Cholesterol confers resistance to Apatinib-mediated ferroptosis in gastric cancer.

Background
Gastric cancer (GC) is the fifth most common cancer worldwide and ranks as the fifth leading cause of cancer-related deaths globally [1]. Early-stage gastric cancer is often asymptomatic, consequently, most GC patients are diagnosed at an advanced stage, leading to poor prognoses [2]. Despite advancements in standard-of-care treatments, the high heterogeneity and malignancy of GC contribute to acquired resistance during therapy, resulting in limited survival benefits [3, 4]. Therefore, it is imperative to investigate strategies to overcome treatment resistance to improving patient outcomes.
Apatinib, a VEGFR2-targeting anti-angiogenic drug, has been approved for third-line treatment of advanced gastric cancer [5–7]. It has been shown to enhance the efficacy of chemotherapy and immune checkpoint inhibitors in several Phase I/II studies as first-line treatment for advanced GC [8–12]. Meanwhile, increasing evidence suggests that Apatinib also exerts direct anti-tumor effects by inducing various forms of tumor cell death, such as apoptosis [13–15], autophagy [16, 17], necrosis [18], and ferroptosis [19–21]. Even though it can effectively improve the prognosis of patients with advanced GC, acquired resistance still poses a significant obstacle in its clinical practice [22]. However, safe and effective strategies to overcome Apatinib resistance remain elusive. Therefore, it is essential to explore a novel and effective strategy to overcome Apatinib resistance.
Recent works have established that cholesterol metabolism is profoundly dysregulated in many tumors and plays multifaceted roles in cancer progression and therapy [23]. Tumor cells often upregulate cholesterol uptake and de novo synthesis to support membrane biogenesis and sustain oncogenic signaling [24]. Cholesterol and its metabolites can directly activate pro-tumor signaling pathways, such as PI3K/AKT, Sonic hedgehog and AMPK, that enhance proliferation, stemness and metastasis [25–27]. Elevated cellular or mitochondrial cholesterol also promotes drug resistance by inhibiting apoptosis and altering membrane transporter function [28]. These mechanistic insights have spurred interest in targeting cholesterol metabolism for cancer therapy. Statins, inhibitors targeting 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGCR), can inhibit tumor cell proliferation and metastasis by inducing apoptosis or ferroptosis [29–31]. It has been shown that statins could synergize with standard therapies to prolong survival in colorectal, pancreatic and other cancer patients [32–34]​. However, whether statins affect Apatinib resistance in GC remains unknown.
In this study, we found that the cholesterol biosynthesis pathway was upregulated in Apatinib-resistant GC cells, leading to increased intracellular cholesterol content and resistance to Apatinib-mediated ferroptosis. HMGCR, the rate-limiting enzyme in this pathway, was significantly overexpressed in Apatinib-resistant GC cells. Both in vitro and in vivo inhibition of HMGCR via Simvastatin treatment or genetic knockdown enhanced Apatinib-induced ferroptosis in Apatinib-resistant GC cells. Mechanistically, the expression of ETS Homologous Factor (EHF) was elevated in Apatinib-resistant GC cells. As a transcription factor, EHF promoted HMGCR transcription, which in turn increased cholesterol synthesis and lipid raft levels, ultimately protecting GC cells from Apatinib-mediated ferroptosis. Herein, we demonstrated that Simvastatin could enhance Apatinib-mediated ferroptosis, suggesting that this combination might serve as a promising and safe strategy to overcome Apatinib resistance in GC.

Method

Cell culture
Human gastric adenocarcinoma cell lines, MGC803 and AGS, were obtained from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). The cells were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM; C11995500BT, Gibco) supplemented with 10% fetal bovine serum (FBS; 164210-50, Procell) and 1% penicillin-streptomycin (ATPS-10001-100, OriCell). All cultures were incubated in a 37 °C humidified atmosphere with 5% CO₂. Apatinib-resistant cell lines (MGC803/AR and AGS/AR) were developed from their respective parental lines (MGC803/WT and AGS/WT) through exposure to gradually increasing concentrations of Apatinib (HY-13342, MedChemExpress). To maintain the Apatinib-resistant phenotype, MGC803/AR and AGS/AR cells were continuously cultured in medium containing 50 µM Apatinib. For experimental purposes, Apatinib-resistant cells were cultured without Apatinib for seven days prior to experiment.

