Herpes simplex virus replication compartments can form by coalescence of smaller compartments

Herpes simplex virus replication compartments can formby coalescence of smaller compartments

Travis J Taylor, Elizabeth E. McNamee, Cheryl Day, and David M. Knipe*Department of Microbiology and Molecular Genetics, and the Program in Virology, Harvard Medical School,

200 Longwood Avenue, Boston, MA 02115, USA

Received 11 November 2002; returned to author for revision 20 December 2002; accepted 14 February 2003

Abstract

Herpes simplex virus (HSV) uses intranuclear compartmentalization to concentrate the viral and cellular factors required for theprogression of the viral life cycle. Processes as varied as viral DNA replication, late gene expression, and capsid assembly take place withindiscrete structures within the nucleus called replication compartments. Replication compartments are hypothesized to mature from a fewdistinct structures, called prereplicative sites, that form adjacent to cellular nuclear matrix-associated ND10 sites. During productiveinfection, the HSV single-stranded DNA-binding protein ICP8 localizes to replication compartments. To further the understanding ofreplication compartment maturation, we have constructed and characterized a recombinant HSV-1 strain that expresses an ICP8 moleculewith green fluorescent protein (GFP) fused to its C terminus. In transfected Vero cells that were infected with HSV, the ICP8–GFP proteinlocalized to prereplicative sites in the presence of the viral DNA synthesis inhibitor phosphonoacetic acid (PAA) or to replicationcompartments in the absence of PAA. A recombinant HSV-1 strain expressing the ICP8–GFP virus replicated in Vero cells, but the yieldwas increased by 150-fold in an ICP8-complementing cell line. Using the ICP8–GFP protein as a marker for replication compartments, weshow here that these structures start as punctate structures early in infection and grow into large, globular structures that eventually fill thenucleus. Large replication compartments were formed by small structures that either moved through the nucleus to merge with adjacentcompartments or remained relatively stationary within the nucleus and grew by accretion and fused with neighboring structures.© 2003 Elsevier Science (USA). All rights reserved.

Introduction

Herpes simplex virus (HSV) is a large double-strandedDNA virus that replicates within the nucleus of the infectedcell. As such, many viral processes must occur within thenucleus during the course of infection including viral tran-scription (Knipe et al., 1987; Leopardi et al., 1997; Phelanet al., 1997; Randall and Dinwoodie, 1986; Rice et al.,1994), viral DNA synthesis (de Bruyn Kops and Knipe,1988; Phelan et al., 1997; Randall and Dinwoodie, 1986),and virion assembly and DNA packaging (de Bruyn Kops etal., 1998; Lamberti and Weller, 1998; Ward et al., 1996; Yuand Weller, 1998). While these processes are temporallyregulated by the viral cascade of immediate-early, early, andlate gene expression (Roizman and Knipe, 2001), there is an

additional level of spatial control imposed within the nu-cleus. Viral processes do not take place diffusely within thenucleus, but within globular intranuclear structures calledreplication compartments (Bush et al., 1991; de Bruyn Kopsand Knipe, 1988; de Bruyn Kops et al., 1998; Knipe et al.,1987; Phelan et al., 1997; Quinlan et al., 1984).

HSV employs the strategy of intranuclear compartmen-talization to organize the processes required for viral repli-cation, which allows for the local accumulation of diffuse orlimiting factors, be they individual proteins, large proteincomplexes, or small biological molecules, to promote abiological process. The mammalian cell uses compartmen-talization to organize intranuclear operations, with subcom-partments associated with various cellular processes. Themajority of the nucleus is composed of the chromatin com-partment, which is subdivided into transcriptionally activeeuchromatin or transcriptionally silent heterochromatin. In-dividual chromosomes occupy defined regions within the

* Corresponding author. Fax: �1-617-432-0223.E-mail address: [email protected] (D.M. Knipe).

R

Available online at www.sciencedirect.com

Virology 309 (2003) 232–247 www.elsevier.com/locate/yviro

0042-6822/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved.doi:10.1016/S0042-6822(03)00107-7

nucleus called “chromosome territories,” with euchromatinmore centrally located within the nucleus and the hetero-chromatin located near the nuclear periphery (Abney et al.,1997; Carmo-Fonseca, 2002; Chubb et al., 2002; Ferreira etal., 1997; Manders et al., 1999; Qumsiyeh, 1999; Sadoni etal., 1999; Zink et al., 1998).

Running throughout the interchromosomal spaces arechannels or streams of nucleoplasm in which processesinvolved in RNA metabolism and trafficking are found(Politz and Pederson, 2000). Other nuclear compartmentswithin the interchromosomal spaces include the nucleolus,which is the major site of ribosome biogenesis (Carmo-Fonseca et al., 2000), the nuclear speckles that house spli-ceosomal components (Misteli et al., 1997), the Cajal bod-ies (formerly known as coiled bodies) that are involved inthe generation of snRNPs (Gall, 2000), and the ND10 sites,which may serve as depots for transcriptional regulators(Ascoli and Maul, 1991; Everett et al., 1999; Hodges et al.,1998; Maul et al., 2000; Zhong et al., 2000). Many of thesestructures, such as chromatin and the ND10 sites, are an-chored within the nucleus by direct association with thenuclear matrix (Ascoli and Maul, 1991; Chubb et al., 2002;Hodges et al., 1998), while others are associated with spe-cific genomic loci, as is the case with Cajal bodies andhistone gene clusters (Gall, 2000). Within this well-orga-nized intranuclear environment, the virus must initiate itslife cycle by accumulating the required cellular and viralfactors and the incoming parental genome at the same sitewithin the host nucleus. The assembling of these sites isinfluenced and possibly limited by the preexisting nucleararchitecture (de Bruyn Kops and Knipe, 1994). For exam-ple, in binucleate CV-1 cells infected with HSV, similarpatterns of replication compartments, some even mirrorimages, were observed in the adjacent sister nuclei (deBruyn Kops and Knipe, 1994).

Replication compartments were initially defined by thepresence of the HSV single-stranded DNA-binding protein,ICP8 (Quinlan et al., 1984). Other viral proteins known toaccumulate within replication compartments include the or-igin-binding protein (UL9), the heterotrimeric helicase–pri-mase complex (UL5/UL8/UL52), polymerase and its acces-sory factor (UL30/UL42), the immediate-early ICP27protein, the ICP4 viral transactivator protein, and the ICP5capsid protein (Bush et al., 1991; de Bruyn Kops et al.,1998; Knipe et al., 1987; Liptak et al., 1996; Lukonis andWeller, 1996; Phelan et al., 1997). Cellular proteins that areinvolved in DNA replication, such as PCNA, DNA ligase,and RP-A, are also recruited to replication compartments(de Bruyn Kops and Knipe, 1988; Wilcock and Lane, 1991),but their participation in viral replication has not been de-termined. The above-mentioned viral proteins are containedwithin subcompartments within replication compartments,with some proteins, such as ICP8, distributed to punctatefoci within compartments, and others, such as ICP4, distrib-uted diffusely within compartments (de Bruyn Kops et al.,1998). Viral DNA synthesis, as detected by BrdU incorpo-

ration (de Bruyn Kops and Knipe, 1988), and viral tran-scription, as detected by RNA pulse-labeling and by thepresence of the cellular RNA polymerase II (Leopardi et al.,1997; Phelan et al., 1997; Rice et al., 1994), both occurwithin replication compartments.

In the absence of viral replication, as with the treatmentwith the viral DNA synthesis inhibitor phosphonoaceticacid (PAA), the viral replication proteins including ICP8,the origin-binding protein, and the helicase–primase com-plex are distributed within the nucleus to punctate structurescalled prereplicative sites (de Bruyn Kops and Knipe, 1988;Liptak et al., 1996; Lukonis and Weller, 1996; Quinlan etal., 1984). Because the replication proteins associate withthese structures in the absence of viral DNA synthesis, theyare hypothesized to be an early stage of protein accumula-tion and, thus, the precursors to replication compartments.However, in the presence of PAA, two populations of pre-replicative sites are observed (Lukonis et al., 1997; Upri-chard and Knipe, 1997). One population, the S-phase-de-pendent prereplicative sites, co-localize with sites ofcellular DNA synthesis, as shown by BrdU incorporation(Uprichard and Knipe, 1997). In contrast, the other popula-tion, the S-phase-independent prereplicative sites, form ad-jacent to cellular ND10 sites (Uprichard and Knipe, 1997).The function of ND10 sites is unknown, but it has beensuggested that they are involved in processes as varied astranscriptional control, cell growth and regulation, and an-tiviral defense (Ascoli and Maul, 1991; Maul et al., 2000;Negorev and Maul, 2001; Zhong et al., 2000). The “nucleardomain 10” or ND10 structures vary in number in mamma-lian cells in the range of 10–30 per nucleus, and are gen-erally 0.2–1.0 �m in diameter (Everett et al., 1999; Hodgeset al., 1998; Zhong et al., 2000). The size and number ofND10 structures change during the cell cycle, which varieswith the presence of hormones or interferon, possibly via apathway regulated by the promyelocytic leukemia protein(PML) (Everett et al., 1999; Hodges et al., 1998; Negorevand Maul, 2001; Zhong et al., 2000). While ND10 structureshave been found in numerous cell types in cultured cells orin vivo, they have yet to be detected in cells of a neuronallineage (Negorev and Maul, 2001). The role that ND10structures play in viral infection is also unclear as thesestructures are disrupted soon after infection by the viralimmediate-early gene product ICP0, possibly via a protea-some-dependent mechanism (Everett et al., 1998; Everettand Maul, 1994). It has previously been shown that incom-ing genomes of various viruses including HSV, SV40, andadenovirus are deposited near ND10 structures in the ab-sence of viral protein production via an unknown mecha-nism (Ishov and Maul, 1996; Maul et al., 1996). However,the cellular machinery that one might imagine that a viruswould require to replicate its genome is not present at thesesites and must be recruited to the parental genome(s). Incontrast, the S-phase-dependent sites are regions of activecellular DNA synthesis and presumably contain the cellularmachinery needed for viral DNA synthesis (de Bruyn Kops

233T.J Taylor et al. / Virology 309 (2003) 232–247

and Knipe, 1988; Uprichard and Knipe, 1997). It is possiblethat the virus may usurp these preformed replication com-plexes for its own purposes and that these sites are theprecursors of viral replication compartments.

