Epstein–Barr virus envelope glycoprotein 110 inhibits NF-κB activation by interacting with NF-κB subunit p65

Abstract

Epstein–Barr virus (EBV) is a member of the lymphotropic virus family and is highly correlated with some human malignant tumors. It has been reported that envelope glycoprotein 110 (gp110) plays an essential role in viral fusion, DNA replication, and nucleocapsid assembly of EBV. However, it has not been established whether gp110 is involved in regulating the host's innate immunity. In this study, we found that gp110 inhibits tumor necrosis factor α–mediated NF- κB promoter activity and the downstream production of NF- κB-regulated cytokines under physiological conditions. Using dual-luciferase reporter assays, we showed that gp110 might impede the NF-κB promoter activation downstream of NF-κB transactivational subunit p65. Subsequently, we used coimmunoprecipitation assays to demonstrate that gp110 interacts with p65 during EBV lytic infection, and that the C-terminal cytoplasmic region of gp110 is the key interaction domain with p65. Furthermore, we determined that gp110 can bind to the N-terminal Rel homologous and C-terminal domains of p65. Alternatively, gp110 might not disturb the association of p65 with nontransactivational subunit p50, but we showed it restrains activational phosphorylation (at Ser536) and nuclear translocation of p65, which we also found to be executed by the C-terminal cytoplasmic region of gp110. Altogether, these data suggest that the surface protein gp110 may be a vital component for EBV to antagonize the host's innate immune response, which is also helpful for revealing the infectivity and pathogenesis of EBV.

The signal cascade responses initiated by the innate immune system begin with recognizing pathogen-associated molecular patterns by different pattern recognition receptors. When the host is subjected to pathogen invasion, a class of receptor molecules on the cell surface can sense the pathogens, then provoke a series of signaling pathways to activate interferon regulatory factor 3 (IRF3), IRF7, and/or NF-κB. Finally, the expression of interferon-stimulated genes and proinflammatory factors are upregulated (1). NF-κB is a substantial factor in gene transcription regulation relevant to immunity, inflammatory response, and cell differentiation (2, 3).

Epstein–Barr virus (EBV) is a member of the lymphotropic virus genus of Herpesviridae. The virus particle is spherical and has a diameter of 120 to 180 nm with a fundamental structure that is the same as other herpes viruses, principally composed of double-stranded linear DNA, stereosymmetric capsid, and glycoproteins decorated envelope (4, 5). EBV is primarily transmitted through saliva into oropharyngeal epithelial cells, and then the proliferative virus invades B lymphocytes and induces infectious monocytosis in children and youngsters (6). The EBV can be carried for life after primary infection and maintains latent in human lymphatic tissues for a long time (6, 7), which is associated with many diseases, including Hodgkin's disease, nasopharyngeal carcinoma, gastric cancer, childhood lymphoma, and other human malignant tumors (7, 8, 9).

Envelope glycoproteins are conserved in members of the herpesvirus family, and they are essential in many vital processes, including specific binding to the cell surface, the fusion of the viral envelope with the cell membrane, assembly, and release of virions (10). For example, Kaposi's sarcoma–associated herpesvirus gB is indispensable for virus–cell fusion (11). Pseudorabies virus gB is required for virus entry into the target cell and transmission between cells (12, 13). Herpes simplex virus 1 (HSV-1) gB is also essential for virus–cell fusion (14, 15, 16), and it can stimulate the NF-κB signaling pathway by interplaying with Toll-like receptor 2 (TLR2) (17, 18, 19).

gp110, also known as gB, the gene product of EBV BALF4, is the glycoprotein recognized by the lymphocyte surface EBV receptor, the type 2 complement receptor, CR2 (CD21) (20), which takes significant roles in the process of viral fusion with the cell membrane (21, 22, 23, 24) and replication and assembly of the virion (10, 25). However, it is unknown whether gp110 is also involved in regulating the host's innate immunity. In the present study, it is the first time that the surface protein gp110 was found to abolish NF-κB activation and downstream cytokine production by inhibiting the phosphorylation (Ser536) and nuclear translocation of p65 via its C-terminal cytoplasmic region interaction with p65.

Results

gp110 and its C-terminal cytoplasmic region inhibit tumor necrosis factor-α–mediated NF-κB promoter activity

The classical NF-κB signaling pathway can be triggered by many stimuli. Among them, tumor necrosis factor α (TNF-α) is an important stimulating factor (26). To assess whether gp110 possesses the ability to inhibit the canonical NF-κB, human embryonic kidney 293T (HEK293T) cells were cotransfected with NF-κB-Luc reporter, internal reference plasmid, and gp110-HA or BGLF4-HA expression plasmid, and cells were treated without or with TNF-α, then dual-luciferase reporter (DLR) was performed. Compared with the positive control BGLF4 (27), TNF-α stimulation strongly promoted NF-κB promoter activity, but this activation was constrained in the presence of gp110-HA (Fig. 1A). Moreover, gp110 could dose-dependently hamper TNF-α-mediated NF-κB activity (Fig. 1B).

Figure 1.

gp110 and its functional domain inhibit TNF-α-mediated NF-κB promoter activation.A, HEK293T cells were cotransfected with gp110-HA expression plasmid (500 ng), HA vector (500 ng), or positive control BGLF4-HA expression plasmid (500 ng), together with a pNF-κB-Luc reporter plasmid (100 ng) and pRL-TK reference plasmid (10 ng). After transfection for 24 h, cells were stimulated without or with TNF-α (20 ng/ml) for 6 h, then DLR was performed. B, HEK293T cells were cotransfected with different concentrations (500 and 1000 ng) of gp110-HA expression plasmid or HA vector (500 ng), along with a pNF-κB-Luc reporter plasmid (100 ng) and pRL-TK reference plasmid (10 ng), then DLR was followed as indicated in A. C, the structural domain diagram of gp110 and its truncated mutant expression plasmid construction. Numbers indicate the amino acid position. gp110(1–685)del means deletion of the gp110 N-terminal extracellular region, gp110(686–753)del means deletion of the gp110 hydrophobic region in the intermediate region, and gp110(754–857)del means deletion of the gp110 C-terminal cytoplasmic region. D, HEK293T cells were cotransfected with different truncated mutant expression plasmids of gp110 (500 ng), HA vector (500 ng), or BGLF4-HA expression plasmid (500 ng), along with a pNF-κB-Luc reporter plasmid (100 ng) and pRL-TK reference plasmid (10 ng), then DLR was followed as indicated in A. The expression of gp110 or BGLF4 was probed by WB using the indicated Abs, and β-actin was applied to verify equal loading of protein in each lane. Data were shown as means ± SD from three independent experiments. Statistical analysis was performed using the Student's t test. ns, not significant; ∗∗p < 0.01; ∗∗∗p < 0.001; and ∗∗∗∗p < 0.0001. Ab, antibody; DLR, dual-luciferase reporter; gp110, glycoprotein 110; HEK293T, human embryonic kidney 293T cell line; TNF-α, tumor necrosis factor α; WB, Western blot.

