Abstract Background Clear cell renal cell carcinoma (ccRCC) represents the most prevalent subtype, accounting for nearly 80% of all RCC cases. Recent research has shown that high expression of circular non-coding RNA (circRNA) is associated with poor prognosis in patients with renal clear cell carcinoma (ccRCC), however, the underlying mechanism remains unclear. Methods After analysing self-sequenced renal cancer and paracancer circRNA sequencing data and comparing it with the GEO public database, we discovered that circASAP1 expression was significantly up-regulated in renal cancers. We also tested circASAP1 levels in 102 renal cancer patients and found that high expression of circASAP1 was associated with poor prognosis and metastasis. The interaction between circASAP1, HNRNPC and their downstream target genes was confirmed through experiments such as RNA pull-down, RIP and fluorescence in situ hybridisation. A series of in vitro and in vivo functional experiments were performed to verify the effects of circASAP1 on RCC proliferation and metastasis. Results Circular RNA sequencing analysis revealed that circASAP1 expression was markedly elevated in ccRCC, with a significant association observed between elevated circASAP1 expression and poor prognosis and metastasis. Actinomycin D, RNase R, as well as fluorescence in situ hybridization (FISH) analyses revealed the ring structure and cytoplasmic localization of circASAP1. High circASAP1 expression was associated with ccRCC cell proliferative viability, invasion, and metastasis in CCK-8, transwell, plate cloning, and EdU experiments. Interaction of circASAP1 with HNRNPC and their downstream target genes was confirmed by RNA pull-down, RNA immunoprecipitation, FISH, silver staining, and mass spectrometry. Experiments using truncated isoforms demonstrated that amino acids 16–87 of HNRNPC bound circASAP1. Proteins altered by circASAP1 were enriched in the ferroptosis pathway on the Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis. Conclusions The relationship between circRNA and the ASAP1/HNRNPC/GPX4 axis was demonstrated by experimental data, which was further confirmed by rescue experiments. circASAP1 influenced tumor growth and ferroptosis in animal experiments and predicted the prognosis of patients with ccRCC. The circASAP1/HNRNPC/GPX4 axis provides novel directions and potential targets for RCC treatment. Supplementary Information The online version contains supplementary material available at 10.1186/s12943-024-02122-8. Keywords: Renal cell carcinoma, CircASAP1, HNRNPC, Ferroptosis Introduction The incidence of renal cell carcinoma (RCC) represents approximately 3% to 4% of malignants and has demonstrated an increasing trend over the past decade [[44]1]. Approximately 90% of RCC is of the clear cell subtype (ccRCC) [[45]2]. Typically, radical nephrectomy is the main treatment for primary RCC; nevertheless, even after surgery, around 30% of individuals experience metastases, leading to a high mortality rate [[46]3, [47]4]. Numerous new biomarkers and the mechanisms behind them have been discovered in RCC, showcasing their importance in clinical settings [[48]5]. However, there is still a lack of complete knowledge regarding the initiation and advancement of RCC, and more research is required to uncover the molecular pathways involved in these processes. Circular non-coding RNAs (circRNAs) are a subclass of small RNAs characterized by being covalently closed loops without 5′ caps or 3′ polyadenylate tails [[49]6, [50]7]. Initially, circRNAs were regarded as mere byproducts of transcription, with no discernible function. [[51]8]. Numerous circRNAs have since been identified in mammalian cells by high-throughput sequencing [[52]9], the majority of these molecules are involved in physiological and pathological processes, including the development and progression of cancer [[53]10]. circRNAs can act as microRNA sponges, RNA binding protein (RBP)-binding molecules, transcriptional regulators, or protein translation templates, and their functions in various tumors have been intensively investigated. The molecular functions of circRNAs are related to their subcellular localization, with those in the nucleus potentially involved in gene transcription regulation and variable RNA splicing [[54]11, [55]12], while those in the cytoplasm often act as RNA sponges or interact with RBPs to participate in the post-transcriptional regulation of tumor-related genes [[56]13]; however, the potential roles of circRNAs in RCC remain elusive. Ferroptosis is a new type of programmed cell death that has been reported in the past decade or so. It is characterized by the accumulation of iron in cells, which leads to lipid peroxidation and cell death [[57]14, [58]15]. Pathophysiologically distinct from other forms of cell death (e.g., apoptosis, necrosis), ferroptosis has important roles in various diseases (e.g., cancer, cardiovascular diseases, neurodegenerative diseases, etc.) [[59]16]. In cancer, ferroptosis regulation is closely related to drug resistance, tumor growth, and metastasis, among other processes; hence, studies targeting ferroptosis are relevant to cancer therapy. circRNAs dysregulation may be associated with the pathogenesis, diagnosis, and treatment of various malignant tumors, including ccRCC [[60]17]. For example, it has been demonstrated that circRNA, circPOLR2A, facilitates the progression of ccRCC by inducing UBE3C-catalyzed ubiquitination of PEBP1, which in turn activates ERK signaling [[61]18], while the circRNA, circ-TNPO3, inhibits ccRCC metastasis by binding to IGF2BP2 and destabilizing SERPINH1 mRNA [[62]19]. circRNAs also have significant roles in regulating RCC and ferroptosis. For example, Hsa_circ_0057105 regulates the balance between epithelial-mesenchymal transition and ferroptosis in RCC [[63]20]. Hence circRNAs are of interest in the context of RCC and may assist in identifying potential therapeutic targets. Here, we conducted bioinformatics analysis of ccRCC and paracancer circRNA sequencing data. Our findings indicate that the expression of circASAP1 was significantly up-regulated in ccRCC. Further, high circASAP1 levels in RCC surgical specimens were associated with poor prognosis and metastasis. We found that circASAP1 can bind the RBP protein, HNRNPC, and regulate its levels through the ubiquitin-proteasome pathway, leading to decreased GPX4 protein levels and ferroptosis pathway activation. Our data reveal the important role of circASAP1 in RCC and provide strong evidence of its potential as a therapeutic target in this context. Materials and methods Cell culture and cell lines The Chinese Academy of Sciences provided human RCC cell lines (A498, 786-O, 769P, and Caki-1), a human renal proximal tubular epithelial cell line (HK2), and a human embryonic kidney cell line (HEK293T). 