Human embryonic kidney 293T (HEK293T) cells, porcine kidney cells (PK15), and Sp2/0 myeloma cells were obtained from the American Type Culture Collection (ATCC, USA). The HEK293T cells were incubated in Dulbecco’s modified eagle medium (DMEM), supplemented with 10% fetal bovine serum (FBS) (Gibco, USA). The HEK293T and PK15 cells were maintained at 37°C with 5% CO₂. The eukaryotic expression plasmid pcDNA3.1-CD2v-HA containing a hemagglutinin (HA)-tagged full-length coding sequence of the EP402R gene (encoding CD2v) from the ASFV SY-1 strain (GenBank: OM161110.1; Chinese strain; genotype II ASFV) was obtained from GenScript Biotech Co, Ltd. (Nanjing, China). The pEGFP-N1-CD2v-GFP plasmid containing a green fluorescent protein (GFP)-tagged full-length coding sequence of the EP402R gene was constructed by our laboratory based on pcDNA3.1-CD2v-HA. From Beijing Vital River Laboratory Animal Technology Co., Ltd (Beijing, China), we purchased female specific-pathogen-free BALB/c mice. The animal experiments were conducted in accordance with national guidelines and were approved by the Ethics Committee of the Academy of Beijing Institute of Pharmacology and Toxicology (IACUC-DWZX-2021–621).

We purchased restriction endonucleases XbaI and NotI from New England Biolabs Inc. (USA). Roswell Park Memorial Institute (RPMI) 1640 medium, DMEM, MEM, hypoxanthine–aminopterin–thymidine (HAT) selective medium, An Expi293 expression system kit (including Expi293F cells, Expi293 expression medium, ExpiFectamine 293 transfection kit, and Opti-MEM I antiserum medium), Alexa 594-labeled Goat anti-mouse IgG, and mouse IgG (mIgG) as the isotype control, etc. were purchased from Thermo Fisher Scientific (USA). Freund’s complete adjuvant, Freund’s incomplete adjuvant, tunicamycin, and Triton X-100 were purchased from Sigma-Aldrich (USA). Enhanced chemiluminescence (ECL) solution was purchased from Cytiva Co., Ltd. (USA). Radioimmunoprecipitation assay (RIPA) cell lysis buffer was purchased from Solarbio technology Co., Ltd, (Beijing, China). Human and mouse CD2 with His tag (hCD2-His and mCD2-His) were purchased from ACROBiosystems (USA). Recombinant ASFV-P30, P72 and pp62 protein with His tag were sourced from East-Mab Bio Technology (Suzhou, China). Rabbit anti-6×HA tag antibody was purchased from R&D Systems (USA). Rabbit anti-His tag antibody was purchased from Abcam (USA). Horseradish peroxidase (HRP)-labeled goat anti-mouse and anti-rabbit IgG secondary antibodies and 4′,6­diamidino­2′­phenylindole (DAPI) were purchased from ZSGB Biotechnology Co., Ltd. (Beijing, China). Rabbit antibodies against phosphorylated NF-κB p65 (p-NF-κB p65), NF-κB p65, glyceraldehyde 3-phosphate dehydrogenase (GAPDH), and β-actin were purchased from Cell Signaling Technology Inc. (USA).

To construct the eukaryotic expression plasmid expressing the CD2v extracellular domain (amino acids 16–206), we amplified the target fragment using the pcDNA3.1-CD2v-HA plasmid as a template. To enhance the expression and secretion efficiency, we replaced the original signal peptide (amino acids 1–15) with a signal peptide sequence derived from IL-10 (MHSSALLCCLVLLTGVRA). Additionally, to enable convenient downstream purification, we introduced a C-terminal 6×His tag. After digestion with the NotI and XbaI restriction enzymes, the gene fragment corresponding to the CD2v extracellular domain was directly cloned into the pcDNA3.1(+) eukaryotic expression plasmid, which was designated pcDNA3.1-N-CD2v-His. The recombinant expression plasmid was verified by double enzyme digestion and sequencing.

