IntroductionInsulin resistance (IR) crucially dominates the development of type 2 diabetes mellitus(T2DM), influencing the disease throughout its entire progression. It is considered as a significant risk factor for varying comorbid conditions such as hypertension, obesity, and cardiovascular diseases1,2,3. According to relevant researches, IR often manifests before the formal diagnosis of T2DM, frequently leading to a higher occurrence of complications associated with diabetes. This connection highlights the importance of addressing IR in the early stages to mitigate the progression of diabetes and its related health issues4,5. IR is particularly prevalent in liver tissue, where it has a profound impact on both glucose production and utilization. This issue not only disrupts the body’s ability to manage glucose effectively but also interferes with lipid metabolism and synthesis. The consequences of these metabolic disturbances are significant, further complicating the health of individuals at risk for or already diagnosed with type 2 diabetes4,6. Current treatment options for managing type 2 diabetes primarily include metformin, insulin sensitizers, and thiazolidinediones, among others. While these medications are widely used, it is important to acknowledge their potential side effects, e.g. gastrointestinal issues and higher risk of cardiac events. These concerns necessitate the development of safer and more reliable hypoglycemic agents. Focusing research efforts on creating new treatment options is critical for enhancing the management of insulin resistance and improving patient outcomes7.Given the potential adverse effects of current pharmaceuticals, exploring safe natural products from natural sources represents a viable strategy for IR prevention or treatment. Evodiamine is a primary indolequinazoline alkaloid extracted from the dried near-ripe fruit of Evodia rutaecarpa, a traditional Chinese medicine. It is one of the key medicinal ingredients of Evodia officinalis8. According to numerous studies, Evo possesses diverse pharmacological properties, including anti-tumor, lipid-lowering, cardioprotective, and anti-infective effects9. It is a natural drug with broad development prospects. Previous research has demonstrated that Evo decreases blood glucose levels, oxidative stress, and inflammation in T2DM animal models by modulating metabolic pathways10,11. As an essential downstream signaling cascade affected by IRS-1, the PI3K/AKT signal significantly influences glucose metabolism through enhancing glycogen synthesis and glucose uptake12,13. The activation of this signaling pathway contributes substantially to maintaining glucose homeostasis14,15 and has a recognized correlation with IR pathogenesis16,17.Although it has been shown that Evo can improve glucose tolerance and slow the progression of insulin resistance associated with diabetes and obesity by reducing IRS-1 serine phosphorylation and inhibiting mTOR-S6K signaling in adipocytes18. However, the effects of insulin resistance on other body systems, such as liver and muscle, are still poorly understood. Therefore, by integrating network pharmacology, molecular docking, and cellular experiments, the present study engaged in exploring the effect of Evo against hepatic IR, and to explore its regulation of GLUT4 in depth, so as to provide a new potential therapeutic modality for hepatic IR.Materials and methodsMaterials, reagents, and chemicalsMedChemExpress Co., Ltd (New Jersey, USA) provided Evo with a purity level reaching 99.97%. High-glucose DMEM and FBS were provided by HyClone (USA). Beyotime Biotechnology (Shanghai, China) supplied the primary antibodies targeting IRS-1, phosphorylated IRS-1 (Ser307), AKT, phosphorylated AKT, PI3K, phosphorylated PI3K, and GLUT4, along with the matching secondary antibodies. PVDF membranes were acquired from Millipore (MA, USA), and all remaining chemicals were analytical grade reagents.Pharmacological network analysisTCMSP was utilized to predict Evo’s potential targets19 and SwissTarget databases20. IR-associated genes were retrieved from GeneCards21, TTD22 and OMIM23. Evo’s candidate targets for IR therapy emerged from the intersection of drug and disease targets; database specifics are listed in Table 1.The criteria for target screening are as follows: When obtaining the targets related to Evo from the TCMSP database, the compound screening criteria of this platform are applied, namely oral bioavailability (OB) ≥ 30% and drug-likeness (DL) ≥ 0.18. For the targets obtained from the SwissTarget database, we retain all targets with a predicted probability value > 0 to maximize the candidate range. For disease targets, we search in the GeneCards database using the keyword ‘Insulin Resistance’ and set the relevance score (Relevance score) of 0 as the threshold to retain genes with a stronger association with the disease. All targets obtained from the OMIM and TTD databases are retained. Finally, the unique drug predicted targets are intersected with the disease targets to obtain the common targets for subsequent analysis.Table 1 Database Information and Retrieval URLs.Full size tableNetwork construction of common targets between evo and IRWe used R language and the Venn Diagram package to comparatively analyze gene targets related to Evo and IR. The intersection targets were potential targets for Evo in managing IR. Guided by the relatonships among Evo, intersection targets, and IR, the network of Evo in treating IR was constructed via Cytoscape 3.6.0 software24.PPI network constructionThe study imported common targets of Evo and IR into the STRING database (version 12.0), and restricted the species to Homo sapiens. The Minimum required interaction score threshold was set to 0.400 to obtain high-confidence protein-protein interactions and hide isolated nodes in the network. Subsequently, the obtained CSV format interaction data underwent visualization after being imported into Cytoscape software (version 3.6.0), and its built-in ‘Network Analyzer’ tool was applied to the calculation of the topological parameters (Degree value) of nodes to identify the core targets in the network25.Analysis of enrichment in GO and KEGG PathwaysEnrichment analysis focusing on the overlapping targets of Evo and IR was implemented through the Bioconductor package within the R statistical software environment26. For controlling the False Discovery Rate (FDR) resulting from the simultaneous examination of a large number of functions/paths, the original p-values were corrected with the Benjamini-Hochberg (BH) method. The p-values after correction (p.adjust)