ICP8 is a 125-kDa multifunctional zinc metalloproteinencoded by the UL29 gene (Gao et al., 1988; Gupte et al.,1991; Quinn and McGeoch, 1985) and is expressed withearly kinetics (Honess and Roizman, 1974). ICP8 exhibits apreference for binding single-stranded DNA, but can binddouble-stranded DNA (Lee and Knipe, 1985; Ruyechan,1983; Ruyechan and Weir, 1984). ICP8 binds nonspecifi-cally to DNA in a cooperative manner (Lee and Knipe,1985). ICP8 also has the ability to destabilize double-stranded DNA (Boehmer and Lehman, 1993a) and can pro-mote DNA strand transfer (Bortner et al., 1993) and rena-turation of complementary DNA strands (Dutch andLehman, 1993), which suggests that ICP8 may play a role inHSV genome recombination. In addition to its role in viralDNA synthesis, ICP8 has also been shown to regulate viralgene expression by repressing transcription from the paren-tal genome (Godowski and Knipe, 1983; Godowski andKnipe, 1985; Godowski and Knipe, 1986) and stimulatinglate gene expression from progeny genomes (Chen andKnipe, 1996; Gao and Knipe, 1991).

The requirements for the targeting of ICP8 within thenucleus may depend on the conditions examined. In trans-fected cells, the co-expression of the three components ofthe helicase–primase complex (UL5, UL8, and UL52) issufficient to localize ICP8 to punctate prereplicative sites(Liptak et al., 1996; Lukonis and Weller, 1996). In infectedcells, the requirements for ICP8 intranuclear localizationmay be different because, in addition to the helicase–pri-mase complex, the origin-binding protein (UL9) is alsorequired (Liptak et al., 1996; Lukonis and Weller, 1996).ICP8 has been shown to interact either physically orfunctionally with several of the other viral replication pro-teins including the origin-binding protein (Boehmer, 1998;Boehmer and Lehman, 1993b; He, 2001; Lee and Lehman,1997; Makhov et al., 1996), the helicase–primase complex(Falkenberg, Elias, and Lehman, 1998; Hamatake et al.,1997; Le Gac, 1996), and polymerase (Chiou et al., 1985;Hernandez and Lehman, 1990; O’Donnell et al., 1987;Ruyechan and Weir, 1984). As ICP8 is required for prerep-licative site and replication compartment formation (deBruyn Kops and Knipe, 1988; Liptak, Uprichard, andKnipe, 1996; Quinlan, Chen, and Knipe, 1984), and becauseits intranuclear distribution is well characterized, we choseICP8 as a marker for viral replication structures and to focuson the intranuclear movement of ICP8 as a model system tostudy replication compartment formation in infected cells.To achieve this, we fused the coding sequence from greenfluorescent protein (GFP) to the C-terminal ICP8 sequenceto analyze the formation of prereplicative sites and replica-tion compartments in living cells.

Results

The number of viral functions occurring within replica-tion compartments speaks to the significance of these struc-tures in the viral replication cycle, but little is known abouthow these structures form and mature to occupy almost theentire volume of the nucleus (Monier et al., 2000). Weinitiated these studies to determine if replication compart-ments originated from a few distinct structures in the nu-cleus or if they formed from the coalescence of numeroussites within the nucleus. It is likely that the few S-phase-independent prereplicative sites observed adjacent to cellu-lar ND10 sites (Uprichard and Knipe, 1997) and not thenumerous S-phase-dependent prereplicative sites are theprecursors to the large replication compartments observedlater during infection (de Bruyn Kops and Knipe, 1988),based on the evidence that the incoming viral genome andviral proteins accumulate at sites adjacent to ND10 sites(Ishov and Maul, 1996; Maul et al., 1996; Uprichard andKnipe, 1997). However, it was unclear how these few smallsites eventually form the large replication compartmentsobserved later in infection. To analyze this, we constructeda recombinant HSV-1 strain expressing an ICP8–GFP fu-sion protein to examine prereplicative site and replicationcompartment formation in cultured cells.

Construction of an ICP8–GFP fusion gene

We first constructed vectors that expressed an ICP8 mol-ecule with GFP fused to the C- or N-terminus. The C-terminal ICP8–GFP fusion protein consists of the entire1196 amino acid residues of ICP8, an 18-amino-acid residuelinker region, and the 239 amino acid residues of GFP. Thepredicted size of the C-terminal ICP8–GFP fusion proteinwas 157 kDa. The N-terminal GFP–ICP8 protein consistedof the 239 residues of GFP fused to the 1196 residues ofICP8 with 3 residues as a linker region. We constructedplasmids that expressed a C-terminal ICP8–GFP fusionprotein under the control of the CMV promoter pCMV-8GFP (Fig. 1A), and an N-terminal ICP8–GFP fusion pro-tein under the control of the CMV promoter pCMV-GFP8(Fig. 1A) or the ICP8 promoter p8-8GFP (Fig. 1B).

Localization of the ICP8–GFP protein in transfected cells

When expressed in the absence of other viral proteins intransfected Vero cells, the ICP8–GFP fusion proteinshowed a diffuse distribution within the nucleus (Fig. 2A).The ICP8–GFP fusion protein localized to viral replicationcompartments after infection with the replication-defectiveHD-2 ICP8 mutant virus (Fig. 2C), which expresses theN-terminal 250 amino acid residues of ICP8 fused in-frameto lacZ. When the specific viral DNA synthesis inhibitorPAA was present during HD-2 infection, the ICP8–GFPfusion protein localized to punctate prereplicative sites (Fig.2E). Similar results were observed with wild-type KOS 1.1

234 T.J Taylor et al. / Virology 309 (2003) 232–247

infection of transfected Vero cells (data not shown). Theaddition of GFP to the C terminus of ICP8 did not alter itslocalization to prereplicative sites or replication compart-ments in transfected Vero cells infected with HSV. There-fore, the ICP8–GFP fusion protein behaved as describedpreviously for wild-type ICP8 under similar conditions. Theaddition of GFP to the N terminus of ICP8 resulted in aprotein that did not localize to replication compartments(data not shown). The N-terminal ICP8–GFP fusion proteinhad a tendency to aggregate nonspecifically in the nuclei oftransfected cells (data not shown). For these reasons, weused only the C-terminal ICP8–GFP fusion protein for ourexperiments.

Complementation of the HD-2 mutant virus

We next tested the ability of the C-terminal ICP8–GFPfusion protein to complement the HD-2 ICP8 mutant virus,which expresses the first 250 amino acid residues of ICP8fused to lacZ. Vero cells transfected with the empty pCIvector plasmid or plasmids expressing either wild-typeICP8 or the ICP8–GFP fusion proteins were infected withHD-2 virus. At 24 h postinfection, progeny virus was har-vested, and the viral yield was determined on an ICP8-complementing cell line. The ICP8–GFP fusion proteincould complement the growth of HD-2 in Vero cells by50-fold over the negative control, but this was 40-fold lessthan the level observed with the wild-type ICP8 protein(Table 1). These results suggested that the addition of GFPto the C terminus of ICP8 altered one or more functions ofICP8 or that the ICP8–GFP fusion protein was not ex-pressed to the levels required for efficient complementation

of the HD-2 virus. The N-terminal fusion protein failed tocomplement an ICP8 mutant virus (Taylor, 2002).

Construction of an ICP8–GFP recombinant virus

We constructed a recombinant HSV strain expressing theICP8–GFP fusion protein via homologous recombination.Purified HD-2 viral DNA and linearized pICP8–GFP plas-mid DNA were co-transfected into the V827-complement-ing cell line. At 4 days posttransfection, the progeny viruswas harvested and plated onto V827 cells. Fluorescentplaques were identified and triply purified on V827 cells.Stocks were grown on ICP8-complementing cell lines.

Fig. 2. Localization of the C-terminal ICP8–GFP fusion protein in trans-fected cells. Vero cells were transfected with the p8-8GFP plasmid andsubsequently mock-infected or infected with the HD-2 ICP8 mutant virus.Images were obtained at 6 h postinfection. (A) Mock-infected; (B) phase of(A); (C) HD-2 infected; (D) phase of (C); (E) HD-2-infected in thepresence of PAA; (f) phase of (E).

Fig. 1. Diagram of the plasmids used in this study. (A) pCMV–8GFP andpCMV-G8 plasmids, which express the C- (ICP8–GFP) and N-terminal(GFP–ICP8) fusion proteins, respectively. (B) Top: HSV-1 genome, withthe location of the UL29 (ICP8) gene indicated. Bottom: p8-8GFP plasmidthat was used to generate the recombinant ICP8-GFP virus.

Table 1Complementation of the HD-2 ICP8 mutant virus with ICP8–GFPfusion protein

DNA transfecteda Yield (PFU/ml)b

pCI plasmid vector 1.4 � 102, 1.4 � 102

pCMV–ICP8 3.1 � 105, 5.6 � 105

pCMV–ICP8–GFP 8.0 � 103, 5.6 � 103

a Three micrograms of DNA was transfected and, at 24 h posttransfec-tion, the cells were infected with HD-2 virus.

b Yield from duplicate samples determined on an ICP8-complementingcell line.


Detection of the ICP8–GFP fusion protein in infectedcells

The ICP8–GFP fusion protein was detected in [35S]me-thionine- and cysteine-labeled infected Vero cells at 3.5,6.5, and 9.5 h postinfection (Fig. 3A, lanes 3, 5, and 7).While the ICP8–GFP was expressed with kinetics similar tothose of wild-type ICP8, it was expressed to lower levels.

