1 Introduction
Ovarian cancer is one of the malignant tumors with a high prevalence rate among women and also the main cause of death related to gynecological cancers globally. In 2020, there were 314 000 new cases and 207 000 deaths [
1,
2]. Due to the absence of typical symptoms in the early stage of ovarian cancer and the lack of effective screening methods, patients are often in an advanced stage when diagnosed. Currently, a satisfactory debulking surgery in combination with platinum-based chemotherapy is the standard treatment for ovarian cancer [
1,
3,
4]. With the development of new drugs and the conduct of clinical studies, PARP (poly ADP-ribose polymerase) inhibitors have been approved for the maintenance treatment of ovarian cancer patients. However, the therapeutic effect of ovarian cancer is not satisfactory. Approximately 70% of ovarian cancer patients appear recurrence and platinum resistance, which is the crucial factor leading to the death of patients [
5,
6], and the 10-year survival rate is merely 17% [
7–
9]. Given the close correlation between platinum resistance and poor prognosis of patients, it is extremely urgent to elucidate the mechanism of platinum resistance and identify new targets for the treatment of ovarian cancer patients with platinum resistance.
The mechanism by which ovarian cancer cells to chemotherapy resistance is relatively complex, including DNA damage repair, multi-drug resistance, cell metabolism, cell cycle regulation, cancer stem cells, cellular immunity, autophagy, abnormal signaling pathways, and so on [
10]. Among them, DNA damage repair is one of the most important mechanisms of platinum resistance in ovarian cancer [
11]. Studies have shown that more than 50% of epithelial ovarian cancer have defects in homologous recombination repair (HRR), which is mainly related to germline inheritance and epigenetic alterations of genes related with the HRR pathway [
12,
13].
BRCA1/2 play a crucial role in the HRR pathway, and maintain the normal functions of cells by repairing DNA damage with high fidelity [
14]. BRCA upregulation results in the enhancement of DNA damage repair; therefore, cancer cells are highly sensitive to drugs, including platinum and PARP inhibitors [
11,
15]. In ovarian cancer patients with BRCA mutations, reversion mutations or overexpression of BRCA are key mechanisms leading to platinum resistance [
16]. These reversion mutations may be induced by DNA damage caused by chemotherapy or genetic instability, and usually transform tumor cells from HRR-deficient to HRR wild-type through restoring open reading frame (ORF) to encode functional proteins [
17]. In malignant tumors, reversion mutations of the HRR pathway, including BRCA1, BRCA2, PALB2, RAD51C, and RAD51D, repair the DNA damage induced by platinum-based drugs or PARP inhibitors, disrupt synthetic lethality, enable tumor cells to survive, and ultimately lead to resistance to platinum-based drugs and PARP inhibitors [
18].
Inhibiting DNA damage repair can reverse the resistance of tumor cells to platinum-based drugs [
19]. However, real-world studies show that in ovarian cancer patients, especially those with the high-grade serous carcinoma subtype, germline BRCA mutations can only be detected in approximately 14%–18% of cases, which greatly limits the application of PARP inhibitors, resulting in a high mortality rate among ovarian cancer patients [
4,
20]. Therefore, developing new targets aimed at BRCA protein expression to enhance the sensitivity of ovarian cancer cells to platinum-based drugs is the main hot spot currently.
It is reported that the ways to influence the protein expression level of BRCA1 include epigenetic modifications of DNA, regulations at the transcriptional, post-transcriptional, and translational levels, as well as epigenetic modifications after translation. Among these, protein–protein interactions have very important significance. Proteins that bind to BRCA1 can also play a role in regulating the expression level of BRCA1 [
21]. Through proteomics we have screened out the protein SFPQ that binds to BRCA1 with high abundance. SFPQ, also known as PSF, is a DNA/RNA-binding protein consisting of 707 amino acids. Current research indicates that SFPQ is not only involved in mRNA splicing, maturation, and translation, but also participates in the progression of malignant tumors, neural development and regeneration [
22,
23]. In DNA damage repair, SFPQ has been proven to directly regulate the homologous pairing repair function of RAD51 [
24]. In addition, the SFPQ/NONO heterodimer has been identified as a novel telomere regulator to maintain genomic stability. The absence of SFPQ/NONO leads to a significant increase in telomere recombination and rapid changes in telomere length [
25]. However, whether SFPQ affects DNA damage repair in ovarian cancer through BRCA1 and the regulatory mechanism therein has not been reported at home and abroad.
