The human GC cell lines AGS and HGC-27 were procured from the Cell Resource Center of the Chinese Academy of Medical Sciences in collaboration with Peking Union Medical College. The cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM; Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Wisent, Inc.). The cells were maintained at 37˚C in a humidified incubator with 5% CO₂ (Thermo Fisher Scientific, Inc.).

Hesperadin (cat. no. HY-12054; MedChemExpress) and cisplatin (cat. no. 232120; MilliporeSigma) were dissolved in dimethyl sulfoxide (DMSO) and normal saline, respectively, and stored at -20˚C in light-protected conditions. The NOX1 inhibitor ML171 (2-acetylphenothiazine; cat. no. HY-12805; MedChemExpress) was also dissolved in DMSO before use. The Cell Counting Kit-8 (CCK-8; cat. no. C0038), Annexin V-FITC apoptosis detection kit (cat. no. C1062M), ROS detection kit (cat. no. S0033S), JC-1 mitochondrial membrane potential (MMP; cat. no. C2006) assay kit and DNA damage detection kit (cat. no. C2035S) were obtained from Beyotime Institute of Biotechnology.

To evaluate cytotoxicity, AGS and HGC-27 GC cells were seeded in 96-well plates at a density of 6,000 cells per well and exposed to Hesperadin (0-90 nM) or cisplatin (0-120 µM) for 24 h. After 48 h of treatment, CCK-8 reagent (10 µl per well) was added and incubated at 37˚C for 2 h. The absorbance at 450 nm was then measured using a SpectraMax iD5 microplate reader (Molecular Devices, LLC) to determine the half-maximal inhibitory concentration (IC₅₀) values using statistical software.

For the proliferation inhibition assay, cells were seeded at the same density and treated with Hesperadin alone or in combination with cisplatin for 48 h. The procedures for CCK-8 reagent incubation and absorbance measurement at 450 nm were performed as described for the cytotoxicity assay.

RNA extraction, library preparation, and sequencing were performed by Shanghai Applied Protein Technology Co., Ltd. Briefly, total RNA was extracted from AGS cells using TRIzol^® Reagent (cat. no. R0016; Beyotime Institute of Biotechnology). RNA integrity was verified (RIN >8.0). Sequencing libraries were constructed using the VAHTS Universal V6 RNA-seq Library Prep Kit for Illumina (Vazyme Biotech Co., Ltd.) and sequenced on an Illumina NovaSeq 6000 platform to generate 150 bp paired-end reads. Subsequent bioinformatic analysis was conducted as follows: Raw reads were quality-filtered using Fastp. De novo assembly was performed with Trinity. Transcripts were annotated against NR (https://www.ncbi.nlm.nih.gov/refseq/about/nonredundantproteins/), SwissProt (https://www.uniprot.org/uniprotkb?facets=reviewed%3Atrue), Pfam (http://pfam.xfam.org/), Gene Ontology (GO; http://geneontology.org/) and Kyoto Encyclopedia of Genes and Genomes (KEGG; https://www.genome.jp/kegg/) databases. Coding sequences were predicted using TransDecoder. Gene expression (FPKM) was quantified, and differential expression analysis was performed using DESeq2 (threshold: |log2FC|>1, padj <0.05). GO and KEGG enrichment analyses were conducted using clusterProfiler (version 4.16.0; Bioconductor).

With |log2-fold change (FC)|>1 and P<0.05 as screening criteria, DEGs from bulk sequencing were identified via the ‘Limma’ package (version 3.64.3; Bioconductor) of R. DEGs were illustrated via volcano maps and heatmaps, which were plotted via the ‘ggplot2’ and ‘Pheatmap’ packages of R, respectively. In addition, KEGG and GO analyses were performed via the ‘clusterProfiler’ R package, which aims to assess gene-related biological processes (BP), molecular functions, cellular components, and gene-related signaling pathways (KEGG). The conditions for screening significantly enriched GO terms and KEGG pathways were Min overlap=three and Min enrichment=1.5. The enrichment significance threshold was set to an adjusted P-value below 0.05. In addition, gene set variation analysis (GSVA) is a non-parametric and unsupervised method for estimating changes in specific gene sets. The activities of the hallmark pathways were quantified with the GSVA R package according to the ‘Hallmarks geneset’ downloaded from the MsigDB database. Finally, gene set enrichment analysis (GSEA) was employed to identify DEG functions on the basis of P<0.05 and a false discovery rate (FDR) value of 0.25.

