A parapoxviral virion protein inhibits NF-κB signaling early in infection

Poxviruses have evolved unique proteins and mechanisms to counteract the nuclear factor κB (NF-κB) signaling pathway, which is an essential regulatory pathway of host innate immune responses. Here, we describe a NF-κB inhibitory virion protein of orf virus (ORFV), ORFV073, which functions very early in infected cells. Infection with ORFV073 gene deletion virus (OV-IA82Δ073) led to increased accumulation of NF-κB essential modulator (NEMO), marked phosphorylation of IκB kinase (IKK) subunits IKKα and IKKβ, IκBα and NF-κB subunit p65 (NF-κB-p65), and to early nuclear translocation of NF-κB-p65 in virus-infected cells (≤ 30 min post infection). Expression of ORFV073 alone was sufficient to inhibit TNFα induced activation of the NF-κB signaling in uninfected cells. Consistent with observed inhibition of IKK complex activation, ORFV073 interacted with the regulatory subunit of the IKK complex NEMO. Infection of sheep with OV-IA82Δ073 led to virus attenuation, indicating that ORFV073 is a virulence determinant in the natural host. Notably, ORFV073 represents the first poxviral virion-associated NF-κB inhibitor described, highlighting the significance of viral inhibition of NF-κB signaling very early in infection.


Introduction
Orf virus (ORFV), the prototype member of the genus Parapoxvirus (PPV) of the Poxviridae, is the etiologic agent of contagious pustular dermatitis or orf, a ubiquitous disease of sheep and goats [1]. Orf is characterized by inflammatory, often proliferative lesions affecting the skin and the oral mucosa [2]. Lesions evolve through the stages of erythema, pustules and scabs, and are usually restricted to areas surrounding the virus entry sites [1,2]. If not complicated by secondary infections, orf lesions usually resolve in 6 to 8 weeks [3].
ORFV is highly epitheliotropic, and only keratinocytes or their counterparts in the oral mucosa have been shown to support viral replication in vivo [4,5]. Keratinocytes provide the first physical barrier to invading pathogens, and function as immune sentinels initiating inflammation and promoting skin healing after injury [6]. Keratinocytes express different cytokine receptors, such as tumor necrosis factor (TNF) receptor 1 (TNFR1) and interleukin-1 receptor (IL-1R), and multiple pattern recognition receptors (PRRs) such as toll-like receptors (TLRs) for recognition of pathogen-associated molecular patterns (PAMPs) of bacterial or viral origin [7]. Additional PRRs, such as the cyclic GMP-AMP Synthase (cGAS), retinoic acid -inducible gene 1 (RIG-I)-like receptors and NOD-like receptors (NLRs) recognize viral nucleic acid in the cytoplasm [8]. Engagement of these receptors initiates downstream proinflammatory signaling cascades [6,7], including the nuclear factor-kappa B (NF-κB) signaling pathway, which mediates innate immune responses and contributes to skin homeostasis [9,10].
Several poxviral proteins specifically target the IKK complex, a bottleneck for most NF-κB activating signals, including those involved in nucleic acid sensing and response to infection [19]. Two VACV proteins prevent phosphorylation and subsequent activation of IKK complex. VACV B14 directly interacts and inhibits the activity of IKKβ, while VACV N1L interacts with multiple subunits of the IKK complex [21,23]. MCV FLICE-like proteins (vFLIPs) MC159 and MC160 also target the IKK complex, with MC159 interacting with NEMO and preventing activation of IKKβ and MC160 inducing cytoplasmic degradation of IKKα [33,34].
In general, multiple NF-κB inhibitors encoded by a given poxvirus function at different levels of the NF-κB signaling pathway; however, viruses encoding inhibitors acting at the same level have been described [19]. While targeting multiple branches of the NF-κB pathway, poxviral inhibitors are not completely redundant in vivo as viruses harboring single gene deletions affecting NF-κB inhibitors have been shown to influence aspects of disease [35].
Here, we show that ORFV073, a virion protein unique to parapoxviruses, is an inhibitor of NF-κB signaling that prevents activation of the IKK complex and subsequent nuclear translocation of NF-κB-p65 at early times post-infection. Notably, ORFV073 represents the first poxviral virion-associated NF-κB inhibitor described, highlighting the significance of viral inhibition of NF-κB signaling very early in infection.
