Apatinib combined with olaparib induces ferroptosis via a p53-dependent manner in ovarian cancer

PARP inhibitors combined with antiangiogenic drugs have been reported to improve outcomes in BRCA wild-type ovarian cancer patients, the mechanism of the combination is unclear. In this study, we explored the mechanism of apatinib combined with olaparib in the treatment of ovarian cancer. In this study, human ovarian cancer cell lines A2780 and OVCAR3 were used as experimental objects, and the expression of ferroptosis-related protein GPX4 after treatment with apatinib and olaparib was detected by Western blot. The SuperPred database was used to predict the target of the combined action of apatinib and olaparib, and the predicted results were verified by Western blot experiment to explore the mechanism of ferroptosis induced by apatinib and olaparib. Apatinib combined with olaparib-induced ferroptosis in p53 wild-type cells, and p53 mutant cells developed drug resistance. The p53 activator RITA sensitized drug-resistant cells to ferroptosis induced by apatinib combined with olaparib. Apatinib combined with olaparib-induced ferroptosis via a p53-dependent manner in ovarian cancer. Further studies showed that apatinib combined with olaparib-induced ferroptosis by inhibiting the expression of Nrf2 and autophagy, thereby inhibiting the expression of GPX4. The Nrf2 activator RTA408 and the autophagy activator rapamycin rescued the combination drug-induced ferroptosis. This discovery revealed the specific mechanism of ferroptosis induced by apatinib combined with olaparib in p53 wild-type ovarian cancer cells and provided a theoretical basis for the clinical combined use of apatinib and olaparib in p53 wild-type ovarian cancer patients.


Introduction
Ovarian cancer (OC) is one of the three common gynecological cancers among women all over the world (Sung et al. 2021). It is also the most common cause of death from female reproductive system cancer in United States (Siegel et al. 2022). Among them, highly epithelial ovarian cancer (HG-EOC) accounts for 70-80% of all histological subtypes of ovarian cancer (Wang et al. 2020). Most of patients are in advanced stage at the first diagnosis. Despite the development of active first-line treatment, including surgery and chemotherapy, the 5-year survival rate of ovarian cancer in women diagnosed with ovarian cancer is less than 50% which bring huge burden on patients and the society . Therefore, it is necessary to find new treatments for patients with ovarian cancer.
Ferroptosis is defined by Dixon in 2012 which is a form of iron-dependent regulated cell death. It is characterized by accumulation of reactive oxygen species (ROS) and lipid peroxidation which is distinct from apoptosis, necrosis and autophagy (Dixon et al. 2012). Recent studies have shown that induced ferroptosis can inhibit the growth of a variety of Wang Yue and Gu Yupeng contributed equally to this work and share first authorship.
1 3 cancer cells, including ovarian cancer cells (Sun et al. 2021). Inducing ferroptosis may be an effective treatment strategy.
Olaparib is the most classic and effective PARP inhibitor, which has been approved by the U.S. Food and Drug Administration (FDA) for the first-line maintenance treatment of advanced ovarian cancer patients with BRCA1/2 mutation (Foo et al. 2021). Apatinib is a novel small-molecule tyrosine kinase inhibitor (TKI) that selectively binds to vascular endothelial growth factor receptor-2 (VEGFR-2) to inhibit angiogenesis (Ding et al. 2019). Apatinib has shown encouraging clinical results in the treatment of various solid tumors (Qin et al. 2021;Ruan et al. 2021;Lin et al. 2022). Clinical studies have found that olaparib combined with antiangiogenic drugs such as cediranib can improve the OS and PFS of patients in BRCA wild-type patients, and can enhance the efficacy of drugs, but the mechanism is still unclear (Liu et al. 2014). Therefore, we explore the mechanism of the combination of the olaparib and apatinib in the treatment of ovarian cancer, to find a new method for the treatment of ovarian cancer. Studies have found that olaparib can inhibit the proliferation of ovarian cancer cells by inducing ferroptosis in BRCA wild-type ovarian cancer (Hong et al. 2021), and apatinib can also inhibit the proliferation of cancer cells by inducing ferroptosis in a variety of cancer cells Tian et al. 2021). Herein, it is very important to understand the role of ferroptosis in the treatment of ovarian cancer cells by olaparib combined with apatinib.
Tumor suppressor gene p53 plays a key role in tumor inhibition. Although it is generally believed that the role of p53 in regulating apoptosis and cell cycle arrest contributes to the role of p53 in tumor inhibition, new evidence shows that p53 can also inhibit tumor growth by inducing ferroptosis (Kang et al. 2019). Recently studies demonstrated that p53 can promote the accumulation of lipid ROS and induce ferroptosis by transcriptional repression of glutamate/cystine antiporter solute carrier family 7 member 11 (SLC7A11) (Jiang et al. 2015). In addition, p53 can directly inhibit DPP4 in a transcriptionindependent manner which finally results ferroptosis is limited (Xie et al. 2017). The mechanism of p53 regulating ferroptosis is still unclear which deserve further investigation.
In this study, we found that in p53 wild-type ovarian cancer cells, apatinib combined with olaparib can inhibit the proliferation of ovarian cancer cells by inducing ferroptosis. Furthermore, we explored the mechanism of ferroptosis induced by the combination of the two drugs.