This may have been the result of lower levels of synthesis ordue to a less stable protein as a result of the addition of GFP.When the fusion protein was detected by Western blot ofinfected Vero cell lysates with the 3–83 anti-ICP8 poly-clonal antibody, it migrated near the approximate predictedmolecular size of 157 kDa (Fig. 3B, lane 2). The levels ofICP8–GFP expressed in ICP8-complementing cells weresimilar to those observed in Vero cells, suggesting that thepresence of wild-type ICP8 does not stabilize or otherwiseincrease the expression of the ICP8–GFP fusion protein(data not shown).

Growth properties of the ICP8–GFP recombinant virus

To determine if there was a growth defect associatedwith the addition of GFP to the C terminus of ICP8, weexamined the kinetics of ICP8–GFP viral growth in bothICP8-complementing V529 cells and noncomplementingVero cells (Fig. 4). Monolayers of cells infected at a m.o.i.of 3 were harvested at various times postinfection. The viralyield was determined on the ICP8-complementing cell lineV529, and the ICP8–GFP recombinant virus replicated onVero cells, but the yield was increased by 150-fold on theV529-complementing cell line (Fig. 4). The growth differ-ence of the ICP8–GFP recombinant virus may be due to thelower levels of ICP8–GFP produced during infection com-pared with wild-type ICP8 (Fig. 3A). It was possible that themajority of Vero cells infected with the ICP8–GFP recom-binant virus did not produce progeny virus and that only asmall percentage of infected cells produced progeny, thusgiving rise to a much lower yield. If only a few Vero cellswere producing the bulk of the progeny virus, it would beexpected that only those few Vero cells would containreplication compartments. However, the number of Verocells containing replication compartments was not dramat-

Fig. 3. Analysis of protein expression in Vero cells. (A) At the indicatedtimes postinfection, Vero cells were labeled for 30 min with [35S]methi-onine and harvested, and the proteins were separated by SDS–PAGE on a9.25% gel. Vero cells were mock-infected (lane 1) or infected with wild-type KOS 1.1 (lanes 2, 4, 6) or the ICP8–GFP virus (3, 5, 7). The positionsof several viral proteins are shown at the right. (B) Western blot analysisof lysates from Vero cells infected either with wild-type KOS1.1 (lane 1)or the ICP8–GFP (lane 2) virus.

Fig. 4. Viral growth curve. Vero or V529 cells were infected with either wild-type (wt) or ICP8–GFP (8GFP) virus. At 2, 12, and 24 h postinfection, thecells were harvested, and the yields of duplicate samples were determined by titration on V529 cells.


ically different from the number of V529 cells containingreplication compartments (Table 2), but the replicationcompartments were smaller and in greater number in Verocells. Taken together, these results suggested that the ma-jority of Vero cells are replicating viral DNA and producingprogeny virions, but possibly to a lower extent than in thecomplementing cell line. Large stocks of the ICP8–GFPrecombinant virus at titers greater than 108 PFU/ml weregrown on an ICP8-complementing cell lines. Thus the ad-dition of GFP to the C terminus of ICP8 did affect somefunction(s) required for viral replication. The presence ofGFP at the C terminus may be acting in a steric manner bypartially blocking an interaction required for viral replica-tion. However, the phenotype defect of ICP8–GFP could becomplemented by wild-type ICP8.

Viral DNA synthesis by the ICP8–GFP virus

ICP8 is required for viral DNA synthesis, so the growthdefect observed in Vero cells may be due in part to de-creased levels of DNA synthesis in Vero cells. We infectedVero cells with either wild-type virus or the ICP8–GFPrecombinant virus in the presence and absence of the spe-cific viral DNA synthesis inhibitor PAA. At 16 h postinfec-tion, the cells were harvested, and the total cellular DNAwas isolated. We blotted the DNA in fivefold dilutions to amembrane and probed with an HSV-specific 32P-labeledplasmid to quantify the relative amounts of viral DNA byPhosphoImager analysis. The values were corrected for theamount of input viral DNA, as measured by the signal fromthe PAA-infection samples. The amount of viral DNA at16 h postinfection in Vero cells infected with the ICP8–GFP virus was 3% of the level of DNA synthesized inwild-type virus-infected Vero cells (Fig. 5), as measured byPhosphoImager analysis. Thus, a defect in viral DNA syn-thesis was at least partially responsible for the growth dif-ferences observed in Vero and ICP8-complementing cells.

UL9 binding assay

To determine if the decrease in viral replication was dueto the GFP moiety blocking an important protein interactionwith ICP8, we tested the ability of the ICP8–GFP fusionprotein to bind to a Staphylococcus aureus protein A–UL9fusion protein that had previously been used to map the UL9sites required for ICP8 binding (Arbuckle and Stow, 1993;Boehmer et al., 1994). As a negative control, we used amutant protein A–UL9 fusion protein, D27, which was in-capable of binding to ICP8 (Boehmer et al., 1994).

Approximately 90% of the wild-type ICP8 bound to theprotein A–UL9 fusion protein (Fig. 6, lane 6), as comparedwith 10% of the ICP8–GFP fusion protein (Fig. 6, lane 3),as measured by densitometry. Neither the wild-type ICP8nor the ICP8–GFP protein was retained by the mutant D27protein A–UL9 fusion protein (data not shown). This indi-cated that the ICP8–GFP fusion protein was defective forbinding to UL9, possibly because the binding site(s) may beoccluded by the GFP moiety.

Characterization of intranuclear structures containingICP8–GFP

We first compared the distribution of ICP8 and ICP8–GFP in Vero cells and ICP8-complementing V529 cells

Table 2Comparison of replication compartments in Vero and V827 cells

Number of replicationcompartmentsa

% of cells with indicated numberof replication compartmentsb

V827 Vero

1 or 2 42 173 or 4 33 22�5 3.5 37Diffuse 9.0 9.0Punctate 12 15

a The numbers of replication compartments were determined at 7.5 hpostinfection with the recombinant ICP8–GFP virus. Cells contained arange of compartments with 1 or 2 large compartments to 5 or more smallcompartments.

b Fifty-seven V827 cells were counted; 54 Vero cells were counted.

Fig. 5. Viral DNA synthesis at 16 h postinfection. Vero cells were mock-infected (M), or infected with either wild-type KOS1.1 (WT) or therecombinant ICP8-GFP virus (8GFP) in the presence or absence of PAA.Total DNA was isolated, vacuum-blotted to a filter in five-fold dilutions,and probed with an HSV-1-specific 32P-labeled probe.

Fig. 6. Binding of ICP8 or ICP8–GFP to a protein A-UL9 fusion proteinpreabsorbed to IgG Sepharose. Protein extracts obtained from transfected293 cells were mixed batchwise with a protein A-UL9 fusion proteinimmobilized to IgG Sepharose. After the unbound fraction (FT) wascollected, the IgG Sepharose was washed with 100 mM NaCl low-saltbuffer (LS) and then 1 M NaCl high-salt buffer (HS). A portion of eachfraction (1/10 volume) was separated by SDS–PAGE, blotted to mem-brane, and probed with the 3-83 polyclonal anti-ICP8 antibody.


using standard immunofluorescence techniques. ICP8 andICP8–GFP were found in similar globular structures inVero cells (Figs. 7A and C, respectively) and in V529 cells(Figs. 7B and E, respectively), suggesting that the structuresformed by ICP8–GFP in Vero and V529 cells were repli-cation compartments.

Wild-type ICP8 reacts with the 39S conformation-spe-cific monoclonal antibody only when it is localized to pre-replicative sites or replication compartments (S.L. Upri-chard, D.M. Knipe, submitted for publication). As theaddition of GFP may have altered the conformation of ICP8,we used immunofluorescence to determine if the ICP8–GFPfusion protein contained within replication compartmentswas detected with the 39S monoclonal antibody. We in-fected Vero cells with recombinant ICP8–GFP virus, and at5.5 h postinfection we processed the cells for immunofluo-rescence. ICP8–GFP was detected by the 39S monoclonalantibody (Fig. 8B), suggesting that the addition of GFP didnot significantly alter the conformation of ICP8 and that theICP8–GFP fusion protein adopted a conformation similar tothat of wild-type ICP8 within replication compartments.

To further demonstrate that the ICP8–GFP-containingstructures in the nucleus were similar to wild-type replica-tion compartments and not nonspecific aggregates of theICP8–GFP fusion protein, we determined if other viralreplication compartment proteins were present in infectedVero cells within the ICP8–GFP-containing structures viaindirect immunofluorescence. Both the major viral transac-tivator ICP4 (Fig. 8D) and the polymerase processivityfactor UL42 (Fig. 8G), previously shown to localize toreplication compartments (de Bruyn Kops et al., 1998;Knipe et al., 1987; Liptak et al., 1996; Randall and Din-woodie, 1986), co-localized with the ICP8–GFP fusion pro-tein in replication compartments within the nucleus (Figs.8F and I, respectively). Therefore, the ICP8–GFP-contain-ing structures were similar in protein composition to wild-type replication compartments and not simply aggregates ofthe ICP8–GFP fusion protein; thus, ICP8–GFP localized tonormal replication structures in infected cells. These condi-tions provided a system to study ICP8 localization in livingcells.

ICP8-GFP distribution in fixed cells

We next observed replication compartment formation inVero cells and the ICP8-complementing cell line V827 at 4and 6.5 h postinfection with the recombinant ICP8–GFPvirus. At 4 h postinfection, the majority of ICP8–GFP wasdistributed diffusely within the nucleus, but several smallstructures similar to prereplicative sites were observed (Fig.9A). These small sites were composed of only one or twofoci that contained the ICP8–GFP fusion protein (Fig. 9G).In the V827-complementing cells, the infection in many of

the cells had progressed beyond the individual foci or dou-blets and these cells contained what appeared to be smallcompartments each consisting of several ICP8–GFP foci(Fig. 9B).