786-O and 769P cells were grown in RPMI 1640 (Gibco, China), with the addition of 10% FBS (PAN-Seratech, Germany).A498, Caki-1, HK2, and HEK-293 T cells were grown in DMEM (Gibco, China) with the addition of 10% FBS (PAN-Seratech, Germany). Incubation occurred at a temperature of 37 degrees Celsius in an environment containing 5% carbon dioxide. Mycoplasma infection in cells was regularly monitored throughout cell culture (Beyotime, China). ccRCC patient samples and follow-up For this study, a group of 102 patients who underwent radical nephrectomy between January 2002 and December 2012 at the First Affiliated Hospital of Sun Yat-sen University (Guangzhou, China) were recruited. It is worth noting that none of the patients received neoadjuvant chemotherapy or radiotherapy. To maintain consistent monitoring of patient health status, the hospital carried out periodic follow-ups for a median duration of up to 99.0 months until 31 December 2023. The overall survival (OS) was defined as the time from surgery until the patient passed away for any reason. In order to conduct a thorough study of RNA expression, the hospital procured formalin-fixed, paraffin-embedded (FFPE) samples containing both tumor tissue and adjacent normal tissue from the patients. Total RNA was extracted from FFPE tissue samples using a nucleic acid isolation kit designed for this purpose (ThermoFisher, U.S.A.). The use of these tissue samples in this study was approved by the Medical Ethics Committee of the First Affiliated Hospital of Sun Yat-sen University to ensure compliance with ethical standards. RNA and gDNA extraction RNA samples were extracted from cells using Trizol (Invitrogen, USA) following the manufacturer’s instructions. Genomic DNA (gDNA) was isolated using a genomic DNA isolation kit (Sangon Biotech, China). All operations are performed in an enzyme-free environment. RNase R treatment, cDNA synthesis, and PCR In order to obtain purified RNA product, 2 μg aliquots of total RNA were subjected to treatment with 3 U/μg RNase R (Lucigen, USA) for 10 minutes at 37℃, followed by reverse transcription to cDNA using a 4× RT Master Mix (EZBioscience, USA) containing random and oligo (dT) primers. The RNA was purified and isolated prior to reverse transcription to cDNA using 4× RT Master Mix (EZBioscience, USA) containing random and oligo (dT) primers. Subsequently, qRT-PCR was performed using 2× SYBR Green qPCR Master Mix (EZBioscience, USA). Primers and siRNA sequences are listed in Additional file 2: Table S4. Agarose gel electrophoresis Agarose gels containing 1.2% concentration were prepared using 1× TAE buffer (Biosharp, China) and nucleic acid dye SYBR Green II (Biosharp, China). DNA nucleic acid samples were electrophoresed at 120 V for 25-35 min. The gel bands were then observed and photographed using a UV gel imaging system (UVP GelStudio PLUS touch, Germany), with the position of the bands indicated using DL2000 DNA Marker Reagent (Tsingke, China) as a DNA marker. Fluorescence in situ hybridization (FISH) The circASAP1 oligonucleotide probe with Cy3 fluorescent labeling was synthesized by RiboBio, China. The probe was hybridised into cells using a fluorescence in situ hybridisation kit (RiboBio, China). Images were taken by a confocal laser scanning microscope (FV1000, Olympus, Japan). Analysis of cell proliferation The CCK-8 kit (MCE, China) was used to detect cell viability. Briefly, 2 × 10^3 cells per well were inoculated in 96-well plates with 3 replicates per group. On days 0, 1, 2, and 3, 10 μL of CCK-8 solution was added to each well and incubated at 37°C for 2 h. The cell viability was measured by measuring the absorbance at 450 nm. Western blot The cellular precipitate was initially obtained via trypsin digestion, followed by cell lysis using a protein lysis buffer (ThermoFisher, USA) containing protease inhibitors (Beyotime, China) on ice. The resulting mixture was thoroughly homogenized and incubated on ice for 15 minutes. Subsequent centrifugation at 13,000 RPM and 4°C for 15 minutes yielded the supernatant, from which the protein concentration was determined using a bicinchoninic acid (BCA) protein quantification kit (ThermoFisher, USA). Thereafter, each lane of the SDS-PAGE gel was loaded with 20 µg of protein sample. The proteins were subsequently transferred to a polyvinylidene fluoride (PVDF) membrane and blocked with a 5% skimmed milk solution. Following this step, the membranes were subjected to overnight incubation at 4°C with the primary antibody and then co-incubated with the secondary antibody for one hour at room temperature on the following day. Hybridizations were detected using a western blot substrate kit (Tanon, China) on a FluorChem E System (ProteinSimple, USA). The antibodies used in the western blots were incubated for 1 hour at room temperature. Hybridizations were detected using a western blot substrate kit (Tanon, China) on a FluorChem E System (ProteinSimple, USA). The western blot was then developed using a substrate kit (Tanon, China) and imaged on a FluorChem E System (ProteinSimple, USA). Antibodies were applied to the western blot and incubated for 1 hour at room temperature. EdU experiment The experiment involved seeding cells into 6-well plates (with coverslips) in amounts of 3×10^5 cells/well and incubating them for 48 hours after treatment with different substances. The cultured cells were then permeabilized with EdU (5-ethynyl-2'-deoxyuridine) using the BeyoClick EdU-594 Cell Proliferation Assay Kit (Beyotime, China). The final concentration of EdU was adjusted to 10 μM (working solution). RCC cells were treated with the working solution and incubated for 1-2 hours. After that, the cells were fixed with 4% paraformaldehyde (Beyotime, China) by immersion for 15 minutes, and rinsed twice with 1×PBS solution for 5 minutes each time at room temperature. Then, 0.3% Triton X-100 PBS was then used to permeabilize the cells. The cells were rinsed twice with 1×PBS solution for 5 min each time at room temperature, followed by incubation with the reaction solution for 30 min. The reaction solution was added to the cells and incubation was 30 minutes. DAPI dye was used to stain the cell nuclei, and the wavelength of 594 nm was used to detect EdU. The images were photographed using an automated inverted fluorescence microscope (IX83, Olympus, Japan). Wound-healing assay siRNA-transfected cells are seeded into 6-well plates. After incubating for 48-72 hours, A 1000 μL pipette tip was used to create an equal-width scratch in the center of each well. To exclude the proliferative effect of FBS on the cells, they were cultured in a serum-free medium. It should be noted that an inverted microscope (Olympus, Japan) was used to capture images at 0 and 24 hours. The migration and motility of the cells were determined by measuring the change in the width of the scratches. Cell migration and invasion assays The cells were starved by changing the medium to a serum-free medium six hours before the experiment. Subsequently, the cells were resuspended in serum-free medium, adjusted to contain 1 × 105 cells in 100 μL, and plated into 24-well transwell plates (Corning, USA). In the invasion assay, the substrate was covered by 2% Matrigel (Corning, USA), and the bottom layer was coated with medium with 10% FBS as chemoattractant. The cells were cultured for 12-18 hours for migration and invasion assays. Subsequently, the upper chamber was fixed with 4% paraformaldehyde (Beyotime, China) and air-dried at room temperature. After that, it was stained with 4% crystalline violet (Beyotime, China) for 15 minutes, washed with PBS, and air-dried. Cells in the upper chamber were removed with a cotton swab and photographed in randomly selected fields using an automatic inverted fluorescence microscope (IX83, Olympus, Japan). The number and percentage of migrating or invading cells were then calculated. RNA pull-down assay RNA pull-down experiments were performed using the Pierce Magnetic RNA-Protein Pull-Down Kit (Millipore, USA). A biotinylated probe for the circASAP1 reverse junction site was designed by RiboBio (China), while an oligo probe (RiboBio, China) was applied as a negative control. Mix the two probes with streptavidin-coated magnetic beads (ThermoFisher, China) and incubate together at room temperature for 2 hours. After washing to remove any unbound or non-specific binding beads, the cell lysis products were mixed with RNA streptavidin magnetic beads and incubated at 4°C overnight. RNA-binding proteins (RBPs) were separated using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), and differential bands were visualised through silver staining (Beyotime, China). Subsequently, the differential proteins were subjected to analysis via mass spectrometry (BGI, China). RNA