|Home | About | Journals | Submit | Contact Us | Français|
Herpes simplex virus (HSV) establishes latency in sensory nerve ganglia during acute infection and may later periodically reactivate to cause recurrent disease. HSV type 1 (HSV-1) reactivates more efficiently than HSV-2 from trigeminal ganglia while HSV-2 reactivates more efficiently than HSV-1 from lumbosacral dorsal root ganglia (DRG) to cause recurrent orofacial and genital herpes, respectively. In a previous study, a chimeric HSV-2 that expressed the latency-associated transcript (LAT) from HSV-1 reactivated similarly to wild-type HSV-1, suggesting that the LAT influences the type-specific reactivation phenotype of HSV-2. To further define the LAT region essential for type-specific reactivation, we constructed additional chimeric HSV-2 viruses by replacing the HSV-2 LAT promoter (HSV2-LAT-P1) or 2.5 kb of the HSV-2 LAT sequence (HSV2-LAT-S1) with the corresponding regions from HSV-1. HSV2-LAT-S1 was impaired for reactivation in the guinea pig genital model, while its rescuant and HSV2-LAT-P1 reactivated with a wild-type HSV-2 phenotype. Moreover, recurrences of HSV-2-LAT-S1 were frequently fatal, in contrast to the relatively mild recurrences of the other viruses. During recurrences, HSV2-LAT-S1 DNA increased more in the sacral cord compared to its rescuant or HSV-2. Thus, the LAT sequence region, not the LAT promoter region, provides essential elements for type-specific reactivation of HSV-2 and also plays a role in viral neurotropism. HSV-1 DNA, as quantified by real-time PCR, was more abundant in the lumbar spinal cord, while HSV-2 DNA was more abundant in the sacral spinal cord, which may provide insights into the mechanism for type-specific reactivation and different patterns of central nervous system infection of HSV-1 and HSV-2.
During an initial infection, herpes simplex virus (HSV) establishes latency in the sensory nerve ganglia innervating the peripheral site of inoculation. In response to various stimuli, the virus can reactivate from the sensory neurons to cause recurrent disease at or near the original site of inoculation. HSV type 1 (HSV-1) reactivates preferentially from the trigeminal ganglia to cause recurrent orofacial herpes, while HSV-2 reactivates preferentially from the lumbosacral dorsal root ganglia (DRG) to cause recurrent genital herpes (19).
The trigeminal (TG) and DRG sensory neurons in which HSV establishes latency have a single bifurcated axon, with one branch extending to the periphery and the second branch extending to the spinal cord. During primary infection or reactivation, the virus can reach the central nervous system (CNS) via the spinal cord branch of the axon. CNS complications of infection may occur during either acute or recurrent infections. Encephalitis is more prevalent with HSV-1, and recurrent meningitis is more prevalent with HSV-2 (6, 15, 25, 33, 37), although the mechanism for this type-specific difference is not understood.
Previous studies have shown that the HSV latency-associated transcript (LAT) plays an important role in viral latency. HSV-1 and HSV-2 LAT deletion mutants are impaired for recurrent disease (16, 20, 30). HSV-1 LAT is believed to enhance the establishment of latency, either by repression of lytic gene expression (7, 9, 22) or inhibition of apoptosis of neuronal cells (1, 4, 11, 14, 29, 32). Either of these mechanisms could influence neuronal cell survival, increasing the establishment of latency and providing a greater latent pool from which the virus could reactivate. While this may partially explain a role for LAT in reactivation in a general sense, the specific mechanism of type-specific reactivation of HSV-1 and HSV-2 is not understood.
We previously reported on a chimeric virus in which HSV-2 expressed the LAT from HSV-1, including the promoter and sequences extending to near the 3′ end of the LAT intron (38). This chimeric virus, HSV-2 333/LAT1, exhibited a recurrence phenotype more similar to HSV-1 than to HSV-2, with a reduced reactivation frequency from the DRG in the guinea pig genital model and an increased reactivation frequency from TG in the rabbit eye model relative to wild-type HSV-2. Therefore, the LAT region encompassing the promoter, 5′ exon, and intron provides the essential elements for type-specific reactivation of HSV-2.
