Human OC cell lines (OVCAR3, SKOV3, ES2) from Procell (Wuhan, China), IGROV-1 from Cellverse (Shanghai, China), and normal ovarian epithelial cells (IOSE80) from Stemrecell (Shanghai, China) were cultured in RPMI 1640 (Gibco, Thermo Scientific, MA, USA) or McCoy’s 5A (Gibco) medium supplemented with 10% fetal bovine serum (FBS, Gibco) and 1% penicillin/streptomycin (Meilunbio, Dalian, China) at 37°C. Subculturing was performed using trypsin-EDTA (Meilunbio) upon reaching 80–90% confluency. Cells were passaged at a ratio of 1:3 and maintained in T25 flasks or multiwell plates as required for subsequent experiments. Media were refreshed every 48 hours to ensure optimal growth conditions.

Stable RNLS-knockdown (sh-RNLS) and overexpression (ov-RNLS) OVCAR3 cell models were generated using third-generation lentiviral vectors. For viral packaging, 293T cells (ATCC^® CRL-3216™) were co-transfected with transfer plasmids (pLKO.1-puro for shRNA or pLVX-IRES-ZsGreen1 for overexpression, Addgene, MA, USA) and packaging plasmids (psPAX2 and pMD2.G, Addgene) using Lipofectamine 3000 (Invitrogen, Thermo Scientific) at a ratio of 4:3:1 (plasmid: psPAX2: pMD2.G). Viral supernatants were harvested at 48 and 72 hours post-transfection, filtered through 0.45 μm PVDF membranes (Beyotime, Shanghai, China), and concentrated via ultracentrifugation (50,000 ×g, 4°C, 2 hours). Viral titers were determined by ELISA quantification of p24 antigen (1×10⁸ TU/mL, Takara, Dalian, China). For RNLS knockdown, three independent shRNA sequences (sh-RNLS-1: 5′-AGCTGTTGCTGTTACTTTA-3′; sh-RNLS-2: 5′-GGATGTGTCCTTGAATCAA-3′; sh-RNLS-3: 5′-GCTTCGTCTCCATTGATAA-3′) targeting distinct regions of the RNLS transcript were cloned into pLKO.1-puro, with scrambled shRNA (sh-NC) as the negative control. For overexpression, the full-length human RNLS cDNA (NM_001031709.3) was subcloned into pLVX-IRES-ZsGreen1 under the CMV promoter, with an empty vector (ov-NC) serving as the control. OVCAR3 cells at 20% confluency in 6-well plates were transduced with concentrated lentiviral particles at a multiplicity of infection (MOI) of 20 in RPMI 1640 containing 8 μg/mL Polybrene (Beyotime). After 6 hours of incubation at 37°C, the virus-containing medium was replaced with fresh complete medium. Transduction efficiency was validated by quantitative real-time PCR (qPCR) analysis of RNLS expression.

Cell viability was assessed using the CCK-8 assay (Meilunbio). OVCAR3 cells (5×10⁴ cells/mL) were seeded into 96-well plates and incubated for 24 hours. After treatment with STAT3 siRNA on sh/ov RNLS stable OVCAR3 cells, 10% CCK-8 reagent was added to each well and incubated for 40 minutes at 37°C. Post-treatment, fresh medium was replenished and cells were cultured until designated timepoints (24/48/72 hours) prior to CCK-8 assay. Absorbance at 450 nm was measured using a microplate reader (Thermo Scientific, Multiskan 3K). Data were normalized to untreated controls, and viability rates were calculated from triplicate wells across three independent experiments.

Matrigel matrix (Corning, NY, USA) was thawed overnight at 4°C, diluted 1:8 in pre-chilled serum-free medium, and 50 μL aliquots (~1 mg/mL final concentration) were polymerized in Transwell chambers at 37°C for 1 hour. For invasion assays, OVCAR3 cells (2.5×10⁵ cells/mL) in serum-free medium were seeded into the upper chambers of 24-well Transwell inserts (Corning). The lower chambers contained 20% FBS. After 24 hours, non-invading cells were removed with a cotton swab, and invaded cells were fixed with 4% paraformaldehyde, stained with 0.1% crystal violet (Meilunbio), and imaged under a microscope (Ts2-FL, Nikon, Japan). For wound healing assays, confluent monolayers in 6-well plates were scratched with a sterile pipette tip, washed with PBS, and cultured in 2% FBS medium. Five representative fields (×100 magnification) per wound were imaged at identical locations immediately post-scratching (0h) and at 24h. Migration distances were quantified at 0 and 24 hours using ImageJ software (National Institutes of Health [NIH], MD, USA).

