In Vivo Changes in the Patterns of Chromatin Structure Associated with the Latent Herpes Simplex Virus Type 1 Genome in Mouse Trigeminal Ganglia Can Be Detected at Early Times after Butyrate Treatment

Abstract

During herpes simplex virus type 1 (HSV-1) latency in mouse dorsal root ganglia (DRG), chromatin associated with the latency-associated transcript (LAT) region of the viral genome is hyperacetylated at lysines 9 and 14 of histone 3 [H3(K9, K14)], while lytic genes are hypoacetylated. Explanted DRG exhibit a pattern of deacetylation of the LAT enhancer followed by acetylation of the ICP0 promoter at early times postexplant. Recently, we reported that sodium butyrate induced in vivo reactivation of HSV-1 in latent mice. In this study, we assessed the effect of sodium butyrate on the chromatin patterns of latent and butyrate-treated mouse trigeminal ganglia (TG) via chromatin immunoprecipitation (ChIP). We detected deacetylation of acetyl H3(K9, K14) of the LAT promoter and LAT enhancer regions as early as 0.5 h post-butyrate treatment, and this deacetylation corresponded to an increase in the acetylation of the lytic promoters ICP0 and ICP4 at 0.5 h and 1 h post-butyrate treatment, respectively. This is the first study to combine in vivo reactivation with the examination of the HSV-1 genome through ChIP assays at early times after the introduction of in vivo reactivation stimuli.

During latency, the herpes simplex virus type 1 (HSV-1) genome exists as a circular episome associated with histones (5, 14, 19), and the only abundant transcription that occurs is that of the latency-associated transcript (LAT). LAT has been implicated in numerous viral functions which are currently under investigation (4, 16-18, 20, 21). Furthermore, the structure and function of the HSV-1 genome is being actively explored. As early as 1989, Deshmane and Fraser first demonstrated that during latency, the HSV-1 genome is associated with nucleosomes in a chromatin-like structure similar to that of cellular chromatin (5).

Cellular chromatin consists of actively transcribed permissive regions (euchromatin) and repressed nonpermissive regions (heterochromatin) with respect to RNA polymerase II-mediated transcription (7, 11, 23). The histone composition of euchromatin and heterochromatin varies depending, in part, on the nature of the specific histone tail modifications. For example, posttranslational histone hyperacetylation at lysines 9 and 14 of histone 3 [H3(K9, K14)] is indicative of transcriptional permissiveness, and underenrichment of this modification is usually associated with transcriptional repression. In vitro studies examining the viral genome during the acute HSV-1 infection show that H3 is associated with HSV-1 DNA in the initial infection and further suggest that productive infections maintain covalent histone modifications representative of transcribed cellular genes (6, 8). More recently, Kubat et al. (9, 10) used mouse dorsal root ganglia (DRG) latently infected with HSV-1 strain KOS to demonstrate that during latency the LAT promoter and the LAT 5′ exon (enhancer) regions of the HSV-1 genome are significantly enriched in acetyl H3(K9, K14), whereas the lytic promoters ICP0 and ICP4 are underenriched in this modification. Furthermore, Wang et al. (22) demonstrated that lytic genes are associated with repressive marks (H3 K9 dimethylated histones) during latency. Collectively, these data indicate that at the chromatin level, the HSV-1 genome can be segregated into distinct transcriptional domains, analogous to the heterochromatic and euchromatic domains observed in cellular chromatin, and that these domains are separated by insulator elements that may act as chromatin boundaries (2). The association of specific histone modifications with the transcriptional status of HSV-1 genes during the lytic infection and latency strongly suggests that chromatin modulatory enzymes, such as histone deacetylases, could play key roles in the regulation of HSV-1 gene expression. To illustrate this point, a recent study performed by Amelio et al. (1) reported that explant-induced reactivation results in a rapid deacetylation of the LAT enhancer, followed by a dramatic drop in LAT RNA, and the subsequent remodeling of the ICP0 promoter to an acetylated state.

