Main

Epstein–Barr virus (EBV) is the first oncogenic herpesvirus that targets epithelial cells (ECs) and B lymphocytes1. It infects ~95% of the population worldwide2 and is associated with a spectrum of severe diseases, especially mononucleosis and various forms of cancer, including nasopharyngeal carcinoma (NPC), gastric cancer, colorectal cancer and B cell lymphoma3,4,5,6. Effective prevention of EBV infection is a crucial public health issue.

During EBV infection, viral glycoproteins collaborate with host envelope proteins to enable membrane fusion and EBV entry into target host cells7. Core to this process is the glycoprotein B (gB) homotrimer and the glycoprotein H/L (gH/gL) heterodimer, facilitating fusion in ECs and B lymphocytes8,9,10,11. Over the past several decades, extensive research has been conducted on B cell receptors involved in EBV entry11,12,13,14,15. As to ECs, integrins were initially identified to be the primary receptors for EBV entry16,17, but they were later confirmed to be accessory receptors18,19. More recently, following the discovery of Neuropilin 1 (NRP1) and Non-muscle myosin IIA (NMHC-IIA, also known as MYH9)20,47,Western blotting

Cell lysates were collected in RIPA buffer (Thermo Fisher, 89900) or cell lysis buffer and supplemented with a commercial protease inhibitor cocktail (Thermo Fisher, 78442) before use. Total protein was obtained by centrifugation and the protein concentration was determined by a Pierce BCA Protein Assay kit (Thermo Fisher, 23225). An equal amount of total protein was fractionated by 8～12% SDS–PAGE and transferred onto a 0.22 μm PVDF membrane (Millipore, ISEQ 00010). Membranes were immunoblotted with the indicated primary antibodies and then incubated with HRP-conjugated secondary antibody. Primary antibodies were as follows: mouse anti-IFITM1 (60074-1, Proteintech, 5B5E2, KD/KO validated, 1/1,000), rabbit anti-EphA2 (6997, CST, 1/1,000), rabbit anti-EGFR (18986-1, Proteintech, KD/KO validated, 1/1,000), rabbit anti-DDX5 (ab126730, Abcam, EPR7239, KD/KO validated, 1/1,000), rabbit anti-DDX6 (ab174277, Abcam, EPR12146, KD/KO validated, 1/1,000), rabbit anti-DDX17 (ab180190, Abcam, EPR13807(B), KD/KO validated, 1/1,000), rabbit anti-Actin (YT0096, ImmunoWay, 1/5,000) and rabbit anti-GAPDH (ab9485, Abcam, 1/5,000). Secondary antibodies were HRP-conjugated goat anti-rabbit IgG (SA00001-2, Proteintech, 1/5,000) and HRP-conjugated goat anti-mouse IgG (SA00001-1, Proteintech, 1/5,000). Blots were incubated with ECL substrate (BioRad, 1705061) and imaged with the ECL detection system (ChemiDoc, BioRad).

Flow cytometry

To determine EBV infection rates, 1 × 106 cells incubated with EBV were collected and washed using 1×PBS containing 0.2% bovine serum albumin (BSA). Cells were then resuspended in 300 μl of 1×PBS containing 0.2% BSA. Data were acquired using an LE-SA3800 Spectral Analyzer (Sony) and FlowJo software was used for analysis.

