CircPPP1CB subtype, hsa_circ_0007439, promotes nasopharyngeal carcinoma progression by upregulating KRT1

Abstract

Background

Nasopharyngeal carcinoma (NPC) is characterized by pronounced metastatic and invasive properties. Emerging research has elucidated that circular RNAs (circRNAs) are intricately associated with the pathogenesis of NPC, potentially serving as critical mechanisms in tumorigenesis and as promising therapeutic targets for NPC.

Methods

The expression levels of circRNAs were analyzed in NPC using reverse transcription quantitative PCR (RT-qPCR). Wound healing, transwell, and clone formation assays were conducted to examine the metastatic potential of NPC. Cell proliferation and apoptosis assays were performed to evaluate cell proliferation and survival. 3D spheroid cultures, RNA pull-down, LC-MS, RNA-sequencing, and luciferase reporter assays were carried out to examine the mechanisms of circPPP1CB in NPC progression. NPC xenograft tumor models were estimated to verify the mechanisms of circPPP1CB in tumor growth and metastasis of NPC in vivo.

Results

CircPPP1CB subtype, hsa_circ_0007439, that was more highly expressed in NPC patients with distant metastasis than in those without distant metastasis. Hsa_circ_0007439 overexpression specifically promotes proliferation and inhibits the apoptosis of NPC cells. Further experiments indicated that hsa_circ_0007439 overexpression was correlated with increased tumor growth and distant metastasis of NPC cells. Mechanistically, hsa_circ_0007439 upregulated KRT1 expression by acting as a sponge for miR-1275 and miR-1293. This led to increased proliferation and metastatic progression in NPC by activating the JAK/STATs signaling pathway.

Conclusions

This study is the first to demonstrate that hsa_circ_0007439 functions as a competitive endogenous RNA to upregulate KRT1 expression, thereby promoting the metastasis of NPC, suggesting that hsa_circ_0007439 serve as a potential diagnostic biomarker and therapeutic target for NPC.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12672-025-03888-z.

Introduction

Nasopharyngeal carcinoma (NPC) is a common malignant tumor of the head and neck characterized by Epstein-Barr virus (EBV) infection, poor differentiation of tumour cells, and high metastatic and invasive nature [1, 2]. Patients with NPC often present with advanced diseases at first diagnosis, as NPC lack specificity during the initial stage [3, 4]. Despite advancements in radiotherapy and chemotherapy leading to improved local-regional control, metastasis continues to be the primary cause of treatment failure and cancer-related mortality in NPC [5, 6]. Distant metastasis and recurrence account for the main cause of treatment failure in NPC and remain as the significant obstacle to achieving higher cure rates in clinical practice. Recurrent NPC caused by distant metastasis has even become the most severe challenge in NPC treatment [7]. Hence, finding reliable biomarkers and understanding of the molecular mechanisms underlying metastasis are crucial for developing more targeted clinical strategies to treat metastatic or recurrent NPC.

Circular RNAs (circRNAs) are a kind of single-stranded, closed endogenous RNAs generated by reverse splicing of precursor RNA following transcription and exhibiting notable stability in mammals [8, 9]. In contrast to linear RNAs, circRNAs demonstrate significant conservation and abundance due to their closed circular structure, which confers resistance to RNA exoenzymes. This distinctive attribute suggests that circRNAs have obvious potential as biomarkers of various clinical diseases [10, 11]. Extensive research on circRNAs has elucidated their multifaceted roles, including functioning as miRNA or RNA-binding protein sponges, protein localization baits, and templates for peptide translation, etc [12, 13]. Advances in high-throughput RNA sequencing have led to identification of circRNAs involved in tumorigenesis, progression, invasion, and metastasis [14]. For instance, circular RNA hsa_circ_0003823 functions as a miR-607 sponge to facilitate tumor progression, metastasis, and apatinib resistance in esophageal squamous cell carcinoma through the upregulation of CRISP3¹⁵. circPOFUT1 directly sponges miR-488-3p to upregulate PLAG1 and ATG12 in gastric cancer, thereby enhancing malignant phenotypes and autophagy-associated chemoresistance [16]. CircRNF13 enhances SUMO2 levels, leading to GLUT1 degradation and regulation of the AMPK-mTOR pathway, which inhibits glycolysis and promotes tumor proliferation and metastasis [1, 17]. In addition, extracellular vesicles contain mature and stable transcriptomes enriched in circRNAs, making circRNAs an excellent candidate to be used as tumor biomarkers [18]. Notably, studies have indicated that more than 80% of circRNAs originate from exonic regions and exhibit identical sequences to their linear RNA counterparts. These exonic circRNAs tend to localize and function predominantly in the cytoplasm, where they can influence translational processes though competitive endogenous RNA (ceRNA) mechanism or interact directly with proteins to modulate protein localization and function [12, 19, 20]. For example, circular RNA circFOXO3 functions as a ceRNA for acid-sensing ion channel subunit 1 mediating oxeiptosis [21]. Circ-Ccnb1 could interact with both Ccnb1 and Cdk1 proteins and dissociate the Ccnb1/Cdk1 complex, suppressing cell invasion and tumorigenesis [22].

More importantly, although a large number of circular RNAs have been identified with the advancement of RNA sequencing, the biological functions of most circular RNAs have not yet been elucidated, especially in NPC. Recently studies have demonstrated the involvement of circRNAs, such as circRILPL1 and circCOL1A1, in the processes of proliferation, metastasis, and metabolic alterations in NPC by acting as ceRNA [23, 24], suggesting a potential significance of circRNAs in the progression of NPC. However, further comprehensive research is required to pinpoint specific circRNAs that could serve as biomarkers for the detection and monitoring of tumor progression and metastasis, as well as for guiding therapeutic strategies in patients with NPC.

Protein phosphatase 1 catalytic subunit β isoform (PPP1CB) is a catalytic subunit of protein phosphatase-1 and exhibits elevated expression levels in liver cancer, pancreatic cancer and melanoma. Numerous studies have demonstrated the functions of PPP1CB participating in regulating cell division, migration, and skeleton remodeling by catalyzing serine or threonine dephosphorylation [25–27]. circPPP1CB is a set of circRNAs formed by reverse splicing of different exons and introns in PPP1CB precursor RNA. While 16 circPPP1CB subtypes have been identified, only the pathological role of hsa_circ_0119704 subtype in the growth, metastasis, and epithelial-mesenchymal transition (EMT) of bladder cancer by regulating the miR-1307-3p/SMG1 axis has been clarified [28]. The biological role of the remaining circPPP1CB subtypes in tumorigenesis remain unclear.

