Proteolytic Cleavage of the p65-RelA Subunit of NF-κB during Poliovirus Infection*

Activation of NF-κB during viral infection is one of the critical elements in innate immune response. Several virus-specific factors, such as double-stranded RNA, can trigger host defense mechanisms by inducing NF-κB-mediated expression of cytokines and interferons. Early stages of poliovirus infection are also associated with degradation of IκBα and translocation of NF-κB into the nucleus. However, at later stages of poliovirus replication the p65-RelA component of the NF-κB complex undergoes a specific cleavage that coincides with the onset of intensive poliovirus protein synthesis and the appearance of the activity of poliovirus protease 3C. Indeed, the p65-RelA amino acid sequence contains the recognition site for 3C, and recombinant protein 3C was shown to be capable of proteolytic cleavage of p65-RelA, generating truncated product similar to that observed during poliovirus infection. Cleavage of p65-RelA occurs during replication of ECHO-1 and rhinovirus 14, suggesting that inactivation of NF-κB function by proteolytic cleavage of p65-RelA is the common mechanism by which picornaviruses suppress the innate immune response.

The innate immune response to viral and bacterial infections involves the production of immune molecules, including cytokines, interferons, chemokines, and major histocompatibility complex proteins that act in combination to suppress infectious agents (1). Expression of the genes encoding the majority of these proteins is regulated by NF-B. NF-B activation was reported during many viral infections, such as cytomegalovirus, human immunodeficiency virus, rhinovirus, and measles virus (2)(3)(4). Viruses can activate NF-B through doublestranded RNA (dsRNA) 1 intermediate (2,5,6) and viral proteins (7,8) via activation of Toll-like receptors (TLRs) (9). In particular, TLR3 and dsRNA-dependent protein kinase are involved in activation of a variety of pro-inflammatory cellular genes through the NF-B pathway upon recognition of dsRNA, a replication intermediate of RNA-containing viruses, and the products of transcription of DNA-containing viruses (10 -12).
We found that early stages of poliovirus infection also involve activation of NF-B. This seems to be a common property of picornaviruses because a similar effect was detected during rhinovirus and Theiler murine encephalomyelitis (TME) virus infection (10,13), suggesting that the infected cell makes an attempt to develop innate immune response. However, as we further demonstrated, p65-RelA subunit of the NF-B complex undergoes rapid site-specific proteolytic cleavage by poliovirus protease 3C. We found that representatives of ECHO and rhinoviruses use a similar mechanism of suppression of NF-B response, suggesting that it is common for Picornaviridae. This mechanism, along with suppression of host cap-dependent translation (14), inhibition of transcription of cellular genes (15), and abrogation of secretion of cytokines and of presentation of their receptors on cell surface (16,17), acts to make picornavirus infection resistant to the innate immune response. These activities may explain why poliovirus infection does not stimulate inflammation (16).

EXPERIMENTAL PROCEDURES
Cell Culture and Virus Infections-HeLa and HeLaBcl-2 cell lines were cultured in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 10% fetal calf serum. The infection of HeLa and HeLaBcl-2 cells was done by poliovirus strain Mahoney with multiplicity of infection 10 (17). HeLa cells were infected by T7 RNA polymerase-expressing vaccinia virus (strain Ankara) with multiplicity of infection 1 (18).
Plasmids-Expression vector for FLAG-and His-tagged poliovirus protein 3C was generated by PCR cloning into pCR2.1 plasmid (Invitrogen). This plasmid contains the T7 RNA polymerase-specific promoter. The primers specific for the protein 3C coding sequence were used for PCR (5Ј-GGGCCTGGGTTTGACTATGCA-3Ј and 5Ј-TTGGCTCTGAGT-GAAGTAGA-3Ј). The FLAG-coding sequence, the ATG codon to the 5Ј-end of protein 3C cDNA, the polyhistidine tag-coding sequence, and the termination codon to the 3Ј-end of protein 3C cDNA were added by PCR. The sequencing proved the integrity of the tagged protein 3C cDNA molecules. The ability of the vectors to express proteins was tested by Western blotting.