Indirect immunofluorescence assay

To detect the co-localization of IFITM1 and EphA2 or IFITM1 and EGFR, cells were plated on glass-bottom cell culture plates (801006, NEST) for 24 h. After brief washing twice with 1×PBS, cells were fixed with 4% paraformaldehyde (BioSharp) and permeabilized with 0.2% Triton X-100 (Thermo Fisher, 89900). Plates were washed gently and blocked with 3% normal goat serum (C0265, Beyotime). Subsequently, the cells were incubated with the primary antibody pairs overnight, followed by incubation with respective fluorophore-conjugated secondary antibodies (Alexa Fluor 488 goat anti-mouse IgG H&L and Alexa Fluor 647 goat anti-rabbit IgG H&L; 150113, 150079, Abcam) for 1 h and counterstained with DAPI (C1005, Beyotime) for 10 min at room temperature. To avoid false-positive cross-reactivity, primary antibody pairs were chosen from different species (mouse anti-IFITM1 and rabbit anti-EphA2, mouse anti-IFITM1 and rabbit anti-EGFR). Fluorescence images were recorded using the High Content Analysis System (CQ1, Yokogawa).

Co-immunoprecipitation

Immunoprecipitation was performed using an IP kit (abs955, Absin) following supplier instructions. Briefly, whole-cell proteins were isolated in lysis buffer and the lysate was centrifuged; the supernatant was subsequently collected. To lower the background, 500 µl of supernatant containing ~500–1,000 µg protein was incubated with 5 µl of protein A and 5 µl of protein G for 1 h at 4 °C. Total protein lysates that removed unspecific binding proteins were obtained after centrifugation. For protein binding, 1–5 µg of the corresponding antibodies (to IFITM1 or EphA2, as mentioned above) and homologous IgG (eBiosciences) was added to pre-cleaned protein lysates. Immediately afterwards, samples were incubated at 4 °C overnight under gentle rotation. To precipitate the target proteins, 5 µl of protein A and 5 µl of protein G were added to bind the antigen–antibody complexes; the reaction was maintained at 4 °C for 3 h. After gently washing three times with wash buffer, the unbound proteins were removed and the pellet (agarose–antibody–antigen complex) was resolved in 20–40 µl of SDS-loading buffer for further western blotting with the indicated antibodies. For exogenous co-IP of IFITM1, EphA2, gH/gL and gB, HEK293 cells were transfected with the corresponding combination of Myc-gH/gL, Myc-gB, pSMCV-IFITM1, pSMCV-EphA2 and empty vector (see figure legends for details).

Protein expression and purification

IFITM1, gB, gH/gL and EphA2 DNAs were constructed into pGEX6p-1-GST/His vector and recombinant fusion proteins were purified in the Rosetta strain of E. coli. Briefly, after the constructed plasmids were cloned into pET28a vector, E. coli Rosetta star pLysS cells were transformed with these plasmids. Colonies were inoculated into 25 ml of Luria-Bertani (LB) media containing ampicillin (100 μg ml−1) and grown for 16 h at 37 °C; then 25 ml cultures were transferred into 500 ml of fresh LB medium and grown at 37 °C for 4 h, adding isopropyl-β-d-thiogalactopyranoside (IPTG, 1 mM) to continue cultivation for 24 h at 16 °C. Bacterial cells were collected by centrifugation and 6×His- or GST-tagged target proteins were purified through His-tag or GST-tag affinity chromatography (Qiagen) and desalted using an Amersham column.

Competitive binding assays

First, 96-well microtitre plates were coated with 200 ng of GST-EphA2 overnight at 4 °C, followed by blocking with 5% BSA in 1×PBS for 2 h at room temperature. Then, gradient concentrations of 6×His-IFITM1 and 6×His-gB or 6×His-gH/gL proteins were added and incubated for 2 h at room temperature. After washing with 1×BST (0.05% Tween-20 in 1×PBS), the plate was incubated with a rabbit anti-GST antibody or mouse anti-His antibody (1:4,000 dilution) for 2 h at room temperature. The plate was then washed and incubated at 37 °C for 1 h with HRP-conjugated goat anti-rabbit antibody or HRP-conjugated goat anti-mouse antibody. After adding the substrate tetramethyl-benzidine, the absorbance was measured at 450 nm using the Nanodrop One spectrophotometer (Thermo Fisher).

