circDHX33 suppresses glycolysis, malignant proliferation, and metastasis in prostate cancer by interacting with RNA-binding protein IGF2BP2 to destabilize its protein

Abstract

Prostate cancer (PCa) is a malignant tumor characterized by dependence on androgens and enhanced glycolytic processes in response to the energy demands of rapid proliferation. This study delved into the role of circDHX33 interacting with IGF2BP2 in regulating the malignant behavior of PCa. circRNA expression data from PCa tissues and normal tissues were selected from the GEO database, and differentially expressed circRNAs were screened out. circDHX33 expression in clinical PCa samples was verified by RT-qPCR. Cellular experiments included cell culture, RNA interference and overexpression assays, as well as the use of Transwell migration invasion assay and EdU cell proliferation assay to assess the effect of circDHX33 on the proliferation and migration of PC-3 cells. In addition, its regulatory effect on energy metabolism in tumor cells was assessed by glycolysis assay. FISH assay, RNA pull-down, silver staining assay, and RIP were used to evaluate the interaction between circDHX33 and IGF2BP2. circDHX33 expression was significantly reduced in PCa tissues relative to normal tissues. Overexpression of circDHX33 significantly inhibited the glycolytic activity, proliferative capacity, and migratory and invasive abilities of PC-3 cells, and this effect was closely related to its reduction of IGF2BP2 protein stability. Knockdown of IGF2BP2 could reverse the malignant behavior of cells enhanced by circDHX33 knockdown. In addition, the direct intracellular interaction between circDHX33 and IGF2BP2 was verified. circDHX33 inhibits glycolysis and malignant proliferation in PCa through interaction with IGF2BP2, suggesting its potential as a potential therapeutic target.

Introduction

Prostate cancer (PCa) is a malignancy originating primarily from the epithelial cells of the prostate gland, characterized by androgen-dependent proliferation (Sekhoacha et al. 2022). At the metabolic level, PCa cells adapt to the heightened energy demands associated with rapid proliferation by upregulating glycolysis, a hallmark of the Warburg effect. This metabolic reprogramming not only supports rapid cell growth but is frequently correlated with enhanced survival under hypoxic conditions (Song et al. 2022; Koo et al. 2023). Malignant proliferation is driven by alterations in the expression of cell cycle regulators and the inactivation of tumor suppressor genes, such as PTEN, which collectively promote uncontrolled cell division and growth (Chaudagar et al. 2023). Furthermore, the aggressive nature of PCa is manifested in its capacity to remodel the extracellular matrix and invade adjacent tissues through the secretion of proteolytic enzymes, such as matrix metalloproteinases, facilitating distant metastasis via the circulatory or lymphatic systems. Common metastatic sites include bone and other distant organs, with these characteristics often being associated with advanced disease stages and poor prognosis (Zhu et al. 2024; Bian et al. 2023). These integrated mechanisms underscore the complex biology of PCa malignancy and its therapeutic challenges.

Back-splicing of precursor mRNA forms circular RNA molecules, which have covalently closed loop structures. These molecules are highly stable and play diverse roles in cellular processes (Chen et al. 2021; Zhu et al. 2021). In PCa, circRNAs primarily function as sponges for microRNAs (miRNAs), thereby modulating gene expression and influencing tumor cell behaviors such as proliferation, migration, and invasion. For instance, circCEMIP directly targets miR-1248 to enhance protective autophagy, thereby promoting resistance to anoikis in PCa cells (Yu et al. 2022). CircGNG4, by sequestering miR-223, upregulates EYA3/c-myc, thereby promoting malignant proliferation, invasion, and metastasis in PCa (Xu et al. 2021). Additionally, circRNAs regulate the stability and translation of other RNA species and can even participate in protein coding. For example, circEXOC6B, induced by short reverse repeat sequences, suppresses PCa metastasis by enhancing the interaction between RBMS1 and HuR (Zhang et al. 2023). Circ_0004087 interacts with SND1 to enhance mitotic checkpoint proofreading, contributing to docetaxel resistance in PCa cells (Chen et al. 2022). Due to these multifaceted roles, circRNAs are critical in the initiation, progression, and microenvironmental modulation of PCa, positioning them as potential diagnostic biomarkers and therapeutic targets. However, the interaction between circRNAs and RNA-binding proteins (RBPs) in the context of PCa remains insufficiently explored. A comprehensive understanding of these interactions and their impact on PCa malignancy is essential for improving the diagnostic and therapeutic strategies for this disease.

This study aims to elucidate the molecular mechanisms by which circDHX33 interacts with IGF2BP2 and assesses the regulatory role of this interaction in the malignant behaviors of PCa. The findings from this investigation will contribute to a deeper understanding of the functional roles of circRNAs in PCa progression and provide a foundation for the development of novel diagnostic biomarkers and therapeutic interventions.

