INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Virions of alphaherpesviruses are structurally composed of membrane, tegument, nucleocapsid, and core (24). The tegument is an amorphous structure between the nucleocapsid and membrane. In addition to being essential virion structural components, tegument proteins are important for releasing viral genomic DNA early in infection, nucleocapsid formation, viral DNA packaging, and regulation of viral gene expression (24). However, the process of tegument assembly and the precise functions of most tegument proteins are still unclear.

Analysis of the genome of Bovine herpesvirus 1 (BHV-1), an alphaherpesvirus, reveals that there are at least 16 proteins known or presumed to be present in the tegument. A short open reading frame (ORF) in the BHV-1 genome designated the U_L3.5 gene encodes a 13-kDa tegument protein expressed late in infection (23). Unlike most alphaherpesviral proteins, U_L3.5 is not conserved throughout the alphaherpesvirus family. Homologs have been found only in Pseudorabies virus (PrV) (6), Varicella zoster virus (VZV) (5), Equine herpes virus 1 (EHV-1) (25), and Infectious laryngotracheitis virus (11). Herpes simplex virus type 1 (HSV-1) and HSV-2 do not have U_L3.5 homologs (18, 19). Moreover, homologs of U_L3.5 differ in size (from 71 amino acids [aa] for VZV to 220 aa for PrV) (5, 6, 13) and have overall 20 to 30% amino acid sequence homology that is restricted mostly to the N-terminal 50 aa (13). The roles of U_L3.5 homologs in virus replication are apparently different. PrV U_L3.5 is required for virus egress. A PrV mutant lacking U_L3.5 replicates very poorly in the one-step replication and plaque assays (10). On the other hand, VZV lacking gene ORF57, the homolog of the U_L3.5 gene, grows in cell culture at a same rate as wild-type virus (3). Neither the need for nor the function of U_L3.5 in BHV-1 replication has been determined. However, BHV-1 U_L3.5 rescued a PrV U_L3.5 deletion mutant (9) implying that BHV-1 U_L3.5 may also participate in virus egress.

BHV-1 -transinducing factor (BTIF), encoded by the U_L48 gene, is a virion component that transactivates immediate-early gene promoters during viral lytic infection (20). Homologs are present in HSV-1, EHV-1, and VZV (1, 5, 15, 22) and probably all other Alphaherpesviruses. In addition to being a transcription factor, the HSV homolog, TIF, also known as VP16 or Vwm65, is a tegument protein indispensable for virion assembly (27). No structural function of other TIF homologs has been reported.

One approach to understanding tegument assembly and the function of tegument proteins is to identify and characterize proteins that associate with previously characterized tegument proteins. In this study, we sought to identify proteins that interact with BHV-1 U_L3.5. Five methods showed a specific interaction between BHV-1 U_L3.5 and BTIF. (i) BTIF specifically attached to His-tagged BHV-1 U_L3.5 in an in vitro pull-down assay. (ii) A complex containing U_L3.5 and BTIF immunoprecipitated from BHV-1-infected cell lysates with either anti-U_L3.5 or anti-BTIF polyclonal antibodies. (iii) Confocal microscopy showed that BHV-1 U_L3.5 and BTIF colocalized in a Golgi-like subcellular compartment late in infection. (iv) In a transient expression system, U_L3.5 and BTIF colocalized in the nucleus. (v) Deletion of the N-terminal 20 aa, but not the C-terminal 40 aa, of U_L3.5 abolished the binding of BTIF in the pull-down assay and the colocalization in the transient expression system. The interaction between U_L3.5 and BTIF may have roles in virion assembly and regulation of BTIF transactivation activity.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Virus, cells, and media. BHV-1 (Cooper strain; ATCC VR-864) was replicated in Madin-Darby bovine kidney (MDBK; ATCC CCL22) cells in minimum essential medium (MEM) (Gibco Laboratories, Life Technologies, Inc.) supplemented with 5% fetal bovine serum (FBS) (Hyclone) at 35°C in a 5% CO₂ humidified atmosphere. Vero cells were cultured in MEM supplemented with 10% FBS at 35°C in a 5% CO₂ humidified atmosphere.

