CHD4 regulates platinum sensitivity through MDR1 expression in ovarian cancer: A potential role of CHD4 inhibition as a combination therapy with platinum agents

Platinum sensitivity is an important prognostic factor in patients with ovarian cancer. Chromodomain-helicase-DNA-binding protein 4 (CHD4) is a core member of the nucleosome remodeling and deacetylase complex, which functions as a chromatin remodeler. Emerging evidence indicates that CHD4 could be a potential therapeutic target for cancer therapy. The purpose of this study was to clarify the role of CHD4 in ovarian cancer and investigate its therapeutic potential focusing on platinum sensitivity. In an analysis of the Cancer Genome Atlas ovarian cancer dataset, CHD4 gene amplification was associated with worse overall survival. CHD4 mRNA expression was significantly higher in platinum-resistant samples in a subsequent clinical sample analysis, suggesting that CHD4 overexpression conferred platinum resistance to ovarian cancer cells, resulting in poor patient survival. In concordance with these findings, CHD4 knockdown enhanced the induction of apoptosis mediated by cisplatin in ovarian cancer cells TOV21G and increased cisplatin sensitivity in multiple ovarian cancer cells derived from different subtypes. However, CHD4 knockdown did not affect the expression of RAD51 or p21, the known targets of CHD4 in other cancer types that can modulate platinum sensitivity. Knockdown and overexpression assays revealed that CHD4 positively regulated the expression of multi-drug transporter MDR1 and its coding protein p-glycoprotein. In addition, a first-in-class CHD4/SMARCA5 inhibitor ED2-AD101 showed synergistic interactions with cisplatin. Our findings suggest that CHD4 mediates platinum sensitivity by modulating MDR1 expression in ovarian cancer. Further, CHD4 suppression has a potential to be a novel therapeutic strategy in combination with platinum agents.


Introduction
More than half of the ovarian cancers are diagnosed in advanced stages, and present with carcinomatous peritonitis [1]. The therapeutic strategy for advanced ovarian cancer is based on a combination of chemotherapy and debulking surgery [2]. Although several molecular-targeted drugs such as VEGF antibodies and PARP inhibitors have become available, for more than a decade frontline chemotherapy has been based on platinum agents, regardless of the subtype [3][4][5]. Therefore, the sensitivity to platinum agents is the most important prognostic factor for patient survival [6]. Platinum agents are the major DNA damaging agents that exert anticancer effects mainly by forming intrastrand and interstrand DNA crosslinks [7,8]. Although about 70% of the patients with ovarian cancer initially respond to platinum-based chemotherapy, some develop resistance during the primary treatment [9,10]. Recurrence after remission is almost inevitable in most of the other cases, and patient prognosis is poor once they have developed platinum resistance [9,11]. Thus, a strategy to overcome platinum resistance is an important concern in ovarian cancer treatment.
Chromodomain-helicase-DNA-binding protein 4 (CHD4) is a component of the nucleosome remodeling and deacetylase (NuRD) complex. As the name suggests, NuRD complex acts as a chromatin remodeler and CHD4 is considered to play a pivotal role by conferring ATPase activity on the complex [12][13][14]. The NuRD complex is believed to act as a transcription repressor, but recent research indicated that it can also promote gene transcription globally [14,15]. NuRD or CHD4 is reported to be involved in several important biological functions in cancer cells, such as cell differentiation, cell cycle regulation, and DNA damage repair [16][17][18][19][20][21][22]. It has also been reported that CHD4 expression was correlated with patient prognosis in a variety of malignancies [23][24][25][26]. These findings suggest that CHD4 could be a potential therapeutic target in cancer therapy. However, the importance of CHD4 in ovarian cancer has not been fully studied.
The purpose of this study was to elucidate the potential of CHD4 as a new therapeutic target for the treatment of ovarian cancer, with a particular focus on its interaction with platinum agents.

Cell lines and cell culture conditions
TOV21G was obtained from the American Type Culture Collection. KURAMOCHI was obtained from the Japanese Cancer Research Resources Bank. JHOS2, and JHOC5 were obtained from RIKEN. A2780 and A2780cis were obtained from Sigma-Aldrich. OSE2 and OSE4 were kindly provided by Dr. Makoto Nitta (University of Kumamoto, Kumamoto, Japan) [27]. TOV21G and KURAMOCHI were maintained in RPMI1640 with 10% fetal bovine serum (FBS). JHOS2 and JHOC5 were cultured in Dulbecco's modified Eagle's medium Nutrient Mixture F-12 (DMEM/F-12) supplemented with 10% FBS and MEM non-essential amino acid. A2780, A2780cis, OSE2, and OSE4 were cultured in DMEM supplemented with 10% FBS. All cells were cultured in 5% CO2 at 37˚C. All of the cells were validated as mycoplasma negative at the time of arrival and cell aliquots were cryopreserved after a couple of passages for subsequent experiments.

