Inhibition of Nuclear Factor kB Activation by a Virus-encoded IkB-like Protein*

Certain viruses have evolved mechanisms to counteract innate immunity, a host response in which nuclear factor kB (NF-kB) transcription factors play a central role. African swine fever virus encodes a protein of 28.2 kDa containing ankyrin repeats similar to those of cellular IkB proteins, which are inhibitors of NF-kB. Transfection of the African swine fever virus IkB gene inhibited tumor necrosis factoror phorbol ester-induced activation of kBbut not AP-1-driven reporter genes. Moreover, African swine fever virus IkB co-immunoprecipitated with p65 NF-kB, and the purified recombinant protein prevented the binding of p65-p50 NF-kB proteins to their target sequences in the DNA. NF-kB activation induced by tumor necrosis factor, as detected by mobility shift assays or by transfection of kB-driven reporter genes, is impaired in African swine fever virusinfected cells. These results indicate that the African swine fever virus IkB gene homologue interferes with NF-kB activation, likely representing a new mechanism to evade the immune response during viral infection.

One of the hallmarks of innate immunity consists of the ability of the infectious agent to generate cytokines that help mount the inflammatory response and recruit immune cells to the site of infection. Work carried out in the past several years has identified nuclear factor B (NF-B) 1 as one of the most important elements coordinating such responses (1). NF-B controls the transcription of many different genes involved in many aspects of the inflammatory and immunological responses. Of particular importance for innate immunity are cytokines, cytokine receptors, chemokines, adhesion molecules, and acute-phase response genes (2)(3)(4)(5).
NF-B inducible transcriptional activators are a family of transcription factors that includes the dorsal gene of Drosophila and the mammalian genes nfkb-1 (p105-p50), c-rel, relA (p65), nfkb-2 (p100-p52), and relB, all involved in the regulation of gene transcription (2,3). NF-B is composed of homo-or heterodimers, with the subunit composition of the different complexes defining the fine specificity of binding to the target sequences and their transactivating activity (6). All members share a homologous 300-amino acid Rel region containing the three essential domains for their activity: the DNA-binding, dimerization, and nuclear localization domains (5).
In response to a variety of activators, IB-␣ proteins undergo phosphorylation of Ser 32 and Ser 36 (20,21), rendering IB susceptible to proteolysis via the ubiquitin-proteasome pathway (22,23). This unmasks the nuclear localization sequence of the transactivating heterodimers, allowing translocation of active NF-B complexes to the nucleus. Furthermore, it has been recently shown that phosphorylation of tyrosine 42 of human IB-␣ leads to activation of NF-B without degradation of the IB protein (24). In addition to retaining NF-B in the cytoplasm, IB-␣ prevents p65 and c-Rel binding to DNA in vitro (10,25).
African swine fever virus (ASFV) is a large DNA virus that infects different species of suids, causing an acute and frequently fatal disease (26). ASFV mainly replicates in macrophages and monocytes, and this may have major effects on the pathogenicity of the disease (27). Infection by ASFV is characterized by the absence of a neutralizing immune response, which has prevented the development of a conventional vaccine. Moreover, it has been speculated that ASFV might have mechanisms to counteract the host's immune response, as in the case of other viruses (28,29). The genome of ASFV is a double-stranded DNA with a size ranging from 170 to 190 kilobase pairs, depending on the virus strain. Recently, the entire genome of the ASFV BA71V isolate, a strain adapted to grow in Vero cells, was completely sequenced, and 151 open reading frames (ORFs) were detected (30). One of these genes, A238L, contains ankyrin repeats homologous to those found in the IB family. We show in this report that the protein encoded by the A238L gene, expressed in Escherichia coli and purified, behaves as a bona fide IB-␣ viral homologue since it binds p65 NF-B and prevents the binding of p65-p50 NF-B dimers to their target sequence in the DNA. Our work also demonstrates that ASFV IB specifically inhibits B-driven reporter genes and that ASFV-infected cells have an impaired ability to activate NF-B at the protein and functional levels.

