The Transcriptional Activation Domain of VP16 Is Required for Efficient Infection and Establishment of Latency by HSV-1 in the Murine Peripheral and Central Nervous Systems

Virology 259, 20–33 (1999) Article ID viro.1999.9756, available online at http://www.idealibrary.com on

The Transcriptional Activation Domain of VP16 Is Required for Efficient Infection and Establishment of Latency by HSV-1 in the Murine Peripheral and Central Nervous Systems Ruth Tal-Singer,* ,1 Rath Pichyangkura,† ,2 Eugene Chung,† ,3 Todd M. Lasner,* ,‡ ,1 Bruce P. Randazzo,* ,§ ,4 John Q. Trojanowski, ¶ Nigel W. Fraser,* ,‡ ,5 and Steven J. Triezenberg† ,6 *The Wistar Institute, 3601 Spruce Street, Philadelphia, Pennsylvania 19104; †Department of Biochemistry, Michigan State University, East Lansing, Michigan 48824–1319; ‡Division of Neurosurgery, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania 19104; and §Department of Dermatology and ¶Department of Pathology and Laboratory Medicine, Division of Anatomic Pathology, University of Pennsylvania, Philadelphia, Pennsylvania 19104 Received January 28, 1999; returned to author for revision March 15, 1999; accepted April 8, 1999 The herpes simplex virus (HSV) transactivator VP16 is a structural component of the virion that activates immediate-early viral gene expression. The HSV-1 mutant in1814, which contains a 12-bp insertion that compromises the transcriptional function of VP16, replicated to a low level if at all in the trigeminal ganglia of mice (I. Steiner, J. G. Spivack, S. L. Deshmane, C. I. Ace, C. M. Preston, and N. W. Fraser (1990). J. Virol. 64, 1630–1638; Valyi-Nagy et al., unpublished data). However, in1814 did establish a latent infection in the ganglia after corneal inoculation from which it could be reactivated. In this study, several HSV-1 strains were constructed with deletions in the VP16 transcriptional activation domain. These viruses were viable in cell culture, although some were significantly reduced in their ability to initiate infection. A deletion mutant completely lacking the activation domain of VP16 (RP5) was unable to replicate to any detectable level or to efficiently establish latent infections in the peripheral and central nervous systems of immunocompetent mice. However, similar to in1814, RP5 formed a slowly progressing persistent infection in immunocompromised nude mice. Thus RP5 is severely neuroattenuated in the murine model of HSV infection. However, the activation domain of VP16 is not essential for replication in the nervous system, since we observed a slow progressive infection persisting in the absence of an immune response. © 1999 Academic Press

Fraser and Valyi-Nagy, 1993; Whitley and Gnann, 1993; Whitley and Lakeman, 1995). HSV infects the human host through the epithelium that lines the skin or mucous membranes. This initial infection results in a regulated cascade of viral gene expression, beginning with activation of viral immediateearly (IE or a) genes, the products of which in turn activate the early (E or b) genes. Many E gene products are involved in viral DNA replication, which is followed by high level expression of the late (L or g) genes (Roizman and Sears, 1996). VP16 (also known as a-TIF, ICP25, and Vmw65), a 490-amino-acid phosphoprotein component of the virion tegument (Batterson and Roizman, 1983), activates transcription of the immediate-early genes (reviewed in Ward and Roizman, 1994; Roizman and Sears, 1996). VP16 mediates transinduction through specific promoter elements containing TAATGARAT (R 5 purine) and GCGGAA sequences (Mackem and Roizman, 1982; Gaffney et al., 1985; Triezenberg et al., 1988b). VP16 does not bind directly to either of these conserved motifs, although it makes minimal contact with the DNA (Marsden et al., 1987; Stern and Herr, 1991). At the TAATGARAT elements, a domain of VP16 located between amino acids 1–400 forms a multicomponent DNA-binding complex with the

INTRODUCTION The nervous system plays a central role in the pathogenesis of herpes simplex virus (HSV) infections. Both serotypes of the virus, HSV-1 (the oral form) and HSV-2 (the genital form) are capable of life-long persistence within sensory peripheral ganglia in a nonreplicative latent state. In response to environmental or cellular changes, both serotypes can reactivate to produce recurrent epithelial eruptions and virus shedding. In addition, the central nervous system is a major target for HSV-related morbidity and mortality resulting from encephalitis (for reviews, see Hill, 1985; Fraser et al., 1991;

1 Present address: Department of Molecular Virology and Host Defense, SmithKline Beecham Pharmaceuticals, Collegeville, PA 19426. 2 Present address: Department of Biochemistry, Faculty of Science, Chulalongkorn University, Payathai Road, Bangkok 10330 Thailand. 3 Present address: Thomas Jefferson Medical School, Philadephia, PA. 4 Present address: Jansen Research Foundation, 1125 Trenton-Harbortown Road, Titusville, NJ 08560. 5 Present address: Department of Microbiology, Medical School, University of Pennsylvania, 319 Johnson Pavilion, 3601 Hamilton Walk, Philadelphia, PA 19104-6076. 6 To whom reprint requests should be addressed. Fax: 517-353-9334. E-mail: [email protected].

0042-6822/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

20

NEUROPATHOGENICITY OF VP16 ACTIVATION DOMAIN MUTANTS OF HSV-1

cellular protein Oct-1 as the major determinant of DNA binding (O’Hare and Goding, 1988; Preston et al., 1988; Pruijn et al., 1988; Kristie et al., 1989; Stern et al., 1989) and with HCF (also called VCAF-1 and C1), which facilitates the interaction between VP16 and DNA-bound Oct-1 (Katan et al., 1990; Xiao and Capone, 1990; Kristie and Sharp, 1993; Wilson et al., 1993). VP16 contributes to this complex a potent transcriptional activation domain, located between amino acids 410 and 490, that interacts with targets within the basal transcription machinery (Cousens et al., 1989; Triezenberg et al., 1988a; Triezenberg, 1995). The second promoter element, comprising the GCGGAA motif, binds a heterodimeric protein known as GABP (LaMarco et al., 1991; Thompson et al., 1991) or NRF-2 (Virbasius et al., 1993), but no physical association of VP16 with this DNA-binding protein has been reported. In addition to its transcriptional activation function, VP16 interacts with the virion-associated virion host shutoff protein (vhs) and restrains its function (Smibert et al., 1994; Lam et al., 1996; Schmelter et al., 1996). Several studies have explored the properties of HSV variants bearing mutations in VP16. The HSV-1 variant 8MA (derived from HSV-1 KOS) is a VP16-null mutant that requires propagation on cells that provide VP16 in trans (Weinheimer et al., 1992). An HSV-2 strain (ts13) reportedly bears a mutation in the HSV-2 VP16 gene that renders the virion thermolabile in vitro but had no effect on IE gene activation (Moss, 1989). Other temperaturesensitive mutations in VP16 were constructed by replacing certain Cys residues with Gly residues (Poon and Roizman, 1995). The temperature-sensitive effect on viruses expressing these VP16 genes was traced to events late in the replicative cycle and not to the IE stage, despite the fact that the mutant VP16 proteins were defective in forming complexes with HCF and Oct-1 at IE promoters. The replication-competent mutant in1814 (derived from HSV-1 strain 17 1), which contains a 4-aminoacid insertion at codon 397 of VP16, produces a VP16 protein that is incorporated into mature virions but lacks transinducing activity (Ace et al., 1989). No detectable replication of strain in1814 occurred in the trigeminal ganglia (TG) of mice after corneal inoculation (Steiner et al., 1990). However, subsequent and more sensitive experiments detected a low level of replication at early times postinfection (Valyi-Nagy et al., unpublished observations). Nevertheless, in1814 established a reactivationcompetent latent infection in the TG (Steiner et al., 1990; Valyi-Nagy et al., 1991a). Truncation of the VP16 gene at codon 422 (deleting most of the transcriptional activation domain) resulted in a virus with reduced infectivity in cultured cells (Lam et al., 1996) and reduced IE gene activation, both of which could be compensated by the presence of hexamethylene bisacetamide, properties shared with in1814 (Smiley and Duncan, 1997). The properties of this virus in in vivo infections using an animal model system have not yet been described.