The objective of our current study was to further explore the possibility that chromatin remodeling occurs early in in vivo reactivation, subsequent to the introduction of reactivation stimuli. Specifically, we sought to determine if deacetylation of the LAT region is specific only to explanted tissue or is an initial step in in vivo reactivation and whether changes in the acetylation profile of chromatin associated with the immediate early lytic promoters ICP0 and ICP4 can be detected in the initial stages of in vivo reactivation. In order to explore these questions, we used a rapid in vivo method of HSV-1 reactivation which utilizes the histone deacetylase inhibitor sodium butyrate (NaB). We previously reported that intraperitoneal (i.p.) injections of NaB yielded reactivation frequencies of >75% in the mouse trigeminal ganglia (TG) model (15). To our knowledge, this is the first report that combines a systemic in vivo reactivation method with exploration of the chromatin structures associated with HSV-1 soon after the introduction of the reactivation stimuli in the mouse.

Four- to five-week-old female BALB/c mice (Charles River Laboratories Inc., Wilmington, MA) ocularly infected with HSV-1 strain 17syn+ were used in all studies presented. Viral inoculations and slit lamp examinations (SLE) were performed as previously reported (13). Briefly, viral inoculation titers of 100,000 PFU/eye were used on eyes scarified with a 3 by 3 cross-hatch pattern. Infections were confirmed through SLE of the corneas, and mice were considered latent at 28 days postinfection at which time no lesions were apparent by SLE. Latent mice were separated into groups of four, and NaB/phosphate-buffered saline (1,200 mg/kg of body weight in a 100-μl dose) was administered i.p. as previously reported (15). Controls were latent mice sacrificed without NaB treatment (0 h), and experimental mice were sacrificed at 0.5 h, 1 h, or 2 h post-butyrate treatment. All chromatin immunoprecipitation (ChIP) assays (Upstate Biotechnology, Charlottesville, VA) were performed using anti-acetyl H3 (Upstate catalog no. 06-599) as described by Kubat et al. (9) with minor modifications made to accommodate the TG. Validations of the efficiencies of ChIP assays were performed as previously described. Primer and probe sequences used for viral and host genes were identical to those reported by Kubat et al. (9). A standard curve was generated using 10-fold serial dilutions of the purified 17syn+ virus to control for differences in primer sensitivities. All assays were normalized to the host gene APRT, a constitutively expressed cellular gene, for comparison between different experimental groups of mice.

ChIP analysis of TG from mice latently infected with HSV-1 strain 17syn+ (0 h) demonstrated that the LAT promoter and enhancer regions (5′ exon) maintained a hyperacetylated state relative to the lytic promoters of ICP0 and ICP4 (Fig. 1). The relative bound/input ratios calculated from the real-time PCR and normalized to cellular controls for the mouse TG showed that the LAT enhancer and promoter regions of 17syn+ displayed no significant difference in the n-fold enrichment between these two regions (P = 0.201; Fig. 1). Our results are consistent with previously published results regarding hyperacetylation of LAT regions relative to the lytic promoters ICP0 and ICP4, from experiments that utilized the DRG latently infected with HSV-1 strain KOS (9, 10).

FIG. 1.

Histone H3(K9, K14) acetylation of 17syn+ during latency: analysis of three independent ChIP assays using mouse TG latent with 17syn+. ChIP analyses were performed for the 0-h time point (latency). TG from four mice per assay were pooled. All samples were precipitated with anti-acetyl H3(K9, K14) and analyzed by TaqMan PCR with previously reported primers and probes specific for the regions analyzed (9). Bound/input ratios were normalized to the cellular control, APRT. Average bound/input ratios from the three independent assays for each gene region analyzed are represented by a horizontal bar. The LAT promoter and LAT enhancer (5′ exon) regions were significantly enriched in acetyl H3, relative to the lytic promoters ICP0 and ICP4 (P of <0.004 and <0.001, respectively).

