FerroOrange probes were purchased from Dojindo (Kumamoto, Japan). Reactive oxygen species (ROS) and BODIPY493/503 assay kits were purchased from Beyotime (Shanghai, China). Oil Red O solutions were purchased from Solarbio (Beijing, China). Malondialdehyde (MDA), glutathione (GSH), triglyceride (TG) and cholesterol (TC) assay kits were purchased from Nanjing Jiancheng (Nanjing, China). Glutathione peroxidase 4 (GPX4) monoclonal antibody, SLC7A11/xCT polyclonal antibody, SCD polyclonal antibody, fatty acid synthase (FASN) polyclonal antibody, acetyl coenzyme A carboxylase 1 (ACC1) polyclonal antibody, β-catenin polyclonal antibody, and β-actin monoclonal antibody were purchased from Proteintech Group, Inc. (Wuhan, China). Anti-rabbit and anti-mouse antibodies were sourced from CST (Boston, United States). A939572 and ferrostatin-1 (fer-1) were purchased from MedChemExpress (shanghai, China).

The herbal granules of JPYZXZ were purchased from Tianjiang pharmaceutical Co., Ltd., which consist of 60 g Astragalus membranaceus (Fisch.) Bunge (Huang Qi) (catalog number: 21062531), 30 g Codonopsis pilosula (Franch.) Nannf. (Dang Shen) (catalog number: 21080721), 10 g Atractylodes macrocephala Koidz. (Bai Zhu) (catalog number: 21062681), 10 g Angelica sinensis (Oliv.) Diels (Dang Gui) (catalog number: 21082261), 10 g Paeonia lactiflora Pall. (Bai Shao) (catalog number: 21061851), 30 g Sparganium stoloniferum (Graebn.) (San Leng) (catalog number: 21041051), 30 g Curcuma zedoaria (Christm.) Roscoe (E Zhu) (catalog number: 21071571), 10 g Vladimiria souliei (Franch.) (Mu Xiang) (catalog number: 21072701), 10 g Citrus reticulata Blanco (Chen Pi) (catalog number: 21061701), 15 g Scleromitrion diffusum (Wild.) (Bai Hua She She Cao) (catalog number: 21074161), 15 g Salvia chinensis Benth. (Shi Jian Chuan) (catalog number: 21060621) and 3 g Glycyrrhiza uralensis Fisch. (Gan Cao) (catalog number: 21073621). All granules were dissolved in double-distilled water, boiled to concentrate to 1 g/mL, and then filtered through a 0.22 μm filter. The extracts were stored at −20°C until use.

The active compounds of JPYZXZ were derived from the TCMSP database (https://old.tcmsp-e.com). OB ≥ 30% and DL ≥ 0.05 were applied to screen the active compounds of JPYZXZ in this study (Song et al., 2022).

The PubChem database (https://pubchem.ncbi.nlm.nih.gov) and the Swiss Target Prediction website (www.swisstargetprediction.ch) were used to forecast all potential targets of the active compounds in JPYZXZ. The relevant targets in GC were obtained using the DisGeNET, GeneCards, and OMIM databases. We searched for targets related to ferroptosis in the FerrDB database (www.zhounan.org/ferrdb/). All potential targets and their corresponding active compounds of JPYZXZ were imported into Cytoscape in order to construct the “JPYZXZ active compound-target” network.

To elucidate the molecular mechanisms underlying JPYZXZ-induced ferroptosis in GC, we utilized the Metascape online tool (https://metascape.org/gp/index.html) to perform GO and KEGG enrichment analyses on the identified targets. The analysis generated results across three categories: Biological Process (BP), Cellular Component (CC), and Molecular Function (MF). The enrichment analysis was executed with the following settings: minimum overlap of 3, p-value cutoff of 0.01, and minimum enrichment of 1.5. The top 20 results were graphically represented in bar charts, providing a visual representation of the enriched terms.

