Background

Ovarian cancer is the deadliest gynaecological cancer, with a 5-year survival rate of approximately 30% [1, 2]. Due to the high heterogeneity and complex pathogenesis of ovarian cancer, conventional chemotherapies are far from perfect due to their high toxicity, insufficient efficacy, as well as the risk of drug resistance. This highlights the urgent need for the development of novel therapeutic strategies to address the unmet needs in ovarian cancer treatment.

Poly (ADP-ribose) polymerase (PARP) inhibitors (PARPis), a novel class of targeted drugs, show remarkable efficacy against ovarian cancer by specifically targeting PARPs, key enzymes involved in DNA damage repair. This breakthrough in targeted therapy has transformed the concept of “synthetic lethality” from theory into clinical reality. Notably, several PARPis have gained approval for ovarian cancer treatment, including olaparib, rucaparib, niraparib, fluzoparib, pamiparib, etc. [3,4,5]. These PARP inhibitors are actively tested preclinically and clinically in homologous recombination (HR) deficiency (HRD)-positive cancers, in which they may cause synthetic lethality and thus achieve better efficacies [6,7,8]. However, approximately 70% of ovarian cancers are HRD-negative, which poses a significant challenge to the widespread clinical use of PARP inhibitors.

PTEN is a canonical dual-specificity phosphatase that dephosphorylates lipids and proteins [9]. It also plays crucial roles in regulating chromosome stabilities, cell cycle arrest, and DNA repair [10,11,12]. Intriguingly, several studies have demonstrated that PTEN deficiency or mutations in tumour cells may cause impaired DNA double-strand break (DSB) repair, rendering them more susceptible to PARP inhibitors both in vitro and in vivo [13,14,15].

Till now, PTEN, as a well-established tumour suppressor [16, 17], has rarely been attempted as a potential therapeutic target for cancer treatment. However, in our preliminary study, we discovered that the PTEN inhibitor (PTENi) VO-OHpic exhibited excellent anti-cancer activity against ovarian cancer cells, while also increasing DSBs within these cells. These encouraging findings prompted us to investigate the possibility of intentionally inducing HRD in ovarian cancer cells through PTEN inhibition, thereby broadening the potential application of PARPis. Subsequently, we conducted further research to elucidate the underlying mechanisms behind these observed phenomena.

Methods

Cell lines

The ovarian surface epithelial cell line HOSEpiC and five human ovarian cancer cell lines (SKOV3, HO8910, OVCAR5, OVCAR8, and ES2) were obtained from the American Type Culture Collection (ATCC, Gaithersburg, MD, USA), and cultured in DMEM or RPMI-1640 (Thermo Fisher Scientific, Waltham, MA, USA) with 10% fetal bovine serum (FBS; Thermo Fisher Scientific) and penicillin/ streptomycin (100 U/ml and 100 μg/ml, respectively; Thermo Fisher Scientific). All cells were maintained at 37 °C and 5% CO2 atmosphere, and tested for mycoplasma contamination before experiments.

Antibodies and reagents

The antibodies used in the current study were listed in Table S2. VO-OHpic and Olaparib were purchased from Selleck Chemicals (Houston, TX, USA). All other chemicals were purchased from Sinopharm (Shanghai, China) and Sigma-Aldrich (St. Louis, MO, USA) unless otherwise specified.

Cell viability and colony formation assays

MTT assays were performed as previously described [18]. IC50 values were calculated using the dose-response Inhibition model ([Inhibitor] vs. normalised response-variable slope) via GraphPad Prism (San Diego, CA, USA; version 9.5). For colony formation assays, cells were plated in 6-well plates at 1000 cells/well, and incubated at 37 °C and 5% CO2 for 2 weeks. After that, cells were stained with 0.1% crystal violet (Beyotime, Shanghai, China) and photographed with a digital camera (MS60, Mshot, Guangdong, China). Each experiment was repeated at least three times independently.

Drug synergy testing

MTT assays were used to determine cell viabilities. Cells were plated in 96-well plates at 4000 cells/well in triplicate. After attached, cells were treated with combination of olaparib and VO-OHpic as indicated. Cell viabilities were measured 72 h later. Combination index (CI) values were calculated by the Chou-Talalay method using the CompuSyn software package (ComboSyn, Paramus, NJ, USA). Each experiment was repeated at least three times independently.

Immunofluorescence assays

Immunofluorescence assays were performed as previously described [19]. SKOV3 and ES2 cells were treated with VO-OHpic or doxorubicin or both as indicated. After that, cells were fixed with freshly prepared 4% paraformaldehyde at ambient temperature for 10 min, permeated with 0.5% Triton X-100 for 5 min, and blocked with 3% of BSA. Next, cells were incubated with primary antibodies against γH2AX, 53BP1, MRE11, RAD50, PTEN or E2F1 in 3% of BSA, subsequently incubated with secondary antibodies, and then stained with 1 μg/μl of DAPI, and observed using Nikon Eclipse fluorescence microscope (Tokyo, Japan). Stained cells (with >5 foci per cell) in five different high-power fields from each slide were quantified as positive cells for γH2AX or 53BP1 foci. The fluorescence intensity and colocalization analysis was performed using Image-Pro Plus software (Rockville, MD, USA). The scale bars represented 10 μm. Each experiment was repeated at least three times independently.

