Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Virology. Author manuscript; available in PMC 2010 August 1.
Published in final edited form as:
PMCID: PMC2741392

Early viral protein synthesis is necessary for NF-κB activation in modified vaccinia Ankara (MVA)-infected 293T fibroblast cells


Modified vaccinia Ankara (MVA) is an attenuated vaccinia virus, and is a promising vaccine vector for variola and monkeypox viruses, as well as for other pathogens. The MVA determinants important for vaccine efficacy and immunogenicity are poorly defined. MVA infection of fibroblast cells activates NF-κB, a characteristic not ascribed to wild-type vaccinia viruses. Thus, NF-κB activation, and the subsequent upregulation of host immune molecules, could be one of the determinants for MVA’s immunogenicity. We report that ERK2 phosphorylation, an event preceding and required for NF-κB activation, occurred rapidly after virus infection. ERK2 and NF-κB remained inert when virus endocytosis was prevented, suggesting virus-host cell interactions were insufficient for activating NF-κB. Inhibition of viral protein synthesis decreased NF-κB activation, and elimination of intermediate and late gene expression did not alter MVA-induced NF-κB activation. Thus, early gene expression activates NF-κB.

Keywords: vaccinia, NF-kappaB, MVA, vaccine


Nuclear factor-κB (NF-κB) is a conserved eukaryotic transcription factor that is important for the immune system; it controls the expression of cytokines, MHC proteins, apoptotic factors, and cell adhesion molecules (Hoffmann, Natoli, and Ghosh, 2006). The canonical pathway triggering NF-κB activation is well-characterized (Gilmore, 2006). Prior to activation, the p65/p50 heterodimeric NF-κB complex is inactive in the cytoplasm, bound to its inhibitory protein, IκBα. Diverse events, such as virus infection, will activate the I kappa kinase complex (IKK) complex. Once IKK phosphorylates IκBα(DiDonato et al., 1997; Mercurio et al., 1997), IκBαis degraded (Baldi et al., 1996; Chen et al., 1995). The NF-κB NLS is then exposed and NF-κB relocates to the nucleus, where the transcription factor binds to a conserved DNA sequence to regulate gene expression.

The control of NF-κB activation during a natural virus infection can have great consequences (Hiscott et al., 2006). NF-κB inhibition, for example, is detrimental for a host, by virtue of viral proteins dampening components of the innate and specific immune responses. In contrast, virus-induced NF-κB activation may be beneficial to the host because anti-viral immune responses to neutralize virus infections are activated. The effect of the virus-mediated NF-κB activation is well-documented for cells infected with vaccinia virus, a virus that is highly similar to variola virus and an effective vaccine against smallpox (Moss, 2007). An attenuated vaccinia virus, such as the modified vaccinia Ankara (MVA) strain (Mayr, Hochstein-Mintzel, and Stickl, 1975), induces NF-κB activation in host cells, whereas wild-type vaccinia strains do not (Oie and Pickup, 2001; Shisler and Jin, 2004).

Wild-type vaccinia viruses utilize myriad mechanisms to inhibit NF-κB activation, including the synthesis of proteins that inhibit Toll-like-, interleukin-1β receptor- and tumor necrosis factor receptor-induced signaling (Alcami et al., 1999; Alcami and Smith, 1992; Bowie et al., 2000). However, the molecular mechanism responsible for MVA-induced NF-κB activation is unknown. The elucidation of this viral mechanism is important information when understanding how MVA regulates immune responses in vivo, and in considering the design of future MVA-based vaccines against smallpox or other infectious diseases. To this end, the stage of the MVA lifecycle responsible for activating NF-κB was identified in the 293T fibroblast cell line. Using molecular and pharmacological inhibitors, it was found that virus binding to or entry into host cells was not sufficient to trigger the activation of this cellular transcription factor, nor did Toll-like receptor signaling appear to be important for the MVA phenotype. Instead, viral protein expression was required for the NF-κB nuclear translocation in the 293T fibroblast cell line. Poxviral gene expression is temporal, and ORFs are grouped into early, intermediate or late genes, based on promoter sequences and time of expression during infection (Moss, 2007). MVA infections under conditions preventing intermediate or late gene expression did not ablate NF-κB or ERK2 activation, suggesting early gene expression is responsible for this phenotype.


NF-κB and ERK2 activation occur early during MVA infection, and EKR2 phosphorylation precedes NF-κB activation

It was shown previously that MVA infection of 293T cells results in host cell NF-κB activation (Oie and Pickup, 2001; Shisler and Jin, 2004). To determine when this event occurred during poxvirus infection, we detected NF-κB nuclear translocation, a hallmark of NF-κB activation, in virus—infected cells by using EMSA. When NF-κB is present in the nuclear extracts, and therefore active, it binds to radiolabeled oligonucleotides and retards their mobility. While there were multiple bands present in all lanes, one unique band was present in the lane containing extracts from infected cells harvested at 2-12 hours post-infection (Figure 1A). This same band increased in intensity until 4 hours post-infection, remained relatively high until 10 hours post infection, and then decreased in intensity at 12 hours post-infection (Figure 1A). It was verified that the unique mobility-shifted band indeed contained the NF-κB complex: when antibody recognizing the p65 subunit of NF-κB was present, and presumably binding to NF-κB, the unique band disappeared and a higher-mobility band appeared. The NF-κB-containing band disappeared in reactions containing of non-labeled oliogonucleotides possessing NF-κB binding sites, but remained when non-labeled oligonucleotides instead possessing the AP1 binding site were present.