Plasmids. The U_L3.5 complete ORF was amplified by PCR with Pfu polymerase by using the N-terminal primer CGGGATCCGCCATGGCCCGCGTGCGCGCCG and the C-terminal primer GTGAATTCTTATTGGAACGTGCGGTAATTG (viral sequences underlined) from plasmid pSD72 (17). The PCR product was digested with BamHI and EcoRI and cloned into both the His-tagged fusion protein expression vector pRSET (Invitrogen) and the eukaryotic expression vector pcDNA3 (Invitrogen). The resulting plasmids were called pU_L3.5SET and pU_L3.5cDNA3, respectively.

Antibodies. Production of rabbit anti-U_L3.5 polyclonal antibody was previously described (23). To produce anti-BTIF, a bacterially expressed GST-BTIF fusion protein was produced as an antigen. The plasmid pGST-BTIF was transformed into E. coli BL21, and GST-BTIF was induced by isopropyl--D-thiogalactopyranoside (IPTG) at a final concentration of 0.1 mM for 7 h with gentle shaking at 26°C. The cells were suspended in phosphate-buffered saline (PBS) containing 0.25% Tween 20, 1 mM phenylmethylsulfonyl fluoride (PMSF), and 0.1 mM chymostatin and lysed by sonication. The lysate was centrifuged, and GST-BTIF was collected from the supernatant with glutathione-Sepharose 4B (Pharmacia) according to the manufacturer's instructions. The fusion protein was eluted in buffer (10 mM glutathione, 50 mM Tris-HCl [pH 8.0], 150 mM NaCl, 0.2% Triton X-100). The purified GST-BTIF was emulsified in Freund's complete adjuvant and injected intraperitoneally into BALB/c mice. Mice were boosted twice at 3-week intervals with GST-BTIF emulsified with Freund's incomplete adjuvant. Sera were sampled 2 weeks following each boost.

Radioimmunoprecipitation. Radiolabeled uninfected and BHV-1-infected MDBK cells were prepared as described by Marshall et al. (16). MDBK cells were infected at a multiplicity of infection (MOI) of 10 and labeled with [³H]leucine (ICN Pharmaceuticals Inc.) from 6 to 18 h after infection. The labeled cells were lysed in NET buffer (150 mM NaCl, 5 mM EDTA, 50 mM Tris [pH 8]) containing 1 mM PMSF and 0.5% Triton X-100. Immunoprecipitations were done with 10 µl of antiserum for each 10⁶ cells. Proteins were immunoprecipitated from in vitro translation reaction mixtures or from lysates of BHV-1-infected and uninfected MDBK cells on protein A-Sepharose (Sigma) coated with rabbit anti-U_L3.5 or mouse anti-BTIF polyclonal antibodies.

Purification of His-tagged fusion proteins and in vitro pull-down assay. Cells from a 2-ml overnight culture of E. coli BL21(DE3)pLysS containing the His-tagged fusion protein vector pU_L3.5SET or deletion mutants of this same plasmid were collected by centrifugation and resuspended with an equal volume of Luria broth (LB) and inoculated 1:100 into LB containing 100 µg of ampicillin per ml. The culture was incubated at 37°C with shaking until the optical density at 600 nm reached 1.0, IPTG was added to a final concentration of 0.5 mM, and the culture was incubated for an additional 2 h at 37°C with shaking. The cells were suspended with lysis buffer (50 mM Tris-HCl [pH 8.0], 500 mM NaCl, 1 mM imidazole, 1 mM PMSF) and lysed by sonication. A 50% slurry of Ni-nitrilotriacetic acid (NTA) agarose (Qiagen) was added to the supernatant of the bacterial lysate, and the mixture was incubated with gentle agitation at 4°C for 1 h. The Ni-NTA agarose pellet was washed two times with lysis buffer containing 20 mM imidazole and two times with lysis buffer containing 40 mM imidazole and was finally equilibrated in TN buffer (50 mM Tris-HCl [pH 8.0], 100 mM NaCl, 1 mM PMSF) before being applied to a pull-down assay. To prepare a control pRSET/Ni-NTA agarose, exactly the same protocol was used, except for starting with a culture of E. coli containing vector pRSET.

