Construction of a Herpes Simplex Virus Type 1 Mutant with Only a Three-Nucleotide Change in the Branchpoint Region of the Latency-Associated Transcript (LAT) and the Stability of Its Two-Kilobase LAT Intron

Abstract

Previous studies using a eukaryotic expression system indicated that the unusual stability of the latency-associated transcript (LAT) intron was due to its nonconsensus branchpoint sequence (T.-T Wu, Y.-H. Su, T. M. Block, and J. M. Taylor, Virology, 243:140-149, 1998). The present study investigated the role of the branchpoint sequence in the stability of the intron expressed from the herpes simplex virus type 1 (HSV-1) genome and the role of LAT intron stability in the HSV-1 life cycle. A branchpoint mutant called Sy2 and the corresponding rescued viruses, SyRA and SyRB, were constructed. To preserve the coding sequence of the immediate early gene icp0, which overlaps with the branchpoint region of the 2-kb LAT, a 3-nucleotide mutation into the branchpoint region of the 2-kb LAT was introduced, resulting in a branchpoint that is 85% identical to the consensus intron branchpoint sequence of eukaryotic cells. As anticipated, there was a 90- to 96-fold reduction in 2-kb LAT accumulation following productive infection in tissue culture and latent infection in mice with Sy2, as determined by Northern blot analysis. These results clearly suggest that the accumulation of the 2-kb intron in tissue culture and in vivo is, at least in part, due to the nonconsensus branchpoint sequence of the LAT intron. Interestingly, a failure to accumulate LAT was associated with greater progeny production of Sy2 at a low multiplicity of infection (0.01) in tissue culture, but not in mice. However, the ability of mutant Sy2 to reactivate from trigeminal ganglia (TG) derived from latently infected mice was indistinguishable from that of wild-type virus, as assayed in the mouse TG explant reactivation system.

Herpes simplex virus type 1 (HSV-1) is a common human pathogen that can cause various painful clinical syndromes by way of primary infection and recurrent disease. After active infection with HSV-1, the human host retains the genetic information of the virus in a latent form for the remainder of the host's lifetime (53). Latent HSV-1 resides within neurons of the peripheral nervous system in the absence of overt disease, but reactivation of HSV-1, with resultant clinical presentation, is always possible (53, 67).

During latency, HSV expresses an unusual set of latency-associated transcripts (LATs) that correspond to the intron sequences derived from an 8.3-kb primary transcript (21), which appears to exist in only low abundance during latency (17, 72). The 2-kb LAT introns have been widely studied because (i) they are the only abundant family of transcripts found during the latent phase of an HSV-1 infection (51, 59, 61) and (ii) they are believed to play a key role in HSV-1 latency.

One approach to study the function of the 2-kb LAT has been by generating a family of LAT mutant constructs. Results from these studies indicate that LATs are pleiotropic molecules possibly functioning in several ways to influence HSV-1 latency (43, 56, 64), reactivation (5, 8, 25, 37, 45), or neuronal survival (46, 65), as summarized below. First, the 2-kb LAT has been shown to down-regulate productive infections in sensory neurons and suppress viral replication (12, 22, 39), thus promoting the establishment of latent infection in these neurons (43, 56, 64). Second, it has been suggested that LAT functions in maintenance of viral latency, possibly by acting as antisense RNA to the immediate early gene icp0, thus blocking viral replication (21, 61). However, Burton et al. (10) recently suggested that the LATs are unlikely to involve a direct-acting anti-icp0 antisense mechanism, suggesting rather that the LAT region could affect icp0 mRNA expression from the viral genome, acting as a cis element. Third, LAT facilitates reactivation of HSV-1 in the rabbit spontaneous (25, 44)- and induced-reactivation (7, 66) models and the induced-murine-reactivation model (6, 31, 37, 38, 48, 56, 60). However, other studies (19, 30) more specifically suggested that the LAT's reactivation function does not require the 2-kb intron in the rabbit reactivation model. Furthermore, for a murine model, several studies also reported that mutants unable to transcribe LAT reactivate similarly to the wild-type virus, implying that LAT did not play a significant role in the murine-explant-induced-reactivation model (6, 13, 15, 32). Recent studies by Perng et al. (48) suggest that the mouse strain appeared consequential in the HSV-1 explant-induced-reactivation phenotype in mice. Fourth, it has also been suggested that the 2-kb LAT possesses antiapoptotic characteristics, thereby facilitating neuronal survival (1, 28, 46, 65) and resulting in more latently infected cells able to be reactivated.

Despite extensive characterization of LAT function, its role, if any, in the pathobiology of HSV, is unclear. It is difficult to produce viruses that are “pure” LAT-null mutants. The genome of HSV-1 is extremely compact; the large number of overlapping genes (icp0, icp4, and γ-34.5) and newly described transcripts ORF P and ORF O (9) and LS/Ts (70) in the region, as well as 0.7-, 0.9-/1.1-kb, and 1.8-kb transcripts (71) and AL-RNA (47), overlapping the promoter region of the LAT, often prevents a direct assignment of the observed phenotypes to LAT.

