Search tips
Search criteria 


Logo of jvirolPermissionsJournals.ASM.orgJournalJV ArticleJournal InfoAuthorsReviewers
J Virol. 2012 May; 86(10): 5905–5914.
PMCID: PMC3347259

Formation of Orthopoxvirus Cytoplasmic A-Type Inclusion Bodies and Embedding of Virions Are Dynamic Processes Requiring Microtubules


In cells infected with some orthopoxviruses, numerous mature virions (MVs) become embedded within large, cytoplasmic A-type inclusions (ATIs) that can protect infectivity after cell lysis. ATIs are composed of an abundant viral protein called ATIp, which is truncated in orthopoxviruses such as vaccinia virus (VACV) that do not form ATIs. To study ATI formation and occlusion of MVs within ATIs, we used recombinant VACVs that express the cowpox full-length ATIp or we transfected plasmids encoding ATIp into cells infected with VACV, enabling ATI formation. ATI enlargement and MV embedment required continued protein synthesis and an intact microtubular network. For live imaging of ATIs and MVs, plasmids expressing mCherry fluorescent protein fused to ATIp were transfected into cells infected with VACV expressing the viral core protein A4 fused to yellow fluorescent protein. ATIs appeared as dynamic, mobile bodies that enlarged by multiple coalescence events, which could be prevented by disrupting microtubules. Coalescence of ATIs was confirmed in cells infected with cowpox virus. MVs were predominantly at the periphery of ATIs early in infection. We determined that coalescence contributed to the distribution of MVs within ATIs and that microtubule-disrupting drugs abrogated coalescence-mediated MV embedment. In addition, MVs were shown to move from viral factories at speeds consistent with microtubular transport to the peripheries of ATIs, whereas disruption of microtubules prevented such trafficking. The data indicate an important role for microtubules in the coalescence of ATIs into larger structures, transport of MVs to ATIs, and embedment of MVs within the ATI matrix.


Viral factories (VFs) are cytoplasmic juxtanuclear bodies (also called B-type inclusions) that organize poxvirus DNA replication, intermediate and late transcription, protein synthesis, and assembly of mature virions (MVs). Poxvirus MVs are infectious particles containing genomic DNA and enzymes necessary for early transcription packaged in a multiprotein core surrounded by a lipoprotein membrane (5). During infection with orthopoxviruses (OPXVs), a population of MVs segregates by moving along microtubules to the trans-Golgi network or late endosomes, where they acquire an additional double membrane to form triple-membrane wrapped virions (WVs). WVs are transported on microtubules to the periphery of the cell and exocytosed, losing the outermost membrane to become double-enveloped extracellular virions (EVs). Most MVs, however, are not destined to become EVs and are retained in the cytoplasm until cell lysis. The MVs of some OPXVs, including strains of cowpox virus (CXPV), ectromelia virus, fowlpox virus, and raccoonpox virus (but not vaccinia virus [VACV], monkeypox virus, or variola virus), are embedded in cytoplasmic, proteinaceous matrices called A-type inclusions (ATIs) (8, 18, 21, 28), which are thought to protect infectious particles after release into the environment.

ATIs are composed of multiple copies of a single ATIp viral polypeptide (26). ATIp is a large protein (160 kDa in CPXV) containing C-terminal, hydrophobic repeats (10, 26) that are required for assembly of inclusion bodies (14). ATIs are localized throughout the cytoplasm and increase in size during infection. The embedding of MVs within ATIs that leads to their occlusion requires the A26 protein (14, 22). A26 is an MV-specific protein anchored to the MV membrane by interactions with the A27-A17 complex (3, 4, 13).

OPXVs commandeer the cytoskeleton for entry, intracellular transport, and exit from the host cell. Modulation of the actin cytoskeleton facilitates virion entry (20, 23, 24), after which viral cores associate with microtubules, allowing migration to a juxtanuclear position (2, 27). In addition, progeny MVs traffic on microtubules to sites of wrapping (30, 31), and WVs use microtubules to move to the plasma membrane, where exocytosis of EVs occurs (12, 29, 32). Actin polymerization then propels many of the EVs at the tips of cytoplasmic projections to neighboring cells (6). Given their size (approximately 360 by 270 by 250 nm) and the consequent low rate of diffusion through the cytoplasm, it seems likely that the association of MVs with inclusion bodies is dependent on cytoskeletal transport, although this has not been demonstrated. McKelvey et al. (22) proposed that A26 might have a role in directing transport of MVs from VFs to ATIs during occlusion.

Some OPXVs, such as VACV, the best-characterized member of the OPXV genus, encode a C-terminally truncated ATIp that does not self-associate to form ATIs (1, 7). We demonstrated previously that VACV could be enabled to make ATIs and occlude MVs if the CPXV ATI gene were substituted for the shorter VACV A25 homolog (14). Thus, VACV encodes all of the necessary proteins for this process except for the full-length ATIp. As much more is known about the molecular biology of VACV than about that of CPXV or other OPXVs that make ATIs naturally, we primarily used the reconstructed VACV system to study ATI formation and the mechanism of occlusion. We previously demonstrated that the direct association of A26 with both the A27-A17 complex and the ATIp is required to mediate occlusion (14). In the present study, we showed that ATI enlargement was dependent on both continued protein synthesis and coalescence of smaller ATIs, which was documented by live imaging and confirmed with CPXV. Additionally, the coalescence of ATIs contributed to the redistribution of MVs from the exterior to the interior of the ATI matrix. Both the coalescence of ATIs and occlusion required intact microtubules and MVs moved from virus factories to ATIs at speeds characteristic of microtubular transport.


