An Sp1/Sp3 Site in the Downstream Region of Varicella-Zoster Virus (VZV) oriS Influences Origin-Dependent DNA Replication and Flanking Gene Transcription and Is Important for VZV Replication In Vitro and in Human Skin

Abstract

The distribution and orientation of origin-binding protein (OBP) sites are the main architectural contrasts between varicella-zoster virus (VZV) and herpes simplex virus (HSV) origins of DNA replication (oriS). One important difference is the absence of a downstream OBP site in VZV, raising the possibility that an alternative cis element may replace its function. Our previous work established that Sp1, Sp3, and YY1 bind to specific sites within the downstream region of VZV oriS; we hypothesize that one or both of these sites may be the alternative cis element(s). Here, we show that the mutation of the Sp1/Sp3 site decreases DNA replication and transcription from the adjacent ORF62 and ORF63 promoters following superinfection with VZV. In contrast, in the absence of DNA replication or in transfection experiments with ORF62, only ORF63 transcription is affected. YY1 site mutations had no significant effect on either process. Recombinant viruses containing these mutations were then constructed. The Sp1/Sp3 site mutant exhibited a significant decrease in virus growth in MeWo cells and in human skin xenografts, while the YY1 site mutant virus grew as well as the wild type in MeWo cells, even showing a late increase in VZV replication in skin xenografts following infection. These results suggest that the Sp1/Sp3 site plays an important role in both VZV origin-dependent DNA replication and ORF62 and ORF63 transcription and that, in contrast to HSV, these events are linked during virus replication.

INTRODUCTION

Varicella-zoster virus (VZV) is a ubiquitous human alphaherpesvirus and is the causative agent of two diseases, varicella (chickenpox) and herpes zoster (shingles). The VZV genome (125 kb) contains at least 71 genes and contains two origins of DNA replication (oriS) flanked by the ORF62 and ORF63 genes (7). VZV oriS contains a 46-bp AT-rich palindrome and three consensus binding sites for the VZV origin-binding protein (OBP) (ORF51). All three OBP-binding sites (boxes A, B, and C) for the VZV OBP [5′-C(G/A)TTCGCACT-3′] are upstream of the palindrome and in the same orientation (Fig. 1). This is in contrast to the structure of herpes simplex virus (HSV) oriS, where the OBP-binding sites (boxes I, II, and III) are located both upstream and downstream of the AT-rich element. These binding sites are also oriented in different directions and use different DNA strands (42).

Fig 1.

Description of the architectures of VZV and HSV-1 oriS origins of DNA replication and structure of the wild-type pLitmus R62/63F plasmid used in DpnI replication and reporter gene assays. The positions of the OBP-binding-site boxes and their orientations are indicated. The vertical numbers under the VZV oriS structure represent the nucleotide positions of the beginning and end of the oriS structure of the Dumas strain IRs copy of the origin of replication; these locations are duplicated in TRs.

Stow et al. (43) showed previously that the A site is absolutely required for DNA replication and that the deletion of the C site results in a decreased level of replication in VZV. In contrast, the deletion of the B site showed no effect on the extent of replication. Thus, the minimal VZV origin consists of the A site and the AT-rich palindrome. In contrast, all three OBP-binding sites in HSV are required for efficient DNA replication (1, 9, 13, 23, 31, 41, 46). These data suggest that the OPB binding pattern, the unwinding of the origin, and the development of the replication fork may be different in VZV and HSV.

One possibility that might resolve this apparent difference is that there is an alternative binding sequence for the VZV OBP in this region, resulting in an overall mechanism similar to that observed for HSV. There is a partial OBP-binding site in the downstream region adjacent to a GA-rich region unique to the VZV origin (Fig. 1). However, previous gel shift assays with recombinant OBP did not demonstrate an interaction (5), and this site was not protected in DNase I protection assays with recombinant OBP (43). In our previous work, we designated this sequence box D and showed that the mutation of this site resulted in increased DNA replication levels in both the presence and the absence of flanking gene expression (17). These findings essentially exclude the possibility of an alternative OBP site downstream from VZV oriS that plays the role of box II in HSV oriS.

An alternative possibility is the presence of a cellular protein that binds downstream of the AT-rich stretch in VZV oriS and compensates for the absence of the downstream OBP-binding site. Indeed, it has been shown for other herpesviruses that cellular transcription factors can influence DNA replication. Both Epstein-Barr virus (EBV) and HSV require additional cellular and/or viral proteins that specifically activate origin-dependent replication. In the case of EBV, the viral protein BZLF1, a transcription factor, is involved in DNA replication, and two cellular transcription factors, Sp1 and ZBP-89, interact with the viral DNA polymerase and its processivity factor to stimulate replication (2). In the case of HSV, in addition to the requirement for the seven viral DNA replication factors, the cellular Sp1 and Sp3 proteins are involved in the DNA replication process (29). Mutations of the Sp1/Sp3-binding site in HSV oriS decreased the DNA replication efficiency by approximately 60%.

In previous work (16), we identified the binding of the cellular transcription factors Sp1, Sp3, and YY1 to portions of the downstream region of VZV oriS. YY1 appeared to have no effect on origin-dependent DNA replication. However, the mutation of the Sp1/Sp3 site (formerly GC box 1) ablated the binding of these factors, resulting in an increase in the level of replication in the absence of the adjacent transcribable genes. This indicates that Sp1 and/or Sp3 acts to suppress VZV oriS-dependent DNA replication in the absence of flanking gene expression.