Figure 1
Temporal NF-κB nuclear translocation and ERK1/2 phosphorylation during vaccinia infection

Activation of the ERK2 protein is necessary for MVA-induced NF-κB nuclear translocation in 293T cells (Gedey et al., 2006). Therefore, ERK2 phosphorylation should temporally precede MVA-induced NF-κB activation. To verify this, the presence of the active, phosphorylated (phospho-) ERK2 was detected at different times post-infection by using immunoblotting. Indeed, ERK2 activation occurred before NF-κB activation; a 42 kDa phospho-ERK2 band was detected in cells harvested as early as 30 minutes post-infection (Figure 1B). ERK2 activation, as measured by the intensity of the phospho-ERK2-containing band, peaked between 3 and 4 h post-infection, paralleling maximal NF-κB nuclear translocation in MVA-infected cells. Finally, ERK2 phosphorylation waned 6 h post-infection, and was no longer detectable 8-12 h post-infection (Figure 1B), even though NF-κB was still present in the nucleus of virus-infected cells at these times (Figure 1A). The anti-phospho-ERK2 antibody also cross-reacts with the highly similar phospho-ERK1 protein, as is evidenced by the presence of a higher molecular weight, 44 kDa band in most samples (Figure 1B). The differences in phosphorylated ERK2 levels were not due to uneven protein loading since similar amounts of ERK1 and ERK2 proteins were detected in each reaction (Figure 1B). The density of phospho-ERK2-containing bands was quantified by using densitometry. Values for phosph-ERK2—specific signals were computed by dividing the value of the signal obtained from lanes containing lysates from virus-infected cells by the value of the signal obtained from the lane containing lysates from mock-infected cells. The differences in phospho-ERK2 band intensities was represented graphically, and mimicked the trends observed in the immunoblot in Figure 1B. When comparing the phospho-ERK2 levels, the density of the phospho-ERK2 band the lane containing lysates from mock-infected cells was set at a value of “1,” and the value for other treatments (virus infection) were presented as fold-increases over the value obtained for mock-infected cells. For this particular immunoblot, the phospho-ERK2 levels in cells at 3 h post-infection is approximately 9-fold higher than that observed during mock infection.

MVA does not utilize the MyD88 accessory protein to activate NF-κB in 293T cells

MVA activation of dendritic cells (DCs) requires, in part, the Toll-like receptor 9 (TLR9), an intracellular molecule known to trigger NF-κB activation (Samuelsson et al., 2008). Since NF-κB activation is necessary for DC activation (Rowe et al., 2008), we queried if the TLR signal transduction pathway was responsible for MVA-induced NF-κB activation in 293T cells. To this end, 293T cells ectopically over-expressing dominant negative mutant MyD88 proteins were infected with MVA, and NF-κB activation was measured via a luciferase-based reporter assay, in which a luciferase gene is under the transcriptional control of an NF-κB promoter. Results are shown in Figure 2. Luciferase activity levels were increased about 7-fold in lysates from MVA- versus mock-infected cells, inferring that MVA activated NF-κB. As previously published (Gedey et al., 2006), the over-expression of a dominant negative ERK2 protein diminished luciferase activity. In contrast, luciferase activity levels in pMyD88DN-transfected cells were higher than those observed for either pERK2DN or untransfected cells, suggesting that MVA-induced NF-κB activation did not require MyD88-dependent TLR activation. To confirm that the dominant negative MyD88 protein was expressed at levels that were biologically significant, a separate set of pMyD88-transfected cells were incubated with interleukin-1β(IL-1β), a cellular cytokine that utilizes the MyD88 accessory protein to activate NF-κB. As would be expected, IL-1β-mediated luciferase activity was greatly diminished in pMyD88DN-transfected cells versus pcDNA3.1-transfected cells. Notably, MVA-induced luciferase activity remained unaffected in pMyD88DN-transfected cells, inferring that the ERK2 activation pathway remained intact in these cells.

Figure 2
MVA-induced NF-κB activation in cells expressing a dominant negative mutant MyD88 protein

Virus binding to the cell membrane and viral endocytosis is not sufficient to activate NF-κB

Since NF-κB activation and ERK2 phosphorylation occurred relatively early during virus infection, one of the following events was responsible: virus binding to the host cell, virus entry into the cell, early viral protein expression, or DNA replication. To test if virus binding to the host cell was sufficient to activate NF-κB, the virus adsorption stage of infection occurred at 4°C and in the presence of anti-L1R antibodies. Infections then proceeded for 4 h at 37°C. This treatment allows virus binding to the host cell, but prevents entry (Ichihashi, Takahashi, and Oie, 1994; Ramsey-Ewing and Moss, 1998). Under these conditions, there was a dramatic decrease in MVA-mediated NF-κB nuclear translocation as compared to MVA-infected cells incubated in medium lacking antibodies or in medium with antibodies not specific to vaccinia virus (Figure 3A). As would be expected, NF-κB remained inactive if non-specific antibodies were present during mock infection or during infection with WR, a wild-type strain of vaccinia virus that inhibits host cell NF-κB activation (Shisler and Jin, 2004). Similar to Figure 1A, the addition of anti-p65 antibodies to reactions containing extracts from MVA-infected cells super-shifted the NF-κB-containing band, verifying the presence of NF-κB in the observed band.