One objective of our current study was to establish whether rapid deacetylation of the LAT promoter and LAT enhancer regions of the 17syn+ HSV-1 genome occurred after the introduction of an in vivo reactivation stimulus (in our case, NaB). Mice latently infected with 17syn+ were treated with NaB (1200 mg/kg, i.p.) and sacrificed at 0.5 h, 1 h, and 2 h posttreatment, and the harvested TG were subjected to ChIP analyses using the anti-acetyl H3(K9, K14) antibody. Subsequent real-time PCR analyses of the ChIP fractions indicated a significant reduction in acetyl H3(K9, K14) enrichments of both the LAT promoter and LAT enhancer regions over a 2-h period of time after the reactivation stimulus, NaB, was introduced, with the most dramatic decrease in acetylation occurring by 1 h post-NaB treatment (Fig. 2, panels 1 and 2). It is important to note that while the data obtained indicate that acetylation of the LAT promoter region in experiment 2 increased between the 0-h and 0.5-h NaB treatment times, by 1 h an overall decrease in the acetylation enrichment of this region could clearly be observed (Fig. 2B, panel 1). In contrast, the deacetylation of the enhancer (LAT 5′ exon) region among the three independent experiments was consistent over 1 h (Fig. 2, panel 2). Interestingly, by 2 h post-NaB treatment, the acetylation profiles of the LAT enhancer and promoter regions in two of three experiments increased relative to the 1-h time point (Fig. 2A and B, panel 1), perhaps indicating that the LAT region of 17syn+ was reestablishing its latent-state acetyl enrichment profile; further studies are under way to establish a kinetic pattern associated with this observation. Nevertheless, the data indicate that both the LAT promoter and enhancer regions of HSV-1 strain 17syn+ underwent deacetylation very early after the in vivo reactivation stimulus was introduced, a finding consistent with reported data obtained in explants of HSV-1 strain KOS (1).

FIG. 2.

Mice were given an i.p. injection of NaB (1,200 mg/kg) and sacrificed at 0.5, 1, or 2 h postinjection. All changes were calculated as a percent change relative to the 0-h (latent) time point, which was set to 100%. Subsequent real-time PCR revealed that the LAT promoter (panel 1) and the LAT enhancer (panel 2) regions underwent rapid deacetylation as early at 0.5 h post-NaB treatment. This decrease corresponded to a 3- to 10-fold increase in the acetylation of the ICP0 promoter (panel 3) at 0.5 h post-NaB treatment. Subsequent analysis of the ICP4 promoter (panel 4) showed an increase in the acetylation of this promoter, with the peak at 1 h post-NaB treatment. All samples in each experiment were analyzed in triplicate for real-time PCR. Four mice per ChIP assay (total of eight TG) were used in all three experiments, and all samples were normalized to the cellular control, APRT. B/I, bound/input.

To determine if deacetylation of the LAT region could be correlated to an increase in acetyl H3(K9, K14) enrichment of ICP0 and/or ICP4, the ICP0 and ICP4 promoter regions of 17syn+ were analyzed at 0 h and at 0.5 h, 1 h, and 2 h post-NaB treatment using the same three independent ChIP assays. We observed a 3- to 10-fold increase in the acetylation enrichment of the ICP0 promoter at 0.5 h post-NaB treatment compared to the level with the control (Fig. 2, panel 3). Interestingly, in all three experiments, we observed deacetylation of the ICP0 promoter by 1 h post-NaB treatment. Furthermore, an increase of acetyl H3(K9, K14) enrichment at the ICP4 promoter relative to the control could also be observed in all three experiments at 0.5 h and 1 h post-NaB treatment (Fig. 2, panel 4), consistent with the observed deacetylation pattern of the ICP0 promoter during this time frame (Fig. 2, panel 3). It is important to note that, while in experiment 3 the peak acetyl H3(K9, K14) enrichment for the ICP4 promoter was seen at 2 h post-NaB treatment (Fig. 2C, panel 4), the data are still consistent with the observed deacetylation pattern of the ICP0 promoter (Fig. 2C, panel 3).