Apoptosis was analyzed using the Annexin V-APC/PI Apoptosis Kit (AP107, Lianke, Hangzhou, China). Cells were harvested, washed with PBS, and resuspended in 1× binding buffer. After staining with Annexin V-APC and PI for 15 minutes in the dark, samples were analyzed on a BD FACSCalibur flow cytometer (BD Biosciences, CA, USA). For ROS detection, cells were incubated with 10 μM dihydroethidium (DHE, Beyotime) for 30 minutes, washed, and analyzed via flow cytometry at 488 nm excitation/610 nm emission. Data were processed using FlowJo software (version 12.0; FlowJo LLC), with gating based on unstained and single-stained controls.

Total proteins were extracted from cells or tissue using RIPA lysis buffer (Meilunbio). Protein concentrations were determined via BCA assay (Meilunbio). Equal amounts of protein (20 μg) were separated on 10% SDS-PAGE gels and transferred to PVDF membranes (Beyotime). Membranes were blocked with 5% non-fat milk (Beyotime) and probed overnight at 4°C with primary antibodies against RNLS (1:500, 15003-1-AP, Proteintech, Wuhan, China), STAT3 (1:2000, 10253-2-AP, Proteintech), p-STAT3 Tyr705 (1:1000, ab267373, Abcam, Cambridge, UK), p-STAT3 Ser727 (1:1000, ab32143, Abcam), PI3K (1:5000, 60225-1-Ig, Proteintech), p-PI3K (1:500, bs-6417R, Bioss, Beijing, China), AKT (1:5000, 60203-2-Ig, Proteintech), p-AKT (1:2000, 66444-1-Ig, Proteintech), and GAPDH (1:50000, 60004-1-Ig, Proteintech). Membranes were washed thrice in TBST (0.1% Tween-20, 10 min/wash), incubated with HRP-conjugated secondary antibodies (1:2000, SA00001-2, Proteintech), and developed using ECL kit (Meilunbio). Band intensities were quantified via ImageJ densitometry, with target protein expression normalized against GAPDH.

Total RNA was extracted from cells or tissue with RNAiso Plus (Takara) and reverse-transcribed into cDNA using the NovoScript^® Plus cDNA Synthesis Kit (Novoprotein, Suzhou, China). qPCR was performed on an ABI 7300 system with NovoStart^® SYBR SuperMix (Novoprotein). The qPCR protocol included an initial denaturation at 95°C for 1 minute, followed by 40 cycles of 20 s at 95°C for denaturation, 20 s at 56°C for annealing, and 38 s at 72°C for extension. Primers for RNLS (Mice, F: 5′-AGAAGTCTCCCTCAAGCACTG-3′, R: 5′-GCATGGTGAGGATGACAAGGT-3′; Human, F: 5′-GCTCCCCTGAGCAGTTTGAT-3′, R: 5′-TTGCTGCCTTTGGCATTCAC-3′) and STAT3 (Mice, F: 5′-AGCTGGACACACGCTACCT-3′, R: 5′-AGGAATCGGCTATATTGCTGGT-3′; Human, F: 5′-AGAAGGACATCAGCGGTAAGA-3′, R: 5′-GGATAGAGATAGACCAGTGGAGAC-3′) were synthesized by Shangya Biotechnology (Zhejiang, China). GAPDH (Mice, F: 5′-GGTGAAGGTCGGTGTGAACG-3′, R: 5′-CTCGCTCCTGGAAGATGGTG-3′; Human, F: 5′-GCTCATTTGCAGGGGGGAG-3′, R: 5′-GTTGGTGGTGCAGGAGGCA-3′) served as the endogenous control. Relative expression was calculated using the 2^−ΔΔCt method (21).

STAT3 subcellular localization was analyzed using IF. OVCAR3 cells grown on glass coverslips were fixed with 4% paraformaldehyde for 20 min, permeabilized with 0.5% Triton X-100 (Beyotime) for 15 min. Cells were incubated overnight at 4°C with STAT3 primary antibody (1:50), followed by Coralite594-conjugated goat anti-rabbit IgG (1:100, SA00013-4, Proteintech) for 1 h at 37°C. Nuclei were counterstained with DAPI (1 μg/mL, Meilunbio) for 5 min. Coverslips were mounted with anti-fade reagent (Meilunbio) and imaged using a Nikon Ts2-FL fluorescence microscope with excitation/emission filters set for 594 nm (red) and 358 nm (blue).