RNA- sequencing
RNA was purified from total RNA using polyT to isolate mRNA, which was then fragmented into 300–350 bp pieces. First-strand cDNA was synthesized using the fragmented RNA and dNTPs (dATP, dTTP, dCTP, and dGTP), followed by second-strand synthesis. The overhangs of the double-stranded cDNA were converted into blunt ends using exonuclease/polymerase activities. After adenylation of the 3’ ends, sequencing adapters were ligated to the cDNA, and the library was purified. The library was then enriched by PCR, and the final product was purified. Differential expression analysis was performed using EdgeR, and P-values were adjusted with the Benjamini-Hochberg method to control the false discovery rate.

GSEA analysis
The GMT files of REACTOME (version 2023.2) and WikiPathway (version 2023.2) were downloaded from the Gene Set Enrichment Analysis (GSEA) website(https://www.gsea-msigdb.org/gsea/index.jsp). Subsequently, GSEA analysis of the transcriptome between AGS/AR cells and AGS/WT cells was carried out using the R platform (version 4.3.2). The data analysis and processing involved several R packages, including “limma” (version 3.60.3), “org.Hs.eg.db” (version 3.19.1), “DOSE” (version 3.30.1), and “clusterProfiler” (version 4.12.0). Visualizations were created with the “enrichplot” package (version 1.24.0).

Quantitative real-time polymerase chain reaction (qRT‒PCR)
Total RNA was extracted from cells with TRIzol (15596026CN, Invitrogen). Subsequently, cDNA was synthesized using the SweScript RT II First Strand cDNA Synthesis Kit (With gDNA Remover) (G3333, Servicebio). Then, qPCR was conducted with the 2×Universal Blue SYBR Green qPCR Master Mix (G3326, Servicebio) on the 7500 Real-Time PCR System Platform (Thermo Fisher, America). The relative gene expression was calculated by the 2 − ΔΔCt method, with GAPDH serving as the internal control. All primers were synthesized by Tsingke Biotech (Beijing, China) and are presented in Supplementary Table 1.

Intracellular cholesterol measurement
To measure the intracellular cholesterol contents, cells were fixed with 4% paraformaldehyde (PFA) for 15 min and then washed three times with phosphate-buffered saline (PBS). Subsequently, the staining sections were washed three times with PBS and incubated with 1.5 mg/mL glycine (ST085, Beyotime) for 10 min at room temperature. Next, the sections were stained with 0.05 mg/mL Filipin complex (HY-N6716, MedChemExpress) in PBS containing 10% fetal bovine serum (FBS) for 2 h and then rinsed three times with PBS. After the incubation, the staining was detected using the EVOS M5000 (Life, America). In each experiment, all images were acquired with identical laser outputs, gains, and offsets. The intracellular cholesterol contents were further verified using a Tissue Total Cholesterol Assay Kit (E1015, Applygen) following the manufacturer’s instructions. The data were normalized to the total protein level in each sample.

Transwell assay
Cells were suspended in serum-free DMEM and seeded into the inserts of transwell chambers (5 × 104 cells per insert) for 24 h incubation. The lower chambers were filled with DMEM supplemented with 10% FBS. For the invasion assay, the inserts were pre-coated with Matrigel (356237, BD Biosciences). The cells that had invaded were stained with crystal violet. Statistical analysis was performed using ImageJ software.

Cell viability assay
Cells were plated in 96-well plates (3 × 103 cells/well) and treated with or without combinations of Apatinib (1-100 µM), Cholesterol (Water Soluble) (5 mg/ml, HY-N0322A, MedChemExpress), 1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (10 µM, HY-130462, MedChemExpress), Methyl-beta-cyclodextrin (MβCD) (3 mM, HY-101461, MedChemExpress), Simvastatin (2–5µM, HY-17502, MedChemExpress), RSL3 (0.5–32 µM, HY-100218 A, MedChemExpress), Oxaliplatin(1-100 µM, HY-17371, MedChemExpress), 5-Fluorouracil(5-FU) (1-160 µM, HY-90006, MedChemExpress) for 24, 48–72 h. At the indicated time points, cell viability was examined by using the CCK8 kit (GK10001, Glpbio).

Colony-formation assay
Cells were plated in 6-well plates (1200 cells/well) and treated with or without combinations of Apatinib (20 µM), Cholesterol (5 mg/ml), Simvastatin (3 µM) and RSL3 (2 µM) for 48 h and MβCD (3 mM) for 1 h on the following day, after which the medium was replaced with fresh medium and continued to culture for 5 to 8 days. When the control group was observed that most of the colonies contain more than 50 cell clones, the colonies were fixed with 4% paraformaldehyde and stained with crystal violet and photographed by EVOS M5000 (Life, America). ImageJ was used to perform the statistical analysis of colonies.

