Inhibition of RNA Polymerase II Phosphorylation by a Viral Interferon Antagonist*

Many viruses subvert the cellular interferon (IFN) system with so-called IFN antagonists. Bunyamwera virus (BUNV) belongs to the family Bunyaviridae and is transmitted by arthropods. We have recently identified the nonstructural protein NSs of BUNV as a virulence factor that inhibits IFN- (cid:1) gene expression in the mammalian host. Here, we demonstrate that NSs targets the RNA polymerase II (RNAP II) complex. The C-terminal domain (CTD) of RNAP II consists of 52 repeats of the consensus sequence YSPTSPS. Phosphorylation at serine 5 is required for efficient initiation of transcription, and subsequent phosphorylation at serine 2 is required for mRNA elongation and 3 (cid:1) -end processing. In BUNV-infected mammalian cells, serine 5 phosphorylation occurred normally. Furthermore, RNAP II was able to bind to the IFN- (cid:1) gene promoter as revealed by chromatin immunoprecipitation analysis, indicating that the initiation of transcription was not disturbed by NSs. However, NSs prevented CTD phosphorylation at serine 2, suggesting a block in transition from initiation to elongation. Surprisingly, no interference with CTD and 10% tryptose phosphate broth (Difco). The wt BUNV and the BUNdelNSs virus stocks used in this study were plaque-purified in BHK-21 cells, and working stocks were grown in VeroE6 cells as de- scribed (21).

During the transcription cycle, the C-terminal domain (CTD) of the large subunit of RNAP II is reversibly phosphorylated (4). In mammalian cells, the CTD is composed of 52 repeats of the consensus sequence YSPTSPS. RNAP II containing unphosphorylated CTD is recruited to the promoter (5), whereas the CTD hyperphosphorylated form is involved in active transcription (6). Phosphorylation occurs at two sites within the heptapeptide repeat, at serine 5 and serine 2. Serine 5 phosphorylation is confined to promoter regions and is necessary for the initiation of transcription, whereas serine 2 phosphorylation is important for mRNA elongation and 3Ј-end processing (7)(8)(9)(10)(11).
Viruses have evolved numerous mechanisms for evading the IFN system. Specific viral gene products have been identified that inhibit either IFN synthesis, IFN signaling, or IFN action and are therefore referred to as IFN antagonists (2,3,12).
Bunyamwera virus (BUNV) is a member of the family Bunyaviridae, representing a large group of negative strand RNA viruses that are mainly transmitted by arthropods (13). Several bunyaviruses cause encephalitis or hemorrhagic fevers in humans, e.g. the Hantaan virus, the Sin Nombre virus, the Rift Valley fever virus, the La Crosse virus, and the Crimean-Congo hemorrhagic fever virus (14). Like other arboviruses, BUNV can replicate both in mammals and insects. Depending on the host, however, the outcome of infection is different. In mammalian cells, infection is lytic and causes host cell shutoff and cell death. In insect cells, however, infection is non-cytolytic and leads to long-term viral persistence (15)(16)(17).
We have demonstrated previously that the nonstructural NSs protein of BUNV acts as an IFN antagonist (18). A mutant virus with a deleted NSs gene (BUNdelNSs virus) was generated (19) and found to be highly attenuated in immunocompetent mice. However, in mice with a defective IFN system this NSs-deficient virus was as virulent as the wild-type virus (wt BUNV). Further analyses revealed that NSs inhibited transcriptional activation of the IFN-␤ gene without interfering with the activation of IRF-3 (18,20). These findings suggested that NSs blocks a transcriptional step downstream of IRF-3. Here, we further investigated the mechanism and show that NSs inhibits a basic process in host cell gene transcription by targeting the CTD of RNAP II.