Materials and methods

Bioinformatics analysis

The circRNA expression profiles and corresponding tissue information were obtained from the open-access GSE113153 dataset on the NCBI Gene Expression Omnibus (GEO). The GSE113153 project provides circRNA expression data for five low-grade and five high-grade PCa tissue samples. Differentially expressed genes (DEGs) were analyzed using R software. Genes with a Log2 fold change > 1 and P < 0.05 were considered significant. The expression pattern of IGF2BP2 in PCa was retrieved from the GEPIA database (http://gepia.cancer-pku.cn/index.html).

Clinical sample collection

Between April 2019 and September 2022, 49 pairs of PCa tissues and adjacent normal tissues (more than 5 cm from the cancerous tissue) were collected from The Fourth Hospital of Hebei Medical University. Liquid nitrogen was administered immediately to the samples, followed by long-term storage at −80 °C. Each sample was independently assessed by two different pathologists. None of the patients had received preoperative radiation or chemotherapy. Informed consent was obtained from all participants, and the study protocol was approved by the Ethics Committee of The Fourth Hospital of Hebei Medical University (No. 201801HB78).

Quantitative real-time PCR (RT-qPCR)

Total RNA was extracted from tissues and cells using TRIzol reagent (15596018, Thermo Fisher Scientific). cDNA synthesis was performed using PrimeScript RT Master Mix (RR058A, Takara, Japan). Quantitative PCR was conducted using SYBR Green qPCR Super Mix-UDG (Thermo Fisher) with GAPDH as the endogenous control. Primer sequences (Table 1) were taken to estimate gene expression using the 2^−ΔΔCT method.

Table 1.

PCR sequences

Genes	Sequences (5′–3′)
CircDHX33	Forward: 5'- CAGATCCACCAGGTTGGCTA-3'
	Reverse: 5'- TGCAGCTTTCACCACTCCAA-3'
IGF2BP2	Forward: 5'- GACAGGTCCTGCTGAAGTCC-3'
	Reverse: 5'- TGTTGACTTGTTCCACATTCTCC-3'
GAPDH	Forward: 5'- CACCCACTCCTCCACCTTTG-3'
	Reverse: 5'- CCACCACCCTGTTGCTGTAG-3'

IGF2BP2 insulin-like growth factor 2 mRNA binding protein 2; GAPDH glyceraldehyde 3-phosphate dehydrogenase

Cell culture

The RWPE-1 human prostate epithelial cell line, along with PCa cell lines 22Rv1, PC-3, and DU145, were acquired from the Chinese Academy of Sciences Cell Bank in Shanghai, China. Every PCa cell line was grown in RPMI-1640 medium, enriched with 10% fetal bovine serum. RWPE-1 cells underwent cultivation in prostate epithelial cell medium provided by ScienCell, CA, USA. Every cell was kept in an incubator set at 37 °C, containing 5% CO₂.

Actinomycin D treatment

PC-3 cells were seeded in 6-well plates at a density of 5 × 10⁵ cells per well. After 24 h, cells were treated with 2 μg/ml Actinomycin D (Sigma) for the indicated time points. Total RNA was collected, and the stability of linear GAPDH mRNA was analyzed by RT-qPCR.

RNase R treatment

RNA from PC-3 cells (10 μg) was treated with RNase R (3 U/μg, Epicenter) at 37 °C for 30 min. Subsequently, circDHX33 and linear GAPDH mRNA were evaluated by RT-qPCR.

Agarose gel electrophoresis

A 1% agarose gel was prepared using 50 ml 1 × TAE buffer and 0.5 g agarose, which was heated to boiling. After cooling to 70–80 °C, 5 μl of 4S GelRed (Biotechnology, Shanghai, China) was added and mixed. After solidification of the gel, it was placed in an electrophoresis tank filled with 1 × TAE buffer. DNA samples (10 μl per well) mixed with loading buffer were electrophoresed at 120 V for 30 min. Gel bands were visualized using a UV imaging system.

Cell transfection

siRNA targeting circDHX33 and IGF2BP2, as well as the pcDNA 3.1 overexpression vector and corresponding negative controls, were purchased from GenePharma (Shanghai, China). Transfection was carried out using Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA). Forty-eight hours post-transfection, knockdown and overexpression efficiency were assessed by RT-qPCR and Western blot analysis.

Colony formation assay

Approximately 600 PC-3 cells were cultivated in 6-well plates over a period of 2 weeks, with the media being replaced every 3 days. By the 14th day, the colonies had been stabilized using 4% paraformaldehyde (Sigma-Aldrich, MO, USA) and then colored with crystal violet (Sigma-Aldrich). Using a light microscope, the colonies were tallied and imaged.