In the work presented here, we investigate in detail the role played by these cellular factor-binding sites in VZV origin-dependent DNA replication and in the expression of the oriS-flanking genes ORF62 and ORF63. The results of VZV superinfection demonstrated that the Sp1/Sp3 site mutation caused a decrease in the level of DNA replication in the presence of flanking gene expression and decreases in ORF62 and ORF63 gene transcription levels in the absence of the DNA replication inhibitor phosphonoacetic acid (PAA). In contrast, the YY1 site mutation had no statistically significant effects on DNA replication or on flanking gene expression. However, in experiments done in the presence of PAA or with ORF62 transfection, the Sp1/Sp3 site mutation inhibited only the expression of the ORF63 gene. A recombinant virus with the Sp1/Sp3 mutation showed a significant decrease in virus growth in MeWo cells, while a YY1 mutant virus grew as well as the wild-type virus. Also, the Sp1/Sp3 site mutation decreased virus growth in human skin xenografts in the SCID mouse model significantly at day 10 postinfection. These results suggest that the Sp1/Sp3 site plays important roles in both VZV origin-dependent DNA replication and flanking gene transcription as well as in pathogenesis in human skin.

MATERIALS AND METHODS

Cells and viruses.

MeWo cells, a human melanoma cell line that supports the replication of VZV, were grown in Eagle's minimal essential medium supplemented with 10% fetal bovine serum, as previously described (40). VZV strain MSP (VZV-MSP) (11) and VZV strain pOka (10) were propagated in MeWo cell monolayers, as described previously by Lynch et al. (22) and Peng et al. (32).

Preparation of whole-cell lysates and immunoblot analysis.

Whole-cell lysates of VZV-infected MeWo cells were prepared as previously described (32). Cells were grown to confluence in 100-mm petri dishes and infected with wild-type and mutant pOka VZVs at a ratio of infected to uninfected cells of 0.4:1. The infected cells were washed with phosphate-buffered saline (PBS) and then suspended in lysis buffer (50 mM Tris-HCl [pH 7.5], 0.15 M NaCl, 1 mM EDTA, 0.1% Triton X-100, and a protease inhibitor cocktail [Roche, Mannheim, Germany], added according to the manufacturer's instructions). The lysates were collected and centrifuged for 5 min, and supernatants were stored at −70°C for subsequent use.

Whole-cell lysates of VZV-infected MeWo cells were analyzed by 10% SDS-PAGE and immunoblotted for the presence of IE62 and IE63, as described previously by Yang et al. (47). Antisera against full-length IE62 (40) and antisera against full-length IE63 (48) were used as described previously. Antibody against β-tubulin was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Reactive bands were visualized by using goat anti-rabbit immunoglobulin G conjugated with horseradish peroxidase (Chemicon, Temecula, CA) in conjunction with Supersignal West Pico chemiluminescence substrate (Pierce, Rockford, IL). The quantification of the relative amounts of IE62 and IE63 normalized to the amounts of β-tubulin in loading controls was performed by using a Bio-Rad GS700 imaging densitometer (Bio-Rad, Hercules, CA). Statistical significance was determined by a one-way analysis of variance (ANOVA) followed by Tukey's post hoc test.

Plasmids.

Plasmid pLitmus R62/63F and the parental plasmid pLitmus Ren/FF were constructed as described previously by Jones et al. (15). pLitmus R62/63F contains the complete 1.5-kb intergenic region of VZV DNA between the ORF62 and ORF63 genes of strain pOka, including the VZV oriS structure, inserted between genes encoding Renilla and firefly luciferases (Promega), respectively, so that the luciferase genes acted as reporters of ORF62 and ORF63 transcription. Plasmids containing Sp1/Sp3, YY1, and box A site-specific mutations within the oriS region were generated by mutating wild-type pLitmus R62/63F plasmids using the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) and the following primer sets: 5′-ATGTCGCGGTTTTATGGGGTGTTCGCGGGCTTTTCACAGAATATA-3′ and 5′-TATATTCTGTGAAAAGCCCGCGAACACCCCATAAAACCGCGACAT-3′ for the Sp1/Sp3 site mutation, 5′-ACAGAATATATATATTCCAAATTTAGCGGCAGGCTTTTTAAAATC-3′ and 5′-GATTTTAAAAAGCCTGCCGCTAAATTTGGAATATATATATTCTGT-3′ for the YY1 site mutation, and 5′-GGCATGTGTCCAACCACCGTTAAAACTTTCTTTCTATATATATAT-3′ and 5′-ATATATATATAGAAAGAAAGTTTTAACGGTGGTTGGACACATGCC-3′ for the box A mutation (the underlined nucleotides are the mutated nucleotides in each primer). All primers were synthesized by IDT (Coralville, IA). The positions and sequences of the mutations in the VZV oriS sequence contained within the pLitmus R62/63F plasmids used for transfections were verified by sequencing at the Roswell Park Cancer Institute sequencing facility in Buffalo, NY.

The cloning of the pCMV62 plasmid expressing wild-type ORF62 under the control of the cytomegalovirus (CMV) immediate-early (IE) promoter was described previously (33–35). The ORF62 gene was derived from the EcoRI E fragment of the genome of the low-passage-number North American clinical isolate Scott (44).

DpnI replication assays.

MeWo cells were transfected with Lipofectamine reagent (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. Four microliters of Lipofectamine reagent was used per microgram of transfected DNA in each transfection. In transfections, which were performed with 100-mm-diameter petri dishes, 2.1 × 10⁶ MeWo cells per dish were seeded into 12 ml of complete growth medium. The cells were 80% confluent at the time of transfection. Three hours before transfection, the medium was replaced with fresh medium. Origin-dependent DNA replication experiments were performed as described previously by Stow and McMonagle (41), Stow and Davison (42), and Khalil et al. (16). The cells were transfected with 5 μg of wild-type or mutant pLitmus R62/63F plasmids. At 6 h posttransfection, cells were superinfected with VZV strain MSP (11) by the addition of a ratio of 0.4 infected cells per 1 uninfected cell to each monolayer. Total cellular DNA was prepared at 48 h after superinfection, and the DNA was isolated by phenol-chloroform extraction followed by ethanol precipitation. The DNA was digested with DpnI and EcoRI, as described previously by Stow and Davison (42), and analyzed by Southern blot hybridization. Transfers were done by using TurboBlotter kits obtained from Whatman, Inc. (Sanford, ME). The blots were probed with a 476-bp PCR product prepared from pLitmus R62/63F by using primers 5′-TAGGCCACCACTTCAAGAACTCTGT-3′ and 5′-AGCAAAAGGCCAGCAAAAGGCCAGG-3′, and the probe was end labeled with[α-³²P]ATP by using T4 kinase (Invitrogen, Carlsbad, CA).

