The Interaction between Tegument Proteins ORF33 and ORF45 Plays an Essential Role in Cytoplasmic Virion Maturation of a Gammaherpesvirus

ABSTRACT

Tegument, which occupies the space between the nucleocapsid and the envelope, is a unique structure of a herpesvirion. Tegument proteins are major components of tegument and play critical roles in virus life cycle. Murine gammaherpesvirus 68 (MHV-68), a member of the gammaherpesvirus subfamily, is closely related to two human herpesviruses, Kaposi's sarcoma-associated herpesvirus (KSHV) and Epstein-Barr virus (EBV). We have previously shown that MHV-68 ORF33, conserved among all herpesviruses, encodes a tegument protein that is associated with intranuclear capsids and is essential for virion morphogenesis and egress. Another tegument protein ORF45, which is conserved only among gammaherpesviruses, also plays an essential role in virion morphogenesis of MHV-68. In this study, we investigated the underlying mechanism and showed that these two proteins colocalize and interact with each other during virus infection. We mapped the ORF33-interacting domain to the conserved carboxyl-terminal 23 amino acids (C23) of ORF45. Deletion of the C23 coding sequence in the context of viral genome abolished the production of infectious virions. Transmission electron microscopy results demonstrated that C23 of ORF45 are essential for virion tegumentation in the cytoplasm. We further mapped the ORF45-interacting domain to the N-terminal 17 amino acids (N17) of ORF33. Deletion of the N17 coding sequence in the context of viral genome also abolished production of infectious virions, and N17 of ORF33 are also essential for virion tegumentation in the cytoplasm. Taken together, our data strongly indicate that the interaction between ORF45 and ORF33 plays an essential role in cytoplasmic maturation of MHV-68 virions.

IMPORTANCE A critical step in viral lytic replication is the assembly of progeny viral particles. Herpesviruses are important pathogens. A herpesvirus particle comprises, from inside to outside, four layers: DNA core, capsid, tegument, and envelope. The tegument layer contains dozens of virally encoded tegument proteins, which play critical roles in virus assembly. Murine gammaherpesvirus 68 (MHV-68) is a tumor-associated herpesvirus and is closely related to Kaposi's sarcoma-associated herpesvirus and Epstein-Barr virus. We previously found that the absence of either tegument protein ORF33 or ORF45 inhibits the translocation of nucleocapsids to the cytoplasm and blocks virion maturation, but the underlying mechanism remained unclear. Here, we showed that ORF33 interacts with ORF45. We mapped their interaction domains and constructed viral mutants with defects in ORF33-ORF45 interaction. Transmission electron microscopy data demonstrated that the assembly of these viral mutants in the cytoplasm is blocked. Our results indicate that ORF33-ORF45 interaction is essential for gammaherpesvirus replication.

INTRODUCTION

The herpesvirus family consists of three subfamilies, alpha-, beta-, and gamma-herpesvirinae. Members of the gammaherpesvirus subfamily are widespread in nature and infect a variety of mammalian species, including humans (1). Murine gamma-herpesvirus 68 (MHV-68) is a member of the gammaherpesvirus subfamily and is genetically related to Kaposi’s sarcoma-associated herpesvirus (KSHV) and Epstein-Barr virus (EBV), which cause malignant neoplasms (2). A herpesvirion consists of four distinct structures: genomic DNA, capsid, tegument, and envelope (3). The tegument, which occupies the place between the nucleocapsid and the envelope, is a unique structure in herpesvirus families and plays important roles in virus life cycle (4, 5). Herpesvirus morphogenesis is a highly complicated process that can be roughly divided into four stages: capsid formation in the nuclei, primary envelopment by budding at the inner nuclear membrane and deenvelopment by fusing with the outer nuclear membrane, tegumentation and secondary envelopment in the cytoplasm, and egress from the cell (6, 7). Tegumentation is a crucial step for virion maturation, during which tegument proteins interact with not only nucleocapsids but also each other (4, 8). Although specific tegument protein interactions play important roles in herpesvirus tegumentation, the underlying molecular mechanisms remain largely unclear.

We previously found that MHV-68 tegument protein open reading frame 33 (ORF33) is expressed with true late kinetics. In the absence of ORF33, egress of capsids from the nucleus into the cytoplasm is markedly inhibited, and maturation of virions in the cytoplasm is arrested in a partially tegumented stage (9). ORF33 is accumulated in the nucleus at discrete areas and associated with capsids in the nucleus prior to primary envelopment, consistent with its role in nuclear egress (10). Furthermore, ORF33 interacts with ORF38, and they colocalize at the trans-Golgi network (TGN) (11).

A previous report on genome-wide protein interaction network of MHV-68 showed that ORF33 interacts with tegument protein ORF45 (12). ORF45 is a virion-associated protein (13), and is conserved among gammaherpesviruses, but the homology is limited and mostly resides at the carboxyl-terminal ends. Exogenously supplemented MHV-68 ORF45 in ORF45-null virions plays an essential role during the immediate early phase of viral infection (14). By using Transmission Electron Microscopy (TEM), we showed that ORF45 is functionally involved in virus assembly and egress and is essential for infectious virus production (15). Its homologue in KSHV is a multifunctional protein which has been identified as an immediate early gene and also encodes a tegument protein (16, 17). It can associate with other viral tegument proteins, such as ORF33 and ORF36, and contribute to efficient production of viral particles (18, 19). In addition, KSHV ORF45 interacts with kinesin-2 motor protein KIF3A that transports cargoes along microtubules to cell periphery; this interaction helps transportation of newly assembled KSHV virions along microtubules, indicating a role of ORF45 in viral egress (20). KSHV ORF45 also associates with lipid rafts to facilitate the maturation and release of virion particles (21).

In this study, we investigated the interaction between ORF33 and ORF45 in detail and elucidated the functional role of this interaction in virion assembly. We mapped the ORF33-binding domain of MHV-68 to the highly conserved carboxyl-terminal 23 aa (C23) of ORF45 and the ORF45-binding domain to the N-terminal 17 aa (N17) of ORF33. Both ORF45 C23 domain and ORF33 N17 domain are essential for virion maturation in the cytoplasm, indicating that the interaction between ORF45 and ORF33 plays an essential role in virion tegumentation in the cytoplasm and hence production of progeny virions.

RESULTS

ORF33 colocalizes and interacts with ORF45 during viral infection.