To obtain OV-IA82Δ073, OFTu cells were infected with OV-IA82 and transfected with recombination vector pUC57-073LF-GFP-073RF. Similarly, to obtain OV-IA82RV073 Flag , cells were infected with OV-IA82Δ073 and transfected with recombination vector pUC57-073LF-0733xflag-RFP-073RF. Fluorescent plaques indicative of recombinant virus replication were selected and subjected to virus purification by limiting dilution and plaque assays as previously described [37]. Integrity and fidelity of sequences in recombinant viruses were confirmed by PCR and DNA sequencing.

Virus purification and characterization
To obtain semi-purified ORFV for infection experiments, OFTu cells cultured in five T175 were infected with OV-IA82, OV-IA82Δ073 or OV-IA82RV073 Flag (multiplicity of infection, MOI = 0.1) and harvested at 90-95% cytopathic effect (CPE). Cultures were freeze and thawed three times, spun down (1500 rpm, 5 min) to remove cellular debris, and then ultracentrifuged (25000 rpm, 1 h) to pellet virus. Virions were resuspended in MEM, and viral titers were determined by the Spearman and Karber's method and expressed as tissue culture infectious dose 50 (TCID50)/ml.
For virion protein studies, OV-IA82Δ073 and OV-IA82RV073 Flag were purified by double sucrose gradients with modifications [41]. OFTu cells (10 T175 flasks) were infected with OV-IA82Δ073 or OV-IA82RV073 Flag (MOI = 0.1), harvested at advanced CPE, and centrifuged to obtain supernatant and cell pellet fractions. Supernatants were ultracentrifuged to pellet extracellular enveloped virus (EEV) as described above and cell pellets were freeze and thawed three times to release intracellular mature virus (IMV) and centrifuged to remove cellular debris. Both EEV and IMV preparations were centrifuged through a sucrose cushion followed by double sucrose gradient purification. EEV and IMV-containing bands were collected and resuspended in 250 μl 10 mM Tris Hcl. Whole cell lysates (10 μg) from mock and OV-IA82RV073 Flag infected cells (MOI = 10) (24 h p.i) and purified OV-IA82Δ073 and OV-IA82RV073 Flag EEV and IMV virion proteins (10 μg) were resolved by SDS-PAGE, blotted to nitrocellulose membrane and probed with primary antibody against flag (Catalog no. A00187-200; Genscript) or ORFV086 structural protein [42]. Blots were developed with HRPconjugated goat anti mouse secondary antibody (sc-2031; Santa Cruz) and chemiluminescent reagent (Super Signal West Femto, Thermo Fischer).
Establishment of ORFV073 expressing stable cell lines A retroviral expression system (pLNCX2; Clontech) was used to construct HeLa cells constitutively expressing GFP (GFP/HeLa) or ORFV073-GFP (ORFV073GFP/HeLa) fusion protein. GFP or ORFV073-GFP DNA sequences were cloned into plasmid pLNCX2 and transfected into the packaging cell line GP2-293 using Lipofectamine 2000. After 48 h, supernatants containing GFP or ORFV073-GFP-encoding recombinant retrovirus particles were harvested and used to infect HeLa cells. Selection, amplification and maintenance of the individual clones were performed in the presence of G418 (500 μg/ml; Gibco). Expression of control GFP or ORFV073-GFP was monitored by fluorescence microscopy and Western blot using antibody against GFP (sc-9996; Santa Cruz Biotechnology).
