Structural and Functional Studies of the Abundant Tegument Protein ORF52 from Murine Gammaherpesvirus 68*

The tegument is a layer of proteins between the nucleocapsid and the envelope of herpesviruses. The functions of most tegument proteins are still poorly understood. In murine gammaherpesvirus 68, ORF52 is an abundant tegument protein of 135 residues that is required for the assembly and release of infectious virus particles. To help understand the molecular basis for the function of this protein, we have determined its crystal structure at 2.1 Å resolution. The structure reveals a dimeric association of this protein. Interestingly, an N-terminal α-helix that assumes different conformation in the two monomers of the dimer mediates the formation of an asymmetrical tetramer and contains many highly conserved residues. Structural and sequence analyses suggest that this helix is more likely involved in interactions with other components of the tegument or nucleocapsid of the virus and that ORF52 functions as a symmetrical dimer. The asymmetrical tetramer of ORF52 may be a “latent” form of the protein, when it is not involved in virion assembly. The self-association of ORF52 has been confirmed by co-immunoprecipitation and fluorescence resonance energy transfer experiments. Deletion of the N-terminal α-helix, as well as mutation of the conserved Arg95 residue, abolished the function of ORF52. The results of the functional studies are fully consistent with the structural observations and indicate that the N-terminal α-helix is a crucial site of interaction for ORF52.

Herpesviruses are large, enveloped viruses that carry a double-stranded DNA genome of ϳ110 -230 kb. Three subfamilies belong to the Herpesviridae: alpha-, beta-, and gammaherpesviruses. A unique feature of these viruses is the presence of a tegument layer between the nucleocapsid and the glycoprotein envelope. The tegument is composed predominantly of proteins. Tegument proteins of alpha-, beta-, and gammaherpesviruses have been found to be involved in at least three essential functions in viral replication: (i) the assembly and egress of virions (1-3); (ii) the entry of virions into naive cells, including the translocation of nucleocapsids to the nucleus; and (iii) other effects during the immediate-early phase of infection, including the transactivation of viral immediate-early genes and the possible modulation of host cell gene expression, innate immune mechanisms, and signal transduction (3)(4)(5)(6)(7). Recently, proteins in the tegument of various herpesviruses have been identified with the development of mass spectrometry technology (8 -13). However, the functions of many tegument proteins have not been defined.
Gammaherpesviruses can infect and establish latency in lymphocytes and cause malignancies, especially in immunocompromised patients (14). Murine gammaherpesvirus 68 (MHV-68) 4 is a natural pathogen of wild rodents (15)(16)(17). It is related to human Kaposi's sarcoma-associated virus and Epstein-Barr virus (18 -20), and its infection in mice has been developed as an animal model for investigating the biology and pathogenesis of gammaherpesviruses in humans (21). The functions of many MHV-68 proteins can be studied in the context of viral infection by mutagenesis, taking advantage of the herpesvirus genome cloned as bacterial artificial chromosomes (22)(23)(24)(25)(26). Although the major proteins in tegument of the gammaherpesviruses have been reported, little is known about the structure and function of these tegument proteins (8,12).
MHV-68 ORF52 encodes a capsid-associated tegument protein that is present in abundance in virions (8). It is a highly expressed gene with true late kinetics, activated after viral DNA replication (27,28). This protein is well conserved among the gammaherpesviruses (Fig. 1), but a homolog has not been found in the alpha-and betaherpesviruses. A number of tegument proteins including ORF52 are unique to each subfamily of herpesvirus and may have distinct roles in the formation and egress of virions (1,29,30). Mutation of ORF52 in the MHV-68/BAC genome leads to arrest at the lytic phase of infection after viral genome replication, late gene expression, viral DNA cleavage/ packaging, and nucleocapsid assembly in the nucleus but prior to complete virion tegumentation and envelopment in the cytoplasm and egress of infectious virions from the cell (31). Partially tegumented capsids produced by the ORF52-null mutant contain conserved capsid proteins, the ORF64 and ORF67 tegument proteins, but virtually no ORF45 tegument protein. Enhanced green fluorescent protein fusions to ORF52 localize in the cytoplasm to a distinct compartment reminiscent of the secretory pathway (31). ORF52 is essential for the tegumentation and egress of infectious MHV-68 particles in the cytoplasm, suggesting an important function in gammaherpesvirus virion morphogenesis.