immunoprecipitation(RIP) The Magna RIP RNA Binding Protein Immunoprecipitation Kit (Millipore, USA) was used to detect RIP. Initially, the target protein HNRNPC was tagged with a Flag tag by transfecting the plasmid into the cells. This enabled us to capture the tagged HNRNPC using Flag magnetic beads without the need for an antibody. Subsequently, RNA was extracted and purified to detect circASAP1 by qRT-PCR. Ferroptosis detection assay The Reactive Oxygen Species(ROS) assay was conducted using BODIPY 581/591 C11 (Invitrogen, USA). ROS were assessed with a 488 nm laser on a CytoFLEX flow cytometer (Beckman, USA) detector. The flow cytometry data were analysed using FlowJo v10 software (BD, USA). The concentration of MDA was evaluated by means of the Malondialdehyde (MDA) Assay Kit (No.BC0025, Solarbio, China). The Mitochondrial Membrane Potential Assay Kit (JC-10) (No. CA1310, Solarbio, China) is an effective method for accurately detecting changes in mitochondrial membrane potential within cells. FerroOrange, a novel fluorescent probe from Dojindo in Japan, can be used for fluorescence imaging of Fe^2+ in living cells. Assay kits(No.BC1170、No.BC1185, Solarbio, China) for the detection of reduced glutathione (GSH) and oxidised glutathione (GSSG) were utilised. Animal experiment All animal ethical and experimental procedures were conducted in accordance with the National Institutes of Health guidelines and approved by the Institutional Animal Care and Use Committee of Sun Yat-sen University. BALB/c nude mice, aged between four and five weeks, were sourced from GemPharmatech (Guangdong, China). To construct a stable strain of cells with sh-NC, sh-RNA of 786-O was injected subcutaneously from mice via the left axilla, using an equal amount of 786-O cells. In this study, tumour volume (V) was calculated using the formula V=(L×W^2)/2. Following this, RSL3 was injected into Group 3 mice via intratumoral injection every 4 days. Tumour diameter (L) and width (W) were measured every 7 days starting from day 7 after inoculation until the maximum diameter of the tumour reached 1.5 cm. Following a period of approximately five weeks, the mice bearing tumors were euthanized, and samples of the tumors were collected. Immunohistochemistry (IHC) was performed on the collected tumours. For lung metastasis experiments, BALB/c nude mice, aged between four and five weeks, were sourced from GemPharmatech (Guangdong, China). A total of 1 × 106 cells of the 786-O stable strain, were injected into each of the mice via the tail vein. Starting four weeks later, D-Luciferin Sodium Salt D (fluorescein sodium salt) (Yeasen, China) was injected intraperitoneally for 5-10 minutes. Images were then taken using an Olympus IX83 inverted microscope (Japan) and the software LivingImage was used to quantify the fluorescence intensity of lung metastases in each sample. After five weeks, all mice were euthanised, and lung tissues were stained with HE. Statistical analysis All statistical analyses were performed using GraphPad Prism version 9.0 with R (version 4.3.1) ([64]https://www.rproject.org/). Comparisons between the two experimental groups were made using student’s t-test. One-way ANOVA and Bonferroni test were performed for three or more groups. Pearson coefficient was used to assess correlation. Survival analysis was performed using the Kaplan-Meier curve and statistical significance was calculated using the logrank test. Univariate and multivariate hazard ratios (HRs) were calculated using Cox regression. All quantitative experimental data were obtained from at least three replicates and expressed as mean ± SD. Data are mean±SD, n=3. *, P<0.05; **, P<0.01; ***, P<0.001. Results Discovery and characterization of the oncogenic role of circASAP1 in ccRCC To identify potentially oncogenic circRNAs in RCC, we first performed in-depth bioinformatic analysis of circRNA sequencing data from cancer and paracancerous tissues from five patients with RCC. Based on the stringent screening criteria of |Log 2fold-change (FC)| > 1 and P < 0.05, The analysis revealed that 53 circular RNAs (circRNAs) exhibited markedly elevated expression levels in cancerous tissues compared to those of paracancerous samples (Fig. [65]1A, Additional file 2: Table. S1). We also conducted bioinformatic analysis of the [66]GSE108735 dataset from the GEO public database, which was derived from seven tumor and matched adjacent normal tissue samples, and identified 580 circRNAs significantly highly expressed using the same screening criteria (Fig. [67]1B, Additional file 2: Table. S2). A comparison of these two datasets demonstrated that 9 circRNAs showed significantly increased expression in RCC in both datasets (Fig. [68]1C). After reviewing the relevant literature, we found that circASAP1 (also known as circASAP1) has been relatively understudied in RCC, which provides an opportunity to further use it as a research target. Therefore, we assessed the relationship of circASAP1 expression levels with treatment outcomes and prognosis in patients with ccRCC; the expression of circASAP1 exhibited a positive correlation with poor overall survival (Additional file 1: Fig. S1), and circASAP1 expression level was correlated with TNM stage (Table [69]1). Therefore, we chose to conduct further analysis of circASAP1 in ccRCC. circASAP1 is a 550 bp transcript derived from the ASAP1 gene on chromosome 8, formed by reverse splicing of exons 8, 9, 10, 11, 12, and 13. To ensure the accuracy of circASAP1 reverse splicing, we validated our findings by Sanger sequencing (Fig. [70]1D). Fig. 1. [71]Fig. 1 [72]Open in a new tab Discovery and characterization of the oncogenic role of circASAP1 in ccRCC. A–B Volcano plots comparing circRNA levels in ccRCC and adjacent normal tissues sequenced in-house and the [73]GSE108735 dataset. Red, green, and grey dots represent up-regulated, down-regulated, and non-significantly different circRNAs, respectively. C Venn diagram of in-house and [74]GSE108735 sequence datasets. Overlapping circRNAs |log[2] (fold-change, FC)| ≥ 1 and P <0.05 were selected. D Schematic representation of the formation of circASAP1 from the ASAP1 gene on chromosome 8. The reverse splice site back splicing junction (BSJ) was verified by Sanger sequencing. Black arrows indicate specific junctions. E circASAP1 expression levels were detected by qRT-PCR in the normal renal cell line, HK2, and various renal cancer cell lines. F Assessment of circASAP1 and ASAP1 mRNA stability by qRT-PCR following treatment with actinomycin D (5 µg/ml) for various periods of time. G Assessment of circASAP1 and ASAP1 mRNA stability by qRT-PCR after treatment with RNase R (3U/μg RNA, 10 min, 37°C). H qRT-PCR using convergent and divergent primers, followed by agarose gel electrophoresis confirming the presence of circASAP1. circASAP1 could only be amplified from cDNA. β-actin served as a control. Red arrows, divergent primers. Black arrows, convergent primers. I Localization of circASAP1 in cells detected by fluorescence in situ hybridization (FISH). Nuclei were labelled with DAPI dye. The positive controls, 18s, and U6, were predominantly present in the cytoplasm and nucleus, respectively. CircASAP1 was mainly located in the cytoplasm. Data are mean±SD, n=3. *, P<0.05; **, P<0.01; ***, P<0.001 Table 1. Association of circASAP1 expression with clinicopathological characteristics in 102 patients with ccRCC Parameter Total circASAP1 expression P value High Low Age (years)  < 60 65 32 33 0.837  ≥ 60 37 19 18 Sex  Male 75 38 37 0.916  Female 27 14 13 TNM stage  I 74 32 42 0.027  II–III 28 19 9 Tumor  < 5 cm 53 26 27 0.843  ≥ 5 cm 49 25 24 [75]Open in a new tab Chi-square test was used to compare clinicopathological characteristics between high/low circASAP1 groups of patients with ccRCC (n = 102). Median