|Home | About | Journals | Submit | Contact Us | Français|
Poxvirus acquires its primary envelope through a process that is distinct from those of other enveloped viruses. The molecular mechanism of this process is poorly understood, but several poxvirus proteins essential for the process have been identified in studies of vaccinia virus (VACV), the prototypical poxvirus. Previously, we identified VACV A6 as an essential factor for virion morphogenesis by studying a temperature-sensitive mutant with a lesion in A6. Here, we further studied A6 by constructing and characterizing an inducible virus (iA6) that could more stringently repress A6 expression. When A6 expression was induced by the inducer isopropyl-β-d-thiogalactoside (IPTG), iA6 replicated normally, and membrane proteins of mature virions (MVs) predominantly localized in viral factories where virions were assembled. However, when A6 expression was repressed, electron microscopy of infected cells showed the accumulation of large viroplasm inclusions containing virion core proteins but no viral membranes. Immunofluorescence and cell fractionation studies showed that the major MV membrane proteins A13, A14, D8, and H3 did not localize to viral factories but instead accumulated in the secretory compartments, including the endoplasmic reticulum. Overall, our results show that A6 is an additional VACV protein that participates in an early step of virion membrane biogenesis. Furthermore, A6 is required for MV membrane protein localization to sites of virion assembly, suggesting that MV membrane proteins or precursors of MV membranes are trafficked to sites of virion assembly through an active, virus-mediated process that requires A6.
The majority of enveloped viruses obtain their lipid envelope via budding through a cellular organelle or plasma membrane. Poxviruses, however, acquire the primary envelope of their virions through a distinct yet poorly understood process. Poxviruses are a family of large, complex viruses that replicate entirely in the cytoplasm (29). The best-characterized family member is vaccinia virus (VACV), which encodes more than 200 proteins in a 190-kb genome. The process of VACV virion assembly involves a series of intermediate steps that are discernible by electron microscopy (EM) (reviewed in reference 2). The virions are assembled in areas of the cytoplasm referred to as viral factories, which are defined by cytoplasmic DNA staining in fluorescence microscopy and by a unique electron density and exclusion of cellular organelles in EM. The viral structures that appear first in the factories are electron-dense viroplasms containing viral core proteins. Crescent-shaped membranes consisting of a single lipid bilayer stabilized with a lattice of VACV D13 protein (11, 41) then develop at the peripheries of viroplasms. The crescent membranes engulf part of the viroplasm and circularize to form the spherical immature virions (IV). The viral genome is encapsidated in IV before the IV membrane completely closes off, forming IV with an electron-dense nucleoid (IVN). Concomitant with proteolytic processing of several major virion core proteins, including A10 (17, 36), IVNs mature into the brick-shaped intracellular mature virions (MV), which are the majority of infectious particles produced during infection.
Many details, as well as underlying mechanisms, of the virion morphogenesis process remain enigmatic. A longstanding question has been the origin and biogenesis of the crescent-shaped membranes that ultimately become the primary envelope of VACV. The apparent lack of continuity of crescent membranes with cellular membranes initially prompted a model in which the crescent membranes are synthesized de novo (5). More recent models, however, suggest that the crescent membranes are acquired from a cellular organelle, either the intermediate compartment between the endoplasmic reticulum and Golgi apparatus (ERGIC) (33, 40) or the endoplasmic reticulum (ER) (13). Consistent with these models, several membrane proteins destined for the MV envelope are synthesized on the ER (13, 37), and a pathway exists for the trafficking of MV membrane proteins from the ER to IV (13). However, it is unclear whether precursors of MV membranes or MV membrane proteins need to be actively trafficked from the ER to IV through a virus-mediated process, as MV membrane proteins are believed to be synthesized in viral factories (16) and no specific signal is required for MV membrane protein A9 to be incorporated into IV (14).
Several VACV proteins have been identified through studies of conditional-lethal mutants as essential for viral membrane biogenesis. F10 (42, 43), A11 (32), H7 (38), and L2 (21) are required at an early stage, as repression of their individual expression resulted in factories that accumulated large masses of viroplasm but lacked any membrane structure. A14 (35, 44) and A17 (34, 47) are required at a later stage, as repression of A14 or A17 expression resulted in accumulation of vesicular or tubular membrane structures at the boundaries of large viroplasm inclusions. A defect in crescent membrane formation was also observed in some temperature-sensitive (ts) mutants with a lesion in G5 (3) or H5 (7). However, G5 does not directly participate in crescent membrane biogenesis, as normal IV formation was observed in a G5R deletion mutant (39). H5 may also affect membrane formation indirectly, as it plays a role in the transcription of viral genes and replication of the viral genome (6, 19).
