The coactivator host cell factor-1 mediates Set1 and MLL1 H3K4 trimethylation at herpesvirus immediate early promoters for initiation of infection

HCF-1-Dependent Coactivation of IE62 Target Genes.

Previously it was demonstrated that HCF-1 was an essential component required for the combinatorial regulation of the expression of the HSV and VZV α-herpesvirus IE genes by the respective viral transactivators VP16, ORF10, and IE62 (8). As illustrated in Fig. 1A, the VZV IE62 promoter contains enhancer core elements that assemble the multiprotein Oct-1/HCF-1/ORF10 complex as well as sites for costimulatory factors such as GABP and Sp1 (25). In addition, this promoter contains an IE62 binding site that responds to the viral transactivator (26, 27). As previously demonstrated, IE62 stimulated the expression of an IE62 promoter-reporter gene and this induction was severely compromised in HCF-1-depleted cells (Fig. 1B).

Similarly, an IE62 promoter-reporter construct containing only the IE62 binding site, a TATA element, and an adjacent Sp1 binding site was induced by IE62 in an HCF-1-dependent manner. In contrast, deletion of the TATA proximal Sp1 binding site abrogated IE62-mediated induction, thereby defining the minimal promoter requirements for IE62-mediated activation and HCF-1 coactivation. As shown in supporting information (SI) Fig. 8, the requirement for Sp1 for IE62 activation of the IEP-61 promoter may be explained by the lack of stable IE62 binding in the absence of the Sp1 binding site.

In addition to its autoregulatory function, IE62 also induces transcription of the early and late VZV gene classes (28). As shown in SI Fig. 9, promoter–reporter constructs representing early and late VZV gene classes were induced in the presence of IE62. In each case, depletion of HCF-1 resulted in a significant reduction in the IE62-dependent induction, indicating that the requirement for HCF-1 in IE62 transcriptional activation is a general one and is not restricted to the regulation of the IE genes.

IE62 Recruitment of the Coactivator HCF-1 to the IE62 Minimal Model Promoter.

The HCF-1 dependence of IE62-mediated activation suggested that IE62 might directly recruit HCF-1 to the minimally responsive promoter target gene. Therefore, HCF-1 promoter occupancy was determined by ChIP assays in the presence and absence of IE62. As shown in Fig. 2, both cotransfected V5-tagged HCF-1 (Fig. 2 Upper) and endogenous HCF-1 (Fig. 2 Lower) were recruited to the IE62P-61 promoter in the presence of IE62. Although a low level of endogenous HCF-1 could be detected in the absence of IE62, this occupancy was strongly stimulated in the presence of IE62 (8.5-fold).

IE62-Mediated Assembly of the Basal Transcription Complex Is Not HCF-1-Dependent.

To investigate the biochemical mechanism(s) by which IE62/HCF-1 stimulates transcription, ChIP assays were done to assess the promoter occupancy of basal transcription factors in the presence and absence of IE62 or HCF-1. As shown in Fig. 3A, in the absence of IE62, low levels of Sp1 and the basal factors TBP, TFIIB, TFIIH, and RNAPII could be detected. However, a strong stimulation of promoter occupancy of all factors was found in the presence of IE62. In particular, stimulation of Sp1 occupancy in the presence of IE62 underscores the cooperative DNA binding suggested previously (SI Fig. 8) (29).

To determine whether HCF-1 was required for the assembly of the RNAPII transcription complex, occupancy was also assessed in HCF-1-positive and HCF-1-depleted cells (Fig. 3B). Strikingly, the depletion of HCF-1 had no effect on the recruitment of the RNAPII complex to the minimal promoter, indicating that the IE62-Sp1 activators were solely responsible for this assembled but transcriptionally inactive complex.

Transcriptional Induction by IE62 Results in HCF-1-Dependent Set1- and MLL1-Mediated Trimethylation of Histone H3K4.

