Abstract Ovarian cancer (OC) is a leading cause of mortality among gynecologic malignancies. Cisplatin (DDP) is a first-line chemotherapy agent, but resistance to DDP often develops, compromising its efficacy. Astragalus membranaceus (AS), a traditional Chinese medicine, has shown promise in enhancing chemotherapy sensitivity due to its anti-inflammatory and immunomodulatory properties. This study investigates the potential of AS to overcome DDP resistance in OC. We integrated multiple independent DDP-resistant OC datasets identified 337 DDP resistance-associated targets. Network pharmacology identified 20 active compounds in AS, with 22 potential targets related to DDP resistance. GO and KEGG analyses revealed enrichment in pathways involving inflammation and cell adhesion. Survival analysis indicated nine genes significantly associated with OC prognosis and immune infiltration. Molecular docking showed strong binding affinities between AS compounds and these targets. In vitro, assays demonstrated that AS combined with DDP significantly inhibited cell proliferation and migration while inducing apoptosis in DDP-resistant OC cells. Western blot analysis confirmed significant changes in critical proteins (IL1B, IL1A, SERPINE1, ITGA2, and AXL) with combined treatment. AS combined with DDP significantly enhances the inhibition of cell proliferation and migration while promoting apoptosis in DDP-resistant OC cells. These findings suggest that AS could be a valuable adjunct to DDP in overcoming chemoresistance in OC, potentially improving patient outcomes. Supplementary Information The online version contains supplementary material available at 10.1186/s12906-025-05066-8. Keywords: Ovarian cancer, Cisplatin resistance, Astragalus Membranaceus, Network pharmacology analysis Introduction Ovarian cancer (OC) is one of the most lethal malignancies of the female reproductive system, posing a significant threat to women’s health and quality of life worldwide [[30]1]. In 2019, there were 2,944,422 new cases of OC globally, with 198,412 deaths resulting from the disease’s progression [[31]2]. In clinical practice, OC is challenging to diagnose early, and most patients are diagnosed at an advanced stage. At this stage, cancer cells are prone to metastasize to multiple organs such as the fallopian tubes, uterus, and intestines, making it difficult for patients to achieve a good prognosis [[32]3], resulting in a low five-year survival rate [[33]4]. Cisplatin (DDP) binds to DNA as a first-line chemotherapeutic agent, disrupting its normal replication and transcription processes and inhibiting cell growth and division. It has been widely used in chemotherapy for various cancers, including OC [[34]5, [35]6]. However, as treatment progresses, patients often develop resistance to DDP, significantly reducing its efficacy and leading to relapse and progression. In specific clinical settings, DDP remains the primary treatment option; thus, enhancing or maintaining the sensitivity of cancer cells to DDP is of significant clinical importance [[36]7, [37]8]. Although various phytochemicals have been proven effective in reducing DDP-resistant by inhibiting resistance mechanisms or enhancing cytotoxic efficacy [[38]9–[39]11], the high heterogeneity of OC necessitates the discovery of new supplements for personalized treatment. Astragalus membranaceus (AS), a traditional Chinese medicine, has been widely used for many years and exhibits good biosafety. Its active components have been reported to possess multiple pharmacological effects, including anti-inflammatory, immunomodulatory, and antitumor activities [[40]12–[41]16]. These unique pharmacological properties suggest that AS combined with DDP may be advantageous in treating OC. Although the anticancer effects of AS have been preliminarily validated in some studies, the mechanisms by which it affects DDP-resistant in OC remain unclear. This study explores the potential and mechanisms of AS combined with DDP in treating DDP-resistant OC. Here, we integrated multiple independent DDP-resistant OC datasets to identify genes associated with DDP resistance and used network pharmacology analysis to determine the active components of AS and their potential targets. We also conducted further cellular experiments to validate the effects of AS combined with DDP, evaluating its impact on DDP-resistant cells’ proliferation, migration, and apoptosis. Methods Identification of DDP resistance-associated targets in OC The Gene Expression Omnibus (GEO; [42]https://www.ncbi.nlm.nih.gov/geo/) database was searched for DDP resistance-associated datasets in OC. Inclusion criteria were: (i) mRNA high-throughput data; (ii) inclusion of both DDP-sensitive and DDP-resistant OC cells; and (iii) at least three replicates per group. Based on these criteria, three datasets—[43]GSE15372, [44]GSE45553, and [45]GSE58470—were selected for analysis (Table [46]1). Table 1. The datasets of DDP resistance in OC GEO dataset Cell type Platform Number of samples Reference Sensitive Resistant [47]GSE15372 A2780 [48]GPL570 5 5 Li et al. [[49]17] [50]GSE45553 OVCAR-8 [51]GPL6244 4 4 Chowanadisai et al. [[52]18] [53]GSE58470 IGROV-1 [54]GPL6947 3 3 Arrighetti et al. [[55]19] [56]Open in a new tab Each dataset was annotated according to the probe label information provided by GEO. For each dataset, the degree of similarity between samples was counted based on Pearson correlation coefficients; the principal component analysis (PCA) was used to differentiate between groups of samples; and the differentially expressed genes between groups were identified using the R package “limma” (adj. P < 0.05, |log[2]FC| ≥ 1) The above results were visualized using the R package “ggplot2”. To identify robust DDP resistance-related targets across datasets, we applied the RobustRankAggreg (RRA) method, an established probabilistic ranking algorithm that aggregates gene rankings across multiple datasets to prioritize consistently significant genes. Unlike methods that require uniform expression trends, RRA tolerates platform heterogeneity and modest inconsistencies, thus yielding biologically meaningful markers across datasets. This method has been widely used in cancer biomarker discovery [[57]20]. In our study, the genes identified by RRA were considered DDP resistance-associated genes and used for downstream network pharmacology analysis. Collection of active components and potential targets of AS The Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP, [58]https://old.tcmsp-e.com) collects information on a wide range of Chinese herbal medicines, including their constituents and potential targets [[59]21]. We selected the active ingredients of AS based on TCMSP. The screening criteria are oral utilization (OB) ≥ 30% and drug-like properties (DL) ≥ 0.18. In addition, the Swiss Target Prediction, CTD, and STITCH databases were used to complement the potential targets of the active ingredients. The potential targets from different databases were harmonized based on the UniProt database. Construction of the interaction network affecting OC sensitivity to DDP by AS We utilized Cytoscape 3.10.0 to construct and visualize the interaction network between AS, active compounds, potential targets, and DDP resistance. Some targets were shared between AS active compounds and DDP resistance, which can be considered potential therapeutic targets to enhance OC sensitivity to DDP. Protein-protein interaction (PPI) networks for potential therapeutic targets were generated based on the STRING online website. The Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses of these potential therapeutic targets were conducted using the DAVID online platform [[60]22]. Subsequently, Cytoscape was used to construct the AS active compounds-potential therapeutic targets-signaling pathways-DDP resistance network. Survival analysis Correlations between potential therapeutic targets expression and OV progression-free survival (PFS) and overall survival (OS) were analyzed by Kaplan-Meier plotter cancer database ([61]http://kmplot.com). Significance was determined by the hazard ratio (HR) with 95% confidence intervals and log-rank P-values. Immune infiltration analysis The Tumor IMmune Estimation Resource (TIMER) database has been widely used to analyze the correlation between gene expression and the abundance of immune cell infiltration [[62]23]. We examined the correlation between potential therapeutic targets and OV immune cell (including B cells, CD8^+ T cells, CD4^+ T cells, macrophages, neutrophils, and dendritic cells) infiltration by the TIMER database. Their correlation scores, including P-values and Pearson’s correlation coefficients, are presented as heatmaps. Molecular docking Molecular docking is a method of analyzing the binding capacity and strength between two substances using computer simulations. This method is commonly used to predict the degree of binding between small molecule compounds and target proteins. We downloaded the 3D structures of AS active compounds and potential therapeutic targets and subsequently predicted the binding capacity between them by the Swissdock online program. The interaction patterns and details were visualized by the Pymol. Cell culture DDP-resistant OC cells (SKOV3-DDP and A2780-DDP) were obtained from the American Type Culture Collection (ATCC, USA). These cell lines were cultured in RPMI-1640 medium (Qisi Biological Technology, China; QS-S402) supplemented with 1% Penicillin-Streptomycin Solution (Qisi Biological Technology, China; QS-S402) and 10% Fetal Bovine Serum (Gibco, China; F800821) and maintained at 37 °C in a 5% CO[2] incubator. Subculturing was performed when the cell density reached 80%. Cell