VACV A6 is a minor component of the virion core (1) and is conserved in all vertebrate poxviruses. Previously, we showed that A6 plays an essential role in virion morphogenesis by studying a temperature-sensitive A6 mutant (24). Because the temperature-sensitive mutant synthesizes a small amount of A6 at a nonpermissive temperature, we constructed an inducible A6 mutant that more stringently represses A6 expression. The characterization of the inducible A6 mutant showed that A6 is required for an early step in viral membrane biogenesis. Furthermore, we found that A6 is required for localization of several major MV membrane proteins to viral factories, suggesting that MV membrane proteins or precursors of MV membranes are trafficked to viral factories through a virus-mediated process.
BS-C-1 cells were maintained in minimum essential medium with Earle's salts supplemented with 10% fetal bovine serum (FBS). Baby hamster kidney (BHK) cells and HeLa cells were maintained in Dulbecco's modified Eagle's medium (DMEM) with 10% FBS. Murine monoclonal antibodies (MAbs) against V5 (Sigma-Aldrich), PDI (Enzo Life Sciences), ERGIC53 (Enzo Life Sciences), and Golgin 97(Invitrogen) and rabbit polyclonal antibody against calnexin (Enzo Life Sciences) were purchased from the vendors. Murine MAbs against VACV proteins A10 (26) (BG3 clone), A13 (48) (11F7), A14 (26) (FE11), D8 (26) (BD6), H3 (23) (number 41), WR148 (26) (HE7), and B5 (26) (LA10) were described previously. The murine MAb against I1 was derived from a mouse infected with VACV, similar to that described previously (48). Polyclonal mouse sera against L1 or F9 were obtained from mice immunized with recombinant L1 fused with glutathione S-transferase (GST) or F9 fused with maltose-binding protein (MBP), respectively.
The plasmid pVote-A6L-V5 was constructed by PCR amplification of the coding sequences for A6 with a C-terminal V5 epitope tag from pLJ2 (24), followed by insertion into NcoI and BamHI sites of pVote1 (46). vYB21, a VACV with the inducible A6L-V5 gene at the A56R locus, in addition to the endogenous A6L gene, was constructed through homologous recombination of pVote-A6L-V5 with vT7LacOI (provided by Bernard Moss ). vYB21 was isolated after three rounds of plaque purification under selection conditions for guanine phosphoribosyl transferase (GPT) according to a standard protocol (8).
pA5-GFP-A7, the transfer plasmid for replacing the endogenous A6L gene with a green fluorescent protein (GFP) gene, was constructed as follows. Approximately 500-bp left and right flanking sequences of A6L were PCR amplified from genomic DNA of WR virus with primer pairs 5′-CGGGATCCTTAGTTGTTTAATTTATTTGTGC-3′ plus 5′-CCCGATAAGCTTTACGAACTACATCTGATATTATT-3′ and 5′-CGGGAGCTCATCTTAAACATATAGGGAATCATAT-3′ plus 5′-ATAAGAATGCGGCCGCTGTTCAAAGTCTTATCAAATTCA-3′, respectively. The PCR products were then sequentially cloned into pYW31 (25) by using the restriction enzyme sites underlined in the primer sequences (BamHI, HindIII, SacI, and NotI) to flank the open reading frame (ORF) of GFP under the control of the P11 late promoter. Isopropyl-β-d-thiogalactoside (IPTG)-inducible, conditional-lethal A6L virus (iA6)-GFP was constructed through homologous recombination of pA5-GFP-A7 with vYB21 and was isolated after three rounds of plaque purification of GFP-expressing virus in the presence of 500 μM IPTG.
pA5-GUS-A7, the transfer plasmid for replacing the endogenous A6L gene with β-glucuronidase (GUS), was constructed by replacing the GFP cassette between BamHI and NotI sites in pA5-GFP-A7 with a cassette of GUS between BamHI and NotI sites of pBSgptgus (15). iA6-GUS was constructed through homologous recombination of pA5-GUS-A7 with vYB21 and was isolated after three rounds of plaque purification of GUS-expressing viruses in the presence of 500 μM IPTG, according to a standard protocol (8). iA6-GFP and iA6-Gus were propagated on BS-C-1 cells in the presence of 100 μM IPTG.