To further define the region of LAT most important for HSV-2 type-specific reactivation, we constructed two additional chimeric HSV-2 viruses. We divided the region expressed in our previous HSV-2 333/LAT1 chimera into the promoter and sequence regions and replaced these regions in HSV-2 with the corresponding sequences from HSV-1. After in vitro characterization of the chimeric viruses, we tested the phenotypes of these viruses in the guinea pig genital model of HSV infection. We also performed molecular studies of tissues from latently infected animals to evaluate differences in the distribution of viral DNA.
These studies demonstrate that the LAT sequence containing the 5′ exon and the LAT intron contains the essential elements for type-specific reactivation of HSV-2. This LAT region also contributes to the virulence of the virus. We also provide evidence that HSV-1 and HSV-2 preferentially spread to different regions of the nervous system and that LAT influences the efficiency of HSV-2 replication in these different regions.
Vero cells were obtained from the American Type Culture Collection (Manassas, VA) and maintained in minimum essential medium with 10% inactivated fetal bovine serum and 1% penicillin, streptomycin, and l-glutamine (Quality Biologicals, Gaithersburg, MD). HSV-2 strain 333 was obtained from Gary Hayward (Johns Hopkins University, Baltimore, MD). HSV-1 strain 17+ was obtained from John Hay (SUNY-Buffalo, Buffalo, NY). Virus stocks were produced in Vero cells and plaque titered in duplicate. To compare one-step growth characteristics of wild-type and mutant viruses, ~106 Vero cells were inoculated in duplicate at time zero with a multiplicity of infection of approximately 0.1 PFU/cell for each virus. Medium was added after a 2-hour adsorption period. At 0, 2, 5, 12, and 20 h postinoculation, cells were scraped, freeze-thawed three times, and plaque titered in duplicate.
Genome locations for HSV-2 are given relative to the sequence of strain hg52 (8) and for HSV-1 are given relative to the sequence of strain 17+ (24). An AvrII-AluI fragment of HSV-1 strain 17+ and an SphI-BamHI fragment of HSV-2 strain 333 were each cloned into previously described vectors Avr-AluΔXho and Sph-Sal-Bam (38). A single base pair change, which did not influence the ICP0 reading frame or amino acid sequence, was introduced by site-directed mutagenesis, adding an XhoI site to the Avr-Alu clone at position 121259. For chimeric virus HSV2-LAT-P1 (see Fig. Fig.1),1), the HSV-1 promoter region from NotI to PvuI sites (118439 to 118802) was cloned from the HSV-1 Avr-AluΔXho plasmid into the HSV-2 Sph-Sal-Bam plasmid, replacing the HSV-2 LAT promoter region from NotI to PvuI sites (119108 to 119519). For chimeric virus HSV2-LAT-S1, the HSV-1 region from the PvuI site to the added XhoI site (118802 to 121259) was cloned into the HSV-2 Sph-Sal-Bam plasmid, replacing the HSV-2 LAT sequence from the PvuI site to its native XhoI site (119519 to 122276). These two plasmids were used to construct chimeric HSV-2 viruses HSV2-LAT-P1 (promoter swap) and HSV2-LAT-S1 (sequence swap) by homologous recombination after cotransfection of plasmid DNA and parent virus DNA (HSV-2 strain 333) into Vero cells using Lipofectamine reagents (Invitrogen, Carlsbad, CA) in accordance with the manufacturer's protocol. After identification of mutant virus, plaque purification was performed until no parent DNA could be detected by Southern hybridization. Three additional blind plaque purifications were performed, purity was verified by Southern hybridization, and mutant virus stocks were grown and plaque titered in Vero cells. Sequencing of the mutation junctions further validated the mutant virus sequence. A rescuant designated HSV2-LAT-S1-R was made from HSV2-LAT-S1 using the same method, by cotransfecting mutant viral DNA with the wild-type HSV-2 Sph-Sal-Bam plasmid DNA.
