Integrated network pharmacology and experimental verification to reveal the mechanisms of curcumin in the treatment of colorectal cancer

Background: Curcumin is a natural compound with potent anti-tumorigenic properties, demonstrating significant efficacy in treating colorectal cancer (CRC). However, the mechanisms underlying this anti-tumor effect remain unelucidated. This study aimed to explore curcumin's potential mechanisms in the CRC treatment via integrated network pharmacology, bioinformatics and experimental validation. Methods: Curcumin and CRC targets were obtained from public databases, with differentially expressed genes (DEGs) from RNA-seq. Network pharmacology-based prediction was employed to elucidate the potential mechanisms by which curcumin exerts its anti-CRC effects. Single-cell analysis was used to explore the expression of hub genes in CRC's tumor microenvironment (TME). Least absolute shrinkage and selection operator (LASSO) Cox analyses were used to construct a prognostic model. Molecular docking was employed to investigate the interactions between curcumin and hub genes. Molecular dynamics (MD) simulation was carried out to provide further verification of the findings. In vitro and xenograft mouse model experiments were conducted to validate the effects of curcumin. Results: A total of 46 potential targets were obtained. Functional enrichment analysis revealed that the potential gene set was significantly enriched in ferroptosis and the Wnt/β-catenin signaling pathway. 11 hub genes were identified from PPI network. Single-cell analysis of the hub genes indicated that their aberrant expression profiles was associated with the TME of CRC. A four-gene prognostic model, including SIRT1, SERPINE1, MMP3 and WNT5A, was constructed from the hub genes. Curcumin exerts regulatory effects on mast cells, fibroblasts, and plasma components within the context of immune modulation. Molecular docking studies showed that curcumin exhibits strong binding affinity to the hub targets with high docking scores (binding energies ≤ -6 kcal/mol), which was further confirmed by MD simulation. In addition, curcumin treatment promoted accumulation of lipid ROS (p