Lipid ROS and intracellular ROS assay
For Flow cytometry analysis, lipid peroxidation was measured by incubating cells in prewarmed (37 °C) 10 µM BODIPY™ 581/591 C11 (D3861, Invitrogen) and intracellular ROS was measured by incubating in 10 µM 2’,7’-Dichlorodihydrofluorescein diacetate (H2DCFDA) (HY-D0940, MedChemExpress) for 30 min at 37 °C. Then cells were washed three times with PBS, collected to analyzed by Attune NxT Flow Cytometer (Invitrogen, CA, USA) and used a confocal laser scanning microscopy (Nikon A1 LFOV, Nikon, Japan) for image.

GSH, MDA, ferrous iron and lipid raft assay
We used an reduced Glutathione(GSH) Content Assay Kit (BC1175, Solarbio) to detect GSH content, Malondialdehyde (MDA) Colorimetric Assay Kit (E-BC-K028-M, Elabscience) to detect MDA content, Cell Ferrous Iron Colorimetric Assay Kit (E-BC-K881-M, Elabscience) to detect ferrous iron content, and Human Flotillin-1(FLOT1) ELISA kit (CSB-EL008727HU, CUSABIO) to detect lipid raft levels, according to the manufacturer’s instruction. Determinations were normalized with protein content.

Gene knockdown using SiRNA
Small interfering (si) RNA were obtained from GeneCreate Biotech (Wuhan, China) and are presented in Supplementary Table 2. Cells were transfected with 50 nM siRNA or scrambled negative control (si-NC) siRNA using Lipo8000™ Transfection Reagent (C0533FT, Beyotime) and incubated for 48 h for further experiments.

Construction of stably knockdown/overexpression cell lines
To produce lentivirus, 293T cells were transfected with pGLVU6 along with the packaging vectors pMD2.G and psPAX2. The virus-containing medium was harvested 48 hours after transfection. Subsequently, the viral supernatant was utilized to infect the target cells. Stable cell lines were selected by treating with 4 µg/mL puromycin 48 hours post-transfection. We generated lentiviruses harboring control-shRNA (shNC) and HMGCR-shRNA to establish stable knockdown cell lines. The shRNA hairpin sequence of HMGCR-shRNA1 was 5’-GGTTCTAAAGGACTAACATAA-3’, and that of HMGCR-shRNA2 was 5’-CTATGATTGAGGTCAACATTA-3’. The LV5-control and LV5-HMGCR lentiviruses were procured to establish overexpression cell lines. The efficiency was evaluated by western blot and qPCR.

ChIP assay
Cells were cross-linked with 1% formaldehyde and then quenched in 125 mM glycine. Chromatin was isolated by using a BeyoChIP™ Enzymatic Chromatin Immunoprecipitation (ChIP) Assay Kit with Protein A/G Magnetic Beads (P2083S, Beyotime). Subsequently, it was sonicated and catalyzed by MNase to a length ranging from 200 to 400 bp. The chromatin was then subjected to immunoprecipitation (IP) with either EHF (27195-1-AP, Proteintech) or IgG (30000-0-AP, Proteintech) antibody. Quantitative reverse transcription-polymerase chain reaction (qRT‒PCR) was carried out to quantify the immunoprecipitated genomic DNA regions. The primers used are presented in Supplementary Table 1.

Promoter luciferase assay
AGS/AR cell lines were transfected with the Lipo8000™ transfection reagent in a 24-well plate. The transfection was carried out using the EHF overexpression plasmid, with the pcDNA3.1 plasmid serving as a control, and the process lasted for 24 h. Subsequently, the AGS/AR cell lines were transfected again with the HMGCR promoter-reporter sequences or vector pEZX-FR01 plasmid (ZX001, GeneCopoeia), which was used as a control for transfection efficiency, also using the Lipo8000™ transfection reagent. The luciferase assay was then performed by employing the dual-luciferase reporter assay kit (LF004, GeneCopoeia) in accordance with the manufacturer’s protocol.