The resulting bands were quantified by PhosphorImager (Molecular Dynamics, Sunnyvale, CA) analysis. The ratio of replicated plasmid to input plasmid represents the replication efficiency of the test plasmid. The data from representative experiments are presented as the means of data from triplicate DpnI replication assays. Statistical significance was determined by a one-way ANOVA followed by Tukey's post hoc test.

Total virus DNA level determination.

MeWo cells were grown to confluence in 100-mm petri dishes and infected with wild-type and mutant pOka VZVs at a ratio of infected to uninfected cells of 0.4:1. Total cellular DNA was prepared at 24 h after superinfection, and the DNA was isolated by phenol-chloroform extraction followed by ethanol precipitation. For total virus DNA level determinations, the DNA was digested with HindIII and EcoRI and analyzed by Southern blot hybridization. The blots were probed with a 337-bp PCR product containing the intergenic region between VZV ORF3 and ORF4 and prepared by using the following primers: 5′-ATTAAACGTTCGGTACACGTCTGGT-3′ and 5′-AAATAAAAAATACCTTTTTCATGCT-3′. For the control experiment to assess total applied cellular DNA, the DNA was digested with XbaI and EcoRI and analyzed by Southern blot hybridization. The blots were probed with a 744-bp PCR product containing a portion of the coding sequence of the E2F3 gene and prepared by using the following primers: 5′-CAATAAATACGGCATTACATTATGA-3′ and 5′-GCAGCGGCCATCTCCCACTGGGAAT-3′. Both probes were end labeled with [α-³²P]ATP by using T4 kinase (Invitrogen, Carlsbad, CA). Transfers were done by using TurboBlotter kits obtained from Whatman, Inc. (Sanford, ME).

The resulting bands were quantified by PhosphorImager (Molecular Dynamics, Sunnyvale, CA) analysis. The total virus DNA level was normalized to the total applied cellular DNA level. Data from representative experiments are presented as the means of data from triplicate experiments. Statistical significance was determined by a one-way ANOVA followed by Tukey's post hoc test.

Reporter gene assays.

Luciferase reporter gene assays were performed as previously described (47), using MeWo cells. Transfections were performed by using 12-well plates. Briefly, 2 × 10⁵ cells were seeded into each well 24 h before transfection. The pCMV · SPORT · β-Gal (β-galactosidase) vector (Gibco, Carlsbad, CA) was used as an internal control reporter for transfections. One microgram of each reporter vector (pLitmus R62/63F) and 0.4 μg of a β-galactosidase-expressing plasmid were transfected for each assay with Lipofectamine reagent (Invitrogen, Carlsbad, CA), according to the manufacturer's instructions, to act as a control for the transfection efficiency. The cells were superinfected at 24 h after transfection with VZV strain MSP (11) by the addition of a ratio of 0.4 infected cells per 1 uninfected cell to each monolayer when experiments with VZV superinfection were required.

In the experiments done in the context of ORF62 transfection, the reporter plasmids were cotransfected along with 0.005 to 0.02 μg of pCMV-ORF62 plasmids. Dual-luciferase activities were normalized to the beta-galactosidase activities. Various amounts of the pcDNA empty cloning vector were transfected along with plasmid pCMV62 to equalize the amounts of both total DNA and the CMV promoter in each set of transfections.

The cells were collected at 48 h after superinfection or transfection and lysed in 250 μl of lysis buffer (50 mM HEPES [pH 7.4], 250 mM NaCl, 1% NP-40, 1 mM EDTA). Control experiments without infection were done for each plasmid to determine the basal expression levels. Dual-luciferase assays were performed by using a dual-luciferase reporter assay system (Promega) with 20 μl of cell extract, 50 μl of LARII reagent, and 50 μl of Stop & Glo reagent in each assay mixture. β-Gal assays were performed by using a β-Gal assay kit (Invitrogen) according to the microtiter plate protocol recommended by the manufacturer. Transfection experiments were repeated at least three times, and each set of transfection conditions in a given experiment was performed in triplicate. The data from representative experiments are presented as the means of data from triplicate gene reporter assays. Error bars indicate standard errors.

Generation of pOka recombinant viruses with Sp1/Sp3 and YY1 site mutations.

Recombinant viruses were generated by using cosmids derived from pOka (30). The entire pOka genome is covered in four overlapping cosmids, designated Fsp73 (pOka nucleotides [nt] 1 to 33128), Spe14 (pOka nt 21795 to 61868), Pme2 (pOka nt 53755 to 96035), and Spe23 (pOka nt 94055 to 125124). The oriS is located in the internal repeat sequence/terminal repeat sequence (IRs/TRs) regions in cosmid Spe23. Mutations of the Sp1/Sp3 and YY1 sites were first introduced into pLitmus(ORF59-65) and pLitmus(ORF66-71) and were sequenced to verify the mutations. To insert the promoter mutations into the pSpe23 cosmid, an NheI/AvrII fragment was excised from plasmid pLitmus(ORF59-65) and ligated into pSpe23. In the final step, to insert the mutation into the other copy of the promoter, an AscI/AvrII fragment from plasmid pLitmus(ORF66-71) was ligated into the corresponding mutant pSpe23 cosmid.