BS-C-1 cells in a 12-well tissue culture plate were infected with iA6-GFP, iA6-GUS, or vT7lacOI at a multiplicity of infection (MOI) of 5 PFU per cell. After 1 h of adsorption at room temperature, the inoculum was removed and the cells were washed with DMEM plus 1% FBS. The infected cells were incubated with medium in the presence or absence of various concentrations of IPTG. At various times after infection, cells were harvested and viral titers in the cell lysates were determined by plaque assay in the presence of 100 μM IPTG.
Western blot analysis was performed as previously described (25). Briefly, the samples were solubilized in sodium dodecyl sulfate (SDS) sample buffer, resolved by SDS-polyacrylamide gel electrophoresis (PAGE), transferred to nitrocellulose membranes, and blocked with Tris-buffered saline supplemented with 5% nonfat dried milk and 0.05% Tween 20 for 1 h at room temperature. Subsequently, the membranes were incubated with the antibodies and analyzed with chemiluminescence reagent (Pierce).
BHK or HeLa cells grown on coverslips were infected at 0.5 PFU/cell in the presence or absence of 100 μM IPTG. At 8 h postinfection (p.i.), the cells were fixed with 4% paraformaldehyde for 20 min, permeabilized with 0.5% saponin (Sigma-Aldrich) for 5 min, blocked with 10% FBS for 60 min, and stained with various antibodies for 1 h and an appropriate secondary antibody for an additional hour. The DNA was stained with DAPI (4′,6-diamidino-2-phenylindole) (Invitrogen). The coverslips were imaged with an Olympus FV-1000 laser scanning confocal system.
Subcellular fractionation was performed similarly to the method described by Earley et al. (9). HeLa cells grown in a 150-mm dish were infected with the virus at an MOI of 5 PFU per cell. After 2 h of adsorption at room temperature, the inoculum was replaced with fresh medium with or without 100 μM IPTG. At 8 h p.i., the cells were harvested by scraping them into medium, collected by centrifugation at 1,000 × g for 5 min, resuspended in 1 ml of homogenization buffer (0.25 M sucrose, 1 mM EDTA, 10 mM HEPES, pH 7.4) containing 0.2 mM phenylmethylsulfonyl fluoride and protease inhibitor cocktail tablets (Roche Diagnostics), and broken by passage through a 27-gauge syringe needle 10 times. The nuclei and unbroken cells were removed by centrifugation at 2,400 × g for 5 min. The postnuclear supernatant was loaded onto a 10.5-ml preformed 5 to 25% continuous iodixanol gradient (Axis Shield) and centrifuged at 200,000 × g for 2.5 h at 4°C. After centrifugation, 0.8-ml fractions were collected from the top of the tube. The protein in each fraction was precipitated with trichloroacetic acid and washed twice with acetone. Precipitated proteins were resuspended in SDS sample buffer, resolved by SDS-PAGE, and processed for Western blotting as described above.
For transmission electron microscopy, BS-C-1 cells in 60-mm-diameter dishes were infected with 1 PFU/cell of iA6L-GFP. At 24 h p.i., the cells were fixed with 4% formaldehyde and 1% glutaraldehyde for 1 h at 4°C. The cells were then scraped off and prepared for transmission electron microscopy by the Electron Microscopy Core Laboratory at the University of Texas Health Science Center (HSC) at San Antonio. The thin sections were examined with a JEOL 1230 transmission electron microscope.