Mass spectrometry protein identification. To prepare the sample for protein identification, a preparative pull-down was performed using a lysate of 8 × 10⁷ unlabeled BHV-1-infected MDBK cells and about 100 µg of His-tagged U_L3.5. To monitor the assay, a parallel experiment using [³⁵S]methionine- and [³⁵S]cysteine-labeled virus-infected cell lysate was done. After SDS-PAGE, the gel was stained with 0.1% Coomassie blue and destained in 10% acetic acid-50% methanol-40% H₂O. A strip of the gel containing the radiolabeled sample was dried and autoradiographed to identify the band of interest. A prominent band at about 60 kDa containing about 1 µg of protein was excised, frozen at 80°C, and sent for matrix-assisted laser desorption ionization mass spectrometry (MALDI-MS) analysis. MALDI-MS was performed by the W. M. Keck Facility, Yale University.

In vitro transcription and translation of U_L3.5 and BTIF. In vitro transcriptions were carried out according to the AmpliScribe protocol (Epicentre Technologies). Briefly, 1 µg of pU_L3.5cDNA3 or pBTIFcDNA3 was linearized with XhoI and mixed with 2 µl of 10× T7 reaction buffer, 1.5 µl of 100 mM (each) ATP, CTP, GTP, and UTP, 2 µl of 100 mM dithiothreitol, and 2 µl of AmpliScribe T7 enzyme. The reaction mixture was incubated at 37°C for 2 h. Transcripts were translated using Red Nova reticulocyte lysate kit (Novagen) as described by the manufacturer. To optimize the translation efficiency, final salt concentrations in the reaction mixture were adjusted to 50 mM for potassium acetate and 0.25 mM for magnesium acetate. The U_L3.5 protein was labeled with [³H]leucine, and BTIF was labeled with [³⁵S]methionine and [³⁵S]cysteine.

Western blotting. Western blotting was carried out as described previously (28). Proteins were separated by SDS-PAGE and transferred to nitrocellulose paper (Bio-Rad). The nitrocellulose was blocked for 30 min with blocking buffer (20 mM Tris-HCl [pH 7.5], 150 mM NaCl, 0.05% Tween 20, 5% powdered skim milk), incubated with appropriate primary antibodies diluted in blocking buffer, and then reacted with peroxidase-labeled anti-rabbit or anti-mouse immunoglobulin (Ig) antibody (1:3,000; Amersham). The results were visualized by an enhanced chemiluminescence reaction (Amersham).

Transfections and confocal microscopy. The transfections were performed with Lipofectamine Plus reagent (Gibco) as described by the manufacturer. The day before transfection, Vero cells (4 × 10⁵) were plated into six-well plates containing one coverslip per well. Each well of cells was transfected with 2 µg of expression plasmid made up to 5 µg with pcDNA3 DNA. Forty hours after transfection, the cells were fixed with 4% paraformaldehyde for 15 min at room temperature and permeabilized with PBS containing 0.5% Triton X-100 for 15 min. For immunofluorescence analysis of virus-infected cells, MDBK cells were grown to 75% confluency on coverslips and infected with BHV-1 at an MOI of 10. At different times after infection, virus-infected cells were fixed and permeabilized as described above. For both transfected cells and BHV-1-infected cells, the monolayers were blocked with PBS containing 5% bovine serum albumin (BSA) at room temperature for 30 min and then incubated with rabbit anti-U_L3.5 (1:150) and mouse anti-BTIF(1:300) polyclonal antibodies diluted in PBS containing 5% BSA at room temperature for 1 h. The monolayers were washed extensively with PBS containing 0.2% Tween 20 and incubated with Texas Red-conjugated anti-rabbit IgG (Molecular Probes) and fluorescein isothiocyanate (FITC)-conjugated anti-mouse IgG (Gibco), both at a final concentration of 10 µg/ml, in PBS containing 5% BSA at room temperature for 45 min. Again, the monolayers were extensively washed with PBS containing 0.2% Tween 20. Coverslips were mounted onto slides with mounting medium (10% 10× PBS, 90% glycerol, 1 mg of p-phenylenediamine/ml), sealed with fingernail polish, and kept in the dark at 4°C until examination. The mounted coverslips were examined in two channels with a Bio-Rad MRC-1024 confocal microscope.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

BHV-1 U_L3.5 interacts with BTIF. To investigate the role of BHV-1 U_L3.5 in virus replication, we decided to first identify the proteins that associate with U_L3.5. Previously, Schikora et al. developed a polyclonal antibody that specifically recognizes the BHV-1 U_L3.5 protein (23). In the present experiment, that same anti-U_L3.5 polyclonal antibody immunoprecipitated U_L3.5 from both BHV-1-infected MDBK cells and partially purified BHV-1 virions but not from mock-infected MDBK cells (data not shown and Fig. 2, lanes 4 to 6). Strikingly, a 60-kDa protein also was immunoprecipitated by U_L3.5 antiserum, but not preimmune serum, from virus-infected MDBK cells and BHV-1 virions.