21

In this study, we investigated the in vitro and in vivo properties of HSV-1 mutants with deletions of various portions of the activation domain of VP16. These viruses were all capable of replicating in tissue culture cells, although infectivity was significantly reduced when the activation domain was deleted (strain RP5). RP5 failed to activate either transfected IE reporter genes or the IE genes in the viral genome. In contrast to in1814, RP5 was unable to establish a latent infection in neuronal cells following corneal inoculation of immunocompetent mice. Furthermore, the ability of RP5 to cause clinical disease and to establish latency after intracranial inoculation in immunocompetent mice was markedly reduced compared with in1814. These data support the hypothesis that the insertion in in1814 does not completely destroy the transactivation function and that the transcriptional activation domain in VP16 is a neurovirulence determinant. RESULTS Construction of HSV strains bearing deletions of the VP16 activation domain Previous reports have described an HSV-1 strain termed in1814 bearing a four-codon insertion in the VP16 gene (Ace et al., 1989; Steiner et al., 1990; Valyi-Nagy et al., 1991b). This insertion at codon 397 abolished the ability of VP16 to trans-activate reporter plasmids with IE gene promoters, apparently by disrupting the association of VP16 with the cellular transcription factor Oct-1 through which VP16 is tethered to IE promoters. This virus strain was viable in cell culture but infected cells inefficiently at low multiplicities of infection (Ace et al., 1989). This virus also showed reduced virulence in mice (Steiner et al., 1990; Valyi-Nagy et al., 1991b). Although no replication of this virus was detected in murine TG after corneal inoculation, the virus did establish and reactivate from latency as efficiently as wild-type viruses (Steiner et al., 1990). Because the attenuated phenotype of in1814 was most apparent only at low m.o.i.s, we wished to determine whether its properties in vivo (ability to replicate at the periphery, ability to establish latent infections) might be due to partial activity of the transcriptional activation domain, which remained intact despite the observed defects in IE gene activation. Therefore we constructed by homologous recombination a set of viral mutants in which the activation domain of VP16 (residues 413–490) was either partially or entirely deleted (Fig. 1A). RP1 is a control strain bearing an additional BamHI restriction site and a short oligonucleotide segment bearing stop codons in all reading frames, located 19 bp downstream of the natural stop codon for the VP16 open reading frame. The strain designated RP3 is truncated at codon 456 and thus lacks the second subregion of the VP16 activation domain termed H2 (Walker et al., 1993) or

22

TAL-SINGER ET AL.

FIG. 1. Virus strains used in this study. (A) Schematic representation of HSV-1 genome. The unique regions (U L and U S) are each flanked by inverted repeats denoted as blank boxes. In HSV-1 strain 8MA the VP16 gene has been replaced by the E. coli lacZ gene regulated by the HSV-1 ICP6 promoter. Strain RP1 is a wild-type revertant of 8MA bearing a BamHI site located 19 bp downstream of the VP16 gene. Strain RP3 bears a deletion of codons 456–490 of the VP16 gene, RP4 lacks codons 413–452, and RP5 lacks codons 413–490 of the VP16 gene. RP3 and RP5 also lack 19 bp of 39 untranslated sequence. The wild-type fragment from the HSV-1 KOS genome, used to rescue RP3, RP4, and RP5, is shown at bottom. (B) Southern blot of virion DNAs from wild-type and recombinant HSV strains. DNA extracted from equivalent numbers of virions for each strain was digested with BamHI, electrophoresed in agarose gels, blotted to nitrocellulose, and hybridized with a radiolabeled probe from the VP16 gene. M indicates molecular weight markers. (C) Southern blot of virion DNAs from mutant and rescued viruses, performed as for (B). (D) Immunoblot of VP16 proteins present in mutant virions. Equivalent numbers of virions (;1 3 10 8) were lysed and electrophoresed in denaturing acrylamide gels, electroblotted to nitrocellulose, and probed with a rabbit serum (C8–5; Triezenberg et al., 1988a) directed against purified virion VP16 protein. RP5/16–8 designates viral strain RP5 grown in the cell line 16–8, which expressed wild-type VP16 protein. In control experiments, serial dilutions of virion samples gave proportional signals in immunoblots (data not shown).

VP16C (Sullivan et al., 1998). Strain RP4 lacks the first subregion of the activation domain (H1 or VP16N), corresponding to amino acids 413–452. Strain RP5 lacks the entire activation region (residues 413–490). For each of the deletion mutants, rescued viruses were also constructed using a corresponding fragment from the wildtype KOS strain. Southern blots of virion DNA clearly demonstrate the identities of each of the mutant and rescued strains (Figs. 1B and 1C). Immunoblot analysis of virion proteins, developed using an antibody directed against VP16, revealed that for each virus strain the corresponding protein is incorporated into virions at approximately wild-type levels (Fig. 1D). Notably, RP5 is viable in noncomplementing Vero cells, with efficient production of virions using standard protocols for prep-

aration of virus stocks (Fig. 1D and Table 1). When this strain was grown in 16–8 cells, which express a stably transformed VP16 gene, the resultant virions bore both truncated and full-length VP16 at roughly a 2:1 ratio (Fig. 1D, RP5/16–8). The deletions in the VP16 activation domain affected virus infectivity but not the eventual yield of virions from monolayers of infected cells (Table 1). The RP3 virus, which lacks the C-terminal activation subregion, was not obviously affected in its in vitro growth properties compared with the wild-type counterpart RP1. RP4 showed a modest reduction in infectivity, with ratio of particles to plaque-forming units (PFU) approximately 40 times higher than the wild-type virus. RP5, lacking the VP16 activation domain altogether, was more seriously debil-

NEUROPATHOGENICITY OF VP16 ACTIVATION DOMAIN MUTANTS OF HSV-1 TABLE 1 Titers and Virion Concentrations of Recombinant Viral Strains Grown in Vero Cells a Virus

Titer (PFU/ml)

Virions per ml

Virion/PFU ratio

RP1 RP3 RP4 RP5 RP4R RP5R

3.9 3 10 8 5.8 3 10 8 3.3 3 10 7 3.1 3 10 6 1.1 3 10 9 1.1 3 10 9

3.4 3 10 9 5.3 3 10 9 1.2 3 10 10 3.6 3 10 9 6.1 3 10 9 6.3 3 10 9

8.6 9.1 360 1200 5.5 5.7

a

Viral strains were titered on Vero cells by plaque assays. Virions were counted by electron microscopy using latex bead concentration standards. Results reflect triplicate assays of a single representative stock of each virus.

itated, yielding virus stocks whoseinfectivity was reduced by more than two orders of magnitude. This deficit was more pronounced at lower m.o.i.s, asrevealed by the growth curves shown in Fig. 2A. The defects in infectivity shown by RP4 and RP5 were not overcome by the VP16 protein expressed in the complementing cell line 16–8;

23

titers of virus stocks produced in Vero cells were the same when assayed on 16–8 cells as on Vero cells (Fig. 2B). This result implies that the virion-borne VP16 must have properties important for infection that the stably integrated VP16 gene cannot provide. However, when the RP4 and RP5 strains were grown in 16–8 cells, the resulting viruses had high titers indistinguishable from those of wild-type virus (data not shown). The defects in infectivity of RP4and RP5 cannot be attributed to binding, entry, or nuclear localization of the viral genome because Southern blots of DNA isolated from nuclei of infected cells showed equivalent amounts of viral DNA when equivalent particle numbers were used to initiate the infection (Fig. 2C). Despite the differences in infectivity, when monolayers were allowed toproceed to full cytopathic effect (which took considerably longer for the RP4 and RP5 viruses), the numbers of virions produced by RP4 and RP5 were equivalent to the numbers produced from wild-type viruses (Table 1), and the amounts of viral DNA and of VP16 protein contained in those virions were also approximately equivalent (Figs. 1B and 1D). These observations suggest that the RP4 and RP5 viruses have

FIG. 2. Infectivity of HSV-1 strains bearing deletions in the VP16 activation domain. (A) Growth curves of HSV-1 strains RP1 (represented by diamond symbols), RP5 (squares), and RP5 grown in 16–8 cells (circles). Vero cells were infected at either high multiplicity (solid symbols and lines; 10 PFU per cell for RP1 and RP5/16–8, 1 PFU per cell for RP5) or low multiplicity (open symbols and broken lines; 0.1 PFU per cell). At various times after infection, cells and supernatants were harvested, cells were lysed by sonication, and the titer of virus in the extract was determined by plaque assays in Vero cells. (B) Stocks of the recombinant viruses RP1, RP3, RP4, and RP5 were prepared by infecting Vero cell monolayers at low multiplicity and harvesting virus after extensive cytopathic effect was evident. The titers of these stocks were assessed on both Vero cells and the VP16-expressing cell line 16–8 by plaque assay. (C) Viral DNAs present in preparations of nuclear DNA from infected cells. Four hours after infection of Vero cells, nuclei were isolated and DNA was extracted. Southern blot hybridizations were performed as in Fig. 1.

24

TAL-SINGER ET AL.