In this study, we constructed a correlation network analysis of 16 circPPP1CB subtypes included in the circBase and found that 5 of them were associated with multiple tumors. Subsequent validation through RT-qPCR demonstrated that only hsa_circ_0007439 subtype was highly expressed in both highly metastatic NPC cells and NPC tissues with distant metastasis. We further revealed the molecular mechanism by which hsa_circ_0007439 sponges miR-1275 and miR-1293 to upregulate KRT1, ultimately leading to enhanced proliferation and metastasis while suppressing apoptosis in NPC. These findings not only improve our understanding of circRNAs in NPC progression but also uncover prognostic biomarker and promising therapy target for NPC.

Materials and methods

Ethics approval

Ethical approval for this study was obtained from the Ethical Review Board of Sun Yat-sen University Cancer Center (approval number: B2025-340-01, 2025), ensuring adherence to the ethical standards for research involving human participants as outlined in the Declaration of Helsinki. All patients provided written and oral informed consent. The clinicopathological characteristics of the patients are summarized in Table S1. All experiments involving animals were conducted according to the ethical policies and procedures approved by the Institutional Animal Care and Use Committees of the Sun Yat-sen University (approval number: SYSU-LS-IACUC-2023-0096). All animal experiments were strictly implemented in compliance with the ARRIVE guidelines.

Cell lines and human NPC samples

Human EBV-negative NPC cells (6-10B, 5–8 F, S26, and S18) were kindly provided by the Sun Yat-sen University Cancer Center and cultured in RPMI medium supplemented with 10% fetal bovine serum (FBS, Gibco, USA). HEK293T cells were procured from ATCC and were cultured in DMEM medium supplemented with 10% FBS. Cell authentication was performed using short tandem repeat profiling, and all cells were cultured at 37℃ in a 5% CO₂ environment. Tumor tissues were gathered from NPC patients for subsequent reverse transcription quantitative PCR (RT-qPCR) analysis of circRNA expression using standard procedures.

Plasmid construction, RNAi and cell transfection

The overexpression vectors for hsa_circ_0007439 were constructed by inserting the full-length hsa_circ_0007439 sequence into the pCD5-ciR, pLO5-ciR, or pLC5-ciR-copGFP vectors, each of which contained a front and back circular frame. A mock vector devoid of the circRNA sequence was used as a control. To knockdown hsa_circ_0007439, three siRNAs and antisense oligonucleotides (ASOs) targeting back splice junction of hsa_circ_0007439 were synthesized (Ruibo Biotech, Guangzhou, China). To construct KRT1 overexpression vector, the CDS sequence of human KRT1 was inserted into the pCDH-CMV-MCS-EF1-copGFP vector. The knockout vectors for KRT1 were constructed by inserting guide RNA (gRNA) targeting KRT1 into the lentiCRISPR v2 vectors. Hilymax (Dojindo, Japan) was used for plasmids transfection. The siRNA/ASO transfection was conducted using riboFECT™ CP reagent (RiboBio). All overexpression stable cell lines were generated by transfecting NPC cells with hsa_circ_0007439- or KRT1-expressing lentivirus generated with recombinant lentiviral expression vectors with psPAX2 and pMD2.G in HEK293T cells. KRT1 knockout NPC cells were generated using the CRISPR/Cas9 system, as described previously [29]. To confirm overexpression and target knockdown/knockout, cells were collected for RT-qPCR or Western blot analysis. The sequences of primers, siRNAs, ASOs, and sgRNAs in this study were listed in Table S2 and Table S3.

RT-qPCR and RT-PCR

Total RNA was extracted from NPC cells and NPC tissues using HP Total RNA Kit (Omega, USA). RNA was transcribed into cDNA using the PrimeScript RT reagent kit (Takara) for RT-qPCR with a SYBR Green qPCR Master Mix (Bimake, China). Relative expression levels of genes were determined using the ΔΔCT method. The primer sequences for RT-qPCR analysis were provided in Table S4. In the RT-PCR assay, NPC cells were subjected to extraction of total RNA and genomic DNA using the HP Total RNA Kit and genomic DNA extraction kit (Omega), respectively. The presence of hsa_circ_0007439 in cDNA or genomic DNA was confirmed using convergent and divergent primers in conjunction with KOD-Plus-NEO polymerase (TOYOBO, Japan).

RNase R and actinomycin D treatment and nuclear-cytoplasmic fractionation

For RNase R digestion, total RNA (5 µg) was incubated with or without 20 U/µl RNase R at 37 °C for 30 min. RNA was reverse transcribed into cDNA. RT-qPCR was performed to quantify hsa_circ_0007439 and linear PPP1CB as described above. For the actinomycin D (ActD) assay, NPC cells were treated with ActD (2 µg/mL, Selleck, USA) for 24 h to block transcription. Total RNA was extracted and used for RT-qPCR analysis as described above. For nuclear-cytoplasmic fractionation, RNAs from nucleus and cytoplasm of NPC cells were separated by the Nuclear and Cytoplasmic RNA Purification Kit (Beyotime, China) following the manufacturer’s instructions.

Fluorescence in situ hybridization (FISH)

FISH analysis was performed using the Fluorescent in Situ Hybridization Kit (RiboBi, China) as previously described [30]. Cy3-labeled hsa_circ_0007439 probes targeting the back splice junction were synthesized by RiboBio (RiboBi, China). NPC cells were washed with PBS and fixed in 4% formaldehyde, followed by permeabilization 0.5% Triton X-100. Subsequently, cells were blocked in pre-hybridization buffer at 37℃ for 30 min. Hybridization was conducted overnight at 37℃ using the hsa_circ_0007439 FISH Probe Mix under light-free conditions. Following washes with 4×, 2×, and 1×SSC buffer, cells were visualized using an LSM880 confocal microscope (Zeiss, Germany).

Cell proliferation assays

For 5-Ethynyl-20-deoxyuridine (EdU) cell proliferation assay, the EdU staining was performed according to the manufacturer’s protocol of the EdU Proliferation Kit (RiboBio, China). Briefly, NPC control cells, hsa_circ_0007439-overexpressed cells, and hsa_circ_0007439 knockdown NPC cells were seeded onto glass coverslips and incubated at 37℃ for 48 h. Following this, cells were treated with EdU for 2 h, fixed with 4% paraformaldehyde for 30 min, permeabilized with 0.5% Triton X-100 for 10 min, and incubated with Apollo reaction buffer for 30 min at room temperature in the absence of light. Subsequently, cells were stained with Hoechst 33342 for 10 min and imaged using an LSM 880 confocal microscope (Zeiss). For cell counting kit-8 (CCK-8) assay, growth factor-starved NPC cells were seeded in 96-well plates at a concentration of 5000 cells/well and cultured at 37℃ for 48 h. Subsequently, the cells were incubated with CCK-8 (Dojindo, Japan) for 3 h at 37℃, and the OD values at 450 nm were measured to determine the relative cell proliferation.