EXPERIMENTAL PROCEDURES
Cells, Viruses, and Reagents-Vero (African green monkey) and Jurkat (human) cells were obtained from the American Type Culture Collection and grown in Dulbecco's modified Eagle's medium containing 10% newborn calf serum. The Vero-adapted ASFV strain BA71V was propagated and titrated as described previously (31). Recombinant TNF (10 7 units/mg) was a generous gift from Pharmacia Spain. All other reagents were obtained from Sigma, except where indicated.
Plasmids-pCMV-IB-␣ was a generous gift of Dr. F. Arenzana-Seisdedos and contains full-length IB-␣ under control of the cytomegalovirus (CMV) immediate-early promoter. The NF-B-dependent plasmid p3ConA-Luc vector is driven by three synthetic copies of the NF-B consensus sequence of the immunoglobulin chain promoter and was a generous gift of Dr. J. Alcamí. The AP-1-dependent plasmid AP-1-Luc vector is driven by three synthetic copies of the AP-1 consensus sequence and was a generous gift of Dr. J. Moscat.
The A238L gene was cloned under control of a CMV early promoter into the pRc/CMV expression vector (Invitrogen), which also carries the prokaryotic phage T7 RNA polymerase promoter. Briefly, the A238L gene was amplified by polymerase chain reaction using oligonucleotides 5Ј-GCGCGCAAGCTTATGGAACACATGTTTCAAG-3Ј and 5Ј-CGCGC-GTCTAGATTACTTTCCATACTTGTTC-3Ј as primers. The first primer was designed with a GCGCGC tail and a HindIII site, and the second contains a CGCGCG tail and an XbaI site. The polymerase chain reaction product was digested with HindIII and XbaI and cloned into the pRc/CMV vector.
Expression of the A238L Gene in E. coli-The A238L ORF lacking the first 8 nucleotides was cloned in the expression vector pRSET-A. Plasmid p2SIЈ (32), containing the A238L ORF, was digested with AflIII and treated with Klenow fragment. A 1.4-kilobase AflIII restriction fragment was then purified from agarose gels by electroelution. The pR-SET-A vector was digested with PvuII, treated with calf intestinal phosphatase, and ligated to the 1.4-kilobase blunt-end AflIII fragment obtained from plasmid p2SIЈ. E. coli strain BL21 cells were transformed with the recombinant plasmid pRSET-A238L.
Affinity Purification of Recombinant ASFV IB-A 1-liter culture of E. coli cells harboring the pRSET-A238L plasmid was induced with 0.4 mM isopropyl-␤-D-thiogalactopyranoside for 2 h at 37°C. The fusion protein present in the cell lysate was purified under denaturing conditions using a 5-ml Ni 2ϩ affinity chromatography column (QIAGEN Inc.) according to the manufacturer's instructions. The ASFV IB protein (eluted from the column with 8 M urea, 0.1 M NaH 2 PO 4 , and 0.01 M Tris-HCl, pH 4.5) was renatured by dialysis against 25 mM Tris-HCl, pH 8.0, containing 1 mM EDTA, 1 M NaCl, 5% glycerol, and 0.5% Nonidet P-40.
Preparation of Antibodies-To prepare anti-peptide antibodies specific for the ASFV IB protein, a 15-amino acid peptide (GGSG-GCVKKLNKYGK) was synthesized. The last 9 amino acids of this peptide (underlined) correspond to the carboxyl-terminal region of ASFV IB, which is not conserved in cellular IB-␣. The peptide was conjugated to keyhole limpet hemocyanin and used to immunize rabbits. The immune serum obtained recognized the recombinant ASFV IB protein on Western blots.