View larger version (62K): [in this window] [in a new window]   FIG. 1.   Analysis of proteins that bind to His-tagged UL3.5. MDBK cells were infected with BHV-1 at an MOI of 10, labeled with [35S]Met and [35]Cys for 12 h beginning 6 h after infection, and lysed with Triton X-100. HisUL3.5/Ni-NTA agarose and the control pRSET/Ni-NTA agarose were prepared as described in Materials and Methods. The agarose was incubated with the infected cell lysate and washed. Unbound flowthrough (FT) and bound proteins were analyzed by SDS-PAGE on a 13% gel and autoradiographed. Arrows, 60- and 85-kDa proteins.

U_L3.5 and BTIF coimmunoprecipitate. To biochemically verify the MALDI-MS result, a coimmunoprecipitation study was performed. Anti-BTIF polyclonal antibody was generated by immunizing mice with GST-BTIF. BHV-1-infected MDBK cell lysates were immunoprecipitated by either anti-U_L3.5 or anti-BTIF polyclonal antibodies. To test the specificity of both polyclonal antibodies, immunoprecipitation of in vitro-translated U_L3.5 and BTIF was also performed. Figure 2A shows the immunoprecipitation results. Anti-U_L3.5 and anti-BTIF, but not their preimmune sera, specifically recognized in vitro-translated U_L3.5 and BTIF, respectively (lanes 1, 3, 10, and 12). Moreover, anti-U_L3.5 did not precipitate in vitro-translated BTIF and vice versa (lanes 2 and 11). Two bands, 60 and 13 kDa, were precipitated by anti-BTIF specifically from BHV-1-infected MDBK cells (lane 7) but not from mock-infected cells (lane 8). The 60-kDa protein precipitated from virus-infected cells by anti-BTIF comigrated in SDS-PAGE gel with in vitro-translated BTIF (lane 10). Again, the anti-U_L3.5 antibody precipitated two proteins, 13 and 60 kDa, from BHV-1-infected cells (lane 6) but not from mock-infected cells (lane 5). The slight difference in migration rates between the in vitro-synthesized U_L3.5 and the native U_L3.5 may be due to posttranslational modification of U_L3.5 in infected cells. The 13-kDa proteins, as well as the 60-kDa proteins, which were precipitated from virus-infected cells by either anti-U_L3.5 or anti-BTIF antibodies, comigrated in SDS-PAGE gel (lanes 6 and 7), suggesting that they were identical.

View larger version (39K): [in this window] [in a new window]   FIG. 2.   Coimmunoprecipitation of UL3.5 and BTIF from virus-infected cells. (A) Cells were infected with BHV-1 at an MOI of 10 or were mock infected and labeled with [3H]leucine for 12 h beginning at 6 h pi. Proteins from BHV-1-infected (VC) and mock-infected (Mock) cells were precipitated with anti-UL3.5 antibodies (lanes 5 and 6) or anti-BTIF antibodies (lanes 7 and 8). Proteins from BHV-1-infected cells were also precipitated by UL3.5 and BTIF preimmune sera (lanes 4 and 9). The UL3.5 protein was synthesized in vitro in the presence of [3H]leucine and was immunoprecipitated with anti-UL3.5, anti-BTIF, and UL3.5 preimmune sera (lanes 3, 2, and 1, respectively). [35S]Met- and [35S]Cys-labeled in vitro-synthesized BTIF was immunoprecipitated with anti-BTIF, anti-UL3.5, and BTIF preimmune serum (lanes 10 to 12, respectively). The precipitated proteins were analyzed by SDS-PAGE on a 10 to 20% gradient gel and fluorographed. (B) One-fifth of samples 4 through 9 from panel A were separated by SDS-PAGE on a 13% gel. The proteins were transferred onto a nitrocellulose membrane and probed with either anti-UL3.5 or anti-BTIF antibodies as indicated. The bound antibodies were detected by enhanced chemiluminescence.

