TRIM47 drives metabolic reprogramming and tumor progression in Nasopharyngeal Carcinoma via K48-linked ubiquitination and degradation of SDHB

Abstract

Nasopharyngeal carcinoma (NPC) remains a therapeutically challenging malignancy due to its late diagnosis and limited treatment efficacy. Although metabolic reprogramming is a hallmark of cancer, the ubiquitin-mediated mechanisms underlying NPC progression are incompletely understood. Here, we demonstrate that tripartite motif-containing protein 47 (TRIM47) is significantly upregulated in NPC tissues and drives tumor aggressiveness. Through integrated in vitro and in vivo approaches, we found that TRIM47 promotes proliferation, migration, epithelial–mesenchymal transition (EMT), and tumor growth. Mechanistically, TRIM47 directly interacts with succinate dehydrogenase subunit B (SDHB)—a key component of mitochondrial complex II—via its B30.2/SPRY domain and catalyzes K48-linked polyubiquitination, leading to SDHB proteasomal degradation. This degradation induces metabolic reprogramming characterized by enhanced aerobic glycolysis, as evidenced by increased glucose consumption and lactate production. Critically, the oncogenic effects of TRIM47 were reversed by SDHB reconstitution. Moreover, supplementation with succinate, the enzymatic product of SDH, counteracted the tumor-suppressive effects of TRIM47 knockdown. Furthermore, exploiting the metabolic vulnerability induced by TRIM47, ascorbate treatment effectively suppressed TRIM47-driven tumor growth. Our results identify TRIM47 as a novel E3 ligase responsible for SDHB ubiquitination and degradation, thereby promoting Warburg-like metabolism and NPC progression. These findings unveil the TRIM47–SDHB axis as a promising therapeutic target and support metabolic intervention with ascorbate as a potential precision strategy for NPC treatment.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s00018-026-06271-5.

Introduction

Nasopharyngeal carcinoma (NPC) is a head and neck malignancy arising from the columnar epithelium of the nasopharyngeal mucosa, demonstrating distinct geographical prevalence with particularly high incidence rates in Southern China and Southeast Asia, where it remains one of the leading lethal malignancies [1, 2]. The clinical management of NPC presents significant challenges due to its anatomically concealed location and asymptomatic early-stage presentation, resulting in approximately 80% of patients being diagnosed at advanced stages [3]. Current therapeutic protocols require multimodal approaches combining radiotherapy, chemotherapy, and immunotherapy, which are frequently associated with prolonged treatment duration, substantial financial burden, adverse side effects, and suboptimal therapeutic outcomes [3]. The precise pathogenesis of NPC remains incompletely understood. Emerging evidence supports a multifactorial etiology involving Epstein-Barr virus (EBV) infection, genetic predisposition, and exposure to chemical carcinogens [4]. Mechanistically, NPC progression involves an intricate interplay of immune evasion mechanisms, dysregulated gene expression, aberrant signaling pathways, and remodeling of the tumor microenvironment, collectively forming a complex regulatory network that drives tumorigenesis [5]. Given these persistent challenges, elucidating the molecular underpinnings of NPC pathology and identifying novel biomarkers represent critical priorities for enhancing diagnostic precision and developing targeted therapeutic strategies.

The ubiquitin-proteasome system (UPS) serves as a critical post-translational regulator of protein homeostasis, governing degradation, functional modulation, and subcellular trafficking through coordinated actions of ubiquitin-activating enzymes (E1s), ubiquitin-conjugating enzymes (E2s), and ubiquitin ligases (E3s) [6]. Dysregulation of E3 ubiquitin ligases has been increasingly recognized as pivotal in carcinogenesis and tumor progression across malignancies [7]. Tripartite motif-containing protein 47 (TRIM47), a zinc-finger RING-type E3 ligase within the TRIM family, is characterized by conserved N-terminal RING finger, B-box, and coiled-coil domains, and maps to chromosomal locus 17q24-25 [8, 9]. Emerging evidence demonstrates that TRIM47 exhibits oncogenic properties through distinct ubiquitination substrates, driving malignant progression and therapeutic resistance in diverse cancers [10–14].

In hepatocellular carcinoma (HCC), it facilitates K48-linked ubiquitination-mediated degradation of cysteine dioxygenase type 1 (CDO1), thereby promoting proliferation and metastasis [10]. Pancreatic cancer models reveal its role in accelerating aerobic glycolysis and tumor growth through fructose-1,6-bisphosphatase (FBP1) ubiquitination [11]. In breast cancer, TRIM47 induces endocrine resistance via K27-linked polyubiquitination and stabilization of PKCε [12]. Head and neck squamous cell carcinoma studies demonstrate its suppression of apoptosis and autophagy through XAF1 ubiquitination and degradation [13]. Colorectal cancer progression is potentiated by TRIM47-mediated Smad4 ubiquitination and proteasomal elimination [14]. However, the functional significance and molecular mechanisms of TRIM47 in NPC remain elusive.

Herein, we identified TRIM47 as a critical oncogenic driver in NPC, demonstrating its tumor-promoting effects through both in vitro and in vivo models. Specifically, TRIM47 exhibited significant upregulation in NPC tissues and functionally enhanced cancer cell proliferation and migration. Xenograft assays further confirmed its capacity to accelerate tumor growth. Mechanistically, we uncovered a novel regulatory axis wherein TRIM47 directly binds to succinate dehydrogenase subunit B (SDHB) and promotes its K48-linked ubiquitination-mediated proteasomal degradation. This SDHB destabilization induced metabolic reprogramming characterized by enhanced aerobic glycolysis. Our findings not only elucidate the TRIM47/SDHB axis as a pivotal metabolic switch driving NPC pathogenesis but also provide preclinical rationale for targeting this pathway–small-molecule inhibitors of TRIM47 or SDHB-stabilizing agents could represent novel therapeutic strategies.

Materials and methods

Cell lines

Human nasopharyngeal carcinoma (NPC) cell lines (5–8F, 6-10B, CNE-1, CNE-2Z, HK-1) were purchased from IMMOCELL (Xiamen, China), immortalized normal nasopharyngeal epithelial cells (NP69) were provided by Xiaopeng Xiong (Jiangxi Cancer Hospital). Human NPC cell lines (SUNE1 and HONE1) were obtained from Qiaoli Lv (Jiangxi Cancer Hospital). HEK293T were gifts from Dr. Chao Shen and Dr. Congyi Zheng (College of Life Sciences, Wuhan University). Mycoplasma contamination tests were carried out routinely in all listed cells using Mycoplasma Detection Kit (Servicebio, Wuhan, China). NP69 cells were grown in K-SFM (Kinlogix, Guangzhou, China) supplemented with 2% FBS and 0.2ng/mL EGF. The other cells were maintained routinely in RPMI 1640 medium or DMEM (Servicebio, Wuhan, China) supplemented with 10% FBS (ExCell Bio, Taicang, China) and 100 units/mL penicillin and streptomycin under an atmosphere of 5% CO₂ at 37 °C.