We used the vT7LacOI and pVote system developed by Ward and coworkers (46) to construct two recombinant viruses that expressed A6 only in the presence of inducer IPTG. The recombinant viruses were constructed in two steps. First, an intermediate recombinant virus that contains both an inducible A6L gene and the endogenous A6L gene was constructed. Then, the endogenous A6L gene of the intermediate virus was replaced with either Green fluorescent protein (GFP) or β-glucuronidase (GUS) by homologous recombination in the presence of IPTG, resulting in two viruses referred to, respectively, as iA6-GFP and iA6-GUS. The inducible A6L viruses expressed A6 in an IPTG dose-dependent manner, as determined by Western blotting with an antibody against the V5 epitope tag appended to the C terminus of A6 (Fig. 1A). When the IPTG concentration was greater than 100 μM, in addition to the full-length A6 protein, a smaller protein of approximately 25 kDa was also detected by the anti-V5 antibody. This protein was present at a very low level, so it may represent some minor degradation of A6 in the cells or during sample preparation. In the absence of IPTG, the inducible viruses did not express a detectable level of A6 (Fig. 1A), produce any plaque (Fig. 1B), or grow in titer over 48 h of infection (Fig. 1C and D). The yield of inducible viruses at 24 h p.i. depended on the IPTG concentration in the medium, with the maximum yield at 50 μM IPTG (Fig. 1B). In the presence of 100 μM IPTG, the growth kinetics of the inducible viruses was similar to that of the parental vT7LacOI in one-step growth experiments, achieving maximum amplification in titers (~2 log) at 24 h p.i. (Fig. 1D). The replication defect of the inducible viruses in the absence of IPTG was due specifically to repression of A6, as it could be rescued by transfection of a plasmid containing A6L under the control of its natural promoter (data not shown). Metabolic labeling of infected cells with [35S]methionine and [35S]cysteine showed no gross difference in viral protein synthesis in the presence or absence of IPTG (data not shown), consistent with previous observations with the temperature-sensitive A6L mutant (24).
We then studied the effect of A6 repression on virion morphogenesis with transmission electron microscopy. In cells infected with iA6-GFP in the presence of IPTG, the entire spectrum of normal morphological forms of VACV, including crescents, IV, MVs, and wrapped virions, was observed at 24 h postinfection (Fig. 2A and B). In contrast, no crescents or IV were seen in cells infected with iA6-GFP for 24 h in the absence of IPTG (Fig. 2C and D). Instead, electron-dense viroplasm accumulated in the cells as inclusions that measured up to 2 to 3 μm in diameter. The majority of the viroplasm inclusions were not surrounded by any membrane (Fig. 2C and D), but some inclusions had short membrane arcs at their peripheries (Fig. 2E and F). Similar small membrane arcs were also observed when H7 or L2 expression was repressed (21, 38).
Recently, it was reported that a subset of MV membrane proteins became unstable when virion membrane formation was blocked by repression of L2, A11, or A17 (22). This subset of membrane proteins is involved in entry-fusion of VACV and includes F9 and L1. We thus assessed the effect of A6 repression on specific viral-protein levels in the cells. Cells were infected with iA6-GFP in the presence or absence of IPTG for 8 h. Western blots were then performed with antibodies against various VACV proteins. F9 and L1 proteins were detected in IPTG-treated cells, but they were undetectable in cells not supplemented with IPTG (Fig. 3E and F), indicating that, similar to L2, A11, and A17, A6 is required for the stability of L1 and F9. Based on densitometry, almost half of the A10 proteins (46%) were proteolytically processed in IPTG-treated cells, but 85% of the A10 proteins remained in the precursor form when A6 was repressed (Fig. 3G), consistent with a defect in virion morphogenesis, which is coupled with proteolytic processing of virion core proteins. Two forms of A14 protein, a 15-kDa form and an 18-kDa form, were detected in infected cells (Fig. 3B); the latter was previously shown to result from N-glycosylation (28). There was a significant increase in the level of A14 glycosylation when A6 was repressed (Fig. 3B). This is also consistent with a previous finding that A14 glycosylation was increased when virion membrane biogenesis was blocked due to the lack of a functional F10, H5, A17 (28), or A11 (32). There was no obvious change in protein level or posttranslational modification for major MV membrane proteins A13, D8, and H3 when A6 was repressed (Fig. 3A, C, and D).
