Patients and tissue samples

For the analysis of circ_ARHGEF28 expression in cisplatin-resistant and sensitive ovarian cancer, samples were collected from 150 female patients at Sun Yat-Sen Memorial Hospital, Sun Yat-Sen University, between September 2008 and May 2021. This study was conducted in accordance with the STARD reporting guidelines and the Helsinki Declaration. Informed consent was obtained from all donors, adhering to the International Ethical Guidelines for Biomedical Research Involving Human Subjects (CIOMS). This retrospective study, involving identified data, received approval from the Institutional Review Board (IRB) of Sun Yat-Sen Memorial Hospital [IRB no. SYSKY-2023-114-01], with a waiver for written consent.

Cells and cell culture

A2780 cells, sourced from the American Type Culture Collection (ATCC), were cultured in DMEM supplemented with 10% FBS, following standard protocols. Cisplatin-resistant A2780 cells (A2780R) were developed through treatment with 10 nmol cisplatin until cell death rates fell below 5%. A2780R cells were maintained in DMEM with 10% FBS, mirroring the conditions for parental A2780 cells.

RNA preparation and high-throughput sequencing analysis

Total RNA from A2780, cisplatin-resistant A2780R, and ovarian cancer samples (both cisplatin-resistant and sensitive) was extracted using Trizol reagent. For RNA sample preparations, 3 µg of RNA per sample was used. Ribosomal RNA was removed using the Epicentre Ribo-Zero™ Gold Kit (Human/Mouse/Rat) (Epicentre, USA). Following the manufacturer’s recommendations, sequencing libraries were generated using the NEBNext® Ultra™ Directional RNA Library Prep Kit for Illumina (NEB, Ipswich, USA). Library construction entailed the following steps: ribosomal RNA removal, RNA fragmentation under elevated temperatures, and synthesis of the first cDNA strand using random hexamer primers with RNA fragments as templates. The second strand of cDNA was synthesized using a mix of buffer, dNTPs, DNA polymerase I, and RNase R. After purification with QiaQuick PCR kits and EB buffer elution, library fragments underwent end repair, poly(A) addition, and adapter ligation. cDNA fragments of approximately 300 bp were selected through agarose gel electrophoresis, followed by UNG enzyme digestion of the second cDNA strand. PCR amplification was conducted, and the final library was retrieved through agarose gel electrophoresis.

RNA sequencing was performed by OE Biotech Co., Ltd. (Shanghai, China).

siRNA / shRNA and constructs

Two siRNA/shRNA duplexes, produced by Guangzhou Nuohao Biotechnology Co., Ltd., were utilized as detailed in Supplemental Table 3. The pLKO-Tet-On “all-in-one” plasmid facilitated inducible expression of shcirc_ARHGEF28 and sh-control. For ectopic expression in cells, MST1 and MST2 (full-length and truncated versions), along with circ_ARHGEF28, were cloned into the siRNA/shRNA or pCDNA3.1 plasmid. Transfections were performed using Lipofectamine 3000 (Invitrogen).

Quantitative RT-PCR

1 µg total RNA, extracted from cultured cells following standard protocols, was reverse transcribed into cDNA using the Superscript First-Strand cDNA Synthesis Kit (18080-051, Invitrogen, Carlsbad, CA). Quantitative RT-PCR was conducted using the SYBR Premix Ex Taq II kit (DRR081A, TAKARA, Otsu, Shiga, Japan) on a Light Cycler 480 System (Roche, Basel, Switzerland).

Colony formation assay

A total of 1,000 cells were seeded into 6-well plates and cultured for 14 days. Post fixation with 4% formaldehyde for 15 min, colonies were stained with 1% crystal violet.

IC50 and MTS cell viability assay

For each assay, 2,000 cells were plated in 96-well plates and treated with varying concentrations of cisplatin. At designated time points, cells were incubated with a sterile MTS mix (1:9 in culture medium) for 2 h at 37 °C in darkness. Absorbance was measured at 490 nm.

BaseScope and fluorescence in situ hybridization (FISH)

BaseScope (BaseScope™ Reagent Kit v2-RED, Advanced Cell Diagnostics China, #323,900) was employed for in situ detection of circ_ARHGEF28 expression in ovarian tumor samples. Deparaffinized paraffin-embedded samples were treated with 1×BaseScope solution (Target Retrieval) and hybridized with probes in a humidified box at 40℃ for 2 h. Hybridization was followed by sequential incubation with BaseScope v2 AMP 1 to 8 and washing with RNAscope wash buffer. Signals were detected using BaseScope™ Fast RED and counterstained with hematoxylin. Positivity was identified as red dots or clusters.