Recombinant viruses were isolated by the transfection of melanoma cells with either the wild-type or the mutated Spe23 cosmid and the other three intact cosmids, Fsp73, Spe14, and Pme2. To confirm the targeted mutations in the Sp1/Sp3 and YY1 sites, genomic DNA was extracted from virus-infected melanoma cells or HELF with DNAzol reagent (Invitrogen, Carlsbad, CA). A PCR fragment covering the mutated region was amplified from genomic DNA by using Pfu polymerase (Stratagene, La Jolla, CA), gel purified with a QIAquick gel extraction kit (Qiagen, Inc., Valencia, CA), and sequenced (Elim Biopharm, Inc., Hayward, CA).

Growth kinetics of pOka recombinant viruses with Sp1/Sp3 and YY1 site mutations.

The replication kinetics of recombinant viruses were assessed by an infectious-focus assay with immunostaining to detect plaques, as previously described (4, 24).

Briefly, 6-well assay plates and 24-well titer plates were seeded with MeWo cells. Assay plates were incubated for various times, and several dilutions of the samples were taken for infectious-focus assays. Titer plates were incubated for 4 days and then fixed with 4% paraformaldehyde for immunohistochemical staining using anti-VZV human serum. Statistical differences in growth kinetics were determined by Student's t test.

Infection of human skin xenografts in SCIDhu mice.

Skin xenografts were made in homozygous CB-17^scid/scid mice, using human fetal tissue supplied by Advanced Bioscience Resources (Alameda, CA) according to federal and state regulations (24, 25). Animal use was in accordance with the Animal Welfare Act and was approved by the Stanford University Administrative Panel on Laboratory Animal Care. Wild-type pOka and pOka Sp1/Sp3 and YY1 mutant viruses were passed three times in primary HELF before the inoculation of the xenografts. The infectious-virus titer was determined for each inoculum at the time of inoculation. Skin xenografts were harvested at 10 and 21 days postinoculation, and virus was titrated on a MeWo cell monolayer and analyzed by an infectious-focus assay and immunohistochemistry with polyclonal anti-VZV human immune serum, as described above and as reported previously by Moffat et al. (24). Virus recovered from the tissues was tested by PCR and sequencing to confirm the expected mutations.

RESULTS

The Sp1/Sp3 site influences origin-dependent DNA replication.

In our previous work (15), we established the binding of the cellular transcription factors Sp1, Sp3, and YY1 to portions of the downstream region of VZV oriS. The binding of Sp1 and Sp3 to the Sp1/Sp3 site (formerly GC box 1) led to the suppression of VZV origin-dependent DNA replication in DpnI replication assays, while the presence or absence of YY1 had no significant effect, based on experiments with Sp1/Sp3 and YY1 site mutations. In these experiments, we used pVO2 plasmids, which lack the promoter elements of the flanking ORF62 and ORF63 genes (8, 26, 27). The protocol involved the transfection of the pVO2 plasmid containing the VZV oriS sequence into MeWo cells, followed by VZV-MSP (11) superinfection, as described previously (16, 17).

In this paper, we extend our previous study and examine the influence of the presence of Sp1/Sp3 and YY1 sites on origin-dependent DNA replication in the presence of flanking gene transcription, looking for links between these two viral functions. For these experiments, in place of pVO2, we employed pLitmus R62/63F plasmids, which contain the entire region between the ORF62 and ORF63 genes, including VZV oriS and the promoters for the two genes (15) (Fig. 1). The ORF62 and ORF63 genes themselves are replaced by two reporters, the Renilla (R) and firefly (F) luciferase genes, respectively. Analogous experiments were performed by the transfection of the pLitmus R62/63F plasmid into MeWo cells, followed by VZV strain MSP superinfection. As presented in Fig. 2B, the CGC mutation to AAA in box A showed no replicated DNA band, indistinguishable from the negative control, as we reported previously (17).

Fig 2.

Results of DpnI replication assays performed using the pLitmus R62/63F plasmid and VZV-MSP superinfection. (A) Sequences of wild-type and mutant Sp1/Sp3 and YY1 sites in the downstream region of VZV oriS. Mutated nucleotides are in italic type. (B) Typical Southern blot analysis of the effects of site-specific mutations of Sp1/Sp3 and YY1 sites. Wild type indicates the pLitmus R62/63F plasmid containing the full intergenic region between the coding sequences of ORF62 and ORF63, including the full oriS. Sp1/Sp3 mutation indicates pLitmus R62/63F containing a point mutation in the Sp1/Sp3 site. YY1 mutation indicates pLitmus R62/63F containing a point mutation in the YY1 site. Box A mutation indicates pLitmus R62/63F containing a triple point mutation of the core CGC triplet within box A. The upper band (R) indicates the position of DpnI-resistant DNA resulting from replication in MeWo cells. The lower band (U) indicates the position of the unreplicated input plasmid. (C) Histogram summarizing the data from three independent DpnI replication assays analyzed at 48 h after superinfection by VZV-MSP. Error bars indicate standard errors. Statistical significance was determined by a one-way ANOVA followed by Tukey's post hoc test.

However, in contrast to our previously reported data using the pVO2 plasmid, the Sp1/Sp3 site mutation (Fig. 2A) in the context of the pLitmus plasmid showed a substantial, significant decrease in the DNA replication efficiency compared to the wild-type level. On the other hand, the YY1 site mutation (Fig. 2A) showed an increase that was not significant (P = 0.087) (Fig. 2B and C). The decrease in the DNA replication efficiency seen with the Sp1/Sp3 site mutation in the pLitmus plasmid implies the involvement of this sequence in the regulation of VZV origin-dependent DNA replication under these experimental conditions.

