DNA Cleavage and Packaging Proteins Encoded by Genes UL28, UL15, and UL33 of Herpes Simplex Virus Type 1 Form a Complex in Infected Cells

Previous studies have indicated that the U_L6, U_L15, U_L17, U_L28,U_L32, and U_L33 genes are required for the cleavage and packagingof herpes simplex viral DNA. To identify proteins that interactwith the U_L28-encoded DNA binding protein of herpes simplexvirus type 1 (HSV-1), a previously undescribed rabbit polyclonalantibody directed against the U_L28 protein fused to glutathioneS-transferase was used to immunopurify U_L28 and the proteinswith which it associated. It was found that the antibody specificallycoimmunoprecipitated proteins encoded by the genes U_L28, U_L15,and U_L33 from lysates of both HEp-2 cells infected with HSV-1(F)and insect cells infected with recombinant baculoviruses expressingthese three proteins. In reciprocal reactions, antibodies directedagainst the U_L15- or U_L33-encoded proteins also coimmunoprecipitatedthe U_L28 protein. The coimmunoprecipitation of the three proteinsfrom HSV-infected cells confirms earlier reports of an associationbetween the U_L28 and U_L15 proteins and represents the firstevidence of the involvement of the U_L33 protein in this complex.

INTRODUCTION

Top

Introduction

The cleavage and packaging of newly synthesized herpesvirusDNA is a highly conserved process occurring late in viral replication.During herpesvirus infection, double-stranded viral DNA genomesaccumulate in cell nuclei as head-to-tail concatemers whichare subsequently cleaved into single genome lengths and packagedinto preformed viral capsids (reviewed in reference 19). Sixgenes of herpes simplex virus type 1 (HSV-1) are known to beessential for this process, namely, U_L6, U_L15, U_L17, U_L28, U_L32,and U_L33. In cells infected with viruses individually lackingthese genes, capsids appear normal and are readily detected,but viral DNA is neither cleaved nor packaged (2, 3, 5, 9, 12,25, 26, 30, 31, 35, 37, 40–42, 44).

The cleavage and packaging process of herpesviruses is believedto be similar to that employed by double-stranded DNA bacteriophages;consequently, it is useful to consider the mechanisms used bythese bacteriophages as a model for the cleavage and packagingevents employed by herpesviruses. Such models propose a centralrole for the viral "terminase," a complex of at least two proteinsthat (i) binds the viral DNA and links it with the empty viralcapsid (HSV-1) or prohead (bacteriophages); (ii) cleaves theviral DNA at precise internal sites, resulting in the separationof unit-length genomes from concatameric DNA; and (iii) hydrolyzesATP, providing the energy required to drive the DNA into thecapsid (reviewed in reference 10). With this paradigm in mind,efforts have been expended to identify herpesvirus gene productsthat perform functions expected of the viral terminase, especiallyDNA binding, ATP hydrolysis, and at least transient associationwith capsids.

There is a growing body of both direct and indirect evidencesuggesting that the U_L15 and U_L28 gene products comprise twosubunits of the HSV-1 terminase. The U_L28 gene product has beenshown to bind specifically to the HSV-1 DNA sequence pac1, whichis found in the a sequence of the genome and is known to beessential for the generation of correct genomic termini (4,38). The U_L15 protein has been hypothesized to hydrolyze ATP,based on limited homology with a putative nucleotide bindingmotif comprised of Walker boxes A and B within gp17, the largersubunit of the T4 bacteriophage terminase (15). Although theATPase activity of the U_L15 protein has yet to be directly demonstrated,a mutation within the Walker box precludes viral DNA cleavageand packaging (46). The U_L28 and U_L15 proteins have also beenshown to interact in transient-expression assays (1, 22, 23,45) and in coimmunoprecipitation experiments using nuclear extractsof infected cells (23).

This paper describes the isolation of a protein complex consistingof the U_L28, U_L15, and U_L33 proteins by immunoprecipitationfrom lysates of infected cells. This suggests for the firsttime that the U_L33 gene product may function as a third subunitof the putative viral terminase.