TCGA data analysis
Data on genomic alterations and patient survival among 489 ovarian cancer patients registered to The Cancer Genome Atlas Network (accession number: phs000178) were downloaded via cBioPortal (https://www.cbioportal.org/) [28,29]. Overall survival was analyzed using the Kaplan-Meier method and survival difference was compared using the log-rank test.

Clinical patient data analysis
Histologically confirmed primary ovarian high-grade serous or clear cell carcinoma specimens from patients who underwent surgery at Tohoku University Hospital between January 2011 and December 2017 were subjected to analysis with waived informed consent under the approval of the ethics committee at Tohoku University Graduate School of Medicine (approval number: 2020-1-238.). The investigator accessed the medical record and the stored samples during August 2020 to September 2020. The data were not fully anonymized before the investigators accessed them. To assess CHD4 RNA expression, only those patients whose specimens were stored at -80˚C after embedding in OCT compound were eligible (Sakura Finetechnical Co. Ltd., Tokyo). Samples were homogenized, and RNA was extracted and quantified as described later. Clinical information, including overall survival and platinum sensitivity, were also collected. In this study, a specimen was considered to be a platinum-sensitive one only if clinically detectable tumor remained after the surgery and the patients responded to the subsequent three cycles of platinum-based chemotherapy immediately after the surgery. In contrast, a specimen was considered to be platinum resistant if the detectable tumor showed progressive or non-regressive behavior toward subsequent chemotherapy after primary suboptimal surgery. The platinum sensitivity was considered to be unknown in cases where the tumor was completely or almost completely resected by surgery. The correlation between CHD4 mRNA expression with overall survival and platinum sensitivity was assessed using the log-rank test and Mann-Whitney test, respectively.

Immunohistochemistry
Formalin-fixed, paraffin-embedded tissue sections were treated with 3% hydrogen peroxide diluted in methanol after deparaffinization and heat-induced epitope retrieval. Tissue sections were then incubated with Anti-CHD4 antibody (ab70469, Abcam), which was diluted to the concentration of 1:200 at 4˚C overnight, and subsequently treated with Histofine 1 Simple Stain MAX-PO(M) (Nichirei) at room temperature for 30 minutes. Antigen-antibody complex was visualized with 3,3⁻diaminobenzidine solution. After counterstaining with hematoxylin, immunohistochemical reaction was evaluated through a light microscope.

RNA extraction, cDNA synthesis, and quantitative RT-PCR
Total RNA was extracted using the RNeasy Mini Kit (QIAGEN) or ISOGENII (NIPPON GENE). Reverse transcription was performed using ReverTra Ace qPCR RT Master Mix with gDNA Remover (FSQ-301, TOYOBO) to obtain complementary DNA. Quantitative PCR was performed using the StepOne Plus Real-Time PCR System with TaqMan1 Fast Advanced Master Mix (Applied Biosystems) and a TaqMan probe specific for each gene. Relative gene expression was measured using the ΔΔCT method. The expression of CHD4, ABCC1, ABCC2, and MDR1 was normalized to GAPDH mRNA expression. The following TaqMan probes were used in this study: CHD4 (Hs00172349_m1), MDR1 (ABCB1) (Hs00184500_m1), ABCC1 (Hs01561512_m1), ABCC2 (Hs00166123_m1), and GAPDH (NM_002046.3) (all from Thermo Fisher Scientific).
Clonogenic assay. After washing with phosphate buffered saline (PBS), cells plated on a 6-well plate were fixed with 100% methanol at -20˚C for 10 minutes. The cells were then stained with 0.5% crystal violet for 10 minutes at room temperature.

Flow cytometry
After the treatment, cells were washed with PBS and stained in Annexin V Binding Buffer with SYTOX™ AADvanced™ Dead Cell Stain Kit (S10349, Thermo Fisher Scientific) and annexin V, Pacific Blue conjugate (A35122, Thermo Fisher Scientific) following the manufacturer's instructions. The fraction of Annexin V-positive and Pacific blue positive cells were measured using flow cytometry with FACS Canto™ II (BD) and analyzed using FACS Diva software (BD).