EXPERIMENTAL PROCEDURES
Cells and Viruses-Simian VeroE6 cells and human 293 cells were grown at 37°C in Dulbecco's modified Eagle's medium containing 10% fetal calf serum. Aedes albopictus C6/36 cells were maintained at 33°C in L-15 medium (Invitrogen) supplemented with 5% fetal calf serum and 10% tryptose phosphate broth (Difco). The wt BUNV and the BUNdelNSs virus stocks used in this study were plaque-purified in BHK-21 cells, and working stocks were grown in VeroE6 cells as described (21). * This work was supported by grants from the Deutsche Forschungsgemeinschaft (to F. W.) and the Wellcome Trust (to R. M. E.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Antibodies-Monoclonal antibodies H5 and H14 (Covance) specifically recognize phosphoserine 2 and 5, respectively, on the CTD (22). Rabbit polyclonal antibody N-20 (Santa Cruz Biotechnologies) recognizes the N terminus of the large subunit of RNAP II. Rabbit polyclonal serum against the N protein of BUNV has been described previously (23). Mouse monoclonal antibody M2 (Sigma) and a rabbit polyclonal antiserum (Sigma) were used to detect the FLAG epitope. Rabbit polyclonal antibody FL-425 (Santa Cruz Biotechnologies) recognizes IRF-3. Cy2-and Cy3-conjugated goat antisera (Alexis) were used as secondary antibodies, and mouse monoclonal antibody B-10 (Santa Cruz Biotechnologies), which recognizes protein kinase R, was used as an IgG control.
Reporter Assays-Subconfluent monolayers of 293 cells were transfected with 0.5 g luciferase reporter plasmid and 0.5 g expression plasmid in 200 l of OptiMEM (Invitrogen) containing 2 l of DAC-30 (Eurogentec). After 5 h at 37°C, the liposome-DNA mixture was removed, and cells were treated with a specific inducer (10 g of the synthetic double-stranded RNA poly(I⅐C) (Sigma) prepared with 10 l of DAC-30 liposomes, 50 pg/ml tumor necrosis factor ␣ (Sigma), or 1 mM N 6 -benzoyl-cAMP (Calbiochem)) or left untreated. At 24 h post-transfection, cells were harvested and lysed in 200 l of reporter lysis buffer (Promega). An aliquot of 20 l of lysate was used to measure luciferase activities as described by the manufacturer (Promega).
Western Blot and Immunofluorescence Analyses-Cells were lysed in radioimmune precipitation assay buffer (50 mM Tris, pH 7.6, 150 mM NaCl, and 1% Nonidet P-40) containing protease inhibitors (Complete protease inhibitor; Roche Applied Science) and phosphatase inhibitors (phosphatase inhibitor mixture II; Calbiochem). A total of 10 g protein was separated on a 5% polyacrylamide gel by electrophoresis and blotted onto an Immobilon-P membrane (Millipore) following the manufacturer's instructions (Bio-Rad). The membrane was incubated with a 1:500 dilution of mouse antibodies H5 or H14 or rabbit anti-BUNV N serum. Protein bands were visualized using the ECL Plus method (Amersham Biosciences).
For immunofluorescence analysis, cells were grown on coverslips to 30 -50% confluency and transfected or infected as indicated. Cells were fixed with 3% paraformaldehyde and permeabilized with 0.5% Triton X-100 in phosphate-buffered saline (PBS). After washing three times with PBS, cells were incubated with the primary antibodies polyclonal rabbit anti-BUNV N and mouse H5 or H14 diluted 1:1000 and 1:200 in PBS, respectively. After incubation at room temperature for 1 h, the coverslips were washed three times in PBS and then treated with the secondary antibodies at a dilution of 1:200 each. Cells were again washed three times in PBS and mounted using Fluorsave solution (Calbiochem). Stained cell samples were examined using a Leica confocal laser-scanning microscope with a 63ϫ NA1.4 objective.
