Circular RNA circR-4600 inhibits hepatocellular carcinoma progression by suppressing autophagy through DVL3 and DAB2 interaction

Circular RNA microarray screening for target circular RNAs

Arraystar Human circRNA Array analysis was conducted on 6 cases of HCC tissues and corresponding adjacent non-tumor tissues. Total RNA from each sample was quantified using NanoDrop ND-1000. Sample preparation and microarray hybridization were performed following Arraystar's standard protocol. Briefly, total RNA was digested with Rnase R (Epicentre, Inc.) to remove linear RNA and enrich circular RNA. Subsequently, the enriched circular RNA was amplified and transcribed into fluorescent cRNA using the random priming method (Arraystar Super RNA Labeling Kit; Arraystar). The labeled cRNAs were purified using the RNeasy Mini Kit (Qiagen) to measure the concentration and specific activity of labeled cRNAs using NanoDrop ND-1000. Each labeled cRNA (1 μg) was fragmented by adding 5 μl 10 × Blocking Agent and 1 μl of 25 × Fragmentation Buffer, heating the mixture at 60℃ for 30 minutes, and then adding 25 μl 2 × Hybridization Buffer to dilute the labeled cRNA. Fifty microliters of hybridization solution were dispersed onto the circRNA expression microarray slide and assembled. The slide was incubated at 65℃ for 17 hours in an Agilent Hybridization Oven. After hybridization, the array was washed, fixed, and scanned using the Agilent Scanner G2505C. Subsequently, scanned images were imported into Agilent Feature Extraction software for raw data extraction. Quantile normalization and subsequent data processing of the raw data were performed using the R software package. After quantile normalization of the raw data, low-intensity filtering was performed, and circRNAs with “P” or “M” (“All Targets Value”) flags present in at least 6 samples were retained for further analysis. When comparing the differences in profiles between HCC and control groups, the “fold change” (i.e., the ratio of group averages) for each circRNA was calculated. The statistical significance of differences was estimated using t-tests. CircRNAs with fold changes greater than or equal to 1.5 and P-values less than 0.05 were selected as significantly differentially expressed.

RT-qPCR detection

In this study, tissue samples weighing between 50 and 100 mg were homogenized using the TRIZOL reagent (Invitrogen). Chloroform was then added to effectively separate RNA from other components, ensuring complete transfer of RNA to the aqueous phase. The RNA was precipitated using isopropanol and the precipitate was cleaned with 75% ethanol, prepared with DEPC-treated water. The RNA precipitate was dissolved in RNase-free water and stored at -80℃ for future use. The concentration and purity of the RNA were measured using the NanoDrop® ND-1000 instrument, and the quality of the RNA was evaluated through denaturing agarose gel electrophoresis. cDNA synthesis was carried out using SuperScript^TM III reverse transcriptase (Invitrogen), an RNase inhibitor (Epicentre), 5 × RT buffer (Invitrogen), and N6 random primers. For the reverse transcription of mRNA, a reverse transcription kit (TAKARA, Japan) was utilized. The RT-qPCR analysis was conducted using a 2 × PCR master mix (Arraystar), and the PCR reaction was completed using the ViiA 7 Real-time PCR System (Applied Biosystems). The sequences of the primers are provided in Supplementary Table 1. The relative levels of RNA were calculated using the comparative CT (2^{−ΔΔ Ct}) method.

Fluorescence in situ hybridization (FISH) localization of circR-4600 in HCC cells

