Enterovirus 71 2C Protein Inhibits NF-κB Activation by Binding to RelA(p65)

Viruses evolve multiple ways to interfere with NF-κB signaling, a key regulator of innate and adaptive immunity. Enterovirus 71 (EV71) is one of primary pathogens that cause hand-foot-mouth disease. Here, we identify RelA(p65) as a novel binding partner for EV71 2C protein from yeast two-hybrid screen. By interaction with IPT domain of p65, 2C reduces the formation of heterodimer p65/p50, the predominant form of NF-κB. We also show that picornavirus 2C family proteins inhibit NF-κB activation and associate with p65 and IKKβ. Our findings provide a novel mechanism how EV71 antagonizes innate immunity.


Results
Identification of host protein p65 as a binding partner for EV71 2C. To explore the mechanism of 2C in the pathogenicity of EV71 infection, we screened a human Spleen Matchmaker cDNA library (Clontech, Mountain View, CA, USA) fused to the GAL4 activating domain vector using EV71 2C as a bait in AH109 yeast two-hybrid system. The positive colonies were selected on high stringency plates Scientific RepoRts | 5:14302 | DOi: 10.1038/srep14302 (lacking tryptophan, leucine, adenine and histidine) and were incubated until colonies appeared, leading to the identification of 12 host proteins that potentially interact with 2C: ATCG1, CES1, CFP, CORO1A, CRLF3, DOK1, FLT, GPBAR1, LTBP4, PIAS3, PKM, RELA (Fig. 1A). Interestingly, RelA/p65, the most abundant member of NF-κ B family was found as one of the candidates to interact with 2C.
To further confirm the interaction between 2C and p65, we performed an in vitro GST pull-down assay with GST-fused 2C expressed in bacteria. GST-2C, but not GST, was able to pull down FLAG-p65 (Fig. 1B). To validate the interaction between the endogenous p65 and 2C in the context of EV71 infection, we performed immunoprecipitation experiment in RD cells infected with EV71 using anti-2C or anti-p65. In both cases, 2C was revealed to interact with p65 (Fig. 1C,D).
To test whether 187-273 and 187-290 of p65 are sufficient to bind 2C, 2C-GST or GST immobilized on glutathione-Sepharose beads were incubated with lysates from 293T cells transfected with 187-273 or 187-290 of p65-FLAG plasmids. As shown in Fig. 2E, 187-273 and 187-290 of p65 associated with 2C-GST. IPT domain of p65 is 194-290 and we found that GST-fused IPT interacted with GFP-2C (Fig. 2F). Taken together, EV71 2C protein interacted with IPT domain of p65. (B) EV71 2C interacts with p65. 2C-GST or GST immobilized on glutathione-Sepharose beads were incubated with lysates from 293T cells transfected with p65-FLAG plasmid. The bound proteins were subjected to Western blots using indicated antibodies. (C) Co-immunoprecipitation confirms that the EV71 2C binds to p65. RD cells were infected with EV71 for 24 h. Co-IP analysis was performed with anti-2C antibody or control serum followed by Western blot. (D) Co-immunoprecipitation confirms that p65 binds to EV71 2C. RD cells were infected with EV71 for 24 h. Co-IP analysis was performed with anti-p65 antibody or control serum followed by Western blot.
Because 1-104 of 2C didn't bind to IPT-GST while 1-125 of 2C did, we hypothesized that the IPT-associated region was narrowed down to 105-125 of 2C. Next, we constructed different truncated 2C-GST immobilized on glutathione-Sepharose beads were incubated with lysates from 293T cells transfected with p65-FLAG or truncated p65-FLAG plasmids. The bound proteins were subjected to Western blots using indicated antibodies. (E) 188-273 and 188-290 of p65 interacts with 2C. 2C-GST or GST immobilized on glutathione-Sepharose beads were incubated with lysates from 293T cells transfected with indicated truncated p65-FLAG plasmids. The bound proteins were subjected to Western blots using indicated antibodies. (F) p65 IPT interacts with 2C. IPT-GST or GST immobilized on glutathione-Sepharose beads were incubated with lysates from 293T cells transfected with GFP-2C plasmid. The bound proteins were subjected to Western blots using indicated antibodies. (G) 2C inhibits p65/p50 dimerization. 293T cells transfected with p65, p50, 2C or GFP constructs were harvested and analyzed by coimmunoprecipitation and Western blots using indicated antibodies.