Cells and virus infection.

Human HeLa cells were maintained in minimum essential medium (Quality Biological, Gaithersburg, MD) containing 2 mM l-glutamine, 100 units of penicillin/ml, 100 μg of streptomycin/ml, and 10% fetal bovine serum (FBS). Virus infection of cells was carried out in the medium described except that 2.5% serum was used. Recombinant viruses were propagated as previously described (9).

Recombinant viruses and plasmids.

The recombinant VACVs used in this study were derived from the WR strain and described previously (14). Briefly, vATI+A26+.A4:YFP is a recombinant VACV expressing a C-terminally hemagglutinin (HA)-tagged CPXV strain, Brighton Red (CPXV-BR) ATIp (ATIpHA), and a C-terminally V5-tagged VACV, A26 (A26V5). The vA25+A26+.A4:YFP strain expresses HA-tagged VACV A25 (A25HA) and A26V5. The vA25+A26.A4:YFP strain expresses A25HA but with the A26 gene deleted. In the viruses described above, A4:YFP denotes the fusion of the A4 gene in frame with the gene expressing yellow fluorescent protein (YFP). vA25A26+ is an A25 deletion virus that expresses A26V5. vA25+A26+, vATI+A26+, and vATI+A26 are as described above except that the A4 gene was not fused to YFP.

The pCherry:ATIHA DNA was assembled by amplifiying the mCherry coding sequence with a forward primer that introduced the ATI promoter at the 5′ end and a reverse primer that generated a 20-nucleotide (nt) sequence overlapping the 5′ end of the ATI gene. The resulting PCR product lacked the mCherry stop and ATI start codons such that the respective genes were fused in frame. The ATI sequence was amplified using a forward primer containing the ATI sequence just 3′ of the start site and a reverse primer coding for the HA epitope tag. The two PCR products were joined by overlap PCR using Accuprime Pfx (Invitrogen, Carlsbad, CA) and the mCherry forward and ATI reverse primers and blunt-end ligated into pCR-BluntII-TOPO (Invitrogen).

Antibodies for immunofluorescence.

Rabbit polyclonal antibodies and mouse monoclonal antibodies (Covance, Princeton, NJ) that recognize the HA tag (YPYDVPDYA) were used for immunofluorescence studies at a 1:250 dilution. All Alexa Fluor-conjugated anti-IgGs were purchased from Invitrogen and used at a dilution of 1:250.

Preparation for confocal microscopy.

For confocal microscopic imaging of fixed cells, 24-well dishes containing 12-mm-diameter coverslips seeded with HeLa cells were infected at a multiplicity of infection of 0.1 PFU per cell in minimum essential medium and incubated with rocking at room temperature for 1 h. Following adsorption, the inoculum was removed and the cells were incubated at 37°C. Cells were washed twice in Dulbecco's phosphate-buffered saline (PBS) and then fixed with 4% paraformaldehyde for 15 min. Cells were washed with PBS, permeabilized with 0.2% Triton X-100–PBS for 7 min, washed (4 times for 5 min each time) in PBS, and blocked with 10% FBS diluted in PBS. Primary antibodies and secondary Alexa Fluor-conjugated IgG were diluted as described above in 10% FBS–PBS and incubated with cells for 1 h and 45 min, respectively, at room temperature. Coverslips were mounted using ProLong Gold mounting medium containing 4′,6′-diamidino-2-phenylindole (DAPI) from Invitrogen.

Preparation for live imaging.

For live imaging of virion movement, 35-mm-diameter glass bottom dishes (ibidi GmbH, Munich, Germany) were seeded with HeLa cells and infected at a multiplicity of 0.1 PFU/cell for 1 h. The inoculum was removed, and medium containing 10 μM ST-246 {4-trifluoromethyl-N-(3,3a,4,4a,5,5a,6,6a-octahydro-1,3-dioxo-4,6-ethenocycloprop[f]isoindol-2(1H)-yl)-benzamide} (a kind gift of Siga Technologies) was added to prevent formation of WVs. For live imaging of inclusion body enlargement and occlusion over time, Lab-Tek 8-chambered borosilicate cover glasses (Thermo Scientific, Rochester, NY) were seeded with HeLa cells, infected at a multiplicity of infection of 1 PFU/cell for 1 h at room temperature, and transfected with plasmids expressing pCherry:ATIHA. At 8 h after transfection, medium containing the transfection mix was removed and replaced with untreated medium or medium containing 30 μM nocodazole (NOC) with or without cycloheximide (CHX) (10 μg/ml). Live-imaging experiments examining the coalescence of ATIs in CPXV-infected cells were performed as described above except that the transfection mix was replaced after 5 h of incubation with untreated medium.

Confocal microscopy and image processing.