Cell apoptosis assays

For flow cytometric analysis of apoptosis, a total of 1 × 10⁶ NPC cells were labeled with Annexin V-633 and PI (Dojindo) or Annexin V-FITC and 7-AAD (Biolegend, USA) for 15 min in the dark and analyzed using an CytoFLEX flow cytometer (Beckman, Germany). For detection of caspase 3 activity, a total of 1 × 10⁶ NPC cells were lysed and 50 µl of supernatants with a total protein concentration of 2 mg/mL were transferred into a 96-well plate following the manufacturer’s instructions (Dojindo). Subsequently, a detection substrate and analysis buffer were added, and the plate was incubated at 37℃ for 6 h. The OD values at 405 nm were then recorded to determine the level of caspase 3 activity.

Wound healing, cell migration, invasion, and clone formation assays

Western blot, wound healing, transwell cell migration, invasion, and clone formation assays were performed as previously reported [29, 31].

Immunofluorescence assay

Hsa_circ_0007439 overexpressed NPC cells were grown on µ-Dish (Ibidi, Germany) and transfected with miR-NC, hsa-miR-1275, or hsa-miR-1293 mimics for 48 h. Following fixation in 4% paraformaldehyde and permeabilization in 0.1% Triton X-100, cells were exposed to primary antibodies overnight. Subsequently, fluorescence-labelled secondary antibodies were used for protein labeling for 1 h. Cell nuclei were counterstained with Hoechst 33342 and visualized using confocal microscope (Zeiss).

Liquid chromatography-mass spectrometry (LC-MS) analysis

The analysis of differential protein abundances via LC-MS was conducted as previously described [32]. Briefly, total protein samples were prepared using standard procedures. The peptide powder was dissolved in 0.1% FA and subsequently analyzed with a Q-Exactive HF mass spectrometer (Thermo Scientific). The mobile-phase buffer comprised two components: solution A (99.9% H2O and 0.1% FA) and solution B (100% methanol). The mass spectrometer operated within a scan range of 100–900 m/z. Mass spectrometry data were processed using Proteome Discoverer 2.4 referred to the UniProt Human database. The list of identified proteins was summarized in Table S5.

RNA sequencing

NPC cells were transfected with either an empty vector or a vector overexpressing hsa_circ_0007439 for 48 h. Total RNA was subsequently extracted utilizing the TRIzol reagent (Invitrogen, USA). RNA sequencing libraries were then prepared, sequenced, and analyzed using the Illumina HiSeq 2500 platform, as detailed in our previous publication, by the Lianchuan Biological Company (Hangzhou, China) [29, 31]. The list of identified DEGs was summarized in Table S6.

3D spheroid cultures

The 3D spheroid cultures were performed as described previously [29]. Briefly, NPC cells overexpressing hsa_circ_0007439 and control NPC cells were resuspended in RPMI with 30% FBS and mixed with VitroGel Hydrogel Matrix (TheWell, Bioscence, USA). The mixtures were placed in a 24-well plate and covered with medium containing 10% FBS. After a 10-day incubation, spheroids were extracted using VitroGel Cell Recovery Solution (TheWell, Bioscience) for immunofluorescence analysis.

Ago2-RNA Immunoprecipitation

NPC cells were transfected with hsa_circ_0007439 overexpression vectors for 48 h, followed by lysis using RIP lysis buffer at 4℃ for 30 min. The lysates were centrifuged at 14,000×g for 10 min at 4℃ and incubated overnight with magnetic beads coated with either Ago2 or IgG antibodies at 4℃ for magnetic labeling and magnetic separation. Subsequently, the coprecipitated Ago2 was then detected via Western blot, while the coprecipitated circRNAs were extracted using TRIzol reagent (Invitrogen) and quantified by RT-qPCR.

RNA pull-down

NPC cells were transfected with hsa_circ_0007439 overexpression vectors for 48 h, lysed with RIP buffer at 4℃ for 30 min, and centrifuged at 14,000×g for 10 min at 4 ℃. The streptavidin magnetic beads were incubated with biotin-labeled hsa_circ_0007439 or negative control probes for 1 h. Subsequently, lysates were incubated overnight with the corresponding biotin-labeled probes for magnetic labeling and magnetic separation. The precipitated hsa_circ_0007439 was then extracted using TRIzol and quantified by RT-qPCR. Concurrently, the coprecipitated miRNAs were sequenced and analyzed by Lianchuan Biological Company (Hangzhou, China), and non-specific miRNAs were excluded using negative probes according to standard protocols. The sequences of hsa_circ_0007439 and negative control probes were provided in Table S7.

Luciferase reporter assay

The luciferase reporter assay was conducted as previously described [33]. Briefly, the sequences of hsa_circ_0007439 and KRT1 3’UTR were cloned into the pmir-GLO vector. miR-NC, hsa-miR-1275, and hsa-miR-1293 mimics were synthesized by TSINGKE (Guangzhou, China). HEK293T cells were co-transfected with the respective pmir-GLO recombinant plasmids and miR-NC, hsa-miR-1275, or hsa-miR-1293 mimics for 24 h. Luciferase assays were conducted using the Dual-Specific Luciferase Assay Kit (Promega, USA). The firefly luciferase activity was normalized to the Renilla luciferase activity.

Animal experiments

Eighteen five-week-old BALB/c nude mice were maintained in the Laboratory Animal Center of Sun Yat-sen University and evenly divided into two groups. The mice were subcutaneously inoculated with NPC cells (1 × 10⁶ cells in 100 ul) that were stably infected with hsa_circ_0007439 overexpression or mock lentiviruses, with or without GFP labeling. Mice were monitored for tumor growth every other day and euthanized with CO₂. Tumor volume and weight were recorded. The lungs, liver, spleen, kidneys, and heart were excised and imaged using an in vivo imaging system (Vieworks, South Korea). Tumor tissues were embedded in paraffin and sectioned for immunohistochemistry and immunofluorescence using standard procedures. The lungs and liver were processed for hematoxylin and eosin (H&E) staining, and metastatic nodules were quantified microscopically. The maximal tumor size permitted by the Institutional Animal Care and Use Committees of Sun Yat-sen University was 1500 mm³, and this study did not exceed that limit.