Expression of the CD2v extracellular domain fusion protein (N-CD2v-His) was conducted following the guidelines provided in the Expi293 expression system kit (Thermo Fisher Scientific, USA). Briefly, a 50-mL cell culture was prepared in a 125-mL cell shake flask. Expi293F cells were added to the flask and diluted to a concentration of 3 × 10⁶ cells/mL. The expression plasmid (pcDNA3.1-N-CD2v-His) and ExpiFectamine 293 transfection reagents were diluted with Opti-MEM. The transfection reagents were mixed with the diluted plasmid and added to the flask. After incubation for 18–22 h, ExpiFectamine 293 transfection enhancers were added. The cells were then cultured for an additional 5–7 days. The culture supernatant, which contained N-CD2v-His, was filtered using a 0.45-μm filter membrane. Purification of the protein was carried out using a Ni-NTA His Bind protein purification kit (GenScript Biotech Co, Ltd., Nanjing, China). Following elution, the protein was concentrated by ultrafiltration, and its concentration was determined using a bicinchoninic acid (BCA) protein concentration determination kit (Thermo Fisher Scientific, USA). Finally, the purified protein was analyzed using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), followed by detection using western blotting with a rabbit anti-His tag antibody.

Anti-CD2v mAbs were generated following a previously described method (41). Anti-CD2v mAbs were generated as described (15). Female BALB/c mice were subcutaneously immunized with 100 µg purified N-CD2v-His protein per mouse in combination with Freund’s adjuvant, followed by two biweekly booster immunizations using the same dose and incomplete adjuvant. Antibody titer in serum were measured by ELISA. Mice received a final hyperimmunization dose with 100 µg/mouse intraperitoneally, and were euthanized 3 days later. Splenocytes were fused with Sp2/0 myeloma cells using PEG 1500. Positive hybridomas were subcloned and high-producing lines were injected into mice to induce ascites. Purification was achieved using a HiTrap Protein G HP column (GE Healthcare, Cytiva, USA).

To perform mAb subclass analysis, a mouse antibody isotype detection ELISA kit (Thermo Fisher Scientific, USA) was used following the manufacturer’s instructions. Hybridoma cell total RNA was isolated using an RNeasy Plus Micro Kit (QIAGEN, Germany) as instructed by the manufacturer. The isolated RNA was then reverse transcribed into cDNA using SMARTScribe Reverse Transcriptase (TaKaRa, Japan) according to the manufacturer’s instructions. Heavy and light chain antibody fragments were amplified using a previously reported protocol (42) and sent to GenScript Biotech Co., Ltd. (Nanjing, China) for sequencing. The complementarity-determining region 3 (CDR3) sequences were verified using the IMGT online software (https://imgt.org/IMGT_vquest/input).

ELISA plates were coated with 0.2 μg/well of N-CD2v-His and incubated overnight at 4°C. The plates were then blocked with 4% skim milk in PBST (PBS with 0.1% Tween-20) at 37°C for 1 hour. Gradient dilutions of purified mAbs and a negative control mouse IgG isotype were added to the wells and incubated at 37°C for 2 hours. After washing with PBST, HRP-labeled goat anti-mouse IgG secondary antibody (1:3000 dilution) was added and incubated at 37°C for 45 minutes. Tetramethylbenzidine (TMB) solution (Kangwei Century Biotechnology Co., Ltd., Beijing, China) was then added for visualization. The optical density (OD) at 450 nm was measured using an ELISA microplate reader.

To determine the affinity of anti-CD2v mAbs, we utilized a ForteBio-Octet molecular interaction analyzer with the BLI technique. First, N-CD2v-His was immobilized onto Ni-NTA biosensors. Then, the mAbs were diluted and added to the biosensors to bind to N-CD2v-His. The binding and dissociation times were 200 and 400 s respectively. To regenerate the biosensor surfaces, a pH 1.7 solution of 10 mM Gly-HCl was applied in 5-s pulses, repeated three times. The data obtained were analyzed using a 1:1 binding model to calculate the equilibrium dissociation constant (K_D).