Female Hartley guinea pigs (Charles River, Wilmington, MA) were inoculated intravaginally with 2 × 105 PFU of each virus. Guinea pigs were monitored and scored daily during acute infection (14 days) for lesion severity around the vulva (on a scale from 0 to 4, where 0 means no disease, 1 means redness/swelling, 2 means one or two lesions, 3 means three to five lesions, and 4 means six or more lesions or coalescence of lesions) and urinary tract dysfunction manifested by lack of micturition due to nervous system involvement. Urinary tract dysfunction was determined by presence of gross blood manually expressed from the bladder. Recurrences, defined as vesicular lesions, were enumerated during the latent phase from day 15 through day 60 postinoculation (p.i.) and graphed as cumulative recurrences per guinea pig in each group. Two independent investigators, masked to the identity of the inoculated viruses, observed the guinea pigs, and composite scores are presented in the figures. Three separate animal experiments were performed: experiment 1, HSV-2 (n = 8), HSV2-LAT-P1 (n = 5), and HSV2-LAT-S1 (n = 6); experiment 2, HSV-2 (n = 6) and HSV2-LAT-P1 (n = 7); experiment 3, HSV-2 (n = 7), HSV-1 (n = 6), HSV2-LAT-S1 (n = 9), and HSV2-LAT-S1-R (n = 8). All graphs are composites of all three experiments. Animals were housed in an FDA/CBER animal facility, an American Association for Accreditation of Laboratory Animal Care-approved facility, and cared for in accordance with institutional guidelines. Lumbosacral DRG and sacral and lumbar spinal cord tissues were collected from each animal immediately after sacrifice and snap-frozen on dry ice. DNA was isolated from ganglia and spinal cord tissues using the QIAGEN DNA Minikit (Valencia, CA) after homogenization with a rotor-stator homogenizer (Omni International, Marietta, GA). Tissues from uninfected control animals were isolated in parallel.
The copy numbers of viral genomes and the quantity of guinea pig DNA in DRG and spinal cords from infected and uninfected control animals were determined by quantitative real-time PCR using an ABI 7700 Taqman PCR system (Applied Biosystems, Foster City, CA). Primers and probes were specific for sequences within HSV-2 glycoprotein D (gD) and HSV-1 glycoprotein G (gG). The HSV-2 gD primers were TCAGAGGATAACCTGGGA and GGGAGAGCGTACTTGCAGGA (250 nM), and the probe was 6-carboxyfluorescein (FAM)-CCAGTCGTTTTCTTCACTAGCCGCAG-6-carboxytetramethylrhodamine (TAMRA) (200 nM) (36). The HSV-1 gG primers were CTGTTCTCGTTCCTCACTGCCT and CAAAAACGATAAGGTGTGGATGAC (1,000 nM), and the probe was FAM-CCCTGGACACCCTCTTCGTCGTCAG-TAMRA (250 nM). The PCR mixture contained Taqman Universal PCR Master Mix (Applied Biosystems, Foster City, CA), specific primers and probe, and 50 ng of tissue DNA. Cycle conditions were 50°C for 2 min and 95°C for 10 min, followed by 40 cycles of 95°C for 20 s and 60°C for 1 min. A standard curve based on 10-fold dilutions of known amounts of plasmid DNA containing gD or gG coding sequences was used to determine copy numbers. Viral genome copy numbers were normalized to the gene for the 18S rRNA using commercial primers and probe (Applied Biosystems, Foster City, CA).
Statistics were performed using nonparametric analyses with SPSS version 11.5.0, including the Kruskal-Wallis and Mann-Whitney tests (LEAD Technologies). Comparisons of acute infections were based on total area under the lesion severity curve for days 1 to 14 after inoculation. Comparisons of recurrent infections were based on cumulative recurrences per guinea pig, adjusted for the number of days of observation. Viral DNA quantities were compared by analysis of variance and post hoc least significant difference analyses. Error bars represent standard errors of the means for each group.