Reagents

MG132 (HY-13259), Cycloheximide (CHX, HY-12320), 3-Methyladenine (3-MA, HY-19312) and L-Ascorbic acid (Vitamin C, HY-B0166) were obtained from MCE (Monmouth Junction, NJ, USA). Succinic acid (S9512) was purchased from Sigma-Aldrich (St. Louis, MO, USA). K48-linkage specific Ub (8081S) and Ub (3936S) were purchased from Cell Signaling Technology (Danvers, MA, USA). TRIM47 (26885-1-AP), SDHB (10620-1-AP, 67600-1-Ig), GFP (66002-1-Ig, 50430-2-AP), Flag (66008-4-Ig, 20543-1-Ig), HA (66006-2-Ig, 51064-2-Ig), Ki67 (27309-1-AP), PCNA (10205-2-AP), E-cadherin (20874-1-AP), N-cadherin (22018-1-AP), Vimentin (10366-1-AP), GAPDH (60004-1-Ig), β-actin (66009-1-Ig) antibodies and recombinant human SDHB protein (Ag29868) were from Proteintech (Wuhan, China). Anti-TRIM47 (H00091107-M02) antibody was purchased from Abnova (Taibei, China). The lentiviruses pLV-Scramble [shRNA-NC] (sequence: 5’-CCTAAGGTTAAGTCGCCCTCG-3’), pLV-shTRIM47 [shRNA-1] (sequence: 5’-TGAAGCTCCCAGGGACTATTT-3’) and pLV-shTRIM47 [shRNA-2] (sequence: 5’-TACTGGGAGGTGGAGATTATC-3’) were purchased from Cyagen Biosciences (Suzhou, China). pLV3-U6-TRIM47-sgRNA1 (sequence: 5’- AGCGCTCGAGCGGCCGTAGG-3’), pLV3-U6-TRIM47-sgRNA2 (sequence: 5’- CCACAAGAGCCCTACGACCG-3’) and pLV3-U6-NT-sgRNA (non-targeting sgRNA, sequence: 5’- CAACAAGATGAAGAGCACCA-3’) were provided by MiaoLing Plasmid Platform (Wuhan, China). Q-PCR primers were from TSINGKE Biological Technology (Beijing, China).

Cell proliferation assay

The cell proliferation was detected using the Cell counting kit-8 (CCK8), low serum assay and colony formation assay. Specifically, the CCK8 assay (Dojindo, Kumamoto, Japan) was employed, wherein cells were inoculated at a density of 5000 cells per 96-well plate. On the second day (at least 16 h), 10 µL of CCK8 solution was added to each well and incubated for 1–2 h at 37℃, after which the OD450 was measured with a microplate reader. For the low serum assay, the indicated cells were seeded in 12-well plates (10⁵ cells per well) with three replicates, which were grown in full growth medium with 10% FBS. On the following day, the number of cells was calculated and recorded as day 0, and the medium was replaced to 1% FBS medium every other day. Cell density was counted at day 2, 4 and 6, respectively. In the colony formation assay, cells were plated into 6-well plates at a density of 10³ per well and grown in 5% FBS medium for 14–20 days when visible colonies appeared. Subsequently, the plates were fixed in 4% PFA and stained with 2% crystal violet. The number of clones containing more than 50 cells was counted and photographed.

Cell migration assay

The cell migration assay was detected using the scratch wound healing and transwell migration assay, respectively. To conduct the scratch wound healing assay, the indicated cells were seeded into 6-well plates to obtain a monolayer of 90–100% confluence. Thereafter, a linear scratch was created by means of a pipette tip, thereby simulating the desired injury to the monolayer. After washing with PBS, the cells were subject to photographic documentation at 12-hour intervals until the “scratch” was fulfilled. The transwell migration assay was performed as described using 8 μm pore size transwell chambers manufactured by Corning (NY, USA). A total of 10⁵ cells were resuspended in 0.2 mL of 1% FBS medium and subsequently seeded into the upper chamber. Meanwhile, 0.5 mL of 10% FBS medium was added to the lower chamber. Following a 10-hours incubation period at 37 °C, the cells located on the upper surface of the membrane were removed, while the cells on the lower surface were fixed, stained with 1% crystal violet and photographed.

Western blot and quantitative real-time PCR (qPCR)

Whole-cell protein lysates from the indicated cells or tumors were extracted by RIPA lysis buffer (Beyotime, Shanghai, China). Following BCA quantitation via the BCA Protein Assay Kit (Glpbio, Shanghai, China), the protein lysates were analyzed through western blot as described previously [15]. The density of each band was then analyzed with Gelpro 32 software and normalized to that of endogenous housekeeping protein. Quantitative PCR (qPCR) was performed on a ViiA-7 real-time PCR system instrument (ABI, Waltham, MA, USA) using the 2^−ΔΔCT method, with the β-actin transcript serving as the internal standard. The primer sequences used are listed in Table 1.

Table 1.

Primer sequences used in RT-qPCR experiment

Primer names	Sequence
TRIM47 Forward Primer	5’-CAGTCCAAAGTCCTGAGCG-3’
TRIM47 Reverse Primer	5’-GCTACGGCTGCACTCTTGAT-3’
SDHB Forward Primer	5’- TGTGATAAAGGATCTTGTTCCCGA-3’
SDHB Reverse Primer	5’- CGTAGAGCCCGTCCAGTTTC-3’
β-ACTIN Forward Primer	5’-ACGTTGCTATCCAGGCTGTG-3’
β-ACTIN Reverse Primer	5’-GAGGGCATACCCCTCGTAGA-3’

Tumor xenografts

In order to establish a tumor xenograft model, 5 × 10⁶ CNE-1 stable cells (empty vector group as control and TRIM47 overexpression group) resuspended in 50 µL of PBS/Matrigel (1:1) were subcutaneously injected into both flanks of 4-6-week-old male Balb/c-nu mice (SPF Biotechnology, Beijing, China). For the TRIM47 knockdown experiments, stable transfected CNE-2Z cells (control shRNA group and sh-TRIM47 group) were collected and resuspended in 50 µL of PBS/Matrigel (1:1) at a total amount of 5 × 10⁶ cells. The injection procedures were the same as listed above. Tumor sizes were measured with calipers every three days, and the volumes of tumors were calculated via the formula (Volume=Length×Wideth²/2). On the day of sacrifice, the xenografts were weighed and harvested for HE and IHC analysis.