Our previous study showed that a high expression level of SFPQ is associated with platinum resistance and poor prognosis of ovarian cancer. Up to now, it has not been reported that SFPQ binds to BRCA1 and participates in platinum resistance in ovarian cancer. Therefore, we further demonstrated that silencing SFPQ enhanced the sensitivity to platinum of ovarian cancer cells in vivo and in vitro. Mechanistically, SFPQ binds to BRCA1 and inhibits its ubiquitination and degradation, upregulates the protein expression level of BRCA1, enhances DNA damage repair, and promotes platinum resistance. In addition, we discovered that hsa_circSFPQ_008, derived from the parent of SFPQ recruits HDAC1 to deacetylate the H3K27Ac region of the SFPQ promoter to regulate its expression. Overall, our study suggested that SFPQ may be a potential target for overcoming platinum resistance in ovarian cancer and provide a new theory for individualized and precise treatment in ovarian cancer.
2 Materials and methods
2.1 Patients and samples
OC tissues were collected from patients who did not receive therapy prior to tumor debulking operation from the Third Affiliated Hospital of Guangzhou Medical University. All patients were diagnosed with serous ovarian carcinoma, and their postoperative histopathological types were reviewed according to WHO criteria by two experienced gynecological pathologists.
2.2 Cell culture and treatment
Human embryonic kidney (HEK293T) cells were obtained from the American Type Culture Collection (ATCC, Manassas, USA), and human OC cell lines A2780, Skov3 were obtained from Shanghai Bioresource Collection Center (SHBCC, Shanghai, China). A2780-PR (A2780-platinum resistant) cells were developed from the parental cells by sequential exposure to increasing concentrations of carboplatin. A2780 was cultured in high-glucose Dulbecco’s Modified Eagle’s Medium (DMEM, #C11995500BT, GIBCO) supplemented with 10% fetal bovine serum (FBS, #A3161002C, GIBCO) and 1% penicillin-streptomycin (#15070063, GIBCO), while Skov3 was cultured in McCoy’s 5A (#PM150710, Procell) supplemented with 10% FBS and 1% penicillin-streptomycin at 37 °C in an incubator with 5% CO2. All cell lines were authenticated using short-tandem repeat profiling when this project was initiated, and the cells were not cultured for more than 2 months.
2.3 RNA inference
Short interfering RNA designated specifically against SFPQ, BRCA1, HDAC1, hsa_circSFPQ_008, and their corresponding scrambled siRNA were purchased from Ribobio Co., Ltd. (Guangzhou, China). The sequences were summarized in Table S4. For transfection, 1 μg of siRNA was dissolved in 125 μL of Opti-MEM media (#31985070, GIBCO). In another tube, the transfection reagent Lipofectamine 3000 (#L3000015, Invitrogen) was dissolved in 125 μL of Opti-MEM media. The two mixtures were then combined and incubated for 20 min at room temperature (RT). The mixture was added to cells that had been plated in a 6-well plate with 2 mL of fresh media. At 48 h after transfection, cells were harvested, and the effect of gene silencing was examined by Western blot.
2.4 Plasmids, virus production, and infection
pCDNA3.1-SFPQ-3xHA, pCDH-BRCA1-3xFlag, pCDNA3.1-BRCA1-3xFlag, and its four truncated BRCA1 plasmids, PLP1, PLP2, and PLP/VSVG plasmids were obtained from Guangzhou Hanyi Biotechnology Co., Ltd. (Guangzhou, China). pLV-hsa_circ_0011532 plasmid was purchased from Beijing SyngenTech Co., Ltd. (Beijing, China). His-Ubiquitin plasmid was kindly provided by Prof. Min Zheng from Sun Yat-sen University Cancer Center.
To produce retrovirus, HEK293T cells were co-transfected with packaging plasmids and pCDNA3.1-SFPQ-3xHA or pCDH-BRCA1-3xFlag plasmid using Lipofectamine 3000. At 48 h post-transfection, the supernatants were collected and filtered using a 0.45-μm cellulose acetate filter (#YA0673, Solarbio). Subsequently, cells were incubated in culture medium with supernatants containing virus particles and supplemented with 8 μg/mL polybrene (#TR-1003-G, Millipore Sigma) for 12 h, followed by replacement with fresh medium. At 48 h post-infection, puromycin (2 μg/mL, #P8833, MilliporeSigma) was added to screen the infected cells.
2.5 Cell apoptosis assay
Apoptosis assays were conducted using an Annexin V-FITC/PI Apoptosis Detection Kit (#BB-4101-3, BestBio). Cells were treated with different concentrations of carboplatin (#A8321, APExBIO) for 24 h. Apoptotic cells were then collected by trypsin without EDTA (#BL526A, Biosharp), washed three times with phosphate-buffered saline (PBS, #BL302A, Biosharp), and stained with Annexin V-FITC and propidium iodide (PI) according to the manufacturer’s protocol. Early apoptotic cells (Annexin V-positive, PI-negative) and late apoptotic cells (Annexin V-positive, PI-positive) were determined by flow cytometry (Attune NxT, Thermo Fisher Scientific Inc., Waltham, USA), and the results were analyzed using FlowJo 10 software (Tree Star, Ashland, USA).