To identify the key functional domain of gp110 for inhibiting NF-κB activity, three truncated mutant expression plasmids of gp110, gp110(1–685)del, gp110(686–753)del, and gp110(754–857)del, were constructed (Fig. 1C), which contains a deletion of the N-terminal extracellular region, deletion of the hydrophobic region in the intermediate region, and deletion of the C-terminal cytoplasmic region, respectively. Then, DLR was performed in the presence of these gp110 mutant expression plasmids. As a result, gp110(1–685)del and gp110(686–753)del, but not gp110(754–857)del, could inhibit NF-κB promoter activity (Fig. 1D), suggesting the C-terminal cytoplasmic region of gp110 plays a vital role in this process.

Knockdown of gp110 promotes NF-κB activity during EBV lytic infection

To further verify the inhibitory effect of gp110 mentioned previously, three expression plasmids of shBALF4 (shBALF4-1, shBALF4-2, and shBALF4-3) that specifically target gp110 were first constructed, then pSuper-Retro vector, shRadom, or shBALF4 was cotransfected with gp110-HA expression plasmid into HEK293T cells to analyze the interference efficiency of shBALF4. Compared with the pSuper-Retro vector and shRadom, all the plasmids of shBALF4 could efficiently downregulate the expression of gp110-HA, and shBALF4-2 had the highest knockdown effect of gp110-HA, but shBALF4-1 and shBALF4-3 had a relative low knockdown effect (Fig. 2A). Therefore, shBALF4-2 was used in the following knockdown-related experiments. Subsequently, the expression of gp110 during EBV lytic infection induced in EBV-positive Hone1 cells was detected by Western blot (WB), and the result found that the expression of gp110 reached a peak at 24 h after EBV reactivation (Fig. 2B), confirming the latent EBV was successfully induced into a lytic infection. Then, DLR was performed as indicated previously in EBV-positive Hone1 cells, except for shBALF4-2 was used to downregulate the expression of gp110, and 12-O-tetradecanoylphorbol-13-acetate (TPA) and sodium butyrate (NaB) were applied to induce the lytic cycle of EBV. As shown in Figure 2C, gp110 was efficiently expressed after TPA and NaB treatment, again confirming EBV was successfully induced into a lytic infection. Compared with the pSuper-Retro vector, stimulation of TPA and NaB activated NF-κB in EBV-positive Hone1 cells, and this activation increased when gp110 was knocked down after EBV lytic infection (Fig. 2C), indicating gp110 certainly inhibits NF-κB activity in the course of EBV lytic infection.

Figure 2.

Knockdown of gp110 during EBV lytic infection increases the activity of NF-κB.A, HEK293T cells were cotransfected with the expression plasmids of gp110-HA (1000 ng) and pSuper-Retro vector, shRandom, shBALF4-1, shBALF4-2, or shBALF4-3 (1 μg). At 24 h post-transfection, cell lysates were harvested, and WB analysis was carried out. The expression of gp110-HA was detected by anti-HA mAb, and β-actin was used as a loading control. B, EBV-positive Hone1 cells were treated without or with TPA (40 ng/ml) and NaB (3 mM) for 0, 2, 4, 8, 12, and 24 h to induce EBV lytic infection, and then the expression of gp110 was detected by WB using mouse anti-gp110 mAb, and β-actin was applied to verify equal loading of protein in each lane. C, the pSuper-Retro vector or shBALF4-2 expression plasmid (1000 ng) was cotransfected with a pNF-κB-Luc reporter plasmid (100 ng) and pRL-TK internal reference plasmid (10 ng) into EBV-positive Hone1 cells for 24 h. Then, luciferase activity was measured after 24 h stimulation without or with TPA (40 ng/ml) and NaB (3 mM) to induce EBV reactivation. The expression of gp110 was detected to show the lytic cycle of EBV by mouse anti-gp110 mAb, and β-actin was applied to verify equal loading of protein in each lane. Data were shown as means ± SD from three independent experiments. Statistical analysis was performed using the Student's t test. ∗∗∗∗p < 0.0001. EBV, Epstein–Barr virus; gp110, glycoprotein 110; HEK293T, human embryonic kidney 293T cell line; mAb, monoclonal antibody; NaB, sodium butyrate; TPA, 12-O-tetradecanoylphorbol-13-acetate; WB, Western blot.

gp110 suppresses the production of NF-κB-regulated cytokines

To continue to determine whether gp110 downregulates NF-κB-regulated cytokines, HeLa cells were transfected with the positive control BGLF4 expression plasmid (27), gp110-HA expression plasmid or HA vector, then cells were treated without or with TNF-α, and quantitative PCR (qPCR) was performed to measure the mRNA expression of gp110, BGLF4, and cytokines interleukin (IL)-1β, IL-6, and IL-8. Compared with the positive control BGLF4, TNF-α induced a high level of diverse cytokines. However, the mRNA expression of these cytokines was decreased in the presence of gp110 (Fig. 3A). To further confirm this result, the expression of gp110 was knocked down during EBV lytic cycle induced from EBV-positive Hone1 cells. Here, gp110 expression confirmed the lytic infection of EBV after TPA and NaB treatment (Fig. 3B). Compared with the vector control pSuper-Retro, shBALF4-2 increased the mRNA and protein expressions of these cytokines during EBV replication (Fig. 3, C and D). These results suggested that gp110 hampers NF-κB activity and thus abolishes the production of downstream cytokines.

Figure 3.

gp110 inhibits the mRNA and protein expressions of NF-κB-regulated downstream cytokines. A, expression plasmid of gp110-HA, BGLF4-HA, or HA vector was transfected into HeLa cells, at 24 h post-transfection, cells were stimulated without or with TNF-α (20 ng/ml) for 6 h. Total RNA was then extracted from the cells and reverse transcribed into cDNA, then qPCR amplification was performed using the primers for gp110, BGLF4, IL-1β, IL-6, IL-8, and the housekeeping gene GAPDH. B and C, expression plasmid of shBALF4-2 or pSuper-Retro vector was transfected into EBV-positive Hone1 cells for 24 h, and cells were stimulated without or with TPA (40 ng/ml) and NaB (3 mM) for 24 h to induce the lytic infection of EBV, then qPCR amplification was performed using the primers for gp110, IL-1β, IL-6, IL-8, and GAPDH (C). The expression of gp110 was detected to show the lytic cycle of EBV by mouse anti-gp110 mAb, and β-actin was applied to verify equal loading of protein in each lane (B). D, cell medium from shBALF4-2 or pSuper-Retro vector transfected EBV-positive Hone1 cells in C was isolated and analyzed by ELISA for inflammatory cytokine IL-1β, IL-6, or IL-8 secretion, as described in the Experimental procedures section. Data were shown as means ± SD from three independent experiments. Statistical analysis was performed using the Student's t test. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; and ∗∗∗∗p < 0.0001. cDNA, complementary DNA; EBV, Epstein–Barr virus; gp110, glycoprotein 110; IL, interleukin; NaB, sodium butyrate; qPCR, quantitative PCR; TNF-α, tumor necrosis factor α; TPA, 12-O-tetradecanoylphorbol-13-acetate.