Apoptosis was assessed using an Annexin V-FITC/propidium iodide (PI) apoptosis detection kit (cat. no. C1062M; Beyotime Institute of Biotechnology). Briefly, after drug treatment, the cells were collected, washed twice with ice-cold PBS, and resuspended in 1X binding buffer at a density of 1x10⁶ cells/ml. Subsequently, cells were stained with Annexin V-FITC and PI for 15 min at room temperature in the dark according to the manufacturer's instructions. To establish appropriate gating and compensation, unstained cells and single-stained controls (cells stained with Annexin V-FITC only or PI only) were included in each experiment. Flow cytometric analysis was immediately performed on a Beckman CytoFLEX flow cytometer (Beckman Coulter, Inc.). The data were analyzed using FlowJo software (version 10.8.1; FlowJo LLC). The percentages of cells in early apoptosis (Annexin V⁺/PI^-) and late apoptosis/necrosis (Annexin V⁺/PI⁺) were quantified.

For ROS detection, following the respective treatments, cells were collected, washed twice with pre-warmed phosphate-buffered saline (PBS), and then incubated with 10 µM CM-H2DCFDA diluted in serum-free medium at 37˚C in the dark for 25 min. After incubation, the cells were washed to remove the excess probe and analyzed using a fluorescence microscope (MShot MF53-N, FITC channel), and the captured images were quantified using ImageJ software (version 1.53k; National Institutes of Health). The fluorescence intensity was quantified via flow cytometry (FITC channel).

The MMP was assessed via a JC-1 assay kit (cat. no. C2006; Beyotime Institute of Biotechnology). The cells were incubated with JC-1 staining solution for 20 min at 37˚C in the dark for 20 min. The fluorescence intensity ratios (red/green) were subsequently analyzed via fluorescence microscopy (MShot MF53-N). JC-1 aggregates (red fluorescence) in healthy mitochondria with high MMP and JC-1 monomers (green fluorescence) in depolarized mitochondria with low MMP were detected using the TRITC and FITC channels, respectively. For quantitative analysis, the fluorescence intensity of the red and green channels from the captured images was measured using ImageJ software (version 1.53k; National Institutes of Health).

The cells were fixed for 15 min and subsequently incubated overnight at 4˚C with the γ-H2AX primary antibody (cat. no. C2035S; Beyotime Institute of Biotechnology). After three gentle rinses with phosphate-buffered saline (PBS), the cells were incubated for 1 h with an Alexa Fluor 488-conjugated goat anti-rabbit secondary antibody. Nuclei were counterstained with DAPI (10 µg/ml) for 5 min. Fluorescence images were captured via a fluorescence microscope (MShot MF53-N). The number of γ-H2AX foci per cell was quantified via the ‘Analyze Particles’ function in ImageJ software (version 1.53k).

Total RNA was extracted from cultured AGS and HGC-27 cells using an RNA extraction kit (Omega Bio-Tek, Inc.). cDNA was synthesized from 1 µg RNA using the SweScript RT SuperMix (Wuhan Servicebio Technology Co., Ltd.). with the following thermal protocol: 25˚C for 5 min, 42˚C for 15-30 min, and 85˚C for 5 sec. qPCR was performed on a QuantStudio 5 system (Applied Biosystems; Thermo Fisher Scientific, Inc.) using Servicebio SYBR Green Master Mix. The thermocycling protocol was as follows: 95˚C for 10 min; 40 cycles of 95˚C for 15 sec and 60˚C for 1 min. The relative expression of NOX1 was normalized to GAPDH and calculated using the 2^-ΔΔCq method (17). The sequences of primers used were as follows: NOX1 forward, 5'-TCTGGTTGTTTGGTTAGGGCT-3' and reverse, 5'-CGGCTGCAAAACCCAAGGA-3'; and GAPDH forward, 5'-CGGAGTCAACGGATTTGGTCGTAT-3' and reverse, 5'-AGCCTTCTCCATGGTGGTGAAGAC-3'.

The GC cell lines were seeded into culture plates. After adhering, the cells were treated with the designated concentrations for 10 days, followed by staining with 0.1% crystal violet for 30 min. The plates were rinsed with ultrapure water, air-dried, and scanned. Colonies containing more than 50 cells were automatically quantified using ImageJ software (version 1.53k; National Institutes of Health).