Figure 3
The effect of inhibiting virus entry, but not attachment, on MVA-induced NF-κB activation and ERK2 phosphorylation

The effectiveness of anti-L1 antibody treatment to prevent virus entry was measured indirectly, by detecting viral protein synthesis in lysates from infected cells (Figure 3B). When probing cytoplasmic extracts from WR- or MVA-infected cells with polyclonal antiserum raised against whole vaccinia virions, several unique bands (110 kDa, 35 kDa, 27 kDa, 15 kDa) were present, indicating that the polyclonal anti-vaccinia virus antiserum recognized multiple vaccinia virus proteins. In comparison, these bands’ intensities were greatly decreased when assaying lysates from anti-L1, MVA-infected cells (Figure 3B), inferring that the anti-L1 treatment indeed impeded virus entry into cells. Since actin levels were similar in all samples, the differences in viral protein levels were not due to unequal protein levels in each reaction.

It was reported recently that vaccinia virus penetration of the host cell occurs via a dynamin-dependent endocytosis pathway, an event inhibited by dynasore (Huang et al., 2008). As shown in Figure 3C, when virus infections occurred in the presence of dynasore, NF-κB no longer translocated to the nucleus, indicating that virus binding and endocytosis was not sufficient for virus-induced NF-κB activation. Detection of viral E3 protein, an early vaccinia protein expressed only after viral endocytosis, was utilized as a mechanism to indirectly confirm the effectiveness of dynasore treatment (Figure 3D). Indeed, the intensity of band representing the 20 kDa E3 product was dramatically decreased when infections proceeded in the presence versus the absence of dynasore.

It was expected that ERK2 would remain inactive when MVA infections occurred in the presence of dynasore. As predicted, there was a decrease in MVA-mediated ERK2 activation in the presence of dynasore, as measured by detection of phosphorylated ERK2 by immunoblotting (Figure 3D). When comparing the intensity of the phospho-ERK2 bands in this immunoblot by using densitometry, it was observed that the phospho-ERK2 levels in cells treated with 40 μM dynasore was similar (approximately 1), regardless of whether cells were infected or mock-infected. While there was a slight increase in the phospho-ERK2 levels in MVA- versus mock-infected cells in conditions in which cells were treated with 80 μM dynasore (1 versus 0.5), levels were still lower than those observed for MVA infection (approximately 2) in the absence of drug. For this immunoblot, the fold-increase in phospho-ERK2 levels in MVA-infected cells was lower than that observed for Figure 1, and this may be due to the presence of DMSO. Detection of similar amounts of ERK1 and ERK2 proteins in each sample negated the possibility that differences in phospho-ERK2 levels were due to decreased amounts of proteins in each sample.

Viral protein expression is necessary for MVA-induced NF-κB activation

Cordycepin is a nucleoside adenosine derivative that prevents the translation of poxviral mRNA (Esteban and Metz, 1973). Thus, its presence during infection would assess the role of viral proteins synthesis on MVA-induced NF-κB activation. Since NF-κB remained inactive when MVA infection occurred in the presence of cordycepin (Figure 4A), it was concluded that viral protein synthesis was necessary for NF-κB activation. At the concentration used for this experiment, cordycepin was effective at inhibiting viral protein expression: immunoblotted viral proteins detected with anti-vaccinia or anti-E3 antiserum were decreased when infections occurred in the presence of cordycepin versus its absence (Figure 4B). Actin levels in each of the cytoplasmic lysates were equivalent, inferring that the differences in observed viral protein levels were not due to differences in protein levels loaded in each sample.

Figure 4
The virus-mediated NF-κB activation state in the presence of cordycepin

To confirm the necessity of viral protein expression for MVA-induced NF-κB activation, a separate experiment, using MVA virions pre-exposed to ultraviolet (UV) light and psoralen, was performed. The use of these two treatments results in virions that still bind to and enter the host cell, yet a majority of the viral genome is not expressed or replicated (Tsung et al., 1996). In comparison to cells infected with untreated MVA, UV/psoralen-treated MVA no longer activated NF-κB (Figure 5A), confirming data in Figure 4A. When luciferase reporter assays were used as an alternative method to indirectly measure NF-κB activation, similar results were obtained. As shown in Figure 5B, MVA infection of cells induced an increase in firefly luciferase activity approximately 7-fold greater than mock-infected cells, presumably reflecting NF-κB activation. However, luciferase activity was decreased when cells were instead infected with psoralen/UV-inactivated MVA virions (Figure 5B), inferring that NF-κB was inactive under these conditions.

Figure 5
The effect of either Ara C treatment of cells or psoralen/UV treatment of virions on MVA-induced NF-κB activation and ERK2 phosphorylation

Phospho-ERK2 was greatly decreased in cells infected with the treated virions as compared to those incubated with unmodified MVA (Figure 5C), a result that was predicted. Equivalent levels of unmodified ERK2, and the highly similar ERK1, were detected in each sample, indicating that the differences observed for phospho-ERK2 levels were not due to differing protein levels present in each sample. The UV/psoralen treatment of MVA virions was effective for decreasing viral protein synthesis. As shown in Figure 5C, when expressing phospho-ERK2 levels graphically, by using densitometry to measure phospho-ERK2, it was observed that phospho-ERK2 levels were 30-fold higher in MVA-infected cells versus mock-infected cells. In contrast, phospho-ERK2 levels were only 5-fold about mock-infected cells when MVA virions were pre-treated with UV and psoralen. UV/psoralen treatment of MVA dramatically decreased the synthesis of the early E3L protein, and the L1R late protein was no longer detected under these conditions, indicating that the UV/psoralen treatment of virions was sufficient to greatly decrease early gene expression.