Western blot

Western blot experiments were performed as previously described [19, 20]. Briefly, whole protein lysates were obtained using ice-cold RIPA lysis buffer and centrifuged at 10,000 × g for 1 min at 4 °C; whereas nuclear proteins were extracted using NE-PERTM reagent following the manufacture’s manual (Thermo Fisher Scientific). Protein concentrations were quantified using the BCA protein assay kit (Beyotime). Thirty μg of protein samples were resolved on 8–12% SDS-PAGE gels, and then transferred to PVDF membranes (Millipore, Billerica, MA, USA). Blots were visualised with ECL plus chemiluminescence kit (Thermo Fisher Scientific). To extract cytoplasmic and nuclear proteins separately, the experiments were performed as previously described [21]. Briefly, cells were lysed with the buffer (10 mM HEPES–KOH (pH = 7.5), 10 mM KCl, 1.5 mM MgCl2, 0.34 M sucrose, 10% glycerol, 0.1% Triton X-100) for 5 min, and the supernatants (cytoplasmic proteins) were collected after centrifugation at 1300 × g for 5 min at 4 °C. Then, the pellets were resuspended and lysed with the lysis buffer (50 mM Tris–HCl (pH 7.4), 150 mM NaCl, 1 mM PMSF, 1 mM EDTA, 1% Triton X-100, 1% Halt Phosphatase Inhibitor Cocktail EDTA free 100X) for 10 min, and the supernatants (nuclear proteins) were collected after centrifugation at 10,000 × g for 1 min at 4 °C. Each experiment was repeated at least three times independently.

Cell apoptosis and cell cycle assays

Cell apoptosis and cell cycle assays were performed as previously described [18]. Cells were seeded in 6-well plates with 4 × 105 cells/well, and then exposed to the indicated concentration of VO-OHpic for 48 h for apoptosis assays or 24 h for cell cycle assays. According to the cell viability curves, 1, 7, and 20 μM of VO-OHpic were used to treat SKOV3 cells, while 10, 25, and 60 μM of VO-OHpic were used to treat ES2 cells. Cells were stained with Annexin V-FITC/PI detection kit (KeyGen Biotech, Nanjing, Jiangsu, China) for apoptosis analysis, and were stained with propidium iodide (PI, Sigma-Aldrich) for cell cycle assays. Experiments were performed in the flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA) and data were analysed using FlowJo software (Ashland, OR, USA). Each experiment was repeated at least three times independently.

Correlation analyses using GEPIA, Kaplan–Meier plotter, and UALCAN

Correlations between expression of PTEN and key genes in KEGG HR signalling pathway hsa03440 in clinical tissue specimens were analysed in TCGA and GTEx datasets using GEPIA platform (http://gepia2.cancer-pku.cn/#index) [22]. Correlation coefficient (R) > 0.3 and p value < 0.05 were set as the significance thresholds. Correlations between expression of MRN complex-related genes and prognosis of ovarian cancer patients were analysed using Kaplan–Meier plotter (http://kmplot.com/analysis/) with logrank p value < 0.05 set as the significance threshold [23]. Differential expression of MRN complex-related genes in ovarian cancer specimens versus normal ones was analysed using CPTAC proteomics dataset from UALCAN database (http://ualcan.path.uab.edu/) [24], and p value < 0.05 was set as the significance threshold.

Quantitative real-time PCR (RT-PCR)

qRT-PCR experiments were performed as previously described [20]. Total RNAs were extracted using TRIzol reagent (Thermo Fisher Scientific). Then reverse transcription was performed using the HiScript™ Reverse Transcriptase kit, and amplification was carried out using AceQ qPCR kit. mRNA levels of MRE11 and RAD50 were quantified using qRT-PCR with SYBR Green mix in QuantStudio 3 Real-time PCR System (Thermo Fisher Scientific). All above kits were purchased from Vazyme (Nanjing, Jiangsu, China) unless otherwise specified. β-actin was used as the internal control. The primers used were as follows: MRE11 (forward: 5’-CAGCAACCAACAAAGGAAGAGGC-3’; reverse: 5’-GAGTTCCTGCTACGGGTAGAAG-3’); RAD50 (forward: 5’-GGAAGAG- CAGTTGTCCAGTTACG-3’; reverse: 5’-GAGTAAACTGCTGTGGCTCCAG-3’); β-actin (forward: 5’-AGAGCTACGAGCTGCCTGAC-3’; reverse: 5’-AGCACTGTGTTGGCGTACAG-3’). Each experiment was repeated at least three times independently.

Construction and transduction of shRNAs targeting PTEN

Two shRNAs targeting PTEN, including shPTEN-1 and shPTEN-2, were cloned into lentivirus vector pLKO.1-TRC control (shControl, Addgene, Cambridge, MA, USA). The primers for shPTEN-1 were 5’-CCGGGTCTGCCAGCTAAAGGTGAAGATATACTCGA-GTATATCTTCACCTTTAGCTGGCAGACTTTTTG-3’ (forward primer) and 5’-AATTC-AAAAAGTCTGCCAGCTAAAGGTGAAGATATACTCGAGTATATCTTCACCTTTAGCTGGCAGAC-3’ (reverse primer); the primers for shPTEN-2 were 5’-CCGGGAGCGTGCAGATAATGACAAGGAATACTCGAGTATTCCTTGTCATTATC- TGCACGCTCTTTTTG-3’ (forward primer) and 5’-AATTCAAAAAGAGCGTGCAGAT- AATGACAAGGAATACTCGAGTATTCCTTGTCATTATCTGCACGCTC-3’ (reverse primer). Plasmids of shControl, shPTEN-1 and shPTEN-2 were transfected into 293T cells to generate lentivirus particles, and then these particles were collected and transduced into SKOV3 and ES2 cells using 5 µg/ml polybrene. After that, 1 µg/ml puromycin was used for screening cells which were stably transducted with lentiviruses.

Chromatin immunoprecipitation (ChIP) assays

Chromatin immunoprecipitation in SKOV3 and ES2 cells transducted with shControl, shPTEN-1 or shPTEN-2 lentivirus particles was performed at day 10 as previously described [25] using a monoclonal anti-E2F1 antibody or an isotype-matched IgG control. Similarly, chromatin immunoprecipitation in SKOV3 and ES2 cells treated with vehicle or VO-OHpic was performed after 48 h of drug treatment. Immunoprecipitated DNA was analysed using SYBR green-based quantitative PCR (Vazyme) against the human RAD50 gene promoter region containing E2F1 binding sites using the forward primer 5’-GCTTCCCTGTTCCCGAGTC-3’ and reverse primer 5’-CGGAACTTGCGAAAGACGC-3’. Immunoprecipitated DNA was also analysed against the human MRE11 gene promoter region containing E2F1 binding sites using the forward primer 5’-GGCGCCATCTAAGAACTCCT-3’ and reverse primer 5’-CTTCCAAACAGCCAAGAGCG-3’. Each experiment was repeated at least three times independently.