ORF33 and ORF45 are both tegument proteins that are expressed during the late stage of MHV-68 lytic infection cycle. We found that ORF33 interacts with ORF45 during cotransfection using coimmunoprecipitation (co-IP) assay (Fig. 1A). We also determined the localization of ORF33 and ORF45 in transfected cells. When expressed individually, ORF33 was localized in both the cytoplasm (on average ~60%, quantitation of 10 cells) and the nucleus (~40%), while ORF45 was distributed mostly in the nucleus, consistent with our previous reports (9, 15). However, when coexpressed, ORF33 was distributed predominantly in the nucleus (~89%) and colocalized with ORF45, indicating that the interaction between ORF33 and ORF45 may contribute to the translocation of cytoplasmic ORF33 to the nucleus (Fig. 1B). To further confirm that ORF33 interacts with ORF45 in the context of viral infection, we infected BHK-21 cells with MHV-68 and harvested cells at 24 h postinfection (hpi). Co-IP assays using either anti-ORF33 antibody or anti-ORF45 antibody demonstrated that these two proteins interacted with each other during virus infection (Fig. 1C). When examining their subcellular localization in MHV-68 infected BHK-21 cells by three-dimensional Structured Illumination Microscopy (SIM), we found colocalization of ORF33 and ORF45 in both the nucleus and the cytoplasm. In the nucleus, both ORF33 and ORF45 showed a diffused distribution (Fig. 1D), while in the cytoplasm they mainly formed big bright agglomerate, suggesting they gathered at an area assumed to be “tegument deposit” (Fig. 1D).

FIG 1.

ORF33 colocalizes and interacts with ORF45 during viral infection. (A) 293T cells were transfected with pHA33, pFlag45 or both, respectively. At 48 hpt, cells were harvested, subjected to Co-IP using FLAG beads, and analyzed by Western blotting. (B) 293T cells were transfected with pHA33, pFlag45 or both. At 24 hpt, cells were fixed and stained with anti-HA (red channel) or anti-Flag (green channel) antibody, or DAPI (blue channel). Samples were observed with a confocal microscopy. (C) BHK-21 cells were infected with MHV-68 at an MOI of 3 for 24 h. Cells were harvested, and cell lysates were either immunoprecipitated with anti-ORF33 antibody or mouse IgG (left panel) or immunoprecipitated with anti-ORF45 antibody or rabbit IgG (right panel). Cell lysates and precipitates were then subjected to Western blotting with antibodies as indicated below. (D) BHK-21 cells were infected with MHV-68 (MOI = 3), and at 24 hpi, cells were examined by indirect immunofluorescence. ORF45 was detected using a rabbit anti-ORF45 polyclonal antibody, followed by an Alexa Fluor 488-conjugated secondary antibody (green channel). ORF33 was detected using a mouse anti-ORF33 monoclonal antibody, followed by an Alexa Fluor 568-conjugated secondary antibody (red channel). Nuclei were stained with DAPI (blue channel). Samples were observed with a three-dimensional Structured Illumination Microscope (Delta Vision OMX).

The carboxyl-terminal tail of ORF45 is critical for its interaction with ORF33.

To investigate the functional role of ORF33-ORF45 interaction in virus assembly, we sought to first identify their interaction domains. To map the sequence(s) in ORF45 protein that mediate its interaction with ORF33, we constructed several N-terminal and C-terminal deletion mutants of ORF45 (Fig. 2A). We then individually cotransfected each of the full-length (pFlag45) or truncated ORF45 expression plasmid plus the ORF33 expression plasmid (pHA33) into 293T cells. Co-IP results showed that ORF33 was pulled down by full-length or N-terminal deletion mutants of ORF45, but not by ORF45 mutants lacking the C-terminal 23 amino acids (aa) (named ORF45DC23) or C-terminal 41 aa (named ORF45DC41), indicating that the C-terminal 23 aa of ORF45 are required for its binding to ORF33 (Fig. 2B). Since the result from Fig. 1 showed that ORF45 induced the translocation of ORF33 from the cytoplasm to the nucleus, we further determined whether the translocation of ORF33 depends on its interaction with ORF45. Confocal microscopy results showed that ORF45DC23 by itself was distributed mainly in the nucleus, similar to full-length ORF45. However, when ORF33 was coexpressed with ORF45DC23, ORF33 retained its distribution in both the nucleus (~51%) and the cytoplasm (~49%) (Fig. 2C), in contrast to nuclear localization in the presence of full-length ORF45 (Fig. 1B). These data suggested that the C-terminal 23 aa of ORF45 are responsible for its interaction with ORF33 and efficient translocation of ORF33 to the nucleus.

FIG 2.

The carboxyl-terminal tail of ORF45 is critical for its binding to ORF33. (A) Schematic diagram of ORF45 truncations. (B) 293T cells were individually cotransfected with pHA33 plus expression plasmid for FLAG-tagged full-length ORF45 or truncated ORF45. At 48 hpt, cells were harvested, subjected to Co-IP using FLAG beads, and analyzed by Western blotting. (C) 293T cells were transfected with pFlag45DC23 plasmid or together with pHA33. At 24 hpt, cells were fixed and stained with anti-HA (red channel) or anti-Flag (green channel) antibody, or DAPI (blue channel). Samples were observed with a confocal microscopy.

The carboxyl-terminal 23 aa of ORF45 are essential for infectious virus production.

To investigate the functional significance of the C-terminal 23 aa of ORF45 in MHV-68 lytic phase, we introduced triple-stop codons and a Bgl II site upstream of the coding sequences for the C-terminal 23 aa of ORF45 in viral genome using the MHV-68 BAC system, to generate a BAC named ORF45DC23 BAC (Fig. 3A). A revertant was also generated and named ORF45DC23.R BAC (Fig. 3B). Viral BACs were examined by restriction enzyme digestions. As expected, digestion of ORF45DC23 BAC with Bgl II resulted in two restriction fragments of 3.7-kb and 2.0-kb, compared to a 5.7-kb restriction fragment generated by wild type (WT) or ORF45DC23.R BAC (Fig. 3C). Digestion of ORF45DC23 BAC with BamH I yielded a pattern similar to that of WT or ORF45DC23.R BAC, suggesting that no undesired rearrangement was introduced during recombination (Fig. 3C). These results were further confirmed by PCR amplification and DNA sequencing.