Intermediate or late gene expression, but not early gene expression, is dispensable for MVA-induced activation of NF-κB and ERK2

Poxvirus gene expression is temporal, with early genes expressed before viral DNA replication, and intermediate and late genes expressed after viral DNA replication (Moss, 2007). To determine if intermediate and late proteins were dispensable for MVA-induced NF-κB activation, we infected 293T cells in the presence of cytosine arabinoside (AraC), a drug that prevents viral DNA replication, and subsequent intermediate and late gene expression.

Under these conditions, the lack of intermediate and late protein expression did not affect the ability of MVA to activate host cell NF-κB at 4 h post-infection, since the nuclear translocation of NF-κB was still detected in nuclear extracts (Figure 5A). Furthermore, NF-κB remained effective as a transcriptional activator of a luciferase reporter gene in MVA-infected cells, regardless of whether AraC was present or absent during infection (Figure 5B). Not surprisingly, virus-induced ERK2 phosphorylation was unaffected by the presence of AraC during infection (Figure 5C). The effectiveness of the drug in inhibiting intermediate and late protein expression is demonstrated in Figure 5C, when immunoblotted lysates were probed for the presence of either an early product (the E3L product) or a late product (the L1R product). When comparing lysates from cells infected in AraC-containing versus complete medium, there was a decrease in the amount of L1R product present, indicating a decrease in late protein synthesis. E3L protein expression was not affected by the presence of Ara C, as would be predicted.

As confirmation that late viral protein expression was dispensable for NF-κB activation, 293T cells were instead infected with MVAΔE3L, a mutant MVA construct deleted for the early E3L ORF. As reported by Ludwig et al (Ludwig et al., 2006), infection of 293T cells with MVAΔE3L results in an abortive infection, in which candidate late proteins are not produced. Under these circumstances, NF-κB remained active in MVAΔE3L-infected cells, as detected by an EMSA (Figure 6A). Importantly, the temporal regulation of NF-κB during MVAΔE3L infection was similar to that observed during infection with MVA, implying that the extended NF-κB activation observed during MVA infection was solely due to early gene expression. Since ERK2 activation is required for NF-κB activation (Gedey et al., 2006), it was not surprising to observe that phosphorylated ERK2 was detected in extracts from MVAΔE3L-infected cells (Figure 6B). As would be expected, the temporal activation of ERK2 was similar in MVAΔE3L- and MVA-infected cells. The density of phospho-ERK2-containing bands was quantified by using densitometry. Values for phosph-ERK2—specific signals were computed by dividing the value of the signal obtained from lanes containing lysates from virus-infected cells by the value of the signal obtained from the lane containing lysates from mock-infected cells. In this experiment, phospho-ERK2 levels were highest at 1-3 h post-infection, with MVA infection inducing phospho-ERK2 levels 3-5 fold above mock-infected cells. While this activation profile was slightly different than that observed for Figure 1, we still observed ERK2 activation early during infection, and then waning at times later in infection for both experiments. Phospho-ERK2 levels in MVAΔE3L-infected cells were slightly lower than those of MVA-infected cells, inducing 2-4-fold above mock-infected cells. Regardless of virus used, phospho-ERK2 levels were similar to mock-infected cells by 4 h post-infection.

Figure 6
The NF-κB activation and ERK2 phosphorylation states in cells infected with MVA viruses either lacking or containing the E3L ORF


Modified vaccinia virus Ankara infection, unlike wild-type vaccinia strains, activates the cellular NF-κB transcription factor (Gedey et al., 2006; Oie and Pickup, 2001). The viral mechanism(s) responsible for this phenotype was heretofore unknown. Using several approaches, we identified that viral gene expression was required for this phenotype, with the expression of the early class of vaccinia ORFs correlating with MVA-induced NF-κB activation. Two possibilities for this phenotype existed. First, the concomitant expression of all early proteins was activating NF-κB via mechanisms such the endoplasmic reticulum overload response (EOR) pathway. However, since MVA infection of 293T cells did not trigger calcium release (data not shown), an event occurring during EOR activation (Kuang et al., 2005), this possibility was unfavorable. An alternative hypothesis is that the expression of an individual early viral protein activates NF-κB either directly or indirectly. Experimental approaches to test each of these theories are the subject of ongoing investigations.