gp110 restrains the NF-κB signaling pathway downstream of p65

To explore at what level gp110 impedes NF-κB activity, HEK293T cells were cotransfected with the gp110-HA expression plasmid and the classical NF-κB signaling adaptor expression plasmid, including TNFR-associated death domain (TRADD), TNF receptor–associated factor-2 (TRAF2), TGF-β-activated kinase-1 (TAK1), receptor-interacting protein-1 (RIP1), IκB kinase alpha/beta (IKKα/β), or the subunit of NF-κB-p65, along with the NF-κB-Luc reporter. As a result, all these components led to marked activation of the NF-κB-Luc reporter (28) (Fig. 4), which was inhibited in the presence of gp110 (Fig. 4, A–G), indicating gp110 may disrupt the NF-κB signaling at the downstream of p65.

Figure 4.

gp110 inhibits the NF-κB signaling pathway downstream of p65. pNF-κB-Luc (100 ng) and pRL-TK internal reference plasmid (10 ng) were cotransfected with 100 ng of pTRADD-FLAG (A), pTRAF2-FLAG (B), pRIP1-FLAG (C), pTAK1-FLAG (D), pIKKα-HA (E), pIKKβ-FLAG (F), or p65-FLAG (G) expression plasmid into HEK293T cells, along with 500 ng of gp110-HA expression plasmid or HA vector. At 24 h post-transfection, the cell lysates were collected and divided into two aliquots, one aliquot was used for DLR detection for NF-κB-Luc activity, and the other was used for WB analysis with tag-specific Abs to detect the expression of TRADD, TRAF2, RIP1, TAK1, IKKα, IKKβ, p65, and gp110. β-actin was used to verify equal loading of protein in each lane. Data were shown as means ± SD from three independent experiments. Statistical analysis was performed using the Student's t test. ∗∗p < 0.05; ∗∗∗p < 0.001; and ∗∗∗∗p < 0.0001. Ab, antibody; DLR, dual-luciferase reporter; gp110, glycoprotein 110; HEK293T, human embryonic kidney 293T cell line; IKK, IκB kinase; RIP1, receptor-interacting protein-1; TAK1, TGF-β-activated kinase-1; TRADD, TNFR-associated death domain; TRAF2, TNF receptor–associated factor-2; WB, Western blot.

gp110 interacts with p65, and its C-terminal cytoplasmic region is the key interaction domain

To probe the molecular mechanism of the gp110 evading NF-κB signaling pathway, the potential interaction between gp110 and p65 was investigated. gp110-HA/p65-FLAG or gp110-HA/MAVS-FLAG (the adaptor of RIG-I like receptor signaling pathway) expression plasmids were cotransfected into HEK293T cells, then coimmunoprecipitation (co-IP) analysis was carried out with anti-HA monoclonal antibody (mAb). The result showed that p65-FLAG, but not the negative control MAVS-FLAG, was immunoprecipitated with gp110-HA (Fig. 5A), which eliminated the nonspecificity of anti-HA mAb from co-IP assay, indicating gp110 may interact with p65. For further verification, gp110-HA expression plasmid was transfected into HEK293T cells, and then co-IP assay was performed with anti-HA mAb to detect the interaction of gp110-HA with endogenous p65. As shown in Figure 5B, endogenous p65 was associated with gp110-HA. To check the gp110 and p65 interaction under physiological conditions, EBV-positive Hone1 cells were treated with TPA and NaB to induce EBV lytic cycle, then co-IP assay was performed with anti-p65 pAb, and the result showed that gp110 was successfully immunoprecipitated with p65 (Fig. 5C), uncovering gp110 interacts with endogenous p65 during EBV lytic infection.

Figure 5.

gp110 interacts with p65, and the C-terminal cytoplasmic region plays an important role in this process.A, D–F, HEK293T cells were cotransfected with the expression plasmids combination of gp110-HA/p65-FLAG (A), gp110-HA/MAVS-FLAG (A), gp110(1–685)del-HA/p65-Myc (D), IKKα-HA/p65-Myc (D), gp110(686–753)del-HA/p65-Myc (E), or gp110(754–857)del-HA/p65-Myc (F). At 24 h post-transfection, cells were collected for lysis. The samples were subjected to co-IP analysis using anti-HA mAb or nonspecific IgG. Then, samples were resolved by 10% SDS-PAGE. WB analysis was probed with anti-HA, anti-FLAG, or anti-Myc mAb. B, gp110-HA expression plasmid was transfected into HEK293T cells, at 24 h post-transfection, cells were stimulated with TNF-α (20 ng/ml) for 6 h, and the cells were collected for lysis. The samples were then subjected to co-IP assay using anti-HA mAb or nonspecific IgG. WB analysis was performed with anti-HA mAb and anti-p65 pAb. C, EBV-positive Hone1 cells were treated with TPA (40 ng/ml) and NaB (3 mM) for 24 h to induce EBV lytic infection, and the cells were collected for lysis. The samples were then subjected to co-IP assay using anti-p65 pAb or nonspecific IgG. WB analysis was performed with anti-gp110 mAb and anti-p65 pAb. gp110, glycoprotein 110; HEK293T, human embryonic kidney 293T cell line; co-IP, coimmunoprecipitation; EBV, Epstein–Barr virus; IgG, immunoglobulin G; mAb, monoclonal antibody; NaB, sodium butyrate; TNF-α, tumor necrosis factor α; TPA, 12-O-tetradecanoylphorbol-13-acetate; WB, Western blot.