FIG. 3. Activation of transfected reporter genes by recombinant viruses bearing deletions in the VP16 activation domain. Mouse L cells were transfected with reporter gene plasmids expressing the bacterial chloramphenicol acetyltransferase (CAT) gene under the control of the ICP0 promoter (A), the ICP4 promoter (B), the ICP22 promoter (C), or the ICP27 promoter (D). Forty hours after transfection, cells were infected by each of the recombinant virus strains shown in the presence of cycloheximide to prevent synthesis of viral IE proteins. Three hours later, cycloheximide was removed and actinomycin D was added to permit the translation of expressed mRNAs without allowing further transcription. Two hours later, cell extracts were prepared and assayed for CAT enzymatic activity by the fluor-diffusion method. Bars represent means of at least three independently transfected and infected cell cultures. Primer extension experiments (E) were also performed by simultaneously transfecting L cells with three reporter gene plasmids expressing the HSV thymidine kinase gene under the control of the ICP0 promoter, the ICP4 promoter, and the Moloney sarcoma virus long terminal repeat. Forty hours after transfection, cells were infected with the recombinant virus strains in the presence of cycloheximide. After 2 h, total cell RNA was isolated and assayed by primer extension using a primer specific to the tk gene transcript. The positions of the primer extension products corresponding to transcripts from each of the reporter plasmids are indicated at the right.

no significant deficit in virion formation or packaging of viral DNA. Transcriptional activation by VP16 mutant viruses An obvious hypothesis is that the defects in infectivity result from reduced transcriptional activation of IE genes by the mutant VP16 proteins. To assess how well these viruses can activate IE gene promoters, plasmids expressing the CAT reporter gene under control of each of the IE promoters were transfected into mouse L cells. These cells were then infected with equivalent numbers of virions of each recombinant virus. Figure 3 shows that each of the various reporter plasmids responded to VP16 activation by the wild-type virus with effects ranging from 1.5-fold activation (ICP22-CAT) to 160-fold (ICP27-CAT). Truncation of VP16 at residue 456 (RP3) had little or no effect on transcriptional activation, except on the ICP27CAT reporter where activation fell from 160- to 60-fold. However, neither RP4 nor RP5 effectively activated any of the IE-CAT reporter genes. These results were confirmed using a different reporter gene system, in which the ICP0 and ICP4 promoters were linked to the thymidine kinase gene. Primer extension assays (Fig. 3E) demonstrated that RP1 and RP3 effectively activated expression from both promoters. In contrast, neither RP4 nor RP5 activated expression from either reporter gene, although the RP5R rescued strain and the RP5 virus grown in the complementing cell line 16–8 activated expression as effectively as the wild-type viruses. The failure of RP4 to activate transfected reporter plasmids was surprising given that the RP4 virus grows

reasonably well in culture and the transcriptional activation subregion present on the RP4 VP16 protein can function in similar transfection experiments when fused to the Gal4 DNA binding domain (Sullivan et al., 1998). To assess the ability of RP4 and the other recombinant viruses to activate the IE promoters in the context of the virus itself, we assayed viral IE gene expression by reverse transcription followed by polymerase chain reaction amplification (RT–PCR). At various times after infection, total RNA was harvested from infected cells. After reverse transcription using oligo-dT as primer, PCR reactions were performed using sets of primers specific to each IE gene. As an internal control for effective PCR amplification, a set of primers specific for the cellular cyclophilin gene were included in each reaction. Figure 4 shows the results for a single representative time point (4 h p.i.) where it appears that RP1, RP3, and RP4 all activated expression from each of the viral IE promoters, whereas RP5 activated expression of these genes weakly if at all. RP5 does not replicate efficiently in murine eyes The in vitro infection properties of RP5 are similar in several respects to the properties of strain in1814 (Ace et al., 1989). However, the following experiments demonstrate that the in vivo properties of RP5 differ from those of in1814 with respect to virulence and latency in infected mice. Replication of virus in infected eyes was examined after corneal inoculation of BALB/c mice with 10 5 PFU/ eye of wild-type viruses (RP1, 17 1, or the rescued virus

NEUROPATHOGENICITY OF VP16 ACTIVATION DOMAIN MUTANTS OF HSV-1

25

viruses can be distinguished in the assay, the results clearly point to an inability of the RP5 mutant to replicate in neuronal tissues. RP5 does not establish latent infection in murine TG

FIG. 4. Expression of viral IE genes in infection by recombinant HSV strains bearing deletions in the VP16 activation domain. Vero cells were infected with the indicated HSV-1 strains. At various times after infection, total cell RNA was harvested; this figure shows duplicate samples from 4 h p.i. After synthesis of cDNA by reverse transcriptase, PCR reactions were performed with primer sets specific to the ICP0, ICP4, ICP22, or ICP27 genes. PCR products were electrophoresed in 4% NuSieve agarose gels and stained with ethidium bromide. The positions of the IE-specific RT-PCR products are indicated by arrows on left. The positions of PCR products from a primer set specific to the endogenous cyclophilin gene, included in each reaction as an internal control, are indicated by arrows on right.

RP5R), or the mutant viruses (RP5 or in1814). As expected, increasing titers of infectious virus were detected in eye homogenates up to 96 h after infection with the wild-type viruses (Fig. 5A). In contrast, no infectious virus was detected in RP5-infected eyes after 24 h p.i. Similar to previous observations (Steiner et al., 1990), in1814 titers at all time points were three orders of magnitude lower than those of wild-type viruses. To ensure that viral titers obtained were not due to inoculum, early time points were also analyzed. At 12 h p.i., no infectious virus was detected in homogenates of eyes infected with RP5 or in1814, suggesting that the virus obtained after 24 h was the product of viral replication. Thus RP5 replicates poorly in mouse eyes and shows less replication than the VP16 mutant in1814.

We previously reported that in1814 establishes latent infection in murine TG as determined by the presence of reactivation-competent virus (Steiner et al., 1990) and by the presence of the latency-associated transcript (LAT) in the TG at 4 weeks after infection (Valyi-Nagy et al., 1992). In this study, multiple LAT-positive neurons were observed by in situ hybridization in BALB/c mice infected with RP1 (Fig. 6A). In contrast, no LAT expression was detected in TG of RP5-infected animals (Fig. 6B). Moreover, in explant cocultivation experiments, 100% of RP1infected ganglia reactivated by 13 days postexplantation, but no infectious virus was detected in TG explants obtained from RP5-infected animals (Fig. 7). Thus RP5 is different from in1814 in its reduced ability to establish reactivatable latent infection in murine TG after corneal inoculation. RP5 infection is severely attenuated in the CNS The inability to detect latent virus in the TG of RP5-infected mice may be due to limited viral replication in eyes (Leib et al., 1989; Steiner et al., 1990). To address this question, BALB/c mice were inoculated intracranially as described under Materials and Methods with 10 5 PFU of RP1, RP5R, or RP5 (grown on Vero cells). RP1- and RP5R-infected mice (5/5 and 3/3, respectively) died within the first 4 days after infection (Table 2). HSV-1 gene expression was detected in multiple brain regions, as determined by immunostaining using a polyclonal antibody specific to HSV (Fig.

RP5 does not replicate in TG from immunocompetent mice Viral replication in trigeminal ganglia of mice after corneal inoculation was also assessed (Fig. 5B). Although wild-type viruses showed increasing viral titers in TG homogenates, no infectious virus was detected in TG homogenates from mice infected with either RP5 or in1814. This observation is consistent with our previous report indicating lack of replication of in1814 in TG. Although no differences between the two VP16 mutant

FIG. 5. Viral titers in BALB/c eyes and TG during acute infection with RP1, RP5R, in1814, and RP5. Each point represents the geometric mean titer determined from six individual eye or trigeminal ganglia homogenates at the indicated time (hours) postinfection. Virus titrations were done on 16–8 cell monolayers in duplicate. (A) Eyes. (B) Trigeminal ganglia.

26

TAL-SINGER ET AL.

FIG. 6. Detection of HSV-1 gene expression in infected tissues by immunostaining and in situ hybridization for LAT. (A and B) BALB/c mice were inoculated with RP1 (A) or RP5 (B) via the corneal route. At 40 days postinfection, trigeminal ganglia sections were assayed by in situ hybridization (ISH) for LAT RNA expression. (C–H) BALB/c mice were infected intracranially with RP1 (C and D) or RP5 (E–H). Sections prepared from the brain of an RP1-infected animal that died 3 days after infection were analyzed by immunostaining (C) or ISH (D) for viral gene expression. At 4 days postinfection with RP5, brains were analyzed by immunostaining for HSV-1 antigen (E and F). Antigen-positive cells in the ependyma are denoted as (e). (G and H) At 40 days after intracranial infection with RP5, mice were sacrificed and brain sections were analyzed by ISH. The ventricle is denoted as (v) (G); the arrow denotes a single LAT-positive cell in the subependymal region (H). Magnification 5 3400 (A, B, and H), 3200 (D–G), and 3100 (C).