Cell cycle analysis
After the treatment, cells were washed with PBS, fixed with 70% ethanol, and stained with Propidium Iodide (P3566, Thermo Fisher Scientific). Cell cycle analysis was performed by flow cytometry with FACS Canto™ II and analyzed by FACS Diva software.

Drug interaction assessment using Bliss independence model
The SMARCA5/CHD4 inhibitor ED2-AD101 was kindly provided by Dr. Saunthararajah [31]. Cells were plated in 96-well plates 24 h prior to treatment. Cells were treated with the designated concentrations of ED2-AD101 and cisplatin. Cell viability was assessed 72 h after drug administration and cell viability was determined as mentioned above. The percentage of cell growth inhibition compared to mock treatment was determined. A Bliss independence model was employed to evaluate the drug interaction effects [32]. Briefly, the Bliss independence model was used to estimate growth inhibition if the two agents were independent. When the cell growth inhibition (%) by each drug A or B is EA and EB, respectively, the Bliss index (Ebliss), which is the calculated value when the combined action of drugs A and B is an additive effect, is calculated using the following formula: Ebliss = E + EB (EA×EB). In this study, the difference between the observed growth inhibition and expected growth inhibition, which was calculated using the Bliss independence model was defined as the Bliss independence value (BI value). BI values of more than 0 and less than 0 suggest antagonistic and synergistic interactions, respectively.

Vector transfection
TOV21G cells were plated as 3×10 5 cells per well on a 6-well plate. After 24 h, the cells were transfected with 1 μg/well of human CHD4 overexpression vector (FHC11086 from Kazusa DNA Research Institute) or the corresponding pFN21A-and-pFN21K-HaloTag-CMV-Flexi empty vector (G282A from Promega) using Lipofectamine 2000 transfection reagent (Life Technologies) according to the manufacturer's instructions.

Statistical analysis
Statistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software Inc.). If not mentioned otherwise, the values in the figures represent the means of three independent assays. Error bars in each figure represent the standard deviation (SD). Statistical analysis was performed using Student's t-test, except for the log-rank test and Mann-Whitney test in the clinical data set analyses. P-values less than 0.05 were considered statistically significant.

CHD4 amplification or overexpression is associated with poor survival in patients with ovarian cancer
To investigate the role of CHD4 in the outcome among patients with ovarian cancer, we first referred to the TCGA dataset and analyzed the relationship between CHD4 gene alterations and overall survival of the patients. Mutations in CHD4 were found in 2.5% of the patients, and we noticed that a small number of patients harbored CHD4 amplification. Among the 485 cases, in which the information on putative copy number was available, 27 cases had CHD4 amplification while no deletion was detected. Importantly, patients harboring CHD4 amplification showed significantly worse overall survival than those without CHD4 copy number alteration. (p = 0.0475) (Fig 1A).
To verify the above findings in our clinical database, we accessed the clinical dataset and tissue bank at the Tohoku University Hospital. We focused on high-grade serous cancer and clear cell carcinoma as these are the most and second most frequent histologic subtypes in our nation. Patient characteristics are shown in Table 1. Twenty-nine patients with high-grade serous carcinomas and 32 with clear-cell carcinomas were enrolled. Based on the definition described above, 20 cases were considered as platinum sensitive, 6 as resistant, and 35 as unknown platinum sensitivity. As platinum-resistance is one of the most critical prognostic factors in patients with ovarian cancer, the information on platinum sensitivity as well as on overall survival in each patient was obtained. Patient characteristics are summarized in Table 1. RNA was extracted from the stored samples and CHD4 mRNA expression was quantified using qPCR. With regard to patient survival, there was no statistically significant difference between the two groups divided at median CHD4 mRNA expression. (p = 0.084, S1A Fig). Interestingly, however, CHD4 expression was significantly higher in platinum-resistant cases (p = 0.033), suggesting that CHD4 overexpression had the potential to lead platinum resistance in patients with ovarian cancer (Fig 1B). Immunohistochemistry was additionally performed to further investigate the interaction between CHD4 expression and platinum sensitivity. Majority of the specimens were positive for CHD4 immunostaining. Interestingly, the six platinum resistant samples were all positive for CHD4 expression, and all of the samples negative for CHD4 immunostaining belonged to the platinum sensitive group although no statistically significant difference was observed. (S1B and S1C Fig). The results from the clinical data analysis led us to hypothesize that CHD4 expression possibly influences patient prognosis by regulating platinum sensitivity in ovarian cancer.