Since MV membrane formation was blocked when A6 was repressed, we were curious about the localization of MV membrane proteins in infected cells. We infected BHK cells with iA6-GFP in the presence of absence of IPTG for 8 h and then processed the cells for immunofluorescence microscopy by using various MAbs against VACV proteins. In the presence of IPTG, A13, A14, D8, and H3 predominantly localized to viral-DNA-containing factories (labeled “F” in Fig. 4), which were stained with DAPI. Weak staining of these proteins was sometimes observed outside the factories. GFP, which was under the control of the P11 late promoter, was diffused throughout the cell. In the absence of IPTG, however, inclusions of GFP, with diameters from 1.4 to 3 μm, accumulated in and around DNA factories (Fig. 4). Similar GFP inclusions were previously observed when H7 expression was repressed (38), and they are thought to correspond to the large, dense masses of viroplasm seen under EM. Consistent with this idea, the GFP inclusions were also stained with antibodies against A10 (Fig. 4E), a core protein that was previously shown to associate with viroplasm. Most interestingly, most of the A13, A14, D8, and H3 proteins localized to cytoplasmic areas outside the viral factories (Fig. 4A to D). A13 also accumulated in a perinuclear area, which was clearly devoid of any viral DNA (arrow in Fig. 4A). Similar changes in the localization of MV membrane proteins were also observed at 16 and 24 h postinfection and in HeLa cells (data not shown). In contrast to altered localization of MV membrane proteins, the distribution of B5, a membrane protein of extracellular enveloped virions (EV), and I1, a DNA binding protein (18), did not change. In the presence or absence of IPTG, I1 localized to DNA factories (Fig. 4F), while B5 localized to structures outside the factories that were consistent with the Golgi apparatus (data not shown).
To find out whether MV membrane proteins localized to specific cellular organelles when A6 expression was repressed, we infected BHK cells with iA6-GUS for 8 h in the presence or absence of IPTG and costained the cells with antibodies against viral proteins and markers for cellular organelles. iA6-GUS was used here instead of iA6-GFP because the former did not express GFP and thus was more convenient for double labeling in immunofluorescence analysis. In the presence of IPTG, A13 and A14 did not colocalize with either protein disulfide isomerase (PDI), a resident ER protein, or ERGIC53, a marker for ERGIC (Fig. 5). In the absence of IPTG, A13 and A14 colocalized extensively with PDI. The area where A13 accumulated was also stained with the antibody against ERGIC53 (arrow in Fig. 5B). In an additional confirmatory experiment, D8 and H3, as well as A13 and A14, were found to colocalize extensively with calnexin, an integral membrane protein of the ER, when A6 expression was repressed (Fig. 6).
Altogether, the immunofluorescence results in Fig. 4 to to66 showed that, when A6 expression was repressed, major MV membrane proteins predominantly localized to cellular secretory compartments outside the viral factories.
To confirm the immunofluorescence results and to assess biochemically the membrane structures with which MV membrane proteins were associated, we also examined the subcellular localization of VACV proteins by fractionating the cells and assessing the distribution of viral proteins in the cell fractions. HeLa cells were infected with iA6-GFP in the presence or absence of IPTG for 8 h. The cells were then lysed and fractionated with a continuous iodixanol gradient. Proteins from gradient fractions were precipitated and analyzed by Western blotting (Fig. 7). In the presence of IPTG, A13, D8, unglycosylated A14, and a mature form of A10 were present in the denser parts of the gradient (fractions 12 to 15 in Fig. 7), consistent with their presence in MV virions, which are heavier than most cellular organelles. Some A13 and D8 and a small amount of unglycosylated A14 were also present in lighter fractions that also contained PDI (fractions 6 to 11). Since only unglycosylated A14 is incorporated into the virions (27) while glycosylation of A14 indicates its trafficking through the ER, the separation of glycosylated A14 from unglycosylated A14 in the gradient served as further evidence that the gradient was able to separate the ER from the virions. A6 was predominantly present at the top of the gradient (fractions 2 to 4), consistent with its localization, not to viral factories, but to the cytoplasm in immunofluorescence (24). WR148 and full-length A10 were also present at the top of the gradient (fractions 1 to 4). In the absence of IPTG, WR148 and full-length A10 remained at the top of the gradient. However, only small amounts of A13, A14, D8, and the mature form of A10 were found in the denser part of the gradient, where virions sedimented (fractions 12 to 15). A14, which was mostly glycosylated, cosedimented closely with PDI (fractions 6 to 12). A13 and D8 were present in lighter fractions that contained the Golgi marker Golgin 97, as well as in fractions that contained PDI. There was a very small amount of A6 at the bottom of the gradient, which may represent A6 that was brought in by the inoculum. Altogether, the cell fractionation data showed that when A6 expression was repressed, major MV membrane proteins were predominantly associated with membrane structures that were similar in density to the ER.
