Counterregulation of chromatin deacetylation and histone deacetylase occupancy at the integrated promoter of human immunodeficiency virus type 1 (HIV-1) by the HIV-1 repressor YY1 and HIV-1 activator Tat

Abstract

Repression of human immunodeficiency virus type 1 (HIV-1) transcription may contribute to the establishment or maintenance of proviral quiescence in infected CD4(+) cells. The host factors YY1 and LSF cooperatively recruit histone deacetylase 1 (HDAC1) to the HIV-1 long terminal repeat (LTR) and inhibit transcription. We demonstrate here regulation of occupancy of HDAC1 at a positioned nucleosome (nuc 1) near the transcription start site of integrated LTR. We find that expression of YY1 increases occupancy by HDAC1, decreases acetylation at nuc 1, and downregulates LTR expression. HDAC1 recruitment and histone hypoacetylation were also seen when Tat activation was inhibited by the overexpression of YY1. A YY1 mutant without an HDAC1 interaction domain and incompetent to inhibit LTR activation fails to recruit HDAC1 to LTR or decrease nuc 1 acetylation. Further, expression of a dominant-negative mutant of LSF (dnLSF), which inhibits LSF occupancy and LTR repression, results in acetylation and decreased HDAC1 occupancy at nuc 1. Conversely, exposure of cells to the histone deacetylase inhibitor trichostatin A or activation of LTR expression by HIV-1 Tat results in the displacement of HDAC1 from nuc 1, in association with increased acetylation of histone H4. Recruitment of HDAC1 to the LTR nuc 1 can counteract Tat activation and repress LTR expression. Significantly, when repression is overcome, LTR activation is associated with decreased HDAC1 occupancy. Since the persistence of integrated HIV-1 genomes despite potent suppression of viral replication is a major obstacle for current antiretroviral therapy, strategies to selectively disrupt the quiescence of chromosomal provirus may play a role in the future treatment of AIDS.

Figures

FIG. 1.

The HDAC inhibitor TSA increases acetylation of histone H4 at the nuc 1 region of HIV-1 LTR and activates the integrated LTR. A HeLa cell line with an integrated copy of a LTR-CAT reporter gene was grown in 0.5% serum overnight, refed with complete medium, and stimulated with TSA for 2 h. Cell extracts were subjected to ChIP assays with antibody directed against acetylated histone H4. Total RNA was isolated from a portion of cells and analyzed by RT-PCR. (A) Semiquantitative duplex PCR was performed to amplify both LTR and β-actin promoter from genomic DNA extracted from HeLa LTR CAT cells. Serial dilutions of input were subjected to PCR amplification in parallel with each experiment. The relative intensity of the PCR product bands correlated with the amount of input DNA. These results were used to validate the fold change in quantification. (B) Duplex PCR amplifications of whole-cell extract (Input), nonspecific IgG immunoprecipitate (IgG), and α-acetylated histone H4 immunoprecipitate (IP) were carried out to show changes of acetylated histone H4 at nuc 1 and β-actin promoter after TSA treatment. The abundance of DNAs associated with acetylated histone H4 (IP) is shown as PCR products. (C) RT-PCR assay compared CAT transcripts after TSA treatment. β-Actin transcripts were amplified to ensure the equivalent input of cDNA.

FIG. 2.

HDAC1 is localized at the nuc 1 region of HIV-1 LTR in vivo. (A) A ChIP assay was performed with antibody directed against HDAC1. Formaldehyde cross-linking retained HDAC1 at the LTR. HDAC1 occupancy was not detected at the constitutive β-actin promoter. (B) ChIP assays with antibodies against HDAC1 detected the displacement of HDAC1 occupancy at nuc 1 after LTR activation by TSA. (C). Transient TSA treatment did not alter the expression of HDAC1 as shown by Western blot and RT-PCR. (D) TSA does not decrease HDAC1 occupancy by decreasing LSF occupancy.

FIG. 3.

Changes of histone H4 acetylation and HDAC1 occupancy at nuc 1 correlate with Tat activation of LTR. HeLa LTR-CAT cells were cotransfected with GFP and empty CMV vector or with GFP and CMV-Tat plasmids. After 24 h, transfected cells were separated into GFP-positive and -negative populations by FACS. Extracts were prepared for CAT and ChIP assays with antibodies against acetylated histone H4, HDAC1, or LSF. (A) Tat-activated integrated LTR-CAT in HeLa reporter line as shown by CAT assay. (B) A CAT assay showed cotransfection exclusively delivered Tat plasmids into GFP-positive cells. (C) ChIP assays showed increased acetylation of histone H4 and decreased HDAC1 occupancy at nuc 1 of LTR upon Tat activation in GFP-positive cells. (D) Overexpression of Tat did not affect LSF occupancy at nuc 1.

FIG. 4.

Binding of LSF to LTR correlates with HDAC1 recruitment and hypoacetylation at the nuc 1 region of LTR. HeLa LTR-CAT cells cotransfected with GFP plus empty CMV vector, GFP plus CMV-LSF, or GFP plus CMV-dnLSF were sorted by FACS and ChIP assays performed with antibodies against acetylated histone H4 or HDAC1. (A) Overexpression of dnLSF decreased LSF occupancy at nuc 1. (B) Overexpression of dnLSF increased acetylation of histone H4 (upper panel) and decreased HDAC1 occupancy (middle panel) at nuc 1.

FIG. 5.

YY1 recruits HDAC1 and inhibits histone H4 acetylation at the nuc 1 region of LTR. HeLa LTR-CAT cells were cotransfected with GFP plus empty CMV vector or with GFP plus CMV-YY1 and then sorted by FACS and ChIP assays performed with antibodies against acetylated histone H4, HDAC1, and LSF. (A) YY1 increased HDAC1 occupancy and decreased acetylation of histone H4 at nuc 1. (B) Overexpression of YY1 did not change LSF occupancy at nuc 1.

FIG. 6.

HDAC1 recruitment by YY1 is required for the repression of LTR. Cell extracts were prepared from GFP-positive cells cotransfected with empty CMV vector, Tat, Tat plus wild-type YY1, and Tat plus mutant YY1 lacking the HDAC1 interaction domain. CAT assays (upper panel) were performed to correlate the level of LTR expression with the acetylation of histone H4 and HDAC1 occupancy (middle panel) at nuc 1, as measured by ChIP assays. A YY1 mutant with a deletion of HDAC1 interaction domain was unable to repress the LTR, increase HDAC1 occupancy, or significantly decrease H4 acetylation at nuc 1.