CHD4 knockdown restores the sensitivity to cisplatin in ovarian cancer cells
To experimentally prove the hypothesis that CHD4 regulates platinum sensitivity in ovarian cancer, ovarian cancer cell lines were subjected to in vitro experiments. First, the expression of CHD4 in several ovarian cancer cell lines was compared by quantitative PCR and Western blotting to examine whether there was a significant difference among ovarian cancer cells including A2780 and A2780cis, a pair of parental cells and experimentally acquired platinumresistant derivatives. Although quantitative PCR indicated that CHD4 expression tends to be higher in KURAMOCHI, A2780, and A2780cis, obvious difference was not identified across the cell lines by Western blotting (S2 Fig). Thus, we decided to select two clear-cell subtype cells (TOV21G, JHOC5) and two high-grade serous subtype cells (JHOS2, KURAMOCHI), the biomolecular characteristics of whose original subtypes are well conserved [33][34][35]. The cells were treated with the designated concentrations of cisplatin for 72 h after transfection either with siRNA targeting CHD4 or negative control. The efficiency of CHD4 knockdown was verified using quantitative PCR in all cells. Cell viability was assessed 72 h after cisplatin administration to obtain dose-response curves and determine IC50 (Fig 2 and S1 Table). Of the four cell lines, CHD4 knockdown significantly decreased the IC50 of cisplatin in TOV21G, JHOC5, and KURAMOCHI. While the IC50 was not significantly different in the three independent assays, the same trend was observed in JHOS2 (Fig 2 and S1 Table). To validate the influence of CHD4 knockdown on cisplatin sensitivity in detail, TOV21G, JHOC5 and KURA-MOCHI were treated with 0.5, 1, 2, 5, and 10 μM of cisplatin. CHD4 knockdown significantly increased the sensitivity to cisplatin at the multiple concentration points as demonstrated in S3 OSE2 and OSE4, two immortalized cell lines derived from normal ovarian surface epithelium, were additionally tested to compare the difference between cancer and non-cancerous cells.
The sensitivity to cisplatin did not significantly differ between CHD4 knockdown and control in these non-cancerous cell lines, which indicates the cancer specific involvement of CHD4 in the regulation of platinum sensitivity (S5 Fig). To further understand the direct influence of CHD4 knockdown on cell viability independent of cisplatin treatment, the cell viability was compared after 72 to 120 h after CHD4 knockdown in the four ovarian cancer cells. The response varied among cell lines, indicating that the combination of treatment with platinum agents and CHD4 suppression was a more promising therapeutic strategy for ovarian cancer than CHD4 inhibition alone (S6 Fig).

Novel SMARCA5/CHD4 inhibitor ED2-AD101 synergizes with cisplatin
Although not specific to CHD4, Dr. Saunthararajah and his colleagues established a SMARCA5/ CHD4 dual inhibitor ED2-AD101 that potently suppressed AML cell growth in cell-based assays [31], which motivated us to investigate whether ED2-AD101 increased the sensitivity of ovarian cancer cells to platinum agents. IC20 of ED2-AD101 for each cell was preliminarily determined during 72 h of dose-response assays (5 μM for TOV21G and KURAMOCHI, 7.5 μM for JHOC5; S7 Fig). As it was difficult to determine IC20 for JHOS2 based on the dose response curve, which was the least sensitive cell line to ED2-AD101 treatment, 10 uM of ED2-AD101 was administered to JHOS2 instead. These cells were treated with the designated concentrations of cisplatin in the presence of IC20 of ED2-AD101 or vehicle for 72 h. Among the four cells, IC50 was significantly lower in TOV21G, JHOC5, and KURAMOCHI, consistent with the results shown in Fig 2. The increase of platinum sensitivity was not observed in JHOS2 (Fig 3A). To further assess the synergistic interaction of ED2-AD101 and cisplatin in detail, a Bliss independence (BI) model was employed. The left panels in Fig 3B show the observed percentage viability at the indicated concentrations of ED2-AD101 and cisplatin. The right panel shows the color-coded BI values, where red, white, and blue indicate synergistic, additive, and antagonistic interactions, respectively. The results clarified that ED2-AD101 and cisplatin demonstrated synergistic reactions at most of the points in TOV21G, JHOC5 and KURAMOCHI. Interestingly, synergistic interaction was also suggested at almost all concentration points in JHOS2 as well. The results with ED2-AD101 successfully validated the potential of CHD4 as a novel therapeutic target in ovarian cancer cells in combination with cisplatin.