MATERIALS AND METHODS

Top

Materials and Methods

Cells and viruses. Vero and HEp-2 cells and transformed cell lines were maintainedin Dulbecco's modified Eagle's medium (DMEM) supplemented with10% newborn calf serum, penicillin, and streptomycin. Sf21 (Spodopterafrugiperda) cells were maintained in Grace's insect cell culturemedium (GibcoBRL) supplemented with 10% fetal bovine serum andgentamicin (10 µg/ml). The wild-type virus HSV-1(F) andthe mutant HSV-7202 have been previously described (6, 16),and their titers were determined on Vero cell monolayers. Thetiter of the U_L28 deletion virus HSV-gCB was determined by propagationon the transformed cell line C1 (41). The mutant viruses lackingU_L15 and U_L33 have been described previously (5, 9).

Recombinant baculoviruses encoding the HSV-1 U_L28, U_L15, andU_L33 genes were generated by cotransfecting Sf21 cells withbaculovirus (Autographa californica multicapsid nucleopolyhedrosisvirus [Ac MNPV]) DNA (Invitrogen) and plasmid DNA encoding theentire open reading frame of U_L15, U_L28, or U_L33 under the controlof the baculovirus polyhedron promoter in the pBlueBacIII vector(Invitrogen), followed by plaque purification of recombinantviruses.

Antiserum production. Plasmid pJB179, predicted to encode the full-length U_L28 proteinfused to glutathione S-transferase (GST), was generated by ligatinga 2.5-kb DNA fragment containing bases 55,682 to 58,158 (27)of HSV-1(F) to the EcoRI site of pGEX-4T-1 (Pharmacia). Theplasmid was sequenced to confirm that the U_L28-GST junctionmaintained the open reading frames of the U_L28 and GST proteins(data not shown). Escherichia coli BL-21 cells transformed withpJB179 were cultured at 37°C for 1 h followed by inductionof the fusion protein with 0.1 mM isopropyl-ß-D-thiogalactopyranosideat 25°C for 3 h. The induced U_L28-GST fusion protein wasisolated from the bacteria as described previously (18) withthe following modifications: Triton X-100 (2% [vol/vol]) wasadded to the clarified bacterial lysate, and the lysate wasincubated with glutathione-Sepharose beads overnight at 4°C.Bound protein was eluted from the beads by boiling in buffercontaining 2% sodium dodecyl sulfate (SDS) and 5% ß-mercaptoethanol,and the U_L28-GST fusion protein was separated from contaminatingbacterial proteins by SDS-polyacrylamide gel electrophoresis.The separated proteins were visualized by lightly staining withCoomassie blue, and the band containing U_L28-GST was removedand minced. To emulsify the acrylamide (containing the fusionprotein) for injection, sterile phosphate-buffered saline (PBS)was added to the minced gel and it was passed through successivelysmaller-gauge needles.

Determination of antibody specificity. Flasks (25 cm²) of Vero cells were infected at a multiplicityof infection (MOI) of 5.0 with either HSV-1(F) or the mutantgCB virus lacking U_L28. Twenty hours postinfection the cellswere harvested, pelleted, and resuspended in a buffer consistingof PBS with 0.5% Triton and 0.5% sodium deoxycholate. They werethen sonicated for 5 s at low power and clarified at high speedin a microcentrifuge. Fifty microliters of the supernatant wasthen added to 20 µl of loading buffer, and the samplewas boiled for 3 min and electrophoretically separated on anSDS-polyacrylamide gel. The proteins from the gel were transferredto nitrocellulose membranes and probed with either polyclonalrabbit antiserum (diluted in PBS plus 1% bovine serum albuminand 1% Tween 20) directed against ICP8 (36) or the antiserumdirected against U_L28-GST.