Mouse models

All procedures involving mice were approved by the Institutional Animal Care and Use Committee of Jiangsu University. Six-week-old nude BALB/c female mice (7 mice/group) were obtained from GemPharmatech (Nanjing, Jiangsu, China), acclimatised on a standard diet for 1 week before the experiments, and randomly grouped. VO-OHpic was dissolved in 10/18 DRD solution (10% DMSO, 18% Cremophor RH 40, 3.6% dextrose, and 68.4% water) before injection; olaparib was dissolved in a solution containing 4% DMSO, 30% PEG300, and 66% water. For xenograft mouse models, 2 × 106 ES2 cells were injected subcutaneously in the middle of the left flank of mice. On the 7th day, tumour-bearing mice were randomly divided into four groups with seven mice per group, and then administrated with either vehicle, 40 mg/kg of VO-OHpic every other day by tail vein injection, or 200 mg/kg of olaparib every day by intraperitoneal injection, or both VO-OHpic and olaparib at the above mentioned doses and times for single reagents, and then sacrificed on the 21st day. Tumour xenografts and organs were removed and weighted; and tumour tissues were collected for immunohistochemistry (IHC) staining.

Immunohistochemistry (IHC) assays

IHC assays were performed as already described [19]. Tissue sections were dewaxed with xylene, then hydrated and incubated with citrated buffer solution (pH = 6.0), subsequently immersed in 3% H2O2, and then incubated with primary antibodies anti-γH2AX or anti-53BP1. After that, the sections were incubated with secondary antibodies, and visualised using 3-3-diamiobenzidine and counterstained using haematoxylin. Expression of γH2AX and 53BP1 in situ were observed using a bright field microscope and calculated using histological score (H-score). IHC staining was independently scored by two pathologists blinded to the tissue groups. The percentage of positive cells ranged from 0–100% was scored as 0 (0%), 1 (1–25%), 2 (26–50%), 3 (51–75%), and 4 (76–100%); and the intensity of staining varied from weak to strong was scored as 0 (negative), 1 (weak, light yellow), 2 (moderate, yellowish brown), and 3 (strong, brown). Each experiment was repeated at least three times independently.

Luciferase assay

Luciferase assays were performed as previously described [26]. The proximal promoter regions of RAD50 and MRE11 were synthesised and cloned into pGL3 Luciferase Reporter Vector (Promega, Madison, WI, USA). SKOV3 or ES2 cells were transfected with pGL3-derived plasmids or vectors together with plasmids of pcDNA3.1-E2F1 or vectors, or cells were transfected with pGL3-derived plasmids or vectors together with/without VO-OHpic treatment as indicated. After 48 h, luciferase activities were assayed. The relative luciferase activities were obtained by normalising the values against those of the control group as indicated. Each experiment was repeated at least three times independently.

Statistical analysis

Data were presented as mean ± standard deviation (SD), unless otherwise stated. Statistically significant differences were analysed between two groups by Student’s t tests (two-tailed) and were analysed among multiple groups by one-way ANOVA with Tukey multiple comparison test using GraphPad Prism (San Diego, CA, USA; version 9.5), unless otherwise specified. p < 0.05 was considered statistically significant (*p < 0.05; **p < 0.01; ***p < 0.001).

Results

Combination of PTENi VO-OHpic with PARPi olaparib synergistically inhibits proliferation of ovarian cancer cells

To evaluate the effects of PTEN inhibition on ovarian cancer, a small molecule inhibitor of PTEN, VO-OHpic, was used to treat the normal ovarian surface epithelial cell line HOSEpiC and four epithelial ovarian cancer cell lines, including serous ovarian cancer cell lines SKOV3, OVCAR8, OVCAR5 and clear cell ovarian cancer cell line ES2. Results demonstrated significant inhibition on the growth of these cells with VO-OHpic treatment in a dose-dependent manner (Fig. 1a). Among these ovarian cancer cell lines, SKOV3 was the most sensitive one to VO-OHpic with IC50 at 7.15 μM; whereas ES2 was the least sensitive one with IC50 at 25.24 μM (Fig. 1a). And their IC50 values were all significantly lower than that of the normal ovarian surface epithelial cell line HOSEpiC. Furthermore, long-term effects of VO-OHpic on viability of ovarian cancer cells were tested at their corresponding IC25 values from Fig. 1a using colony formation assays and results indicated that VO-OHpic greatly reduced the viability of all four ovarian cancer cell lines for a prolonged period of time (Fig. 1b). These data suggested that VO-OHpic significantly suppressed proliferation of multiple ovarian cancer cell lines. To find out how VO-OHpic affected proliferation, cell-cycle distribution of SKOV3 and ES2 cells after VO-OHpic treatment was analysed. As shown in Fig. S1, exposure of these cells to VO-OHpic resulted in a significant increase of G2/M-phase cells, which was accompanied by a decrease in G1/G0-phase cells. Additionally, the results showed that there was a significant increase in the percentages of apoptotic cells in both early and late apoptosis stages with VO-OHpic treatment in both SKOV3 (Fig. S2A, B) and ES2 cells (Fig. S2C, D).

