CRISP3, a Potential Tumor Suppressor, Inhibits the Progression of High-Grade Serous Ovarian Carcinoma by Modulating the PI3K/AKT Pathway

Abstract

Background: Ovarian cancer (OC) remains the most common cause of gynecological cancer-related death, and about 70% of these deaths are from advanced high-grade serous ovarian cancer (HGSOC). Cysteine-rich secretory protein 3 (CRISP3) is related to various human diseases; however, the roles and mechanisms of CRISP3 in HGSOC remain unclear. Methods: The clinical significance of CRISP3 in patients with OC was analyzed using the Cancer Genome Atlas (TCGA) and Gene Expression Omnibus databases. CRISP3 expression in OC tissues was validated by RNA-sequencing (RNA-seq), quantitative PCR, and immunohistochemistry. Furthermore, we explored the effect of CRISP3 expression modulation on the biological behavior of HGSOC through CCK-8, EdU, and Transwell assays in vitro, and the differences in CRISP3 during the progression of HGSOC in vivo. We utilized RNA-seq, GSEA and Western blotting to detect CRISP3’s regulatory mechanisms. Finally, we employed data from the IMvigor210 cohort and TCGA to assess the correlation of CRISP3 with clinical response to immunotherapy, and the landscape of immune cell infiltration. Results: CRISP3 expression was markedly reduced in HGSOC. In vitro studies demonstrated that CRISP3 knockdown significantly enhanced proliferation, migration, and invasion of HGSOC cells, whereas its overexpression suppressed these malignant phenotypes. Moreover, CRISP3 expression was found to be downregulated during OC progression in vivo. Mechanistically, CRISP3 acted as a tumor suppressor through the PI3K/AKT signaling pathway to inhibit the progression and metastasis of HGSOC. Additionally, we observed an association between CRISP3 expression and CD8⁺ T cell, macrophage, neutrophil and Th1 cell infiltration. Conclusions: We demonstrate that CRISP3 suppresses tumorigenesis in HGSOC by regulating the PI3K/AKT pathway, and that alterations in its expression correlate with disease progression, supporting its utility as a biomarker.

1. Introduction

Ovarian cancer ranks as the second leading cause of gynecological cancer mortality worldwide. In 2022, there were an estimated 324,398 new cases and 206,839 ovarian cancer-related deaths globally [1]. Epithelial ovarian cancer (EOC) accounts for over 90% of ovarian tumors, with HGSOC being the most prevalent subtype, typically originating from the distal fallopian tube [2]. Although most patients initially respond to treatment, drug resistance and relapse are common, leading to limited therapeutic efficacy and making ovarian cancer a major cause of death among gynecological malignancies [3]. Therefore, elucidating the molecular mechanisms underlying HGSOC progression and identifying novel therapeutic targets have become urgent priorities in current research.

In this context, secretory protein families have attracted considerable attention due to their critical roles in intercellular communication and the regulation of the tumor microenvironment. Among them, the cysteine-rich secretory protein (CRISP) family represents a class of evolutionarily conserved secretory proteins widely involved in reproduction, immunity, and tumor progression [4]. Studies have shown that multiple CRISP members are aberrantly expressed in various cancers and influence tumor behavior; for example, CRISP2 is significantly associated with head and neck squamous cell carcinoma [5]. Moreover, accumulating expression data indicate that dysregulation of specific CAP domain proteins—particularly CRISP3—is linked to cancer initiation and progression and may play a functional role in these processes [4]. These findings suggest that the CRISP family plays an important and broad role in tumor regulation, warranting further exploration in HGSOC.

The human CRISP3 gene is located on chromosomal region 6p21.3 and is significantly upregulated in various pathological conditions, including prostate cancer [6,7], tongue squamous cell carcinoma [8], and Sjögren’s syndrome [9,10]. It is also overexpressed in multiple types of immune cells [11,12]. However, the role of CRISP3 exhibits tissue specificity in cancer research: in prostate cancer [13], breast cancer [14], and non-small-cell lung cancer [15], CRISP3 has been shown to promote tumor invasion and progression, with its high expression associated with poor prognosis. In contrast, its expression is downregulated in cervical cancer, where low expression correlates with unfavorable outcomes [16]. Similarly, earlier studies suggested that CRISP3 expression is reduced in malignant ovarian tissues, although not as prominently as its interacting partner β-microseminoprotein [17], implying that CRISP3 may play a distinct regulatory role in ovarian cancer compared to other malignancies, warranting further investigation. Moreover, recent studies have revealed that CRISP3 can modulate the tumor microenvironment, influencing immune cell infiltration [18]. Therefore, elucidating the specific functions of CRISP3 in HGSOC development and its interaction with the immune microenvironment is of significant importance.

2. Materials and Methods

2.1. Bioinformatic Analysis

Analysis of CRISP3 expression was performed utilizing publicly available data from TCGA and the GEO database, incorporating the datasets GSE69428, and GSE69429. Additionally, protein expression profiles visualized by immunohistochemistry (IHC) for normal fallopian tube epithelium and ovarian tumor tissues were acquired from the Human Protein Atlas (HPA).