FIG 3.

Construction of an ORF45DC23 BAC. (A) Nucleotide sequence of the region containing the mutation in the ORF45DC23 MHV-68 mutant. A Bgl II site and triple stop codons were inserted between nucleotides 549 and 550 of ORF45 to create ORF45DC23 MHV-68 mutant. The introduced Bgl II site is underlined, and the triple stop codons are boxed. (B) Schematic representation of ORF45 and its flanking ORFs in WT BAC, ORF45DC23 BAC, and ORF45DC23.R BAC. The sizes of predicted restriction fragments are indicated. (C) Enzymatic digestion of WT BAC, ORF45DC23 BAC, and ORF45DC23.R BAC by BamH I or Bgl II. WT BAC DNA, ORF45DC23 BAC DNA, and ORF45DC23.R BAC DNA were isolated from E. coli strain SHG68 (GS1783). The DNAs were digested with BamH I or Bgl II as indicated on the top. The asterisks mark the positions corresponding to the predicted fragments of Bgl II-digested WT or ORF45DC23.R DNAs, and the arrow marks the position corresponding to that of ORF45DC23 DNA.

We first investigated the functional consequence of deleting the C-terminal 23 aa from ORF45. We individually transfected BHK-21 cells with WT, ORF45DC23, or ORF45DC23.R BAC and checked the transfectants daily. Severe cytopathic effect (CPE) was readily observed in WT and ORF45DC23.R BAC samples at 48 hpt; however, no CPE was detected in ORF45DC23 BAC sample up to 90 hpt (Fig. 4A). For quantitation, supernatant from BAC transfected 293T cells at 70 hpt were collected, and virus titers were determined by plaque assays. Virus titers of WT and ORF45DC23.R samples reached 10⁵ PFU/mL; however, no progeny virion was detected in ORF45DC23 sample (Fig. 4B). We also conducted indirect immunofluorescence assay on BAC transfected cells, to examine the expression of capsid protein ORF65 and tegument protein ORF33. Consistent with results from the plaque assays, we observed high percentage of cells expressing ORF33 and ORF65 in WT or ORF45DC23.R sample, indicating that progeny virions were produced in the original transfectants to infect nearby cells and express viral lytic genes. In contrast, fluorescent signals were only observed in single cells in ORF45DC23 sample, indicating a blockage in viral lytic replication and hence production of progeny virions (Fig. 4C). Western blot analyses of the cell lysates at 70 hpt showed that in ORF45DC23 sample, ORF45DC23 was expressed at the expected, smaller sizes as a doublet, whereas ORF33 and other viral proteins examined were expressed at the correct sizes (Fig. 4D). To further characterize the mutant BAC, we examined the subcellular distribution of ORF33 and ORF45 (or mutant ORF45) in BAC-transfected 293T cells. Consistent with our plasmids cotransfection data, ORF33 and ORF45 colocalized and mainly resided in the nucleus (~88% and ~89%, respectively) in WT BAC transfected cells. However, in ORF45DC23 BAC transfected cells, about 59% ORF33 and 84% ORF45 were localized to the nucleus, respectively, and ORF33 showed additional, evidently cytoplasmic localization (~41%) (Fig. 4E). Taken together, these data demonstrated that C-terminal 23 aa of ORF45 are essential for efficient translocation of cytoplasmic ORF33 to the nucleus and production of infectious virus.

FIG 4.

The carboxyl-terminal 23 aa of ORF45 are essential for infectious virus production. (A) BHK-21 cells were individually transfected with WT BAC, ORF45DC23 BAC, or ORF45DC23.R BAC, and at 90 hpt, cell morphology was visualized under a microscope. (B) Supernatants from 293T cells individually transfected with WT BAC, ORF45DC23 BAC, or ORF45DC23.R BAC were collected at 72 hpt and passaged onto fresh cells. After two rounds of total passages, supernatants were harvested, and virus titers were determined by plaque assays in BHK-21 cells in duplicates. (C) BHK-21 cells were individually transfected with WT BAC, ORF45DC23 BAC, or ORF45DC23.R BAC, and indirect immunofluorescence assay was performed at 58 hpt. Samples were observed with a fluorescence microscope (FV1200). The capsid protein ORF65 was detected using an anti-ORF65 antibody followed by an Alexa Fluor 488-conjugated secondary antibody (green channel). The tegument protein ORF33 was detected using an anti-ORF33 antibody followed by an Alexa Fluor 568-conjugated secondary antibody (red channel). DAPI staining was shown by blue channel. (D) 293T cells were transfected with WT, ORF45DC23, or ORF45DC23.R BAC respectively. At 70 hpt, cells were collected and subjected to Western blotting. Viral proteins were detected with specific antibodies as indicated to the right. (E) 293T cells were transfected with WT BAC or ORF45DC23 BAC, respectively. At 32 hpt, cells were fixed and stained with anti-ORF33 (red channel) and anti-ORF45 (green channel) antibodies. Nuclei were stained with DAPI (blue channel). Samples were observed with a confocal microscopy.

The carboxyl-terminal 23 aa of ORF45 are essential for virion tegumentation in the cytoplasm.

We reasoned that truncation of ORF45 may lead to a defect(s) in virus assembly since ORF45 plays an essential role in virion maturation (15). We therefore analyzed the ultrastructural phenotype of WT BAC-, ORF45DC23 BAC-, or ORF45DC23.R BAC-transfected 293T cells by transmission electron microscopy (TEM). Based on the current model of herpesvirus morphogenesis (primary envelopment - de-envelopment - secondary envelopment), we checked for newly assembled viral particles in each stage of virus maturation. In the nuclei, all three types of capsids were observed in WT- as well as ORF45DC23- and ORF45DC23.R-transfected cells (Fig. 5, A1, B1 and C1): a capsids (empty), b capsids (containing scaffold proteins), and c capsids (filled with viral genomic DNA). Notably, capsids in ORF45DC23 BAC-transfected cells were capable of budding through the nuclear membrane into the cytoplasm, as we observed primary enveloped virion in the perinuclear space (Fig. 5, B2). In the cytoplasm of ORF45DC23 BAC-transfected cells, large numbers of immature viral particles were surrounded by high-density materials considered to be tegument deposits (Fig. 5, B3), however, no mature virion was formed in this area or in cytoplasmic vesicles, compared to WT or ORF45DC23.R BAC-transfected cells (Fig. 5, A3, B3 and C3). Moreover, no viruses were found to be associated with cell surface (Fig. 5, B4). In contrast, in WT or ORF45DC23.R BAC-transfected cells, fully enveloped virions inside cytoplasmic vesicles (Fig. 5, A3 and C3) and extracellular mature virions associated with or adjacent to the cellular plasma membrane (Fig. 5, A4 and C4) were observed. These results indicated that the C-terminal 23 aa of ORF45 are essential for virion maturation in the cytoplasm.

