|Home | About | Journals | Submit | Contact Us | Français|
A multisubunit RNA polymerase (RPO) encoded by vaccinia virus (VACV), in conjunction with specific factors, transcribes early, intermediate, and late viral genes. However, an additional virus-encoded polypeptide referred to as the RPO-associated protein of 94 kDa (RAP94) is tightly bound to the RPO for the transcription of early genes. Unlike the eight RPO core subunits, RAP94 is synthesized exclusively at late times after infection. Furthermore, RAP94 is necessary for the packaging of RPO and other components needed for early transcription in assembling virus particles. The direct association of RAP94 with NPH I, a DNA-dependent ATPase required for transcription termination, and the multifunctional poly(A) polymerase small subunit/2′-O-methyltransferase/elongation factor was previously demonstrated. That RAP94 provides a structural and functional link between the core RPO and the VACV early transcription factor (VETF) has been suspected but not previously demonstrated. Using VACV recombinants that constitutively or inducibly express VETF subunits and RAP94 with affinity tags, we showed that (i) VETF associates only with RPO containing RAP94 in vivo and in vitro, (ii) the association of RAP94 with VETF requires both subunits of the latter, (iii) neither viral DNA nor other virus-encoded late proteins are required for the interaction of RAP94 with VETF and core RPO subunits, (iv) different domains of RAP94 bind VETF and core subunits of RPO, and (v) NPH I and VETF bind independently and possibly simultaneously to the N-terminal region of RAP94. Thus, RAP94 provides the bridge between the RPO and proteins needed for transcription initiation, elongation, and termination.
The ability of poxviruses to replicate in the cytoplasm depends on the presence within the infectious particle of a system for the synthesis of translatable early mRNAs (29). The protein components of this system have been isolated from vaccinia virus (VACV) and comprise a multisubunit DNA-dependent RNA polymerase (RPO), RPO-associated protein of 94 kDa (RAP94), the VACV early transcription factor (VETF), capping and methylating enzymes, poly(A) polymerase, two nucleic acid-dependent ATPases, and topoisomerase. The products of the early mRNAs include proteins needed for the replication of the viral genome and intermediate gene transcription factors, thereby allowing the next stage of gene expression. Newly synthesized viral DNA serves as a template for the successive transcription of intermediate genes encoding late transcription factors and late genes encoding early transcription factors. In this way, the orderly expression of early, intermediate, and late genes is tightly programmed.
Previous studies suggested that the components of the early transcription system are associated in a labile complex. The complex, which could initiate and terminate transcription from an exogenous DNA template containing an early promoter, was released by the disruption of purified virions and isolated by glycerol gradient sedimentation (8). After further treatments and purification, the early promoter binding protein VETF and capping enzyme (which also serves as a termination factor) could be separated from the RPO (9, 24, 34). RAP94, encoded by the H4 gene, appears to be the lynchpin for the complex since the RPO and several other enzymes involved in early transcription fail to be packaged into virions when the expression of H4 is repressed (39). The finding that VETF is packaged under the latter conditions, possibly by binding to promoter sequences, suggested a model whereby RAP94 links VETF with RPO and other enzymes. RAP94 is tightly associated with RPO and is required for early transcription by contributing to promoter specificity (2, 3, 14, 21) as well as for transcription termination (28). NPH I, a DNA-dependent ATPase (31) required for the termination of early transcripts (12, 13), associates directly with RAP94 (27, 32). The multifunctional J3 protein, with poly(A) polymerase-stimulatory (17), cap nucleoside-2′-O-methyltransferase (33), and transcription elongation (23) activities, also interacts with RAP94 (26). The interaction of VETF with other protein components of the early transcription complex remained to be investigated. Here we provide evidence that VETF interacts directly with RAP94.
BS-C-1 cells (ATCC CCL-26) were maintained in minimum essential medium with Earle's balanced salts (Quality Biological, Gaithersburg, MD) supplemented with 10% fetal bovine serum, 2 mM l-glutamine, and 100 μg penicillin-streptomycin per ml at 37°C with 5% CO2. HeLa S3 suspension cells (ATCC CCL-2.2) were maintained in minimum essential medium with spinner modification (Quality Biological) supplemented with 5% equine serum at 37°C in a 5% CO2 atmosphere. The procedures for the preparation and titration of virus stocks were described previously (15). The Western Reserve (WR) strain of VACV (ATCC VR-1354; GenBank accession number AY243312) was used as the wild-type virus and for the construction of recombinant viruses.