Electrophoretic Mobility Shift Assays (EMSA)-The binding assays with nuclear extracts from Vero or Jurkat cells were performed as reported (33), using as 32 P-labeled probe either a B oligonucleotide (5Ј-AGTTGAGGGGACTTTCCCAGGC-3Ј) from the immunoglobulin chain enhancer or an Oct-1 oligonucleotide (5Ј-TGTCGAATGCAAAT-CACTAGA-3Ј). The binding complexes were separated on a 5% acrylamide gel, and their specificity was determined by competition with a 50-fold molar excess of the same unlabeled oligonucleotide. The protein composition of each DNA-protein complex was identified by supershifting assays using specific polyclonal rabbit anti-NF-B antisera. These assays were performed as normal EMSAs, except that 1 l of the different anti-NF-B antisera was added to the nuclear extract prior to the addition of the probe. The anti-c-Rel, anti-p50, and anti-p65 antisera were kindly provided by Drs. Nancy Rice and Alain Israël. In some experiments, purified ASFV IB was added to the nuclear extracts.
Transfection Assays-Vero or Jurkat cells (1 ϫ 10 7 ) resuspended in 0.5 ml of Dulbecco's modified Eagle's medium containing 10 g of linearized plasmid DNA were subjected to electroporation in a Bio-Rad Gene Pulser (960 microfarads, 250 V). For those experiments with infected cells, 2 h before electroporation, the cells were mock-infected or infected with ASFV or vaccinia virus at a multiplicity of 10 plaqueforming units/cell. After electroporation, the cells were resuspended in 10 ml of fresh Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum and cultured overnight at 37°C before being activated with phorbol 12-myristate 13-acetate (PMA; 10 ng/ml) or TNF (30 ng/ml). Cells were incubated for an additional 4-h period, harvested, and lysed. Luciferase activity was measured in a luminometer and expressed as relative luciferase units, calculated as light emission from the experimental sample divided by light emission from untransfected cells per 10 6 cells.
Immunoprecipitation-COS cells (1 ϫ 10 7 ), resuspended in Dulbecco's modified Eagle's medium containing 10 mM Hepes and 10% fetal calf serum, were transfected by electroporation with 10 g of pCMV, pCMV-IB-ASFV, or pCMV-IB-␣ linearized plasmids, together with 10 g of pCMV-p65, as described above. After electroporation, the cells were resuspended in fresh medium supplemented with 10% fetal calf serum and cultured overnight at 37°C. The cells were then labeled with a mixture of [ 35 S]methionine and [ 35 S]cysteine (50 Ci/1 ϫ 10 6 cells; Amersham Corp.) for 24 h. Lysates were prepared by resuspending the cells at 1 ϫ 10 7 cells/ml in 1% digitonin lysis buffer containing protease inhibitors (1% digitonin, 10 mM triethanolamine, pH 8.0, 150 mM NaCl, 1 mM EDTA, 1 g/ml leupeptin, 1 g/ml aprotinin, and 1 g/ml pepstatin) as described (34). The cell extracts obtained from 1 ϫ 10 7 cells for each immunoprecipitation were precleared three times with protein A-Sepharose and then immunoprecipitated as described (34) using 30 g of ASFV IB-specific anti-peptide antiserum or polyclonal anti-IB-␣ antiserum (Santa Cruz Biotechnology, Inc.) and protein A-Sepharose. For re-precipitation analysis, bound proteins were solubilized by boiling in 400 l of SDS-containing buffer (0.4% SDS, 50 mM triethanolamine, 100 mM NaCl, 2 mM EDTA, and 2 mM 2-mercaptoethanol). Subsequently, 100 l of 10% Triton X-100 and 10 mM iodoacetamide were added to the supernatants. The supernatants were precleared twice with protein A-Sepharose, diluted with 1 volume of 1% digitonin lysis buffer, and immunoprecipitated with 20 g of polyclonal anti-p65 antiserum (Santa Cruz Biotechnology, Inc.) and protein A-Sepharose. Bound proteins were eluted by boiling with electrophoresis sample buffer and resolved by SDS-polyacrylamide gel electrophoresis. For autoradiography, the gel was exposed on a Fujifilm BAS-MP 20405 imaging plate at room temperature. The exposed imaging plate was analyzed with a Fuji BAS 1500 analyzer.