proliferation To ensure effective drug concentrations without causing excessive cell death, the IC30 (30% of the half-maximal inhibitory concentration) was chosen as the treatment concentration for subsequent experiments [[63]24, [64]25]. The IC30 of DDP and AS for SKOV3-DDP and A2780-DDP cells were determined using the Cell Counting Kit-8 (CCK8, Qisi Biological Technology, China; QS-S321). Briefly, SKOV3-DDP and A2780-DDP cells in logarithmic growth phase were seeded at a density of 3 × 10^3 cells per well in 96-well plates with fresh culture medium. After 12 h of incubation, cells were treated with various concentrations of DDP and AS for 24 h. Subsequently, 10 µL of CCK8 working solution was added to each well, and the cells were incubated for an additional 2 h. Absorbance at 450 nm was measured using a microplate reader. For subsequent experiments, cells were seeded similarly at a density of 3 × 10^3 cells per well in 96-well plates with fresh culture medium. After 12 h of incubation, cells were randomly divided into four groups based on the previously determined IC30 values: (a) untreated control group, (b) DDP-treated group, (c) AS-treated group, and (d) combined DDP and AS-treated group. After 24 h of treatment, cell proliferation was assessed using the CCK8 assay. Absorbance at 450 nm was measured, and cell proliferation rates were normalized to the untreated control group. Cell migration To evaluate the effect of different treatments on cell migration ability, wound healing assays were performed using SKOV3-DDP and A2780-DDP cells. Cells in the logarithmic growth phase were seeded at a density of 5 × 10^5 cells per well in 6-well culture plates. Cells were incubated overnight at 37 °C in a 5% CO[2] incubator to allow attachment and formation of a monolayer. Following the grouping scheme outlined in Method 2.8, cells were treated for 24 h. After removing the culture medium, cells were washed with PBS. A 200 µL pipette tip was used to scratch the cell monolayer in each well, ensuring straight scratches consistently across all wells. Cells were then washed with PBS three times to remove detached cells. A Serum-free culture medium was added to each well, and initial images of the scratch were captured under a microscope (0 h). Plates were incubated at 37 °C in 5% CO[2] conditions, and pictures of the scratch were captured after 24 h to assess cell migration. Cell apoptosis Following the grouping scheme outlined in Method 2.8, cells were treated with respective drugs and concentrations for 24 h. SKOV3-DDP and A2780-DDP cells were harvested by trypsinization, washed twice with PBS, and centrifuged at 1200 rpm for 5 min. Apoptosis was detected using the Annexin V-FITC/PI Apoptosis Detection Kit (Qisi Biological Technology, China; QS-S306) according to the manufacturer’s instructions. Briefly, the cell pellet was resuspended in 500 µL of Binding Buffer. Subsequently, 5 µL of Annexin V-FITC was added and gently mixed, followed by 5 µL of PI. The mixture was incubated at room temperature in the dark for 5–15 min. Untreated cells were negative controls and processed similarly without Annexin V-FITC and PI. Samples were analyzed using flow cytometry to determine the percentage of apoptotic cells. Western blot Following treatment, cells were treated with respective drugs and concentrations for 24 h according to the grouping scheme outlined in Method 2.8. Total protein was extracted for Western blot analysis to detect IL-1α, IL-1β, SERPINE1, ITGA2, and AXL. To ensure accurate detection and quantification, IL-1α, SERPINE1, and ITGA2 were probed on one membrane, while IL-1β and AXL were analyzed separately on another membrane. This separation was necessary because IL-1α and IL-1β share similar molecular weights (~ 30 kDa), and ITGA2 and AXL are also close in size (~ 150 kDa vs. ~140 kDa), which could lead to signal overlap or band merging if co-probed. Protein extracts were separated by SDS-PAGE and transferred to PVDF membranes. Membranes were blocked with 5% non-fat milk in TBST (Tris-buffered saline with 0.1% Tween 20) for 1 h at room temperature, followed by overnight incubation at 4 °C with primary antibodies: IL-1α (30 kDa; Wuhan Sanying Biotechnology, China; 16765-1-AP,), IL-1β (30 kDa; Abcam, USA; [65]Ab254360), SERPINE1 (45 kDa; Wuhan Sanying Biotechnology, China; 13801-1-AP), ITGA2 (150 kDa; ABclonal, China; A19068), and AXL (140 kDa; Wuhan Sanying Biotechnology, China; 13196-1-AP). After washing with TBST, membranes were incubated with appropriate HRP-conjugated secondary antibodies for 1 h at room temperature. Protein bands were visualized using an enhanced chemiluminescence detection system. Band intensities were quantified using ImageJ software and normalized to GAPDH as an internal control. Statistical analyses Statistical analyses were performed using GraphPad Prism 8.0 software. Data were presented