V5-tagged full-length RAP94 plasmid pV5RAP94 and truncated RAP94 mutants included pV5RAP94(1-100), pV5RAP94(1-150), pV5RAP94(1-195), pV5RAP94(1-256), and pV5RAP94(257-C) (base pairs are indicated in parentheses, with C indicating the C-terminal amino acid). To construct these plasmids, DNA segments were amplified by PCR using AccuPrime Pfx (Invitrogen) and specific primers. The primer for the N terminus of the proteins contained the VACV late p11 promoter and V5 epitope tag sequences. The DNA fragments were ligated into the Zero-Blunt TOPO vector (Invitrogen, Carlsbad, CA). Plasmids encoding RAP94 with a V5 tag, NPH I with a Myc tag, and D6 with a 3× Flag tag connected by the cotranslational “self-cleavage” T2A peptide (36) to A7 with a hemagglutinin (HA) tag, each regulated by the VACV synthetic intermediate G8R promoter (5), were also constructed by inserting the PCR-amplified DNA segments into the Zero-Blunt TOPO vector. All inserts were analyzed by DNA sequencing. The transfection of plasmids was carried out with Lipofectamine 2000 (Invitrogen) according to the manufacturer's recommendations.
The recombinant viruses prepared for this study were vA7-3×Flag, vD6-3×Flag, v3×Flag-RAP94, vRAP94iA7-3×Flag, vD6iA7-3×Flag, and vA7iD6-3×Flag. In these recombinant viruses, “v” represents VACV, “i” indicates an IPTG (isopropyl-β-d-thiogalactopyranoside)-inducible gene, and “3×Flag” indicates three copies of the Flag epitope (DYKDHDGDYKDHDIDYKDDDDK). DNA for vA7-3×Flag and vD6-3×Flag was assembled by overlapping PCR in the following order: (i) approximately 500 bp of DNA upstream of the stop codon of the A7 or D6 gene, (ii) 69 bp DNA encoding 3× Flag followed by a stop codon, (iii) the enhanced green fluorescent protein open reading frame (ORF) controlled by the viral late p11 promoter, and (iv) approximately 500 bp of DNA downstream of the A7 or D6 ORF.
The DNA for v3×Flag-RAP94 was assembled by overlapping PCR in the following arrangement: (i) approximately 500 bp of DNA upstream of the H4 ORF, (ii) the green fluorescent protein ORF controlled by the p11 promoter, (iii) H4 promoter DNA, and (iv) DNA of the first methionine of the H4 ORF, followed by the DNA sequence for the 3× Flag epitope and the approximately 500-bp DNA sequence of the H4 ORF.
vRAP94iA7-3×Flag was constructed as described above for vA7-3×Flag except that vRAP94i, which contains an IPTG-inducible H4 gene (39), was used as the parental virus instead of VACV WR.
vD6i and vA7i express inducible D6 and A7 genes, respectively, and were derived from vT7LacOi, a recombinant VACV with an Escherichia coli lac repressor gene and an IPTG-inducible T7 RPO gene (37). The inserted DNA was assembled by overlapping PCR and contained (i) approximately 500 bp of DNA upstream of the D6 or A7 start codon, (ii) the ORF of red fluorescent protein controlled by the p11 promoter, (iii) a lac operator-regulated T7 promoter followed by a consensus sequence for the initiation of translation (CGAAATTAATACGACTCACTATAGGGAATTGTGAGCGCTCACAATTCCCGCCGCCACCATG), and (iv) approximately 500 bp of DNA downstream of the D6 or A7 gene start codon. vD6i and vA7i were further modified by adding a 3× Flag tag to the C terminus of A7 of vD6i and to the C terminus of D6 for vA7i as described above for vA7-3×Flag and vD6-3×Flag.
Homologous recombination was achieved by infecting BS-C-1 cells in 24-well plates with 0.5 PFU per cell of the parental virus followed after 1 h by transfection of 0.3 μg of a PCR product. The cells were harvested 48 h later and lysed by three freeze-thaw cycles. The suspension was diluted and plated onto BS-C-1 monolayers. Recombinant viruses exhibiting green or red fluorescence were clonally purified by three or four rounds of plaque isolation (16). The medium contained IPTG for producing and propagating inducible viruses.