We previously reported on a chimeric HSV-2, HSV-2 333/LAT1, that expressed the LAT from HSV-1, from the NotI restriction site upstream of the promoter to an XhoI restriction site near the 3′ end of the LAT intron (Fig. (Fig.1).1). HSV-2 333/LAT1 reactivated from TG and lumbosacral DRG with a phenotype similar to HSV-1, and its rescuant had a wild-type HSV-2 phenotype (38). For the present studies, we constructed two additional chimeric HSV-2 viruses to further define the region of LAT essential for type-specific reactivation of HSV-2. We divided the region expressed in our previous HSV-2 333/LAT1 chimera into the promoter region (NotI-PvuI) and the sequence region (PvuI-XhoI) and replaced both copies of these regions in HSV-2 strain 333 with the corresponding sequences from HSV-1 strain 17+. Figure Figure11 shows the relevant genomic regions and endonuclease cleavage sites. A rescuant of HSV2-LAT-S1, designated HSV2-LAT-S1-R, was also generated.
The construction of the chimeric viruses HSV2-LAT-P1 and HSV2-LAT-S1, as well as the rescuant virus HSV2-LAT-S1-R, was verified by Southern hybridization (using three sets of restriction endonucleases) and sequencing of the mutation junction sites. HSV2-LAT-S1-R was indistinguishable from wild-type HSV-2, and each virus yielded the expected bands and sequence (not shown). One-step growth characteristics of each virus were compared in Vero cells. Each of the mutant viruses replicated with wild-type kinetics in cell culture (data not shown).
To determine whether the promoter or the sequence region of LAT provides the essential elements for type-specific reactivation of HSV-2, the viruses were evaluated in the guinea pig genital model. Female guinea pigs were inoculated intravaginally with 2 × 105 PFU of wild-type HSV-2, wild-type HSV-1, chimeric viruses HSV2-LAT-P1 and HSV2-LAT-S1, and the rescuant HSV2-LAT-S1-R. The severities of lesions were compared during the acute phase of infection through day 14 p.i. (Fig. (Fig.2).2). The mean lesion scores for HSV2-LAT-S1, its rescuant, and the wild-type viruses were similar. The acute infection with HSV2-LAT-P1 was less severe than that with wild-type HSV-2 (P = 0.022 by the Mann-Whitney test).
During acute HSV infection, some guinea pigs experience urinary tract dysfunction, indicative of autonomic nervous system involvement. In the present study, the guinea pigs were evaluated for presence or absence of urinary tract dysfunction during the acute phase of infection. A significantly greater percentage of guinea pigs in the group infected with HSV2-LAT-S1 displayed urinary tract dysfunction during the acute phase of disease compared to the group infected with its rescuant, HSV2-LAT-S1-R (P = 0.035), wild-type HSV-2 (P = 0.004), and HSV2-LAT-P1 (P < 0.001) (Fig. (Fig.33).
During the latent phase of infection, the reactivation frequency of HSV2-LAT-P1, the promoter mutant, was similar to that of wild-type HSV-2 (Fig. (Fig.4).4). In contrast, HSV2-LAT-S1, the LAT sequence mutant, reactivated inefficiently in the guinea pig genital model, similarly to wild-type HSV-1 (P = 0.005 compared to HSV-2 and P = 0.692 compared to HSV-1). The rescuant HSV2-LAT-S1-R had a wild-type HSV-2 reactivation phenotype. These data imply that the LAT sequence downstream of the PvuI restriction site, rather than the promoter region, provides the essential elements for type-specific reactivation of HSV-2.
HSV2-LAT-S1, the sequence chimera, unexpectedly produced a greater mortality rate than the other viruses (Fig. (Fig.5).5). Of the 20 guinea pigs inoculated with HSV2-LAT-S1, 15 developed acute symptoms. Two of those 15 died during the acute phase of infection, which is similar to the number observed with wild-type HSV-2 (2 of 21), wild-type HSV-1 (1 of 6), or the rescuant of HSV2-LAT-S1 (0 of 8). Eight of the remaining 13 guinea pigs died following recurrences, which were marked by severe and progressive clinical disease. This is in contrast to wild-type HSV-2 and the rescuant HSV2-LAT-S1-R, which typically caused mild recurrent peripheral lesions that resolved quickly (Kaplan-Meier; P < 0.001 compared to HSV-2 or HSV2-LAT-S1-R).