Production of radiolabeled cell lysates. HEp-2 cell monolayers were infected at an MOI of 5.0 PFU percell with either HSV-1(F) or the R7202 mutant virus, which isderived from HSV-1(F) but lacks the capacity to produce glycoproteinE (6). After 4 h of incubation, the medium was replaced withlabeling medium containing DMEM with 10% of the normal amountsof methionine and cysteine and 50 µCi of [³⁵S]-labeledcysteine and methionine (Translabel; ICN) per ml. Twenty hoursafter infection, cells were washed four times with ice-coldPBS, scraped off the flask, collected, pelleted and, if necessary,stored at -80°C until use.

Immunoprecipitations. Immunoprecipitations were performed using radiolabeled lysatesof infected HEp-2 cells immediately after thawing on ice (seeabove). Alternatively, HEp-2 cells were infected with HSV-1(F)or the gE-null mutant R7202 at an MOI of 5.0 PFU per cell andincubated at 37°C for 20 to 24 h. The cells were washedfour times with ice-cold PBS, scraped off the flask, collected,pelleted and, if required, stored at -80°C until use.

Cells from the equivalent of a 12.5-cm² flask were sonicatedfor 5 s at low power in a 1-ml volume of buffer consisting ofPBS containing 1% Tween 20 (PBSTw). Cells were then clarifiedin an Eppendorf Microfuge for 20 min at 14,000 x g, and thesupernatants were transferred to new microcentrifuge tubes.Three microliters of antiserum was added, and the tubes wereleft on wet ice for 2 h. Fifty microliters of a 50% slurry ofGammabind G-Sepharose beads (Amersham Pharmacia Biotech) inPBS were washed three times in PSBTw and added to each celllysate, and the tube was rotated at 4°C for 1 h. The beadswere then washed four times in excess PBSTw and boiled in 2%SDS, and eluted proteins were separated by electrophoresis ona denaturing polyacrylamide gel. The gel was then used eitherfor immunoblotting or fluorography. For the latter, the gelwas incubated for 30 min in a 1.25 M sodium salicylate solution,vacuum dried, and placed next to scientific imaging film (X-Omat;Kodak) at -80°C for a minimum of 24 h, after which the filmwas developed. The method for immunoblotting has been previouslydescribed (6); the antibodies used are listed in Table 1 anddetailed below in the text.

View this table:
[in this window]
[in a new window]
TABLE 1. Provenance of antibodies used in this study

In some experiments, Sf21 cells were infected at an MOI of 5.0PFU per cell with recombinant baculoviruses expressing the U_L15,U_L28, or U_L33 proteins. The cells were incubated at 30°Cfor 40 h, after which time the cells were washed four timeswith ice-cold PBS and stored at -80°C until use. The remainderof the immunoprecipitation was carried out as described above.

RESULTS

Top

Results

U_L28 and U_L15 proteins form a complex in HSV-1(F) infected cells. A polyclonal antiserum recognizing the HSV-1 protein U_L28 wasproduced by immunizing rabbits with full-length U_L28 fused tothe gene encoding GST (see Materials and Methods). To test thespecificity of the antiserum, HEp-2 cells were infected withHSV-1(F) (wild-type virus) or the U_L28-null mutant HSV-1 virusgCB (41) and harvested after 20 h. Proteins were denatured byboiling in SDS, separated on a denaturing polyacrylamide gel,transferred to a nitrocellulose membrane, and probed with antiseradirected against either ICP8, a virally encoded single-strandedDNA binding protein, or the U_L28 protein. Figure 1A shows thepresence of ICP8 in lysates of cells infected with either virus,indicating that the cells were infected. The antiserum directedagainst the U_L28 protein (Fig. 1B) recognized an 85,000 apparentM_r protein in the HSV-1(F)-infected cells but not in cells infectedwith gCB, thus indicating recognition of the U_L28 protein.

View larger version (44K):
[in this window]
[in a new window]
FIG. 1. Digitally scanned images of immunoblots probed with anti-ICP8 antibody (A) and anti-U_L28 antibody (B). Cells were infected with either HSV-1(F) or the U_L28 mutant virus gCB, and the proteins were separated on an 8% polyacrylamide gel before being transferred to a nitrocellulose membrane and probed with antiserum. Protein sizes are indicated on the left in thousands.