Prediction of potential targets
Using PubChem database (https:// pubch em. ncbi. nlm. nih. gov) to obtain the canonical simplified molecular input entry specification (SMILES) information of olaparib and apatinib. The SMILES information of olaparib and apatinib imported into SuperPred website (https:// predi ction. chari te. de/), which predict the potential targets of compounds based on the similar property principle.

Wound healing migration assay
Ovarian cancer cells were seeded in 6-well plates at a density of 5 × 10 5 cells per well. When at 90% confluency, a 200 μl pipette tip was used to scratch though the middle of monolayer, creating a wound. Photomicrographs at 10 × objective magnification were taken at 0 h and 48 h to assess cell migration. Image J software was used to measure and quantify the degree of cell migration under different treatment conditions.

Colony formation assay
A2780 and OVCAR3 cells were seeded in 6-well plates at the density of 1000 cells per well. Cells were treated with the combination of olaparib and apatinb with or without ferrostatin-1 for 48 h and incubated for approximate 14 days, cells were fixed with 4% formaldehyde and stained with crystal violet staining solution. The colonies were imaged by the camera and quantified by Image J software.

GEPIA analysis
Gene Expression Profiling Interactive Analysis 2.0 (GEPIA2, http:// gepia2. cancer-pku. cn/# index) is an online open database for gene expression analysis based on tumor and normal samples in The Cancer Genome Atlas (TCGA) and Genotype-Tissue Expression (GTEx) databases. In this study, GEPIA2 was used to evaluate the correlation between molecules and the expression of Nrf2 in ovarian cancer.

Statistical analysis
All assays were performed in at least three independent experiments. SPSS software (version 22.0) was used for statistical analysis of all data. Statistical comparison of multiple groups was analyzed by one-way analysis of variance (ANOVA). When p < 0.05, the difference was considered statistically significant.

Apatinib combined with olaparib induces cell death in A2780 cell line through ferroptosis but not in OVCAR3 cells
Our previous study showed that apatinib combined with olaparib treatment could significantly inhibit the proliferation of A2780 cells but not OVCAR3 cells (Wei 2022). To evaluate the role of ferroptosis behind the different effect in these two cell lines, Western blot was used to detect the level of ferroptosis-related protein. It was found that the expression level of GPX4, a negative target of ferroptosis, decreased in A2780 and increased in OVCAR3 cells after treatment with apatinib and olaparib for 48 h (Fig. 1A, B). Then to evaluate the effect of ferroptosis on cell survival, colony formation assay was performed. The ferroptosis inhibitor, Fer-1, rescued cell death induced by apatinib combined with olaparib in A2780 but not in OVCAR3 cells (Fig. 1C, D). Furthermore, wound-healing assays were performed to investigate whether ferroptosis affects cell migration. In A2780 cells, the use of Fer-1 reduced the migration of cells which was promoted by the combination of two drugs, but there is no such change in OVCAR3 cells ( Fig. 1E-F). Summing up, these data suggest that apatinib combined with olaparib suppresses cell viability and migration by inducing ferroptosis in A2780 cells. However, in OVCAR3 cells, ferroptosis was inhibited. Ferroptosis is essential for the efficacy of combination of apatinib and olaparib in ovarian cancer cells.