Fig. 1: Combination of PTENi VO-OHpic with PARPi olaparib synergistically inhibited proliferation of ovarian cancer cells.
figure 1

a Ovarian cancer cell lines and the normal ovarian surface epithelial cell line HOSEpiC were treated with VO-OHpic at the indicated concentration for 72 h, and cell viabilities were evaluated by MTT assays. IC50 and IC25 values were calculated using the dose-response Inhibition model ([Inhibitor] vs. normalised response-variable slope) via GraphPad Prism (version 9.5). b Ovarian cancer cell lines and the normal ovarian surface epithelial cell line HOSEpiC were treated with VO-OHpic at their corresponding IC25 values from (a) for 15 days, respectively, and stained with crystal violet. c Ovarian cancer cell lines and the normal ovarian surface epithelial cell line HOSEpiC were treated with olaparib at the indicated concentrations for 72 h, and cell viabilities were evaluated by MTT assays. IC50 and IC25 values were calculated using the dose-response Inhibition model ([Inhibitor] vs. normalised response-variable slope) via GraphPad Prism (version 9.5). d Ovarian cancer cell lines and the normal ovarian surface epithelial cell line HOSEpiC were treated with olaparib at their corresponding IC25 values from (c) for 15 days, respectively, and stained with crystal violet. e ES2 cells were treated at the indicated concentrations of PTENi VO-OHpic and PARPis olaparib, rucaparib, or veliparib for 72 h, and cell viabilities were evaluated by MTT assays. Data shown are the mean values (±S.D.) from three independent experiments. Statistically significant differences were analysed among multiple groups in (a) and (c) by one-way ANOVA with Tukey multiple comparison test and were analysed between two groups in (b) and (d) by Student’s t tests (two-tailed) using GraphPad Prism (version 9.5) with p < 0.05 were considered significant (#p > 0.05, *p < 0.05, ***p < 0.001).

Full size image

Olaparib was the first PARPi approved for ovarian cancer treatment, and has been approved for patients with BRCA mutations and actively tested in patients with HRD [27, 28]. MTT results showed that olaparib significantly reduced viabilities of multiple ovarian cancer cell lines, including SKOV3, OVCAR5, OVCAR8, and ES2 (Fig. 1c). Among them, SKOV3 was the most sensitive one to olaparib, with IC50 at 8.12 μM; whereas ES2 was the least sensitive one to olaparib, with IC50 at 31.29 μM (Fig. 1c). And their IC50 values were all significantly lower than that of the normal ovarian surface epithelial cell line HOSEpiC. Furthermore, results from colony formation assays demonstrated that olaparib greatly decreased colony formation capabilities of all four ovarian cancer cell lines (Fig. 1d).

In addition to monotherapy, combination strategies using PARPis with other agents are actively evaluated pre-clinically and clinically to improve their efficacies combating various cancers [29]. Accordingly, we wanted to explore the inhibitory effects of PARPis on ovarian cancer cells for the potential combination therapy together with VO-OHpic. Since ES2 was the least sensitive cell line to either VO-OHpic or olaparib as monotherapy, ES2 cells were chosen to be evaluated to take advantages of the combination therapies. MTT results showed that the combination of VO-OHpic with PARPis olaparib, rucaparib, and veliparib achieved synergistic inhibitory effects on ES2 cells, with much better efficacy than those of monotherapies with VO-OHpic or PARPis alone (Fig. 1e and Table S1). The calculated CI values of the combination therapies smaller than 1 indicated synergism between VO-OHpic and PARPis.

VO-OHpic treatment causes DSB accumulation in ovarian cancer cells

To better understand the mechanism underlying the synergistic effects of the combination of VO-OHpic with olaparib observed in ovarian cancer cells, the effects of VO-OHpic on DNA damage repair were evaluated, and DSB markers phosphorylated histone 2A family member X (γ-H2AX) and p53 binding protein 1 (53BP1) were therefore examined in SKOV3 and ES2 cells treated with VO-OHpic using immunofluorescence (IF) and western blot. Results showed that VO-OHpic treatment significantly increased foci and expression of γ-H2AX and 53BP1 in both SKOV3 (Fig. 2a–d) and ES2 cells (Fig. 2e–h). These data indicated that DSBs were accumulated upon VO-OHpic treatment in these ovarian cancer cells.

Fig. 2: VO-OHpic treatment caused DSB accumulation in ovarian cancer cells.
figure 2

Ovarian cancer cells SKOV3 and ES2 were treated with VO-OHpic at their corresponding IC50 and IC75 values for 24 h or 48 h as indicated, respectively, and expression levels of DNA damage markers γH2AX and 53BP1 were detected by immunofluorescence and western blot. Representative images of γH2AX (a) and 53BP1 (b) in SKOV3 cells with VO-OHpic treatment by immunofluorescence. c Statistical analysis results of γH2AX and 53BP1 positive cells in (a) and (b). d Protein levels of γH2AX and 53BP1 in SKOV3 cells with VO-OHpic treatment by western blot. Representative images of γH2AX (e) and 53BP1 (f) in ES2 cells with VO-OHpic treatment by immunofluorescence. g Statistical analysis results of γH2AX and 53BP1 positive cells in (e) and (f). h Protein levels of γH2AX and 53BP1 in ES2 cells with VO-OHpic treatment by western blot. γH2AX or 53BP1 foci positive cells (with >5 foci per cell) were quantified in five different high-power fields from each well, and presented as percentage of total cells. Data shown are the mean values (±S.D.) from three independent experiments. Statistically significant differences were analysed among multiple groups by one-way ANOVA with Tukey multiple comparison test using GraphPad Prism (version 9.5) with p < 0.05 were considered significant (*p < 0.05, **p < 0.01, ***p < 0.001). Scale bar = 10 μm.

Full size image

VO-OHpic treatment reduces protein levels of PTEN in nuclei

To gain insight into the mechanisms by which PTEN inhibition modulates DNA damage, the expression of PTEN in ovarian cancer cells treated with VO-OHpic was investigated. Our results revealed a discernible decrease in PTEN expression in total cell lysates of SKOV3 and ES2 cells following treatment with VO-OHpic, as illustrated in Fig. 3a, b. Given that PTEN primarily regulates DNA damage repair within the nuclei [11, 12], the expression of PTEN in both cytoplasm and nuclei of these ovarian cancer cells was examined. Our findings indicated that the expression of PTEN in the cytoplasm remained unchanged following treatment with PTEN; however, nuclear PTEN was significantly decreased in SKOV3 and ES2 cells with VO-OHpic treatment (Fig. 3a, b). Together with results from Fig. 2, these data suggest that reduced protein levels of PTEN, especially in nuclei, may contribute to its function in regulating DNA damage and repair. Since the combination of PTENi VO-OHpic with PARPis synergistically inhibited proliferation of ovarian cancer cells, these data indicated that the decreased expression of PTEN regulated certain proteins that are critical for the HR process.

