Introduction

Epstein–Barr virus (EBV) is a large double-stranded DNA virus that belongs to the gammaherpesvirinae subfamily [1]. As the first identified human oncogenic virus, EBV contributes to approximately 1.5% of all cases of cancer worldwide, including lymphoid and epithelial cancers [2,3,4]. The malignancy that is most closely associated with EBV infection is undifferentiated nasopharyngeal carcinoma (NPC), which occurs in the epithelial lining of the nasopharynx [5,6,7]. Chemoradiotherapy is the fundamental treatment strategy for NPC. However, the treatment of NPC still faces great challenges due to chemoresistance [8].

EBV is consistently detected in NPC patients from both endemic and non-endemic area. Delineating the cellular processes targeted by EBV is essential to understand its role in tumor initiation and progression and may contribute to the discovery of new therapeutic targets. Increasing evidence has emerged to reveal the functions of viral proteins (e.g., EBVN1 and LMP1) and small RNAs that may contribute to EBV-associated cancers [9, 10]. However, the role of EBV infection in the mechanism of chemoresistance has not been fully elucidated.

Ferroptosis, a nonapoptotic form of cell death, is featured by the excessive iron-dependent accumulation of lipid reactive oxygen species (ROS) [11,12,13]. The distinctive characteristics of cells undergoing ferroptosis include a series of morphological abnormalities of mitochondria, such as organelle shrinkage, condensed organelle membrane, and lessened cristae. Ferroptosis is precisely controlled via a regulatory network involving the inhibitory role of glutathione peroxidase 4 (GPX4) and cystine transporter SLC7A11 (system Xc-, xCT). GPX4 helps to clear the toxic lipid hydroperoxides (LOOH), which prevents cellular damage from oxidative stress and maintains redox homeostasis [14], while the inhibition of GPX4 promotes lipid ROS-dependent ferroptosis. Cancer cells are known to develop iron addiction, which increases the ROS production as a result of cellular transformation and tumorigenesis [15]. In addition to their increased antioxidant capacity, cancer cells are rendered more susceptible to ferroptosis due to their altered redox environment. Therefore, cancer cells are more dependent on GPX4, especially following epithelial-to-mesenchymal transition (EMT) [16, 17]. Directly targeting GPX4 may serve as an efficient strategy to induce ferroptosis in cancer cells in vivo and provide a new approach for ROS manipulation-based cancer therapy.

To the best of our knowledge, the effect of EBV infection on the ferroptosis sensitivity of host cells has not been studied to date. Here, we show that EBV infection activates the p62-Keap1-NRF2 signaling axis, leading to upregulation of GPX4 and SLC7A11, and effectively reduces the ferroptosis sensitivity of NPC cells. Inhibition of GPX4 leads to enhanced chemosensitivity in EBV-infected NPC cells. Additionally, GPX4 knockdown significantly suppresses tumor cell proliferation in vitro and in vivo. We further demonstrated that GPX4 interacts with the TAK1-TAB1/TAB3 complex, regulates TAK1 kinase activity, and activates the downstream MAPK-JNK and NFκB pathways. Altogether, our findings uncovered a novel mechanism by which chemoresistance induced by EBV infection facilitates the evasion of ferroptosis, identifying GPX4 is a potential therapeutic target in NPC.

Results

EBV infection reduces ferroptosis in NPC cells

To explore the effect of EBV infection on the ferroptosis sensitivity of nasopharyngeal epithelial cells, we developed EBV-infected NPC cell lines as previously described [Full size image

TAK1 is a member of the mitogen-activated protein (MAP) kinase kinase kinase (MAP3K) family [30], and it functions as an intracellular hub that modulates not only MAP kinase, but also the nuclear factor-κB (NF-κB), thereby regulating multiple vital biological progresses [