Computer Analysis-Computer analyses of DNA and protein sequences were performed with the software package of the University of Wisconsin Genetics Computer Group. Data base searches were done with the programs FASTA and TFASTA. Protein patterns were searched using the MacPattern program and the PROSITE and BLOCKS data bases. Multiple alignment of protein sequences was performed with the PILEUP program. To identify ankyrin repeats in the ASFV A238L ORF, the method described by Bork (35) was followed.

RESULTS
Homology of A238L to IB-␣ Proteins-One of the ORFs of ASFV, denoted A238L, encodes a protein of 238 amino acids with a predicted molecular mass of 28.2 kDa, which was previously shown to have homology to the IB family of inhibitors of NF-B (30). When the A238L deduced protein sequence was compared with IB-␣ from different species (human, rat, chicken, and pig) following the procedure described by Bork (35), it was shown that it contains several regions that can be aligned with the ankyrin repeats of IB-␣ (Fig. 1). The similarity between the ASFV and cellular proteins is 20 -24% overall, but is greater in the central part of the molecules. The cellular IB proteins contain six ankyrin repeats (35), each of which is composed of 32-36 amino acids. Only five repeats of 30 -33 amino acids were identified in the viral protein. The three most carboxyl-terminal repeats co-align with the corresponding ankyrin repeats of the cellular IB proteins, while the second repeat from the amino terminus of the viral protein encompasses the second and third ankyrin motifs of the cellular proteins. On the other hand, the first repeat of protein A238L, which has been identified with a low score with the PROFILESEARCH program, is located in the multiple alignment between the first and second repeats of the cellular IB proteins.
Inhibition of NF-B Activation by the ASFV IB Gene-The above analysis suggested a putative role for the A238L gene product (hereafter named ASFV IB) as an inhibitor of NF-B. To test this hypothesis, we isolated and inserted the gene into pCMV under control of the immediate-early CMV promoter, and this plasmid (pCMV-IB-ASFV) was transfected by electroporation into Vero or Jurkat cells, together with a B-driven reporter gene. When Vero cells, a cell line susceptible to infec-tion with ASFV, were transfected with pCMV-IB-ASFV, an almost complete inhibition of the basal activity of the B-driven reporter gene was observed in unstimulated cells. More interestingly, pCMV-IB-ASFV completely prevented the increase in reporter activity observed when the cells were treated with TNF for 4 h ( Fig. 2A).
Similarly, transfection of pCMV-IB-ASFV into Jurkat cells induced a substantial decrease in the basal activity of the B-driven reporter gene and completely prevented the increase in the activity of the reporter after TNF or PMA treatment for 4 h (Fig. 2B). Unlike the induction by PMA, NF-B activation by TNF involves a signal transduction pathway different from protein kinase C (36). However, both were blocked by ASFV IB, suggesting that this gene product is acting either on a common point downstream from both pathways or directly on NF-B itself. As a positive control, a plasmid containing the human IB-␣ gene under control of the same CMV promoter was used. This plasmid was also able to suppress B-dependent activation induced by PMA or TNF. Moreover, the effect of pCMV-IB-ASFV was specific for the B-dependent promoter since a similar plasmid containing luciferase under control of an AP-1 site was not affected at all by pCMV-IB-ASFV (Fig.  2B).