Fig. 3: VO-OHpic treatment reduced the MRN complex in ovarian cancer cells.
figure 3

Ovarian cancer cells SKOV3 and ES2 were treated with VO-OHpic at IC50 and IC75 for 48 h, respectively. Representative western blot images of PTEN expression in whole cell lysates, nuclei, and cytoplasm fractions of SKOV3 (a) and ES2 cells (b). c Flow chart for screening key genes which are associated with PTEN in KEGG HR signalling pathways. d Positive correlation between PTEN and key genes in KEGG HR signalling pathways, including RAD54, TOPBP1, RAD50, MRE11, and TOP3, in TCGA and GTEx datasets analysed by online tool GEPIA. SKOV3 (e) and ES2 cells (f) were treated with VO-OHpic at IC75 for indicated time periods, and mRNA levels of MRE11, RAD50, RAD54 and TOPBP1 were examined by qRT-PCR. SKOV3 (g) and ES2 cells (h) were treated with VO-OHpic at IC50 and IC75 for 48 h, and protein levels of MRE11 and RAD50 were examined by western blots. Data shown are the mean values (±S.D.) from three independent experiments. Statistically significant differences were analysed among multiple groups by one-way ANOVA with Tukey multiple comparison test using GraphPad Prism (version 9.5) with p < 0.05 were considered significant (*p < 0.05, **p < 0.01, ***p < 0.001).

Full size image

VO-OHpic treatment reduces the MRE11-RAD50-NBN (MRN) complex in ovarian cancer cells

Previous studies mainly focused on the roles of PTEN in the dephosphorylation of its substrates in the cytoplasm [30,31,32], therefore, the protein levels of PDK1 and AKT, the major substrates of PTEN, were tested (Fig. S3). The results of Western blots demonstrated that VO-OHpic treatment didn’t significantly affect the expression of either phosphorylated or total proteins of PDK1 or AKT in SKOV3 or ES2 cells. Based on the hypothesis that the decreased expression of PTEN regulated proteins that are essential for the HR process, we thus examined the correlation between the expression of PTEN and that of the key genes in KEGG HR signalling pathways in ovarian cancer using TCGA and GTEx datasets via the online tool GEIPA to search for the potential signalling pathways affected by PTEN inhibition in ovarian cancers (Fig. 3c). The results showed that the expression of the following five genes in HR signalling pathways, including RAD54, TOPBP1, RAD50, MRE11, and TOP3, was most positively correlated with that of PTEN in ovarian cancer (R > 0.3 and p < 0.001) (Figs. 3d and S4). Furthermore, whether expression of these genes can be modulated by VO-OHpic treatment was investigated by quantitative real-time PCR (qRT-PCR). While mRNA levels of TOP3 was too low to be detected, the results showed that only the mRNA expression of MRE11 and RAD50 were consistently reduced in both SKOV3 and ES2 cells with VO-OHpic treatment (Fig. 3e, f). So, MRE11 and RAD50 were focused on, and western blot results demonstrated that their protein levels were also significantly reduced by VO-OHpic treatment in both cell lines (Fig. 3g, h).

Upon conducting an extensive literature review, we were surprised to discover that MRE11 and RAD50 are the members of the MRE11-RAD50-NBN (MRN) complex, known for its pivotal role in sensing and repairing DSBs. Moreover, this complex primarily regulates the HR repair in mammals [33,34,35]. In addition, the correlation analyses between the expression of these genes and prognosis of ovarian cancer patients by Kaplan–Meier plotter and UALCAN indicated that the expression of MRE11 correlated well with the overall survival (OS) of ovarian cancer patients, and the expression of MRE11 and RAD50, correlated well with progression free survival (PFS) of ovarian cancer patients inversely (Fig. S5A). Combining expression data of these genes, the expression of them also correlated very well with both OS and PFS with higher expression of these genes predicting worse prognosis (Fig. S5B). Additionally, expression of MRE11 and RAD50 proteins in ovarian cancer specimens (n = 100) was significantly higher than in normal specimens (n = 25) in CPTAC proteomics dataset via UALCAN platform (Fig. S5C).

Accumulation of DSB is accompanied with reduction of the MRN complex with VO-OHpic treatment

To enhance the detection of MRN complexes at resting state, where their expression is limited and immunofluorescence (IF) signals of MRN proteins may be insufficient, doxorubicin (DOX) was used to induce a certain level of DSBs in cells. As shown in Fig. 4a, b, the increased fluorescence intensities of γ-H2AX confirmed successful DSB induction by DOX. In the absence of VO-OHpic treatment, the elevated fluorescence intensities of MRE11 and RAD50 indicated the normal functionality of the MRN complex in dealing with DSBs. However, upon additional treatment with VO-OHpic, the intensities of γ-H2AX were further escalated, suggesting a greater accumulation of DSBs in cells with VO-OHpic treatment. Simultaneously, this treatment substantially reduced protein levels of MRE11 and RAD50 induced by doxorubicin treatment in both SKOV3 and ES2 cells (Fig. 4a, b). These results signify that VO-OHpic treatment potentially represses HR repair pathways by reducing the expression of the MRN complex in ovarian cancer cells.