as mean ± standard deviation (SD). The one-way analysis of variance (ANOVA) was used to compare the significant difference of multiple experimental groups. P < 0.05 was classified as statistically significant. Results Integration of multiple datasets to identify DDP resistance-associated targets in OC OC is a highly heterogeneous disease, and studies on a single cell line limit our understanding of its molecular pathology. We integrated three independent DDP-resistant OC datasets ([66]GSE15372, [67]GSE45553, and [68]GSE58470) to identify more universally applicable DDP resistance-associated targets. Pearson correlation and PCA analysis of samples from these datasets revealed significant differences in expression profiles of DDP-resistant cells (Fig. [69]1A-B). Subsequent differential expression analysis showed that, compared to parental cells, a large number of genes were significantly altered in DDP-resistant cells ([70]GSE15372: 309 significantly upregulated, 828 significantly downregulated; [71]GSE45553: 1360 significantly upregulated, 1365 significantly downregulated; [72]GSE58470: 734 significantly upregulated, 848 downregulated considerably) (Fig. [73]1C). To identify robust and representative DDP resistance-associated targets, we used the RRA algorithm to integrate DEG rankings from the three datasets. Based on aggregate ranking scores, we identified 337 candidate targets, including 246 upregulated and 91 downregulated genes (Table S1). Fig. 1. [74]Fig. 1 [75]Open in a new tab DDP-resistant cells in OC exhibit significant differential gene expression. A Pearson correlation analysis of gene expression profiles between DDP-resistant and parental OC cells across three datasets ([76]GSE15372, [77]GSE45553, and [78]GSE58470). B PCA shows distinct clustering of DDP-resistant and parental OC cells in three datasets ([79]GSE15372, [80]GSE45553, and [81]GSE58470). C Volcano plots of differentially expressed genes in DDP-resistant OC cells from three datasets ([82]GSE15372, [83]GSE45553, and [84]GSE58470). D The top 20 upregulated and downregulated genes in DDP-resistant OC cells were identified across the three datasets ([85]GSE15372, [86]GSE45553, and [87]GSE58470) Figure [88]1D shows the top 20 up- and down-regulated genes in the three datasets, along with their fold changes. Many of these genes showed consistent expression trends (either up- or down-regulated) across all datasets. For example, FMA107A, ANKRD13A, and CALD1 were all up-regulated, while TEDDM1, LRFN5, and SLC27A6 were consistently down-regulated. Some of these genes showed variation in one dataset but remained strongly regulated in the others, reflecting inter-platform variability while maintaining overall statistical robustness—an expected property based on RRA integration. Identification of AS regulating DDP resistance-related targets of OC via network pharmacology analysis Using the network pharmacology analysis methods, we screened for active compounds based on OB (≥ 30%) and DL (≥ 0.18). As a result, we identified 20 active compounds in AS (Table [89]2). Database searches revealed 386 potential targets for AS (Fig. [90]2A). Subsequently, we used Cytoscape to construct an AS-active compound-potential targets-DDP resistance interaction network. As shown in Fig. [91]2A and 22 potential targets were shared between AS and DDP resistance, suggesting these targets as potential therapeutic targets to enhance OC sensitivity to DDP by AS. The relationships among these candidate targets are depicted in the PPI network in Fig. [92]2B, where IL6, IL1B, CCL2, PTGS2, IL1A, ICAM1, PLAU, SERPINE1, MMP1, and MMP3 have more edges, indicating their potential critical roles in AS regulation of DDP resistance. Table 2. Active compounds of AS Mol ID Molecule Name OB (%) DL MOL000211 Mairin 55.38 0.78 MOL000239 Jaranol 50.83 0.29 MOL000296 hederagenin 36.91 0.75 MOL000033 (3 S,8 S,9 S,10R,13R,14 S,17R)−10,13-dimethyl-17-[(2R,5 S)−5-propan-2-yloctan-2-yl]−2,3,4,7,8,9,11,12,14,15,16,17-dodecahydro-1 H-cyclopenta[a]phenanthren-3-ol 36.23 0.78 MOL000354 isorhamnetin 49.6 0.31 MOL000371 3,9-di-O-methylnissolin 53.74 0.48 MOL000374 5’-hydroxyiso-muronulatol-2’,5’-di-O-glucoside 41.72 0.69 MOL000378 7-O-methylisomucronulatol 74.69 0.3 MOL000379 9,10-dimethoxypterocarpan-3-O-β-D-glucoside 36.74 0.92 MOL000380 (6aR,11aR)−9,10-dimethoxy-6a,11a-dihydro-6 H-benzofurano[3,2-c]chromen-3-ol 64.26 0.42 MOL000387 Bifendate 31.1 0.67 MOL000392 formononetin 69.67 0.21 MOL000398 isoflavanone 109.99 0.3 MOL000417 Calycosin 47.75 0.24 MOL000422 kaempferol 41.88 0.24 MOL000433 FA 68.96 0.71 MOL000438 (3R)−3-(2-hydroxy-3,4-dimethoxyphenyl)chroman-7-ol 67.67 0.26 MOL000439 isomucronulatol-7,2’-di-O-glucosiole 49.28 0.62 MOL000442 1,7-Dihydroxy-3,9-dimethoxy pterocarpene 39.05 0.48 MOL000098 quercetin 46.43 0.28 [93]Open in a new tab Fig. 2. [94]Fig. 2 [95]Open in a new tab