Previous studies by us (68) and others (52) show that the 2-kb LAT accumulates in a nonlinear conformation and exhibits a lariat appearance, produced during splicing. Surprisingly and perhaps of importance, the half-life of the 2-kb LAT is measured in hours, instead of seconds (63, 69). The latter is the half-life unit associated with most introns, whose lariats are normally efficiently debranched and then degraded (57). Since the half-life of an intron is usually very short, the half-life of the LAT intron is unusual, suggesting very inefficient debranching.

In the course of natural RNA processing, introns are removed from mRNA by a succession of molecular events involving specialized factors, such as small nuclear ribonucleoproteins (RNPs) (41). Essential steps in this process include proper recognition of (i) splicing signals, such as splice donor and acceptor sites, and (ii) branchpoint sequences (23, 50). In addition to the branchpoint itself, other factors such as surrounding regions (11, 18), RNA secondary structures (20, 24, 58), and polypyrimidine tract sequences (42, 55) can influence branchpoint selection, thus affecting the efficiency of splicing. Mutation in a lariat branchpoint sequence can cause intron retention and is the basis for at least one inherited human disorder (36).

Based on sequence analysis of the consensus correlations involved in eukaryotic RNA processing, the splice donor, the acceptor site (at −1) (21), and the branchpoint (at −29) (69) of this unique 2-kb LAT intron were previously suggested and confirmed (2, 3, 33, 35). This unusually stable lariat-shaped intron fails to follow the typical debranching kinetics in vivo. It has been suggested that the basis for this is either a defect in the branchpoint sequence or the RNA secondary structure (35, 69).

Our previous findings demonstrate that mutagenesis of a LAT expression cassette with a 3- or 4-nucleotide (nt) change at the region of the branchpoint eliminated at least 90% of the Northern blot-detectable 2-kb LAT (69). These mutants also offer a novel way to reduce or abolish the accumulation of LATs with only minimal changes in the HSV-1 genome. In this report, we introduced a 3-nt mutation into the HSV-1 genome that not only eliminated LAT accumulation but also did not alter the coding sequences for the overlapping icp0 gene. Thus, the mutation permitted us to use both cell cultures and an animal model to study the consequences of eliminating the 2-kb LAT intron accumulation.

MATERIALS AND METHODS

Cells, media, and virus.

CV-1 cells (African green monkey kidney fibroblasts) were maintained as monolayers in an Eagle medium supplemented with 5% newborn calf serum at 37°C and 5% CO₂. HSV-1 viruses were propagated by inoculating CV-1 cells. DNA was isolated from virions as described by Pignatti et al. (49).

Plasmids, DNA transfections, and isolation of recombinant virus.

Plasmid p6/7 DNA was used to construct branchpoint mutant Sy2 from the HSV-1 wild-type strain 17+. Plasmid p6/7 contains the desired 3-nt mutation in the branchpoint region of the 2-kb LAT intron, and this plasmid is derived from pSVL (containing a 2.4-kb fragment spanning the 2-kb LAT region of the genome from HSV-1 wild-type strain 17+), as described previously (69). This plasmid was linearized and cotransfected with the HSV-1 wild-type strain 17+ genomic DNA into monolayers of CV-1 cells by using a calcium phosphate method to construct the branchpoint mutant Sy2. To construct rescued viruses SyRA and SyRB, the plasmid pSVL was used for cotransfection with the mutant Sy2 genome. On day 1 posttransfection, the transfected CV-1 cells were overlaid with the methylcellulose medium. Individual plaques were isolated by using a pipette.

Screening for recombinants by performing PCR on lysate from infected cells.

Each individual isolated plaque from the transfected culture was centrifuged for 1 min at 12,000 × g. The cell pellet was resuspended in 100 μl of 0.3 M NaOH and incubated at 37°C for 1 h. Infected-cell lysate was then neutralized with 150 μl of 0.3 M Tris (pH 7.5). PCR screening was conducted by reacting 1 μl of the cell lysate with two sets of primers (Fig. 1B). The first set of primers, 242 (positions 120877 to 120898; 5′CCTTCTTGTTTTCCCTCGTCCC3′) and 243m (positions 121384 to 121405; 5′AAGAGGAAACGTCACTGGGCC3′) amplified only the 3-nt-mutated sequence and detected only the mutant recombinant (Fig. 1B). The second set of primers, 242 and 243 (positions 121437 to 121416; 5′CTCGGGCGGGGCCGTCGGTGCC3′) amplified both the wild-type (HSV-1 strain 17+) and the 3-nt-mutated (p6/7 and Sy2) sequences (Fig. 1B). PCR was performed as follows: a 20-μl reaction mixture consisting of 1 μl of the cell lysate, 1 mM MgCl₂ and 0.5 U of Taq (Fisher Scientific, Pittsburgh, Pa.) in 1× PCR buffer B (without MgCl₂), 200 μM deoxynucleoside triphosphate, and 0.1 μM primers was mixed, denatured at 95°C for 4 min, and then subjected to 35 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 30 s, followed by 1 cycle of 72°C for 5 min. To screen for the rescued viruses, the PCR products derived from individual plaques with primers 242 and 243 were digested with restriction endonuclease ApaI. This approach distinguished the genomes of the rescued virus from the mutant Sy2 genome, since only the PCR product derived from the mutant genome contained an ApaI site (Fig. 1B). When the desired recombinants or rescued viruses were identified, isolates were plaque purified three times before virus stock was prepared.