Recent studies have investigated the effects of MVA infection on DC activation. Dendritic cells (DCs) are potent antigen presenting cells to activate naïve CD4+ and CD8+ T cells, making them important for the development of adaptive immune responses (Guermonprez et al., 2002). NF-κB activation is necessary for DC activation (Rowe et al., 2008). One prediction from our results is that early protein synthesis is also required for MVA-induced DC activation. This model is in agreement with a report that UV-inactivation of MVA does not enhance virus-induced DC activation (Samuelsson et al., 2008). However, DC activation with UV-treated MVA has also been documented (Drillien, Spehner, and Hanau, 2004; Waibler et al., 2007), in contrast to our findings reported here. Additionally, MVA activation of DCs utilizes the TLR9, an intracellular molecule known to trigger NF-κB activation (Samuelsson et al., 2008). However, the inhibition of TLR9-induced NF-κB activation, via over-expression of a dominant negative MyD88 protein, did not decrease MVA-induced NF-κB activation. Whether MVA possesses two different mechanisms for activating NF-κB in fibroblasts (early protein synthesis required, TLR-independent) versus immune cells (early protein synthesis not required, TLR-dependent) is the subject of future studies. Yet another possibility is that UV irradiation of MVA is not as efficient for cross-linking the MVA genome as UV/psoralen treatment of MVA (Tsung et al., 1996). Thus, UV-irradiated MVA virions might allow for the transcription of viral ORFs, an event that would not occur with UV/psoralen treated MVA virions. Finally, these differences could be due to varying TLR protein expression levels in fibroblast versus immune cells.

ERK2 and NF-κB activation occurred rapidly after virus infection, raising the possibility that proteins in the mature virion may activate NF-κB when the virus is intracellular. However, we observed two experimental conditions in which the inhibition of viral protein synthesis rendered NF-κB inactive in virus-infected cells (UV/psoralen treatment of virions and incubation of cells with cordycepin). These data support our position that nascent viral proteins are responsible for NF-κB activation.

Curiously, NF-κB remained active in MVA-infected cells for an extended period of time. Although MVA infection of human monocyte-derived DCs up regulates IκBα gene expression (Guerra et al., 2007), it is likely that virus infection in 293T cells sequesters the auto-regulatory feedback loop for NF-κB. Whether these patterns remain present in all fibroblasts versus immune cells, and whether this trait possesses physiological relevance, is unknown. Equally unlikely is a late gene product prolonging NF-κB activation: cells infected with MVAΔE3L, a deletion mutant virus the no longer expresses intermediate or late genes in 293T cells (Ludwig et al., 2005; Ludwig et al., 2006), have an NF-κB activation profile that is identical to that of MVA-infected cells. A current favored model for this prolonged activation is that a viral product actively synthesized throughout infection induces this phenotype. Yet another possibility is that an early protein that is stable throughout infection can activate NF-κB throughout infection. Data obtained using the MVAΔE3L virus for infection cannot rule out either of the above possibilites.

Poxviruses are complex viruses that produce multiple immunoevasion products, with many of these products important for virulence (Moss, 2007). Data shown here indicate that vaccinia also expresses proteins to stimulate the immune response. Why the wild-type virus maintains multiple genes that encode proteins with opposing actions is an important conundrum, given the traditional idea that viruses are evolutionarily driven to minimize the number of genes they encode. This finding may indicate that temporal regulation of NF-κB activation is more important than overall inhibition of NF-κB during vaccinia virus pathogenesis, and leads to a new appreciation for the regulation of this transcription factor during a poxvirus lifecycle. Notably other viruses, such as herpes viruses, also express proteins that either activate or inhibit NF-κB (Hiscott et al., 2006), suggesting that control of this transcription factor by complex viruses is a common strategy for survival and pathogenesis.

Materials and Methods

Cells and viruses

The human kidney fibroblast HEK293T (293T) and baby hamster kidney (BHK-21) cell lines were obtained from the American Type Culture Collection (ATCC). All cells were cultivated in Eagle’s minimal essential medium (EMEM) supplemented with 10% fetal calf serum (FCS; Hyclone), referred to as complete medium. MVA is an attenuated strain of vaccinia virus (Mayr, Hochstein-Mintzel, and Stickl, 1975). MVAΔE3L, an MVA-based virus in which the E3L ORF has been deleted (Ludwig et al., 2005), was a gift from Dr. Gerd Sutter (Paul-Ehrlich Institut). MVAΔE3L was amplified in BHK-21 cells, while MVA was amplified in primary CEF cells (Charles Rivers Laboratories). The Western Reserve (WR) strain of vaccinia virus, which was amplified in BHK-21 cells, is considered to be a wild-type strain of vaccinia virus.

Conditions for virus infections

Subconfluent 293T cell monolayers in 6-well plates were either mock-infected or infected with either WR, MVA or MVAΔE3L at 10 plaque forming units (PFU)/cell, unless otherwise indicated. During this absorption phase, cellular monolayers and viruses were incubated in EMEM containing 1% FCS. After a 1 h incubation at 37°C, virus-containing medium was removed, and infected-cell monolayers were incubated in fresh, complete medium. In experiments where the activated form of ERK2 was detected, cells were instead incubated in medium containing 1% FCS for 24 h before infection, during the absorption stage of infection, and for the duration of infection. Infections proceeded for indicated amounts of time, and cytoplasmic and nuclear proteins were extracted as described below.