To examine the possibility of localization of ORFV073 in endosomes, co-localization of ORFV073 with endosomal marker (Caveolin-1) was performed. OFTu cells mock infected or infected with OV-IA82RV073 Flag (MOI = 10) were fixed at 16 and 24 h p.i, permeabilized and blocked as describe above. Cells were sequentially incubated with primary mouse monoclonal antibody against flag (no. A00187-200; Genscript) and rabbit polyclonal antibody against Caveolin-1 (no. sc-894, Santa Cruz), and Alexa Fluor 488-labeled secondary goat anti mouse antibody (no. A-11001; Thermo Fisher Scientific) and Alexa Fluor 647-labeled secondary goat anti rabbit antibody (no. A-21244; Thermo Fisher Scientific). Cells were then counterstained with DAPI and examined by confocal microscopy (A1, Nikon) To examine co-localization of ORFV073 with ORFV086, OFTu cells mock infected or infected with OV-IA82RV073 Flag (MOI = 10) were fixed at 16 and 24 h p.i, permeabilized and blocked as describe above. Cells were sequentially incubated with primary rabbit monoclonal antibody against flag (no. 14793, Cell Signaling) and mouse monoclonal antibody against ORFV086, and Alexa Fluor 488-labeled secondary goat anti rabbit antibody (no. A-11008; Thermo Fisher Scientific) and Alexa Fluor 647-labeled secondary goat anti mouse antibody (no. A-21236; Thermo Fisher Scientific). Cells were then counterstained with DAPI and examined by confocal microscopy (A1, Nikon).

Growth curves
The replication characteristics of OV-IA82Δ073 was assessed in OFTu cells. Cells cultured in 6-well plates were infected with OV-IA82 or OV-IA82Δ073 using MOI 0.1 (multi-step growth curve) or 10 (single-step growth curve) and harvested at 6, 12, 24, 36, 48, 72 and/or 96 h p.i. Virus titers at each time point were determined as described above. To compare the cytopathic effect (CPE) induced by OV-IA82 and OV-IA82Δ073, OFTu cells were infected with OV-IA82 or OV-IA82Δ073 (MOI = 10) and evaluated under an inverted light microscope at 48 h p.i. (Leica DMI 4000B; 20X).

Real-time PCR analysis
To assess the effect of ORFV073 on NF-κB regulated gene transcription, OFTu cells were mock infected or infected with OV-IA82, OV-IA82Δ073 or OV-IA82RV073 Flag (MOI = 10) and harvested at 1 and 2 h p.i. in the presence of Trizol reagent (Thermo Fisher, Waltham, MA), and RNA samples were processed and reverse transcribed as previously described [36]. The expression of interleukin-8 (IL-8), prostaglandin endoperoxide synthase 2 (PTGS2), C-C chemokine ligand 20 (CCL20) and NF-κB inhibitor alpha (NF-κBIA) genes was assessed using Custom Plus TaqMan Gene Expression Assays (Applied Biosystems) based on ovine gene sequences in GenBank. Real-time PCR and data analysis were performed as previously described [36]. Statistical analysis of the data was performed by using Student's t test.

NF-κB-p65 nuclear translocation assay
To investigate the effect of ORFV073 on nuclear translocation of NF-κB-p65 following ORFV infection, OFTu cells were mock infected or infected (MOI = 10) with OV-IA82, OV-IA82Δ073 or OV-IA82RV073 Flag . Cells were fixed at 30 min and 1, 2, 4, 6, 8, 12 and 24 h p.i. as described above, sequentially incubated with antibody against NF-κB-p65 (no. 8242; Cell Signaling) and with Alexa Fluor 488-labeled goat anti rabbit antibody, counterstained with DAPI, and examined by confocal microscopy. Cells (n = approximately 300 per sample) from randomly selected fields were scored for nuclear NF-κB-p65 and results depicted as the mean percentage of cells expressing nuclear NF-κB-p65 for each time point. Statistical analysis of data was performed by using Student's t test.
To examine the effect of ORFV073 expression on TNFα-induced nuclear translocation of NF-κB-p65, HeLa cells stably expressing GFP (GFP/HeLa) or ORFV073-GFP fusion protein (073GFP/HeLa) were treated with 20 ng/ml of TNFα (Cell Signaling, Danvers, MA). Cells were fixed at 30 min and 1 h post-treatment, sequentially incubated with primary antibody against NF-κB-p65, and Alexa Fluor 594-labeled goat anti rabbit secondary antibody (no. A-11037, Thermo Fisher Scientific), counterstained with DAPI, and examined by confocal microscopy. Cells (n = approximately 200 per sample) from randomly selected fields were scored for nuclear NF-κB-p65 and results depicted as the mean percentage of GFP/073GFP expressing cells containing nuclear NF-κB-p65 for each time point. Statistical analysis of data was performed by using Student's t test.