The spatial distribution of circR-4600 within SMMC7721 cells was delineated utilizing FISH technology. Before the commencement of the experiment, preparations were made for reagents, including the circR-4600 probe, Anti-Dig-red, hybridization fluid, DAPI, Rubber Cement, and a 3% BSA sealing solution. Concurrently, solutions such as 0.1M HCl, 0.5% TritonX-100, and a fixative were prepared. The sequence of the probe was established as 5’-3’ GGCCTTCAATTTCCCACTCTTCTTTAGAGAGAGAAGTTCTACAAT. The cells were initially fixed using a fixative, formulated by combining methanol and glacial acetic acid in a 3 : 1 ratio, followed by a wash with a 2 × SSC solution containing 0.1% NP-40. A 0.5% TritonX-100 solution was employed for cell permeabilization, facilitating probe penetration. Subsequently, the cells were once again fixed with a 4% paraformaldehyde solution to preserve cellular structural integrity. The probe was denatured at 88℃ for 5 minutes, swiftly cooled to 4℃ to form a single-strand structure, and then diluted in a 1 : 100 ratio in a preheated (55℃) hybridization solution. The hybridization reaction was conducted overnight in a humid box at 37℃, ensuring optimal probe-target RNA binding. Post-hybridization, the cells underwent multiple washes with a 1 × PBS solution containing 0.1% DEPC, effectively removing non-specifically bound probes. The cells were then treated with a 3% BSA sealing solution for 30 minutes to obstruct non-specific binding sites. Subsequently, Anti-Dig-red was employed for probe detection. This detection reagent was diluted in a 1:100 ratio with 1% BSA, added to the cells, and incubated in the dark at 37℃ for 1 hour. The cells were then stained with DAPI, covered with a cover slip, and placed in the dark for 20 minutes to minimize photobleaching. Ultimately, image results were acquired via a CSM 700 laser confocal microscope.

Analysis of the correlation between circR-4600 and clinical prognosis

An analysis was conducted to examine the correlation between circR-4600 and clinical prognosis. The study incorporated 222 patients, postoperatively diagnosed with HBV-associated HCC. Because this was a retrospective observational study, the sample size was determined by the number of consecutive eligible patients available during the study period. Histopathological examinations were performed on all patients, ensuring adherence to the diagnostic standards set by the National Comprehensive Cancer Network (NCCN) for tumor clinical practice. To acquire comprehensive patient data, telephonic or in-patient follow-ups were conducted until the patient’s demise or the final follow-up. The median follow-up duration extended to 17 months, with a range of 3 to 32 months, thereby ensuring data accuracy. During the screening phase, meticulous attention was given to confirm the absence of prior cancer diagnoses among patients, thereby eliminating potential confounding factors. Detailed patient records were maintained, encompassing age, gender, pathological grading, tumor biological behavior, serum alpha-fetoprotein (AFP) levels, cirrhosis status, and whether radical resection was performed, among other clinicopathological characteristics. To gain a comprehensive understanding of the patient’s tumor status, the Barcelona Clinical Liver Cancer (BCLC) staging system was employed for tumor classification. This system considers various factors, including the patient’s liver function, tumor size and number, and the presence of vascular invasion. The patient enrollment flowchart is shown in Supplementary Figure 1.

Plasmid construction and lentiviral packaging

The construction of plasmids and the packaging of lentiviruses were carried out as follows. The flanking sequences of the mature circR-4600 sequence were compared in the circBase database. An additional 365 bp was appended to the 5’ end, and 416 bp to the 3’ end, resulting in a total sequence of 1160 bp for the vector, inclusive of the 379 bp mature circR-4600 sequence. The full-length 1160 bp sequence was synthesized using the whole gene synthesis method and subsequently incorporated into the pCDH-CMV-MCS-EF1-GFP+Puro (CD513B-1) vector via EcoRI and BamHI linkages. Upon completion of the linkage, the recombinant plasmid was screened and verified. Total RNA was extracted from the cell samples and subjected to RT-qPCR detection. This process involved the use of an internal reference, circular RNA-specific primers (Divergent primers), and corresponding linear sequence primers (Convergent primers). Sequencing primers were also designed, and post-PCR, the bands were recovered and sequenced to verify the junction of the circularization site. The sequences of these primers are provided in Supplementary Table 2. Subsequently, lentivirus packaging was conducted using 293T cells. Small or short interfering RNA (siRNA) inhibition technology was employed to construct DVL3 and ATG5 inhibitory lentiviruses. siRNA sequences, specifically targeting DVL3 and ATG5, were designed. These sequences were capable of precisely binding to the target mRNA, thereby inhibiting its translation process. The corresponding siRNAs for DVL3 and ATG5 are as follows: DVL3 -homo: CCAGGAGAACCUGGACAAUTT; ATG5-homo: CCAUCAAUCGGAAACUCAUTT. These siRNA sequences were then individually incorporated into the lentivirus genome, resulting in the construction of DVL3 and ATG5 inhibitory lentiviruses. The effectiveness and specificity of these lentiviruses were confirmed by infecting various HCC cells and assessing the expression levels of DVL3 and ATG5.