FIG. 1.

Schematic depiction of the 3-nt mutation in the branchpoint of the 2-kb LAT intron and location of the LAT probe and primers for screening for recombinants. (A) Locations of the 2.4-kb ApaI fragment, the 3′ end of the 2-kb LAT, the 5′ end of coding sequence of icp0, and the Hpa-Hpa LAT probe as related to the HSV-1 genome (corresponding to GenBank nucleotide sequence accession number 14112). (B) Primers used in this study. The nucleic acid sequences for the primers 242, 243, and 243m and their relative locations in the LAT intron are indicated. The predicted branchpoint for the 2-kb LAT, the 3-nt mutation (italicized) introduced into the predicted branchpoint region, and the new ApaI site (boxed) are shown.

In vitro growth kinetics.

Viral growth kinetics were analyzed with CV-1 monolayers inoculated at a multiplicity of infection (MOI) of either 0.1 or 0.01. After infection at 37°C for 1 h, infected CV-1 cells were treated with a citrate buffer (pH 3) to inactivate viruses remaining in the medium. The buffer-treated cells were incubated with 1 ml of culture medium at 37°C and were then harvested into culture medium at 0, 6, 16, 24, and 48 h after infection. The infected CV-1 cells were subjected to three cycles of freezing and thawing, followed by assays to determine viral titers using CV-1 monolayers.

Northern blot analysis and reverse transcriptase PCR (RT-PCR) of LAT RNA.

To analyze the LAT RNA from infected-tissue culture, CV-1 cells were infected with HSV-1 at a MOI of 5. After 1 h of inoculation, medium was added to the infected culture and the culture was incubated at 37°C for 24 h. To harvest RNA, the infected-cell monolayer was lysed by 1 ml of Trizol reagent (Invitrogen, Grand Island, N.Y.). The cell lysate was then subjected to RNA isolation according to the manufacturer's specifications. To analyze the LAT RNA from murine trigeminal ganglia (TG) that were latently infected with HSV-1, the TG were harvested 30 days after the primary inoculation as described in the next section. Harvested TG were immediately homogenized with 1 ml of Trizol reagent by a homogenizer. The TG lysate was subjected to RNA isolation according to the manufacturer's specifications. RNA was quantified spectrophotometrically at 260 nm.

To perform Northern blot analysis, approximately 2 μg of RNA was resolved by electrophoresis in a 1% agarose-formaldehyde gel and then transferred to a Gene Screen Plus membrane (Dupont, NEN Research Products, Boston, Mass.). The membrane was subjected to hybridization involving the selected ³²P-radiolabeled probe of interest. The thymidine kinase (tk) gene probe (a 3.6-kb BamHI fragment) and the Hpa-Hpa LAT probe were synthesized by the RadPrime labeling kit (Invitrogen, Grand Island, N.Y.) with [³²P]dCTP, followed by purification with G50 columns (Shelton Scientific, Shelton, Conn.). The membrane was hybridized with the probe of interest in RapidHy buffer (Amersham Biosciences, Piscatawav, N.J.) at 68°C. The hybridized membrane was washed at 70°C for the Hpa-Hpa LAT or at 68°C for the tk gene probe. An autoradiographic image was generated and quantified by analysis with a phosphorimager (Bio-Rad, Hercules, Calif.). To reprobe the membrane, the blots were stripped by a 15-min wash three times with 0.2 M NaOH at 42°C, followed by one wash with 2× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate) for 10 min.

To perform RT-PCR, 0.5 μg of RNA was treated with DNase I (Invitrogen) to eliminate DNA contamination and then subjected to cDNA synthesis with the SuperScript preamplification system (Invitrogen) according to the manufacturer's instructions in a total volume of 20 μl. PCR amplifications were done with 2.5 U of Taq polymerase (Fisher Scientific), 0.2 μM primers specific for the LAT 5′ exon (16) or icp27 (62), and 1 μl of cDNA or 0.025 μg of RNA (control). PCR amplifications consisted of 40 cycles of denaturation at 94°C for 30 s, annealing at 60°C for 30 s, and extension at 72°C for 30 s. PCR products were resolved by 2% agarose gel electrophoresis and visualized by ethidium bromide staining.

Inoculation of mice, in vivo growth kinetics, and reactivation.

After corneal scarification, female BALB/c mice (Charles River Breeding Laboratories, Kingston, N.Y.), 4 to 6 weeks old, were inoculated in each eye with approximately 10⁶ PFU of HSV-1 wild-type strain 17+, mutant strain Sy2, or rescued virus SyRA or SyRB. Animals were sacrificed at days 0, 1, 3, and 5 after infection, and infected eyes and TG were removed, snap-frozen, and stored at −80°C. These tissues were homogenized in 1 ml of minimum essential medium, clarified for 5 min at 5,000 × g, and assayed for infectious titer on CV-1 monolayers.

The remaining mice were maintained for the reactivation study. At 30 days postinfection, the mice were sacrificed. The TG were removed, and each one was transferred onto one well of a six-well tray containing CV-1 cells. Each culture medium was collected daily and assayed for the appearance of reactivated infectious virus on CV-1 monolayers.