Inhibition of Binding of p65-p50 Complexes to B DNA Sequences by the ASFV IB Protein-To study the function of the protein, recombinant ASFV IB was purified as described under "Experimental Procedures." Coomassie Blue staining of the purified protein after SDS-polyacrylamide gel electrophoresis is shown in Fig. 3. The recombinant ASFV IB protein was added to a nuclear extract from Vero cells either untreated or treated with TNF, and EMSAs were performed. In unstimulated cells, a specific complex that binds to a B oligonucleotide was detected (Fig. 4A, lane 1). Upon treatment with TNF, an additional, slowly migrating complex was observed (Fig. 4A,  lane 2). Those two bands were specifically competed by an excess of the unlabeled B oligonucleotide (Fig. 4A, lane 3), but not by a mutant B oligonucleotide (data not shown). Interestingly, when purified recombinant ASFV IB protein was added to nuclear extracts from TNF-stimulated Vero cells, a dose-dependent displacement of the slowly migrating complex was observed, whereas the inhibition of the faster migrating band was much less pronounced and only observed at the higher concentration (Fig. 4A, lanes 4 and 5). The two complexes detected were identified by supershifting with specific antibodies against the different members of the Rel family (Fig. 4B). The fastest migrating complex present in unstimulated Vero cells was supershifted only by anti-p50 antibodies (Fig. 4B, lane 2) suggesting that it was composed of p50 homodimers. However, the complex induced after TNF treatment was supershifted by anti-p50 and anti-p65 antibodies, but not by anti-c-Rel antibodies (Fig. 4B, lanes 2-4). This may indicate that it contains p65-p50 heterodimers, the most commonly induced NF-B complex in many cell systems (2)(3)(4)(5). ASFV IB was also able to displace the nuclear factors from a preformed complex with DNA (Fig. 4C). Again, it can be seen that the p65-p50 heterodimer was much more efficiently displaced from the com- plex than the p50-p50 homodimer (Fig. 4C, lanes 3 and 4).
This effect of the ASFV IB protein was also seen by using nuclear extracts of TNF-treated Jurkat cells. This cell line is much more sensitive to TNF-mediated activation of NF-B and contained larger amounts of p65-p50 complexes (Fig. 4D, lane  2). The inhibition by ASFV IB (Fig. 4D, lane 4) was specific since the addition of bovine serum albumin (lane 5) or several other proteins (data not shown) did not result in the displacement of any of the NF-B complexes. Moreover, the binding of the Oct-1 nuclear factor to an Oct-1 oligonucleotide using nuclear extracts from the same stimulated Jurkat cells was not affected (Fig. 4E).
A summary from four experiments on the dose-response inhibition by the ASFV IB protein of the binding of nuclear  factors to their specific oligonucleotides is shown in Fig. 5. Whereas the binding of p65-p50 complexes was highly sensitive to ASFV IB inhibition, the binding of p50 homodimers was weakly affected. The binding of Oct-1 factors was unaffected even at concentrations much higher than those that completely displaced the p65-p50 complex. The above results indicate that the ASFV IB protein behaves as an IB-␣ protein, specifically interacting with p65-containing NF-B complexes. To confirm this, COS cells were cotransfected with pCMV vectors expressing p65 and ASFV IB. Cotransfection of the cells with pCMV expressing human IB-␣ or with an empty plasmid containing no IB gene (as positive and negative controls, respectively) was also performed. Cell extracts were prepared and subjected to immunoprecipitation with IB-␣-or ASFV IB-specific antibodies. As shown in Fig. 6 (lanes 2 and 3), each antibody specifically immunoprecipitated (in addition to the corresponding IB protein) a band migrating at a position around 65 kDa. This protein was not observed in immunoprecipitates from cells transfected with the empty pCMV vector (Fig. 6, lane 1). To corroborate that the 65-kDa band represented p65 NF-B, the immunoprecipitates were treated with an SDS-containing buffer and re-precipitated with a polyclonal antibody against p65. In both cases, the specific 65-kDa band was again immunoprecipitated from ASFV IB or IB-␣ precipitates (Fig. 6,  lanes 5 and 6).
Inhibition of NF-B by ASFV Infection-To confirm the role of this ASFV IB protein in infection, Vero cells were infected with ASFV, and the activity of NF-B was studied. The ASFV IB protein can be detected in infected cells by Western blotting at very early times after infection and remained up to 16 h post-infection (data not shown). Moreover, its synthesis was insensitive to cytosine arabinoside treatment, indicating that the A238L gene is an ASFV early gene (37). As in the case of many other viral infections (4, 38), a p65-p50 NF-B complex was detected in the nucleus of unstimulated ASFV-infected cells (Fig. 7). The amount of this complex, as determined by densitometric scanning of the films, was always below 20% of that obtained in the presence of TNF in mock-infected cells and did not further increase with time post-infection (data not shown). Interestingly, no further increase in active p65-p50 NF-B complex able to bind to the B probe was observed in the nucleus of TNF-treated ASFV-infected cells (Fig. 7). This TNF unresponsiveness was observed at any time post-infection (data not shown).
