CircNFIX stimulates the proliferation, invasion, and stemness properties of ovarian cancer cells by enhancing SH3RF3 mRNA stability via binding LIN28B

Abstract

We aimed to study the regulatory roles and mechanism of circular nuclear factor IX (circNFIX) in cancer growth and stemness properties of ovarian cancer (OC). CircNFIX and SH3RF3 levels in OC tissues and cells were tested by quantitative real‐time PCR. RNase R treatment quantified circNFIX RNA stability. Molecular interaction among circNFIX, LIN28B, and SH3RF3 was predicted by bioinformatics software and validated through RNA immunoprecipitation (RIP) assay. The gain‐ or loss‐experiments of circNFIX on capabilities of metastasis and stemness in vitro were assessed using Cell Counting Kit‐8, Transwell, western blot, and sphere‐formation assays. CircNFIX and SH3RF3 were markedly elevated in OC tissues and OC cells. Knocking down circNFIX repressed the proliferation, migration, invasion, and stemness properties of A2780 and SKOV3 cells. The RIP assay verified the direct binding relationship between LIN28B, circNFIX, and SH3RF3. Additionally, overexpression of circNFIX elevated the SH3RF3 expression, while this effect was reversed by LIN28B silence. Rescue experiments demonstrated that the overexpression of SH3RF3 reversed the knockdown of circNFIX on OC cells' proliferation, metastasis, and stemness properties. CircNFIX improved the mRNA stability and translation of SH3RF3 via recruiting LIN28B, thus promoting the proliferation, invasion, and stemness properties of OC cells in vitro.

Abbreviations

ANOVA

one‐way analysis of variance

ATCC

American type culture collection

CCK‐8

Cell Counting Kit‐8

circRNA

circular RNA

FBS

fetal bovine serum

NFIX

nuclear factor IX

OC

ovarian cancer

POSH2

plenty of SH3 domains protein 2

qPCR

quantitative real‐time PCR

RBP

RNA‐binding protein

RIP

RNA immunoprecipitation

RNase R

Ribonuclease R

SD

standard deviation

SFE

sphere formation efficiency

SH3

Src homology 3

SH3RF3

SH3 domain containing ring finger 3

shRNA

short hairpin RNA

WB

western blot

1. INTRODUCTION

Ovarian cancer (OC) is an important reason for cancer‐related death in women worldwide and is one of the most malignant cancers. ¹ OC may originate from any histological part of the ovary, including epithelium, stroma, or germ cells, ² and is associated with a high incidence of metastasis, recurrence, and chemoresistance. ³ The surgery and platinum‐based chemotherapy are the mainstays of treatment for OC. However, patients always relapse because of chemotherapy resistance within a few years of initial therapy. ⁴ Residual cancer stem cells after chemotherapy are the initiators of cancer resistance because drug‐resistant subsets of cancer stem cells drive cancer progression, metastasis, and ultimately disease recurrence. ⁵ Thus, exploring new molecular mechanisms of OC cell stemness properties and drug resistance regulation is of great value in finding effective targets to reverse drug resistance and improve chemotherapy efficacy.

Circular RNA (circRNA) is an endogenous, highly stable, single‐stranded closed circular RNA molecule that lacks a terminal 5′ cap and 3′ poly (A) tail. ⁶ CircRNAs can act as microRNAs and protein sponges, interact with translational protein regulators, and compete with pre‐mRNA splicing. ⁷ CircRNAs have received increasing scientific attention due to their key regulatory roles in various cancers' occurrence and development. In OC, Chen et al. reported that hsa_circ_0006404 and hsa_circ_0000735 modulated OC response to docetaxel treatment by modulating p‐GP expression. ⁸ Wang et al. showed circ_0000745 elevated by IGF2BP2 could facilitate the invasiveness and stemness of OC cells via miR‐3187‐3p/ERBB4/PI3K/AKT axis. ⁹ These studies suggested that circRNAs have unique functions in the diagnosis and treatment of OC. hsa_circ_0049658, with a full length of 695 bp, located at chromosome 19 (13,183,860–13,192,669), is back spliced by host gene nuclear factor IX (NFIX) mRNA exons 3–8 (10 exons in total) become. Previous studies have suggested circNFIX was elevated in non‐small cell lung cancer, liver cancer, and human glioma to function as an oncogene, and was involved in cancer metabolism, chemotherapy resistance, migration, and invasion. ¹⁰ , ¹¹ , ¹² However, there are few kinds of literature about circNFIX function and mechanism in OC stemness development.