Western blotting
Equal amounts of protein samples were resolved via sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene difluoride (PVDF) membranes. Membranes were blocked with 5% nonfat dry milk in TBST for 1 h, then incubated with primary antibodies overnight at 4 °C. The details of the primary antibodies used were as follows: HMGCR at a dilution of 1:1000 (A16875, ABclonal), EHF at a dilution of 1:250 (27195-1-AP, Proteintech), SREBP2 at a dilution of 1:1000 (YP-Ab-04920, UpingBio) and β-actin at a dilution of 1:10000 (20536-1-AP, Proteintech). Subsequently, the membranes were washed thoroughly three times with TBST and then incubated with HRP-conjugated secondary antibodies for 1 h at room temperature. The secondary antibody employed was Goat anti-Rabbit IgG-HRP at a dilution of 1:5000 (M21002, Abmart). Finally, the signals were detected using the ECL Chemiluminescent Substrate (BL523A, Biosharp) and analyzed by the GeneGnome XRQ system (SYNGENE, Cambridge, UK). The band grayscale values were analyzed using ImageJ.

In vivo experiment
Five-week-old male immunodeficient BALB/c nude mice were obtained from Guangdong Zhiyuan Biology (Guangzhou, China). All procedures involving mice and their care were performed in accordance with protocols approved by the Animal Ethics Committee of Nanfang Hospital (application No: IACUC-LAC-20240808-006). In the first experiment, AGS/AR or AGS/WT cells (at a concentration of 5 × 10⁶ cells in 100 µL of a PBS: Matrigel = 1:1 mixture per mouse) were subcutaneously implanted into the right flanks of the mice. In the second experiment, MGC803/AR cells were implanted using the same procedure. When tumor volume reached approximately 60 mm³, the mice were randomly assigned to groups (n = 5 per group). The first exprement: DMSO (control); Apatinib (oral administration, 50 mg/kg daily). The second experiment: DMSO (control); Simvastatin (intraperitoneal injection, 10 mg/kg every 3 days); Apatinib (oral administration, 50 mg/kg daily); Simvastatin (10 mg/kg intraperitoneal injection every 3 days) + Apatinib (50 mg/kg oral daily), as shown in Fig. 7A. In all experiments, The body weights of the mice were measured every 3 days. Throughout all experiments, mouse body weights and tumor sizes were measured every 3 days using digital calipers. Tumor volume was calculated with the formula: Volume = (width2 × length)/2. The treatment period lasted 15 days. At the end of the experiments, mice were euthanized, and tumors, livers, lungs, hearts, spleens, and kidneys were harvested for subsequent analyses.

H&E staining
Briefly, immerse the samples in Xylene I and Xylene II for 10 min each, followed by immersion in absolute ethanol I and absolute ethanol II for 5 min each. Then, soak in 95%, 80%, and 60% ethanol for 3 min each. Rinse thoroughly with distilled water. Submerge the samples in hematoxylin solution for 10 min, then rinse with distilled water. Perform differentiation using 1% hydrochloric acid alcohol solution, followed by another rinse with distilled water. Stain with eosin solution for 10 min, and rinse with distilled water to remove residual stain. After dehydration steps, mount the samples using neutral resin and photograph them by EVOS M5000 (Life, America).

Ki67 staining
Deparaffinize paraffin sections following the dewaxing procedure for HE staining. Perform antigen retrieval using EDTA antigen retrieval solution (pH 9.0). Immerse the sections in 3% hydrogen peroxide solution, protected from light, for 15 min to eliminate endogenous peroxidase activity. Block with 3% BSA, then incubate with anti-Ki67 antibody (27309-1-AP, Proteintech) at room temperature for 2 h. Wash the sections three times with PBS, then incubate with the secondary antibody (PV-6001, zsgbbio) at room temperature for 30 min. Wash again four times with PBS, followed by the addition of Diaminobenzidine(DAB) (PR30010, Proteintech) substrate for chromogenic development. Counterstain with Mayer’s hematoxylin and perform bluing. Finally, dehydrate and mount the sections and photograph by EVOS M5000 (Life, America).

Statistics
Statistical analyses were carried out using GraphPad Prism 10.0 software. All data are presented in the form of the mean ± standard error of the mean. The significance of differences between groups was analyzed by means of Student’s t tests, one-way analysis of variance (ANOVA), or two-way ANOVA. A p < 0.05 was regarded as indicative of a statistically significant difference. The levels of significance are presented as *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. The experimental reports in this study were reproducibly and reliably obtained in at least three independent replicates.

