Intracellular Distribution of Capsid-Associated pUL77 of Human Cytomegalovirus and Interactions with Packaging Proteins and pUL93

ABSTRACT

DNA packaging into procapsids is a common multistep process during viral maturation in herpesviruses. In human cytomegalovirus (HCMV), the proteins involved in this process are terminase subunits pUL56 and pUL89, which are responsible for site-specific cleavage and insertion of the DNA into the procapsid via portal protein pUL104. However, additional viral proteins are required for the DNA packaging process. We have shown previously that the plasmid that encodes capsid-associated pUL77 encodes another potential player during capsid maturation. Pulse-chase experiments revealed that pUL77 is stably expressed during HCMV infection. Time course analysis demonstrated that pUL77 is expressed in the early late part of the infectious cycle. The sequence of pUL77 was analyzed to find nuclear localization sequences (NLSs), revealing monopartite NLSm at the N terminus and bipartite NLSb in the middle of pUL77. The potential NLSs were inserted into plasmid pHM829, which encodes a chimeric protein with β-galactosidase and green fluorescent protein. In contrast to pUL56, neither NLSm nor NLSb was sufficient for nuclear import. Furthermore, we investigated by coimmunoprecipitation whether packaging proteins, as well as pUL93, the homologue protein of herpes simplex virus 1 pUL17, are interaction partners of pUL77. The interactions between pUL77 and packaging proteins, as well as pUL93, were verified.

IMPORTANCE We showed that the capsid-associated pUL77 is another potential player during capsid maturation of HCMV. Protein UL77 (pUL77) is a conserved core protein of HCMV. This study demonstrates for the first time that pUL77 has early-late expression kinetics during the infectious cycle and an intrinsic potential for nuclear translocation. According to its proposed functions in stabilization of the capsid and anchoring of the encapsidated DNA during packaging, interaction with further DNA packaging proteins is required. We identified physical interactions with terminase subunits pUL56 and pUL89 and another postulated packaging protein, pUL93, in infected, as well as transfected, cells.

INTRODUCTION

Human cytomegalovirus (HCMV) belongs to the betaherpesviruses and is characterized by its narrow host range and prolonged replication cycle (1–3). Viral DNA packaging is a key step in HCMV replication. Enzymes involved in the viral DNA packaging process, so-called terminases, are responsible for site-specific duplex nicking and insertion of the DNA into the procapsids, (4, 5). Terminases were first described for double-stranded DNA bacteriophages and are highly conserved throughout many double-stranded DNA (dsDNA) viruses, e.g., herpesviruses and adenoviruses (6–13).

The HCMV enzymes required for this process, terminase subunits pUL56 and pUL89 (14–18), have already been identified. Together with the portal protein pUL104, they form the most powerful biological nanomotor (19, 20). Terminase subunit pUL56 provides the ATP-hydrolyzing activity required for DNA translocation (21, 22). The toroidal structure of pUL56 is in accordance with its function as a DNA-metabolizing protein. The nuclease activity necessary to process the concatemers into unit length genomes is mediated by terminase subunit pUL89. Binding to the portal protein is a prerequisite for DNA translocation. It is most likely that several additional viral proteins are required for the complex DNA packaging process. Recently Borst et al. (23) have shown that the HCMV homolog of herpes simplex virus 1 (HSV-1) pUL33, pUL51, is essential for cleavage and packaging.

In HSV-1, it has been shown that three additional proteins (pUL32, pUL17, and pUL25) are essential for packaging and retention of viral DNA. Two proteins, pUL17 and pUL25, are capsid associated and are found in mature virions (24, 25). They make up the capsid vertex-specific component (26, 27), an astral density on the exterior of capsids. Five copies of this complex surround the pentons at each capsid vertex. We have provided the first evidence that HCMV protein UL77 (pUL77), the homologue of HSV-1 pUL25, is a capsid-associated DNA packaging protein (28). Here we report the first characterization of HCMV pUL77 in infected cells regarding its intracellular localization, nuclear localization signals (NLSs), and interaction with terminase subunits pUL56 and pUL89 as well as pUL93, a homolog of HSV-1 pUL17.

MATERIALS AND METHODS

Cells and virus.

