Human gastric cancer (GC) cell lines HGC-27 (Cat NO., CL-0107), SNU-1 (Cat NO., CL-0474), and human gastric mucosa epithelial cells GES-1 (CL-0563) were supplied by Procell Life Science & Technology (Wuhan, China). Mycoplasma testing was negative for all cell lines that underwent validation by Short Tandem Repeats (STR) profiling. At 37 °C in a 5% CO₂ atmosphere, cells were cultured in Roswell Park Memorial Institute (RPMI) 1640 supplemented with 10% fetal bovine serum (FBS), streptomycin (100 µg/mL), penicillin (100 U/mL).

We used a Multiskan™ FC microplate reader (Thermo Fisher, Carlsbad, CA, USA) to measure the optical density at 450 nm after exposing GC cells to Ole at various doses (0, 2.5, 5, 10, 20, 40, 80, and 160 nM) [17] for 24 hours after overnight culture in 96-well plates. Then, we tested the cells using a CCK-8 assay.

Following the injection of 500 GC cells into plates for a duration of 24 hours, the cells underwent a ten-day treatment with Ole at several doses (0, 10, 20, 40 nM). Before fixation with 4% paraformaldehyde, the cells were washed twice with PBS (0.2 M, pH 7.8). Subsequently, they were subjected to staining with 1% crystal violet for 15 minutes at ambient temperature. Colonies comprising over fifty cells were enumerated and analyzed. The colony formation rate can be determined by dividing the number of treatments by the number of controls and subsequently multiplying the result by 100%.

After seeding GC cells onto 96-well plates, they were treated with Ole at multiple concentrations (0, 10, 20, 40 nM) for a duration of 24 hours. In order to stain the cells, calcein AM and PI were diluted a thousand times with the working solution. Olympus Corporation’s BX41 fluorescence microscope (Tokyo, Japan) was used to examine and capture images of the fluorescence.

GC cells were cultured for 24 hours with gradient concentration of Ole (0, 10, 20, 40 nM) or 3-methyladenine (3-MA) (30 µM). The cells were incubated with 75% ethanol at 4 °C overnight, and the following day, they were stained with PI for cell cycle analysis. Apoptosis assessment was conducted on cells treated with Ole and stained with Annexin V-Fluorescein Isothiocyanate (FITC) and Propidium Iodide (PI).

Ole (0, 10, 20, 40 nM) or 3-MA (30 µM) were applied to GC cells during 24 hours. Various proteins involved in cell cycle regulation, intrinsic and extrinsic apoptosis, and autophagy were examined using western blotting. To summarize, the proteins from these cells were isolated by lysing them on ice with Radio Immuno Precipitation Assay (RIPA) lysis buffer. Beyotime Biotech’s Bicinchoninic Acid (BCA) Protein Assay Kit (P0011, Shanghai, China) was used to quantify the supernatant. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was used to extract proteins of the same grade, which were then transferred to nitrocellulose membranes. Skim milk in PBS (0.2 M, pH 7.8) was used to block the membranes for 2 hours. The next step was to incubate the primary antibodies (rabbit anti-human cell division cyclin 25 homolog C (Cdc25c) (1:1000, cat. No. 4866S), rabbit anti-human Cdk1(1:1000, cat. No.77055S), rabbit anti-human Cyclin B1 (1:1000, cat. No. 12231S), rabbit anti-human Caspase 3 (1:1000, cat. No. 9662S), rabbit anti-human Bak (1:1000, cat. No. 12105S), rabbit anti-human Bax (1:1000, cat. No. 2774S), rabbit anti-human Bcl-2(1:1000, cat. No. 4223S), rabbit anti-human PARP (1:1000, cat. No. 9532S), rabbit anti-human Caspase 8 (1:1000, cat. No. 4790S), rabbit anti-human Fas (1:1000, cat. No. 4233S), rabbit anti-human Fas ligand (FasL) (1:1000, cat. No. 68405S), rabbit anti-human microtubule-associated protein1 light chain3B (LC3B) (1:1000, cat. No. 3868S), rabbit anti-human Beclin (1:1000, cat. No. 34) at 4 °C throughout the night. Membranes were incubated at room temperature for 1 hour with HRP-conjugated secondary antibodies (Proteintech, SA00001-2, Chicago, IL, USA) not long ago. Alpha’s FluorChem FC2 (92-13779-00, Center Valley, CA, USA) was used to detect the blots. Using GAPDH, the data was standardized.