HEK293T cells were transfected with pcDNA3.1-CD2v-HA or empty plasmid pcDNA3.1 (control), using jetPRIME transfection reagent (Polyplus, France) for 12 h. The cells were then incubated with or without 10 μg/ml tunicamycin for 24 h (22, 43). To extract the cellular proteins, RIPA buffer containing a protease inhibitor cocktail (R&D, USA) was used for cell lysis. The protein lysates were separated by SDS-PAGE on 12% gels and transferred onto polyvinylidene fluoride (PVDF) membranes. Subsequently, the membranes were submerged in blocking buffer (5% skim milk in PBST) for 1h at 37°C, incubated with mAbs or a rabbit anti-HA tag primary antibody (positive control) at 4°C overnight, washed with PBST, and then incubated for 1 h at room temperature with HRP- labeled goat anti-rabbit IgG secondary antibody (diluted in blocking buffer). After washing with PBST again, the bound antibodies were detected using an ECL western blotting detection system.

HEK293T cells were inoculated in 35mm confocal dishes (1 × 10⁵ cells/dish). When the cells grew to about 70% confluence, a pEGFP-N1-CD2v-GFP plasmid or an empty pEGFP-N1 plasmid (negative control) were transfected into the cells. After 36 h, 4% paraformaldehyde was added for fixation for 24 h at 4°C. Next, absorbed paraformaldehyde and dialysate (1% bovine serum albumin [BSA]+0.1% Triton X-100/PBS buffer) were added for permeabilization for 20 min at room temperature. Dialysate containing 5% goat serum was then added and incubated for 1 h at 37°C. Thereafter, 1 mg/mL of each of the selected mAbs was separately added and incubated overnight at 4°C. Alexa 594-labeled goat anti-mouse IgG secondary antibody was added and incubated at room temperature in the dark for 1 h. Next, 1 drop of DAPI was added and the fluorescence localization in cells was observed by confocal microscopy.

To evaluate the overlapping epitopes recognized by different anti-CD2v mAbs, competitive ELISA was conducted. The selected mAbs (2B25, 3J25, and 8G1) were biotinylated using an EZ-Link™ Sulfo-NHS-Biotinylation kit (ThermoFisher, USA). N-CD2v-His protein (0.2 µg/well) was coated onto 96-well ELISA plates overnight at 4°C. The plates were then blocked with 4% skim milk in PBST at 37°C for 1 hour. Varying concentrations (35.2 ng/mL to 36 μg/mL) of a competitor mAb (2B25, 3J25, or 8G1) and 1 μg/mL of a biotinylated detection mAb (2B25-biotin, 3J25-biotin, or 8G1-biotin) were added to each well and incubated at 37°C for 1 hour. After washing, streptavidin–HRP conjugate (1:4000 dilution, Thermo Fisher, USA) was added and incubated at 37°C for 45 minutes. TMB solution was added, and the reaction was stopped by adding H2SO4. The OD at 450 nm was measured as before.

To develop a sensitive sandwich ELISA for CD2v detection, the reactivity of the anti-CD2v mAbs as capture and detection antibodies was evaluated. Each mAb (1 μg/mL) was coated separately onto an ELISA plate. Serially diluted N-CD2v-His protein (1.9 ng/mL to 2 μg/mL) was added to the plate. After washing, each biotinylated mAb (1 μg/mL) was added to the appropriate wells. After washing again, streptavidin–HRP conjugate (1:8000 dilution in blocking buffer) was added and incubated at 37°C for 45 minutes. TMB solution was added, and the reaction was stopped using 1N H₂SO₄. The OD at 450 nm was measured.