circASAP1 expression was used as the cut-off point Next we assessed circASAP1 expression in a panel of renal cancer cell lines using qRT-PCR and found that it was significantly increased in two RCC cell lines, 786-O and Caki-1, relative to those in the normal renal epithelial cell line, HK2; however, increased circASAP1 expression was not significant in two other RCC cell lines, 769-P and A498 (Fig. [76]1E). To further investigate the stability of circASAP1 in comparison to ASAP1 mRNA, we initially exposed the 786-O and Caki-1 cell lines to actinomycin D (5 µg/ml) for varying durations. Subsequently, quantitative reverse transcription polymerase chain reaction (qRT-PCR) analysis revealed that the half-life of circASAP1 was notably longer than that of ASAP1 mRNA in both cell lines (Fig. [77]1F). In addition, cell treatment with RNase R showed that circASAP1 exhibited greater resistance to RNase R and was less susceptible to degradation, whereas ASAP1 mRNA was less resistant to RNase R and significantly degraded (Fig. [78]1G). Together, These findings collectively illustrate that circular circASAP1 exhibits greater stability than linear ASAP1 mRNA, with a reduced propensity to undergo degradation within cells. To confirm the presence of circASAP1 in tumor cells, we designed convergent and divergent primers, where convergent primers could amplify genomic DNA, linear RNA, and cyclic RNA, while divergent primers could only amplify cyclic RNA. qRT-PCR using these primers showed that amplification of large numbers of PCR products using divergent primers was only achieved after reverse transcription of RNA extracted from the cells (Fig. [79]1H), confirming the presence of circASAP1 in tumor cells. Finally, to study the distribution of circASAP1 within the cells, we designed circASAP1-specific fluorescence in situ hybridization (FISH) probes, labeled with Cy3 fluorescent dye, and stained the cell nuclei with DAPI. These experiments demonstrated that circASAP1 was mainly localized in the cytoplasm of tumor cells (Fig. [80]1I), providing important clues for our subsequent analyses of circASAP1 function and mechanism of action in renal cancer. circASAP1 promotes RCC proliferation and invasion After confirming the presence of circASAP1 and its high expression in RCC cells, we next investigated its potential biological roles and their underlying mechanisms. To control circASAP1 expression levels, we constructed two siRNAs specifically targeting the reverse splice junction, which successfully knocked down circASAP1 expression when transfected into the circASAP1-overexpressing RCC cell lines, 786-O and Caki-1, while expression of linear ASAP1 mRNA was almost unaffected (Fig. [81]2A). Fig. 2. [82]Fig. 2 [83]Open in a new tab Knockdown of circASAP1 inhibits RCC cell proliferation and invasion. A Levels of circASAP1 in 786-O and Caki-1 cells transfected with siRNA or siRNA-NC determined by qRT-PCR. GAPDH was used as an internal reference. B EdU assay to assess the proliferation of 786-O and Caki-1 cells transfected with circASAP1 siRNAs or siNC. Scale bar, 200 μm. C CCK8 assay assessing the proliferation of 786-O and Caki-1 cells transfected with circASAP1 siRNAs or siNC. The proliferation rate was normalized to day 0. D Plate colony formation assay to assess the proliferative capacity of 786-O and Caki-1 cells transfected with circASAP1 shRNA or shNC lentivirus. E Transwell assay to assess the migration and invasion abilities of 786-O and Caki-1 cells transfected with circASAP1 siRNA or siNC for 48 h. Photographed by inverted microscope with 10× objective. F Wound healing assay to assess the migration potential of 786-O and Caki-1 cells transfected with circASAP1 siRNAs or siNC. Photographed by inverted microscope with 10× objective at 0 and 24 h. The scratch width was normalized to 0 has the base point. G qRT-PCR to detect circASAP1 expression levels in 769-P cells transfected with circASAP1 overexpression plasmid or control vector. GAPDH was used as an internal reference. H CCK8 assay to assess 769-P cell proliferation after transfection with circASAP1 overexpression plasmid or control vector. The proliferation rate was normalized to day 0. I Transwell assay to assess the migration and invasion abilities of 769-P cell transfected with circASAP1 overexpression plasmid or control vector. Photographed by inverted microscope with 10× objective. Data are mean±SD, n=3. *, P<0.05; **, P<0.01; ***, P<0.001 Then, EdU and CCK-8 assays were conducted to compare the proliferation of RCC cells in the circASAP1 siRNA (knockdown) and siRNA-NC (negative control) groups. The proliferation of 786-O and Caki-1 cells was significantly lower in the circASAP1 knockdown group than in the control group (Fig. [84]2B, C). Plate colony formation assay further verified that RCC cell proliferation was reduced on circASAP1 knockdown (Fig. [85]2D). The effects of circASAP1 expression on RCC cell migration and invasion were investigated using transwell and scratch assays. Both migration and invasion were significantly lower in the circASAP1 knockdown group than those in the control group (Fig. [86]2E, F). Next, we designed and constructed a circASAP1 overexpression vector and transfected it into the 769-P RCC cell line, which had relatively low circASAP1 levels. qRT-PCR assay revealed that circASAP1 expression was significantly upregulated in 769-P cells after transfection of the overexpression plasmid (Fig. [87]2G). Next, CCK-8 and EdU assays were conducted to investigate the effect of high circASAP1 expression on RCC cell proliferation. Proliferation of the circASAP1 overexpression group was significantly higher than that of the negative control group (Fig. [88]2H, Additional file 1: Fig. S2A). Finally, we conducted transwell and scratch assays to determine the effect of high circASAP1 expression on RCC cell migration and invasion, and found that the migration and invasion of RCC cells in the circASAP1 overexpression group were significantly enhanced (Fig. [89]2I, Additional file 1: Fig. S2B). circASAP1 binds to HNRNPC in RCC cells circRNAs are reported to interact with RBPs [[90]21, [91]22]; therefore, we designed and synthesized a streptavidin probe against circASAP1, based on the circRNA reverse shear site, and used an RNA Pull-down kit to successfully capture circASAP1 together with its bound proteins. SDS-PAGE and silver staining demonstrated that, compared with bands in the antisense probe control lane, bands of approximately 37 kDa were markedly enriched in the circASAP1 probe lane of the silver-stained gel (Fig. [92]3A), suggesting that circASAP1 may interact with some specific proteins. Fig. 3. [93]Fig. 3 [94]Open in a new tab circASAP1 binds to HNRNPC in RCC cells. A Protein complex pulled down by the circASAP1 probe was detected by silver staining. Bands indicated by arrows correspond to the positional size of the HNRNPC protein. B Venn diagrams showing the RNA binding proteins (RBPs) that can potentially bind circASAP1 predicted using three databases (circAtlas, RBPmap, and RBPsuite), as well as the proteins detected by mass spectrometry (MS). C HNRNPC secondary spectra obtained by MS. D HNRNPC protein structure obtained using SWISS-MODEL, a protein structure homology modelling server. E Western blot showing that HNRNPC was captured by the circASAP1 probe in pull-down experiments. F After transfection of plasmid encoding HNRNPC with 3× Flag tags into cells, HNRNPC proteins expressing Flag tags were successfully captured in RIP experiments using anti-Flag magnetic beads, while HNRNPC proteins not expressing Flag tags were not captured. G qRT-PCR detection of RNA samples obtained in RIP experiments revealed that circASAP1 was captured by HNRNPC. RNA samples obtained from IgG magnetic beads served as controls. H Co-localization of circASAP1 