At 6.5 h postinfection, larger compartments were ob-served in both the Vero and the complementing V827 cells(Figs. 9C and D, respectively). The smaller compartmentsize observed in the Vero cells compared with the V827cells may be due in part to the lower level of ICP8–GFPviral DNA synthesis in Vero cells or the lower levels ofICP8–GFP fusion protein compared with wild-type ICP8protein. In both cell types, when PAA was added, theexpected punctate prereplicative sites were observed (Figs.9E and F). Because the replication compartments werelarger in the ICP8-complementing cell line, we used thesecells in all future experiments for visualizing replicationcompartment formation.

Time-lapse microscopy of replication compartmentformation

To follow the maturation of replication compartments incultured cells, we infected V827 cells with the ICP8–GFPrecombinant virus and acquired images at various intervalsof time. Intranuclear movement was observed on two scales.On a small scale, the individual foci containing ICP8–GFPthat made up the replication compartments were not static,but appeared to rotate in place or move relatively shortdistances (less than the diameter of the foci) within a timeframe of seconds (not shown). In addition, on a larger scale,small replication compartments moved through the nucleusto merge or fuse with neighboring structures. This move-ment took place over a time frame of minutes and greaterintranuclear distances (over 1 �m). At 4.5 h postinfection,the majority of ICP8–GFP was distributed diffusely in thenucleus, with only a few punctate prereplicative sites (Fig.10A) or small replication compartments (Fig. 10B) ob-served. The three arrows in Fig. 10A at 4 h 12 min postin-fection indicate three structures that coalesced to form thelarger compartment indicated with the arrow by 5 h 30 minpostinfection. In other cells, small replication compartmentsincreased in size until merging with neighboring structures(Fig. 10B). Large replication compartments merged withneighboring structures to eventually occupy the entire nu-cleus (Fig. 11). The movie versions of the data may be seenon our web site at http://knipelab.med.harvard.edu.

Discussion

ICP8-GFP fusion protein

In this study, we describe the construction and use of arecombinant HSV-1 strain that expresses an ICP8–GFPfusion protein to examine the dynamic intranuclear move-ment of ICP8 into prereplicative sites and replication com-


http://knipelab.med.harvard.edu.

Fig. 7. Comparison of ICP8 and ICP8–GFP distribution patterns. Vero or V529 cells were infected with either wt or the ICP8–GFP recombinant virus. At5.5 h postinfection, the cells were processed for immunofluorescence. The 39S MAb was used to detect both ICP8 and ICP8–GFP with a rhodamine-conjugated secondary antibody. The ICP8–GFP fusion protein was also detected by GFP fluorescence. (A) Distribution of ICP8 in a Vero cell; (B) distributionof ICP8 in a V529 cell; (C) distribution of ICP8–GFP in a Vero cell as detected by the 39S MAb; (D) GFP fluorescence of (C), (E) distribution of ICP8–GFPin a V529 cell as detected by the 39S MAb, (F) GFP fluorescence of (E).Fig. 8. Reactivity of the ICP8–GFP fusion protein with the 39S conformation-specific monoclonal antibody and co-localization with other viral replicationproteins. Vero cells were infected with the ICP8–GFP virus and processed for immunofluorescence as described under Material and Methods. (A) ICP8–GFPdetected by the 39S MAb; (B) ICP8–GFP detected by GFP fluorescence; (C) merge; (D) ICP4 detected by the 58S MAb; (E) ICP8–GFP detected by GFPfluorescence; (F) merge; (G) UL42 detected by the 5H11D6 MAb; (H) ICP8–GFP detected by GFP fluorescence; (I) merge.


partments. The addition of GFP to the carboxy terminus ofICP8 resulted in a fusion protein that was partially func-tional; however, the defect could be complemented by thepresence of wild-type ICP8 supplied in trans by infectionwith wild-type HSV or by an ICP8-expressing cell line. It ispossible that the GFP portion blocks or restricts a region ofICP8 that is required for the interaction with viral or cellularproteins or that lower levels of ICP8–GFP made duringinfection may contribute to the observed growth defect. Ifthis is the case, then the decreased level of ICP8–GFP maynot be sufficient to perform all the required activities of

ICP8 during infection. The presence of wild-type ICP8 incomplementing cells may stimulate viral DNA synthesis oract in concert with the ICP8–GFP fusion protein to recruitviral or cellular proteins required for replication or it maysimply increase the overall amount of ICP8 expressed dur-ing the infection. The growth defect in Vero cells could beaccounted for in part by a 30-fold decrease in viral DNAsynthesis, which was probably due in part to the greater than10-fold decrease in the ability of the ICP8–GFP fusionprotein to bind the UL9 origin-binding protein. UL9 is re-quired for origin-dependent viral DNA synthesis (Chall-berg, 1986), and it has been shown previously that theICP8–UL9 interaction is required for efficient origin-depen-dent DNA synthesis (Boehmer et al., 1994). In an insectcell-based, HSV-1 origin-dependent DNA replication assay,the UL9 mutant UL9DM27, which no longer interacts withICP8 but retains origin binding and helicase activities, dis-played only 5% of the levels of origin-dependent DNAsynthesis compared with wild-type UL9 controls (Boehmeret al., 1994). This level of DNA synthesis is comparable tothe amount that we observed after infection of Vero cellswith the ICP8–GFP virus, indicating that the decreasedbinding of UL9 by ICP8–GFP may account for the lowerlevels of DNA replication observed in Vero cells. It ispossible that the GFP portion blocks or restricts a region ofICP8 that is required for the interaction with UL9. In ICP8-complementing cells infected with the recombinant ICP8–GFP virus, the wild-type ICP8 may interact with UL9, thusinitiating viral DNA synthesis. This does not imply that theonly interaction disrupted by the addition of GFP is withUL9; moreover, the GFP moiety may block interactions in alarger replication complex that includes ICP8 or other in-teractions that have not yet been described. Despite thelowered level of viral DNA replication detected in infectedVero cells, prereplicative sites and replication compart-ments were observed, suggesting that the decreased level ofDNA synthesis was sufficient to promote replication com-partment formation. This suggests that the ICP8–UL9 inter-action may not be directly required for ICP8 intranucleartargeting and for the formation of replication compartments.However, it is possible that the weaker binding of ICP8–GFP to UL9 may be sufficient for intranuclear localizationto prereplicative sites. Previous studies have shown that intransfected Vero cells, co-expression of the three subunits ofthe helicase–primase complex with wild-type ICP8 wassufficient to localize ICP8 to prereplicative sites (Liptak etal., 1996), suggesting, that in certain circumstances, UL9 isnot absolutely required for the intranuclear localization ofICP8.

The intranuclear structures containing the ICP8–GFPfusion protein had properties similar to those previouslyreported for wild-type ICP8-containing replication compart-ments. For example, the ICP8–GFP fusion protein co-lo-calized with the ICP4 and UL42 proteins, both of which

Fig. 9. Distribution of the ICP8–GFP fusion protein at various timespostinfection in fixed cells. Vero cells (A, C, E) or V827 cells (B, D, F, G)were infected with the recombinant ICP8–GFP virus and processed at theindicated times postinfection. (A) Vero cell at 4 h postinfection (1000�mag.); (B) V827 cell at 4 h postinfection (1000� mag.); (C) Vero cells at6.5 h postinfection (630� mag.); (D) V827 cells at 6.5 h postinfection(630� mag.); (E) same as (C) in the presence of PAA; (F) same as (D) inthe presence of PAA. (G) Closeup of ICP8–GFP containing doublets at 4 hpostinfection.


localize to replication compartments. The ICP8–GFP pro-tein intranuclear distribution was also similar to that ob-served for wild-type ICP8, which consisted of punctate foci

within the boundaries of the replication compartment. Thesedata supported the use of the ICP8–GFP protein as a modelfor intranuclear targeting of ICP8.

Fig. 10. Small replication compartments merge to form larger compartments. Time-lapse analysis of replication compartment formation in cultured cells.V827 cells were infected with the ICP8–GFP virus. At 4 h postinfection, images were acquired at 3-min intervals. The composite images shown weregenerated using selected images from the indicated times postinfection (hour:minute). (A) The three arrows at 4:12 point to three structures that merged toform the larger compartment indicated with an arrow at 5:30. (B) The three arrows at 4:30 show three replication compartments that enlarged to form thecompartment indicated with the arrow at 4:54.

Fig. 11. Replication compartments fuse with neighboring structures to fill the nucleus. Time-lapse analsysis of replication compartment formation in culturedcells. V827 cells were infected with the ICP8–GFP virus. At 5.5 h postinfection, images were acquired at 3-min intervals. The composite image shown wasgenerated using selected images from the indicated times postinfection (hour:minute).


Intranuclear movement of viral replication structures

We show here that the HSV prereplicative sites andreplication compartments initiate at a few distinct structuresand then at least some traverse through the intranuclearenvironment to fuse with one another. Early during infec-tion, the ICP8–GFP fusion protein localized to severalsmall structures within the nucleus. It is unknown if theICP8–GFP fusion protein is targeting to a preexisting cel-lular structure that is adjacent to ND10 or if viral proteinsaccumulate at a new intranuclear structure required for viralDNA replication or gene expression. It has recently beenreported that cellular sites of DNA synthesis initiate at a fewsites within the nucleus (Kennedy et al., 2000). It is possiblethat viral proteins, including ICP8–GFP, localize to a sub-population of cellular replication sites that form adjacent toND10. Further studies are required to determine if ICP8 orother viral proteins co-localize with mammalian origin-associated proteins, such as those in the ORC or MCMfamilies. If this is true, then, it is possible that the virus isusurping preformed cellular DNA replication complexes toaid in its own DNA replication.

The coalescence of smaller, distant sites or the mergingof neighboring replication compartments formed largercompartments that eventually filled the nucleus. We ob-served movement on two scales. The first was a rapid,small-scale local movement of individual ICP8–GFP fociwithin replication compartments that occurred over shortdistances. The second type involved movement over greaterintranuclear distances but on a slower time frame on theorder of minutes. To summarize, once prereplicative sites orsmall compartments formed, large replication compart-ments formed by two mechanisms: (1) distant, small struc-tures moved to coalesce into larger compartments (Fig.12A), and (2) the compartments grew by accretion whilemaintaining the same relative position within the nucleusand larger compartments formed when the boundaries ofadjacent compartments merged (Fig. 12B).