After a 24-hour Ole treatment at 20 nM, GC cells were collected and fixed with paraformaldehyde. Cell blocking and permeabilization were done with 5% BSA and Triton X-100 (Beyotime Biotech, Shanghai, China). Overnight, the cells were exposed to anti-microtuble-associated protein light chain 3 (LC3) antibody from Cell Signaling Technology (A32731, A11072, Cambridge, MA, USA) at 4 °C. Nuclear DNA was counterstained with DAPI for 15 minutes in the dark after cells were incubated with Alexa-Fluor 488/594-conjugated secondary antibodies (Thermo Fisher, Carlsbad, CA, USA) in 1% bovine serum for 1 hour at 37 °C. A confocal laser scanning microscope (Olympus Corporation, Center Valley, CA, USA) was used to take the pictures.

1 $\times$ 10⁶ GC cells were treated with 20 nM Ole for 24 hours and then rinsed twice with pre-cooled PBS (0.2 M, pH 7.8). After removing the cells and fixing them with 2.5% glutaraldehyde at 4 °C for the night, electron microscopy slices were prepared. Using an ultramicrotome (50–60 nm), the fixed cells were embedded, sectioned, and then dried. The slices were then double-stained with 3% uranyl acetate and lead citrate. Ultimately, cellular ultrastructures were examined and captured using HITACHI transmission electron microscopy (HT7700-SS, Tokyo, Japan).

All animal experiments adhered to the National Institutes of Health (Bethesda, MD, USA) guidelines for laboratory animal care and utilization. Weighing 90–120 grams, 10 female Balb/c nude mice were supplied by the Laboratory Animal Center of Xi’an Jiaotong University. The mice were aged 4–6 weeks. All mice were fed normal diet during the test period. A subcutaneous injection was made into the right flank of nude Balb/c mice containing luciferase-positive HGC-27 cell suspensions (5 $\times$ 10⁶ cells per mouse). Two groups of mice were randomly assigned after the tumor volume reached 50 mm³. Mice in the Ole group got 50 µg/kg Ole intraperitoneally every three days, while the control group received 1% Dimethyl Sulfoxide (DMSO) in PBS (pH 0.2 M, pH 7.8). Mouse weight, breadth (W) and tumor length (L) were measured every three days. On days 6 and 18, we used Perkin Elmer’s IVIS imaging equipment (Perkin Elmer, Waltham, MA, USA) for bioluminescence imaging. Mice were intraperitoneally given 100 µL of PBS (0.2 M, pH 7.8) with 25 mM D-luciferin (Caliper Life Sciences, Hopkinton, MA, USA) ten minutes before luciferase detection. The formula for tumor volume is L $\times$ (W)²/2. On the 21st day after Ole injection, all mice were euthanized by inhaling 5% isoflurane for 5 minutes, and their tumors, hearts, spleens, livers, kidneys, and lungs were removed for immunohistochemistry and H&E staining.

After being fixed in 4% formaldehyde, the tumor tissues were subjected to alcohol gradient drying, embedded in paraffin at 60 °C for 3 hours, sawn into pieces, and then baked. At ambient temperature, the antigen was repaired for 15 minutes in a citrate buffer (0.01 M, pH 6.0), and then, to inhibit endogenous peroxidase, it was incubated for another 15 minutes in 3% hydrogen peroxide. For 30 minutes, we let the serum block at room temperature. After being treated overnight with anti-Ki67 antibodies (1:500, Proteintech, 98224-1-RR, Wuhan, China), the slices were incubated with HRP-conjugated goat anti-rabbity IgG (1:2000, Sigma-Aldrich, AP-307P, Darmstadt, Germany) for 2 hours at room temperature. Using 3, 3-diaminobenzidine (DAB), the chromogen may be observed under a microscope. A phase contrast microscope (CKX53, Olympus, Tokyo, Japan) was used to examine the slices after dehydrating them and staining them with hematoxylin for three minutes.