Chromatin Immunoprecipitation Assays-Chromatin immunoprecipitation (ChIP) assays were performed essentially as described by the protocol accompanying the ChIP assay kit (Upstate Biotechnology). Briefly, 10 7 human 293 cells, previously infected for 8 h, were incubated for 10 min at 37°C in culture medium containing 1% formaldehyde. Cells were then washed with PBS and collected by scraping with 2 ml of SDS lysis buffer (50 mM Tris/HCl, pH 8.1, 10 mM EDTA, and 1% SDS) supplemented with phosphatase inhibitors and protease inhibitors. Cell lysates were then sonicated with a Branson sonifier 250 (10 ϫ 10 pulses with power set to 10 and duty cycle set to 90) to yield chromatin fragments of ϳ500 bp. The sonicated extracts were diluted 1:10 in dilution buffer (16.7 mM Tris-HCl, 0.167 M NaCl, 1.2 mM EDTA, 0.01% SDS, and 1.1% Triton X-100) supplemented with phosphatase and protease inhibitors and precleared for 1 h at 4°with 80 l of protein A-Sepharose beads saturated with salmon sperm DNA. Immunoprecipitation was then carried out with the appropriate antibodies overnight at 4°C. Immune complexes were isolated by incubation with 60 l of protein A-Sepharose/salmon sperm DNA for 1 h at 4°C. The beads were then washed successively in low salt immune complex buffer (20 mM Tris-HCl, pH 8, 2 mM EDTA, 150 mM NaCl, 0.1% SDS, and 1% Triton X-100), high salt immune complex buffer (20 mM Tris-HCl, pH 8, 2 mM EDTA, 500 mM NaCl, 0,1% SDS, and 1% Triton X-100), LiCl buffer (10 mM Tris-HCl, pH 8, 1 mM EDTA, 250 mM LiCl, 1% Nonidet P-40, and 1% deoxycholate), and TE buffer (10 mM Tris-HCl, pH 8, and 1 mM EDTA). Proteins were eluted by incubating the beads two times in 250 l of elution buffer (1% SDS and 100 mM NaHCO 3 ), and cross-links were reversed by adding 20 l of 5 M NaCl to the combined eluates (500 l) and incubating at 65°C for 4 h. To remove the proteins, 10 l of protease K buffer was added (0.5 M EDTA, 20 l of 1 M Tris-HCl, pH 6.5, and 2 l of 10 mg/ml protease K) followed by incubation for 1 h at 45°C. DNA was extracted with phenol/chloroform, ethanol-precipitated using 5 g of glycogen (Roche Applied Science) as carrier, and dissolved in 20 l of H 2 O. Input and precipitated DNAs were analyzed by PCR using primers specific for the human IFN-␤ promoter from nucleotides Ϫ138 to Ϫ19 (26).

NSs Represses Reporter Gene Expression in a Promoter-independent Manner-
We investigated the effect of NSs on RNAP II transcription using reporter assays. Cells in parallel dishes were transfected with an NSs-expressing cDNA construct together with various reporter plasmids containing the FF-Luc or Renilla luciferase genes under control of different inducible or constitutively active promoters. To activate the inducible promoters, cells were treated with either double-stranded RNA (IFN-␤ promoter), TNF-␣ (NF-B-inducible promoter), or cyclic AMP (AP-1-inducible promoter). After overnight incubation, luciferase activity was determined as a measure of promoter stimulation. As expected, NSs inhibited induction of the IFN-␤ promoter, whereas a control construct (CTRL) encoding a polypeptide of a similar size as that of NSs had no effect (Fig.  1A, columns 1 and 2). Interestingly, NF-B-and AP-1-inducible promoter activity was similarly blocked by NSs. (Fig. 1A, columns [3][4][5][6]. Surprisingly, the constitutively active thymidine kinase promoter and the simian virus 40 immediate early promoter were also inhibited ( Fig. 1B), indicating that NSs had a general effect on cellular transcription. It should be noted that, in these experiments, the expression of NSs was under the control of the CMV immediate early promoter. This resulted in almost undetectable levels of NSs (Fig. 1C, right panel), presumably due to autocrine self-inhibition. In contrast, the CTRL protein was expressed from the same vector to high amounts, as expected ( Fig 1C, left panel). Thus, even small amounts of NSs seem to be able to repress transcription from cellular or viral RNAP II promoters.