The cBioPortal database (id—Ovary/Fallopian (CPTAC GDC, 2025)) provided HGSOC expression profiles spanning various stages of disease progression. Patient overall survival (OS) was assessed using Kaplan–Meier survival analysis, with a p < 0.05 obtained by the log-rank test considered statistically significant. Subsequently, the correlation between CRISP3 expression and both immune cell infiltration levels and the overall extent of immune/stromal infiltration in ovarian serous cystadenocarcinoma was inferred using ssGSEA and ESTIMATE algorithms implemented in the “ESTIMATE” and “GSVA” R packages [19,20].

2.2. Patients and Specimens

We collected clinical samples from two patient cohorts. The first cohort comprised 19 normal tissues and 56 ovarian cancer tissues. A subset of these samples was used for RNA sequencing (RNA-seq), while another subset (including all 19 normal tissues and 23 HGSOC tissues) was subjected to immunohistochemical (IHC) examination. The second cohort included 12 normal fallopian tube tissues and 12 HGSOC tissues, which were utilized for qRT-PCR validation of CRISP3 expression. All patients with OC included in this study with ovarian cancer received treatment at Shanghai First Maternity and Infant Hospital. Enrolled patients had no prior history of other malignancies and had not undergone any radiotherapy or chemotherapy.

The study involving human participants was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the Shanghai First Maternity and Infant Hospital (Approval No. KS22335). Written informed consent was obtained from all participants prior to sample collection.

2.3. Cell Lines

Cell lines for HGSOC-HEY, OVCAR8, and OV90 were obtained from ATCC; IOSE and ID8 (murine ovarian cancer cell line) were purchased from Meisen CTCC (Hangzhou, China). HEY and ID8 cells were cultured in DMEM (Gibco, Waltham, MA, USA), while OVCAR8 and IOSE cells were maintained in RPMI-1640 medium (Gibco, USA). All media were supplemented with 10% Fetal Bovine Serum (FBS, Gibco, USA) and 1% penicillin-streptomycin. OV90 cells were grown in a 1:1 mixture of MCDB 105 and Medium 199 (Gibco, USA), supplemented with 15% FBS and 1% penicillin-streptomycin. Among them, IOSE is a normal ovarian epithelial cell line, and was employed as the benign epithelial reference in this study. All cell lines were maintained at 37 °C in a 5% CO₂ atmosphere and were mycoplasma-free (YEASEN Mycoplasma Detection Kit, Shanghai, China). All cell lines were authenticated shortly before use by the STR technique, carried out by Sangon Biotech (Shanghai, China).

2.4. Immunohistochemistry and Multiplex Immunohistochemical Evaluation

IHC staining was performed using an IHC kit (Abcam, Cambridge, UK). Briefly, tissue sections on slides were immersed in Tris-EDTA buffer (pH 9.0) and heated in a microwave at medium power for 10 min for antigen retrieval. They was incubated with the anti-CRISP3 antibody (1:600, Proteintech, Wuhan, China) at room temperature for 1.5 h, followed by incubation with an ample volume of peroxidase-labeled polymer-conjugated secondary antibody for 30 min at room temperature. Following processing, tissue sections were mounted with coverslips for microscopic observation. Two investigators, blinded to the experimental groups, scored the staining intensity (0: negative; 1: weak; 2: moderate; 3: strong) and the percentage of positive areas (0: <5%; 1: 5–25%; 2: 25–50%; 3: 50–75%; 4: >75%). A final composite score was derived by multiplying the intensity and percentage scores.

Multiplex immunohistochemical (mIF) staining was performed in a sequential manner, with each round detecting a single antigen. The procedure for each cycle consisted of primary antibody incubation, secondary antibody incubation, and tyramine signal amplification (TSA) visualization. Following this, the same protocol was repeated for epitope retrieval and protein blocking prior to the application of the next antibody. The following markers were sequentially examined using immunofluorescence (IF) staining: CD8 (Abclonal, Wuhan, China, A23305PM, dilution 1:3000), CD68 (CST, Danvers, MA, USA, 76437S, dilution 1:4000), and MPO (Abcam, ab208670, dilution 1:6000). Cell nuclei were counterstained with DAPI (blue). TSA-based visualization was conducted using the Opal 6-Color multiplex IHC kit (Absin, Shanghai, China, abs50030) in strict accordance with the manufacturer’s instructions. Quantitative analysis included enumeration of positive cells using ImageJ software (ImageJ 1.53e) and measurement of mean fluorescence intensity with SlideViewer software (2.6.0).

2.5. RNA Extraction and Quantitative Real-Time Polymerase Chain Reaction

Total RNA was extracted from cells or human tissue using Trizol reagent (Takara, Shiga, Japan), and reverse transcription was carried out using the ABScript III RT kit (Abclonal, Wuhan, China) to synthesize cDNA. qRT-PCR was performed using Bright Cycle Universal SYBR Green qPCR Mix (Abclonal, Wuhan, China) according to the manufacturer’s instructions. ACTB was used as an internal control.

The primer sequences were as follows:

ACTB forward: CATGTACGTTGCTATCCAGGC.

ACTB reverse: CTCCTTAATGTCACGCACGAT.

CRISP3 forward: AAATCATGGAAAATAAGGGAATCCT.

CRISP3 reverse: CCAAGAAGCACATTGCATTTG.

TXNIP forward: ATATGGGTGTGTAGACTACTGGG.

TXNIP reverse: GACATCCACCAGATCCACTACT.