3D structure and spatial binding site prediction

Two main methods, I-TASSER from the University of Michigan and SWISS-model from the European Center for Bioinformatics, were used to select the model consistent with the local structure published in ref. 41. After selecting a single predicted structure, the 3D structures of the complex formed by IFITM1, EphA2 and gH/gL were calculated. The top 100 models were selected from the 2,000 rigid docking models optimized by geometry and electrostatics. On the basis of the accurately predicted binding sites and side-chain flexibility of the binding interface amino acids, 200,000 models were generated, and the 10 best models were selected after clustering and scoring. After comparing with the model published in ref. 41, the first model was selected for subsequent calculations.

Site-specific mutation

IFITM1 mutant primers were designed by the homologous recombination method, and IFITM1 wild-type plasmids were amplified by PCR using mutant-specific primers and the KOD-Plus-Neo amplification kit (Toyobo). After mutation, the PCR products were recombined at 37 °C for 30 min. The top 10 competent cells were added to the recombinant products, and the transformation mixture was evenly spread onto an LB plate containing ampicillin and incubated overnight at 37 °C. Single clones were selected for Sanger sequencing.

MeRIP-seq and MeRIP−qPCR

m6A RNA immunoprecipitation was performed according to the Magna MeRIP m6A kit (17-10499, Merck) instructions. Total RNA was extracted by the Trizol method and then interrupted to form fragments of ~100 nt. Beads with anti-m6A antibodies were added and incubated at 4 °C for 8 h to form m6A RNA–antibody–magnetic bead complexes, which were adsorbed using a magnetic rack and washed several times to remove impurities. The m6A RNA was eluted by competitive binding and submitted for sequence analysis by RT−qPCR.

RIP-seq and RIP−qPCR

RNA immunoprecipitation (RIP) was performed according to the instructions of the Magna RIP RNA-Binding Protein Immunoprecipitation kit (Merck). Cells lysed by lysis buffer were incubated with anti-YTHDF3 antibody and magnetic beads at 4 °C for 8 h. The magnetic bead–antibody–target protein–RNA complex was adsorbed with a magnetic rack and cleaned 5 times with wash buffer to remove impurities. RNA was extracted by Trizol, and the purified RNA was analysed by RT−qPCR or submitted for sequence analysis.

Tandem affinity purification pull-down and mass spectrometry

Wide-type full-length YTHDF3 and its double m6A binding site defective mutants, designated YTHDF3WT and YTHDF3DM, were constructed into a tandem affinity purification vector pLVpuro-TAP (SBP-3HA). Then, three stably expressing TAP-Vector, TAP-YTHDF3WT OE and TAP-YTHDF3DM OE HK1 cell lines were subjected to 2 µg ml−1 puromycin selection for a week. The TAP affinities of the exogenously expressed TAP-Vector, -YTHDF3WT and -YTHDF3DM cells were extracted from the three cell lysates by excess streptavidin resin; then the endogenously expressed YTHDF3 partners were immunoprecipitated by YTHDF3 antibodies coupled to protein A/G beads. Finally, the YTHDF3 exogenously expressed TAP affinities and endogenously expressed immunoprecipitates were separately identified by mass spectrometry.

RNA stability assay

To assess IFITM1 RNA stability, cells were incubated with actinomycin (ActD) to terminate transcription. Briefly, HK1 cells were incubated with ActD (5 μg ml−1) for 0, 30 and 60 min, and collected. Total RNA was extracted and the IFITM1 RNA expression was determined by RT−qPCR.

Statistical analyses

Data are presented as mean ± s.e.m. derived from at least three independent experiments. All statistical analyses, including Pearson’s correlation coefficients (r), t-tests and so on were two-tailed and executed using GraphPad Prism 8. Although the data conformed to the assumptions of the statistical tests utilized, normality of data distribution was presumed but was not rigorously tested. In addition, the collection and analysis of data were carried out in a blinded manner relative to experimental conditions and no animal or data point was excluded from the analysis for any reason.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.