Human foreskin fibroblasts (HFFs), human embryonic kidney 293T (HEK293T) cells (ATTC, Manassas, VA), and human telomerase reverse transcriptase-immortalized retinoblastoma cells (RPE-1; ATCC CRL4000) were grown in Dulbecco's minimal essential medium supplemented with 10% fetal bovine serum, 2 mM glutamine, penicillin (5 U/ml), and streptomycin (50 μg/ml). Experiments were carried out with confluent cell monolayers (1.5 × 10⁷ cells), and HFFs were used at passages 8 to 15. HCMV strain AD169 stocks were prepared after infection of HFFs at a multiplicity of infection (MOI) of 0.1. The MOI was determined by plaque assays with staining for HCMV IE1.

Immunofluorescence assays.

For immunofluorescence assays, HFF or HEK293T cells were grown on coverslips. At the appropriate time point, cells were fixed with 3% paraformaldehyde as described previously (29).

Detection of HCMV pUL77 was carried out with pAbUL77 (28) for 45 min at room temperature prior to further incubation for 45 min with Cy3-labeled goat anti-human F(ab′)2 fragments. For staining of immediate-early, early, and late proteins, antibodies mAb63-27, specific for IE1 (provided by T. Stamminger); mAbUL44, specific for pUL44 (CMVpp52; Santa Cruz); mAb28-4, specific for the major capsid protein (MCP; kindly provided by B. Plachter); pAbUL104, specific for pUL104 (19); and M23, specific for the HCMV pUL112/113 gene products (30; kindly donated by K. Hirai) were used. Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI). After staining, the samples were mounted in Fluoroprep (bioMérieux, France) containing 2.5% (wt/vol) 1,4-diazabicylo[2,2,2]octane and examined by immunofluorescence microscopy with an Olympus BX60. Images were captured with the Olympus Cell-D program. Immunofluorescence of phosphonoacetic acid (PAA)-treated cells was performed with a Carl Zeiss LSM 510/ConfoCor2 laser scanning microscope, and images were processed with the Axioplan software program (Carl Zeiss). Images of the subcellular distribution of green fluorescent protein (GFP)-NLS chimeric proteins were acquired with a Nikon Eclipse A1 laser scanning microscope and the NIS software package (Nikon).

Pulse-chase experiments.

For pulse-chase analyses, at 60 h postinfection (p.i.), infected or mock-infected cells (25-cm² T-25 flasks; 2 × 10⁶ cells) were radiolabeled with 50 μCi/ml [³⁵S]methionine for a 1-h pulse after a 2-h deprivation period with a medium lacking methionine and cysteine. For the subsequent chase of infected cells, unlabeled methionine was added in excess for various times. Total cell extracts were prepared from labeled cultures by solubilization in immunoprecipitation buffer (20 mM Tris-HCl [pH 7.5], 100 mM NaCl, 1% NP-40, 5 mM EDTA, 25 mM iodoacetamide, 0.4% sodium deoxycholate) containing cOmplete protease inhibitor mix (Roche Diagnostics) prior to ultrasonic treatment. Insoluble material was sedimented for 30 min at 100,000 × g. Comparable amounts of extracts and pUL77-specific antibody (1:20) were used for precipitation as described previously (21).

Plasmid construction.

Plasmids encoding the putative pUL77 NLS were fused to the reporter protein β-galactosidase (β-Gal)–green fluorescent protein (GFP).

In order to generate the chimera pHM829_UL77NLSm or pHM829_77NLSb, plasmid pHM829 was digested with restriction endonucleases XbaI and BamHI. Gene fragments of the mono- and bipartite NLS were generated by PCR amplification from plasmid pHM829 with the following pairs of oligonucleotides (restriction sites are underlined, and NLS sequences are italic): 5′-GCTCTAGACGGGTGCGCAAACGGTATCTGCGTCAGGATATCGGGCCCGGGCCTAGAGCTCTCG-3′ and 5′-CGGGATCCATCCTCAATGTTGTGTCTGATCTTG-3′ for pHM829_UL77NLSm and pHM829_UL77NLSm2x and 5′-GCTCTAGACTCAAGCGGGGGTTGTACACGCAACCACGGTGGAAACGAGGGCCCGGGCCTAGAGCTCTC-3′ and 5′-CGGGATCCATCCTCAATGTTGTGTCTGATCTTG-3′ for pHM829_UL77NLSb.