Materials and methods

Clinical specimens

To analyze GPX4 expression and GPX4-related survival, we retrospectively collected 181 paraffin-embedded NPC specimens from Sun Yat-sen University Cancer Centre between November 2010 and November 2011. All patients were diagnosed with nonmetastatic NPC, and none of the patients had received radiotherapy or chemotherapy before biopsy. The 7th edition of the AJCC Cancer Staging Manual was used to reclassify TNM stages. The median follow-up period was 62 months (range, 4–84 months). The pathologic type was WHO III in all cases. All patients received radiotherapy, and patients with stage III–IV NPC received inducing chemotherapy plus concurrent platinum-based chemotherapy or chemotherapy alone. The detailed clinicopathological characteristics are shown in Supplementary Table S1. This study was approved by the Institutional Ethical Review Board of the Sun Yat-sen University Cancer Centre (No. GZR2020-090 for human cancer specimens and NO. L102012020000N for in vivo mouse experiments). Written informed consent was obtained from all patients.

Cell culture

Akata-EBV-GFP is an Akata Burkitt lymphoma cell line carrying the Akata bacterial artificial chromosome (BAC) with a GFP tag and was cultured in RPMI 1640 medium with 10% foetal bovine serum (FBS) (HyClone). NPC cell lines (HK1, HK1-EBV, CNE2, CNE2-EBV) were maintained in Roswell Park Memorial Institute (RPMI) 1640 medium (Life Technologies, Carlsbad, CA, USA) supplemented with 5% foetal bovine serum (FBS) in a humidified atmosphere comprising 5% CO2 at 37 °C. The EBV-positive Akata cell line and all NPC cell lines, which had been authenticated, were kindly provided by Dr. Mu-sheng Zeng (Sun Yat-sen University Cancer Centre). 293FT cells obtained from ATCC were grown in DMEM (Invitrogen) with 10% FBS. All cells were authenticated using short-tandem repeat profiling, tested for mycoplasma contamination, and cultured for less than 2 months.

Antibodies and reagents

The following antibodies were used in this study: anti-4-hydroxynonenal (4-HNE), Abcam (ab46545), for immunoblotting and IHC; anti-p62, Abcam (ab109012), for immunoblotting; anti-NRF2, Abcam (ab62352), for immunoblotting and Cell Signaling Technology (#12721), for immunofluorescence staining; anti-Keap1, Abcam (ab227828), for immunoblotting; anti-SLC7A11, Abcam (ab175186), for immunoblotting; anti-GPX4, Abcam (ab125066), for immunoblotting and IHC; anti-GAPDH, Cell Signaling Technology, #5174, for immunoblotting; anti-β-actin, Abcam (ab8226), for immunoblotting; anti-TAK1, Abcam (ab109526), for immunoblotting; anti-p-TAK1(T187), Cell Signaling Technology, #4536, for immunoblotting; anti-TAB1, Abcam (ab76412), for immunoblotting and immunofluorescence staining; anti-TAB3, Abcam (ab124723), for immunoblotting and immunofluorescence staining; anti-Flag, Sigma (F1804), for immunoblotting and immunofluorescence staining; anti-IKKα, Cell Signaling Technology, #11930, for immunoblotting; anti-IKKβ, Cell Signaling Technology, #8943, for immunoblotting; anti-p-IKKα/β(Ser176/180), Cell Signaling Technology, #2697, for immunoblotting; anti-p-NFκB (Ser536), Cell Signaling Technology, #3033, for immunoblotting; anti-NFκB, Cell Signaling Technology, #8242, for immunoblotting; anti-p-IκBα (Ser32), Cell Signaling Technology, #2859, for immunoblotting; anti-IκBα (Ser32), Cell Signaling Technology, #4814, for immunoblotting; anti-p-JNK(Thr183/Tyr185), Cell Signaling Technology, #4668, for immunoblotting; anti-JNK, Cell Signaling Technology, #9252, for immunoblotting; anti-p-p38(Thr180/Tyr182), Cell Signaling Technology, #4511, for immunoblotting; and anti-p38, Cell Signaling Technology, #8690, for immunoblotting. BODIPY 581/591 C11 (Lipid Peroxidation Sensor), Invitrogen (D3861); erastin, Sigma (E7781); Ferrostatin-1, Sigma (SML0583); 1 S,3R-RSL3, Sigma (SML2234); and SYTOX Kit, Invitrogen (S34862) were also utilized.