EdU incorporation assay

The evaluation of cell proliferation utilized the Cell-Light™ EdU Apollo567 In Vitro Imaging Kit (RiboBio, Guangzhou, China). PC-3 cells, at 1 × 10⁴ cells per well in 96-well plates, underwent incubation with EdU, Apollo567, and DAPI solutions. Using a fluorescence microscope (ACCEXP Eclipse TI2-U, ACCEXP), fluorescence images were taken and examined for EdU-positive cell rate through ImageJ software.

Flow cytometry

Apoptosis was detected using the Annexin V-FITC/PI Apoptosis Detection Kit (Vazyme). Flow cytometry was performed on a BD Accuri® C6 Plus (BD Biosciences, NJ, USA) and analyzed by FlowJo. Briefly, cells were washed with pre-cooled PBS and resuspended in 100 μl of 1 × binding buffer and incubated with 5 µl of Annexin V-FIFC and PI staining solution for 10 min. Finally, 400 μl of 1 × binding buffer was added, and cells were analyzed by flow cytometry within 1 h. The FITC channel detects Annexin V-FITC signaling for identification of early apoptotic cells, and the PI channel is used to detect PI signaling for identification of late apoptotic cells.

Co-localization of CircDHX33 and IGF2BP2

Fluorescence microscopy was used to examine the co-localization of CircDHX33 and IGF2BP2 in PC-3 cells. A total of 2000 cells were planted on coverslips within 48-well plates and achieved 70–90% confluence. The cells were stabilized with 4% paraformaldehyde, underwent permeabilization using 0.5% Triton X-100, maintained at 4 °C for 10 min. The pre-hybridization process took place at 37 °C for half an hour with a pre-hybridization buffer, succeeded by an overnight incubation at 37 °C using a Cy3-tagged CircDHX33-FISH probe (RiboBio, Guangzhou, China). Post permeabilization using 0.5% Triton X-100, the cells underwent an overnight incubation at 4 °C with the anti-IGF2BP2 antibody (ab124930, Abcam). Subsequently, the cells underwent a PBS wash followed by a 15-min DAPI staining. The confocal fluorescence microscope (Carl Zeiss AG, Jena, Germany) was employed to capture the images.

RNA pull-down assay

RNA pull-down assays were performed to verify the interaction between CircDHX33 and IGF2BP2 using biotin-labeled CircDHX33 probes and control probes (GenePharma, Suzhou, China). Approximately 1 × 10⁷ PC-3 cells were lysed and sonicated in a 4 °C water bath for 30 min. RNA was extracted and incubated with 100 nmol of specific or non-specific probes at 70 °C for 5 min. Streptavidin magnetic beads (MCE, Monmouth Junction, NJ, USA) were added to the lysate and the mixture was washed with wash buffer. The washed beads were resuspended in 1 ml wash buffer, and RNA and protein were purified for subsequent analysis by RT-qPCR and Western blotting.

Silver staining

Proteins from the RNA pull-down experiment were separated on a 10% SDS-PAGE gel. Following electrophoresis, gels were stained with a Silver Staining Kit (BL620A, Biosharp, Beijing, China) according to the manufacturer’s instructions.

RNA immunoprecipitation (RIP)

To further confirm the interaction between CircDHX33 and IGF2BP2, RIP was performed using a RIP Kit (Millipore, MA, USA). Approximately 2 × 10⁷ PC-3 cells were lysed with RIP lysis buffer, and the RNA–protein complexes were immunoprecipitated using anti-IGF2BP2 antibody (ab5832, Abcam) or IgG antibody (7074, Cell Signaling Technology). Co-precipitated RNA was extracted using TRIzol reagent and analyzed by RT-qPCR.

Xenograft mouse model

Subcutaneous xenograft mouse models were used to assess tumor formation by PC-3 cells with stable knockdown of CircDHX33. The animal experiments were approved by The Fourth Hospital of Hebei Medical University Animal Care and Use Committee (ethics approval number: 201807HB25). Four-week-old (18–22 g) male BALB/c nude mice were purchased from Vital River Laboratory Animal Technology (Beijing, China). Stable CircDHX33-knockdown or control PC-3 cells (3 × 10⁶) were resuspended in 100 μl PBS and injected subcutaneously into the bilateral axillary regions of 5 female BALB/c nude mice per group. Every seven days, the dimensions of the tumor were gauged using calipers, and the volume was determined by the equation: (length × width²)/2. Following a period of 4 weeks, the mice were humanely put to sleep and their tumors were measured for weight. IGF2BP2 in tumors was assessed using immunohistochemical methods. Images were quantitatively analyzed using ImageJ software to assess IGF2BP2 positive expression. For assessing the effect of CircDHX33 on distant metastasis, stable CircDHX33-knockdown or control PC-3 cells (2.5 × 10⁶) were injected into the tail vein of nude mice. After 7 weeks, lungs were excised, paraffin-embedded, and subjected to H&E staining.