To evaluate the effect of protein synthesis inhibitor cycloheximide (CHX) on nuclear translocation of NF-κB-p65 during ORFV infection, OFTu cells mock treated or treated with CHX (50 μg/ml) (Sigma-Aldrich, St. Louis, MO) for 30 min were mock infected or infected with OV-IA82, OV-IA82Δ073 or OV-IA82RV073 Flag (MOI = 10) in absence or presence of CHX (50 μg/ml) for 1 h. Nuclear translocation assays and data analysis were performed as described above. As a control for CHX activity, OFTu cells mock treated or treated with CHX (50 μg/ml) for 30 min were mock infected or infected with OV-IA82RV073 Flag and harvested at 30 min and 1 h p.i. Whole cell protein extracts (50 μg) were resolved by SDS-PAGE, and transferred to nitrocellulose membranes and probed with antibody against p53 (sc-6243; Santa Cruz) and actin (sc-8432; Santa Cruz) as described above.
To assess the effect of ORFV073 on NEMO levels, OFTu cells, mock infected or infected with OV-IA82 or OV-IA82Δ073 (MOI = 10) were harvested at 30 min, 45 min, 1 h, 1 h 30 min and 2 h p.i., and cytoplasmic protein extracts were prepared using NE-PER Nuclear and Cytoplasmic Extraction Reagents following manufacturer's protocol (Thermo Fisher, Waltham, MA). Extracts (50 μg) were resolved by SDS-PAGE, blotted to nitrocellulose membranes, probed with NEMO (sc-8330, Santa Cruz) and GAPDH antibodies, and developed as described above. Densitometric and statistical analysis of the blots was performed as described above.

Animal inoculations
To evaluate the effect of ORFV073 on ORFV virulence in the natural host, five-month-old lambs were randomly allocated to three experimental groups, OV-IA82Δ073 (n = 4), OV-IA82RV073 Flag (n = 4) and mock (n = 3). Following anesthesia, the mucocutaneous junction of the lip near the right labial commissure and the inner sides of hind limbs were scarified along 2 cm and 5 cm-long lines, respectively, and virus inoculum (0.5 ml) containing 10 7.5 TCID50/ml was applied topically to each inoculation site using cotton swabs. The scarified areas of the lips were monitored for 21 days for the presence of characteristic orf lesions. Criteria assessed were extent of erythema, papules, pustules, and attached scab. Each criterion was scored according to the width of the lesion along the line of scarification: 1, lesion < 0.5 cm across; 2, lesion 0.5 cm-1 cm across; 3, lesion > 1 cm across, and the total daily score for each lamb was the sum of scores of the four lesion types. Skin biopsy specimens were collected at days 2, 5, 8, 12 and 21 p.i., fixed in 10% buffered formalin, embedded in paraffin, sectioned, and stained with hematoxylin and eosin using standard methods.

Ethics statement
All animal procedures were approved by University of Nebraska-Lincoln Institutional Animal Care and Use Committee (IACUC; protocol 1318) and were performed in accordance with the Guide for the Care and Use of Agricultural Animals in Agricultural Research and Teaching.

ORFV073 is a conserved PPV protein that is expressed late during virus infection and localizes in the nucleus
ORFV073 encodes for an arginine-rich 188-amino acid, basic protein with predicted molecular weight of 21.9 kDa. ORFV073 is highly conserved among ORFV isolates exhibiting 95%-99% amino acid identity (aa id), and less similar to orthologs in pseudocowpox virus (PCPV, 89% aa id), parapoxvirus of the Red Deer (PPV-RD, 70% aa id), and bovine papular stomatitis virus (BPSV, 63% aa id). Notably, PCPV contains two PPV073 paralogs arranged back to back in the genome (PCPV073 and PCPV073.5; 45% aa id), which are likely the result of gene duplication followed by divergent evolution [43]. A divergent ORFV073 homolog (SQPV0840, 36% aa id) is found in squirrelpox virus, a member of a novel chordopoxvirus genus closely related to PPV. Interestingly, mouse betaherpesvirus 1 (i.e. murid cytomegalovirus, a virus that circulates in wild mice) encodes a protein of unknown function (m170) similar in size to PPV073 and with a region of approximately 50 residues sharing 56% aa id to PPV073 (OV-IA82 amino acid positions 71-122) (Fig 1). While PPV073 orthologs contain a predicted nuclear localization signal (NLS) at their carboxyl-termini (OV-IA82 amino acid positions 149-182; underlined in Fig 1), no NLS was predicted for SQPV0840 and m170.