Assessment of cell proliferation using BrdU and CCK-8 assays

Two commonly used methods for assessing cell proliferation are the BrdU (5-bromo-2’-deoxyuridine) assay and the CCK-8 (Cell Counting Kit-8) assay. In this study, we employed these techniques to investigate cell growth in HCC cells.

BrdU Assay: HCC cells were suspended in DMEM medium supplemented with 10% fetal bovine serum (FBS, Gibco, USA) and a 1% mixture of penicillin-streptomycin (Solarbio, China). After re-suspension, the cells were cultured at 37℃ with 5% CO2. Transient transfection was performed in 6-well plates. The day before transfection, 5 × 10^5 cells were seeded in culture plates and incubated with 2 ml complete growth medium to achieve 70-90% confluence. Plasmid DNA (3 μg) was gently mixed with 100 μl serum-free medium. Lipofectamine 3000 reagent was diluted with 100 μl serum-free medium containing 4 μl Lipofectamine 3000 and incubated at room temperature for 5 minutes. The diluted plasmid and Lipofectamine 3000 were combined and incubated for 20 minutes to form the plasmid-Lipofectamine 3000 complex. The complex was added to cell wells containing 800 μl serum-free medium, followed by gentle shaking. After 5-6 hours of incubation at 37℃ with 5% CO2, the transfection medium was aspirated, and complete growth medium was replaced.

CCK-8 Assay: The CCK-8 assay (Beyotime, China) was used to assess cell viability. Briefly, 10000 cells were seeded in 100 μl DMEM medium per well. Next, 10 μl of CCK-8 solution was added to each well, followed by a 2-hour incubation. Absorbance was measured at 450 nm using a microplate reader.

Cell migration, invasion, and clonogenic assays

Cell migration experiments (scratch assays) primarily involve creating scratches on a cell culture plate and observing how cells fill these scratches to assess their migration ability. First, evenly mark horizontal lines on the back of a 6-well plate using a marker. Then, add fibronectin to each well. After seeding the cells, treat them with cytosine arabinoside for one hour to inhibit cell division. Next, create scratches on the cell culture plate and wash with PBS to remove dislodged cells. Finally, use ImageJ software to measure scratch width at eight arbitrary locations for each cell group at the same time point and calculate cell migration distance.

The detection of cell invasion ability typically employs the Transwell assay. Initially, dissolve matrigel matrix gel at 4℃ and dilute it with pre-chilled serum-free medium. Add this mixture to pre-chilled Transwell chambers and incubate at 37℃ to allow matrigel to solidify. Subsequently, introduce cells into the upper chamber of the Transwell. After 24 and 48 hours of incubation at 37℃ with 5% CO2, remove the chambers, wipe off cells from the upper chamber using a cotton swab, fix with 4% paraformaldehyde, wash with PBS, and stain with crystal violet for 10 minutes. Finally, assess whether cells have penetrated the pores by capturing images.

The plate clonogenic assay evaluates cell clonogenic formation ability by seeding a small number of cells in a culture plate and observing their growth and division. Digest cells in logarithmic growth phase with trypsin, resuspend them in complete culture medium, and count. Then, seed 1000 cells in a 6-well culture plate. Continue cultivation for 14 days or until the majority of single-cell clones contain more than 50 cells. During this period, change the medium every 3 days and monitor cell status. After completing clonogenic growth, photograph the cells under a microscope, wash with PBS, fix with 4% paraformaldehyde for 30 minutes, stain with crystal violet for 20 minutes, rinse with PBS, air-dry, and capture images using a camera.