Data analysis

Data were presented as the mean ± standard deviation (SD). All experiments were performed with at least three biological replicates. Statistical analysis was conducted using GraphPad Prism 9.0 (GraphPad Software, CA, USA). Comparing two groups was conducted using Student’s t-tests, and comparing multiple groups was carried out using one-way ANOVA. *P < 0.05 was considered statistically significant.

Results

CircDHX33 negatively correlates with malignant progression and distant metastasis in PCa

To investigate the aberrantly expressed circRNAs in PCa, we selected the GSE113153 dataset from the GEO and performed differential circRNA expression analysis using R software. Compared to high-grade PCa tissues, a total of 1202 circRNAs with significantly reduced expression and 861 circRNAs with significantly increased expression were identified in low-grade PCa tissues (Fig. 1A). The heatmap of the top 500 most abundant circRNAs revealed a clear distinction in circRNA expression profiles between high-grade and low-grade PCa tissue samples (Fig. 1B). Among the top five significantly differentially expressed circRNAs, we focused on CircDHX33, which not only exhibited significant differences between different tumor grades but also between normal and PCa tissues (Fig. 1C and D). Further analysis using circPrism software revealed that hsa_circ_0000740 is a circRNA formed by exons 3 and 4 of the DHX33 gene, located at chr17: 5364258–5365866, strand: -, with a length of 399 bp (Fig. 1E). Additionally, we observed that CircDHX33 expression in three PCa cell lines was lower than that in the RWPE-1 normal prostate epithelial cell line (Fig. 1F). To assess the circular nature of CircDHX33, we conducted further stability and structure analyses. As shown in Fig. 1G, treatment with actinomycin D did not affect the stability of CircDHX33 but significantly reduced the stability of GAPDH mRNA. Moreover, RNase R digestion failed to degrade CircDHX33, whereas GAPDH mRNA was efficiently digested (Fig. 1H). To confirm that CircDHX33 is generated by back-splicing and not trans-splicing or genomic rearrangement, divergent and convergent primers were designed. PCR results showed that CircDHX33 was detectable only in cDNA, excluding the presence of gDNA as a source of amplification. The primers for GAPDH amplified both cDNA and gDNA, further confirming the results (Fig. 1I). Subsequently, we evaluated the correlation between CircDHX33 expression and clinical pathological features of patients using the chi-square test. Aberrant low expression of CircDHX33 was strongly associated with the TNM stage and distant metastasis of PCa (Table 2). These data suggest that CircDHX33 is aberrantly expressed in PCa and may play a role in regulating the occurrence and progression of the disease.

Fig. 1.

CircDHX33 is negatively correlated with malignant progression and distal metastasis of PCa. A Volcano plot showing differentially expressed circRNAs in low-grade and high-grade PCa tissues; B Heatmap displaying the top 500 most abundant circRNAs; C Bar chart showing the top 5 most highly expressed and lowest expressed differential circRNAs; D RT-qPCR analysis of circDHX33 expression in PCa tissues and adjacent normal tissues; E circPrism software analysis of the genomic features of circDHX33; F RT-qPCR analysis of circDHX33 expression in three PCa cell lines and RWPE-1 cells; G Actinomycin D assay to assess the circular structure stability of circDHX33; H RNase R assay to evaluate the circular structure stability of circDHX33; I Gel electrophoresis detection of circDHX33’s full-length splicing; Data are expressed as mean ± SD (n = 3). *P < 0.05

Table 2.

Relationship between CircDHX33 and clinicopathological characteristics of PCa patients

Characteristic	Cases	The expression of circDHX33		P
	n = 49	Low (n = 24)	High (n = 25)
Age (year)				0.3205
≤ 60	30	13	17
> 60	19	11	8
Tumor size				0.1000
< 3 cm	34	14	20
≥ 3 cm	15	10	5
TNM staging				0.0001*
I/II	28	7	21
II/IV	21	17	4
Lymph node metastasis				0.1085
Positive	17	11	6
Negative	32	13	19
Distant metastasis				< 0.0001*
Positive	23	19	4
Negative	26	5	21