To understand the molecular mechanism for the biological functions of ORF52, we have determined the crystal structure of this protein at 2.1 Å resolution. The structure reveals a dimeric association of this protein, except that the N-terminal ␣-helix does not obey the symmetry of the dimer. This helix contains many highly conserved residues and is more likely involved in interactions with other components of the tegument or nucleocapsid of the virus. We have confirmed the selfassociation of ORF52 by co-immunoprecipitation experiments. Deletion of this N-terminal ␣-helix, as well as mutation of the conserved Arg 95 residue, abolished the function of ORF52.

MATERIALS AND METHODS
Expression and Purification of ORF52-The production of ORF52 protein was carried out using the high throughput protein production platform of the Northeast Structural Genomics Consortium (32). ORF52 corresponds to Northeast Structural Genomics Consortium target MhR28B (www.nesg.org). A truncated portion of the gene ORF52 from murine herpesvirus (68 strain WUMS) encoding residues 1-102 was cloned into a pET21 (Novagen) derivative, generating plasmid pMhR28B-21.3. The resulting recombinant protein contains eight nonnative residues (LEHHHHHH) at the C terminus. The construct was sequence-verified by standard DNA sequence analysis.
Escherichia coli BL21 (DE3) pMGK cells, a rare codon enhanced strain, were transformed with pMhR28B-21.3. A single isolate was cultured in MJ9 minimal medium (33) supplemented with selenomethionine, lysine, phenylalanine, threonine, isoleucine, leucine, and valine for the production of selenomethionine-labeled ORF52 (34). Initial growth was carried out at 37°C until the A 600 of the culture reached 0.6 -0.8 units. The incubation temperature was then decreased to 17°C, and protein expression was induced by the addition of isopropyl-␤-D-thiogalactopyranoside at a final concentration of 1 mM. Following overnight incubation, the cells were harvested by centrifugation.
Selenomethionyl ORF52 was purified using an AKTAxpress (GE Healthcare) based two-step protocol consisting of IMAC (HisTrap HP) and gel filtration (HiLoad 26/60 Superdex 75) chromatography. The purified ORF52 protein was concentrated to 9 mg/ml, flash-frozen in aliquots, and used for crystallization screening. Sample purity (Ͼ97%) and molecular mass (12.3 kDa) were verified by SDS-PAGE and matrix-assisted laser desorption-ionization time-of-flight mass spectrometry, respectively.
Crystallization of ORF52-Crystals of ORF52 were obtained by the hanging drop vapor diffusion method at 4°C. 1 l of protein solution containing ORF52 in 10 mM Tris, pH 7.5, 100 mM NaCl, and 5 mM dithiothreitol was mixed with 1 l of the reservoir solution consisting of 80% (v/v) polyethylene glycol 400, 100 mM MOPS, pH 7.0, and 100 mM NaNO 3 . The crystals were cryo-protected using paratone-n and flash-frozen in liquid propane for data collection at 100 K. They belong to space group P2, with cell parameters of a ϭ 54.9 Å, b ϭ 49.2 Å, c ϭ 88.9 Å, and ␤ ϭ 105.6°. There are two dimers of ORF52 in the asymmetric unit.
Crystals grown in the same crystallization condition but using the mother liquor as cryo-protectant belong to space group C2, with cell parameters of a ϭ 93.5 Å, b ϭ 49.4 Å, c ϭ 56.4 Å, and ␤ ϭ 111.9°. There is one dimer of ORF52 in the crystallographic asymmetric unit.
Structure Determination-A multiple-wavelength anomalous diffraction data set to 2.7 Å resolution was collected at the peak, edge, and remote absorption wavelengths of selenium at the X4A beamline of the National Synchrotron Light Source. The diffraction images were processed with the HKL package (35), and the selenium sites were located with the program BnP (36). The programs SOLVE and RESOLVE (37) were used for phasing the reflections and automated model building, which correctly placed 51% of the residues in the two dimers. An almost complete model of each monomer, except residues 1-6 and 32-38, was built with the program COOT (38) and refined with CNS (39).