2.6 Cell proliferation assay
A2780 or Skov3 cells (1 × 103 cells/well) were seeded into 96-well plates (100 μL cell suspension) and treated with carboplatin. Cell viability was assessed every 24 h by Cell Counting Kit-8 (CCK-8, #CK04, DOJINDO) assays according to the manufacturer’s instructions.
2.7 Immunohistochemistry
Tissue specimens were fixed in 10% formalin for 24 h and paraffin embedded. IHC staining was performed on 4-μm-thick paraffin sections, which were deparaffinized in xylene and alcohol. Antigen retrieval was performed using EDTA buffer (pH 9.0). Sections were blocked in 3% H2O2 for 10 min and incubated with primary antibodies at 4 °C overnight in a moist chamber. Slides were rinsed with PBS and incubated with universal immuno-peroxidase polymer anti-rabbit or anti-mouse antibodies (#H2008, #2004, respectively, Nichirei Biosciences), and visualized by 3,3′-diaminobenzidine tetrahydrochloride substrate (DAB, #K5007, Dako).
For quantification of staining, tissue specimens were scanned and analyzed using Vectra software (PerkinElmer Inc., Waltham, USA). For each core, three images of representative areas were analyzed. IHC scoring was performed using Histoscore (H-score) calculated by the IHC nuclear image analysis system, which included a semiquantitative assessment of both fraction of positive cells and intensity of staining. The intensity score was defined as no staining (score 0), weak (score 1), moderate (score 2), or strong (score 3). The fraction score was based on the proportion of positively stained cells (0%–100%). The intensity and fraction scores were then multiplied to obtain the H-score of 0–3. OC tissues were divided into high- and low-SFPQ expression groups based on the cut-off value of SFPQ expression determined using ROC curve.
2.8 RNA extraction and quantitative real-time PCR
Total RNA was extracted using AG RNAex Pro Reagent (#AG21102, Accurate Biology), and quantified using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific). Total RNA (2 μg) was reversely transcribed using Hifair® III 1st Strand cDNA Synthesis SuperMix for qPCR (gDNA digester plus) (#11141, Yeasen). The mRNA levels were detected by Hieff® qPCR SYBR Green Master Mix (Low Rox Plus) (#11202, Yeasen) and performed on a QuantStudio 3 Real-Time PCR System (Thermo Fisher Scientific Inc., Waltham, USA). GAPDH and U6 were used as a normalization control. Primer sequences are listed in Table S5.
2.9 Western blotting
Cells were collected and lysed in RIPA lysis buffer (#P0013C, Beyotime). A bicinchoninic acid assay (#20200, Yeasen) was used to measure the protein concentrations in the lysate. Approximately 20 μg of total protein was resolved on 8%–15% SDS-PAGE gels, electrophoresed, and transferred to 0.2-μm polyvinylidene difluoride membranes (PVDF, #1620177, Bio-Rad). After blocking with 5% nonfat milk in PBS, the membranes were incubated with specific antibodies at 4 °C overnight, followed by incubation with a HRP-conjugated secondary antibody for 1 h at room temperature. Finally, ECL reagents (#36222, Yeasen) were used to visualize bands and signals were measured.
2.10 Co-immunoprecipitation (Co-IP) assay
For exogenous immunoprecipitation, HEK293T cells were transfected with the indicated plasmids and lysed 48 h after transfection in NP-40 lysis buffer (#P0013F, Beyotime) supplemented with protease inhibitors (#P1045, Beyotime) on ice for 15 min. After centrifugation at 4 °C and 12 000 rpm for 15 min, the supernatants were isolated and incubated with anti-Flag magnetic beads (#HY-K0207, MedChemExpress) at 4 °C overnight. The beads were washed with NP-40 lysis buffer six times before immunoblotting was performed.
For endogenous immunoprecipitation, A2780 or Skov3 cells were harvested and incubated with 2 μg of antibodies overnight at 4 °C. Protein A/G magnetic beads (#HY-K0202, MedChemExpress) were then added for 6 h at 4 °C. Next, the beads were washed with NP-40 buffer six times. Finally, the proteins were dissolved in 1× SDS-PAGE loading buffer, boiled for 10 min at 100 °C, and analyzed by Western blotting.
For ubiquitination assays, HEK293T cells were transfected with the aforementioned plasmids for 48 h. Before the collection of cell lysates, cells were treated with MG132 (10 μmol/L, #HY-13259, MedChemExpress) for 6 h. Cell lysates were then subjected to exogenous immunoprecipitation experiments as described above.
A detailed list of antibodies used can be found in Table S6.