The PCR product and vector pHM829 were separately digested with XbaI and BamHI, and the PCR product was ligated into the vector, yielding pHM829_UL77NLSm, pHM829_UL77NLSb, and pHM829_UL77NLSm2x. Construct pHM829_UL56NLS was designed for a previous study (29).

Plasmid pGEX93 was constructed by PCR amplification of pUL93 coding DNA from cosmid pHM1106 with primers 5′-CAGGATCCATGGAAACGCACCTGTATTCG-3′ and 5′-CACTCGAGCTAAAGATCGTCGAACGGC-3′. (restriction sites are underlined) The BamHI/XhoI-restricted PCR product was cloned into BamHI/XhoI-restricted pGEX5x-1.

Construction of plasmids with mutant pUL77.

Construct pUL77ΔCCM was designed in a previous study (28). To generate plasmids expressing the N- and C-terminal halves of UL77, gene fragments were PCR amplified from pcDNA_UL77 with previously used forward and reverse primers 5′-TCGGCTCGAGCACGACTCACGTCGTCGGC-3′ and 5′-GCGAATTCAGATTGCCGCTTGGGAAGGC-3′ (restriction sites are underlined) (28) and an additional oligonucleotide.

PCR products and vector pcDNA3.1HisC (Invitrogen, Karlsruhe, Germany) were all digested with EcoRI and XhoI, and the PCR products were ligated into pcDNA3.1HisC, resulting in constructs pcDNA_UL77Nt and pcDNA_UL77Ct.

Plasmid pcDNA_UL77-L79R_A86P was constructed stepwise by site-directed mutagenesis PCR with pcDNA_UL77 as the template. Primers 5′-TACTGCGAGGATCGCGAAGGGCG-3′ and 5′-CGCCCTTCGGGATCCTCGCAGTA-3′, containing the mutation L79R, were used in the first step.

Selection against the parental strand was accomplished by DpnI digestion. With generated plasmid pcDNA_UL77-L79R as the template, primers 5′-GTGTCCGAGCCCGAGGCGCTG-3′ and 5′-CAGCGCCTCGGGCTCGGACAC-3′, containing the mutation A86P, were used in the second step. Selection against the parental strand was accomplished by DpnI digestion.

To generate plasmid pcDNA_UL77Δ74-90, mutagenesis was performed by PCR from the template pcDNA_UL77 with primers 5′-GGTCACAGGGTACAGACTAACCAGCAGTGCGAG-3′ and 5′-GCTCGCACTGCTGGTTAGTCTGTACCCTGTG-3′. Selection against the parental strand was accomplished by DpnI digestion.

PAGE and Western blot analysis.

Extracts from HCMV AD169-infected HFFs or transfected HEK293T cells expressing pUL77 or mutant pUL77 constructs, mutant pUL89 or pUL89 constructs, or mutant pUL56 or pUL56 constructs were solubilized in 4× sample buffer (4% [vol/vol] β-mercaptoethanol, 0.01% [wt/vol] bromophenol blue, 4% [wt/vol] glycerol, 4% [wt/vol] SDS, 0.2 M Tris-HCl [pH 6.8]) prior to separation by 8% (wt/vol) SDS-PAGE. Proteins were transferred to nitrocellulose sheets and subjected to Western blot analysis as described previously (29). Antibodies mAbUL89, specific for pUL89 (kindly provided by S. Jonjic [23]; 1:1,000); mAbUL56, specific for pUL56 (kindly provided by S. Jonjic [23]; 1:1,000); and pAbUL77, specific for pUL77 (28) (1:10); and mab63-27, specific for IE1 (kindly provided by T. Stamminger; 1:10), were used as the primary antibodies. For detection of primary antibody binding, horseradish peroxidase-conjugated anti-human or anti-mouse F(ab′)2 fragments (1:5,000 in PBS with 0.3% bovine serum albumin; Abcam, Cambridge, United Kingdom) and the ECL (Super Signal West Pico) reagent were used as recommended by the supplier (Pierce; Thermo Fisher Scientific, Germany).

Transient transfection.