In vivo assessment of tumor formation following lentiviral particle infection with circR-4600 in mice

In this study, we investigated the impact of circR-4600 on tumor formation in vivo using a murine model. Male athymic nude mice (6 weeks old) were procured from the Experimental Animal Center at Guangxi Medical University in Nanning, China. All animals were housed under standard conditions in the center’s animal care facility. Ethical approval for all animal experimental protocols was obtained from the Guangxi Medical University Animal Care and Use Committee, ensuring compliance with relevant ethical guidelines for animal research. To establish the tumor model, SMMC7721 tumor cells stably transfected with either a negative control (NC) or circR-4600 were resuspended in 200 μl of phosphate-buffered saline (PBS) and injected via the tail vein into each mouse (with five mice per group; an additional mouse was used to assess model efficacy). Prior to injection, circR-4600 overexpression was confirmed using reverse transcription quantitative polymerase chain reaction (RT-qPCR). After 3-4 weeks post-injection, humane euthanasia was performed on all mice. Lung metastasis was evaluated through photographic assessment, while hematoxylin and eosin (H&E) staining was employed to assess the engraftment of HCC cells in the lung tissue. Furthermore, we quantified the expression levels of E-cadherin (E-cad), matrix metalloproteinase 9 (MMP9), and Ki-67 in lung tissues using RT-qPCR. Immunohistochemistry and immunoblotting were utilized for validation. Each experiment was repeated three times to ensure the reliability of our results.

Proteomic Analysis of circR-4600 Pulldown

The biotin-labeled RNA was synthesized in vitro using the MEGAscriptTM T7 Kit (Invitrogen, USA), with the incorporation of Bio-16-UTP (10 mM, Ambion) during transcription. Subsequently, 2 μl of DNase I was added, and the reaction mixture was incubated at 37℃ for 15 minutes to remove DNA. Following this, 2 μl of 0.2 M EDTA was added. To facilitate the formation of secondary RNA structures, the biotin-labeled RNA (1 µg) was heated at 95℃ for 2 minutes, followed by incubation on ice for 3 minutes and further incubation at room temperature in RNA structure buffer for 30 minutes. Magnetic beads (Invitrogen, USA) were utilized for RNA binding and enrichment. The folded RNA was then mixed with extracts from HCC cells in 500 μl of RIP wash buffer. The magnetic beads were resuspended in 50 μl of RIP wash buffer, followed by the addition of Dynabeads M-280 streptavidin (60210, Invitrogen), and incubated at 4℃. After centrifugation for 1 minute, the supernatant was discarded. The beads were washed with RIP wash buffer and boiled in SDS buffer. Pulled-down proteins were detected using western blotting and mass spectrometry. The RNA probes used were as follows: circR-4600 sense: taatacgactcactatagggCTGGCCTCGTCATCCGCCCACCTTGGAGATAGGCTCTCGC TCTGTCACC;antisense:taatacgactcactatagggGGAATAGGCTCTCGCTCTGTCACCTGGCCTCGTCATCCGCCCACCTTG.

Western blotting (WB) analysis

WB analysis was performed to detect protein expression levels. Proteins were extracted from both cellular and tissue samples using cell lysis buffer (9803, CST) following the manufacturer's protocol. Protein concentrations were determined using the BCA protein concentration assay kit (Beyotime, China). Subsequently, proteins were separated by SDS-PAGE gel electrophoresis and transferred onto PVDF membranes. After blocking the membranes with 5% non-fat milk, they were incubated overnight at 4℃ with specific primary antibodies (such as GAPDH, Vimentin, E-cadherin, MMP9, DVL3, P62, DAB2, etc., diluted at 1 : 1000). Following primary antibody incubation, membranes were then incubated with appropriate secondary antibodies for 1 hour. Finally, images were captured using the Odyssey CLX dual-color infrared imaging system.