Inhibition of CHD4 enhances cisplatin-induced apoptosis
One of the major mechanisms of platinum agents in cancer therapy is to induce apoptosis by facilitating DNA damage, including DNA double-strand breaks [36,37]. It is also noteworthy that previous studies have shown that CHD4 was involved in DNA damage repair [12,17,21]. This suggests that CHD4 inhibition enhances DNA damage and subsequent apoptosis induced by platinum agents. To investigate whether CHD4 was involved in cisplatin-related apoptosis in ovarian cancer, the change in the phosphorylation at the histone H2AX at Serine 139 (γH2AX) and PARP cleavage were examined using Western blot analysis. TOV21G cells were transfected with either siRNA targeting CHD4 or negative control siRNA for 72 h. Subsequently, the cells were treated with 5μM cisplatin or vehicle for 24 h and protein was harvested. Whereas, CHD4 knockdown alone did not affect the expression of γH2AX, and CHD4 siRNA transfection significantly enhanced γH2AX overexpression induced by cisplatin treatment. With regard to PARP expression, CHD4 knockdown unexpectedly increased both cleaved and non-cleaved forms in the presence of cisplatin, which makes it difficult to interpret these results ( Fig 4A). Thus, flow cytometry was performed to assess the apoptosis. As shown in Fig  4B, CHD4 [36], we further analyzed whether CHD4 inhibition interacts with cisplatin from the perspective of cell cycle control using cell cycle analysis. Approximately, 24 hours of cisplatin treatment strongly facilitated S phase arrest. CHD4 knockdown also led to obvious S phase arrest and mild G2/M arrest as well in TOV21G. Neither S phase nor G2/M phase arrest was synergistically enhanced by the combination of CHD4 knockdown and cisplatin treatment, suggesting CHD4 inhibition and cisplatin work on cell cycle regulation independently (S8 Fig).

CHD4 modulates platinum sensitivity by positively regulating MDR1 expression
Variable factors modulate platinum sensitivity [38]. While it was not tested in ovarian cancer, in glioblastoma and breast cancers, it was demonstrated that knockdown of CHD4 decreases the expression of RAD51 and p21 that are tightly involved in cisplatin sensitivity [39][40][41][42]. Thus, we next examined the expression of RAD51 and p21 under CHD4 knockdown in two ovarian cancer cells, TOV21G, and JHOC5. Different from our expectation, the expression of RAD51 or p21 did not decrease under CHD4 knockdown (Figs 4C and S9A). As RAD51 expression seemed to slightly increase in TOV21G instead, we quantified and statistically compared the relative RAD51 expression between control and CHD4 knockdown but did not find a significant difference (S9B Fig).
Another important factor related to cisplatin sensitivity is the function of multi-drug efflux transporters [38,43]. The influence of CHD4 knockdown on the expression of several known multi-drug efflux transporter ATP-binding cassette (ABC) C1, ABCC2, and ABCB1 or multidrug resistance-1 (MDR1) were examined using quantitative PCR. Interestingly, the expression of MDR1 was significantly decreased when CHD4 expression was suppressed in TOV21G cells (Fig 4D). Western blot analysis confirmed that CHD4 knockdown suppressed P-glycoprotein expression in TOV21G as well as JHOG5 and JHOS2, suggesting that CHD4 was involved in platinum sensitivity by regulating drug efflux via P-glycoprotein in ovarian cancer. (Figs 4E and S9C).
To further confirm the involvement of CHD4 in MDR1 transcription, the CHD4 overexpression vector was introduced to TOV21G. Compared with the empty vector, the overexpression vector introduction increased the expression of MDR1 along with an increase in CHD4 expression, as expected (Fig 4F).