Result

Increasing cholesterol biosynthesis conferred apatinib resistance
In this study, we established Apatinib-resistant gastric cancer cell lines, MGC803/AR and AGS/AR, derived from their respective parental cell lines, MGC803/WT and AGS/WT (Fig. 1A). Compared to their parental counterparts, MGC803/AR and AGS/AR exhibited enhanced colony formation ability when treated with Apatinib (Fig. 1B). To investigate Apatinib resistance in vivo, we established xenograft models using AGS/AR and AGS/WT cells in BALB/c nude mice (Fig. 1C). Although statistical significance was not reached, AGS/AR-derived tumors showed a trend toward faster growth and greater weight compared with AGS/WT-derived tumors. Furthermore, AGS/AR demonstrated greater resistance to Apatinib than AGS/WT in vivo (Fig. 1D-E).

To investigate the molecular mechanisms underlying resistance, we performed transcriptomic analysis (RNA-seq) to compare gene expression profiles between the Apatinib-resistant and wild-type cells (AGS/AR VS AGS/WT) (Figure S1A and B). Gene sets from WikiPathways and Reactome databases were retrieved from the GSEA website and analyzed using Gene Set Enrichment Analysis (GSEA) on the R platform. Interestingly, our analysis revealed significant enrichment of the cholesterol biosynthesis pathway in the Apatinib-resistant cells (Fig. 2A) while other lipid metabolism pathways didn’t have significant changes (Figure S2A). Moreover, genes involved in cholesterol biosynthesis were upregulated in the resistant cells (Fig. 2B), as confirmed by qPCR (Fig. 2C). To quantify intracellular cholesterol levels, we employed Filipin staining, a fluorescent compound that specifically binds cholesterol, serving as a reliable marker for its detection and quantification in cells [35]. Filipin staining demonstrated elevated cholesterol levels in MGC803/AR and AGS/AR cells compared to their parental counterparts (Fig. 2D). Furthermore, using a Cholesterol Assay Kit, we confirmed the increased cholesterol levels in the resistant cells (Fig. 2E).

To determine whether elevated cholesterol levels contributes to Apatinib resistance in GC cells, we modulated intracellular cholesterol levels by Cholesterol supplementation or Methyl-beta-cyclodextrin (MβCD), a compound known to acutely deplete cellular cholesterol (Figure S3A and B) [36]. Cholesterol supplementation increased resistance to Apatinib in MGC803/WT and AGS/WT cells, with a marked effect observed at the 48-hour time point (Fig. 3A). Cholesterol, rather than 1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), a major lipid component of cell membranes [37], could rescue MGC803/WT and AGS/WT from Apatinib treatment (Fig. 3B). Cholesterol supplementation also attenuated the inhibitory effect of Apatinib on the ability of colony formation (Fig. 3C). Conversely, pretreatment with MβCD enhanced the cytotoxic efficacy of Apatinib in MGC803/AR and AGS/AR cells across various time points (Fig. 3D and E). Notably, acute cholesterol depletion by MβCD enhanced the inhibitory effect of Apatinib on the ability of colony formation (Fig. 3F). Knocking down genes in the cholesterol synthesis pathway using siRNA can resensitized MGC803/AR and AGS/AR to Apatinib to varying degrees (Figure S4A and B). These findings highlighted the critical role of cholesterol in mediating resistance to Apatinib in GC cells.

Targeting HMGCR could overcome apatinib resistance
As shown in Figs. 2C and -hydroxy-3-methylglutaryl-CoA reductase (HMGCR) was one of the most significantly upregulated genes in the cholesterol biosynthesis pathway in MGC803/AR and AGS/AR cells, compared to wild-type cells, and was further confirmed by Western blot (Fig. 4A and Figure S5A). To investigate whether HMGCR plays a pivotal role in Apatinib resistance, we generated HMGCR knockdown Apatinib-resistant cell lines (MGC803/AR-shHMGCR1/2 and AGS/AR-shHMGCR1/2) (Figure S5B and C). As expected, HMGCR knockdown resulted in a reduction in intracellular cholesterol levels (Figure S5D and E) and a decrease in cell proliferation (Fig. 4B). Notably, HMGCR knockdown extremely abolished the resistance of GC cells to Apatinib (Fig. 4C and Figure S6A-B). This effect was rescued by cholesterol supplementation (Figure S7A and B). Colony formation ability of MGC803/AR and AGS/AR cells was significantly impaired after HMGCR knockdown and treatment with Apatinib, which could be restored by cholesterol supplementation (Fig. 4D and Figure S7C). Moreover, HMGCR knockdown markedly reduced their abilities of migration and invasion (Fig. 4E and Figure S7D). These findings underscored the crucial role of HMGCR in mediating cholesterol biosynthesis and Apatinib resistance, highlighting it as a potential therapeutic target in Apatinib-resistant GC cells.