Multiple mechanisms may be responsible for replicationcompartment movement and coalescence. To address this,we must first consider the types of intranuclear movementsobserved normally in cells. Many studies suggest that chro-matin movement is restrained during interphase in the mam-malian nucleus, with movements restricted to �0.3–0.5�m, which is the result of Brownian motion or diffusion(Abney, 1997; Manders et al., 1999; Marshall et al., 1997;Shelby et al., 1996; Zink et al., 1998). These mechanismsmay account for the rapid, small-scale movements of theICP8–GFP foci, but possibly not for the large-scale move-ments of replication compartments. Despite the evidence ofa relatively immobile mammalian cell nucleus, at least onereport has shown evidence for cell cycle-regulated intranu-clear chromatin movement (Tumbar and Belmont, 2001).This was demonstrated when a region of transcriptionallysilent chromatin located normally at the nuclear peripherywas permanently redistributed to the nuclear interior on

insertion of a transcriptionally active locus (Tumbar andBelmont, 2001). This large-scale movement took place on atime scale of minutes, but only during G1 or S phase. Moreevidence exists for cell cycle- and developmentally regu-lated intranuclear movement in other organisms such asDrosophila (Gunawardena and Rykowski, 2000; Vazquezet al., 2001) or budding yeast (Heun et al., 2001), but theseorganisms may have a functionally different nuclear archi-tecture compared with mammalian cells due to increasedheterochromatin amounts in mammalian cells, which maylimit intranuclear movements. The mechanisms required forthese examples of intranuclear movements are unknown,and it remains to be determined whether active or passiveforces are responsible.

It is intriguing to suggest that an active mechanism isresponsible for the movements of prereplicative sites andreplication compartments. It is possible that the virus mayassemble an intranuclear molecular motor that can propelcompartments through the interchromosomal spaces. Mo-tors consisting of actin and myosin are well documented inthe cytoplasm, but their presence or action in the nucleushas not been conclusively demonstrated. Recently, it wasproposed that actin filaments play a role in the activity of atranscriptional complex (Zhao, 1998), suggesting that anactin-dependent activity is required in the nucleus. Othershave demonstrated that nuclear myosin 1 interacted andco-localized with RNA polymerase II (Nowak et al., 1997;Pestic-Dragovich et al., 2000), again suggesting that molec-ular motors exist and have a functional role in the nucleus.Future work is required to determine if molecular motorproteins, such as actin and myosin, are required for viralreplication compartment coalescence. However replication

Fig. 12. Models for maturation of replication compartments. The twomodels of compartment maturation are shown. The nucleus is representedby the large oval. (A) Intranuclear movement and coalescence of smallercompartments with neighboring structures. (B) Growth of replication com-partments by accretion and the resulting fusion of adjacent structures.


compartment coalescence is directed, it is clear that this hasa profound effect on the preexisting nuclear architecture. Ithas been shown recently that herpesviruses may haveevolved mechanisms for disrupting the nuclear lamina (Mu-ranyi et al., 2002; Scott and O’Hare, 2001). The disruptionof nuclear laminins, which are an integral part of the nuclearframework, may lead to a higher nuclear plasticity, thusallowing for the movement of viral replication structures. InHSV-infected cells, other compartments within the nucleusmay be altered and the chromatin is marginalized to thenuclear periphery (Darlington and James, 1966; Monier etal., 2000; Schwartz and Roizman, 1969; Sirtori and Bosisio-Besteti, 1967). Channels of nucleoplasm that normally runthroughout the interchromosomal space to the nuclear poresmay be blocked or disrupted by the growth of replicationcompartments, thus possibly inhibiting or slowing normalcellular functions such as mRNA trafficking out to thecytoplasm. This may be another general mechanism for howHSV alters cellular functions to promote its own replicativesuccess.

Materials and methods

Cells and viruses

African green monkey kidney (Vero) and human 293cells were grown and maintained in Dulbecco’s modifiedEagle’s medium (Mediatech, Inc.) with 5% fetal bovineserum � 5% fetal calf serum, glutamine, streptomycin, andpenicillin (DMEM � 10% FCS). The ICP8-complementingcell lines S2 (Gao and Knipe, 1989), V827 (Da Costa et al.,2000), and V529 (Da Costa et al., 2000), which expressICP8, ICP8 and ICP27, and ICP8 and UL5, respectively,were grown and maintained in DMEM � 10% FCS � G418(400 �g/ml).

The herpes simplex 1 KOS 1.1 wild-type strain, origi-nally obtained from M. Levine, was propagated and assayedon Vero cells. The HD-2 mutant virus (Gao and Knipe,1989) is a replication-defective HSV-1 strain that expressesthe first 250 amino acid residues of ICP8 fused to �-galac-tosidase. The ICP8–GFP recombinant virus and the HD-2mutant ICP8 virus were propagated on either S2 or V827 cells.

Plasmids

To construct pICP8–GFP, sequences encoding ICP8 andgreen fluorescent protein (GFP) were first cloned into thepCI (Promega)-derived vector pCI�AflII (Murphy et al.,2000) as follows. A 732-bp NotI fragment containing theGFP open reading frame (ORF) from pGreenLantern(Gibco-BRL) was cloned into the NotI site of pCI�AflII togenerate pCI�A–GFP. A 3.8-kb EcoRI/AvrII fragment con-taining the entire ICP8 ORF from pSV8.2 (Gao and Knipe,1993) was cloned into the EcoRI and XbaI sites of pCI�A–GFP. The fusion protein ORF was generated by mutating

the ICP8 stop codon via polymerase chain reaction (PCR)-mediated site-directed mutagenesis with Pfu DNA polymer-ase (Stratagene) (1 cycle 95°C 2 min; 12 cycles 95°C 30 s,55°C 1 min, 68°C 18 min). The two primers used forchanging the stop codon to an arginine residue were 5�-CAACCCCTCTCAGCATATCCAACG-3� (antisense) and5�-CGTTGGATATGCTGAGAGGGGTTG-3� (sense) (LifeTechnologies). The altered nucleotide is in boldface. AfterPCR amplification, the reaction was digested with DpnI,and 1 �l of the reaction was used to transform bacteria. Theresulting vector was designated pCMV–8GFP.

The pCMV-8.3 plasmid was generated by removing the3.8-kb EcoRI/AvrII fragment containing the ICP8 sequencefrom the pSV8.3 plasmid (Gao and Knipe, 1991) and sub-cloning it into the EcoRI and XbaI sites of pCI�AfIII.

Isolation of a recombinant HSV-1 expressing theICP8–GFP fusion protein

The sequence encoding the C-terminal ICP8–GFP fusionprotein from pCMV–8GFP was cloned into a vector con-taining UL29 flanking sequences as follows. The EcoRI site172 nucleotides upstream from the ICP8 start codon and theHpaI site 147 nucleotides downstream of the GFP stopcodon in pCMV–8GFP were filled in and converted to BglIIrestriction sites by the insertion of a 12-bp BglII oligonu-cleotide linker (5�-GGAAGATCTTCC-3�) (New EnglandBiolabs). The resulting 4.7-kb BglII fragment was clonedinto the BglII site of pICP8PA (M. Chen, D.M. Knipe,unpublished results), a plasmid that contains the ICP8 pro-moter, poly(A), and additional downstream sequence, tomake the plasmid p8-8GFP. To generate the ICP8–GFPrecombinant virus via homologous recombination, linear-ized p8-8GFP and HD-2 viral DNA (Gao and Knipe, 1989)were co-transfected via the calcium phosphate method intothe V827-complementing cell line. HD-2 is a replication-defective virus that expresses a truncated ICP8 fused to the�-gal. Linear p8-8GFP was transfected with various molarratios of HD-2 viral DNA to plasmid DNA (1:10, 1:15,1:20) and brought to a total of 16 �g DNA with salmonsperm DNA. An equal volume of 2� HEPES-bufferedsaline, pH 7.05 (270 mM NaCl, 10 mM KCl, 1.4 mMNa2HPO4, 2% dextrose, 42 mM HEPES) was added to theDNA. The volume was brought up to 600 �l with 1�Hepes-buffered saline, and then 40 �l of 2.0 M CaCl2 wasadded dropwise. The tubes were mixed gently and incubatedat room temperature for 15 min. The suspension was thenadded to flasks of subconfluent V827 cells and incubated atroom temperature for 30 min, after which 5 ml of DMEM�10% FBS was added, and the flasks were placed at 37°C. Onthe following day, the cells were washed twice and overlainwith fresh medium. Four days later, the cells were harvestedby adding 2.5 ml sterile milk and freeze–thawing twice. Thelysate was sonicated three times on ice for 30 s on/ 30 s off.The lysate was then serially diluted and plated on V827cells. An inverted, fluorescent Nikon microscope equipped


with a FITC filter set was used to screen for fluorescentplaques.

Complementation assay

Vero cells in six-well plates were transfected in duplicateusing the Lipofectamine reagent (Gibco-BRL) with 1 �g ofpCMV-8.3 or pCMV–8GFP and 3 �g carrier DNA, or 4 �gcarrier DNA as a negative control. At 24 h posttransfection,the cells were infected at a m.o.i. of 5 with HD-2. At onehour postinfection, the cells were washed (40 mM sodiumcitrate, 135 mM NaCl, 10 mM KCl, pH 3.0) twice for 30 sto inactivate excess extracellular virus. The cells were har-vested 24 h later by freeze–thawing twice and sonicatingthree times on ice for 30 s. The yield was determined bytitration on the S2-complementing cell line.