A Leica SP2 inverted four-channel microscope was used for fixed-cell imaging, and a Leica SP5 inverted 5-channel confocal microscope with a high-speed resonant scanner was used for live imaging. Live cells were imaged on a stage containing an incubation chamber that maintained cells at 37°C with 5% CO2. Representative images and a time series monitoring virion movement were deconvolved with Huygens Essential version 3.5.0, 64-bit software (Scientific Volume Imaging, Hilversum, Netherlands). Z-sections were reconstructed in three dimensions (3D) using the IMARIS Bitplane Scientific software version 7.0 (Saint Paul, MN) Surpass feature in the Maximum Intensity Projection (MIP) mode, which displays the voxels with the maximum intensity of all Z-sections along the viewing plane. Snapshot images of Z-sections viewed in Surpass mode are referred to as 3D reconstructions. The IMARIS Spots feature was used to track MVs in 4D (X, Y, Z, and time [t]) and to extract statistical analyses of particle dynamics. Data extraction of MV embedment and inclusion body enlargement was performed on images in a time series that were not deconvolved. IMARIS Surfaces was used to define individual inclusion bodies as regions of interest using the mCherry channel. Surfaces were generated by threshold calculations based on absolute intensity without additional filters. The autoregressive motion algorithm was selected to track ATI data determined over time using the Surfaces software, and the resulting defined tracks were verified and/or corrected manually by visually tracking ATIs. MV embedment was analyzed based on the sum of the intensities of YFP voxels enclosed within the Surface object comprised of the ATI. Total YFP fluorescence was measured as the intensity sum and compared to the volume of mCherry fluorescence characterized as the intensity mean (intensity sum normalized by volume).


Time course of ATIp synthesis and ATI formation.

In previous studies performed with CPXV, the ATIp was detected starting at 6 h postinfection (hpi) and ATIs at 9 to 12 hpi (16, 26). We employed a recombinant VACV that expresses the full-length CPXV ATI protein with a C-terminal HA tag, in place of the truncated A25 homolog, and the A4 core protein fused to YFP (vATI+A26+.A4:YFP) under the control of their natural VACV promoters. The high sensitivity and high resolution provided by fluorescent confocal microscopy were used to visualize ATIp synthesis with anti-HA antibody. The A4 protein fused to YFP allowed us to simultaneously monitor viral core protein synthesis and MV formation (17). VFs and nuclei were stained with DAPI. HeLa cells were synchronously infected with vATI+A26+.A4:YFP and fixed at 3, 5, 7, 9, and 11 hpi. The VFs appeared as juxtanuclear bodies that were initially compact but became more diffuse after 5 h (Fig. 1A). Numerous punctate bodies containing ATIp were distributed throughout the cytoplasm as early as 3 hpi, whereas A4:YFP was detected in or adjacent to VFs at the same time (Fig. 1A). The ATIs increased in size by 5 hpi and had a characteristic donut appearance due to peripheral staining, presumably because of an inability of the antibody to penetrate the dense structures (Fig. 1A). In contrast to A4:YFP, ATIp was not seen in VFs in any of the infected cells viewed (n = 50). ATIs continued to increase in size throughout infection, although some small ATIs were still seen at 9 hpi (Fig. 1A, arrowhead and inset). MVs appeared as punctate YFP structures, and those associated with ATIs were mostly at their periphery at 7 hpi. At 11 h, MVs were dispersed throughout many of the ATIs, as determined by analyzing Z-sections, whereas other ATI had few MVs (Fig. 1A).

Fig 1
Changes in ATI size and number over time. HeLa cells were infected with VACV strain vATI+A26+ and fixed at 3, 5, 7, 9, and 11 hpi. (A) Representative stacked confocal images of ATIs in infected HeLa cells. Rabbit anti-HA antibodies, followed by Alexa ...

As ATIs increased in size, they decreased in number.

We had the impression that ATIs increased in size as their number decreased. To quantify this relationship, we measured the diameters of ATIs and counted them at consecutive times during the course of infection. HeLa cells were infected with vATI+A26+.A4:YFP and fixed at 5, 7, 9, and 11 hpi. ATIs were visualized by anti-HA antibody staining of ATIp in the fixed cells, which were imaged as Z-stacks. IMARIS Bitplane software was used to measure the diameter of each ATI in the Z-section at which the diameter was greatest. The lowest threshold of ATI diameters was set at 0.3 μm because of the limits of resolution; therefore, ATIs that measured less than 0.3 μm in diameter were not scored. The mean size of ATIs increased from 0.66 ± 0.02 to 1.36 ± 0.04 μm from 5 to 11 hpi (Fig. 1B). During the same time, the mean number of ATIs decreased (Fig. 1B). Plotting the distribution of individual ATI diameters provided further insight. At 5 hpi, ATI diameters distributed as a sharp peak, indicating relative uniformity of sizes (Fig. 1C). By 7 hpi, the major peak shifted, indicating that the population as a whole had larger diameters and that the distribution was more heterogeneous than at 5 hpi. At 9 and 11 hpi, large-diameter ATIs predominated (Fig. 1C). We considered that the increase in ATI size and reduction in number could have been at least partly due to coalescence, although some small ATIs persisted or continued to form, as shown in the inset at the 9-h time point in Fig. 1A.

Enlargement of ATIs is due to continued protein synthesis and ATI coalescence.

Next, we compared the effects of inhibiting protein synthesis and microtubular transport on the formation of ATIs. HeLa cells were infected with vATI+A26+.A4:YFP for 4 h, and then the medium was replaced with (i) drug-free medium; (ii) medium containing CHX, an inhibitor of protein synthesis; (iii) medium containing NOC, which depolymerizes microtubules; or (iv) medium containing both drugs. Infected cells were incubated for an additional 5 h and fixed. ATIs were stained with anti-HA antibody and imaged as Z-stacks using confocal microscopy. ATIs were enumerated and the diameters measured as described above.