CircDHX33 regulates PCa proliferation, glycolysis, and metastasis

To investigate the specific impact of CircDHX33 on tumor behavior, we transiently overexpressed and knocked down CircDHX33 in PC-3 cells through targeted genetic interventions. The experimental results revealed the multifaceted regulatory effects of CircDHX33 on PCa cells. First, RT-qPCR confirmed the transfection efficiency (Fig. 2A). Overexpression of CircDHX33 significantly inhibited PC-3 cell proliferation, as evidenced by a marked reduction in EdU-positive cell proportion and colony formation (Fig. 2B and C). In contrast, knockdown of CircDHX33 significantly promoted these parameters (Fig. 2B and C). This finding was further supported by changes in Ki-67 protein expression (Fig. 2D). In terms of invasion and migration, overexpression of CircDHX33 suppressed PC-3 cell migration and invasion in the Transwell assay, whereas knockdown of CircDHX33 exacerbated these effects (Fig. 2E). Additionally, we assessed the regulation of EMT markers by CircDHX33. We found that CircDHX33 overexpression inhibited N-cadherin expression while increasing E-cadherin levels, whereas knockdown of CircDHX33 induced the opposite effects (Fig. 2F). Furthermore, the impact of CircDHX33 on apoptosis was also prominent. Overexpression of CircDHX33 promoted apoptosis, as indicated by an increase in apoptosis rate and the level of cleaved caspase-3 protein. Conversely, CircDHX33 knockdown decreased these indicators (Fig. 2G and H). Given the central role of glycolysis in tumor energy metabolism, we further explored the regulatory effect of CircDHX33 on glycolytic pathways. Notably, overexpression of CircDHX33 reduced glucose consumption, lactate production, and ATP levels, suggesting its potential to inhibit tumor proliferation through glycolysis suppression (Fig. 2I). Taken together, these findings demonstrate that CircDHX33 precisely regulates key biological processes such as glycolysis, EMT, and cell proliferation.

Fig. 2.

CircDHX33 regulates PCa proliferation, glycolysis, and metastasis. PC-3 cells were transiently transfected with circDHX33 overexpression or knockdown vectors; A RT-qPCR analysis of circDHX33 expression; B Edu assay measuring the percentage of proliferating cells; C Colony formation assay to assess the cell proliferation rate; D Western blot analysis of proliferation marker Ki-67 expression; E Transwell assay measuring the number of invading and migrating cells; F Western blot analysis of EMT-related protein expression, E-cadherin and N-cadherin; G Flow cytometry analysis of cell apoptosis rate; H Western blot analysis of apoptosis marker cleaved caspase-3 expression; I Commercial kits measuring glucose consumption, lactate production, and ATP levels; Data are expressed as mean ± SD (n = 3). *P < 0.05

CircDHX33 interacts with IGF2BP2 in PC-3 cells

In exploring the interaction between circRNAs and RBPs, we focused on the functional role of CircDHX33. Previous studies have demonstrated that many circRNAs can regulate the function and availability of RBPs by binding to them, thereby affecting RNA processing, stability, and translation. In this context, we conducted an in-depth analysis of the potential RBPs interacting with CircDHX33. Using the bioinformatics platform starBase (https://rnasysu.com/encori/index.php), we identified a list of RBPs that may bind to CircDHX33. Among the candidates, IGF2BP2 stood out due to its notable binding potential. IGF2BP2 is not only upregulated in various cancers but also predicted to have specific binding sites with CircDHX33 (Fig. 3A and B). To further investigate this interaction, we first performed RNA pull-down experiments, followed by silver staining, which revealed specific protein bands at 40–70 kDa, indicating that CircDHX33 can bind to specific RBPs (Fig. 3C). We then used RNA pull-down combined with Western blotting to confirm the direct interaction between IGF2BP2 and CircDHX33 (Fig. 3D). Additionally, RIP assays demonstrated that CircDHX33 was highly enriched by anti-IGF2BP2 antibodies, further confirming their binding (Fig. 3E). FISH experiments showed that CircDHX33 and IGF2BP2 co-localized in the cytoplasm of PC-3 cells (Fig. 3F), providing insight into their potential functional domains within the cell. Further investigation into the expression pattern of IGF2BP2 in PCa revealed that IGF2BP2 was significantly upregulated in both PCa tissues and cells compared to normal controls (Fig. 3G and H). Interestingly, overexpression of CircDHX33 in PC-3 cells led to a decrease in IGF2BP2 protein levels, whereas knockdown of CircDHX33 enhanced IGF2BP2 expression. However, overexpression of CircDHX33 did not affect the mRNA levels of IGF2BP2 (Fig. 3I). CircDHX33 might affect PCa development by interfering with IGF2BP2, as suggested by these findings.

Fig. 3.