and HNRNPC in the cytoplasm detected by immunofluorescence-FISH. Nuclei were stained with DAPI dye. I Schematic representation of full-length and truncated HNRNPC isoforms. J Western blot verification that cells transfected with plasmids successfully expressed full-length and truncated HNRNPC with 3× Flag (left) and full-length and truncated HNRNPC obtained by anti-Flag magnetic bead capture in RIP assay (right). K–L qRT-PCR and agarose gel electrophoresis demonstrate that circASAP1 binds to HNRPC amino acids 16–87. Data are mean±SD, n=3. *, P<0.05; **, P<0.01; ***, P<0.001 To further investigate this finding, we conducted mass spectrometry (MS) analysis to identify the proteins obtained in the pull-down assay; proteins shared by intersecting fractions were excluded by rigorous data analysis and screening. Finally, 111 proteins uniquely captured by the circASAP1 probe were identified and screened against the UniProt protein database (Additional file 2: Table. S3). Comparison of the 111 identified unique proteins with molecules predicted to bind to circASAP1 using databases including circAtlas, RBPmap, and RBPsuite, identified the intersecting protein, HNRNPC (Fig. [95]3B). To verify the reliability of the MS results, we also conducted secondary spectral analysis, which further supported our findings (Fig. [96]3C). After reviewing relevant literature reports, we found that the molecular weight of HNRNPC is approximately 37 kDa, consistent with our silver staining results. Furthermore, we obtained HNRNPC protein structure information using the online protein structure homology modelling server, SWISS-MODEL (Fig. [97]3D). Finally, we subjected the pull-down protein samples to SDS-PAGE and performed western blot analysis using an anti-HNRNPC antibody; the circASAP1 probe pulled down HNRNPC protein, while the antisense probe did not (Fig. [98]3E). We next conducted a series of RNA immunoprecipitation (RIP) experiments to validate our finding that the HNRNPC protein was bound by circASAP1. A plasmid encoding HNRNPC with 3× Flag tags was constructed and transfected into 293T cells, and Flag-tagged HNRNPC protein was successfully captured using anti-Flag magnetic beads (Fig. [99]3F). Subsequently, we detected the presence of circASAP1 in RIP experiments using qRT-PCR. Significantly more circASAP1 was detected in the experimental group (Flag) than in the control group (IgG) (Fig. [100]3G), confirming the interaction between circASAP1 and HNRNPC. To visualize the intracellular localization of circASAP1 and HNRNPC, we conducted combined FISH and immunofluorescence analysis using laser confocal microscopy. The results showed that circASAP1 and HNRNPC co-localized in the cytoplasm (Fig. [101]3H), providing visual evidence of their mutual binding. To explore the specific binding sites of circASAP1 and HNRNPC, we reviewed relevant literature on HNRNPC [[102]23] and generated a series of plasmids encoding truncated HNRNPC isoforms based on its structural domains (Fig. [103]3I, Additional file 2: Table. S5). These plasmids were then transfected into 293T cells to express their corresponding proteins (Fig. [104]3J, left), and RIP capture was conducted using anti-Flag magnetic beads. Western blot assay verified the successful capture of HNRNPC truncated isoforms of different molecular weights in the RIP experiments (Fig. [105]3J, right). RNA samples obtained by RIP experiments were amplified by qRT-PCR and separated by agarose gel electrophoresis. The results showed that circASAP1 binds to amino acids 16–87 of HNRNPC (Fig. [106]3K and L). circASAP1 regulates HNRNPC ubiquitination level and promotes its degradation After confirming the binding relationship between circASAP1 and HNRNPC, we investigated whether circASAP1 regulates HNRNPC. We transfected siRNA into 786-O and Caki-1 RCC cells to reduce circASAP1 expression, then extracted proteins and RNA from these cells for analysis. Western blot experiments showed significantly lower HNRNPC levels in the siRNA-1 and siRNA-2 (knockdown) groups than those in the siRNA-NC (negative control) group (Fig. [107]4A); however, qRT-PCR revealed no significant difference in HNRNPC mRNA levels between the circASAP1 knockdown and negative control groups (Fig. [108]4B). Fig. 4. [109]Fig. 4 [110]Open in a new tab circASAP1 regulates HNRNPC ubiquitination levels and promotes its degradation. A Western blot analysis showing changes in HNRNPC protein levels in 786-O and Caki-1 cells transfected with circASAP1 siRNA or siNC. α-Tubulin was used as an internal control. B qRT-PCR to detect changes in HNRNPC mRNA levels in 786-O and Caki-1 cells transfected with circASAP1 siRNA or siNC. GAPDH served as internal control. C HNRNPC mRNA stability in 786-O and Caki-1 cells transfected with circASAP1 siRNA or siNC were examined by qRT-PCR after treatment with actinomycin D (5 µg/ml) for different periods of time. D 786-O and Caki-1 cells transfected with circASAP1 siRNA or siNC were treated with actinomycin ketone (150 µmol/L) for different times, and HNRNPC protein levels were examined by western blot. α-Tubulin was used as an internal control. E 786-O and Caki-1 cells transfected with circASAP1 siRNA or siNC for 48 h were treated with MG132 (20 µM, 8 h). HNRNPC protein levels were examined by western blot. α-Tubulin was used as an internal control. F Flag-tagged HNRNPC proteins were captured by co-immunoprecipitation, and their ubiquitination levels were detected by western blot. α-Tubulin was used as an internal control. G After transfection of different amounts of siRNA, Flag-tagged HNRNPC proteins were captured by co-immunoprecipitation, and their ubiquitination levels were detected by western blot. α-Tubulin was used as an internal control. Data are mean±SD, n=3. *, P<0.05; **, P<0.01; ***, P<0.001 To assess whether circASAP1 influenced HNRNPC mRNA stability, RCC cells transfected with siRNA were treated with actinomycin D to inhibit new RNA synthesis and RNA extracted for qRT-PCR at different times after treatment. HNRNPC mRNA stability was not significantly altered in the siRNA-1- and siRNA-2-treated groups relative to the siRNA-NC group (Fig. [111]4C), suggesting that circASAP1 primarily regulates HNRNPC protein levels, rather than its mRNA level or stability. To investigate the effect of circASAP1 on HNRNPC protein stability, following transfection with siRNA RCC cells were treated with actinomycin ketone to inhibit new protein synthesis. Western blot experiments showed that HNRNPC protein half-life was significantly lower in the siRNA-1- and siRNA-2-treated groups than in the siRNA-NC group (Fig. [112]4D), confirming that circASAP1 knockdown reduced HNRNPC protein stability. As HNRNPC regulation by circASAP1 primarily occurs at the protein level, we hypothesized that circASAP1 may induce HNRNPC degradation through the ubiquitin-proteasome pathway by altering its ubiquitination level. To test this hypothesis, we treated 786-O and Caki-1 RCC cells with siRNA followed by proteasome inhibitor MG132 inhibition. Western blot analysis showed that circASAP1 knockdown after MG132 treatment did not significantly affect the HNRNPC protein level (Fig. [113]4E), supporting the ubiquitin-proteasome pathway hypothesis. Next, we co-transfected Flag-HNRNPC and HA-Ub plasmids into RCC cells and regulated circASAP1 expression by transfection of different siRNAs after MG132 treatment. The degree of HNRNPC protein ubiquitination was significantly higher in the circASAP1 knockdown group than in the control group (Fig. [114]4F). To further verify the effect of circASAP1 knockdown on HNRNPC ubiquitination, we used different doses of siRNA for transfection, to achieve varying degrees of circASAP1 expression knockdown. Co-immunoprecipitation and western blot detected a positive correlation between the degree of HNRNPC ubiquitination and the level of circASAP1 knockdown (Fig. [115]4G). circASAP1 reduces GPX4 mRNA stability via HNRNPC To explore the potential impact of circASAP1 on downstream pathways in RCC cells, we knocked down circASAP1 in 786-O RCC cells by siRNA transfection, followed by sequencing the transcriptome of the knockdown cells and bioinformatics analysis. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis indicated that circASAP1 knockdown significantly enriched ferroptosis pathways (Fig. [116]5A). Following a literature review of publications related to ferroptosis [[117]14–[118]16, [119]24], to identify representative markers, circASAP1 was knocked down in 786-O and Caki-1 RCC cells and proteins extracted. Then, proteins associated with ferroptosis, including SLC7A11, ACSL4, GPX4, and FTH1, were analyzed by western blot. GPX4 protein levels were significantly lower in cells with circASAP1 knocked down than those in the control group (Fig. [120]5B). Fig. 5. [121]Fig. 5 [122]Open in a new tab circASAP1 enhances GPX4 mRNA stability via HNRNPC. A KEGG enrichment analysis showed that ferroptosis pathways were enriched in transcriptomic data from RCC cell lines with circASAP1 knocked down. B Western blot analysis showing changes in ferroptosis-related proteins after circASAP1 knockdown in 786-O and Caki-1 cells. α-Tubulin was used as an internal control. C Western blot analysis showing changes in GPX4 protein after HNRNPC knockdown in 786-O and CaKi-1 cells. α-Tubulin was used as an internal control. D Western blot analysis showing changes in GPX4 protein levels after HNRNPC knockdown/overexpression in rescue experiments using 786-O and CaKi-1 cells. α-Tubulin was used as an internal control. E qRT-PCR analysis showing changes in GPX4 mRNA levels in 786-O and Caki-1 cells transfected with HNRNPC siRNA or siNC. GAPDH was used as an internal control. F GPX4 mRNA stability was assessed by qRT-PCR after treatment of 786-O and Caki-1 cells transfected with circASAP1 siRNA or siRNA-NC with actinomycin D (5 µg/ml) for different periods of time. Data are mean±SD, n=3. *, P<0.05; **, P<0.01; ***, P<0.001 Our data indicated that circASAP1 knockdown impacts the levels of its binding protein, HNRNPC; therefore, we hypothesized a potential link between circASAP1, HNRNPC, and GPX4. To test this hypothesis, siRNAs against HNRNPC were designed and transfected into 786-O and Caki-1 RCC cells to knock down HNRNPC, and then GPX4 protein levels were detected by western blot. GPX4 levels were significantly lower in the HNRNPC knockdown group than those in the control group (Fig. [123]5C). To explore the relationship among circASAP1, HNRNPC, and GPX4, we conducted rescue experiments to determine whether HNRNPC directly regulates GPX4 or if it is regulated by circASAP1. We found that circASAP1 regulates GPX4 via HNRNPC, which directly regulates GPX4. Furthermore, was observed a positive correlation between GPX4 and HNRNPC expression levels (Fig. [124]5D). To further explore the specific mechanism by which HNRNPC regulates GPX4, 786-O, and Caki-1 RCC cells were transfected with siRNA to silence HNRNPC, then GPX4 mRNA expression levels detected by qRT-PCR. A notable decrease in GPX4 mRNA was detected in the HNRNPC knockdown group relative to the control group (Fig. [125]5E), suggesting that HNRNPC controls GPX4 protein quantity by regulating its mRNA levels. Next, 786-O and Caki-1 RCC cells transfected with siRNA were treated with actinomycin D to impede new RNA synthesis. Subsequent qRT-PCR analysis showed that GPX4 mRNA stability was significantly lower in the siRNA-1 and siRNA-2 groups than in the HNRNPC siRNA-NC group at different time points (Fig. [126]4F). These data suggest that circASAP1 functions in regulating GPX4 protein levels in RCC cells via HNRNPC, which reduces GPX4 mRNA stability, thereby affecting its protein levels. circASAP1 knockdown induces RCC cell ferroptosis circASAP1 knockdown resulted in decreased levels of GPX4 protein, which has a crucial role in ferroptosis. To investigate the effect of circASAP1 on ferroptosis, 786-O and Caki-1 RCC cells were divided into experimental groups, as follows: unstained, siRNA-NC, siRNA-1, siRNA-2, and RSL3-treated; cells in the latter four groups were stained using the lipid peroxidation probe, C11, and reactive oxygen species (ROS) in content each subgroup detected by flow cytometry. The results indicated a significant increase in ROS content in cells with circASAP1 knocked down relative to those in the siRNA-NC group (Fig. [127]6A). To further investigate the ferroptosis process, levels of malondialdehyde (MDA), a key product produced during lipid peroxidation, were examined in RCC cells using an MDA kit. Following circASAP1 knockdown, MDA content in 786-O and Caki-1 cells was significantly higher in the circASAP1 knockdown group than that in the control group (Fig. [128]6B). Further, a decrease in mitochondrial membrane potential was observed in cells in the circASAP1 knockdown group (Fig. [129]6C). To better understand the intracellular reduction state, FerroOrange was used to detect ferrous ions. A significant increase in intracellular ferrous ion content was detected in the circASAP1 knockdown group (Fig. [130]6D). In ferroptosis, the antioxidant effect of intracellular reduced glutathione (GSH) is inhibited, resulting in heightened oxidative stress. Therefore, levels of GSH and its oxidation product, oxidized glutathione (GSSG), were simultaneously examined in circASAP1 knockdown cells. GSH content in the circASAP1 knockdown group was significantly lower than that in the control group, while GSSG content was significantly higher (Fig. [131]6E). Fig. 6. [132]Fig. 6 [133]Open in a new tab circASAP1 knockdown induces RCC cell ferroptosis. A ROS detection in 786-O and Caki-1 cells by flow cytometry after staining with BODIPY 581/591 C11 probe. Unstained cells served as a negative control and the RSL3-treated group as a positive control. B Malondialdehyde (MDA) measurement in 786-O and Caki-1 cells. C Mitochondrial membrane potential in 786-O and Caki-1 cells was detected using JC-10. When mitochondrial membrane potential was high, JC-10 aggregated and fluoresced red; when mitochondrial membrane potential was low, JC-10 was monomeric and fluoresced green. D Detection of ferrous ions in 786-O and Caki-1 cells using FerroOrange. E Reduced (GSH) and oxidized (GSSG) glutathione levels were assayed in 786-O and Caki-1 cells. F Mitochondrial structure in 786-O and Caki-1 cells after circASAP1 knockdown observed by transmission electron microscopy. Scale bars, 2 µm at 2000× and 500 nm at 4000×. Red arrows indicate mitochondria. Data are mean±SD, n=3. *, P<0.05; **, P<0.01; ***, P<0.001 Next, we used transmission electron microscopy to investigate the effect of circASAP1 on RCC cell mitochondrial structure. circASAP1 knockdown in 786-O and Caki-1 cells led to a significant reduction in mitochondrial volume, increased bilayer membrane density, wrinkled mitochondrial membrane, and shallow and flat mitochondrial cristae, relative to the control group (Fig. [134]6F). These findings confirm the significant role of circASAP1 in RCC cell ferroptosis. HNRNPC overexpression rescues RCC cell ferroptosis caused by circASAP1 knockdown circASAP1 knockdown resulted in ROS accumulation in RCC cells, which in turn triggered ferroptosis. Further, HNRNPC directly regulates GPX4 via circASAP1. To investigate the effects of circASAP1 and HNRNPC on ferroptosis, 786-O and Caki-1 RCC cells were divided into experimental groups, as follows: control, si-NC+Vector, si-circASAP1+Vector, si-NC+OE, and si-circASAP1+OE. Following staining with the C11 lipid peroxidation probe, ROS levels in each group were quantified by flow cytometry. Cellular ROS content in the si-circASAP1+Vector group was significantly