Given the critical role of HMGCR in Apatinib resistance, we overexpressed the HMGCR gene in MGC803/WT and AGS/WT cells (MGC803/WT-HMGCROE and AGS/WT-HMGCROE) (Figure S8A and B). HMGCR overexpression resulted in an increase in intracellular cholesterol (Figure S8C and D) and an enhancement of proliferative capacity (Figure S8E). HMGCR overexpression also resulted in resistance to Apatinib (Fig. 4F), enhanced colony formation (Figure S8F and G), and stimulated both migration and invasion (Figure S8H). These findings further indicated that HMGCR promoted GC cells proliferation, migration, and invasion, as well as resistance to Apatinib.

Simvastatin, an FDA-approved drug that targets HMGCR, has been shown to inhibit tumor proliferation (Fig. 5A) and enhance therapeutic efficacy [30, 38]. Even at a low concentration of 5 µM, Simvastatin effectively reduced intracellular cholesterol levels (Fig. 5B and C) and resensitized MGC803/AR and AGS/AR cells to Apatinib (Fig. 5D). This effect could be abolished by co-treatment with cholesterol (Fig. 5E and F). These results suggested that Simvastatin may serve as a promising therapeutic strategy to overcome Apatinib resistance in Apatinib-resistant GC cells by targeting cholesterol biosynthesis.

Simvastatin inhibited ferroptosis induced by apatinib
Our previous work has demonstrated that Apatinib can induce ferroptosis in GC cells [19]. Compared to wild-type GC cells, MGC803/AR and AGS/AR cells showed greater resistance to RSL3, a known ferroptosis inducer (Figure S9A). This led us to hypothesize that cholesterol confers Apatinib resistance by protecting GC cells from Apatinib-mediated ferroptosis. To our expectation, treatment with Simvastatin or HMGCR knockdown significantly re-sensitized MGC803/AR and AGS/AR cells to RSL3, which could be rescued by cholesterol supplement (Fig. 6A, Figures S9B and C, Figures S10A and B). Additionally, HMGCR overexpression enhanced the viability and proliferation of GC cells under RSL3 treatment (Figures S9D and E). Cholesterol has been identified as a protector against ferroptosis by promoting the formation of lipid rafts [39]. Consistently, Apatinib-resistant GC cells exhibited a higher levels of lipid rafts compared to wild-type cells, which could be suppressed by Simvastatin treatment (Figure S10C). However, the addition of cholesterol did not change the sensitivity of GC cells to Oxaliplatin and 5-FU which inhibits tumor mainly not by ferroptosis [40] (Figures S10D and E).
HMGCR inhibition significantly elevated Apatinib-induced lipid peroxidation levels, a key marker of ferroptosis. Conversely, HMGCR overexpression or cholesterol pretreatment resulted in a reduction of lipid peroxidation (Fig. 6B and Figure S11A). Additionally, HMGCT Inhibition further augmented Apatinib-mediated increases in intracellular ferrous iron and malondialdehyde (MDA) while concurrently decreasing intracellular glutathione (GSH) concentrations. These effects could be rescued by cholesterol supplementation (Figs. 6C-E and Figure S11B-D). Using flow cytometry, we observed that co-treatment with Simvastatin and Apatinib generated higher levels of lipid ROS and intracellular ROS, which could be rescued by cholesterol supplementation or Ferrostatin-1, a ferroptosis inhibitor (Figs. 6F and G). These results indicated that cholesterol promoted lipid raft formation and protected GC cells from Apatinib-induced ferroptosis, and targeting HMGCR with Simvastatin sensitized Apatinib-resistant cells to ferroptosis by disrupting cholesterol-mediated defenses.

EHF upregulated HMGCR by binding its promotor
Given the critical role of HMGCR in Apatinib-resistant gastric cancer cells, we investigated the factors contributing to its increased transcription. Nevertheless, the expression levels of SREBP2, an well-known transcriptional regulator of HMGCR, did not exhibit significant changes in Apatinib-resistant cells (Fig. 2C and Figure S12A). Therefore, we analyzed the promoter sequence of HMGCR using UCSC Genome Browser (https://genome.ucsc.edu/) and JASPAR (https://jaspar2022.genereg.net/docs/), which identified ETS homologous factor (EHF) as a potential transcription factor regulating HMGCR expression in Apatinib-resistant cells. Further analysis using the GEPIA database (http://gepia2.cancer-pku.cn/#index) revealed a positive correlation between EHF and HMGCR expression (Fig. 7A). Western blotting confirmed that EHF expression was significantly elevated in MGC803/AR and AGS/AR cells (Fig. 7B). To further assess the role of EHF, we used siRNA to knock down EHF expression in MGC803/AR and AGS/AR cells. This led to a significant downregulation of HMGCR expression (Fig. 7C). Importantly, EHF knockdown re-sensitized Apatinib-resistant cells to Apatinib, which could be rescued by cholesterol supplementation (Figs. 7D and E). Additionally, silencing EHF also abolished resistance to the ferroptosis inducer RSL3 (Fig. 7F).