6C) and in situ hybridization for LAT (Fig. 6D). These results indicate that mortality was induced by viral encephalitis. In contrast to these wild-type viruses, the VP16 mutant virus RP5 was nonlethal, and all five mice survived the infection without demonstrating signs of illness (ruffled furs, arched backs, weight loss). In mice sacrificed at 4 days after inoculation, RP5 gene expression was detected primarily in ependymal cells in the epithelial lining of the ventricles (Figs. 6E and 6F). Antigen expression was also detected in a few

cortical neurons (Fig. 6F). Moreover, brain ventricles were enlarged (Figs. 6E and 6F), possibly due to the obstruction of flow of the cerebrospinal fluid due to sloughing of the infected ependymal cells layer. By Day 40, the ependymal layer was destroyed and the ventricles remained enlarged (Fig. 6G). At Day 40, there was no detectable viral antigen expression by immunostaining (not shown), indicating the lack of acute RP5 replication. LAT was detected by in situ hybridization in one to two neurons per brain, mostly in

NEUROPATHOGENICITY OF VP16 ACTIVATION DOMAIN MUTANTS OF HSV-1

FIG. 7. Explant-reactivation from TG of latently-infected mice. BALB/c mice were infected with 10 5 PFU/eye of RP1 or RP5. At 4 weeks after infection, 10 ganglia of RP1 (squares) and 10 ganglia of RP5 (triangles) were explanted and cocultivated with monolayers of VP16-complementing 16–8 cells. Reactivation was scored as positive when CPE was detected. No CPE was observed in RP5 explant cultures. Cultures were observed for 40 days.

27

(Figs. 1 and 2). Thus even RP5 viruses capable of a single round of replication are not lethal after intracranial inoculation. Moreover, these results indicate that RP5 is more attenuated than in1814, which had an LD 50 (minimum dose of virus that kills 50% of infected animals) of 7.4 3 10 3 PFU in intracranially inoculated mice (strain unspecified; Ace et al., 1989). To determine whether RP5 was attenuated in more than one host strain and to compare RP5 and in1814 directly, C57Bl/6 mice were injected intracranially with 10 5 PFU RP5, in1814, or RP5R. Animals infected with RP5R and in1814 died within the first week after infection. in1814-infected animals died from HSV-induced encephalitis, as determined by immunostaining for HSV antigen even at low infecting titers (results not shown). In contrast, only one of five mice infected with RP5 died after 10 days (Table 2). Therefore in1814 is more virulent than RP5. RP5 forms persistent infection in nude mice

subependymal regions (Fig. 6H). Thus RP5 shows a reduced ability to replicate in the brain and to establish latency after intracranial inoculation, similar to the results obtained with corneal inoculation. Long-term survival of RP5-treated mice All RP5-infected BALB/c mice survived longer than 40 days after infection (Table 2). Similar survival results were obtained using 10 6 PFU of RP5 grown on complementing 16–8 cells (Table 2); these virions carry equal proportions of wild-type and truncated VP16 protein and have wild-type infection properties in tissue culture cells

We have previously shown (Valyi-Nagy et al., 1992, 1994b) that in1814 and a thymidine-kinase (TK)-deficient HSV-1 mutant established both latent and slowly progressing infections in the TG after corneal inoculation of mice with severe combined immunodeficiency (SCID) (Bosma et al., 1983). In recent studies (Lasner et al., 1996, 1998), we reported that an HSV-1 variant (strain 1716) bearing a deletion in the g34.5 gene causes toxicity in nude mice; this results from persistent neuronal replication in multiple brain regions. To determine whether RP5 is virulent in immunocompromised animals, we inoculated 20 nude mice with 10 5 PFU/brain. During the first 4

TABLE 2 Survival of Mice After Inoculation with HSV-1

Mouse type

HSV-1 strain

BALB/c

RP1 RP5 (16-8) RP1 RP5 RP5R RP1 RP5 (16-8) in1814 in1814 RP5 (16-8) RP5R RP5 Mock d wt UV-irradiated e

C57Bl/b

Nude

a

Infecting titer PFU

Route of inoculation a

Survival (%) b

10 5 10 5 10 5 10 5 10 5 10 6 10 6 10 6 10 5 10 5 10 5 10 5

Oc Oc IC IC IC IC IC IC IC IC IC IC IC IC

14/14 (100) 15/15 (100) 0/5 (0) 5/5 (100) 0/3 (0) 0/5 (0) 5/5 (100) 0/2 (0) 0/5 (0) 4/5 (80) 0/5 (0) 16/20 (80) 5/5 (100) 5/5 (100)

Mice were inoculated after corneal scarification (Oc) or by intracranial injection (IC). Percentages in parentheses. c Surviving mice were observed for $200 days. d Mock-infected mice were inoculated with 10 microliters of serum-free DMEM. e UV-irradiated wild-type virus equivalent to 10 5 PFU. b

Time of death days post-infection c

3–4 4 3–4 4 4, 5, 6 10 4 26, 31, 40, 46

28

TAL-SINGER ET AL.

weeks, none of the mice showed signs of disease. However, four mice died on Days 26, 28, 40, and 46 after infection (Table 2). RP5 replication was detected by immunostaining in multiple brain regions (not shown). Three surviving mice were sacrificed, and two brains stained positive for HSV antigen. Therefore although RP5 is less virulent than other HSV-1 mutants, it is capable of establishing persistent and sometimes lethal infection in a subpopulation of immunocompromised hosts. DISCUSSION HSV-1 VP16 is a virion tegument protein that stimulates transcription of IE genes during lytic infection (Batterson and Roizman, 1983; Campbell et al., 1984; Preston et al., 1984). The transcriptional activation domain of VP16, which resides in the C-terminal 80 amino acids of the protein (Triezenberg et al., 1988a; Cousens et al., 1989), is an important model for mechanisms of transcriptional activation. Most investigations of the function of the VP16 activation domain have been in heterologous biological or in vitro systems. In this study, we investigated the role of the VP16 activation domain in the infectious processes of HSV-1 by constructing mutant viruses bearing deletions of some or all of the VP16 activation domain. Our results demonstrate the importance of the VP16 activation domain for lytic infection in vitro and in vivo. Replication of a mutant virus, RP5, lacking the activation domain of VP16 was severely attenuated in murine PNS and CNS. RP5 failed to replicate, efficiently establish latency, or reactivate in murine TG after corneal inoculation. Furthermore, RP5 replication was severely attenuated after intracranial inoculation. In contrast to infection with wild-type virus, all immunocompetent mice inoculated with RP5 survived infection without evidence of disease. Our observations of the in vitro infection properties of RP5 are consistent with those reported for V422, a strain in which the VP16 gene is truncated at codon 422 (Lam et al., 1996; Smiley and Duncan, 1997). Both viruses are approximately 100-fold less infectious than their wildtype parental strain or corresponding rescued strains. Both viruses fail to activate viral IE gene expression during lytic infection in cultured cells. The in vivo infection properties of V422, however, have not yet been described. In several respects, the in vitro and in vivo properties of RP5 also resemble those of HSV-1 mutant in1814 (Ace et al., 1989), which contains an insertion in its VP16 gene that affects the transinducing activity of the gene product. However, RP5 differs from in1814 in one significant aspect; whereas in1814 was capable of establishing latency after corneal inoculation, from which virus could be reactivated by explant cocultivation (Steiner et al., 1990; Valyi-Nagy et al., 1992), the RP5 virus showed no clear evidence of efficiently establishing or reactivating from

latency. Thus these two viruses, both defective in IE viral trans-activation, show significant differences in latency and neuropathogenesis. Some contribution to these phenotypic differences may arise from other genetic differences in the parental strains for RP5 (derived from KOS) and in1814 (derived from strain 17 1) because 17 1 is more neurovirulent than KOS. Nonetheless, the phenotypic differences between RP5 and its KOS parent are more pronounced than we have previously observed for in1814 and its parental strain (Steiner et al., 1990; Valyi-Nagy et al., 1992). The differences in neuropathogenesis between in1814 and RP5 may result from either qualitative or subtle quantitative differences in their abilities to activate viral IE transcription. The mutation in in1814 is a 4-amino-acid linker insertion at codon 397 (Ace et al., 1989) that disrupts the interaction of VP16 with the cellular proteins Oct-1 and HCF but preserves the C-terminal acidic activation domain. In contrast, the truncated VP16 protein expressed from RP5 can effectively form a DNA-binding complex with Oct-1 and HCF but lacks the entire transcriptional activation domain. Thus the mutant VP16 protein encoded by in1814 may have a residual transcriptional function which could account for the ability of in1814 to establish latent infections. This residual function may also account for the ability of in1814 to activate IE transcription at wild-type levels in high multiplicities of infection of cultured cells (Ace et al., 1989). Alternatively, if both in1814 and RP5 are fully defective in IE gene activation, then the differences in the neuropathogenesis may be due to differences in some other activity of the respective VP16 proteins. One possibility for this ‘‘other activity’’ is the interaction of VP16 with the vhs protein. Several reports have indicated that VP16 facilitates the persistence of HSV mRNAs in infected cells at late times by interacting with the viral vhs protein and dampening its activity (Smibert et al., 1994; Lam et al., 1996). The domain of VP16 that interacts with vhs lies amino-terminal to residue 369 (Smibert et al., 1994), whereas the transcriptional activation domain resides within 80 amino acids at the C terminus of VP16 (Sadowski et al., 1988; Triezenberg et al., 1988a; Cousens et al., 1989). Therefore the mutation in RP5 should not affect its interaction with vhs. This conclusion is supported by preliminary experiments in which the virion-borne vhs activities of RP1 and RP5 (i.e., at early times during infection) were found to be comparable when analyzed by Northern blots of infected-cell RNA. In contrast, at late stages of infection (10 h p.i.), the vhs activity in RP5infected cells was markedly reduced (Y. Mao and S. Triezenberg, unpublished results). This difference may result from the delay in late gene expression (including vhs expression) in RP5-infected cells that results from the initial inability to transactivate IE gene expression. Deletion of the vhs gene results in virions that are capable of efficient replication in Vero cells but are unable to

