Recruitment of the Transcriptional Coactivator HCF-1 to Viral Immediate-Early Promoters during Initiation of Reactivation from Latency of Herpes Simplex Virus Type 1^{▿ †}

Recruitment of HCF-1 to IE promoters upon reactivation ex vivo.

To further investigate the role of HCF-1 in HSV-1 reactivation, chromatin immunoprecipitations (ChIP) were used to assess HCF-1 occupancy at viral IE promoters during reactivation in neurons of trigeminal ganglia (TG). TG from latently infected mice were removed and immediately processed (0 h) or were explanted into tissue culture for 4 h to stimulate efficient viral reactivation prior to processing as described in the supplemental material. Aliquots of isolated chromatin corresponding to 10 latently infected ganglia were subjected to ChIP using control immunoglobulin G (IgG) or HCF-1 antibodies (Fig. 1A). Due to the limited amount of material and the low frequency of viral reactivation, a nested PCR method was developed to increase the sensitivity of the assay (see the supplemental material). As shown in Fig. 1A, no HCF-1 could be detected at the ICP0 enhancer domain at 0 h (lane 7) but the protein was readily detected after explant for 4 h (lane 9).

In sensory neurons of unstimulated TG, HCF-1 is sequestered in the cytoplasm. Upon stimulation, the percentage of neurons that exhibit nuclear HCF-1 localization increases over the course of 6 h poststimulation (11, 13). To determine whether the occupancy of HCF-1 could be detected at earlier times postinduction of reactivation, the ChIP assays were done using TG explanted for 0, 1, or 4 h. As shown in Fig. 1B and C, HCF-1 occupancy was reproducibly detected as early as 1 h postexplantation and occupancy increased significantly by 4 h at both the ICP0 and ICP27 IE gene enhancer domains. In contrast, no significant occupancy was detected within control regions (ICP27 coding region, UL45 promoter, and glyceraldehyde-3-phosphate dehydrogenase [GAPDH] promoter) (Fig. 1C). Thus, specific HCF-1 occupancy of the viral IE promoter correlates with the previously described time course of transport and nuclear localization of HCF-1.

HCF-1 recruitment correlates with RNAPII occupancy and induction of HSV-1 IE gene transcription.

The recruitment of HCF-1 to the viral IE gene enhancer-promoter domain during the initial stages of reactivation is anticipated to correlate with the recruitment of RNA polymerase II (RNAPII) and the accumulation of IE mRNAs. As shown in Fig. 2, RNAPII occupancy of the ICP0 promoter was not significant relative to that of the IgG control or that of the promoters of the nontranscribed ANT4 gene or the viral late UL45 gene at 0 h poststimulus but was detected at the promoter of the transcribed GAPDH gene. At 4 h poststimulus, RNAPII occupancy of the viral IE promoter was clearly evident. Additionally, a low-level occupancy was reproducibly detected within the coding sequences (Fig. 2, ICP0-Cd). No change in the status of occupancy of the control UL45, ANT4, or GAPDH promoter was detected.

To confirm that HCF-1 and RNAPII occupancy correlates with transcriptional activation and accumulation of viral IE mRNAs, cDNAs were produced from total RNAs isolated from latently infected TG at 0, 2, and 4 h after reactivation explant. The levels of viral (latency-associated transcript [LAT], ICP0, and ICP27) and cellular control (S15, Sp1, and TATA-binding protein [TBP]) mRNAs were determined by nested PCR or quantitative PCR (qPCR). The LAT RNA and cellular control mRNAs were readily detected at zero time (Fig. 3B, C, and F). In contrast, ICP0 and ICP27 mRNAs were not detected at zero time but were detected at 2 h postexplant, and their relative abundance increased by 4 h postexplant (Fig. 3B, D, and E). The cellular controls S15, Sp1, and TBP showed no significant variance between TG harvested at 0, 2, and 4 h postexplant (Fig. 3F).

HCF-1 occupancy localizes to IE distal promoter-enhancer domains.

The promoter-enhancer domains of the HSV-1 IE genes are complex. These regions contain reiterated enhancer core elements that nucleate the assembly of the Oct-1/VP16/HCF-1 enhancer complex during initiation of lytic infection (Fig. 4A). In addition, these domains contain binding sites for other cellular transcription factors, such as GABP, Sp1, and CREB3, that interact directly with HCF-1 and are capable of recruiting the coactivator (6, 8, 12, 16, 17, 29).