NSs Targets the C-terminal Domain of the Mammalian RNA Polymerase II-CTD phosphorylation plays a central role in the regulation of RNAP II (4,11). We therefore investigated whether NSs would affect the phosphorylation of the CTD heptapeptide at serine 5 or serine 2. Cells were infected with either the wt BUNV or the NSs-deficient BUNdelNSs virus or were left uninfected. At 6 h after infection, cells were lysed and subjected to Western blot analysis using phosphorylation statespecific antibodies. Fig. 2A (panel 1) shows that infection with either wt BUNV or BUNdelNSs did not change CTD phosphorylation on serine 5. By contrast, phosphorylation on serine 2 was strongly reduced in wt BUNV-infected cells but not in BUNdelNSs-infected cells ( Fig. 2A, panel 2). We also determined the overall amount of RNAP II in infected cells by using an antiserum that recognizes the N terminus of the RNAP II large subunit. As demonstrated in Fig. 2A (panel 3), infection with wt BUNV shifted the distribution of the RNAP II CTD from the hyper-to the hypophosphorylated state ( Fig. 2A, IIo and IIa, respectively), in agreement with the observed reduction in CTD phosphorylation. The overall levels of RNAP II, however, were comparable between cells infected with either virus. Immunostaining for the viral N protein showed that the cell cultures were infected with similar efficiencies, irrespective of whether the wild-type or the mutant virus was used ( Fig. 2A,  panel 4). In summary, these data demonstrate that infection with wt BUNV reduces CTD serine 2 phosphorylation without affecting CTD serine 5 phosphorylation or the overall levels of RNAP II. To examine this finding at the single-cell level, we performed double immunofluorescence analyses using the phosphorylation state-specific antibodies. Again, no major changes were detected for CTD phosphoserine 5 (Fig. 2B, upper  panel), whereas the phosphoserine 2-specific signal disappeared in wild-type BUNV-infected cells but not in cells infected with the delNSs mutant (Fig. 2B, lower panel).
The only difference between wt BUNV and the BUNdelNSs virus is their capacity to express NSs. We therefore argued that NSs was most likely responsible for the observed effect on CTD phosphorylation. To verify this hypothesis, we expressed recombinant NSs from appropriate plasmids. To bypass the suppressive effect of NSs on RNAP II-driven promoters (see Fig.  1C), we used a T7 system for NSs expression. Cells were transfected with a FLAG-NSs construct containing a T7 promoter, infected with an MVA-T7 helper virus, and fixed and stained at 6 h post-infection. Fig. 3A shows that T7-mediated expression of NSs indeed resulted in much higher protein levels than did CMV-driven expression, which is in agreement with our hypothesis that NSs inhibits the RNAP II. Furthermore, in cells expressing FLAG-NSs the phosphorylation of both serine 5 (Fig. 3, panels 1-3) and serine 2 (Fig. 3, panels 4 -6) was barely detectable, whereas the surrounding untransfected cells were not affected. By contrast, parallel immunostainings with an antiserum that recognizes the N terminus of the RNAP II large subunit did not detect any major changes in RNAP II levels (Fig. 3, panels 7-9), indicating that NSs does not simply induce RNAP II degradation (compare Fig. 2A).
Taken together, these data suggest that NSs represses RNAP II (but not T7 polymerase) by down-regulating the CTD phosphorylation. The main target appears to be phosphoserine 2. However, phosphoserine 5 was also affected in cells overexpressing recombinant NSs. In line with this, wt BUNV but not BUNdelNSs inhibited both serine 5 and serine 2 phosphorylation late in infection (data not shown).
Subcellular Localization of NSs-In the T7 RNA polymerase-mediated high expression system, NSs was found to accumulate mainly in the nucleus (Fig. 3), whereas in the CMV promoter-mediated low expression system, NSs was detected in the cytoplasm (Fig. 1 C). To investigate a possible relationship between NSs expression levels and intracellular localization, we monitored the subcellular distribution of NSs at different time points after the onset of expression by using the T7 system. Cells were transfected with the T7-FLAG-NSs construct, infected with MVA-T7 helper virus, and fixed and stained at 2  (Fig. 4, left panel) similar to the pattern observed with CMV promoter-driven expression (see Fig. 1C). In addition, minor amounts of NSs were localized in the nucleus, as indicated by a faint nuclear staining. At 6 h after the onset of expression, NSs was mainly nuclear, and the cytoplasm showed a faint and more uniform staining (Fig. 4, right panel). This staining pattern was also observed at later time points (data not shown). Because NSs in the CMV promoter-driven expression system was cytoplasmic even 24 h after transfection, we conclude that the subcellular localization of NSs depends on the expression levels, i.e. low amounts of NSs result in a mainly cytoplasmic, punctate pattern, whereas higher amounts localize to the nucleus.