Yeast two-hybrid screening and validation

Screening

The full-length human TRIM47 gene was subcloned into the pGBKT7 vector as the bait. The Mate & Plate Library-Universal Human (Normalized) (TaKaRa, Dalian, China) was used as the prey library, which was composed of Y187 strains pre-transformed with library plasmids. After a 24 h incubation, the bait-transformed Y2HGold cells were mated with the prey strains for 20–24 h to allow diploid formation. The diploid cells were plated onto DDO/X/A (SD/-Leu/-Trp/X-α-Gal/Aba⁺) plates for the first-round screening. Blue colonies were then picked and streaked onto higher stringency QDO/X/A (SD/-Ade/-His/-Leu/-Trp/X-α-Gal/Aba⁺) plates for a second-round screening. Colonies that grew and turned blue on QDO/X/A plates were subjected to PCR and sequencing.

Validation

The indicated combinations of pGBKT7-TRIM47 (bait) and pGADT7-SDHB (prey), along with negative controls (pGBKT7-Lam+pGADT7-T) and positive controls (pGBKT7-53 + pGADT7-T), were co-transformed into Y2HGold yeast strain. Transformants were then seeded onto the DDO (SD/-Leu/-Trp) and QDO/X/A plates. The blue colonies observed on QDO/X/A plates were considered to exhibit interactions.

Co-immunoprecipitation (Co-IP)

For co-IP assays, cells transfected with indicated vectors were collected and lysed with 0.5% Lubrol-Px lysis buffer (50 mM KCl, 2 mM CaCl₂, 20% glycerol and 50 mM Tris-HCl, pH7.4). After centrifugation, the resultant supernatants were harvested and mixed with 1 µg antibody for 2 h at 4℃. Thereafter, the mixture was incubated with 100 µL protein A/G beads (Beyotime, Shanghai, China) at 4℃ overnight with rotation of 10 r/min. On the subsequent day, the bound proteins were subjected to immunoblotting analysis following a thorough wash with PBS, as described previously.

GST pull-down

For the GST pull-down experiment, recombinant GST-TRIM47 or GST protein was produced and purified in E.coli as previously described [16]. Cell lysates from GFP-SDHB transfected HEK293T cells were subjected to an overnight incubation with GST-TRIM47 or GST as negative control overnight at 4℃ with gentle rotation. Protein complexes were precipitated with Glutathione Sepharose 4B Beads (Thermo Scientific, Waltham, MA, USA), followed by SDS-PAGE for Coomassie brilliant blue staining and immunoblotting with the indicated antibodies.

Immunofluorescence

The co-localization of TRIM47 and SDHB was detected through the following procedures. HONE1 cells were seeded onto the glass coverslips and grown overnight until the cell confluence reached approximately 20%, then treated with MG132 for another 4 h. Thereafter, cells were fixed in cold 4% PFA for 1 h and incubated with 0.5% TritonX-100 for 10 min. For IF staining of TRIM47 (1:500, Abnova, mouse) and SDHB (1:500, Proteintech, rabbit), the detailed methods described before were used. Following the addition of a drop of mounting medium with DAPI (Servicebio, Wuhan, China), the slides were photographed using a Zeiss LSM800 confocal microscope.

In vitro ubiquitinoylation assay

The in vitro ubiquitination assay was performed via the Ubiquitinoylation Kit (Enzo, Long Island, USA) following the manufacturer’s instructions. Briefly, His-SDHB was incubated with purified GST-TRIM47, along with E1, E2 enzymes (Ube2d1 + Ube2d2), ubiquitin, ATP, and reaction buffer supplied in the kit. The reaction mixture was incubated at 37℃ for 1 h, then denatured by adding SDS-containing 2×loading buffer and boiled at 100℃ for 8 min. Ubiquitination products were subsequently detected by immunoblotting using specific antibodies.

Immunohistochemistry

Tumor tissue samples were dewaxed with water, and antigen retrieval was subsequently performed. The activity of endogenous peroxidase was then blocked using a 3% hydrogen peroxide solution. After blocking with 5% BSA, the slides were incubated with Ki67, TRIM47 and SDHB primary antibodies and subsequently secondary antibodies, respectively. Finally, DAB staining solution was used for the initiation of color development, with the nuclei being counterstained with hematoxylin. Following the processes of dehydration and sealing, the samples were scanned.

Measurement of glucose consumption and lactate production

Glycolysis metabolic analysis was conducted via the detection of glucose consumption and lactic acid production. In detail, the cells to be tested were seeded into 6-well plates with full growth medium (10% FBS) and cultured overnight. Following a 4 h period of starvation, the culture media were harvested and the floating cells or cell debris were removed by 1000 g centrifugation at 4℃ for 5 min. Then, the supernatant was collected, and both glucose consumption and lactate production were measured using a glucose assay kit (ab65333, Abcam, USA) and L‐lactate assay kit (ab65331, Abcam, USA) according to the manufacturers’ instructions, respectively.

Statistical analysis

Data obtain in this study were expressed as mean ± standard deviations (SD) from experiments based on at least three replicates. The two-tailed unpaired Student’s t test was used to evaluate the differences between the two groups. A P-value less than 0.05 was considered to be significant.

Results

The expression of TRIM47 was elevated in NPC cells

We first investigated the expression pattern of TRIM47 in the clinical samples of NPC. As shown in Fig. 1A and 1B, immunohistochemical (IHC) staining revealed stronger TRIM47 immunoreactivity in NPC tissues than in noncancerous nasopharyngeal (NC) tissues. Western blot analysis of six paired clinical samples further confirmed that TRIM47 protein expression was significantly elevated in NPC tissues compared to NC tissues (Fig. 1C). We next examined TRIM47 expression in a panel of NPC cell lines. Western blot analysis demonstrated that TRIM47 protein levels were markedly higher in multiple NPC cell lines (SUNE-1, HONE-1, 5-8F, CNE-1, 6-10B, CNE-2Z) compared to the immortalized normal nasopharyngeal epithelial cell line NP69 (Fig. 1D). Furthermore, quantitative RT-PCR analysis confirmed that TRIM47 mRNA expression was also significantly upregulated in CNE-1, CNE-2Z and 6-10B NPC cell lines (Fig. 1E). Collectively, these results indicate that TRIM47 is consistently overexpressed in NPC at both the protein and mRNA levels, suggesting its potential involvement in NPC pathogenesis.––

Fig. 1.