SH3 domain containing ring finger 3 (SH3RF3), known as plenty of SH3 domains protein 2 (POSH2), has four Src homology 3 (SH3) domains and one Ring finger domain. ¹³ Zhang et al. studies showed that SH3RF3 promoted JNK‐JUN signaling activation and elevated PTX3 expression, which ultimately enhanced the function and mechanism of breast cancer cells stemness properties, and showed its high expression promoted tumor progression and predicted poor patient prognosis, suggesting that SH3RF3 was the potential diagnostic marker and therapeutic target in clinical breast tumor diagnosis and treatment. ¹⁴ Additionally, in thyroid cancer‐related reports, SH3RF3 expression was elevated, and plays an oncogene function in promoting tumor growth, metastasis, invasion, and epithelial–mesenchymal transition process. ¹⁵ Therefore, we wanted to explore whether SH3RF3 correlated with circNFIX in OC.

Lin‐28 homolog B (LIN28B) is an evolutionarily conserved RNA‐binding protein and a major regulator of embryonic stem cell self‐renewal. ¹⁶ Following various bioinformatics analyses, we discovered that LIN28B has several binding sites on the circNFIX and SH3RF3 RNA sequences. LIN28B has been widely confirmed to be involved in the maintenance of tumor stemness properties, tumor migration, and invasion in current tumor‐related research reports. ¹⁷ , ¹⁸ Based on the above reports, we proposed that circNFIX maintained and promoted the mRNA stability and translation of SH3RF3 by binding to LIN28B, thus facilitating stemness properties, proliferation, and migration of OC cells. Our study might be provided the theoretical basis for OC pathogenesis and new ideas for OC treatment.

2. MATERIALS AND METHODS

2.1. Collection of clinical samples

Cancer and paracancerous tissues from 48 cases of diagnosed OC patients in West China Second University Hospital, Sichuan University were collected. According to the article “Diagnosis and Management of Ovarian Cancer” published by Doubeni et al., ¹⁹ the inclusion criteria for OC patients were determined. None of the patients received drug therapy. After the samples were taken, they were transferred to the −80°C refrigerator for storage. This study obtained informed consent from all participants who provided samples and was approved by the Ethics Committee of West China Second University Hospital, Sichuan University.

2.2. Cell culture and treatment

Normal ovarian epithelial cells HOSEPiCs and OC cell lines 3AO, OVCAR3, A2780, and SKOV3 were provided by American type culture collection (ATCC). HOSEPiCs, OVCAR3, and 3AO were cultured in RMPI‐1640 medium (Procell). SKOV3 and A2780 were cultured in DMEM (Procell). They were all containing 10% fetal bovine serum (FBS; Sigma‐Aldrich) and 1% penicillin/streptomycin. All cells were normally cultured in an incubator with a temperature of 37°C, a humidity of 70%–80%, and 5% CO₂.

2.3. Cell transfection

Short hairpin RNA (shRNA) targeting circNFIX, SH3RF3, and LIN28B (sh‐circNFIX, sh‐SH3RF3, and sh‐LIN28B) as well as circNFIX overexpressing vectors (pcDNA3.1‐circNFIX), and their corresponding negative controls (sh‐NC, pcDNA3.1‐NC) were provided by Genepharma. Lipofectamine 3000 (Invitrogen) was used to perform transfection for 48 h. Then, the transfection efficiency of these RNAs was measured by quantitative real‐time PCR (qPCR).

2.4. RNA extraction and qPCR

TRIzol (Thermo Fisher Scientific) extracted total cellular RNA. cDNA was synthesized by a cDNA reverse transcription kit (Invitrogen). qPCR was conducted on ABI 7900 system using SYBR Premix Ex Taq Kit (TaKaRa). Using GAPDH as an internal reference gene, gene levels were calculated by the 2^−ΔΔCt method. Primers were as follows (5′‐3′): circNFIX (F): TTCCCCTCCACGTCCATCAT, (R): CGTTGGGCAGTGGTTTGATG; NFIX (F): CAGGCTGACAAGGTGTGGC, (R): GGCAGTGGTTTGATGTCCGC; SH3RF3 (F): CGGAATTCATGCTGCTCGGAGCGTCCTGGCTG, (R): CGGGATCCTCTCAGAAGCTCTCGACGAAG; LIN28B (F): GCCCCTTGGATATTCCAGTC, (R): TGACTCAAGGCCTTTGGAAG; GAPDH (F): GAATGGGCAGCCGTTAGGAA, (R): AAAAGCATCACCCGGAGGAG.

