functional target of cypermethrin and suggest that AKR1C3 may be involved, at least in part, in cypermethrin-induced CRC-relevant cellular injury and oxidative stress. This study provides an exploratory framework for identifying exposure-related molecular targets linking foodborne pesticide residues to CRC-associated biological alterations.