The primary active compounds of JPYZXZ were identified and their mol2 format was obtained from the TCMSP database. Additionally, the X-ray crystal structure files for the SCD1 (4ZYO) protein were acquired from the RCSB PDB (http://www.rcsb.org/). PyMOL 3.0.0 was used for protein preparation, removing water molecules and heteroatoms, and saving the file in. pdb format. AutoDock 4.2.6 and PyMOL 2.5.2 software were utilized for conducting docking studies between compounds and SCD1 proteins. The grid box function of AutoDock Tools was utilized to define a specific pocket for the active ingredients of compounds binding to the proteins. Following this, molecular docking analyses were carried out and the findings were visualized using PyMOL.

The MKN-45 cell line was procured from the Japanese Collection of Research Bioresources Cell Bank, whereas the HGC-27 cell line and human gastric epithelial cell GES-1 were sourced from the Chinese Academy of Sciences in Shanghai, China. The cells were propagated in a nutrient-rich RPMI-1640 medium (Gibco, United States) augmented with 10% fetal bovine serum (EVERY GREEN, China) and 1% penicillin-streptomycin solution (Gibco, United States). The cells were then maintained in a controlled environment at 37°C with a 5% CO₂ atmosphere, fostering optimal growth conditions.

Briefly, 5 × 10³ cells were plated in each well of a 96-well plate and incubated at 37°C with 5% CO₂ for 24 h. The cells were then exposed to varying concentrations of JPYZXZ (0, 1, 2, 4, 8, 16 and 32 mg/mL) for 24 h or 48 h. Afterward, 10 μL of CCK-8 working solution (Vazyme, China) was added to a 96-well plate and allowed to react for 2 h. The absorbance of the samples at OD 450 nm was subsequently quantified using the Bio Tek 800 TS (Agilent, United States). The formula for calculating cell viability is expressed as follows (experimental group - blank group)/(control group - blank group) × 100%.

HGC-27 and MKN-45 cells were seeded at a concentration of 500 cells per well in 6-well plates and exposed to various doses of JPYZXZ (0, 2, 4, and 8 mg/mL) in the culture medium. After a 2-week incubation period, the cells were fixed with a 4% paraformaldehyde solution (1 mL) for 15 min, followed by staining with a 0.1% crystal violet solution (500 μL) for 20 min. Subsequently, high-resolution images of each well were captured, and colonies comprising more than 50 cells were subjected to quantitative analysis.

The impact of JPYZXZ on the migratory capacity of GC cells was assessed through a wound-healing experiment. HGC-27 and MKN-45 cells were seeded at a density of 1 × 10⁶ cells/well in 6-well plates and allowed to reach 95% confluency. Subsequently, a 10 μL pipette tip was used to create uniform scratches, followed by the addition of serum-free medium containing JPYZXZ (0, 2, 4, and 8 mg/mL) to each well. The progress of wound closure was monitored at 0, 24, and 48 h intervals using a microscope (Olympus DP74, Japan) at a magnification of ×100, and the scratch area was assessed using ImageJ software.

Approximately 5 × 10⁴ cells were seeded into the upper transwell chambers, which contained 200 μL of serum-free medium supplemented with various concentrations of JPYZXZ (0, 2, 4, and 8 mg/mL). The lower compartment was filled with 500 μL of medium containing 10% fetal bovine serum, serving as a chemoattractant. After a 24-h coculture period, the chambers were fixed with 4% paraformaldehyde for 15 min and then stained with 0.1% crystal violet. An Olympus DP74 microscope (Japan) was used to capture images of the cells that had migrated into the lower chambers. Subsequently, ImageJ software was employed to quantify the total number of cells present in the images.

Intracellular levels of ROS were assessed using a ROS Assay Kit. HGC-27 and MKN-45 cells were inoculated into 12-well plates at a density of 1 × 10⁵ cells/well according to the manufacturer’s instructions. Thereafter, GC cells were cultured with JPYZXZ (0, 2, 4, and 8 mg/mL) or fer-1 (2 μM) for 24 h. Afterwards, the GC cells were rinsed with serum-free medium and then incubated in a dark environment with 10 μM DCFH-DA for 20 min. Subsequently, intracellular ROS levels were assessed using a fluorescence microscope (Olympus DP74, Japan) and ImageJ software.