2.11 Protein half-life assay
To determine the half-life of BRCA1, A2780-Vector and A2780-SFPQ cells were treated with 20 μg/mL cycloheximide (CHX) (#C7698, Sigma-Aldrich) at the indicated time points, and then were collected for Western blotting analysis.
2.12 RNA immunoprecipitation (RIP)-qPCR
Total RNA from A2780 cells was isolated as described in Section RNA extraction and quantitative real-time PCR. Next, 2 μg of HDAC1 antibody (#10197-1-AP, Proteintech) was incubated with protein A/G magnetic beads (#HY-K0202, MedChemExpress) in NP40 buffer supplemented with PMSF and RNase inhibitor overnight at 4 °C. RNA was extracted from the beads and reverse-transcribed into cDNA for further qRT-PCR.
2.13 Orthotropic xenograft models
For xenograft models, 1 × 106 A2780-Vector or A2780-SFPQ cells were resuspended in 0.1 mL of PBS and subcutaneously inoculated into 5-week-old female BALB/c-nu mice (Guangdong Medical Laboratory Animal Center, Guangzhou, China), respectively. After 2 weeks of inoculation, carboplatin (10 mg/kg) was administered intraperitoneally every 3 days for 2 weeks. Tumor volumes were measured every 3–4 days. After 4 weeks of inoculation, all mice were euthanized, and the xenograft tumors were isolated, photographed, and weighed.
2.14 Statistics
Data from all experiments were presented as mean ± standard deviation (SD). Student’s t-test was utilized to compare the differences between two groups. One-way analysis of variance (ANOVA) was used to compare the differences among three or more groups. The numbers of independent experiments were indicated in the figure legends. ROC curve analysis was performed to determine the optimal cutoff values for predicting the prognosis. Pearson’s chi-square test and Fisher’s exact test were performed to determine the relationships between the clinical characteristics and the protein expression level of SFPQ. The Kaplan–Meier method was used to plot the survival curves and the Log-rank test was used to test for differences in survival between the groups. Cox proportional hazards regression analysis was used to determine the independent risk factors for clinical outcomes. All analyses were performed using IBM SPSS Statistics 25.0 (IBM, Armonk, NY, USA) and GraphPad Prism 8 (GraphPad Software Inc., San Diego, CA, USA). P < 0.05 was considered statistically significant.
3 Results
3.1 SFPQ is associated with platinum resistance and poor prognosis in ovarian cancer patients
By analyzing the GSE27651 dataset, it was found that the RNA expression level of SFPQ in ovarian cancer tissues was higher than that in normal ovarian epithelial tissues (P = 0.0041, Fig. 1A). To clarify the clinical significance of SFPQ in ovarian cancer, we performed immunohistochemical staining for SFPQ on 141 cases of ovarian cancer tissues, and summarized the clinicopathological characteristics of the patients (Table S1). Next, we conducted a receiver operating characteristic (ROC) curve based on the protein expression level of SFPQ, and divided the patients into a low SFPQ expression group (SFPQlow, n = 58) and a high SFPQ expression group (SFPQhigh, n = 83) according to the cut-off value. The analysis of the relationship between the protein expression level of SFPQ and clinicopathological parameters showed that a high SFPQ expression level was positively correlated with higher grade pathological differentiation and high recurrence risk (Table S2). Kaplan–Meier survival analysis and Log-rank test further indicated that both the overall survival (OS) and progression-free survival (PFS) of ovarian cancer patients in the SFPQhigh group were worse than those in the SFPQlow group (Fig. 1B). Univariate and multivariate Cox regression analyses showed that a high SFPQ expression level was an independent predictor of poor OS in ovarian cancer patients (Table S3). Next, according to the response to chemotherapy, patients were divided into the platinum-resistant group (n = 27) and the platinum-sensitive group (n = 114). In the platinum-resistant group, survival analysis indicated no significant difference in OS between patients with high and low SFPQ expression (Fig. S1A). However, in the platinum-sensitive group, prognosis analysis showed that patients with high SFPQ expression had worse outcomes in both OS and PFS compared to those with low SFPQ expression (Fig. S1B). These results suggest that in platinum-resistant patients, the predictive value of SFPQ protein expression levels for ovarian cancer prognosis is negative, which is likely due to the small sample size of the resistant group. In the platinum-sensitive group, higher SFPQ protein expression indicated a poorer prognosis. Therefore, we hypothesize that SFPQ may also be involved in the progression beyond chemoresistance in ovarian cancer. In addition, comparison of SFPQ protein expression between the resistant and sensitive groups revealed a statistically significant difference, with significantly higher expression levels in the platinum-resistant group (Fig. 1C). These results suggest that in ovarian cancer, a high expression of SFPQ indicates a poor prognosis and is closely related to recurrence, and it can be used as one of the biomarkers for predicting the prognosis of ovarian cancer patients.