FIG 5.

The carboxyl-terminal 23 aa of ORF45 are essential for virion maturation in the cytoplasm. 293T cells were transfected with WT BAC, ORF45DC23 BAC, or ORF45DC23.R BAC respectively. Approximately 70-nm thin sections were prepared at 72 hpt and examined by a 120-kV transmission electron microscope (FEI, Tecnai Spirit). A capsids (a), B capsids (b), and C capsids (c) were found in the nuclei (Nu) of WT BAC, ORF45DC23 BAC and ORF45DC23.R BAC-transfected cells (A1, B1, C1). Primarily enveloped virions were found in the perinuclear space (A2, B2, C2). In the cytoplasm (Cyto), viral particles, including mature virions, were observed in WT BAC and ORF45DC23.R BAC-transfectants (A3, C3, black arrows). A large number of capsids, but no mature virions, gathered at the tegument deposit in ORF45DC23 BAC-transfectant (B3, black arrowheads). In the extracellular space, WT BAC and ORF45DC23.R BAC-transfectants contained mature virions associated with plasma membranes (A4, C4, black arrows), whereas no mature virion was found in ORF45DC23 BAC-transfectant (B4).

To quantify the defect in virion morphogenesis that results from deleting the C-terminal 23 aa of ORF45, viral particles at various stages of maturation in TEM micrographs were classified and enumerated. As shown in Table 1, WT, ORF45DC23 and ORF45DC23.R BAC-transfectants had similar proportions of capsids to total particles in the nuclei (40.31%, 49.05%, and 54.24%, respectively). In the cytoplasm, WT and ORF45DC23 BAC transfectants also had similar proportions of unenveloped capsids (45.26% and 47.88%). Differences, however, were noted in the proportions of enveloped mature virions in the cytoplasm: in WT and ORF45DC23.R BAC transfectants, the proportions were 5.30% and 6.88%, respectively, whereas in ORF45DC23 BAC transfectant, it was zero. Furthermore, extracellular virions were not found in ORF45DC23 sample, compared to WT (5.72%) and ORF45DC23.R (9.76%) samples. These observations and statistical results demonstrated that the C-terminal 23 aa of ORF45 are essential for virion maturation in the cytoplasm.

TABLE 1.

Distribution of viral particles

BAC	No. of particlesa	Value (%) for:
		Nucleusb	Perinuclear spacec	Cytoplasm		Cell surfacef
				Unenveloped capsidsd	Enveloped virionse
WT	1434	578 (40.31)	49 (3.42)	649 (45.26)	76 (5.30)	82 (5.72)
ORF45 DC23	1111	545 (49.05)	34 (3.06)	532 (47.88)	0 (0)**g	0 (0)*g
ORF45 DC23.R	1250	678 (54.24)	17 (1.36)	347 (27.76)	86 (6.88)	122 (9.76)

^aTotal viral and subviral particles enumerated in 14 selected cells that contained viral particles.

^bValues are calculated as number of nucleocapsids/total number of viral particles.

^cValues are calculated as number of perinucleus capsids/total number of viral particles.

^dValues are calculated as number of immature virions (including capsids) in the cytoplasm/total number of viral particles.

^eValues are calculated as number of mature virions in the cytoplasm/total number of viral particles.

^fValues are calculated as number of fully mature extracellular virions/total number of viral particles.

^g*, P < 0.05; **, P < 0.01 versus WT, unpaired t test.

The N-terminal 17 aa of ORF33 are critical for its interaction with ORF45.

The above results showed that the C-terminal 23 aa of ORF45 interacts with ORF33 (Fig. 2) and are essential for cytoplasmic virion maturation (Fig. 5, Table 1). To determine whether the functional role of ORF45 in cytoplasmic virion maturation is mediated by its interaction with ORF33, but not by interaction with other proteins, we next mapped the region in ORF33 that mediates its interaction with ORF45. We constructed several N-terminal and C-terminal deletion mutants of ORF33 (Fig. 6A). Results showed that ORF33 or ORF33-1-131, but not the vector, ORF33DN17, or ORF33-126-327, was coimmunoprecipitated with ORF45 (Fig. 6B), indicating that the N-terminal 17 aa of ORF33 are required for its binding to ORF45. We also detected the localization of ORF33DN17 in plasmid transfected cells. Like full-length ORF33, ORF33DN17 by itself showed nuclear and cytoplasmic distribution (Fig. 6C). However, when ORF33DN17 was coexpressed with ORF45, its localization seemed to be unaltered (~45% in the nucleus), and no colocalization between these two proteins was observed. In contrast, wild-type ORF33 was translocated to the nucleus (~90%) in the presence of ORF45 (Fig. 6D). These data demonstrated that the ORF33 N-terminal 17 aa are important for its interaction with ORF45 and consequently increased translocation to the nucleus.

FIG 6.

The N-terminal 17 aa of ORF33 are critical for its binding to ORF45. (A) Schematic diagram of ORF33 truncations. (B) 293T cells were individually cotransfected with pFlag45 plus the expression plasmid for HA-tagged full-length ORF33 or truncated ORF33. At forty-eight hours posttransfection, cells were harvested, subjected to Co-IP using FLAG beads, and analyzed by Western blotting. (C) 293T cells were transfected with plasmid pHA33 or pHA33DN17 individually. At 72 hpt, cells were fixed and stained with anti-HA (red channel) antibody. Nuclei were stained with DAPI (blue channel). Samples were observed with a confocal microscopy. (D) 293T cells were cotransfected with pFlag45 plus pHA33, or pFlag45 plus pHA33DN17. After 24 hpt, cells were fixed and stained with anti-HA (red channel) and anti-Flag (green channel) antibodies. Nuclei were stained with DAPI (blue channel). Samples were observed with a confocal microscopy.

The N-terminal 17 aa of ORF33 are essential for infectious virus production.