The tumors, hearts, livers, spleens, lungs, kidneys, and other tissues were preserved by fixing, drying, embedding in paraffin, sectioning, and baking. Xylene was employed for dewaxing and hydration. The slices were subsequently stained with hematoxylin and differentiated using 0.1% hydrochloric acid ethanol. Subsequently, the sequential processes of eosin staining, alcohol dehydration, xylene transparency, and resin sealing were executed. The slices were finally examined and photographed using a phase contrast microscope. For morphological evaluation, slices were stained with terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling (TUNEL).

Data is presented as mean $\pm$ standard error of the mean (SEM). A parametric or nonparametric Student’s t-test was used to determine the statistical significance of two groups, whereas for multiple pairwise comparisons involving more than two groups, one-way ANOVA with Tukey’s post-hoc testing was utilized. In every instance, * p 0.05, ** p 0.01, *** p 0.001. Utilizing GraphPad Prism 9 software (Dotmatics, Boston, MA, USA), statistical analysis was carried out.

The inhibition effect of Ole (chemical structure featured in Fig. 1A) on proliferation of HGC-27 and SNU-1 cells was tested by CCK-8. The results showed that the proliferation of HGC-27 and SNU-1 cells were significantly inhibited by Ole in a dose-dependent manner compared to GES-1 cells (p 0.05, Fig. 1B). The half-maximal inhibitory concentration (IC50) for HGC-27 cell was 6.36 n M at 24 h and 4.13 n M at 48 h, IC50 for AGS cell was 5.94 n M at 24 h and 3.36 n M at 48 h, and IC50 for GES-1 was more than 10 uM (Fig. 1C) both at 24 h and 48 h. According to these results, we choose 0, 10, 20, 40 n M for the following assays. The results of colony formation assay showed that Ole could evidently inhibit the proliferation of HGC-27 and SNU-1 cells even at 10 n M compared with GES-1 cells, and there were remarkable inhibitions at 20 n M and 40 n M (p 0.05, Fig. 1D). As a result, we confirmed that Ole possess an obvious inhibition on proliferation of HGC-27 and SNU-1 cells in a dose-dependent manner at low nanomolar concentrations. Then we tested the effect of Ole on cell cycles of HGC-27 and SNU-1 cells. The flow cytometry essay revealed that Ole could induce cell cycle arresting at G0/G1 phase of HGC-27 and SNU-1 cells in a dose-dependent manner ((p 0.05, Fig. 1E). And Western blot results also showed that cell cycle-related proteins such as cyclin B1, Cdc25c, Cdk1 in HGC-27 and SNU-1 cells treated with Ole were significantly down-regulated depending on Ole dosage (p 0.05, Fig. 1F). These results indicated that Ole could induce gastric cancer cells arresting at G0/G1 phase in a dose dependent manner.

Fig. 1.

Oleandrin (Ole) suppresses the growth of gastric cancer (GC) cells by drawing the cell cycle into a G0/G1 arrest. (A) Chemical structure of oleandrin; Molecular formula: C₃₂H₄₈O₉; Molecular weight: 576.72 g/mol. (B) Ole was applied to HGC-27, SNU-1, and GES-1 cells at different doses for 24 and 48 hours, and the Cell Counting Kit (CCK)-8 assay was used to evaluate the cells’ viability. (C) IC₅₀ of Ole on GC and GES-1 cells at 24 hours and 48 hours. (D) Ole reduced the colony development of GC cells. (E) The impact of Ole at various doses on the cell cycle of GC cells was assessed using flow cytometry. (F) The quantities of cell cycle-associated proteins treated with varying amounts of Ole were assessed by western blotting analysis. In contrast to the control group, *p 0.05, **p 0.01, ***p 0.001.