HEK293T cells at 60% confluence were transfected with the appropriate DNA (1 μg each) and TurboFect (Fermentas) according to the manufacturer's protocol. At 48 h after transfection, cells were harvested and subjected to further analyses.

Immunoprecipitation.

For coimmunoprecipitation, RPE-1 cells (2 × 10⁶) were infected with HCMV AD169 (MOI, 2) or HEK293T cells (2 × 10⁶) were transfected with (i) Myc-UL77 and pcDNA-56 (either the wild type or various mutant constructs), (ii) Myc-UL77 and pcDNA-UL89, (iii) pUL93GST, (iv) His-UL77 (wild type or mutant constructs) and pUL93GST, (v) His-UL77 (wild type or mutant constructs) and pUL56GST, or (vi) His-UL77 (wild type or mutant constructs) and pUL89GST. Total cell extracts were prepared 72 h after infection or 48 h after transfection by solubilization in coimmunoprecipitation buffer (20 mM Tris-HCl [pH 7.5], 150 mM NaCl, 0.5% Tween, 5 mM EDTA) containing the protease inhibitor mix M (Serva, Heidelberg, Germany) prior to ultrasonic treatment. Insoluble material was sedimented for 10 min at 100,000 × g. Comparable amounts of extracts were used for precipitation. According to the proteins used for transfection, pUL77 was precipitated either by a monoclonal antibody (MAb) against glutathione S-transferase (GST; 1:200; GeneTex) or by an anti-His MAb (1:200; Qiagen) in cotransfection experiments. Immunoprecipitates were analyzed by SDS-PAGE, followed by Western blotting with UL56- or UL89-specific antibodies (1:1,000; kindly provided by S. Jonjic), an anti-GFP MAb (1:1,000), an anti-GST MAb (1:1,000), or an anti-His MAb (1:1,000).

RESULTS

pUL77 stability.

To analyze the intracellular stability of pUL77, pulse-chase experiments were performed with AD169-infected HFF (MOI, 1) starting at 60 h p.i. Extracts of mock-infected or infected cells obtained at pulse and chase time points were immunoprecipitated with pAbUL77, and after SDS-PAGE, pUL77 was visualized by autoradiography. Radiolabeled pUL77 was clearly visible after a 1-h pulse in both the monomer and oligomer forms (Fig. 1A). Both polypeptides were still present after a 1-h chase. The proteins were detected after all of the chase periods; however, there was a marked reduction of the proteins after a 3-h chase. After a 5-h chase, the pUL77 signal strength was down by approximately 50%. These data show that pUL77 is stable for several hours.

FIG 1.

(A) Intracellular stability of pUL77. Extracts from radiolabeled mock- or HCMV-infected HFF were immunoprecipitated with pAb-UL77 and analyzed by autoradiography after SDS-PAGE. Molecular mass markers (M) are indicated on the left, and the positions of monomeric and dimeric pUL77 are indicated on the right. (B) Intracellular distribution of pUL77 in infected cells. Mock-infected or HCMV-infected HFF were subjected to immunofluorescence analysis with antibodies to pUL77. (C) Intracellular localization of IE1, pUL44, and the MCP in infected- or mock-infected cells. Cell nuclei were stained with DAPI. The bars in panels B and C are 5 μm.

Subcellular distribution of pUL77.

To characterize pUL77 during viral infection, we analyzed its expression and subcellular distribution by immunofluorescence experiments. Coverslip cultures were infected with laboratory strain AD169 (MOI, 1). At different time points p.i., cells were stained with the pUL77-specific polyclonal antibody (pAbUL77) and the MAb to IE1, pUL44, or the MCP. As shown in Fig. 1B, pUL77 was detected predominantly in the nucleus; however, its distribution changed over time. At early time points, only a few pUL77-positive signals were detected in the nucleus, according to its colocalization with pUL44 in prereplication centers (Fig. 1B and C; 24 and 48 h p.i.). The pattern changed from distribution in prereplication into replication centers (Fig. 1B, 75 and 96 h p.i.). While most of the pUL77 was still localized in the nucleus at late times during infection, some cytoplasmic staining was observed in addition (Fig. 1B, 120 h p.i.). Mock-infected cells served as controls (Fig. 1B).

Kinetics of pUL77 expression.