Rabbit polyclonal antisera for D6, A7, RAP94, and RPO30 were described previously (1, 3, 18). Rabbit polyclonal antiserum for NPH I was obtained from Edward Niles (SUNY, Buffalo, NY). Anti-Flag M2 monoclonal antibody (MAb) was purchased from Stratagene (La Jolla, CA), and MAb to the V5 tag was purchased from Invitrogen.
Cells were harvested and lysed in ice-cold immunoaffinity purification (IP) buffer (50 mM Tris-HCl [pH 7.5], 1 mM EDTA, 150 mM NaCl, 1% Triton X-100, 1× protease inhibitor cocktail [Pierce, Rockford, IL]) for 1 h. After centrifugation at 16,000 × g for 10 min, the cell lysates were incubated overnight at 4°C with 2 to 3 μg of specific antibodies and protein G beads (Amersham, Piscataway, NJ). The beads were washed four times with IP buffer, and the bound proteins were eluted by heating in sample buffer, analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and transferred onto a polyvinylidene difluoride or nylon membrane with an iBlot apparatus (Invitrogen). The membrane was blocked with 5% skim milk in TBS-Tween (TBST) (50 mM Tris-HCl [pH 7.5], 200 mM NaCl, 0.05% Tween 20) at room temperature for 1 h, followed by incubation with the primary antibody in the same TBST-milk buffer for 1 to 2 h at room temperature or overnight at 4°C. The membrane was washed with TBST three times for 10 to 15 min each time, incubated with horseradish peroxidase-conjugated secondary antibody at room temperature for 1 h, washed three times with TBST, and developed with Pico or Femto chemiluminescent substrate (Pierce). Antibodies were stripped from the membrane by Restore (Pierce) for analysis using another antibody.
BS-C-1 cells (108 cells) in minimum essential medium with Earle's balanced salts with 2.5% fetal bovine serum and 100 μg/ml of rifampin (Sigma, St. Louis, MO) were infected at a multiplicity of 5 PFU of virus per cell. The cells were harvested 24 h postinfection, disrupted with ice-cold lysis buffer (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 1× protease inhibitor [Pierce]), and rotated at 4°C for 30 min. The lysate was then treated with 150 U/ml of Benzonase (Sigma) for 1 h at room temperature and centrifuged at 4°C at 20,000 × g for 15 min. The supernatant was collected and incubated with 150 μl anti-Flag MAb agarose beads (Sigma), which had been prewashed twice with lysis buffer. The lysate and beads were rotated overnight at 4°C followed by washing four times with lysis buffer. The bound proteins were eluted with 1× sample buffer or 150 μg/ml 3× Flag peptide and resolved by SDS-PAGE. The gel was stained with Coomassie blue or silver reagent. The bands of interest were excised from the gel, digested with trypsin, and analyzed by mass spectrometry.
BS-C-1 cells infected with vD6-3×Flag or VACV WR (control) were lysed with ice-cold buffer (50 mM Tris-HCl [pH 7.5], 500 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.2% sodium deoxycholate [DOC], 1× protease inhibitor [Pierce]) and rotated at 4°C for 1 h. The lysate was centrifuged at 4°C at 20,000 × g for 15 min. The supernatant was collected and incubated with anti-Flag MAb agarose beads overnight at 4°C followed by washing three times with buffer (50 mM Tris-HCl [pH 7.5], 1 M NaCl, 1 mM EDTA, 1% Triton X-100, 0.2% DOC, 1× protease inhibitor). The bound proteins were eluted and tested for purity by SDS-PAGE, Coomassie blue staining, and Western blotting or left on the beads for in vitro binding assays. RPO with or without RAP94 was purified from BS-C-1 cells infected with vRPO-10×His in the absence or presence of 40 μg/ml cytosine arabinoside (AraC) using Ni-nitrilotriacetic acid agarose gel essentially as described previously (22), followed by dialysis against buffer containing 20 mM Tris-HCl (pH 7.5), 100 mM NaCl, 1 mM EDTA, and 0.2% Triton X-100.
Assays of in vitro binding of VETF and RPO were carried out in buffer (50 mM Tris-HCl [pH 7.5], 250 mM NaCl, 1 mM EDTA, 1% Triton X-100, 1% bovine serum albumin, and 1× protease inhibitor cocktail) overnight at 4°C; the beads were washed three times with the same buffer. Bound materials were eluted by heating in sample buffer and analyzed by SDS-PAGE followed by Western blotting with specific antibodies.