Viral yield assay

Vero or V529 cells in T25 flasks were infected at a m.o.i.of 3 with either wild-type KOS 1.1 or the recombinantICP8–GFP virus. One hour after infection the cells werewashed (40 mM sodium citrate, 135 mM NaCl, 10 mM KCl,pH 3.0) to inactivate extracellular virus. At 2, 12, or 24 hpostinfection, the cells were harvested by freeze–thawingand sonicating three times for 30 s on ice. The yield wasdetermined by titration on V529 cells.

Viral DNA synthesis assays

Vero cell monolayers were infected with either KOS1.1or the ICP8–GFP virus at a m.o.i. of 10 in the presence orabsence of a specific viral DNA inhibitor, phosphonoaceticacid (PAA), at a concentration of 400 �g/ml. At 16 hpostinfection, total DNA was obtained by disrupting thecells in 3 ml of lysis buffer (10 mM Tris–Cl, pH 8.0, 10 mMEDTA, pH 8.0, 2% SDS, 100 �g/ml proteinase K) andincubating overnight at 37°C. After the addition of 0.3 ml of3 M sodium acetate (pH 5.2), lysates were extracted oncewith phenol:chloroform and then once with chloroform.After precipitation with 2 vol of 70% ethanol and resuspen-sion in 700 of �l TE buffer (10 mM Tris–HCl, pH 8.0, 1mM EDTA), the lysate was incubated with RNase A (25�g/ml) for 30 min at 37°C. Seventy microliters of 3 Msodium acetate (pH 5.2) was added, and the samples wereonce again extracted with phenol:chloroform and then oncewith chloroform. The DNA was precipitated on ice for 1 hby the addition of an equal volume of 1.6 M NaCl–13%PEG. The DNA was then resuspended in TE buffer andquantitated. Equivalent amounts of DNA were applied to anitrocellulose filter in fivefold dilutions (2000–16 ng in12� SSC) via a dot-blotter apparatus. The filter was thenprobed with the HSV-1-specific 32P-labeled plasmidpSV8.3. Exposure and quantification of the filter were witha Bio-Rad GS-525 Molecular Imager and Multi-Analystsoftware.

Preparation of protein extracts

Protein extracts were obtained from two 150-cm2 cultureflasks of 293 cells transfected with either pSV8.3 or pCMV–8GFP DNA with the SuperFect reagent (Qiagen) as per themanufacturer’s instructions. At 48 h posttransfection, thecells were washed twice with phosphate-buffered saline,pelleted, and resuspended in 2 ml of high-salt buffer (1.2 MNaCl, 20 mM HEPES–NaOH, pH 7.5, 10% (v/v) glycerol,1 mM DTT, 1 mM EDTA, one Complete-Mini proteaseinhibitor cocktail tablet (Roche) per 7 ml buffer). The ly-sates were sonicated on ice three times for 30 s and thenclarified by centrifugation at 4°C in an Eppendorf mi-crofuge. The lysates were dialyzed overnight against low-salt buffer (100 mM NaCl, 20 mM HEPES–NaOH, pH 7.5,10% (v/v) glycerol, 1 mM DTT, 0.1 mM EDTA), and thencleared by centrifugation at 4°C for 30 min in an Eppendorfmicrofuge. After the addition of 1 M MgCl2 to 20 mM and200 units of DNase I, the lysates were incubated at 4°C for2 h and then stored at �80°C.

UL9 binding assay

Five hundred-milliliter cultures of DH5 Escherichia colicontaining either the pB1 or pD27 plasmid were grown inLuria broth overnight at 37°C. The pelleted cells werewashed three times with 100 ml TE (10 mM Tris–HCl, pH7.6, 1 mM EDTA). After resuspension in 10 ml of lysisbuffer (50 mM Tris–HCl, pH 7.6, 150 mM NaCl, 0.05%Tween 20, 1 mM EDTA, one Complete-Mini protease in-hibitor cocktail tablet per 10 ml buffer), lysozyme wasadded to 200 �g/ml, and the cells were incubated at roomtemperature for 20 min. The cells were sonicated on ice, andthe lysates were clarified by centrifugation at 4°C in aSorvall centrifuge at 15,000 rpm for 1 h. The lysates wereincubated overnight at 4°C after the addition of 1 M MgCl2to 20 mM and 400 units of DNase I. The lysates werealiquoted and stored at �80°C.

One milliliter of B1 or D27 lysate was added to 0.5 ml ofIgG Sepharose Fast Flow 6, equilibrated as per the manu-facturer’s directions (Pharmacia Biotech), batchwise in mi-crofuge tubes. The slurry was incubated at 4°C for 30 minwith gentle agitation. The IgG Sepharose slurry was washedthree times with TST buffer (50 mM Tris–Cl, pH 7.6, 150mM NaCl, 0.05% Tween 20), and then twice with low-saltbuffer (100 mM NaCl, 20 mM HEPES–NaOH, pH 7.5, 10%(v/v) glycerol, 1 mM DTT, 0.1 mM EDTA). After 0.5 ml ofeither wild-type ICP8 or ICP8–GFP extract was added tothe protein A–UL9/IgG Sepharose preparations, the slurrywas incubated for 1 h at 4°C with gentle agitation. Theslurry was centrifuged at 4°C in an Eppendorf microfuge for1 min. The supernatant was removed and kept as the flow-through fraction. The slurry was subsequently washed twicewith 0.5 ml of low-salt buffer and then 0.5 ml of high-saltelution buffer (1 M NaCl, 20 mM HEPES–NaOH, pH 7.5,10% (v/v) glycerol, 1 mM DTT, 0.1 mM EDTA). The two


0.5-ml wash fractions were combined for a final volume of1 ml. A portion of each fraction (1/10 vol) was separated bySDS–PAGE, transferred to a PVDF membrane (Bio-Rad),and probed with the ICP8 3-83 polyclonal antibody. Inputand recovery of protein was monitored by loading decreas-ing twofold dilutions of total lysate equivalent to 1/10th,1/20th, and 1/40th of the lysate on a gel that was preparedin parallel. The amount bound was determined as the per-centage of total protein in the high-salt fraction as measuredby densitometry of the film.

Visualization of ICP8-GFP

Vero or V529 cells were grown on 22 � 22-mm cover-slips in six-well trays. For short-term live cell analysis, thecoverslip was removed at the indicated time postinfection,washed briefly in PBS, and then placed onto a glass slide forobservation. For fixed cell analysis, the coverslip waswashed in PBS, fixed for 5 min in 3.7% formaldehyde inPBS, washed twice in PBS and once in distilled water, andthen mounted onto glass slides with gelatin.

Cells were viewed with an Axioplan 2 (Zeiss) micro-scope equipped with an Orca cooled CCD camera(Hamamatsu Photonics). A Zeiss 40� or 63� plan Apo-chromat objective was used. Images were collected with theOpenLab software package (Improvision). Figures wereprepared with Canvas 5.0 (Deneba).

Indirect immunofluorescence

For indirect immunofluorescence of cultured cells, Verocells were plated on 18-mm coverslips in a 24-well tray. At5.5 h postinfection with the ICP8–GFP virus, the cells werewashed briefly in PBS and then fixed for 5 min in 3.7%formaldehyde in PBS. After being washed with PBS for 5min, the cells were subsequently permeabilized with coldacetone for 2 min. The cells were then rinsed in distilledwater and washed for 5 min in PBS. Primary antibody wasthen applied at a 1:50 dilution of one of the following: the58S anti-ICP4 monoclonal antibody, the 39S anti-ICP8monoclonal antibody, or the 5H11D6 anti-UL42 monoclo-nal antibody. After being washed three times for 5 min inPBS, a 1:100 dilution of a rhodamine-conjugated goat anti-mouse antibody was used as a secondary antibody for de-tection. After a final wash in PBS, the coverslips weremounted onto glass slides with gelatin. The same micro-scope setup used for observing ICP8–GFP was used forindirect immunofluorescence.

Analysis of viral proteins

Vero cells in 25-cm2 culture flasks were mock-infectedor infected at a m.o.i. of 10 with wild-type KOS 1.1, therecombinant ICP8–GFP virus, or the replication-defectivevirus HD-2. At the indicated times postinfection, the cellswere incubated for 30 min at 37°C in labeling medium

(Eagle’s minimum essential medium, without methionine orcysteine (Sigma), supplemented with 1% dialyzed fetal bo-vine serum and 100 �Ci/ml EXPRE35S35S labeling mix(NEN)). The cells were harvested by the addition of 1 ml ofSDS sample buffer (100 mM Tris–HCl, pH 6.8, 4% SDS,20% glycerol, 200 mM DTT, 0.2% bromphenol blue). Theproteins were separated on a 9.25% gel by SDS–PAGE. Thegel was dried and exposed to BioMax MR film (Kodak).

For immunodetection of the ICP8 protein by Westernblot, the 3–83 polyclonal antibody was used at a dilution of1:1000 in TBST (20 mM Tris–HCl, pH 7.5, 150 mM NaCl,1% Tween 20). The membrane was blocked with TBST plus5% milk for 30 min at room temperature before the additionof 3–83 antiserum. The membrane was washed three timesfor 15 min with TBST at room temperature after primaryand secondary antibody incubations. An HRP-conjugatedgoat anti-rabbit secondary antibody (Santa Cruz Biotech-nology) was used at a dilution of 1:5000 in TBST. Enhancedchemiluminescence reagents were used as directed by themanufacturer (Amersham Pharmacia Biotech).

Time-lapse video microscopy

S2 cells for time-lapse analysis were grown on 18-mmglass coverslips in six-well plates. A DeltaVision micro-scope system (Applied Precision Instruments) assembledaround an Axiovert microscope (Zeiss) and a PXL CCDcamera (Photometrics Ltd.) was used to view the cells. AZeiss 63X plan Apochromat objective was used. The tem-perature was held constant at 37°C with an objectivewarmer (Bioptechs, Inc.). Cells were maintained in CO2

independent medium (Gibco-BRL) plus 1% heat-inacti-vated fetal bovine serum for the duration of the time-lapseprocedure.

Acknowledgments

This study was supported by NIH grant CA-26345. Wethank Nigel Stow and Min Gao for providing reagents usedin this study. We also thank Darren Higgins for his micros-copy expertise and Lisa Holik for her aid in preparing thismanuscript.