To gain insight into the component(s) that might be involved in the recruitment of HCF-1 to viral IE genes during reactivation, the occupancy of HCF-1 was localized using primer sets across the ICP0 promoter-enhancer domain. In these experiments, isolated chromatin from 4-h-induced TG was sonicated to approximately 300 bp to increase the resolution of the ChIP. As shown in Fig. 4B, immunoprecipitation with anti-HCF-1 antibodies resulted in the recovery of IE promoter fragments nearly exclusively representing the distal promoter-enhancer region (Fig. 4A). In contrast, no significant levels of fragments representing any other region were detected. The results indicate that stable HCF-1 association is dependent upon factors recognizing elements within the distal enhancer region, a region that minimally contains binding sites for GABP and Sp1 in addition to two enhancer core elements.

HCF-1, a component of the regulatory mechanisms governing HSV-1 reactivation.

HCF-1 is an essential component of the HSV-1 IE regulatory paradigm. The protein is recruited to viral IE promoter-enhancer domains via complex interactions with multiple factors, including viral IE activators, that contribute to the induced expression of these genes during initiation of viral lytic infection. HCF-1 functions, at least in part, by coordination of chromatin modification complexes, such as the Set1/MLL1 histone methyltransferase and the histone demethylase LSD1, to prevent the accumulation of repressive chromatin and promote the installation of positive chromatin marks.

In addition to its role in lytic infection, chromatin plays a significant role in the regulation of HSV-1 latency and reactivation cycles. Activating marks are evident on the transcribed LAT gene while repressive marks are prevalent on the repressed IE gene promoter regions (1, 14, 15, 21, 30). Conversely, activating marks accumulate on the viral IE gene promoter domains during the initial stages of reactivation (1, 21). Strikingly, HCF-1 is sequestered in the cytoplasm of sensory neurons, where the virus establishes latency. The protein is rapidly relocalized to the nucleus upon in vivo and ex vivo stimulation, suggesting that HCF-1 may also play a critical role in reactivation of HSV-1 from latency. However, a direct involvement in viral reactivation has not yet been demonstrated.

In this study, a ChIP approach was used to demonstrate that HCF-1 is recruited to HSV-1 IE gene promoter-enhancer domains upon stimulation of reactivation of HSV-1 from latency in sensory neurons. Recruitment was detected as early as 1 h postinduction of reactivation, and occupancy increased by 4 h postinduction. This time course correlates well with the nuclear transport of HCF-1, where the number of neurons exhibiting nuclear localization increases over the course of 6 h poststimulation. Additionally, HCF-1 occupancy of viral IE promoter domains correlated with occupancy by RNAPII and accumulation of viral IE mRNAs. This study represents the first report of the direct recruitment of a critical component of the IE gene regulatory circuit during viral reactivation.

HCF-1 is not a direct DNA binding component, and its recruitment is dependent upon interactions with site-specific transcription factors. The protein interacts with members of numerous transcription factor families, including several activators (GABP, Sp1, and VP16) that contribute to the induced transcription of viral IE genes during lytic infection. Occupancy of the IE promoter-enhancer domain by HCF-1 during reactivation localizes to the distal promoter region of the enhancer domain. This region contains several defined factor binding sites, including two TAATGARAT enhancer core elements, a consensus GABP, and several Sp1 sites. In contrast, no significant association of HCF-1 with the promoter-proximal enhancer core element or the putative CREB3 site was detected, a factor that was proposed to be involved in HCF-1 recruitment during reactivation (16). While this suggests that the factor(s) responsible for HCF-1 recruitment is localized distal to the promoter domain, this may also be a reflection of multiple interactions that result in a highly stable association at this region. It is important to note that due to the limitations of the ChIP assay, the data do not exclude the possibility that HCF-1 is transiently or nonstably recruited via factors whose recognition sites are within other regions of the IE enhancer-promoter domains.

It is also possible that the factors that recruit HCF-1 could depend upon the specific reactivation stimuli. Of note, a recent study has indicated that the viral IE activator VP16, a protein whose function is dependent upon HCF-1, may be critical for viral IE transcription and reactivation (27). The recruitment of HCF-1 to the region containing two enhancer core elements is consistent with this model. However, it remains to be determined which factors are directly responsible for HCF-1 recruitment and whether those factors reflect distinct signaling pathways. Regardless, the demonstration of the direct recruitment of HCF-1 to viral IE genes during the initial stages of reactivation promotes the model that HCF-1 complexes play a critical role in the viral reactivation process.

Supplementary Material