Statistical analysis

All results are presented as the mean ± SD derived from a minimum of three independent experiments. For comparisons between two groups, Student’s t-test was used. In cases involving comparisons among more than two groups, data were analyzed using one-way analysis of variance (ANOVA). Statistical significance was established at *P < 0.05, **P < 0.01, and ***P < 0.001. All statistical analyses were conducted using GraphPad Prism 7 software.

Results

Hsa_circ_0007439 exhibits elevated expression levels in highly metastatic NPC

Numerous studies have demonstrated that PPP1CB is implicated in various tumorigenesis processes. However, there is a paucity of research concerning the pathological role of its derived circRNA in tumor progression. To identify circPPP1CB subtypes related to NPC metastasis, we conducted a correlation network analysis, revealing that subtypes hsa_circ_0007439, hsa_circ_0053325, hsa_circ_0053330, hsa_circ_0053329, and hsa_circ_0003905 are expressed in various tumor cells (Figure S1). Notably, RT-qPCR validation showed that only hsa_circ_0007439 exhibited a markedly elevated expression in highly metastatic NPC cells 5–8 F and S18 (Fig. 1A and B). Meanwhile, we observed that hsa_circ_0007439 was significantly upregulated in tumor tissues of NPC patients with distant metastases compared to those without distant metastases (Fig. 1C). This finding raises the question of whether the differential expression of hsa_circ_0007439 play a non-redundant regulatory role in the metastasis of NPC.

Fig. 1.

Hsa_circ_0007439 is validated and characterized in NPC cells. A, B Relative expression of circPPP1CB subtypes in NPC cells by RT-qPCR. C Relative expression of hsa_circ_0007439 in NPC tissues with distant metastases (n = 8) and without distant metastases (n = 9) by RT-qPCR. D Schematic illustration of hsa_circ_0007439 formation. E PCR analysis was carried out using divergent (red) and convergent (blue) primers on cDNA and gDNA taken from NPC cells. F Sanger sequencing was used to validate the back-splice junction sequences of hsa_circ_0007439. G The cDNA of hsa_circ_0007439 and linear PPP1CB was reverse transcribed with random 6 primers or oligo (dT) primers and then measured by RT-qPCR. H The abundances of hsa_circ_0007439 and linear PPP1CB were determined using RT-qPCR on NPC cells treated with RNase R. I The abundances of hsa_circ_0007439 and linear PPP1CB were determined using RT-qPCR on NPC cells treated with ActD. J, K Identification of hsa_circ_0007439 cytoplasmic distribution by RT-qPCR analysis and FISH (Scale bar, 10 μm). GAPDH was considered as a cytoplasmic control and U6 was used as a nuclear control

Hsa_circ_0007439 is a 224-nt circRNA generated from exon 5 and exon 6 of PPP1CB gene by back-splicing. To confirm that hsa_circ_0007439 is generated through head-to-tail splicing, specific divergent (red) and convergent (blue) primers were designed to amplify the circular transcripts of hsa_circ_0007439 and the linear transcripts of PPP1CB (Fig. 1D). PCR analysis revealed the exclusive presence of hsa_circ_0007439 in cDNA, thereby excluding its presence in gDNA. In contrast, the convergent primers successfully amplified PPP1CB from both cDNA and gDNA (Fig. 1E). The backspliced junction points of hsa_circ_0007439 were further amplified using outward-facing primers and subsequently validated through Sanger sequencing (Fig. 1F). Following reverse transcription with Oligo (dT) and Random 6 mers primers, RT-qPCR analysis revealed no significant difference in the reverse transcription efficiency for linear PPP1CB across the two primer conditions. Conversely, hsa_circ_0007439 underwent successful reverse transcription exclusively in the presence of the Random 6 mers primer, thereby corroborating the head-to-tail splicing structure of hsa_circ_0007439 (Fig. 1G). Furthermore, our findings indicate that hsa_circ_0007439 exhibits enhanced resistance to RNase R digestion relative to linear PPP1CB transcripts (Fig. 1H). Additionally, upon inhibition of RNA synthesis using ActD, hsa_circ_0007439 exhibited a slower degradation rate (Fig. 1I, and Figure S2), indicating that hsa_circ_0007439 is indeed a circRNA characterized by a stable circular structure. Nucleocytoplasmic separation experiments with RT-qPCR analysis and FISH revealed that hsa_circ_0007439 is significantly localized in both the cytoplasm and nucleus (Fig. 1J and K). Together, these findings identify hsa_circ_0007439 as a circRNA with elevated expression levels in metastatic NPC tissues, suggesting its potential role in the tumor metastasis process.

Hsa_circ_0007439 enhances the proliferation and inhibits apoptosis of NPC cells

To elucidate the potential biological functions of hsa_circ_0007439 in NPC, we generated overexpression vectors for hsa_circ_0007439 and developed siRNA/ASO mixtures targeting the junction sites of hsa_circ_0007439, respectively (Figure S3A and S3B). RT-qPCR analysis demonstrated that the expression of hsa_circ_0007439 was significantly upregulated or downregulated in NPC cells following transfection with the respective overexpression vectors or siRNA/ASO mixtures. However, our findings indicated that neither overexpression nor knockdown of hsa_circ_0007439 affected the expression levels of the linear transcript PPP1CB and other circPPP1CB subtypes (Fig. 2A and B, and Figure S3C, S3D). Next, we investigated the effects of hsa_circ_0007439 on NPC cell proliferation and apoptosis. The EdU assay results indicated that overexpression of hsa_circ_0007439 enhanced the proliferation ability of poorly metastatic NPC cells, whereas hsa_circ_0007439 knockdown resulted in a reduction in cell proliferation in highly metastatic NPC cells (Fig. 2C and D). These findings were further corroborated by the CCK-8 assay (Fig. 2E). Nevertheless, Annexin V/PI double staining showed that overexpression of hsa_circ_0007439 significantly reduced apoptosis in NPC cells, irrespective of ActD treatment, whereas knockdown of hsa_circ_0007439 exhibited the contrary effect (Fig. 2F and G). In addition, while ActD treatment resulted in an upregulation of caspase-3 expression in NPC cells, the overexpression of hsa_circ_0007439 attenuated this effect, whereas the knockdown of hsa_circ_0007439 further enhanced caspase-3 expression (Fig. 2H). These data suggest that hsa_circ_0007439 exerts an oncogenic role in the NPC cells.