U_L3.5 and BTIF colocalize in virus-infected cells. To determine whether U_L3.5 and BTIF interact with each other in vivo, indirect immunofluorescence and confocal microscopy were performed to detect the colocalization of the two proteins in virus-infected cells. MDBK cells were infected with BHV-1 at an MOI of 10. At 6, 8, 10, and 12 hpi, cells were fixed and stained for U_L3.5 and BTIF (Fig. 3). Newly synthesized U_L3.5 was detected in the cytoplasm by 6 hpi and was distributed predominantly in the cytoplasm from 8 through 12 hpi. Starting at 10 hpi, a small portion of U_L3.5 also appeared in specific perinuclear structures. BTIF was mainly in the nucleus at 4 hpi (data not shown), was localized throughout the cells from 6 to 8 hpi, and was concentrated in the cytoplasm from 10 to 12 hpi. Localization of BTIF in perinuclear structures was also observed by 8 hpi. Although both the U_L3.5 and the BTIF were expressed by 6 hpi, colocalization of the proteins was not observed until 10 hpi, was restricted to the perinuclear structures, and increased dramatically up to 12 hpi, the last time point observed. The perinuclear fluorescent dots observed at 10 and 12 hpi resembled those observed in BHV-1-infected MDBK cells stained for gD (4) and in PrV-infected MDBK cells stained for gE (26), suggesting that they might represent part of the Golgi apparatus or virion assembly sites.

View larger version (82K): [in this window] [in a new window]   FIG. 3.   Colocalization of UL3.5 and BTIF in BHV-1-infected MDBK cells. MDBK cells were mock infected or infected with BHV-1 at an MOI of 10. At 6, 8, 10, and 12 hpi, the cells were fixed and incubated with both rabbit anti-UL3.5 and mouse anti-BTIF polyclonal antibodies. Cells were further incubated with Texas Red-conjugated anti-rabbit IgG and FITC-conjugated anti-mouse IgG and examined with a dual-channel confocal microscope.

Interaction of U_L3.5 and BTIF in vitro requires only aa 10 to 86 of U_L3.5. The in vitro pull-down assay was performed to demonstrate the direct interaction between U_L3.5 and BTIF. BTIF was synthesized in an in vitro transcription-translation system in which it was labeled with [³⁵S]methionine and [³⁵S]cysteine. In vitro-synthesized BTIF was incubated with HisU_L3.5/Ni-NTA agarose or pRSET/Ni-NTA agarose. Proteins bound to the agarose after extensive washing were eluted with imidazole and analyzed by SDS-PAGE and autoradiography. The in vitro translation reaction mixture immunoprecipitated with anti-BTIF was also run on the gel as a control (Fig. 4B, lane 2). Figure 4B shows that in vitro-synthesized BTIF bound to HisU_L3.5/Ni-NTA agarose but not to pRSET/Ni-NTA agarose (lanes 4 and 3), demonstrating the specific in vitro interaction between U_L3.5 and BTIF.

View larger version (27K): [in this window] [in a new window]   FIG. 4.   An in vitro pull-down assay shows that only aa 10 to 86 of UL3.5 were required for UL3.5-BTIF interaction. (A) Diagrammatic description of UL3.5 mutants used in the in vitro pull-down assay. The black box represents the fusion partner containing the His tag encoded by the pRSET vector. The white box represents the UL3.5 amino acid sequence. wt, wild type. (B) The BTIF protein was translated in vitro in the presence of [35S]Met and [35S]Cys. In vitro-synthesized BTIF (lane 1) was incubated with anti-BTIF bound to protein A-Sepharose (lane 2), control pRSET/Ni-NTA agarose (lane 3), full-length UL3.5 (lane 4), and truncated UL3.5 (lanes 5 to 8) bound to Ni-NTA agarose. The bound proteins were analyzed by SDS-PAGE and autoradiography. (C) A Coomassie blue-stained gel was used to show that about equal amounts of recombinant His-tagged UL3.5 proteins were loaded in the lanes shown in panel B.