The BaseScope signal was quantified by combining the percentage of positive tumor cells with staining intensity, graded as 0 (no staining), 1 (weak, red dot), 2 (moderate, red dot to cluster), or 3 (strong, red cluster). An H-score was calculated using the formula: [1 × (% cells 1+) + 2 × (% cells 2+) + 3 × (% cells 3+)]. This method assessed circ_ARHGEF28 expression in ovarian cancer samples, with the median value defining circ_ARHGEF28-high and circ_ARHGEF28-low expression.

For FISH, standard protocols were followed using a probe targeting circ_ARHGEF28 (RiboBio Co, Ltd, Guangzhou, China). Confocal microscopy involved overnight incubation with fluorescence-conjugated anti-DIG antibodies (Roche) at 4 °C, followed by DAPI counterstaining and imaging with a laser scanning confocal microscope (A1, Nikon).

Immunoblotting

Cell lysates were prepared using RIPA lysis buffer (Beotime, China) supplemented with protease and phosphatase inhibitors (Life Technologies, USA). Proteins were separated by 8% SDS-PAGE and transferred onto PVDF membranes (Bio-Rad, USA). After blocking with 5% non-fat milk in 0.1% TBST buffer overnight at 4 °C, the membranes were incubated with primary antibodies (dilutions 1:500-1:1000, as listed in Supplemental Table 4). The protein-antibody complexes were detected using HRP-conjugated secondary antibodies and enhanced chemiluminescence. The Image J was used for statistical analysis of the result.

RNA immunoprecipitation (RIP)

RIP assays were conducted using the Magna RIP RNA Binding Protein Immunoprecipitation Kit (Millipore, MA, USA), following the manufacturer’s guidelines. Cell extracts were prepared in lysis buffer containing protease and RNase inhibitors, incubated on ice for 5 min and centrifuged at 10,000 g at 4 °C for 10 min. Magnetic beads were preincubated with 5 µg of IP-grade antibody for 30 min at room temperature with rotation. The supernatant was then added to the bead-antibody complexes in immunoprecipitation buffer and incubated overnight at 4 °C. RNA was subsequently purified and quantified via qRT-PCR. Input controls and normal rabbit IgG controls were included to confirm the specificity of RNA-protein binding.

Ubiquitination assay

Cells were treated with MG132 (20 nM) in DMEM culture medium for 4 h before lysate collection. Lysates were immunoprecipitated with specific antibodies on protein A/G beads (Life Technologies) overnight at 4 °C with rotation, then boiled in SDS buffer. The eluted proteins were detected by immunoblotting.

Circular RNA pull-down assay

Circular RNA pull-down was executed using biotinylated circ_ARHGEF28 (sense) and a negative control (antisense) probes provided by Guangzhou RiboBio Co, Ltd. Initially, cells were lysed in IP lysis buffer at 4 °C for 30 min. Cell supernatants were incubated with circ_ARHGEF28 and control biotinylated probes overnight at 4 °C. The mixtures were then co-incubated with streptavidin magnetic beads at room temperature for 1 h. Following SDS-PAGE separation, proteins were visualized via silver staining. Mass spectrometry analysis was conducted to identify potential binding proteins of circ_ARHGEF28.

Co-immunoprecipitation (Co-IP) assay

Cell lysates from A2780R were immunoprecipitated with specified antibodies on protein A/G beads (Life Technologies) overnight at 4 °C with rotation, followed by boiling in SDS buffer. The eluted proteins were analyzed using immunoblotting techniques.

Animal experiments

All animal studies were performed in compliance with the guidelines of the Institutional Animal Care and Use Committee at Sun Yat-Sen University Medical School. Mice were sourced from Beijing Vital River Laboratory Animal Technology Co., Ltd. Six-week-old female nude mice (n = 5 per group) were subcutaneously inoculated with indicated cells. Upon reaching an average tumor volume of approximately 150 mm3, the mice were administered doxycycline (DOX) to induce DOX-inducible shRNA-circ_ARHGEF28 expression, followed by cisplatin treatment (4 mg/kg intraperitoneally) daily for one week.

Statistics

In vitro, experimental data were presented as the mean ± standard deviation (S.D.) derived from three independent experiments. Statistical analyses were conducted using the SPSS 16.0 software package (SPSS, Chicago, IL, USA). The Student’s t-test and one-way ANOVA were applied to assess differences in cell viability, colony formation, apoptosis rates, and tumor volumes under various treatment conditions. The Chi-square test was utilized to examine the association between circ_ARHGEF28 expression and clinicopathological characteristics. Survival outcomes, including overall survival (OS) and progression-free survival (PFS), were compared across different patient groups using Kaplan-Meier curves and log-rank tests. Statistical significance was denoted as follows: *P < 0.05, **P < 0.01, and ***P < 0.001.