NEUROPATHOGENICITY OF VP16 ACTIVATION DOMAIN MUTANTS OF HSV-1

establish acute or reactivation-competent latent infections in immunocompetent CD-1 mice after corneal inoculations (Strelow and Leib, 1995). Therefore the reduced vhs activity observed at late times in RP5 infection may similarly contribute to the severely attenuated phenotype of RP5 in vivo. It would be of interest to determine whether vhs activity is preserved in in1814-infected cells. Intracranial injection of in1814 into immunocompetent C57Bl/6 mice was lethal at infectious titers as low as 10 3 PFU/brain (Table 2). These results were similar to a previous observation that the LD 50 of in1814 in BALB/c mice was approximately 7.5 3 10 3 PFU/brain (Ace et al., 1989). In contrast, RP5 was nonneurovirulent upon intracranial inoculation. BALB/c (100%) and C57Bl/6 (80%) mice survived infections with as much as 10 6 PFU/brain (Table 2). Although there was some evidence for infection of nonneuronal ependymal cells during the acute stage of infection (Fig. 6F), no symptomatic disease was observed. This phenotype resulted from the deletion in the VP16 gene because RP1 and the rescued virus RP5R caused lethal encephalitis at those infecting titers (Table 2). Moreover, we found no evidence of latent infection in BALB/c brains after intracranial inoculation. Therefore, as observed in the PNS, RP5 is profoundly defective in its ability to establish efficient acute or latent infections in the CNS of immunocompetent animals. These results suggest that the transactivation domain in VP16 is important for neurovirulence. In this study we found that RP5 replication was mostly limited to ependymal cells as previously observed using other neuroattenuated HSV-1 mutants such as strains deficient in ICP34.5 (g34.5) (Thompson et al., 1993; Lasner et al., 1996). This indicates that ependymal cells are highly permissive to HSV infection; however, they are unable to transmit virus efficiently to the surrounding cells. It is possible that ependymal cells possess cellular factors that are strong viral activators capable of compensating for different viral defects. We have previously found (Valyi-Nagy et al., 1992, 1994b) that in1814 and a thymidine-kinase (TK)-deficient HSV-1 mutant established both latent and slowly progressing infections in the TG after corneal inoculation of mice with severe combined immunodeficiency (SCID) (Bosma et al., 1983). These infections resulted in occasional deaths (Valyi-Nagy et al., 1992, 1994b). In contrast, an HSV-1 mutant lacking the g34.5 gene (1716) was unable to establish persistent infection after corneal or intracranial inoculation into SCID mice (Valyi-Nagy et al., 1994a). This mutant formed persistent lethal infections in a subpopulation of nude mice after intracranial inoculation (Lasner et al., 1996, 1997). In this study, we found evidence for a similar low-level lethal replication in the brains of nude mice after intracranial inoculation with RP5 (Table 2). These results suggest that in the absence of mature T cells and an efficient immune response, RP5

29

is capable of low-level acute replication in CNS neurons that does not require the activity of VP16. HSV-1 mutants that preferentially replicate in dividing cells but not in normal postmitotic and/or quiescent brain cells have been investigated as potential agents for brain tumor therapy (Chambers et al., 1995; Martuza et al., 1991; Markert et al., 1993). We have described the use of the g34.5 deletion mutant strain 1716 (MacLean et al., 1991) in mouse models of primary human brain tumors (Kesari et al., 1995b; Lasner et al., 1996) and intracranial melanoma (Randazzo et al., 1995). Because RP5 has a similar in vivo phenotype to 1716, we wished to determine whether RP5 may be useful for brain tumor therapy. In preliminary experiments, we have determined that RP5 was capable of replication in human neuroblastoma D283 (Friedman et al., 1985) tumor cells and in D283 tumors in nude mice (T. M. Lasner, unpublished results). Therefore, RP5 is replication competent in rapidly dividing cells in vivo and, like 1716, may serve as a therapeutic agent. Furthermore, 1716 is capable of efficiently establishing latent infection in both immunocompetent and immunodeficient mice, whereas RP5 does not establish latency. Therefore RP5 may be a better candidate for evaluation as a tumor therapeutic agent. Studies to determine whether RP5 kills tumor cells more efficiently than 1716 are in progress. MATERIALS AND METHODS Animals Female BALB/c BYJ and C57Bl/6 mice (4–6 weeks old) were obtained from Jackson Laboratories. Female homozygous nude mice (4–6 weeks old) were obtained from Taconic (Germantown, NY). The animal studies were approved by the University of Pennsylvania Institutional Animal Care and Use Committee. Cell cultures Vero, 16–8, BHK, and L cells were cultured in DMEM supplemented with either 5 or 10% fetal calf serum (FCS). Virus stocks HSV-1 strains wild-type 17 1 (Brown et al., 1973) and insertion mutant in1814 (Ace et al., 1989) were grown on baby hamster kidney 21 clone 13 (BHK) cells. All other virus strains used in this study are shown in Fig. 1. HSV-1 strain 8MA, a gift from S. Weinheimer, was derived from strain KOS by replacing the VP16 gene with the Escherichia coli lacZ gene regulated by the HSV-1 ICP6 promoter (Weinheimer et al., 1992). This strain was propagated in a Vero-derived cell line (designated 16–8) that expresses the VP16 gene product (Weinheimer et al., 1992). Strain RP1 was constructed by homologous recombination in Vero cells using viral genomic DNA from strain 8MA and a XhoI–PstI restriction fragment encom-

30

TAL-SINGER ET AL.

passing the VP16 gene. This fragment bore a BamHI site located 19 bp downstream of the VP16 gene followed by an oligonucleotide with stop codons in all three reading frames (59-TAGTTAATTGA). Strains RP3, RP4, and RP5 were constructed similarly using DNA fragments lacking codons 456–490, 413–452, or 413–490 (respectively) of the VP16 gene. The mutations in RP3, RP4, and RP5 were rescued by homologous recombination using viral genomic DNA from the mutant strains and a wild-type XhoI–PstI DNA fragment from strain KOS, yielding viruses designated RP3R, RP4R, and RP5R, respectively. Recombinant viruses were plaque-purified twice, and their identities were confirmed by Southern blot and immunoblot analyses. Details of these constructions are provided in (Pichyangkura, 1996) and are available upon request. Viral stocks were grown and titered on monolayers of Vero cells or in some instances on VP16complementing 16–8 cells (Weinheimer et al., 1992).

various times after infection using Tri-Reagent (Molecular Research Center, Inc., Cincinnati OH) followed by digestion with RNase-free DNaseI (Boehringer). Firststrand cDNA synthesis was primed with oligo-dT and was extended using Superscript II reverse transcriptase (Gibco BRL). Subsequent PCR reactions included primer pairs specific for the HSV-1 ICP0, ICP4, ICP22, or ICP27 genes (details available upon request). Each reaction also included a primer pair specific for the endogenous cyclophilin gene (Tal-Singer et al., 1997). PCR reactions were performed using a ‘‘touchdown’’ strategy (Don et al., 1991) in which the hybridization temperature was decreased by 1°C after every two cycles, beginning at 60°C. A total of 30 cycles of polymerization (15 temperature steps) was followed by seven additional minutes of extension at 72°C. Reaction products were electrophoresed in 4% NuSieve agarose gels (FMC), stained with ethidium bromide and photographed under UV illumination.