The expression kinetics of ORFV073 was assessed by Western blot. ORFV073 was increasingly detected between 10 and 24 hours post-infection (h p.i.) (Fig 2A). Consistent with this observation, ORFV073 transcription was detected only at late times during ORFV infection Results suggest that ORFV073, a late viral protein, may be a virion component.

ORFV073 is nonessential for ORFV replication in primary OFTu cells
The replication kinetics of ORFV073 gene deletion virus (OV-IA82Δ073) was compared with that of wild-type virus (OV-IA82) in primary ovine cells (OFTu). No differences in replication kinetics and viral yields were observed in multi-step or one-step growth curves between the two viruses (Fig 3A and 3B). Also, no differences in cytopathic effect, and plaque size and morphology were observed between the viruses (Fig 3C). These data indicate that ORFV073 is nonessential for ORFV replication in OFTu cells.

Infection of cells with OV-IA82Δ073 leads to increased levels of NEMO and marked phosphorylation of IKKα/β, IκBα and NF-κB-p65 early in infection
To investigate the role of ORFV073 in NF-κB inhibition, OFTu cells were infected with OV-IA82, OV-IA82Δ073 or OV-IA82RV073 Flag for 30 min or 1 h, and phosphorylation of IKKα/β, IκBα and NF-κB-p65 was assessed by Western blot. Infection by OV-IA82Δ073 led to marked and early phosphorylation of IKKα/β (Ser176/180), IκBα (Ser32/36) and NF-κB-p65 (Ser536) (Fig 5A). Densitometric analysis showed that relative fold increases of phosphorylated forms in OV-IA82Δ073-infected cells were 101.7 and 123.7 for IKKα/β (Fig 5B), 54.2 and 33.7 for IκBα (Fig 5C), and 5.5 and 2.5 for NF-κB-p65 (Fig 5D), at 30 min and 1 h p.i., respectively. To assess the effect of ORFV073 on NEMO, OFTu cells were mock infected or infected with OV-IA82 or OV-IA82Δ073 for 30 min, 45 min, 1 h, 1 h 30 min and 2 h, and expression of NEMO was assessed by Western blot. Virus infection resulted in a significant increase in NEMO levels in wild-type virus-infected cells at 30 min (2.0 fold) and OV-IA82Δ073 infected cells at 30 min (3.04 fold), and 1 h p.i. (3.39 fold) compared to mock infected cells (Fig 6A and 6B). However, NEMO levels in OV-IA82Δ073 infected cells were significantly higher at 30 min (1.53 fold), 45 min (1.41 fold) and 1 h (1.31 fold) than those observed in wild-type virus-infected cells (Fig 6A and 6C). Together, results indicate that ORFV073 prevents NF-κB activation early in infection by inhibiting activation of the IKK complex. This is likely the result of a ORFV073-dependent event that leads to reduced accumulation of NEMO in wild-type virus-infected cells compared to levels found in OV-IA82Δ073 infected cells.  ORFV073 is a virion protein and its early NF-κB inhibitory activity does not involve de novo viral protein synthesis The early inhibitory effect of ORFV073 on NF-κB signaling is at variance with it being expressed at late times p.i. This observation, together with ORFV073 staining small circular to ovoid structures in infected cells (Fig 2B and S2A Fig)   experiments) were detected in the EEV fraction. These ORFV073 specific bands were not observed in western blots of OV-IA82Δ073 virions or uninfected cell lysates (Fig 7A, top panel). Higher molecular weight forms of ORFV073 in EEV suggest possible covalent modification of virion-incorporated ORFV073 during particle maturation and morphogenesis. As a control, the virion core protein ORFV086 was detected as a predominant 21 kDa band together with previously described higher molecular weight forms in both EEV and IMV fractions [42] (Fig 7A, bottom panel).