CTD Phosphorylation in Insect Cells-Most bunyaviruses have the ability to replicate both in mammals and in arthropods (13). Infection of mammalian cells is lytic and results in a shutoff of host cell protein synthesis, whereas insect cells are persistently infected without deleterious effects on cell function (15)(16)(17). This difference in the outcome of infection is not easily explained, especially because NSs is expressed in BUNV-infected mosquito cells (17). Therefore, we investigated whether  1-3) and CTD phosphoserine 2 (panels 4 -6) were detected with specific mouse antibodies. BUNV N protein (panels 1-6) was stained with a rabbit antiserum. Cy3-labeled goat anti-mouse and Cy2-labeled goat anti-rabbit were used as secondary antibodies. Infection was at a multiplicity of 5.

FIG. 3. CTD phosphorylation and RNAP II stability in mammalian cells expressing BUNV NSs.
VeroE6 cells were transfected with expression plasmid for FLAG-tagged BUNV NSs. Expression was under control of the T7 promoter and an internal ribosome entry site. At 5 h post-transfection, cells were infected with the T7-expressing helper virus MVA-T7 and fixed and stained at 6 h post-infection. For detection of RNAP II, mouse antibodies recognizing CTD phosphoserine 5 (panels 1-3), CTD phosphoserine 2 (panels 4 -6), or rabbit antiserum recognizing the N terminus of the large subunit of RNAP II (panels 7-9) were used. The FLAG epitope was detected with a rabbit antiserum (panels 1-6) or the mouse monoclonal antibody M2 (panels 7-9). Secondary antibodies were Cy3-labeled goat anti-mouse and Cy2-labeled goat antirabbit (panels 1-6) or the converse combination (panels 7-9). NSs also affects CTD phosphorylation in insect cells. C6/36 mosquito cells were infected for 24 h, and CTD serine phosphorylation was investigated by immunofluorescence analyses. No apparent difference in serine 5 phosphorylation was observed between wt BUNV-infected and uninfected cells (Fig. 5, panels  1 and 2). Similarly, CTD serine 2 phosphorylation was unaffected (Fig. 5, panels 4 and 5). Infection with the BUNdelNSs virus also did not impair CTD phosphorylation, as expected (Fig. 5, panels 3 and 6). These results imply that NSs does not target the CTD of RNAP II in insect cells, allowing cell survival and persistent infection.
Promoter Binding by RNAP II-Current models of RNAP II transcription postulate that CTD phosphorylation at serine 5 is confined to promoter regions and is necessary for initiating mRNA synthesis (8,9,11). Subsequent phosphorylation at serine 2 is required for mRNA elongation and 3Ј-end processing. We determined the promoter binding capacity of RNAP II in infected cells using ChIP assays. We focused on the IFN-␤ promoter because the biological function of NSs is to block the antiviral IFN-␣/␤ system (18). Cells were infected with wt BUNV or the BUNdelNSs virus or were left uninfected. At 8 h after infection, protein-DNA complexes were cross-linked with formaldehyde, cells were lysed, and the lysates were sonicated to obtain short DNA fragments. Phospho-CTD-specific antibodies were used to immunoprecipitate RNAP II-DNA complexes. After removing the proteins, the DNA in the immunoprecipitates was analyzed by PCR amplification for the presence of IFN-␤ promoter sequences. A positive PCR signal would indicate that the phosphorylated RNAP II had bound to the IFN-␤ promoter. Fig. 6A (panel 1) shows that usage of the phospho-CTD serine 5-specific antibody resulted in a PCR product in all cases, irrespective of whether the cells were left uninfected or were infected with wt BUNV or the BUNdelNSs virus. By contrast, no signal was obtained in control experiments with an irrelevant antibody (Fig. 6A, panel 3). The input control showed that equal amounts of DNA were used (Fig. 6A, panel 4). These data indicate that the binding of serine 5-phosphorylated RNAP II to the IFN-␤ promoter occurred unimpeded in infected cells and suggest that NSs affects a step after RNAP II has bound to the promoter. We also tested whether the bound RNAP II was phosphorylated at serine 2. As expected, in wt BUNV-infected cells no PCR signal was detected. In contrast, a strong signal was detectable in cells infected with the NSsdeficient mutant, indicating serine 2-phosphorylation of RNAP II (Fig. 6A, panel 2).