To evaluate whether latent viral DNA load was responsible for the differences in reactivation, DNA was extracted from tissues harvested from the infected guinea pigs at the time of death. Viral DNA copy numbers were quantified by Taqman real-time PCR assay and normalized to the quantity of the gene for 18S rRNA (Fig. (Fig.6).6). To evaluate differences between the viruses during latent infection and symptomatic recurrences, each group of infected guinea pigs was divided based on the absence or presence of symptoms at the time of tissue harvest. Comparisons during latent infection demonstrate differences in the distribution of the establishment of latency in neuronal tissues. One would expect virus to be actively replicating during symptomatic recurrences; thus, differences between viral DNA loads in animals whose tissues were harvested during symptomatic recurrences would provide an indication of the efficiency of replication in the various regions of the nervous system after viral reactivation.
In asymptomatic latently infected guinea pigs (Fig. (Fig.6A),6A), viral DNA levels of all of the HSV-2 viruses were similar, with the largest quantities of viral DNA found in the sacral spinal cord. However, HSV-1 DNA was primarily found in the lumbar spinal cord (P < 0.0005 compared to all of the HSV-2 viruses), suggesting that HSV-1 is capable of establishing latency in the lumbar cord while HSV-2 is capable of establishing latency in the sacral cord in addition to the DRG.
In guinea pigs with symptoms at the time of tissue extraction, there was no significant difference between the viral DNA copy numbers of HSV-1 and HSV-2 in the DRG (P = 0.527) (Fig. (Fig.6B).6B). However, DNA of all of the wild-type and chimeric HSV-2 viruses was present at significantly higher levels in the sacral spinal cord compared to HSV-1 (P < 0.022). HSV-1 DNA was primarily found in the lumbar spinal cord (P < 0.0005 compared to the HSV-2 viruses). Thus, HSV-1 and HSV-2 replicate more efficiently in different regions of the CNS during symptomatic recurrences, which may provide further insight into the mechanism for type-specific reactivation of HSV-1 and HSV-2 and into different patterns of CNS infection. During recurrences, HSV-2 viral DNA quantities increased in the sacral spinal cord and the DRG compared to levels present in asymptomatic animals, while HSV-1 DNA quantities increased in the lumbar cord and the DRG (compare Fig. 6A and B); this suggests involvement of the spinal cord during reactivation of both HSV-1 and HSV-2.
During latent infection (Fig. (Fig.6A),6A), the viral DNA quantities of HSV2-LAT-S1, its rescuant, HSV2-LAT-P1, and wild-type HSV-2 were similar to each other in the DRG, the sacral spinal cord, and the lumbar spinal cord. During recurrences, guinea pigs infected with HSV2-LAT-S1 had less viral DNA in their DRG than animals infected with wild-type HSV-2 (P = 0.021), although HSV2-LAT-S1 was not significantly different from the rescuant (P = 0.513) (Fig. (Fig.6B),6B), implying that differences in the DRG may not have a significant impact on the behavior of the virus. In the sacral spinal cord, animals infected with the more virulent HSV2-LAT-S1 produced higher levels of viral DNA than those infected with wild-type HSV-2 and the rescuant HSV2-LAT-S1-R during periods of active viral replication (P = 0.007 compared to either HSV-2 or HSV2-LAT-S1-R) (Fig. (Fig.6B).6B). These data suggest that the LAT sequence contributes to the ability of the virus to either spread to or replicate in different regions of the nervous system during reactivations, although the LAT sequences are not the sole determinants of viral spread.