CircDHX33 interacts with IGF2BP2 in PC-3 cells. A Bioinformatics prediction of potential binding sites between CircDHX33 and IGF2BP2 using the StarBase platform; B GEPIA database analysis of IGF2BP2 expression in different cancer tissues; C Silver staining confirming the presence of specific RBPs interacting with CircDHX33; D RNA pull-down assay validating the direct interaction between CircDHX33 and IGF2BP2; E RIP assay analyzing the direct interaction between CircDHX33 and IGF2BP2; F FISH assay showing co-localization of CircDHX33 and IGF2BP2 in PC-3 cells; G Western blot analysis of IGF2BP2 expression in PCa tissues and adjacent normal tissues; H Western blot analysis of IGF2BP2 expression in three PCa cell lines and RWPE-1 cells; I RT-qPCR and Western blot analysis of the effect of CircDHX33 overexpression or knockdown on IGF2BP2 mRNA and protein expression; Data are expressed as mean ± SD (n = 3). *P < 0.05

Knockdown of IGF2BP2 inhibits malignant behaviors of PC-3 cells

To further explore the role of IGF2BP2 in PCa cells, we transfected PC-3 cells with siRNA targeting IGF2BP2 to knock down its expression. Western blot analysis confirmed a significant reduction in IGF2BP2 expression (Fig. 4A). We then assessed the impact of IGF2BP2 knockdown on cell proliferation using EdU incorporation and colony formation assays. Both assays revealed a marked decrease in EdU-positive cell percentages and colony numbers (Figs. 4B and C), which was further supported by a reduction in Ki-67 protein expression (Fig. 4D), indicating that cell proliferative capacity was impaired. In terms of cell invasion and migration, Transwell assays showed that IGF2BP2 knockdown significantly reduced the invasive and migratory abilities of the cells (Fig. 4E). Moreover, we assessed EMT markers, revealing that IGF2BP2 knockdown suppressed N-cadherin expression and enhanced E-cadherin expression (Fig. 4F), suggesting its role in regulating the EMT process. Further flow cytometry analysis demonstrated that IGF2BP2 knockdown notably increased the apoptosis rate (Fig. 4G), and elevated levels of cleaved caspase-3 protein confirmed the activation of the apoptotic pathway (Fig. 4H). Additionally, we examined glycolysis-related parameters and found that knockdown of IGF2BP2 reduced glucose consumption, lactate production, and ATP levels, indicating its involvement in regulating tumor cell energy metabolism (Fig. 4I). Our findings demonstrate that IGF2BP2 is a key regulator of PCa cell proliferation, invasion, migration, EMT, and glycolysis. Knockdown of IGF2BP2 significantly suppresses these tumor-promoting processes.

Fig. 4.

Knockdown of IGF2BP2 inhibits malignant behavior of PC-3 cells. PC-3 cells were transfected with siRNA targeting IGF2BP2 for knockdown; A Western blot analysis of IGF2BP2 expression; B Edu assay measuring the percentage of proliferating cells; C Colony formation assay to assess the cell proliferation rate; D Western blot analysis of proliferation marker Ki-67 expression; E Transwell assay measuring the number of invading and migrating cells; F Western blot analysis of EMT-related protein expression, E-cadherin and N-cadherin; G Flow cytometry analysis of cell apoptosis rate; H Western blot analysis of apoptosis marker cleaved caspase-3 expression; I Commercial kits measuring glucose consumption, lactate production, and ATP levels; Data are expressed as mean ± SD (n = 3). *P < 0.05

circDHX33 modulates PC-3 malignant behaviors via regulation of IGF2BP2

To further elucidate the interaction between circDHX33 and IGF2BP2 and their biological effects in PCa cells, we simultaneously knocked down circDHX33 and IGF2BP2 to analyze their combined impact on PC-3 cell behaviors. Co-transfection of si-circDHX33 and si-IGF2BP2 into PC-3 cells revealed that knockdown of circDHX33 increased the protein expression of IGF2BP2, while the presence of si-IGF2BP2 reversed this effect (Fig. 5A). In terms of cell proliferation, both EdU and colony formation assays demonstrated that circDHX33 knockdown significantly enhanced the percentage of EdU-positive cells and colony formation. However, this effect was reversed when IGF2BP2 was also knocked down (Figs. 5B and C). Ki-67 protein expression analysis further supported these observations, showing that knockdown of circDHX33 increased Ki-67 expression, which was prevented by IGF2BP2 knockdown (Fig. 5D). Regarding cell invasion and migration, Transwell assays indicated that silencing circDHX33 increased the number of invading and migrating cells, but this effect was effectively reversed by IGF2BP2 knockdown (Fig. 5E). Additionally, analysis of EMT-related protein expression revealed that circDHX33 knockdown promoted N-cadherin and suppressed E-cadherin expression, while IGF2BP2 knockdown blocked these changes (Fig. 5F). Flow cytometry results showed that silencing circDHX33 reduced the apoptosis rate, whereas IGF2BP2 knockdown increased apoptosis, and this was further confirmed by elevated levels of cleaved caspase-3 (Figs. 5G and H). Finally, metabolic studies showed that circDHX33 knockdown increased glucose consumption, lactate production, and ATP levels, and these changes were reversed by IGF2BP2 knockdown (Fig. 5I). The interaction between circDHX33 and IGF2BP2 appears to be crucial to the proliferation, invasion, migration, EMT process, and apoptosis of PCa cells.

Fig. 5.