TRIM47 is upregulated in NPC tissues and cell lines. A. Representative images of IHC staining of TRIM47 in noncancerous nasopharyngeal (NC) tissues and NPC tissues. Scale bar, 50 μm. B. Perform optical density analysis on the immunohistochemistry results of NC (n = 10) and NPC (n = 96) tissues, and perform unpaired t-test analysis. C. Western blot analysis showing TRIM47 protein expression in 6 paired NC and NPC tissue samples. GAPDH was used as a loading control. Quantitative analysis of TRIM47 protein levels in NPC tissues (n = 6) relative to NC tissues (n = 6) using Gelpro 32. D. Western blot analysis of TRIM47 protein levels in a panel of NPC cell lines (SUNE-1, HONE-1, 5–8F, CNE-1, 6-10B, CNE-2Z) and an immortalized normal nasopharyngeal epithelial cell line (NP69). GAPDH served as a loading control. Quantitative analysis of TRIM47 protein levels in various NPC cell lines relative to NP69 cells using GelPro. E. Quantitative real-time PCR (qRT-PCR) analysis of TRIM47 mRNA levels in the indicated NPC cell lines. Expression levels were normalized to β-actin and are presented relative to NP69 cells. Data are presented as mean ± standard deviation (SD); n.s. p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

TRIM47 deficiency suppresses proliferation and migration of NPC cells

To investigate the functional role of TRIM47 in NPC progression, we first established TRIM47-knockdown stable cell lines using 6-10B and CNE-2Z, which exhibit relatively high endogenous TRIM47 protein levels (Fig. 1D and Supplementary Fig. 1A). Knockdown of TRIM47 significantly inhibited NPC cell proliferation, as evidenced by both CCK-8 and low serum assays (Supplementary Fig. 1B, C). Consistent with this result, TRIM47 downregulation also markedly reduced colony formation capacity (Supplementary Fig. 1D). We next examined whether TRIM47 influences NPC cell migration. Wound healing and Transwell assays revealed that TRIM47 knockdown significantly decreased the number of migrated cells in both tested cell lines (Supplementary Fig. 1E). Furthermore, wound closure was delayed in TRIM47-depleted cells compared to controls (Supplementary Fig. 1F, G). To explore a potential mechanism underlying these phenotypic changes, we analyzed the expression of epithelial–mesenchymal transition (EMT) markers. TRIM47 knockdown increased the protein level of E-cadherin, an epithelial marker [17], while decreasing the expression of the mesenchymal markers N-cadherin and Vimentin [18] (Supplementary Fig. 1H and Fig. 2).

To circumvent off-target effects associated with shRNA, we generated TRIM47-knockout 6-10B and CNE-2Z cells using a mixture of two specific sgRNAs (sgRNA1 and sgRNA2) targeting TRIM47. Successful knockout was confirmed by immunoblotting for TRIM47 protein (Fig. 2A). Consistently, TRIM47 knockout significantly inhibited the proliferation (Fig. 2B, C) and colony formation (Fig. 2D) of NPC cells. Moreover, TRIM47 deficiency suppressed tumor cell migration, as demonstrated by wound healing and Transwell assays (Fig. 2E–G). As expected, TRIM47 knockout increased E-cadherin protein levels while decreasing the expression of N-cadherin and Vimentin (Fig. 2H). Collectively, these findings demonstrate that TRIM47 knockout effectively suppresses the proliferation and migration of NPC cells, potentially by reversing the EMT.

Fig. 2.

TRIM47 deficiency inhibits proliferation and migration of NPC cells. (A) To generate stable cell lines, 6-10B and CNE-2Z cells were infected with sgRNAs (mixture of sgRNA-1 and sgRNA-2) targeting TRIM47 or the control lentivirus (NT sgRNA). The efficiency of knockout was detected via western blot. (B) The indicated NPC cells were seeded into a 96-well plate and subsequently subjected to CCK8 analysis. (C) The indicated NPC cells were identified through the implementation of a low-serum growth assay in a 1% FBS medium, and cell counts were conducted at 0, 2, 4, and 6 days. The data are expressed as the mean ± standard deviation (n = 3). (D) 1 × 10³ cells were seeded in each well of a 6-well plate for the execution of the colony formation assay. Following a 14-20 day period, the cells were subjected to fixation, staining, and photographic documentation (see left image). The number of clones was meticulously enumerated and subjected to statistical analysis (right image). (E) Transwell migration assay. A total of 1 × 10⁵ cells as indicated were seeded in the upper chamber with 1% FBS medium and grown for about 10 h. Then, cells were fixed, stained and photographed (left). The number of migrated cells were calculated (right). Scale bar: 200 μm. F-G. Wound healing assay. 6-10B cells in 6-well plate were scratched with a pipette tip and photographed at 24-hour intervals (F). Scale bar: 200 μm. The cell migration rate was calculated and analyzed statistically (G). H. The protein levels of key EMT markers in TRIM47 knockout cells or control ones were detected using immunoblotting as indicated. Relative protein density was determined by Gelpro, which was normalized to GAPDH. The data are shown as the means ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, compared with control group

Overexpression of TRIM47 promotes the proliferation and migration of NPC cells

To further validate the oncogenic role of TRIM47, we generated TRIM47-overexpressing stable cell lines using 5–8F and CNE-1 cells, which exhibit low endogenous TRIM47 levels (Fig. 1D and Supplementary Fig. 3A). Consistent with the loss-of-function results, TRIM47 overexpression significantly enhanced NPC cell proliferation, as measured by CCK-8 and low serum assays (Supplementary Fig. 3B, C). Colony formation ability was also increased upon TRIM47 upregulation (Supplementary Fig. 3D). Moreover, TRIM47 overexpression markedly promoted cell migration, as evidenced by accelerated wound closure (Supplementary Fig. 4A, B) and increased number of migrating cells in Transwell assays (Supplementary Fig. 4C, D). Furthermore, ectopic expression of TRIM47 induced EMT, characterized by downregulation of E-cadherin and upregulation of N-cadherin and Vimentin (Supplementary Fig. 4E, F). Taken together, these gain-of-function results demonstrate that TRIM47 overexpression enhances the proliferative and migratory capacities of NPC cells, further supporting its role as a promoter of malignant progression.