Discussion
In this study, we clarified that CHD4 was involved in platinum sensitivity in ovarian cancer by positively regulating MDR1 expression. We have also successfully demonstrated the potential of the combination therapy of platinum agents and CHD4 inhibitor by introducing the firstin-class SMARCA5/CHD4 inhibitor ED2-AD101 in ovarian cancer cells.
There are several known biological mechanisms regulating platinum sensitivity [38]. Of these, the involvement of DNA repair systems such as nucleotide excision repair, mismatch repair, homologous recombination (HR), and non-homologous end binding has been well investigated [38,44]. For example, cisplatin sensitivity can be increased by inhibiting the function of the molecules implicated in the HR pathway, such as BRCA1, BRCA2, and RAD51 [45][46][47]. Dysfunction of p21, which leads to G1 cell cycle arrest in response to DNA damage in a p53-dependent manner, is also involved in HR and is known to increase platinum sensitivity [37,48,49]. Another important mechanism related to platinum sensitivity is the regulation of influx or efflux via multi-drug transporters [43]. Among these, P-glycoprotein encoded by MDR1 is one of the most studied cell multi-drug transporters that are highly expressed on the membrane of multidrug-resistant cancer cells [50,51]. As the name implies, MDR1/P- glycoprotein is reported to be involved in the resistance of various drugs including anti-cancer agents such as paclitaxel, doxorubicin, PARP inhibitors, and importantly, cisplatin [52].
Contrary to our expectation, the expression of RAD51 or p21, which are the known targets of CHD4 as it was shown in other cancer types [40,53,54], was not clearly affected by CHD4 knockdown in the ovarian cancer cells we tested. In contrast, our experimental results and supporting literature indicated that reduced platinum efflux via P-glycoprotein was one of the mechanisms of action of CHD4 suppression that enhanced the effect of cisplatin. It should be noted that CHD4 possibly regulates platinum sensitivity through more variable and complex mechanisms. Further investigation is necessary to determine whether the regulation of MDR1 expression is a pivotal mechanism. At the same time, the results of this study imply the cancer type-specific function of CHD4. The common and cancer type-specific gene transcription regulation should also be further assessed to unveil the detailed function of CHD4 as well. Another question to be addressed is the mechanism of regulation of MDR1 expression by CHD4. MDR1 expression is complicatedly regulated by several important cancer-associated pathways such as RAS/Raf/MEK, JAK/STAT, PI3K/AKT, as well as epigenetic regulators such as histone acetyltransferases, histone deacetylases, etc. [55] As CHD4/NuRD complex functions as a chromatin remodeler that regulates the accessibility of variable complexes related to gene transcription, and a chromodomain motif in CHD4 does not directly bind to a specific DNA-sequence, CHD4 is more likely to regulate MDR1 expression via complex indirect transcriptional regulation rather than functions as direct transcription factor or enhancer [14,56].Suppression of ERBB signal cascade by CHD4 inhibition could be an example of several potent mechanisms [57].
Ovarian cancer is now recognized as a heterogeneous disease [28,58]. The genomic alteration profiles are quite distinct depending on histologic subtypes or even among the same subtypes [59]. As we observed increased platinum sensitivity by CHD4 knockdown consistently in multiple ovarian cancer cells derived from HGSC and clear cell carcinoma, it is plausible to consider that CHD4 suppression can be beneficial to the majority of ovarian cancer cases regardless of the subtype or genomic alterations. However, it is important to note that Guillemette et al. reported that CHD4 conferred cisplatin resistance in BRCA2-mutant cancer [60]. As BRCA2 mutation is found in 11-30% of HGSC, the indication of CHD4 inhibition in combination with cisplatin might be genetically tailored [61,62].
We have successfully demonstrated the potential of the novel SMARCA5/CHD4 inhibitor ED2-AD10 in combination with cisplatin for the first time in ovarian cancer cells. Along with the emerging literature reporting CHD4 as a candidate therapeutic target, our findings highlight the importance of developing clinically available CHD4 inhibitors.
A major limitation of this study is that we could not assess whether CHD4 suppression actually sensitizes clinically chemo-resistant cancer to platinum agents. Although it is difficult to evaluate this aspect appropriately with conventional cancer cell models, patient-oriented models such as patient-derived xenograft mice or organoids established from clinically proven platinum-resistant tumor tissue will help us address this question in the future. Another possibility is that the combination therapy with CHD4 inhibitor might be able to make platinum agents effective at lower concentrations, resulting in reduction of adverse events, if not overcoming resistance.

Conclusions
In conclusion, we clarified the role of CHD4 in regulation of platinum sensitivity in ovarian cancer cells. We further showed the potential of CHD4 as a candidate therapeutic target in combination with platinum agents for ovarian cancer. Although the feasibility should be assessed further, CHD4 has the potential to bring benefits to patients with ovarian cancer.