The average n-fold differences from the three ChIP assays indicate that by 0.5 h post-NaB treatment, the ICP0 and ICP4 promoter regions of 17syn+ were no longer hypoacetylated relative to the LAT promoter (average ratio is 1:1), while the enhancer region ratio approaches 3:1 (enhancer:lytic promoters; Fig. 3A and B). Furthermore, by 1 h post-NaB treatment, the ICP4 promoter region acetyl H3(K9, K14) n-fold enrichment was an average of two times higher than acetyl enrichment of the LAT promoter and enhancer regions of HSV-1 17syn+ (Fig. 3C). The HSV-1 17syn+ acetyl H3(K9, K14) enrichment profile at the 2-h post-NaB treatment time indicated that reestablishment of a latent-like acetyl H3 profile may have been occurring (Fig. 3D). Studies are currently under way to determine if these increases in acetylation of ICP0 and ICP4 correspond to increased transcription at early times, as well as at 12 and 24 h post-NaB treatment.

FIG. 3.

The average percent n-fold difference in the four HSV-1 regions analyzed for all three independent assays at four different time points. Percent n-fold differences were calculated relative to the LAT promoter region, which was normalized to 100%. The LAT promoter region is significantly enriched in acetyl H3(K9, K14) relative to the ICP0 and ICP4 promoter regions (P of <0.009 and <0.01, respectively) during latency (0 h). In contrast, no significant enrichment of the LAT promoter region compared to the ICP0 and ICP4 promoter regions is observed at 0.5 h (P = 0.366, P = 0.388), 1 h (P = 0.394, P = 0.183), and 2 h (P = 0.126, P = 0.112) post-NaB treatment.

Our results showed that there was a significant change in the enrichment profiles of acetyl H3(K9, K14) associated with the LAT promoter and enhancer regions of HSV-1 and the lytic promoters ICP0 and ICP4 in the HSV-1 high phenotypic reactivator 17syn+ at very early times after the in vivo reactivation stimulus NaB was introduced. Our studies corroborate the finding that posttranslational chromatin hyperacetylation of the LAT region (relative to lytic promoters ICP0 and ICP4) of the HSV-1 genome can be correlated with the maintenance of latency (22). Consistent with previously published reports (9), we found hyperacetylation of the LAT region relative to ICP0 and ICP4 during latency. Our studies also showed that disruption of this hyperacetylation status of the LAT region, using the previously described reactivation stimulus NaB (15), can also be observed at very early times after NaB treatment. The same deacetylation pattern of the LAT region has been previously reported in explanted ganglia subjected to ChIP analysis (with a peak at 1 h postexplant) (1). The deacetylation of the LAT region after exposure to reactivation stimuli in vivo appears to be a more-rapid process, where the peak deacetylation of the LAT region after administration of NaB is 0.5 h. This corresponds to increased acetylation found in both the ICP0 and ICP4 promoter regions at 1 h post-NaB treatment (Fig. 3). In contrast, ChIP analysis of explanted ganglia revealed no increases in ICP0 promoter acetylation until 4 h postexplant (1).

To assess whether the changes observed in the acetylation profile of the LAT region could be correlated to a decrease in the LAT transcription, RNA was isolated from the TG of mice latently infected with HSV-1 17syn+ at 0, 1, 2, 3, and 4 h post-NaB administration. The RNA was reverse transcribed using random decamers, and the cDNA was analyzed by real-time PCR. Relative quantities of LAT were normalized to host APRT. A significant (P < 0.05) decrease in the LAT accumulation was observed at 1 h post-NaB treatment (Fig. 4) and corresponds to the deacetylation patterns observed for this region at 1 h post-NaB treatment. In contrast, we detected no significant changes in the accumulation of ICP0 at any times through 4 h post-NaB treatment (data not shown). A more-extensive kinetic study of other lytic transcriptions during in vivo reactivation is currently ongoing.