ID2 forward: GCTATACAACATGAACGACTGCT.

ID2 reverse: AATAGTGGGATGCGAGTCCAG.

2.6. Plasmid Constructs, Cell Transfection, and Stable Transfection Strain Construction

The overexpression plasmid was constructed as follows: First, the full-length coding sequence (CDS) of CRISP3 was amplified by RT-PCR from the total RNA of OVCAR8 cells. The resulting PCR product was purified and cloned into an expression vector. The successfully constructed recombinant plasmid was then subjected to full-length Sanger sequencing. Finally, the sequencing results were aligned with the NCBI reference sequence (NM_006061.4) and Ensembl identifier ENSG00000096006 for verification. Then the target fragment was amplified by PCR, digested with the pLVX-IRES-Puro lentiviral expression vector (qincheng BIO, QCP0249), and ligated with T4 DNA ligase to clone the full-length human CRISP3, which was transformed into host bacteria to verify the recombinant clone. The shRNA plasmids targeting CRISP3 were designed and synthesized by Genomeditech (Shanghai, China). The specific shRNA target sequences used were as follows:

sh1# GAAGCTCACATTAACCTGTAA TTACAGGTTAATGTGAGCTTC.

sh5# GCCTCCTGCAATTGTTCAAAC GTTTGAACAATTGCAGGAGG.

293T cells, seeded in 10 cm culture dishes and grown to 60–70% confluence, were first pre-conditioned by replacing the existing medium with 4 mL of pre-warmed, serum-free DMEM, followed by a return to the incubator. For the transfection complex, two solutions were prepared in sterile microcentrifuge tubes. Solution A contained 500 μL of serum-free DMEM combined with 8 μg target plasmid + 2.67 μg pMD2.G + 5.34 μg psPAX2, and solution B was prepared by diluting 48 μL of PEI MAX (Polysciences, Warrington, PA, USA) in serum-free DMEM, and the solutions were mixed and incubated at room temperature for 5 min. Solution B was then added to Solution A and mixed thoroughly, and the combined mixture was incubated at room temperature for 30 min to allow for complex formation. The resulting transfection complexes were added dropwise evenly onto the pre-conditioned 293T cells. After 48 h of incubation, the viral supernatant was collected and filtered through a 0.45 μm filter into sterile ultracentrifuge tubes. The filtered supernatant was centrifuged at 15,000 rpm for 2 h at 4 °C. The supernatant was carefully discarded, and the viral pellet was resuspended in 2 mL of DMEM. Cells were infected with a medium containing lentivirus and supplemented with 1.2 μg/mL (HEY), 1.8 μg/mL (OVCAR8), and 1.6 μg/mL (OV90) polybrene by Genomeditech (Shanghai, China) for 48 h. After successful transfection, cell lines were screened with puromycin for 30 days.

2.7. RNA Sequencing Analysis

Transcriptomic profiling via RNA sequencing (RNA-seq) was carried out on HEY cells following transfection with OENC-CRISP3 (control) and OE-CRISP3 (overexpression) constructs. The total amount of RNA used for RNA sequencing was 1000 ng. All steps for sample preparation adhered strictly to the manufacturer’s protocol (Novogene Co., Ltd., Tianjin, China). Total RNA served as the starting material for library preparation. Polyadenylated mRNA was enriched using Oligo(dT) magnetic beads and subsequently fragmented randomly in Fragmentation Buffer with divalent cations. The purified double-stranded cDNA was then subjected to end repair, A-tailing, and ligation to sequencing adapters. cDNA fragments of approximately 370–420 bp were selected using AMPure XP beads, followed by PCR amplification. The PCR products were further purified with AMPure XP beads to obtain the final library.

The reference genome index was constructed using HISAT2 (v2.0.5), and the paired-end clean reads were then aligned to this reference genome using the same software. Subsequently, featureCounts (v1.5.0-p3) was employed, with a dilution of 1:600, to count the reads mapped to each gene. Differential expression analysis between comparative groups was performed using the DESeq2 software (v1.20.0). The resulting p-values were adjusted using the Benjamini and Hochberg’s method to control the false discovery rate (FDR). A significant differential expression threshold was set at an adjusted p-value (padj) ≤ 0.05 and an absolute log₂ fold change ≥ 1. Finally, Gene Ontology (GO) enrichment analysis of the differentially expressed genes (DEGs) was conducted using the clusterProfiler package (v3.8.1). The same package was also utilized for statistical enrichment analysis of the DEGs in the KEGG, Reactome, Disease Ontology (DO), and DisGeNET pathways.

The RNA-seq data from this study of CRISP3-overexpressing HEY cells were submitted to the Gene Expression Omnibus with sequence number GSE316844.

