David J. Clark, PhD, Head, Unit on Chromatin and Gene Expression
Peter Eriksson, PhD, Research Fellow
Neil McLaughlin, PhD, Postdoctoral Fellow
Geetu Mendiratta, PhD, Postdoctoral Fellow
Naoe Kotomura, PhD, Volunteer

Gene activation involves the regulated recruitment of factors to a promoter in response to appropriate signals, ultimately resulting in the formation of an initiation complex by RNA polymerase II and transcript elongation. These events must occur in the presence of nucleosomes, which are compact structures capable of blocking transcription at every step. To circumvent the chromatin block, eukaryotic cells possess a set of chromatin-remodeling and nucleosome-modifying complexes. The former (e.g., the SWI/SNF complex) use ATP to drive conformational changes in nucleosomes and to slide nucleosomes along DNA. The latter contain enzymatic activities (e.g., histone acetylases) that modify histones post-translationally to mark them for recognition by other complexes. Geneticists have described many interesting connections between chromatin components and transcription, but a system to investigate the structural basis of the connections has been lacking. We have developed such a model system, involving native plasmid chromatin purified from the yeast Saccharomyces cerevisiae, to perform high-resolution studies of the chromatin structures of active and inactive genes. Remarkably, we found that activation correlates with large-scale movements of nucleosomes and conformational changes within nucleosomes over entire genes. Our current work focuses on the roles of chromatin remodeling and histone acetylation in gene regulation.

Activation of yeast HIS3 results in Gcn4p- and SWI/SNF-dependent mobilization of nucleosomes over the entire gene

Kim, ¹ McLaughlin; in collaboration with Lutter, Tsukiyama

We chose budding yeast as a model organism because it permits biochemical studies of chromatin structure to be combined with molecular genetics. Current models for the role of the SWI/SNF ATP-dependent chromatin remodeling complex in gene regulation focus on promoters, where the most obvious changes in chromatin structure occur. However, by using our plasmid model system with yeast HIS3, a SWI/SNF-regulated gene, we discovered that induction of HIS3 creates a domain of remodeled chromatin structure that extends far beyond the promoter to include the entire gene (Kim and Clark, Proc Natl Acad Sci USA 2002;99:15381). We addressed the effects of transcriptional activation on the chromatin structure of the yeast HIS3 gene by mapping the precise positions of nucleosomes in uninduced and induced chromatin. In the absence of the Gcn4p activator, the HIS3 gene is organized into a dominant nucleosomal array. In wild-type chromatin, the array is disrupted, and several alternative, overlapping nucleosomal arrays are formed. Disruption of the dominant array also requires the SWI/SNF remodeling machine, indicating that the SWI/SNF complex plays an important role in nucleosome mobilization over the entire HIS3 gene. The Isw1 remodeling complex plays a more subtle role in determining nucleosome positions on HIS3, favoring positions that differ from those preferred by the SWI/SNF complex. We propose that Gcn4p stimulates nucleosome mobilization over the entire HIS3 gene by the SWI/SNF complex. We suggest that the net effect of interplay among remodeling machines at HIS3 is to create a highly dynamic chromatin structure (Kim et al., under revision). Our current work aims at elucidating the structure of the remodeled nucleosome. Thus far, we have identified at least two possibilities: unstable nucleosomes (remodeled such that they fall apart easily) and nucleosomes with a dramatically altered conformation. Our work on CUP1 and HIS3 indicates that, at least for these two genes, the target of remodeling complexes is a domain rather than just the promoter. This is an important finding because it suggests that remodeling complexes act on chromatin domains. As to the function of domain remodeling, we speculate that remodeling entire genes might facilitate elongation through nucleosomes by RNA polymerase II.

In a collaboration with Len Lutter, we compared the structures of yeast and mammalian chromatin, leading to the suggestion that yeast chromatin is composed of relatively ordered arrays of closely spaced nucleosomes separated by substantial gaps, possibly corresponding to regulatory regions.

The yeast Spt10 protein is a sequence-specific activator of histone genes, possessing a DNA-binding domain fused to a putative histone acetylase domain.

Eriksson, Mendiratta, McLaughlin; in collaboration with Landsman, Shen

Our early studies showed that induction of CUP1 by copper results in targeted acetylation of nucleosomes at the CUP1 promoter. This acetylation is dependent on SPT10, which encodes a putative histone acetylase (HAT). SPT10 has been implicated as a global regulator of core promoter activity. We confirmed SPT10's role by expression microarray analysis and then addressed the mechanism of global regulation (Eriksson et al., 2005). Using the chromatin-immunoprecipitation (ChIP) assay, we were unable to detect Spt10p at any of the most strongly affected genes in vivo, but we confirmed its presence at the core histone gene promoters, which it activates. We proposed that, in the absence of Spt10p, a shortage of histones might occur, resulting in defective chromatin structure and consequent activation of basal promoters. Consistent with this hypothesis, the spt10Δ phenotype can be rescued by extra copies of the histone genes, and chromatin is poorly assembled in spt10Δ cells, as shown by irregular nucleosome spacing. Furthermore, we find that Spt10p binds specifically and highly cooperatively to pairs of UAS elements in the core histone promoters (consensus: (G/A)TTCCN6TTCNC), consistent with a direct role in histone gene regulation. No other high-affinity sites are predicted in the yeast genome. Our observations are consistent with the idea that the global changes in gene expression in spt10Δ cells are actually the indirect effects of defective regulation of the core histone genes. Thus, Spt10p is a sequence-specific activator of histone genes, possessing a DNA-binding domain fused to a likely HAT domain. We have now identified the DNA-binding domain of Spt10p: it contains an unusual zinc finger (His2-Cys2), which has homology to the DNA integrase of foamy retroviruses. We proposed that this integrase might also be a sequence-specific DNA-binding protein (Mendiratta et al., 2006). We have also shown that Spt10p is a dimer and that the N-terminal domain is required for dimer formation (Mendiratta et al., submitted). Our current work aims to (1) demonstrate the putative histone/protein acetylase activity of Spt10p; (2) identify proteins that interact with Spt10p; and (3) determine whether human foamy virus (HFV) integrase is indeed a sequence-specific DNA-binding protein.

¹ Yeonjung Kim, PhD, former Visiting Fellow

COLLABORATORS

David Landsman, PhD, National Center for Biotechnology Information, NLM, Bethesda, MD
Len Lutter, PhD, Henry Ford Hospital, Detroit, MI
Chang-Hui Shen, PhD, City University of New York Staten Island, New York, NY
Toshio Tsukiyama, PhD,Fred Hutchinson Institute for Cancer Research, Seattle, WA

For further information, contact clarkda@mail.nih.gov.