Timosaponin AIII inhibits gastric cancer by causing oxidative stress and blocking autophagic flux.

Introduction
Gastric cancer (GC) is one of the most common malignancies and a leading cause of cancer-related mortality in recent decades [1]. Due to the lack of distinct symptoms in the early stages, the majority of patients are often diagnosed as late-stage, while chemotherapy remains the main treatment for advanced-stage GC patients [2]. Considering that chemical treatments frequently cause adverse effects and drug resistance, it is crucial to investigate novel therapies and uncover new therapeutic drugs.
Oxidative stress (OS) is a redox homeostasis imbalance dominated by peroxidation that is characterized by the production of substantial reactive oxygen species (ROS) [3]. The main sources of ROS come from two pathways: one is the production and release of NADPH oxidase complexes, and the other is the massive production of ROS through mitochondrial dependent pathways during cellular metabolic stress, necrosis, and apoptotic damage [4]. If cancer cells are exposed to excessive amounts of ROS via drug or molecular intervention, excessive ROS can induce apoptosis, ferroptosis, damaging mitochondrial function, and impairing cell membrane integrity, leading to fatal cytotoxic effects [5,6]. Cisplatin and resveratrol, as ROS-inducing agents, can induce the production of ROS through various signaling pathways. Cisplatin primarily generates ROS through DNA damage and mitochondrial dysfunction [7], while resveratrol exerts its ROS-inducing effects by modulating cellular signaling and mitochondrial activity [8]. Resveratrol can enhance the production of ROS, thereby sensitizing cancer cells to chemotherapy and contributing to its potential as an adjunctive therapeutic agent in cancer treatment. For example, Cisplatin and resveratrol combination therapy can cause GC cell death via inducing OS and cellular aging [9]. Through the Hsc70 pathway, Gastrokine 2 enhances oxidative stress-induced GC cell death [10]. Nuclear factor erythroid-derived 2-like 2 (Nrf2) is considered to be a crucial transcription factor involved in cancer cells' defense against OS [11]. Under physiological conditions, Kelch-like ECH-associated protein 1 (Keap1) binds to Nrf2 and promotes degradation via the ubiquitin proteasome pathway [12]. When confronted with OS, the Keap1-Nrf2 pathway is frequently activated, resulting in decreased Keap1 expression and increased Nrf2 expression, and it typically plays a protective role in cancer cells [13]. Luteolin significantly sensitizes A549 cells to anticancer drugs oxaliplatin and doxorubicin by inhibiting Nrf2 expression [14]. Family with sequence similarity 117, member B (FAM117B) promotes the growth and drug resistance of GC cells by activating the Keap1-Nrf2 pathway [15]. Therefore, inducing oxidative stress and inhibiting the Keap1-Nrf2 pathway have emerged as promising therapeutic targets.
Autophagy is a lysosome-dependent degradation and metabolic process of cellular contents [16]. The autophagy process consists of five stages: initiation, elongation and maturation of isolation membranes; autophagy‒lysosome fusion; and degradation of autophagic lysosome contents [17], this dynamic process is known as "autophagic flux".
Autophagy plays a complex and controversial role in cancer, as it can inhibit cancer cell proliferation and progression by inducing oxidative stress and DNA damage, as well as promote tumor cell survival by providing energy and essential substances under various stress conditions [18]. For example, Ubiquitin-conjugating enzyme E2C (UBE2C) regulates the Keap1-Nrf2 signaling pathway to promote the growth of GC by inhibiting autophagy [19]. Nobiletin induces autophagy-dependent cell death in GC cells through PI3K-mTOR signaling pathway [20]. However, if the fusion and degradation process between autophagosomes and lysosomes is disrupted, leading to impaired autophagic flux, autophagy cannot proceed. This phenomenon is referred to as "autophagic flux blockage " or "incomplete autophagy " [21]. Recent studies has shown that blocking autophagic flux in cancer cells can result in metabolic imbalance and an abnormal accumulation of autophagosomes, ultimately leading to cell death without promoting growth [18]. Polyphyllin D triggers lysosomal damage by inhibiting sphingomyelinase, leading to blockage of autophagic flux in hepatocellular carcinoma and enhancing its sensitivity to sorafenib [22]. Silencing polypyrimidine tract-binding protein 1 (PTBP1) resulted in the abnormal accumulation of autophagosomes, which inhibited the survival of GC cells. Mechanistically, interference with PTBP1 enhanced the stability of thioredoxin-interacting protein (TXNIP), leading to increased TXNIP-mediated oxidative stress [23]. Chloroquine (CQ) exhibits anti-tumor effects by blocking the fusion of autophagosomes and lysosomes and damaging lysosome function, and can enhance the sensitivity of various cancer cells, including GC cells, to chemotherapy [24]. Therefore, autophagic flux blockage is expected to become one of the targets for cancer treatment.
Timosaponin AIII (Tim AIII) is a natural steroid saponin extracted from the plant Anemarrhena asphodeloides. Research on various tumors has demonstrated its significant therapeutic effects. For example, it induces ferroptosis in non-small cell lung cancer [25], inhibits the migration of cervical cancer cells by influencing the MAPK pathway [26], and increases the efficacy of doxorubicin in treating colorectal cancer [27]. Tim AIII inhibits the progression of colorectal cancer by inducing ROS production and lipid peroxidation [28]. However, the effect of Tim AIII on gastric cancer (GC) remains unexplored. The purpose of this study was to investigate whether Tim AIII possesses anti-GC capabilities, its impact on gastric cancer cell growth and proliferation, as well as its influence on oxidative stress and autophagic flux in GC cells.

Materials and methods

Cell culture
The human gastric mucosal cell (GES-1) and gastric cancer cell (AGS, HCG27) were obtained from the Cell Bank of the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). The identification numbers for GES-1, AGS and HGC27 are CVCL_EQ22, CVCL_0139 and CVCL_1279 respectively. All cells were sourced through legitimate channels and were subjected to STR identification. The specialized culture media used for the GES-1, AGS, and HGC27 cells were obtained from Procell Life Science & Technology Co., Ltd. (Wuhan, China). The cells were cultivated in an environment that maintained a temperature of 37 °C, a CO2 concentration of 5 %, and sufficient humidity.