To validate the kinetics of HCMV pUL77 expression, experiments in the presence of PAA and cycloheximide (CHX) in combination with actinomycin D (ActD) were performed.

While PAA inhibits the viral DNA polymerase, leading to an arrest of viral gene expression in the early phase, CHX blocks protein translation and ActD inhibits transcription.

To examine the kinetics of pUL77 expression, cells were either only treated with CHX for 6 h p.i. or incubated with ActD for 18 h after CHX treatment. As controls, staining of immediate-early protein IE1, pUL104, and pUL112/113 was used. In cells treated only with CHX, no staining was observed (Fig. 2A; 6 h p.i. + CHX). While both pUL104 and IE1, which are expressed with immediate-early kinetics, could be detected in cells treated with CHX for 6 h p.i., followed by 18 h of treatment with ActD, neither pUL112/113 nor pUL77 could be seen (Fig. 2A).

FIG 2.

Kinetics of pUL77 expression in infected cells. (A) Intracellular localization of pUL77, IE1, pUL112/113, and pUL104 in infected or mock-infected cells after the addition of CHX at 6 or 24 h p.i. (6 h CHX + 18 h ActD) or untreated cells. Cell nuclei were stained with DAPI. Bars, 20 μm. (B) Confocal microscopy of the effect of PAA on pUL77 expression. HCMV-infected HFF were treated with PAA prior to immunofluorescence analysis with antibodies to pUL77 and pUL112/113.

To determine whether pUL77 is expressed with early or late kinetics, cells were infected in the presence of 200 μg/ml (wt/vol) PAA and analyzed at 72 h p.i. by double staining with pAbUL77 and MAb M23 against pUL112/113. The fluorescence pattern of PAA-treated cells (72 h p.i. + PAA) equaled that of untreated cells at 24 h p.i. (Fig. 2B). Taken together, these results indicate early-late expression of HCMV pUL77 during viral replication in the nucleus of the infected cell.

Nuclear translocation of pUL77 is mediated by the N-terminal part.

To further investigate the nuclear localization of pUL77, HEK293T cells were transfected with pcDNA_UL77, pcDNA_UL77Nt, pcDNA_UL77Ct, pcDNA_UL77ΔCCM, and pcDNA_UL77Δ74-90 (Fig. 3A). As shown in Fig. 3B, transiently expressed full-length pUL77, as well as pUL77Nt and pUL77Δ74-90, was detected in the nucleus. In contrast, pUL77Ct and pUL77ΔCCM, both lacking the N terminus, had exclusive cytoplasmic staining (Fig. 3B). HEK293T served as a control (Fig. 3B). These data indicate that pUL77 is translocated into the nucleus in the absence of other viral proteins and that this nuclear translocation is mediated by the N-terminal part of the protein.

FIG 3.

Nuclear localization of pUL77. (A) Schematic drawings of wild-type pUL77 and mutant constructs pcDNA_UL77Nt, pcDNA_UL77Ct, pcDNA_UL77ΔCCM, and pcDNA_UL77Δ74-90. (B) Intracellular localization in HEK293T cells transfected with pcDNA_UL77, pcDNA_UL77Nt, pcDNA_UL77Ct, pcDNA_UL77ΔCCM, or pcDNA_UL77Δ74-90. Cell nuclei were stained with DAPI. Bars, 5 μm.

Analysis of putative NLSs of pUL77.

In order to validate the nuclear localization of HCMV pUL77, in silico analyses of putative NLSs were performed by using the program NLStradamus (31). Two putative NLSs in the N-terminal part of pUL77 were predicted, a monopartite NLS (amino acids [aa] 55 to 63) and a bipartite NLS (aa 219 to 231) (Fig. 4A).

FIG 4.

Analysis of the predicted NLS of pUL77. (A) Schematic drawings of identified mono- and bipartite NLSs. (B) β-Gal- and GFP-specific protein segments. The residues of pUL77NLSm, pUL77NLSm2x, pUL77NLSb, or pUL56NLS are shown. (C) Subcellular distribution of pUL77 NLSm (e to f), pUL77NLSb (g, h), or pUL77NLSm2x (k, l) fused to the β-Gal–GFP reporter examined at 48 h after transfection of HEK293T cells by detection of GFP fluorescence. As a control, the fusion protein with pUL56NLS (b, c) was used. Cell nuclei were stained with DAPI. The merged image and the GFP channel are also shown.