To further elucidate how the transcription factor EHF regulates HMGCR expression, we constructed dual-luciferase reporter plasmids containing different lengths of the HMGCR promoter sequence. The dual-luciferase assay revealed that EHF binds to the HMGCR promoter sequence and promotes its transcription, with the binding site located between positions − 1953 bp and − 1084 bp (Fig. 7G). Analysis using the JASPAR database predicted that the EHF-recognized binding sequence is (C/A)CTTCCT(G/C), located between − 1415 bp and − 1407 bp (Fig. 7H). To confirm the importance of this binding site, we generated a reporter plasmid with a mutation in the predicted binding sequence (Fig. 7I). The mutation significantly reduced the binding efficiency of EHF, as shown by the luciferase assay (Fig. 7J). Chromatin immunoprecipitation (ChIP) further validated that EHF binds to this specific region of the HMGCR promoter (Fig. 7K). These findings suggested that EHF acted as a key transcriptional regulator of HMGCR in Apatinib-resistant cells, contributing to their resistance to Apatinib and ferroptosis.

Simvastatin promoted the anti-tumor ability of apatinib in vivo
To evaluate the effect of Simvastatin in overcoming Apatinib resistance in vivo, we established xenografts with MGC803/AR cells in BALB/c nude mice. After subcutaneous injection of tumor cells, the mice were randomly divided into four groups: Control, Simvastatin(SIM), Apatinib(AP), and Simvastatin + Apatinib(SIM + AP) (Fig. 8A). After treatment, the mice were euthanized and tumor tissues were collected for further analysis (Figs. 8B and C). Simvastatin alone had no significant effect on tumor volume or weight compared to the control group. However, co-administration of Simvastatin and Apatinib significantly reduced both tumor volume and weight compared to Apatinib treatment alone (Figs. 8D and E). Importantly, the combination treatment of oral Apatinib and intraperitoneal Simvastatin injection did not result in significant changes in mice body weight, indicating good tolerability. (Fig. 8F). Immunohistochemical staining for Ki67 demonstrated that the combination of Simvastatin and Apatinib inhibited tumor cell proliferation more effectively compared to either treatment alone (Fig. 8G). Furthermore, Hematoxylin and eosin (H&E) staining revealed no significant pathological damage to major organs, including the heart, spleen, lungs, and kidneys, across all treatment and control groups (Fig. 8H). Collectively, these findings suggest that Simvastatin represents a safe and effective in vivo strategy to overcome Apatinib resistance, highlighting its potential as a promising combination therapy for gastric cancer.