The Sp1/Sp3 site is involved in expression of the ORF62 and ORF63 genes.

To explore the involvement of the Sp1/Sp3 site in gene expression, we carried out a series of luciferase reporter gene assays to examine the effects of Sp1/Sp3 and YY1 site mutations on the expression of the ORF62 and ORF63 genes. We studied the effects of these mutations in the context of VZV superinfection as well as following the transfection of the pLitmus R62/63F plasmid in the presence and absence of 400 μg/ml phosphonoacetic acid (PAA). DpnI DNA replication assays using the wild-type plasmid confirmed that the presence of 400 μg/ml of PAA completely inhibited DNA replication (data not shown). The results of the luciferase experiments are shown in Fig. 3A and B. The activities of the promoter in the presence of VZV infection are reported as the induction (n-fold) of luciferase activities compared to the basal activity level observed without infection.

Fig 3.

Effects of DNA replication and Sp1/Sp3 and YY1 site mutations on flanking gene expression in the context of VZV-MSP superinfection. (A) Results of triplicate assays comparing the effects of the presence of the Sp1/Sp3 and YY1 site mutations on the expression levels of the Renilla luciferase reporter gene present at the position of the ORF62 gene. (B) Results of triplicate assays comparing the effects of the presence of the Sp1/Sp3 and YY1 site mutations on the expression levels of the firefly luciferase reporter gene present at the position of the ORF63 gene. The promoter activities resulting from the presence of VZV-MSP superinfection are reported as the induction (n-fold) of luciferase activity over the basal level (without infection). +PAA indicates experiments done in the presence of 400 μg/ml PAA. Statistical significance was determined by a one-way ANOVA followed by Tukey's post hoc test. Error bars indicate standard errors.

In the absence of PAA, the Sp1/Sp3 site mutation significantly decreased the expression levels of the ORF62 and ORF63 reporter genes (∼2- and 3-fold, respectively) compared to the wild-type level, while in the presence of PAA, the levels of activity of the ORF62 and ORF63 reporters of the wild-type plasmid were reduced 7- and 5-fold, respectively, compared to their levels in the absence of PAA. Also, the Sp1/Sp3 site mutation reduced the level of activity by an additional 50% compared to the wild-type level in the presence of PAA but only with the ORF63 reporter (not ORF62). In contrast, the YY1 mutation did not significantly affect the expression levels of the two genes in the absence of PAA and had no effect in the presence of PAA. These data indicate that the Sp1/Sp3 site influences the expression levels of both the ORF62 and ORF63 genes during viral replication and confirm our previously reported finding that origin-dependent DNA replication is important for high levels of flanking gene transcription (17). The statistically significant decrease in the ORF63 expression level seen in the presence of PAA using the Sp1/Sp3 mutation also suggests that the involvement of the Sp1/Sp3 site in ORF63 expression is separate from its enhancement of origin-dependent DNA replication.

In the next set of experiments, to explore which elements of the VZV infection process might be responsible for the effects that we had seen, we examined the influence of Sp1/Sp3 and YY1 site mutations on the expression levels of the ORF62 and ORF63 genes in the context of ORF62 transfection alone. IE62, encoded by ORF62, is the major VZV transactivator, which has the ability to activate most, and perhaps all, of the VZV genes (6, 26, 33–35, 38). IE62 was shown previously to activate both the ORF62 and ORF63 genes in the absence of any other viral protein (18–20, 27). The reporter vectors were cotransfected along with an ORF62-expressing plasmid and the pCMV β-Gal-expressing control plasmid into MeWo cell monolayers; at 48 h posttransfection, the cells were lysed, and measurements of luciferase and β-galactosidase were taken. The results of these assays are shown in Fig. 4A and B. Levels of luciferase activity obtained from each pLitmus reporter plasmid in the absence of ORF62 transfection represented the basal levels from this plasmid and were normalized to a value of 1. The activities of the reporters in the presence of ORF62 transfection are reported as the induction (n-fold) of luciferase activities in reference to the basal level of activity without ORF62 transfection.

Fig 4.

Effects of Sp1/Sp3 and YY1 site mutations on flanking gene expression in the context of ORF62 transfection. (A) Results of triplicate assays assessing the effects of the presence of the Sp1/Sp3 and YY1 site mutations on the expression levels of the Renilla luciferase reporter gene present at the position of the ORF62 gene. (B) Results of triplicate assays assessing the effects of the presence of the Sp1/Sp3 and YY1 site mutations on the expression levels of the firefly luciferase reporter gene present at the position of the ORF63 gene. (C) Absorbance at 405 nm representing the β-galactosidase activity used as a normalizer for the triplicate reporter gene assays done. The promoter activities resulting from the presence of transfected ORF62 are reported as the induction (n-fold) of luciferase activity over the basal level (without ORF62 transfection). Statistical significance was determined by a one-way ANOVA followed by Tukey's post hoc test. Error bars indicate standard errors.

Our first finding from these experiments was that the ORF62 expression level using the wild-type pLitmus reporter in the context of ORF62 transfection is very low compared to the level in the context of VZV superinfection (Fig. 3). These data suggest, not surprisingly, the involvement of viral factors other than ORF62 in the expression of the ORF62 gene during normal VZV replication and also reflect the ability of IE62 to autoregulate its promoter. Second, and similarly to the reporter gene assays done in the presence of PAA, the Sp1/Sp3 site mutation resulted in a decrease in the expression level of ORF63 only and not of ORF62. On the other hand, the YY1 site mutation had no effect on the expression of either ORF62 or ORF63 in the context of ORF62 transfection. There was no significant change in the expression level of β-Gal (used as a normalizer) by increasing amounts of IE62 (Fig. 4C). These data imply the utilization of the cellular transcription factors Sp1 and Sp3 by ORF62 for ORF63 promoter activation.