Immunoprecipitations were performed to identify proteins thatdirectly or indirectly associated with the U_L28 protein. Briefly,cells were infected with HSV-1(F) and infected cell proteinswere radiolabeled with [³⁵S]methionine and [³⁵S]cysteine asdetailed in Materials and Methods. Since the HSV-1 glycoproteinsE (gE) and I form an Fc receptor and are normally present inimmune complexes derived from HSV-infected cells (20), radiolabeledproteins were also immunoprecipitated from cells infected withthe gE-null mutant virus R7202 (6). The infected cells werelysed in PBS supplemented with 0.5% Tween 20, briefly sonicated,and clarified extensively by centrifugation. The lysates werethen reacted with anti-U_L28 serum as detailed in Materials andMethods. Immune complexes were purified, denatured in SDS, andelectrophoretically separated on denaturing polyacrylamide gels,followed by fluorography.

In both the wild-type- and gE^- virus-infected cells, the anti-U_L28-GSTantibody consistently immunoprecipitated two proteins of approximately80,000 and 85,000 apparent M_r (Fig. 2). On the basis of theirrespective sizes, these proteins were hypothesized to be derivedfrom U_L15 and U_L28, respectively. To address this possibility,nonradiolabeled cell lysates were harvested 24 h after infectionwith HSV-1(F) and immunoprecipitated with the U_L28-GST antiserum.The immunoprecipitated proteins were transferred to nitrocelluloseand probed with antibodies directed against U_L28-GST or theN terminus of U_L15 fused to GST (34). As shown in Fig. 3A andB, both the U_L28 and U_L15 proteins were detected in immune complexesderived from reaction of the anti-U_L28-GST antibody with HSV-1(F)-infectedcell lysates.

View larger version (38K):
[in this window]
[in a new window]
FIG. 2. Scanned digital image of fluorograph of radiolabeled proteins electrophoretically separated on an 8% polyacrylamide gel. HEp-2 cells were infected with either HSV-1(F) (lanes 1 and 2) or the gE-mutant virus strain 7202 (lanes 3 and 4) and immunoprecipitated as described in Materials and Methods by using either preimmune sera (lanes 1 and 3) or anti-U_L28 antibody (lanes 2 and 4). Protein sizes are indicated on the left in thousands. Arrowheads indicate proteins of 85,000 and 80,000 apparent M_r, which correspond to the U_L28 and U_L15 proteins, respectively.

View larger version (29K):
[in this window]
[in a new window]
FIG. 3. Scanned digital images of immunoblots probed with antisera directed against the U_L28 (A), U_L15 (B), or U_L33 (C) proteins. Cell lysates were immunoprecipitated with preimmune serum ( ${alpha}$ pi) or antibody directed against U_L28 ( ${alpha}$ 28), U_L15 ( ${alpha}$ 15), U_L33 ( ${alpha}$ 33), or U_L14 ( ${alpha}$ 14), as described in the text. The position and size of molecular weight standards are indicated on the left in thousands. The heavychain of the immunoprecipitating antibody can be seen as a dark band around 49,000 (arrowhead). (A and B) 8% polyacrylamide gels; (C) 15% polyacrylamide gel.

In order to confirm the putative interaction between U_L15 andU_L28 proteins, a reciprocal experiment was done. Cell lysateswere subjected to immunoprecipitation with the antibody directedagainst the N terminus of the U_L15 protein, followed by immunoblottingwith the anti-U_L28-GST serum. As shown in Fig. 3, the antiserumdirected against the U_L15 protein immunoprecipitated both theU_L15 and U_L28 proteins (Fig. 3A and B). Control reactions containinglysates of HSV-1(F)-infected cells mixed with preimmune sera(Fig. 3) failed to immunoprecipitate either the U_L15 or U_L28proteins. These data are therefore consistent with previouslypublished work showing that the U_L15 and U_L28 proteins interactin HSV-1(F)-infected cells (23).