Fig. 2.

Hsa_circ_0007439 promotes the proliferation and suppresses the apoptosis of NPC cells. A Relative expression of hsa_circ_0007439 was measured by RT-qPCR in NPC cells transfected with hsa_circ_0007439 overexpression plasmids. B Relative expression of hsa_circ_0007439 was measured by RT-qPCR in NPC cells transfected with siRNA/ASO. C, D Cell proliferation was evaluated by EdU assay on NPC cells (Scale bar, 50 μm). E CCK-8 assay was used to measure the growth curves of NPC cells. F, G Apoptosis rate was tested using flow cytometry following the double staining of NPC cells with Annexin V and PI. H Relative expression of Caspase3 was measured by a microplate reader in NPC cells treated with or without AtcD

Hsa_circ_0007439 promotes the metastatic potential of NPC cells

Given the elevated levels of hsa_circ_0007439 in highly metastatic NPC cells, we conducted further investigations to determine its potential association with NPC metastasis. Wound healing assays suggested that overexpression of hsa_circ_0007439 significantly enhanced the migratory capacity of poorly metastatic NPC cells. Conversely, knockdown of hsa_circ_0007439 led to a marked inhibition of cell migration in highly metastatic NPC cells (Fig. 3A and B). Transwell cell migration and invasion assays also demonstrated that the overexpression of hsa_circ_0007439 substantially augmented the migratory and invasive capabilities of NPC cells, whereas the knockdown of hsa_circ_0007439 resulted in a pronounced reduction in cell migration and invasion (Fig. 3C-E). Similarly, we found that overexpression of hsa_circ_0007439 promoted the clonogenic abilities of NPC cells by clone formation assays, while knockdown of hsa_circ_0007439 revealed an opposite effect (Fig. 3F and G). Taken together, these data show that hsa_circ_0007439 is functionally relevant to NPC metastasis.

Fig. 3.

Hsa_circ_0007439 promotes migration, invasion, colony formation, and EMT process of NPC cells. A, B Migration abilities of NPC cells were measured by wound healing assays. C-E Migration and invasion abilities of NPC cells were assessed using a Transwell invasion assay (Scale bar, 183.37 μm). F, G Clone formation capacity of NPC cells was measured by clone formation assays

KRT1 is the main downstream gene upregulated by hsa_circ_0007439

To investigate the underlying molecular mechanisms of hsa_circ_0007439, comparative proteome and transcriptome analyses were initially conducted to identify potential downstream genes regulated by hsa_circ_0007439. Differential gene expression analysis revealed a significant overlap in the expression of genes involved in tumor metastasis, proliferation, and apoptosis regulation, with alterations observed at both the protein and transcript levels following hsa_circ_0007439 overexpression (Fig. 4A-C). The expression levels of genes associated with tumor metastasis promotion (e.g., VIM and CDH2), proliferation promotion (e.g., EGFR and FLT4), and apoptosis suppression (e.g., BCL2) were significantly elevated in NPC cells overexpressing hsa_circ_0007439. Conversely, the expression levels of genes involved in tumor metastasis suppression (e.g., CDH1), proliferation suppression (e.g., CTNNB1), and apoptosis promotion (e.g., CASP3 and BID) were markedly reduced in NPC cells overexpressing hsa_circ_0007439 compared to those transfected with an empty vector (Fig. 4A-C). Notably, a total of 25 genes showed upregulation at both the protein and transcript levels in NPC cells overexpressing hsa_circ_0007439. Among these, KRT1 expression demonstrated the most pronounced upregulation (Fig. 4D). Subsequently, the top 5 upregulated candidates, KRT1, GSTZ1, MAP1S, PRSS1, and MPST, were validated by RT-qPCR. The results further corroborated that KRT1 was the most distinctly upregulated gene in hsa_circ_0007439-overexpressing NPC cells (Fig. 4E). Using 3D spheroid cultures coupled with immunofluorescence analysis, we observed that the overexpression of hsa_circ_0007439 resulted in the upregulation of KRT1, which was concomitant with an increase in spheroid dimensions (Fig. 4F). In addition, an in-depth analysis of the GEPIA2 databases [34] indicated a significant elevation in KRT1 expression in head and neck squamous cell carcinoma (HNSCC) relative to normal tissue (Figure S4). Analysis of the RNA-Seq dataset GSE68799 also identified an upregulation of KRT1 in tumor tissues of NPC patients compared to adjacent non-cancerous tissues (Fig. 4G and H). Altogether, these findings collectively indicate that KRT1 is the main downstream gene upregulated by hsa_circ_0007439 and therefore, have a novel metastasis-promoting effect in NPC.

Fig. 4.

KRT1 is the main downstream gene upregulated by hsa_circ_0007439. A Volcanic map of differentially expressed genes (DEGs) between hsa_circ_0007439-overexpressed and empty vector-transfected NPC cells by LC-MS/MS. B Heatmap of DEGs correlated with tumor metastasis and proliferation in hsa_circ_0007439-overexpressed NPC cells by RNA-Seq. C Heatmap of grouping clustering in DEGs correlated with EMT, proliferation, and apoptosis in hsa_circ_0007439-overexpressed NPC cells at transcript and protein levels. D Expression of potential genes regulated by hsa_circ_0007439 in NPC cells. The numbers indicating the quantity of genes in each subset of DEGs. E Bar chart showing the top 5 upregulated genes by hsa_circ_0007439 in NPC cells at both the transcript and protein levels. F Representative immunofluorescence of KRT1 in 6-10B spheroids released from 3D spheroid culture (Scale bar, 100 pixel). G, H KRT1 expression in tumor tissues of NPC patients and adjacent non-cancerous tissues from RNA-Seq dataset GSE68799