To assess whether early inhibition of NF-κB-p65 nuclear translocation by ORFV073 involves de novo viral protein synthesis in the infected cells, OFTu cells were pre-treated with the protein synthesis inhibitor cycloheximide (CHX) for 30 min followed by infection with OV-IA82, OV-IA82Δ073 or OV-IA82RV073 Flag for 1 h in presence of the drug. Under these conditions expression of p53, a cellular protein with short half-life, was inhibited (S3 Fig). NF-κB-p65 nuclear translocation was inhibited in both OV-IA82 and OV-IA82RV073 Flag -infected cells regardless of CHX treatment (Fig 7B and S4 Fig). Together, these results indicate that ORFV073 is a virion component responsible for early inhibition of NF-κB signaling.

ORFV073 interacts with NEMO in NF-κB signaling pathway
Results above demonstrated that ORFV073 functions at or upstream of IKK complex in NF-κB signaling pathway. To examine the potential mechanism(s) underlying ORFV073 function, reciprocal co-immunoprecipitation of ORFV073 with various mediators of the TNFα-induced NF-κB signaling pathway was performed. OFTu cells were co-transfected with control plasmid or pORFV073-His together with pNEMO, pRIPK1, or pTRAF6. Cells were harvested 24 h post-transfection and nuclear extracts obtained as described in Material and Methods. Reciprocal interaction was observed between ORFV073 and NEMO following either anti-His or anti-NEMO antibody pull downs (Fig 10). Reciprocal co-immunoprecipitation of ORFV073 with RIPK1 and TRAF6 were not observed. These results show that ORFV073 interacts with NEMO, the regulatory subunit of the IKK complex. Interaction of ORFV073 with NEMO in uninfected cells and elevated levels of NEMO in cells infected with OV-IA82Δ073 early during infection suggest that ORFV073 interferes with assembly and/or activation of the IKK complex thus affecting subsequent activation of NF-κB signaling.

ORFV073 contributes to ORFV virulence in the natural host
The effect of ORFV073 in virus virulence was investigated in sheep, a natural ORFV host. Animals were inoculated with OV-IA82Δ073 (n = 4), OV-IA82RV073 Flag (n = 4) or PBS (control group, n = 3) in the right labial commissure and the inner side of the thighs, and disease course was monitored for 21 days. All virus-inoculated animals developed clinical orf (Fig 11A). However, clinical disease was less severe in sheep infected with OV-IA82Δ073 (Fig 11B). No significant differences in disease onset and time to lesion resolution between animal groups were observed. By day 5 p.i., lesions in all four OV-IA82RV073 Flag -infected sheep exhibited scabby tissue deposition and pustules at the lesion margins. In OV-IA82Δ073 -infected sheep, however, pustule development was not observed and deposition of scabby tissue was seen in only one animal at this time point (sheep #62; Fig 11A, 5 dpi). Lesions in two OV-IA82RV073 Flag -infected sheep continued to evolve by further scabby tissue deposition during the following week (sheep 21 and 124), while scabs in the other two animals were shed leaving pustules exposed. In contrast, changes in sheep inoculated with mutant virus progressed modestly and a clinical pustular stage was never observed (Fig 11A, Day 9 p.i.). Lesions started to regress by day 12 p.i., with one animal per group exhibiting scabby lesions at day 16 p.i. (sheep 62 and 124), and clinical resolution was complete in all virus-infected sheep by day 21 p.i.
Punch biopsies were collected from inoculation sites in the thighs at various times postinfection and processed for histology. By 2 dpi, skin samples from all animals showed epidermal hyperplasia, active re-epithelialization, and various degrees of inflammatory cell infiltration. All OV-IA82RV073 Flag -infected sheep showed foci of ballooned degenerated keratinocytes, a morphological indication of advanced virus replication. In contrast, none of the OV-IA82Δ073 -infected sheep exhibited ballooned degeneration by this time. (Fig 12, left panels). By day 5 p.i., with the exception of sheep 62, samples from all infected animals exhibited ballooning degeneration of keratinocytes. Congruent with the gross pathology, OV-IA82RV073 Flag -infected sheep samples showed large, often broken and hemorrhagic pustules. In contrast, lesions in OV-IA82Δ073-infected sheep contained small, intact micropustules contained by a mildly hyperkeratotic stratum corneum (Fig 12, right panels). These pustules never developed further beyond this stage. Data indicate that infection of sheep with OV-IA82Δ073, a virus lacking ORFV073, resulted in delayed infection of keratinocytes and absence of a clinical pustular stage.