To validate both putative NLSs, the amino acid motifs were fused to the reporter pHM829, which enables expression of the protein fragments that are fused to β-Gal and GFP (Fig. 4B). The fusion constructs were expressed transiently in HEK293T cells (Fig. 4C). To determine whether the inserted amino acid sequences were expressed properly, the subcellular localization of the chimeras was analyzed by GFP signals and DAPI staining for visualization of the nuclei by confocal microscopy. Interestingly, the analyses showed that neither pUL77NLSm nor pUL77NLSb is sufficient to translocate the reporter protein into the nucleus; however, a small amount is detectable in the nucleus (Fig. 4C, c to f). Interestingly, with plasmid pHM829_UL77NLSm2x, encoding a chimeric protein with two NLSm motifs, the nuclear localization of the reporter protein increased up to complete translocation into the nucleus (Fig. 4C, k and l) As a negative control, the expression of the reporter pHM829 alone revealed retention of the reporter proteins in the cytoplasm (Fig. 4C). As a positive control, plasmid pHM829_UL56NLS, containing the NLS of terminase subunit pUL56, was used. Expression of this plasmid showed a clear translocation of the reporter proteins into the nucleus (Fig. 4C). These results indicated that none of the newly identified NLS sequences of HCMV pUL77 could serve as a nuclear targeting signal by itself.

pUL77 interacts with terminase subunits pUL56 and pUL89.

To analyze whether pUL77 interacts with the terminase subunits, HFFs were infected with HCMV AD169 (MOI, 1) and subjected to coimmunoprecipitation with monospecific antibody pAb-UL77 at 72 h p.i. In contrast to our further investigation here, a MAb to pUL56, mAbUL56, instead of polyclonal antibody pAbUL56 (28), was used. The following immunostaining detected full-size pUL56 (mAbUL56; Fig. 5A, pUL56), as well as full size 70-kDa pUL89 (mAbUL89; Fig. 5A, pUL89) and pUL77 (Fig. 5A, pUL77), in both HCMV-infected cell extracts and coimmunoprecipitates. Immunostaining of IE1 was used as a negative control (Fig. 5A, IE1).

FIG 5.

Interaction of pUL77 with terminase subunits pUL56 and pUL89. (A) HFFs were infected with HCMV AD169 prior to coimmunoprecipitation (Co-IP) with an antibody specific to pUL77. Precipitated proteins were subjected to Western blot analyses with antibodies to pUL77, IE1, pUL56, and pUL89. (B) HEK293T cells were cotransfected with pEYFP_UL56 and pcDNA_UL77 or two mutant variants of pcDNA_UL77. Immunoprecipitation (IP) was performed with antibodies to the His tag. Cell extracts were subjected to Western blot analyses with antibodies to GFP. (C) HEK293T cells were cotransfected with pGEX_UL89 and pcDNA_UL77 or mutant variants of pcDNA_UL77. Immunoprecipitation was performed with antibodies to the His tag. Cell extracts were subjected to Western blot analyses with antibodies to GST and the His tag. Molecular mass markers (M) are indicated on the left.

To further investigate whether interactions between pUL77 and the terminase subunits are obtained with solely expressed proteins, coimmunoprecipitations were performed with HEK293T-transfected cells. We have shown before that pUL77 contains a coiled-coil motif (CCM) that is important for oligomerization, as well as for DNA binding properties (28). In this experiment, besides full-length pUL77, we used mutant constructs with a deletion of the CCM or a destroyed CCM (pUL77L79R_A86P).