Hsa_circ_0007439 upregulates KRT1 expression by sponging miR-1275 and miR-1293

We then investigated the mechanism through which hsa_circ_0007439 upregulates KRT1 expression. Given our previous findings that hsa_circ_0007439 can localize within the cytoplasm, we hypothesized that hsa_circ_0007439 might be involved in translational regulation by functioning as competing endogenous RNA. To investigate whether hsa_circ_0007439 acts as a miRNA sponge, we performed argonaute 2 (Ago2) crosslinking immunoprecipitation in NPC cell lysates (Fig. 5A). Our results demonstrated a significant enrichment of hsa_circ_0007439 in Ago2-containing immunoprecipitates relative to the isotype control (Fig. 5B), thereby confirming that hsa_circ_0007439 participate in the translational regulation of downstream genes by acting as a competing endogenous RNA. Therefore, we proceeded to identify the miRNAs that interact with hsa_circ_0007439. To this end, we employed RNA pull-down assays using biotinylated hsa_circ_0007439. This approach led to the identification of 95 miRNAs that specifically interact with hsa_circ_0007439 (Fig. 5C and D). Subsequently, the TargetScan database was employed to refine the list of miRNAs with potential interactions with hsa_circ_0007439. From this intersection, 6 miRNAs were identified, and the top 3 were selected as candidates for further validation (Fig. 5E and F). Additionally, conserved sites within the KRT1 3’UTR and hsa_circ_0007439, which may function as miRNA response elements for miR-1293, miR-582-3p, and miR-1275, were predicted using TargetScan, respectively (Figure S5). To validate this prediction, we individually inserted the KRT1 3’UTR and hsa_circ_0007439, which contain the binding sites for the three miRNAs, into the pmir-GLO luciferase reporter vector (Fig. 5G). Luciferase assays confirmed that KRT1 binds to miR-1293, miR-582-3p, and miR-1275 sequence, significantly reducing luciferase activity. Meanwhile, hsa_circ_0007439 was found to interact with miR-1275 and miR-1293, but not with miR-582-3p, also resulting in reduced luciferase activity (Fig. 5H), suggesting that hsa_circ_0007439 competitively binds with miR-1275 and miR-1293, thereby upregulating KRT1 expression.

Fig. 5.

Hsa_circ_0007439 upregulates KRT1 by sponging miR-1275 and miR-1293. A Schematic diagram of Ago2-RIP assay. B Ago2-RIP experiment showed the combination of Ago2 protein with hsa_circ_0007439 in NPC cells. RT-qPCR assessing the enrichment of hsa_circ_0007439 in anti-Ago2 group and anti-IgG group.(C RNA pull-down assay analysis of hsa_circ_0007439-binding miRNAs. RT-qPCR was used to assess the enrichment of hsa_circ_0007439 using a hsa_circ_0007439 probe. D Circular barplot illustrating the miRNAs enriched by hsa_circ_0007439. E, F Venn diagram and bar chart displaying the overlap of hsa_circ_0007439 target miRNAs predicted by Targetscan and identified through RNA pull-down assay. G Schematic diagram of dual-luciferase assay. H Dual-luciferase assays showing the luciferase activity of the pmir-GLO-hsa_circ_0007439 and pmir-GLO-KRT1 3’UTR vector in HEK293T cells cotransfected with target miRNA mimics. I Immunofluorescence assay for KRT1 expression in hsa_circ_0007439 overexpression combined with miR-1275 mimic or miR-1293 mimic transfected NPC cells

To verify this speculation, we transfected NPC cells with hsa_circ_0007439, followed by the addition of miR-1275 or miR-1293 mimics. Subsequent immunofluorescence staining revealed a significant decrease in KRT1 expression post-transfection with either miR-1275 or miR-1293 mimics. However, overexpression of hsa_circ_0007439 mitigated the inhibitory effects of miR-1275 and miR-1293 on KRT1 expression (Fig. 5I). In conclusion, the findings substantiate that hsa_circ_0007439 can act as a competing endogenous RNA, thereby upregulating KRT1 expression by sponging miR-1275 and miR-1293.

Hsa_circ_0007439 promotes NPC progression by upregulating KRT1

To elucidate the relationship between the upregulation of KRT1 by hsa_circ_0007439 and the proliferation, apoptosis, and metastasis of NPC cells, we generated KRT1 overexpression and knockout NPC cell lines. Unsurprisingly, KRT1 overexpression resulted in increased proliferation and decreased apoptosis in NPC cells, whereas KRT1 knockout produced the opposite effects. Furthermore, transfection with miR-1275 or miR-1293 mimics mitigated the proliferation-promoting and apoptosis-suppressing effects of KRT1 in NPC cells (Fig. 6A-C, and Figure S6A, S6B). Notably, while our previous experiments have shown that the upregulation of hsa_circ_0007439 promotes NPC cell proliferation and inhibits apoptosis, the knockout of KRT1 attenuated the proliferation-enhancing and apoptosis-inhibiting effects of hsa_circ_0007439 on NPC cells (Fig. 6A-C, and Figure S6A, S6B).

Fig. 6.

Hsa_circ_0007439 promotes NPC progression by upregulating KRT1. A Cell proliferation of NPC cells was measured by CCK-8 assay. B, C Apoptosis rate was tested using flow cytometry following the double staining of NPC cells with Annexin V and 7-AAD. D-F Wound healing assay (D), transwell invasion assay (E), and clone formation assay (F) were used to assess the migration abilities (Scale bar, 200 μm), invasion abilities (Scale bar, 200 μm), and clone formation abilities of KRT1 overexpressed or knockout, hsa_circ_0007439 overexpression combined with KRT1 knockout, and KRT1 overexpression combined with miR-1275 mimic or miR-1293 mimic transfected NPC cells

In addition, although KRT1 overexpression led to enhanced migration, invasion, and colony formation in NPC cells, transfection with miR-1275 and miR-1293 mimics attenuated these promoting effects of KRT1. Moreover, the knockout of KRT1 in NPC cells abolished the facilitatory impact of hsa_circ_0007439 on migration, invasion, and colony formation (Fig. 6D-F). These results indicate that hsa_circ_0007439 promotes NPC proliferation and metastasis while inhibiting apoptosis through the upregulation of KRT1 via the hsa_circ_0007439/miR-1275/miR-1293/KRT1 axis.

KRT1 promotes NPC progression by activating the JAK/STATs pathway

To elucidate the specific mechanism through which hsa_circ_0007439-upregulated KRT1 facilitates the progression of NPC, we conducted signaling pathway enrichment analysis to investigate the potential signaling pathways. Our analysis revealed a significant enrichment of the JAK/STATs signaling pathway in NPC cells overexpressing hsa_circ_0007439, indicating that KRT1 upregulation, driven by hsa_circ_0007439, may promote NPC progression via the JAK/STATs pathway (Fig. 7A). We further validated the activation of the JAK/STATs pathway in NPC cells with elevated KRT1 expression, as opposed to other pathways such as PI3K/AKT and MAPK/ERK. However, this activation was abolished by KRT1-KO in NPC cells (Fig. 7B). We subsequently used a small-molecule inhibitor targeting STATs (SH-4-54) to inhibit the JAK/STATs signaling pathways activated by KRT1 in NPC cells, with the aim of exploring the relationship between tumor progression and KRT1-induced pathways. The CCK-8 assay results indicated that overexpression of KRT1 enhanced the proliferative ability of NPC cells, but this activation was abolished by SH-4-54 treatment (Fig. 7C). In parallel, wound healing, invasion, and colony formation assays revealed that KRT1 overexpression dramatically increased the migration, invasion, and colony formation of NPC cells. However, SH-4-54 treatment significantly inhibited KRT1-induced migration, invasion, and colony formation in NPC cells (Fig. 7D-F). Therefore, these results suggest that hsa_circ_0007439 upregulates KRT1 in NPC cells, thereby triggering the activation of the JAK/STATs pathway to promote progression in NPC.