Sp1/Sp3 and YY1 site mutant viruses influence VZV replication in MeWo cells and in skin xenografts.

Since we had shown significant effects of these mutations in biochemical assays, it was now important to study their effects on viral replication by using mutant viruses. Two mutants were constructed, one with the Sp1/Sp3 site mutation and one with the YY1 site mutation. pSpe23 cosmids containing either the Sp1/Sp3 or YY1 mutation were transfected with the intact pFsp73, pPme2, and pSpe14 cosmids in melanoma cells; pSpe23 containing the wild-type sequence was transfected along with the other cosmids as a control (recombinant wild-type virus). Recombinant viruses, designated the pOka Sp1/Sp3 mutant and the pOka YY1 mutant, were recovered from the transfections. The growth kinetics of wild-type pOka and the Sp1/Sp3 and YY1 mutant viruses in MeWo cells were evaluated (Fig. 5A). The titer of the Sp1/Sp3 mutant virus was significantly lower than that of wild-type strain pOka for the length of the study (days 1 to 5). In contrast, the titer of the YY1 mutant virus was similar to that of wild-type strain pOka over the period of infection. These results indicate that the Sp1/Sp3 site mutation, which influenced origin-dependent DNA replication as well as ORF62 and ORF63 expression levels, also impaired VZV replication in vitro. The YY1 mutation had no effect on the virus in this system, reflecting its behavior in biochemical assays.

Fig 5.

Effects of VZV pOka Sp1/Sp3 and YY1 mutations on virus replication in MeWo cells and in skin xenografts in the SCID mouse model. (A) Growth kinetics of pOka, the pOka Sp1/Sp3 mutant, or the pOka YY1 mutant virus in MeWo cells. Cells were inoculated with wild-type and mutant viruses at 10³ PFU/ml, and infectious-virus yields were determined for 5 days after inoculation. (B) Skin xenografts were inoculated with pOka, the pOka Sp1/Sp3 mutant, or the pOka YY1 mutant. Day 0 indicates the inoculum titers of each virus. Bars at day 10 and day 21 represent the mean titers of infectious virus ± standard deviations recovered from at least three xenografts harvested at these time points after virus inoculation. ** represents a highly significant increase (P < 0.01), and *** represents a more highly significant increase (P < 0.001).

To determine whether Sp1/Sp3 and YY1 site mutations affected VZV pathogenesis, the growths of the Sp1/Sp3 and YY1 pOka mutants were next compared to that of wild-type pOka in human skin xenografts (Fig. 5B). The titers of all virus inocula were similar. Wild-type pOka titers were 3.66 × 10³ PFU/implant at day 10 and 3.79 × 10³ PFU/implant at day 21. The titer of the pOka Sp1/Sp3 site mutant was 2.72 × 10³ PFU/implant at day 10, which was statistically significantly lower than that of wild-type pOka. The titer was 3.43 × 10³ PFU/implant at day 21, which is slightly lower than that of wild-type strain pOka but not statistically significant. The titer of the pOka YY1 site mutant, however (3.44 × 10³ PFU/implant), was not statistically significantly different from the titer of wild-type pOka at day 10. At day 21, however, the titer was 4.40 × 10³ PFU/implant at day 21; this is statistically significantly higher than that of wild-type pOka. The YY1-based difference at day 21 may reflect a growth advantage over wild-type pOka or perhaps slower growth initially, which leaves more tissue available for infection at later times. These data indicate that the Sp1/Sp3 site is important for VZV replication in vivo and that the YY1 site may also affect growth in vivo.

Sp1/Sp3 and YY1 site mutant viruses show low IE63 expression and total virus DNA levels.

To assess if Sp1/Sp3 and YY1 site mutations affected the expression levels of the ORF62 and ORF63 genes flanking the oriS, the kinetics and the levels of IE62 and IE63 protein expression in MeWo cells following mutant virus infection were also tested. MeWo cells were infected with pOka and the Sp1/Sp3 and YY1 mutant viruses and analyzed for IE62 and IE63 expression levels. The expression level of IE63 with the Sp1/Sp3 mutant virus was reduced significantly (about 3-fold, based on triplicate densitometer analyses), at both 12 and 36 h postinfection, compared to wild-type pOka (Fig. 6A and B). In contrast, the level of IE62 expression did not change at these two time points (12 and 36 h) with the Sp1/Sp3 mutant virus (Fig. 6A and B). With the YY1 mutant virus, neither IE62 nor IE63 expression levels changed significantly at 12 and 36 h postinfection (Fig. 6A and B).

Fig 6.

Effects of Sp1/Sp3 and YY1 site mutations on IE62 and IE63 expression levels in MeWo cells during the course of VZV infection. Western blot analyses show the expression levels of IE62, IE63, and β-tubulin, and the histograms summarize data from triplicate experiments done for each virus at 12 h (A) and 36 h (B) postinfection. β-Tubulin was used as a loading control in the experiments. The blots were scanned by densitometry to obtain quantitative data (in triplicate). Statistical significance was determined by a one-way ANOVA followed by Tukey's post hoc test.

Finally, the influence of Sp1/Sp3 and YY1 site mutations on the virus DNA level was tested by using Southern blot analyses. Total cellular DNA was prepared from MeWo cells infected with the wild-type and the Sp1/Sp3 and YY1 mutant pOka viruses and probed for total virus DNA. As shown in Fig. 7A and B, the virus DNA level decreased significantly (about 2-fold) in the Sp1/Sp3 site mutant virus at 24 h postinfection compared to the wild-type level. On the other hand, there was a significant increase in the total virus DNA level (by about 60%) with the YY1 site mutant virus.

Fig 7.