Virus preparation and infection

EBV with integrated EGFP (EBV-EGFP) was prepared in Akata cells as previously described [43]. The Akata-EBV-GFP cell line was used to produce EBV in accordance with the following procedure: 0.8% (v/v) goat anti-human IgG was used to treat the cells for 6 h to induce EBV into a lytic cycle, and then the cells were cultured in fresh medium for 3 days. Next, supernatants of the cells were centrifuged at 800×g for 30 min to remove cellular debris. Then, the supernatants were centrifuged at 50,000×g for 2 h to obtain EBV pellets. EBV pellets were resuspended in RPMI-1640 and used to infect NPC cells. EBV-infected CNE2 and HK1 cells were sorted by flow cytometry and cultured in RPMI-1640 medium using a various concentration (300–700 μg mL−1) of G418 (Sigma–Aldrich). EBV infection efficiency was verified by EBER detection using in situ hybridization (ISH) according to the manufacturer’s instructions of the ISH kit for EBER (Zhongshan **qiao Bio. Co.) or confocal laser-scanning microscopy (Olympus FV1000).

Constructs and CRISPR/Cas9-mediated EBNA1 deletion

For EBNA1 knockout cell line construction, the plasmid lentiCRISPR-v2-EBNA1 for EBNA deletion was kindly supplied by Prof. **ang (**ang Tong, Sun Yat-Sen University Cancer Centre) and has been previously described [44]. For lentiviral packaging, the lentiviral vector and packaging plasmids (pMD2.G and psPAX2) were both transfected into HEK293T cells; 48 h later, the virus in the supernatant was collected and used to infect EBV-positive cells. Puromycin (2 μg mL−1) was added to select stable EBNA1-deleted cells. GFP-expressing cells were detected to assess EBV clearance efficiency.

Cell death analysis

Cell death was detected using the SYTOX dead cell stain sampler kit (#S34862, Invitrogen) and the Annexin V-PI apoptosis assay kit (#556547, BD). Briefly, the indicated cells were plated in 12-well plates and treated with different concentrations of cytotoxic compounds for the indicated times. Then, SYTOX stain was added to the cell supernatants, and the cells were incubated at room temperature for 10 min, followed by observation and imaging using a fluorescence microscope. The Annexin V-PI apoptosis assay was quantified using flow cytometry.

Lipid ROS assay

Lipid ROS production was detected by flow cytometry using C11-BODIPY dye (#D-3861, Life Technologies, Grand Island, NY, USA) according to the manufacturer’s instructions. Briefly, cells were seeded into 12-well plates and cultured in a 37 °C incubator with 5% CO2. After the cells were treated with different cytotoxic reagents for the indicated times, C11-BODIPY was added to the cell supernatants, and the cells were cultured for more than 30 min before ROS detection. ROS can oxidize the polyunsaturated butadienyl portion of C11-BODIPY, and then the fluorescence emission peak of the dye shifts from ~590 nm to ~510 nm.

Oligonucleotides, plasmids, and stable cell lines

The pEZ-Lv105-puro-vector and GPX4 plasmids were purchased from GeneCopoeia, Inc., USA. The sequence of the human GPX4 gene was synthesized and cloned into the lentiviral plasmid pEZ-Lv105-puro-vector. The primers used for amplification were as follows:

GPX4-Forward 5′-ATGAGCCTCGGCCGCCTTTG-3′;

GPX4-Reverse 5′-CCCACAAGGTAGCCAGGGGT-3′.