Fig. 4: Accumulation of DSB was accompanied with reduction of MRN complex and expression of MRN complex probably via decreasing PTEN/E2F1-mediated transcription with VO-OHpic treatment.
figure 4

SKOV3 (a) and ES2 cells (b) were treated with DNA damaging inducer doxorubicin at 5 μM, with or without VO-OHpic at IC75 for 24 h. Representative immunofluorescence images of DNA damage marker γH2AX and components of MRN complex, including MRE11 and RAD50, as well as their statistical analysis graphs were presented. Relative luciferase activities of pGL3-RAD50-promotor or pGL3-MRE11-promotor with/without plasmids of pcDNA3.1-E2F1 or vectors or with/without VO-OHpic treatment in SKOV3 (c) or ES2 cells (d) as indicated. Luciferase activities were assayed after 48 h. PTEN depletion induced by VO-OHpic at their corresponding IC50 values for 48 h in SKOV3 (e) and ES2 cells (f) (7 and 25 μM, respectively) significantly decreased the recruitment of E2F1 at both RAD50 and MRE11 with ChIP assays. Representative immunofluorescence images, statistical analysis, and colocalization results of expression of PTEN and E2F1 in SKOV3 (g) and ES2 cells (h). Rr. Mander’s overlap coefficients (MOC) between expression of PTEN and E2F1 were presented as R. Ovarian cancer cells SKOV3 and ES2 were treated with VO-OHpic at IC50 and IC75 for 48 h. Data shown are the mean values (±S.D.) from three independent experiments. Statistically significant differences were analysed among multiple groups in (a, b) and (g, h) by one-way ANOVA with Tukey multiple comparison test and were analysed between two groups in (c–f) by Student’s t tests (two-tailed) using GraphPad Prism (version 9.5) with p < 0.05 were considered significant (#p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001). Scale bar = 10 μm.

Full size image

VO-OHpic treatment decreased expression of MRN complex probably via decreasing PTEN/E2F1-mediated transcription of MRN complex

It was reported that the transcription factor E2F1 can bind to the promoters of both MRE11 and RAD50 to regulate their transcription [36], while PTEN may act as an accessory factor to enhance the transcription activity of E2F1 [10, 36]. To verify whether E2F1 can promote expression of the MRN complex, SKOV3 or ES2 cells were transfected with plasmids of pGL3-RAD50 or pGL3-MRE11 or empty vectors, together with plasmids of pcDNA3.1-E2F1 or empty vectors, and luciferase activities were assayed (Fig. 4c, d). The results demonstrated that E2F1 can in fact promote expression of the MRN complex in these ovarian cancer cells. Additionally, to explore whether PTEN inhibition may affect E2F1 to promote expression of MRN complex, SKOV3 or ES2 cells were transfected with plasmids of pGL3-RAD50 or pGL3-MRE11 or empty vectors, with/without VO-OHpic treatment, and luciferase activities were assayed (Fig. 4c, d). The results exhibited that PTEN inhibition can decrease expression of MRN complex. To further confirm PTEN regulated MRN complex via E2F1, ChIP assays were performed to check the recruitment of E2F1 at the MRE11 and RAD50 promoters. The results of ChIP assays demonstrated that PTEN depletion induced by VO-OHpic significantly decreased the recruitment of E2F1 at both RAD50 and MRE11 in SKOV3 (Fig. 4e) and ES2 cells (Fig. 4f). To verify whether PTEN can work together with E2F1 to promote expression of MRN complex, expression of PTEN and E2F1 in nuclei, where they regulate transcription, were evaluated. Results demonstrated that in both SKOV3 and ES2 cells, exposure to DNA damage inducer DOX significantly promoted nuclear expression of PTEN and E2F1; while VO-OHpic treatment reduced their nuclear expression (Fig. 4g, h). It is noteworthy that typical colocalization coefficients Pearson correlation coefficient (PCC) and Mander’s overlap coefficient (MOC) demonstrated that expression of PTEN were positively correlated with that of E2F1 in nuclei with the addition of DNA damage inducer DOX or with the addition of both DOX and VO-OHpic, confirming that they work together to regulate transcription (Fig. 4g, h). These results indicated that VO-OHpic may inhibit the expression of the MRN complex by reducing nuclear expression of PTEN and decreasing E2F1/PTEN-mediated transcription of MRE11 and RAD50.

PTEN knockdown increases olaparib sensitivities, inhibits the expression of the MRN complex, and increases DSBs via decreasing PTEN/E2F1-mediated transcription in ovarian cancer cells

To rule out the potential off-target effects of VO-OHpic, endogenous PTEN levels in ovarian cancer cells were suppressed using shPTEN-1 and shPTEN-2. The expression of PTEN at both mRNA and protein levels in SKOV3 and ES2 cells were effectively diminished and demonstrated with qRT-PCR and western blot results (Fig. 5a, b). Though shPTEN-1 and shPTEN-2 effectively diminished the expression of PTEN at both mRNA and protein levels in SKOV3 cells, significant reduction in PTEN protein levels had only a slight impact on the proliferation of SKOV3 cells (Fig. 5c). Similar results were also observed in ES2 cells. Additionally, in both SKOV3 and ES2 cells, the combination of shPTENs with olaparib exhibited notably inhibitory effects on ovarian cancer cell proliferation compared to the control group (Fig. 5d). Consistently, the combination of shPTENs with olaparib also significantly enhanced the formation of γ-H2AX and 53BP1 foci in SKOV3 (Fig. 5e) and ES2 cells (Fig. 5f) compared to the control or olaparib alone groups. To further confirm PTEN regulated MRN complex via E2F1, ChIP assays were performed to check the recruitment of E2F1 at the MRE11 and RAD50 promoters. The results of ChIP assays demonstrated that PTEN depletion induced by shPTEN-1 or shPTEN-2 significantly decreased the recruitment of E2F1 at both RAD50 and MRE11 in SKOV3 (Fig. 5g) and ES2 cells (Fig. 5h). Western blot results suggested a corresponding decrease in the expression of the MRN complex upon PTEN knockdown (Fig. 5i). These data indicated that PTEN knockdown, similar to VO-OHpic treatment, may inhibit HR repair in ovarian cancer cells by decreasing the expression of the MRN complex via decreasing PTEN/E2F1-mediated transcription, thereby increasing the sensitivity to olaparib.