To investigate the functional significance of the N-terminal 17 aa of ORF33 and its interaction with ORF45 in MHV-68 lytic phase, we first engineered mutations in the context of viral genome so that the N-terminal 17 aa from ORF33 are not expressed. Analysis of the viral genome sequence revealed that ORF33 overlaps with ORF32 by 41 bp, but their coding frames were different. We, therefore, mutated the ATG of ORF33 into ACG, and introduced an Acl I site and a new ATG codon immediately downstream of the coding sequences for the ORF33 N-terminal 17 aa, generating a mutant BAC named ORF33DN17 BAC (Fig. 7A). A revertant, ORF33DN17.R BAC, was also generated (Fig. 7B). Viral BACs were examined by restriction enzyme digestions. As expected, digestion of ORF33DN17 BAC with Acl I resulted in a new restriction fragment of 2.3-kb, compared to that of WT or ORF33DN17.R BAC. Digestion of ORF33DN17 BAC with BamH I yielded a pattern similar to that of WT or ORF33DN17.R BAC, suggesting that no undesired rearrangement was introduced during recombination (Fig. 7C). The sequences surrounding the insertion site in ORF33DN17 BAC were further PCR amplified and verified by DNA sequencing.

FIG 7.

Construction of an ORF33DN17 MHV-68 BAC. (A) Nucleotide sequence of the region containing the mutation in the ORF33DN17 MHV-68 mutant. The original ATG was mutated to ACG, and an Acl I site and a new ATG codon were inserted between nucleotides 51 and 52 of ORF33 to create ORF33DN17 MHV-68 mutant. The introduced Acl I site is boxed, and the new ATG codon is underlined. (B) Predicted restriction fragment of WT, ORF33DN17, and ORF33DN17.R BAC by Acl I digestion. (C) Enzymatic digestion of WT BAC, ORF33DN17 BAC, and ORF33DN17.R BAC by Acl I or BamH I. WT BAC DNA, ORF33DN17 BAC DNA, and ORF33DN17.R BAC DNA were isolated from E. coli strain SHG68 (GS1783). The DNAs were digested with Acl I or BamH I as indicated on the top. The asterisks mark the positions corresponding to the predicted 2.3-kb fragment of Acl I-digested viral DNAs.

To investigate the effect of deleting the N-terminal 17 aa on ORF33’s function and viral replication, we individually transfected 293T cells with WT BAC, ORF33DN17 BAC, or ORF33DN17 BAC plus pHA33 plasmid and observed the transfectants daily up to 72 h. WT or ORF33DN17 BAC plus pHA33 samples showed severe CPE, however, no CPE was detected in ORF33DN17 sample (Fig. 8A). For quantitation, supernatants from BAC transfected 293T cells at 72 hpt were collected, and virus titers were determined by plaque assays. Virus titer of WT sample reached 5.7 × 10⁵ PFU/mL, however, no progeny virion was detected in the ORF33DN17 sample. Although progeny mutant virions can be produced from the complementation sample (i.e., ORF33DN17 BAC plus pHA33), the genome within the mutant virions can only dictate the synthesis of truncated ORF33 (i.e., ORF33DN17) in the next round of de novo infection. Therefore, the mutant virions cannot complete lytic replication and no plaque was formed (Fig. 8B). We also conducted indirect-immunofluorescence assay on BAC transfected 293T cells and examined the expression of capsid protein ORF65 as an indicator of viral lytic replication. Consistent with results from the plaque assays, we observed agglomerates of fluorescence signals in WT or ORF33DN17.R sample at 48 hpt, indicating that progeny virions were produced in the original transfectants to infect nearby cells and express viral lytic genes in new cells. In contrast, fluorescent signals were only observed in single cell pattern in ORF33DN17 sample at early (24 h) and late (48 h) time points, indicating a blockage in viral lytic replication (Fig. 8C). Western blot analyses of the cell lysates at 29 hpt showed that in ORF33DN17 sample, the size of mutant ORF33 ran smaller than that of WT ORF33, as expected, but the sizes of all other viral proteins examined remained the same (Fig. 8D). In ORF33DN17 BAC transfected 293T cells, a large proportion of ORF33DN17 showed cytoplasmic localization (~40%), in contrast to wild-type ORF33, which predominantly localized in nucleus (~92%), confirming that the interaction between ORF33 and ORF45 plays a crucial role in ORF33 localization during viral lytic replication (Fig. 8E). Taken together, these data demonstrated that the N-terminal 17 aa of ORF33 are essential for mediating its appropriate localization and infectious virus production.

FIG 8.

The N-terminal 17 aa of ORF33 are essential for infectious virus production. (A) 293T cells were transfected with WT BAC, ORF33DN17 BAC, or ORF33DN17 BAC plus pHA33, and at 72 hpt cell morphology was visualized under a microscope. (B) Supernatants from 293T cells transfected with WT BAC or ORF33DN17 BAC were collected at 3 days posttransfection and passaged onto fresh cells. After two rounds of passages, the supernatants were collected, and virus titers were determined by plaque assays in BHK-21 cells in duplicates. (C) 293T cells were transfected with WT BAC, ORF33DN17 BAC, or ORF33DN17.R BAC, and indirect immunofluorescence was performed at 24 or 48 hpt. Samples were observed with a fluorescence microscope (Nikon-EclipseTi). ORF65 was detected using an anti-ORF65 antibody followed by an Alexa Fluor 488-conjugated secondary antibody (green channel). DAPI staining was shown in blue channel. (D) 293T cells were transfected with WT or ORF33DN17 BAC. At 29 hpt, cells were collected and subjected to Western blotting. Viral proteins were detected with specific antibodies as indicated to the right. (E) 293T cells were transfected with WT BAC or ORF33DN17 BAC respectively. At 32 hpt, cells were fixed and stained with anti-ORF33 (red channel) and anti-ORF45 (green channel) antibodies. Nuclei were stained with DAPI (blue channel). Samples were observed with a confocal microscopy.

The N-terminal 17 aa of ORF33 are essential for virion maturation in the cytoplasm.