A second data set was collected to 2.1 Å and belonged to C2 space group. The program COMO (40) was used to solve this structure with a dimer from the P2 crystal as the model. The crystallographic information on this model is summarized in Table 1.
Co-immunoprecipitation Experiments-HA-tagged ORF52 was expressed using the pCMV-HA vector (Clontech). Fulllength ORF52 and an N-terminal deletion mutant lacking the first 33 residues (⌬N33) were amplified from wild-type MHV-68/BAC by PCR. The R95A single-site mutant of ORF52 was created using an oligonucleotide-directed twostep PCR mutagenesis method. The PCR fragments were cloned into the EcoRI and KpnI sites of pCMV-HA. The clones were verified by sequencing.
293T cells were seeded onto a 6-cm plate (0.8 ϫ 10 5 /plate) 24 h before transfection. 4 g of HA-tagged expression plasmids for ORF52 or its mutants were individually co-transfected with 4 g of pFLAG-ORF52 (31) by the calcium phosphate method. 48 h after transfection, the cells were washed once with ice-cold phosphate-buffered saline, and then solubilized in EBC buffer (50 mM Tris-Cl, pH 7.4, 150 mM NaCl, 1% Nonidet P-40, 1 mM EDTA) with protease inhibitors. The lysates were clarified by centrifugation (13,000 rpm, twice for 15 min). Five percent of the supernatant was used as an input control; the rest was incubated overnight at 4°C with anti-FLAG M2 agarose beads (Sigma) that were washed five times with EBC buffer before use. Immune complexes were washed five times in NETN buffer (20 mM Tris-Cl, pH 8.0, 1 mM EDTA, 0.5% Nonidet P-40, 100 mM NaCl, or 1 M NaCl), and supernatant was depleted. Bound proteins were recovered by boiling in SDS sample buffer for 10 min. Immunoprecipitated proteins were analyzed with primary antibodies specific for HA tag and extracted proteins with antibodies specific for FLAG tag using Western blotting.
Quantitative Real Time PCR-293T cells were seeded onto 24-well plates the day before transfection. 600 ng of BAC (wildtype or 52-null) and 200 ng of plasmid expressing ORF52 or its mutants were co-transfected into each well. After 4 days, the supernatant was collected, and the viral genome was extracted. 200 l of supernatant was incubated with 100 g/ml proteinase K and 20 g/ml RNase A at 37°C for 15 min, and the reaction was stopped by adding 1/10 volume of 0.5 M EDTA, pH 8.0, followed by incubation at 70°C for 10 min. An equal volume of 2ϫ lysis buffer (200 mM NaCl, 20 mM Tris-Cl, pH 8.0, 50 mM

RESULTS AND DISCUSSION
Structure Determination-The crystal structure of MHV-68 ORF52 was determined at 2.7 Å resolution by the selenomethionyl multiple-wavelength anomalous diffraction method (41) and refined at 2.1 Å resolution using a second crystal form ( Table 1). The structures in the two crystal forms are essentially the same, with root mean square distance of 0.5 Å among their equivalent C␣ atoms. The structure at 2.1 Å resolution will be used in further descriptions here. The atomic model has low R values and excellent agreement with expected geometric parameters ( Table 1). None of the residues of the protein is located in the disallowed region, whereas 97% of the residues are located in the most favored region of the Ramachandran plot (data not shown).
The current atomic model contains residues 6 -31 and 41-102 for monomer A and residues 6 -102 for monomer B of ORF52 in the crystallographic asymmetric unit. The expression construct covers residues 1-102 of the protein. Additional residues at the C terminus (103-135) are highly hydrophilic in nature and are unlikely to be ordered on their own. A polyethylene glycol 400 (3,6,9,12,15,18,21,24octaoxahexacosan-1-ol) molecule is bound to each monomer, although they are not related by the 2-fold symmetry of the dimer.