2.5. RNase R treatment

Total RNA (1 μg) extracted from OC cells was incubated with/without RNase R (3 U/μg, Epicenter) at 37 °C for 20 min. And the treated RNA was reverse transcribed into cDNA, and circNFIX and its host gene NFIX expressions were tested by PCR and qPCR.

2.6. Western blot

RIPA (Beyotime) extracted total protein from cells, and protein quantification was performed based on the BCA protein assay kit (#BL521A, Biosharp). And equal proteins were uploaded to 12% SDS‐PAGE (#MB2479, Meilunbio) for protein isolation, and the protein was followed and moved to the PVDF membrane. Next, the membrane was blocked with 5% nonfat milk solution for 1 h at room temperature. SH3RF3 (#ab184977, Abcam), LIN28B (#11965, CST), N‐cadherin (#66219‐1‐Ig, Proteintech), E‐cadherin (#20874‐1‐AP, Proteintech), SOX2 (#11064‐1‐AP, Proteintech), OCT4 (#60242‐1‐Ig, Proteintech), and GAPDH (#10494‐1‐AP, Proteintech) were incubated with the membrane at 4°C overnight. Next, the membrane was incubated with HRP‐conjugated secondary antibody (#ab7153, Abcam) at room temperature for 2 h. Afterward, bands were visualized with an immobilon ECL substrate kit (Millipore), imaged and analyzed with Chemiscope6100 software.

2.7. Cell Counting Kit‐8 assay

After indicated treatment, cells were seeded in a 96‐well plate (1 × 10⁴ cells/well), and cultured for 0, 12, 24, 48, and 72 h. Then Cell Counting Kit‐8 (CCK‐8) solution (10 μl/well, Dojindo) was added. Absorbance (450 nm) was analyzed with the Bio‐Tek microplate reader (Heales) after incubation for 4 h.

2.8. Transwell assays

The migration of A2780 and SKOV3 cells was determined via transwell chambers (8 μm, Corning). After transfection, the cells were harvested and resuspended in a serum‐free medium. Cell suspension (200 μl) was pipetted into the upper chamber, and a complete medium with 10% FBS was added to the lower chamber. After 48 h, the culture medium in the chamber was discarded and washed twice with PBS. Cells on the upper chamber surface were wiped with a damp cotton swab. Then cells were fixed with 4% paraformaldehyde for 60 min and stained with 0.5% crystal violet for 20 min. Cells on the outer surface of the upper chamber were observed and photographed under an inverted microscope (Beijing Cnmicro). Matrigel‐coated transwell chambers (BD Biosciences) were used for invasion experiments, and the experimental procedure was the same as for migration experiments.

2.9. Sphere‐formation assay

After indicated transfection, OC cells were collected and washed twice with PBS and resuspended by DEME medium with FBS (1%), penicillin/streptomycin (1%), recombinant basic fibroblast growth factor (10 ng/ml), and recombinant epidermal growth factor (20 ng/ml). Cells were seeded in an ultra‐low adsorption cell culture plate (approximately 1000 cells/well in a six‐well plate) after counting. After culturing for about 15 days, the spheroid formation was observed and captured by a microscope.

2.10. Subcellular localization assay

Nuclear and cytoplasmic fractions were isolated through PARIS™ Kit (Life Technologies) in line with the working instruction. RNA was obtained using a TRIzol reagent, and circNFIX levels in cytoplasmic and nuclear levels were detected via qPCR. GAPDH and U6 acted as cytoplasmic and nuclear internal controls, respectively.

2.11. RNA immunoprecipitation assay

The binding relationship between LIN28B and circNFIX or SH3RF was verified through the EZMagna RNA immunoprecipitation (RIP) kit (Millipore). Shortly, cells were lysed with RIP lysis buffer at 4°C for 30 min and incubated with RIP buffer with magnetic beads coupled antibodies against IgG (#14708, CST) or LIN28B (#11965, CST) overnight, respectively. Afterward, the precipitated RNAs were isolated by TRIzol reagent (Thermo Fisher Scientific). The abundances of circNFIX and SH3RF were analyzed by qPCR.