GPX4 shRNA sequences were as follows:

shGPX4 1#: GCTACAACGTCAAATTCGA

shGPX4 2#: GAGGCAAGACCGAAGTAAA

To generate stably transfected cell lines, 293FT cells were cotransfected with the lentivirus packaging vector and shRNA knockdown plasmids. The lentiviral particles were subsequently harvested and used to infect NPC cells 48 h later. Stable clones were then selected using 1 mg/mL puromycin (Sigma–Aldrich), and real-time RT–PCR or western blot assays were used to validate the infection efficiency.

siRNA transfection and qRT–PCR

Effective siRNA oligonucleotides targeting TAK1 and NRF2 were obtained from RiboBio with the following sequences:

siTAK1-#1: GGAGTGGCTTATCTTCACA;

siTAK1-#2: GGCTTATCTTACACTGGAT;

siNRF2-1#: GAGAAAGAATTGCCTGTAA;

siNRF2-2#: GCTACGTGATGAAGATGGA;

siNRF2-3#: GCCCTCACCTGCTACTTTA.

The negative control (siCtrl) was nonhomologous to any human genome sequence and purchased from RiboBio Co., Ltd. Predetermined cells were grown in 6-well plates for 12 h and then transfected with 20 nM siRNA mixed with 5 μL of Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer’s instructions. Thirty-six to 48 h after transfection, the cells were harvested for further analysis.

Total RNA was extracted using TRIzol (Life Technology), reverse transcribed into cDNA using the GoScriptTM Reverse Transcription System (Promega) and analyzed by real-time qPCR on a BIO-RAD Real-time PCR machine using iTaq™ Universal SYBR® Green Super mix (Bio–Rad). The results were normalized to β-actin, and relative values were calculated using the 2[-(CT gene -CT reference)] method. The gene-specific primers used for qPCR are listed in Supplementary Table 3.

Western blot assay

Total protein was obtained using RIPA buffer (Beyotime Biotechnology) containing EDTA-free Protease Inhibitor Cocktail (Roche). Protein extracts were separated on 10–12% acrylamide SDS–PAGE and transferred to polyvinylidene fluoride membranes (Merck Millipore). The membranes were blocked in 5% non-fat milk and incubated with primary antibodies overnight at 4 °C. The membranes were then incubated with horseradish peroxidase-conjugated secondary antibodies (anti-mouse or anti-rabbit; 1:3,000; Thermo) at room temperature for 45 min. Finally, the target protein bands were detected using an enhanced chemiluminescence system (Thermo Fisher Scientific).

Coimmunoprecipitation (co-IP) and mass spectrometry

For the immunoprecipitation (IP) assay, HK1 cells transfected with the empty vector or Flag-GPX4 expression plasmid were lysed in IP lysis buffer. ANTI-FLAG® M2 Affinity Gel (Sigma, A2220-10 ML) was incubated with the lysates overnight at 4 °C and then washed and collected according to the manufacturer’s protocol. Mass spectrometry was performed by Huijun Biotechnology. For the co-IP assay, western blotting was performed to determine protein levels. Cells were rinsed in PBS and then lysed in IP lysis buffer. The protein extracts were subsequently incubated with the ANTI-FLAG® M2 Affinity Gel. The precipitated proteins were separated and detected by western blotting using the indicated antibodies. Finally, the blots were visualized using a chemiluminescence system.

GST pull-down assay

For the GST pull-down assay, recombinant human GPX4 protein expressed in Escherichia coli was purchased from Abcam company (ab82660), and GST-TAK1 (full length 1-606 aa or N-terminal 1-305 aa) was expressed in Rossita. GST-TAK1 or GST-vector and GPX4 were incubated in lysis buffer overnight at 4 °C. GST-agarose was added to lysis buffer, incubated for 4 h at 4 °C, washed four times with lysis buffer and analyzed by immunoblotting.