Recently, HCF-1 was identified as a component of the MLL family of HMTs (22, 23). As HCF-1 is critical for IE62-mediated transcriptional induction yet it is not required for recruitment/assembly of the RNAPII complex, the state of selected chromatin modifications was determined in HCF-1-positive and HCF-1-depleted cells (Fig. 4A). In the absence of IE62, the IE62 responsive promoter was marked by the repressive dimethyl-H3K9. However, in the presence of IE62, the activating trimethyl-H3K4 was significantly enhanced, whereas dimethyl-H3K9 was significantly reduced. In addition, IE62 specifically resulted in significant recruitment of the H3K4 HMTs Set1 and MLL1, relative to the HMT Set7/9. Interestingly, IE62-mediated occupancy by MLL1 increased only 4-fold. In contrast, IE62 resulted in a 30-fold increase in the promoter occupancy by Set1, suggesting that IE62 preferentially recruits this H3K4 HMT.

Most significantly, in the absence of HCF-1 (Fig. 4B), the recruitment of Set1 and MLL1 and the resulting H3K4 trimethylation were all severely reduced (9.5%, 18.4%, and 5.7% of HCF-1+, respectively). This reduction was not due to reduced levels of the critical protein components in HCF-1-depleted cells as demonstrated by quantitative Western blot analysis (SI Fig. 10). Thus, recruitment of the coactivator HCF-1 results in Set1- and MLL1-dependent H3K4 trimethylation.

Sp1 Is Required to Bridge/Stabilize the IE62-HCF-1 Interaction.

Sp1 is clearly an important component of the IE62-HCF-1-mediated transcriptional induction as shown by the removal of the TATA proximal Sp1 site, which abrogates stable IE62 binding (SI Fig. 8) and by the increased promoter occupancy by Sp1 in the presence of IE62 (Fig. 3). Interestingly, both IE62 and HCF-1 interact directly with Sp1 (12, 29), leading to the possibility that, in addition to cooperative DNA binding interactions with IE62, Sp1 may also function to mediate or stabilize the interaction of IE62 with HCF-1. As shown in Fig. 5A, both endogenous HCF-1 and Sp1 were coimmunoprecipitated with IE62 (lane 4). However, RNAi-mediated depletion of Sp1 reduced the coimmunoprecipitation of HCF-1 to background levels (Fig. 5B, compare lanes 4 and 5 with lane 6). Thus, Sp1 is required to bridge or stabilize the interaction of the activator IE62 with its coactivator HCF-1.

Recruitment of HCF-1 and Chromatin Modifications at the VZV IE62 Promoter During VZV Lytic Infection.

The data presented here using a model IE62-responsive promoter argue that HCF-1-mediated modifications would be represented at the viral IE promoter during the early stages of infection. Therefore, cells were infected with VZV, and 4 h after infection, the occupancy of the viral IE62 promoter by IE62, Sp1, HCF-1, Set1/MLL1 HMTs, and trimethyl-H3K4 was assessed. As shown in Fig. 6, results comparable with those obtained by using the transfected model promoter were found at the viral genomic IE promoter, including clear occupancy of IE62, Sp1, HCF-1, Set1, and MLL1. As anticipated, based on the HCF-1 and Set/MLL occupancy, the promoter exhibited a strong trimethyl-H3K4 signal. In contrast, the IE62 coding region did not reveal any significant HCF-1 occupancy or Set1/MLL1-mediated H3K4 trimethylation.

The requirement of HCF-1 in mediating these modifications during a viral infection was similarly addressed by infection of a cell culture that had been partially depleted of HCF-1. In these cells, a 52% HCF-1 depletion resulted in similar decreases in HCF-1, Set1, MLL1, and trimethyl-H3K4 signals at the viral IE promoter, whereas no significant change in the promoter occupancy by the viral IE62 protein was found (data not shown). The results indicate that HCF-1-mediated chromatin modulation of the IE gene promoter domain is likely to play a critical role in the initiation of the viral infectious cycle.

Cell Culture and Virus.

HeLa and BS-C-1 cells were maintained according to standard procedures. VZV (Ellen) was obtained from American Type Culture Collection (Manassas, VA). VZV viral infections were done by overlaying cell-associated virus on naive BS-C-1 cells. Cells were harvested for ChIP assays 4 h after infection.

Reporter Assays.