Previously, we analyzed a ts mutant with a lesion in A6 and showed that A6 was essential for VACV virion morphogenesis (24). The A6 mutation in the ts mutant reduced A6 protein stability in a temperature-dependent manner. As a result, cells infected by the ts mutant at the higher nonpermissive temperature showed reduced numbers of IV but no IVN or MV (24). Since there was still a low level of A6 protein at the nonpermissive temperature, we had speculated that A6 might be required for IV formation but that the low level of A6 protein at 42°C might allow the formation of a small number of IV (24). Therefore, to further define the role of A6 in virion morphogenesis, we constructed an IPTG-inducible A6L mutant that showed more stringent repression of A6 expression at normal temperature. In the absence of the inducer IPTG, the A6L inducible mutant did not express any detectable amount of A6, and there was no growth in the viral titer over 48 h of infection (Fig. 1). Previously, we had repeatedly tried and failed to generate an IPTG-inducible A6 mutant by utilizing a modified pVote transfer vector. In our previous attempts, A6L was cloned in pVot (10), which was derived from pVote by removing the encephalomyocarditis virus (EMCV) internal ribosome entry site (IRES) sequence and was used previously to make an inducible A22R virus (10). Our intention in using pVot was to generate a virus that would repress A6 expression more stringently than a pVote-derived virus, as the former would not allow any cap-independent translation of A6 from potential read-through transcripts derived from promoters upstream of the inducible T7 promoter. We were able to use a pVot-derived transfer vector to generate an intermediate virus with both the endogenous and the inducible A6L gene. However, we could not delete the endogenous A6L gene from the intermediate virus under IPTG-inducible conditions for unknown reasons. In retrospect, we think that the EMCV IRES sequence might be important for inducible expression of essential proteins that are needed at higher levels for viral replication.
Characterization of the A6L inducible mutant showed that A6 belongs to a class of viral proteins that play an essential role in an early step of virion membrane biogenesis. Direct evidence for such a role was obtained by ultrastructural analysis of infected cells with EM. When A6 expression was repressed, large, dense viroplasm inclusions accumulated in the cells with no sign of a crescent membrane (Fig. 2), similar to the phenotypes associated with the lack of F10, A11, H7, or L2 (21, 27, 32, 38, 42). Furthermore, A6 repression caused a dramatic increase in the glycosylation of MV membrane protein A14 and a great reduction in the levels of MV membrane proteins F9 and L1 (Fig. 3). Both of these phenotypes are associated with VACV mutants that are defective in virion membrane biogenesis. A14 is one of the few VACV MV membrane proteins that has an N-glycosylation site, but A14 is glycosylated at a low level during normal viral replication (28). A14 glycosylation was increased, however, when virion membrane biogenesis was blocked, presumably because A14 was retained or diverted to the ER in the absence of viral membranes. When A6 expression was repressed, almost all A14 protein was glycosylated (Fig. 3 and and7).7). This level of A14 glycosylation appeared to be as much as, if not more than, that in the absence of a functional A11 (32), A17, F10, or H5 (28). Some VACV MV membrane proteins that are part of the entry-fusion complex (EFC) became unstable when viral membrane formation was blocked by repression of L2, A11, H7, or A17 (22, 38). Similarly, EFC proteins F9 and L1 were undetectable in the absence of A6 (Fig. 3). Thus, based on similarity in EM morphology and the fate of several MV membrane proteins under nonpermissive conditions, A6 can be classified together with F10, A11, H7, and L2 in a class of viral proteins that play essential roles in an early step of virion membrane biogenesis. Although the molecular functions of this group of VACV proteins are largely unknown, it is critical to identify all the viral proteins that participate in the virion membrane biogenesis process in order to gain a molecular understanding of the unique mechanism used by the poxviruses to acquire their primary envelope.