The U_L28 protein forms a complex with the U_L15 and U_L33 proteins. The U_L28 and U_L15 proteins were the most readily detected radiolabeledproteins in the above immunoprecipitations; however, it remainedpossible that other proteins were coimmunoprecipitated but escapeddetection either because they were masked by comigrating nonspecificallyimmunoprecipitated radiolabeled proteins or because they didnot radiolabel efficiently under the experimental conditions.Therefore, the U_L28/U_L15 complex was immunopurified, and theimmunoprecipitated proteins were transferred to nitrocelluloseand probed directly with antisera directed against the HSV proteinsencoded by U_L6, U_L14, U_L15, U_L16, U_L17, U_L18, U_L19, U_L21, U_L25,U_L26.5, U_L28, U_L29, U_L31, U_L32, U_L33, or U_L49 or the host proteinactin (see Table 1 for antibody sources). Of these proteins,only the proteins encoded by U_L15 and U_L33 coimmunoprecipitatedwith the U_L28 protein, as shown in Fig. 3B and C. To confirmthat the U_L33 protein associated with the U_L28/U_L15 proteincomplex, antibody directed against the U_L15 protein was usedto immunoprecipitate the U_L33 protein from infected cell lysates(Fig. 3C), and anti-U_L33 antibody was used successfully to immunoprecipitateboth the U_L28 and U_L15 proteins (Fig. 3A and B). All experimentsincluded two negative controls: immunoprecipitations with preimmunesera and antibody directed against the HSV-1 protein U_L14 (Fig.3). In neither case were the U_L28, U_L15, or U_L33 proteins detectedin immunoprecipitated material.

The U_L28/U_L15/U_L33 protein complex forms in lysates of recombinant baculovirus-infected insect cells. To determine if the three proteins were sufficient to interactwith one another in the absence of other HSV-1-encoded proteins,Sf21 cells were infected or coinfected with recombinant baculovirusesexpressing the U_L28, U_L15, and U_L33 proteins. In the first experiment,insect cells were coinfected with recombinant baculovirusesexpressing all three proteins, and lysates of the infected cellswere reacted with antibody directed against either the U_L28,U_L15, or U_L33 protein. Immunoprecipitated material was thendenatured and analyzed on immunoblots probed with antibodiesdirected against U_L28-encoded protein. It was found that allthree antibodies immunoprecipitated the U_L28 protein, thus showingthat the three proteins can form a complex in the absence ofother HSV proteins when coexpressed in insect cells (Fig. 4).

View larger version (58K):
[in this window]
[in a new window]
FIG. 4. Digitally scanned image of an immunoblot showing the detection of the U_L28/U_L15/U_L33 protein complex from baculovirus-infected Sf21 cells. Cell lysates were immunoprecipitated as described in Materials and Methods with antibody directed against the U_L28 ( ${alpha}$ 28), U_L15 ( ${alpha}$ 15), or U_L33 ( ${alpha}$ 33) proteins or preimmune serum ( ${alpha}$ pi). The position of the U_L28 protein is indicated with an arrow. The heavy chain of the antibody can be seen as a dark band at approximately 50,000 (arrowhead). Molecular weight standards are marked at the right in thousands.

To identify which proteins within the U_L28/U_L15/U_L33 proteincomplex directly interact, insect cells were coinfected withrecombinant baculoviruses expressing either (i) U_L28 and U_L33,(ii) U_L15 and U_L33, or (iii) U_L28 and U_L15. Lysates of the coinfectedinsect cells were then subjected to immunoprecipitation withantisera as described in Materials and Methods. As had beenreported previously (1), U_L15 and U_L28 were coimmunoprecipitatedby the anti-U_L28 antibody in the absence of U_L33 (Fig. 5C).The anti-U_L33 antibody was also found to coimmunoprecipitatethe U_L28 protein in the absence of U_L15 and the U_L15 proteinin the absence of U_L28 protein (Fig. 5A and B). The additionof anti-U_L33 antibody to Sf21 cells infected with only the U_L28-and U_L15-expressing baculoviruses did not result in the coimmunoprecipitationof either protein in the absence of the U_L33 protein (data notshown). Further controls included in these experiments were(i) immunoprecipitations using preimmune sera followed by immunoblottingwith antisera directed against the U_L28, U_L15, and U_L33 proteins,and (ii) immunoprecipitations from lysates of Sf21 cells infectedwith wild-type baculovirus followed by immunoblotting with therespective antisera. These samples were uniformly negative,confirming that U_L28, U_L15, and U_L33 form a specific complexwhen coexpressed by infection of Sf21 cells with recombinantbaculoviruses.