NF-B Electrophoretic Mobility Shift Assay-The protocol for these experiments was described in our previous publication (19). The cytoplasmic and nuclear protein extracts were purified from 5 ϫ 10 6 HeLa cells, according to the Dignam protocol (20), after 1, 2, 3, 4, and 6 h of poliovirus infection. The NF-B activation by TNF was initiated at 1 h prior to protein purification.

Polyhistidine Affinity Protein Purification and Protease 3C Assay-
The polyhistidine-containing protease 3C was affinity purified on nickel-nitrilotriacetic acid beads (Dynal). 3 ϫ 10 6 HeLa cells were infected by T7 RNA polymerase-expressing vaccinia virus for 2 h and then transfected with expression vector that contained a FLAG-3C-His. The cells were collected on ice and resuspended in 500 l of lysis buffer (50 mM sodium phosphate buffer, pH 7.5, 50 mM NaCl, 10 mM imidazole, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml aprotinin). The cytoplasmic protein extract was purified according to the Dignam protocol (20)

FIG. 1. Poliovirus infection stimulates translocation of NF-B into the nucleus.
A, electrophoretic mobility shift assay of protein extracts from the cytoplasm and the nuclei of poliovirus-infected HeLa cells. 10 g of protein extracts from cytoplasm (1)(2)(3)(4)(5) and from the nuclei (6 -11) were used for assay with a 32 P-labeled NF-B-specific oligonucleotide probe. Arrows show the position of the p65/p50 dimers. B, Western blotting analysis is shown of the protein extracts from the cytoplasm and the nuclei of poliovirus-infected HeLa cells. 10 g of cytoplasmic (1-5) and nuclear (6 -10) protein extracts were analyzed with anti-p65 antibodies specific to the C terminus of the protein. Control for poliovirus infection was done with anti-protein 3A antibodies. These antibodies recognize poliovirus proteins 3AB and 3A. Coomassie blue staining was used as a protein loading control. C, the activation of NF-B by poliovirus is accompanied by IB␣ degradation. Western blotting analysis of the cytoplasmic protein extracts from poliovirus-infected HeLa cells with anti-IB␣ antibodies. IB␣ degradation between 2 and 3 h postinfection coincided with the time of NF-B activation. Western blotting with anti-actin antibodies was used as a protein loading control. D, Western blotting analysis is shown of the cytoplasmic protein extracts from TNF/CHItreated and TNF-treated HeLa cells with anti-IB␣ antibodies. IB␣ reappeared after degradation because of the NF-B-specific synthesis. IB␣ reappearance after degradation is affected in TNF/CHI-treated cells because of the blockade of translation by CHI. and was fractionated on nickel-nitrilotriacetic acid beads according to the manufacturer's protocol (Dynal). The proteins from the beads were analyzed by immunoblotting. In vitro cleavage assay was performed with affinity-purified protease FLAG-3C-His. 100 ng of column-purified protease 3C was incubated for 30 min at 37°C with 20 g of cytoplasmic protein extract from HeLa cells in cleavage buffer (100 mM NaCl, 5 mM MgCl 2 , and 10 mM HEPES-KOH, pH 7.4). The reaction was terminated by addition of loading buffer, and proteins were separated by SDS-PAGE.

RESULTS
Poliovirus Infection Leads to Activation of NF-B-The effect of poliovirus infection on NF-B function was analyzed by monitoring translocation of the NF-B complex into the nucleus using electrophoretic mobility shift assay (Fig. 1A) and Western immunoblotting (Fig. 1B). According to both assays, the poliovirus infection stimulated translocation into the nucleus of the NF-B complex at 2-3 h postinfection. The translocation of NF-B into the nucleus was accompanied by the degradation of IB␣ (Fig. 1C). The scale of NF-B activation in poliovirus-infected HeLa cells was similar to that caused by TNF treatment (Fig. 1A). However, the reappearance of IB␣ in TNF-treated cells was not observed in poliovirus-infected cells, presumably as a result of poliovirus-mediated block of cap-dependent translation (Fig. 1, C and D).