Reagents and materials
The primary antibodies used in western blotting assay are as follows: Protein Kinase B (AKT) (4691), P-AKT (4060), Phosphatidylinositol-3-kinase (PI3K) (4257), AMP-activated protein kinase (AMPK) (5831), P-AMPK (2535), Mammalian target of rapamycin (mTOR) (2983), P-mTOR (5536), which were purchased from Cell Signaling Technology; Sequestosome 1 (P62) (ab207305), microtubule-associated protein 1 light chain 3 alpha (LC3) (ab192890), Nuclear factor erythroid 2-related factor 2 (Nrf2) (ab62352), Kelch-like ECH-associated protein 1 (Keap1) (ab227828), Heme Oxygenase-1 (HO-1) (ab52947), Glutamate-cysteine ligase modifier subunit (GCLM) (ab126704), NQO1 (ab80588), and Glutathione Peroxidase 4 (GPX4) (ab125066), which were purchased from Abcam; and Rabbit polyclonal antibody to GAPDH (#AF7021), which was purchased from Affinity. The secondary antibody used in western blotting is Goat Anti-Rabbit IgG (H + L) HRP (S0001), purchased from Affinity.
In immunofluorescence assay, p62 antibody (ab207305) and Goat Anti Rabbit IgG H&L (ab150077) were purchased from abcam.
Timosaponin AIII (Tim AIII; T3395, Topscience), 3-Methyladenine (3MA; M129496, Aladdin), Chloroquine (CQ; C193834, Aladdin), 3-(4,5-dimethylthiazolyl-2)-2,5-diphenyltetrazolium bromide (MTT; M158055, Aladdin), Hoechst 33,342 and propidium iodide (P0137; Beyotime), lactate dehydrogenase (LDH) Cytotoxicity Assay Kit (C0016, Beyotime), Crystal violet (C110703, Aladdin), Cell Cycle and Apoptosis Analysis Kit (C1052, Beyotime), 2′,7′-dichlorodihydrofluorescein diacetate (DCF; D6883, Sigma-Aldrich), JC-1 Mitochondrial Membrane Potential Assay Kit (40706ES60, Yeasen Biotechnology), Malondildehide (MDA) Content Assay Kit (BC0025, Solarbio), Glutathione (GSH) Content Assay Kit (BC1175, Solarbio), Cell lysis buffer for Western and IP (P0013, Beyotime), BCA protein colorimetric assay kit (E-BC-K318-M, Elabscience), enhanced ECL chemiluminescence substrate kit (36,222, Yeasen Biotechnology), Enhanced immunization permeabilization buffer (P0097, Beyotime), QuickBlock™ blocking buffer (P0228, Beyotime), Antifade Mounting Medium with DAPI (P0131, Beyotime), Ad-mCherry-GFP-LC3B (C3011, Beyotime), Lyso-Tracker Red (C1046, Beyotime).

Cell viability assay
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) was used to determine the cell survival rate. Different concentrations of Tim AIII were used to treat GES-1, HGC27, and AGS cells for 24 and 48 h. Once the processing time was complete, the MTT working solution was subsequently added, and the mixture was incubated for 2 h at 37 °C. After the incubation, the MTT working solution was removed, and DMSO was added to dissolve the cells. Finally, the absorbance was measured at 490 nm.

Hoechst 33,342/ propidium iodide (PI) staining
For the experiment, cells were initially plated on coverslips in a 24-well plate, with the same number of AGS, HGC27, and GES-1 cells added as in the MTT assay. After 24 h of cell growth, the cells were treated with different concentrations of Timosaponin AIII for 48 h. After treatment, cells were washed twice with PBS and fixed with 4 % paraformaldehyde for 20 min. The fixed cells were then stained with an anti-fade mounting medium containing Hoechst 33,342/PI. After 5 min of staining, the cells were examined under a fluorescence microscope, and images were captured using red and blue fluorescence channels.

lactate dehydrogenase (LDH) release experiment
LDH Cytotoxicity Assay Kit was used to detect the toxic effect of Tim AIII. After AGS, HGC27, and GES-1 cells were treated with Tim AIII, the cell supernatant was collected, and 120 μL was added to a 96-well plate. Then, the prepared LDH detection working solution was added. The 96-well plates were wrapped in tin foil and incubated at room temperature on a shaker for 30 min. The absorbance was measured at 490 nm.

Colony formation assays
HGC27 and AGS cells were seeded in 6-well plates at cell counts of 800 cells per well. Fresh culture media containing varying concentrations of Tim AIII were added after 72 h of steady cell growth, and the media was replenished every three days. After drug treatment for 14 days, HGC27 and AGS cells were fixed with 4 % paraformaldehyde and stained for 2 h with 0.1 % crystal violet.

Cell cycle assays
HGC27 and AGS cells (1.5 × 105) were seeded in a 6-well plate and treated with different concentrations of Tim AIII for 48 h. After trypsinization, the cells were collected and centrifuged at 4 °C. The pellet was fixed with 1 mL of 70 % ethanol overnight at 4 °C, then centrifuged at 1500 rpm for 5 min. The supernatant was discarded, and the cells were washed with 1 mL PBS, centrifuged again, and resuspended in 100 μL PBS. RNA was removed by adding 2 μL of RNase A (10 mg/mL) and incubating at 37 °C for 30 min. The cells were then stained with 100 μL of PI solution (100 μg/mL) for 10 min in the dark. Cell cycle distribution was analyzed by flow cytometry with excitation at 488 nm and emission at 585 ± 21 nm, and the results were analyzed using Modfit software.