FerroOrange was utilized for the quantification of intracellular Fe²⁺ concentrations in GC cells. Following the manufacturer’s instructions, HGC-27 and MKN-45 cells were seeded at a density of 1 × 10⁵ cells/well in 12-well plates. After adherence, the cells were treated with JPYZXZ (0, 2, 4 and 8 mg/mL) or fer-1 (2 μM) for 24 h, followed by three washes with serum-free medium. Subsequently, the cells were incubated with a working solution of FerroOrange at a concentration of 1 μM in a light-free environment for 30 min. Finally, the fluorescence intensity was measured using a fluorescence microscope (Olympus DP74, Japan) and ImageJ software.

MDA, GSH, TC and TG levels in GC cells were assessed using corresponding assay kits. HGC-27 and MKN-45 cells were seeded at a density of 1 × 10⁶ cells/well in 3 cm plates. Subsequently, varying concentrations of JPYZXZ (0, 2, 4, and 8 mg/mL) or fer-1 (2 μM) were administered to the cells for 24 h. Following this, the cells were harvested and lysed using ultrasonication. The resulting cell lysates were then treated with appropriate working solutions and MDA, GSH, TC and TG levels were determined using a Bio Tek 800 TS (Agilent, United States).

Oil Red O and BODIPY493/503 staining kits were utilized for the evaluation of neutral lipid content in GC Cells. HGC-27 and MKN45 cells, treated with JPYZXZ (0, 2, 4, and 8 mg/mL) or fer-1 (2 μM), were fixed and stained with the respective work solutions. Subsequently, neutral lipids were determined using a light microscope (Olympus DP74, Japan) and ImageJ software.

Wb assay was performed to assess the protein expression levels in GC cells. The pre-treated cells were lysed using RIPA buffer containing PMFS, phosphatase and protease inhibitors. Subsequently, the protein samples were quantified using a BCA Kit and denatured by heating in boiling water for 10 min. Each sample was then separated on 10% SDS-PAGE gels and transferred onto a PVDF membrane using the wet transfer method. Following blocking with a blocking buffer at room temperature for 15 min, the membranes were immunoblotted overnight at 4°C with primary antibodies targeting SCD1, ACC1, FASN, β-catenin, GPX4, and xCT. After washing three times for 10 min each with TBST, the membranes were incubated with corresponding secondary antibodies for 2 h. Subsequently, the protein signals were detected using the LI-COR 2802 imaging system (LI-COR, United States). Image Studio software was utilized to quantify the protein expression levels.

Two small interfering RNAs (siRNAs) targeting SCD1 or a negative control siRNA were designed by Morzan Biotech (Shanghai, China) (Supplementary Table S1). Transfection was performed using Lipofectamine 3000 (Invitrogen, California, United States) according to the manufacturer’s protocol. The transfected cells were cultured for an additional 24 h. Subsequently, the medium was replaced with serum containing JPYZXZ (0 or 8 mg/mL) for both the control and experimental groups. After a further incubation period of 24 h, the cells were harvested for subsequent analyses.

The animal experiments were performed in accordance with relevant guidelines and regulations, which strictly adhering to the ARRIVE guidelines. The ethical approval for the animal study was granted by the Institutional Animal Care and Use Committee of the Affiliated Hospital of Nanjing University of Traditional Chinese Medicine (Approval ID: 2022DW-44-02). Four-week-old male BALB/c nude mice, weighing between 18-20 g, were procured from Beijing Vital River Laboratory Animal Technology Co., Ltd., a reputable supplier in Beijing, China. Each mouse received an injection of 4 × 10⁶ MKN45 cells in the right axilla. Once subcutaneous tumors developed, the mice were randomly assigned to one of five groups: control group (0.2 mL saline administered by gavage daily), JPYZXZ-L group (15 g/kg administered by gavage daily), JPYZXZ-M group (30 g/kg administered by gavage daily), JPYZXZ-H group (45 g/kg administered by gavage daily), and 5-fluorouracil (5-Fu) group (25 mg/kg, 0.02 mL/g administered by intraperitoneal injection every other day). After 14 consecutive days of intervention, the mice were euthanized by cervical dislocation after pentobarbital sodium anesthesia and their tumors were collected for subsequent experiments.

