Histone deacetylase 3, not histone deacetylase 2, interacts with the major immediate early locus of human cytomegalovirus
Evidence suggests that genome chromatinization and the posttranslational modification of histones are involved in the regulation of viral gene expression, including the human cytomegalovirus (HCMV). We performed a ChIP-on-Chip assay to determine whether histone deacetylases (HDACs) interact with HCMV genomic DNA on a global level. Surprisingly, we found that HDAC3, but not HDAC2, interacts not only with the major immediate early (MIE) promoter but also with the entire MIE locus, suggesting a heterogeneous interaction of HDAC3 with HCMV DNA. The interaction of HDAC3 with the MIE region is related to inhibition of viral replication because HDAC3 inhibitors enhanced HCMV replication.
KeywordsViral Replication HDAC Inhibitor Viral Gene Expression HCMV Infection Human Foreskin Fibroblast
(major immediate early)
(cytomegalic inclusion disease)
(small ubiquitin-like modifier)
(human foreskin fibroblast cells)
Human cytomegalovirus (HCMV) is a ubiquitous virus and infects a majority of the general population (50-90%) . The fact that the incidence of cytomegalic inclusion disease (CID) is intimately related to viral burden suggests that the inhibition of viral production by the specific repression of viral gene expression will reduce the occurrence of CID . Understanding the mechanism of HCMV gene regulation is the pre-requisite for developing drugs that interfere with viral replication by repressing viral gene expression. The HCMV major immediate early (MIE) gene products, IE1 and IE2, are among the first de novo-expressed viral proteins that are required for subsequent viral gene expression and hence viral replication . IE1 and IE2 mRNAs are encoded by the major IE locus that spans from 169 to175 kbp of the viral genome and are produced by alternative splicing and differential polyadenylation [4, 5, 6]. IE1 and IE2 share the first 85 amino acids . MIE genes are controlled by a strong promoter/enhancer that contains many regulatory elements [3, 7].
Different cellular mechanisms have been found to play roles in inhibiting viral gene expression, and one of the most prominent ones is gene silencing through viral DNA chromatinization (also called chromatin remodeling) , a procedure carried out by histone or histone-related proteins, such as histone acetylase (HAT) and histone deacetylase (HDAC). Several posttranslational modifications of histone proteins have been defined to be involved in chromatin remodeling, including acetylation by HAT, deacetylation by HDAC, SUMOylation by SUMO (Small Ubiquitin-like Modifier)-related pathway, deSUMOylation by SENP (a SUMO-specific protease) family enzymes, phosphorylation by kinase pathways, and methylation via methylases [9, 10, 11, 12, 13]. Those enzymatic pathways orchestrate to regulate cellular gene transcription and are termed as epigenetic codes [10, 13]. Viral gene transcription requires cellular machinery, which is probably also regulated by cellular gene regulatory pathways. Histones are abundant nuclear proteins and have been shown to bind with HCMV genomic DNA . Therefore, it was reasonable to propose that chromatin remodeling of viral DNA takes place in the nucleus, which speculation was validated when the fact that HDAC inhibitors can promote cytomegalovirus production was also confirmed [15, 16, 17].
How HCMV strips off the cellular proteins in order for the virus to replicate its own DNA is not fully understood. Recent studies have shown that HCMV and murine CMV (MCMV) major immediate early proteins, IE1 and IE2 (or IE3 for MCMV), interact with HDAC1, 2, and 3, and HDAC inhibitors enhance viral production [17, 18, 19, 20, 21], and dynamic chromatin modification of the MIE promoter and other viral promoters has been shown. However, interaction of HDACs with the viral genome has not been clearly demonstrated . In this study, we performed chromatin immunoprecipitation followed by microarray on an HCMV DNA chip (ChIP-on-chip) assay to demonstrate the interaction of HCMV DNA with HDACs. To our surprise, we found that HDAC3, but not HDAC2, interacts specifically with the MIE locus, which suggests a heterogeneous interaction of HDAC3 with HCMV genomic DNA. In addition, we found that the interactions of HDAC3 with the MIE locus might relate to the modulation of viral replication because HDAC3 inhibitors can significantly enhance viral growth.
The chromatinization of viral DNA after its having entered the nucleus has been noted not only in latently infected viruses such as EBV and KSHV (the genomes of which are tethered to cellular chromosomes) but also in the lytic infection of HCMV [8, 23, 24]. On the other hand, histone proteins have not been found in herpesvirus virons . Therefore, the chromatinization of HCMV DNA must be temporary and dynamic. We wonder 1) whether the HDACs are bound to the HCMV DNA, and 2) if so, where they interact and whether the interaction is homogenous or heterogeneous. In order to answer these questions, we performed a ChIP-on-chip assay.
The human foreskin fibroblast cells (HFF) were infected with HCMV at an MOI of 5. The cells were fixed at 24 hours postinfection with 1% paraformaldehyde. The chromatin immunoprecipitation (ChIP) inputs were prepared and performed using the commercial kit (EZ ChIP, Upstate Cell Signal Solutions), according to the manufacturer's protocol. The antibodies used for ChIP assays include anti-HDAC2 (clone 3F3), anti-HDAC3 (clone 3G6, Upstate USA, Inc.), and normal IgG (as a negative control).
Since chromatin immunoprecipitation yields a small amount of DNA, PCR amplifications are required before labeling steps. First, precipitated input DNA was randomly primed with a sequence-tagged oligonucleotide (GTTTCCCAGTCACGATCNNNN NNNNN) to generate templates with a specific tag at both ends for subsequent PCR. Then, a specific primer (GTTTCCCAGTCACGATC) was used to amplify the templates previously generated. The final amplification consisted of additional PCR cycles to incorporate aminoallyl-dUTP. The amplified products were then labeled with Alexa Flour 647 or Alexa Flour 555 and applied to a DNA array of HCMV. The hybridization was performed according to manufacture's instruction (Corning UltraGAPS™).