Wound healing assay
A total of 5 × 105 AGS and HGC27 cells were seeded into a 6-well plate. A straight horizontal line was drawn when the cells were 90 % confluent. After the wound was created, the cells were treated for 48 h with different concentrations of Tim AIII, and images were captured at 0 and 48 h.

Intracellular ROS detection
2′,7′-dichlorodihydrofluorescein diacetate (DCF) fluorescent probe was used for intracellular ROS staining. Configure DCF into a storage solution with a concentration of 10mM/L using DMSO. The storage solution was added to the cells treated with Tim AIII for incubation after diluting the mixture 1:1000 with basic culture media. After incubating for 30 min, images of the cells were taken using a fluorescence microscope to determine the presence of green fluorescence.

Mitochondrial membrane potential detection
Mitochondrial membrane potential damage was detected using JC-1 probe. After treatment with Tim AIII, AGS, HGC27, and GES-1 cells were incubated for 20 min in JC-1 (10 μM working solution) staining solution. Observed the changes in mitochondrial red and green fluorescence under a fluorescence microscope.

Malondildehide (MDA) assay
The intracellular MDA content was determined using Malondildehide Content Assay Kit. At least 5 million AGS, HGC27, and GES-1 cells were collected from each group and centrifuged to obtain cell precipitates. After the MDA extraction solution was added, the cells were subjected to repeated ultrasonic lysis. Drop 100 μL samples into the EP tube, then add the MDA detection working solution and proceed according to directions. Finally, the cells were transferred to a 96-well plate, and a microplate reader was used to measure the absorbance at 532 nm.

Glutathione (GSH) assay
Intracellular GSH levels were evaluated using Glutathione Content Assay Kit. To obtain cell precipitates, at least 5 million cells were collected from each group and centrifuged. The GSH reagent mixture was added, and the mixture was freeze‒thawed three times in a water bath at 37 °C and in liquid nitrogen. After centrifuging the samples to obtain the supernatant, the GSH detection working solution was added in accordance with the manufacturer’s directions. A microplate reader was used to measure the absorbance at 412 nm.

Western blotting
HGC27, AGS, and GES-1 cells were treated with different concentrations of Tim AIII for 48 h. The culture medium was discarded, and the cells were washed twice with PBS. Cells were collected using a cell scraper into 1.5 mL centrifuge tubes. The cells were centrifuged at 3000 rpm for 5 min at 4 °C to obtain the cell pellet. Cell lysis buffer for Western, along with appropriate Phenylmethanesulfonyl fluoride (PMSF), was added to the pellet, and the cells were lysed for 45 min. After lysis, the cell lysates were centrifuged at 12,000 rpm for 15 min at 4 °C to obtain the protein supernatant. The protein concentration in the supernatant obtained after 4 °C centrifugation was measured using a BCA protein colorimetric assay kit. The proteins were separated via electrophoresis on a polyacrylamide gel for 1.5 h and subsequently transferred to a polyvinylidene fluoride (PVDF) membrane, and then blocked by 5 % skim milk for 2 h. After washing Tris-buffered saline and Tween 20 (TBST) three times, the protein was incubated overnight on a shaking table at 4 °C with the corresponding antibodies. After the primary antibody was recycled, the protein was washed three times with TBST. The corresponding secondary antibiotics were added to incubate at room temperature for 1.5 h. After washing TBST three times, the enhanced ECL chemiluminescence substrate kit was used for development on the imaging system.

RNA sequencing and proteomics analysis
The RNA sequencing experimental group was divided into two groups: the treatment group included AGS cells treated with Tim AIII at a concentration of 3 µM for 48 h, while the control group received no intervention. Following the preparation, add TRIzol Reagent (Seville Biotechnology, China) for total RNA extraction. The extracted samples were subsequently sent to Beijing Easyresearch Co., Ltd., for machine sequencing and subsequent mRNA repository construction. After sample quality inspection, the samples were examined, and the completed data were evaluated via transcriptome gene analysis.
The experimental grouping for proteomics is the same as RNA sequencing. Using a clean cell scraper, the cells were carefully scraped into EP tubes. Cell precipitates were produced following 5 min centrifugation at 3000 rpm/min. The samples were subsequently propagated to Beijing Easyresearch Co., Ltd. Protein separation, protein identification, protein functional analysis, data processing, and analysis were the primary steps of the whole process.

Transmission electron microscopy
Collect cell pellets from the control group and Tim AIII group, and fix them with 2.5 % glutaraldehyde for 24 h, washed in PBS for three times. Then the cells were post-fixed with 1 % osmium tetroxide in 0.1 M PBS for 1.5 h at room temperature. After dehydration in a gradient series of Acetone, the cells were embedded in pure EMBed 812 and cut into 70 nm thick ultrathin sections. After uranyl acetate and lead citrate staining, the samples were examined by transmission electron microscope (HITACHI).

Immunofluorescence
After treatment with Tim AIII for 48 h, AGS, HGC27, and GES-1 cells were washed twice with PBS before being fixed for 20 min with 4 % paraformaldehyde. Enhanced immunization permeabilization buffer was added to treat cells for 10 min. After washing with PBS three times, QuickBlock™ blocking buffer was added for 20 min. Subsequently, an immunofluorescence primary antibody was added, and the samples were incubated overnight at 4 °C. Recycled the primary antibody, washed with PBS three times, added the secondary antibody, and incubated at room temperature for 1 h. Finally, Antifade mounting medium with DAPI was used. After sample preparation, the samples were observed and photographed using a confocal laser scanning microscope.

Ad-mCherry-GFP-LC3B staining
Ad-mCherry-GFP-LC3B was used to infect AGS, HGC27, and GES-1 cells at a density of approximately 30 %. To each sample intended for transfection, 675 µL of new complete culture media and 25 µL of adenovirus solution were added. After the target cells were infected for 24 h, the culture medium that contained the viral solution was discarded, and the cells were washed with PBS. The experimental groups were subsequently divided into a control group, Tim AIII group, CQ group, and rapamycin group. AGS and HGC27 cells were treated for 48 h and photographed under a confocal laser scanning microscope.