The tissues were embedded in paraffin, sectioned, and then subjected to deparaffinization using xylene and ethanol. Next, the sections underwent hematoxylin staining for 3 min, followed by washing with water, dehydration with alcohol, and staining with eosin solution for 10 s. Afterward, the sections were dehydrated and made transparent using anhydrous ethanol and xylene. Finally, the stained sections were examined and imaged under a microscope for further analysis.

Paraffin-embedded tumor specimens were subjected to sectioning and deparaffinization through a series of solvents, including xylene, ethanol, and distilled water. Subsequent antigen retrieval was performed utilizing a microwave oven. To prevent endogenous peroxidase interference, the sections were treated with a 3% hydrogen peroxide solution, followed by a 30-min incubation with 3% bovine serum albumin at ambient temperature. Primary antibodies against SCD1, xCT, GPX4, FASN, ACC1, and β-catenin were added and the sections were incubated overnight at 4°C. Subsequently, HRP-labeled secondary antibody was applied for 50 min at room temperature. DAB chromogen solution was used to visualize the staining while hematoxylin re-stained the nuclei. Dehydration and sealing of the slides preceded imaging under a microscope.

GraphPad Prism 10.1.2 software was utilized for the analysis of quantitative data. All experimental data were presented as mean ± SD and the differences among groups were assessed using one-way ANOVA. A p value <0.05 was considered to indicate statistical significance.

The CCK-8 assay was used to assess the inhibitory effect of JPYZXZ on HGC-27 and MKN-45 cells. Following exposure to various concentrations of JPYZXZ, the IC50 values for HGC-27 and MKN-45 cells were determined as 4.51 and 5.38 mg/mL at 24 h, and 3.31 and 3.93 mg/mL at 48 h, respectively (Figure 1A). Additionally, JPYZXZ exhibited a significant dose- and time-dependent inhibition of GC cells viability, but no obvious toxicity to GES-1 cells (Supplementary Figure S1A). Therefore, concentrations of JPYZXZ (0, 2, 4, and 8 mg/mL) were selected for subsequent experiments. The colony formation assay was employed to further assess the impact of JPYZXZ on GC cells growth, revealing a significant decrease in the number of colonies with increasing concentration of JPYZXZ (Figure 1B). Furthermore, wound healing assay and transwell assay demonstrated that JPYZXZ significantly suppressed the migration of HGC-27 and MKN-45 cells at concentrations of 4 and 8 mg/mL but not at 2 mg/mL (Figures 1C–F). Therefore, we can conclude that JPYZXZ possesses the capability to inhibit the proliferation and migration of PC-GC cell.