Discussion
In this study, we discovered that EV71 2C inhibited NF-κ B activation through two different mechanisms. 105-125 and 126-263 of 2C suppressed p65/p50 dimerization probably by competing p65 IPT domain with association of p50. 1-104 and 105-121 of 2C inhibited NF-κ B activation through association with IKKβ . These results have important implications in the understanding of the innate immune antagonism strategies by EV71.
Numerous studies have investigated the innate immune evasion by EV71. EV71 3C suppressed the induction of type I interferon responses through cleavage of RIG-I 16 , TLR3 17 , and IRF7 18 . EV71 2A targets IFNAR1 and MAVS to antagonize type I IFN responses and type I IFN signaling 19,20 . EV71-induced miR-146a targets IRAK1 and TRAF6 involved in TLR signaling and type I interferon production 21 . EV71 2C associated IKKβ and suppressed its phosphorylation to inhibit NF-κ B activation 5 . Together with this, the findings presented here demonstrated that 2C targets two components of NF-κ B pathway. EV71 2C is capable to reduce p65/p50 dimer formation, which will shed important insights on other p65 associated viral proteins. The p65/p50 heterodimer is the most abundant form of the NFκ B dimers. By interacting with IPT domain of p65, EV71 2C is capable to disrupt the p65/p50 heterodimer, resulting in the suppression of NF-κ B activation.
in the future in order to delineate the precise amino acids in 2C for interacting with p65 and IKKβ . One of strategies is to determine the structure of 122-263 of 2C with IPT. While structures of many EV71 proteins have been reported, no structure for any picornavirus 2C proteins is solved. Interaction between 2C and IPT region of p65 will provide a valuable opportunity to solve the structure of protein complex containing IPT and 2C or partial 2C.
In summary, our study provides mechanistic evidences that EV71 2C could inhibit NF-κ B activation by association with p65. Two components of NF-κ B pathway including p65 and IKKβ associated with 2C, suggesting that multiple mechanisms are involved for 2C to suppress the NF-κ B activation. Our 105-125aa peptides from polivirus type I, polivirus type II, coxsackievirus B1, and EV68 inhibit NF-κ B activation. 293T cells were transfected with pNF-κ B, pRL-TK, and indicated 2C truncated constructs for 24 hours, and then treated with TNF (10 ng/ml) for 6 hours. The cells were harvested and assayed for dual luciferase activity. Asterisks indicated significant differences between groups, data statistics were used student t-test (mean ± SD, *** indicated p < 0.001). (B) 2C 105-125aa peptides from polivirus type I, polivirus type II, coxsackievirus B1, and EV68 interact with p65 IPT domain. IPT-GST immobilized on glutathione-Sepharose beads were incubated with lysates from 293T cells transfected with indicated 2C truncated plasmids. The bound proteins were subjected to Western blots using indicated antibodies. (C) 2C 105-125aa peptides from polivirus type I, polivirus type II, coxsackievirus B1, and EV68 interact with IKKβ . Lysates from 293T cells transfected with IKKβ and 2C 1-125aa truncated constructs were analyzed by coimmunoprecipitation and Western blots using indicated antibodies. (D) 2C proteins from polivirus type I, polivirus type II, coxsackievirus B1, and EV68 inhibit NF-κ B activation. 293T cells were transfected with pNF-κ B, pRL-TK, and indicated 2C truncated constructs for 24 hours, and then treated with TNF (10 ng/ml) for 6 hours. The cells were harvested and assayed for dual luciferase activity. Asterisks indicated significant differences between groups, data statistics were used student t-test (mean ± SD, *** indicated p < 0.001).
finding will not only further elucidate the mechanism of NF-κ B activation antagonism by EV71 2C, but also advance a general understanding of picornavirus 2C proteins, as key mechanisms are likely to be conserved across all picornavirus.