CircDHX33 regulates PC-3 cell malignant behavior through IGF2BP2. PC-3 cells were co-transfected with si-circDHX33 and si-IGF2BP2; A Western blot analysis of IGF2BP2 expression; B Edu assay measuring the percentage of proliferating cells; C Colony formation assay to assess the cell proliferation rate; D Western blot analysis of proliferation marker Ki-67 expression; E Transwell assay measuring the number of invading and migrating cells; F Western blot analysis of EMT-related protein expression, E-cadherin and N-cadherin; G Flow cytometry analysis of cell apoptosis rate; H Western blot analysis of apoptosis marker cleaved caspase-3 expression; I Commercial kits measuring glucose consumption, lactate production, and ATP levels; Data are expressed as mean ± SD (n = 3). *P < 0.05

circDHX33 suppresses PCa tumor growth and metastasis in vivo

To further investigate the effects of circDHX33 on PCa tumor growth and metastasis in vivo, we established a PCa tumor model overexpressing circDHX33. Tumor volume and weight were reduced in the circDHX33 overexpression group (Fig. 6A–C), indicating that circDHX33 effectively inhibits tumor growth. Further IHC staining revealed that IGF2BP2 was markedly suppressed in tumors overexpressing circDHX33 (Fig. 6D). This suggests that circDHX33 may influence tumor cell behavior by regulating IGF2BP2. Additionally, we assessed the metastatic potential of the tumors. Analyses of lung metastatic nodules indicated that circDHX33 overexpression significantly reduced lung metastases (Fig. 6E). These results further support the role of circDHX33 in inhibiting PCa metastasis.

Fig. 6.

CircDHX33 inhibits PCa tumor growth and metastasis in vivo. A Representative tumor images from mice; B Tumor volume measurements in mice; C Tumor weight measurements in mice; D IHC staining showing IGF2BP2 levels in mouse tumor tissues; E HE staining showing the number of metastatic lung nodules in mice; Data are expressed as mean ± SD (n = 5). *P < 0.05

Discussion

PCa is a malignant tumor primarily originating from the epithelial cells of the prostate gland. It is characterized by androgen-dependent proliferation, enhanced glycolysis to meet energy demands, and dysregulated cell cycle control leading to malignant proliferation (Matsushita et al. 2020; Graham et al. 2023). In this context, the interaction between circRNAs and RBPs plays a critical role in regulating PCa cell behavior and shaping the tumor microenvironment (Cheng et al. 2024; Zhang et al. 2022). However, existing studies often lack systematic approaches to identify and validate direct interactions between circRNAs and RBPs, limiting a comprehensive understanding of their roles in PCa progression. Investigating these interactions could provide insights into the molecular mechanisms of PCa and offer potential new targets for diagnosis and therapy, ultimately helping to combat this highly invasive and metastatic cancer.

This study identified circDHX33, which interacts directly with the RBP IGF2BP2 and, through this mechanism, inhibits the malignant proliferation, invasion, migration, and metastasis of PCa. Through GEO database analysis, we observed that circDHX33 level was lower in high-grade tumor tissues compared to low-grade tumor tissues, and higher in tumor tissues compared to adjacent normal tissues. The differential expression of circDHX33 across various grades of tumor and normal tissues suggests that it may serve as a valuable biomarker for both the diagnosis and disease monitoring of PCa. Particularly, the contrast between tumor and adjacent normal tissues provides a potential clinical application for distinguishing tumor tissues from neighboring non-tumor tissues. Notably, previous studies have suggested that circDHX33 could promote malignant behavior in renal clear cell carcinoma by targeting the miR-489-3p/MEK1 axis (Wang et al. 2020). circDHX33, however, appeared to function as a tumor suppressor in PCa. circDHX33 overexpression significantly reduced the proliferation and metastasis of PCa cells. Not only circDHX33 but also other circRNAs may have context-dependent pro- or anti-cancer roles in different cancers (Wang et al. 2023; Jin et al. 2020). These results imply that the function of circDHX33 may depend on the specific cell type and microenvironment, which could be driven by cell-specific transcriptional regulatory networks, microenvironmental factors, or different interacting molecular partners. Therefore, therapeutic strategies targeting circDHX33 should take into account its role in specific cancers.