higher than that in the si-NC+Vector group (Fig. [135]7A). Conversely, no significant difference in ROS content was detected between the si-NC+OE and the si-circASAP1+OE groups. Next, we quantified MDA content in RCC cells from each group. The si-circASAP1+Vector group had significantly higher MDA content than that in the control group (Fig. [136]7B), and HNRNPC overexpression reversed this effect. Furthermore, the mitochondrial membrane potential assay demonstrated that mitochondrial membrane potential was significantly reduced in the si-circASAP1+Vector group (Fig. [137]7C), while HNRNPC overexpression rescued this state. To evaluate the intracellular reduction state, FerroOrange was used to detect ferrous ions, and demonstrated that HNRNPC reversed the elevation of intracellular ferrous ion content induced by circASAP1 knockdown (Fig. [138]7D). Furthermore, levels of GSH and its oxidation product, GSSG, were also evaluated and GSH content was significantly lower in the si-circASAP1+Vector group than in the control group, while GSSG content was significantly higher (Fig. [139]7E); however, HNRNPC overexpression alleviated this oxidative stress state. Fig. 7. [140]Fig. 7 [141]Open in a new tab HNRNPC overexpression rescues ferroptosis caused by circASAP1 knockdown in RCC cells. A Representative images (upper) and quantification data (lower) of ROS detection in 786-O/Caki-1 cell rescue experiments. B MDA detection in 786-O/Caki-1 cell rescue experiments. C Representative images (upper) and quantification data (lower) of mitochondrial membrane potential detection in JC-10 of 786-O/Caki-1 cell rescue experiments. D Representative images (left) and quantification data (right) of ferrous ion detection in 786-O/Caki-1 cell rescue experiments. E Reduced glutathione (GSH) and oxidized glutathione (GSSG) levels in 786-O and Caki-1 cell rescue experiments. Data are mean±SD, n=3. *, P<0.05; **, P<0.01; ***, P<0.001 Knockdown of circASAP1 inhibits RCC tumor proliferation and metastasis in vivo To assess the impact of circASAP1 on cancer cell proliferation and metastasis in vivo, we devised a range of animal experiments. Three stable 786-O cell lines were generated, namely: circASAP1 Control (negative control), circASAP1 shRNA (knockdown), circASAP1 Vector (negative control), and circASAP1 OE (overexpression) groups. To form subcutaneous tumors, 1×10^6 786-O cells were subcutaneously injected through the left axilla of nude mice. To observe the effects of RSL3, half of the mice in the control and knockdown groups were randomly selected and intraperitoneally injected with RSL3 (100 mg/kg) every four days and labelled. Tumor volume was measured weekly. Mice were euthanized when the maximum diameter of the largest tumor was around 1.5 cm, and the tumors were dissected and weighed. Subcutaneous tumor growth in nude mice was significantly reduced when circASAP1 was knocked down or RSL3 was injected. Further, combined circASAP1 knockdown and RSL3 treatment resulted in significantly lower tumor weight and volume (Fig. [142]8A, B, C). Moreover, circASAP1 overexpression increased tumor growth relative to the Vector control group (see Fig. [143]8D, E, F). To examine the effect of circASAP1 on renal cancer metastasis in vivo, experiments were conducted using a lung metastasis model. Equal numbers of 786-O cells (1×10^6) from the circASAP1 control, circASAP1 shRNA, circASAP1 Vector, and circASAP1 OE groups were injected into the tail veins of nude mice. Following euthanasia, lung tissues were subjected to hematoxylin and eosin (HE) staining. Nude mice in the circASAP1 knockdown group had fewer lung metastatic foci than the control group, while those in the circASAP1 overexpression group had more metastatic foci in the lungs (Fig. [144]8G). These findings were supported by the results of HE staining (Fig. [145]8H). Immunohistochemical analysis of subcutaneous tumors showed reduced levels of Ki67, HNRNPC, and GPX4 expression in the circASAP1 knockdown group, in contrast to those in the circASAP1 overexpression group (Fig. S3A, S3B). These results suggest that circASAP1 knockdown in vivo inhibits RCC tumor progression. Fig. 8. [146]Fig. 8 [147]Open in a new tab Knockdown of circASAP1 inhibits RCC tumor proliferation and metastasis in vivo. A Subcutaneous tumor formation in nude mice in different treatment groups: circASAP1 Control, circASAP1 shRNA, Control+RSL3-treated, shRNA+RSL3-treated. Injection: intratumoral injection, 100 mg/kg, every four days. B Weight of subcutaneous tumors in each group (n=5 nude mice per group). C Length and width of subcutaneous tumors in nude mice were measured once a week to calculate the tumor volume, as follows: V = (L × W^2) / 2, where V denotes tumor volume, L indicates tumor length, and W denotes tumor width. D Subcutaneous tumor formation in nude mice in different treatment groups: circASAP1 Vector, circASAP1 OE. E Weight of subcutaneous tumors in each group (n=5 nude mice per group). F Length and width of subcutaneous tumors in nude mice were measured once a week to calculate the tumor volume. G Images of lung metastasis in live animals in different treatment groups. Images were captured using an Olympus IX83 inverted microscope after intratumoral injection of D-Luciferin for 5–10 min, and fluorescence intensity was quantitatively analyzed using LivingImage software. H Hematoxylin and eosin staining of lung tissue specimens from euthanized mice. Arrows indicate lung metastases. I Mechanism of action of circASAP1 induces renal clear cell carcinoma ferroptosis by binding to HNRNPC and thereby regulating GPX4. Data are mean±SD, n=3. *, P<0.05; **, P<0.01; ***, P<0.001 Discussion RCC is a prevalent malignant tumor of the urinary system that has gradually increased in incidence worldwide in recent years [[148]25]. RCC originates from the tubular epithelial cells of the kidney and can be divided into several pathological types, with ccRCC the most common [[149]1, [150]26]. The occurrence of RCC is closely associated with various genetic, environmental, and lifestyle factors [[151]27] and it is commonly treated by surgical resection, drug therapy, and radiotherapy, with surgical resection generally preferred for patients with early-stage RCC. In cases where the cancer has advanced or metastasized, drug therapy and radiotherapy are often important adjuvant treatments. Continuous developments in immunotherapy [[152]28, [153]29] and targeted therapy [[154]30, [155]31] have significantly improved the efficacy of kidney cancer treatment in recent years; however, despite the increasing therapeutic options for renal cancer, prognosis is influenced by various factors, including patient age, tumor stage, pathological type, and treatment modality. Improving patient survival rates and quality of life is a crucial goal, and prevention and screening of RCC can contribute to these aims. The roles of non-coding RNAs in tumorigenesis and development are increasingly recognized. circRNAs are non-coding RNAs with a unique circular structure and stable expression properties [[156]17, [157]32], and have garnered significant interest due to their potential diagnostic and therapeutic value in the study of RCC. circRNAs are formed from precursor mRNA molecules through reverse splicing and are not affected by RNA exonucleases. Additionally, they exhibit abundant variety and quantity. Abnormal expression of circRNAs has been observed in different types of tumors, and they function in regulating biological processes including tumor cell proliferation, apoptosis, migration, and invasion. Consequently, circRNAs have been investigated as a novel tumor markers and potential therapeutic targets [[158]13]. Further, several studies have reported specific mechanisms involving circRNA activity in RCC, where circRNAs can act as sponges for