Apatinib exhibits synergistic effect with pyrotinib and reverses acquired pyrotinib resistance in HER2-positive gastric cancer via stem cell factor/c-kit signaling and its downstream pathways

Human GC cell lines (NCI-N87, SGC7901, MGC-803, and MKN45) were purchased from the Cell Bank of Chinese Academy of Sciences (Shanghai, China). Human HER2-positive GC SNU216 cells were a generous gift from Professor Ruihua Xu, of Sun Yat-sen University Cancer Center (Guangzhou, China). NCI-N87-AR is a pyrotinib-resistant cell line obtained by continuous exposure of NCI-N87 cells to pyrotinib. All cells were cultured in RPMI-1640 (Hyclone, Provo, UT) containing 10% fetal bovine serum (FBS; Gibco, MD, USA) in an incubator with 5% CO₂ and proper humidity, at 37 ℃. Pyrotinib (SHR1258) and apatinib (YN968D1) were obtained from Hengrui Medicine Co., Ltd (Lianyungang, China). Reagents were dissolved in dimethyl sulfoxide (DMSO) and stored at − 20 ℃. Recombinant human SCF was purchased from MedChemExpress LLC. (New Jersey, USA) and maintained at − 20 ℃ until further use.

Cells (4000–8000 cells/100μL per well) were seeded in 96-well plates and incubated at 37 °C overnight. After 72 h of treatments, the Cell Counting Kit (CCK)-8 reagent (Promotor, Wuhan, China) was added at a concentration of 10%, and the absorbance of the samples was measured at 450 nm using a microplate reader (BioTek, VT, USA). The experiment was performed in triplicate and data were representative of three separate experiments.

Cells were cultured in 6-well plates at a density of 800–1000 cells/well. The culture medium was renewed to normal after 72 h of different treatments and incubated for 2 weeks. The colonies were fixed and stained with 0.1% Crystal Violet, and colony (> 50 cells/colony) counts of the samples were then compared to that of the control group. The experiments were performed in triplicate and data were representative of three separate experiments.

Cells (5 × 10⁴) suspended in 200 µL serum-free medium were transferred into the upper chamber of a 24-well insert containing a membrane with an 8-μm pore size (Corning, New York, USA), and 500 μL medium containing 20% FBS was added to the bottom chamber. After incubation for 72 h, the cells that crossed the membrane were fixed and stained with 0.1% Crystal Violet for 30 min. For each sample, 10 fields of view were counted and the average was calculated (100 ×). The final data were obtained from three independent experiments.

5 × 10⁵ cells were obtained, washed twice with PBS, and re-suspended in 200 μL of 1 × binding buffer (BD Biosciences, NJ, USA), stained with 5 µL of FITC Annexin V, and incubated in the dark for 15 min, at 37 ℃. 5 µL of propidium iodide (PI) and 300 μL of 1 × binding buffer were then added 5 min prior to testing in the dark, at room temperature. Samples were measured by flow cytometry on an LSRFortessa cell analyzer (BD Biosciences), and the results were analyzed using FlowJo software (v.10.0; TreeStar, CA, USA). The final data were obtained from three independent experiments (Table 1).

Total RNA was extracted using RNAiso Plus (Takara, Dalian, China), and reverse-transcribed to cDNA with the RevertAid first-strand cDNA synthesis kit (Thermo Fisher) according to a standard protocol. Real-Time qPCR was performed using the 7900HT Fast real-time PCR system (Thermo Fisher). GAPDH and β-actin were used as control and the 2^−ΔΔC_t method was used to quantify the relative mRNA expression of the genes. The sequences of primers are enlisted in the Supplementary Table 2. The experiments were performed in triplicate and the data were obtained from three separate experiments.

Total protein was extracted with a radioimmunoprecipitation assay (RIPA) buffer (Beyotime, Shanghai, China) supplemented with phenylmethylsulphonyl fluoride (PMSF) and phosphatase inhibitors (Servicebio, Wuhan, China). Protein quantification was then performed using the BCA reagent (Beyotime), following the manufacturer's instructions. The extracted proteins were loaded into 10% SDS-PAGE for separation, and transferred to a 0.45 μm polyvinylidene fluoride (PVDF) membrane (Millipore, Massachusetts, USA). Primary antibodies were added, and the membrane was incubated at 4 ℃ overnight. Subsequently, the membrane was incubated with secondary antibodies (Promotor) at room temperature for 1 h. Post incubation, the SuperSignal West Pico Chemiluminescent Substrate (Thermo Fisher) was added. The immuno-reactive bands obtained were analyzed using the G: BOX Chemi X system (Syngene, Cambridge, UK). The primary antibodies used are enlisted in the Supplementary Table 3.