RESULTS

Construction of an HSV-1 LAT branchpoint mutant with a 3-nt mutation.

Previously, we proposed that the accumulation of 2-kb LAT in tissue culture cells and murine TG infected with HSV-1 is due to inefficient degradation of this LAT intron after splicing (69). The inefficient degradation was reasoned to be a consequence of the lack of a consensus 7-nt branchpoint sequence in the 2-kb intron. In a previous study (69) we introduced a branchpoint where 6 out of 7 nt matched the consensus (6-7) and one where there was a perfect match (7-7) into the LAT intron and demonstrated that this matched-branchpoint mutation could substantially facilitate degradation of the LAT intron after splicing when LAT was transcribed and spliced from a eukaryotic expression vector. At least a 90-fold decrease in the amount of the 2-kb lariat relative to wild type accumulated in transfected Huh7 cells 3 days after transfection was observed.

To further investigate whether the consensus branchpoint sequence of the LAT intron could prevent accumulation of the 2-kb LAT in HSV-1 in vitro and in vivo, a LAT branchpoint mutant virus was constructed. As mentioned, the branchpoint region of the 2-kb LAT overlaps with the open reading frame of the immediate early gene icp0 and two previously constructed plasmids, one containing the LAT intron with a perfect matched consensus branchpoint sequence (7-7) and the other containing the LAT intron with an imperfect matched branchpoint sequence (6-7; plasmid p6/7) (69). Both plasmids specified a phenotype of an unstable LAT intron in an in vitro transcription system. The p6/7 plasmid was chosen to construct the desired LAT branchpoint mutation because its 3-nt mutation does not alter the amino acid sequence encoded by icp0. In addition, introduction of this 3-nt mutation into the branchpoint region of the 2-kb LAT creates an ApaI site, which was used to verify the success of introducing the mutation into the parental genome.

To construct the LAT branchpoint mutant, the genomic DNA from the HSV-1 parental strain 17+ and linearized plasmid p6/7 DNA were cotransfected into a CV-1 culture and the culture was incubated with medium containing methylcellulose. After plaques appeared in the transfected culture, the individual plaques were picked and lysed and the lysate was subjected to PCR screening as described in Materials and Methods. The 3-nt recombinant mutant HSV-1, designated Sy2, was isolated and plaque purified. The mutation was detected by restriction endonuclease ApaI digestion of the PCR products generated by primers 242 and 243, as shown in Fig. 2B. The presence of the mutation was confirmed by nucleotide sequencing (data not shown).

FIG. 2.

PCR screening and mapping of branchpoint mutant Sy2 and rescued viruses, SyRA and SyRB. (A) Specificity of primers 242 and 243 and 242 and 243m. DNA from plasmid p6/7 (containing the 3-nt branchpoint mutation) and DNA from HSV-1 wild-type strain 17+ were used to verify primer specificity. Left, 560-bp PCR products amplified from p6/7 and 17+ DNA templates by primers 242 and 243 and 528-bp PCR products amplified from only the p6/7 plasmid template by primers 242 and 243m; right, ApaI digestion of PCR products generated by using primers 242 and 243. (B) Genetic analysis of HSV-1 17+, its branchpoint mutant Sy2, and the rescued viruses, SyRA and SyRB. PCRs with two sets of primers, 242 and 243 and 242 and 243m, were performed as described in Materials and Methods. One-half of the PCR products derived from primers 242 and 243 were digested to completion with the restriction endonuclease ApaI. PCR products derived from primers 242 and 243 with and without ApaI digestion (top) and from primers 242 and 243m (bottom) and molecular weight markers (MW) were resolved in a 1.5% agarose gel, stained with ethidium bromide, and photographed.

To prove that the phenotype of interest associated with mutant Sy2 was due to the change at the expected branchpoint site, it was necessary to obtain a rescued virus. Independent rescued viruses SyRA and SyRB were constructed and isolated, followed by plaque purifications (three times). Figure 2B shows the detection of a novel ApaI site in the PCR products amplified from the Sy2 branchpoint region by using primers 242 and 243. This ApaI site was absent in the branchpoint region of the genome from the parental HSV-1 strain (17+). Thus, as in the genome of the control, this ApaI site was also absent in the genomes from the rescued viruses SyRA and SyRB. Furthermore, the primers 242 and 243m could amplify only the genome for Sy2 and the positive control, p6/7. No corresponding PCR product could be detected from the templates of the parental strain 17+ or from the rescued virus strains, SyRA and SyRB. Thus, the LAT branchpoint mutant Sy2 and its rescued viruses, SyRA and SyRB, were successfully constructed.

Effect of the 3-nt branchpoint mutation on the stability of the LAT intron.