The ATI diameters seen at 9 hpi in cells incubated with CHX were smaller than those in untreated cells at the same time point but larger than ATI diameters in cells fixed at 4 hpi (time zero) (Fig. 2A). The impression that new protein synthesis contributed to ATI size was corroborated by measurements showing that the peak of ATI diameters in cells incubated with CHX did not shift as far toward a value representing larger sizes as the peak of ATI in untreated cells (Fig. 2B). In addition to some increase in ATI sizes, the average ATI number per cell decreased in the presence of CHX (Fig. 2C).

Fig 2
Contribution of new protein synthesis and microtubular transport to ATI size and number. (A) Stacked confocal images. CHX (10 μg/ml) or NOC (30 μm) or both or untreated medium (No drug) was added to vATI+A26+.A4:YFP-infected HeLa cells ...

We considered that the increase in size and decrease in number of ATIs during infection even in the presence of CHX might have been due to coalescence and might require microtubules. NOC treatment depolymerized microtubules, as shown in Fig. 2A, and resulted in a relatively narrow peak of ATIs representing structures of greater uniformity and smaller diameter than occurred without drug (Fig. 2B). Furthermore, the number of ATIs in NOC-treated cells was greater than in untreated cells (Fig. 2C), consistent with an inhibition of coalescence.

The ATI diameters after incubation with both NOC and CHX were more drastically reduced and resembled the peak of diameters in cells fixed at 4 hpi (time zero) (Fig. 2A and B). With CHX and NOC treatment, the mean number of ATIs per cell was slightly reduced compared to the number seen at time zero, likely reflecting the effects of the combination of inhibition of new ATIp synthesis and prevention of coalescence (Fig. 2C). The data suggested that both new protein synthesis and intact microtubules contributed to ATI enlargement.

Visualization of ATI coalescence.

The finding that ATIs enlarged while the mean number per cell decreased with time suggested that inclusion bodies were merging and coalescing. Moreover, the inhibition by NOC implied a role for microtubules. To visualize ATI dynamics over time, we transfected plasmids expressing mCherry fused to the N terminus of ATIp (pCherry:ATIHA) into cells infected with VACV expressing YFP fused to the A4 core protein. We tracked individual mCherry-fluorescing ATIs and YFP-fluorescing MVs by collecting Z-stack data with a confocal microscope at 15- to 30-min intervals between 12 and 16 hpi. Figure 3A shows an image of a 3D reconstruction of Z-stacks at 12 hpi. The lower ATI shown in the box is filled with MVs, whereas the upper ATI has MVs mostly at the periphery. Figure 3B contains individual sectional images taken at 30-min intervals, showing merger of the two ATIs. An intermediate bilobed structure was evident at 12.5 hpi. The YFP fluorescence gradually redistributed within the merged ATIs over the 4-h interval. The box in the right panel of Fig. 3A shows an image of a 3D reconstruction of the merged ATI at 16 h.

Fig 3
Visualization of coalescing ATIs. (A) Images of confocal Z-sections reconstructed in 3D, showing a portion of a vA25A26+.A4:YFP-infected cell at 12 h posttransfection (left) and 16 h posttransfection (right) with plasmids expressing mCherry fused ...

We conducted similar experiments to determine if ATIs coalesce in CPXV-infected cells. Cells infected with CPXV-BR or vATI+A26+ were transfected with pCherry:ATIHA to visualize ATIs and imaged from 8 to 18 hpi. A total of 20 ATIs from each infection in two separate experiments were selected randomly and monitored for coalescence. As shown in Fig. S1 in the supplemental material, 16 and 14 ATIs in CPXV and recombinant VACV infections, respectively, were observed to coalesce with at least one other ATI. Anti-ATI antibodies and mCherry colabeled ATIs in cells infected with CPXV and transfected with pCherry:ATIHA in fixed-cell experiments (data not shown). Since CPXV-BR lacks a functional A26 protein, MVs were not associated with ATIs. Nevertheless, coalescence of ATI occurred, indicating that MVs were not required for this event.

Live imaging of cells infected with a vATI+A26+.A4:YFP was performed with a confocoal microscope at 8 hpi. At that time, MVs began to associate with the peripheries of ATIs, which could be recognized as spherical structures studded with YFP-fluorescing MVs. In contrast, VFs containing A4:YFP appeared as amorphous structures and were easily distinguishable from ATIs. This distinction was confirmed by analysis of cells infected with a similar virus expressing the truncated ATIp, which did not contain the spherical structures (data not shown), and in experiments using fluorescently labeled ATIp, as shown subsequently (Fig. 4A, right panel). As shown in Fig. 3C and Movie S1 in the supplemental material, two mobile ATIs studded with MVs on their exteriors contacted and merged into a larger, spherical ATI.

Fig 4
Contribution of coalescence to MV distribution within ATI. (A) Representative images of filled and unfilled ATIs. HeLa cells were infected with vA25A26+.A4:YFP and transfected with plasmids expressing mCherry fused to ATIp. Images shown represent ...

Correlation of ATI coalescence with increased numbers of MVs within ATIs.