2.12. Statistical analysis

Statistical analysis was conducted through Graphpad 8.0, and the data of three independent experiments were expressed as mean ± standard deviation (SD). A Student's t‐test was performed to analyze differences between two groups, and one‐way analysis of variance (ANOVA) was utilized to compare differences among multiple groups. The Pearson coefficient was used to analyze the correlation between circNFIX and SH3RF3. A p value of <0.05 was considered statistically significant.

3. RESULTS

3.1. CircNFIX expression was increased in OC tissues and cells

First, we analyzed the detailed information on circNFIX from circinteractome database (https://circinteractome.nia.nih.gov/index.html). Results showed that circNFIX was located in chromosome 19 (13,183,860–13,192,669), which was back‐spliced by exons 3–8 and had a length of 695 bp (Figure 1A). By designing full‐length PCR primers, we were able to obtain the full length of circNFIX (about 695 bp) amplification products, which were not degraded by RNAseR (Figure 1B). According to the qPCR results of divergent primer, circNFIX was more stable than linear NFIX mRNA (Figure 1C). Then, we detected circNFIX expression in clinical OC tissue samples. The result suggested that circNFIX level was elevated in OC tumor tissues than in  adjacent tissues (Figure 1D), and positively correlated to tumor metastasis (Figure 1E). According to the mean value of circNFIX expression, OC patients were allocated into low‐ and high‐ circNFIX expression groups. Clinical pathological features analysis was performed, and results were presented in Table 1, which revealed high‐ circNFIX level was positively correlated with higher tumor pathological grade and advanced FIGO stage. Furthermore, circNFIX level in OC cells (3AO, OVCAR3, A2780, and SKOV3) was elevated than normal ovarian epithelial cells HOSEPiCs, especially in A2780 and SKOV3 cells (Figure 1F). These results suggested that circNFIX was an abnormally overexpressed circRNA in OC.

FIGURE 1.

CircNFIX expression was increased in OC tissues and cells. (A) The structure of circNFIX. (B) PCR product of full‐length circNFIX was verified by agarose gel electrophoresis. (C) qPCR detected circNFIX and linear NFIX mRNA levels in A2780 and SKOV3 cells with or without RNase R treatment. (D) CircNFIX level in cancer and adjacent tissues from OC patients was examined by qPCR (N = 48). (E) CircNFIX level in cancer tissues with/without metastasis in OC patients was measured by qPCR. (F) CircNFIX level in normal ovarian epithelial cells HOSEPiCs and OC cells was tested by qPCR. circNFIX, circular nuclear factor IX; OC, ovarian cancer; qPCR, quantitative real‐time PCR. *p < 0.05, **p < 0.01, ***p < 0.001. N = 3.

TABLE 1.

Correlation of circNFIX and SH3RF3 levels with clinicopathological characteristics

Characteristics	circNFIX level		p Value	SH3RF3 level		p Value
	Low (n = 21)	High (n = 27)		Low (n = 24)	High (n = 24)
Age (years)
<50	15	13	0.1437	13	10	0.5639
≥50	6	14		11	14
FIGO stage
I/II	13	8	0.0401	15	6	0.0189
III/IV	8	19		9	18
Pathological grade
G1	3	8	0.0200	8	7	0.0297
G2	7	15		10	3
G3	11	4		6	14

Abbreviation: circNFIX, circular nuclear factor IX.

3.2. Knockdown of circNFIX inhibited OC cells proliferation, invasion, and stemness properties

To investigate circNFIX function in OC cell behaviors, we knocked down circNFIX by transfection with sh‐circNFIX in A2780 and SKOV3 cells. As evidenced by the qPCR assay, we found that sh‐circNFIX transfection obviously reduced circNFIX expression compared to the sh‐NC group (Figure 2A). Cellular functional experiments showed knocking down circNFIX remarkably reduced the proliferation, migration, and invasion abilities of A2780 and SKOV3 cells (Figure 2B–D). Moreover, the western blot assay further verified knocking down circNFIX‐promoted E‐cadherin levels and decreased N‐cadherin levels (Figure 2E). Consistently, the sphere‐formation assay showed that knocking down circNFIX markedly repressed stemness properties of A2780 and SKOV3 cells (Figure 2F). Meanwhile, circNFIX downregulation reduced the protein expressions of stem cell markers SOX2 and OCT4 (Figure 2G). These results indicated that knocking down circNFIX alleviated the proliferation, migration, invasion, and tumor stemness of OC cells.