RIP-qPCR detects the interaction between DVL3 and circR-4600

The RIP (RNA-Binding Protein Immunoprecipitation) experiment was conducted using the RIP RNA-Binding Protein Immunoprecipitation kit (Sigma, USA) according to the manufacturer’s instructions. Initially, cells were collected from the cell culture plate and transferred to centrifuge tubes. Subsequently, 100 μl of RIP lysis buffer was added to the cell pellet. The lysate was then incubated on ice for 5 minutes followed by overnight incubation at 4℃. Upon thawing, the lysate was centrifuged at 14,000 rpm for 10 minutes at 4℃. Next, 100 μl of the supernatant was mixed with 900 μl of RIP immunoprecipitation buffer (comprising 860 μl of RIP wash buffer, 35 μl of 0.5 M EDTA, and 5 μl of RNase inhibitor). DVL3 antibody (Abcam) was bound to magnetic beads and then incubated overnight at 4℃ with the cell lysate under gentle rotation. Following five washes, purified precipitated RNA and proteins were obtained for analysis. WB was employed to confirm the specificity of the DVL3 antibody. RT-qPCR was utilized to detect RNA enriched by DVL3 to elucidate its interaction with circR-4600.

Analysis of the relationship between DVL3 expression and prognosis in HCC using The Cancer Genome Atlas (TCGA) database

In this study, mRNA expression data and clinical information of HCC patients were retrieved from TCGA, and differentially expressed genes associated with OS were identified, with a specific focus on the DVL3 gene. Differential expression of DVL3 in HCC patients was visualized using box plots. Based on the expression levels of DVL3, HCC patients were stratified into high-expression and low-expression groups, and Kaplan-Meier survival curves were employed for visualization to assess the relationship between DVL3 and the prognosis as well as OS rate of HCC.

Immunofluorescence analysis

The expression levels of LC3B and P62 were examined using immunofluorescence. Cell slides were placed in 6-well plates and cells were seeded onto them, then cultured at 37℃ with 5% CO2 until reaching appropriate density. After cell attachment, fixation was performed using 4% paraformaldehyde at room temperature for 10 minutes, followed by permeabilization with PBS containing 0.25% Triton X-100 for 10 minutes. Subsequently, nonspecific binding sites were blocked with PBS solution containing 1% BSA for 30 minutes. Then, primary antibodies against LC3B (3868, CST) and P62 (23214, CST) were incubated overnight at 4℃ at a dilution ratio of 1 : 1000. The next day, cells were washed three times with PBS for 5 minutes each, followed by incubation with corresponding fluorescently labeled secondary antibodies (diluted at 1:500) for 1 hour. A 50 μl drop of anti-fade mounting medium (containing DAPI) (S2110-5-solarbiol) was placed on a coverslip, and cells from the slides were positioned in the mounting medium to avoid air bubbles. Finally, fluorescent microscopy was employed to observe and capture images for the analysis of LC3B and P62 expression levels.

Statistical analysis

Quantitative data were categorized based on their distribution into normal and non-normal distributions. Data with a normal distribution were presented as mean ± standard deviation (SD), while non-normally distributed data were presented as median and interquartile range. For normally distributed data, comparisons between two groups were conducted using Student’s t-test. For non-normally distributed data, comparisons were performed using the Wilcoxon rank-sum test. For comparisons among multiple groups, one-way analysis of variance (ANOVA) followed by Tukey’s honest significant difference test. Total survival rates were computed using the Kaplan-Meier method, with differences in survival assessed using the log-rank test. All statistical analyses and visualizations were performed using GraphPad Prism 9, SPSS 23.0, and R Studio software. In statistical analysis, a significance level of P < 0.05 indicated statistical significance, P < 0.01 denoted highly significant differences, and P < 0.001 signified extremely significant differences.

HCC patients with low circR-4600 expression present poor prognosis

To further screen for significantly differentially expressed circRNAs in HCC, circRNA microarray analyses were conducted on six pairs of HCC and adjacent non-tumor tissues to delineate the aberrantly regulated circRNA expression profile in HCC (Figure 1A). To validate the microarray analysis results, several promising circRNAs were selected for RT-qPCR validation, and the results were generally consistent with those of the microarray analysis. Particularly, hsa_circRNA_104600 (also known as hsa_circ_0005927, circR-4600) exhibited significantly downregulated expression in HCC tissues compared to adjacent non-tumor tissues (Figure 1B). The parental gene of circR-4600 is VDAC3, a protein located in the outer mitochondrial membrane, crucial for ion channel function, especially in cellular metabolism and energy generation processes. Additionally, VDAC3 plays a critical role in maintaining mitochondrial function under oxidative stress conditions