2.8. Western Blotting

Cells were washed three times with ice-cold PBS. Subsequently, an appropriate volume of RIPA lysis buffer (Beyotime, Shanghai, China) (containing 1× protease and phosphatase inhibitors (NCM Biotech, Suzhou, China)) was added. The supernatant was carefully transferred to a new pre-chilled centrifuge tube. Protein concentration was determined using the BCA assay (Takara, Japan), and samples containing 20 µg of total protein were prepared. An appropriate volume of 5× SDS-PAGE loading buffer was added, followed by denaturation at 95–100 °C for 10 min using a metal bath or water bath. After brief centrifugation, the denatured samples were separated on a 10% gel and electrophoretically transferred onto a polyvinylidene fluoride (PVDF) membrane (Merck, Rahway, NJ, USA). After blocking with 5% non-fat milk in TBST at room temperature (RT) for 1 h, membranes were incubated with the indicated primary antibodies overnight at 4 °C. After three washes with TBST, the membranes were probed with an HRP-conjugated goat anti-rabbit IgG (1:10,000, Proteintech, Wuhan, China) for 1 h at room temperature, with subsequent visualization on a Tanon chemiluminescence imager (Shanghai, China). GAPDH and Vinculin were the internal controls. The antibodies used were CRISP3 antibody (#14847-1-AP, 1:2000, Proteintech), GAPDH antibody (#60004-1-Ig. 1:10,000, Proteintech), Vinculin (#26520-1-AP, 1:10,000, Proteintech), AKT antibody (#sc-81434, 1:500, Santa Cruz, Dallas, TX, USA), and p-AKT antibody (#sc-514032, 1:500, Santa Cruz).

2.9. Cell Proliferation and EdU Assays

HEY-oeNC vs. HEY-OE (1.5 × 10³), OVCAR8-oeNC vs. OVCAR8-OE (2 × 10³), HEY-shNC vs. HEY-sh1 and HEY-sh5 (1.5 × 10³), OVCAR8-shNC vs. OVCAR8-sh1 and OVCAR8-sh5 (2 × 10³), OV90-shNC vs. OV90-sh1 and OV90-sh5 (2.5 × 10³) cells were seeded into 96-wellplates (Corning, Corning city, NY, USA) and incubated at 37 °C. At each indicated time (0 h, 24 h, 48 h, 72 h or 96 h), cell viability was assessed using CCK-8 (Med Chem Express, Monmouth Junction, NJ, USA) following an additional 2 h of incubation. The absorbance was measured at 450 nm. Cell proliferation was tested according to the manufacturer’s protocol. The EdU kit (CX003, Beyotime Shanghai, China) was used for EdU detection according to the manufacturer’s guidelines.

2.10. Transwell and Cell Wound-Healing Assays

Cell migration and invasion assays were performed using Transwell chambers (8 μm pore, Corning, NY, USA). For the migration assay, the upper chamber was left uncoated, whereas for the invasion assay, it was pre-coated with 70 µL of Matrigel (Corning, USA). The constructed ovarian cancer cell lines with stable overexpression or knockdown of CRISP3, including HEY (2 × 10⁴), OVCAR8 (4 × 10⁴) and OV90 (6 × 10⁴), were resuspended in serum-free medium. A 200 µL aliquot of the cell suspension was seeded into the upper chamber, and the lower chamber was filled with 600 µL of complete medium containing 10% FBS. After 24 h of incubation, cells on the lower surface were fixed with 4% paraformaldehyde and stained with 0.1% crystal violet (Beyotime, Shanghai, China). All images were captured using a Nikon Eclipse Ti2 inverted microscope under bright-field mode with a 10× objective lens. Image analysis was systematically performed using ImageJ software (National Institutes of Health, MD, USA). The number of transmigrated cells in each field was counted via the automatic or semi-automatic counting function of the software. The final data, expressed as the average number of transmigrated cells per treatment group, were statistically analyzed and used for graphical presentation.

A wound healing assay was performed using HEY and OVCAR8 cells to evaluate cell migration. Cells were seeded in 6-well plates (2 × 10⁵ cells/well), scratched with a 200 µL pipette tip after adhesion, and then cultured in serum-free medium for 48 h. Wound closure was imaged at 0 h and 48 h and quantified using ImageJ.

2.11. Cell Viability Assay

HEY (2.5 × 10³ cells/well) and OVCAR8 (3 × 10³ cells/well) cells were seeded onto 96-well plates. After 24 h of culture, the cells were treated with various concentrations of Esculetin. Following 48 h of treatment, 10 μL of CCK-8 reagent was added to each well, and the plates were incubated at 37 °C in a 5% CO₂ incubator for 2 h. The absorbance at 450 nm was finally measured using a microplate reader.

2.12. Animal Study

Mice: Female C57BL/6 mice aged 5–6 weeks were purchased fromJSJ Laboratory Animal Co., Ltd. (Shanghai, China). The animals were acclimatized for at least one week prior to experiments at the Laboratory Animal Center of Tongji University. A total of 16 mice were randomly housed in groups of four per cage within individually ventilated cages. All animals were maintained under specific pathogen-free (SPF) conditions, with a temperature of 22.5 ± 1.5 °C, relative humidity of 50 ± 10%, and a 12 h light/dark cycle. Mice had ad libitum access to water and standard rodent diet (Catalog No.: 1010011, Xietong Pharmaceutical Bioengineering Co., Ltd. (Nanjing, China)). Environmental enrichment consisted of corncob bedding, paper tube tunnels, and a large tissue sheet for nesting. Animal health and welfare were monitored daily. All animal care and experimental procedures were conducted following NIH guidelines and approved by the Animal Ethics Committee of Tongji University.