We next investigated the effect of deleting the N-terminal 17 aa of ORF33 on virion morphogenesis. We analyzed the ultrastructural phenotype of WT BAC, ORF33DN17 BAC, or ORF33DN17.R BAC-transfected 293T cells by TEM. We checked for newly assembled viral particles in each stage of virus maturation. In the nuclei, viral capsids were observed in WT, ORF33DN17 and ORF33DN17.R BAC-transfected cells (Fig. 9, A1, B1 and C1), indicating that N-terminal 17 aa of ORF33 are not required for capsid assembly. Furthermore, capsids in ORF33DN17 BAC-transfected cells were capable of budding through the nuclear membrane into the cytoplasm, like those in WT and ORF33DN17.R BAC-transfected cells, as we observed primary enveloped virion in the perinuclear space (Fig. 9, A2, B2 and C2). In the cytoplasm of ORF33DN17 BAC-transfected cells, a large number of immature viral particles gathered around the tegument deposit area, however, no mature virion was formed in this area nor in cytoplasmic vesicles (Fig. 9, B3). Moreover, no viruses were found to be associated with the cell surface (Fig. 9, B4). In contrast, in WT or ORF33DN17.R BAC-transfected cells, fully enveloped virions inside cytoplasmic vesicles (Fig. 9, A3 and C3) and extracellular mature virions associated with or adjacent to the cellular plasma membrane (Fig. 9, A4 and C4) were observed. These results indicated that the N-terminal 17 aa of ORF33 are essential for virion maturation in the cytoplasm.

FIG 9.

The N-terminal 17 aa of ORF33 are essential for virion maturation in the cytoplasm. 293T cells were transfected with WT BAC, ORF33DN17 BAC, or ORF33DN17.R BAC, respectively. Approximately 70-nm thin sections were prepared at 48 hpt and examined by a 120-kV transmission electron microscope (FEI, Tecnai Spirit). All three types of capsids were found in the nuclei (Nu) of WT BAC, ORF33DN17 BAC, and ORF33DN17.R BAC-transfected cells (A1, B1, C1). Primarily enveloped virions were found in the perinuclear space (A2, B2, C2). In the cytoplasm (Cyto), viral particles, including mature virions, were observed in WT and ORF33DN17.R BAC-transfectants (A3, C3, black arrows). A large number of capsids, but no mature virions, gathered at the tegument deposit in ORF33DN17 BAC-transfectant (B3, black arrowheads). In the extracellular space, WT and ORF33DN17.R BAC-transfectants contained mature virions associated with plasma membranes (A4, C4, black arrows), whereas no mature virion was found in ORF33DN17 BAC-transfectant (B4).

To quantify the defect in virion morphogenesis that results from deleting the N-terminal 17 aa of ORF33, the distribution of virus particles (unenveloped capsids and mature virions) in different cellular compartments (nucleus, cytoplasm, and cell surface) were classified and enumerated for each sample. As shown in Table 2, WT, ORF33DN17 and ORF33DN17.R BAC-transfectants had similar proportions of capsids to total particles in the nuclei (42.43%, 51.24% and 54.03%, respectively). In the cytoplasm, WT and ORF33DN17 BAC transfectants also had similar proportions of unenveloped capsids (34.16% and 46.93%). Differences, however, were noted in the proportions of enveloped mature virions in the cytoplasm: in WT and ORF33DN17.R BAC-transfectants, the proportions were 16.61% and 17.20%, respectively, whereas in ORF33DN17 BAC-transfectant, it was zero. Furthermore, extracellular virions were not found in ORF33DN17 sample, compared to WT (5.87%) and ORF33DN17.R (2.20%) samples. These observations and statistical results are similar to those from the ORF45DC23 BAC-transfectant sample and further demonstrated that the interaction between ORF45 and ORF33 plays a key role in virion maturation in the cytoplasm.

TABLE 2.

Distribution of viral particles

BAC	No. of particlesa	Value (%) for:
		Nucleusb	Perinuclear spacec	Cytoplasm		Cell surfacef
				Unenveloped capsidsd	Enveloped virionse
WT	1294	549 (42.43)	12 (0.93)	442 (34.16)	215 (16.61)	76 (5.87)
ORF33 DN17	1255	643 (51.24)	23 (1.83)	589 (46.93)	0 (0)****g	0 (0)****g
ORF33 DN17.R	1227	663 (54.03)	20 (1.63)	306 (24.94)	211 (17.20)	27 (2.20)*g

^aTotal viral and subviral particles enumerated in 14 selected cells that contained viral particles.

^bValues are calculated as number of nucleocapsids/total number of viral particles.

^cValues are calculated as number of perinucleus capsids/total number of viral particles.

^dValues are calculated as number of immature virions (including capsids) in the cytoplasm/total number of viral particles.

^eValues are calculated as number of mature virions in the cytoplasm/total number of viral particles.

^fValues are calculated as number of fully mature extracellular virions/total number of viral particles.

^g*, P < 0.05; ****, P < 0.0001 versus WT, unpaired t test.

DISCUSSION

Herpesvirus morphogenesis follows a primary envelopment - de-envelopment – tegumentation - secondary envelopment process (22). A herpesvirion contains dozens of tegument proteins, and the interactions among various tegument proteins are key to the tegumentation process as well as final envelopment in the cytoplasm (23–26). However, the underlying mechanisms remain largely enigmatic. In this study, we investigated in detail the functional role of the interaction between two tegument proteins, ORF33 and ORF45, of a gammaherpesvirus. We confirmed that these two proteins interact with each other in both transfected and MHV-68 infected cells (Fig. 1). We further mapped the critical interaction domains to the C-terminal 23 aa of ORF45 and N-terminal 17 aa of ORF33 (Fig. 2 and Fig. 6). By deleting the sequences encoding the C-terminal 23 aa of ORF45 or N-terminal 17 aa of ORF33 from MHV-68 genome, we found that both the C-terminal 23 aa of ORF45 and the N-terminal 17 aa of ORF33 are essential for virion tegumentation in the cytoplasm and therefore production of infectious virions (Fig. 3 to 5, Fig. 7 to 9, and Table 1–2). Collectively, our results demonstrated that the interaction between ORF33 and ORF45 is essential for virion morphogenesis in the cytoplasm.