Lysosome red fluorescence probe assay
AGS, HGC27, and GES-1 cells were treated for 48 h with the specified concentrations of Tim AIII, CQ, or rapamycin. Following therapy, the culture medium was discarded, and Lyso-Tracker Red was diluted 1:20,000 with basic culture medium. The cells were incubated with the prepared working solution for 10 min. Afterward, the cells were washed twice with PBS and photographed using a laser confocal microscope.

Subcutaneous tumor formation in nude mice
Five-week-old male BALB/c nude mice were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. All nude mice were housed in a specific pathogen-free (SPF) animal facility. This study received approval from the Laboratory Animal Welfare and Ethics Committee of the Ministry of Medicine. We confirm that all experiments were performed in accordance with relevant named guidelines and regulation sand complied with the ARRIVE guidelines. The nude mice were injected subcutaneously on the right side with 150 μL of PBS containing 6 × 105 AGS cells. After 7 days of tumor growth, the nude mice were randomly divided into four groups, with four mice in each group. This study included four groups: the control group (PBS+DMSO) as the negative control, Tim-AIII (low-dose group, 5 mg/kg), Tim-AIII (high-dose group, 10 mg/kg), and 5-FU (30 mg/kg) as the positive control group. Control group (PBS+DMSO), Tim-AIII (5 mg/kg), Tim-AIII (10 mg/kg), and 5-FU (30 mg/kg) were injected intraperitoneally every two days. Starting from the day of drug administration, the weight of mice was recorded using a scale, and tumor growth was monitored by measuring tumor volumes using calipers at regular intervals (every 2 days). Tumor volume was calculated using the formula: (short diameter)2 × (long diameter)/2.

Statistical analysis
GraphPad Prism 8 was used to statistically analyze the data. The data are presented as the mean ± standard deviation (SD). Student's t-test was used to assess the differences between two groups. One-way ANOVA was used to compare the various groups. P < 0.05 was considered to indicate statistical significance. * P < 0. 05, ** P < 0. 01, *** P < 0.001.

Results

Tim AIII has significant cytotoxic effects on GC cells
The molecular structure of Tim AIII (MW, 740.92) used in this study (Fig. 1A). This study selected gastric cancer AGS and HGC27 cells for parallel validation, and evaluated drug safety using human gastric mucosal GES1 cells. Tim AIII had no significant influence on the survival rate of GES1 cells after 24 h or 48 h at a concentration of 0–5 μM, showing that GES1 cells could be safely tolerated in this concentration range (Fig. 1B). Tim AIII dramatically decreased the survival rate of AGS and HGC27 cells at lower concentrations in a time- and concentration-dependent manner (Fig. 1C-D). After Tim AIII treatment, variations in the quantity of AGS, HGC27 and GES-1 cells were apparent under a microscope (Fig. 1E-F, Supplementary Figure. 1A). Based on the above data, the Tim AIII concentrations used in AGS, HGC27 and GES-1 cells were 0, 1, 2, and 3 μM, for a total treatment time of 48 h. The production of LDH is positively correlated with the concentration of Tim AIII in AGS and HGC27 cells (Fig. 1G-H). The higher extracellular LDH levels reflect cellular metabolic arrest and the loss of cellular membrane integrity of AGS and HGC27 cells after Tim AIII treatment. Hoechst/PI staining were utilized to assess membrane damage caused by Tim AIII. Propidium iodide (PI) can permeate damaged cell membranes and cause red fluorescence, and the quantity of red fluorescence is positively correlated with the concentration of Tim AIII in AGS and HGC27 cells (Fig. 1I-J). There were no significant differences in LDH and PI staining between the control group and the Tim AIII-treated group of GES-1 cells (Supplementary Figure. 1B-C). These results indicate that Tim AIII inhibits the viability of GC cells and induces cytotoxic effects, whereas under the same treatment time and concentration, Tim AIII does not affect the viability of GES-1 cells nor induce cytotoxic effects.

Tim AIII inhibits GC cells migration and proliferation in vitro
Next, we evaluated whether Tim AIII affects the migration and proliferation of GC cells. The results of the Wound healing assay show that the migration of AGS and HGC27 cells was dramatically inhibited with increasing Tim AIII concentrations (Fig. 2A-B). The number and size of colonies formed by AGS and HGC27 cells were significantly inhibited by Tim AIII treatment (Fig. 2C). Quantitative analysis of the absorbance of cell colonies dissolved in 30 % acetic acid was performed in AGS and HGC27 cells (Fig. 2D-E). The cell cycle distribution of HGC27 and AGS cells was subsequently evaluated via flow cytometry. AGS and HGC27 cells in the S phase exhibited gradual shortening compared to those in the control group (Fig. 2F-G). These findings showed that Tim AIII can inhibit the migration and proliferation of AGS and HGC27 cells.

Tim AIII induces ROS production and damages the mitochondrial membrane potential
Based on the research findings of Tim AIII in other tumors, we first chose to investigate whether oxidative stress is involved in the mechanism of Tim AIII inhibition of GC cells. DCF fluorescent probes were used to investigate whether ROS levels increase after Tim AIII treatment. ROS (green fluorescence) were significantly accumulated in AGS and HGC27 cells and were positively correlated with the Tim AIII concentration (Fig. 3A-D), whereas Tim AIII did not induce ROS production in GES-1 cells (Supplementary Figure. 2A). Excessive ROS can lead to mitochondrial depolarization and damage the mitochondrial membrane potential (MMP) [29]. JC-1 is an effective fluorescent probe for detecting MMP: it emits red fluorescence when the MMP is normal. When the MMP is disrupted, it exhibits green fluorescence. The findings demonstrated that the green fluorescence intensity of AGS and HGC27 cells increased gradually following JC-1 probing and was positively linked with the Tim AIII concentration (Fig. 3E-F). No obvious green fluorescence was observed in Tim AIII treated GES-1 cells using JC-1 probe, but bright red fluorescence was detected (Supplementary Figure. 2B). These results confirmed that Tim AIII can induce ROS production and disrupt mitochondrial homeostasis in GC cells, but not in GES-1 cells.