RBPs play diverse roles in cancer by regulating mRNA processing, stability, transport, and translation, thereby directly influencing tumor cell proliferation, survival, and invasion. IGF2BP2, a member of the RBP family, has been implicated in promoting cancer in PCa. For example, IGF2BP2 stabilizes IGF1R, CCAR1, and HOXC6 mRNAs, promoting bone metastasis in PCa (Lang et al. 2021; He et al. 2024; Huang et al. 2023). However, the upstream regulatory mechanisms of IGF2BP2 in PCa, particularly its interactions within the broader regulatory network, remain poorly understood. Our findings contribute to filling this knowledge gap by revealing a novel mechanism through which circDHX33 negatively regulates IGF2BP2 via direct interaction. Interestingly, we found that overexpression of circDHX33 reduced IGF2BP2 protein expression without affecting its mRNA levels. This suggests that circDHX33 may regulate IGF2BP2 through post-translational modifications or protein stability, rather than by altering mRNA stability or synthesis. In summary, our data provide a theoretical foundation for the development of novel therapeutic strategies targeting IGF2BP2 and its regulatory network. Furthermore, this underscores the complexity of RNA-RNA and RNA–protein interactions within the cell and their potential roles in regulating cancer behavior.

Importantly, our findings highlight the potential clinical relevance of circDHX33 in PCa. The unique properties of circRNAs—particularly their covalently closed loop structure and relative resistance to exonucleases—confer higher stability in circulation compared to linear RNAs, making them promising biomarkers in liquid biopsies for cancer diagnosis and prognosis. Notably, the U.S. Food and Drug Administration and/or the European Medicines Agency (EMA) have approved 11 RNA-based therapeutics targeting multiple patient tissues and organs (Winkle et al. 2021). Synthetic circRNAs have attracted more attention for their robust and stable translation in eukaryotic cells (Zhang et al. 2016). Recently, Li et al. developed a novel circRNA vaccine platform to stimulate robust innate and adaptive immune responses for anti-tumor effects in multiple mouse tumor models (Li et al. 2022). Therefore, restoring circDHX33 expression or disrupting its pathological interactions in PCa might be an attractive strategy to inhibit tumor progression, particularly for patients with advanced disease or those who are resistant to conventional treatments. However, despite significant progress in circRNA research and applications, most candidates are currently in the discovery or preclinical stage. To date, no circRNA therapeutic candidates have entered clinical trials. Nevertheless, RNA therapies, such as mRNA-based therapies, can be translated into circRNA-based therapies and provide valuable insights into the development of circRNAs as therapeutic agents. Moreover, targeting the circDHX33–IGF2BP2 axis might provide additional benefits by reducing the metastatic potential of PCa, as IGF2BP2 has been implicated in multiple pro-oncogenic pathways. Future in vivo studies and clinical trials are warranted to rigorously assess the therapeutic efficacy of circDHX33-based interventions and to determine the most effective means of delivering such therapies. Collectively, these clinical implications underscore the importance of further elucidating the molecular details of circDHX33’s tumor-suppressive role, including its impact on post-translational modifications and protein stability, to refine our strategies for translational applications.

This study reveals the functional role of circDHX33 in PCa and its mechanism of regulating tumor behavior through IGF2BP2, using both public databases and in vitro experimental models. However, the generalizability of these results may be limited by sample representativeness and quantity, and the heterogeneity of data as well as the limitations of in vitro conditions may affect the interpretation of results. Additionally, the specific molecular mechanisms underlying the influence of circDHX33 on IGF2BP2 protein stability have yet to be fully elucidated, and the complex interactions within the tumor microenvironment have not been thoroughly explored. Therefore, while these findings provide potential targets for molecularly targeted therapy in PCa, further studies are needed to validate these mechanisms in preclinical models and with a broader range of samples, as well as to assess their biological implications.

Conclusion

In conclusion, our study systematically analyzes the expression patterns and biological functions of circDHX33 in PCa. These findings highlight the potential of circDHX33 as both a biomarker and a therapeutic target in the management of PCa. However, given that the study mainly relies on in vitro models and public databases, future research should validate the generalizability and clinical relevance of these mechanisms in larger clinical cohorts and in vivo models. Additionally, further investigation into the molecular mechanisms of the interaction between circDHX33 and IGF2BP2 will provide a more detailed scientific basis for the development of targeted therapies in PCa. Overall, our study offers a new perspective on the molecular regulatory network of PCa and lays the foundation for therapeutic approaches targeting circRNAs.

Author contributions

XDL designed the research study. XDL and XLT performed the research. LP and CHD provided help and advice. XLT and LP analyzed the data. XDL wrote the manuscript. CHD reviewed and edited the manuscript. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript.

Funding

Medical Science Research Project of Hebei (NO.20221242).

Data availability

The datasets used and/or analyzed during the present study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval

The present study was approved by the Ethics Committee of The Fourth Hospital of Hebei Medical University (No. 201801HB78) and written informed consent was provided by all patients prior to the study start. All procedures were performed in accordance with the ethical standards of the Institutional Review Board and The Declaration of Helsinki, and its later amendments or comparable ethical standards.

Research involving human and animal participants

All animal experiments were complied with the ARRIVE guidelines and performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. The experiments were approved by the Institutional Animal Care and Use Committee of The Fourth Hospital of Hebei Medical University (No. 201807HB25).