Luciferase reporter genes contained ORF29, ORF28, and gI VZV promoter sequences cloned into pGL3-basic (SI Fig. 9). IE62P-61 and IE62P-39 were derived from IE62PR that contained IE62 promoter sequences from −269 to +73. pCMV-IE62 expresses the IE62 activator under control of the CMV IE promoter and has been previously described (8). For all reporter assays, 4 × 10⁴ HeLa cells were transfected with 300 ng of pU6-Si-HCF-1 or pU6-Si control RNAi using FuGENE (Roche Diagnostics Corporation) according to the manufacturer's recommendations. At 48 h after transfection, the cells were cotransfected with the reporter constructs and increasing amounts of pCMV-IE62, or control vector. The reporter gene activity was measured 24 h later by using a Dual-Luciferase assay kit (Promega, Madison, WI) in a luminometer (Berthold, Bad Wildbad, Germany). All activity units were normalized by protein concentration and by the activity of the control vector. The data presented are representative of three independent experiments.

ChIP Assays.

Unless otherwise indicated, 1.3 × 10⁶ HeLa cells were transfected with 1 μg of the appropriate reporter plasmids (IE62P-61 or IE62P-39) or cotransfected with 2 μg of the IE62 expression plasmid (pCMV-IE62). For HCF-1 recruitment, 2 μg of HA-tagged IE62 (pCMV-HA-IE62) and 4 μg of V5-tagged HCF-1 (pHCF-1-V5) (18) were cotransfected with the IE62P-61 reporter. For depletion of HCF-1, cells were transfected with 3 μg of pU6-HCF-RNAi or pU6-control RNAi and were retransfected with reporter and IE62 expression plasmids 48 h later. ChIP assays were done essentially as previously described (39–41). Cell lysates were sonicated to obtain DNA fragments ranging from 300 to 700 bp. Chromatin from 7 × 10⁶ cells was used for each immunoprecipitation with the following antibodies: IE62 (29); HCF-1 (2); IgG, RNAPII, dimethyl-H3K4, trimethyl-H3K4, monomethyl-H3K9, and dimethyl-H3K9 (catalog nos. 12-370, 05-952, 07-030, 05-745, 07-450, and 07-441, respectively; Upstate Biotechnology, Lake Placid, NY); Sp1, TBP, TFIIB, TFIIH, and HA (catalog nos. SC-59, SC-273, SC-274, SC-6857/6859, and SC-805, respectively; Santa Cruz Biotechnology, Santa Cruz, CA); MLL1 and Set1 (catalog nos. 1408/1289 and 1193, respectively; Bethyl Laboratories, Montgomery, TX); and Set7/9 (catalog no. 13731; Abcam, Cambridge, MA). Immunoprecipitates were washed and eluted, and the cross-linking was reversed. Recovered DNA was subjected to standard PCR with dilutions of input DNA to ensure linearity, resolved in ethidium bromide agarose gels, and the signal intensities were quantitated by using a 4000MM Image Station (Kodak, Rochester, NY). The signal intensities of individual bands were calculated as a percentage of the intensity of the input extract after subtraction of the appropriate background antibody controls. Where indicated, samples were also subjected, in triplicate, to real-time PCR using SYBR Green (Qiagen, Valencia, CA) on a Prism 7900 system (Applied Biosystems, Foster City, CA) and analyzed with SDS 2.2.2 software. The data presented are the quantity means. All ChIP data shown are derived from a single ChIP assay but are representative of at least two independent experiments. The location and sequence of primer sets used are shown in SI Fig. 11.

Coimmunoprecipitations.

HeLa cells (1.3 × 10⁶) were transfected with 5 μg of pHA-IE62 expression plasmid or control vector DNA. For depletion of Sp1, 5 μg of pU6-Sp1 vector (Panomics, Redwood City, CA) or control RNAi was transfected on day 1 and retransfected on day 2 with 2.5 μg of pU6-Sp1 or control and 48 h later pHA-IE62 expression vector. Extracts were made 48 h later as described in ref. 18, sonicated briefly, and clarified by centrifugation. Protein extract was incubated for 2 h at 4°C with HA-Sepharose beads, washed five times with binding buffer, eluted in SDS sample buffer, and resolved in 4–20% Tris-glycine gels. Western blot analyses of resolved extracts and immunoprecipitates were done with the indicated antibodies (HA, HCF-1 AB2131, and Sp1), developed for chemiluminescence by using Super Signal Dura (Pierce, Rockford, IL), and quantitated by using a Kodak 4000MM Image Station.