He continues to develop a broad range of biochemical methods designed to probe key features of chromatin structure common to all eukaryotes. At Janelia Farm, he is working collaboratively with colleagues using advanced microscopy to image histone and chromatin dynamics in vitro and in live cells.
Nucleosome-free, DNase I hypersensitive promoters
Chromatin is the complex of DNA, histone proteins and associated macromolecules that not only packages eukaryotic genomes within the confines of the cell nucleus but also physically masks the sequence of DNA bases in the double helix, thereby preventing indiscriminate access to genetic information. This organization plays a key role in controlling genome activities throughout life, from the onset of embryonic development to cell, tissue and organ differentiation. Aside from its fundamental importance, chromatin is causally implicated in gene misregulation in many human diseases, including cancer.
The organization of chromatin starts at the level of nucleosomes, the primary histone-DNA units of compaction, through intermediate levels of folding of nucleosome arrays to the most compacted chromosome forms visible under the light microscope. However, at cis-regulatory DNA elements, chromatin folding is interrupted by structural changes in nucleosome organization linked to early steps of gene expression. Initial studies from our laboratory suggested that creation of such a poised chromatin state, revealed by DNase I hypersensitivity, is necessary for subsequent induction of transcription. Mapping of DNase I hypersensitivity has been a longstanding assay in the chromatin toolbox, enabling discoveries of long distance enhancers and locus control regions, and has evolved into current genome-wide protocols to detect nucleosome-free and nucleosome-depleted sites at cis-regulatory elements throughout eukaryotic genomes.
ATP-dependent chromatin remodeling enzymes
We also developed an exonuclease protection assay that revealed the first in vivo footprints of eukaryotic transcription factors, leading to purification by DNA affinity chromatography and characterization of the heat shock transcription factor HSF, which undergoes heat shock-induced trimerization and high affinity DNA binding. Subsequently, development of a DNase I hypersensitivity reconstitution assay led to the discovery that promoter-specific chromatin remodeling requires concerted actions of sequence-specific GAGA factor and a novel ATP-dependent enzyme activity in Drosophila extracts, named NURF (Nucleosome Remodeling Factor). Unbiased, multi-step biochemical purification of NURF identified the four-subunit complex and its catalytic component ISWI, related to the SWI/SNF ATPase previously identified genetically as a co-regulator of transcription. NURF repositions nucleosomes by ATP-dependent ‘sliding’ of histone octamers on DNA, and additional genetic and molecular studies showed its requirement for activation or repression of several hundred genes, and its essential function in embryonic and post-embryonic fly and mouse development.
Histone H2A.Z replacement at promoters
Currently we are studying budding yeast SWR1, a 14-subunit enzyme whose catalytic ATPase is related to ISWI and SWI/SNF. The conserved SWR1 complex remodels nucleosomes by a mechanism distinct from all other chromatin remodelers. SWR1 evicts a histone H2A-H2B dimer from a conventional nucleosome and replaces it with histone variant H2A.Z-H2B. Histone H2A.Z is highly conserved from yeast to human, is universally localized to eukaryotic promoters and enhancers, and is also implicated in transcription. To elucidate the structural and biochemical basis of enzyme activity, we are systematically dissecting the histone H2A.Z replacement reaction. We have found that histone replacement occurs uni-directionally from H2A to H2A.Z in a step-wise fashion, one dimer at a time, and is dependent on activation of the Swr1 ATPase by its two natural substratesthe conventional nucleosome and the H2A.Z-H2B dimer. Ongoing studies probe other steps in the histone H2A.Z exchange pathway, and their relationship and dynamic interplay with the assembly and function of the transcription machinery.
We are also investigating the centromere-specific histone variant, CenH3, which forms the foundation of the kinetochore connecting daughter chromosomes to the spindle apparatus for faithful segregation of duplicated genomes. We have recently identified a yeast CenH3-specfic histone chaperone, Scm3, and are studying the assembly and maintenance of the centromere-specific nucleosome. We are particularly interested in how this unique variant nucleosome performs its specialized function as the key link to kinetochore proteins during chromosome segregation.