Effects of Sp1/Sp3 and YY1 site mutations on the total virus DNA level in MeWo cells 24 h after VZV infection. (A) Southern blot analyses showing the total virus DNA levels for wild-type, Sp1/Sp3 mutant, and YY1 site mutant virus infections of MeWo cells. (B) Histograms summarizing data from triplicate analyses done for each virus. An ORF3 promoter probe was used to determine the total virus DNA level. An E2F3 probe was used in the loading control experiments. Error bars indicate standard errors. Statistical significance was determined by a one-way ANOVA followed by Tukey's post hoc test.

DISCUSSION

The architectures of the origins of DNA replication in the genomes of the two closely related human alphaherpesviruses VZV and HSV exhibit significant differences (Fig. 1). While both viruses contain AT-rich stretches, the numbers and positions of sites or potential sites for interactions with the VZV or HSV OBPs relative to these AT-rich stretches differ with respect to their numbers, their relative orientations, and the requirements for their presence in origin-dependent replication. This is despite the fact that the VZV and HSV OBPs recognize essentially the same sequence. Specifically, the box I-, box II-, and box III-binding sites are all necessary for efficient origin-dependent DNA replication in HSV-1 replication (1, 9, 13, 23, 31, 41, 43).

In contrast, only box A is required for VZV origin-dependent replication, with box B being dispensable and box C playing an auxiliary role (43). In addition, the arrangement and orientation of the three upstream OBP-binding sites in VZV preclude the formation of a hairpin, which is involved in OBP binding and origin activation in HSV-1 (31). There also appear to be significant differences in the influence of DNA sequences within or near the origin on the expression levels of flanking genes. The presence of HSV-1 origins was reported previously to have no influence on the expression levels of flanking genes (29, 45). In the case of VZV, however, previous work by Jones et al. (15), using an identical experimental strategy involving recombinant viruses, showed that the presence of VZV origin sequences had a significant effect on flanking gene expression and that some elements within or near the minimal origin may act as cis elements in the expression of flanking genes.

The function or functions of the sequence(s) downstream of the AT-rich stretch within VZV oriS and the possibility that these sequences can compensate for the absence of a downstream OBP site were unknown prior to the work presented here. The VZV OBP does not bind to this sequence (5, 43), but the plasmids used in the original identification of the minimal VZV oriS contained this 125-bp downstream sequence, which included the Sp1/Sp3 and YY1 sites. We therefore wished to determine if the Sp1/Sp3- and YY1-binding sites that we studied previously (16) were involved in origin-dependent DNA replication and/or in the expressions of the ORF62 and ORF63 genes, which flank VZV oriS.

The results of the DpnI replication assays using the pLitmus plasmids containing the intergenic region between ORF62 and ORF63 indicated that the Sp1/Sp3 site acts as an enhancer of origin-dependent DNA replication, while the YY1 site has no significant role in the process. This differs from our previous findings with the pVO2 plasmid lacking the promoter elements of the flanking genes (16); there could be several reasons for this. First, the oriS structure in the pVO2 plasmid is from the Dumas strain, while plasmid pLitmus is from the pOka strain. There is, however, no difference in the sequences of the Sp1/Sp3 and YY1 sites within the oriS downstream region between these two strains (10), excluding the possibility of simple strain variation. However, in the pLitmus plasmid, there are the promoter elements of the ORF62 and ORF63 genes, which are not present in the pVO2 plasmid. This suggests that the Sp1/Sp3 site (and, possibly, flanking gene transcription) may compensate for the absence of an OBP site in the downstream region of VZV oriS.

We observed previously (17) that the VZV DNA polymerase inhibitor PAA inhibited the expressions of ORF62 and ORF63 in a nonlinear correlation with the decrease in DNA replication seen. In confirmation, in this study, we showed that the presence of PAA caused 7- and 5-fold decreases in the expression levels of the ORF62 and ORF63 reporters, respectively. If the decreases were due solely to less available total plasmid, they would be of the order of 25 to 30%. These results support and extend our previous findings that origin-dependent DNA replication and flanking gene transcription appear to be coupled in VZV.

We followed these assays with the wild-type virus by using Sp1/Sp3 and YY1 mutant constructs. The mutation of the Sp1/Sp3-binding site within the oriS structure of HSV does not affect flanking gene transcription (29). In contrast, our results with the pLitmus dual-luciferase reporter plasmids showed that the mutation of the Sp1/Sp3 site negatively affected the transcription of both the ORF62 and ORF63 reporters under active DNA replication (the absence of PAA). Expression from the ORF62 and ORF63 reporters should be inhibited by about 10% if the sole cause was a lower DNA replication efficiency leading to less available total plasmid. The data showed that the decrease in the expression level was ∼2- to 3-fold, suggesting that transcription regulated by the Sp1/Sp3 site is coupled to viral DNA replication; this regulation involves both sides of oriS in the presence of active DNA replication. The data again confirm that active DNA replication is required for efficient ORF62 and ORF63 expressions.

Interestingly, in the absence of viral DNA replication (in the presence of PAA or in the context of ORF62 transfection alone), there was a statistically significant effect of the Sp1/Sp3 mutation on ORF63 transcription only and not on ORF62 transcription. This fits our data reported previously for the downstream box D site, where we showed effects on both origin-dependent DNA replication and flanking gene transcription (17). Furthermore, it appears that the influence of the Sp1/Sp3 mutation on ORF62 expression is due solely to the involvement of the Sp1/Sp3 site in origin-dependent DNA replication and that origin-dependent DNA replication is required for efficient ORF62 expression.