Although in uninfected cells the IFN-␤ gene is inactive, we detected CTD-phosphorylated RNAP II bound to the IFN-␤ promoter region (Fig. 6A, panels 1 and 2, compare lanes m and  delNSs). Binding of fully phosphorylated RNAP II to the inactive IFN-␤ promoter was also observed in human A549 cells and in mouse embryo fibroblasts (data not shown), ruling out cell-specific effects. To elucidate the differences between the uninduced and induced state of the IFN-␤ promoter, we determined promoter binding of the IFN-specific transcription factor IRF-3, again using ChIP assays. In uninfected cells, IRF-3 did not bind to the IFN-␤ promoter (Fig. 6B, panel 1). By contrast, an IRF-3-specific signal was obtained in cells infected with either bunyavirus (Fig. 6B, panel 3). This result is in agreement with our previous finding that IRF-3 activation is not affected by NSs (20). Taken together, these data suggest that RNAP II can be fully phosphorylated and recruited to the IFN-␤ promoter even in the absence of virus infection but requires IRF-3 for efficient transcriptional activity. Moreover, NSs affects RNAP II after it has bound to the promoter. DISCUSSION In this study, we demonstrate that the IFN antagonist NSs of BUNV profoundly impairs the function of the mammalian RNAP II. NSs targets a step in the cellular transcription cycle that occurs after the RNAP II has bound to the promoter. NSs inhibits phosphorylation of CTD residue serine 2 but not of  1 and 4) or cells infected with wt BUNV (panels 2 and 5) or the BUNdelNSs virus (panel 3 and 6) were fixed 24 h after infection. The CTD phosphoserine 5 (panels 1-3) and CTD phosphoserine 2 (panels 4 -6) were detected with specific mouse antibodies, and the BUNV N protein (panels 1-6) was stained with a rabbit antiserum. Cy3-labeled goat anti-mouse and Cy2-labeled goat anti-rabbit were used as secondary antibodies. Infection was at a multiplicity of 5. CTD serine 5. As a consequence, RNAP II function is disturbed, leading to the down-regulation of host cell mRNA synthesis.
CTD phosphorylation is required for the eukaryotic RNAP II to overcome rate-limiting steps in transcription. Serine 5 phosphorylation initiates transcription of a short mRNA fragment and recruitment of a capping enzyme (9), whereas the subsequent phosphorylation at serine 2 is important for mRNA elongation and 3Ј-end processing (7,8,10,11). Serine 5 phosphorylation is mainly mediated by cyclin-dependent kinase 7, a subunit of the general transcription factor TFIIH (9). The main kinase responsible for CTD serine 2 phosphorylation is called P-TEFb, a complex consisting of cyclin-dependent kinase 9 and cyclin T (27). Recently, TLK-1 has been identified as a further serine 2-specific CTD kinase that is essential for transcription (28). After termination of transcription, dephosphorylation of both serine 5 and serine 2 is necessary to recycle the RNAP II. Three CTD phosphatases have been identified to date, namely FCP1 (29) and PP1 (30), which dephosphorylate both serine 5 and serine 2 indiscriminately, and SCP1, which preferentially dephosphorylates serine 5 (31). Given the preferred interference of NSs with phosphorylation at serine 2, it is more likely that NSs acts by deregulating one of the serine 2 kinases rather than by activating a CTD phosphatase. It is also conceivable that NSs influences some component of the RNAP II complex such that transcription does not proceed to the step at which CTD serine 2 phosphorylation would occur. Because NSs is located in the cytoplasm early after the onset of expression (see Fig. 4), it is possible that NSs sequesters a nuclear factor necessary for CTD serine 2 phosphorylation. Experiments are under way to explore the exact molecular mechanism and to identify the cellular interaction partners of NSs.
It is becoming increasingly clear that CTD phosphorylation represents an important step in the regulation of host gene expression. For example, cellular regulators like granulin (32), hnRNPU (33), PIE-1 (34), and the immunosuppressive glucocorticoid receptor (35) specifically inhibit CTD phosphorylation. This leads to a repression of RNAP II elongation. On the other hand, the transactivator protein Tat of human immunodeficiency virus has the opposite effect and stimulates phosphorylation of CTD serine 2, thus promoting RNAP II elongation (36,37). Similarly, heat shock (38) and chemical RNAP II inhibitors (39) increase CTD serine 2 phosphorylation. Our data indicate that the bunyaviruses also take advantage of the regulatory role of the CTD.