¹⁹

. The VDAC3 gene is located on human chromosome 8, and its transcript is a 1160 bp mRNA. Within this mRNA, a 379 bp sequence forms the circular structure of circR-4600 through back-splicing junction (BSJ) (Figure 1C). To delve deeper into the expression of circR-4600 in different HCC cell lines, RT-qPCR analysis was performed on L02, QGY7703, SMMC7721, HepG2, and Huh7 cell lines. The results showed a significant downregulation of circR-4600 expression in QGY7703, SMMC7721, and HepG2 cell lines compared to L02 cells (P < 0.05), as depicted in Figure 1D. Subsequently, 33 pairs of HCC and corresponding adjacent non-tumor tissue samples were collected, revealing that circR-4600 expression levels were generally lower in HCC tissues than in adjacent non-tumor tissues, consistent with the observed expression pattern in cell lines (Figure 1A). Furthermore, FISH detection confirmed that circR-4600 primarily localized to the cytoplasm (Figure 1F).

Following the initial identification and molecular characterization of circR-4600, we further examined its clinical significance regarding dysregulation. We analyzed circR-4600 expression in a well-characterized HCC patient cohort to assess its correlation with tumor progression and clinical outcomes. 222 pairs of HCC samples and corresponding adjacent non-tumor tissues were collected. RT-qPCR was utilized for quantitative analysis of circR-4600 expression levels. Based on the median relative expression level of circR-4600 in tumor tissues, samples were divided into high-expression and low-expression groups (Figure 2A). The study results revealed a significant decrease in circR-4600 expression in HCC tissues compared to adjacent non-tumor tissues (P < 0.001, Figure 2B). Prognostic evaluation of HCC patients using the Kaplan-Meier method indicated that low circR-4600 expression portended poorer prognosis compared to the high-expression group, manifested by significantly lower OS rates (P < 0.05, Figure 2C). The baseline characteristics of the 222 HCC patients are presented in Tables 1 and 2. These findings suggest that low expression of circR-4600 in HCC tissues may be closely associated with adverse patient prognosis.

circR-4600 inhibits HCC cell proliferation, migration, and invasion

To investigate the role of circR-4600 in HCC cell proliferation, migration, and invasion, a circR-4600 overexpression system was constructed using the pCDH-CMV-MCS-EF1-GFP+Puro (CD513B-1) vector. Control (non-transfected cells), negative control (NC, i.e., pCDH group), and circR-4600 overexpression plasmids were transfected into 293T cells, and total RNA was extracted for RT-qPCR analysis of circR-4600 expression levels. The results demonstrated significantly higher expression levels in the circR-4600 overexpression group compared to the control and NC groups, thereby validating the successful construction of the overexpression system (Figure 3A). The specific primers for circR-4600 (divergent primers) were as follows: Forward: 5’-tagaacttctctctctaaaga-3’, Reverse: 5’-cgaaattattctgtgacagtt-3’. To confirm that the amplified product was circR-4600, sequencing analysis was performed to detect reads matching the BSJ site. The results showed that the amplified product matched the circR-4600 sequence and contained the BSJ site, further confirming the accuracy of the overexpression system (Figure 3B). Subsequently, the overexpression system of circR-4600 was applied to HepG2 and SMMC7721 cells. BrdU assay showed that the proliferation ability of cells overexpressing circR-4600 was significantly lower than that of the control and NC groups, especially in SMMC7721 cells (Figure 3C-D). CCK-8 assay yielded similar results (Figure 3E). Plate clone formation assay further confirmed a significant reduction in the number of clones in SMMC7721 cells after overexpression of circR-4600 (Figure 3F). These data indicate that overexpression of circR-4600 inhibits HCC cell proliferation. To further investigate the impact of circR-4600 on the migration and invasion ability of HCC cells, transwell invasion assay and scratch wound healing assay were conducted. The results showed that the invasion ability of cells overexpressing circR-4600 was significantly reduced, with a decrease in the number of cells crossing the matrigel and membrane (Figure 3G). The scratch assay also demonstrated that the migration speed of cells overexpressing circR-4600 was significantly slower than that of the control and NC groups (Figure 3H). Therefore, by constructing and applying the circR-4600 overexpression system in HCC cells, we found that overexpression of circR-4600 significantly inhibits HCC cell proliferation, migration, and invasion. These results indicate that circR-4600 plays a suppressive role in HCC cell proliferation, migration, and invasion.