To dissect which domain of gp110 can interact with p65, p65-Myc expression plasmid was cotransfected with the gp110 truncated mutant expression plasmids gp110(1–685)del-HA, gp110(686–753)del-HA, gp110(754–857)del-HA, or the negative control IKKα-HA (29), and co-IP assay was completed with anti-HA mAb. As a result, p65-Myc was immunoprecipitated with gp110(1–685)del-HA (Fig. 5D) and gp110(686–753)del-HA (Fig. 5E), but not gp110(754–857)del-HA (Fig. 5F) or the negative control IKKα-HA (Fig. 5D), indicating the C-terminal cytoplasmic region of gp110 is the key domain for gp110 and p65 interaction.

gp110 interacts with the Rel homology domain and C-terminal domain of p65, but it does not affect p65 binding to p50

p65 harbors 551 amino acids (aa), its N terminus contains a Rel homology domain (RHD, aa 21–187), which is predominantly responsible for binding to DNA-specific sequences, whereas the IPT (Ig-like, plexins, and transcription factors) domain (aa 188–290) of p65 is mainly responsible for homologous dimerization or heterodimerization, and the C terminus possesses a transcription activation domain (aa 426–550) (Fig. 6A) (30). To inspect which domain of p65 interplays with gp110, the potential interaction of gp110 with the N terminus and/or C terminus of p65 was first analyzed. p65(aa1-290)-FLAG or p65(aa291–551)-FLAG truncated expression plasmid was cotransfected into HEK293T cells with gp110-HA expression plasmid, and then samples were 0subjected to co-IP analysis using anti-HA mAb. Surprisingly, gp110 simultaneously interacted with aa 1–290 and aa 291–551 of p65 (Fig. 6, B and C), suggesting more than one interaction site exists between gp110 and p65. Next, the potential interaction between gp110 and the RHD (aa 21–186) domain or IPT domain of p65 was examined. Expression plasmids combination of gp110-HA/p65(RHD)-FLAG, gp110-HA/p65-EYFP, or gp110-HA/p65(IPT)-GFP were cotransfected into HEK293T cells, and then co-IP assay was carried out using anti-HA mAb. The results showed that p65(RHD)-FLAG (Fig. 6D) and the positive control p65-EYFP (Fig. 6E), but not p65(IPT)-GFP (Fig. 6E), was efficiently immunoprecipitated with gp110-HA, which again erased the nonspecificity of anti-HA mAb from co-IP assay. Consequently, these data indicated that gp110 interacts with the N-terminal RHD domain and C terminus of p65.

Figure 6.

gp110 interacts with the RHD and C-terminal domain of p65, but it may not affect p65 binding to p50.A, the structural domain diagram of p65 and its truncated mutant expression plasmids. Numbers indicate the amino acid position. B–E, gp110-HA expression plasmid was cotransfected with p65(1–290)-FLAG (B), p65(291–551)-FLAG (C), p65(RHD)-FLAG (D), p65-EYFP, or p65(IPT)-GFP (E), expression plasmid into HEK293T cells, and cells were harvested for lysis at 24 h post-transfection. Then, samples were subjected to co-IP analysis using anti-HA mAb or nonspecific IgG. WB analysis was performed with anti-HA, anti-FLAG, or anti-GFP mAb. F, gp110-HA expression plasmid or HA vector was cotransfected with p65-Myc and p50-FLAG expression plasmids into HEK293T cells. Cells were harvested at 24 h post-transfection, and then samples were subjected to co-IP analysis using anti-Myc mAb. WB analysis was performed with anti-FLAG, anti-Myc, or anti-HA mAb. β-actin was used to verify equal loading of protein in each lane. co-IP, coimmunoprecipitation; EYFP, enhanced YFP; GFP, green fluorescent protein; gp110, glycoprotein 110; HEK293T, human embryonic kidney 293T cell line; IgG, immunoglobulin G; IPT, Ig-like, plexin, and transcription factor; mAb, monoclonal antibody; RHD, Rel homology domain.

The aforementioned results showed that gp110 interacts with p65(RHD) but not p65(IPT). We, therefore, continued to examine whether gp110 affects the heterodimerization of p65 and p50. gp110-HA expression plasmid or HA vector was cotransfected into HEK293T cells with p50-FLAG and p65-Myc expression plasmids. Then, co-IP was performed with anti-Myc mAb. As shown in Figure 6F, p65-Myc was associated with p50-FLAG and gp110-HA. However, the presence of gp110 did not alter the interaction between p65 and p50, indicating that gp110 may not compete with p50 to bind to the corresponding sites of p65 and diminish the p65 and p50 heterodimerization.

gp110 and its C-terminal cytoplasmic region restrict the phosphorylation of p65 (Ser536)

Post-translational modification of NF-κB takes a significant role in its function. The p65 subunit of NF-κB is only functional after phosphorylation during the activation of the TNF-α-induced NF-κB signaling pathway. Phosphorylation of p65 can happen at various sites to influence diverse functions of NF-κB. Among them, phosphorylation at C-terminal Ser536 changes the nucleation of NF-κB dimer (31), whereas phosphorylation at N-terminal Ser276 affects the binding of NF-κB to nuclear transcriptional coactivators (32).

As mentioned previously, gp110 interacts with the N-terminal RHD and C-terminal domain of p65. Next, we continued to investigate whether gp110 disrupts the phosphorylation of p65. gp110-HA expression plasmid or HA vector was transfected into HeLa cells, and then cells were stimulated with TNF-α at different times. As shown in Figure 7, A and B, when cells were stimulated with TNF-α for 30 and 60 min, phosphorylated p65 (Ser536) and p65 (Ser276) were enhanced. However, the expression of phosphorylated p65 (Ser536) (Fig. 7A), but not phosphorylated p65 (Ser276) (Fig. 7B), was confined at both 30 and 60 min in the presence of gp110.

Figure 7.

gp110 inhibits the phosphorylation of p65 (Ser536), which is executed by its C-terminal cytoplasmic region.A and B, D–F, HA vector, gp110-HA expression plasmid (A and B) or different truncated mutants of gp110 expression plasmids (D–F), was transfected into HeLa cells, at 24 h post-transfection, the cells were stimulated without or with TNF-α (20 ng/ml) for 0, 30, and 60 min, then cells were harvested for WB analysis with anti-p65 pAb, anti-HA mAb, phospho-NF-κB-p65 (Ser536) (A, D–F) or phospho-NF-κB-p65 (Ser276) (B). β-actin was used to verify equal loading of protein in each lane. Here, the vectors showed in D–F were actually all from the same vector sample, since only one vector control was set for the experiments of D–F. C, expression plasmid of shBALF4-2, shRandom, or pSuper-Retro vector was transfected into the EBV-positive Hone1 cells, followed cells were stimulated with TPA (40 ng/ml) and NaB (3 mM) for 24 h, then cells were stimulated without or with TNF-α (20 ng/ml) for 60 min, and the cells were harvested for WB analysis, as indicated in A. EBV, Epstein–Barr virus; gp110, glycoprotein 110; NaB, sodium butyrate; TNF-α, tumor necrosis factor α; TPA, 12-O-tetradecanoylphorbol-13-acetate; WB, Western blot.

An interference experiment was executed in EBV-positive Hone1 cells to confirm the aforementioned result further. pSuper-Retro vector, shBALF4-2, or shRandom expression plasmid was transfected into EBV-positive Hone1 cells, and TPA and NaB were added for 24 h to stimulate EBV into the lytic cycle. Compared with the pSuper-Retro vector and shRandom, the expression of phosphorylated p65 (Ser536) increased in EBV-positive Hone1 cells stimulated with TNF-α, and its expression further elevated when gp110 was knocked down by shBALF4-2 (Fig. 7C). Altogether, these results disclosed that gp110 inhibits the phosphorylation of p65(Ser536) during EBV lytic infection.