Calcein/PI and flow cytometry was performed to verify the apoptosis effect of Ole on gastric cancer cells. The results shown that active cells stained by Calcein Acetoxymethyl Ester significantly decreased and apoptotic cells stained by Propidium Iodide increased depending on dosage in HGC-27 and SNU-1 cells (p 0.05, Fig. 2A). The flow cytometry also exhibited that apoptotic cells increased in a dose-dependently manner of HGC-27 and SNU-1 cells (p 0.05, Fig. 2B). These results suggested that Ole could induce apoptosis of gastric cancer cells depend on concentration. Further more, Western blot results revealed that cle-Caspase 3, Bax, Bak, cle-PARP were up-regulated, and Bcl-2 was down-regulated in a Ole dosage-dependent manner (p 0.05, Fig. 3A). However, no significant changes existed in cle-Caspase 8, Fas, FasL depending on Ole dosage (p $>$ 0.05, Fig. 3B). Cle-Caspase 3, Bax, Bak, cle-PARP and Bcl-2 are endogenous apoptotic-related proteins, whereas cle-Caspase 8, Fas, FasL are extrinsic apoptosis-related proteins. These results indicated that Ole induces gastric cancer cells apoptosis with an intracellular mechanism.

Fig. 2.

Ole triggers apoptosis in GC cells. (A) Calcein/PI assays was employed to clarify the apoptotic potential of GC cells exposed to varying doses of Ole. (Scale bars = 200 µm) (magnification: $\times$200). (B) Flow cytometry was employed to evaluate the apoptotic status. In contrast to the control group, *p 0.05, **p 0.01, ***p 0.001.

Fig. 3.

Ole triggers intrinsic apoptosis in GC cells. (A) Western blotting analysis of the intrinsic apoptosis proteins in GC cells triggered by Ole at varying doses. (B) Western blotting analysis of the extrinsic apoptosis proteins in GC cells. In contrast to the control group, *p 0.05, **p 0.01, ***p 0.001. Bak, Bcl-2 Antagonist Killer; Bax, Bcl-2 Associated X Protein1; PARP, Poly ADP Ribose Polymerase3.

The variation of autophagy treated with Ole in HGC-27 and SNU-1 cells were investigated. Transmission electron microscopy revealed that the formation of autophagic vacuoles in HGC-27 and SNU-1 cells treated with Ole was obviously increased compared to control group (Fig. 4A). And the confocal microscopy also showed that Ole lead to aggregation of autophagosomes in HGC-27 and SNU-1 cells compared to control group (Fig. 4B). Western blot results also showed that autophagy related proteins such as LC3BⅡ/LC3BⅠ ratio and beclin-1 were increased and p62 was decreased (p 0.051, Fig. 4C). These results suggested that Ole could promote autophagy accumulation in gastric cancer cells. 3-MA, an autophagy inhibitor, was used to identify autophagy flux induced by Ole in gastric cancer cells. The flow cytometry revealed that the apoptosis of HGC-27 and SNU-1 cells in Ole-treated group was increased compared to Ole+3-MA treated group (p 0.05, Fig. 5A), and the autophagy-related proteins LC3BⅡ/LC3BⅠ ratio was up-regulated and p62 was down-regulated in Ole-treated group compared to Ole+3-MA treated group. Meanwhile, the endogenous apoptosis-related proteins cle-Caspase 3, Bax and cle-PARP were significantly up-regulated in Ole-treated group compared to Ole+3-MA treated group (p 0.05). Whereas extrinsic apoptotic associated proteins (Cle-Caspase 8, FasL) had no significant change in Ole-treated group compared to Ole-3-MA treated group (p $>$ 0.05) (Fig. 5B). These results suggested that Ole induces gastric cancer cells apoptosis in an endogenous mechanism depending on activation of autophagy.

Fig. 4.