ORF33 is a tegument protein conserved among all herpesvirus subfamilies. Although its interaction with ORF45 has been demonstrated for multiple gammaherpesviruses (18, 27), differences have also been noted. MHV-68 ORF33 is colocalized with ORF45 in both the nucleus and the cytoplasm during virus late infection (Fig. 1D). When expressed individually, MHV-68 ORF33 was distributed in both the nucleus and the cytoplasm, whereas ORF45 was localized mostly in the nucleus. When coexpressed with ORF45, ORF33 was distributed mostly in the nucleus and colocalized well with ORF45 (Fig. 1B). Moreover, cytoplasmic ORF33 translocated to the nucleus in the presence of ORF45, but not ORF45DC23, during plasmid transfection (Fig. 2C) or BAC transfection (Fig. 4E). In contrast, the cytoplasmic portion of mutant ORF33DN17 did not show evident translocation even in the presence of WT ORF45 (Fig. 6C, Fig. 6D, and Fig. 8E). These results demonstrated that the interaction between ORF33 and ORF45 is responsible for the translocation of cytoplasmic ORF33 to the nucleus in MHV-68. In EBV, when expressed alone, the ORF33 homologue BGLF2 was mostly located in the cytoplasm whereas the ORF45 homologue BKRF4 was predominantly localized in the nucleus in a punctate pattern and in the perinuclear region (27). However, when coexpressed with BKRF4, BGLF2 translocated to the nucleus and colocalized with BKRF4 in both the nucleus and perinuclear region. This phenomenon was to some extent similar to our observation in this study. In contrast, in KSHV, ORF33 and ORF45 were reported to distribute differently. When expressed individually, ORF33 was mostly localized in the nucleus, whereas ORF45 was observed in both the nucleus and the cytoplasm. When coexpressed, a portion of the nuclear ORF33 translocated to the cytoplasm (18). Although the functional consequences of ORF33 trafficking remain to be revealed, for all these gammaherpesviruses, subcellular distribution of ORF33 is altered in the presence of ORF45 through their interaction.

Another difference is the stability of ORF33 in relation to ORF45. In MHV-68, the expression level of ORF33 was not affected by the presence or absence of ORF45 or their interaction (Fig. 2B, Fig. 4E). However, in KSHV, ORF33 is stabilized by ORF45. USP7, a ubiquitin-specific cellular protease, is required for ORF45-induced stabilization of ORF33. ORF45 interacts with ORF33 via its C-terminal 19aa (aa 383 to 407) and recruits USP7 via its central region (aa 219–229), thus bringing ORF33 and USP7 into close proximity and resulting in decreased ORF33 ubiquitination and increased ORF33 stability (18). It is worth noting that KSHV ORF45 is almost double the size of its homologues in other gammaherpesviruses, and the USP7-binding motif is not present in MHV-68 ORF45, consistent with our observation that MHV-68 ORF45 has no effect on the stability of ORF33.

We previously revealed that both ORF33 and ORF45 are associated with the capsids in the nucleus as well as in the cytoplasm. ORF33 and ORF45 nuclear localizations clearly play critical roles in nuclear egress, as ORF33-null or ORF45-null virus manifested severe defect in nuclear egress (10, 15). However, since TEM pictures are individual static snapshots, it is unclear whether ORF33 and ORF45 associated with the capsids in the nucleus are retained throughout the subsequent nuclear egress and tegumentation process. It is very likely that more ORF33 and ORF45 are recruited to the capsids during the tegumentation process in the cytoplasm, and along this line, the interaction between ORF33 and ORF45 plays an essential role in virion maturation in the cytoplasm, as demonstrated by this study.

In KSHV, deleting the C-terminal 19aa of ORF45 in the context of KSHV genome abolished virion production to an extent comparable to that with the ORF45-null viral mutant (18). However, because deleting the C-terminal 19aa of ORF45 also led to a drastic reduction of ORF33 protein level, it is difficult to conclude whether the defect in KSHV virus production can be solely attributed to ORF33-ORF45 interaction, as ORF33 protein also plays a critical role in both the early and late stage of virus replication (28, 29). Upon de novo infection, ORF33 is delivered into infected cell as virion component and enhances recruitment of host protein phosphatase PPM1G to decrease the phosphorylation levels of STING and MAVS, leading to suppressed innate immunity and improved virus replication (29). During virus late infection, newly synthesized ORF33 is assembled into virions via multiple protein-protein interactions (5, 17, 18, 28). In addition, the relatively low replication efficiency of KSHV makes it difficult to conduct ultrastructural studies of virion morphogenesis. In this study, by deleting the sequences encoding the C-terminal 23 aa of ORF45 or N-terminal 17 aa of ORF33 from the MHV-68 genome, we unequivocally demonstrated the essential role of ORF33-ORF45 interaction in virus production (Fig. 5, Fig. 9, and Tables 1 and 2). Mechanistically, our TEM data showed that disrupting ORF33-ORF45 interaction caused deficiency in virion tegumentation in the cytoplasm (Fig. 5, B3, and Fig. 9, B3). Collectively, our results demonstrate that the interaction between ORF33 and ORF45 plays an essential role in MHV-68 virion tegumentation and maturation in the cytoplasm. A recent report showed that mutating the key residues (W403A or W405A) in the C-terminal 19aa of KSHV ORF45 or using a competing cell penetration peptide TAT-C19, dramatically reduced virus production (30). These two residues are conserved in MHV-68 ORF45 and EBV BKRF4, suggesting that targeting ORF33-ORF45 interaction may be a general strategy to combat gammaherpesvirus infections.

Our previous immunogold labeling TEM results showed ORF33 is enriched in cytoplasmic electron-dense areas as well as within viral particles in the cytoplasm (ref. 10). ORF45 is also localized both in this electron-dense proteinaceous structure and within the tegument of mature virion (15). In WT BAC transfected cells, we observed an electron-dense proteinaceous structure containing virions inside, termed “tegument deposit” (Fig. 5, A3). Complicated protein-protein interactions taking place in this compartment are thought to contribute to the budding of partially-tegumented or tegumented viral particles into host-derived vesicles for secondary envelopment. Further studies are required to dissect the detailed mechanisms involved in virion assembly.

MATERIALS AND METHODS

Cells and viruses.

293T cells or BHK-21 cells were cultured in complete Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum, 1% penicillin and streptomycin. WT MHV-68 was originally obtained from Ren Sun (University of California, Los Angeles). The working stock was generated by infecting BHK-21 cells at a multiplicity of infection (MOI) of 0.03 PFU per cell. To infect BHK-21 cells, viral inoculum in DMEM was incubated with cells for 1 h with occasional swirling. The inoculum was then removed and replaced with fresh DMEM plus 10% fetal bovine serum. Virus titer was measured by plaque assay. Briefly, monolayer of BHK-21 cells were infected with virus for 1 h and overlaid with 1% methylcellulose for 5 days. Plaques were then counted to determine the virus titer.