Tim AIII affects redox substances and inhibits the Keap1-Nrf2 pathway
MDA, which reflects the degree of lipid oxidation, and GSH, which are important intracellular antioxidants, are both important indicators of oxidative stress. The results showed that as the concentration of Tim AIII increased, the GSH content decreased and the MDA content increased in AGS and HGC27 cells (Fig. 4A-D). In contrast, Tim AIII did not reduce the GSH content in GES-1 cells, nor did it increase the MDA levels (Supplementary Figure. 3A-B). Consequently, we detected alterations in the Keap1-Nrf2 pathway, one of the most significant signaling pathways associated with OS. Keap1 and Nrf2 separate each other in response to OS, and Nrf2 subsequently activates and transfers to the nucleus, triggering the transcription and translation of antagonist response elements (AREs), such as glutamate-cysteine ligase (GCLM), glutathione peroxidase 4 (GPX4), heme oxygenase 1 (HO-1) and NAD(P)H dehydrogenase (NQO1). Nonetheless, a growing number of studies on cancer have demonstrated that Keap1-Nrf2 is a pathway protecting cancer cells, and damaging this pathway can exert anti-cancer effects [30]. The Western blotting results showed an increase in Keap1 expression, while Nrf2, HO-1, GCLM, NQO1 and GPX4 expression showed a downward trend in AGS and HGC27 cells (Fig. 4E-G). Western blotting results of GES-1 cells treated with Tim AIII revealed that Tim AIII did not affect the protein levels of Keap1, Nrf2, NQO1, GCLM, and GPX4 (Supplementary Figure. 3C). Based on the above results, it can be concluded that Tim AIII can regulate the content of intracellular redox substances and inhibit the Keap1-Nrf2 pathway in GC cells, with no significant effect on the redox homeostasis of GES-1 cells.

RNA sequencing results are enriched in autophagy related pathways
RNA sequencing was subsequently used to elucidate the molecular mechanism by which Tim AIII inhibits GC progression. Total RNA was extracted from the control and Tim AIII (3 μM) groups after 48 h. Genes (Q value < 0.05, |fold change| > 1.5) that were upregulated or downregulated were identified. The data demonstrated that, compared to those in the control group, 954 Differentially expressed genes (DEGs) were differentially expressed, with 503 upregulated and 451 downregulated genes (Fig. 5A). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were used to investigate the effects of the DEGs on GC cells. The molecular function (MF) category was enriched mainly in protein binding and 1-phosphatidylinositol-3-kinase (PI3K) regulator activity (Fig. 5B). Biological process (BP) and cellular component (CC) category were enriched mainly in the regulation of PI3K and lysosomal-related functions (Fig. 5C-D). According to the KEGG analysis, these DEGs were strongly enriched in many signaling pathways, including the lysosome, phagosome, Ras signaling pathway, and AMPK signaling pathway (Fig. 5E). The gene set enrichment analysis (GSEA) results indicated that the DEGs are associated with genes upregulated through activation of the mTORC1 complex (Fig. 5F). Lysosomes, autophagosomes, PI3K, AMPK, and mTOR are pivotal intracellular regulatory factors that play critical roles in modulating cellular growth, oxidative stress, and autophagy processes through intricate interactions and precise regulation. In summary, RNA sequencing results indicate that in addition to oxidative stress, autophagy is also involved in the mechanism of Tim AIII acting on GC cells.

Proteomic analysis further confirms that the mechanism of Tim AIII is related to autophagy
To further explore the underlying molecular mechanisms involved, proteomic analysis was performed on AGS cells, and genes differentially expressed between Tim AIII-treated and untreated cells were identified. Differentially expressed proteins were screened using statistical methods, with the selection criteria for differential expression based on the chi-square test (P-value 〈 0.05, |fold change| 〉 1.5). The chi-square test calculates the P-value based on frequency analysis, with a P-value < 0.05 considered statistically significant. There were significant differences in the expression of 861 proteins (464 upregulated and 397 downregulated) (Fig. 6A). Clusters of Orthologous Groups (COG) analysis revealed enrichment in energy production and conversion and signal transduction mechanisms; these findings indicate that Tim AIII is closely related to cell growth, function, and regulation (Fig. 6B). The DEGs were enriched in the CC (early endosome, autophagosome, autolysosome, and late endosome), MF (DNA and protein kinase binding), and BP (DNA replication, autophagy, and cellular response to external stimulus) categories (Fig. 6C). KEGG pathway enrichment analysis was performed to identify the biological pathways that are significantly associated with the differentially expressed proteins identified in our proteomic analysis. Functional enrichment analysis of DEGs was performed based on the Fisher's exact test algorithm. The results are presented as p-values, with a p-value <0.05 indicating significant functional enrichment. A smaller p-value indicates more significant functional enrichment. The results indicated several pathways that were significantly enriched, including lysosome, autophagy and mitophagy pathways (Fig. 6D). These changes in protein levels further confirmed that the effect of Tim AIII on GC cells is related to autophagy.

Tim AIII induces autophagosome formation and regulates the PI3K/AKT and AMPK pathways in GC cells
It is not surprising that the differentially expressed genes identified through RNA sequencing and proteomics are highly enriched in autophagy, as oxidative stress is closely related to autophagy [31], OS can induce autophagy by activating the AMPK pathway or by inhibiting the PI3K-AKT pathway [32]. Therefore, next focus was on investigating whether Tim AIII regulated autophagy in GC cells and what role autophagy played. First, under transmission electron microscopy, a large number of autophagosomes and autophagic vacuoles were found in the Tim-AIII group compared to the control group (Fig. 7A-B). Then, we detected proteins that regulate autophagy initiation, such as mTOR, p-mTOR, PI3K, AKT, p-AKT, AMPK, and p-AMPK. Activated mTOR (PI3K-AKT or MAPK signal transduction) can inhibit autophagy, while negative regulation of mTOR (AMPK or p53 signal transduction) promotes autophagy [33]. Western blotting results revealed decreases in PI3K, p-AKT, and p-mTOR levels and activation of p-AMPK in AGS and HGC27 cells (Fig. 7C-E). Western blotting results for GES-1 cells showed that Tim AIII does not affect the expression of the PI3K-AKT-mTOR and AMPK-mTOR pathways in GES-1 cells (Supplementary Figure. 4A). The above results indicate that Tim AIII activates the autophagy pathway and induces autophagosome production in GC cells.