The slides were scanned using an Axon 4000A scanner. Images were analyzed using GenePix Pro 5.1. A normalization factor was determined using the ratio of the median signal intensity of the cy5 and cy3 channels, and all data sets were normalized to a single medium intensity. A log2 signal intensity ratio for each fragment was calculated for the ChIP samples that were pulled down by monoclonal antibody against HDAC3 (or HDAC2 or histone H3) to that by normal IgG. To correct for the biases in the ChIP procedure, all log ratios were normalized by subtracting the average log ratio across the whole array. We considered that a spot was enriched if the signal intensity from ChIP sample of anti-HDAC3 was 4-fold greater than the corrected signal from the IgG ChIP sample.
To further confirm the ChIP-on-chip results, we performed a series PCR reactions using primers amplifying HCMV genomic fragments 113 to 140 (Figure 2, Additional file 1, Table S1). As shown in Figure 2B, all HCMV genomic fragments were present in the ChIP input sample; IgG and HDAC2 did not specifically interact with the HCMV genome; histone H3 interacted with the entire viral DNA tested; and most importantly, HDAC3 interacted with fragment 47-53 kb of the Toledo genome corresponding to the HCMV MIE locus (Figure 2C, green). To confirm that the anti-HDAC2 antibody is effective for the ChIP assay, we performed another ChIP assay using anti-HDAC2, -HDAC3 antibodies to bind ChIP samples made from uninfected Mrc-5 cells. The eluted DNA samples were applied for running PCR against hTERT promoter as reported by Straat et al. As shown in Figure 2D, HDAC2 bound to hTERT promoter, which is consistent with the report .
We demonstrated that HDAC3 but not HDAC2 interacts with the HCMV MIE locus. The reason that HDAC2 was not detected in our assay could be due to the fact that HDAC2 is known to interact with IE2, and this binding suppresses the repressive effect of HDAC2 . At 24 hrs postinfection, the genomic region around the MIE may need to stay repressed, which could be accomplished by the action of HDAC3. This binding of HDAC3 to not only the MIE locus but also the genes upstream of MIE is intriguing. The region between the MIE and UL127 has been named as a unique region without any known function. This region is known to have binding sites for various cellular repressor proteins that help to repress the transcription of UL127. It is known that the promoter of UL127 gene is silenced during productive infection in fibroblasts . The UL126 gene has been reported as a latency-associated gene . UL124 is a putative membrane glycoprotein that may be expressed late in the infection process . It will be interesting to see how the HDACs bind with those genomic regions at later times of infection.
HDAC3 has been shown to be a repressor of the viral MIE promoter . A significant increase in viral replication in the presence of the HDAC inhibitor demonstrates that the binding of HDAC3 causes inhibition of viral replication. This inhibition seems to be more pronounced at a low MOI. Host cells contain several proteins including gene expression suppressors that form a defensive arm against viral infection . HCMV has evolved strategies against cellular defense. We speculate that HCMV infection must reduce HDAC activity. Both IE1 and IE2 of HCMV were reported to functionally interact with HDACs [14, 18, 19, 20], and IE1 of MCMV is particularly adept at reducing HDAC activity . These observations and our finding of an increase in viral replication in the presence of TSA further confirm the IE2-mediated repression of the MIE promoter by the recruitment of chromatin remodeling factors.
This work was supported by NIH grant AI050709 (H.Z.), a Pilot grant from the Research Center for Minority Institutes (RCMI) Program (2G12RR003050-24) (Q.T.), and American Cancer Society grant RSG-090289-01-MPC (Q.T). We are grateful to Bob Ritchie (RCMI, 2G12RR003050-24) for English editing.
- 1.Mocarski E: Cytomegaloviruses and Their Replication. In Fields Virology. Volume 2. 5th edition. Edited by: Knipe DMaH. P.M. Philadelphia: Lippincott Williams & Wilkins; 2007:2629-2673.Google Scholar
- 15.Michaelis M, Suhan T, Reinisch A, Reisenauer A, Fleckenstein C, Eikel D, Gumbel H, Doerr HW, Nau H, Cinatl J Jr: Increased replication of human cytomegalovirus in retinal pigment epithelial cells by valproic acid depends on histone deacetylase inhibition. Invest Ophthalmol Vis Sci 2005, 46: 3451-3457. 10.1167/iovs.05-0369CrossRefPubMedGoogle Scholar
- 19.Reeves M, Murphy J, Greaves R, Fairley J, Brehm A, Sinclair J: Autorepression of the human cytomegalovirus major immediate-early promoter/enhancer at late times of infection is mediated by the recruitment of chromatin remodeling enzymes by IE86. J Virol 2006, 80: 9998-10009. 10.1128/JVI.01297-06PubMedCentralCrossRefPubMedGoogle Scholar
- 21.Martinez FP, Cosme RS, Tang Q: Murine Cytomegalovirus Major Immediate Early Protein 3 (IE3) Interacts with Cellular and Viral Proteins in the Viral DNA Replication Compartments and Is Important for Early Gene Activation. J Gen Virol 2010,91(Pt 11):2664-76. 10.1099/vir.0.022301-0PubMedCentralCrossRefPubMedGoogle Scholar
- 25.Varnum SM, Streblow DN, Monroe ME, Smith P, Auberry KJ, Pasa-Tolic L, Wang D, Camp DG, Rodland K, Wiley S, et al.: Identification of proteins in human cytomegalovirus (HCMV) particles: the HCMV proteome. J Virol 2004, 78: 10960-10966. 10.1128/JVI.78.20.10960-10966.2004PubMedCentralCrossRefPubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.