Fig. 5: PTEN knockdown increased olaparib sensitivities, inhibited the expression of MRN complex, and increased DSBs via decreasing PTEN/E2F1-mediated transcription in ovarian cancer cells.
figure 5

Ovarian cancer cells successfully transducted with shControl, shPTEN-1, or shPTEN-2 lentiviruses were selected for further qRT-PCR, western blot, MTT and immunofluorescence experiments. mRNA and protein expression of PTEN in SKOV3 (a) and ES2 cells (b) transducted with shPTENs were significantly reduced. c Relative cell viabilities of SKOV3 and ES2 cells transducted with shControl, shPTEN-1, or shPTEN-2. d SKOV3 and ES2 cells were transducted with shControl, shPTEN-1, or shPTEN-2 and treated with vehicle or olaparib for 72 h as indicated and their viabilities were evaluated by MTT assays. Statistical analysis results of γH2AX and 53BP1 foci positive cells in SKOV3 (e) and ES2 cells (f) transducted with shControl, shPTEN-1, or shPTEN-2 and treated with vehicle or olaparib as indicated by immunofluorescence assays. PTEN depletion induced by shPTEN-1 or shPTEN-2 significantly decreased the recruitment of E2F1 at both RAD50 and MRE11 in SKOV3 (g) and ES2 cells (h) with ChIP assays. i Protein levels of MRN complex in SKOV3 and ES2 cells transducted with shPTENs. Data shown are the mean values (±S.D.) from three independent experiments. Statistically significant differences were analysed among multiple groups by one-way ANOVA with Tukey multiple comparison test using GraphPad Prism (version 9.5) with p < 0.05 were considered significant (#p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, as compared to the results of the group shControl + vehicle or shControl; $p < 0.05, $$p < 0.01, $$$p < 0.001, as compared to the results of the group shControl+ olaparib).

Full size image

Combination of VO-OHpic with olaparib exerts enhanced inhibitory effects on tumour growth of ES2 cells in vivo

Furthermore, the therapeutic effects of combining VO-OHpic with olaparib were evaluated using a xenograft tumour model of ES2 cells. Results showed that though when administered individually at suboptimal dosages of 40 mg/kg for VO-OHpic and 200 mg/kg for olaparib, neither monotherapy significantly reduced tumour growth, the combination therapy of these two agents at the same dosage exhibited a significant reduction in tumour weights (Fig. 6a). Notably, the body weights and major organ weights of the mice were not significantly affected, indicating an absence of obvious toxicities associated with these treatments (Fig. S6). Immunohistochemistry staining of γH2AX in tumour sections of the mice treated with VO-OHpic, olaparib, and particularly with both reagents indicated a significant increase compared to the vehicle or monotherapy group (Fig. 6b, c). Similar results were also observed in immunohistochemistry staining of 53BP1. Additionally, immunohistochemistry results showed significant decreases of MRE11 and RAD50 protein levels in tumour sections of the mice treated with both reagents compared to the vehicle or olaparib monotherapy group. Immunohistochemistry staining of p-AKT in tumour sections of the mice treated with VO-OHpic, olaparib, and with both reagents did not show significant difference, consistent with the in vitro results (Fig. S3). Immunohistochemistry results showed significant decreases of PTEN levels in tumour sections of the mice treated with both reagents or VO-OHpic monotherapy group compared to the vehicle, consistent with the in vitro results in Fig. 3. It appeared that VO-OHpic inhibited the expression of MRE11 and RAD50 via suppressing the accumulation of PTEN in nuclei. Collectively, these findings confirm that the combination therapy with VO-OHpic and olaparib effectively inhibits the tumour growth of the insensitive ES2 cells by taking advantages of synergistic enhancement of DNA damages by suppressing the MRN complex via VO-OHpic treatment, thereby sensitising these cells to the PARP inhibitor, olaparib.

Fig. 6: Combination of VO-OHpic with olaparib exerted enhanced inhibitory effects on tumour growth of ES2 cells in vivo.
figure 6

ES2 cells were injected subcutaneously into the middle of the left flank of mice (7 mice/group) for in vivo experiments. From 7th to 20th day, these mice were administrated with vehicle, 40 mg/kg of VO-OHpic every other day, 200 mg/kg of olaparib every day, or both VO-OHpic and olaparib at the same doses and schedule, and then sacrificed and tumour weights were measured on the 21st day. a Statistical analysis results of tumour weights. b Representative immunohistochemistry images of γH2AX, 53BP1, MRE11, RAD50, p-AKT, and PTEN in tumour tissues as indicated. c Statistical analysis results of γH2AX, 53BP1, MRE11, RAD50, p-AKT, and PTEN in tumour tissues as indicated. d Schematic illustration of the proposed mechanisms of PTEN inhibitor VO-OHpic to enhance the sensitivity of ovarian cancer cells to PARP inhibitor olaparib. VO-OHpic inhibits the nuclear expression of PTEN and binding to E2F1 in nuclei in ovarian cancer cells, which lead to reduced expression of MRE11-RAD50-NBN (MRN) complex and DSB accumulation. Hence, VO-OHpic caused HR impairment and sensitised ovarian cancer cells to olaparib, resulting in cancer cell death. Data shown are the mean values (±S.D.) from three independent experiments. Statistically significant differences were analysed among multiple groups by one-way ANOVA with Tukey multiple comparison test using GraphPad Prism (version 9.5) with p < 0.05 were considered significant (#p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001). Scale bar = 50 μm.