FIGURE 2.

Knocking down circNFIX repressed the proliferation, invasion, and stemness properties of OC cells. Both sh‐circNFIX and sh‐NC were transfected into A2780 and SKOV3 cells. (A) CircNFIX level was evaluated by qPCR. (B) CCK‐8 detection of cell proliferation ability. (C, D) Transwell detection of cell migration and invasion abilities. (E) Western blot evaluated E‐cadherin and N‐cadherin levels. (F) Sphere‐formation assay examined stemness properties of A2780 and SKOV3 cells. (G) Western blot quantified the expressions of stem cell markers SOX2 and OCT4. CCK‐8, Cell Counting Kit‐8; circNFIX, circular nuclear factor IX; OC, ovarian cancer; qPCR, quantitative real‐time PCR. *p < 0.05, **p < 0.01, ***p < 0.001. N = 3.

3.3. CircNFIX maintained SH3RF3 expression by binding to LIN28B

On this part, we warranted exploring the potential mechanism between circNFIX and SH3RF3. First, the expression level of SH3RF3 in clinical OC tissues was examined by qPCR. As suggested in Figure 3A, SH3RF3 was greatly upregulated in OC tumor tissues than in adjacent tissues, and the high level of SH3RF3 was positively associated with OC metastasis (Figure 3B). Pearson analysis described that SH3RF3 expression presented a positive correlation with circNFIX in clinical OC tissues (Figure 3C). We also observed that SH3RF3 expression was suppressed in circNFIX downregulated cells (Figure 3D,E). Then, we further verified the subcellular localization of cirNFIX in OC cell lines. As indicated in Figure 3F, circNFIX was expressed both in cell nuclear and cytoplasmic extract. Venn diagram showed the circNFIX and SH3RF3 mRNA‐interacting RNA‐binding proteins, including DGCR8, FUS, FXR2, IGF2BP1/2/3, LIN28B, and TAF15, as analyzed by RBPsuite (http://www.csbio.sjtu.edu.cn/bioinf/RBPsuite/) and Starbase databases (http://starbase.sysu.edu.cn/) (Figure 3G). LIN28B has been widely confirmed to be involved in the maintenance of tumor stemness properties, tumor migration, and invasion in current tumor‐related research reports, ¹⁷ , ¹⁸ so we wanted to explore whether LIN28B was involved in the association between circNFIX and SH3RF3. Initially, we discovered that LIN28B mRNA and protein expression in OC cells was also increased, particularly in A2780 and SKOV3 cells (Figure S1–S2). RIP experiments revealed that both circNFIX and SH3RF3 were enriched in RNA complexes precipitated by LIN28B‐specific antibodies compared with the IgG group (Figure 3H), which suggested that LIN28B could bind to circNFIX and SH3RF3. Next, we overexpressed circNFIX and knocked down LIN28B in A2780 and SKOV3 cells by transfection pcDNA 3.1‐circNFIX vectors and sh‐LIN28B, respectively. After validation by qPCR assay, we found that pcDNA3.1‐circNFIX transfection increased circNFIX expression (Figure 3I), and sh‐LIN28B transfection dramatically reduced LIN28B mRNA and protein expression (Figure 3J,K). Knockdown of LIN28B also reduced SH3RF3 mRNA and protein expression (Figure S3–S4). Moreover, overexpressing circNFIX notably promoted SH3RF3 expression, while knockdown of LIN28B markedly reversed this effect of circNFIX overexpression (Figure 3L). Taken together, circNFIX could maintain SH3RF3 expression through binding to LIN28B.

FIGURE 3.

CircNFIX maintained SH3RF3 expression by binding to LIN28B. (A) qPCR determined SH3RF3 level in cancer and adjacent tissues from OC patients (N = 48). (B) qPCR tested SH3RF3 level in cancer tissues obtained from OC patients' metastases and those without metastases. (C) Pearson correlation analysis between circNFIX and SH3RF3. (D, E). SH3RF3 level in A2780 and SKOV3 cells with or without circNFIX knocked down was tested by qPCR and western blot, respectively. (F) Subcellular localization of circNFIX was tested by qPCR. (G) RBPsuite (http://www.csbio.sjtu.edu.cn/bioinf/RBPsuite/) and Starbase (http://starbase.sysu.edu.cn/) database analysis of RNA‐binding proteins interacting with circNFIX and SH3RF3 mRNA. (H) RIP detected the binding relationship between LIN28B and circNFIX or SH3RF3, respectively. (I) The overexpressing efficiency of circNFIX was evaluated by qPCR. (J, K) Knocked‐down efficiency of LIN28B was tested by qPCR and western blot. (L) SH3RF3 expression in A2780 and SKOV3 cells with circNFIX overexpression or circNFIX overexpression plus LIN28B silencing was evaluated by qPCR. circNFIX, circular nuclear factor IX; OC, ovarian cancer; qPCR, quantitative real‐time PCR; RIP, RNA immunoprecipitation. *p < 0.05, **p < 0.01, ***p < 0.001. N = 3.