3.2 High expression of SFPQ promotes platinum resistance in ovarian cancer
To verify the role of SFPQ in ovarian cancer, we first used the SFPQ-overexpressed plasmid to alter the protein expression level of SFPQ in A2780 and Skov3 cells (Figs. 2A and S2A). Cells were treated with different concentrations of carboplatin, and the CCK-8 assay was performed to detect the drug sensitivity of the cells. The results showed that the IC50 of two cell lines increased compared with that of the control group after SFPQ overexpression (P = 0.0384, Fig. 2B; P = 0.0141, Fig. S2B). The flow cytometry assay showed that the apoptosis rate of cancer cells decreased significantly (Figs. 2C and S2C). To further confirm the function of SFPQ, we used A2780-SFPQ cell to perform subcutaneous implantation in nude mice and observed the growth of the transplanted tumors. As shown in Fig. 2D and 2E, after carboplatin administration, the tumors of A2780-SFPQ group were significantly larger than those in the control group. In summary, these results indicated that SFPQ induced platinum resistance in ovarian cancer cells, and silencing SFPQ could enhance the platinum sensitivity of ovarian cancer cells.
3.3 SFPQ interacts with BRCA1, and SFPQ promotes the protein expression of BRCA1 by inhibiting its ubiquitination and degradation
Next, we wondered how SFPQ exerted its function of promoting platinum resistance in ovarian cancer. It is known that BRCA1 plays a very important role in the occurrence and development of ovarian cancer. In our previous study, we used two BRCA1 wild-type ovarian cancer cell lines, A2780 and Skov3, to perform proteomic sequencing, screened out the proteins that bind to BRCA1 with high abundance, and intersected them with the proteins predicted to interact with BRCA1 on the BioGRID website. A total of 83 genes were screened out (Fig. 3A). Then, we used Co-IP experiments to verify and confirm the combination between SFPQ and BRCA1 (Figs. 3B and S3). We also used immunohistochemistry to analyze the protein expression level of BRCA1 in the same ovarian cancer tissues, which showed that the protein level of BRCA1 was positively correlated with that of SFPQ (Fig. 3C). Besides, Gene Set Enrichment Analysis (GSEA) of signaling pathway suggested that SFPQ participated in the DNA double-strand break repair signaling pathway (Fig. 3D). By knocking down or overexpressing SFPQ, we found that the protein level of BRCA1 changed alone with that of SFPQ (Fig. 4A). Furthermore, when BRCA1 was knocked down in cells with SFPQ overexpression, the increase in the IC50 of ovarian cancer cells was reversed, and the apoptosis rate of cancer cells increased again (Figs. 4B, 4C, S2D, and S2E). The above results indicate that BRCA1 is a downstream molecule of SFPQ, and inhibiting BRCA1 could weaken the effect of SFPQ in promoting platinum resistance in ovarian cancer.
According to BRCA1 molecular structure, we constructed four truncated BRCA1 plasmids, each harboring mutations in one of the four domains: RING, NLS, BRCT1, and BRCT2 (Figs. S4 and S5A). Co-IP assays revealed that the interaction between SFPQ and BRCA1 was disrupted when either the BRCT1 or BRCT2 domain was mutated (Fig. S5B). After mutating the BRCA1 BRCT domains, overexpression of SFPQ failed to significantly increase BRCA1 protein level, and enhanced the sensitivity to platinum-based drugs (Fig. S5C and S5D). These results suggest that SFPQ binds to the BRCT domains of BRCA1, and mutation of this binding region reverses the platinum sensitivity of ovarian cancer cells.
However, we found that SFPQ did not significantly affect the mRNA level of BRCA1, suggesting that SFPQ might regulate the protein stability of BRCA1 (Fig. 5A). As is well known, ubiquitination is the most common pathway for protein regulation. Therefore, we speculated whether SFPQ affects the ubiquitination of BRCA1. We used the proteasome inhibitor MG132 and observed that after knocking down SFPQ, the ability of inhibiting the expression of BRCA1 was weakened (Fig. 5B). At the same time, the protein synthesis inhibitor, CHX, was used to detect the degradation rate of BRCA1. We found that after overexpressing SFPQ, the protein degradation rate of BRCA1 slowed down (Fig. 5C). Then, we co-transfected the Flag-BRCA1 and His-Ub plasmids or si-SFPQ into HEK293T cells and found that the ubiquitination level of BRCA1 was enhanced after knocking down SFPQ (Fig. 5D). These results indicate that SFPQ binds to BRCA1 and inhibits the ubiquitination and degradation of BRCA1, enhances the protein stability of BRCA1, increases its protein expression level to strengthen the DNA damage repair ability of ovarian cancer cells, and induces the occurrence of platinum resistance.