Previously in our laboratory, a chimeric HSV-2 expressing the LAT from HSV-1 failed to reactivate efficiently in the guinea pig genital model of HSV infection (38). One goal of the present study was to determine whether the LAT promoter or the LAT sequence provides the essential elements for type-specific reactivation of HSV-2. The HSV LAT region substituted in our previous chimera was divided into the promoter region (from NotI to PvuI sites) and the sequence region (from PvuI to XhoI sites), and additional chimeric HSV-2 viruses were constructed. HSV2-LAT-P1 is HSV-2 that expresses the native HSV-2 LAT sequence under the control of the HSV-1 LAT promoter. HSV2-LAT-S1 is HSV-2 that expresses HSV-1 LAT sequences, including the 5′ exon and most of the LAT intron, under the control of the native HSV-2 LAT promoter. HSV2-LAT-P1 reactivated efficiently in the guinea pig genital model of infection, although the acute infection was attenuated in both lesion severity and urinary tract dysfunction. The chimeric HSV2-LAT-S1 virus and its rescuant produced acute infections with lesion severities similar to those for wild-type HSV-2 as well as for wild-type HSV-1. However, HSV2-LAT-S1 reactivated inefficiently in the guinea pig genital model of infection, with a reactivation frequency similar to that of wild-type HSV-1, and a wild-type recurrence phenotype was restored with its rescuant, HSV2-LAT-S1-R. Thus, the LAT sequence including the 5′ exon and intron, rather than the LAT promoter, provides the essential elements for efficient genital reactivation of HSV-2. This correlates well with studies on HSV-1, which imply that the region between the promoter and the LAT intron contains the reactivation-critical region (3, 12, 23) and enhancer activities (2, 21) for HSV-1. Since this region of HSV-1 has known enhancer activity, we hypothesize that enhancer activity within this region in both HSV-1 and HSV-2 may play a role in determining site-specific recurrence phenotypes.
Infections with the HSV2-LAT-S1 chimera were characterized by a dramatic increase in virulence and mortality relative to those with wild-type HSV-2, the rescuant HSV2-LAT-S1-R, HSV2-LAT-P1, and wild-type HSV-1. During acute disease, a greater percentage of animals infected with HSV2-LAT-S1 than with HSV-2 developed urinary tract dysfunction, suggestive of more pronounced autonomic nervous system involvement. During latency, HSV-1 and HSV2-LAT-S1 did not reactivate efficiently. During its rare recurrences, HSV-1 produced a single vesicular lesion lasting 1 to 2 days, similar in character to recurrences produced by HSV-2, HSV2-LAT-P1, and HSV2-LAT-S1-R. However, in guinea pigs infected with the HSV2-LAT-S1 chimera, recurrent disease often progressed from a single lesion to multiple lesions coinciding with neurological symptoms, followed by rapid deterioration and death. The most likely explanation for this progression of symptoms is that the HSV2-LAT-S1 chimera was able to spread more efficiently through the nervous system via a change in neurotropism, mediated by its HSV-1 LAT sequences. In a previous report on HSV-1 LAT region deletion mutants, virulence either increased or decreased depending on the size of the deletion and the species infected (31), which suggests that the LAT region has multiple functional elements and that specific neuronal factors may interact with LAT sequences in HSV-1 to regulate viral replication. Fatal recurrences were not observed with the previously tested HSV-2 333/LAT1 chimeric virus, which contained both the HSV-1 LAT promoter and sequence, implying that an interaction of the HSV-2 LAT promoter and HSV-1 LAT sequence may have permitted unusual spread of the HSV2-LAT-S1 virus, contributing to its increased virulence. Alternatively, a cis-acting regulatory element located at the mutation junction site may have been disrupted, permitting unusual replicative characteristics. Although the LAT promoters are fairly well conserved, there is little discernible similarity between the HSV-1 and HSV-2 LAT regions downstream of the promoter. This region in both HSV-1 and HSV-2 is densely populated by putative transcription factor binding sites and enhancer elements, which likely function cooperatively to modulate the activity of the LAT region. LAT sequences could play a role in down-regulation of viral replication, which may fail when non-type-specific sequences are introduced or when specific sequences are deleted, or these LAT sequences could facilitate growth in different types of neurons, which could also give rise to an unusual neurotropic phenotype.
The acute infection caused by HSV2-LAT-P1 was attenuated in both lesion severity and urinary tract dysfunction. However, since HSV2-LAT-P1 demonstrated a wild-type reactivation phenotype, a rescuant of this chimeric virus was not constructed. Any secondary mutation that may have occurred during the construction of the chimera did not affect the reactivation phenotype, although construction of a rescuant would be necessary to ascertain whether the modest differences observed during acute infection with HSV2-LAT-P1 compared to that with HSV-2 were attributable to the LAT promoter substitution.