Network pharmacology analysis of active compounds and their influence on DDP resistance in OC. A Interaction network of AS’ active compounds and their potential targets related to DDP resistance. B PPI network of potential therapeutic targets. C-D GO (C) and KEGG (D) pathway enrichment analysis of potential therapeutic targets. E Visualization of the potential relationships between AS’ active compounds, potential therapeutic targets, and the DDP resistance-related signaling pathways. Each element in the network diagram is represented as a node. Different node attributes are distinguished by shapes: active compounds are shown as yellow diamonds, targets as blue circles, AS as red squares, DDP resistance as blue squares, and pathways as octagons. Nodes are connected by edges, indicating interactions between them To further understand the functions involved with these targets, we performed GO enrichment analysis, including biological processes, cellular components, and molecular functions (Fig. [96]2C and Table S2). In biological processes, the candidate targets were mainly associated with “response to external stimulus,” “positive regulation of response to stimulus,” “response to organic substance,” and “cell migration.” In cellular components, the targets were primarily involved in the “extracellular region part,” “extracellular region,” “extracellular space,” and “cell surface.” In molecular functions, most targets were enriched in “receptor binding,” “serine-type endopeptidase activity,” “serine-type peptidase activity,” and “serine hydrolase activity.” Additionally, KEGG enrichment analysis indicated that they were primarily involved in pathways regulating inflammation and cell adhesion functions, including the IL-17 signaling pathway (7 candidate targets, adjust.P = 2.03E-12), TNF signaling pathway (7 candidate targets, adjust.P = 4.71E-12), NF-kappa B signaling pathway (5 candidate targets, adjust.P = 2.56E-08), Focal adhesion (3 candidate targets, adjust.P = 3.72E-04), and Cell adhesion molecules (CAMs) (2 candidate targets, adjust.P = 4.65E-03) (Fig. [97]2D and Table S3). Figure [98]2E and Table S3 showed the potential relationships through which active compounds from AS influence DDP resistance via candidate target-mediated signaling pathways. Potential therapeutic targets are also associated with survival and immune infiltration in OC To identify potential therapeutic targets significantly associated with the prognosis of OC patients, we performed survival analysis to evaluate the relationship between the expression levels of these genes and patients’ PFS and OS. The results indicate that nine genes—IL1B, IL1A, PLAU, SERPINE1, CAV1, ITGA2, NT5E, AXL, and PTGES—have significant associations with the prognosis of OC patients (Table [99]3; Fig. [100]3A-I). Specifically, high expression of IL1B is associated with poorer PFS but better OS. High expressions of PLAU, SERPINE1, CAV1, NT5E, and AXL are associated with poorer PFS and OS, whereas high expressions of IL1A, ITGA2, and PTGES are associated with better PFS and OS. Additionally, Fig. [101]3J illustrates the correlation between these genes and immune infiltration in OC. For instance, the expression levels of PLAU, NT5E, and AXL (high expression indicating poorer PFS and OS) are significantly positively correlated with the infiltration levels of immune cells (CD8^+ and CD4^+ T cells, macrophages, neutrophils, and dendritic cells) in OC. This finding suggests that the expression levels of these genes might influence the infiltration of immune cells in the tumor microenvironment (TME), thereby affecting patient prognosis. Table 3. Survival analysis of potential therapeutic targets in OC patients PFS OS P HR P HR MMP1 0.37 1.07 (0.93 − 1.23) 0.04 1.15 (1.01 − 1.31) CAV1 0.0095 1.18 (1.04 − 1.35) 0.00097 1.24 (1.09 − 1.41) ITGA2 0.0062 0.82 (0.71 − 0.95) 0.034 0.87 (0.76 − 0.99) MMP3 0.1 1.11 (0.98 − 1.26) 0.14 0.91 (0.79 − 1.03) SERPINE1 0.0000011 1.38 (1.21 − 1.57) 0.00056 1.27 (1.11 − 1.46) AKR1C3 0.0027 0.81 (0.71 − 0.93) 0.11 1.13 (0.97 − 1.31) PLAT 0.051 1.15 (1 − 1.31) 0.073 1.13 (0.99 − 1.29) PTGS2 0.068 1.15 (0.99 − 1.33) 0.083 0.88 (0.77 − 1.02) F3 0.036 1.16 (1.01 − 1.34) 0.16 0.91 (0.8 − 1.04) CXCL2 0.17 0.92 (0.81 − 1.04) 0.069 0.89 (0.78 − 1.01) ICAM1 0.058 1.15 (1 − 1.32) 0.25 1.09 (0.94 − 1.26) IL1A 0.000096 0.78 (0.68 − 0.88) 0.00023 0.76 (0.66 − 0.88) NT5E 0.013 1.17 (1.03 − 1.33) 0.015 1.18 (1.03 − 1.34) IL6 0.11 0.89 (0.78 − 1.03) 0.11 1.11 (0.98 − 1.27) AXL 0.0000000000089 1.6 (1.4 − 1.84) 0.000013 1.37 (1.19 − 1.58) PLAU 0.0016 1.22 (1.08 − 1.39) 0.000041 1.34 (1.17 − 1.55) IL1B 0.02 1.18 (1.03 − 1.37) 0.015 0.85 (0.75 − 0.97) CCL2 0.0013 1.24 (1.09 − 1.41) 0.064 0.87 (0.74 − 1.01) FLNA 0.00028 1.26 (1.11 − 1.43) 0.12 0.89 (0.76 − 1.03) PTGES 0.023 0.86 (0.75 − 0.98) 0.013 0.85 (0.74 − 0.97) AOX1 0.12 1.12 (0.97 − 1.29) 