Degradation of p65-RelA at Later Stages of Poliovirus Infection-Poliovirus-induced accumulation of the NF-B complex in the nucleus occurred between 2 and 3 h after infection and was followed by the degradation of p65-RelA and the appearance of a shorter fragment. These events were consistent with proteolytic cleavage of the protein (Fig. 1B and Fig. 2A). Because the truncation of the p65-RelA structure could be functionally significant, we followed this observation in more detail. Western blot analysis of p65-RelA with antibodies specific for the C terminus of the protein showed a rapid and profound decrease of p65-RelA in the cytoplasm of infected cells between 2 and 4 h after infection and complete disappearance from the cytoplasm and the nucleus by 6 h after infection ( Fig. 1B and Fig. 2A). Antibodies specific for the N terminus of p65-RelA revealed the accumulation of a shorter product, named p65dC (Fig. 2B). The total protein levels did not show a significant decline even at late stages of poliovirus infection (Fig. 2C). These results indicate that p65-RelA undergoes proteolytic cleavage during poliovirus infection resulting in removal of the C-terminal fragment of the protein containing a transactivation domain. In contrast to p65-RelA subunit, the amounts of the p50 component of the NF-B complex only slightly decreased during infection, probably as a result of general protein synthesis shutdown ( Fig. 2A), with no indications of protein truncation.
Truncation of p65-RelA Is Not the Result of Caspase-mediated Cleavage-It was previously shown that caspase-3 and caspase-6, activated during apoptosis, could cleave p65-RelA close to the C terminus (23)(24)(25). This cleavage results in creation of the products that fail to act as the transactivators and turn into the inhibitors of transcription (25).
According to previous reports, enterovirus infection does not involve the appearance of the markers of apoptosis (26); on the contrary, it was shown to be associated with the suppression of apoptosis (27). However, we examined the possibility that p65dC in poliovirus-infected cells appears as a result of caspase-mediated cleavage (28,29). We analyzed whether poliovirus infection stimulates activation of caspase-3 from its inactive form (30) and the appearance of the products of caspase-mediated cleavage of keratin 18 (21) and lamin A (31) in infected cells. In contrast to apoptotic cells, these proteins did not show apoptosis-specific cleavage in poliovirus-infected cells (Fig. 3, A-C), indicating a lack of caspase activation.
Lack of involvement of apoptotic mechanisms in p65-RelA degradation was also evident from the results of analysis of HeLaBcl-2, the poliovirus-infected variant of HeLa cells overexpressing anti-apoptotic protein Bcl-2 (32). These cells became resistant to Fas-mediated apoptosis (Fig. 4A) and did not show apoptosis-specific cleavage of p65-RelA (Fig. 4B). At the same time, they were as susceptible to poliovirus infection as the original HeLa cells (Fig. 4A), including the poliovirus-specific cleavage of p65-RelA (Fig. 4B). The pattern of p65-RelA cleavage in HeLaBcl-2 poliovirus-infected cells was not changed (Fig. 4C).
Finally, we examined the effect of the pan caspase inhibitor z-VAD-fmk on p65-RelA cleavage during poliovirus infection. Although z-VAD-fmk protected HeLa cells from Fas-induced apoptosis and from apoptosis-specific cleavage of p65-RelA (Fig. 5A), it had no effect on poliovirus infection and on poliovirus-specific cleavage of p65-RelA (Fig. 5A). Moreover, the cleavage products of p65-RelA in poliovirus-infected HeLa cells and in apoptotic HeLa cells had different molecular masses (Fig. 5B). All these observations indicate that generation of p65dC during poliovirus infection has no relation to apoptosis and caspases and is likely to be carried out by virus-specific proteins. NetPicornaRNA web-based software (37) revealed the amino acid sequence specific for the protease 3C cleavage close to the C terminus of the p65-RelA protein (Fig. 2D). The size of the poliovirus-specific truncated form of p53-RelA, p65dC, was consistent with that expected from the protease 3C cleavage site on p65-RelA (Fig. 2D). Moreover, the initiation of cleavage of p65-RelA during infection coincided with the appearance of 3C protease activity detected by cleavage of the poliovirus polypeptide (Fig. 2, A and B).