In addition, the inhibition of TNF-induced NF-B activity caused by ASFV infection in Vero cells was confirmed in cells transfected with a B-driven luciferase plasmid. In mock-or vaccinia virus-infected cells, treatment with TNF induced average 2-and 4-fold increases, respectively, in the activity of the B-driven reporter gene. By contrast, in ASFV-infected cells, basically no reporter activity was observed in either unstimulated or TNF-stimulated cells (Fig. 8). More important, this drastic inhibition took place even when the cells were cultured in the presence of cytosine arabinoside, which, as mentioned above, allows the expression of a limited number of the total ORFs of ASFV and blocks viral replication and the concomitant shutoff of host protein synthesis. These results seem to rule out an unspecific effect of ASFV infection on the B-dependent promoter and clearly confirm that ASFV-infected cells have a defect in NF-B activation. DISCUSSION Our results indicate that the ASFV genome encodes a protein with homology to the ankyrin repeats of IB, which behaves as an effective inhibitor of NF-B activity. IB-␣ proteins serve many functions. 1) They prevent binding of NF-B to B DNA sequences. 2) They prevent nuclear translocation of active NF-B. 3) They dissociate NF-B from DNA. 4) Some IB proteins, such as Bcl-3, also serve as transcriptional activators (7,39). As deduced from our studies, ASFV IB is able at least to prevent NF-B binding to DNA and to dissociate already active p65-p50 NF-B dimers from it.
Comparison of sequences indicates that ASFV IB is more homologous to the ankyrin repeats present in IB than to those present in other proteins. Several residues and motifs have been identified in cellular IB-␣ as essential in the various processes that lead to its degradation. Amino-terminal serines (20 -22) and lysines (40) are the sites of inducible phosphorylation and ubiquitination, respectively. The carboxyl-terminal PEST-like sequence also plays a role in signal-dependent proteolysis (41). Moreover, the carboxyl-terminal segment between amino acids 269 and 287 of MAD3 renders the protein constitutively unstable, diminishes its interaction with p65, and is involved in regulating protein half-life (42). In addition, it has been shown that phosphorylation of Tyr 42 leads to activation of NF-B without proteolytic degradation of the inhibitory protein (24). All these residues and motifs are absent in the ASFV IB protein, suggesting that the viral protein cannot be regulated and therefore is a natural, constitutive, and potent suppressor of NF-B activity.
ASFV IB specifically interacts with p65 NF-B since it mainly affects the binding of p65-p50 heterodimers to DNA and co-immunoprecipitates with the p65 protein. The p65-p50 heterodimer is the most common transactivating form of the NF-B/Rel factors and is induced by proinflammatory cytokines (TNF and interleukin-1) (2)(3)(4)(5). The binding of p50 homodimers to DNA was weakly affected by ASFV IB, whereas the binding of other nuclear factors was totally unaffected. In this regard, it is worth mentioning that IB-␣ is also able to bind most of the transcriptionally active heterodimers (p65-p50, c-Rel-p50, p65-p52, RelB-p50, p65-p65, c-Rel-c-Rel), but not inactive p50-p50 and p52-p52 (39).
In addition to retaining NF-B in the cytoplasm, IB-␣ is able to prevent in vitro DNA binding of the p65 and c-Rel subunits and displace them from the DNA (10,25). Binding of p65 to IB-␣ or DNA is mutually exclusive (43). Moreover, IB-␣ can be found in the nucleus, suggesting a role for IB-␣ in negatively regulating transcription (9). ASFV IB might go to the nucleus, thus sparing the need for the ankyrin repeat(s) involved in blocking the nuclear localization system of p65 and providing an explanation for the weak homology detected in one of the five ankyrin repeats.
