Integrated analysis of network pharmacology, machine learning, and cell experiments revealed the mechanism by which Neoastilbin (a component of Xiaobi Pill) alleviates rheumatoid arthritis in vitro via targeting the MMP13/NF-κB pathway

Background In recent years, various Chinese medicines have been widely applied in the treatment of rheumatoid arthritis (RA). This study integrated network pharmacology, machine learning, and cell experiments, aiming to explore the action mechanism of Xiaobi Pill in alleviating RA. Objectives To identify key targets of Xiaobi Pill in RA and validate the underlying mechanism Methods Network pharmacology analysis was performed through screening of active ingredients from Xiaobi Pill, online target prediction, Kyoto Encyclopedia of Genes and Genomes (KEGG), and Gene Ontology (GO) enrichment analysis. Candidate targets were subsequently screened by machine learning (LASSO and RF algorithms). Molecular docking was completed using AutoDock Vina. Tumor necrosis factor-alpha (TNF-α) was used for in vitro RA induction in human fibroblast-like synoviocytes (HFLS). Cell viability was detected by Cell Counting Kit-8 (CCK-8) assay. The protein expression was measured using Western blotting. Inflammatory cytokines were examined using Enzyme-Linked Immunosorbent Assays (ELISA). Cell apoptosis was assessed via Terminal-deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) assay. Results This study predicted 518 targets from ingredients of herbs in Xiaobi Pill, and 506 RA-associated targets. A total of 77 targets were common to Xiaobi Pill and RA. KEGG and GO enrichment analysis results showed the associated pathways and functional orientation of these 77 genes. Machine learning screened matrix metallopeptidase 13 (MMP13) and spleen tyrosine kinase (SYK) as two candidate targets, and molecular docking indicated the interactions between some ingredients and MMP13 or SYK protein. Cell experiments indicated that Neoastilbin relieved TNF-α-induced inflammation, enhanced apoptosis, and suppressed the nuclear factor kappa-B (NF-κB) pathway in HFLS cells. These regulatory effects of Neoastilbin on HFLS cells were attributed to MMP13 downregulation and subsequent NF-κB signaling inhibition. Conclusion MMP13 was screened as a key target of Neoastilbin (a component of Xiaobi Pill) for RA alleviation through network pharmacology and machine learning. Cell validation suggested that Neoastilbin mitigated inflammatory response and promoted apoptosis of HFLS in an in vitro RA model via targeting the MMP13/NF-κB pathway.