Structure of ORF52 Monomer and Dimer-The structure of ORF52 monomer consists of three ␣-helices (␣1-␣3) followed by one ␤strand (␤1) near the C terminus ( Fig. 2A). The ORF52 dimer exhibits 2-fold symmetry (except for the N-terminal helix ␣1; see below) and contains an extensive hydrophobic core between an anti-parallel ␤-sheet formed by the two ␤1 strands and the ␣2 and ␣3 helices of the two monomers (Fig. 2, A and B). Approximately 1,900 Å 2 of the surface area of each monomer is buried in the dimer, and residues in this interface are highly conserved among ORF52 homologs (Fig. 1).
Helix ␣1 at the N terminus does not obey the 2-fold symmetry of the dimer. The helix in monomer A is in contact with the rest of the dimer, but the helix in monomer B is splayed away (Fig. 2A). The loop between ␣1 and ␣2 is disordered in monomer A. This segment is weakly conserved among ORF52 homologs (Fig. 1). Excluding the ␣1 helix, the structures of the two monomers are essentially the same, with root mean square distance of 0.5 Å among their equivalent C␣ positions.
Structure of ORF52 Tetramer-An asymmetric tetramer of ORF52 is formed from the dimer by the crystallographic 2-fold axis, and the ␣1 helices of the two dimers are intertwined in this tetramer interface (Fig. 3A). The ␣1 helix that is splayed away from the dimer interacts with the ␣1 helix that is in contact with the other dimer. The two ␣1 helices in the tetramer interface are arranged in a parallel fashion and are related by a local 2-fold symmetry operation along their length (Fig. 3B). This helix/helix interaction provides most of the 2,500 Å 2 of surface area burial in the tetramer interface. The structural information therefore suggests that this asymmetric tetramer of ORF52 may be stable. Our static light scattering studies showed that ORF52 (residues 1-108) is actually a tetramer in solution. 5 Several hydrophobic residues are buried at the interface between the ␣1 helices of the four monomers, including the highly conserved residues Tyr 10 , Met 13 , Val 17 , Leu 20 , and Leu 27 (Fig. 1). In addition, Glu 23 in one monomer is ionpaired with Lys 28 and hydrogen-bonded to Asn 24 of the other monomer. These residues are also highly conserved among the ORF52 homologs (Fig. 1). Besides interactions between the ␣1 helices, several residues from ␣2 are also located in the tetramer interface, although most of them are hydrophilic in nature (Fig. 1).
Functional Implications-Our structural data suggest two possible models for the active form of ORF52. In one model, ORF52 functions as a dimer (monomers are unlikely to be stable structurally because of the lack of a hydrophobic core). With such a dimer (Fig. 2C), the ␣2 and ␣3 helices and the ␤1 strand form a scaffold, and the ␣1 helix could be projected away from this scaffold to interact with other tegument and/or capsid proteins. The ␣1 helix is connected to the rest of the protein through a long, flexible loop (residues 30 -45), which is poorly conserved among ORF52 homologs (Fig. 1). The conformation of ␣1 observed here is stabilized by the tetramer and may not be the same in the isolated dimer. This model is supported by the fact that four of the five strictly conserved residues in ORF52 (Leu 20 , Glu 23 , Asn 24 , and Leu 27 ) are located in the ␣1 helix (Fig. 1). Moreover, several other positions in this helix are conserved to be hydrophobic residues (Met 13 , Val 14 , Val 17 , and Leu 22 ) (Fig.  1). Therefore, once the ␣1 helix moves away from the rest of the dimer, several conserved, highly hydrophobic patches will be exposed (Fig. 2D), which would be optimal for mediating interactions with its partners. The fifth conserved residue, Arg 95 , is located in strand ␤1 ( Fig. 2A). Its side chain is pointed away from the dimer (Fig. 2D). In the structure of ORF52, it forms a salt bridge with Asp 97 , although this residue is not conserved (Fig. 1).
In the second model, ORF52 functions in the tegument as the asymmetric tetramer observed here and interacts with other proteins in the virus. The Arg 95 residues from the four monomers are clustered together on one face (Fig. 3C), although there is a large, electronegative groove in the opposite face of this tetramer (Fig. 3D). A possible weakness of this tetramer model is that the only highly conserved surface patch for this tetramer corresponds to that for the Arg 95 residues (Fig. 3E). A larger epitope could be expected to mediate interactions with the other tegument and/or capsid proteins, and such a larger interface is observed in the dimer form of ORF52 (Fig. 2D).