To check whether poliovirus-specific protease 3C can induce proteolytic cleavage of p65-RelA, we generated recombinant protein 3C with a FLAG tag on its N terminus and a His tag on its C terminus expressed from the T7 RNA polymerase/vaccinia virus expression system. FLAG-3C-His purified by affinity chromatography on Ni beads was incubated in vitro with cyto-plasmic protein extract from HeLa cells. Western immunoblotting with the antibodies specific to the N terminus of p65-RelA showed the cleavage of p65-RelA protein in the protein extract incubated with protease 3C (Fig. 5C) similar to that occurring during poliovirus infection.
Infection with ECHO-1 and Rhino-14 Viruses Stimulates p65-ReLA Cleavage-Is the ability to stimulate p65-RelA proteolytic cleavage unique for the poliovirus infection, or does it occur during infection by other picornaviruses? To address this question, we analyzed the status of p65-RelA during infection with ECHO-1 and rhino-14 viruses. Rhinovirus was especially interesting because activation of NF-B was previously reported for this virus infection (13). The patterns of p65-RelA cleavage were very similar during poliovirus and ECHO-1 virus infections (Fig. 6, A and B). Rhinovirus 14 infection also  5. The cleavage of p65-RelA is protease 3C-specific. A, The z-VAD-fmk inhibitor of caspases did not prevent poliovirus-specific cleavage of p65-RelA. HeLa cells were infected with poliovirus for 5 h or treated with anti-Fas antibodies (Fas) for 7 h. 10 g of total protein extracts from these cells were analyzed by Western blotting with anti-p65-RelA (C terminus-specific) antibodies. Western blotting with antiactin antibodies was used as a loading control. The z-VAD-fmk inhibitor was added 2 h before infection or anti-Fas treatment. Caspase inhibitor protected from apoptosis-specific p65 cleavage but did not protect from poliovirus-specific cleavage of p65-RelA. B, cleavage of p65-RelA during poliovirus infection and during apoptosis generated the truncated forms of different molecular mass. 10 g of total protein extracts from HeLa cells treated with anti-Fas antibodies (Fas) for 5 h, HeLa cells infected with poliovirus for 3 or 4 h, and control HeLa cells were analyzed by Western blotting with anti-p65-RelA antibodies (N terminus-specific). Poliovirus-specific cleavage generated the truncated form of p65-RelA of higher molecular mass than the truncated forms of p65-RelA generated by caspase-specific cleavage during apoptosis. C, the presence of poliovirus-specific protease 3C stimulates p65-RelA cleavage. FLAG-3C-His protein was purified by Ni affinity chromatography. The cytoplasmic protein extracts from HeLa cells were incubated with FLAG-3C-His for 30 min at 37°C. Western blotting analysis was done with anti-p65-RelA N terminus-specific antibodies and anti-FLAG antibodies. The product of p65-RelA cleavage with poliovirus protease 3C has the same molecular mass as the p65-RelA truncated form from poliovirus-infected cells. stimulated p65-RelA cleavage in HeLa cells, but the cleavage started later than in poliovirus-infected cells, consistent with the slower growth cycle of this virus (Fig. 6, C and D). DISCUSSION Organisms react to infection through the innate immune response. It starts from the activation of NF-B; as a result, the expression of genes encoding pro-inflammatory cytokines is stimulated, including TNF, interleukin 6, interleukin 8, chemokines, and adhesion molecules (38). These molecules are involved in the immunity process by recruiting the immune cells to the site of infection. NF-B is also a component of the activation of the important anti-viral cytokine, interferon-␤ (39). The expression of MHCI protein genes is regulated by NF-B as well (40). New anti-viral activity of NF-B was reported recently for cells infected by HPIV-3 (41).
RNA-containing viruses activate the NF-B-specific response and interferon synthesis via activation of Toll-like receptor 3 by dsRNA (7,42). Activation of NF-B was described for several picornaviruses, including rhinovirus and TME virus infections (10,13). In this study, we report that poliovirus infection also initiates NF-B activation that takes place between 2 and 3 h after infection. This is in contrast with rhinovirus and TME virus infections, which initiate NF-B activation after infection either later (8 h) or earlier (30 min) (10, 13). The time of poliovirus-specific activation of NF-B correlated well with the time of dsRNA synthesis (43), suggesting it may be the trigger for NF-B activation. As happens with NF-B activation by many other factors, such as cytokines and dsRNA (44), the poliovirus-specific activation of NF-B was preceded by the degradation of IB␣ (Fig. 1C), indicating that it goes through the conventional mechanism.