0.29 1.07 (0.94 − 1.22) CHRM3 0.18 1.14 (0.97 − 1.37) 0.024 1.12 (1.03 − 1.55) [102]Open in a new tab Fig. 3. [103]Fig. 3 [104]Open in a new tab Impact of potential therapeutic targets on the prognosis of OC patients. A-I Kaplan-Meier survival curves for the nine candidate targets (IL1B, IL1A, PLAU, SERPINE1, CAV1, ITGA2, NT5E, AXL, PTGES) significantly associated with the prognosis of OC patients, showing PFS (left) and OS (right). J Correlation analysis between these targets and immune infiltration in OC, with color indicating the strength of correlation: blue for negative correlation, red for positive correlation Molecular docking simulation between active compounds of AS and potential therapeutic targets significantly affecting the prognosis of OC patients To further validate the potential impact of AS active compounds on DDP resistance in OC. We performed molecular docking simulations between AS’ active compounds and potential therapeutic targets significantly associated with patient prognosis. As shown in Fig. [105]4, these active compounds exhibited strong affinity for the targets, with ΔG values all below − 5 kJ/mol, indicating stable binding states of the molecular docking complexes. Those results suggest that these active compounds may influence DDP resistance by directly interacting with the targets. Specifically, the main active compounds in AS, such as quercetin, FA, and kaempferol, formed intermolecular hydrogen bonds with amino acid residues of IL1B, IL1A, PLAU, SERPINE1, CAV1, ITGA2, NT5E, AXL, and PTGES (Fig. [106]4A-J). Fig. 4. [107]Fig. 4 [108]Open in a new tab Molecular docking of AS’ active compounds with potential DDP resistance targets. A-J show the binding modes and affinities of AS’ active compounds (quercetin, FA, and kaempferol) with potential DDP resistance targets (IL1B, IL1A, PLAU, SERPINE1, CAV1, ITGA2, NT5E, AXL, and PTGES). The left panel illustrates the overall view of the binding, while the right panel provides detailed binding interactions. The receptor and ligand used in molecular docking and the binding energy (ΔG; kJ/mol) are indicated in the top left corner. Active compounds (green) and amino acid residues of the potential targets (blue) are represented as stick structures, with hydrogen bonds depicted as yellow dashed lines. Bond energies are annotated in white numbers beside the interactions AS combined with DDP significantly inhibits proliferation and migration but increases apoptosis in DDP-resistant cells The bioinformatics and molecular docking analyses above suggest that AS may influence DDP resistance by modulating DDP resistance-related targets. To validate this hypothesis, we examined the effects of AS and DDP—alone or in combination—on the proliferation, migration, and apoptosis of two DDP-resistant OC cell lines. First, we determined the IC₃₀ concentrations of AS and DDP using CCK-8 assays across a range of concentrations. The calculated IC₃₀ values of AS were 2.7 mg/mL for SKOV3-DDP and 2.083 mg/mL for A2780-DDP cells, while the IC₃₀ values of DDP were 54.96 µM and 19.245 µM, respectively (Fig. [109]5A–B). These concentrations were subsequently used in all downstream functional assays. Fig. 5. [110]Fig. 5 [111]Open in a new tab Effects of AS and DDP, alone or in combination, on proliferation, migration, and apoptosis of DDP-resistant cell lines. A-B Cell proliferation curves of DDP-resistant cells treated with varying concentrations of astragalus aqueous extract (A) or DDP (B) for 24 h. C CCK-8 assay results show the proliferation rates of A2780 and SKOV3 cell lines under different treatment conditions. D Scratch assay results demonstrate the migration ability of the two cell lines after treatment. E Flow cytometry analysis indicates the apoptosis rates in the cell lines following treatment As shown in Fig. [112]5C, both AS and DDP monotherapy significantly inhibited cell proliferation compared to the untreated control. Similarly, wound healing assays demonstrated that either AS or DDP alone markedly suppressed the migratory ability of SKOV3-DDP and A2780-DDP cells (Fig. [113]5D). Flow cytometry analysis further revealed a significant increase in apoptosis upon treatment with either AS or DDP alone (Fig. [114]5E). Notably, the combination of AS and DDP produced the most pronounced effects—exhibiting the greatest inhibition of proliferation and migration, along with the highest apoptosis induction (Fig. [115]5C–E). These findings suggest that AS synergizes with DDP to overcome chemoresistance by enhancing cytotoxicity and promoting apoptosis, offering a promising therapeutic strategy for treating DDP-resistant OC. AS combined with DDP significantly alters protein levels of potential therapeutic targets To further validate the molecular docking results, we conducted a Western blot analysis on selected potential therapeutic targets. Although nine targets (IL1B, IL1A, PLAU, SERPINE1, CAV1, ITGA2, NT5E, AXL, and PTGES) were identified as significantly associated with OC prognosis, NT5E and PTGES were excluded because they were not enriched in the pathway (Fig. [116]2E). Based on the combined ranking of differential genes in the three datasets, five genes, IL1B, IL1A, SERPINE1, ITGA2, and AXL were ranked high and enriched in inflammation, proliferation, and apoptosis-associated pathways, and we focused on these five key targets for further validation (Fig. [117]6A and Table S1). The results showed that the protein levels of SERPINE1, IL1A, and IL1B were significantly upregulated after treatment with AS or DDP alone, and this upregulation was more pronounced with combined treatment. Conversely, ITGA2 and AXL protein levels were downregulated under the same conditions, with the greatest downregulation observed in the combined treatment group (Fig. [118]6A-F). These findings support the hypothesis that AS and DDP can modulate the expression of key proteins involved in DDP resistance, with combined treatment demonstrating a more pronounced effect. Fig. 6. [119]Fig. 6 [120]Open in a new tab Western blot analysis of key proteins in DDP-resistant cell lines under different treatments. Western blot analysis of SKOV3-DDP and A2780-DDP cell lines treated with AS, DDP, or a combination of both (A), showing the expression levels of (B) ITGA2, (C) SERPINE1, (D) IL1A, (E) AXL, and (F) IL1B. Protein expression levels were quantified using GAPDH as a loading control and normalized to the untreated group Discussion OC is the deadliest gynecologic cancer, and DDP is its first-line chemotherapy drug, exerting its effects through inducing DNA damage and promoting apoptosis [[121]26, [122]27]. However, OC cells often develop resistance to DDP via various mechanisms, leading to chemotherapy failure, including enhanced DNA repair capacity, dysregulation of apoptotic pathways, upregulation of drug efflux proteins, and alterations in cellular signaling pathways [[123]28]. Therefore, understanding and mitigating DDP resistance is crucial for improving the prognosis of OC patients. In this study, based on genomic sequencing data from multiple DDP-resistant OC cell Lines, we identified 337 genes associated with DDP resistance. These genes exhibit broader applicability compared to studies focusing on single-cell lines, potentially providing insights for early clinical screening and therapeutic research. Traditional Chinese herbs, with a long history and high safety profile, have been shown to enhance the sensitivity of resistant cells to DDP [[124]29–[125]32]. However, the precise mechanisms of action of AS and its synergistic effects with DDP in OC remain unclear. Utilizing network pharmacology, molecular docking, and experimental validation, we revealed potential molecular mechanisms. We identified 20 active compounds of AS and constructed interaction networks to identify 22 potential targets associated with DDP resistance in OC. These targets primarily involve inflammation (IL-17 signaling pathway, TNF signaling pathway, and NF-κB signaling pathway), cell adhesion (cell adhesion and cell adhesion molecules), and signaling pathways (apoptosis, HIF-1 signaling pathway, PI3K-Akt signaling pathway, etc.). Survival prognosis analysis of these potential targets revealed significant associations with the survival of OC patients. For example, IL1B and IL1A are critical initiators of inflammatory responses in the TME, which may foster an immunosuppressive milieu when chronically activated. Elevated IL1B has been shown to activate the NF-κB/STAT3 axis, promoting tumor cell survival, angiogenesis, and immune evasion, ultimately correlating with worse clinical outcomes [[126]33, [127]34]. Similarly, SERPINE1 (PAI-1) has pro-angiogenic and anti-apoptotic properties. In OC, hypermethylation-induced overexpression of SERPINE1 contributes to EMT, which is a hallmark of therapy resistance and tumor aggressiveness [[128]35]. Therefore, its high expression likely reflects both enhanced invasiveness and resistance to DDP-induced cell death, explaining its adverse prognostic impact. ITGA2, a cell adhesion molecule, facilitates tumor–extracellular matrix (ECM) interactions that promote cell survival under therapeutic stress via the CAM-DR (cell adhesion-mediated drug resistance) mechanism. Its upregulation sustains integrin-FAK-ERK signaling, leading to apoptosis resistance and persistent proliferation [[129]36]. AXL, a receptor tyrosine kinase, has been implicated in PI3K-Akt activation and epithelial-mesenchymal transition (EMT). High AXL expression predicts resistance to multiple agents, including platinum compounds, and is associated with immune checkpoint inhibitor failure. Mechanistically, AXL promotes cell survival, invasiveness, and immune escape [[130]37]. These mechanisms likely underpin