Inhibition of Nrf2 enhances the sensitivity of ovarian cancer cells to treatment with apatinib and olaparib
To study the underlying mechanism of the effect of combined apatinib with olaprib in inhibiting proliferation of ovarian cells, the superpred website (http:// predi ction. chari te. de/) was employed. Figure 2A showed that Nrf2 was noticed as a target of both apatinib and olaprib. Furthermore, Western blot assay also suggested that the expression of Nrf2 was inhibited in A2780 cell line following treatment with apatinib and olaprib (Fig. 2B). The expression of Nrf2 was activated in OVCAR3 cell line following treatment with apatinib and olaprib (Fig. 2C). We next examined the effect of Nrf2 on cell migration using wound-healing assay. After activating Nrf2 by RTA-408 (Nrf2 activator), the wound distance was distinctly shorter compared to two drug combination group in A2780 cell line. On the contrary, inhibition of Nrf2 by ML385 (a Keap1-Nrf2 inhibitor) the wound distance was distinctly longer compared to two drug combination group in OVCAR3 cell line (Fig. 2D-E). As expected, apatinib combined with olaparib affect migration of ovarian cancer cells partly via Nrf2. These observations reveal that inhibition of Nrf2 sensitized ovarian cancer cell to apatinib and olaparib treatment.

Nrf2 induces autophagy and inhibits ferroptosis
Previously, we showed that co-administration of apatinib and olaparib had effects on autophagy (Wei 2022). Daniela et al. found that Nrf2 system maintains cell homeostasis by positively regulating autophagy (Frias et al. 2020). To explore whether Nrf2 is involved in the effect of combined drugs on autophagy, the relationship between the Nrf2 expression level and autophagy-related protein LC3 was analyzed using GEPIA2 database. Nrf2 expression has been positively correlated with LC3 in ovarian cancer (Fig. 3A). As indicated in Fig. 3B, C, compared with the combination group, the use of Nrf2 activator RTA-408 can remarkably upregulated LC3II/I, increased the decline in p62 expression induced by apatinib and olaprib in A2780 cells. The use of Nrf2 inhibitor ML385 significantly inhibited the expression of LC3II/I, and increased the expression of p62 in OVCAR3 cells. Additionally, we found that compared with co-treatment group, improved the expression of Nrf2 led to an increase of GPX4 expression in A2780 cells. Inhibited the expression of Nrf2 led to an decrease of GPX4 in OVCAR3 cells. These results above illustrated that Nrf2 can promote autophagy and inhibit ferroptosis.

Autophagy inhibits ferroptosis induced by combination of two drugs in A2780 cell line
Recent studies have suggested a link between ferroptosis and autophagy . To further detect the role of autophagy in apatinb combined with olaparib-induced ferroptosis, A2780 cells were treated with the autophagy inducer, rapa and OVCAR3 cells were treated with autophagy inhibitor, 3MA. The results indicated that combination therapy with rapa treatment can attenuate ferroptosis by enhancing autophagy in A2780. Meanwhile, 3MA treatment inhibited autophagy induced by combination drugs and promoted ferroptosis in OVCAR3 cells (Fig. 4A, B). All of these results suggest that inhibition of autophagy promoted ferroptosis induced by apatinib combined with olaparib. At the same time, we also found that rapa promoted the expression of Nrf2, 3MA suppress the expression of Nrf2. There is reciprocal regulation between Nrf2 and autophagy.