microRNAs, adsorbing and inhibiting their function, which in turn neutralizes their inhibitory effects on target genes [[159]9, [160]33]. Further, RBPs influence circRNA formation, stability, and functional regulation by binding to them. RCC cells may undergo malignant transformation and progression due to aberrant expression of certain RBPs and their interaction with circRNAs [[161]34, [162]35]. Additionally, circRNAs and RBPs can regulate signaling pathways, gene transcription, and protein translation, which impact renal carcinogenesis and progression [[163]18, [164]21, [165]22], and the insights provided by determining such regulatory mechanisms can lead to a better understanding of RCC pathogenesis [[166]36, [167]37]. A close link between ferroptosis and tumors has been discovered in recent years [[168]16, [169]38, [170]39]. Ferroptosis is a novel mode of cell death characterized by increased oxidative stress, resulting from intracellular iron ion overload [[171]14, [172]16]. Inducing or inhibiting RCC cell ferroptosis can significantly affect biological behaviors, including cell proliferation, migration, and invasion [[173]40, [174]41]. Abnormalities in iron metabolism are often detected in RCC cells, leading to an overload of intracellular iron ions, which can result in the generation of high ROS levels through the Fenton reaction, causing oxidative stress and cellular damage, and if the damage exceeds the repair capacity of cells, it can trigger ferroptosis. Regulating iron metabolism in renal cancer cells can induce ferroptosis, which may inhibit RCC growth and spread. Furthermore, alterations in the expression of certain genes and proteins can impact the sensitivity of RCC cells to ferroptosis, and precise ferroptosis regulation in RCC cells could potentially be achieved by regulating the expression of these key molecules. In summary, circRNAs and ferroptosis may play important roles in RCC occurrence and development. The aims of this study were to provide new ideas and methods for RCC treatment through the in-depth study of circRNA, renal cancer, and ferroptosis. We identified the novel circRNA, circASAP1, as highly expressed in renal cancer using both in-house sequencing data and a publicly available circRNA sequencing dataset. The presence and localization of circASAP1 in renal cancer cells were subsequently confirmed through Sanger sequencing, RNase R treatment, actinomycin D treatment, and FISH experiments. Moreover, gene silencing and overexpression experiments confirmed that circASAP1 promotes malignant biological behaviors, including cell proliferation and metastasis, in RCC cells. This study sheds new light on the mechanism underlying circASAP1 activity in RCC, which has not been previously investigated in this context. CircRNAs have various functions in both normal and cancer cells, including interactions with RBPs, acting as microRNA sponges, and coding for novel proteins. Interactions between circRNAs and RBPs are particularly unique and essential, as circRNA-RBP interactions can regulate downstream gene expression. Additionally, some RBPs can directly bind to pre-RNAs to assist in circRNA back-pruning and synthesis. To explore RBPs that could interact with circASAP1, we conducted MS analysis of RNA pull-down samples, as well as protein cross-predictions using various databases, including circAtlas, RBPmap, and RBPsuite. Further experiments confirmed the interaction between circASAP1 and HNRNPC, where HNRNPC is an RBP expressed in human cells that has been suggested to contribute to tumor growth regulation in various types of cancer, including RCC. The role of circPPAP2B in promoting ccRCC proliferation and metastasis through HNRNPC-dependent alternative splicing and the miR-182-5p/CYP1B1 axis was previously investigated in RCC [[175]42]. Further, the lncRNA, CYTOR, stabilized ZEB1 mRNA in oral squamous carcinoma by inhibiting non-ubiquitylated degradation of HNRNPC, which in turn regulates mitochondrial metabolism and glycolysis [[176]43]. Additionally, in hepatocellular carcinoma, HNRNPC induces hepatocarcinogenesis by regulating miR-21-5p synthesis [[177]44]. According to our experimental studies, circASAP1 appears to impact HNRNPC protein stability, leading to its degradation via the ubiquitin-proteasome pathway. Consequent changes in HNRNPC protein levels may influence GPX4 expression. The GPX4 enzyme is an important moderator of ROS levels; it regulates the supply of GSH, which provides a constant supply of GSH reductase activity, and has a detoxifying effect on hydroperoxides and lipid peroxides, thereby reducing oxidative damage to DNA, proteins, and lipids [[178]45, [179]46]. GPX4 is also reported to have inhibitory effects on the development of renal, gastric, and pancreatic cancers [[180]47, [181]48]. Based on our experimental data, circASAP1 may be a significant biological target with potential for application in RCC diagnosis, treatment, and prognosis prediction. While this study has made significant progress, it has some issues that require attention. Importantly, the upstream regulatory mechanism of circASAP1 remains unclear. Additionally, the impact of circASAP1 on HNRNPC protein degradation is yet to be clarified, and an investigation of whether it is associated with E3 ubiquitin ligase is warranted. Moreover, additional research is needed to fully elucidate the mechanism by which HNRNPC controls GPX4 mRNA stability and impacts its expression levels. Further investigation is necessary to gain a more thorough and precise understanding. Conclusions In this study, we identified nine shared circRNAs significantly highly expressed in two separate sequencing datasets from patients with RCC. After reviewing the relevant literature, we found that circASAP1 has been relatively understudied in RCC, which provides an opportunity to further use it as a research target. The presence and location of circASAP1 in renal cancer cells were verified through various experiments. We observed that circASAP1 expression of was higher in tumor samples from patients with RCC with recurrence and metastasis than in those without these disease features. Further, patients with high circASAP1 expression had worse overall and disease-free survival. These findings are based on objective evaluations and were not influenced by subjective opinions. Our results suggest that circASAP1 may have pro-cancer effects and could be linked to tumor metastasis, which is a significant factor contributing to poor prognosis of patients with RCC. Knockdown of circASAP1 in RCC cells significantly reduced RCC cell proliferation and cells with low circASAP1 expression also had reduced migration and invasion abilities. Further, we demonstrate that HNRNPC is a binding partner of circASAP1 in the cytoplasm, interacting with circASAP1 through amino acids 16–87. Further, circASAP1 knockdown influenced protein, but not RNA levels, of HNRNPC through the ubiquitin-proteasome pathway. Protein-coding genes altered in response to changes in circASAP1 expression, including GPX4, were enriched in the ferroptosis pathway. Our data demonstrate that HNRNPC directly regulates GPX4 by affecting the GPX4 mRNA stability, and consequently its protein levels, while reduced circASAP1 expression led to a ferroptosis phenotype in RCC cells and reduced subcutaneous tumor growth and lung metastasis in nude mice. Overall, our data suggest that low circASAP1 expression may inhibit RCC proliferation and metastasis, while inducing ferroptosis through regulating HNRNPC ubiquitination levels. These findings offer a new perspective for future research into renal cancer treatment. Supplementary Information [182]Supplementary Material 1.^ (3.9MB, docx) [183]Supplementary Material 2.^ (420.7KB, zip) [184]Supplementary Material 3.^ (4MB, docx) Acknowledgements