To test our hypothesis that accumulation of the 2-kb LAT intron in HSV-1-infected cells was due to its nonconsensus (or lack of a consensus) branchpoint sequence, we evaluated the accumulation of the LAT intron when the CV-1 cells were infected with the 6-7-matched-consensus-branchpoint mutant Sy2. CV-1 cultures were inoculated with the wild-type parental HSV-1 strain (17+), the branchpoint mutant Sy2, or its rescued viruses, SyRA and SyRB, at a MOI of 5. Infected cultures were harvested at 6 and 24 h after infection. Total RNA was isolated, and analyzed by Northern blot hybridization and RT-PCR. Figure 3A shows an image of the agarose gel stained with ethidium bromide before it was transferred to the nylon membrane to show both the integrity and the relative amounts of RNA loaded into each lane. The RNA membrane was analyzed with LAT-specific probe Hpa-Hpa (Fig. 1A) labeled with ³²P. As predicted for RNA derived from cells infected with parental strain 17+ virus, the 2-kb LAT intron accumulated and was readily apparent 24 h after infection (Fig. 3B). In contrast, with RNA derived from the culture cells infected with the LAT branchpoint mutant Sy2, there was very little LAT intron detected, even though the amount of RNA loaded was approximately twice that used for strain 17+ (as determined by densitometry analysis of the ethidium bromide-stained gel; Fig. 3A). By phosphorimager analysis, the signal intensity for the 2-kb LAT was approximately 4% of the signal intensity from parental strain 17+ in this experiment.

FIG. 3.

Detection of the 2-kb LAT intron in cells infected with wild-type and recombinant HSV-1 virus. (A) CV-1 cultures were either mock infected or infected with HSV-1 strain 17+, Sy2, SyRA, or SyRB at a MOI of 5. At 6 and 24 h postinfection (PI), infected cultures were harvested, total RNA was isolated from each infected culture, and the amount of RNA was determined and analyzed by Northern blot hybridization or RT-PCR, as described in Materials and Methods. To detect the 2-kb intron of LAT transcript, total RNA from each sample was loaded and resolved by electrophoresis in a 1% agarose gel and photographed after staining with ethidium bromide to show the integrity of the RNA and the relative amount of RNA loaded in each lane. (B) RNA was transferred onto a membrane and hybridized with the Hpa-Hpa LAT probe (as described for Fig. 1A) to detect the LAT intron. (C) The membrane was washed and rehybridized with the tk gene probe. To detect the primary transcript of LAT, RT-PCR reactivations with primers specific for the 5′ exon of the LAT transcript or the icp27 transcript were performed. (D) PCR products derived from the cDNA (+RT) or RNA (−RT) were visualized in an ethidium bromide-stained gel. Results are from three independent experiments.

The wild-type LAT phenotype (accumulation of the 2-kb LAT intron) was restored to the Sy2 rescued viruses, SyRA or SyRB, as indicated by RNA analyzed from cells infected with the rescued viruses. A hybridized signal was observed above the LAT band in the RNA isolated from the cultures infected with all four viruses. The species corresponding to this signal is not clear as yet; the signal must be virus specific since it did not appear in the mock-infected-RNA lane. Similar experiments were repeated two more times to quantify the reduction in the amount of the 2-kb LAT intron in the Sy2 mutant-infected cultures relative to that for wild-type strain 17+ and showed a range of 90- to 96-fold reduction in three experiments.

As a positive control for successful infection of Sy2 into CV-1 cells, the same membrane was washed and reanalyzed with a probe consisting of a 3.6-kb tk gene BamHI fragment (Fig. 3C). The tk gene is an early gene that is down-regulated at the late stage of infection (26, 54, 67). As expected, 6 h after infection, the tk gene transcripts were readily detected from the cultures infected with either wild-type strain 17+ or the mutant Sy2. The amount of tk gene transcripts decreased 24 h after cultures were infected for all four viral strains used in this study. Therefore, the fact that the LAT transcript was nearly undetectable in the Sy2-infected culture was not due to insufficient viral infection. Rather, within the context of the HSV-1 genome, the 2-kb LAT intron was greatly reduced (a 90- to 96-fold reduction) by this 3-nt mutation that converts the LAT's nonconsensus branchpoint sequence to the 6-7 matched consensus sequence.

To rule out the possibility that the lack of detectable 2-kb LAT intron in the Sy2-infected culture is due to the effect of the 3-nt mutation on the transcription of LAT rather than the stability of the intron, RT-PCR with primers specific for the LAT 5′ exon (PCR product positions 118888 to 119036 [16]) was performed to detect the LAT primary transcript, as shown in Fig. 3D. The LAT primary transcripts were readily detected in the cDNA from cultures infected with all four different virus strains, including LAT branchpoint mutant Sy2, 24 h postinfection. Thus, the 3-nt mutation did not significantly reduce the transcription of LAT.

In vitro and in vivo growth analysis of the LAT branchpoint mutant Sy2.

To study how accumulating the LAT intron affects the life cycle of HSV-1, we determined the growth kinetics for the branchpoint mutant Sy2 by inoculating CV-1 cells and murine tissues (eye and TG). First, the in vitro growth kinetics of LAT branchpoint mutant Sy2 were determined by inoculating a CV-1 monolayer with the parental strain 17+, mutant Sy2, or one of its rescued viruses, SyRA or SyRB, at a MOI of 0.1. At 6 and 24 h after inoculation, infected cultures were harvested and assayed for the titer of infectious virus by using the standard plaque assay on CV-1 cultures. After 24 h of infection, all four viruses replicated and generated infectious virus particles to similar levels (∼10⁵ PFU per culture; data not shown). No significant difference in growth kinetics in CV-1 culture at a MOI of 0.1 was observed for these four viruses.