To allow for virus attachment to, but prevent entry into, host cells, a previously described protocol was followed (Ramsey-Ewing and Moss, 1998). Briefly, the 1 h virus absorption (MOI = 10) occurred at 4°C, and cells were then incubated in pre-chilled medium containing anti-L1R antibodies (1:5,000 dilution; a gift from Dr. Jay Hooper, United States Army Medical Research Institute for Infectious Diseases) or antibodies not reactive to the L1 protein (murine IgG; 1:5,000 dilution; Sigma-Aldrich). After an additional 15 min incubation at 4°C, cells were incubated pre-warmed complete medium at 37°C for 4 h. As an alternative approach to prevent virus endocytosis, cells were instead incubated in 1% FCS EMEM containing dynasore (40 mM or 80 mM; Sigma Aldrich) for 30 min prior to infection. The drug remained present in medium during the adsorption phase of virus infection (MOI = 10) and for the duration of infection. Cytoplasmic and nuclear proteins were extracted as described below.

To decrease viral gene expression, MVA particles were incubated in PBS containing 10 μg psoralen/ml (Sigma Aldrich) and subsequently exposed to ultraviolet (UV) radiation for 5 min. These are conditions, according to a previous report, would greatly inhibit viral gene expression (Tsung et al., 1996). Next, treated virions were incubated with cellular monolayers for 1 h at 37°C at an MOI of 10. Virus-containing medium was removed and replaced with EMEM. An alternative approach to inhibit viral protein expression was to incubate cells with cordycepin (40 μg/ml; Sigma Aldrich) for 30 min before infection. This drug was also present during the adsorption stage of infection and for the duration of the infection. Cytoplasmic and nuclear proteins were extracted as described below.

To inhibit intermediate and late gene expression, cells were incubated in medium containing 40 μg/ml cytosine arabinoside (Ara C; Sigma Aldrich) for 30 min. The drug remained present in medium during the adsorption phase of virus infection, in which cells were infected with 10 PFU/ cell, and for the duration of infection. Cytoplasmic and nuclear proteins were extracted as described below.

Immunoblotting of cytoplasmic-extracted proteins

For all experiments, infected cellular monolayers were dislodged by scraping. Next, cells were collected by centrifugation (30 sec at 10,000 × g). Supernatants were removed, and cells were lysed, and cytoplasmic proteins were collected as described previously (Gedey et al., 2006). The protein concentration of each sample was determined by using the BCA assay (Pierce). An equivalent amount of protein (20 μg) from each sample was loaded into separate wells of SDS-12% polyacrylamide gels.

Proteins were separated electrophoretically and then transferred to a polyvinyl difluoride membrane (PVDF; Millipore). Membranes were incubated in blocking buffer (TBS containing 5% milk and 0.1% Tween-20) for at least one hour. Next, membranes were incubated at 4°C overnight in TBS-T (TBS containing 0.1% Tween-20 and 0.5% milk) containing the primary antibody below. After washing membranes in TBS-T to remove unbound primary antibodies, blots were incubated with secondary antibody consisting of either horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (1:10,000; Fisher Scientific) or HRP-conjugated goat anti-mouse IgG (1:5,000; Fisher Scientific). The antigen-antibody reactions on each immunoblot (IB) were detected by using chemilluminescence and autoradiography. For all IBs, the Pierce Supersignal West Pico substrate was utilized, according to the manufacturer’s directions. Primary antibodies included: rabbit anti-phospho-ERK1/2 antiserum (1:1000; Santa Cruz Biotechnology), which interacts with the phosphorylated form of ERK2 and the highly similar ERK1 protein; rabbit anti-ERK1/2 antiserum (1:1,000; Santa Cruz Biotechnology), which recognizes the ERK2 and highly similar ERK1 protein; rabbit anti-vaccinia antiserum (1:1,000; Accurate Antibody), which was raised against vaccinia virus strain Lister particles, and contains polyclonal antiserum recognizing vaccinia proteins that are present in a mature virions; rabbit anti-actin antiserum (1:1,000, Fisher Scientific); mouse anti-E3L antiserum (Weaver et al., 2007)(1:1,000), a gift from Dr. Stuart Isaacs (University of Pennsylvania).

For quantification of the phospho-ERK2 signals, X-ray films were digitized and analyzed by using densitometry, using the Scion Image software (Scion Corporation). The signals were quantified by using the “analyzing one-dimensional electrophoresis gels” program of Scion Image, using the “GelPlot2” macro. This method was used to compare the ratio of the band density for phospho-ERK2 only within each film image, where the film exposure time to the immunoblot was identical. A value for the phospho-ERK2-specific signal was computed by diving the value of signal obtained from lanes containing virus-infected cellular lysate by the value of signal from lanes containing lysates of mock-infected cells. In this fashion, the value obtained for mock-infected cells was equal to one, while values of signals from MVA-infected cells were expressed as a fold-increase of phospho-ERK2 over uninfected cells.

Gel electromobility shift assays (EMSAs) of nuclear-extracted proteins

The method for isolating nuclear proteins from virus-infected cells was described previously (Gedey et al., 2006). The concentration of protein in each sample was determined by using the BCA Protein Assay Kit (Pierce). Five μg of each extract was incubated with 0.35 pmol of 32P-labeled double-stranded oligonucleotides containing binding sites for the NF-κB transcription factor (Promega) in Gel Shift Assay System Binding buffer (Promega) as per manufacturer’s directions. Some reactions also included excess molar amounts of non-radiolabeled oligonucleotides that either contained or lacked NF-κB binding sites, or 1 μg monoclonal anti-p65 antiserum (Santa Cruz Biotechnology). Following incubation, reactions were resolved electrophoretically in a 6% acrylamide gel (Invitrogen) under non-denaturing conditions, using TBE buffer. Afterward, the gels were dried onto filter paper, exposed to a phosphorimager plate (Molecular Devices), and images were developed and analyzed using the ImageGauge and ImageReader programs, respectively (Fuji).