Similar to most proteins that function in early steps of virion membrane formation, A6 does not appear to be a structural protein of MV. Mass spectrometry analysis previously identified A6, and an A6 ortholog in myxoma virus, as a minor virion component (1, 31, 49). Western blot analysis further demonstrated that A6 was present in the virion core (24). Immunofluorescence analysis, however, showed that, unlike MV structural proteins, A6 did not specifically localize to viral factories but instead localized throughout the cytoplasm (24). Indeed, cell fractionation with a density gradient showed that A6 was predominantly present in the lighter cytoplasmic fractions and only a small amount of A6 was present in the heavier virion fractions (Fig. 7). It is likely that A6 mainly functions in the cytoplasm and is only packaged into virions nonspecifically. Similarly, L2 appears to be nonspecifically incorporated into virions (22), while H7 and A11 are not packaged in virions. F10 is packaged in the virion core (20), but its role in membrane biogenesis is probably not a structural one, as it depends on the protein kinase activity of F10 (42). H7 is similar to A6 in that it localizes to the cytoplasm during infection (38). In contrast, L2 is a membrane protein that localizes to the ER (22), while A11 localizes to the factories (32). The intracellular localization of F10 is unclear, as one report suggested that it localizes to viral factories (42) while another report suggested that it localizes to the ER and ERGIC (30).
The A6L inducible mutant also showed a novel phenotype that has not been previously reported for VACV mutants that are defective in virion membrane biogenesis. During normal VACV infection, MV membrane proteins predominantly localize to viral factories, where virion assembly occurs. The intracellular localization of MV membrane proteins in the absence of A11, H7, or L2 was previously studied by immunofluorescence analysis of MV membrane protein L1 or A17, and no defect was observed (21, 38). In contrast, when A6L expression was repressed, there was a profound defect in localization of MV membrane proteins to viral factories. This defect was demonstrated with four different MV membrane proteins that were inserted into membranes through two different mechanisms. A13 and A14 were presumably synthesized on rough ER and cotranslationally inserted into membranes (37), while H3 and presumably D8 were synthesized on free ribosomes and posttranslationally inserted into membranes (4). The absence of A6 did not affect the levels of these four MV proteins in the cells but dramatically affected their intracellular localization. Immunofluorescence analysis showed that they no longer localized to viral factories but instead colocalized extensively with ER markers PDI and calnexin outside the factories (Fig. 4, ,5,5, and and6).6). A13 also accumulated in areas that were costained for ERGIC53, a marker for ERGIC. Furthermore, cell fractionation studies showed that, in the absence of A6, only a small amount of these MV proteins sedimented with virions, while the majority of the proteins sedimented at densities that were lighter than virions and were similar to that of the ER (Fig. 7).
EM analysis of VACV-infected cells previously showed that some MV membrane proteins were associated with ER or ERGIC tubules in cell sections (40). However, analysis of the entire cells by immunofluorescence microscopy showed that MV membrane proteins predominantly localized to viral factories and rarely colocalized with any cellular organelle markers (40). Similarly, despite the presence of a natural N-glycosylation site, the majority of the A14 proteins are unglycosylated, and minor glycosylated A14 proteins are not packaged in the virions (27). It is thus unprecedented to find that, in the absence of A6, the majority of major MV membrane proteins are present in secretory compartments outside viral factories and almost all A14 proteins are glycosylated. This is unlikely to be the result of MV membrane proteins being diverted to the ER located outside the factories from their normal trafficking to the viral membrane, which could have explained the association of low levels of MV membrane proteins with the ER or ERGIC in previous reports. It is more likely that, in the absence of A6, MV membrane proteins are synthesized outside the factories and fail to be recruited to the factories. This further suggests that ER components are not passively incorporated into viral factories but are actively recruited to viral factories through a virus-mediated process that requires at least A6. Thus, when A6 is absent, no ER components are recruited to the factories, and MV membrane proteins are inappropriately synthesized in the ER outside the viral factories. Alternatively, MV membrane proteins could be naturally synthesized in the ER outside the factories and then specifically recruited, along with the underlying membranes, to factories through an active, virus-mediated process that requires A6. Either way, this virus-mediated membrane recruitment process is likely to be quite different from intracellular trafficking of membrane vesicles from the ER to the Golgi apparatus, since blocking ER-to-Golgi vesicle transport with brefeldin A or a dominant-negative Sar-1 GTPase failed to inhibit the formation of IV (12, 45). Determining the molecular mechanism by which A6 functions may provide an intriguing story of how viruses manipulate intracellular membrane trafficking.
This work was supported by a grant from NIAID to Y. Xiang (AI079217).
Published ahead of print 7 March 2012