TRIM47 promotes NPC tumor growth in vivo

To evaluate the role of TRIM47 in NPC tumor growth in vivo, we established xenograft models in nude mice using CNE-2Z cells with TRIM47 knockdown and CNE-1 cells with TRIM47 overexpression. Knockdown of TRIM47 significantly suppressed tumor growth, as shown by reduced tumor volume and weight (Fig. 3A–D). Consistent with these findings, IHC and western blot analyses revealed decreased expression of the proliferation markers Ki67 and PCNA in TRIM47-knockdown tumors (Fig. 3E, F). Conversely, TRIM47 overexpression in CNE-1 cells markedly increased both tumor volume and weight (Fig. 3G–J), and elevated the protein levels of Ki67 and PCNA (Fig. 3K, L). Collectively, these in vivo results demonstrate that TRIM47 plays a crucial role in promoting NPC tumor growth.

Fig. 3.

TRIM47 promotes NPC tumor growth in vivo. A-D. TRIM47 knockdown led to the inhibition of tumor formation with CNE-2Z cells in vivo. Six nude mice were injected subcutaneously with 5 × 10⁶ TRIM47-knockdown stable cell lines or control cells/mouse. The tumor volume was measured every 3 days from day 21, and the mice were euthanized on day 33. The growth curves of subcutaneous tumor xenografts were plotted individually for each mouse (C) and grouped for statistical analysis (B). Tumors were isolated, measured and weighed. Images of the tumors are present in A, while data of the tumor volume and weight are listed in D. Data are presented as the means ± SD (n = 6). E. On the day of sacrifice, the tumors were isolated, fixed, and subjected to HE and IHC assays for Ki67 and TRIM47. F. The expression of proliferation marker PCNA was detected in the tumors via immunoblotting. G-J. Overexpression of TRIM47 promoted the growth of CNE-1 cells in a xenograft model. Nude mice were injected subcutaneously with 5 × 10⁶ NPC cells/mouse stably expressing the empty vector (EV) or TRIM47 (OE-TRIM47). 14 days later, the tumor volume was measured every 3 days, and the mice were sacrificed on day 26. The growth curves of subcutaneous tumor xenografts were plotted individually for each mouse (I) and grouped for statistical analysis (H). Tumors were isolated, measured and weighed. The images of the tumors are shown in G, while the tumor volume and weight data are listed in J. The data are presented as the means ± SD (n = 6). K. Representative HE and IHC images of TRIM47 overexpressing tumors. L. PCNA protein levels in the tumors were detected by western blot analysis. *p < 0.05, **p < 0.01, ***p < 0.001 by Student’s t-test

SDHB is identified as a novel binding partner of TRIM47

To elucidate the molecular mechanism by which TRIM47 exerts its oncogenic functions in NPC, we performed a yeast two-hybrid screen using full-length human TRIM47 as bait. This screen identified SDHB—an iron-sulfur protein subunit of the succinate dehydrogenase (SDH) complex with essential roles in mitochondrial respiration [19, 20]—as a potential interacting partner. The TRIM47–SDHB interaction was confirmed by yeast two-hybrid assay (Fig. 4A), and further validated through co-immunoprecipitation (co-IP) in HEK293T cells overexpressing both proteins (Fig. 4B). Endogenous co-IP in 5–8F NPC cells also demonstrated a direct interaction between TRIM47 and SDHB (Fig. 4C). Additionally, GST pull-down assays confirmed the physical binding, wherein SDHB was specifically pulled down by GST-TRIM47, but not by GST alone (Fig. 4D). Subcellular localization studies in HONE1 cells revealed that TRIM47 and SDHB were mainly localized in the cytoplasm, and after treatment with MG132, they exhibited strong co localization signals (Fig. 4E).

Fig. 4.

TRIM47 interacted with SDHB. (A) SDHB was screened to bind with TRIM47 by yeast-two-hybrid assays. Their interaction was confirmed in yeast using a co-transfection method on various selection plates. (B) Co-immunoprecipitation analysis of the interaction between exogenous Flag-TRIM47 and HA-SDHB in HEK293T cells. (C) The endogenous interaction between TRIM47 and SDHB was detected in 5–8F cells via co-IP analysis. (D) The purified GST or GST-TRIM47 protein was subjected to an incubation with the lysates of HA-SDHB-transfected HEK293T cells. The expression of the GST and GST-TRIM47 proteins was confirmed by Coomassie brilliant blue staining. After the GST pull-down experiment, the eluted protein and cell lysates were analyzed via western blot. (E) Co-localization of TRIM47 and SDHB in HONE1 cells with MG132. The SDHB proteins were stained with a green fluorescent antibody (Alexa Fluor 488), with red fluorescence (Alexa Fluor 594) representing TRIM47 and blue representing the nucleus. (F) Detection of TRIM47 binding domain with SDHB was determined by co-IP assays of HEK293T cells co-transfected with SDHB and TRIM47 (FL: full-length) or its deletion mutants as indicated. (G) Schematic representation of TRIM47 deletion mutants with Flag tag. (H) Determination of SDHB binding domain with TRIM47 was performed using co-IP assay (FL: full-length). (I) Schematic representation of SDHB deletion mutants with GFP tag

To map the interaction domains, we constructed a series of deletion mutants of both TRIM47 and SDHB (Fig. 4G, I). Co-IP assays following MG132 treatment showed that only the C-terminal B30.2/SPRY domain (amino acids 410–638) of TRIM47 interacted with SDHB (Fig. 4F). For SDHB, both N- and C-terminal regions were dispensable for binding to TRIM47 (Fig. 4H). These results indicate that TRIM47 binds to SDHB specifically through its B30.2/SPRY domain.

TRIM47 promotes K48-linked ubiquitination and proteasomal degradation of SDHB

Given that TRIM47 is a known E3 ubiquitin ligase [8], we investigated whether it ubiquitinates SDHB and mediates its degradation. Overexpression of TRIM47 significantly enhanced the ubiquitination of SDHB (Fig. 5A). To identify the specific ubiquitin linkage type, we utilized a series of ubiquitin mutants in which all lysine residues except one were substituted with arginine (Supplementary Fig. 5A). TRIM47 selectively promoted K48-linked polyubiquitination of SDHB, as ubiquitination was observed only in the presence of the wide-type ubiquitin or K48-only ubiquitin mutant (K48O) (Fig. 5B). Consistently, using a K48R ubiquitin mutant abrogated TRIM47-induced ubiquitination of SDHB (Fig. 5C and Supplementary Fig. 5B). The enhancement of K48-linked ubiquitination by TRIM47 was further confirmed using a K48-linkage specific antibody (Fig. 5D). Furthermore, overexpression of full-length Flag-TRIM47, but not the TRIM47 mutant (M) lacking the RING finger domain, markedly enhanced SDHB ubiquitination, indicating that the E3 ligase activity of TRIM47 is essential for this modification (Fig. 5E). In addition, endogenous immunoprecipitation assays also showed increased ubiquitination of SDHB in TRIM47-overexpressing cells compared to controls (Fig. 5F). Finally, an in vitro ubiquitination assay demonstrated that TRIM47 can directly induce SDHB ubiquitination in a cell-free system (Fig. 5G).