To increase the assay sensitivity, growth experiments were repeated at a lower MOI of 0.01. Interestingly, as shown in Fig. 4A, mutant Sy2 replicated to a significantly higher titers at 16 and 24 h after infection than wild-type strain 17 and rescued viruses, as shown in standard deviations and as analyzed by the Kruskal-Wallis one-way analysis of variance (P < 0.05). However after 48 h, titers of all four viruses peaked to similar levels (∼10⁷ PFU per culture), and the differences among them were not significant. This suggests that the accumulation of the 2-kb LAT intron exerted a modest inhibitory effect on the productive cycle of HSV-1 in CV-1 cells.

FIG. 4.

Effect of LAT intron stability on HSV-1 replication in vitro and in vivo. (A) In vitro growth of the branchpoint mutant. CV-1 cultures were infected with HSV-1 strain 17+, Sy2, SyRA, or SyRB, as described in Materials and Methods. Infected cultures were harvested at the indicated times after infection; the amount of infectious virus on CV-1 monolayers was determined. Each symbol represents the average of the viral titers from three experiments, and the standard deviations are indicated by the error bars. (B and C) The branchpoint mutation's effect on growth kinetics for HSV-1 in murine eyes (B) and TG (C). Murine eyes were infected with 10⁶ PFU of HSV-1 strain 17+, Sy2, SyRA, or SyRB/eye after corneal scarification, as described in Materials and Methods. Infected eyes and TG were harvested at the postinfection days indicated. Tissues were then homogenized and measured for the content of infectious virus. Each symbol represents the viral content from individual tissue. The bar in each column represents the average of the viral content for each of the four groups during each indicated assay day after infection. Results are from two independent experiments.

We then studied the growth of Sy2 in vivo. Five-week-old BALB/c female mice were infected with the parental strain 17+, LAT branchpoint mutant Sy2, or either of its rescued viruses, SyRA and SyRB, as described in Materials and Methods. Infected eyes and TG were harvested and homogenized at days 0, 1, 3, and 5 postinfection; the titer of infectious virus from each homogenized sample was determined by using the standard plaque assay on CV-1 cultures. Figure 4B and C show the Sy2 growth patterns in the infected eye and TG, compared with those for the parental HSV-1 strain 17+ and for Sy2's rescued viruses (SyRA and SyRB). No significant difference for TG was observed for all the time points studied. Therefore, within the limit of assay sensitivity and animal variation, the stability of the 2-kb LAT intron did not appear to affect the capability of HSV-1 to replicate in the eye and TG of the BALB/c mice.

The effect of a branchpoint mutation on LAT accumulation during latent infection of HSV-1 and the role of this mutation during viral reactivation.

One feature regarding HSV-1 latency is the accumulation of abundant 2-kb LAT transcripts in latently infected TG. We used the murine model for HSV-1 latency to investigate whether the mutation that eliminates the accumulation of the 2-kb LAT intron in tissue culture has the same effect in vivo on both the accumulation of the LAT during the latent state and the kinetics of viral reactivation in explant TG. Murine eyes were infected with HSV-1 wild-type strain 17+, the Sy2 mutant, or rescued virus SyRA or SyRB, as described in Materials and Methods. The animals were sacrificed 30 days after infection. Latently infected TG were harvested and analyzed for (i) the accumulation of the LAT transcript by using Northern blot hybridization and (ii) the reactivation of the virus.

Total RNA was isolated from murine TG, as described in Materials and Methods. Two micrograms of total RNA was resolved by electrophoresis in a 1% agarose gel. The gel was then stained with ethidium bromide and photographed (Fig. 5A). The RNA gel was transferred onto a nylon membrane and hybridized with the LAT Hpa-Hpa probe (Fig. 5B) labeled with ³²P. As anticipated, the 2-kb LAT intron was readily detected in the RNA isolated from TG that were latently infected with either strain 17+ or the rescued virus SyRA. No LAT intron was detected from the TG that were latently infected with the branchpoint mutant Sy2. Since the reactivation experiments (see below) show that the Sy2 mutant established latent infection in murine TG, the lack of detectable LAT introns by Northern blot analysis (Fig. 5) was not due to the absence of infection by the Sy2 strain of HSV-1 in TG. Thus these results suggest that the 3-nt mutation at the branchpoint of the 2-kb LAT intron destroys the unusual stability of this LAT intron, which is then undetectable in murine TG latently infected with Sy2.

FIG. 5.

A 3-nt mutation in the branchpoint region of the 2-kb LAT intron eliminates accumulated LAT in latently infected murine TG. Mice were either not inoculated or were infected with 10⁶ PFU of HSV-1 strain 17+, Sy2, or the rescued virus SyRA/eye. After 30 days of infection, TG from infected mice were removed and subjected to total RNA isolation as described in Materials and Methods. Approximately 2 μg of total TG RNA was analyzed by electrophoresis in a 1% agarose gel. The gel was photographed after staining with ethidium bromide (A) and transferred to a membrane and hybridized with the Hpa-Hpa LAT probe for the LAT intron (B). These data are results from two independent experiments.