HSV-1 spread to the lumbar spinal cord more efficiently than HSV-2, while all of the HSV-2-based viruses spread more efficiently than HSV-1 to the sacral spinal cord. The autonomic nervous system innervates the genitalia and bladder through both sympathetic and parasympathetic fibers extending from the paracervical ganglia to the pelvic organs. Presynaptic parasympathetic fibers extend from the sacral spinal cord to the paracervical ganglia, while sympathetic fibers originate in the lumbar cord (13, 26, 27). While HSV-1 has been shown to replicate and establish latency in sympathetic neurons (10, 17, 18, 34), HSV-2 appears to be limited in its ability to replicate or establish latency in sympathetic neurons (28). HSV-1 and HSV-2 are both able to gain access to parasympathetic (5, 28, 35) and sensory nerve fibers. It is thus possible that the difference between lumbar and sacral quantities of HSV-1 and HSV-2 DNA could be related to different autonomic pathways for reaching the cord, potentially more likely to be sympathetic for HSV-1 and parasympathetic for HSV-2. During recurrences, HSV-2 viral DNA quantities increased in both the sacral cord and the DRG compared with levels present during latent infection (compare Fig. 6A to B), suggesting that HSV-2 viral replication may occur in both the sacral cord and the DRG during reactivation. HSV-1 exhibited increases in viral DNA in both the lumbar cord and the DRG during recurrences, suggesting that HSV-1 may also replicate in the lumbar spinal cord as well as the DRG during reactivation. However, it is unclear whether the increased DNA in the spinal cord during symptomatic recurrences is the result of viral spread from the DRG or reactivation originating in neurons of the spinal cord. Since genital reactivation of viruses with greater quantities of viral DNA in the sacral cord was more efficient, this also raises the possibility that some proportion of HSV reactivation may occur through autonomic pathways, which may provide further insight into type-specific differences in reactivation.
Substitution of HSV-1 LAT sequences into the HSV-2 LAT region did not eliminate the HSV-2 preference for the sacral spinal cord but did alter the quantity of DNA found in the sacral cord and the DRG during recurrences. The more virulent sequence chimera, HSV2-LAT-S1, had higher levels of viral DNA in the sacral cord and lower levels in the DRG than the wild-type HSV-2 during recurrences. Thus, the LAT sequence is not the major factor that permits HSV-2-based viruses to more efficiently reach the sacral spinal cord, but it does influence this process, potentially via a mechanism involving spread through or replication in autonomic neurons. The increased efficiency of HSV2-LAT-S1 in reaching and replicating in the sacral spinal cord further supports the conclusion that LAT region sequences are involved in neurotropism, or the ability of the virus to infect and replicate within specific neuronal subtypes.
Consistent with previous experiments, HSV-1 failed to reactivate efficiently in the guinea pig genital model. Wild-type HSV-2, HSV2-LAT-P1, and HSV2-LAT-S1-R reactivated efficiently, and these viruses also showed relatively higher DNA levels in the sacral spinal cord than HSV-1. This suggests that efficient replication in the sacral spinal cord and the ability to appropriately down-regulate viral replication are correlated with efficient genital reactivation during latency. Although the HSV2-LAT-S1 chimera efficiently reached the sacral spinal cord, its reduced-recurrence phenotype and high degree of virulence after reactivation may have been due to an inability to down-regulate the productive cycle upon reactivation, thus destroying its host instead of reestablishing a latent state from which it could again reactivate.
HSV-1 and HSV-2, while similar in many ways, exhibit significant differences in patterns of latency and reactivation, manifested clinically by site-specific recurrence patterns and differences in CNS manifestations. In the present study, we show that type-specific recurrence of HSV-2 depends on the presence of appropriate LAT region sequences downstream of the promoter. This same region of the LAT sequence appears to influence neurotropism within the CNS as well as the autonomic nervous system, which in turn influences reactivation as well as the pathogenesis of the virus, suggesting that some portion of the LAT phenotype may be due to an effect on viral replication and/or establishment of latency in different types of neurons.
We thank Kening Wang for help with Taqman PCR assays, and Lesley Pesnicak for assistance with the guinea pig model. We also thank Cara Olsen for help with statistical analyses.
Published ahead of print on 4 April 2007.