To counteract the effect of NF-B, some viruses develop activities that interfere with virus-specific NF-B activation. For instance, NS1 protein of the influenza A virus prevents the activation of NF-B (45). The African swine fever virus has two homologues of IB with potent anti-inflammatory effect (46). Adenovirus E1A protein affects TNF-specific NF-B stimulation (47). Finally, many members of the Poxvirus genus encode the proteins that interfere with the positive regulation of NF-B activation (11,48). All these examples show that anti-inflammatory activities are common among viruses of different origin and that these activities are important for viral infection.
In this communication, we report a new mechanism used by picornaviruses to mitigate the NF-B-dependent innate immune response. This mechanism consists of rapid and efficient proteolytic cleavage of the p65-RelA component of the NF-B complex by the virus-encoded protease 3C. Similar proteolytic cleavage of the C terminus of p65-RelA was shown to inactivate the protein (25). This cleavage of p65-RelA starts ϳ3 h after poliovirus infection. After 4 h of infection, most of the p65-RelA molecules in poliovirus-infected cells are truncated. The beginning of the degradation coincides with a time of intensive poliovirus protein synthesis and does not involve caspases (Figs. 3A and 5A). The ability of protease 3C to cleave p65-RelA may explain the pro-apoptotic activity of this protease described previously (49), because the activation of NF-B in the majority of cases has a strong anti-apoptotic effect (50). Based on these data, we suggest that picornaviruses developed the mechanism of p65-RelA degradation. This mechanism is especially important for them because the inhibition of cellular mRNA translation destroys the IB␣-dependent regulatory loop in infected cells (Fig. 1C), thus abrogating the natural mechanism of negative regulation of NF-B response. The proteolytic cleavage of p65-RelA resolves this problem and limits the time of NF-B activation to less than 2 h in HeLa cells.
Why would picornaviruses benefit from the specific suppression of NF-B if they have a more general mechanism suppressing cap-dependent translation that should affect NF-B-mediated protein expression? Indeed, poliovirus-infected cells fail to restore their IB␣ levels (Fig. 2). However, some of the cellular mRNAs for NF-B-responsive genes, unlike IB␣ mRNA, can be efficiently translated regardless of poliovirus infection (51). These include such well characterized NF-B targets as interleukin 6, interleukin 8, and c-Myc (51). Hence, poliovirus protein 2A-mediated inhibition of host translation is not enough to completely shut down the innate immune response.
Poliovirus is known to inhibit not only translation but also the transcription of cellular genes acting against RNA polymerase II (52)(53)(54) or specific transcription factors, such as cAMPresponse element-binding protein (15,55). In this article, we report a new target for poliovirus-specific inhibition of cellular gene transcription, p65-RelA protein, the critical component of the NF-B complex.
The cleavage of the C terminus of p65-RelA can be also detected in cells infected by other picornaviruses such as ECHO-1 and rhinovirus 14, thus reflecting the common nature of the discovered phenomenon. Together with previously described activities (inhibition of translation, transcription, nuclear shuttling, protein trafficking, and receptor presentation), this mechanism is used by the virus to modify cellular processes to its own benefit and is served as part of a viral strategy FIG. 6. Infection by ECHO-1 and rhino-14 viruses stimulated p65-RelA cleavage. Western blotting analysis of total protein extracts from ECHO virus 1-infected HeLa cells with p65-RelA C terminusspecific antibodies (A) or with p65-RelA N terminus-specific antibodies (B) represented a pattern of p65 cleavage similar to that observed in poliovirus-infected cells. C and D, the rhinovirus infection stimulated p65-RelA C terminus cleavage. The analysis is shown of total cellular protein extracts from poliovirus-and rhinovirus-infected HeLa cells with C terminus-specific (C) and N terminus-specific (D) anti-p65-RelA antibodies. The molecular masses of the p65 truncated forms (p65/dC) were similar during the infection with poliovirus, ECHO virus, and rhinovirus.

p65-RelA Cleavage in Picornavirus Infection
of effective replication and abrogation of defense mechanisms of the host.