Next, we studied the effect of the unstable 2-kb LAT on HSV-1 reactivation in a murine TG explant reactivation model as described in Materials and Methods. Groups of 12 TG were used to calculate the percentage of reactivated cultures (Fig. 6A). As expected, no virus was detected for the first 3 days after explantation. Reactivated viruses were detected 4 days after explantation of the culture containing TG infected by the wild-type strain 17+, mutant strain Sy2, or rescued virus SyRA. After 5 days of cocultivation, latent virus was reactivated in at least 40% of cultures for all four groups. Reactivation reached 90 to 100% within 6 days of cocultivation for all four viral strains. Thus, within the sensitivity range of our assay, we noted no significant difference in the kinetics of reactivation among the four viral strains tested, including the branchpoint mutant Sy2.

FIG. 6.

Kinetics of reactivation and genetic analysis of reactivated virus. (A) Mice were infected with HSV-1 strain 17+, Sy2, SyRA, or SyRB, as described in Materials and Methods. After 30 days of infection, infected TG were removed and individual explant TG were placed on a CV-1 monolayer for the reactivation study. To detect reactivated viruses, cocultivated medium was collected daily and assayed for the appearance of infectious progeny. The percentage of reactivation was calculated daily. There were 12 TG in each group. (B) To genetically compare reactivated virus with the HSV strains used as the original inocula, the medium collected from two individual explanted latently infected TG on day 9 after cocultivation (a and b; randomly chosen from each group) was centrifuged, lysed, and subjected to PCR with primers 242 and 243. The PCR product from each reaction was digested with ApaI (+) or left undigested (−) and then analyzed by electrophoresis on a 1.5% agarose gel. The gel was photographed after staining with ethidium bromide, as shown in panel B.

Genetic characterization of reactivated virus.

To verify that the reactivated viruses compared genetically to the original virus used to inoculate each mouse, reactivated viruses were randomly chosen from two individual reactivation cultures on day 11 postexplantation. These randomly chosen specimens were centrifuged and subjected to the PCRs with primers 242 and 243 to amplify the region covering the branchpoint sequence (Fig. 1B). As mentioned, the Sy2 genome contained a novel ApaI site introduced into the branchpoint sequence of its 2-kb LAT intron. After the completion of the PCRs, one-half of the PCR products were digested with restriction endonuclease ApaI and analyzed by electrophoresis on a 1.5% agarose gel (Fig. 6B). As expected, no ApaI site was detected in the branchpoints of the LATs from the reactivated viruses in the murine TG infected with wild-type strain 17+ or rescued virus SyRA or SyRB. The novel ApaI site was clearly detected in the branchpoint of the LAT intron for the Sy2 reactivated virus (Fig. 6B). No revertant virus was detected in this assay. Thus, the 3-nt mutation introduced into the LAT's branchpoint was stable in the HSV-1 genome.

DISCUSSION

We have previously shown that the unusual stability of the LAT intron could be abolished by introducing a consensus branchpoint sequence into this LAT intron in a eukaryotic transcription system, while its efficiency for transcription and splicing was unaffected (69). Here we introduced a 3-nt mutation, resulting in a sequence that was an 85% (instead of a 100%) match for the consensus branchpoint sequence into the HSV-1 genome (Sy2 mutant), in order to preserve the amino acid sequence encoded by icp0, which overlaps with the branchpoint sequence of the 2-kb LAT, and clearly demonstrate that this 3-nt mutation destroyed the unusual stability of the 2-kb intron both in tissue culture and in mice, thus producing the LAT-null phenotype. Northern blot analysis indicates that the 3-nt branchpoint mutation eliminates the phenotype for accumulating the 2-kb LAT intron in both productive and latent infections. This once again suggests that the unusual stability of the 2-kb LAT intron is, at least in part, due to its nonconsensus branchpoint sequence.

Although there are many previous modifications made to the 2-kb LAT region, such as deletion (4, 5), insertion (6, 8), and ectopic genomic rearrangement (45), this is the first replication-competent mutant virus constructed without deletions or insertions of foreign DNA inserts. The Sy2 mutant here contains only a 3-nt mutation resulting in a LAT-null phenotype, with no alteration in the amino acid sequence encoded by the icp0 gene. Furthermore, this mutation does not alter the sequences of any known transcripts that have been found around this region. Thus, besides being important for determining the importance of the branchpoint sequence for 2-kb LAT intron stability, as reported here, this branchpoint mutant is extremely valuable for investigating the role of this 2-kb LAT intron in HSV latency and antiapoptotic function. Due to the scope of this study, phenotypic characterization of this mutant was centered on the stability of 2-kb LAT intron both in vitro and in vivo and the role of the 2-kb LAT in HSV-1 replication in vitro and in vivo. It was of interest to only briefly examine the role of the 2-kb LAT in reactivation efficiency in the HSV-1 murine TG explant-induced-reactivation model.