Apatinib combined with olaparib represses GPX4 expression in a p53-dependent manner
As a tumor suppressor, p53 can inhibit tumor cells by triggering ferroptosis (Jiang et al. 2015). Our previous experiments showed that p53 is involved in the effects of apatinib and olaparib on autophagy. To explore whether p53 is involved in Nrf2-dependent ferroptosis. The correlations between gene expression levels were analyzed using Fig. 1 The role of ferroptosis in olaparib combined with apatinib. A, B The protein level of GPX4 was analyzed by western blotting in A2780 and OVCAR3 cells treated with apatinib (6 μM) or olaparib (1 μM) alone or in combination for 48 h. C, D Representative images of clonogenic assay in A2780 and OVCAR3 cells treated with DMSO, or combined with apatinib (2 μM) and olaparib (1 μM) or in combination with ferrostatin-1 (5 μM). E Effects on the wound healing in A2780 cells were determined for 0 and 48 h (at magnification, × 100). F Effects on the wound healing in OVCAR3 cells were determined for 0 and 48 h (magnification, × 100). *P < 0.05 GEPIA2. P53 expression has been positively correlated with Nrf2 in ovarian cancer (Fig. 5A). To test wondered whether p53 activation is involved in Nrf2-dependent ferroptosis, OVCAR3 cells, in which p53 was deficient were treated with RITA, the p53 activator. As shown in (Fig. 5B), RITA activates p53 to downregulate the expression of Nrf2 and GPX4 induced by apatinib and olaprib in OVCAR3 cells. In summary, our experimental results show that p53 negatively C Western blot analysis of Nrf2 in OVCAR3 cells treated as indicated for 48 h. *P < 0.05. D Cell migration potential was determined using a wound-healing assay. Wound distance was assessed in A2780 cells treated with dual-drug combination or combined with RTA-408 (20 nM). Images of the wound areas were shown at 0 and 48 h (at magnification × 100). E Cell migration potential was determined using a wound-healing assay. Wound distance was assessed in OVCAR3 cells treated with dual-drug combination or combined with ML385 (5 μM). Images of the wound areas were shown at 0 and 48 h (at magnification × 100). *P < 0.05 Fig. 3 Nrf2 contributions to apatinib combined with olaparib-induced autophagy and ferroptosis. A Nrf2 had a significant positive relationship with LC3 based on the GEPIA2 database. B Western blot-ting analysis shows the protein levels of Nrf2, GPX4, LC3 and p62 after the combination treatment with RTA408 or ML385 (C) for 48 h. *P < 0.05, **P < 0.01 regulates Nrf2, GPX4 and plays a promoting role in apatinb and olaparib-induced ferroptosis in ovarian cancer cells.