Full size image

Discussion

PTEN, a well-established tumour suppressor gene, has been a subject of extensive research since its discovery in 1997 [37, 38]. Initially, investigations primarily focused on its role in phosphorylation regulation of cytoplasmic substrates, such as PDK1, AKT, FAK, and IRF3 [30,31,32]. However, there was no significant up-regulation of p-AKT with VO-OHpic treatments observed in Fig. S3. After we delved into the study by Rosivatz et al. [39], we found that the study demonstrated that under serum starvation, which ensures the phosphorylation levels of AKT remain minimal, VO-OHpic treatment significantly impacts the phosphorylation of AKT. In our experiments, cell cultures were consistently maintained in a nutrient-rich environment (10% FBS), ensuring that AKT was highly phosphorylated throughout our experiments. Such conditions likely diminished the impact of VO-OHpic treatment on the phosphorylation of AKT in our study. Furthermore, that study employed mouse fibroblast cells and human urinary bladder carcinoma cells, which could also result in varying outcomes. A groundbreaking study by Bassi et al. revealed that PTEN in the nuclei can regulate HR repair of DSBs [40, 41]. Interestingly, we found nuclear PTEN was significantly decreased in SKOV3 and ES2 cells with VO-OHpic treatment (Fig. 3a, b), therefore VO-OHpic treatment could potentially affect HR repair of DSBs. We investigated further how VO-OHpic treatment could change nuclear PTEN. Our study revealed that VO-OHpic treatment decreased the levels of total PTEN, while such treatment did not significantly change the levels of phosphorylated PTEN (Fig. S7A, B). It indicates the levels of non-phosphorylated PTEN were decreased with VO-OHpic treatment, a finding that is consistent with the study conducted by Shi et al., wherein VO-OHpic treatment specifically diminished the levels of non-phosphorylated PTEN [42]. The treatment with VO-OHpic could potentially induce structural alterations in PTEN or disrupt its metabolic pathways, leading to a reduction in the levels of non-phosphorylated PTEN. Furthermore, through cellular component separation and Western blot experiments, we found that nuclear PTEN is predominantly in its non-phosphorylated form (Fig. S7C, D). Literatures support that dephosphorylated PTEN is transported into and retained within the nuclei via diffusion or active transportation mechanisms [43,44,45]. Taking into account these findings, we propose that VO-OHpic treatment reduces the amounts of dephosphorylated PTEN, thereby decreasing the accumulation of PTEN in the nuclei.

Concurrently, over the past two decades, significant advancements have been made in understanding of the therapeutic potential of PARP inhibitors for treating BRCA-associated breast and ovarian cancers, with PARP inhibitors now commonly used as first-line maintenance therapy drugs. Despite approximately half of high-grade serous ovarian cancer harbouring HRD [46, 47], the majority of ovarian cancer patients do not carry HRD. Therefore, our study intentionally employed PTEN inhibitors to induce HRD status in ovarian cancer cells, making them susceptible to treatment with PARP inhibitors. Our results demonstrated that the combination of the PTEN inhibitor, VO-OHpic, with olaparib synergistically inhibited proliferation of ovarian cancer cells. Furthermore, our study showed that VO-OHpic effectively suppressed HR repair signalling pathways by modulating the MRN complex. Consistent findings were obtained using PTEN-targeting shRNAs. In vivo experiments confirmed that the combination therapy of VO-OHpic and olaparib effectively inhibited tumour growth of ES2 cells, which displayed the least sensitivity to either monotherapy throughout the study.

MRN complex, the earliest identified sensing complex of intracellular DSBs, plays a crucial role in the mammalian HR repair [33,34,35]. Consequently, researchers began investigating the roles of the MRN complex in tumorigenesis and its potential clinical applications. Although some studies have suggested that defects in the MRN complex might render certain ovarian cancers susceptible to treatment with PARP inhibitors [48, 49], some large-scale retrospective studies have indicated no association between the MRN complex and increased ovarian cancer risk [34, 50]. Scientists were initially disappointed at these findings. However, two proof-of-concept studies utilising genetic engineering techniques (shRNAs and dominant-negative NBN expression constructs) demonstrated that inhibiting the MRN complex can enhance the sensitivity of cancer cells to PARP inhibitors [51, 52]. These studies inspired us to explore the potential of MRN complex inhibition to facilitate PARP inhibitor treatments, even though MRN complex defects are uncommon in ovarian cancers. In our study, we effectively reduced MRN complex expression by decreasing PTEN levels in the nuclei, leading to increased sensitivity of ovarian cancer cells to the PARP inhibitor olaparib when combined with the PTEN inhibitor, VO-OHpic. In addition, resistance to PARP inhibitors often arises due to reverse mutations of HR genes or reactivation of their promoters [53, 54]. Given that the MRN complex sits at the apex of the DSB repair pathway, inhibiting MRN may offer a more efficient means to prevent PARP inhibitor resistance by blocking the entire HR pathway.

It is widely accepted that HR and non-homologous end joining (NHEJ) are the primary pathways for repairing DSBs, operating cooperatively and competitively [35, 53]. Inhibiting the MRN complex often increases the NHEJ pathway activity [55, 56]. A recent study by Sergey et al. showed that sensitivity of cancer cells to PARP inhibitors can be enhanced by improving the NHEJ pathway activity [57]. This finding makes the anti-cancer strategy of inhibiting the MRN complex in combination with PARP inhibitors more appealing. In our current study, we observed a significant increase in the foci and protein levels of 53BP1, a key player in NHEJ, following PTEN inhibition. Our results demonstrated that by suppressing the HR pathway, PTEN inhibition compels cells to rely on the error-prone NHEJ pathway, which can lead to apoptosis.

Noticeably, the study conducted by Piscitello et al. revealed a similar yet distinct mechanism in colon carcinoma cells. They discovered that PTEN loss in these cells led to the down-regulation of MRE11, triggered by an overactive AKT pathway, while the latter facilitated the protein degradation of MRE11 through p70S6 [58]. Such discrepancy in findings may stem from variations in cell lines derived from different tissues or varying degrees of PTEN reduction. In Piscitello et al.’s research, cells lacking PTEN expression were utilised, whereas in our study, PTEN was only partially decreased in the nuclei using the PTEN inhibitor or shRNAs.

Overall, our study revealed that inhibiting PTEN has dual functions in ovarian cancer treatment: acting as a tumour suppressor and PARPi sensitiser. Through the inhibition of PTEN nuclear expression, treatment with VO-OHpic disrupted the PTEN-E2F1 interaction, leading to a reduction in the transcription and protein levels of MRE11 and RAD50. Consequently, the HR repair pathway was effectively suppressed, resulting in the DSB accumulation, reduced cell proliferation, and inhibited tumour growth. Additionally, this treatment increased the sensitivity of ovarian cancer cells to the PARP inhibitor olaparib (Fig. 6d). By targeting the expression of the MRN complex, which serves as the most upstream sensor of the HR pathway, PTEN inhibitors offer a near-perfect application scenario for PARP inhibitors in HR proficient cells. This combination strategy has the potential to significantly expand the application of PARP inhibitors and benefit a larger population of ovarian cancer patients.