By Northern blot analysis we showed that introduction of a branchpoint with 85% of the consensus branchpoint sequence into the LAT intron of the HSV-1 genome results in a 90- to 96-fold reduction in LAT intron accumulation. Interestingly, a similar reduction (90-fold) was observed in the previous studies using a 2.4-kb ApaI fragment (positions 119269 to 121565) in a eukaryotic transcription system in vitro (69). These data further suggest that the region outside of this 2.4-kb ApaI region of the HSV-1 genome appears to play no detectable role in the stability of the 2-kb LAT intron. Within the limit of the Northern blot analysis, the residual 2-kb LAT in TG infected with mutant Sy2 was not detected. Although no quantification can be done, the reduction of 2-kb LAT intron accumulation due to the 3-nt mutation during latency in vivo is apparent. Although no analysis of the influence of this branchpoint mutation on the stability of 1.5-kb LAT intron has been done, we expect that the stability of 1.5-kb intron of the Sy2 mutant is greatly reduced as well, since the same branchpoint region is shared by these two LAT introns.

As mentioned earlier, results from extensive studies of the intron branchpoint sequence indicate that this sequence plays an important role in the efficiency of splicing or debranching in vitro (27, 29). Interestingly, our previous studies (69) indicate that this sequence appears to influence only debranching efficiency, not splicing efficiency. Support for this theory includes results indicating that the 2-kb LAT intron was relatively resistant to the debranching enzyme (35, 52, 68). Also, by introduction of a matched consensus branchpoint sequence, the accumulation of the LAT intron was eliminated, and no significant effect on efficiency of transcription and splicing was observed (69). Here we introduced the same matched consensus branchpoint sequence in the context of the complete HSV-1 genome and once again show that this mutation eliminates accumulation of the LAT intron. We have not directly shown that the low or undetectable level of the LAT intron in the cultures or mouse TG infected with the branchpoint mutant Sy2 is due to increased debranching efficiency rather than inefficient splicing of the primary transcript. However, the latter is less likely because no accumulation of unprocessed primary transcript was detected (Fig. 3B) and the amount of primary transcript in the Sy2-infected culture was not significantly less than that of 17+-infected cultures (Fig. 3D).

As mentioned previously, this branchpoint LAT-null mutant was constructed with only a 3-nt mutation and no alteration in the amino acid sequence encoded by icp0 and, theoretically, the mutation does not alter any other known transcripts encoded in this complex region of the HSV-1 genome. Interestingly, when this LAT-null mutant was characterized in tissue culture, the accumulation of 2-kb LAT intron appeared to inhibit the growth of HSV-1 modestly at a low MOI (0.01), although this modest inhibitory effect was not observed in vivo in mouse tissues. This might be due to its function in inhibiting immediate early and early phases of viral gene expression, as suggested in various studies (34, 54). Due to the animal variation and errors from experimental handling, the sensitivity for detecting this modest overall inhibition in viral growth might be too low for detection in vivo.

The most widely used method to screen for the desired recombinant HSV is to isolate the individual plaques and then amplify the virus in tissue cultures to provide enough viral DNA for the Southern hybridization and thus identify isolates possessing the desired mutation. To facilitate screening for the branchpoint mutation containing a change of only 3 nt, we developed a novel approach to directly identify the mutant virus from isolated plaques by using specific primers to amplify the mutant sequence (Fig. 1B). The time from appearance of individual plaques in transfection cultures until identification of the desired mutant could be as short as several hours. This approach has been successfully used to identify the desired branchpoint mutant Sy2 and its rescued viruses, SyRA and SyRB.

This report focuses on the importance of the branchpoint sequence for the stability of 2-kb LAT intron. However, it was of interest to study the reactivation phenotype of the branchpoint LAT mutant because of the controversial results obtained regarding the reactivation phenotypes of various LAT mutants in the murine TG explant-induced-reactivation model. As reported here, the branchpoint LAT mutant Sy2 appeared to reactivate with the wild-type phenotype in murine TG explant induction model. This was not a surprise since the data agree with our previous studies with 17 BstE mutants (14, 40), which indicate that the 2-kb LAT intron is not important for reactivation in the murine TG explant reactivation model. Although other studies have suggested that the 2-kb LAT was important for a wild-type reactivation phenotype in murine TG explant-induced-reactivation model (13, 37, 40), the deletion or insertion resulting in the LAT-null phenotype in these studies might also interfere with other known or unknown transcripts, which was not detectable at the time of the studies. Thus it is uncertain whether the phenotypes observed in these LAT mutants are attributable to the disruption of the 2-kb LAT intron or other transcripts. Perng et al. (48) have suggested that mouse strain, age, and dose of infection could be factors that contribute to the confusion in the function of the LAT in reactivation when this was assayed in murine TG explant-induced-reactivation model.

In conclusion, we have demonstrated that a mutation to the consensus branchpoint sequence destroys the unusual stability of the 2-kb LAT intron, as observed in both productive and latent infections. These results indicate the important role of the branchpoint sequence in the stability of HSV-1 intron. A modest inhibitory effect of 2-kb LAT intron accumulation on HSV-1 replication in tissue culture was observed at a low MOI. However, no substantial effect was observed in terms of viral growth in vivo or effect on reactivation, as determined by using the mouse TG explant model. Although there was no attempt to dissect out the role of 2-kb intron in the regulation of HSV-1 gene expression and the mechanisms of action in this study, the Sy2 mutant is a valuable tool for reexamining the antisense role with respect to icp0 and its role in the regulation of other HSV-1 genes (such as icp4 and the tk gene, etc). Current studies are using a rabbit reactivation model to further analyze the role of LAT intron stability in the establishment and reactivation of latency and its regulatory role in the HSV-1 life cycle.