circR-4600 inhibits in vivo lung metastasis

To investigate the role of circR-4600 in HCC cell metastasis, a mouse lung metastasis model was established by tail vein injection of SMMC7721 cells stably overexpressing circR-4600 into 6-week-old athymic male nude mice (Figure 4A). Mice in the NC group were examined at the 3rd and 4th weeks to observe liver and lung metastasis. Compared to the NC group, mice in the circR-4600 overexpression group exhibited significantly reduced lung metastatic lesions, indicating an inhibitory effect of circR-4600 on intrahepatic cholangiocarcinoma cell metastasis in vivo (Figure 4B). Subsequently, H&E staining was performed to assess tumor cell lung metastasis. Tumor tissues in the NC group showed pronounced heterogeneity, disordered arrangement of cancer cells, and loss of normal structure and polarity, suggesting a higher incidence of tumor cell invasion compared to the circR-4600 group (Figure 4C). RT-qPCR results revealed higher levels of E-cad and lower levels of Ki67 and MMP9 in the circR-4600 overexpression group (Figure 4D). IHC staining and WB were performed to specifically label Ki67, MMP9, and E-cad proteins. In lung metastatic lesions of mice overexpressing circR-4600, the number of Ki67-positive cells (indicating actively proliferating tumor cells) was significantly reduced, while the expression of MMP9 (a proteinase closely associated with tumor invasion and metastasis) was suppressed. Concurrently, the expression of E-cad (a protein involved in maintaining intercellular adhesion) was markedly enhanced (Figure 4E-F). These findings collectively suggest that circR-4600 inhibits the proliferation, invasion, and metastasis of HCC cells in vivo.

circR-4600 regulates HCC progression through interaction with DVL3

Based on the results from FISH analysis, we observed the expression of circR-4600 in the cytoplasm. This finding suggests that circR-4600 may interact with cytoplasmic proteins, potentially influencing their biological functions. To further investigate the interaction between circR-4600 and cytoplasmic proteins, we conducted a circR-4600 RNA pulldown experiment. In this experiment, we used sense and antisense probes for circR-4600 and identified relevant proteins using ESI-FT-ICR-MS technology (Figure 5A). Among the identified interacting proteins, DVL3 exhibited significant binding activity with circR-4600. Notably, DVL3 plays a crucial role in regulating malignant behaviors of tumors, including cell proliferation, migration, epithelial-mesenchymal transition (EMT), and stemness

^20,

²¹

. This discovery provides a novel perspective on the functional role of circR-4600 in cells. Furthermore, RIP-qPCR experiments using DVL3 antibody confirmed the interaction between DVL3 and circR-4600. RT-qPCR analysis of precipitated RNA validated the presence of circR-4600 (Figure 5B). Additionally, RT-qPCR analysis revealed the impact of circR-4600 overexpression on DVL3 mRNA levels, showing a significant decrease with increasing circR-4600 expression. This was further confirmed by WB experiments, demonstrating that circR-4600 overexpression led to a decrease in DVL3 protein levels (Figure 5C). Collectively, these results indicate a negative regulatory relationship between circR-4600 and DVL3. Furthermore, we constructed DVL3 inhibition plasmids and found that both circR-4600 overexpression and DVL3 inhibition decreased DVL3 expression levels. When cells were simultaneously overexpressed with circR-4600 and inhibited DVL3, DVL3 levels were reduced to the lowest extent (Figure 5D). Overall, these data emphasize the significant interaction between DVL3 protein and circR-4600, providing important clues for understanding their cellular function.