Based on the sequence characteristics of the CD2v extracellular region, obtained from the Swiss Model database (https://swissmodel.expasy.org), we utilized secondary structure discrimination and potential epitope prediction analysis to design, express, and purify five truncation mutants (CD2v D1–D5) of the CD2v extracellular domain. To identify the CD2v regions recognized by mAbs, we initially conducted an indirect ELISA. Based on these results, we designed 11 overlapping peptides that spanned the CD2v extracellular domain (specifically, D1 and D4). To facilitate better coupling, cysteine residues were added at the C-terminus of the shorter peptides. These peptides were synthesized and conjugated to Bovine Serum Albumin (BSA) by Chinese Peptide (Hangzhou, China), ensuring their purity was confirmed to be ≥95%. To further investigate the binding sites, a peptide-based ELISA was performed. The 96-well plates were coated with the peptides (0.1 μg/well) at 4°C overnight and then blocked with 4% skim milk at 37°C for 1 h. Each mAb (0.1 μg/well) was added separately and incubated for 37°C for 1 h. Next, an HRP-labeled goat anti-mouse IgG secondary antibody (1:3000 dilution) was added and incubated at 37°C for 45 min. Subsequently, the plates were processed following the same steps as described above for indirect ELISA. To construct the theoretical 3D structure of the CD2v extracellular domain, we used the model with the highest sequence similarity (PDB code: 2dru). The optimized model was visualized using PyMol software.

To evaluate the conserved nature of epitope sequences among different ASFV strains, 15 Chinese epidemic strains and 27 representative globally circulating strains (comprising 5 genotypes) were selected, with strain information presented in Supplementary Table S1 . The conservation of mAb-binding CD2v epitope amino acid sequences in different ASFV strains was assessed by multiple sequence alignment in Jalview v2.11.3.2 (44). A maximum likelihood phylogenetic tree was constructed based on the CD2v amino acid sequence of the 42 representative ASFV strains in MEGA v11.0 (45) and visualized utilizing ggtree v3.10.0 (46).

In a time course experiment, PK-15 cells were incubated with 500 ng of N-CD2v-His and PBS (negative control) for 15–120 minutes to study CD2v-induced NF-κB activation. Cell lysis was performed using RIPA buffer containing a protease and phosphatase inhibitor cocktail (R&D, USA), followed by western blotting using specific rabbit primary antibodies to detect phosphorylated NF-κB p65 (p-NF-κB p65), total NF-κB p65, and the reference protein GAPDH. To assess the inhibitory activity of mAbs on CD2V-induced NF-κB signaling activation, different concentrations of each mAb (2–16 μg) were preincubated with 500 ng of N-CD2v-His protein for 30 min at 37°C. The mAb and CD2v mixture were then added to PK-15 cells and incubated for 90 min at 37°C. After cell lysis, western blotting was performed to detect phosphorylated NF-κB p65, total NF-κB p65, and the reference protein β-actin using specific rabbit primary antibodies.

Statistical analyses were conducted using Prism v8.0 (GraphPad, San Diego, CA, USA). The experimental data are presented as mean ± standard deviation (SD). Unpaired t-tests were used for statistical analyses, and differences were considered significant at p < 0.05.

Anti-CD2v mAbs were generated using the hybridoma technique through a series of selection and cloning processes ( Figure 1A ; Supplementary Figure S1 ). Initially, the recombinant CD2v extracellular domain was purified, and then SDS-PAGE and western blotting confirmed that it was highly glycosylated, with an approximate molecular weight of 55–80 kDa, with the main band at around 70 kDa ( Figures 1B, C ). This is consistent with previous findings (22, 23, 43). Subsequently, five BALB/c mice were immunized with the purified protein, and serum samples were collected for antibody titer detection using indirect ELISA. The serum from all five mice exhibited high titers against the N-CD2v-His glycosylated protein, indicating strong immunogenicity ( Figure 1D ). Hybridoma cell lines were generated by fusing mouse splenocytes from the three mice with the highest titers with Sp2/0 cells. Initially, a total 158 hybridoma cell lines that showed affinity to CD2v were obtained. Following two rounds of subcloning, sequencing analysis, and the exclusion of clones with repetitive sequences, we carefully selected four hybridoma cell lines (2B25, 3J25, 7B1, and 8G1) based on their superior binding affinity for further investigation. The purified antibodies were identified using SDS-PAGE ( Supplementary Figure S2 ). The isotypes of mAbs were determined to be IgG2b for 2B25 and IgG1 for the other three mAbs. Moreover, all antibodies had kappa light chains ( Table 3 ; Supplementary Figure S3 ). The V gene subclasses and the CDR3 sequence for each mAb are presented in Table 3 .