Luciferase reporter assay

A luciferase reporter assay was utilized to quantitate NF-κB activation in cells that were infected with poxviruses, as previously described (Shisler and Jin, 2004). For these assays, 50 ng pRL-null (Promega) and 450 ng pNF-κBluc (Stratagene) were transfected into subconfluent 293T cell monolayers in 12-well plates by using FuGene 6 transfection reagent (Roche) according to the manufacturer’s protocol. At 24 h post-transfection, 293T cell monolayers were either mock-infected or infected with MVA (MOI = 10). Alternatively, cells were infected with MVA virions pre-treated with psoralen and UV, as described above. To inhibit intermediate and late viral gene expression, infections instead occurred in the presence of Ara C, as described above. In a separate experiment to detect the effect of Toll-like receptor signaling on MVA-induced NF-κB activation, cells were co-transfected 50 ng pRL-null, 450 ng pNF-κBluc, and 1000 ng either pERK2DN (Gedey et al., 2006) or pMyD88DN (Nichols and Shisler, 2006). At 24 h post-transfection, cells were infected with MVA (MOI = 10) or incubated in medium containing either 10 or 20 ng IL-1β/ml medium (Sigma). Regardless of these variations, all cells were harvested 4 h post-infection by scraping, collected by centrifugation (30 sec at 14,000 × g), and resuspended in passive lysis buffer (Promega).

For all luciferase assays, the resultant lysates were assessed for luciferase activities as described previously (Shisler and Jin, 2004). All assays were performed in triplicate. For each experimental point, firefly luciferase activity was divided by sea pansy luciferase activity to correct for differences in transfection efficiencies. Resultant ratios were averaged and normalized to the appropriate control cells (mock-infected cells or cells co-transfected with pcDNA3.1, pNF-κBluc and pRL-null). The value for the control cells was taken as 1. Results were displayed as relative fold change in luciferase activity as compared to either mock-infected cells or pcDNA3.1-transfected cells. Statistical significance was determined by using the Student’s t-test.


We thank Drs. Gail Scherba and Amy MacNeill for critical review of the manuscript. We thank Drs. Jay Hooper, Stuart Isaacs and Gerd Sutter for generosity with reagents. This work was funded by NIH grant number AI055530 (JLS). SM was supported by NIH Cellular and Molecular Biology Training Grant T32 GM007283.


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.