Mitochondrial ultrastructure was examined by TEM. Cells were fixed in 2.5% glutaraldehyde overnight at 4°C, post-fixed with 1% osmium tetroxide for 2 h, and dehydrated through an ethanol gradient (50%-100%). Samples were embedded in Epon 812 resin and sectioned into 70-nm ultrathin slices using a ultramicrotome (Leica UC7, Leica, Germany). Sections were stained with uranyl acetate and lead citrate for 15 min each. Images were acquired using a Hitachi HT7800 TEM operated at 120 kV, with mitochondrial cristae structure and membrane integrity analyzed using ImageJ software.

Male BALB/c nude mice (5-week-old) were procured from Sibeifu Biotechnology Co., Ltd. and housed under specific pathogen-free conditions with a 12-hour light/dark cycle. Mice were randomized into 4 groups (n=6 per group): sh-NC, sh-RNLS, ov-NC, and ov-RNLS. Subcutaneous tumors were induced by injecting 5.5 × 10⁶ OVCAR3 cells (suspended in 100 μL PBS) into the right flank. Tumor dimensions (length and width) were measured every 5 days using digital calipers, and volumes were calculated as V = 1/2×(length×width²), with timepoints recorded at days 5, 10, 15, 20, and 25 post-inoculation. At experimental endpoint (day 25), humane euthanasia was achieved through tail vein administration of sodium pentobarbital (200 mg/kg). Excised tumors underwent gravimetric analysis using a precision balance (BS210S, Sartorius) followed by bifurcated processing: 4% paraformaldehyde fixation (48h) for histopathological evaluation versus cryopreservation in liquid nitrogen for downstream molecular profiling. All experimental were approved by the Institutional Animal Care and Use Committee of Fujian Cancer Hospital (approval No. IACUC FJABR 2024080101).

Tumor tissues from xenograft models were fixed in 4% paraformaldehyde, paraffin-embedded, and sectioned at 4 μm. After deparaffinization and antigen retrieval with citrate buffer, sections were blocked with 5% BSA and incubated overnight with antibodies against Ki-67 (1:1000, 28074-1-AP, Proteintech), RNLS (1:50), or STAT3 (1:20). HRP-conjugated secondary antibodies and DAB substrate were applied for signal detection. Slides were counterstained with hematoxylin (Beyotime), dehydrated, and mounted with neutral resin. Images were captured using a microscope (E200, Nikon) and observe via Fully automatic digital slide scanner (version GCell-60, GCell, Guangzhou, China).

Reduced GSH and MDA levels in tumor tissues were quantified using commercial kits (Jiancheng Bioengineering Institute, Nanjing, China). Tissue samples (50 mg) were homogenized in ice-cold PBS (1:9 w/v), centrifuged at 12,000×g (4°C, 15 min), and supernatants collected. For GSH, homogenates were mixed with precipitating reagent and chromogenic substrates, and absorbance was measured at 405 nm. MDA was assessed via thiobarbituric acid reactive substances assay at 532 nm. GPX4 activity was determined using an ELISA kit (MM-60328H2, Enzyme Immunoassay). Protein concentrations were normalized using BCA assay results.

Data are presented as mean ± SD from three independent experiments. Comparisons between groups were analyzed using one-way ANOVA followed by Tukey’s post hoc test in GraphPad Prism 8.0. A p-value <0.05 was considered statistically significant.

OC cell lines relative to normal ovarian epithelial cells. Among these, the OVCAR3 line exhibited the most marked RNLS upregulation ( Figure 1A ). To interrogate RNLS functionality, stable OVCAR3 derivatives with either RNLS knockdown (sh-RNLS) or overexpression (ov-RNLS) were generated via lentiviral transduction. Subsequent analyses confirmed successful transcriptional and translational modulation of RNLS in these engineered lines ( Figures 1B, C ). Western blot analysis revealed that RNLS and STAT3 protein levels were suppressed by sh-RNLS and promoted by ov-RNLS ( Figure 1D ). These results indicate that the expression of RNLS affects the transcription and translation of STAT3, which is a key factor influencing the OC process.