References

Abney, J.R., Cutler, B., Fillbach, M.L., Axelrod, D., Scalettar, B.A., 1997.Chromatin dynamics in interphase nuclei and its implication for nuclearstructure. J. Cell Biol. 137, 1459–1468.

Arbuckle, M.I., Stow, N.D., 1993. A mutational analysis of the DNA-binding domain of the herpes simplex virus type 1 UL9 protein. J. Gen.Virol. 74, 1349–1355 (Pt. 7).

Ascoli, C.A., Maul, G.G., 1991. Identification of a novel nuclear site.J. Cell Biol. 112, 785–795.

Boehmer, P.E., 1998. The herpes simplex virus type-1 single-strandedDNA-binding protein, ICP8, increases the processivity of the UL9protein DNA helicase. J. Biol. Chem. 273, 2676–2683.


Boehmer, P.E., Craigie, M.C., Stow, N.D., Lehman, I.R., 1994. Associa-tion of origin binding protein and single strand DNA-binding protein,ICP8, during herpes simplex virus type 1 DNA replication in vivo.J. Biol. Chem. 269, 29329–29334.

Boehmer, P.E., Lehman, I.R., 1993a. Herpes simplex virus type 1 ICP8:helix-destabilizing properties. J. Virol. 67, 711–715.

Boehmer, P.E., Lehman, I.R., 1993b. Physical interaction between theherpes simplex virus 1 origin-binding protein and single-strandedDNA-binding protein ICP8. Proc. Natl. Acad. Sci. USA 90, 8444–8448.

Bortner, C., Hernandez, T.R., Lehman, I.R., Griffith, J., 1993. Herpessimplex virus 1 single-strand DNA-binding protein (ICP8) will promotehomologous pairing and strand transfer. J. Mol. Biol. 231, 241–250.

Bush, M., Yager, D.R., Gao, M., Weisshart, K., Marcy, A.I., Coen, D.M.,Knipe, D. M., 1991. Correct intranuclear localization of herpes simplexvirus DNA polymerase requires the viral ICP8 DNA-binding protein.J. Virol. 65, 1082–1089.

Carmo-Fonseca, M., 2002. The contribution of nuclear compartmentaliza-tion to gene regulation. Cell 108, 513–521.

Carmo-Fonseca, M., Mendes-Soares, L., Campos, I., 2000. To be or not toin the nucleolus. Nat. Cell Biol. 2, E107–E112.

Challberg, M.D., 1986. A method for identifying the viral genes requiredfor herpesvirus DNA replication. Proc. Natl. Acad. Sci. USA 83,9094–9098.

Chen, Y.M., Knipe, D.M., 1996. A dominant mutant form of the herpessimplex virus ICP8 protein decreases viral late gene transcription.Virology 221, 281–290.

Chiou, H.C., Weller, S.K., Coen, D.M., 1985. Mutations in the herpessimplex virus major DNA-binding protein gene leading to alteredsensitivity to DNA polymerase inhibitors. Virology 145, 213–226.

Chubb, J.R., Boyle, S., Perry, P., Bickmore, W.A., 2002. Chromatin mo-tion is constrained by association with nuclear compartments in humancells. Curr. Biol. 12, 439–445.

Da Costa, X.J., Kramer, M.F., Zhu, J., Brockman, M.A., Knipe, D.M.,2000. Construction, phenotypic analysis, and immunogenicity of aUL5/UL29 double deletion mutant of herpes simplex virus 2. J. Virol.74, 7963–7971.

Darlington, R.W., James, C., 1966. Biological and morphological aspectsof the growth of equine abortion virus. J. Bacteriol. 92, 250–257.

de Bruyn Kops, A., Knipe, D.M., 1988. Formation of DNA replicationstructures in herpes virus-infected cells requires a viral DNA bindingprotein. Cell 55, 857–868.

de Bruyn Kops, A., Knipe, D.M., 1994. Preexisting nuclear architecturedefines the intranuclear location of herpesvirus DNA replication struc-tures. J. Virol. 68, 3512–3526.

de Bruyn Kops, A., Uprichard, S.L., Chen, M., Knipe, D.M., 1998. Com-parison of the intranuclear distributions of herpes simplex virus pro-teins involved in different viral functions. Virology 252, 162–178.

Dutch, R.E., Lehman, I.R., 1993. Renaturation of complementary DNAstrands by herpes simplex virus type 1 ICP8. J. Virol. 67, 6945–6949.

Everett, R.D., Freemont, P., Saitoh, H., Dasso, M., Orr, A., Kathoria, M.,Parkinson, J., 1998. The disruption of ND10 during herpes simplexvirus infection correlates with the Vmw110- and proteasome-depen-dent loss of several PML isoforms. J. Virol. 72, 6581–6591.

Everett, R.D., Lomonte, P., Sternsdorf, T., van Driel, R., Orr, A., 1999.Cell cycle regulation of PML modification and ND10 composition.J. Cell Sci. 112, 4581–4588.

Everett, R.D., Maul, G.G., 1994. HSV-1 IE protein VMW 110 causesredistribution of PML. EMBO J. 13, 5062–5069.

Falkenberg, M., Elias, P., Lehman, I.R., 1998. The herpes simplex virustype 1 helicase–primase. J. Biol. Chem. 273, 32154–32157.

Ferreira, J., Paolella, G., Ramos, C., Lamond, A., 1997. Spatial organiza-tion of large-scale chromatin domains in the nucleus: a magnified viewof single chromosome territories. J. Cell Biol. 139, 1597–1610.

Gall, J.G., 2000. Cajal bodies: the first 100 years. Annu. Rev. Cell Dev.Biol. 16, 273–300.

Gao, M., Bouchey, J., Curtin, K., Knipe, D.M., 1988. Genetic identification

of a portion of the herpes simplex virus ICP8 protein required forDNA-binding. Virology 163, 319–329.

Gao, M., Knipe, D.M., 1989. Genetic evidence for multiple nuclear func-tions of the herpes simplex virus ICP8 DNA-binding protein. J. Virol.63, 5258–5267.

Gao, M., Knipe, D.M., 1991. Potential role for herpes simplex virus ICP8DNA replication protein in stimulation of late gene expression. J. Virol.65, 2666–2675.

Gao, M., Knipe, D.M., 1993. Intragenic complementation of herpes sim-plex virus ICP8 DNA-binding protein mutants. J. Virol. 67, 876–885.

Godowski, P.J., Knipe, D.M., 1983. Mutations in the major DNA-bindingprotein gene of herpes simplex virus type 1 result in increased levels ofviral gene expression. J. Virol. 47, 478–486.

Godowski, P.J., Knipe, D.M., 1985. Identification of a herpes simplex virusfunction that represses late gene expression from parental viral ge-nomes. J. Virol. 55, 357–365.

Godowski, P.J., Knipe, D.M., 1986. Transcriptional control of herpesvirusgene expression: gene functions required for positive and negativeregulation. Proc. Natl. Acad. Sci. USA 83, 256–260.

Gunawardena, S., Rykowski, M.C., 2000. Direct evidence for interphasechromosome movement during the mid-blastula transition in Drosoph-ila. Curr. Biol. 10, 285–288.

Gupte, S.S., Olson, J.W., Ruyechan, W.T., 1991. The major herpes simplexvirus type-1 DNA-binding protein is a zinc metalloprotein. J. Biol.Chem. 266, 11413–11416.

Hamatake, R.K., Bifano, M., Hurlburt, W.W., Tenney, D.J., 1997. Afunctional interaction of ICP8, the herpes simplex virus single-strandedDNA-binding protein, and the helicase–primase complex that is depen-dent on the presence of the UL8 subunit. . J. Gen. Virol. 78, 857–865.

He, X., Lehman, I.R., 2001. An initial ATP-independent step in the un-winding of a herpes simplex virus type 1 origin of replication by acomplex of the viral origin-binding protein and single-strand DNA-binding protein. Proc. Natl. Acad. Sci. USA 98, 3024–3028.

Hernandez, T.R., Lehman, I.R., 1990. Functional interaction between theherpes simplex-1 DNA polymerase and UL42 protein. J. Biol. Chem.265, 11227–11232.

Heun, P., Laroche, T., Raghuraman, M.K., Gasser, S.M., 2001. The posi-tioning and dynamics of origins of replication in the budding yeastnucleus. J. Cell Biol. 152, 385–400.

Hodges, M., Tissot, C., Howe, K., Grimwade, D., Freemont, P.S., 1998.Structure, organization, and dynamics of promyelocytic leukemia pro-tein nuclear bodies. Am. J. Hum. Genet. 63, 297–304.

Honess, R.W., Roizman, B., 1974. Regulation of herpesvirus macro-mo-lecular synthesis. I. Cascade regulation of the synthesis of three groupsof viral proteins. J. Virol. 14, 8–19.

Ishov, A.M., Maul, G.G., 1996. The periphery of nuclear domain 10(ND10) as site of DNA virus deposition. J. Cell Biol. 134, 815–826.

Kennedy, B.R., Barbie, D.A., Classon, M., Dyson, N., Harlow, E., 2000.Nuclear organization of nuclear replication in primary mammaliancells. Genes Dev. 14, 2855–2868.

Knipe, D.M., Senechek, D., Rice, S.A., Smith, J.L., 1987. Stages in thenuclear association of the herpes simplex virus transcriptional activatorprotein ICP4. J. Virol. 61, 276–284.

Lamberti, C., Weller, S.K., 1998. The herpes simplex virus type 1 cleav-age/packaging protein, UL32, is involved in efficient localization ofcapsids to replication compartments. J. Virol. 72, 2463–2473.

Le Gac, N.T., Villani, G., Hoffmann, J.S., Boehmer, P.E., 1996. The UL8subunit of the herpes simplex virus type-1 DNA helicase–primase opti-mizes utilization of DNA templates covered by the homologous single-strand DNA-binding protein ICP8. J. Biol. Chem. 271, 21645–21651.