Fig. 5.

TRIM47 promotes K48-linked ubiquitination and proteasomal degradation of SDHB. (A) Overexpression of TRIM47 promoted the ubiquitination of SDHB in HEK293T cells. After 24 h, the transient co-transfected cell lysates (HA-Ub/GFP-SDHB/Flag-TRIM47) as depicted were harvested, immunoprecipitated with anti-GFP antibody, and immunoblotted with the indicated antibodies. (B) The detailed linkage way of poly-ubiquitination of SDHB mediated by TRIM47 was determined using co-IP with seven different HA-ubiquitin (HA-Ub) mutants (K6O, indicating K6 only, which means that Ub ligation only occurs via the K6-linked pathway) as indicated. (C) The K48-linked ubiquitination of SDHB mediated by TRIM47 was confirmed via co-IP with seven different HA-ubiquitin (HA-Ub) K→R mutants as indicated. (D) TRIM47 promoted the K48-linked ubiquitination of SDHB were further validated using specific K48 Ub antibody. (E) TRIM47 promoted the ubiquitination of SDHB in a manner dependent on its E3 ligase activity. (F) The ubiquitination level of the SDHB protein was detected after immunoprecipitation with a SDHB antibody in TRIM47-overexpressing 5–8F cells and control cells. (G) In vitro ubiquitination assay. His-SDHB proteins were used in an in vitro ubiquitination reaction in the presence of E1, E2 (Ube2d1 + Ube2d2), Ub and purified GST-TRIM47 proteins. (H) The effect of TRIM47 overexpressing on the endogenous protein levels of SDHB were detected in 5–8F cells transfected with different amount of Flag-TRIM47. (I) TRIM47 mediated degradation of endogenous SDHB was found to be contingent upon the UPS, a finding that was corroborated through incubation with the proteasome inhibitor MG132, the lysosome inhibitor chloroquine (CQ), and the autophagy inhibitor 3-MA. The data are shown as the means ± SD. n.s. p > 0.05, **p < 0.01, ***p < 0.001, compared with control group. # P < 0.05, compared with Flag-TRIM47 (1 µg) group

Since K48-linked ubiquitination typically targets substrates for proteasomal degradation [6], we examined SDHB protein levels. TRIM47 overexpression reduced endogenous SDHB abundance in a dose-dependent manner in 5–8F cells (Fig. 5H). This degradation was rescued by the proteasome inhibitor MG132, but not by inhibitors of lysosomal degradtion (chloroquine, CQ) or autophagy (3-methyladenine, 3-MA) [21] (Fig. 5I). These results demonstrate that TRIM47 promotes K48-linked polyubiquitination of SDHB, leading to its proteasomal degradation.

TRIM47 promotes the degradation of SDHB

Building on the finding that TRIM47 mediates UPS-dependent degradation of SDHB, we further investigated their regulatory relationship in NPC. Endogenous protein levels of SDHB were inversely correlated with TRIM47 expression across multiple NPC cell lines (Fig. 6A). Consistent with this, TRIM47 overexpression decreased, while its knockdown increased, SDHB protein levels in stable cell lines (Fig. 6B). This negative correlation was also observed in xenograft tumor tissues by both western blot and IHC analysis (Fig. 6C, E), but not at the mRNA level (Fig. 6D), supporting post-translational regulation. Cycloheximide chase assays confirmed that TRIM47 overexpression shortened the half-life of SDHB, whereas TRIM47 knockdown extended it (Fig. 6F, G). These results establish SDHB as a proteasomal degradation substrate of TRIM47.

Fig. 6.

The protein level of TRIM47 negatively correlated with that of SDHB in NPC cells and tumors. A-C. The protein levels of TRIM47 and SDHB were detected using western blot analysis in NPC cells (A), TRIM47 stably overexpressing or knocking-down NPC cells (B), as well as tumors from CNE-1/CNE-2Z xenografts (C). D. The effect of TRIM47 knockdown on the mRNA levels of SDHB was detected by RT-qPCR. E. Representative IHC staining images of TRIM47 and SDHB in CNE-1/CNE-2Z xenograft tumors. F-G. TRIM47-knockdown stable 6-10B cells/control cells (F) or TRIM47-overexpressing stable 5–8F cells/control cells (G) were treated with 100 µg/mL CHX for 0, 2, 4, 6, 8 h. The protein expression levels of TRIM47 and SDHB in the aforementioned cells were detected using western blot analysis. Subsequently, Gelpro32 software was employed for gray quantification analysis. The relative protein density of TRIM47/SDHB was normalized to that of β-actin. The data are shown as the means ± SD. n.s. p > 0.05, *p < 0.05, **p < 0.01, compared with control group

TRIM47 drives malignant phenotypes in an SDHB-dependent manner

To determine whether SDHB downregulation mediates the oncogenic effects of TRIM47, we performed rescue experiments. TRIM47 overexpression enhanced proliferation and colony formation of 5–8F cells, and these effects were reversed by co-expression of SDHB (Fig. 7A–D). Since SDHB is a key subunit of the SDH complex that catalyzes the oxidation of succinate to fumarate [22], we assessed metabolic changes. TRIM47 overexpression increased glucose consumption and lactate production, indicating enhanced aerobic glycolysis; this effect was also abolished by SDHB restoration (Fig. 7E, F). Similarly, SDHB overexpression counteracted TRIM47-induced cell migration (Fig. 7G, H). These findings demonstrate that TRIM47 promotes malignant phenotypes in NPC primarily through SDHB degradation.

Fig. 7.