Orthotopic Ovarian Tumor Model: To investigate the metastasis of ovarian cancer in vivo, animal experiments were performed. Adherent cells were first separated into single cell suspensions by trypsin. After counting, the cells were resuspended in sterile phosphate-buffered saline (PBS) and placed on ice. Mice were anesthetized via intraperitoneal injection of 1.25% Avertin (10 mL/kg). Following sterilization, a laparotomy was performed via a dorsal incision to expose both ovaries. A suspension of 50 µL containing 2.0 × 10⁶ ID8 cells (mouse ovarian epithelial cancer cell line) [21,22] mixed with 50 µL Matrigel (Corning Inc., USA) was injected orthotopically into each ovary; the specific operation was as follows: a mixed cell suspension of 100 µL was drawn using an insulin syringe fitted with a 30 G needle. The needle was then inserted into the ovarian parenchyma at an angle of 20 degrees to a depth of approximately 1–2 mm. The injection was performed very slowly. Upon completion, the needle was kept in place for 30 s to allow the Matrigel to solidify. The surgical incision was closed using sterile suture. Mice were randomly assigned to one of the following four treatment groups (n = 4 per group) using a computer-generated randomization sequence: (A) normal control group, (B) group sacrificed for sampling on day 10, (C) group sacrificed for sampling on day 20, and (D) group sacrificed for sampling on day 30. The activity status, general health condition, and body weight changes in the animals were monitored daily. Body weight was measured by two independent researchers using a microelectronic balance. The researchers were blinded to the treatment group assignments during outcome assessments.

Four mice were sacrificed at 10 (early), 20 (middle) and 30 (late) days after surgery. Mice were placed in an empty euthanasia chamber, and carbon dioxide was introduced at a flow rate displacing 20–30% of the chamber volume per minute, allowing for a gradual increase in concentration. The mice were then maintained in a ≥70% carbon dioxide atmosphere for 2 min to ensure complete respiratory arrest. Following chamber removal, death was confirmed by cervical dislocation to prevent recovery of consciousness. Following confirmation of death, normal tissues and tumor tissues from the corresponding sites were harvested for subsequent qRT-PCR, H&E staining, and immunohistochemical analyses. GraphPad Prism software (Version 8.0.2, GraphPad Software, Inc., San Diego, CA, USA) was used for statistics.

Ethical Considerations: The housing and care of animals complied with the regulations for the protection of animals used for scientific purposes at Tongji University and were in accordance with the recommendations in the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH). The study protocol was approved by the Animal Experiment Ethics Committee of Tongji University (Approval No. TJBG19826101). In vivo experiments are reported in compliance with the ARRIVE guidelines [23].

2.13. Statistical Analysis

All data are presented as the mean ± S.D. from at least 3 independent experiments using GraphPad Prism software (Version 8.0.2, GraphPad Software, Inc., USA). Analysis of variance, two-tailed Student’s t-tests, One-way ANOVA and Wilcoxon rank sum tests were used to calculate p values. A p < 0.05 was considered statistically significant.

All uncropped full blots of Western blotting have been provided in Section S2 in Supplementary Materials.

3. Results

3.1. CRISP3 Is Downregulated and Correlated with Clinical Characteristics in HGSOC

We first found that CRISP3 was significantly downregulated in HGSOC tissues by analyzing the GSE69428 database (Figure 1A), and also found that CRISP3 protein expression was downregulated in ovarian cancer tissues by analyzing the TCGA database (Figure 1B). Further, based on a heatmap of the GSE69429 database, we found that CRISP3 expression was gradually decreased from the fallopian tube epithelium (FT) to HGSOC tissues with the development of HGSOC (Figure 1C). We also found, through the analysis of the cBioPortal database, that the expression of CRISP3 was the highest in normal fallopian tubes and gradually decreased with the occurrence and progression of HGSOC (Figure 1D). CRISP3 also decreased in various cancers, such as head and neck squamous cell carcinoma (HNSC), kidney chromophobe carcinoma (KICH), lung adenocarcinoma (LUAD), thyroid carcinoma (THCA), and uterine corpus endometrial carcinoma (UCEC) (Figure S1A).

RNA-seq sequencing data of 16 normal tissues and 56 ovarian cancer tissues collected in our hospital were analyzed to verify the expression level of CRISP3 in ovarian cancer tissues. The results revealed that CRISP3 was markedly downregulated in tumor tissues compared to normal controls (Figure 1E). qRT-PCR further confirmed a significant reduction in CRISP3 expression levels in HGSOC tissues (Figure 1F). Consistent with these findings, immunohistochemistry demonstrated significantly decreased CRISP3 protein expression in HGSOC tissues (Figure 1G–I). We subsequently investigated the relationship between CRISP3 expression, patient prognosis across various clinical characteristics, and age at diagnosis in the TCGA cohort. The analysis indicated that high CRISP3 expression predicted a more favorable prognosis in patients without lymph node invasion (Figure S1B) or with venous invasion (Figure S1C). Furthermore, CRISP3 expression was significantly lower in patients initially diagnosed with ovarian cancer at ≥60 years of age compared to those diagnosed at <60 years (Figure S1D), suggesting potential clinical utility for CRISP3 as a prognostic biomarker.

