Le rôle de la structure chromatinienne dans la régulation transcriptionnelle de l'expression des gènes

In: Bulletin de la Classe des Sciences de l'Académie Royale de Sciences, des Lettres et des Beaux-Arts de Belgique, Band 10, Heft 7, S. 303-338

Abstract

Chromatin is an important and dynamic participant in the regulation of transcription initiation and elongation. In the nucleosome (the basic sub-unit of chromatin), DNA is tightly wound around an octamer of histone proteins, thereby restricting access of the DNA-binding factors to the DNA. Moreover, the histone N-terminal tails are subject to several cova-lent post-translational modifications including phosphorylation, acetyla-tion, methylation and poly-ADP-ribosylation ; each of these modifications can potentially alter the interaction between DNA and the histone octamer and thus modulate transcription factor access to DNA. Chromatin is heterogeneous in the nucleus : small regions of the genome are extremely sensitive to digestion by nucleases (such as DNase I) and are called nuclease-hypersensitive sites. This increase in nuclease sensitivity is thought to result from the local disruption of the packaging of DNA into nucleosomes by DNA-bound regulatory proteins. DNase I-hypersen-sitive sites ("nucleo some-free regions") are found associated with regions of the genome that are important for the regulation of gene expression (such as enhancers, promoters, silencers, origins of replication). Moreover, changes in nuclease sensitivity secondary to the disrup¬ tion of specific nucleosomes have been observed during transcriptional activation or silencing. Consequently, understanding the mechanisms responsible for the positioning and for the disruption of specific nucleosomes is a crucial step in understanding the transcriptional regulation of the corresponding genes. Recent studies have identified a large group of proteins whose primary function is to help activate or repress transcrip¬ tion by altering chromatin.
Retroviruses in general, and HIV (Human Immunodeficiency Virus) in particular, are confronted with a unique problem in terms of transcriptional regulation and packaging into chromatin. They can integrate in many different sites within the host cell genome, each site with its own properties susceptible of influencing the degree of viral expression. The focus of our laboratory is understanding HIV-1 transcription within its natural context, which is represented by the provirus integrated into the cellular genome and packaged into chromatin. We have previously stud¬ ied the chromatin structure of HIV-1 integrated into several chronically infected cell lines and identified five major DNase I-hypersensitive sites in the complete proviral genome. Three hypersensitive sites were identified in the HIV 5' LTR where transcription initiation takes place : two of these sites (HS2 and HS3 ) correspond to the viral promo ter ! enhancer region, and the third major site (HS4) corresponds to a new domain immediately downstream of the 5' LTR. A single major DNase I-hyper-sensitive site is associated with the LTR located at the 3' portion of the virus where transcripts are polyadenylated (HS8). In addition, we have identified a single major DNase I-hypersensitive site in the 8-kb region located between the two LTRs. This site (HS7) maps to the pol gene of HIV and corresponds to a new positive transcriptional regulatory element.
We have defined the precise positions of nucleosomes in the 5' LTR of HIV-1 integrated in several chronically infected cell lines. We have observed that only one change occurs in this chromatin organization during transcriptional activation of the HIV-1 promoter : a single nucle-osome (called nuc-1 and positioned immediately after the transcription start site) is specifically disrupted ! displaced in response to several bio logical modifiers including TNF (Tumor Necrosis Factor) and phorbol esters. The position of nuc-1 immediately after the transcription start site and its disruption during transcriptional activation suggest that nuc-1 might play a direct role in the suppression of HIV-1 transcription during latency and that transcriptional activation might proceed in part through chromatin modification(s). To directly test this model, we examined the effect of a chromatin modification (histone acetylation) on HIV-1 gene expression. Chromatin is an active component of transcriptional regulation and histone acetylation/deacetylation is critical in this modulating role. Acetylation of histones takes place on their amino-ter-minal tail and is the result of a dynamic equilibrium between histone acetyltransferases and histone deacetylases. Trapoxin (TPX) and Tri-chostatin A (TSA) are two specific inhibitors of histone deacetylase(s) and cause a global hyperacetylation of cellular histones. Treatment of cell lines latently infected with HIV-1 with these inhibitors causes the transcriptional activation of HIV gene expression. This activation occurs in the absence of NF-kB stimulation. Using RNA differential display to screen a large number of genes, we have shown that the expression of a small fraction of cellular genes (~2%o of all genes ) is changed in response to histone hyperacetylation, indicating that the activation of HIV-1 transcription in response to TSA or TPX demonstrates a speci ficity and does not result from a global derepression of all cellular genes. Remarkably, despite the global histone hyperacetylation observed fol lowing treatment with these drugs, we found that the sole detectable modification at the level of HIV chromatin was the disruption of nuc-1 in the HIV promoter. These results are consistent with a model in which the HIV promoter is under the constitutive negative control of one or several histone deacetylases, which are specifically targeted to nuc-1 by a direct or indirect interaction with a DNA-binding factor (s).
The functional role of nuc-1 in latency and the mechanism of nuc-1 disruption in response to TNF or to histone hyperacetylation remain to be determined. A potential mechanism is that nuc-1 could block the bind ing of a transcription factor necessary for the assembly of the initiation complex. Alternatively, nuc-1, by its position immediately after the tran scription start site, could impede the progression of RNA polymerase II (by accentuating a natural pausing site), resulting in inefficient elonga tion and in the accumulation of short attenuated transcripts detected in vivo. Since the viral trans -regulatory protein Tat binds to TAR in a region close to nuc-1 and since Tat activity is critical for transcriptional elongation through a region corresponding to nuc-1, our results point to a role for Tat or for a Tat co-factor in the disruption of nuc-1 through a post-translational modification such as acetylation.
Overall, these studies provide fundamental new insights into the pro¬ cess of HIV-1 transcriptional latency and reactivation and ultimately should contribute to an increased understanding of AIDS pathogenesis. These aspects of HIV-1 biology might define novel targets for therapy aimed at interfering with HIV-1 replication by maintaining cells in the latent state. Additionally, since chromatin influences the transcriptional regulation of many genes besides HIV-1, the observations made in this system will bear relevance to the regulation of other genes whose expres¬ sion is altered in human diseases.