Using indirect ELISA, the N-CD2v-His binding activity of the mAbs was confirmed. All four mAbs (2B25, 3J25, 8G1, and 7B1) showed dose-dependent binding, with half maximal effective concentration (EC₅₀) values of 33.71, 56.32, 59.23, and 539.10 ng/mL, respectively. The mouse IgG isotype control antibody exhibited no binding ( Figure 2A ). To assess the specificity of the four mAbs for CD2v, known for its high similarity to CD2 found in T cells (18), their binding to hCD2 and mCD2 was evaluated using ELISA. The results clearly revealed that all four mAbs specifically bound to N-CD2v-His but no binding to hCD2-His or mCD2-His ( Figure 2B ). Furthermore, we investigated the binding interactions of the four mAbs with ASFV structural proteins P30, P72, and pp62, respectively. The results revealed that 2B25, 3J25, and 8G1 did not exhibit significant binding to any of the three ASFV structural proteins. Conversely, antibody 7B1 showed no binding to P30 and pp62, but displayed partial cross-reactivity with P72 ( Figure 2C ). These findings indicate that the screened mAbs, 2B25, 3J25, and 8G1, are specifically targeted towards the CD2v antigen.

The affinity of the four mAbs was evaluated using BLI (47, 48). The results demonstrated favorable association and dissociation curves for all four mAbs with the CD2v antigen at the nanomolar level ( Figure 3 ). The results demonstrated that 3J25 exhibited the lowest K_D value (9.94 nM), indicating the highest affinity among the four mAbs, while 7B1 exhibited the highest K_D value (32.7 nM), indicating the lowest affinity. In the subsequent experiments, we primarily focused on the three mAbs with higher affinity to the CD2v antigen (2B25, 3J25, and 8G1).

In the western blotting assay of the binding of the three mAbs (2B25, 3J25, and 8G1) to exogenous CD2v-HA, HEK293T cells were transfected with a pcDNA3.1-CD2v-HA plasmid encoding the full-length CD2v protein fused with a C-terminal HA tag ( Figure 4A ). The results revealed that all three mAbs effectively recognized the glycosylated CD2v-HA full-length protein (about 100 kDa), but only 3J25 and 8G1 recognized the deglycosylated CD2v protein (about 43 kDa). Notably, none of the mAbs recognized the intracellular C-terminal region of CD2v (26 kDa) ( Figure 4A ). These findings highlight the dependence of the recognition of CD2v by 2B25 on the presence of the glycosylated epitope on the surface of CD2v. Furthermore, the findings suggest that the three of mAbs could target linear epitopes on CD2v.

In the IFA of the cellular localization of CD2v, we transfected HEK293T cells with a pEGFP-N1-CD2v-GFP plasmid expressing full-length CD2v with a C-terminal GFP tag. The results showed that CD2v-GFP exhibited green fluorescence and predominantly localized to the cytoplasmic and perinuclear regions, as well as the cytomembranes ( Figure 4B ). This localization pattern confirms previous findings (22, 23, 25, 43). Simultaneously, the mAbs, employed as primary antibodies, displayed red fluorescence upon detection using Alexa 594-labeled secondary antibody. Notably, there was significant colocalization between the red signal from the mAbs and the green signal from CD2v-GFP in the merged image ( Figure 4B ). These results provide evidence that all three mAbs specifically recognize the naturally spatially structured CD2v-GFP protein expressed in HEK293T cells.