To continue examining which domain of gp110 can restrict p65 phosphorylation, HA vector or diverse truncated mutant expression plasmids gp110(1–685)del-HA, gp110(686–753)del-HA, or gp110(754–857)del-HA was transfected into HeLa cells, and experiments were performed as indicated in Figure 7A. Compared with the HA vector, gp110(1–685)del-HA (Fig. 7D) and gp110(686–753)del-HA (Fig. 7E), but not gp110(754–857)del-HA (Fig. 7F), could inhibit the phosphorylation of p65(Ser536) at the detected times, which was again consistent with the results of Figures 1D and 5.

gp110 and its C-terminal cytoplasmic region prevent the nuclear translocation of p65

The NF-κB dimer p65/p50 is present in the cytoplasm in an inactive form bound by the NF-κB inhibitor IκBα, which blocks the nuclear localization signal of p65. Upon stimulation, IκBα is degraded through the ubiquitination-proteasome system. Subsequently, p65/p50 dimer is released into the nucleus to induce transcriptional activation (33). Previous studies have shown that gp110 is mainly located in the nuclear membrane and cytoplasm of the plasmid-transfected and EBV-infected cells (34,35), suggesting gp110 may hijack p65 in the cytoplasm. To verify this speculation, HeLa cells were transfected with gp110-HA expression plasmid or HA vector and treated without or with TNF-α, and indirect immunofluorescence assay (IFA) was performed to detect the subcellular distribution of p65. The results showed that p65 was located in the cytoplasm without TNF-α stimulation, whereas p65 entered the nucleus when cells were stimulated with TNF-α. Upon TNF-α stimulation, p65 remained in the cytoplasm in the presence of gp110 (Fig. 8A).

Figure 8.

gp110 and its C-terminal cytoplasmic region inhibit TNF-α-mediated nuclear translocation of p65. HA vector, gp110-HA expression plasmid (A), or different truncated mutants of gp110 expression plasmids (B–D) was transfected into HeLa cells, at 24 h post-transfection, cells were treated without or with TNF-α (20 ng/ml) for 30 min. Cells were then fixed with paraformaldehyde for 30 min, permeated with 0.1% Triton X-100 for 15 min, and blocked with 5% BSA for 1 h. Then, cells were stained with primary Abs anti-HA mAb, anti-p65 pAb, and secondary Abs FITC-conjugated goat anti-rabbit IgG (green) and Cy5-conjugated goat anti-mouse IgG (red). Subsequently, cell nuclei were stained with Hoechst 33342 (blue), and the images were obtained by laser scanning confocal microscopy (SP8; Leica Microsystems). All scales indicate 10 μM. Ab, antibody; BSA, bovine serum albumin; gp110, glycoprotein 110; mAb, monoclonal antibody; TNF-α, tumor necrosis factor α.

Finally, to explore which domain of gp110 can inhibit the nuclear accumulation of p65, HA vector or different truncated mutant expression plasmids gp110(1–685)del-HA, gp110(686–753)del-HA, or gp110(754–857)del-HA was transfected into HeLa cells, and IFA was carried out as indicated in Figure 8A, and results again showed that gp110(1–685)del-HA (Fig. 8B) and gp110(686–753)del-HA (Fig. 8C), but not gp110(754–857)del-HA (Fig. 8D), could impede the nuclear trafficking of p65(Ser536) as gp110. These results revealed that gp110 interacts with p65 and hijacks it in the cytoplasm to inhibit TNF-α-induced NF-κB activation, and the C-terminal cytoplasmic region is the indispensable domain for gp110 to accomplish this function.

Discussion

Innate immunity is the first line of host defense against pathogen invasion, but many viruses, including EBV, have evolved distinct strategies to hamper the host's innate immune response for maintaining their replications, such as impeding the RLR/TLR-mediated signal transduction pathways (36). Zta, an early lytic transactivator of EBV, stimulates p65 nuclear translocation but obstructs its transcriptional activity (37), which also directly binds to the TNF-α promoter to inhibit NF-κB activation (38). Deubiquitinase BPLF1 bolsters the deubiquitination of TRAF6 and constrains its activity, thus suppressing the TLR-mediated NF-κB activation (39). In addition, BGLF4 represses NF-κB activity by targeting the coactivator UxT of NF-κB (27). However, the relationship between host innate immunity and EBV is not simply hostile.

NF-κB is an influential immune and inflammatory transcription factor that plays a critical role in the life cycle and tumorigenesis of EBV (40, 41). NF-κB restricts the replication of EBV by restraining the expression of lytic inducers ZTA and Rta (27, 42, 43). The NF-κB pathway is appropriately activated when EBV is required to maintain its latent infection. At the same time, NF-κB signaling facilitates the growth and survival of infected cells, which must be sustained if the virus is to hold its infection. LMP1 recruits TRAF1, TRAF2, TRAF3, and TRAF5 to induce the IKK-dependent NF-κB pathway through its CTAR1 domain, whereas its CTAR2 domain indirectly recruits TRAF6 to activate the IKK and IKK-dependent NF-κB pathway (44, 45, 46). The G protein–coupled receptor BILF1 also stimulates the activation of NF-κB (47). EBV generally endorses the activation of NF-κB during incubation to retain cell growth, but it intercepts NF-κB signaling during the lytic cycle. Therefore, the finale of EBV infection is determined by the subtle interaction between antiviral innate immunity and EBV infection.

gp110 is an EBV-encoded lytic protein that encompasses three domains. The extracellular domain contains nine potential N-junction glycosylation sites and a predicted cleavage signal sequence (23). It also possesses a fragment of the Furin recognition sequence, cleaved by the cell protease (22). The intermediate transmembrane domain contains three hydrophobic motifs (aa 733–753), indispensable for membrane anchoring. Besides, the N-terminal cytoplasmic tail region (aa 802–816) is fundamental for the subcellular localization and functions of gp110 (48, 49). It is well known that gB is the most conserved glycoprotein in all herpes viruses, and EBV gp110 and HSV-1 gB have 29% sequence homology and 43% similarity (10). EBV gp110 has a comparable 3D structure with HSV-1 gB (21). gp110 can mediate the fusion of the virus and host cell membrane (21, 22, 23, 24). It also participates in the replication and capsid assembly of viruses in some cell lines (10, 25), thus contributing to the maturation of virions. HSV-1 gB is also involved in fusing virus and host cells, but it is not required for viral maturation (50). In addition, HSV-1 gB can activate NF-κB through the TLR signaling pathway (17, 19), but it is not known whether the surface protein gp110 of EBV is also implicated in regulating antiviral innate immunity.