3.4 HDAC1 modifies the H3K27Ac region of the SFPQ promoter through deacetylation to regulate its expression
Due to the lack of SFPQ inhibitors, we explored the upstream mechanism regulating SFPQ expression to extend our research findings to clinical applications. It is known that epigenetics plays an important role in gene regulation, and the acetylation modification of chromatin histones is one of the important drivers of gene expression regulation and is currently a research hotspot [
26]. We first evaluated the modification of the
SFPQ promoter in the UCSC Genome Browse database. As shown in Fig. 6A, the
SFPQ promoter region is highly enriched with acetylated histone H3K27 (H3K27Ac), indicating that the expression of SFPQ is regulated by chromatin acetylation. It is known that HDAC1 is one of the members of the histone deacetylase family, and it can deacetylate the lysine of histones, increase the positive charge of histone tails, and promote the high-affinity binding between histones and the DNA, thereby regulating gene transcription and translation [
27]. In 2022, Ding
et al. reported that DNTTIP1 maintained the deacetylation of H3K27 on the
DUSP2 promoter by recruiting HDAC1, thus inhibiting the expression of DUSP2 and promoting the invasion and metastasis of nasopharyngeal carcinoma [
28]. Another report showed that in prostate cancer cells, KLF5 bound with HDAC1 and recruited it to the promoter of
IGF1, inhibiting the transcription of IGF1 and participating in the occurrence and development of prostate cancer [
29]. Previous studies have confirmed that the histone deacetylase HDAC1 deacetylates H3K27, thereby downregulating the transcriptional activity of target genes [
30]. Through the GSE18520 dataset, the results showed that there was a negative correlation between the mRNA expression levels of SFPQ and HDAC1 in ovarian cancer tissues (Fig. 6B). Therefore, we speculated that HDAC1 may regulate the expression of SFPQ in ovarian cancer through epigenetic modification. We detected the mRNA level and protein expression of
SFPQ in ovarian cancer cell lines after knocking down the expression of HDAC1, and found that both the mRNA level and the protein expression of
SFPQ were upregulated (Fig. 6C and 6D). These results indicate that the deacetylation modification of H3K27 on the
SFPQ promoter mediated by HDAC1 partially elucidates the mechanism of the overexpression of SFPQ in ovarian cancer.
3.5 hsa_circSFPQ_008 binds HDAC1 to deacetylate and modify the SFPQ promoter
However, the mechanism that regulates the modification function of HDAC1 has not been fully clarified. A large number of studies have demonstrated that although most of the parent-homologous non-coding RNAs do not have the function of encoding proteins, they play a crucial role in the proteins encoded by their parent genes, consisting of the growth and development of organisms, signal responses, and the occurrence and development of diseases [
31–
33]. It is known that circRNAs can perform the following functions, including adsorbing miRNAs competitively, regulating alternative splicing of transcripts, serving as templates for translating and encoding proteins, interacting with RNA binding proteins, and regulating transcription by interacting with RNA polymerase II directly or indirectly [
31]. In 2021, Ma
et al. reported that in breast cancer, the antisense circSCRIB promotes cancer development by inhibiting the splicing and translation of the pre-mRNA of the parent gene
SCRIB [
34]. Another research showed that circEIF3J and circPAIP2 promoted the transcription and expression of the parent genes
EIF3J and
PAIP2 by interacting with U1 small nuclear ribonucleoproteins (snRNPs) and RNA polymerase II (Pol II) at the promoters of the parent genes
EIF3J and
PAIP2 [
35]. In this study, we wondered whether HDAC1 was recruited by the non-coding RNA homologous to the parent of SFPQ to regulate the expression of SFPQ. First, we screened the circRNAs derived from the parent of SFPQ, and determined that HDAC1 bound to circSFPQ_008 through the RIP experiment (Fig. 7A). We also detected the expression level of circSFPQ_008 in a total of 82 cases of ovarian cancer and normal ovarian epithelial tissues, and found that the expression level of circSFPQ_008 in ovarian cancer tissues was lower than that in normal ovarian tissues (
P = 0.0288, Fig. 7B). After inhibiting circSFPQ_008, the protein expression levels of SFPQ and BRCA1 and the acetylation level of H3K27 were all upregulated, the protein expression level of HDAC1 was downregulated, the IC
50 of platinum in ovarian cancer cells increased, and the chemosensitivity decreased. When HDAC1 was overexpressed after inhibiting circSFPQ_008, the acetylation level of H3K27 was inhibited, the protein levels of SFPQ and BRCA1 were downregulated, and the chemosensitivity of ovarian cancer cells was enhanced (Fig. 7C and 7D). In platinum-resistant ovarian cancer cells A2780-PR (Fig. S6A), we overexpressed circSFPQ_008 or knocked down SFPQ expression in A2780-PR cells, the results showed that the IC
50 values of A2780-PR cells decreased, and the proportion of apoptotic ovarian cancer cells increased (Fig. S6B–S6D). These data suggest that circSFPQ_008 regulates the protein expression of SFPQ and BRCA1 by binding to HDAC1, thus participating in platinum resistance in ovarian cancer.