TRIM47 drives malignant phenotypes in an SDHB-dependent manner. A. Western blot analysis of 4 groups of co-transfected 5–8F cells (Flag + HA, TRIM47 + HA, Flag + SDHB, and TRIM47 + SDHB). B-C. 5–8F cells were co-transfected with the indicated vectors. Following a 48-hour period, the cells were transferred to 96-well plates and subsequently subjected to CCK8 analysis (B). The aforementioned cells were identified through the implementation of a low-serum growth assay in a 1% FBS medium, and cell counts were conducted at 0, 2, 4, and 6 days (C). The data are expressed as the mean ± SD (n = 3). D. Colony formation assay. A total of 5 × 10³ cells from each group were plated in 6-well plates containing 5% FBS medium. Following a 14–20 day period, the cells were subjected to fixation and staining with 2% crystal violet. E-F. The glucose consumption (E) and lactate production (F) of the indicated cells were detected using the supernatants. G-H. The migration of those cells in the four groups was detected via the Transwell migration assay (G) and the wound healing assay (H), respectively. The data are presented as the means ± SD. n.s. p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, compared with control group unless indicated

TRIM47 drives malignant phenotypes dependent of its E3 ligase activity

TRIM47 typically exerts its biological functions through its E3 ubiquitin ligase activity. Therefore, we proceeded to investigate whether this enzymatic activity is required for TRIM47-induced malignant behaviors in NPC cells. Upregulation of SDHB inhibited the proliferation and colony formation of 5–8F cells, which can be reversed by co-expression of wildtype TRIM47, but not by the RING finger domain deletion mutant (TRIM47 M) (Fig. 8A-D). Consistently, only wildtype TRIM47 rather than TRIM47 M, rescued the suppression of SDHB-induced cell migration (Fig. 8G, H). Similar results were observed for glucose consumption and lactate production (Fig. 8E, F). Collectively, these results proved that the E3 ligase activity of TRIM47 is indispensable for both SDHB degradation and the promotion of malignant phenotypes in NPC cells.

Fig. 8.

E3 ligase activity is essential for the oncogenic function of TRIM47 in NPC cells. A. Western blot analysis of 4 groups of co-transfected 5-8F cells (Flag + GFP, Flag + SDHB, TRIM47 + SDHB, and TRIM47 M (ΔRING mutant) + SDHB). B. CCK-8 assay was employed to detect the proliferation of the indicated cells of the four groups in 96-well plates. C. The aforementioned cells were identified through the implementation of a low-serum growth assay in a 1% FBS medium, and cell counts were conducted at 0, 2, 4, and 6 days. The data are expressed as the mean ± SD (n = 3). D. Colony formation assay. 5×103 cells from each group were plated in 6-well plates containing 5% FBS medium. Following a 14–20-day period, the cells were subjected to fixation and staining with 2% crystal violet. E-F. The glucose consumption (E) and lactate production (F) were detected using the supernatants of these cells. G-H. The migration of those cells in the four groups was detected via the Transwell migration assay (G) and the wound healing assay (H), respectively. The data are presented as the means ± SD. n.s. p >0.05, *p <0.05, ***p <0.001, ****p <0.0001, compared with control group unless indicated

TRIM47 drives NPC progression by impairing SDHB-mediated succinate metabolism

To determine whether the catalytic activity of SDHB is critical for TRIM47-induced malignancy, we modulated metabolic and redox pathways. Given that ascorbic acid (Vitamin C, VC) exacerbates oxidative stress in SDHB-deficient tumors [23], we treated TRIM47-overexpressing cells with VC. VC administration significantly attenuated TRIM47-enhanced proliferation, colony formation, migration, and glycolytic flux (increased glucose consumption and lactate production) (Fig. 9A–G and Supplementary Fig. 6A-C).

Fig. 9.

TRIM47 drives NPC progression by impairing SDHB-mediated succinate metabolism. A-G. 5–8F cells were transfected with TRIM47 for 24 h, and then treated with or without ascorbic acid (VC) for another 24 h. The aforementioned cells were then subjected to cell proliferation and migration detection using CCK-8 (A), low serum assay (B), colony formation (E), Transwell migration (F) and wound healing assays (G) as indicated. The glucose consumption (C) and lactate production (D) of the above cells were detected using the supernatants, respectively. The data are presented as the means ± SD. *p < 0.05, **p < 0.01, compared with control group. ^##p < 0.01, compared with TRIM47 group. H-N. 6-10B cells were transfected with sh-TRIM47 for 24 h, and then treated with or without succinate (su) for another 24 h. The above cells were then subjected to cell proliferation and migration detection using CCK-8 (H), low serum assay (I), colony formation (L), transwell migration (M) and wound healing assays (N), respectively. The glucose consumption (J) and lactate production (K) were detected using the supernatants as well. The data are presented as the means ± SD. *p < 0.05, compared with control group. *p < 0.05, ***p < 0.001, compared with control group. ^#p < 0.05, ^##p < 0.01, compared with sh-TRIM47 group

Since the primary function of the SDH complex is the conversion of succinate to fumarate [20], we next examined whether succinate supplementation could rescue the tumor-suppressive effects of TRIM47 knockdown. Indeed, succinate treatment reversed the inhibition of proliferation, colony formation, and migration caused by TRIM47 silencing (Fig. 9H-I, L-N and Supplementary Fig. 6D-F). It also restored glycolytic activity, as indicated by increased glucose consumption and lactate production (Fig. 9J-K). Succinate supplementation counteracts the metabolic and phenotypic consequences of TRIM47 knockdown, consistent with the model that TRIM47 promotes progression by degrading SDHB and impairing SDH enzymatic function.

These complementary experiments—using both VC to aggravate metabolic vulnerability in TRIM47-high cells and succinate to restore metabolism in TRIM47-low cells—collectively demonstrate that TRIM47 may promotes NPC progression through the succinate metabolism.

Discussion

This study identifies TRIM47 as a pivotal oncogenic driver in NPC, establishing a novel TRIM47/SDHB metabolic axis as a central mechanism of tumor progression. We demonstrate that TRIM47, significantly upregulated in NPC, directly binds the SDHB, a core component of mitochondrial complex II [22], and targets it for K48-linked ubiquitination and proteasomal degradation. This degradation disrupts succinate-to-fumarate conversion, triggering metabolic reprogramming characterized by enhanced aerobic glycolysis. Functionally, this axis is indispensable for TRIM47-mediated promotion of NPC cell proliferation, migration, and in vivo tumor growth. Our findings position TRIM47 as a master regulator of cancer metabolism in NPC and highlight its therapeutic potential.