The matrix of one of the two ATIs coalescing in Fig. 3B was filled with MVs whereas the other had MVs mainly on the periphery. Immediately after coalescence, the MVs remained in the portion of the matrix that the filled ATI contributed to the merger. Gradually, however, the MVs became more evenly distributed within the merged ATI (Fig. 3B). The latter image and others like it suggested that coalescence drives MVs into the interior of the ATI matrix. To further assess the effects of coalescence on MV embedment, we used live-cell microscopy to follow individual ATIs over time. We infected cells with vA25A26+.A4:YFP and transfected them with plasmids expressing mCherry fused to ATIp. Figure 4A illustrates a filled ATI with MVs in the interior (left panel) and an unfilled ATI with MVs only at the surface (right panel). Note that a Z-stack analysis is necessary to distinguish the two types of ATIs. A time series of images was made from 10 to 18 h after transfection. We randomly picked filled and unfilled ATIs present at 18 h and then traced them back in time. Eight ATIs from each group that were not filled with MVs at 10 h were analyzed over the entire 8-h time period. Only two of the ATIs that remained unfilled at 18 h had coalesced with at least one other ATI during the 8-h time course, whereas seven of the eight filled ATIs merged with at least one other ATI (Table 1). The filled and unfilled ATIs were defined as regions of interest using the IMARIS Surfaces feature that allows measurements for the mCherry and YFP channels within those ATIs to be extracted over time (see Movie S2 in the supplemental material). The time series for the 16 ATIs scored for volume (calculated as the sum of the mCherry pixels within the ATI) and MV content (calculated as the YFP fluorescent intensity sum) are shown in Fig. 4B and C. The ATIs observed visually to have been filled also had the greatest increase in YFP fluorescence over time (Fig. 4C). Most filled inclusion bodies underwent several coalescence events, as is reflected in the overall greater increase in volume of the filled inclusion bodies relative to the unfilled ATIs (Fig. 4B). An exception, filled ATI 7 (F_7) was not observed to coalesce (Table 1), and we did not observe an increase in volume with ATI F_7 relative to the unfilled ATIs (Fig. 4B). F_6 and F_2 were graphed separately, because the increases in volume and YFP fluorescent intensities were out of range relative to the data determined for the other ATIs (Fig. 4D). With both F_2 and F_6, steep increases of YFP fluorescence correlated with coalescence with ATIs containing MV, as indicated by the concomitant increase in ATI volume. The cumulative mean YFP intensity at 18 h posttransfection of coalesced ATIs was higher than that of noncoalesced ATIs (Fig. 4E).

Table 1
Incidence of coalescence of unfilled and filled ATIsa

Microtubules are important for MV embedment within ATIs.

Thus far, our focus had been on changes in the size of ATI, and we showed that the changes resulted from enlargement and coalescence due to protein synthesis and microtubules, respectively. In addition, coalescence contributed to the internalization of MVs from the periphery of the ATIs. Further experiments were carried out to more directly assess the contributions of microtubules and protein synthesis to embedment of MVs. HeLa cells were infected with vA25A26+.A4:YFP and transfected with plasmids expressing mCherry fused to ATIp. At 8 hpi, cells were treated with NOC with or without CHX or with CHX alone. After an additional 2 h, we selected ATIs containing MVs only at the peripheries and collected Z-stacks at 20-min intervals for 8 h. We then compared the effects of the drugs on MV embedment within the matrix of 10 ATIs at 18 h posttransfection. The mean fluorescent intensity of YFP in ATIs determined as described in the preceding section was lower in cells treated with NOC or CHX alone or together compared to that seen in ATIs in untreated cells (Fig. 5A), demonstrating that the microtubular network and new protein synthesis are important for the embedment of MVs within ATIs.

Fig 5
Effects of NOC and CHX on occlusion. (A) HeLa cells were infected with vA25A26+.A4:YFP and transfected with plasmids expressing mCherry fused to ATIp. Bars indicate the mean intensities of YFP at 10 h posttransfection within ATIs in cells treated ...

We also compared the effects of NOC and cytochalasin D, an inhibitor of actin polymerization, on the association of MVs with ATIs following virion assembly. HeLa cells were infected with vATI+A26+.A4:YFP in the presence of rifampin, which reversibly prevents virus assembly (11, 25) without blocking ATI formation (14, 16). After 10 h, the rifampin was washed out and the cells were incubated for an additional 10 h in the presence of either NOC or cytochalasin D. The cells were fixed and stained with anti-HA antibodies to detect ATIp; occlusion of MVs was monitored using A4p:YFP fluorescence (Fig. 5B). The effects of the drugs on microtubules and actin were visualized using both mouse anti-α- and anti-β-tubulin antibody staining and 594-phalloidin staining, respectively (data not shown). ATIs in randomly selected cells from three separate experiments were scored as positive (+) if MVs were present at the periphery or within the matrix or negative (−) for occlusion if MVs were unassociated. Cytochalasin D caused cells to shrink and have a misshapen appearance but did not have much of an effect on occlusion (Fig. 5B). NOC, however, decreased the number of occluded ATIs by more than 60% (Fig. 5B).

MVs use microtubules for transport to ATIs.