3.4. Overexpression of SH3RF3 reversed the anti‐tumor effects mediated by circNFIX silencing

To verify the biological roles of SH3RF3 in circNFIX‐silenced OC cells, we co‐transfected sh‐circNFIX and pcDNA3.1‐SH3RF3 vectors into A2780 and SKOV3 cells. Results highlighted that SH3RF3 was inhibited in circNFIX‐silenced OC cells, and co‐transfection of pcDNA3.1‐SH3RF3 vectors greatly restored SH3RF3 expression (Figure 4A,B). Cell functional experiments showed that repression of circNFIX repressed the proliferation, migration, and invasion abilities of A2780 and SKOV3 cells, whereas these changes were reversed by overexpression of SH3RF3 (Figure 4C–E). Moreover, repression of circNFIX mediated the upregulation of E‐cadherin and the reduction of N‐cadherin were remarkably abolished by SH3RF3 overexpression (Figure 4F). Meanwhile, the sphere‐formation assay showed that SH3RF3 overexpression greatly enhanced stemness properties of A2780 and SKOV3 cells, which markedly diminished the inhibitory roles mediated by knockdown of circNFIX (Figure 4G). Western blot analysis further confirmed that the downregulated SOX2 and OCT4 expressions in circNFIX‐silenced cells were remarkably restored after SH3RF3 overexpression (Figure 4H). Taken together, SH3RF3 was involved in circNFIX‐mediated regulation of OC cells proliferation, metastasis, and stemness properties.

FIGURE 4.

Overexpression of SH3RF3 reversed the anti‐cancer effects mediated by circNFIX silencing. A2780 and SKOV3 cells were transfected with sh‐circNFIX alone or co‐transfected with sh‐circNFIX plus pcDNA3.1‐SH3RF3 vectors, respectively. (A, B) qPCR and western blot were applied to monitor SH3RF3 level. (C) CCK‐8 measured cell proliferation. (D, E) Transwell detected cell migration and invasion. (F) Western blot quantified E‐cadherin and N‐cadherin levels. (G) Sphere‐formation assay to detect stemness properties of A2780 and SKOV3 cells. (H) Western blot detected stem cell markers SOX2 and OCT4 expressions. circNFIX, circular nuclear factor IX; OC, ovarian cancer; qPCR, quantitative real‐time PCR; RIP, RNA immunoprecipitation. *p < 0.05, **p < 0.01, ***p < 0.001. N = 3.

4. DISCUSSION

OC is a primary ovarian malignant cancer with the highest mortality rate among fatal gynecological cancers. Chemoresistance and metastasis are major challenges in current OC treatment. ²⁰ Resistance to conventional platinum‐based therapy and metastasis has been attributed to a population of cells within the cancer called cancer stem cells. ²¹ Cancer stem cells interact with the microenvironment and release various inflammatory cytokines, chemokines, and matrix metalloproteinase, inducing invasion and spreading to distant organs of the body. ²² Therefore, understanding the different mechanisms that promote cancer stem cell survival and reproduction is vital to improve the therapeutic outcome of OC patients. In vitro results revealed that circNFIX in combination with LIN28B improves SH3RF3 mRNA stability and translation while also promoting OC cell proliferation, invasion, and stemness. This is the first time we have reported on the mechanism of circNFIX/LIN28B/SH3RF3 in OC.