Upregulation of DVL3 correlates with enhanced proliferation, invasion abilities, and altered autophagy levels in HCC cells

We collected 35 samples of HCC tissues along with their corresponding adjacent tissues. The results demonstrated a significant upregulation of DVL3 expression in HCC tissues compared to adjacent tissues, consistent with the analysis from TCGA database (Figure 6A). Survival analysis further revealed a significant decrease in OS in HCC patients with high DVL3 expression levels compared to those with low expression (P < 0.05, Figure 6A). In vitro experiments showed that DVL3 inhibition markedly attenuated the proliferation capability of HCC cells (Figure 6B). Transwell invasion assays further confirmed that DVL3 inhibition weakened the invasive ability of HCC cells, as evidenced by a significant decrease in the number of cells passing through matrigel-coated membranes (Figure 6C). Additionally, DVL3 inhibition reduced the expression of Vimentin and increased E-cad expression, suggesting that DVL3 inhibition may suppress the invasive and migratory abilities of HCC cells and EMT process (Figure 6D). Previous studies have indicated that autophagy-related genes WDFY3 and ALFY can clear DVL3 aggregates through the autophagy pathway, thereby playing a role in disease progression

^22,

²³

. Based on these findings, we examined the effects of DVL3 inhibition on HCC cell autophagy using immunofluorescence and WB techniques. DVL3 inhibition significantly reduced LC3B expression and increased P62 levels, suggesting inhibited autophagic activity (Figure 6E). Further investigation revealed that overexpression of circR-4600 or inhibition of DVL3 could both suppress the expression of LC3B and promote P62 accumulation. When cells were simultaneously overexpressed with circR-4600 and inhibited DVL3, autophagic activity decreased to even lower levels (Figure 6F), providing new insights into the role of DVL3 in HCC development.

ATG5 plays a crucial role in cellular autophagy, participating not only in the formation of autophagosomes and fusion with lysosomes but also regulating the autophagic process, providing an efficient mechanism for cellular recycling and stress response. As a core member of the autophagy-related protein family, ATG5, together with other proteins, forms the basic framework of autophagy. In this study, by constructing ATG5 inhibition plasmids, we found that ATG5 inhibition significantly reduced the migration and invasion capabilities of HCC cells (Figure 7A). Moreover, ATG5 inhibition decreased autophagy levels, manifested by reduced LC3B and increased P62 levels, while also inhibiting the invasive abilities of HCC cells and the EMT process, indicated by downregulation of E-cad and upregulation of Vimentin (Figure 7B). When ATG5 and DVL3 were simultaneously inhibited, autophagic activity was further reduced, accompanied by increased E-cad and decreased Vimentin, suggesting reduced invasive and migratory abilities of HCC cells and restoration of epithelial-like characteristics (Figure 7C).

DAB2 protein participates in HCC Cell invasion, metastasis, and autophagy regulation

Previous studies have revealed that downregulation of DAB2 protein expression is closely associated with the occurrence and progression of various human malignancies. Additionally, DAB2 can interact with Beclin-1, promoting Beclin-1 phosphorylation to regulate autophagosome assembly, thereby influencing the autophagic process

^24,

²⁵

. Interestingly, in our previous circR-4600 RNA pulldown experiments, DAB2 protein exhibited binding activity with circR-4600. Based on this, we performed RIP-qPCR experiments using DVL3 antibody in HCC cells overexpressing DVL3 to elucidate the interaction between DVL3 and DAB2. RT-qPCR analysis of precipitated RNA confirmed the presence of DAB2 (Figure 8A). Furthermore, in DVL3-inhibited HCC cells, DAB2 expression was significantly increased (Figure 8B). In the 35 HCC tissue samples collected, DAB2 expression levels were significantly elevated in adjacent tissues, consistent with the analysis from TCGA database (Figure 8C). Survival analysis further indicated a significant decrease in OS in patients with high DAB2 expression levels compared to those with low DAB2 expression (P < 0.05, Figure 8C). Overall, in HCC cells, circR-4600 overexpression leads to the inhibition of DVL3 expression through its interaction with both DVL3 and DAB2. Consequently, this inhibition suppresses ATG5-mediated cellular autophagy, as evidenced by reduced LC3B levels and increased P62 levels. As a result, HCC cell invasion and metastatic capabilities decline, and epithelial-like features are restored. Additionally, DAB2 may participate in autophagy regulation by interacting with Beclin-1, further influencing HCC cell invasion, metastasis, and the EMT process (Figure 8D).