the negative prognostic impact of AXL in OC. Collectively, these results suggest that high expression of these targets is not only a biomarker of chemoresistance, but also reflects a shift toward an aggressive, immune-evasive tumor phenotype, which contributes to poor patient prognosis. Immune infiltration analysis results also indicate that these targets may modulate the TME to influence chemotherapy response. Recent studies have highlighted the importance of immune cell infiltration—particularly tumor-associated macrophages (TAMs), T cells, and myeloid-derived suppressor cells (MDSCs)—in determining chemotherapeutic efficacy and tumor progression in OC. In our analysis, several AS-associated targets (such as IL1A, IL1B, and AXL) were closely linked to immune infiltration patterns, suggesting a potential immunomodulatory role of AS. Specifically, IL1B is a key cytokine in inflammasome signaling and can influence macrophage polarization [[131]38]. Upregulation of IL1B, as seen in our combination treatment group, may stimulate pro-inflammatory responses that attract cytotoxic lymphocytes into the TME, potentially enhancing immunogenicity and response to DDP. Moreover, AXL expression has been associated with the recruitment of immunosuppressive cell subsets such as M2 macrophages and regulatory T cells (Tregs), and its downregulation through AS treatment may help “reprogram” the immune contexture toward a more pro-inflammatory, anti-tumor phenotype [[132]39]. This modulation of immune cell dynamics may contribute to breaking the immune escape often observed in platinum-resistant tumors. Compounds such as quercetin and naringin—both identified as active components in AS—have been reported to reduce immunosuppressive cytokine secretion, inhibit M2 macrophage polarization, and enhance CD8⁺ T cell responses in other tumor models [[133]40, [134]41]. Their presence in AS may therefore facilitate a more immunologically active TME, thereby synergizing with DDP to promote tumor cell clearance. Molecular docking results demonstrate that active components of AS (quercetin, ferulic acid, and naringin) exhibit binding solid affinity with these key targets, further supporting the potential of AS in enhancing DDP efficacy. In vitro experiments confirmed that AS combined with DDP significantly promotes apoptosis and reduces the proliferation and migration of DDP-resistant cell lines. Significant changes in protein expression of IL1B, IL1A, SERPINE1, ITGA2, and AXL (key targets involved in inflammation, apoptosis, and proliferation signaling pathways) were observed in cells treated with the combination of both drugs compared to either drug alone. These findings suggest that AS synergistically enhances DDP efficacy by modulating the expression of DDP resistance-related genes. This study’s main contribution lies in using network pharmacology to screen and validate the potential of AS in modulating DDP resistance, offering insights into potential molecular mechanisms. However, the study has limitations, such as the need for more animal experiments to validate our findings. Future studies should focus on preclinical validation of the proposed mechanisms. Specifically, in vivo animal models, such as patient-derived xenografts (PDX) or genetically engineered mouse models (GEMMs), are essential to evaluate the therapeutic efficacy, immune-modulatory effects, and toxicity profile of AS in combination with DDP. Such models would allow assessment of pharmacokinetics and pharmacodynamics in a TME that better mimics clinical reality [[135]42]. In addition, multi-omics approaches such as transcriptomics, single-cell RNA sequencing, and proteomics could help decipher the dynamic responses of immune and tumor cells to AS-DDP combination therapy. These technologies may identify new resistance escape pathways or synergistic biomarkers [[136]43]. For clinical translation, it is also crucial to explore AS and its active components in well-designed early-phase clinical trials or prospective cohort studies, focusing on DDP-resistant OC patients. Integrating network pharmacology predictions with clinical pharmacology and immunoprofiling would enhance the translatability of herbal medicine-based strategies in modern oncology. In summary, this study demonstrates that AS enhances DDP’s cytotoxicity against OC by modulating several key targets associated with DDP resistance. Future in vivo and translational studies will be critical to validate these findings and facilitate the clinical development of AS as a novel adjuvant to platinum-based chemotherapy. These findings provide new insights into the application of traditional Chinese herbs in cancer treatment and offer potential directions for future research. Supplementary Information [137]Supplementary Material 1.^ (84KB, xlsx) [138]Supplementary Material 2.^ (5.4MB, pdf) Acknowledgements