VETF contains two subunits, D6 and A7, with masses of 74 and 82 kDa, respectively. In order to study the association of VETF with RPO, a 3× Flag tag was added to the C terminus of the ORFs encoding D6 and A7, in separate recombinant viruses, without altering the promoter or location within the genome. The recombinant viruses, called vD6-3×Flag and vA7-3×Flag, were replication competent, although they formed slightly smaller plaques than the wild-type VACV parent (Fig. (Fig.1A).1A). In each case, the epitope tag did not prevent an association with the untagged subunit, as shown by IP (Fig. (Fig.1B).1B). Because of the additional mass, the epitope-tagged protein migrated slightly slower on SDS-PAGE gels than the untagged protein.
Proteins associated with VETF were isolated by IP and identified by mass spectrometry. BS-C-1 cells were infected with wild-type VACV as a control or with vD6-3×Flag or vA7-3×Flag in the presence of 100 μg of rifampin to prevent the assembly of progeny virions. The infected cells were lysed with buffer containing Triton X-100, and the extracts were treated with the pannuclease Benzonase to digest DNA and RNA. After centrifugation, the supernatants were incubated with anti-Flag MAb agarose beads, which were extensively washed to remove weakly bound proteins. The remaining bound proteins were eluted, resolved by SDS-PAGE, and stained with Coomassie blue. Bands observed in the D6-3×Flag and A7-3×Flag samples but absent from the control (Fig. (Fig.2)2) were excised and digested with trypsin for mass spectrometry analysis. As expected, A7 was associated with D6-3×Flag, and D6 was associated with A7-3×Flag. The large subunits RPO147 and RPO132 of RPO were clearly visible, whereas bands corresponding to smaller subunits were not, presumably due to their smaller mass and weaker staining. RAP94 was detected as a faint band that comigrated with untagged A7; a faint band was found to contain the large subunit of capping enzyme, and NPH I was detected in a band that also contained D6. The specificity of the protein associations was demonstrated by mass spectrometry of the corresponding positions of the gels containing proteins from wild-type VACV-infected cells.
Western blotting was performed to confirm the association of D6 and A7 with RAP94 and RPO, using RPO30 as a representative subunit for which we had a specific antibody. Cells were infected with vD6-3×Flag, vA7-3×Flag, or VACV WR as a control. Analysis of the “input” prior to binding to antibody-coupled beads indicated that each of the proteins was expressed (Fig. (Fig.3A).3A). The Flag MAb detected the appropriate-sized D6 as well as a higher-molecular-weight form that coincidentally migrated near the position of A7-3×Flag (Fig. (Fig.3A).3A). Both forms of D6-3×Flag and A7-3×Flag were captured by the MAb beads (Fig. (Fig.3A).3A). Surprisingly, treatment with N-glycosidase F indicated that the upper band is glycosylated; however, as glycosylated D6 was not associated with A7 or packaged in virions, there is no evidence that this modification has biological significance (Z. Yang, unpublished data). Importantly, RAP94 and RPO30 were associated with the VETF subunits (Fig. (Fig.3A).3A). Only a trace of RAP94 and no RPO30 was bound to control beads that had been incubated with the lysate of cells infected with VACV WR (Fig. (Fig.3A).3A). The association of RAP94 and RPO with VETF was quite stable since almost all of the RAP94 and RPO30 remained bound to A7-3×Flag when the beads were extensively washed with buffer containing concentrations of NaCl up to 0.5 M and 0.4% deoxycholate (Fig. (Fig.3B3B).
The construction of another recombinant VACV with a 3× Flag tag on the N terminus of RAP94 allowed us to perform a reciprocal IP and mass spectrometry analysis to identify proteins associated with RAP94. As shown in Table Table1,1, peptides corresponding to the A7 subunit of VETF were detected, as well as RPO147, RPO132, RPO30, and RPO19 subunits and the small subunit of capping enzyme (a gel segment corresponding to the size of the large capping enzyme subunit was not submitted for analysis in this experiment), whereas there were no or few peptides identified in the parallel-treated control VACV WR sample. The association of the D6 subunit of VETF with tagged RAP94 was demonstrated by Western blotting (Fig. (Fig.3C).3C). Taken together, the SDS-PAGE, mass spectrometry, and Western blotting results indicated that VETF was associated with RPO containing RAP94.