Although microtubules could enhance MV embedment by increasing coalescence of ATIs, they might also have a more direct role in MV movement to ATIs. To facilitate the analysis of MV movement, we used the drug ST-246 to suppress the formation of WVs (33), which are known to move on microtubules. HeLa cells were infected with vATI+A26+.A4:YFP in the presence of ST-246 and were imaged live at 8 hpi, a time at which ATIs encrusted with YFP-fluorescing MVs could be recognized by their characteristically spherical pattern. These ATIs were easily distinguishable from clusters of virions, as verified by imaging cells infected with VACV expressing a truncated ATIp and A4:YFP (data not shown). MVs were tracked by imaging them in the X, Y, and Z planes at 1 frame/1.573 s. The individual Z-stacks were typically about 4 μm thick. Some MVs were observed to associate with an ATI and then disassociate and move away from inclusion bodies. MVs were tracked with the IMARIS Spots function, and those that visibly adhered to inclusion bodies corresponding to at least 10 time points following contact were used for calculations. MVs moved to inclusion bodies at an average speed of 0.501 ± 0.637 μm/s (Table 2), which is consistent with transport on microtubules (31). Particle velocities alternated between high and low, but particles did not pause as often as the MVs previously observed moving between VFs and sites of wrapping (31).

Table 2
Frame-by-frame measurements of movement of four individual virionsa

A26 is not required for MV movement but contributes to particle velocity.

A26 associates with MVs and is required for their occlusion, suggesting that A26 might be a movement protein (22). We therefore compared MV particle dynamics in the presence and absence of A26 to see if there was a difference in movement. VACVs that expressed the truncated form of ATIp (A25) instead of full-length ATIp were used to prevent ATI formation, since this would lead to occlusion of A26+ but not A26 particles. HeLa cells were infected with a VACV expressing an HA-tagged A25 and a V5-tagged A26 (vA25+A26+.A4:YFP) or an A26 deletion VACV expressing an HA-tagged A25 (vA25+A26.A4:YFP). Following adsorption, the medium was replaced with medium containing ST-246 to prevent formation and movement of WVs. At 7 to 8 hpi, MVs were imaged live in regions of cells that were distal from factories and from virion aggregates in order to allow tracking of individual particles. Tracking was carried out using IMARIS Spots software, and the movement of MVs in the presence and absence of A26 was analyzed. MV tracks are shown in Fig. 6A and the track lengths and speeds in Fig. 6B. The deletion of A26 did not affect overall track lengths; however, mean particle speeds were significantly lower.

Fig 6
MV movement dynamics in the absence of A26. (A) Images of tracks in two representative cells are shown with color coding indicating speed. HeLa cells were infected with vA25+A26+.A4:YFP or vA25+A26.A4:YFP in media containing 10 μm ST-246 ...


The formation of dense cytoplasmic inclusions containing embedded virions is a characteristic feature of several orthopoxvirus species that has not been thoroughly investigated. The present studies showed that ATI formation and the occlusion of MVs are dynamic processes that involve the microtubular network at multiple steps. In previous phase-contrast microscopy experiments, ATIs were first visualized at 9 hpi as large, quasispherical masses (16). Using fluorescent confocal microscopy, we detected ATI as small puncti throughout the cytoplasm as early as 3 hpi in cells infected with a recombinant VACV expressing an HA-tagged ATIp. The enhanced detection allowed us to follow the evolution of ATIs from their inception and the embedment of MVs with precision. The ATIs enlarged considerably to quasispherical structures by 5 to 7 hpi, and frequency distributions indicated relative uniformity of ATI diameters. By 9 and 11 hpi, however, ATIs in a large range of sizes existed and their number actually decreased. Further insight into ATI development came from studies performed with inhibitors. Addition of CHX at 4 or 5 hpi prevented the appearance of new, small ATIs that evidently normally continue to arise throughout infection. Surprisingly, while treatment with CHX decreased the number of ATIs, the size of ATIs still increased, albeit to a lesser extent than in the absence of drug. This finding suggested that enlargement of ATIs does not depend solely on the incorporation of newly synthesized ATIp. We considered that the merger of ATIs could contribute to their enlargement and reduction in number.

Imaging of ATI dynamics was achieved by engineering a chimeric protein consisting of mCherry fused to ATIp. This allowed us to track individual ATIs within living infected cells by time-lapse microscopy and demonstrate their frequent coalescence. We also observed coalescence of ATI in cells infected with CPXV-BR, indicating that this event was not unique to the VACV recombinant system used in this study and did not require MV embedment, since CPXV-BR is deficient in this step. Bilobed coalescence intermediates often appeared prior to acquiring characteristic quasispherical shapes. Optical sectioning of adjacent, coalescing ATIs showed a continuum between the bodies. Moreover, we observed ATI mobility and the coalescence of individual spherical ATIs, identifiable by YFP-fluorescing MVs studding their peripheries, in real time. In view of the very large size of ATIs, it seemed likely that the cytoskeletal network was necessary for their mobility. The idea of a key role for microtubules was supported by the inhibitory effect of NOC on ATI enlargement. VFs have also been shown to coalesce, but the role of the cytoskeleton has not been determined (19).

We investigated the mechanisms contributing to the embedment of MVs within ATIs by using a virus that expresses YFP fused to an MV core protein. At early times in the experiment, MVs were mostly at the surface of the dense ATI bodies. We observed the internalization of MVs during the coalescence and reorganization of ATI. Moreover, the majority of ATIs that were filled with MVs by 18 hpi were shown by time-lapse microscopy to have undergone one or more coalescence events. In addition, coalesced ATIs on average contained higher numbers of YFP MVs than ATIs that were not observed to coalesce. Further studies indicated that microtubules and new protein synthesis contributed to the embedment of MV within ATIs.