Tim AIII simultaneously upregulates the expression of P62 and LC3B in GC cells
This study next aimed to further investigate the autophagic process using the two most commonly used autophagosome markers, P62 and LC3 proteins [34]. P62 is a selective autophagy receptor that interacts with LC3 and facilitates the degradation of ubiquitinated substrates via autophagy. P62 as autophagosome marker because its accumulation is often used as an indicator of impaired autophagic degradation, since it is not efficiently cleared when autophagic flux is blocked or dysfunctional. The presence of high levels of P62 correlates with defects in autophagic degradation, which makes it a valuable marker for studying autophagic dysfunction. LC3B is a commonly used marker for autophagy, it is a well-established indicator of autophagic activity, primarily reflecting the formation of autophagosomes. When P62 decreases and LC3B increases at the same time, autophagy proceeds normally. However, Simultaneous overexpression of P62 and LC3B is a marker for late-stage autophagic flux blockage [18]. Western blotting for P62 and LC3B results exhibited the simultaneous increase in LC3B and P62 in AGS and HGC27 cells (Fig. 8A-B). Proteomic data also suggested upregulation of P62 and LC3B protein expression (Fig. 8C). Furthermore, immunofluorescence results indicated that the expression of P62 increased in HGC27 and AGS cells treated with Tim AIII, and was positively correlated with the concentration of Tim AIII (Fig. 8D-E). Correspondingly, Western blotting results did not reveal any changes in the expression of P62 and LC3 proteins in GES-1 cells treated with Tim AIII (Supplementary Figure. 4B). Additionally, immunofluorescence further confirmed that P62 did not accumulate in GES-1 cells after Tim AIII treatment (Supplementary Figure. 4C). Therefore, we hypothesized that Tim AIII induces autophagy in GC cells but blocks the progression of autophagic flux at later stages

Tim AIII inhibits late-stage autophagy by blocking autophagy‒lysosome fusion and interfering with lysosomal function
Tim AIII was used in combination with two common autophagy inhibitors in AGS and HGC27 cells, after which the changes in the cell survival rate were verified. 3-Methyladenine (3-MA) is a frequent early inhibitor of autophagy that can prevent the development of autophagosomes [35]. CQ, an extensively used late-stage autophagy inhibitor, undermines autophagy‒lysosome fusion, limiting autophagy substrate degradation and increasing LC3B accumulation [36]. The combination of Tim AIII and 3-MA did not kill AGS or HGC27 cells; instead, it showed some mild cellular viability recovery effects (Fig. 9A-B). The Tim AIII and CQ combination significantly increased the cytotoxic effect on AGS and HGC27 cells (Fig. 9C-D). Similarly, when Tim AIII was combined with 3-MA or CQ and applied to GES-1 cells, the MTT results showed that neither the combination of Tim AIII with 3-MA nor with CQ had any inhibitory effect on cell viability in GES-1 cells (Supplementary Figure. 5A-B). Based on the results above, Tim AIII is thought to influence the later phases of autophagy rather than the initiation and occurrence of early autophagy in GC cells. An adenovirus expressing the mCherry-GFP-LC3B fusion protein effectively expressed the LC3B fusion protein red fluorescent protein (mCherry) and green fluorescent protein (GFP) after infecting cells. During the fusion process between autophagosomes and lysosomes, the acidic environment within the lysosome can cause fluorescence quenching of mCherry. In the case of smooth autophagic flux (smooth fusion of autophagosomes and lysosomes), red spots appear due to quenching of GFP fluorescence. When autophagic flux is blocked, mCherry-LC3B and GFP-LC3B coexist and exhibit yellow fluorescent spots. Rapamycin (RAP) is an effective and specific mTOR inhibitor commonly used as an autophagy activator. To confirm the impact of Tim AIII on autophagic flux, the Tim AIII group was compared to the CQ group and the RAP group. The findings demonstrated that autophagy-lysosome fusion was inhibited following treatment with Tim AIII in AGS and HGC27 cells, as evidenced by a considerable increase in yellow-spot fluorescence and a significant decrease in red-spot fluorescence (Fig. 9E-F). The Ad-mCherry-GFP-LC3B staining in GES-1 cells results showed that Tim AIII does not induce autophagy in GES-1 cells nor affect the progression of autophagic flux (Supplementary Figure. 5C). Lysosome swelling and enlargement may be associated with decreased lysosomal division and impaired autophagic flux [37]. Lyso-Tracker Red is a cell membrane-permeable probe used in living cells for lysosomal-specific fluorescence staining. In AGS and HGC27 cells, compared to those in the control group, the lysosomes in the RAP group did not significantly change, while the lysosomes in the CQ group swelled. There were more swollen lysosomes in the Tim AIII group than in the CQ group, indicating that lysosomal function was significantly compromised (Fig. 9G-H). In contrast, Tim AIII does not induce lysosome swelling and enlargement in GES-1 cells (Supplementary Figure. 5D). In conclusion, our findings show that Tim AIII inhibits autophagy-lysosome fusion and impairs lysosomal function, suggesting that Tim AIII is a late-stage autophagy inhibitor that induces GC cell death.