Cells were lysed in RIPA buffer (cat. no. P0013B; Beyotime Institute of Biotechnology) supplemented with 1% PMSF. The protein concentration was determined using a BCA Protein Assay Kit (cat. no. P0010; Beyotime Institute of Biotechnology). In total, ~20-30 µg of total protein per lane was separated on 12.5% SDS-PAGE gels (Epizyme, Inc.) and transferred onto polyvinylidene fluoride (PVDF) membranes (MilliporeSigma). The membranes were blocked with 5% non-fat milk in TBST (0.1% Tween-20) for 2 h at room temperature and then incubated overnight at 4˚C with the following primary antibodies: anti-BCL-2 (1:2,000; cat. no. 12789-1-AP; Proteintech Group, Inc.), anti-BAX (1:20,000; cat. no. 50599-2-Ig; Proteintech Group, Inc.), anti-caspase-3/cleaved caspase-3 (1:1,000; cat. no. 19677-1-AP; Proteintech Group, Inc.), anti-NOX1 (1:1,000; cat. no. 671300; Chengdu Zhengneng Biotechnology, Co., Ltd.), anti-β-ACTIN (1:50,000; cat. no. 66009-1-Ig; Proteintech Group, Inc.) and anti-GAPDH (1:20,000; cat. no. 10494-1-AP; Proteintech Group, Inc.). After washing, the membranes were incubated with HRP-conjugated Goat Anti-Rabbit IgG (1:5,000; cat. no. SA00001-2; Proteintech Group, Inc.) or HRP-conjugated Goat Anti-Mouse IgG (1:5,000; cat. no. SA00001-1; Proteintech Group, Inc.) secondary antibodies for 1 h at room temperature. Protein bands were visualized using an Enhanced Chemiluminescence (ECL) Substrate (cat. no. WBKLS0500; MilliporeSigma) and captured digitally. The band intensities were quantified by densitometry using ImageJ software (version 1.53k; National Institutes of Health).

For siRNA transfection, cells were cultured in 6-well plates until they reached ~70-90% confluence. Specific siRNAs were then transfected via Lipofectamine 2000 reagent (cat. no. 11668-027; Thermo Fisher Scientific, Inc.). A final concentration of 20 µM of siRNA was used for each transfection. The transfection was carried out in serum-free Opti-MEM medium and incubated with the cells at 37˚C for 4 h, after which the medium was replaced with complete growth medium. The efficiency of siRNA-mediated interference was verified 48 h post-transfection via western blotting to ensure effective knockdown of the target genes. All siRNAs, including the non-targeting negative control (si-NC), were designed and synthesized by Anhui General Biotech Co., Ltd. The siRNA sequences utilized in the present study were as follows: siNOX1#1 (homo-ID27035 siRNA-318) forward, 5'-AUGAGAAGGCCGACAAAUATT-3' and reverse, 5'-AUUUGCGCAGGCUCUUUGCTT-3'; siNOX1#2 (homo-ID27035 siRNA-1315) forward, 5'-CAUAAGGGCUUUCGAACAATT-3' and reverse, 5'-UUGUUCGAAAGCCCUUAUGTT-3'; and si-NC forward, 5'-UUCUCCGAACGUGUCACGUTT-3' and reverse, 5'-ACGUGACACGUUCGGAGAATT-3'.

The data are presented as the mean ± SDs of three independent biological replicates. Statistical comparisons between two groups were performed using an unpaired Student's t test. For comparisons among more than two groups, one-way analysis of variance (ANOVA) was employed, followed by Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant difference. All statistical analyses were performed via GraphPad Prism software (version 9.5.1; Dotmatics).

To assess the antitumor activity of Hesperadin, its IC50 in GC cell lines was first determined. The IC50 values of Hesperadin in AGS and HGC-27 cells were 52.49 nM and 46.37 nM, respectively (Fig. 1A). Next, concentrations of 10 and 20 nM were selected for AGS cells and 8 and 16 nM for HGC-27 cells, all of which are below the IC50 values for subsequent experiments. Compared with the control, Hesperadin significantly inhibited the proliferation of GC cells in a concentration-dependent manner (Fig. 1B).