While coalescence contributes to the redistribution of MVs from the surface of ATIs to their interior, we were intrigued by how MVs localize to ATI sites, which are distant from VFs. MV movement to ATIs was investigated using a confocal microscope for live imaging at 1.57 frames/s. ST-246, an inhibitor of MV wrapping, was used to prevent formation and subsequent movement of WVs on microtubules, since such activity might otherwise make it difficult to distinguish MVs and WVs. The mean speeds of about 5 μm/s for individually tracked MVs were similar to speeds seen for MV movement to sites of wrapping along microtubules (31). Further support for the idea of microtubular movement came from inhibitor studies. Cytochalasin D, which disrupts actin fibers, did not interfere with occlusion, whereas NOC drastically reduced MV embedment.

A26 is required for MV occlusion, but the mechanism has not been completely defined (22). In our present live-imaging studies, we noted that some MVs tend to associate and disassociate multiple times before stably associating with ATIs. In view of the ability of A26 to interact with ATIp (14), this interaction could mediate stable association of MVs with inclusions. An additional possibility is that A26 serves as a movement factor for MV occlusion (22). The primary observation supporting this hypothesis came from electron microscopic studies that showed MVs distant from ATIs when A26 was absent or nonfunctional. To test whether A26 modulates MV movement, we compared the particle dynamics of a VACV A26 deletion mutant with that of a VACV encoding A26 in the absence of ATI formation. Cells were infected in the presence of ST-246 to arrest WV formation, and particle movement over time was recorded. Although MV lacking A26 moved with statistically lower velocities, the number of particles moving in cells infected with the A26 deletion VACV appeared similar to the number moving in cells infected with VACV expressing A26. The deletion of A26 also did not affect total track length. Further studies are needed to assess whether A26 has multiple functions in MV occlusion.

Taken together, our data suggest important roles for microtubules in the enlargement of ATIs and occlusion of MVs within them. First, MVs require microtubules to move to and associate with the peripheries of ATIs. Second, the microtubular network is required for coalescence of ATIs, which contributes to internalization of MVs in ATIs. A block in new protein synthesis also hinders ATI enlargement and occlusion, presumably due in large part to a block in ATIp synthesis. An interesting issue is how the ATIp gets to the cytoplasmic locations where inclusions form and enlarge. During our studies, we never detected ATIp in VFs, suggesting that it either was rapidly transported out or was not synthesized there. Synthesis outside the VF would be exceptional, since transcription and translation of intermediate and late VACV gene products have been shown to occur within specialized domains of the VF (17). Nevertheless, Ichihashi and Dales (15) showed large numbers of polyribosomes associated with the periphery of ATIs. We are currently investigating the subcellular localization of ATIp synthesis and exploring possible mechanisms for mRNA transport from the VF.

Supplementary Material

Supplemental material:


We thank Catherine Cotter for cells and Cedric Boularan for providing a plasmid containing the mCherry gene. Steven Becker, Matthew Gastinger, and Juraj Kabat and the NIAID Research and Technologies Branch were extremely helpful with the acquisition of live-imaging data and subsequent analyses. Kady Honeychurch and Dennis Hruby of SIGA technologies kindly provided ST-246.

The work was supported by the Division of Intramural Research, NIAID, NIH.