Fig. 7.

KRT1 promotes NPC progression by activating the JAK/STATs pathway. A KEGG pathway enrichment analysis of the DEGs revealed the top 20 enriched pathways in hsa_circ_0007439-overexpressing NPC cells. B Western blot of phospho-STAT3, STAT3, phospho-AKT, AKT, phospho-ERK1/2, and ERK1/2 in KRT1-overexpressed and knockout NPC cells. C-F CCK-8 assay (C), wound healing assay (D), transwell invasion assay (E), and clone formation assay (F) were used to assess the cell proliferation, migration abilities (scale bar, 200 μm), invasion abilities (scale bar, 200 μm), and clone formation abilities of KRT1-overexpressed NPC cells, as well as those treated with SH-4-54 in combination with KRT1 overexpression

Hsa_circ_0007439 accelerates xenograft NPC growth and metastasis

We further addressed the effect of hsa_circ_0007439 in vivo using BALB/c nude mice. BALB/c nude mice were subcutaneously injected with 1 × 10⁶ 6-10B cells with or without hsa_circ_0007439 overexpression to establish NPC xenograft model (Fig. 8A). Upon the largest tumor diameter reaching 15 mm, the mice were euthanized in accordance with ethical guidelines. Our data demonstrated a significant increase in tumor size and volume in the hsa_circ_0007439-overexpressing group compared to the control group (Fig. 8B-D), indicating that hsa_circ_0007439 accelerates the growth of NPC cells in a xenograft mouse model.

Fig. 8.

Overexpression of hsa_circ_0007439 leads to increased NPC growth and metastasis in xenograft mice. A Experimental design of cell transplantation with NPC cells by subcutaneous (SC) injection and subsequent in vivo studies. B-D The representative images of xenograft tumor (B), tumor volume (C), and tumor weight (D) in each group were displayed. E Immunohistochemical analysis of KRT1, Ki-67, E-cadherin, and Vimentin in xenograft tumors (Scale bar, 10 μm). F Immunofluorescence detection of KRT1 and Vimentin (Scale bar, 10 μm). G In vivo tissue imaging of lungs, liver, spleen, kidney, heart at the experimental endpoint. H H&E staining of the livers (Scale bar, 20 μm) and lungs (Scale bar, 200 μm) were showed. Metastatic nodules of the lungs were enlarged and indicated. I, J Number of metastatic nodules in the lung and liver

Subsequently, an immunohistochemical analysis was performed on KRT1, Ki-67, E-cadherin, and Vimentin within tumor tissue. The results indicated that hsa_circ_0007439 overexpression led to an upregulation of KRT1, Ki-67, and Vimentin in NPC tumors, while concurrently downregulating E-cadherin (Fig. 8E). These findings were corroborated by immunofluorescent staining for KRT1 and Vimentin (Fig. 8F). In vivo fluorescence imaging notably revealed that the overexpression of hsa_circ_0007439 facilitated the metastasis of NPC cells to the lung and liver (Fig. 8G). Furthermore, H&E staining indicated that the upregulation of hsa_circ_0007439 significantly enhanced lung and liver metastasis of NPC, as evidenced by an increased number of metastatic nodules, elevated lung weight, and substantial disruption of pulmonary and hepatic architecture in mice with NPC xenografts relative to the control group (Fig. 8H-J, and Figure S7). Collectively, these data, in conjunction with the aforementioned in vitro and in vivo findings, suggest that hsa_circ_0007439 promotes NPC proliferation and metastasis while inhibiting apoptosis by sponging miR-1275 and miR-1293 to upregulate KRT1.

Discussion

NPC exhibits the highest metastasis rate among head and neck cancers, with metastasis being the primary cause of treatment failure in NPC patients [35]. Recently, the burgeoning interest in the role of circRNAs in oncogenesis and cancer progression has attracted considerable attention [36]. With the advancement of sequencing techniques, numerous circRNAs closely associated with tumor metastasis have been identified, thereby expediting research aimed at elucidating the mechanisms underlying tumor metastasis [37, 38]. Nonetheless, the specific circRNAs and their molecular mechanisms implicated in NPC metastasis remain to be comprehensively explored. In this study, we discovered a circRNA, hsa_circ_0007439, which specifically influences NPC progression and whose knockdown significantly impairs the metastasis of NPC cells. Importantly, we demonstrated that hsa_circ_0007439 functions as a molecular sponge for miR-1275 and miR-1293, resulting in the upregulation of KRT1, which in turn facilitates NPC proliferation and metastasis while inhibiting apoptosis. This discovery not only underscores the extensive role of circRNAs in cancer progression but also indicates that hsa_circ_0007439 may serve as a critical oncogenic driver in NPC. Given its evolutionary conservation, stability, and looped structure, hsa_circ_0007439 emerges as a promising candidate for diagnostic and prognostic biomarkers, as well as a potential novel therapeutic target.