To further examine the growth-inhibitory effects of Hesperadin on GC cells, Annexin V-FITC/PI double staining followed by flow cytometry was performed to evaluate its effect on cell death. Flow cytometry revealed increased cell death after 48 h of treatment (Fig. 1C). Western blot analysis demonstrated that hesperadin significantly downregulated the expression of the antiapoptotic protein BCL-2 while upregulating the expression of the proapoptotic proteins BAX and cleaved caspase-3. These findings indicated that Hesperadin promotes cell death through activation of the mitochondrial apoptosis pathway (Fig. 1D). Taken together, these results demonstrated that Hesperadin suppresses GC cell proliferation and induces cell death in a dose-dependent manner via activating the mitochondrial apoptosis pathway.

Inducing the apoptosis of tumor cells is one of the important ways to inhibit cell proliferation (18). To elucidate the molecular mechanisms driving Hesperadin-induced cell death in GC, RNA-seq was conducted on AGS cells treated with 10 nM Hesperadin for 48 h. A complete list of all DEGs is provided in Table SII. The DEGs identified from the RNA-seq results are illustrated in Fig. 2A and B. GO, KEGG, GSEA and GSVA enrichment analyses revealed that the DEGs were significantly enriched in BP such as apoptosis and the oxidative stress response (Fig. 2C-F). While the top 10 upregulated DEGs were associated with key cancer pathways such as cell cycle arrest (for example, CDKN1A), p53 signaling (for example, BTG2) and cell adhesion (for example, PTPRU) (Table SI), the pronounced enrichment in oxidative stress pathways particularly captured our attention. Specifically, GO analysis identified significant enrichment in the ‘cellular response to oxidative stress’ (GO: 0045489) (Fig. 2C) and KEGG analysis confirmed enrichment in the ‘Oxidative phosphorylation’ pathway (hsa00190) (Fig. 2D). This transcriptional profile aligns with the documented mechanism of Hesperadin, which induces mitochondrial dysfunction and a ROS surge in other cancer models (16). Consequently, DEGs within these oxidative stress-related pathways were prioritized for further investigation. Among them, NADPH Oxidase 1 (NOX1) emerged as a compelling candidate. Although its fold-change ranked outside the top 10, NOX1 encodes a critical catalytic subunit of the NADPH oxidase complex, a dedicated superoxide generator and a master regulator of oxidative stress (19,20). Its significant upregulation (log₂FC >1.5, P<0.01) following Hesperadin treatment (Fig. 2A) suggested a functionally central role in mediating the observed oxidative stress phenotype. NOX1 was therefore selected for subsequent functional validation.

Excessive production of ROS induces oxidative stress and mitochondrial dysfunction, ultimately leading to cell death (18,19). To investigate the impact of Hesperadin on ROS generation and oxidative stress pathways in GC cells, intracellular ROS levels in AGS and HGC-27 cells were measured following Hesperadin treatment. The results showed a significant, concentration-dependent increase in ROS fluorescence intensity in the Hesperadin-treated cells compared with the control group (Fig. 3A). Further quantitative analysis by flow cytometry revealed that Hesperadin increased ROS levels in GC cells in a dose-dependent manner (Fig. 3B). In summary, it can be concluded that Hesperadin induces cell death in GCs by increasing the generation of ROS, which leads to oxidative stress and activates the mitochondrial apoptotic pathway.

Cisplatin is a key chemotherapeutic agent widely used in the treatment of GC (21,22). Considering the mechanistic similarities between cisplatin and Hesperadin in modulating tumor-associated pathways, the present study aimed to evaluate the potential therapeutic benefits of combining both agents in GC cell models. To investigate the potential combinatorial effect of Hesperadin and cisplatin, the IC₅₀ for each agent alone was first determined in AGS and HGC-27 cells (Figs. 1A and 4A). To distinguish true synergy from additive toxicity, sub-IC₅₀ concentrations were selected for all combination treatments. Based on dose-response profiles and preliminary validation, the chosen concentrations (20 nM Hesperadin and 10 µM cisplatin for AGS cells, and 16 nM Hesperadin and 8 µM cisplatin for HGC-27 cells) represent ~30-40% of their respective single-agent IC₅₀ values. These doses induce partial biological effects (Fig. 1B and C) and moderate proliferation inhibition without causing overwhelming cell death, thereby providing a clear window to observe enhanced efficacy upon combination. Under these conditions, the combinatorial effects were evaluated. The results revealed that when AGS and HGC-27 cells were treated with a combination of Hesperadin and cisplatin, the IC₅₀ values for cisplatin decreased from 33.66 and 26.88 µM to 24.99 and 20.47 µM, respectively (Fig. 4A). Furthermore, the GC cell survival rate in the Hesperadin-cisplatin combination treatment group was significantly lower than that in the control and single-drug groups (Fig. 4B), whereas the clonogenic ability was significantly reduced (Fig. 4C), and the mortality rate was significantly elevated (Fig. 4D). Additionally, western blot analysis revealed that the expression levels of the proapoptotic proteins BAX and cleaved caspase-3 were increased, whereas the expression of the antiapoptotic protein BCL-2 was decreased in the combination treatment group (Fig. 4E). These results indicated that Hesperadin enhances the antiproliferative effect of cisplatin, promotes apoptosis, and thereby potentiates the cytotoxic effect of cisplatin on GC cells.