  • Alcami A, Khanna A, Paul NL, Smith GL. Vaccinia virus strains Lister, USSR and Evans express soluble and cell-surface tumour necrosis factor receptors. J Gen Virol. 1999;80(Pt 4):949–959. [PubMed]
  • Alcami A, Smith GL. A soluble receptor for interleukin-1 beta encoded by vaccinia virus: a novel mechanism of virus modulation of the host response to infection. Cell. 1992;71(1):153–167. [PubMed]
  • Baldi L, Brown K, Franzoso G, Siebenlist U. Critical role for lysines 21 and 22 in signal-induced, ubiquitin-mediated proteolysis of I kappa B-alpha. J Biol Chem. 1996;271(1):376–9. [PubMed]
  • Bowie A, Kiss-Toth E, Symons JA, Smith GL, Dower SK, O’Neill LA. A46R and A52R from vaccinia virus are antagonists of host IL-1 and toll-like receptor signaling. Proc Natl Acad Sci U S A. 2000;97(18):10162–7. [PubMed]
  • Chen Z, Hagler J, Palombella VJ, Melandri F, Scherer D, Ballard D, Maniatis T. Signal-induced site-specific phosphorylation targets I kappa B alpha to the ubiquitin-proteasome pathway. Genes Dev. 1995;9(13):1586–97. [PubMed]
  • DiDonato JA, Hayakawa M, Rothwarf DM, Zandi E, Karin M. A cytokine-responsive IkappaB kinase that activates the transcription factor NF-kappaB. Nature. 1997;388(6642):548–54. [PubMed]
  • Drillien R, Spehner D, Hanau D. Modified vaccinia virus Ankara induces moderate activation of human dendritic cells. J Gen Virol. 2004;85(Pt 8):2167–75. [PubMed]
  • Esteban M, Metz DH. Early virus protein synthesis in vaccinia virus-infected cells. J Gen Virol. 1973;19(2):201–6. [PubMed]
  • Gedey R, Jin XL, Hinthong O, Shisler JL. Poxviral regulation of the host NF-kappaB response: the vaccinia virus M2L protein inhibits induction of NF-kappaB activation via an ERK2 pathway in virus-infected human embryonic kidney cells. J Virol. 2006;80(17):8676–85. [PMC free article] [PubMed]
  • Gilmore TD. Introduction to NF-kappaB: players, pathways, perspectives. Oncogene. 2006;25(51):6680–4. [PubMed]
  • Guermonprez P, Valladeau J, Zitvogel L, Thery C, Amigorena S. Antigen presentation and T cell stimulation by dendritic cells. Annu Rev Immunol. 2002;20:621–67. [PubMed]
  • Guerra S, Najera JL, Gonzalez JM, Lopez-Fernandez LA, Climent N, Gatell JM, Gallart T, Esteban M. Distinct gene expression profiling after infection of immature human monocyte-derived dendritic cells by the attenuated poxvirus vectors MVA and NYVAC. J Virol. 2007;81(16):8707–21. [PMC free article] [PubMed]
  • Hiscott J, Nguyen TL, Arguello M, Nakhaei P, Paz S. Manipulation of the nuclear factor-kappaB pathway and the innate immune response by viruses. Oncogene. 2006;25(51):6844–67. [PubMed]
  • Hoffmann A, Natoli G, Ghosh G. Transcriptional regulation via the NF-kappaB signaling module. Oncogene. 2006;25(51):6706–16. [PubMed]
  • Huang CY, Lu TY, Bair CH, Chang YS, Jwo JK, Chang W. A novel cellular protein, VPEF, facilitates vaccinia virus penetration into HeLa cells through fluid phase endocytosis. J Virol. 2008;82(16):7988–99. [PMC free article] [PubMed]
  • Ichihashi Y, Takahashi T, Oie M. Identification of a vaccinia virus penetration protein. Virology. 1994;202(2):834–43. [PubMed]
  • Kuang E, Wan Q, Li X, Xu H, Liu Q, Qi Y. ER Ca2+ depletion triggers apoptotic signals for endoplasmic reticulum (ER) overload response induced by overexpressed reticulon 3 (RTN3/HAP) J Cell Physiol. 2005;204(2):549–59. [PubMed]
  • Ludwig H, Mages J, Staib C, Lehmann MH, Lang R, Sutter G. Role of viral factor E3L in modified vaccinia virus ankara infection of human HeLa Cells: regulation of the virus life cycle and identification of differentially expressed host genes. J Virol. 2005;79(4):2584–96. [PMC free article] [PubMed]
  • Ludwig H, Suezer Y, Waibler Z, Kalinke U, Schnierle BS, Sutter G. Double-stranded RNA-binding protein E3 controls translation of viral intermediate RNA, marking an essential step in the life cycle of modified vaccinia virus Ankara. J Gen Virol. 2006;87(Pt 5):1145–55. [PubMed]
  • Mayr A, Hochstein-Mintzel V, Stickl H. Creation of an attenuated strain of Ankara, MVA. Infection. 1975;3:6–14.
  • Mercurio F, Zhu H, Murray BW, Shevchenko A, Bennett BL, Li J, Young DB, Barbosa M, Mann M, Manning A, Rao A. IKK-1 and IKK-2: cytokine-activated IkappaB kinases essential for NF-kappaB activation. Science. 1997;278(5339):860–6. [PubMed]
  • Moss B. Poxviridae: The Viruses and Their replication. In: Knipe D, Howley P, editors. Fields Virology. Fifth edition. Vol. 2. Lippincott Williams & Wilkins; Philadelphia: 2007. pp. 2905–2946. 2 vols.
  • Nichols DB, Shisler JL. The MC160 protein expressed by the dermatotropic poxvirus molluscum contagiosum virus prevents tumor necrosis factor alpha-induced NF-kappaB activation via inhibition of I kappa kinase complex formation. J Virol. 2006;80(2):578–86. [PMC free article] [PubMed]
  • Oie KL, Pickup DJ. Cowpox virus and other members of the orthopoxvirus genus interfere with the regulation of NF-kappaB activation. Virology. 2001;288(1):175–87. [PubMed]
  • Ramsey-Ewing A, Moss B. Apoptosis induced by a postbinding step of vaccinia virus entry into Chinese hamster ovary cells. Virology. 1998;242(1):138–49. [PubMed]
  • Rowe HM, Lopes L, Brown N, Efklidou S, Smallie T, Karrar S, Kaye PM, Collins MK. Expression of vFLIP in a lentiviral vaccine vector activates NF-{kappa}B, matures dendritic cells and increases CD8+ T cell responses. J Virol. 2008 [PMC free article] [PubMed]
  • Samuelsson C, Hausmann J, Lauterbach H, Schmidt M, Akira S, Wagner H, Chaplin P, Suter M, O’Keeffe M, Hochrein H. Survival of lethal poxvirus infection in mice depends on TLR9, and therapeutic vaccination provides protection. J Clin Invest. 2008;118(5):1776–84. [PubMed]
  • Shisler JL, Jin XL. The vaccinia virus K1L gene product inhibits host NF-kappaB activation by preventing IkappaBalpha degradation. J Virol. 2004;78(7):3553–60. [PMC free article] [PubMed]
  • Tsung K, Yim JH, Marti W, Buller RM, Norton JA. Gene expression and cytopathic effect of vaccinia virus inactivated by psoralen and long-wave UV light. J Virol. 1996;70(1):165–71. [PMC free article] [PubMed]
  • Waibler Z, Anzaghe M, Ludwig H, Akira S, Weiss S, Sutter G, Kalinke U. Modified vaccinia virus Ankara induces Toll-like receptor-independent type I interferon responses. J Virol. 2007;81(22):12102–10. [PMC free article] [PubMed]
  • Weaver JR, Shamim M, Alexander E, Davies DH, Felgner PL, Isaacs SN. The identification and characterization of a monoclonal antibody to the vaccinia virus E3 protein. Virus Res. 2007;130(12):269–74. [PMC free article] [PubMed]