Methods
Yeast strains, yeast plasmid, cDNA library. S. cerevisiae AH109 cultivated in YPD liquid (Clontech, Mountain View, CA, USA) or agar medium at 30 °C. Yeast transformant strains were cultured in lacking tryptophan, or lacking tryptophan and leucine, or lacking tryptophan, leucine, adenine, and histidine supplement medium (Clontech, Mountain View, CA, USA) at 30 °C. GAL4 binding domain vector pGBKT7 and GAL4 activating domain vector pPGADT7 were from Clontech (Mountain View, CA, USA). The vector pGBKT7 and pPGADT7 carried the tryptophan and the leucine nutritional maker for selection in yeast, respectively.
Yeast two-hybrid screening. Competent yeast strain AH109 was transformed with bait plasmid pGBKT7-2C, according to the yeast transform system 2 manual (Clontech, Mountain View, CA, USA). After verifying that the bait plasmid pGBKT7-2C was expressed in the AH109 yeast strain and that did not activate the reporter gene, the AH109 yeast strain was transformed with human Spleen Matchmaker cDNA library (Clontech, Mountain View, CA, USA) fused to the GAL4 activating domain vector. Transformants were plated to low stringency plates (lacking tryptophan and leucine) and high stringency plates (lacking tryptophan, leucine, adenine, and histidine) until colonies appeared. DNA constructs. The DNA sequence encoding EV71 2C protein was cloned into the yeast plasmid pGBKT7, containing GAL4 binding domain of to generate PGBKT7-2C as bait for yeast two-hybrid screening. We transformed pEYFP-N1 vector for enzymes digestion EGFP tag and changed GFP/FLAG tags to generate pad vectors. The full-length Picornavirus 2C and 2C truncated mutations were inserted into pEGFP-C1 or padGFP vector. The full-length p65 and p65 truncations were cloned into pad-FLAG vector. EV71 2C or p65 IPT domain was cloned into the PGEX-4T-1 vector. Western blots. After electrophoresis, protein samples were transferred to 0.22 μ m PVDF membranes (Bio-Rad, Hercules, CA, USA). The PVDF membranes were blocked with 5% non-fat dry milk (Bio-Rad, Hercules, CA, USA) and then probed with indicated primary antibodies and HRP conjugate secondary antibodies. The ECL Western Blotting Detection Kit (Applygen, Beijing, China) was used to detect chemiluminescent signals.

Antibodies. Primary mouse antibodies included
Luciferase reporter assays. HEK293T cells seeded in 24-well plate were co-transfected with the plasmids pNF-κ B-luc (0.1 μ g/well) expressing Firefly luciferase, pRL-TK (0.02 μ g/well) expressing Renilla luciferase as an internal control, and indicated expression plasmids using lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA). At 24 hours post transfection, cells were treated with TNF (10 ng/ml) or mock treated for 6 hours. Firefly and Renilla luciferase activities were assessed by the dual-luciferase reporter assay system (Promega, Madison, WI, USA). To present relative fold change, we first calculated the normalized luciferase activity by divided the Firefly luciferase activity by Renilla luciferase activity. Then we divided the normalized luciferase activity with TNF by the normalized luciferase activity without TNF.
GST pull down. GST fusion proteins were purified using GST-Sepharose (GE health, Madison, WI, USA) according to the manufacturer's protocol. HEK293T cell lysates were extracted in lysis buffer containing 50 mM Tris, pH8.0, 100 mM NaCl, 1 mM EDTA, 1% Triton X-100 with proteinase inhibitor cocktail (Roche, Indianapolis, IN, USA). GST fusion proteins immobilized on GST-Sepharose beads were incubated with HEK293T cell lysates at 4 °C for 1 h. Beads were washed three times with lysis buffer.