Histone variant H2A.Z-containing nucleosomes exist at most eukaryotic promoters and play important roles in gene transcription and genome stability. The multisubunit nucleosome-remodeling enzyme complex SWR1, conserved from yeast to mammals, catalyzes the ATP-dependent replacement of histone H2A in canonical nucleosomes with H2A.Z. How SWR1 catalyzes the replacement reaction is largely unknown. Here, we determined the crystal structure of the N-terminal region (599-627) of the catalytic subunit Swr1, termed Swr1-Z domain, in complex with the H2A.Z-H2B dimer at 1.78 Å resolution. The Swr1-Z domain forms a 310 helix and an irregular chain. A conserved LxxLF motif in the Swr1-Z 310 helix specifically recognizes the αC helix of H2A.Z. Our results show that the Swr1-Z domain can deliver the H2A.Z-H2B dimer to the DNA-(H3-H4)2 tetrasome to form the nucleosome by a histone chaperone mechanism.
The budding yeast centromere contains Cse4, a specialized histone H3 variant. Fluorescence pulse-chase analysis of an internally tagged Cse4 reveals that it is replaced with newly synthesized molecules in S phase, remaining stably associated with centromeres thereafter. In contrast, C-terminally-tagged Cse4 is functionally impaired, showing slow cell growth, cell lethality at elevated temperatures and extra-centromeric nuclear accumulation. Recent studies using such strains gave conflicting findings regarding the centromeric abundance and cell cycle dynamics of Cse4. Our findings indicate that internally tagged Cse4 is a better reporter of the biology of this histone variant. Furthermore, the size of centromeric Cse4 clusters was precisely mapped with a new 3D-PALM method, revealing substantial compaction during anaphase. Cse4-specific chaperone Scm3 displays steady-state, stoichiometric co-localization with Cse4 at centromeres throughout the cell cycle, while undergoing exchange with a nuclear pool. These findings suggest that a stable Cse4 nucleosome is maintained by dynamic chaperone-in-residence Scm3.
The histone variant H2A.Z is a genome-wide signature of nucleosomes proximal to eukaryotic regulatory DNA. Whereas the multisubunit chromatin remodeler SWR1 is known to catalyze ATP-dependent deposition of H2A.Z, the mechanism of SWR1 recruitment to S. cerevisiae promoters has been unclear. A sensitive assay for competitive binding of dinucleosome substrates revealed that SWR1 preferentially binds long nucleosome-free DNA and the adjoining nucleosome core particle, allowing discrimination of gene promoters over gene bodies. Analysis of mutants indicates that the conserved Swc2/YL1 subunit and the adenosine triphosphatase domain of Swr1 are mainly responsible for binding to substrate. SWR1 binding is enhanced on nucleosomes acetylated by the NuA4 histone acetyltransferase, but recognition of nucleosome-free and nucleosomal DNA is dominant over interaction with acetylated histones. Such hierarchical cooperation between DNA and histone signals expands the dynamic range of genetic switches, unifying classical gene regulation by DNA-binding factors with ATP-dependent nucleosome remodeling and posttranslational histone modifications.
The ATP-dependent chromatin-remodeling complex SWR1 exchanges a variant histone H2A.Z/H2B dimer for a canonical H2A/H2B dimer at nucleosomes flanking histone-depleted regions, such as promoters. This localization of H2A.Z is conserved throughout eukaryotes. SWR1 is a 1 megadalton complex containing 14 different polypeptides, including the AAA+ ATPases Rvb1 and Rvb2. Using electron microscopy, we obtained the three-dimensional structure of SWR1 and mapped its major functional components. Our data show that SWR1 contains a single heterohexameric Rvb1/Rvb2 ring that, together with the catalytic subunit Swr1, brackets two independently assembled multisubunit modules. We also show that SWR1 undergoes a large conformational change upon engaging a limited region of the nucleosome core particle. Our work suggests an important structural role for the Rvbs and a distinct substrate-handling mode by SWR1, thereby providing a structural framework for understanding the complex dimer-exchange reaction.