Next, we wanted to determine whether VETF was specifically associated with RAP94-containing RPO. For this purpose, we modified a conditional lethal VACV called vRAP94i, in which the expression of RAP94 is regulated by the addition of IPTG (39). This mutant was previously shown to repress RAP94 synthesis by approximately 95% in the absence of inducer. The modification consisted of adding a 3× Flag tag to the C terminus of the A7L ORF (Fig. (Fig.4A).4A). BS-C-1 cells were infected with vRAP94iA7-3×Flag with or without IPTG, and the 3× Flag-tagged A7 was captured with anti-Flag MAb agarose beads. Western blotting indicated that similar amounts of A7-3×Flag were isolated by IP in the presence or absence of IPTG but that only a small amount of RAP94 was made and recovered under the latter condition (Fig. (Fig.4B).4B). Although the levels of RPO detected by Western blotting with antibody to the RPO30 subunit were similar in the presence and absence of IPTG, very little copurified with A7-3×Flag under the latter condition (Fig. (Fig.4B),4B), consistent with the small amount of bound RAP94. This result indicated that RAP94 was required for the association of VETF and RPO in infected cells.
We also investigated the in vitro interactions of VETF with RAP94 and RPO. Because of the difficulty in expressing full-length RAP94 by recombinant methods (2, 14), our strategy was to prepare RPO with and without RAP94 and then examine their interactions with purified VETF. This plan had the added advantage that RAP94 would be in its native state. VETF containing a Flag-tagged subunit was synthesized in cells infected with vD6-3×Flag. We succeeded in isolating VETF uncomplexed to RAP94 and RPO by lysing the cells and binding the proteins to anti-Flag MAb attached to beads in buffer containing 0.5 M NaCl and 0.2% DOC and then washing the beads with buffer containing 1 M NaCl and 0.2% DOC. A Coomassie blue-stained gel of purified VETF is shown in Fig. Fig.5A,5A, and the absence of RAP94 and RPO30 was verified by Western blotting (Fig. (Fig.5B,5B, lane 1). In parallel, cells infected with VACV WR were used for the mock purification of VETF with anti-Flag MAb beads, and neither VETF, RAP94, nor RP30 was detected (Fig. (Fig.5B,5B, lane 2). RPO was purified with Ni-nitrilotriacetic acid agarose from cells infected with vRPO22-10×His (22) in the presence or absence of AraC, which prevents viral DNA replication and late gene expression. Since RAP94 is a late protein (3), RPO made in the presence of AraC contained RPO30 but was nearly devoid of RAP94, as documented in Fig. Fig.5B5B (lane 3). In contrast, RPO made in the absence of AraC contained RPO30 and RAP94 (Fig. (Fig.5B,5B, lane 6). The VETF-containing beads and mock control beads were incubated with RAP94+ RPO and RAP94− RPO. After extensive washing, the bound proteins were eluted, resolved by SDS-PAGE, and analyzed by Western blotting with specific antibodies. As shown in Fig. Fig.5B5B (lanes 4 and 5), neither the VETF beads nor the control beads pulled down RPO30 when incubated with RAP94− RPO. However, both RAP94 and RPO30 were captured by VETF beads when incubated with RAP94+ RPO (Fig. (Fig.5B,5B, lane 7). The control beads pulled down a small amount of RAP94 and a trace amount of RPO30, presumably due to nonspecific binding (Fig. (Fig.5B,5B, lane 8). Therefore, the in vitro binding experiment supported the requirement of RAP94 for the association of VETF with RPO.
The next experiments were designed to ascertain which subunit of VETF associates with RAP94+ RPO. Previous studies had shown that each subunit could be stably expressed in the absence of the other but that both subunits were required for virus reproduction (19, 20). Two new recombinant viruses were constructed for the present study: vD6iA7-3×Flag has an IPTG-inducible D6 and a 3× Flag-tagged A7 (Fig. (Fig.6A),6A), and vA7iD6-3×Flag has an IPTG-inducible A7 and a 3× Flag-tagged D6 (Fig. (Fig.6A).6A). In each case, plaque formation occurred only in the presence of IPTG (Fig. (Fig.6B).6B). We infected cells with vD6iA7-3×Flag or vA7iD6-3×Flag with or without IPTG. Analysis of the lysates showed that RAP94 and RPO30 as well as A7-3×Flag and D6-3×Flag were made in the presence or absence of the inducer (Fig. (Fig.6C).6C). Importantly, similar amounts of epitope-tagged A7 were captured with anti-Flag MAb agarose beads regardless of whether D6 was made, but RAP94 and RPO30 copurified only in the presence of inducer, i.e., when both VETF subunits were expressed (Fig. (Fig.6C).6C). Conversely, similar amounts of epitope-tagged D6 were captured with anti-Flag MAb agarose beads regardless of whether A7 was made, but RAP94 and RPO30 copurified only in the presence of inducer, which allowed the expression of both VETF subunits (Fig. (Fig.6C).6C). These results indicated that only heterodimeric VETF can stably interact with RAP94+ RPO to form the early transcription complex.