On the other hand, NF-B is a key element in coordinating the inflammatory and immune responses (1). The activation of this factor is a very effective way of initiating the immune response to viral infection since it preexists in the cytoplasm and can be rapidly activated, inducing a large set of genes involved in the immune response. Thus, there are many examples of NF-B activation during viral infection. However, some viruses (human immunodeficiency virus type 1, human T-cell lymphotropic virus type 1, hepatitis B virus, herpes simplex virus type 1, CMV, Newcastle disease viruses, SV40, Sendai virus, Epstein-Barr virus, influenza virus, and adenovirus) take advantage of the activation of NF-B for their own benefit to turn on their own genes (4,38).
By contrast, activation of NF-B may be deleterious for viruses that need to establish latency or that have a long infecting cycle, as would be the case for a large virus, such as ASFV. Moreover, cells of the macrophage/monocyte lineage play a central role in the immune defense, producing cytokines, NO synthase, etc., genes whose expression is controlled by NF-B (3,4). NF-B is induced in macrophages by a variety of cytokines, TNF being one of the most important. As macrophages are the principal host cells for ASFV, it would be evolutionarily advantageous for this virus to inhibit the activity of NF-B in the infected cells.
The number of virus-encoded gene products that influence host mechanisms and the viral strategies to evade immune responses continue to expand (28,29). Cytokines regulate the inflammatory (i.e. interleukin-1 and TNF) and immune responses and may be directly antiviral (i.e. interferon) or destroy virus-infected cells (i.e. TNF). Therefore, cytokines constitute good targets for virus evasion strategies. During evolution, some large viruses, such as poxviruses, have acquired from their mammalian hosts a variety of genes that encode proteins that appear to specifically target cytokines to presumably avoid their antiviral and proinflammatory activity. Thus, they encode homologues of cytokine receptors (interleukin-1 receptor type 2, TNF receptor, interferon-␥ receptor, and interferon-␣/␤ receptor) that block cytokine activity (44,45). The functional significance of these genes in vivo is poorly understood, but there is evidence that they influence pathogenicity.
ASFV is a large virus that has also acquired genes that may affect the cytokine response (30). However, its strategy seems to be different from that used by other viruses. Instead of targeting proinflammatory cytokines individually, it expresses an IB-like protein that blocks NF-B, a key common effector downstream from many of them. The results described in this work may explain previous data from our laboratory showing that TNF has no antiviral activity in ASFV-infected macrophages (46).
Recently, a report showing that ASFV-infected macrophages have an impaired ability to secrete inflammatory cytokines has been published (47). The authors proposed that ASFV IB might be responsible for this effect on the basis that transfection of the ASFV IB gene inhibited luciferase expression under control of a fragment of the interleukin-8 promoter that contains, in addition to the B site, other sequences likely involved in promoter activity. Our work provides a characterization of the molecular basis of this effect, showing for the first time the following. 1) ASFV IB specifically inhibits B-driven reporter genes since the reporter gene used in our experiments is driven by a promoter consisting exclusively of NF-B consensus sequences. 2) The ASFV IB protein specifically interacts with p65 NF-B, preventing the binding of p65-p50 NF-B dimers to their target sequence in the DNA. It is important to note that these experiments have been performed with a purified recombinant ASFV IB protein, which provides direct evidence that the ASFV gene encodes a protein with IB-like activity. 3) In addition, we show that ASFV-infected cells have an impaired ability to activate NF-B at the protein and functional levels.
In summary, our results provide evidence of a possible mechanism of host evasion by viruses, the inhibition of NF-B activity by an IB-like protein encoded by ASFV. This may have important biological implications. ASFV and other viruses have survived under adverse host conditions and have adapted to block the action of a key mediator of the inflammatory and immune responses, NF-B. Studies with ASFV IB may help to clarify the role of NF-B, the structure-function relationships of IB, and eventually the mechanism of TNF signal transduction.