Therefore, structural and sequence analyses suggest that the dimer is more likely to be the active form of ORF52, although there are not sufficient experimental data at the current time to distinguish between the two models. On the other hand, the tetramer observed here could be a "latent" form of this protein, i.e. the dominant structure in the absence of other protein binding partners. Once ORF52 comes into contact with the proper protein partner, the tetramer will dissociate into dimers, exposing the functional ␣1 helices. A consistent feature of both models is the functional importance of the ␣1 helix.
Evidence for ORF52 Self-association in Vivo-Our structural and static light scattering studies show that ORF52 can form dimers and tetramers in solution. We next sought experimental evidence for this self-association in vivo, using co-immunoprecipitation experiments. FLAG-tagged wild-type ORF52 and HA-tagged ORF52 and its mutants were co-expressed in 293T cells (Fig. 4A). FLAG-ORF52 was immobilized on anti-FLAG agarose, and an anti-HA antibody was used to detect whether the HA-ORF52 was co-immunoprecipitated. Our experiments clearly demonstrate the strong self-association of wild-type ORF52 (Fig. 4B). Moreover, mutation of the conserved Arg 95 residue (Fig. 4C) or deletion of the N-terminal 33 residues (Fig. 4D), corresponding to removal of helix ␣1 (Fig. 1), has no effect on this self-association. These observations are in complete agreement with our structural information that the dimeric form of the protein is independent of the ␣1 helix ( Fig. 2A). Although the tetrameric form of the protein should be disrupted if helix ␣1 is deleted (Fig. 3A), the co-immunoprecipitation data cannot distinguish between dimers and tetramers of ORF52.
We also employed fluorescence resonance energy transfer assay to confirm the self-interaction of ORF52. We constructed plasmids expressing fusion protein of ORF52 and cyan fluorescent protein (CFP) or yellow fluorescent protein (YFP). The plasmids pCFP/ORF52 and pYFP/ORF52 were either transfected together or alone into 293T cells. Each sample was excited at 415 nm (CFP excitation wavelength), and the fluorescence emission intensity was monitored continuously from 450 to 550 nm. We observed a 1.5-fold decrease of CFP/ORF52 emission intensity in the presence of YFP/ORF52 and a 2-fold increase of YFP/ORF52 emission in the presence of CFP/ ORF52 when compared with either CFP/ORF52 or YFP/ORF52 expression alone (data not shown). This result demonstrates a fluorescence energy transfer between CFP/ORF52 and YFP/ ORF52 when they are co-expressed, indicating self-interaction of ORF52.
The ␣1 Helix and Arg 95 Are Essential for the Function of ORF52-Our structural study suggests that the ␣1 helix may be essential for the function of ORF52, either mediating interactions with other tegument proteins or driving the formation of the tetramer. To obtain direct experimental evi- dence for the functional importance of this helix, we assessed the ability of a deletion mutant lacking these N-terminal residues in rescuing an ORF52-null mutant virus. An ORF52-null MHV-68/BAC mutant (31) was co-transfected into 293T cells with a wild-type or mutant ORF52 expression plasmid. Four days later, viral particles were collected from supernatants, and viral genomes were extracted and subjected to quantitative real time PCR analysis to examine the ability of each mutant ORF52 to complement the ORF52null MHV-68/BAC mutant in tegumentation and egress of mature viral particles. Although the wild-type ORF52 can readily rescue the mutant phenotype and lead to the production of a large amount of virus particles, the deletion mutant lacking helix ␣1 cannot rescue the ORF52-null virus (Fig. 5). In addition, the conserved Arg 95 residue is also essential for the function of the protein, because the R95A mutant cannot rescue the ORF52-null virus (Fig. 5).
In summary, we have determined the crystal structure of the abundant tegument protein ORF52 of MHV-68 at 2.1 Å resolution. Structural and sequence analyses suggest that an N-terminal ␣-helix may be essential for the functions of this protein, which has been confirmed by mutagenesis and functional studies in vivo. This helix may mediate the interactions between ORF52 and other proteins in the tegument and/or capsid of MHV-68. Efforts at identifying these protein partners of ORF52 are currently in progress.