For precipitation, a His-tagged antibody to pUL77, pUL77ΔCCM, or pUL77L79R_A86P was used prior to immunostaining of the pUL56-EYFP fusion protein (anti-GFP antibody; Fig. 5B) or the pUL89-GST fusion protein (anti-GST antibody; Fig. 5C). In precipitates with wild-type pUL77 and the mutant construct pUL77 L79R_A86P, the large terminase subunit pUL56 (Fig. 5B, pUL56-EYFP) and the small subunit pUL89 (Fig. 5C, pUL89GST) were detected. However, no interactions of both terminase subunits with the CCM deletion mutant form of pUL77 were detected (Fig. 5B and C). In precipitates of vector-transfected cell extracts, no specific proteins were observed (Fig. 5B, EYFP, and C, GST). These data imply (i) a physical interaction between capsid-associated pUL77 and terminase subunits pUL56 and pUL89 and (ii) that the coiled-coil domain is not needed for these interactions, whereas the N-terminal part (aa 1 to 93) of pUL77 is required.

pUL77 interacts with the HCMV homologue of HSV-1 UL17.

To investigate direct interactions between pUL77 and pUL93, the homologue of HSV-1 pUL17, analyses were performed with RPE-1 cells infected with HCMV AD169 (MOI, 1) and transfected with pGEX_UL93 at 48 h p.i. Coimmunoprecipitation was performed with monospecific pAb-UL77, and the following Western blot analyses were performed with antibodies to GST. pUL93GST, with an expected size of 100 kDa (Fig. 6A, pUL93GST), and pUL77 were detected (Fig. 6A, pUL77) in HCMV-infected and transfected RPE-1 cell extracts, as well as in coimmunoprecipitates. Infected RPE cells transfected with the vector pGEX5x-1 served as a negative control (Fig. 6A, GST).

FIG 6.

Interaction of pUL77 with pUL93. (A) RPE-1 cells infected with HCMV AD169 were transfected at 48 h p.i. with pGEX_UL93 or pGEX5x-1. Cells were lysed at 24 h posttransfection, and immunoprecipitation (IP) was performed with antibodies to pUL77. Cell extracts and coimmunoprecipitated (Co-IP) proteins were subjected to SDS-PAGE prior to Western blot analyses with antibodies to pUL77 and GST. (B) HEK293T cells were cotransfected with pGEX_UL93 and pcDNA_UL77 or mutant variants of pcDNA_UL77. Immunoprecipitation was performed with antibodies to GST. Cell extracts and coimmunoprecipitated proteins were subjected to SDS-PAGE prior to Western blot analyses with antibodies to pUL77 and GST. Molecular mass markers (M) are indicated on the left of each panel.

To investigate whether solely expressed pUL77 is able to interact with pUL93, HEK293T cells were cotransfected with pcDNA_UL77, pcDNA_UL77ΔCCM, or pUL77L79R_A86P and pGEX_UL93 or pGEX5X-1. For precipitations, GST antibody was used while detection by immunoblotting was performed with antibodies to the His tag. In precipitates, wild-type pUL77 and the mutant construct pUL77-L79R_A86P were observed (Fig. 6B, pUL77his and pUL77his L79R_A86P). The mutant construct pUL77hisΔCCM was not coprecipitated with pUL93 (Fig. 6B). Cotransfection of pUL77 with the vector pGEX5x-1 served as a negative control (Fig. 6B). These results showed an N terminus-dependent physical interaction of pUL77 with pUL93.

DISCUSSION

Viral replication and DNA packaging into preformed capsids are the key steps of HCMV maturation. It is most likely that several additional viral proteins are required for the complex DNA packaging process. One candidate is pUL77, a conserved core protein of HCMV. The study presented here extends our previous observation on this essential protein (28). In this context, it was of particular interest to examine the kinetics of pUL77 expression, the intrinsic potential of pUL77 for nuclear translocation, and its interactions with the terminase subunits and another postulated packaging protein, pUL93.

We demonstrated that pUL77 has predominantly late expression kinetics and is enriched in the nuclei of infected cells. The protein was first found in prereplication and replication compartments (24 to 72 h p.i.), while its distribution at later time points changed into a homogeneous nuclear pattern (96 to 120 h p.i.). In addition, minor amounts of pUL77 were detected in the cytoplasm at very late stages of infection (120 h p.i.). In immunofluorescence analyses of infected HFFs in the absence or presence of the DNA replication and translation inhibitor PAA or CHX, the cells remained unstained in the presence of both inhibitors. These data validated the idea that pUL77 is an early-late protein. Proteins with late kinetics are structural proteins, including those that are important for capsid assembly. The components of the nanomotor (the portal protein pUL104, terminase subunits pUL56 and pUL89) and the recently described proteins pUL51 and pUL52 (23, 32), which seem to be involved in the cleavage-and-packaging process, were also expressed with early-late or late kinetics. Our observations are consistent with findings on the HSV-1 and PrV pUL77 homologue pUL25, which is also expressed late after infection (33, 34).