In the present study, the eukaryotic expression plasmid and shRNA interfering plasmids of gp110 were first constructed, and then DLR assays were conducted in HEK293T cells. It was found that gp110 inhibited TNF-α-mediated NF-κB promoter activity, which was executed by its C-terminal cytoplasmic region. However, NF-κB activity was recovered after knocking down the expression of gp110 in EBV-positive Hone1 cells. Furthermore, gp110 interfered with the mRNA expression of NF-κB-regulated downstream cytokines, including IL-1β, IL-6, and IL-8, which were upregulated when gp110 was knocked down in EBV-positive Hone1 cells reactivated by TPA and NaB. These results suggested that gp110 indeed suppresses the activation of NF-κB and inhibits the production of NF-κB-regulated cytokines.

The subcellular localization of certain proteins is closely related to its function execution. When the subcellular localization pattern of viral protein in transfected cells is consistent with that of viral infection, the function(s) of viral proteins in the transfected cells can reflect its physiological effect(s) in viral infection. In this study, we originally wanted to use the gp110 antibody (Ab) to analyze whether the subcellular localization of gp110 during EBV lytic infection induced in EBV-positive Hone1 cells is consistent with that of plasmid transfected cells, but the gp110 Ab we purchased is described for application only for WB, but not IFA. Although we tried several times to use the gp110 Ab to detect the localization of gp110 by IFA during EBV lytic infection, we still could not detect the fluorescence of gp110. Nevertheless, our previous study shows that the transfected gp110-EYFP exhibits a prominent cytoplasmic localization pattern and intense perinuclear cytoplasmic patches like the endoplasmic reticulum (35). Gong et al. (10) also demonstrated that the distribution of gp110 in gp110-transfected 3T3 psi am 22b cells shows a localization mode of perinuclear, diffusely cytoplasmic or in the nuclear membrane, or in intense perinuclear cytoplasmic patches, whereas gp110 shows a cytoplasmic, perinuclear, and nuclear membranes, ER and the inner/outer nuclear membranes, plasma membrane, and cell surface localization pattern during EBV lytic infection induced in B95-8 cells (10, 34, 51), which is consistent with the localization pattern of gp110 plasmid transfection. Accordingly, gp110 should have a certain physiological function in EBV-infected physiological cells as in gp110-EYFP plasmid transfection.

Subsequently, gp110 was shown to constrain the activation of NF-κB induced by p65 and other upstream adaptors, indicating that its inhibitory mechanism may occur downstream of p65. We, therefore, speculated that gp110 might interact with p65 to interfere with the activation of the NF-κB signaling pathway. Next, co-IP confirmed that gp110 could interact with p65 in plasmid overexpression cells and EBV lytic infection cells, and the C-terminal cytoplasmic region of gp110 was important for its association with p65. Furthermore, gp110 interacted with the full-length, RHD, and C-terminal domains but not the IPT of p65.

RHD of p65 is principally responsible for DNA binding, whereas IPT is predominantly in charge of dimerization. Meanwhile, the nuclear localization signal of p65 is also localized in RHD (30). However, the co-IP experiment showed that gp110 might not influence the heterodimerization of p65/p50. Mechanically, gp110 inhibited the phosphorylation of p65 at Ser536 during plasmid transfection and EBV infection and the nuclear accumulation of p65, which was again implemented by its C-terminal cytoplasmic region. HSV-1-encoded VP16 interacts with p65 to impede Sendai virus and TNF-induced NF-κB promoter activation and expression of NF-κB-dependent inflammatory cytokines. Moreover, VP16 is associated with IRF3 and precisely resists IRF3-mediated transactivation, which cannot influence the dimerization, nuclear accumulation, or DNA-binding activity of IRF3, but interact with the coactivator CBP to effectively intervene in the formation of IRF3–CBP transcriptional complex (19, 52). However, it is unknown whether gp110 also has the same mechanism as VP16 of interacting with IRF3 and suppressing the IFN-I activation, which must be further verified in future studies.

The initial experimental design of this study is to construct BALF4-encoded gp110 deleted, mutated, and reversed EBV recombinant viruses to analyze the inhibition effect of gp110 on the NF-κB pathway in EBV-infected physiological cells, but gp110 is an essential gene for EBV, which plays fundamental roles during viral fusion with the cell membrane (21, 22, 23, 24), as well as the proliferation and capsid assembly of virions (10, 25). When gp110 is deleted, it is not easy to obtain the progeny virions (20, 53). It has been found that EBV employs different manners to suppress the NF-κB activity by its reported (36) or unreported proteins during viral infection, and there may be a synergistic effect or a complementary effect on impeding the NF-κB activity among different viral proteins. Upon EBV infection, the NF-κB repression mediated by one viral protein may be counterbalanced by other viral proteins when this protein is deleted. Accordingly, the suppressive effect of wildtype EBV on NF-κB activity may not significantly differ from that of gene-deleted EBV.

Based on the aforementioned analysis, the restraint effect of EBV on NF-κB activity may not change when gp110 is deleted. Even if the NF-κB activity is increased after gp110 is deleted, it may not be directly triggered by the gp110-mediated reduction of NF-κB activity after gp110 deletion, but it may be caused by the conspicuous deficiency of EBV proliferation after gp110 deletion, resulting in a severe diminishment of virus titer, which also lessens the EBV-mediated suppression of NF-κB activity, since certain kinds of EBV proteins that exert inhibitory activity on NF-κB are also reduced. Thus, it is challenging for us to determine whether the NF-κB activity rebound (if it happened) after gp110 deletion during EBV infection is directly caused by gp110, which means that we only can use the gp110 knockdown expression plasmid and wildtype EBV-infected nasopharyngeal epithelial cells, but not the gp110-deficient EBV- and EBV- infected physiological cells, to analyze the inhibitory effect of gp110 on NF-κB activity.

Studies have shown that the expression of herpesvirus-encoded proteins is strictly regulated with three stages, immediate-early, early, and late. Some are present in mature progeny virions, but other proteins are only involved in viral replication (54, 55, 56). Moreover, different viral proteins may have synergistic and/or compensatory effects in the context of virus infection. When EBV invades the host cell, some EBV-carried proteins are released into the cytoplasm to shut down or hijack the host's transcriptional and translational systems (57, 58) and suppress the host's innate immunity and adaptive immunity (36, 59, 60, 61). Therefore, we believe that EBV proteins expressed/existed in different phases of EBV replication, including absorption, penetration, uncoating, biosynthesis, assembly, and release, can collaboratively regulate the NF-κB activity to promote EBV proliferation, and each EBV-encoded protein with multiple specific functions in EBV replication is not redundant, that is why herpes virus so smart enough to cosurvive and coevolve with its host. Consequently, although other modulators like Zta (37, 38), BPLF1 (39), and BGLF4 (27) can inhibit the NF-κB activity during EBV infection, we suppose gp110 is also important for EBV to restrain the NF-κB activity, but they may execute this function at diverse phases of EBV life cycle, and gp110 may inhibit the NF-κB activity and cytokines during the late phase of EBV reactivation.