Female nude mice (4–6-weeks old; BALB/c nu-nu) were purchased from the Hunan SJA Laboratory Animal Co., Ltd. (Changsha, China), and raised in a specific pathogen-free laboratory. Cells (5 × 10⁶ cells/100 μL) were injected subcutaneously into the left posterior side of each mouse, and mice were randomized into different groups (n = 6) when the tumors reached about ~ 100 mm³. The size of the tumor and the weight of the nude mice were measured every 3 days, and the tumor volume was calculated as follows: [(short diameter)² × (long diameter)]/2. Mice were sacrificed after 21 days. Tumor samples were collected, photographed, and stained. The fixed dose ratio of apatinib/pyrotinib was determined from the median effective dose (ED₅₀) of the two monotherapies to prevent bias. The combination index (CI) was calculated to determine the combined effect, with CI < 1, CI > 1, and 1 used to defined synergism, antagonism, and additivity, respectively. To determine the statistical relevance of CI < 1 [i.e., ln(CI) < 0], the variance of ln(CI) was calculated according to a previous described method [16]. All institutional and national guidelines for the care and use of laboratory animals were followed.

Chart generation and statistical calculations were performed using the GraphPad Prism software (v8.0; CA, USA), and SPSS (v19.0; NY, USA) was used for statistical analysis. All descriptive statistics were presented as the mean ± standard deviation (SD). Statistical comparisons between two experimental groups were performed using Student's t test. If multiple groups were compared, the one-way analysis of variance (ANOVA) was performed first. The least significant difference t test was applied if the overall difference was statistically significant. The main objective of statistical analysis in isobologram studies was to generate an upper confidence limit for the CI to determine whether the observed synergy was statistically relevant (CI < 1). A P < 0.05 was considered statistically significant for all tests.

To establish suitable in vitro HER2-positive GC models, we first examined the expression of HER2 in five GC cell lines, and the NCI-N87 and SNU216 cells were revealed to be HER2-positive. The relative expression of EGFR and VEGFR2 receptors was also examined (Fig. 1a–c). We then evaluated the inhibitory effects of pyrotinib and apatinib on the five GC cells. Compared with the other cells, NCI-N87 and SNU216 cells were more sensitive to pyrotinib (P < 0.01) (Fig. 1d), whereas no significant difference was observed in the inhibitory activity of apatinib (Fig. 1e). The IC₅₀ values of pyrotinib and apatinib for these GC cells are presented in the Supplementary Table 1. To examine the combination effect of pyrotinib and apatinib in HER2-positive GC, NCI-N87 and SNU216 cells were exposed to the two inhibitors, individually or in combination. The combination therapy significantly decreased the cell viability (P < 0.001), cell proliferation (P < 0.01) and invasion (P < 0.05), and increased cell apoptosis rate (P < 0.01) (Fig. 1f–i). To further investigate the associated mechanisms, the activity of the downstream pathways was evaluated. The EGFR/HER2 pathway was downregulated, in a dose-dependent manner, in NCI-N87 and SNU216 cells, on different pyrotinib treatments (Fig. 1j). The combination therapy downregulated the PI3K/AKT and ERK signaling pathways, and up-regulated levels of apoptosis-associated proteins in HER2-positive GC cell lines (Fig. 1k, l).

To evaluate the inhibitory effect of the combination treatment in vivo, the mice were inoculated with NCI-N87 or SNU216 cells and randomized into four groups treated daily with 10 mg/kg pyrotinib, 75 mg/kg apatinib, the combination of both inhibitors and saline as control (Fig. 2a). The combination group exhibited much stronger inhibitory effects in NCI-N87 (P < 0.05) and SNU216 (P < 0.01) xenografts compared to monotherapies (Fig. 2b–d, f–h). No significant difference in weight change was observed between the four groups (Fig. 2e, i). Tissues from the NCI-N87 xenografts were used for immunohistochemistry (IHC) analysis. Decreased Ki67 expression and increased cell apoptosis were observed in the combination group. No significant difference was observed in the expression of CD31 between the apatinib and the combination group (Fig. 2j). To examine the drug-specific interactions between pyrotinib and apatinib in HER2-positive GC, we assessed the tumor-suppressive efficacy of the inhibitors alone or combined in NCI-N87 xenografts (Fig. 2k–n). The ED₅₀ of apatinib and pyrotinib was revealed to be 31.19 mg/kg and 3.95 mg/kg, respectively, and the combination group exhibited an ED₅₀ of 4.54 mg/kg (pyrotinib: 0.54 mg/kg; and apatinib: 4.0 mg/kg). The CI was then estimated at 0.26, which indicated synergy between the two inhibitors (Table 1).