Functional assays demonstrated RNLS’s critical role in driving malignant behaviors. RNLS depletion significantly impaired cell viability in CCK-8 assays ( Figure 2A ), reduced wound closure capacity in migration assays ( Figure 2B ), and attenuated cellular invasion through Matrigel in Transwell assays ( Figure 2C ). Conversely, RNLS overexpression enhanced proliferation, accelerated migration, and increased invasion. Flow cytometry analysis further revealed that RNLS knockdown markedly elevated apoptosis rates, while RNLS overexpression conferred significant protection against apoptosis ( Figure 2D ). These results establish RNLS as a key regulator of OC cell proliferation, motility, invasion, and survival.

To dissect the functional interplay between RNLS and STAT3, STAT3 expression was silenced in OVCAR3 cells using siRNA. Three STAT3-targeting siRNAs were transfected into OVCAR3 cells, with si-STAT3–2 showing the strongest suppression of STAT3 mRNA ( Figure 3A ). qPCR and IF analyses demonstrated that ov-RNLS rescued STAT3 expression suppressed by si-STAT3 ( Figures 3B, C ). Furthermore, Western blot analysis revealed that si-STAT3 transfection did not significantly affect RNLS protein levels but suppressed the expression of STAT3, p-STAT3 (Ser727), and p-STAT3 (Tyr705). In contrast, co-transfection with ov-RNLS promoted the protein levels of RNLS, STAT3, and p-STAT3 (Tyr705), while exhibiting no significant effect on p-STAT3 (Ser727) expression ( Figure 3D ). Furthermore, si-STAT3 inhibited p-PI3K/PI3K and p-AKT/AKT ratios, while ov-RNLS co-transfection mitigated this inhibition ( Figure 3E ). These results indicate that RNLS serves as an upstream regulator of STAT3 and selectively modulates its phosphorylation at the Tyr705 residue.

We next investigated whether RNLS could functionally compensate for STAT3 loss. STAT3 knockdown significantly impaired cell viability ( Figure 4A ), reduced migration and invasion capacities ( Figures 4B, C ), and increased apoptosis and intracellular reactive oxygen species (ROS) levels ( Figures 4D, E ). Strikingly, co-expression of ov-RNLS substantially mitigated these detrimental effects, partially restoring proliferation, migration, and invasion, while concurrently reducing apoptosis and ROS accumulation. TEM revealed that STAT3 deficiency induced severe mitochondrial damage, including cristae disorganization and vacuolization. RNLS co-expression significantly ameliorated these ultrastructural mitochondrial abnormalities ( Figure 4F ), indicating RNLS protects mitochondrial integrity downstream of STAT3.

Given the observed ROS dysregulation, we probed the role of the RNLS-STAT3 axis in redox homeostasis and ferroptosis. We observed that STAT3 suppression significantly increased ROS levels, reduced GPX4 and GSH levels, and promoted MDA accumulation compared to controls ( Figures 5A–D ). These effects were reversed by ov-RNLS co-expression. Treatment with the ferroptosis inhibitor ferrostatin-1 (Fer-1) attenuated STAT3 knockdown-induced ROS and MDA accumulation but failed to restore GPX4 or GSH levels, suggesting RNLS regulates ferroptosis susceptibility through mechanisms distinct from or broader than Fer-1’s action, potentially involving direct GPX4/GSH modulation via the STAT3 pathway.

Subsequently, we verified the pathophysiological significance of RNLS in the subcutaneous xenograft model. Tumors derived from sh-RNLS OVCAR3 cells exhibited significantly slower growth kinetics ( Figure 6A ) and lower final weights ( Figures 6B, C ) compared to control tumors. Conversely, tumors from ov-RNLS cells grew faster and were heavier. Molecular analysis of excised tumors confirmed downregulation of RNLS and STAT3 mRNA in sh-RNLS tumors and upregulation in ov-RNLS tumors ( Figure 6D ). IHC corroborated these findings, showing reduced protein expression of RNLS, STAT3, and the proliferation marker Ki-67 in sh-RNLS tumors, with opposite trends in ov-RNLS tumors ( Figure 7A ). Biochemical assays further revealed elevated MDA content and suppressed GPX4 activity/GSH levels in sh-RNLS tumors, mirroring the in vitro oxidative stress phenotype, while ov-RNLS tumors showed reduced MDA and enhanced GPX4/GSH activity ( Figures 7B–D ). These in vivo data conclusively demonstrate RNLS’s oncogenic role through STAT3-dependent modulation of tumor growth and redox balance.