Lee, C.K., Knipe, D.M., 1985. An immunoassay for the study of DNA-binding activities of herpes simplex virus protein ICP8. J. Virol. 54,731–738.

Lee, S.S., Lehman, I.R., 1997. Unwinding of the box I element of a herpessimplex virus type 1 origin by a complex of the viral origin bindingprotein, single-strand DNA binding protein, and single-stranded DNA.Proc. Natl. Acad. Sci. USA 94, 2838–2842.


Leopardi, R., Ward, P.L., Ogle, W.O., Roizman, B., 1997. Association ofherpes simplex virus regulatory protein ICP22 with transcriptionalcomplexes containing EAP, ICP4, RNA polymerase II, and viral DNArequires posttranslational modification by the U(L)13 protein kinase.J. Virol. 71, 1133–1139.

Liptak, L., Uprichard, S.L., Knipe, D.M., 1996. Functional order of as-sembly of herpes simplex virus DNA replication proteins into prerep-licative site structures. J. Virol. 70, 1759–1767.

Lukonis, C.J., Burkham, J., Weller, S.K., 1997. Herpes simplex virus type1 prereplicative sites are a heterogeneous population: only a subset arelikely to be precursors to replication compartments. J. Virol. 71, 4771–4781.

Lukonis, C.J., Weller, S.K., 1996. Characterization of nuclear structures incells infected with herpes simplex virus type 1 in the absence of viralDNA replication. J. Virol. 70, 1751–1758.

Makhov, A.M., Boehmer, P.E., Lehman, I.R., Griffith, J.D., 1996. Visual-ization of the unwinding of long DNA chains by the herpes simplexvirus type 1 UL9 protein and ICP8. J. Mol. Biol. 258, 789–799.

Manders, E.M., Kimura, H., Cook, P., 1999. Direct imaging of DNA inliving cells reveals the dynamics of chromosome formation. J. CellBiol. 144, 813–821.

Marshall, W.F., Straight, A., Marko, J.F., Swedlow, J., Dernburg, A.,Belmont, A., Murray, A.W., Agard, D.A., Sedat, J.W., 1997. Interphasechromosomes undergo constrained diffusional motion in living cells.Curr. Biol. 7, 930–939.

Maul, G.G., Ishov, A.M., Everett, R.D., 1996. Nuclear domain 10 aspreexisting potential replication start sites of herpes simplex virus type1. Virology 217, 67–75.

Maul, G.G., Negorev, D., Bell, P., Shov, A., 2000. Review: properties andassembly mechanisms of ND10, PML bodies, or PODs. Struct. Biol.129, 278–287.

Misteli, T., Caceres, J.F., Spector, D.L., 1997. The dynamics of pre-mRNAsplicing factor in living cells. Nature 387, 523–527.

Monier, K., Armas, J.C.G., Etteldorf, S., Ghazal, P., Sullivan, K.F., 2000.Annexation of the interchromosomal space during viral infection. Nat.Cell Biol. 2, 661–665.

Muranyi, W., Haas, J., Wagner, M., Krohne, G., Koszinowski, U.H., 2002.Cytomegalovirus recruitment of cellular kinases to dissolve the nuclearlamina. Science 297, 854–857.

Murphy, C.G., Lucas, W.T., Means, R., Czajak, S., Hale, C.L., Lifson,J.D., Kauer, A., Johnson, R.P., Knipe, D.M., Desrosiers, R.C., 2000.Vaccine protection against simian immunodeficiency virus by recom-binant strains of herpes simplex virus. J. Virol. 74, 7745–7753.

Negorev, D., Maul, G.G., 2001. Cellular proteins localized at and inter-acting within ND10/PML nuclear bodies/PODs suggest functions of anuclear depot. Oncogene 20, 7234–7242.

Nowak, G., Pestic-Dragovich, L., Hozak, P., Philimonenko, A., Simerly,C., Schatten, G., de Lanerolle, P., 1997. Evidence for the presence ofmyosin I in the nucleus. J. Biol. Chem. 272, 17176–17181.

O’Donnell, M.E., Elias, P., Funnell, B.E., Lehman, I.R., 1987. Interactionbetween the DNA polymerase and single-stranded DNA-binding pro-tein (infected cell protein 8) of herpes simplex virus. J. Biol. Chem.262, 4260–4266.

Pestic-Dragovich, L., Stojiljkovic, L., Philimonenko, A.A., Nowak, G., Ke,Y., Settlage, R.E., Shabanowitz, J., Hunt, D.F., Hozak, P., de Lanerolle,P., 2000. A myosin I isoform in the nucleus. Science 290, 337–341.

Phelan, A., Dunlop, J., Patel, A.H., Stow, N.D., Clements, J.B., 1997.Nuclear sites of herpes simplex virus type 1 DNA replication andtranscription colocalize at early times postinfection and are largelydistinct from RNA processing factors. J. Virol. 71, 1124–1132.

Politz, J.C., Pederson, T., 2000. Review: movement of mRNA from tran-scription site to nuclear pore. J. Struct. Biol. 129, 252–257.

Quinlan, M.P., Chen, L.B., Knipe, D.M., 1984. The intranuclear location ofa herpes simplex virus DNA-binding protein is determined by the statusof viral DNA replication. . Cell 36, 857–868.

Quinn, J.P., McGeoch, D.J., 1985. DNA sequence of the region in thegenome of herpes simplex virus type 1 containing the genes for DNApolymerase and the major DNA binding protein. Nucleic Acids Res.13, 8143–8163.

Qumsiyeh, M.B., 1999. Structure and function of the nucleus: anatomy andphysiology of chromatin. Cell. Mol. Life Sci. 55, 1129–1140.

Randall, R.E., Dinwoodie, N., 1986. Intranuclear localization of herpessimplex virus immediate-early and delayed-early proteins: evidencethat ICP 4 is associated with progeny virus DNA. J. Gen. Virol. 67,2163–2177.

Rice, S.A., Long, M.C., Lam, V., Spencer, C.A., 1994. RNA polymerase IIis aberrantly phosphorylated and localized to viral replication compart-ments following herpes simplex virus infection. J. Virol. 68, 988–1001.

Roizman, B., Knipe, D.M., 2001. Herpes simplex viruses and their repli-cation, in: Knipe, D.M., Howley, P.M. (Eds.), Fields Virology, 4th ed.Vol. 2. Lippincott Williams & Wilkins, Philadelphia, pp. 2399–2460.

Ruyechan, W.T., 1983. The major herpes simplex virus DNA-bindingprotein holds single-stranded DNA in an extended configuration. J. Vi-rol. 46, 661–666.

Ruyechan, W.T., Weir, A.C., 1984. Interaction with nucleic acids andstimulation of the viral DNA polymerase by the herpes simplex virustype 1 major DNA-binding protein. J. Virol. 52, 727–733.

Sadoni, N., Langer, S., Fauth, C., Bernardi, G., Cremer, T., Turner, B.M.,Zink, D., 1999. Nuclear organization of mammalian genomes: polarchromosome territories build up functionally distinct higher order com-partments. J. Cell Biol. 146, 1211–1226.

Schwartz, J., Roizman, B., 1969. Similarities and differences in the devel-opment of laboratory strains and freshly isolated strains of herpessimplex virus in HEp-2 cells: electron microscopy. J. Virol. 4, 879–889.

Scott, E.S., O’Hare, P., 2001. Fate of the inner nuclear membrane proteinlamin B receptor and nuclear lamins in herpes simplex virus type 1infection. J. Virol. 75, 8818–8830.

Shelby, R.D., Hahn, K.M., Sullivan, K.F., 1996. Dynamic elastic behaviourof alpha-satellite DNA domains visualized in situ in living human cells.J. Cell Biol. 135, 545–557.

Sirtori, C., Bosisio-Besteti, M., 1967. Nucleolar changes in KB tumor cellsinfected with herpes simplex virus. Cancer Res. 27, 367–376.

Taylor, T.J, 2002. Ph.D. thesis, Harvard University, Cambridge, MA.Tumbar, T., Belmont, A.S., 2001. Interphase movements of a DNA chro-

mosome region modulated by VP16 transcriptional activator. Nat. CellBiol. 3, 134–139.

Uprichard, S.L., Knipe, D.M., 1997. Assembly of herpes simplex virusreplication proteins at two distinct intranuclear sites. Virology 229,113–125.

Vazquez, J., Belmont, A.S., Sedat, J.W., 2001. Multiple regimes of con-strained chromosome motion are regulated in the interphase Drosophilanucleus. Curr. Biol. 11, 1227–1239.

Ward, P.L., Ogle, W.O., Roizman, B., 1996. Assemblons: nuclear struc-tures defined by aggregation of immature capsids and some tegumentproteins of herpes simplex virus 1. J. Virol. 70, 4623–4631.

Wilcock, D., Lane, D.P., 1991. Localization of p53, retinoblastoma andhost replication proteins at sites of viral replication in herpes-infectedcells. Nature 349, 429–431.

Yu, D., Weller, S.K., 1998. Genetic analysis of the UL15 gene locus for theputative terminase of herpes simplex virus type 1. Virology 243,32–44.

Zhao, K., Wang, W., Rando, O.J., Xue, Y., Swiderek, K., Kuo, A.,Crabtree, G.R., 1998. Rapid and phosphoinositol-dependent binding ofthe SWI/SNF-like BAF complex to chromatin after T lymphocytereceptor signaling. Cell 95, 625–636.

Zhong, S., Salomoni, P., Pandolfi, P.P., 2000. The transcriptional role ofPML and the nuclear body. Nat. Cell Biol. 2, E85–E90.

Zink, D., Cremer, T., Saffrich, R., Fischer, R., Trendelenburg, M.F.,Ansorge, W., Stelzer, E.H., 1998. Structure and dynamics of humaninterphase chromosome territories in vivo. Hum. Genet. 102, 241–251.


Documents

Herpes simplex virus replication compartments can form by coalescence of smaller compartments