Cisplatin exerts its anticancer effect mainly by binding to DNA to form cross-links and induce the production of ROS, thereby leading to apoptosis (4,5). The present study revealed that Hesperadin significantly inhibits GC cell proliferation and induces cell death by inducing ROS generation (Figs. 1C and 3). Therefore, it was next explored whether Hesperadin could increase the sensitivity of GC cells to cisplatin through oxidative stress. First, the fluorescence intensity in the Hesperadin-cisplatin combination treatment group was significantly greater than that in the single-drug treatment groups (Fig. 5A); the same results were obtained through quantitative flow cytometric analysis (Fig. 5B). Furthermore, it was found that both Hesperadin alone and cisplatin significantly reduced the MMP, whereas the combined treatment caused a further significant reduction in the MMP (Fig. 5C). Additionally, the fluorescence intensity of the DNA damage marker γ-H2AX was measured. The results revealed that γ-H2AX fluorescence intensity in the single-drug treatment groups (Hesperadin or cisplatin) was significantly higher than in the control group, whereas the combination treatment group demonstrated a further significant increase in γ-H2AX fluorescence intensity (Fig. 5D), indicating more extensive DNA damage following combination treatment. Collectively, these findings indicated that Hesperadin enhances the sensitivity of GC cells to cisplatin by promoting ROS generation, disrupting MMP, exacerbating DNA damage, and thereby facilitating cell death.

To directly investigate whether the chemo-sensitizing effect of Hesperadin toward cisplatin was mediated through the NOX1/ROS axis, as suggested by our transcriptomic data, the upregulation of NOX1 expression was first confirmed. Through RT-qPCR analysis, it was found that NOX1 expression was significantly upregulated upon Hesperadin treatment in GC (Fig. 6A). In addition, western blot analysis confirmed that the Hesperadin alone, cisplatin alone, and combination treatments significantly increased the protein expression level of NOX1 in AGS cells. In HGC-27 cells, the increase in NOX1 protein following cisplatin treatment did not reach statistical significance, highlighting a cell-type-dependent difference in response. However, the combination treatment with Hesperadin robustly increased NOX1 levels in both cell lines (Fig. S1C), which was consistent with the observed transcriptional changes. To validate the functional role of NOX1, siRNA was subsequently used to knock down NOX1 expression in AGS and HGC-27 cells. Compared with si-NC transfection, NOX1-specific siRNA transfection significantly reduced NOX1 protein levels (Fig. S1A). ML171, a highly selective NOX1 inhibitor, was subsequently employed to evaluate the role of NOX1 in Hesperadin-induced oxidative stress and its impact on cisplatin sensitivity in GC cells. The results revealed that both ML171 treatment and NOX1 knockdown markedly attenuated the generation of ROS induced by the combination of cisplatin and Hesperadin (Figs. 6B and S1B). Furthermore, NOX1 inhibition rescued the cell death caused by the combined treatment (Fig. 6C). MMP analysis revealed that NOX1 inhibition partially reversed the MMP reduction induced by the combination treatment (Fig. 6D). Additionally, γ-H2AX immunofluorescence analysis revealed that NOX1 inhibition significantly attenuated DNA damage in the combination treatment group (Fig. 6E). To provide molecular substantiation for the flow cytometry results in AGS cells, western blot analysis for Cleaved Caspase-3 was performed using the same treatments. Consistent with the functional data, the combination treatment significantly increased the level of Cleaved Caspase-3, which was significantly reversed by co-treatment with the NOX1 inhibitor ML171 (Fig. 6F). Collectively, these findings indicated that Hesperadin promotes ROS generation through NOX1-mediated oxidative stress in GC cells.