Transient tranfection assays To assess transcriptional activation of IE promoters by the various VP16 mutant viruses, reporter plasmids were transfected into mouse L cells by DEAE-dextran (Lopata et al., 1984) or into Vero cells by calcium phosphate coprecipitation. Plasmids expressing the bacterial chloramphenicol acetyltransferase gene under the control of HSV-1 IE promoters were the generous gift of Dr. P. A. Schaffer and Anh Nguyen-Huynh [pAN5, ICP0; pAN6, ICP4; pAN7, ICP22; and pICP27-CAT (Nguyen-Huynh and Schaffer, 1998)]. Forty hours after transfection, cells were infected with HSV-1 in the presence of cycloheximide (50 mg/ml) to prevent expression of the viral IE gene products. After 3 h, medium containing cycloheximide was replaced with medium containing actinomycin D (10 mg/ ml) to permit translation of the CAT protein from accumulated mRNA. After two additional hours, whole cell protein extracts were prepared and assayed for CAT enzymatic activity using the fluor diffusion method (Neumann et al., 1987; Nielsen et al., 1989). Plasmids expressing the HSV thymidine kinase gene under the control of the ICP4 and Moloney sarcoma virus promoters have been described (Graves et al., 1985; Triezenberg et al., 1988b). A corresponding ICP0-tk plasmid was constructed by cloning a PCR-derived promoter fragment into the same plasmid backbone used for the ICP4-tk reporter gene (P. J. Horn and E. Chung, unpublished). Transfection, subsequent infection, RNA harvest, and assay by primer extension were performed as previously described (Triezenberg et al., 1988b). Viral IE gene expression assays in infected cells To assess the expression of viral IE genes during infection, Vero cells were infected with wild-type or mutant viruses using a number of virions equivalent to 10 PFU of RP1 per cell. Total cell RNA was harvested at

Ocular infection of mice Mice were anesthetized with intraperitoneal injection of ketamine (87 mg/kg)/xylazine (13 mg/kg) then inoculated after corneal scarification with 10 5 PFU per eye of RP1, RP5R, RP5 (grown on 16–8 cells), 17 1, or in1814. At 30 days p.i., some mice were sacrificed by cervical dislocation, and eyes and trigeminal ganglia (TG) were excised. Titration of virus from eyes and TG Other mice were sacrificed 1, 6, 12, 24, 72, or 96 h after corneal inoculation. Tissues from three mice per time point were homogenized in 1 ml serum-free Dulbecco’s modified Eagle medium (DMEM) using a Pyrex Ten Broek tissue grinder (Bellco Glass, Vineland, NJ). The homogenates were frozen and thawed once and cleared by centrifugation at 3000 g for 5 min at 4°C. The supernatant was diluted logarithmically in media and the viral titer was determined on 16–8 cells in 48-well plates by immunoperoxidase assay (Holland et al., 1983; Hung et al., 1992; Tal-Singer et al., 1995) using rabbit anti-HSV-1 polyclonal antibody (American Qualex Antibodies, San Clemente, CA). Titers were determined in duplicate wells. Explant reactivation Groups of six explanted TG from latently infected mice were cocultivated (Stevens and Cook, 1971) with monolayers of 16–8 cells in DMEM supplemented with 5% fetal bovine serum at 37°C. Every 6 days, ganglia were transferred to fresh monolayers of 16–8 cells. Reactivation was considered positive when cytopathic effects were detected in the cell monolayer within 40 days of cocultivation.

NEUROPATHOGENICITY OF VP16 ACTIVATION DOMAIN MUTANTS OF HSV-1

Intracerebral inoculation Four- to 6-week-old female BALB/c, nude, or C57Bl/6 mice were anesthetized, and the heads were cleansed with 70% ethanol. Following a midline incision, the skull was perforated using a 25-gauge needle 2 mm caudal of the bregma and 1 mm left of the midline (Kesari et al., 1995a). The animals were placed in a small-animal stereotactic apparatus (Kopf Instruments, Tujunga, CA), and an appropriate amount of virus (10 5–10 6 PFU of virus in 10 ml) or serum-free DMEM (vehicle) was injected using a 10 ml Hamilton syringe with a 27-gauge needle (Kesari et al., 1996). The injection was performed for 3 min. The needle was left in for 3 min after the injection, then withdrawn slowly over 1 min, as described previously (Kesari et al., 1995a). The superficial skin wound was closed with sutures, and the mice were inspected daily for signs of illness. Immunostaining procedures For the studies assessing replication of virus within the brains, mice were transcardially perfused and fixed with 4% paraformaldehyde (0.1 M phosphate-buffered saline, pH 7.34), and the brains were removed for histological and immunohistochemical analysis. Ganglia used for immunohistochemistry were immersed in 70% ethanol/150 mM NaCl for 24 h and then embedded in paraffin wax. The methods of tissue processing and light microscopic immunohistochemical analysis were similar to those described elsewhere (Randazzo et al., 1995; Trojanowski et al., 1993, 1994). Rabbit polyclonal antisera to HSV-1 (Dako Corp., Carpinteria, CA) was used for detection of viral antigens as described (Adams et al., 1984; Kesari et al., 1995b). Antigen-expressing cells were detected by the indirect avidin-biotin immunoperoxidase method (Vectastain ABC kit, Vector Labs, Burlingame, CA), with 3,39-diaminobenzadine (DAB) as the chromagen (Trojanowski et al., 1993). Spread of virus in all animals was monitored by screening 4 of 30 sections (6-mm thick) through the brain (5–10 sections apart). In situ hybridization for HSV-1 specific gene expression In situ hybridization was performed as previously described (Deatly et al., 1988). Sections of trigeminal ganglia and brain stems from infected mice sacrificed 6 weeks p.i. were mounted on slides. The sections were hybridized with a latency associated transcript (LAT) specificnick-translated[ 35S]-ATP(Amersham)-labeled(BstEII– BstEII) 0.9-kb DNA probe derived from the BamHI B fragment of the HSV genome (Valyi-Nagy et al., 1992). After hybridization, five slides (4–9 sections per slide) for each animal were examined. The positive cells in each section were counted, and the mean number of positive cells per section was determined.

31

ACKNOWLEDGMENTS We thank Yaopan Mao, Søren Jaglo-Ottosen, and Jennifer Unsell for excellent technical assistance, Peter Horn for constructing the ICP0-tk reporter plasmid, Steve Weinheimer for providing HSV strain 8MA and the 16–8 cell line, and Anh Nguyen-Huynh and Priscilla Schaffer for several CAT reporter plasmids. This research was supported by National Institutes of Health Grants NS-33768 (N.W.F.), AI-27323, and AI-01284 (S.J.T.). R.T.-S. was supported by training grant CA-09171 from the NIH. T.M.L. was supported by the Division of Neurosurgery, Hospital of the University of Pennsylvania and a clinical fellowship from the Measey Foundation.

REFERENCES Ace, C. I., McKee, T. A., Ryan, J. M., Cameron, J. M., and Preston, C. M. (1989). Construction and characterization of a herpes simplex virus type 1 mutant unable to transinduce immediate early gene expression. J. Virol 63, 2260–2269. Adams, R. L., Springall, D. R., Levene, M. M., and Bushell, T. E. (1984). The immunocytochemical detection of herpes simplex virus in cervical smears—a valuable technique for routine use. J. Pathol. 143, 241–247. Batterson, W., and Roizman, B. (1983). Characterization of the herpes simplex virion-associated factor responsible for the induction of a genes. J. Virol. 46, 371–377. Bosma, G. C., Custer, R. P., and Bosma, M. J. (1983). A severe combined immunodeficiency mutation in the mouse. Nature 301, 527–530. Brown, S. M., Ritchie, D. A., and Subak-Sharpe, J. (1973). Genetic studies with herpes simplex virus type 1. The isolation of temperature-sensitive mutants, their arrangement into complementation groups and recombination analysis leading to lineage groups. J. Gen. Virol. 18, 329–346. Campbell, M. E. M., Palfreyman, J. W., and Preston, C. M. (1984). Identification of herpes simplex virus DNA sequences which encode a trans-acting polypeptide responsible for stimulation of immediate early transcription. J. Mol. Biol. 180, 1–19. Chambers, R., Gillespie, G. Y., Soroceanu, L., Andreansky, S., Chatterjee, S., Chou, J., Roizman, B., and Whitley, R. J. (1995). Comparison of genetically engineered herpes simplex viruses for the treatment of brain tumors in a scid mouse model of human malignant glioma. Proc. Natl. Acad. Sci. USA 92, 1411–1415. Cousens, D., Greaves, R., Goding, C., and O’Hare, P. (1989). The Cterminal 79 amino acids of the HSV regulatory protein, Vmw65, efficiently activates transcription in yeast and mammalian cells in chimeric DNA-binding proteins. EMBO J. 8, 2337–2342. Deatly, A. M., Spivack, J. G., Lavi, E., O’Boyle, D. R., II, and Fraser, N. W. (1988). Latent herpes simplex virus type 1 transcripts in peripheral and central nervous system tissues of mice map to similar regions of the viral genome. J. Virol. 62, 749–756. Don, R. H., Cox, P. T., Wainwright, B. J., Baker, K., and Mattick, J. S. (1991). ‘‘Touchdown’’ PCR to circumvent spurious priming during gene amplification. Nucleic Acids Res. 19, 4008. Fraser, N. W., Spivack, J. G., Wroblewska, Z., Block, T., Deshmane, S. L., Valyi-Nagy, T., Natarajan, R., and Gesser, R. (1991). A review of the molecular mechanism of HSV-1 latency. Curr. Eye Res. 10, (suppl.), 1–14. Fraser, N. W., and Valyi-Nagy, T. (1993). Viral, neuronal and immune factors which may influence herpes simplex virus (HSV) latency and reactivation. Microb. Pathogen. 15, 83–91. Friedman, H. S., Burger, P. C., Bigner, S. H., Trojanowski, J. Q., Wikstrand, C. J., Halperin, E. C., and Bigner, D. D. (1985). Establishment and characterization of the human medulloblastoma cell line and transplantable xenograft D283 Med. J. Neuropathol. Exp. Neurol. 44, 592–605. Gaffney, D. F., McLauchlan, J., Whitton, J. L., and Clements, J. B. (1985). A modular system for the assay of transcription regulatory signals:

32

TAL-SINGER ET AL.