3.2. CRISP3 Inhibits the Proliferation, Migration, and Invasion of HGSOC Cells

We initially assessed CRISP3 expression in normal ovarian epithelial cells and OC cell lines, showing that its expression was decreased in the latter (Figure S3A,B). Further, we transfected HEY and OVCAR8 cells with two distinct shRNA plasmids targeting CRISP3, and the knockdown efficiency of shCRISP3 was verified by qRT-PCR and Western blotting (Figure 2A,B). The CCK-8 and EdU assay results demonstrated that compared with the control group, HGSOC cells transfected with shCRISP3-1 or shCRISP3-5 exhibited significantly enhanced proliferative vigor and DNA synthesis rates (Figure 2C–E), indicating that CRISP3 knockdown promotes cell proliferation. Colony formation assays revealed that CRISP3 knockdown increased the number of colonies formed (Figure S2A). Furthermore, wound healing scratch assays and Transwell assays showed that downregulation of CRISP3 substantially enhanced the migration and invasion of HGSOC cells (Figure 2F and Figure S2B). To verify the above experimental results, we further introduced shCRISP3-1 or shCRISP3-5 into OV90 cells. Consistent with the above results, CRISP3 promoted the proliferation, migration, invasion, and colony formation of OV90 cells, when CRISP3 was downregulated (Figure S4A–F).

Next, we successfully constructed HEY and OVCAR8 cell lines overexpressing CRISP3 (Figure 2G,H). We found that CRISP3 overexpression significantly suppressed the proliferative capacity and colony-forming ability of both HEY and OVCAR8 cells (Figure 2I and Figure S2C–E). Furthermore, it impaired their migratory and invasive capabilities (Figure 2J–L). Collectively, these findings demonstrate that CRISP3 acts as a suppressor of proliferation, migration, and invasion in HGSOC cells.

3.3. CRISP3 Gradually Decreases with OC Metastasis In Vivo

The schematic of the in vivo experimental procedure is depicted in Figure 3A. Intriguingly, we observed a progressive decrease in CRISP3 expression within primary ovarian tumors over time (Figure 3B). Compared to primary tumors, CRISP3 expression was significantly reduced in metastatic tissues (Figure 3C). Furthermore, we specifically analyzed CRISP3 expression across different metastatic sites and found the lowest levels in colon and omental metastases (Figure 3D–F). Consistent with this, immunohistochemical staining analysis confirmed decreased CRISP3 expression in tumor and metastatic tissues relative to normal tissue (Figure 3G,H). Taken together, our mouse model further validated the downregulation of CRISP3 in both tumoral and ovarian cancer tissues and metastatic tissues.

3.4. CRISP3 Inhibits Cell Proliferation, Migration, and Invasion by Targeting the PI3K/AKT Pathway

To investigate the global impact of CRISP3 overexpression on the transcriptome of HGSOC cells, we stably overexpressed CRISP3 in the HEY cell line (the oeCRISP3 group), using cells transfected with an empty vector as the control (the vector group). Subsequently, RNA-seq analysis was performed on both groups to identify differentially expressed genes (DEGs). A total of 301 significantly differentially expressed genes (DEGs) were identified by RNA-seq analysis, comprising 115 upregulated and 186 downregulated genes (Figure 4A). A heatmap displaying the 25 most significantly upregulated and downregulated genes is presented (Figure S5A), highlighting ID2 as being strongly upregulated. qRT-PCR performed in HEY, OVCAR8 and OV90 cells with stable CRISP3 knockdown or overexpression validated these findings (Figure S5B,C).

Then, the GO enrichment analysis of significantly upregulated differentially expressed genes showed that CMG complex, DNA replication origin binding, and DNA replication were significantly enriched (Figure 4B). Actin-based cell projection and cell–substrate junction were significantly enriched among the downregulated genes (Figure 4C). Subsequently, we performed Reactome analysis of the upregulated DEGs, which revealed that DNA replication correlated with CRISP3 expression (Figure 4D), and DisGeNET analysis of down-regulated differentially expressed genes showed that autism spectrum disorders were associated with the overexpression of CRISP3 (Figure S5D). To identify CRISP3-related signaling pathways in HGSOC, the PI3K/AKT signaling pathway was identified as the most significantly enriched pathway by KEGG and GSEA analyses of genes down-regulated after CRISP3 overexpression (Figure 4E,F). To further validate the role of CRISP3 in regulating the PI3K/AKT pathway, we examined the p-AKT level by Western blotting in HEY and OVCAR8 cells with CRISP3 overexpression or knockdown. The results showed that CRISP3 overexpression led to reduced levels of p-AKT (Figure 4G). Conversely, endogenous knockdown of CRISP3 in HEY cells resulted in elevated p-AKT protein levels, while total AKT levels remained unchanged (Figure 4H). Together, these results show that CRISP3 inhibits the biological phenotype of HGSOC through the PI3K/AKT pathway.