FIG. 4.

A significant decrease (P < 0.05) in the LAT 5′ exon RNA levels is observed at 1 h post-NaB injection. RNA was isolated from mice latently infected with HSV-1 strain 17syn+. Four mice were used for each time point, and the eight TG from the four mice were pooled prior to RNA extraction. cDNA was analyzed by real-time PCR in triplicate using primers and a probe specific for LAT 5′ exon (nucleotides 119326 to 119397) (1). Relative quantities of LAT RNA were normalized to host APRT, and the results of three independent experiments are shown in the graph as percent LAT 5′ exon relative to the 0-h time point.

The LAT 5′ exon (a region of LAT containing enhancer elements) has been suggested to function as a cis-acting DNA element, recruiting proteins, transferases, and modulatory enzymes necessary for the mediation of transcription (2, 3, 6, 12). Numerous studies have suggested that the latent HSV-1 episome resembles cellular chromatin, having permissive and nonpermissive HSV-1 genome regions (6, 9, 10, 22). This model includes the segregation of heterochromatic and euchromatic regions of the HSV-1 genome by insulator elements likely to exist between the U_L and R_L segments and the splice donor of LAT intron and ICP0 (9). Results from the ChIP analyses in this study, using a known reactivation stimulus, indicate that chromatin structures associated with early immediate lytic regions (ICP0 and ICP4) of the viral genome undergo posttranslational modifications (acetylation) as early as 0.5 h after the reactivation stimulus is introduced systemically. Given these data, these modifications could allow the ICP0 and ICP4 regions of HSV-1 to undergo a transition from a repressed state to an open, more transcriptionally permissive state. In such a circumstance, one would expect to find increases in transcriptionally permissive modifications, such as acetylation of the lytic regions ICP0 or ICP4 at early times of induction with respect to the chromatin modifications observed during latency. Considering the results from Amelio et al. (2), suggesting the existence of insulator elements within the HSV-1 genome similar to those found in cellular chromatin, HSV-1 insulators could function in a way comparable to that of cellular chromatin and block enhancer activities at inappropriate times (11, 23). This structure-activity relationship would ensure that the segregation between heterochromatic and euchromatic regions of HSV-1 is maintained in normal biological conditions. Administration of NaB may disrupt this maintenance of insulator function by allowing hyperacetylation of the H3(K9, K14) associated with these lytic regions, which would, in turn, open these genomic regions to transcription.

Alternatively, the LAT region of the HSV-1 genome region may undergo rapid deacetylation resulting from an external stressor (in this case NaB) which disables the insulator region thought to exist between the splice intron and ICP0 regions (1). Enhancer activity that overcomes this insulator element could initiate a cascade effect for other posttranslational histone modifications along the transcriptionally silent region of the genome resulting in the transcription of lytic genes and thus viral reactivation. Real-time PCR from ChIP assays at various times post-NaB treatment indicate that chromatin remodeling of the LAT locus and the ICP0 region are simultaneous events, followed by the change in the chromatin patterns observed for ICP4, and that these molecular events precede clinical reactivation.

Determining the underlying mechanisms and regulators involved in HSV reactivation requires further exploration of the chromatin modifications involved in the viral progression from latency to reactivation. There is a significant difference in the acetylation profile of H3 associated with viral gene promoters during latency and during induced reactivation of the HSV-1 17syn+. Our data suggest that deacetylation of the LAT enhancer could play a key role in the virus progression from latency to NaB-induced viral reactivation by strains of HSV-1 that are high phenotypic reactivators. Given this, the LAT enhancer region of HSV-1 could recruit chromatin-modifying functions that maintain the expression of LAT during latency. To understand the epigenetic factors that are involved in latency and reactivation, more-extensive studies involving other chromatin-modulating factors are needed and experiments are currently under way to resolve these questions.