The sequence TAATGARAT is required for herpes simplex virus immediate early gene activation. Nucleic Acids Res. 13, 7847–7863. Graves, B. J., Eisenman, R. N., and McKnight, S. L. (1985). Delineation of transcriptional control signals within the Moloney murine sarcoma virus long terminal repeat. Mol. Cell. Biol. 5, 1948–1958 Hill, T. J. (1985). Herpes simplex virus latency. In ‘‘The Herpes Viruses’’ (B. Roizman, Ed.), Vol. 4, pp. 175–240. Plenum Publishing Corp., New York. Holland, T. C., Marlin, S. D., Levine, M., and Glorioso, J. (1983). Antigenic variants of herpes simplex virus selected with glycoprotein-specific monoclonal antibodies. J. Virol. 45, 672–682. Hung, S.-L., Srinivasan, S., Friedman, H. M., Eisenberg, R. J., and Cohen, G. H. (1992). The structural basis of C3b binding by glycoprotein C of herpes simplex virus. J. Virol. 66, 4013–4027. Katan, M., Haign, C., Verrijzer, P., van der Vliet, P., and O’Hare, P. (1990). Characterization of a cellular factor which interacts functionally with Oct-1 in the assembly of a multicomponent transcription complex. Nucleic Acids Res. 18, 6871–6880. Kesari, S., Lee, V. M.-Y., Brown, S. M., Trojanowski, J. Q., and Fraser, N. W. (1996). Selective vulnerability of mouse CNS neurons to latent infection with a neuroattenuated herpes simplex virus-1. J. Neurosci. 16, 5644–5653. Kesari, S., Randazzo, B. P., Brown, S. M., MacLean, A. R., Lee, V. M.-Y., Trojanowski, J. Q., and Fraser, N. W. (1995a). Replication of herpes simplex virus-1 gamma 34.5 deletion mutant in murine brain tumors. 25th Annual Society of Neuroscience Conference 838.3 (San Diego, California). Kesari, S., Randazzo, B. P., Valyi-Nagy, T., Huang, Q. S., Brown, S. M., MacLean, A. R., Lee, V. M.-Y., Trojanowski, J. Q., and Fraser, N. W. (1995b). Therapy of experimental human brain tumors using a neuroattenuated herpes simplex virus mutant. Lab. Invest. 73, 636–648. Kristie, T. M., Lebowitz, J. H., and Sharp, P. A. (1989). The octamerbinding proteins form multi-protein DNA complexes with the HSV a TIF regulatory protein. EMBO J. 8, 4229–4238. Kristie, T. M., and Sharp, P. A. (1993). Purification of the C1 factor required for the stable recognition of the Oct-1 homeodomain by the herpes simplex virus a-transinducing factor (VP16). J. Biol. Chem. 268, 6525–6534. Lam, Q., Smibert, C. A., Koop, K. E., Lavery, C., Capone, J. P., Weinheimer, S. P., and Smiley, J. R. (1996). Herpes simplex virus VP16 rescues viral mRNA from destruction by the virion host shutoff function. EMBO J. 15, 2575–2581. LaMarco, K. L., and McKnight, S. L. (1989). Purification of a set of cellular polypeptides that bind to the purine-rich cis-regulatory element of herpes simplex virus immediate early genes. Genes Dev. 3, 1372–1383. LaMarco, K., Thompson, C. C., Byers, B. P., Walton, E. M., and McKnight, S. L. (1991). Identification of Ets- and notch-related subunits in GA binding protein. Science 253, 789–792. Lasner, T. M., Kesari, S., Brown, S. M., Lee, V. M.-Y., Fraser, N. W., and Trojanowski, J. Q. (1996). Therapy of a murine model of pediatric brain tumors using a herpes simplex virus type-1 ICP34.5 mutant and demonstration of viral replication within the CNS. J. Neuropath. Exp. Neurol. 55, 1259–1269. Lasner, T. M., Tal-Singer, R., Kesari, S., Lee, V. M.-Y., Trojanowski, J. Q., and Fraser, N. W. (1998). Toxicity and neuronal infection of a HSV-1 ICP34.5 mutant in nude mice. J. Neurovirol. 4, 100–105. Leib, D. A., Coen, D. M., Bogard, C. L., Hicks, K. A., Yager, D. R., Knipe, D. M., Tyler, K. L., and Schaffer, P. A. (1989). Immediate-early gene mutants define different stages in the establishment and reactivation of herpes simplex virus latency. J. Virol. 63, 759–768. Lopata, M. A., Cleveland, D. W., and Sollner-Webb, B. (1984). High level expression of a chloramphenicol acetyl transferase gene by DEAEdextran mediated DNA transfection coupled with a dimethyl sulfoxide or glycerol shock treatment. Nucleic Acids Res. 12, 5707–5717. Mackem, S., and Roizman, B. (1982). Structural features of the herpes simplex virus a gene 4, 0 and 27 promoter-regulatory sequences

which confer a regulation on chimeric thymidine kinase genes. J. Virol. 44, 939–949. MacLean, A. R., Ul-Fareed, M., Robertson, L., Harland, J., and Brown, S. M. (1991). Herpes simplex virus type 1 deletion variants 1714 and 1716 pinpoint neurovirulence-related sequences in Glasgow strain 171 between immediate early gene 1 and the ‘‘a’’ sequence. J. Gen. Virol. 72, 631–639. Markert, J. M., Malick, A., Coen, D. M., and Martuza, R. L. (1993). Reduction and elimination of encephalitis in an experimental glioma therapy model with attenuated herpes simplex mutants that retain susceptibility to acyclovir. Neurosurgery 32, 597–603. Marsden, H. S., Campbell, M. E. M., Haarr, L., Frame, M. C., Parris, D. S., Murphy, M., Hope, R. G., Muller, M. T., Preston, C. M. (1987). The 65000-Mr DNA-binding and virion trans-inducing proteins of herpes simplex virus type 1. J. Virol. 61, 2428–2437. Martuza, R. L., Malick, A., Markert, J. M., Ruffner, K. I., and Coen, D. M. (1991). Experimental therapy of human glioma by means of a genetically engineered virus mutant. Science 252, 854–856. Moss, H. (1989). Properties of the herpes simplex virus type 2 transinducing factor Vmw65 in wild-type and mutant viruses. J. Gen. Virol. 70, 1579–1585. Nielsen, D. A., Chang, T., and Shapiro, D. J. (1989). A highly sensitive mixed phase assay for chloramphenicol acetyltransferase activity in transfected cells. Anal. Biochem. 179, 12–23. Neumann, J. R., Morency, C. A., and Russian, K. O. (1987). A novel rapid assay for chloramphenicol acetyltransferase gene expression. Biotechniques, 5, 444–447. Nguyen-Huynh, A. T., and Schaffer, P. A. (1998). Cellular transcription factors enhance herpes simplex virus type 1 oriS-dependent DNA replication J. Virol. 72, 3635–3645. O’Hare, P., and Goding, C. R. (1988). Herpes simplex virus regulatory elements and the immunoglobulin octamer bind a common factor and are both targets for virion transactivation. Cell 52, 435–445. Pichyangkura, R. (1996). ‘‘The effect of deletion mutations of the activation domain of VP16 upon herpes simplex virus type 1 lytic infection.’’ Ph.D. Thesis, Michigan State University, East Lansing, MI. Poon, A. P., and Roizman, B. (1995). The phenotype in vitro in infected cells of herpes simplex virus 1 a trans-inducing factor (VP16) carrying temperature-sensitive mutations introduced by substitution of cysteines. J. Virol. 69, 7658–7667. Preston, C. M., Cordingley, M. G., and Stow, N. D. (1984). Analysis of DNA sequences which regulate the transcription of a herpes simplex virus immediate early gene. J. Virol. 50, 708–716. Preston, C. M., Frame, M. C., and Campbell, M. E. M. (1988). A complex formed between cell components and an HSV structural polypeptide binds to a viral immediate early gene regulatory DNA sequence. Cell 52, 425–434. Pruijn, G. J. M., van Miltenburg, R. T., Claessens, J. A. J., and van der Vliet, P. C. (1988). Interaction between the octamer-binding protein nuclear factor III and the adenovirus origin of DNA replication. J. Virol. 62, 3092–3102. Randazzo, B. P., Kesari, S., Gesser, R. M., Alsop, D., Ford, J. C., Brown, S. M., MacLean, A. R., and N. W. Fraser. (1995). Treatment of experimental intracranial murine melanoma with a neuro-attenuated herpes simplex virus-1 mutant. Virology 211, 94–101. Roizman, B., and Sears, A. E. (1996). Herpes simplex viruses and their replication. In ‘‘Fundamental Virology’’ (B. N. Fields, and D. M. Knipe, Eds.), pp. 2231–2296. Raven Press, Ltd., New York. Sadowski, I., Ma, J., Triezenberg, S., and Ptashne, M. (1988). GAL-4VP16 is an unusually potent transcriptional activator. Nature 335, 563–564. Schmelter, J., Knez, J., Smiley, J. R., and Capone, J. P. (1996). Identification and characterization of a small modular domain in the herpes simplex virus host shutoff protein sufficient for interaction with VP16. J. Virol. 70, 2124–2131. Smibert, C. A., Popova, B., Xiao, P., Capone, J. P., and Smiley, J. R. (1994).