Discussion
NF-κB is a key regulator of early host responses against pathogens, playing a critical role in inflammation and integrating many cellular processes including cell proliferation, differentiation, and survival [9,10]. The parapoxvirus ORFV has evolved multiple strategies to counteract activation of the NF-κB signaling pathway, with encoded NF-κB inhibitors targeting both cytoplasmic and nuclear events leading to NF-κB activation [36][37][38]. Here, we describe a ORFV virion protein, ORFV073, that inhibits activation of the IKK complex and subsequent NF-κB signaling at very early times post-infection.
Parapoxviral genes involved in host range, immune modulation/evasion and virulence largely map to the terminal genomic regions [40,44]. Somewhat surprisingly, ORFV073, is located approximately in the center of the central conserved region of the genome, between two highly conserved poxviral genes (ORFV072, which encodes for a transcription termination factor, and ORFV074, which encodes for the small subunit of the mRNA capping enzyme). ORFV073 is highly conserved among the ORFV isolates and, while less similar to orthologs in other parapoxviruses, it still exhibits a higher degree of conservation than that observed for other known parapoxviral host range and immune evasion genes [40]. ORFV073 genomic location and its high degree of conservation may suggest the overall significance of this nonessential gene for viral perpetuation and transmission under selective pressures operating in nature. Interestingly, the finding of a 50 aa region in ORFV073 with homology to a herpesvirus protein of unknown function (mouse cytomegalovirus m170) suggests that yet unmapped ORFV073 functions may extend across virus families.
Notably, ORFV073 is a virion protein that inhibits NF-κB signaling at very early times in infected cells ( 30 min. p.i.) (Figs 4C, 5A, 7A and 7B). Our experiments with virus lacking ORFV073 suggest that early infection events such as virus entry and uncoating are efficiently sensed by PRRs, leading to NF-κB activation. Recently, tumor necrosis factor receptor (TNFR)-associated factor 2 (TRAF2) was reported to be involved in VACV fast entry via plasma membrane fusion [45]. TRAF2 functions downstream of TNFR1 and TNFR2 mediating activation of both canonical and non-canonical NF-κB signaling pathways [46]. If TRAF2 is activated in some manner during virus entry, subsequent activation of intracellular signaling pathways, including the NF-κB pathway, would be the expected outcomes. In the context of wild-type ORFV infection, virion-associated ORFV073 is immediately available to interfere with any potential TRAF2-induced NF-κB activation by inhibiting IKK activation possibly by interaction with NEMO. In contrast, a virus lacking ORFV073 in the virion, such as OV-IA82Δ073 described here, would be unable to block NF-κB early activation and nuclear translocation of NF-κB-p65. While other scenarios are also possible, results here illustrate the importance of preventing NF-κB activation early in infection.
In the context of the virus-infected cell, relatively few poxviral NF-κB inhibitors with clearly defined early functions have been described. VACV K1L protein was shown to prevent degradation of IκBα between 2 and 3 h p.i. in infected cells [25]. Similarly, VACV B14 was shown to reduce phosphorylation of IκBα at 2 and 4 h p.i [47] and VACV M2L was shown to inhibit phosphorylation of extracellular signal-regulated kinase 1 and 2 (ERK1/2) at 2 h p.i. and subsequent activation of NF-κB signaling [24]. Likewise, ORFV ORFV121 and ORFV002 were shown to inhibit NF-κB-p65 phosphorylation and acetylation, respectively, at relatively early times p.i [37,38].
ORFV073 inhibits NF-κB-p65 activation by preventing activation of IKK complex (Figs 5A and 9A). ORFV073 interaction with NEMO, the regulatory subunit of the IKK complex, likely underlies this inhibition (Fig 10). The early inhibition of IKK complex in wild-type ORFVinfected cells, is coincident with the reduced levels of NEMO in wild-type virus-infected cells compared to levels observed in OV-IA82Δ073-infected cells during the first hour p.i., suggesting that altered NEMO protein stability and/or trafficking might occur in the presence of ORFV073 (Fig 6A and 6C). Other poxviral NF-κB inhibitors are reported to specifically target the IKK complex, the bottleneck for most NF-κB-activating signals [19]. ORFV ORFV024 was shown to inhibit activation of IKK complex by preventing phosphorylation of IKK kinase [36]. VACV B14 and N1L were shown to interact with IKKβ and multiple components of IKK complex, respectively inhibiting subsequent activation of IKK complex [23,47]. Similarly, MCV MC159 and MC160, were shown to interact with NEMO preventing IKKβ activation and induce degradation of IKKα, respectively [33,34].