We next sought to define the domains of RAP94 that are necessary for interactions with VETF and RPO. We constructed a series of plasmids with full-length and truncated RAP94 ORFs. In each case, RAP94 had an N-terminal V5 tag, and the VACV p11 late promoter regulated expression. Cells were infected with vRAP94iA7-3×Flag or VACV WR in the absence of IPTG and transfected 1 h later with a full-length or truncated V5RAP94 plasmid or an empty vector control plasmid. The cells were lysed 24 h later, and the 3× Flag-tagged A7 was bound to anti-Flag MAb beads. The lysate and the bound proteins were analyzed by Western blotting using anti-V5 antibody to detect RAP94 proteins. Analysis of the input showed that the truncated proteins were all expressed better than full-length RAP94 (Fig. (Fig.7A).7A). However, only full-length RAP94 and the mutants containing amino acids 1 to 256, 1 to 234, and 1 to 195 efficiently copurified with A7-3×Flag (Fig. (Fig.7A).7A). A small amount of RAP94 containing amino acids 1 to 150 copurified with A7-3×Flag, suggesting a weak interaction, whereas RAP94 mutants containing amino acids 1 to 100 and amino acid 257 to the C terminus failed to copurify with A7-3×Flag (Fig. (Fig.7A7A).
Reciprocal experiments using anti-V5 antibody to capture the RAP94 proteins and anti-Flag antibody to detect A7 by Western blotting yielded compatible results except for a stronger interaction of RAP94 containing amino acids 1 to 150 (Fig. (Fig.7B).7B). These results suggested that the binding of RAP94 to VETF occurred through the N-terminal region of the former. In contrast, we found that RPO30 was detected when full-length RAP94 or the truncated protein containing amino acids 257 to the C terminus was used but not with RAP94 containing amino acids 1 to 56 (Fig. (Fig.7C).7C). No additional mutations of the C-terminal region of RAP94 were made, since the object of this study was to determine the interaction of RAP94 with VETF. The interaction results are summarized in Fig. Fig.7D7D and indicate that RAP94 interacts with VETF and other subunits of RPO via N- and C-terminal domains, respectively. Furthermore, this suggests that RAP94 can bind VETF independently of the RPO, although this might not normally occur with full-length RAP94.
Mohamed and Niles (27) previously reported that NPH I binds to a 195-amino-acid N-terminal segment of RAP94. Since the binding sites of VETF and NPH I may overlap, we needed to determine whether the interaction of VETF with RAP94 was indirect and mediated by NPH I. The following strategy was devised to investigate this: (i) cells were infected with VACV WR in the presence of AraC to prevent the expression of late proteins including RAP94, NPH I, and VETF; (ii) plasmids encoding RAP94 with a V5 tag, NPH I with a Myc tag, or D6 with a 3× Flag tag connected by the “cotranslational cleavage” T2A peptide (36) to A7 with an HA tag, each regulated by the VACV synthetic intermediate G8 promoter to allow expression in the presence of AraC (5), were transfected into infected cells; and (iii) lysates were incubated with anti-Flag MAb beads, and the bound proteins were analyzed by Western blotting. As noted above in Results, D6 migrated as a higher-molecular-mass glycosylated species in addition to the unglycosylated form. When D6 was overexpressed by transfection, the higher-molecular-mass form was predominant and could be resolved into two bands (Fig. (Fig.8).8). As shown by IP of D6-3×Flag, VETF interacted with RAP94 in the absence of NPH I or other viral late proteins, consistent with a direct interaction (Fig, 8). In addition, when all four proteins were expressed, NPH I and RAP94 copurified with D6-3×Flag and A7-HA, suggesting that VETF and NPH I can bind simultaneously to the N-terminal domain of RAP94 (Fig. (Fig.8).8). An alternate possibility is that VETF can interact directly with NPH I as well as RAP94, although the former has not been demonstrated.