Sp1/Sp3 sites have been identified in the replication origins of other herpesviruses (e.g., EBV and HSV-1) (2, 29). Sp1, and possibly other Sp family members, affects origin-dependent EBV DNA replication at the cis-acting element oriLyt (2, 12). This element consists of two essential domains, designated the upstream and downstream components. The upstream component contains DNA-binding motifs for the EBV transcriptional activator BZLF1. The downstream component is known to be the binding site of the cellular transcription factors ZBP-89 and Sp1, which stimulate DNA replication. This stimulation is believed to result from the recruitment of viral DNA replication proteins, since the direct interaction of Sp1 and ZBP-89 with the viral DNA polymerase and its processivity factor was demonstrated previously in protein binding assays (2). The HSV-1 genome contains two oriS core-adjacent regulatory (Oscar) elements, OscarL and OscarR, at the base of the oriS palindrome. The mutation of either element reduced oriS-dependent DNA replication. OscarL contains a consensus binding site for Sp1, and electrophoretic mobility shift assays (EMSAs) and supershift experiments showed the binding of Sp1 and Sp3 to OscarL (29).

Sp1 is involved in lytic VZV replication, likely through its role in the expressions of several important VZV promoters, including those for the viral glycoproteins gI and gE, the VZV major single-strand-binding protein, and the VZV DNA polymerase catalytic subunit (3, 14, 32, 36, 37). Sp1 was shown previously to interact with the VZV major transcription activator IE62 (28, 32), and there is a high incidence of predicted Sp1-binding sites within VZV promoters (37). The role of other Sp family members in VZV infection and replication is still largely unknown, but the demonstration of Sp3 binding to the Sp1/Sp3 site in the downstream region of oriS provides evidence for a role for this transcription factor in the VZV life cycle (16).

In contrast to our results with the Sp1/Sp3 site, the mutation of the YY1 site, which ablated the binding of the YY1 protein to the downstream region of VZV oriS, had no significant effect on either origin-dependent DNA replication or flanking gene transcription. YY1 is a 414-amino-acid zinc finger protein that is capable of the repression and activation of gene transcription in several virus systems, including adeno-associated virus (AAV), human papillomavirus (HPV), parvovirus B19, HSV-1, and murine leukemia virus (MuLV) (21, 39). YY1 was also shown previously to inhibit HPV ori-dependent DNA replication in a cell-free replication system (21).

The role or roles played by YY1 in VZV infection are unknown; thus, the identification of a bona fide YY1-binding site within the oriS downstream region site afforded the potential to identify a VZV-specific function for this ubiquitous cellular factor. The mutation of the YY1 site with the concomitant loss of YY1 binding, however, did not result in a statistically significant effect on replication or gene expression under our experimental conditions. However, this does not eliminate the possibility that YY1 plays an important role in VZV infection. Alternatively, YY1 could directly affect origin-dependent DNA replication and flanking gene transcription in primary cells or in specific tissues infected by VZV in the human host. Our data in support of these possibilities include the significant increase in mutant virus titers in human skin xenografts at day 21 postinfection and the slight but significant increase in the total virus DNA level. The increase in the virus DNA level achieved in the presence of the YY1 mutant virus might explain the increases in virus titers seen in skin xenografts at day 21 postinfection. The results obtained with the Sp1/Sp3 site mutation in the in vitro assays and the in vivo studies indicate that this site is important not only for DNA replication and flanking gene expression but also for VZV growth and pathogenesis. Together, our findings suggest that the Sp1/Sp3 site mutation may be a candidate for a targeted change in pOka that would provide a genetically defined mechanism (low IE63 expression and virus DNA levels) for the further attenuation of vaccine strain Oka.

The data gathered on the ability of Sp1 and Sp3 to bind to the Sp1/Sp3 site in the downstream region of oriS and their influence on oriS-dependent DNA replication and flanking gene expression lead us to propose a model for the role of the Sp1/Sp3 site in VZV infection (Fig. 8). At times immediately after infection, endogenous Sp1 and Sp3 bind to the Sp1/Sp3 site. Once a sufficient amount of IE62 is synthesized, the most active phase of viral gene expression starts, and at this stage, Sp1 and Sp3 interact physically and functionally with IE62 to activate the ORF63 promoter in the absence of active DNA replication. The data supporting this step are the involvement of the Sp1/Sp3 site in ORF63 expression in the presence of PAA (Fig. 3) and in the context of IE62 transfection alone (Fig. 4). The activation of ORF63 expression through the Sp1/Sp3 site during the early stage of infection, coupled with the expressions of the seven VZV replication factors, leads to the enhancement of origin-dependent DNA replication. The data supporting this step are the involvement of the Sp1/Sp3 site in ORF63 expression, which is required for the enhancement of origin-dependent DNA replication (compare our previously reported results [16] to the results shown in Fig. 2). Efficient DNA replication then further enhances ORF62 and ORF63 expression levels during the early and late stages of infection. We note that in the presence of PAA or an Sp1/Sp3 site mutation, the expressions of ORF62 and ORF63 are inhibited.

Fig 8.

Model for the role played by the Sp1/Sp3 site in the downstream region of VZV oriS in origin-dependent DNA replication and flanking gene expression. (A) The formation of Sp1/Sp3 complexes influences ORF63 expression at the immediate-early stage of infection. (B) The presence of Sp1/Sp3 complexes, coupled with ORF63 expression activation, then enhances origin-dependent DNA replication and compensates for the absence of a downstream OBP box at the early stage of infection. (C) The presence of active origin-dependent DNA replication in turn enhances both ORF62 and ORF63 expression levels at the early and late stages of infection.

Finally, one unresolved issue that may be explained by our data is the finding reported in 1986 by Stow and Davidson (42) that HSV can utilize VZV oriS in DNA replication but that VZV cannot use an HSV origin. We have shown that the Sp1/Sp3 site is required for VZV oriS-dependent DNA synthesis. Perhaps, the absence of an equivalent Sp1/Sp3 site in the downstream region of HSV oriS may be a reason why VZV is unable to use the HSV site.