To examine the nuclear localization of pUL77 more precisely, we analyzed the subcellular distribution of pUL77 and different pUL77 mutant constructs in a transient assay. We observed homogeneous intranuclear immunofluorescence signals with wild-type pUL77, pUL77Nt, and pUL77Δ74-90, whereas only cytoplasmic staining was found with the mutant construct pUL77ΔCCM. As expected, intranuclear replication centers were not detectable under these conditions because additional viral proteins and/or viral DNA are needed for this particular localization. The transport of pUL77 into the nucleus in the absence of other viral proteins implied the existence of an endogenous NLS(s) at the N terminus of the protein. NLSs are located mainly at the termini of nucleophile proteins, leading to the assumption that the terminal parts of a protein are rather exposed on the protein surface in order to be able to interact with cellular structures that mediate nuclear transport (35–37). It appeared that both postulated NLSs of pUL77 are not sufficient to transport the cytosolic β-Gal–GFP reporter into the nucleus, although weak fluorescence in the nucleus was shown in both cases. An explanation could be that the cooperation of both NLSs is necessary for complete nuclear localization. In the case of ICP8 of HSV-1, Gao and Knipe (38) reported that NLSs not only at the C terminus but also in distant areas of the N terminus of the protein are important for nuclear localization. In the case of HCMV, analysis was performed with an enlargement of the inserted amino acids by doubling the pUL77-NLSm. Interestingly, the nuclear translocation of this chimeric protein (pUL77-NLSm2x) was increased up to complete nuclear localization. On the basis of these results, it seems that for NLSs that are more complex, the correct conformation and the resulting exposure of the NLS on the protein surface are important for its function. However, it would be interesting to determine the situation in the viral context.

Experiments performed in this study aimed to identify interaction partners of pUL77. By coimmunoprecipitation of infected and transfected cells, we observed a physical interaction of pUL77 with the small terminase subunit, pUL89, and with pUL93, another putative packaging protein. Furthermore, we validated our previous observation of the interaction with the large terminase subunit, pUL56, with a new MAb (28). Interestingly, concerning the interaction of pUL77 with pUL93, Borst and colleagues, whose study is a companion to this paper (39), reported similar observations. Furthermore, by using HCMV deletion mutants, Borst et al. demonstrated that without pUL77 and pUL93, cleavage into unit length genomes and packaging are prevented. These results imply that a stable interaction of pUL77 with the terminase subunits and pUL93 is required for the packaging process.

It turned out that pUL77 interacts with many proteins, such as the major capsid protein, tegument protein pUL71, the components of the nanomotor (pUL56, pUL89, and pUL104), and pUL93, and dsDNA (28). Similar observations of the pUL77 homolog of PrV and HSV-1 were reported (27, 33, 40–43). Therefore, one could conclude that pUL77 is a multifunctional protein that is required for viral DNA packaging and maybe for other processes during viral replication, e.g., transport of the capsid to the nucleus or nuclear egress.

Furthermore, it is clear that the N terminus of pUL77 must be important for the functions of the protein. The protein-protein interaction studies of this work showed that the N-terminally truncated mutant construct pUL77ΔCCM was incapable of interacting with either pUL93 or terminase subunits pUL56 and pUL89.

In sequence comparisons of herpesvirus UL25 homologues, the N terminus is a very poorly conserved area (44) that is nevertheless essential for viral replication (45). Thus, a pUL25 mutant construct lacking the first 45 aa was not sufficient to complement a UL25 deletion mutant virus. Furthermore, it was reported that the first 50 aa of this protein are crucial for the connection of pUL25 to the vertices of the capsid (46). Future experiments will show whether these observations reflect the situation for HCMV pUL77.

In conclusion, the data from the present study demonstrate that (i) pUL77 is expressed with early-late kinetics, (ii) is translocated into the nucleus, and (iii) interacts directly with terminase subunits pUL56 and pUL89 and a new component of the packaging complex, pUL93.