Discussion
In this study, we revealed that an important tumor suppression mechanism which links ferroptosis with the combination of apatinib and olaparib. Specifically, we believe that apatinib combined with olaparib can inhibit autophagy and promote ferroptosis by inhibiting the expression of Nrf2 in p53 wild-type ovarian cancer cells (Fig. 5C). The status of p53 is an important mechanism to regulate the sensitivity of apatinib and olaparib. We further proved that inhibition autophagy can promote ferroptosis. In p53 mutant cells, sensitivity to the combination of the two drugs can be improved by combining Nrf2 inhibitors or autophagy inhibitors.
Ferroptosis is a newly discovered form of regulatory cell death, distinguished from apoptosis, necrosis, and autophagy due to its characteristic of iron dependent and the inhibition of GPX4 (Dixon et al. 2012). Ferroptosis also can be reversed by iron chelators and antioxidants such as ferrostatin-1 or liproxstatin-1 (Cao and Dixon 2016). New evidence suggests that induction of ferroptosis may reverse drug resistance in cancer treatment, especially in tumors resistant to traditional treatments (Zhang et al. 2022). As one of the most widely studied tumor suppressor genes, p53 is involved in controlling cell survival and division under various pressures. In addition to the effects on cell cycle, autophagy and apoptosis, p53 also regulates ferroptosis through transcriptional or post-translational mechanisms (Kang et al. 2019;Liang et al. 2019). Jiang et al. found that p53 promotes ferroptosis by inhibiting the expression of SLC7A11 which is a key component of cystine / glutamate antiporter (Jiang et al. 2015). Our study supports the hypothesis that p53 promotes ferroptosis. On the other hand, p53 inhibits ferroptosis by directly blocking DPP4 (dipeptidyl peptidase 4) activity in a transcription-independent manner or by activating transcriptional target CDKN1A/p21 expression ( (Xie et al. 2017;Tarangelo et al. 2018)). Therefore, it is necessary to further explore the function and mechanism of p53 in ferroptosis in different cell lines. Because p53 in different status has different effects on ferroptosis, this factor should be taken into account when inhibiting tumor cell proliferation by inducing ferroptosis in the future clinical practice. Patients should be grouped according to p53 in different status and different intervention measures should be used.
Nrf2 is a key transcription factor encoding genes of antioxidants, detoxifying enzymes and multidrug resistance, and it also confers anticancer drug resistance. It is proved that Fig. 4 Autophagy confers resistance to ovarian cancer ferroptosis. A Protein expression levels of Nrf2, GPX4, LC3, p62 after the combination treatment with Rapa or 3MA (B) for 48 h. *P < 0.05, **P < 0.01 wild-type p53 inhibits Nrf2 transcription by reducing the binding of SP1 to Nrf2 promoter, which may confer cisplatin sensitivity, and mutant p53 may lead to cisplatin resistance by upregulating Nrf2 expression (Tung et al. 2015). Consistent with previous studies, our findings indicate that p53 inhibits the expression of Nrf2. Autophagy is an evolutionarily conserved way to maintain cellular homeostasis by degrading unnecessary or damaged organelles and proteins (Kocaturk et al. 2019). Autophagy is also a doubleedged sword in the process of tumorigenesis and development, which can promote or inhibit the development of tumor cells (Zhang and Liu 2021). Nrf2 system actively regulates autophagy to maintain cellular homeostasis (Frias et al. 2020). However, in some studies, Nrf2 was associated with autophagy inhibition. Liu et al. displayed that arsenic exposure activated Nrf2 and disrupted autophagy flux in BEAS-2B cells (Lau et al. 2013). Zhu et al. showed that the knockdown of Nrf2 specifically markedly increased the expressions of both lc3I and II in cells exposed to CSE, indicating that Nrf2 activation can negatively regulate autophagy (Zhu et al. 2013). The differences found in these studies suggest that the impact of Nrf2 on the autophagy pathway are not fully understood. Both antioxidant defense and autophagy are participated in maintaining homeostasis and cell growth process, but the possible interactions are different depending on the stress level, suggesting that Nrf2 may be an important factor in cell survival. Here we found that Nrf2 is positively correlated with autophagy-related protein LC3 through GEPIA2 database. Using Nrf2 activator can increase the level of LC3II/I, inhibit the expression of Nrf2 reduce the level of LC3II/I. We also found that inhibiting autophagy can reduce the expression of Nrf2, and promoting autophagy can also increase the expression of Nrf2. The results show that there is a positive feedback loop between Nrf2 and autophagy. Although the exact mechanism of action is still unclear, more research is still needed. At the same time, we found that inhibiting Nrf2 can reduce the expression of GPX4 and promote ferroptosis, and enhancing the expression of Nrf2 can promote the expression of GPX4, inhibit ferroptosis and reduce cell damage, which is consistent with previous studies Deng et al. 2021).
There is increasing evidence that ferroptosis is not a separate phenomenon and is closely linked to other cellular events (Lee et al. 2018;Zhou et al. 2020;Wei et al. 2021). Notably, recent studies have shown that ferroptosis is dependent on autophagy. Autophagy can lead to ferroptosis through the degradation of ferritin, with concomitant accumulation of iron levels and lipid ROS (Hou et al. 2016). Chaperone-mediated autophagy (CMA) is a selective autophagy that degrades intracellular components via lysosomes. Degradation of GPX4 protein by CMA appears to be in response to various ferroptosis activators (Wu et al. 2019). BECN1 is a special protein that promotes autophagy which was recently found that can promote ferroptosis by binding to SLC7A11 to directly block system Xc − activity A diagram elucidating the possible mechanism underlying the modulation of ferroptosis by apatinib combined with olaparib. *P < 0.05, **P < 0.01 (Song et al. 2018). These results all suggest that autophagy promotes ferroptosis. However, recent studies have found that the use of metformin may reverse the production of lipid ROS to induce ferroptosis by inhibiting autophagy . Schut et al. demonstrated that inhibition of autophagy could enhance the cytotoxic effect of sorafenib by induced ferroptosis in desmoid-type fibromatosis cells (Schut et al. 2022). The relationship between ferroptosis and autophagy remains an open question. In our study, it was also found that activation of autophagy using rapa can promote the expression of GPX4 to inhibit ferroptosis, and inhibition of autophagy using 3MA can reduce the expression of GPX4, thereby promoting ferroptosis.
In conclusion, our study described that apatinib combined with olaparib trigger ferroptosis in ovarian cancer cells via p53-mediated Nrf2 inhibition. It provides a theoretical basis for the clinical combined use of apatinib and olaparib. The regulatory relationship between autophagy and ferroptosis deserves further study. Exploring the specific role of autophagy in ferroptosis will help to further explain the anticancer mechanism of apatinib combined with olaparib.
Author contributions CJ and JK designed the study. GY performed the statistical analysis. WY and GY performed the experiments. WY wrote the manuscript. The authors read and approved the final manuscript.
Funding The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Data availability
The datasets of this study are available from the corresponding author on reasonable request.