Tim AIII inhibits subcutaneous tumor growth in vivo
The subcutaneous tumor model in nude mice was utilized to evaluate the potential of Tim AIII to inhibit GC in vivo. Approximately 6 × 105 AGS cells were injected into the subcutaneous region on the right side of the nude mice. After allowing the tumors to grow for 7 days, when the tumor size reached between 40 and 60 mm3, the mice were randomly divided into four groups (n = 4): the control group (PBS+DMSO), Tim AIII (5 mg/kg) group, Tim AIII (10 mg/kg) group, and positive control group treated with 5-FU (30 mg/kg). The drugs were administered intraperitoneally every two days for a total of 7 doses. The timeline of the experiment and tumor photographs were recorded (Fig. 10A-B). Compared to the control group, tumor growth was significantly inhibited in the Tim AIII treatment group, in which the tumor volume and weight were significantly lower (Fig. 10C-D); moreover, there was no significant difference in body weight (Fig. 10E). Histological examination of the hearts, livers, and spleens of nude mice from different groups via HE staining indicated no significant toxic side effects of Tim AIII on important organs. Overall, these results preliminarily confirm the anti-GC activity of Tim AIII in vivo.

Discussion
Numerous studies have demonstrated the severe cytotoxic and inhibitory effects of natural compounds derived from plants on tumor cells [38]. Natural compounds that selectively target cancer cells without harming normal cells within specific concentration ranges are increasingly favored in the prevention and treatment of tumors [39]. Tim AIII, the natural compound used in this research, has been examined in a variety of cancers, including melanoma, lung cancer, liver cancer, and ovarian cancer [40]. Given the research gap regarding Tim AIII in the field of GC, we initially demonstrated its therapeutic potential for GC. Tim AIII exhibits significant inhibition of GC cells based on cell viability, LDH release, and PI staining assays. Clone formation, cell cycle, and wound healing assay support Tim AIII's ability to inhibit GC cells proliferation and migration. Using normal human gastric mucosal epithelial GES-1 cells as a control, treatment with Tim AIII at the same concentrations and treatment times does not inhibit GES-1 cell viability, nor does it cause significant cytotoxicity. The above results suggest that the treatment concentrations and durations of Tim AIII selected in this study have a significant inhibitory effect on GC cells, while being safe and well-tolerated in GES-1 cells. In vivo experiments demonstrated that Tim AIII significantly inhibits tumor growth with no apparent toxic effects.
With the advancement of cancer research, cancer treatment has shifted from traditional targets of proliferation and apoptosis to new targets of oxidative imbalance, autophagy, and metabolism [41]. Tim AIII belongs to natural polyphenolic compounds with a chemical structure that includes numerous phenolic rings or hydroxy groups. This chemical structure's mechanism for inhibiting cancer cells is strongly linked to redox reactions [42]. As a result, in subsequent research into the causes of Tim AIII-induced GC death, oxidative stress is chosen as a target. A large number of studies have demonstrated that natural compounds can inhibit GC by inducing oxidative stress. For example, triptolide can directly bind to and inhibit the expression of the antioxidant enzyme peroxiredoxin-2, thereby inducing the accumulation of ROS, which triggers endoplasmic reticulum stress and mitochondrial dysfunction, ultimately leading to cell apoptosis [43]. Isoliquiritigenin inhibits glucose uptake and mitochondrial oxidative phosphorylation, increasing ROS accumulation, which significantly enhances apoptosis in GC cells [44]. Tanshinone inhibits the function of antioxidant-related protein pathways in gastric cancer cells via the P53 pathway, thereby inducing OS and ferroptosis in gastric cancer cells [45]. In this study, the increase in ROS and MDA levels, as well as the decrease in MMP and GSH levels, indicate a peroxidative state in the GC cells following Tim AIII treatment. This result suggests that Tim AIII induces OS in GC cells. Therefore, we further investigate whether Tim AIII regulates OS-related pathways in GC. The Keap1-Nrf2 pathway, a protective pathway involved in OS, has been shown to be activated under OS conditions. When cancer cells experience OS, Nrf2 dissociates from Keap1, leading to Nrf2 translocation to the nucleus, where it triggers the transcription and translation of downstream antioxidant genes. Research findings have revealed that inhibiting Keap1-Nrf2 pathway can decrease its detoxifying and antioxidant capacities, exhibiting increased anti-cancer potential [46,47]. miR-328–3p is overexpressed inGC cells and is positively correlated with anti-apoptotic and antioxidant stress capabilities. Mechanistically, miR-328–3p upregulates Nrf2 by targeting Keap1, thereby reducing the excessive production of ROS [48]. After treatment with Tim AIII, the expression of Keap1 was upregulated, while the expression of Nrf2, HO-1, GCLM, and NQO1 decreased in GC cells. The results suggest that Tim AIII not only induces oxidative stress in GC, but also inhibits the expression of Keap1-Nrf2 pathway. Similarly, using GES-1 cells treated with Tim AIII as a control, the results showed that the levels of ROS, MMP, MDA and GSH did not change significantly. Likewise, the protein expression levels of Keap1, Nrf2, GCLM, and NQO1 also showed no notable changes. These results suggest that although Tim AIII at this concentration range induces oxidative stress in GC cells, it does not affect the redox status in GES-1 cells.
The initiation of autophagy is closely connected to the energy stress induced by mitochondrial damage and the excessive generation of ROS [49]. Therefore, we further investigated whether Tim AIII has an impact on autophagy. The environment and status of the tumor determine whether autophagy inhibits or promotes growth [50]. Autophagy can result in the production of metabolites and energy sources that support tumor growth, promoting cancer cell survival and growth [51]. Excessive autophagy can also lead to cell death, as observed with sodium selenite, which activates autophagy and induces the production of numerous autophagosomes through the AMPK-mTOR pathway, ultimately resulting in apoptosis in cervical cancer cells [52]. Due to the complexities of autophagy in cancer, the use of targeted autophagy as a cancer treatment strategy is currently being investigated. In this study, according to the enrichment analysis of differentially expressed genes using RNA sequencing and proteomics analysis, Tim AIII is closely related to lysosomes, autophagosomes and autophagy-related pathways. The activation of AMPK and the inhibition of the PI3K-AKT pathway signal the initiation of autophagy. Meanwhile, transmission electron microscopy revealed the presence of numerous autophagosomes and autolysosomes in AGS and HGC27 cells after treatment with Tim AIII. subsequently, the experimental results showed the simultaneous upregulation of P62 and LC3B, we speculated that Tim AIII may act as a late-stage autophagy inhibitor, inducing autophagic flux blockage. Meanwhile, Tim AIII does not affect the expression levels of AMPK and the PI3K-AKT-mTOR pathway in GES-1 cells, nor does it alter the expression levels of P62 and LC3 proteins.
As a late-stage autophagy inhibitor, the antimalarial drug CQ has made significant progress in basic research aimed at inhibiting cancer progression by blocking autophagic flux [53]. However, high concentrations of CQ are required for these effects, which suggests that their selectivity and efficacy are insufficient. Therefore, it is necessary to identify effective autophagy inhibitors to explore new possibilities for cancer treatment. Some natural products exhibit strong anticancer potential by blocking the progression of autophagic flux, such as Tubeimoside [54], N-methylparoxetine [55] and Tetrandrine [56]. Dehydrocostus lactone inhibits the progression of GC by impairing lysosomal function and autophagic flux through the disruption of lysosome-associated membrane proteins 1 and 2 [57]. PTBP1 knockdown impairs autophagic flux and inhibits GC progression through TXNIP‑mediated oxidative stress [23]. The combination of Tim AIII and CQ demonstrated a synergistic inhibition of cell viability in AGS and HGC27 cells, which preliminarily suggests that Tim AIII targets the late-stage progression of autophagy. By comparing the Ad-mCherry-GFP-LC3B staining data among the control, CQ, RAP, and Tim AIII groups, we discovered that Tim AIII inhibited autophagy-lysosome fusion in GC cells. Subsequently, the Lyso-Tracker Red staining data showed that Tim AIII might interfere with lysosomal activity. Correspondingly, experimental results indicate that Tim AIII does not affect the progression of autophagic flux or lysosomal function in GES-1 cells.
Tim AIII has been demonstrated in this study to inhibit the progression of GC by inducing oxidative stress and blocking autophagic flux. Analysis of related intracellular signaling pathways revealed that Tim AIII suppresses the Keap1-Nrf2 and PI3K/AKT/mTOR pathways while activating the AMPK pathway. Using GES-1 cells as normal cell controls, experimental results revealed that under the same concentration and treatment time, Tim AIII does not affect the cell viability, redox status, or autophagic process in GES-1 cells. By comparing the experimental results between GC cells and normal gastric mucosal epithelial cells, it was found that Tim AIII strongly inhibits GC cells within a particular concentration range, while GES-1 cells tolerate it well. This provides a valuable reference for the development of new drugs for GC. However, the mechanism of Tim AIII in GC remains to be further explored in future studies. Increased LDH activity and its elevated extracellular levels are first as a sign of cellular metabolic arrest and second as a verdict for the loss of cellular membrane integrity [4,58]. This study observed significantly elevated LDH release in Tim AIII-treated gastric cancer cells, but lacked further investigation. Therefore, the association between LDH elevation and cellular metabolic arrest, even glucose metabolism and lactylation, are future studies that should be explored. ROS can induce necrosis and enhance apoptotic condition mainly due to mitochondrial dysfunction [59]. The presence of JC-1 in this study indicates the role of ROS-induced mitochondrial dysfunction in this process, however, the lack of data on inducible apoptotic signals cannot distinguish between necrosis-dependent and apoptotic cell death. This is a limitation in this study, and further research is needed to explore and validate the relevant experiments. In this study, Tim AIII is considered as an inhibitor of late-stage autophagy. However, further validation of the molecular mechanisms is necessary. For instance, it remains to be explored whether Tim AIII damages the acidic lysosomal environment, whether it increases lysosomal membrane permeability, and how the interaction between autophagy-related pathways and lysosomal dysfunction occurs.