Previous study has identified the components of JPYZXZ using UPLC-OrbitRAP-MS (Chen M. et al., 2023). We obtained 253 active components of JPYZXZ from the TCMSP database, including 21 Huangqi, 21 Dangshen, 13 Baizhu, 15 Baishao, 20 Danggui, 26 Muxiang, 11 Chenpi, 8 Sanleng, 23 Ezhu, 5 Baihuasheshecao, 4 Shijianchuan, and 86 Gancao. Additionally, a total of 14,251 targets related to GC, as well as 471 targets related to ferroptosis and 1,132 targets related to JPYZXZ were obtained from DisGeNET, GeneCards, OMIM, FerrDB databases and Swiss Target Prediction website. Subsequently, Venny website was utilized to identify 91 overlapped targets (Figure 2A). Following this, JPYZXZ compound-target network and PPI network were constructed (Figures 2B, C). Finally, enrichment analysis of GO/KEGG pathway was performed using Metascape database, and top 20 categories for each were presented. The GO enrichment analysis revealed three primary categories: BP, CC, and MF. Within BP, the dominant themes included cellular responses to oxidative stress, reactive oxygen species, and oxygen levels, as well as metabolic pathways involving unsaturated fatty acids, fatty acid biosynthesis, and lipid oxidation (Figure 2D). Meanwhile, CC were linked to the perinuclear region of cytoplasm, vesicle lumen, membrane raft, and endoplasmic reticulum lumen (Figure 2E). MF were characterized by enzyme activities related to oxidoreduction, fatty acid binding, iron ion interactions, and protein kinase functions (Figure 2F). KEGG pathway analysis revealed involvement in chemical carcinogenesis-reactive oxygen species pathways in cancer, Wnt signaling pathway, ferroptosis, glutathione metabolism, lipid and atherosclerosis (Figure 2G). Both JPYZXZ and Yiqi Huayu decoction are modified versions of the ancient formula “Gui Shao Liu Jun Zi”, which has been used for gastrointestinal diseases for thousands of years. Previous studies have found that Yiqi Huayu plays a crucial role in preventing recurrence and reducing cisplatin resistance in the treatment of GC, both of which are associated with the induction of ferroptosis (Huang et al., 2024; Song et al., 2022). In this study, network pharmacology analysis suggested that the mechanism of JPYZXZ in GC treatment may be associated with ferroptosis, lipid metabolism and Wnt signaling pathway.

In this study, several key experiments related to ferroptosis were conducted to investigate whether JPYZXZ induces ferroptosis in PC-GC cells. The results showed that JPYZXZ increased intracellular ROS and Fe²⁺ levels in HGC-27 and MKN-45 cells in a dose-dependent manner (Figures 3A–D). Additionally, GSH levels decreased (Figure 3E), while MDA levels increased in both cell lines when treated with 4 and 8 mg/mL of JPYZXZ (Figure 3F). Wb assays revealed that protein levels of GPX4 and xCT significantly decreased in HGC-27 and MKN-45 cells after exposure to 4 and 8 mg/mL of JPYZXZ (Figures 3G, H). However, no significant changes were observed when cells were treated with 2 mg/mL of JPYZXZ.

Then the rescue experiments were conducted. Fer-1, a widely used inhibitor of ferroptosis, was utilized as a positive control drug for ferroptosis. HGC-27 and MKN-45 cells were treated with JPYZXZ (0 mg/mL), JPYZXZ (8 mg/mL), fer-1 (2 μM), and JPYZXZ + fer-1 (8 mg/mL JPYZXZ +2 μM fer-1) for 24 h. The results showed that fer-1 was able to reverse JPYZXZ-induced elevation of ROS (Figures 4A, B), Fe²⁺ (Figures 4C, D), and MDA (Figure 4F), as well as the reduction of GSH in GC cells (Figure 4E). Wb experiments also confirmed these results, showing that the decrease of GPX4 and xCT in GC cells induced by JPYZXZ could be rescued with fer-1 (Figures 4G, H). Therefore, these observations indicated that JPYZXZ could induce ferroptosis in PC-GC cell.

To investigate the effect of JPYZXZ on lipid metabolism, we evaluated the levels of neutral lipids in GC cells. Initially, HGC-27 and MKN-45 cells were treated with varying concentrations of JPYZXZ (0, 2, 4, and 8 mg/mL) for 24 h. Oil Red O and BODIPY 493/503 assays were utilized to quantify lipid droplet content, revealing a dose-dependent decrease in lipid droplet formation in GC cells following JPYZXZ treatment (Figures 5A–C). Additionally, JPYZXZ significantly inhibited the levels of TC and TG in GC cells (Figures 5D, E). Western blot analysis revealed a significant decrease in the expression of FASN and ACC1 proteins, which are crucial enzymes involved in lipid metabolism in GC cells following treatment with JPYZXZ (Figures 5F, G). Notably, no significant changes were observed at a lower concentration of JPYZXZ (2 mg/mL).