Discussion
Given Apatinib’s role as a third-line therapy for gastric cancer, the development of resistance leaves patients with few remaining treatment options. However, the underlying mechanisms of resistance have not been thoroughly studied [22]. In this study, we established Apatinib-resistant GC cell lines (MGC803/AR and AGS/AR). We found that active cholesterol biosynthesis conferred resistance to Apatinib which could be reversed by depleting cholesterol with MβCD (Fig. 9). In recent years, accumulating evidence has indicated that cellular lipid metabolism, particularly cholesterol homeostasis, plays a critical role in the development and maintenance of drug resistance in tumor cells [41, 42]. Elevated intracellular cholesterol levels can confer resistance to 5-FU, Gefitinib, Osimertinib, irradiation and Sorafenib by activating PI3K-AKT, EGFR/Src/Erk and Sonic hedgehog signaling pathway in cancer cells [42–46]. Similarly, our study uncovered a mechanistic link between cholesterol metabolism and Apatinib resistance.
In this study, we found that HMGCR knockdown in MGC803/AR and AGS/AR cells significantly reduced intracellular cholesterol levels and promoted Apatinib-mediated ferroptosis. Furthermore, cholesterol supplementation could effectively inhibit ferroptosis, but had a limited effect on rescuing cell viability when treated with 5-FU and oxaliplatin, which primarily induce apoptosis. However, the mechanism connecting cholesterol and ferroptosis are likely intricate [47–49]. Previous studies exploring the mechanism by which cholesterol regulates ferroptosis have shown that cholesterol does not function directly as an antioxidant against lipid peroxidation [50, 51]. Rather, its protective role against ferroptosis may be attributed to its role in regulating the diffusion dynamics of lipid ROS within the cell membrane [39, 52, 53]. Consistently, our study showed that lipid raft content was increased in Apatinib-resistant cells, which could be inhibited by Simvastatin. While our study highlights the crucial role of cholesterol-mediated resistance to ferroptosis, whether cholesterol regulates Apatinib resistance by influencing membrane ROS diffusion remains to be further investigated.
Previous studies have provided valuable insights into Apatinib resistance. For example, paeonol has been shown to overcome Apatinib resistance in gastric cancer by inducing apoptosis [54]. Another study has demonstrated that inhibiting glutamine uptake and catabolism enhances Apatinib’s anti-tumor efficacy in non-small cell lung cancer [55]. Elevated expression of hsa_circ_0003823 and YY1 transcription factor(YY1) has been linked to Apatinib resistance [21, 56]. However, these previous studies are insufficient to provide a safe and effective strategy for clinical application. In this study, we validated that Simvastatin, an FDA-approved HMGCR inhibitor, could resensitize Apatinib-resistant GC cells to Apatinib treatment without notable adverse effects. However, one limitation of our study is the lack of clinical gastric cancer samples to further validate the association between tumor cholesterol levels or simvastatin use and clinical response to Apatinib treatment.
Simvastatin, a specific inhibitor of HMGCR, has been approved by the FDA primarily for dyslipidemia treatment and cardiovascular disease prevention [57]. Recent studies have revealed that simvastatin can inhibit tumor progression and overcome chemoresistance by blocking cholesterol biosynthesis [29, 30, 58]. However, Previous research focused on the direct inhibitory effect of statins at concentrations typically exceeding 10µM. But long-term and high-dose statin administration carries a risk of significant adverse effects such as rhabdomyolysis [59]. In our study, even low-concentration simvastatin (5 µM for 48 h) resensitized resistant cells to Apatinib, despite having no significant impact on cell viability as monotherapy. The clinical antitumor effects of statins remains controversial. Contributing factors include not only the dose-dependent safety concerns but also limited clinical evidence of their cancer-suppressive effects. Our in vivo experiments in AGS/AR xenograft models showed that the combination of Simvastatin and Apatinib yielded superior therapeutic outcomes compared to monotherapy, without causing significant body weight loss or pathological changes in major organs. However, numerous studies indicate that whether statin are administered at a high dose (80 mg) or a low dose (40 mg) in combination with chemotherapy, no significant improvement in clinical prognosis has been observed compared to chemotherapy alone [60–64]. To address these gaps, future research should: (1) identify patient subgroups most likely to benefit from statin treatment, (2) leverage novel technologies (such as nanoparticles) to optimize statin pharmacokinetics, and (3) investigate the impact of statins on the tumor microenvironment [65–69].Therefore, more rigorously designed clinical trials are needed to explore the role of simvastatin in the treatment of tumors.
ETS homologous factor (EHF) is a transcription factor that plays a critical role in regulating cancer progression and chemoresistance. Several studies have showed that EHF expression is elevated in prostate, ovarian, gastric, and colorectal cancers, and its high expression level was associated with poor patient survival [70–73]. In this study, we found that in Apatinib-resistant gastric cancer (GC) cells, EHF promotes the transcription of HMGCR, while EHF inhibition resensitizes these cells to Apatinib-induced ferroptosis. Our findings further reveal that EHF regulates cholesterol biosynthesis, thereby providing a novel mechanism by which EHF mediates resistance to ferroptosis. However, this regulatory effect was validated only in Apatinib-resistant cells. Therefore, future clinical studies are needed to investigate the role of EHF in regulating cholesterol synthesis and ferroptosis in broader contexts.

Conclusion
In summary, our findings revealed that cholesterol biosynthesis conferred resistance to Apatinib-mediated ferroptosis in GC. Importantly, inhibition of HMGCR with Simvastatin resensitized Apatinib-resistant GC cells to Apatinib and exhibited an acceptable safety profile. Thus, Simvastatin may serve as a promising and safe strategy for overcoming Apatinib resistance in GC.

Electronic supplementary material
Below is the link to the electronic supplementary material.