Ole promotes autophagy in GC cells. (A) The autophagosomes of GC cells treated with Ole (20 nM) were examined using transmission electron microscopy (red arrow: autophagosomes). (Scale bars = 1.5 µm, 500 nm, 200 nm). (B) The development of endogenous LC3 puncta in GC cells treated with Ole (20 nM) was assessed using immunofluorescence (magnification: $\times$400) (red arrow: LC3 puncta). (Scale bars = 10 µm). (C) The amounts of autophagy-related proteins in GC cells induced by Ole were assessed by western blot analysis. In contrast to the control group, *p 0.05, **p 0.01, ***p 0.001.

Fig. 5.

Ole promoted endogenous apoptosis by enhancing autophagy in GC cells. (A) Apoptosis of GC cells treated with Ole (20 nM) in the existence or non-existence 3-methyladenine (3-MA) (30 µM) were examined by Flow cytometry. (B) Western blotting analysis of the protein levels of GC cells treated with Ole (20 nM) either with or without 3-MA (30 µM). In contrast to the control group, *p 0.05, **p 0.01. In contrast to 3-MA + Ole (20 nM), #p 0.05, ##p 0.01.

Subcutaneous xenograft model of gastric cancer cells were used to define effect of Ole on tumor growth. The tumour size and weight in Ole (50 μg/kg) group were significantly reduced compared to control group (p 0.05, Fig. 6A–C), and IVIS images exhibit the same changes in tumour size and weight (p 0.05, Fig. 6D). H&E staining shown tumor cells with irregular shape, uneven cell size, large nuclei, and deep cytoplasmic staining in Ole group compared to control group. And Tunel positive staining was increased in Ole group compared to control group (p 0.05, Fig. 6E). Protein expression level of tumor tissue revealed by western blot confirmed that endogenous apoptosis-related proteins (cleaved-Caspase 3, PARP, Bax) were up-regulated and autophagy-related proteins (LC3BⅡ/LC3BⅠ) were up-regulated and p62 were down-regulated in Ole (50 μg/kg) group compared to control group (p 0.05), whereas exogenous apoptosis-related proteins (Cle-Caspase 8, FasL) had no significant change in these two groups (p $>$ 0.05, Fig. 6F). These in vivo results confirmed that Ole could inhibit tumor growth by inducing endogenous apoptosis and activating autophagy in gastric cancer.

Fig. 6.

Ole suppressed tumor growth by promoting apoptosis and autophagy in living organisms. (A) The tumor images in the DMSO group and the Ole (50 µg/kg) group. (B) The tumor volumes of the mice were monitored every three days. (C) Tumor weight of mice was assessed after 21 days of therapy in the DMSO group and the Ole (50 µg/kg) group. (D) IVIS images of murine tumors after 6 days and 18 days in the DMSO group and the Ole (50 µg/kg) group. (E) H&E staining (green framework: Infiltrating inflammatory cells) (magnification: $\times$200), immunohistochemistry for Ki67 (red arrow: Ki67 positive cells) (magnification: $\times$400), and terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling (TUNEL) assay (red arrow: apoptotic cells) (magnification: $\times$400) for apoptosis of tumor tissue in the Ole (50 µg/kg) group compared to the DMSO group. (Scale bars = 50, 20, 20 µm). (F) The expression of apoptosis-related and autophagy-related proteins in tumor tissue of the Ole (50 µg/kg) group compared to the DMSO group. In contrast to the control group, *p 0.05, **p 0.01.

The toxicity of Ole on tumor bearing animals were evaluated. The body weight and liver weight has no significant changes in Ole group compared to control group (p $>$ 0.05, Fig. 7A,B). And H&E staining of different organs uncovered that there was no evident damage on heart, liver, spleen, lung and kidney of mice (magnification: 200$\times$) in Ole group compared to control group (Fig. 7C). These results confirmed that the effective dosage of Ole on gastric tumor is safe for animals.

Fig. 7.

The toxicity of Ole in tumour bearing mice. (A) Mice’s weight was tracked every three days. (B) The mice’s liver weight was assessed after a 21-day period. (C) The toxicity of Ole was assessed by measuring the H&E staining of sections of heart, liver, spleen, lung, and kidney tissue (Scale bars = 100 µm).