Immunofluorescence staining

For immunofluorescence staining, cells were grown on coverslips (Thermo Fisher Scientific). After 24 h, the cells were fixed in 4% paraformaldehyde, permeabilized with 0.1% Triton X-100 in PBS, and incubated with primary anti-NRF2 (1:200; Cell Signaling Technology (#12721)), anti-Flag (1:200; Sigma, F1804), anti-MAP3K7 (1:1000; Abcam, ab109502), anti-TAB1 (1: 400; Abcam, ab76412), and anti-TAB3 (1: 50; Abcam, ab124723) antibodies overnight at 4 °C. The coverslips were then incubated with Alexa Fluor 488- or 594-conjugated goat IgG secondary antibodies (1:1,000; Life Technologies; A-11008 or A-11001) and counterstained with 4′,6-diamidino-2-phenylindole (DAPI). Images were captured using a confocal laser-scanning microscope (Olympus FV1000).

IHC staining

Expression levels of GPX4 in patient tissues were detected by IHC in paraffin-embedded sections. Briefly, the sections were deparaffinized in xylene, rehydrated in a graded alcohol series, incubated in 3% hydrogen peroxide to block endogenous peroxidase activity, heated in antigen retrieval solution, subjected to FBS to block nonspecific binding, and incubated with GPX4 antibody (Abcam, ab125066, 1:100) at 4 °C overnight. Two experienced pathologists validated the scores of all sections.

Measurement of cell death, cell viability and lipid peroxidation

Cell death was analyzed using propidium iodide (Invitrogen, Waltham, MA, USA) or SYTOX Green (Invitrogen) staining followed by microscopy or flow cytometry. Viability was measured using the CCK-8 assay. To analyze lipid peroxidation, cells were stained with 5 μM BODIPY-C11 (Invitrogen) for 30 min at 37 °C followed by flow cytometric analysis. Lipid ROS-positive cells were defined as cells with fluorescence greater than 99% of the cells in the unstained sample.

CCK-8 assay

Cell growth was measured using a CCK-8 kit (Sigma, St Louis, MO, USA) according to the instructions. Briefly, approximately 1500 cells were plated into 96-well plates and cultured in the indicated medium. Cell proliferation measured at 450 nm (A450) was examined every day for five days according to the manufacturer’s protocol. All experiments were performed three times.

Cell cycle analysis and EdU incorporation analysis

The EdU incorporation assay was performed as previously described [45]. Briefly, 5 × 104 cells were seeded onto coverslips in 24-well plates. After reaching 80% confluence, EdU (20 μM) was added to the supernatants of cells and cultured for 1.5 h, followed by the click reaction according to the manufacturer’s instructions. Then, the cells were assessed under a confocal microscope to calculate the positive rate of EdU incorporation. The cell cycle was evaluated using propidium iodide (PI) staining and quantified using a Gallios flow cytometer.

Colony formation assay

Treated cells (400 cells/well) were plated in triplicate into 6-well plates and cultured for 10 days. The colonies were fixed in methanol for 15 min and stained with crystal violet for another 15 min. Finally, the colony number was quantified. All experiments were performed three times.

Statistical analysis

All data are presented as the means ± SD from at least three independent experiments. Two-tailed unpaired Student’s t test was used for statistical analysis involving two group comparisons (ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001). Data with a non-Gaussian distribution were compared using a two-tailed Mann–Whitney test, which was performed for the growth curves of the indicated xenografts. Differences with p < 0.05 were considered statistically significant. Statistical analyses were performed using GraphPad Prism (GraphPad Software, San Diego, CA). The Kaplan–Meier method and log-rank test were used to construct survival curves and compare the differences, respectively. Independent prognostic factors were determined by multivariate analysis using a Cox proportional hazards regression model. A chi-squared (χ2) test was performed to determine the correlation between EBV copy number and GPX4 expression. Statistical analyses were performed using SPSS 22.0 software (IBM).