Published ahead of print 21 March 2012

Supplemental material for this article may be found at


1. Amegadzie BY, Sisler JR, Moss B. 1992. Frame-shift mutations within the vaccinia virus A-type inclusion protein gene. Virology 186:777–782. [PubMed]
2. Carter GC, et al. 2003. Vaccinia virus cores are transported on microtubules. J. Gen. Virol. 84:2443–2458. [PubMed]
3. Ching Y-C, et al. 2009. Disulfide bond formation at the C termini of vaccinia A26 and A27 proteins does not require viral redox enzymes and suppresses glycosaminoglycan-mediated cell fusion. J. Virol. 83:6464–6476. [PMC free article] [PubMed]
4. Chiu W-L, Lin C-L, Yang M-H, Tzou D-L, Chang W. 2007. Vaccinia virus 4c(A26L) protein on intracellular mature virus binds to the extracellular cellular matrix laminin. J. Virol. 81:2149–2157. [PMC free article] [PubMed]
5. Condit RC, Moussatche N, Traktman P. 2006. In a nutshell: structure and assembly of the vaccinia virion. Adv. Virus Res. 66:31–124. [PubMed]
6. Cudmore S, Cossart P, Griffiths G, Way M. 1995. Actin-based motility of vaccinia virus. Nature 378:636–638. [PubMed]
7. de Carlos A, Eduardo P. 1991. Isolation and characterization of mutants of vaccinia virus with a modified 94-kDa inclusion protein. Virology 185:768–778. [PubMed]
8. Downie AW. 1939. A study of the lesions produced experimentally by cowpox virus. J. Pathol. Bacteriol. 48:361–379.
9. Earl PL, Cooper N, Wyatt S, Moss B, Carroll MW. 1998. Preparation of cell cultures and vaccinia virus stocks, p 16.16.1–16.16.3 In Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K, editors. (ed), Current protocols in molecular biology, vol 2 John Wiley and Sons, New York, NY.
10. Funahashi S, Sato T, Shida H. 1988. Cloning and characterization of the gene encoding the major protein of the A-type inclusion body of cowpox virus. J. General Virology. 69(Pt. 1):35–47. [PubMed]
11. Grimley PM, Rosenblum EN, Mims SJ, Moss B. 1970. Interruption by rifampin of an early stage in vaccinia virus morphogenesis: accumulation of membranes which are precursors of virus envelopes. J. Virol. 6:519–533. [PMC free article] [PubMed]
12. Hollinshead M, et al. 2001. Vaccinia virus utilizes microtubules for movement to the cell surface. J. Cell Biol. 154:389–402. [PMC free article] [PubMed]
13. Howard AR, Senkevich TG, Moss B. 2008. Vaccinia virus A26 and A27 proteins form a stable complex tethered to mature virions by association with the A17 transmembrane protein. J. Virol. 82:12384–12391. [PMC free article] [PubMed]
14. Howard AR, Weisberg AS, Moss B. 2010. Congregation of orthopoxvirus virions in cytoplasmic A-type inclusions is mediated by interactions of a bridging protein (A26p) with a matrix protein (ATIp) and a virion membrane-associated protein (A27p). J. Virol. 84:7592–7602. [PMC free article] [PubMed]
15. Ichihashi Y, Dales S. 1973. Biogenesis of poxviruses: relationship between a translation complex and formation of A-type inclusions. Virology 51:297–319. [PubMed]
16. Ichihashi Y, Matsumoto S, Dales S. 1971. Biogenesis of poxviruses: role of A-type inclusions and host cell membranes in virus dissemination. Virology 46:507–532. [PubMed]
17. Katsafanas GC, Moss B. 2007. Colocalization of transcription and translation within cytoplasmic poxvirus factories coordinates viral expression and subjugates host functions. Cell Host Microbe 2:221–228. [PMC free article] [PubMed]
18. Knight JC, et al. 1992. Further analyses of the orthopoxvirus volepox virus and raccoon poxvirus. Virology 190:423–433. [PubMed]
19. Lin YC, Evans DH. 2010. Vaccinia virus particles mix inefficiently, and in a way that would restrict viral recombination, in coinfected cells. J. Virol. 84:2432–2443. [PMC free article] [PubMed]
20. Locker JK, et al. 2000. Entry of the two infectious forms of vaccinia virus at the plasma membrane is signaling-dependent for the IMV but not the EEV. Mol. Biol. Cell 11:2497–2511. [PMC free article] [PubMed]
21. Marchal J. 1930. Infectious ectromelia. A hitherto undescribed virus disease of mice. J. Pathol. Bacteriol. 33:713–728.
22. McKelvey TA, Andrews SC, Miller SE, Ray CA, Pickup DJ. 2002. Identification of the orthopoxvirus p4c gene, which encodes a structural protein that directs intracellular mature virus particles into A-type inclusions. J. Virol. 76:11216–11225. [PMC free article] [PubMed]
23. Mercer J, Helenius A. 2008. Vaccinia virus uses macropinocytosis and apoptotic mimicry to enter host cells. Science 320:531–535. [PubMed]
24. Moser TS, Jones RG, Thompson CB, Coyne CB, Cherry S. 2010. A kinome RNAi screen identified AMPK as promoting poxvirus entry through the control of actin dynamics. PLoS Pathog. 6:e1000954. [PMC free article] [PubMed]
25. Moss B, Rosenblum EN, Katz E, Grimley PM. 1969. Rifampicin: a specific inhibitor of vaccinia virus assembly. Nature 224:1280–1284. [PubMed]
26. Patel D, Pickup D, Joklik W. 1986. Isolation of cowpox virus A-type inclusions and characterization of their major protein component. Virology 149:174–189. [PubMed]
27. Ploubidou A, et al. 2000. Vaccinia virus infection disrupts microtubule organization and centrosome function. EMBO J. 19:3932–3944. [PubMed]
28. Randall CC, Gafford LG, Darlington RW, Hyde J. 1964. Composition of Fowlpox Virus and Inclusion Matrix. J. Bacteriol. 87:939–944. [PMC free article] [PubMed]
29. Rietdorf J, et al. 2001. Kinesin-dependent movement on microtubules precedes actin-based motility of vaccinia virus. Nat. Cell Biol. 3:992–1000. [PubMed]
30. Smith GL, Vanderplasschen A, Law M. 2002. The formation and function of extracellular enveloped vaccinia virus. J. Gen. Virol. 83:2915–2931. [PubMed]
31. Ward BM. 2005. Visualization and characterization of the intracellular movement of vaccinia virus intracellular mature virions. J. Virol. 79:4755–4763. [PMC free article] [PubMed]
32. Ward BM, Moss B. 2001. Vaccinia virus intracellular movement is associated with microtubules and independent of actin tails. J. Virol. 75:11651–11663. [PMC free article] [PubMed]
33. Yang G, et al. 2005. An orally bioavailable antipoxvirus compound (ST-246) inhibits extracellular virus formation and protects mice from lethal orthopoxvirus challenge. J. Virol. 79:13139–13149. [PMC free article] [PubMed]

Articles from Journal of Virology are provided here courtesy of American Society for Microbiology (ASM)