Plasmid construction.

Construction of plasmid pHA33 was described previously (9). Plasmids encoding truncated ORF33 proteins were generated similarly by individual cloning of corresponding PCR fragments into the EcoR I and Xho I sites of pCMV-HA vector. Plasmid pFlag 45 was generated by cloning a PCR fragment of the ORF45 coding sequence into pcDNA3.1flag vector. Plasmids encoding truncated ORF45 proteins were generated similarly by individual cloning of corresponding PCR fragments into the BamH I and HindIII sites of pcDNA3.1flag vector. The sequences of primers used for plasmid construction are available upon request.

Construction of recombinant MHV-68 BAC plasmids.

Recombinant MHV-68 BAC plasmids were generated by allelic exchange in E. coli, using a two-step Red-mediated recombination protocol (31). The construction of ORF45DC23 BAC was accomplished using the GS1783 strain of E. coli, which encodes inducible Red and I-SceI activities. The I-SceI-aphAI cassette from recombinant plasmid pEPkan-S was PCR-amplified with primers containing triple stop codons and a Bgl II site. The 1263-bp PCR product was purified by agarose gel electrophoresis and subsequently introduced by electroporation (0.1 cm cuvette, 1.8 kV, 200 Ω, 25 μF) into the GS1783 E. coli strain harboring MHV-68 BAC. The bacteria were then plated on LB agar plates containing both chloramphenicol and kanamycin. Those in which successful integration of the PCR product into the BAC DNA had occurred were detected by colony PCR. The integrated cassette was then cleaved upon treatment with 1% l-arabinose, and a second recombination event between the duplicated sequences resulted in the loss of Kan^r/I-SceI cassette and recircularization of the MHV-68 BAC DNA, yielding kanamycin-sensitive colonies that were screened by replica plating. The incorporation of stop codons and a Bgl II site in ORF45 were identified by restriction enzyme digestion of the purified DNA. A revertant of ORF45DC23 (ORF45DC23.R) was generated by deleting the stop codons and the Bgl II site using the same method. To construct the ORF33DN17 BAC, the original translation initiation codon ATG in ORF33 was mutated to ACG, and an Acl I site and a new ATG codon were inserted between nucleotides 51 and 52 in ORF33 sequence. A revertant of ORF33DN17 (ORF33DN17.R) was generated by mutating ACG to ATG and deleting the Acl I site and the inserted ATG codon using the same method.

Antibodies, immunoprecipitation, and immunoblotting.

His-ORF45 was expressed in bacteria, purified and provided to the Experimental Animal Center, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, to generate rabbit polyclonal antibody against MHV-68 ORF45. Rabbit polyclonal anti-ORF26 and anti-ORF65 were kind gifts from Ren Sun (University of California, Los Angeles). 293T cells seeded onto 6-cm plates were transfected with 6 μg of total DNA. Forty-eight hours posttransfection, cells were washed once with ice-cold phosphate-buffered saline (PBS) and then solubilized in lysis buffer (50 mM Tris-Cl (pH 7.4), 150 mM NaCl, 1% TritonX-100, 1 mM EDTA) containing protease inhibitors cocktail and phenylmethanesulfonyl fluoride (PMSF). Lysates were cleared by centrifugation for 15 min at 13,000 rpm. 10 percent of the supernatant was used as an input control. Soluble proteins were mixed with 12 μL anti-FLAG M2 agarose (Sigma) and rotated at 4°C for 4 h. Beads were washed three times with lysis buffer before use. Immune complexes were washed five times with lysis buffer and supernatant was depleted. Bound proteins were boiled in SDS sample buffer for 10 min. The protein samples were separated on a 12% SDS-polyacrylamide gel, and proteins were transferred onto a polyvinylidene difluoride membrane (Millipore). The membrane was blocked for 1 h at room temperature in PBS containing 1‰ Tween (PBST) supplemented with 5% nonfat milk powder, incubated for 2 h at room temperature with anti-ORF45 polyclonal antibody or anti-ORF33 monoclonal antibody, washed, and reacted with goat anti-mouse or goat anti-rabbit horseradish peroxidase (HRP)-conjugated IgG secondary antibody. Immunoreactive bands were visualized using Enhanced Chemiluminescence System (Millipore).

Fluorescence microscopy.

BHK-21 cells were seeded onto glass coverslips in 24-well plates. Cells were infected with MHV-68 at an MOI of 3, or individually transfected with WT BAC, ORF45DC23 BAC, ORF45DC23.R BAC, ORF33DN17 BAC, or ORF33DN17.R BAC. At indicated time points, the cells were washed with PBS and fixed with 4% paraformaldehyde for 15 min at room temperature. After another wash, samples were permeabilized with 0.2% TritonX-100 dissolved in PBS for 7 min. The cells were washed twice with PBS and incubated in 5% normal goat serum dissolved in PBS for 30 min. Cells were probed with mouse anti-ORF33 antibody, rabbit anti-ORF65 or anti-ORF45 antibody as indicated for 1 h at room temperature and a conjugated secondary antibody (anti-mouse Alexa Fluor 555 or anti-rabbit Alexa Fluor 488, Invitrogen). Cells were treated with 4’,6-diamidino-2-phenylindole (DAPI) to stain for nuclei. Finally, the coverslips with cells were mounted onto slides and sealed with Fluoromount reagent (Sigma). Images were obtained by using a confocal microscope (Olympus FV1200) or three-dimensional Structured Illumination Microscope (Delta Vision OMX).

Transmission electron microscopy (TEM).

293T cells were transfected with WT BAC, ORF45DC23 BAC, ORF45DC23.R BAC, ORF33DN17 BAC or ORF33DN17.R BAC, respectively. Samples were fixed in 2.5% glutaraldehyde/PBS at 4°C for 12 h, then fixed in 1% OsO₄, dehydrated, and embedded in Epon. Approximately 70-nm thin sections were stained with 2% uranyl acetate and 0.3% lead citrate. The sections were examined with a 120-kV electron microscope (Tecnai Spirit) with CCD camera image acquisition. For quantification, the location and type of all virus particles in 14 cells from each sample were enumerated. Data represent the sum of particles in each category from all 14 cells and the percentage of total for each category.