4 Discussion
Ovarian cancer has the highest mortality rate among gynecological tumors, mainly because it is usually diagnosed at an advanced stage and evolves into chemotherapy resistance under the frequent recurrence-treatment pattern. Platinum-based drugs, especially cisplatin and carboplatin, are the main chemotherapeutic agents for epithelial ovarian cancer [
10]. However, most patients will encounter platinum resistance. Therefore, it is crucial to elucidate the molecular mechanisms of platinum resistance and to improve the survival rate of ovarian cancer patients. In this study, we observed that the high expression of SFPQ was significantly correlated with the resistance to platinum-based chemotherapy and poor prognosis in ovarian cancer patients, which implies that SFPQ should be a target for overcoming platinum resistance.
Recently, SFPQ has been demonstrated to play a complex role in tumor evolution. First, some researchers have found that SFPQ directly binds to Smad4 and inhibits its transcriptional activity through phase separation, thereby interrupting the tumor-suppressive function of the TGF-β signaling pathway and promoting the progression of liver cancer [
36]. In contrast, a study has found that microRNA-1296 promotes the invasion and migration of colorectal cancer by targeting SFPQ, in which SFPQ exerts a tumor-suppressive effect. Colorectal cancer patients with high SFPQ expression have a better prognosis [
37]. Thus, it is necessary to conduct extensive research on SFPQ to understand its functions in malignant tumors. In this study, our results indicate that SFPQ is an independent predictor of poor prognosis in ovarian cancer, and high SFPQ expression is associated with the resistance of ovarian cancer cells to platinum-based chemotherapy. Additionally, we have found both
in vitro and
in vivo that inhibiting the expression of SFPQ can enhance the sensitivity of ovarian cancer cells to platinum, suggesting that SFPQ can serve as a research target for reversing platinum resistance.
To the best of our knowledge, this is the first report demonstrating that SFPQ promotes platinum resistance by regulating the expression of BRCA1, which is completely different from its function as a DNA/RNA binding protein in other studies. In this project, SFPQ have been confirmed to promote platinum resistance by inhibiting the degradation of BRCA1 and enhancing DNA damage repair. Our study reveals the functional diversity of SFPQ and provides new insights into the exploration of SFPQ.
Platinum attacks tumor cells by causing DNA cross-link breaks, including single-strand breaks (SSB) and double-strand breaks (DSB). SSB is mainly repaired by PARP, which recruits downstream proteins to participate in the DNA repair process [
38]. For DSB, HRR is the most precise repairment, and BRCA1 plays a crucial role in this process [
39,
40]. Theoretically, ovarian cancer patients with
BRCA1 mutations (
BRCA1MUT) have deficiency in the DSB repair process. PARP inhibitors take advantage of this self-repair vulnerability to ultimately induce tumor cell death [
41–
43]. In fact, clinical studies have shown that ovarian cancer patients with
BRCA1MUT benefit most significantly from PARP inhibitors [
41]. However,
BRCA wild-type (
BRCAWT) patients account for approximately 80% of all ovarian cancer cases. How to improve the clinical efficacy of these patients is a crucial issue. In this study, inhibiting SFPQ can effectively induce BRCA1 functional deficiency and can effectively create a BRCAness state. Therefore, the combined use of PARP inhibitors and targeted inhibition of SFPQ could achieve a synthetic lethal effect in
BRCA1WT ovarian cancer cells. It is worth noting that these findings provide a novel treatment strategy for ovarian cancer patients with
BRCA1WT who have a poor response to platinum and PARP inhibitors. In addition, previous reports have suggested that the spontaneous reversion of
BRCA mutations restores its ORF, which may be one of the mechanisms of platinum resistance in ovarian cancer patients with
BRCA1MUT [
44,
45]. In these patients, if a BRCAness state can be achieved to weaken the DNA damage repair ability of BRCA through inhibiting SFPQ expression, the chemosensitivity of tumor cells can be reversed.