While TRIM47 has been implicated in diverse tumor contexts through targeting distinct substrates—such as XAF1 degradation in head and neck squamous cell carcinoma [13], vimentin-mediated metastasis in pharyngeal/laryngeal cancer [24], SNAI1 stabilization [25] and CDO1-mediated ferroptosis suppression in hepatocellular carcinoma [10], and SMAD4 degradation in colorectal cancer [14]—its role in mitochondrial metabolic reprogramming remained unexplored. Our study uncovers TRIM47 directly hijacks respiratory chain components to enforce Warburg-like metabolic switching, establishing a previously unrecognized link between TRIM47-mediated ubiquitination and cancer cell bioenergetics. Notably, TRIM47’s E3 ligase activity is domain-specific: its B30.2/SPRY domain mediates the interaction with SDHB, analogous to its role in binding TRAF2 in endothelial inflammation [26]. This versatility in substrate recognition underscores TRIM47’s capacity to hijack diverse signaling nodes across cancers.

SDHB loss is classically linked to paragangliomas and pheochromocytomas due to succinate accumulation and HIF-1α stabilization[27]. Our findings reveal that TRIM47-induced SDHB degradation drives NPC progression via similar pseudohypoxic mechanisms [28]. Reduced SDHB activity elevates the succinate/fumarate ratio, inhibiting α-ketoglutarate-dependent dioxygenases (e.g., prolyl hydroxylases). This stabilizes HIF-1α even under normoxia, amplifying the glycolytic phenotype evidenced by increased glucose consumption and lactate production. Critically, supplementing succinate reversed the suppressive effects of TRIM47 knockdown on proliferation, migration, and glycolysis, confirming that the metabolic dysregulation is central to TRIM47’s oncogenicity. Importantly, SDHB loss may have broader implications: succinate accumulation can inhibit TET DNA demethylases, potentially promoting tumor suppressor hypermethylation [29]. Future work should explore whether TRIM47-mediated SDHB degradation contributes to epigenetic dysregulation in NPC.

The TRIM47-SDHB axis presents actionable targets for NPC therapy. First, inhibiting TRIM47’s E3 ligase activity could stabilize SDHB. Although no TRIM47-specific inhibitors have been reported to date, recent advances demonstrate the feasibility of pharmacologically targeting TRIM47-mediated protein-protein interactions. Hong et al. [30] identified a small molecule, CZ-2401, that disrupts TRIM47 binding to its substrate HNF4α, thereby stabilizing HNF4α and suppressing hepatocellular carcinoma progression. Our preclinical data, demonstrating that TRIM47 knockdown suppresses tumor growth in vivo, strongly supports this strategy. Ascorbate (vitamin C) disrupts redox balance in SDH-deficient tumors [23]. As a well-established low-molecular-weight antioxidant, VC functions as an enzyme cofactor and modulates redox homeostasis beyond its general antioxidant capacity [31]. Notably, high-dose VC exacerbates mitochondrial reactive oxygen species (ROS) production in SDHB-deficient cells, thereby inducing cell death [32]. We show ascorbate selectively reverses the oncogenic effects of TRIM47 overexpression, suggesting a rational combinatorial approach targeting TRIM47 or SDHB loss alongside pro-oxidant therapies. Mechanistically, VC may act as a co-factor for dioxygenases, potentially restoring their enzymatic activity, alleviating the pseudohypoxia state, and exacerbating oxidative stress in metabolically vulnerable cancer cells, thereby inhibiting growth. TRIM47’s upregulation in NPC tissues also suggests diagnostic and prognostic utility, potentially stratifying patients for intensified or targeted therapies. However, caution is warranted due to TRIM47’s dual roles in tumor promotion and stem cell protection [33]; systemic inhibition might compromise tissue repair during genotoxic therapies like radiotherapy.

While our work establishes TRIM47-SDHB as a key axis in NPC, several questions remain to be uncovered. First of all, the upstream mechanisms responsible for TRIM47 overexpression in NPC require further investigation. Potential candidates can be systematically screened via RNA-seq, including oncogenic transcription factors (e.g., MYC), epigenetic modifiers, or Epstein-Barr virus (EBV)-encoded factors, given that EBV is a key etiological agent in NPC. Secondly, it is better to evaluate the therapeutic efficacy of targeting TRIM47-SDHB axis against established tumors. To this end, we plan to develop or screen for TRIM47-specific inhibitors and test their ability to suppress the growth of pre-existing tumors in preclinical models. Thirdly, SDHB loss is known to promote immune evasion through the accumulation of oncometabolites. Therefore, whether TRIM47 shapes the NPC immune microenvironment merits investigation. Ultimately, the direct physical interaction between SDHB and TRIM47 can be further confirmed through orthogonal biochemical experiments, such as surface plasmon resonance (SPR) or biolayer interferometry (BLI). Addressing these questions will not only deepen our mechanistic understanding but also pave the way for translation of this axis into clinical applications.

Conclusion

In summary, we delineate a novel oncogenic pathway in NPC where TRIM47 acts as a master regulator of SDHB stability. This TRIM47/SDHB axis drives tumor progression by linking ubiquitin-dependent proteolysis to profound metabolic reprogramming and pseudohypoxia. Our work significantly expands the understanding of E3 ligases in cancer metabolism and provides a robust preclinical rationale for targeting this axis. The development of TRIM47-specific inhibitors, potentially combined with metabolic modulators or pro-oxidants like ascorbate, represents a promising novel therapeutic strategy for NPC. Future studies should validate these findings in patient-derived models and explore synergies with existing therapies.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

YL and XH conceived and designed the study, prepared the figures, analyzed data, and participated in the paper writing. YL and JYe supervised the study and planned experiments. JY and LD performed the experiments, analyzed the data, and wrote the first draft of the manuscript. YY, TX and JP conducted experiments during the study. QL, YL and XH revised the manuscript. All authors approved and contributed to the final version of the manuscript.

Data availability

All datasets generated or analyzed in this study are incorporated in this article. Additional resources, reagents, and raw data are available upon request from the corresponding authors.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval

Clinical NPC samples were obtained from the Department of Pathology at the First Affiliated Hospital of Nanchang University, Nanchang, China. The use of these clinical samples was reviewed and approved by the Medical Research Ethics Committee of the Affiliated Hospital of Nanchang University (Approval No. (2023)CDYFYYLK(07–013)). Informed consent was obtained from all patients prior to the use of their tissue samples. All animal experiments were approved by The Institutional Animal Care and Use Committee of The First Affiliated Hospital of Nanchang University (Approval No.: CDYFY-IACUC-202307QR010) and conducted in compliance with the ARRIVE guidelines. All methods were carried out in accordance with relevant guidelines and regulations.