With the rapid development of high‐throughput sequencing and bioinformatics analysis technologies, many circRNAs are discovered to be abnormally expressed in diseases. Besides, circRNAs have also become diagnostic and therapeutic targets in human diseases due to their high stability, conservation, and specificity. ²³ It has been reported that downregulation of hsa_circ_0026123 inhibited OC cell metastasis and proliferation through miR‐124‐3p/EZH2 pathway. ²⁴ Furthermore, Wang et al. showed that circ_0000745, upregulated by IGF2BP2, could promote OC cell invasiveness and stemness through the miR‐3187‐3p/ERBB4/PI3K/AKT axis. ⁹ As a newly discovered circRNA, circNFIX has been reported to function as the oncogene and participate in many cancers' progression. For example, circNFIX affected the cell invasion, migration, and proliferation of pituitary adenomas via sponging miR‐34a‐5p/CCNB1 signaling. ²⁵ Furthermore, circNFIX, plays as a competing endogenous RNA that accelerates hepatic carcinoma progression by regulating miR‐3064‐5p/HMGA2 axis, providing a therapeutic strategy for hepatic carcinoma intervention. ¹⁰ In gliomas, circNFIX promoted glioma progression via sponging miR‐34a‐5p/Notch signaling. ²⁶ We found that circNFIX level was facilitated in OC tissues and cells. Additionally, high‐ circNFIX expression was positively correlated with higher cancer pathological grade and later FIGO stage. Furthermore, knocking down circNFIX markedly reduced the proliferation, migration, invasion, and cancer stemness of A2780 and SKOV3 cells. This suggested that circNFIX acted cancer‐promoting roles in OC development.

CircRNAs play vital roles as microRNA sponges, regulators of gene splicing and transcription, RBP sponges, and protein/peptide translators. ²⁷ RBPs are a class of proteins involved in gene transcription and translation. The interaction between circRNAs and RBPs can regulate the occurrence, translation, transcriptional regulation, and extracellularly of target genes. ²⁸ LIN28 has been reported to bind mRNA and further recruit RNA helicase A to facilitate translation. ²⁹ For recent decades, LIN28B has been widely demonstrated to participate in the maintenance of cancer stemness properties, cancer migration, and invasion. ¹⁷ , ¹⁸ Moreover, studies have shown that LIN28B was a cancer stem cell‐related marker and maybe a drug target for treating epithelial OC. ³⁰ Blockade of circ‐SCMH1 has been reported to suppress acquired cisplatin resistance in oral squamous cell carcinoma cells through sponging miR‐338‐3p and regulating LIN28B. ³¹ In addition, circ‐FBXW12 directly targeted LIN28B to aggravate the development of diabetic nephropathy. ³² In this study, we found that LIN28B combined with circNFIX to facilitate SH3RF3 expression, which first elucidated the molecular association among circNFIX, LIN28B, and SH3RF3.

SH3RF3 is a Ring finger E3 ligase with Rac1 binding activity via part of the CRIB domain. ³³ SH3RF3 was upregulated in cancer stem‐like cells of breast cancer clinical cancers and cells and facilitated breast cancer cells' cancer stem‐like cell properties. ¹⁴ In thyroid cancers, SH3RF3 expression was also upregulated and promoted cancer metastasis. ¹⁵ This paper's RIP experiment confirmed the binding relationship between LIN28B and SH3RF3. By overexpressing circNFIX, SH3RF3 expression was upregulated, whereas this trend was reversed when LIN28B was knocked down, which suggested that SH3RF3 was a regulatory target of the circNFIX/LIN28B complex. Moreover, rescue experiments revealed that SH3RF3 overexpression remarkably enhanced cell proliferation, metastasis, and stemness capabilities in OC cells, which dramatically restrained the anti‐cancer effects mediated by circNFIX silence.

From the above findings, our study demonstrated that circNFIX combined with LIN28B to maintain the stabilization and translation of SH3RF3 mRNA, thus aggravating the proliferation, invasion, and stemness properties of OC cells in vitro. Our study provides new opportunities for the treatment of OC and can potentially prolong patient survival.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

Supporting information

FIGURE S1–S2. The mRNA and protein expression of LIN28B in different OC cells were detected by qPCR and western blot

FIGURE S3–S4. The knockdown effect of LIN28B on SH3RF3 mRNA and protein expression was measured by qPCR and western blot

Yu X‐Z, Yang B‐W, Ao M‐Y, Wu Y‐K, Ye H, Wang R‐Y, et al. CircNFIX stimulates the proliferation, invasion, and stemness properties of ovarian cancer cells by enhancing SH3RF3 mRNA stability via binding LIN28B . Kaohsiung J Med Sci. 2023;39(3):234–243. 10.1002/kjm2.12632