Competitive ELISA was employed to elucidate the relationships of the epitopes recognized by the three mAbs. N-CD2v-His was immobilized on ELISA plates, followed by adding an mAb (2B25, 3J25, or 8G1; competitor antibody) at varying concentrations and then a biotinylated mAb (2B25-biotin, 3J25-biotin, or 8G1-biotin; detection antibody), allowing for competition for binding to CD2v. As the concentration of 2B25 increased, the binding of 2B25-biotin to CD2v gradually decreased but minimal changes were observed regarding the binding of 3J25-biotin and 8G1-biotin ( Figure 5A ). This indicates that the epitopes recognized by 2B25 do not overlap with those recognized by 3J25 and 8G1. Similar results were obtained when using 3J25 or 8G1 as the competitor antibody ( Figures 5B, C ). Therefore, it can be concluded that 2B25, 3J25, and 8G1 do not exhibit epitope overlap. To establish a sandwich ELISA, an orthogonal experiment was conducted with these mAbs ( Figures 5D, E ). By utilizing 3J25 as the capture antibody and 2B25-biotin as the detection antibody, successful detection of viral ASFV-CD2v antigen at ng/mL levels was achieved.

Based on the sequence characteristics of the CD2v extracellular region ( Figures 6A, B ) and using secondary structure discrimination and potential epitope prediction analysis, we designed, expressed, and purified five CD2v extracellular domain truncation mutants (CD2v D1–D5; Figure 6C ; Supplementary Figure S4 ). Indirect ELISA ( Figures 6D-F ) showed that 2B25 and 3J25 mainly recognized the D1 region. However, at high CD2v concentrations (1 μg/mL), there was a slight binding of 3J25 to the D1 region compare the negative control; at low CD2v concentrations (0.008 μg/ml), there was a slight lower recognition of the D1 region by 3J25 than 2B25. This indicates that 2B25 and 3J25 have subtle differences regarding their recognition of the fine epitopes within the N-terminal D1 region of CD2v. Additionally, 8G1 predominantly recognized the D4 region, which is different from the other two mAbs.

To distinguish the epitopes recognized by these three mAbs, we designed 11 overlapping polypeptides spanning the D1 and D4 regions ( Figure 7 ). The results revealed that 3J25 primarily recognized the P2 region, while 8G1 recognized the P14 region. Intriguingly, 2B25 did not exhibit binding affinity towards any of the peptides tested, indicating its potential recognition of glycosylated epitopes due to the absence of glycosylation modifications in the peptides. Thus, through epitope analysis, we have identified two novel linear B-cell epitopes (²⁵TIILDSNITNDNN³⁷ and ¹⁴¹LNINDTFVKYTNE¹⁵³). Furthermore, when analyzing the conservation of these epitopes among 42 epidemic strains of ASFV ( Supplementary Table S1 ), we observed high conservation within Chinese epidemic strains and Georgia 2008/1 strains, but differences compared to Benin97/1, Nu1979, E75, BA71V, K-49 strains, among others ( Figure 8 ). Overall, our results underscore distinct epitope recognition patterns exhibited by the three mAbs.

CD2v has been shown to induce activation and nuclear translocation of NF-κB p65 in swine peripheral blood mononuclear cells and macrophages (25). To assess the effects of the mAbs on the NF-κB signaling pathway, we conducted a time course experiment to determine the duration regarding CD2v-induced NF-κB activation. Treatment with purified N-CD2v-His protein resulted in the phosphorylation of NF-κB p65 (70 kDa), which exhibited a continuous increase from 15 to 120 min ( Figure 9A ). Based on these findings, we selected a 90-minute incubation period for PK-15 cells with purified N-CD2v-His protein, either alone or in combination with the anti-CD2v mAbs (2B25, 3J25, or 8G1). All three mAbs exhibited significantly and dose-dependently inhibition of CD2v-mediated NF-κB phosphorylation in PK-15 cells ( Figures 9B-D ). Furthermore, when used individually, neither the three mAbs nor the mIgG isotype control were capable of stimulating NF-κB activation ( Supplementary Figure S5 ). These findings indicate that the anti-CD2v mAbs specifically interfered with the ability of soluble CD2v to induce NF-κB p65 activation in PK-15 cells.