Specific histone tail modification and not DNA methylation is a determinant of herpes simplex virus type 1 latent gene expression

Abstract

During herpes simplex virus type 1 (HSV-1) latency, gene expression is tightly repressed except for the latency-associated transcript (LAT). The mechanistic basis for this repression is unknown, but its global nature suggests regulation by an epigenetic mechanism such as DNA methylation. Previous work demonstrated that latent HSV-1 genomes are not extensively methylated, but these studies lacked the resolution to examine methylation of individual CpGs that could repress transcription from individual promoters during latency. To address this point, we employed established models to predict genomic regions with the highest probability of being methylated and, using bisulfite sequencing, analyzed the methylation profiles of these regions. We found no significant methylation of latent DNA isolated from mouse dorsal root ganglia in any of the regions examined, including the ICP4 and LAT promoters. This analysis indicates that methylation is unlikely to play a major role in regulating HSV-1 latent gene expression. Subsequently we focused on differential histone modification as another epigenetic mechanism that could regulate latent transcription. Chromatin immunoprecipitation analysis of the latent HSV-1 DNA repeat regions demonstrated that a portion of the LAT region is associated with histone H3 acetylated at lysines 9 and 14, consistent with a euchromatic and nonrepressed structure. In contrast, the chromatin associated with the HSV-1 DNA polymerase gene located in the unique long segment was not enriched in H3 acetylated at lysines 9 and 14, suggesting a transcriptionally inactive structure. These data suggest that histone composition may be a major regulatory determinant of HSV latency.

FIG. 1.

Analysis of the HSV-1 genome for specific regions that are predicted to have a high probability of being targets of CpG methylation. The algorithm of Karlin et al. (20) was used (see Materials and Methods) to analyze the HSV-1 genome in 50-bp segments, and the resulting P values are indicated. Random DNA has a P value of 1, whereas P values of ≤0.8 represent a distribution of CpGs with a high probability of being methylated.

FIG. 2.

Regions of the HSV-1 genome selected for bisulfite sequence analysis for DNA methylation based on two models of predicting the potential for CpG methylation. Analyses of the HSV-1 genome indicated that discrete regions within the repeat regions are potential targets of CpG methylation. (A) Diagram of the HSV-1 genome with an expanded view of the internal R_L and R_S regions. The locations of the LAT, ICP0, γ34.5, and ICP4 genes are indicated. (B) P value analysis of the repeat region analyzed in 50-bp segments (see Materials and Methods). Regions containing P value ratios of ≤0.8 sustained over contiguous segments of >400 bp were considered to contain CpG frequencies making these regions probable targets of CpG methylation. Regions with significant P values that were selected for bisulfite analysis are indicated by the pink highlighting. (C) CpG/GpC ratios of the repeat region analyzed in 50-bp segments (see Materials and Methods). Regions containing CpG/GpC ratios of ≥1.5 sustained over contiguous segments of >400 bp were considered to comprise a CpG island and to be potential targets of CpG methylation. These regions encompassing promoters of genes were selected for bisulfite analysis and are indicated by the yellow highlighting. Note that the methods of analysis employed for panels B and C use divergent criteria and that only one region in common was predicted by both models (shown by the orange highlighting in panel A).

FIG. 3.

ChIP analysis of the LAT and DNA polymerase regions of latent HSV-1 DNA, using antiserum specific for anti-acetyl histone H3(K9, K14). DRG from mice latently infected with HSV-1 strain KOS were processed and subjected to ChIP analysis as described in Materials and Methods. The relative enrichment of acetyl histone H3(K9, K14) on a specific region of DNA was determined by PCR analysis of the ChIP fraction relative to dilutions of the input (preimmunoprecipitation) material. (A) ChIPs were validated by performing PCRs with primers to cellular target APRT (actively transcribed in peripheral nerve ganglia) and β-globin (not transcribed in adult DRG) genes. (B) PCRs performed on the same ChIP with primers to the LAT promoter and the HSV-1 DNA polymerase gene. Fold enrichments (which adjust for differences in PCR efficiencies between primer sets) are determined by comparing the band intensities of PCR products generated with ChIP-precipitated DNA and twofold serial dilutions of input (see Results). Lane 6 in panel A and lane 5 in panel B are no-input controls.

FIG. 4.

Precipitation efficiency of the LAT promoter and the HSV-1 polymerase gene acetyl histone H3(K9, K14) ChIPs. (A) Schematic of ChIP efficiency determination, showing fractions of isolated tissue used in the ChIPs with the specific and control antisera and the dilutions of the unbound control fraction used to calculate the IP efficiency. H3Ac, acetyl histone H3(K9, K14). (B) Experimental determination of the relative enrichment of the LAT promoter over the HSV polymerase gene in the acetyl histone H3(K9, K14) ChIP. ChIP 1×, precipitation with 2.25 μg of antiserum; ChIP 2.5×, precipitation with 5 μg of antiserum.

FIG. 5.

Comparison of the histone H3 acetylation status of the LAT promoter with those of two immediate-early genes. PCRs were performed on ChIP products of latently infected mouse DRG with primers specific to the LAT promoter and the α27 (UL54) (A) and α4 (ICP4) (B) promoters. The input lanes represent 3- and 1.5-fold serial dilutions (respectively) of the unbound control fraction. Lanes −, no-input control.