Collectively, the surface protein gp110 of EBV can restrain the activation of NF-κB and the production of its downstream cytokines by suppressing the phosphorylation (Ser536) and nuclear translocation of p65 via the C-terminal cytoplasmic region of gp110 (Fig. 9), whereas HSV-1 gB activates the NF-κB pathway through the TLR2 signaling pathway, suggesting homologs from various subfamilies of herpes virus may be functionally different to adapt their life cycles.

Figure 9.

Schematic diagram of gp110 downregulating the NF-κB signaling pathway. When TNF-α binds to its receptor TNFR, TNFR recruits downstream signaling pathway components RIP1, TRADD, TRAF2, and TAK1, to activate the IKK complex, which results in the degradation of IκBα, then heterodimer of phosphorylated p65 (Ser536) and p50 is released and translocates into the nucleus to eventually induce the activation of NF-κB and series expression of various cytokines. When EBV invades the cell, it can exist in the nucleus as an episome. The EBV-encoded gp110 can downregulate TNF-α-induced NF-κB activity through interaction with p65. Ultimately, gp110 inhibits and prevents phosphorylation (Ser536) and nuclear translocation of p65. The red line (T) shows that gp110 interacts with host molecule p65 to downregulate the TNF-α-mediated NF-κB signaling pathway. EBV, Epstein–Barr virus; gp110, glycoprotein 110; IKK, IκB kinase; TAK1, TGF-β-activated kinase-1; RIP1, receptor-interacting protein-1; TNF-α, tumor necrosis factor α; TNFR, tumor necrosis factor receptor; TRADD, TNFR-associated death domain; TRAF2, TNF receptor–associated factor-2.

Experimental procedures

Cells, antibodies, and cytokines

HEK293T and HeLa cells were grown in Dulbecco's modified Eagle's medium (Gibco-BRL) supplemented with 10% heat-inactivated fetal bovine serum and 100 U/ml of penicillin and streptomycin. EBV-positive nasopharyngeal carcinoma (Hone1-EBV) cells (gift from Prof S.W. Tsao, University of Hong Kong) were cultured in RPMI1640 (Gibco-BRL) supplemented with 10% fetal bovine serum. To induce EBV reactivation in Hone1-EBV cells, TPA and NaB were added to the culture medium at final concentrations of 40 ng/ml and 3 mM, respectively (62, 63). Mouse mAbs anti-DYKDDDDK (FLAG), anti-hemagglutinin (HA), and anti-Myc were purchased from Abmart. Mouse anti-GFP mAb and rabbit polyclonal Ab (pAb) anti-RELA (p65) were provided by Proteintech, and phospho-NF-κB-p65 (Ser276) Ab was offered by Affinity Biosciences. Phospho-NF-κB-p65 (Ser536), nonspecific immunoglobulin G (IgG), alkaline phosphatase–linked goat anti-rabbit IgG and goat anti-mouse IgG were supplied by Cell Signaling Technology. β-actin mAb was purchased from ABclonal, FITC-conjugated donkey anti-rabbit IgG and Cy5-conjugated goat anti-mouse IgG were bought from BBI Life Sciences. Mouse anti-gp110 mAb was acquired from Millipore. Recombinant human TNF-α was obtained from Peprotech.

Plasmids

To generate the eucaryotic expression plasmid of gp110-HA, the open reading frame of gp110 was amplified from bacterial artificial chromosome DNA of EBV, and then the product was digested with EcoRI and BamHI and cloned into the vector pHA-N1 (modified from the pEYFP-N1 vector; Clontech). Moreover, truncated mutant fragments of gp110 (24), gp110(1–685)del (deletion of the N-terminal extracellular region), gp110(686–753)del (deletion of the hydrophobic region in the intermediate region), and gp110(754–857)del (deletion of the C-terminal cytoplasmic region) were cloned into pHA-N1 by a similar method. The primers were available upon request. Three gp110-targeted sequences of shBALF4, 5′-GCA TCT TCC GGG AGT ACA ACT-3′, 5′-GGG CCT GTT GAT GGT GTT TAA-3′, and 5′-GCC ACC GTC CAG ATC CAA TTT-3′, were inserted into pSuper.retro.puro (OligoEngine) to produce pSuper-shBALF4-related expression plasmids, namely shBALF4-1, shBALF4-2, and shBALF4-3. The control shRandom and BGLF4-HA expression plasmids were described previously (64). p65-Myc expression plasmid was yielded by inserting the fragment of p65 into the EcoRI/BamHI sites of the pMyc-N1 vector.

Besides, the pNF-κB-Luc reporter plasmid and pRL-TK (expressing Renilla Luc) internal control were graciously provided by Dr Zhengli Shi (Wuhan Institute of Virology, Chinese Academy of Sciences). TRAF2-FLAG, RIP1-FLAG, TRADD-FLAG, and p65-RHD-FLAG expression plasmids were friendly afforded by Dr Chunfu Zheng (Department of Microbiology, Immunology and Infectious Diseases, University of Calgary) (65). pTAK1-FLAG and MAVS-FLAG were generously offered by Dr Jun Cui (School of Life Sciences, Sun Yat-sen University) (66). pIKKα-HA was genially supplied by Prof Gangmin Hur (College of Medicine, Chungnam National University). pIKKβ-FLAG was nicely furnished by Prof Anjana Rao (Department of Signaling and Gene Expression, La Jolla Institute for Allergy and Immunology). p50-FLAG was a gift from Dr Karl-Klaus Conzelmann (Department of Virology, Max von Pettenkofer Institute and Gene Center, University of Munich). pCMV-p65-FLAG was favorably provided by Prof Katherine A. Fitzgerald (Division of Infectious Diseases and Immunology, University of Massachusetts Medical School). p65-EYFP was a gift from Prof Hong An (Institute of Biology Engineering, Jinan University) (67). Pad-N-p65(1–290)-FLAG and Pad-N-p65(291–551)-FLAG expression plasmids were gifts from Dr Leiliang Zhang (Institute of Pathogen Biology, Chinese Academy of Medical Sciences & Peking Union Medical College).