To investigate the resistant mechanisms of pyrotinib in HER2-positive GC, a pyrotinib-resistant cell line, NCI-N87-AR, was constructed. The IC₅₀ of pyrotinib in NCI-N87-AR cells was 4.15 ± 0.28 μΜ as compared with 0.08 ± 0.01 μΜ in NCI-N87 cells (P < 0.01) (Fig. 3a), and the colony number in NCI-N87-AR cells was significantly higher than that in NCI-N87 cells after pyrotinib treatment (P < 0.001) (Fig. 3b). Then, the RNA sequencing of the cells was performed. Compared to NCI-N87 cells, 736 genes were up-regulated and 795 genes were downregulated in NCI-N87-AR cells (P < 0.05 and > twofold change). The drug resistance-related genes KITLG and PIK3R1 were among the most up-regulated genes, as shown in the heat map (Fig. 3c). The most enriched pathways revealed by the KEGG analysis were the PI3K/AKT and MAPK signaling pathways (Fig. 3d, e). The relative expressions of the genes were then verified by qRT-PCR (Fig. 3f). Comparing the expression of the KITLG and PIK3RI encoded protein (SCF and p85α subunit) and the related pathways, we found that the expression of SCF/c-kit and downstream PI3K/AKT and ERK pathways were significantly up-regulated in NCI-N87-AR cells relative to NCI-N87 cells (Fig. 3g). Furthermore, these pathways were up-regulated, in a dose-dependent manner, in NCI-N87-AR cells after different concentrations of pyrotinib treatment (Fig. 3h). No significant difference was found in the expression of HER2, EGFR, VEGFR2 (including their phosphorylation status) and PTEN levels between the two cell lines (Supplementary Fig. 1). To further determine whether the up-regulation of SCF played a role in pyrotinib resistance, we assessed the effects of SCF in NCI-N87 and NCI-N87-AR cells. As shown in Fig. 3i–k, the proliferation of both cells was significantly increased by SCF (P < 0.001 and P < 0.05), which neutralized pyrotinib-specific effects, with this finding subsequently confirmed by colony formation assays (Fig. 3l, m). Furthermore, on exposure to varying concentrations of SCF, with or without the addition of pyrotinib, the expression of SCF/c-kit, and downstream PI3K/AKT and MAPK pathways, were up-regulated in a dose-dependent manner (Fig. 3n, o).

To verify the involvement of SCF/c-kit signaling pathways in mediating acquired pyrotinib resistance, we evaluated imatinib, a c-kit inhibitor, in acquired pyrotinib resistance in HER2-positive models both in vitro and in vivo [17, 18]. As shown in Fig. 4a, imatinib significantly inhibited SCF-induced c-kit phosphorylation and downstream molecules in the PI3K/AKT and ERK pathways in a dose-dependent manner. We then investigated the inhibitory effects of imatinib and pyrotinib, alone or in combination, in NCI-N87-AR cells. We found that combined therapy significantly decreased cell viability and proliferation, increased cell apoptosis, and downregulated PI3K/AKT and ERK signaling in NCI-N87-AR cells (Fig. 4b–e). To investigate whether combined therapy improves acquired pyrotinib resistance in vivo, mice were inoculated with NCI-N87-AR cells and randomized into four groups treated daily with 10 mg/kg pyrotinib, 100 mg/kg imatinib, the combination of both inhibitors, or saline as a control (Fig. 4f). The group receiving combined therapy exhibited stronger antitumor effects in NCI-N87-AR xenografts relative to those observed from monotherapy (Fig. 4g–i), with no significant difference in weight change observed between the four groups (Fig. 4j). IHC analysis of tissues from xenografts revealed similar SCF levels in the four groups, although decreased Ki67 levels and increased cell apoptosis were observed in the combined-treatment group, and no significant difference was observed in the CD31 levels between groups (Fig. 4k).

To investigate whether apatinib played a role in combating SCF-mediated pyrotinib resistance, the inhibitory effects of apatinib were examined in NCI-N87-AR cells. Apatinib significantly inhibited SCF-induced c-kit phosphorylation and the expression of downstream PI3K/AKT and ERK signaling in a dose-dependent manner (Fig. 5a, b), and enhanced the sensitivity of NCI-N87-AR cells to pyrotinib (Fig. 5c, d). Subsequently, we investigated the inhibitory effects of apatinib and pyrotinib, individually or in combination, in NCI-N87-AR cells, revealing that combined therapy significantly decreased the cell viability, proliferation and invasion, increased cell apoptosis (P < 0.01), as well as levels of apoptosis-associated proteins, and downregulated the PI3K/AKT and ERK signaling in NCI-N87-AR cells (Fig. 5e–j). The similar effects of apatinib also appeared in NCI-N87 and SNU216 cells (Supplementary Fig. 2, Fig. 1k).

To investigate whether apatinib combined with pyrotinib improves acquired pyrotinib resistance in vivo, the mice were inoculated with NCI-N87-AR cells and randomized into four groups treated daily with 10 mg/kg pyrotinib, 75 mg/kg apatinib, the combination of both inhibitors, or saline as a control (Fig. 6a). The combination group exhibited enhanced antitumor effects in NCI-N87-AR xenografts relative to those observed from monotherapies (P < 0.05), with no significant difference in weight change observed between the four groups (Fig. 6b–e). IHC analysis of tissues from xenografts revealed similar SCF levels in the four groups, although decreased Ki67 levels and increased cell apoptosis were observed in the combined-treatment group, with no significant difference observed in CD31 levels between the apatinib and combined-treatment groups (Fig. 6f).