ORFV073 is a late viral protein found predominantly in the nucleus of infected cells at 16 to 24 h p.i. (Fig 2A and 2B). Late expression of ORFV073 in the viral replicative cycle is consistent with it being a virion component and functioning early in the next round of infection; however, the predominant nuclear localization of the protein at late times p.i. suggests it may have additional functions, related or unrelated to the NF-κB signaling pathway. Other poxviral NF-κB inhibitors with nuclear functions have been described. For example, parapoxviral ORFV002 is a nuclear inhibitor of the NF-κB signaling pathway that affects NF-κB-p65-mediated transcription [37]. The myxoma virus virulence factor M150R colocalized with NF-κB-p65 in the nucleus of TNFα-stimulated cells suggesting a potential role in regulation of the NF-κB signaling pathway; however, its effect on NF-κB-mediated gene transcription has not been demonstrated [48]. A nuclear function leading to decreased NF-κB-mediated gene expression was reported for VACV, but the actual viral protein(s) and mechanism(s) responsible for the inhibition are still unknown [49]. Recently, VACV K1 protein was shown to localize in both cytoplasm and nucleus of the cell, and prevent NF-κB-p65 acetylation [50].
Interestingly, ORFV073 interacts with NEMO in the nucleus of ORFV073 transfected cells. In addition to the canonical and non-canonical NF-κB pathway, NEMO is also involved in the atypical NF-κB pathway [51]. In response to genotoxic stress, which conceivably could occur during later stages of virus infection, NEMO translocates to the nucleus where it undergoes ataxia telangiectasia mutated checkpoint kinase (ATM)-mediated ubiquitination. NEMO and ATM are then trafficked to the cytoplasm where they activate IKKβ which results in activation of the canonical NF-κB pathway [11,52]. Although no significant differences in NF-κB-p65 nuclear translocation were observed between wild type-and OV-IA82Δ073-infected cells at late times p.i. (Fig 4C and 4D), possible effects of ORFV073-NEMO interactions on NF-κB signaling in the nucleus cannot be excluded. Other, as yet uncharacterized, late nuclear functions for ORFV073 unrelated to the NF-κB signaling pathway are also possible.
The actual role of poxviral NF-κB inhibitors for aspects of infection biology in vivo remains poorly understood [18,21,36,37,53]. Here, deletion of ORFV073 from the ORFV genome resulted in attenuation of ORFV in sheep, indicating that ORFV073 contributes to ORFV virulence in the natural host. The delayed infection of keratinocytes and absence of a clinical pustular stage in sheep infected with OV-IA82Δ073 likely reflect improved ability of the host to control the infection in the absence of ORFV073. Studies with other ORFV-encoded NF-κB inhibitors have shown that single genes either had no effect on disease pathogenesis, resulting in a wildtype disease phenotype in sheep (ORFV002, ORFV024) [36,37] or, as for ORFV121, a viral protein which binds to-and prevents nuclear translocation of NF-κB-p65, led to a markedly attenuated disease phenotype [38]. Remarkably, single gene deletions of most poxviral NF-κB inhibitors resulted in only modest effects on viral pathogenesis and virulence [22,36,37]. The multiple NF-κB inhibitors encoded by a poxvirus together with the possibility of overlapping or complementing functions may explain this observation. Alternatively, specific poxviral NF-κB inhibitors may exert only subtle and perhaps transient host range effects on specific infected cells or the infected tissue microenvironment. Regardless, the impact of these subtle changes on virus fitness in nature may be difficult to fully ascertain under experimental conditions.
To our knowledge, ORFV073 is the first poxviral NF-κB inhibitor found in virions. As early infection events are likely conserved among poxviruses [54], it is reasonable to speculate that other poxviruses encode yet to be identified virion proteins which inhibit NF-κB activation very early in infection and that early inhibition of NF-κB signaling is of greater biologic significance than currently appreciated.