View larger version (25K):
[in this window]
[in a new window]
FIG. 5. Digitally scanned images of Western blots showing the proteins immunoprecipitated from dual-infected Sf21 cells. Cells were infected with baculoviruses expressing U_L28 and U_L33 (A), U_L15 and U_L33 (B), or U_L28 and U_L15 (C) and processed as described in Materials and Methods. Proteins were immunoprecipitated with antibody directed against the U_L33 protein ( ${alpha}$ 33) or the U_L28 protein ( ${alpha}$ 28), or with preimmune serum ( ${alpha}$ pi). An equivalent amount of lysate was used in each lane. Membranes were blotted with antibody directed against the U_L28 protein (A) or U_L15 protein (B and C). Protein sizes are indicated on the left in thousands.

DISCUSSION

Top

Discussion

This paper describes the immunopurification of a complex ofproteins containing the U_L28, U_L15, and U_L33 gene products.These data support the conclusions of others that the U_L28 andU_L15 proteins interact (1, 22, 23, 45), and they represent thefirst report to show a specific interaction between the U_L33protein and other cleavage and packaging proteins. These resultsare supported by previously reported immunofluorescence assaysthat detailed a similar distribution of all three proteins inboth the cytoplasm and replication compartments of the nucleusin infected cells (8, 23, 32, 45).

The U_L33 protein is the smallest of the seven cleavage and packagingproteins, with an apparent M_r of 19,000. Since it contains onlytwo methionine and two cysteine codons in its open reading frame,the U_L33 protein does not radiolabel to high specific activitiesunder the conditions used in these studies and is thereforedifficult to detect by fluorography. To overcome this problem,immunoprecipitated material was subjected to immunoblottingwith the U_L33-GST antisera. Using this method, the U_L33 proteinwas readily detected in lysates of infected cells immunoprecipitatedwith antiserum directed against the U_L28, U_L15, or U_L33 proteins(Fig. 3C).

The U_L33 protein is expressed as a late gene product and, usingindirect immunofluorescence, it can be identified in the cytoplasmas well as nuclear replication compartments late in infection(32). Viruses lacking a functional U_L33 gene are unable to carryout cleavage and packaging of the HSV genome (5), but untilrecently no evidence revealing a specific role for the proteinin DNA cleavage and packaging reactions had been reported. However,the interaction with the U_L28 and U_L15 proteins described inthis paper suggests that the U_L33 protein may serve in someway that is relevant to terminase function. Although the precisefunction of U_L33 protein remains unknown, some possibilitiesinclude (i) ensuring that the terminase is correctly foldedor assembled, (ii) ensuring that the terminase complex is correctlytransported from the cytoplasm into the nucleus, and (iii) arole as part of the terminase holoenzyme. In light of the smallsize of U_L33, it is intriguing that HSV proteins of 21,000 and22,000 apparent M_r have been shown to bind the a sequence ofHSV-1 DNA (14). Although it is not known if these proteins areencoded by U_L33, the discovery that the U_L33 protein binds tothe U_L28 and U_L15 proteins indicates that further work intoits potential role as a member of the HSV terminase, possiblyas a DNA-binding protein, is warranted.