NEUROPATHOGENICITY OF VP16 ACTIVATION DOMAIN MUTANTS OF HSV-1 Herpes simplex virus VP16 forms a complex with the virion host shutoff protein vhs. J. Virol. 68, 2239–2246. Smiley, J. R., and Duncan, J. (1997). Truncation on the C-terminal acidic transcriptional activation domain of herpes simplex virus VP16 produces a phenotype similar to that of the in1814 linker insertion mutation. J. Virol. 71, 6191–6193. Steiner, I., Spivack, J. G., Deshmane, S. L., Ace, C. I., Preston, C. M., and Fraser, N. W. (1990). A herpes simplex virus type 1 mutant containing a nontransinducing Vmw65 protein establishes latent infection in vivo in the absence of viral replication and reactivates efficiently from explanted trigeminal ganglia. J. Virol. 64, 1630–1638. Stern, S., and Herr, W. (1991). The HSV transactivator VP16 recognizes the Oct-1 homeo domain: Evidence for a homeodomain recognition subdomain. Genes Dev. 5, 2555–2566. Stern, S., Tanaka, M., and Herr, W. (1989). The oct-1 homeodomain directs formation of a multiprotein-DNA complex with the HSV transactivator VP16. Nature 341, 624–630. Stevens, J. G., and Cook, M. L. (1971). Latent herpes simplex virus in spinal ganglia of mice. Science 173, 843–845. Strelow, L. I., and Leib, D. A. (1995). Role of the virion host shutoff (vhs) of herpes simplex virus type 1 in latency and pathogenesis. J. Virol. 69, 6779–6786. Sullivan, S. M., Horn, P. J., Olson, V. A., Koop, A. H., Niu, W., Ebright, R. H., and Triezenberg, S. J. (1998). Mutational analysis of a transcriptional activation region of the VP16 protein of herpes simplex virus. Nucleic Acids Res. 26, 4487–4496. Tal-Singer, R., Peng, C., Leon, M. P. D., Abrams, W. R., Banfield, B. W., Tufaro, F., Cohen, G. H., and Eisenberg, R. J. (1995). The interaction of herpes simplex virus glycoprotein C with mammalian cell surface molecules. J. Virol. 69, 4471–4483. Tal-Singer, R., Lasner, T. M., Podrzucki, W., Skokotas, A., Leary, J. J., Berger, S. L., and Fraser, N. W. (1997). Gene expression during reactivation of herpes simplex virus type 1 from latency in the peripheral nervous system is different from that during lytic infection of tissue cultures. J. Virol. 71, 5268–5276. Thompson, C. C., Brown, T. A., and McKnight, S. L. (1991). Convergence of Ets- and notch-related structural motifs in a heteromeric DNA binding complex. Science 253, 762–768. Thompson, R. L., Wagner, E. K., and Stevens, J. G. (1983). Physical location of a herpes simplex virus type-1 gene function(s) specifically associated with a 10 million-fold increase in HSV neurovirulence. Virology 131, 180–192. Triezenberg, S. J. (1995). Structure and function of transcriptional activation domains. Curr Opin. Genet. Dev. 5, 190–196 Triezenberg, S., Kingsbury, R., and McKnight, S. (1988a). Functional dissection of VP16, the transactivator of HSV immediate early gene expression. Genes Dev. 2, 718–729. Triezenberg, S., LaMarco, K. L., and McKnight, S. (1988b). Evidence of DNA:protein interactions that mediate HSV-1 immediate early gene activation by VP16. Genes Dev. 2, 730–742. Trojanowski, J. Q., Fung, K.-M., Rorke, L. B., Tohyama, T., Yachnis, A. T., and Lee, V.-Y. (1994). In vivo and in vitro models of medulloblastomas and other primitive neuroectodermal brain tumors of childhood. Mol. Chem. Neuropath. 21, 219–239. Trojanowski, J., Mantione, R., Lee, J., Seid, D., You, T., Inge, L., and Lee,

33

V. (1993). Neurons derived from a human teratocarcinoma cell line establishes molecular and structural polarity following transplantation into the rodent brain. Expt. Neurol. 122, 283–294. Valyi-Nagy, T., Deshmane, S. L., Dillner, A. J., and Fraser, N. W. (1991a). Induction of cellular transcription factors in trigeminal ganglia of mice by corneal scarification, herpes simplex virus 1 type infection and explantation of trigeminal ganglia. J. Virol. 65, 4142–4152. Valyi-Nagy, T., Deshmane, S. L., Raengsakulrach, B., Nicosia, M., Gesser, R. M., Wysocka, M., Dillner, A., and Fraser, N. W. (1992). Herpes simplex virus type 1 mutant strain in1814 establishes a unique, slowly progressing infection in SCID mice. J. Virol. 66, 7336– 7345. Valyi-Nagy, T., Deshmane, S. L., Spivack, J. G., Steiner, I., Ace, C. I., Preston, C. M., and Fraser, N. W. (1991b). Investigation of herpes simplex virus type 1 (HSV-1) gene expression and DNA synthesis during the establishment of latent infection by an HSV-1 variant, in1814, that does not replicate in mouse trigeminal ganglia. J. Gen. Virol. 72, 641–649. Valyi-Nagy, T., Fareed, M. U., O’Keefe, J. S., Gesser, R. M., MacLean, A. R., Brown, S. M., Spivack, J. G., and Fraser, N. W. (1994a). The HSV-1 strain 171 gamma 34.5 deletion mutant 1716 is avirulent in SCID mice. J. Gen. Virol. 75, 2059–2063. Valyi-Nagy, T., Gesser, R. M., Raengsakulrach, B., Deshmane, S. L., Randazzo, B. P., Dillner, A. J., and Fraser, N. W. (1994b). A thymidine kinase-negative HSV-1 strain establishes a persistent infection in SCID mice that features uncontrolled peripheral replication but only marginal nervous system involvement. Virology 199, 484–490. Virbasius, J. V., Virbasius, C. A., and Scarpulla, R. C. (1993). Identity of GABP with NRF-2, a multisubunit activator of cytochrome oxidase expression, reveals a cellular role for an ETS domain activator of viral promoters. Genes Dev. 7, 380–392. Walker, S., Greaves, R., and O’Hare, P. (1993). Transcriptional activation by the acidic domain of Vmw65 requires the integrity of the domain and involves additional determinants distinct from those necessary for TFIIB binding. Mol. Cell. Biol. 13, 5233–5244. Ward, P., and Roizman, B. (1994). Herpes simplex genes: The blueprint of a successful human pathogen. Trends Genet. 10, 267–274. Weinheimer, S. P., Boyd, B. A., Durham, S. K., Resnick, J. L., and O’Boyle, D. R. (1992). Deletion of the VP16 open reading frame of herpes simplex virus type 1. J. Virol. 66, 258–269. Whitley, R. J., and Gnann, J. W. (1993). The epidemiology and clinical manifestations of herpes simplex virus infections. In ‘‘The Human Herpesviruses’’ (B. Roizman, R. J. Whitley, and C. Lopez, Eds.), pp. 69–105. Raven Press, New York. Whitley, R. J., and Lakeman, F. (1995). Herpes simplex virus infections of the central nervous system: Therapeutic and diagnostic considerations. Clin. Infect. Dis. 20, 414–420. Wilson, A. C., LaMarco, K., Peterson, M. G., and Herr, W. (1993). The VP16 accessory protein HCF is a family of polypeptides processed from a large precursor protein. Cell 74, 115–125. Xiao, P., and Capone, J. P. (1990). A cellular factor binds to the herpes simplex virus type-1 transactivator Vmw65 and is required for Vmw65-dependent protein-DNA complex assembly with Oct-1. Mol. Cell. Biol. 10, 4974–4977