To further reveal the potential connection between JPYZXZ-induced ferroptosis and lipid metabolism, the other assay was conducted using four groups: JPYZXZ (0 mg/mL), JPYZXZ (8 mg/mL), fer-1 (2 μM), and combination treatment with both substances. The results indicated that fer-1 was able to reverse the decrease in lipid droplet content (Figures 6A–C), as well as TC and TG levels (Figures 6D, E), while also restoring FASN and ACC1 expression caused by high-dose JPYZXZ treatment (Figures 6F, G). These findings suggest that JPYZXZ-induced ferroptosis is linked to alterations in lipid metabolism within PC-GC cell.

Among the 91 overlapping genes, SCD1 stands out as a key rate-limiting enzyme involved in de novo fatty acid synthesis, and its crucial role as a regulator of ferroptosis has been recognized (Sen et al., 2023). Subsequently, 12 compounds from JPYZXZ were selected based on their herbal degree and subjected to molecular docking studies with SCD1 (4ZYO). The results revealed that seven compounds exhibited strong binding affinity: Alpha-Amyrin (Baizhu) = −8.04, 3-Epioleanolic acid (Baihuasheshecao) = −7.95, Glyuranolide (Gancao) = −6.15, alpha-Cyperone (Muxiang) = −6.07, naringenin (Chenpi) = −5.54, Curcumenone (Ezhu) = −5.03 and Senkyunolide E (Danggui) = −5.03 (Figure 7A).

The Wnt/β-catenin signaling pathway is extensively researched and implicated in various aspects of tumorigenesis across multiple cancer types (Zhou et al., 2022). Activation of the Wnt/β-catenin pathway has been shown to inhibit ferroptosis in GC cells by reducing ROS production (Wang et al., 2022). SCD1 has been found to modulate the Wnt/β-catenin pathway in different cancers, leading to upregulation of β-catenin expression in GC cells (Wang et al., 2020). To investigate JPYZXZ-induced ferroptosis via the SCD1/Wnt/β-catenin axis in GC cells, we utilized the SCD1 inhibitor A939572, a small molecule compound known for its ability to inhibit migration, invasion, self-renewal, and chemotherapy resistance of GC cells (Gao et al., 2020). Wb experiments were performed after treating HGC-27 and MKN-45 cells with various concentrations of JPYZXZ (0, 2, 4, and 8 mg/mL) and A939572 (5 μM) for 24 h. The results showed a dose-dependent inhibition of SCD1, β-catenin, GPX4, and xCT expression following treatment with JPYZXZ and A939572 (Figures 7B, C). Then, we established the SCD1 knockdown model of GC and verified the knockdown effect by Wb (Supplementary Figures S1B, C). Then, GC cells with SCD1 knockdown were treated with JPYZXZ. The transfection of si-SCD1 decreased the protein of β-catenin, GPX4 and xCT. However, these proteins of SCD1 knockdown cells had no significant reduced after JPYZXZ treated (Figures 7D, E). These findings led to the conclusion that JPYZXZ induces ferroptosis in PC-GC cell through its interaction with the SCD1/Wnt/β-catenin axis.

To verify the potential therapeutic effect of JPYZXZ against PC-GC in vivo, a subcutaneous xenograft model was established. The tumor volume and weight were significantly decreased in the 5-fluorouracil (5-Fu) group and JPYZXZ groups compared to the control group (Figures 8A–C). In addition, IHC analysis of tumor sections revealed a dose-dependent reduction in the expression of SCD1, β-catenin, GPX4, xCT, FASN, and ACC1 (Figures 8D, E). Finally, body weight of GC xenograft and HE-stained tissue sections of the heart, lung, liver, and kidney were examined to evaluate the in vivo non-toxicity of JPYZXZ (Supplementary Figures S1D,E). Consequently, these findings indicate that JPYZXZ was non-toxic in our PC-GC xenograft model. Hence, we can infer that the effective inhibition of PC-GC growth in vivo by JPYZXZ is linked to ferroptosis, lipid metabolism and SCD1/Wnt/β-catenin axis.