Currently, there is a lack of inhibitors that suppresses SFPQ specifically. Therefore, we expanded the exploration of the upstream mechanisms of SFPQ and found that circSFPQ_008 recruits HDAC1 to deacetylate and modify the H3K27Ac of the
SFPQ promoter, which downregulates its expression, promotes the ubiquitination and degradation of BRCA1, reduces the expression of BRCA1, and weakens the DNA damage repair function of ovarian cancer cells. Nowadays, the biological functions of circRNAs are very diverse. CircRNAs mediate the occurrence and development of tumors through various molecular mechanisms. For example, circRNAs could competitively bind to RNAs (such as microRNAs, miRNAs, and miRs), bind to proteins, regulate transcription and splicing, etc. A few circRNAs could be translated into peptides or proteins under specific conditions [
46,
47]. In some reports, circRNAs are able to classify and transfer proteins to specific locations, and participate in regulating protein–protein and protein–RNA interactions [
47]. In addition, the combination between circRNAs and proteins may have a two-way effect, regulating both the synthesis and degradation of themselves [
48]. In this study, we have discovered for the first time that there is a combination between circSFPQ_008 and HDAC1. However, the specific binding site has not been fully clarified, and it is also worthy of further investigation whether HDAC1 can inversely affect the expression of circSFPQ_008. Another noteworthy point is that previous research has shown that using targeted lipid nanoparticles (tLNPs) to package circRNAs and verifying their potential to effectively improve the treatment effect through
in vivo experiments using mice and other models [
49]. These pioneering studies have provided a foundation for achieving personalized and precise treatment in the future.
Based on the current progress of clinical research, HDAC1 inhibitors have demonstrated favorable anti-tumor efficacy primarily in hematological malignancies, whereas their effectiveness in solid tumors remains suboptimal. In 2024, a multicenter clinical study on peripheral T-cell lymphomas (PTCLs) showed that the combination of the HDAC1 inhibitor chidamide with chemotherapy achieved an overall response rate (ORR) of 53.3% in relapsed or refractory PTCLs [
50]. In ovarian cancer, as early as 2012, researchers evaluated the efficacy of the HDAC1 inhibitor belinostat combined with carboplatin in patients with platinum-resistant ovarian cancer. However, the results showed that the ORR was only 7.4% [
51]. In colorectal cancer and breast cancer, it has been reported that HDAC1 inhibitors epigenetically upregulated CD47 expression in tumor cells, thereby inhibiting the macrophage-mediated killing, leading to a tumor-promoting effect of HDAC1 inhibitors [
52]. Furthermore, the latest research from September 2025 revealed a regulatory mechanism of inositol phosphate on the efficacy of HDAC inhibitors, indicating that specific multiprotein complexes and intracellular molecules such as inositol phosphate, could influence the sensitivity and selectivity of HDAC inhibitors [
53]. These findings suggest that the mechanism and efficacy of HDAC inhibitors are modulated by multiple factors, which ultimately influence the progression of tumor cells. In the present study, suppressing HDAC1 expression increased the protein expression level of SFPQ in ovarian cancer cells, leading to enhancement of platinum resistance. This may be one of the reasons for the limited efficacy of HDAC1 inhibitors in solid tumors. Based on these findings, we suggest incorporating SFPQ expression level as a stratification biomarker. For ovarian cancer patients with SFPQ deficiency, the anti-tumor effect of HDAC1 inhibitor combined with platinum-based drugs may be more effective. Secondly, screening for selective HDAC1 inhibitors that do not regulate SFPQ expression, potentially through modifying the intracellular microenvironment, may improve the therapeutic efficacy of HDAC1 inhibitor and platinum combination therapy in ovarian cancer.
In conclusion, this study has revealed a novel molecular mechanism of the circSFPQ_008/SFPQ/BRCA1 axis in inducing platinum resistance in ovarian cancer (Fig. 8). These findings contribute to the understanding of the potential role of SFPQ in platinum resistance and also provide a theoretical basis for the application of the combination of PARP inhibitors and SFPQ inhibition as personalized treatment for ovarian cancer patients.
Our study also has limitations. First, SFPQ is regulated by circSFPQ_008, mediates the expression of BRCA1, and enhances platinum resistance in ovarian cancer. Currently, there is only one research stating that SFPQ binds to the p54
nrb complex and regulates the activity of the splicing factor SRSF2, reducing the binding of SRSF2 to caspase-9 RNA, which is conducive to the expression of its anti-apoptotic form, and protects cells from platinum-induced death, and ultimately promotes chemoresistance in ovarian cancer [
54]. Besides, it is still unclear whether SFPQ has other different functions in ovarian cancer. Moreover, whether SFPQ plays a similar role in other cancers, especially breast cancer, remains to be investigated. Finally, based on our research, we suggest that the combination of PARP inhibitors and the inhibition of SFPQ could achieve a more effective therapeutic effect in ovarian cancer. This theory requires more experimental evidence and preclinical studies to confirm and support.
4.0.0.0.1 Acknowledgements
This study was funded by the Guangdong Medical Science and Technology Research (No. A2023337), the Guangzhou Science and technology project (Nos. 2024A03J0182 and 2023A03J0375), the Clinical High-tech and Major Technology Projects in Guangzhou (No. 2024PL-ZD06), the Guangzhou Medical University Clinical Research (No. GMUCR2024-01004), and the Natural Science Foundation of Xizang Autonomous Region (No. XZZR202402076(W)).
PDF
(3472KB)