Esculetin, an active compound primarily isolated from the bark of Fraxinus rhynchophylla, functions as a PI3K/Akt pathway inhibitor with demonstrated antioxidant, anti-inflammatory, and anti-tumor activities [24]. Therefore, we first evaluated its effect on cell viability in HEY and OVCAR8 cells using the CCK-8 assay. The 48 h half-maximal inhibitory concentration (IC₅₀) values were determined to be 78.20 μM for HEY cells and 55.81 μM for OVCAR8 cells (Figure S6A). The Esculetin concentrations used in subsequent experiments were selected based on these 48 h IC₅₀ values established in the present study. Further, we investigated the effect of Esculetin on cellular phenotypes in CRISP3-knockdown HEY and OVCAR8 cells. The results showed that CRISP3 knockdown significantly promoted cell proliferation compared to the shNC + DMSO group; however, this promotive effect was markedly attenuated upon Esculetin treatment (Figure S6B). Similarly, in migration and invasion assays, CRISP3 knockdown substantially enhanced migratory and invasive abilities relative to shNC + DMSO controls, whereas the addition of the PI3K/Akt pathway inhibitor Esculetin significantly suppressed these enhanced capacities (Figure S6C,D). In summary, these findings demonstrate that CRISP3 overexpression is closely associated with the activation of the PI3K/AKT signaling pathway, suggesting that CRISP3 may contribute to the progression of HGSOC by modulating this pathway.

3.5. Analysis of CRISP3 Expression in Relation to Tumor Immune Microenvironment and Immunotherapy

The association between CRISP3 expression and the tumor microenvironment (TME) in serous ovarian cystadenocarcinoma was further assessed utilizing the ESTIMATE algorithm. CRISP3 expression exhibited significant positive correlations with both ImmuneScore and ESTIMATEScore (Figure 5A). Consistently, CRISP3 levels were positively associated with the abundance of multiple immune cell populations, such as cytotoxic cells, DCs, macrophages, neutrophils and Th1 cells (Figure 5B). Furthermore, ssGSEA revealed that tumors with high CRISP3 expression displayed greater infiltration of DCs, macrophages, neutrophils and Th1 cells compared to those with low CRISP3 expression (Figure 5C). To further investigate the correlation between CRISP3 expression and immune cell infiltration, multiplex immunofluorescence (mIF) staining was employed to examine the co-expression of CRISP3 with key immune cell markers in both normal and tumor tissues. The selected markers included CD8 for cytotoxic T cells, CD68 for macrophages, and MPO for neutrophils (Figure S7A). Consistent with the immunohistochemistry results, both the number of CRISP3-positive cells and the mean fluorescence intensity of CRISP3 were significantly lower in tumor tissues compared to normal tissues (Figure S7B). Furthermore, the analysis revealed a significant reduction in the infiltration of CRISP3 + CD8+ T cells in tumor tissues relative to normal tissues. Conversely, a marked increase in the infiltration of neutrophils and macrophages was observed in tumor tissues (Figure S7C). Finally, within tumor tissues, CRISP3 expression was found to be predominantly associated with greater macrophage infiltration. Notably, a higher number of CRISP3-positive macrophages was identified in the tumor stroma (Figure S7A,D).

To further explore the correlation between CRISP3 and immunotherapy, we investigated the IMvigor210 cohort [25]. The results showed that CRISP3 upregulation improved survival to some extent (Figure 5D). Further analysis revealed that lower CRISP3 was associated with the immune desert type (Figure 5E). Finally, we observed that high CRISP3 expression in patients had a more positive response to immunotherapy (Figure 5F–H). In summary, these findings indicate that CRISP3 is associated not only with immune cell infiltration in the ovarian cancer microenvironment, but also with responsiveness to immunotherapy. However, its precise mechanistic role within this immune context and its full relevance to ovarian cancer immunotherapy require further investigation.

4. Discussion

Prior studies have established CRISP3 as a secreted, cysteine-rich protein expressed in exocrine tissues and cancer (e.g., prostate, salivary gland) and detectable in plasma and granules of neutrophils, confirming its extracellular nature [26]. Previous studies have shown that CRISP3 acts directly on prostate epithelial cells as a pro-tumor factor in the prostate [27]. Direct transcriptional regulation of CRISP3 by ERG (TMPRSS2-ERG fusion) has also been shown, supporting links between CRISP3 and molecularly defined prostate cancer subtypes [28]. However, the specific mechanism of CRISP3 in HGSOC is still unclear. In the present study, we observed markedly reduced CRISP3 expression in HGSOC tissues, corroborating previous findings [29,30]. Earlier studies have reported intense CRISP3 staining within ovarian vasculature [17]; consistently, our data demonstrated that elevated CRISP3 expression portended a favorable prognosis among patients with serous cystadenocarcinoma, patients lacking lymph vascular invasion or exhibiting venous invasion, and patients younger than 60 years.

CRISP3 downregulation attenuates breast cancer cell migration and invasion [14], and is intimately associated with the initiation, progression, and metastasis of esophageal squamous cell carcinoma and non-small-cell lung cancer [31,32]. Others studies have demonstrated high CRISP3 expression in benign epithelial ovarian tumors and markedly reduced levels in malignant ovarian lesions [17]. Our findings extend these observations by delineating the functional impact of CRISP3 on the biological behavior of HGSOC. In vitro functional assays revealed that CRISP3 silencing significantly enhanced the proliferative, migratory, and invasive capacities of HGSOC cells. Nevertheless, additional in vivo studies are warranted to comprehensively elucidate the tumor-suppressive role of CRISP3.