The presence of a third protein in the U_L28/U_L15 protein complexreveals that, if this tripartite complex does comprise the terminase,it is structurally more complex than its counterpart in bacteriophage ${lambda}$ , where the C-terminal end of gpNu1 simply interacts with theN terminus of gpA (17). The full extent of the HSV terminasemay in fact be larger still. Sixteen HSV proteins and one hostprotein were screened in the present work for inclusion in theputative terminase complex (Table 1). This work represents athorough search of likely interacting candidates, includingall of the cleavage-packaging proteins; however, it is not anexhaustive search of all HSV proteins. It is also possible thatthe antibodies directed against the U_L28 protein in the immunoprecipitationreactions could disrupt interactions with additional proteins,or that the experimental technique is not sensitive enough todetect the presence of additional proteins. Finally, there remainsthe possibility that a host protein may be included in the complex,as suggested by previous workers (21).

A recent report has suggested that the U_L33 protein of HSV-2is transported into the nuclei of cells transiently coexpressingthe U_L14 and U_L33 proteins, implying that the proteins interact(43). Such a conclusion was not corroborated in this study.The results of immunoprecipitation with antibody directed againstthe U_L14 protein followed by immunoblotting with monospecificantisera failed to detect either the U_L28, U_L15, or U_L33 proteinsin immunoprecipitated material (Fig. 3). In addition, the U_L14protein was not detected when material immunoprecipitated bythe U_L28 antibody was probed with U_L14-specific antiserum, whereasthe U_L33 protein was readily detected in this assay.

The anti-U_L28 antibody immunoprecipitated all three proteinsfrom lysates of Sf21 cells coinfected with baculoviruses expressingU_L28, U_L15, and U_L33, thus mirroring the results obtained inlysates of HEp-2 cells infected with HSV-1(F). This indicatedthat other HSV proteins were not necessary for the interactionsof the tripartite complex. When Sf21 cells were infected withsubsets of the recombinant baculoviruses, it was revealed thatnone of the possible interactions were dependent on the presenceof the third protein in the complex, implying direct interactionsbetween all three proteins.

The relatively mild conditions used to produce the HSV-1(F)-infectedcell lysates in these studies suggest that these lysates containpredominantly cytoplasmic rather than nuclear proteins. It istherefore hypothesized that the U_L28/U_L15/U_L33 complex formsin the cytoplasm of infected cells and is subsequently importedinto the nucleus, where all three proteins are present latein infection (8, 23, 32). Further studies are necessary to addressthis possibility. The data also suggest that the putative terminasecomplex forms independently from capsids, since none of themajor or minor capsid proteins, including the putative portalvertex protein encoded by U_L6 (29), were detected within thecomplex. Assembly of the terminase as a distinct process fromassembly of the capsid might serve to (i) promote proper assemblyof the terminase prior to import into the nucleus where DNAcleavage and packaging takes place, (ii) promote diffusion ofthe terminase within the nucleoplasm to increase the likelihoodthat genomic ends are engaged, or (iii) regulate the initiationof DNA cleavage and packaging.

Although the identification of the genes required for HSV-1DNA cleavage and packaging was completed a number of years ago,elucidating the exact functions of these proteins has provedchallenging. While the models of bacteriophage cleavage andpackaging have been useful for heuristic purposes, this andother studies suggest that the analogous processes of herpesvirusesare likely to be more complex.

ACKNOWLEDGMENTS

We gratefully acknowledge the technical assistance from JarekOkulicz-Kozaryn. We also thank J. Blaho for U_L49 antiserum,W. Ruyechan for ICP8 antiserum, F. Homa for the gCB deletionvirus, and G. Cohen and R. Eisenberg for antibodies againstHSV-1 capsid proteins.

These studies were supported by grant R01 GM50740 from the NationalInstitutes of Health.

FOOTNOTES

* Corresponding author. Mailing address: C5143 Veterinary Education Center, Cornell University, Ithaca, NY 14853. Phone: (607) 253-3385. Fax: (607) 253-3384. E-mail: jdb11{at}cornell.edu.

Present address: Department of Veterinary Pathobiology, Universityof Missouri, Columbia, MO 65211.

REFERENCES

Top