In this study, we elucidated the molecular mechanism by which hsa_circ_0007439 acts as ceRNA to regulate the progression of NPC. However, the mechanisms underlying the upregulation of hsa_circ_0007439 in NPC with high metastatic potential remains unclear. Notably, NPC is closely associated with EBV infection, which serves as a significant driver in the genesis and progression of NPC [39]. EBV encodes numerous circRNAs, including circBART2.2 and circRPMS1, which are implicated in the regulation of NPC progression [40, 41]. Despite the EBV-negative status of all cell lines utilized in our study, clinical sample analysis revealed elevated expression levels of hsa_circ_0007439 in NPC tissues exhibiting distant metastasis. Consequently, investigating the potential correlation between EBV infection and the upregulation of hsa_circ_0007439 in clinical NPC samples is crucial for elucidating the molecular mechanisms underlying hsa_circ_0007439 generation. Alternatively, the positive feedback loop mechanism may serve as one of the pathways for the upregulation of circRNAs. For instance, circCAMSAP1 enhances the stability of c-Myc by facilitating the expression of SERPINH1. Concurrently, c-Myc binds to the splicing factor SRSF10, which in turn promotes the formation of circCAMSAP1⁴². Additionally, circRRM2 enhances the expression of IGF2BP1 via the ceRNA mechanism, and the interaction between IGF2BP1 and the transcription factor MYC further induces the upregulation of circRRM2⁴³. In oral squamous cell carcinoma, circUHRF1 acts as a molecular sponge for miR-526b-5p, leading to the upregulation of c-Myc. The transcription factor c-Myc subsequently enhances the transcriptional activity of ESRP1, which promotes the circularization and biogenesis of circUHRF1 by targeting the adjacent intronic regions [44]. In esophageal squamous cell carcinoma, the interaction between circCNTNAP3 and miR-513a-5p is diminished, resulting in the reduced expression of p53 and the p53-induced splicing factor RBM25. This attenuation subsequently impairs the RBM25-mediated cyclization of circCNTNAP3⁴⁵. In NPC, studies have also demonstrated the presence of positive feedback loop mechanism. For example, circPVT1 impedes the proteasomal degradation of c-Myc through its interaction with β-TrCP. As a result, c-Myc transcriptionally upregulates the expression of the RNA splicing factor SRSF1 and facilitates the recruitment of SRSF1 to augment the biosynthesis of circPVT1⁴⁶. These findings suggest that further exploration of whether there are looping mechanisms involved in the formation of hsa_circ_0007439 is crucial for elucidating its role in NPC development.

In recent years, a growing body of evidence has implicated circRNAs in tumor metastasis. Predominantly, studies have demonstrated that circRNAs influence tumor metastasis primarily through mechanisms involving miRNA sponging and protein binding [47]. However, recent research has expanded the functional repertoire of circRNAs, indicating their involvement in transport, transcriptional regulation, and translation into peptides, thereby influencing tumor metastasis through these additional pathways [47, 48]. This study identified that hsa_circ_0007439 can be enriched by Ago2, suggesting its potential role as a miRNA sponge. Specifically, hsa_circ_0007439 interacts with miR-1275 and miR-1293, leading to an upregulation of KRT1 expression and facilitating the metastasis of NPC. However, further research is needed to determine whether hsa_circ_0007439 can promote NPC metastasis through other pathways, including binding proteins and translation peptides. Considering that most circRNAs are relatively low in abundance, a single circRNA or mechanism may limit their application in tumor biology [47, 49]. Therefore, a more comprehensive investigation into additional regulatory networks involving hsa_circ_0007439 could promote its potential application in the detection and treatment of tumor metastasis.

KRT1 is a member of the keratin gene family, which is categorized into type I (KRT9-KRT22) and type II (KRT1-KRT8) [50]. The KRT gene family has been implicated in the pathogenesis and metastatic processes of various malignancies, including lung, breast, and colon cancers [51]. KRT1, classified as a type II keratin, has been demonstrated to be upregulated in various tumor cells and is positively correlated with tumor metastasis [52–54]. However, there are currently no pertinent reports on the biological function of KRT1 in NPC. Our study provides the first evidence that hsa_circ_0007439-upregulated KRT1 facilitates the metastasis in NPC by activating the JAK/STATs pathway. Nonetheless, previous studies have also indicated that KRT1 was involved in the regulation of the Notch signaling pathway [55, 56]. Therefore, future research should focus on investigating the potential role of KRT1 in mediating NPC progression by examining the effects of KRT1 overexpression in conjunction with the inhibition of the Notch signaling pathway. Such studies are of considerable academic importance for elucidating the regulatory mechanisms by which KRT1 influences tumor progression.

In summary, this study identifies a novel circRNA, hsa_circ_0007439, and reveal a hsa_circ_0007439-KRT1 axis involved in the proliferation, survival, and metastasis of NPC. Our work indicates the importance of hsa_circ_0007439 in the progression of NPC and illuminates the possibility that hsa_circ_0007439 is a promising prognostic biomarker and therapeutical target for NPC.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

NPC

Nasopharyngeal carcinoma

EBV

Epstein-Barr virus

circRNA

Circular RNA

ceRNA

Competitive endogenous RNA

PPP1CB

Protein phosphatase 1 catalytic subunit β isoform

EMT

Epithelial-mesenchymal transition

RT-qPCR

Reverse transcription quantitative PCR

ASO

Antisense oligonucleotide

gRNA

Guide RNA

ActD

Actinomycin D

FISH

Fluorescence in situ hybridization

EdU

5-Ethynyl-20-deoxyuridine

LC-MS

Liquid chromatography-mass spectrometry

H&E

Haematoxylin and eosin

Ago2

Argonaut 2

Author contributions

FL, TX, and RS planned and carried out experiments, analysed data, and wrote the manuscript. YH, JFL, HC, JW, JC, PC, QG, QY, WC, CL, MZ, JC, JF, XH, and QW performed the research and analyzed the data. QZ supervised the research, planned experiments, analysed data, and wrote the manuscript. All authors read and approved the final manuscript.

Funding

This work was supported by National Natural Science Foundation of China (31771273, 82273467, and 82403550), China Postdoctoral Science Foundation (2024T171077), Guangdong Province Basic and Applied Basic Research Fund (2022A1515140100), Guangdong Province of China Grant (2021A1515011209), Research Funds for the Shenzhen of China (JCYJ20210324120812035), and the Fundamental Research Funds for the Central Universities, Sun Yat-sen University (24qnpy180).

Data availability

The datasets generated and/or analyzed during the current study are included in this published article and its supplementary information files. The data generated in this study are available upon request from the corresponding author.

Declarations

Ethics approval and consent to participate

Ethical approval for this study was obtained from the Ethical Review Board of Sun Yat-sen University Cancer Center (approval number: B2025-340-01, 2025), ensuring adherence to the ethical standards for research involving human participants as outlined in the Declaration of Helsinki. All experiments involving animals were conducted according to the ethical policies and procedures approved by the Institutional Animal Care and Use Committees of the Sun Yat-sen University (approval number: SYSU-LS-IACUC-2023-0096). All animal experiments were strictly implemented in compliance with the ARRIVE guidelines.

Consent for publication

Not applicable.

Informed consent

All patients provided written and oral informed consent.

Competing interests

The authors declare no competing interests.