U_L3.5 and BTIF colocalize in transfected cells. To show that BHV-1 U_L3.5 and BTIF can interact in vivo in the absence of other viral proteins, their colocalization was tested in a transient coexpression assay. Coding sequences for U_L3.5, its mutants, and BTIF were cloned into the expression vector pcDNA3. The plasmid DNAs were transfected into Vero cells. The cells were stained for BTIF and U_L3.5 40 h after transfection. When transiently expressed alone, BTIF was distributed homogeneously in the nucleus and the cytoplasm of most of the cells (Fig. 5B), whereas the wild-type U_L3.5 was distributed exclusively in the cytoplasm in a net-like pattern resembling those for the endoplasmic reticulum and Golgi apparatus (Fig. 5A). Because there was no green fluorescence detected in cells transfected with only pU_L3.5cDNA3 and no red fluorescence in cells transfected with only pBTIFcDNA3, the signals detected for U_L3.5 and BTIF were specific and not due to the cross-reactivity of the two antibodies. When wild-type U_L3.5 and BTIF were coexpressed, the two proteins colocalized in the nucleus in almost every cell analyzed (Fig. 5C). Colocalization was concentrated in the region near the inner surface of the nuclear membrane. Besides being detected in the nucleus, colocalization was detected in the cytoplasm in some cells (data not shown). However, in cells expressing less U_L3.5 or BTIF, colocalization was exclusively observed in the nucleus, suggesting that colocalization in the cytoplasm in the transient coexpression assay was a result of an excess of both proteins.

View larger version (22K): [in this window] [in a new window]   FIG. 5.   Colocalization of UL3.5 and BTIF in transfected cells. Vero cells were transfected with pUL3.5cDNA3 (A), pBTIFcDNA3 (B), and pUL3.5cDNA3 plus pBTIFcDNA3 (C). The cells were fixed 40 h after transfection, incubated with both anti-UL3.5 and anti-BTIF antibodies, stained with Texas Red- and FITC-conjugated secondary antibodies, and examined with a dual-channel confocal microscope.

View larger version (37K): [in this window] [in a new window]   FIG. 6.   Interaction of UL3.5 mutants with BTIF in transfected Vero cells. Vero cells were transfected with pN10UL3.5cDNA3 (A), pN20UL3.5cDNA3 (B), pC30UL3.5cDNA3 (C), pC40UL3.5cDNA3 (D), pN10UL3.5cDNA3 and pBTIFcDNA3(E), pN20UL3.5cDNA3 and pBTIFcDNA3(F), pC30UL3.5cDNA3 and paBTIFcDNA3(G), or pC40UL3.5cDNA3 and pBTIFcDNA3(H). The cells were stained and examined as described for Fig. 5.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

This research strongly suggests that BHV-1 U_L3.5 binds directly to BTIF both in vitro and in vivo through a domain located between aa 11 and 86 in U_L3.5. Since U_L3.5 does not have a cysteine, the interaction between U_L3.5 and BTIF is likely noncovalent. In our study, we cannot completely rule out the possibility that an intermediary is required for the interaction. However, the absence of stoichiometric amounts of other proteins in complexes immunoprecipitated by antibodies against either U_L3.5 or BTIF suggests that no intermediary is required. Nevertheless, it is possible that the intermediary was not efficiently radiolabeled by [³H]leucine and remained undetected. Because all the assays were done in the presence of mammalian cells or cell lysates, an intermediary, if any, must be common to MDBK cells, African green monkey kidney cells, and rabbit red blood cells.

The sites of U_L3.5 and BTIF colocalization for virus-infected and transfected cells are dramatically different. In virus-infected cells, newly synthesized U_L3.5 and BTIF colocalized at specific perinuclear structures (Fig. 3), whereas in transfected cells they colocalized in the nucleus (Fig. 5). The reason(s) for the difference is not clear. Possibly, colocalization of U_L3.5 and BTIF at the proper intracellular compartment in virus-infected cells requires ongoing synthesis of other viral proteins. Alternatively, downregulation or upregulation of a specific cellular protein(s) resulting from viral infection might account for the difference. Also note that when either U_L3.5 or BTIF was transiently expressed alone, it accumulated in a different pattern than when the two were transiently expressed together or when expressed during viral infection, indicating that both binding together and some other effect in virus-infected cells determined where the proteins localized. The C terminus of U_L3.5 seems to have an effect on the site of colocalization since U_L3.5 C-terminal deletions (Fig. 6G and H), but not the N-terminal 10-aa deletion (Fig. 6E), resulted in colocalization concentrated in the cytoplasm of transfected cells. A mutant BHV-1 containing a U_L3.5 C-terminal deletion would be helpful in answering the question of whether the C terminus of U_L3.5 is required to direct colocalization to the proper site during viral infection.

