Multi-omics and experimental validation reveal the mechanism of DanxiaTiaoban decoction in treating atherosclerosis

Abstract

Background: Atherosclerosis (AS) is a leading risk factor for cardiovascular diseases globally, characterised by the accumulation of lipids and cholesterol in arterial walls, causing vascular narrowing and sclerosis along with chronic inflammation; this leads to increased risk of heart disease and stroke, significantly impacting patients' health. Danxia Tiaoban Decoction (DXTB), a traditional Chinese medicine (TCM) formula, has demonstrated positive clinical effects in treating AS; however, its mechanisms of action remain unclear.

Objective: To explore the potential mechanisms of action of DXTB in treating AS through multi-omics integration and experimental validation.

Method: Active components of DXTB and their targets were identified using the Traditional Chinese Medicine Systems Pharmacology Database (TCMSP), Batman, Herb, and TCM Integrated Database (TCMID). The compounds most closely associated with the active ingredients in DXTB were identified using ultra-high-performance liquid chromatography-high-resolution mass spectrometry (UHPLCHRMS). In addition, targeted quantification of quercetin, luteolin and alisol C in DXTB and in mouse serum collected 2 h after high-dose oral gavage was performed using high-performance liquid chromatography-triple quadrupole tandem mass spectrometry (LC-QQQ-MS/MS). By analysing deCODE plasma protein quantitative trait loci (pQTL) data and multiple Gene Expression Omnibus (GEO) datasets, proteins and differentially expressed genes (DEGs) associated with AS were identified. Twelve machine learning algorithms were employed to select core genes, which were then evaluated using nomogram and shapley additive explanations (SHAP) values to assess their effect on disease risk and model outputs. A drug-component-target network was constructed to identify the core active components of the drug. Mendelian randomization (MR) analysis was used to verify the causal relationship between core genes and AS, while molecular docking and molecular dynamics (MD) simulations were employed to evaluate interactions between DXTB active components and target proteins, which were validated using surface plasmon resonance (SPR) experiments. Additionally, AS was induced in Apolipoprotein E gene knockout (ApoE^-/-) mice fed with high-fat diet and treated with low, medium, and high doses of DXTB by gavage. Aortic tissue pathological changes were examined using haematoxylin and eosin (H&E) staining, transmission electron microscopy (TEM), and Oil Red O staining. Reverse transcription quantitative polymerase chain reaction (RT-qPCR) and western blot were used to verify the expression of core genes and the activation of core pathways, and changes in inflammatory factor levels were measured using enzyme-linked immunosorbent assay (ELISA).

Results: A total of 51 common target genes associated with DXTB and AS were identified, primarily enriched in the lipid and AS and Fc epsilon RI (FcεRI) signalling pathways, with the p38 mitogen-activated protein kinase (MAPK) signalling pathway potentially serving as the key mechanism by which DXTB regulates AS. Six core genes-colony-stimulating factor 1 receptor (CSF1R), dipeptidyl peptidase 4 (DPP4), neutrophil cytosol factor 1 (NCF1), matrix metalloproteinase-9 (MMP9), integrin alpha l (ITGAL), and LYN proto-oncogene (LYN)-were selected using machine learning algorithms. A multivariate logistic regression model was constructed based on these core genes, and their specific contributions and diagnostic value were demonstrated through SHAP value analysis and Nomogram, highlighting the model's potential in enabling clinical decision-making. MR analysis further suggested a causal relationship between CSF1R and AS risk. The AddModuleScore analysis indicated that core gene set expression and MAPK signalling pathway are particularly active in monocytes. Molecular docking, MD simulations, and SPR collectively confirmed a pronounced binding affinity between the active constituents and the core targets, consistent with results from network pharmacology and machine learning analyses. LC-QQQ-MS/MS quantification corroborated these findings by confirming measurable systemic exposure of the core constituents at 2 h post-dose. Furthermore, animal experiments showed that DXTB reduced lipid accumulation, inhibited activation of p38 MAPK signalling pathway, regulated the expression of core genes, and decreased inflammatory factors interleukin (IL)-1β, IL-6, and tumour necrosis factor (TNF)α, leading to reduced plaque area.

Conclusion: This study innovatively integrates multi-omics analyses, advanced machine learning algorithms, and rigorous experimental validation to systematically elucidate the therapeutic mechanisms of DXTB in the treatment of AS. Our findings demonstrate for the first time that DXTB may exert its therapeutic effects by modulating key inflammation- and lipid-associated pathways, particularly the p38 MAPK signalling pathway, as well as core genes including CSF1R, DPP4, and MMP9. In addition, DXTB may alleviate vascular inflammation and lipid accumulation by inhibiting the differentiation of monocytes into macrophages and their subsequent transformation into foam cells. The integrated approach combining bioinformatics with experimental validation provides strong support for the potential clinical efficacy of DXTB and identifies novel candidate targets for AS therapy. These insights enhance the current understanding of DXTB's mechanisms of action and offer a valuable reference for future research and therapeutic development in cardiovascular disease.

Keywords: Atherosclerosis; Danxiatiaoban decoction; MAPK signaling pathway; Machine learning; Molecular docking.