DLR assay

HEK293T cells were cultured on 24-well plates (Corning), then cotransfected with 100 ng of reporter plasmid pNF-κB-Luc, 10 ng of pRL-TK (internal control), and 500 ng of viral protein or related truncated mutant expression plasmid mixed with polyethylenimine transfection reagent (Polysciences). At 24 h post-transfection, cells were treated without or with recombinant human TNF-α (20 ng/ml) for 6 h. About the EBV-positive Hone1 cells, pSuper-Retro vector or shBALF4 expression plasmid was cotransfected with the reporter plasmids as indicated in HEK293T cells for 24 h, and then cells were treated without or with TPA and NaB to induce EBV reactivation. Then, cells were collected for lysis, and the lysates were divided into two aliquots. One aliquot was employed for DLR detection, and the other was used for WB analysis. The DLR was performed as described previously (17, 68, 69). All reporter assays were performed at least in triplicate, and the results were shown as average values ± SD.

WB analysis

WB analysis was executed as previously described (70, 71). Briefly, the cell lysates were resolved in 10% SDS-PAGE and transferred to nitrocellulose filter membrane (Pall), followed by blocking with 5% nonfat milk in Tris-buffered saline and probed with the indicated primary Abs at 4 °C overnight. The membrane was washed with Tris-buffered saline–Tween and then incubated with alkaline phosphatase–conjugated goat anti-mouse IgG or goat anti-rabbit IgG. Finally, protein bands specific to the Ab were developed with 5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium (Biosharp).

RNA isolation and qPCR

qPCR was performed using the Sybr Green procedure and CFX96 Real-Time PCR Detection System (Bio-Rad) (66, 69, 72). According to the manufacturer's operation manual, total RNA was isolated from HeLa cells or EBV-positive Hone1 cells with TRIzol (Invitrogen). Then, samples were subjected to reverse transcription. The obtained complementary DNA was employed as a template for qPCR to investigate the expression of human inflammatory cytokines IL-1β, IL-6, IL-8, gp110, BGLF4, and housekeeping gene GAPDH. The primers for IL-1β were 5′-CAT CAG CAC CTC TCA AGC AG-3′ and 5′-ATA GCC GTA CTC AAA AAC CT-3' (73), for IL-6 were 5′-GAA AGC AGC AAA GAG GCA CT -3′ and 5′-TTT CAC CAG GCA AGT CTC CT-3′, for IL-8 were 5′-GGT GCA GTT TTG CCA AGG AG-3′ and 5′-TTC CTT GGG GTC CAG ACA GA-3′, for GAPDH were 5′-CAT CAT CCC TGC CTC TAC TG-3′ and 5′-GCC TGC TTC ACC ACC TTC-3' (27), for gp110 were 5′-GCT TCG TGA CCA ACA CAA CC-3′ and 5′-GTA ATG GCT TCC TGG CCC TT-3′, and for BGLF4 were 5′-CCA TTA CCT GCG AGT ATC TG-3′ and 5′-CGG CCC TTA GAA GAC TTT AG-3'. All qPCR assays were performed at least in triplicate, and the results were shown as average values ± SD.

ELISA

ELISA was carried out to quantify the cells secreted inflammatory cytokines using human-related ELISA kits (MULTI SCIENCES), according to the manufacturer's recommended instructions. Briefly, EBV-positive Hone1 cells were transfected with shBALF4 expression plasmid or pSuper-Retro vector, at 24 h post-transfection, cells were treated with TPA (40 ng/ml) and NaB (3 mM) for 24 h to induced EBV lytic infection, subsequently, the cell medium was collected and centrifuged to remove cell debris, and 100 μl of cleared supernatants or the inflammatory cytokine standard was loaded on 96-well plates precoated with IL-1β, IL-6, or IL-8 capture Ab. After washing, the bound proteins were detected by adding human IL-1β, IL-6, or IL-8 detection Ab, followed by horseradish peroxidase–conjugated streptavidin, then 3,3′,5,5′-tetramethylbenzidine substrate was added, and the absorbance was read at 450 nm.

Co-IP assay

The co-IP assay was performed as previously described (74, 75, 76, 77). In short, HEK293T cells were cotransfected with 1 μg of the indicated expression plasmids combination of gp110-HA/p65-FLAG, gp110-HA/MAVS-FLAG, gp110 truncated mutant/p65-Myc, IKKα-HA/p65-Myc, gp110-HA/p65 truncated mutant, gp110-HA/p65-EYFP, or gp110-HA/p65-Myc/p50-FLAG. Regarding the endogenous interaction between gp110 and p65 under physiological conditions, EBV-positive Hone1 cells were treated with TPA and NaB to induce EBV reactivation. At 24 h post-transfection or post-treatment, cells were harvested and lysed on ice with 0.6–1 ml lysis buffer. Then, the sample was incubated with 10 μg of the mouse mAb anti-HA, anti-Myc, rabbit pAb anti-p65, or nonspecific IgG and 30 μl of a 1:1 slurry of Protein A/G Plus–agarose for 4 h at 4 °C. The beads were washed at least three times with 1 ml of lysis buffer containing 500 mM NaCl and then subjected to WB analysis.

Phosphorylation analysis

HeLa cells were transfected with HA vector, gp110-HA expression plasmid, or its related truncated mutant expression plasmid for 24 h, and cells were treated without or with TNF-α (20 ng/ml) for 0, 30, and 60 min. About the EBV-positive Hone1 cells, pSuper-Retro vector, shRandom control, or shBALF4 expression plasmid was transfected into the cells for 24 h, then cells were treated without or with TPA and NaB for 24 h to induce EBV reactivation, and cells were stimulated without or with TNF-α (20 ng/ml) for another 60 min. Subsequently, cells were harvested for WB analysis with anti-p65 pAb, anti-HA mAb, phospho-NF-κB-p65 (Ser536), phospho-NF-κB-p65 (Ser276), or anti-gp110 mAb.

IFA

IFA was completed as described previously (69, 72). Briefly, HeLa cells were transfected with HA vector, gp110-HA expression plasmid, or its related truncated mutant expression plasmid for 24 h and treated without or with 20 ng/ml of recombinant human TNF-α for 30 min, then fixed in 4% paraformaldehyde. Cells were incubated with primary Abs rabbit anti-p65 pAb (diluted with 1:100) and mouse anti-HA mAb (diluted with 1:100), followed by incubation with secondary Abs Cy5-conjugated goat anti-mouse IgG and FITC-conjugated goat anti-rabbit IgG. After washing cells with PBS several times, the nuclei were stained with Hoechst 33342. Then, samples were observed with confocal microscopy (Leica). All scales indicate 10 μM.

Statistical analysis

Statistical analyses were conducted using the Student's t test of the GraphPad Prism 6 (GraphPad Software, Inc), with significant differences labeled on the figures. Significance standards were interpreted as ns, not significant, p > 0.05; ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; and ∗∗∗∗p < 0.0001. A p value < 0.05 was considered statistically significant.

Data availability

All data presented are contained within the main article.

Conflict of interest

The authors declare that they have no conflicts of interest with the contents of this article.

Reviewed by members of the JBC Editorial Board. Edited by George M. Carman