Although newly synthesized U_L3.5 and BTIF can be detected throughout the cytoplasm of virus-infected cells at 6 hpi, colocalization restricted to Golgi-like regions was not observed until 10 hpi, suggesting that the interaction may be specifically regulated. What regulates the interaction? Phosphorylation may be a good candidate. Phosphorylation has been recognized as an important regulatory mechanism for other herpesviral tegument proteins. For example, phosphorylation catalyzes herpesvirus tegument disassembly (21). Also, the nonphosphorylated, but not the phosphorylated, form of HSV-1 VP22 is incorporated into the virion (7). Our unpublished results suggest that both U_L3.5 and BTIF are phosphorylated in virus-infected cells. Detailed study of the relationship between U_L3.5 and BTIF interaction and phosphorylation may answer this question.

What is the function of the U_L3.5-BTIF interaction? Both U_L3.5 and BTIF are brought into cells upon virus infection since they are both virion components (20, 23). As shown for the transfected cells (Fig. 5), coexpression of U_L3.5 and BTIF resulted in localization of both proteins to the nucleus. The interaction with U_L3.5 might allow BTIF to enter the nucleus more efficiently at the beginning of infection, thus enhancing intermediate-early gene transactivation. Indeed, our preliminary results (data not shown) obtained with a transient transfection assay did show that cotransfection of U_L3.5 with BTIF enhanced transactivation of BHV-1 immediate-early promoter 1 by BTIF.

Alternatively, an interaction between U_L3.5 and BTIF may be important for viral assembly and egress. Indeed, colocalization of two proteins at probable sites of virion assembly became obvious 10 h after infection, just before infectious particles are released (4, 14). It has been shown that homologs of both U_L3.5 and BTIF have important functions in herpesviral assembly. In cells infected with a PrV mutant lacking U_L3.5, naked nucleocapsids accumulate near the trans-Golgi region (10), implying that PrV U_L3.5 is a key component for virus egress. Although the exact function of BHV-1 U_L3.5 in BHV-1 replication is still unknown, BHV-1 U_L3.5 is able to rescue a PrV U_L3.5-negative virus (9), suggesting that BHV-1 U_L3.5 might play a similar role in BHV-1 viral egress. HSV-1 VP16 is essential for virion assembly (27). Like its HSV counterpart, BTIF is synthesized at the late stage of virus infection (20). Given the shared transcriptional-activation functions of BTIF and VP16, one might speculate that BTIF, like VP16, also is essential in virus assembly. Results from the U_L3.5 deletion mutagenesis study may suggest that the U_L3.5-BTIF interaction is required for virus assembly. The PrV U_L3.5 domain required for viral assembly and egress is apparently located in the N terminus because disruption of the C terminus does not dramatically impair viral growth in cell culture (10). Since the most conserved region between PrV and BHV-1 U_L3.5 proteins is limited to the N-terminal 50 aa (13), it is reasonable to postulate that the N-terminal part of BHV-1 U_L3.5 involved in BTIF binding also contains the essential functional domain required for rescuing the PrV mutant. Thus, the successful rescuing of the PrV U_L3.5-null virus by BHV-1 U_L3.5 might be due to the interaction between the BHV-1 U_L3.5 protein and the PrV homolog of BTIF.

We should note that there are substantial differences among homologs of U_L3.5 and BTIF, so no definite conclusion about the function of the U_L3.5-BTIF interaction is possible from this study. For example, VZV ORF57 (homolog of the U_L3.5 gene) is not required for VZV replication (3). In PrV, there is no virion component that transactivates immediate-early genes (2). Future study on BHV-1 containing different U_L3.5 or BTIF mutants will be helpful to understand the roles of the U_L3.5-BTIF interaction in BHV-1 replication.