Inhibitor of DNA binding 2 (ID2) is a member of the helix–loop–helix (HLH) family of transcriptional regulators that functions primarily by inhibiting basic HLH transcription factors and modulating cell differentiation, proliferation, and tumor progression. ID2 has been implicated in various malignancies, where it can exert context-dependent oncogenic or tumor-suppressive effects [33,34]. Thioredoxin-interacting protein (TXNIP), also known as VDUP1, is a redox-regulating protein that binds to and inhibits thioredoxin, thereby influencing oxidative stress responses, metabolic regulation, and apoptosis. TXNIP is frequently reported to be a tumor suppressor and is involved in multiple signaling pathways, including ROS regulation and inflammasome activation [35,36]. Although both ID2 and TXNIP are well characterized in cancer biology, limited studies have explored their regulation by extracellular or secreted proteins. Therefore, our findings suggest a potential novel regulatory axis linking secreted factor–mediated signaling to intracellular transcriptional and redox control mechanisms.

The PI3K/AKT pathway plays a key role in OC proliferation and progression [37]. Furthermore, evidence suggesting that ID2 inhibits cancer progression and metastasis through this pathway [38]. Integrated RNA-seq and functional analyses consistently demonstrated that genes downregulated following CRISP3 overexpression were significantly enriched in the PI3K/AKT signaling pathway. Furthermore, CRISP3 overexpression suppressed p-AKT levels, whereas CRISP3 knockdown upregulated p-AKT expression. Moreover, the enhanced proliferation, migration, and invasion resulting from CRISP3 knockdown were significantly reversed by treatment with Esculetin, a specific PI3K/AKT pathway inhibitor. The RNA-seq data enrichment analysis also revealed that RAS and MAPK signaling pathways were activated upon CRISP3 overexpression. The simultaneous activation of these key cancer marker pathways and AKT signaling suggests that CRISP3 may contribute to HGSOC progression through a coordinated network involving multiple pro-tumorigenic signaling axes. This highlights the potential complexity of CRISP3-mediated oncogenic effects and warrants further investigation of the specific upstream receptors and downstream effector mechanisms.

However, CRISP3 is a secreted glycoprotein present in exocrine secretions and plasma and can form complexes with plasma proteins (such as alpha1B-glycoprotein), implying roles in extracellular signaling or modulation of the tumor microenvironment rather than direct intracellular action [26]. This extracellular location suggests that observed effects on intracellular signaling (including AKT) are likely indirect, potentially mediated through receptor engagement, extracellular matrix interactions, or paracrine loops. Given CRISP3’s secreted nature, it is unlikely to directly engage intracellular signaling components. However, one recent cancer context (breast cancer) links CRISP3 to an IL-17/AKT axis, where CRISP3 expression correlates with activation of AKT signaling via intermediate extracellular mediators [39]. Secreted matricellular proteins such as periostin and SPARC have been shown to engage integrin and focal adhesion complexes that activate the PI3K/AKT and MAPK pathways [40]. Similarly, extracellular matrix components like tenascin-C and fibronectin modulate intracellular kinase cascades via integrin-mediated mechanotransduction [41]. Paracrine signaling loops involving chemokines such as CXCL12 have been demonstrated to activate AKT and ERK signaling in cancer contexts [42], and extracellular proteases can liberate ECM-sequestered growth factors to augment intracellular pathways [43]. These examples support the notion that the extracellular localization of CRISP3 can indirectly modulate intracellular signaling dynamics, including AKT, through receptor engagement, ECM interactions, and paracrine mechanisms. While our data suggest a functional association with AKT pathway modulation, further studies (e.g., receptor identification, extracellular signaling assays) are needed to confirm the mechanistic basis.

Studies have localized CRISP3 to cells of both the innate and adaptive immune systems, implicating it in antimicrobial defense and immune tolerance [4,11,44], and macrophages and neutrophils are the main immune effector cells with anti-tumor activity in the innate immune system [45]. The tumor immune microenvironment plays a pivotal role in HGSOC progression. Recent research highlights a complex and dynamic interplay wherein immune components, including macrophages, neutrophils, CD8+ T cells, and regulatory T cells, collaborate to drive disease advancement across distinct developmental phases of HGSOC [46,47,48]. Our analyses revealed a significant positive correlation between CRISP3 expression and the abundance of multiple immune infiltrates—including cytotoxic cells, dendritic cells (DCs), macrophages, neutrophils and CD8⁺ T cells. On the basis of the composition and activation state of tumor-infiltrating immune cells, tumors can be broadly classified as immunologically “hot” or “cold” [49]. Further analysis of an immunotherapy cohort in metastatic urothelial carcinoma revealed that CRISP3 expression was associated with an inflamed immunophenotype, and patients with high CRISP3 levels exhibited significantly improved responses to immunotherapy. However, how CRISP3 modulates tumor immune cell infiltration and its role in anti-tumor immunity in OC require further investigation.

Despite our findings that establish CRISP3 as a tumor suppressor in HGSOC, this study has several limitations. The precise molecular mechanism by which CRISP3 may regulate the PI3K/AKT pathway remains to be fully elucidated. In addition, the evidence linking CRISP3 to the immune microenvironment and immunotherapy response is currently confined to online database analyses, lacking concrete experimental and mechanistic investigations. However, addressing these aspects represents a primary direction for our future research and exploration.

5. Conclusions

Our findings establish CRISP3 as a tumor suppressor in HGSOC. CRISP3 expression was also associated with the infiltration of central granulocytes and macrophages in ovarian cancer. These data collectively suggest that CRISP3 represents a promising therapeutic target and prognostic biomarker for HGSOC.