Transcription of bunyaviruses is dependent on host cell mRNAs. In a process called cap snatching, the viral L polymerase cleaves short, capped oligonucleotides from the 5Ј-end of host cell mRNAs and uses them as primers (13). It appears puzzling as to why, by inhibiting RNAP II, bunyaviruses would cut themselves off from the supply of such an important substrate. However, studies with the related La Crosse bunyavirus have shown that viral RNA synthesis occurs unhindered in the presence of the RNAP II inhibitor actinomycin D (40). Thus, RNAP II activity is dispensable for bunyaviruses, and the pool of cytoplasmic mRNAs is sufficient to support bunyavirus transcription. Orthomyxoviruses such as the influenza A virus, by contrast, cleave their capped primers from nascent cellular mRNAs. They replicate in the nucleus and are dependent on ongoing host cell transcription (41). This may explain why the IFN antagonist NS1 of the influenza A virus blocks IFN-␤ transcription at specific pre-and post-transcriptional levels (42-48) instead of targeting RNAP II itself.
The shutoff induced by bunyaviruses appears to be caused by multiple factors. In the nucleus, NSs inhibits the transcriptional activity of RNAP II by interfering with CTD phosphorylation. In the cytoplasm, degradation of cellular capped mRNAs by cap snatching further contributes to the depletion of mRNAs (40). For this reason, a residual shutoff is still detectable in cells infected with the NSs-deficient mutant BUNdelNSs (19). The inhibition of translation by NSs shown in vitro (49) possibly completes the host cell shutoff. In insect cells, bunyavirus infection is non-cytolytic persistent (50), and no apparent shutoff occurs (17). Our data suggest that NSs is not functional in insect cells (Fig. 4). This may contribute to the persistent infection and indicates that the target of NSs is specific for mammalian cells. Interestingly, the C terminus of the mammalian CTD contains 10 unique amino acids that are absent in non-vertebrates (51). Thus, basic differences in CTD regulation appear to exist between mammalian and insect cells, which are exploited by BUNV to regulate its infection cycle.
Despite having a broad effect on mammalian transcription, the biological function of NSs appears to be suppression of IFN synthesis, because NSs confers only a growth advantage for the virus in IFN-competent animals (18). A similar situation has been described for the nonstructural NSs protein of Rift Valley Fever virus, which has no sequence similarity to the BUNV NSs. The NSs of the Rift Valley Fever virus suppresses IFN gene expression by interacting with the p44 subunit of the essential transcription factor TFIIH, thereby preventing the assembly of TFIIH (52,54). 2 Furthermore, the matrix protein of the vesicular stomatitis virus is a potent host cell shutoff factor that inhibits the basal transcription factor TFIID (55), impairs the nuclear/cytoplasmic transport of RNAs and proteins (56), and inactivates translation factors (57). As is the case with NSs, the biological significance of vesicular stomatitis virus matrix protein-mediated shutoff is to suppress the IFN system (58,59). Similarly, subversion of the IFN system is assumed to be the reason for the shutoff of general transcription and translation in picornavirus-infected cells (60). Thus, viral IFN antagonists need not necessarily block the IFN signaling chain in a direct manner but may simply interfere with general gene expression. In contrast, viruses such as the influenza A virus, which depend on RNAP II activity for their productive replication, cannot afford to impair general transcription and, therefore, use alternative strategies.
An outstanding feature of IFN-␣/␤ genes is the absence of introns (1). Considering the fact that viral IFN antagonists often interfere with basic transcription (BUNV NSs, vesicular stomatitis virus matrix protein, and picornaviruses) or RNA splicing (the influenza A virus NS1), it is tempting to speculate that the short primary transcripts of the IFN genes are the result of an evolutionary race. IFN genes may have lost their introns to optimize the escape from viral inhibitors of mRNA synthesis. In line with this, it has been shown that IFN-stimulated promoters can bypass the need for the TATA-binding protein. The TATA-binding protein is normally necessary for general transcription, and picornaviruses induce TATA-binding protein degradation to impose a host cell shutoff (61). IFN-stimulated gene expression, however, is an exception because it can proceed without the TATA-binding protein (53). This finding illustrates that viruses and the host cell transcription machinery indeed are in continuous competition.