Conclusion
Based on the study results, we concluded that Tim AIII is a novel autophagy inhibitor. In particular, we discovered that Tim AIII stimulates autophagy and oxidative stress in GC cells and interferes with autophagy‒lysosome fusion at the late stage of autophagy. Therefore, as a natural compound, Tim AIII may target oxidative stress and autophagic flux as a new opportunity for gastric cancer therapeutic treatment.
Data availability statement
The datasets generated and analyzed during the current investigation are available from the corresponding author upon reasonable request.

Abbreviations
Tim AIII: Timosaponin AIII
OS: oxidative stress
ROS: reactive oxygen species
GSH: glutathione
MMP: mitochondrial membrane potential
MDA: malondialdehyde
Keap1: Kelch-like ECH-associated protein 1
Nrf2: NF-E2-related factor 2
HO-1: haem oxygenase 1
NQO1: NAD(P)H quinone dehydrogenase 1
GCLM: glutamate-cysteine ligase
GPX4: glutathione peroxidase 4
AMPK: AMP-activated protein kinase
AKT: protein kinase B mTOR: mammalian target of rapamycin
CQ: chloroquine
3-MA: 3-Methyladenine

CRediT authorship contribution statement
Chunyang Zhu: Writing – original draft, Visualization, Methodology, Formal analysis, Conceptualization. Shuming Chen: Resources, Methodology, Formal analysis, Conceptualization. Yangyang Lu: Resources, Methodology, Conceptualization. Jialin Song: Resources, Methodology. Shasha Wang: Resources, Formal analysis. Jing Guo: Resources, Data curation. Xiaoxi Han: Methodology, Investigation. YuanYuan Fang: Conceptualization. Siyi Zhang: Software, Resources. Wensheng Qiu: Writing – review & editing. Weiwei Qi: Writing – review & editing, Funding acquisition.

Declaration of competing interest
There are no conflicts of interest by the authors.