An RPO with eight VACV-encoded subunits is utilized for intermediate and late stages of gene expression (30). The same RPO, with the addition of the RAP94 polypeptide, is used for early gene expression (3, 21). Although RAP94 behaves as an RPO subunit, it differs from the other subunits in being expressed at late times after infection, consistent with its exclusive role in early transcription. The RAP94 requirement for early transcription led to a proposal that it forms a bridge between the core RPO and the early transcription factor VETF (3, 14). Further studies showed that RAP94 is present in the ternary complex (2, 14) and remains associated with RPO during elongation (14). Additional experiments indicated that a functional preinitiation complex could be formed by incubating RAP94-containing RPO with an early promoter template containing bound VETF (4). Nevertheless, evidence that RAP94 interacts directly with VETF has been lacking. In the present study, we provided support for such a role by showing that (i) VETF specifically associates with RAP94-containing RPO but not with RPO lacking RAP94 in vivo and in vitro, (ii) the association of RAP94 with VETF requires both subunits of the latter, (iii) neither viral DNA nor other virus-encoded late proteins are required for the interaction of RAP94 with VETF and core RPO subunits, (iv) different domains of RAP94 bind VETF and core subunits of RPO, and (v) NPH I and VETF bind independently and possibly simultaneously to the N-terminal region of RAP94.
A pivotal role for RAP94 in assembling a large multiprotein transcription complex was suggested previously by Zhang et al. (39). When the synthesis of RAP94 was repressed, virions failed to package the core RPO subunits, capping enzyme, DNA-dependent ATPase (NPH I), RNA helicase (NPH II), poly(A) polymerase/methyltransferase, and topoisomerase. Subsequently, evidence for the direct association of NPH I and the multifunctional poly(A) polymerase small subunit/methyltransferase/elongation factor with the N- and C-terminal regions of RAP94, respectively, was presented (26, 28). Here, we show that VETF and RPO bind to the N- and the C-terminal portions of RAP94, respectively. Capping enzyme, NPH II, and topoisomerase, which have not been shown to interact directly with RAP94, may instead interact indirectly via other subunits of RPO. Very little is known regarding the interactions of the RPO subunits with each other.
In contrast to the proteins cited above, the repression of RAP94 did not prevent the incorporation of VETF into virions (39). This was explained by the ability of VETF to bind to early promoters (6, 7, 11). However, substantial but reduced amounts of VETF as well as other enzymes involved in transcription are incorporated into defective virions even when DNA packaging is prevented, so other packaging mechanisms in addition to promoter binding must exist (10). Some VETF may be packaged nonspecifically or by interactions with another virion protein. In the present study, we showed that the interaction of RAP94 and VETF does not depend on viral DNA or other viral late proteins. Although DNA is not required for the assembly or stability of the transcription complex, further studies are needed to determine whether the complex usually forms before or after the promoter binding of VETF. A related question concerns the proportion of VETF in purified infectious virions that is bound to early promoters.
Early transcripts, in contrast to intermediate and late transcripts, are terminated by a mechanism that involves the recognition of a UUUUUNU sequence in the mRNA and is implemented by an ATP-dependent mechanism involving capping enzyme and NPH I (13, 34, 35, 38). RAP94 has a role in early gene transcription termination and release because of its direct interaction with NPH I (25, 27, 28). This interaction involves the same general region (amino acids 1 to 195) of RAP94 that is required for binding VETF. Nevertheless, VETF can bind simultaneously to both RAP94 and NPH I, indicating that the binding sites must differ and suggesting the existence of a single complex rather than separate ones. However, gel mobility shift assays demonstrated that two ternary complexes formed in vitro, both of which contained RAP94 but only one of which contained VETF, indicating that the latter dissociates before or during elongation (14).
The early transcription system of poxviruses is encapsidated within the assembling virus particle, where it is dormant until activated by the infection of a new cell. By binding multiple proteins, RAP94 provides a framework that facilitates the packaging of the system as a complex rather than as individual components. The protein associations may also improve transcription efficiency within the confined space of the virus core.
We thank Eric D. Anderson for the mass spectrometry analysis, Edward Niles for NPH I antibody, Tatiana G. Senkevich for helpful discussions, Catherine Cotter for cell culture, George Katsafanas for providing the recombinant virus vRPO22-10xHis, and other members of our laboratory for their valuable help.
The research was supported by the Division of Intramural Research, NIAID, NIH.
Published ahead of print on 16 September 2009.